Compare commits

..

440 Commits

Author SHA1 Message Date
Yicheng Qin
6335fdc595 *: bump to v2.1.1 2015-07-21 10:41:26 -07:00
Yicheng Qin
ca540b23dc Merge pull request #3158 from yichengq/exp-auth
docs: add experimental notice on auth doc
2015-07-21 10:24:11 -07:00
Yicheng Qin
097ec0f25b docs: add experimental notice on auth doc
Reasons for the notice:
1. No users have reported about their feedback about auth feature so
far.
2. We haven't used it internally.
3. This is the first release that includes auth feature, so it is good
to be more cautious.
2015-07-21 10:23:23 -07:00
Xiang Li
ed444419c0 Merge pull request #3160 from ryandoyle/docs-nss-etcd
docs: glibc NSS module for resolving names in etcd
2015-07-21 08:53:08 +08:00
Ryan Doyle
d373645b8c docs: glibc NSS module for resolving names in etcd 2015-07-21 10:33:23 +10:00
Yicheng Qin
d86e94b824 Merge pull request #3128 from yichengq/doc-watch-api
docs: update watch API doc for clarity
2015-07-20 14:54:26 -07:00
Yicheng Qin
d52cb2e5d9 docs: add watch command and correct the example 2015-07-20 14:49:01 -07:00
Xiang Li
40681bdf03 Merge pull request #3146 from a-robinson/snap
snap: Record the snapshot save duration on success rather than only on error
2015-07-17 06:03:33 +08:00
Yicheng Qin
7d38115cb2 Merge pull request #3148 from yichengq/update-contact
Update contact section in README.md
2015-07-16 15:01:54 -07:00
Barak Michener
f8baa4ebe0 Merge pull request #3138 from barakmich/auth_doc
documentation: Add authentication walkthrough with etcdctl. Fixes #2949
2015-07-16 16:41:15 -04:00
Yicheng Qin
9b962c8350 README: let roadmap point to ROADMAP.md 2015-07-16 12:58:13 -07:00
Yicheng Qin
c1aed32920 README: update irc channel to #etcd 2015-07-16 12:52:26 -07:00
Alex Robinson
57a5520157 snap: Record the snapshot save duration on success rather than only on error.
It makes more sense to record the latency of successes (or all attempts)
than of only a particular failure case.
2015-07-16 10:46:47 -07:00
Barak Michener
452a327334 documentation: Add authentication walkthrough with etcdctl. Fixes #2949 2015-07-15 15:54:26 -04:00
Yicheng Qin
ebbb0caff0 Merge pull request #3136 from yichengq/fix-proxy-doc
docs: fix wrong proxy command
2015-07-15 11:30:07 -07:00
Yicheng Qin
d0e976ad4b docs: fix wrong proxy command 2015-07-15 08:37:10 -07:00
Xiang Li
d0e3e2c992 Merge pull request #3131 from yichengq/remove-header-timeout
discovery: remove ResponseHeaderTimeout when discovery
2015-07-15 07:50:36 +08:00
Yicheng Qin
1db176151b discovery: remove ResponseHeaderTimeout when discovery
The discovery service doesn't return HTTP header early when watch
starts. This may trigger ResponseHeaderTimeout and cause the watch
request failed.

The fix on discovery service may take some time. Remove the
ResponseHeaderTimeout first so it behaves as before.
2015-07-14 16:33:28 -07:00
Michael Schurter
f52db1c08e docs: add back original example w/index=prevIndex 2015-07-13 23:04:39 -07:00
Michael Schurter
b94f6595e6 docs: rewrite existing docs instead of adding a new section
@xiang90 pointed out my earlier commit duplicated a lot of things that
were mentioned earlier in the doc.

This time around I tried just making some gotchas more explicit in the
existing docs instead of just tacking new stuff onto the end.
2015-07-13 23:03:59 -07:00
Xiang Li
953a59d554 Merge pull request #3127 from yanana/emend-error-message
etcdmain: emend configuration error message
2015-07-14 13:46:08 +08:00
Shun Yanaura
d7276d6ace etcdmain: emend configuration error message
etcd shows an odd message on configuration error like this (partially):
```
... discovery or bootstrap flags are setChoose one of ...
                                      ^^^^^^^^^
```
This commit fixes the message format problem.
2015-07-14 14:42:49 +09:00
Yicheng Qin
d80f4c8aa2 Merge pull request #3125 from yichengq/doc-tuning
docs: update tuning.md to match today's situation
2015-07-13 16:52:11 -07:00
Yicheng Qin
8b7c600009 docs: update tuning.md to match today's situation
1. etcd requires that election-timeout >= 5 * heartbeat-interval
2. etcd doesn't have flag -snapshot
2015-07-13 16:35:30 -07:00
Yicheng Qin
7a520bb80b Merge pull request #3121 from yichengq/extend-schedule
pkg/testutil: extend wait schedule time to 10ms
2015-07-13 15:23:36 -07:00
Yicheng Qin
1624235bb3 pkg/testutil: extend wait schedule time to 10ms
Waiting 3ms is not long enough for schedule to work well. The test suite
may fail once per 200 times in travis due to this. Extend this to 10ms
to ensure schedule could work. Now it could run 1000 times successfully
in travis.
2015-07-13 09:05:40 -07:00
Yicheng Qin
5be545b872 Merge pull request #3077 from yichengq/fix-test-sync
etcdserver: init raft internal var early
2015-07-10 14:44:52 -07:00
Xiang Li
c7a949349e Merge pull request #3113 from xiang90/fix_proxy_bt
etcdmain: proxy ignores discovery if it is initialized
2015-07-10 14:12:45 -06:00
Xiang Li
dedabddcb3 etcdmain: proxy ignores discovery if it is initialized 2015-07-10 12:52:24 -07:00
Xiang Li
61e9b99edb Merge pull request #2417 from kelseyhightower/improve-etcdctl-ls-command-help
etcdctl: update the ls subcommand help to match behavior
2015-07-09 11:33:19 -06:00
Xiang Li
4631b727c0 Merge pull request #3105 from xiang90/rd
doc: add rolling upgrade doc for 2.1
2015-07-09 11:27:05 -06:00
Xiang Li
11452585bb doc: add rolling upgrade doc for 2.1 2015-07-07 13:20:41 -07:00
Xiang Li
8ab388fa56 Merge pull request #3001 from mwitkow-io/feature/rich_metrics
Etcd Rich Metrics
2015-07-07 08:12:06 -07:00
Michal Witkowski
7bca757d09 *: add metrics to store and proxy. 2015-07-07 16:01:51 +01:00
Yicheng Qin
573f62f7a5 Merge pull request #3101 from yichengq/check-err
integration: always check error for function calls
2015-07-06 18:10:31 -07:00
Yicheng Qin
e7ed7a7b7a integration: always check error for function calls 2015-07-06 17:44:36 -07:00
Brandon Philips
121ff4684c Merge pull request #3097 from philips/tls-churn-faq
Documentation/security: add FAQ about peer TLS and etcd 2.0.x
2015-07-04 15:30:42 -07:00
Brandon Philips
83fe8187f4 Documentation/security: add FAQ about peer TLS and etcd 2.0.x
etcd 2.0.x TLS can appear not to work on smaller machines with less
horse-power or lots of other work going on. Document the timeout
workaround.
2015-07-04 15:28:47 -07:00
Xiang Li
cbe00e4415 Merge pull request #2967 from webner/feature/proxy-config
proxy: added endpoint refresh and timeout configuration values
2015-07-03 11:51:15 -07:00
Wolfgang Ebner
954e416bf6 proxy: fixed director.go formatting 2015-07-03 14:11:40 +02:00
Xiang Li
883bb47dcf Merge pull request #3074 from xiang90/storage_restore
storage: correctly restore create and ver
2015-06-30 09:20:19 -07:00
Xiang Li
eff67afc60 Merge pull request #3081 from xiang90/storage_fix
storage: fix small issues
2015-06-29 22:05:46 -07:00
Xiang Li
585e74a1b1 Merge pull request #3080 from xiang90/rpc
add gRPC etcd service
2015-06-29 22:04:47 -07:00
Xiang Li
f8b947a00b storage: fix small issues 2015-06-29 22:02:21 -07:00
Xiang Li
2fb8347d36 etcdserver: add rpc proto 2015-06-29 20:00:09 -07:00
Xiang Li
436bacd77a *: introduce grpc dependency 2015-06-29 18:59:00 -07:00
Xiang Li
718cb18ca2 Merge pull request #3079 from xiang90/gogo
*: resolve proto warnings
2015-06-29 18:50:49 -07:00
Xiang Li
581ef05bab *: resolve proto warnings 2015-06-29 18:39:46 -07:00
Xiang Li
621b43bacb Merge pull request #3078 from xiang90/gogo
update gogoprotobuf dependency
2015-06-29 16:59:08 -07:00
Xiang Li
13f44e4b79 *: update generated proto code 2015-06-29 16:45:25 -07:00
Xiang Li
59b479e59b godep: update gogo version 2015-06-29 16:08:04 -07:00
Yicheng Qin
7f95780bfb etcdserver: init raft internal var early
Its `stopped`/`done` should be created always before being used
in defer in server loop.

It fixes the race detected when running TestSyncTrigger.
2015-06-29 15:34:15 -07:00
Yicheng Qin
235aef5365 *: bump to v2.1.0-rc.0+git 2015-06-29 14:02:40 -07:00
Yicheng Qin
00c32ef022 *: bump to v2.1.0-rc.0 2015-06-29 14:02:39 -07:00
Yicheng Qin
9884c9d977 Merge pull request #3075 from yichengq/fix-windows
Godeps/capnslog: bump to 99f6e6b8f8ea30b0f82769c1411691c44a66d015
2015-06-29 14:02:16 -07:00
Yicheng Qin
207b67c72a Godeps/capnslog: bump to 99f6e6b8f8ea30b0f82769c1411691c44a66d015
It fixes windows building problem.
2015-06-29 13:47:21 -07:00
Xiang Li
433f2ee1bc storage: correctly restore create and ver
Add a restore func to correctly restore create reversion and
version of keys for the index.
2015-06-29 13:44:43 -07:00
Xiang Li
8d3e3ff25a Merge pull request #3073 from xiang90/storage_ver
storage: save version
2015-06-29 13:19:02 -07:00
Xiang Li
ccca2b04da storage: save version 2015-06-29 13:15:09 -07:00
Yicheng Qin
bd84e678e6 Merge pull request #3061 from yichengq/fix-stream-test
rafthttp: fix TestStream uses outdated stream
2015-06-29 11:15:29 -07:00
Yicheng Qin
f421eaeff7 Merge pull request #3071 from yichengq/rename-rafthttp-metrics
rafthttp: message_sent_latency metrics: channel -> sendingType
2015-06-29 10:58:36 -07:00
Xiang Li
e01d53b853 Merge pull request #2979 from xiang90/fix_sendapp
raft: fix panic in send app
2015-06-29 10:49:04 -07:00
Yicheng Qin
28342ae097 rafthttp: avoid TestStream to use outdated stream
The original test code before fb4b0b5cf0
doesn't work because reader side may update the
stream, while writer side writes message to the old stream and fails.

This PR removes unnecessary call to set term, and avoids this problem to
happen on term > 1 in the future.
2015-06-29 10:46:54 -07:00
Yicheng Qin
2afa6688ab Merge pull request #3069 from yichengq/init-term
rafthttp: support to init term when adding peer
2015-06-29 10:45:53 -07:00
Yicheng Qin
606876154d rafthttp: message_sent_latency metrics: channel -> sendingType
Better naming.
2015-06-29 10:44:40 -07:00
Yicheng Qin
4430a80c0f Merge pull request #3063 from yichengq/fix-create-root
etcdserver/auth: fix return value when creating root user
2015-06-29 10:29:23 -07:00
Yicheng Qin
bb287fa22e Merge pull request #3051 from yichengq/doc-rafthttp-metrics
docs: doc metrics used in rafthttp package
2015-06-29 10:22:50 -07:00
Yicheng Qin
fb4b0b5cf0 rafthttp: support to init term when adding peer
So it doesn't need to build term-0 stream with the remote first, then update it.
2015-06-29 10:20:48 -07:00
Yicheng Qin
2e41b4f9e1 etcdserver/auth: fix return value when creating root user
Before:

```
$ curl http://127.0.0.1:4001/v2/auth/users/root -XPUT -d '{"user": "root",
"password": "root"}'
{"user":"root","roles":null}
```

After:

```
{"user":"root","roles":["root"]}
```
2015-06-27 23:16:54 -07:00
Xiang Li
c069119abe Merge pull request #3067 from xiang90/storage_created_mod
storage: save created index and modified index
2015-06-27 23:11:05 -07:00
Yicheng Qin
fcdd9779e9 docs: explain label in rafthttp metrics 2015-06-26 15:51:39 -07:00
Xiang Li
4581064060 storage: save created index and modified index 2015-06-26 12:10:26 -07:00
Yicheng Qin
3e455ed104 Merge pull request #3062 from yichengq/fix-auth-doc
docs: fix typos in auth_api.md
2015-06-25 17:54:05 -07:00
Yicheng Qin
9c695dce25 docs: fix typos in auth_api.md 2015-06-25 17:37:16 -07:00
Barak Michener
acca9cc3a9 Merge pull request #3047 from barakmich/auth_cov
auth: improve test coverage
2015-06-25 14:47:22 -04:00
Barak Michener
39c10d1fe4 auth: improve test coverage 2015-06-25 14:25:08 -04:00
Yicheng Qin
3d4642c2c4 Merge pull request #3059 from yichengq/fix-wait-stress-test
pkg/wait: extend timeout to check closed channel
2015-06-25 11:16:54 -07:00
Xiang Li
35d0839909 Merge pull request #3057 from yichengq/fix-snap-test
etcdserver: fix TestTriggerSnap
2015-06-25 10:51:36 -07:00
Xiang Li
a347e1ecf5 Merge pull request #3058 from yichengq/fix-purge
pkg/fileutil: fix TestPurgeFile
2015-06-25 10:50:36 -07:00
Yicheng Qin
eea7f28be4 pkg/wait: extend timeout to check closed channel
It is possible to trigger the time.After case if the timer went off
between time.After setting the timer for its channel and the time that
select looked at the channel. So it needs to be longer.

refer: https://groups.google.com/forum/#!topic/golang-nuts/1tjcV80ccq8
2015-06-25 10:43:12 -07:00
Yicheng Qin
4c8408f92f docs: doc metrics used in rafthttp package 2015-06-25 10:38:36 -07:00
Yicheng Qin
107263ef9f pkg/fileutil: fix TestPurgeFile
It needs to wait longer for file to be detected and removed sometimes.
2015-06-25 10:09:20 -07:00
Yicheng Qin
5d131acfba etcdserver: fix TestTriggerSnap
Before checking, it needs to wait for snapshot goroutine to finish its
work.
2015-06-25 09:58:36 -07:00
Xiang Li
2ace10626d Merge pull request #3050 from yichengq/doc-bench-tool
docs/benchmarks: doc benchmark tool
2015-06-25 08:15:02 -07:00
Chance Zibolski
ca2ea1bc7d Merge pull request #3048 from ecnahc515/documentation_improvements
Documentation: Refer back between name and initial-cluster options
2015-06-24 15:22:46 -07:00
Yicheng Qin
0949cc06e6 docs/benchmarks: doc benchmark tool 2015-06-24 15:11:08 -07:00
Chance Zibolski
ea2c203aee Documentation: Refer back between name and initial-cluster options 2015-06-24 14:10:42 -07:00
Xiang Li
44fda7985a Merge pull request #3046 from xiang90/metrics
refactor metrics
2015-06-24 13:58:28 -07:00
Xiang Li
9aeb181d75 snap: add namespace and subsystem fields for metrics 2015-06-24 13:46:43 -07:00
Xiang Li
c221844d6b Merge pull request #3024 from xiang90/fix_discovery
discovery: add timeouts for discovery client
2015-06-24 13:44:21 -07:00
Xiang Li
52c2a5731f etcdserver: fix typo in metrics.go 2015-06-24 12:42:40 -07:00
Xiang Li
b3cb5f9e4e Merge pull request #3043 from xiang90/update_auth_doc
auth: update the auth doc
2015-06-23 23:19:02 -07:00
Xiang Li
96c0c7a202 Merge pull request #3044 from xiang90/fix_auth_update_role
auth: do not allow update root role
2015-06-23 22:43:28 -07:00
Xiang Li
030d1bbf2d auth: do not allow update root role 2015-06-23 20:15:08 -07:00
Xiang Li
403fad14ae auth: update the auth doc 2015-06-23 20:02:48 -07:00
Xiang Li
c0b5cc6c52 Merge pull request #3041 from xiang90/auth_u
etcdhttp: improve user endpoint validation
2015-06-23 15:58:03 -07:00
Xiang Li
94f8152487 Merge pull request #3042 from yichengq/fix-addr-in-use
integration: fix bind-addr-in-use
2015-06-23 15:57:50 -07:00
Yicheng Qin
88b69a5979 Merge pull request #3030 from yichengq/fix-fallback-case
etcdmain: fix the check in fallback-to-proxy case
2015-06-23 14:48:45 -07:00
Yicheng Qin
8e79fd85cb integration: fix bind-addr-in-use
The bug happens when restarted member wants to listen on its original
port, but finds out that it has been occupied by some client.

Use well-known port instead of ephemeral port, so client cannot occupy
the listen port anymore.
2015-06-23 14:47:21 -07:00
Xiang Li
e291dfd748 etcdhttp: improve user endpoint validation
Giving both roles and grant/revoke is not allowed.
Creating an existing user is not allowed.
Updating a non-existing user is not allowed.
2015-06-23 14:38:44 -07:00
Yicheng Qin
2d426b518a Merge pull request #3035 from yichengq/update-term
rafthttp: update term when AddPeer
2015-06-23 14:05:37 -07:00
Xiang Li
37933cffa4 Merge pull request #3040 from xiang90/fix_auth
Fix auth
2015-06-23 13:47:25 -07:00
Yicheng Qin
cf050ee21d Merge pull request #2943 from yichengq/fix-client-test
client: fix TestSimpleHTTPClientDoCancelContextResponseBodyClosed
2015-06-23 13:43:07 -07:00
Yicheng Qin
e25e368321 rafthttp: update term when AddPeer
Update term when AddPeer, or the term in peer will not be updated until
the term is changed. This fixes the log flood happended when the v2.1
follower applies the snapshot from v2.0 leader:

```
rafthttp: cannot attach out of data stream server [0 / 17]
```
or
```
rafthttp: server streaming to 6e3bd23ae5f1eae0 at term 0 has been
stopped
```
2015-06-23 13:42:21 -07:00
Xiang Li
c8628c8fe5 auth: separate the role create and update path
Giving both permission and grant/revoke is not allowed.
Creating an existing role is not allowed.
Updating a non-existing is not allowed.
2015-06-23 13:15:32 -07:00
Yicheng Qin
36c5fd6265 etcdmain: fix the check in fallback-to-proxy case
advertise-client-urls has to be set if listen-client-urls is set when
fallbacking to proxy, which breaks the behavior. Loosen the check to fix
it.
2015-06-23 13:08:56 -07:00
Xiang Li
bc61056912 etcdhttp: use correct http status const when writing http error 2015-06-23 12:40:30 -07:00
Xiang Li
4f47a6ebfb Merge pull request #3032 from xiang90/refactor_update_role
auth: refactor updateRole
2015-06-23 11:17:45 -07:00
Xiang Li
240e121792 Merge pull request #3039 from xiang90/update_auth
doc: update auth_api.md
2015-06-23 11:12:47 -07:00
Xiang Li
aaf802f321 doc: update auth_api.md 2015-06-23 11:08:04 -07:00
Xiang Li
ad7124599d Merge pull request #3033 from barakmich/strip_pass
etcdhttp: Always strip password hash when returning users
2015-06-22 18:39:50 -07:00
Xiang Li
7f7e2cc79d Merge pull request #3034 from philips/replace-maximal-with-maximum
*: docs and code %s%maximal%maximum%g
2015-06-22 16:24:01 -07:00
Brandon Philips
740187f199 *: docs and code %s%maximal%maximum%g
maximum is a more common word, use it instead
2015-06-22 16:06:57 -07:00
Yicheng Qin
028a1d6dd4 Merge pull request #2994 from webner/feature/cancel-proxy-request
proxy: handle canceled proxy request gracefully
2015-06-22 16:06:05 -07:00
Barak Michener
d5a0e3ac6a etcdhttp: Always strip password hash when returning users 2015-06-22 18:39:16 -04:00
Xiang Li
979f531261 auth: refactor updateRole
We will return error if revoke or grant fails to update the role.
No need to check if revoke or grant is nil or not.
2015-06-22 15:16:10 -07:00
Xiang Li
462baedcd4 Merge pull request #3031 from xiang90/fix_auth
auth: do not allow to grant duplicate role or revoke ungranted role
2015-06-22 15:13:26 -07:00
Xiang Li
3f82e7b116 auth: do not allow to grant duplicate role or revoke ungranted role to a user 2015-06-22 15:11:09 -07:00
Barak Michener
51a65599dd Merge pull request #3021 from xiang90/auth_err
etcdserver: use correct http status code for auth error
2015-06-22 14:58:33 -04:00
Xiang Li
c39aad0e92 etcdserver: use correct http status code for auth error 2015-06-22 09:28:47 -07:00
Xiang Li
3e4479b0cd Merge pull request #3022 from xiang90/aut_type
etcdhttp: fix the response type for auth
2015-06-21 15:06:35 -07:00
Xiang Li
ebd4102578 Merge pull request #3026 from xiang90/better_logging
etcdserver: better log message for url mismatch
2015-06-19 19:39:33 -07:00
Xiang Li
d295d21349 etcdserver: better log message for url mismatch 2015-06-19 19:36:26 -07:00
Xiang Li
1381b44adf discovery: add timeouts for discovery client 2015-06-19 16:50:44 -07:00
Xiang Li
cad757efa0 etcdhttp: fix the response type for auth 2015-06-19 15:19:00 -07:00
Xiang Li
b26b827780 Merge pull request #3020 from xiang90/auth_doc
auth: minor fix for user section
2015-06-19 15:08:51 -07:00
Xiang Li
b1dbab2b6b auth: minor fix for user section 2015-06-19 14:30:04 -07:00
Xiang Li
9f984ea6ae Merge pull request #3015 from xiang90/auth_doc
doc: move enable section to the top in auth_api.md
2015-06-19 14:13:19 -07:00
Xiang Li
4f0f57b322 doc: move enable section to the top in auth_api.md 2015-06-19 14:08:29 -07:00
Brandon Philips
7ee4fb6181 Merge pull request #3011 from philips/fixup-discovery-info-output
discovery: fixup logline
2015-06-19 13:25:08 -04:00
Brandon Philips
e71dc2e565 discovery: fixup logline
before:

```
discovery: duringcluster status checkconnection tohttps://discovery.etcd.iotimed out, retrying in2s
```

after:

```
discovery: cluster status check: connection to https://discovery.etcd.io timed out, retrying in 2s
```
2015-06-19 13:19:09 -04:00
Wolfgang Ebner
a6e6186477 proxy: always set requestClosed flag when client closes the connection prematurely 2015-06-19 08:45:45 +02:00
Yicheng Qin
5787fabe5f Merge pull request #3008 from yichengq/storage-index-test
storage: add range and tombstone test for index
2015-06-18 19:29:31 -07:00
Yicheng Qin
b20598eea0 storage: add range and tombstone test for index 2015-06-18 18:05:37 -07:00
Xiang Li
1a7a5fd45d Merge pull request #3006 from yichengq/storage-kvstore-test
storage: remove unnecessary ForceCommit in kvstore.Close
2015-06-18 13:57:27 -07:00
Yicheng Qin
9f2e4c8a57 storage: remove unnecessary ForceCommit in kvstore.Close
s.b.Close will commit pending ops, so there is no need to FroceCommit
it in kvstore.Close()
2015-06-18 13:36:23 -07:00
Yicheng Qin
789e2f3426 Merge pull request #3003 from yichengq/storage-kvstore-test
storage: add restore test and fix some bug
2015-06-18 12:19:05 -07:00
Yicheng Qin
7cba42fb73 storage: wait for compact goroutine to exit before close backend
If backend is closed, the operations on backend in compact
goroutine will panic. So this PR waits for compact goroutine to exit
before close backend.

This fixes the TestWorkflow failure too.
2015-06-18 12:18:39 -07:00
Wolfgang Ebner
5e31854afd proxy: use atomic operations for requestCanceled flag 2015-06-18 20:56:28 +02:00
Wolfgang Ebner
864ce5f946 proxy: handle canceled proxy request gracefully
when a client of the proxy server cancels a request the proxy should not
set the endpoint state to unavailable
2015-06-18 20:52:52 +02:00
Yicheng Qin
148394f66f storage: fix schedule compaction bug in recover process
It uses wrong schedule compaction reversion before.
2015-06-18 11:11:37 -07:00
Yicheng Qin
26a09d8479 storage: enhance TestRestore and kill TODO 2015-06-18 10:37:12 -07:00
Yicheng Qin
0ef53ee500 Merge pull request #2999 from yichengq/storage-rev-test
storage: add reversion test
2015-06-18 07:39:18 -07:00
Yicheng Qin
74fbf9d6a7 storage: add reversion test 2015-06-17 18:06:42 -07:00
Yicheng Qin
06ca914429 Merge pull request #2998 from yichengq/storage-kvstore-test
storage: add kv range test
2015-06-17 17:49:55 -07:00
Yicheng Qin
80a59f00b7 storage: fix limit mismatch in Range func 2015-06-17 17:43:08 -07:00
Yicheng Qin
93f477944b storage: return ErrFutureRev if rev is a future one 2015-06-17 17:42:43 -07:00
Yicheng Qin
94924d04db storage: add TestRangeBadRev 2015-06-17 16:22:28 -07:00
Yicheng Qin
9ad5e1e64f storage: kill TODO in TestRange 2015-06-17 15:58:28 -07:00
Yicheng Qin
05228729a3 Merge pull request #2996 from yichengq/storage-workflow-test
storage: add TestWorkflow
2015-06-17 15:05:12 -07:00
Yicheng Qin
500894dfe5 storage: add TestWorkflow 2015-06-17 14:38:21 -07:00
Yicheng Qin
7b1a93e1ef storage: put storage info keys into information bucket
They used to be in key bucket, and make recover failed because they
cannot be parsed as normal key.
2015-06-17 14:37:29 -07:00
Yicheng Qin
d0f6432b51 *: bump to v2.1.0-alpha.1+git 2015-06-16 22:02:00 -07:00
Yicheng Qin
c4a5088bbc *: bump to v2.1.0-alpha.1 2015-06-16 22:00:17 -07:00
Xiang Li
2efbc76689 Merge pull request #2993 from xiang90/md
doc: add doc for metrics feature
2015-06-16 14:22:16 -07:00
Xiang Li
c599e81d46 doc: add proposal into glossary.md 2015-06-16 14:19:18 -07:00
Xiang Li
5c1d4544fc doc: add doc for metrics feature 2015-06-16 14:18:22 -07:00
Barak Michener
cdcae2d6a5 Merge pull request #2991 from barakmich/security_rename
*: Rename `security` to `auth`
2015-06-16 14:41:34 -04:00
Yicheng Qin
7716bdf981 client: fix TestSimpleHTTPClientDoCancelContextResponseBodyClosed
This fixes the bug that the test may hang forever because RoundTrip is
blocked. fixes #2449
2015-06-16 11:29:54 -07:00
Wolfgang Ebner
aeeae25d87 proxy: documentation for disabling the proxy timeout 2015-06-16 12:18:16 +02:00
Wolfgang Ebner
5854d0e8a9 proxy: removed unused refreshInterval variable in director structure 2015-06-16 12:17:08 +02:00
Barak Michener
64ec8af91b *: Rename security to auth 2015-06-15 18:18:50 -04:00
Xiang Li
b4022899eb raft: fix panic in send app
sendApp accesses the storage several times. Perviously, we
assume that the storage will not be modified during the read
opeartions. The assumption is not true since the storage can
be compacted between the read operations. If a compaction
causes a read entries error, we should not painc. Instead, we
can simply retry the sendApp logic until succeed.
2015-06-15 14:23:33 -07:00
Xiang Li
e20b487904 Merge pull request #2978 from xiang90/fix_backup
*:fix point-in-time backup
2015-06-15 13:19:29 -07:00
Xiang Li
f59da0e453 *:fix point-in-time backup
Backup process should be able to read all WALs until io.EOF to
generate a point-in-time backup.

Our WAL file is append-only. And the backup process will lock all
files before start reading, which can prevent the gc routine from
removing any files in the middle.
2015-06-15 11:12:28 -07:00
Xiang Li
b69d52e5ac Merge pull request #2988 from xiang90/raft-doc
raft: fix usage section of doc
2015-06-15 10:39:40 -07:00
Xiang Li
2f0169c3ab raft: fix usage section of doc
We recently added a config struct to start raft. Update
our doc accordingly.
2015-06-15 10:26:10 -07:00
Yicheng Qin
5618adff99 Merge pull request #2977 from nikfoundas/patch-1
docs: add etcd-viewer into libraries-and-tools.md
2015-06-14 08:53:06 -07:00
Xiang Li
3fc8d48421 Merge pull request #2982 from aybabtme/etcdserver/wrong-log-func
etcdserver: use Infof to print formatted argument
2015-06-14 06:53:20 -07:00
Antoine Grondin
270487d340 etcdserver: use Infof to print formatted argument 2015-06-14 20:22:21 +07:00
nikfoundas
dadbc03171 docs: add etcd-viewer into libraries-and-tools.md
I've been working on this project for a few weeks and I believe it has some features that could assist maintaining etcd registries. Please check it out and I hope you would like to include it in your list of etcd tools.
Kind regards,
Nikos
2015-06-14 02:25:42 +03:00
Wolfgang Ebner
1264dbe24d proxy: added endpoint refresh and timeout configuration values
the default dial timeout was set to 30 seconds this made the proxy a pain to use
in failure scenarios.

fixes 2862
2015-06-13 09:42:18 +02:00
Xiang Li
8e7fa9e201 Merge pull request #2976 from yichengq/fix-lock-test
pkg/fileutil: wait longer for relock
2015-06-12 15:20:18 -07:00
Yicheng Qin
7723b91c06 pkg/fileutil: wait longer for relock
multiple cpu running makes it slower, so it waits longer for relock.
2015-06-12 15:17:28 -07:00
Yicheng Qin
219d304291 Merge pull request #2968 from yichengq/fix-stream-reader-init
rafthttp: always init streamReader before return from newPeer
2015-06-12 14:51:05 -07:00
Xiang Li
288cce0d76 Merge pull request #2975 from yichengq/fix-purge-test
pkg/fileutil: wait longer before checking purge results
2015-06-12 14:38:55 -07:00
Yicheng Qin
7ff1fa36f2 rafthttp: always init streamReader before return from newPeer
Or etcd will panic if someone calls `setTerm()`, which uses streamReader
internally, before streamReader is inited.
2015-06-12 14:38:14 -07:00
Yicheng Qin
75f91bab5c pkg/fileutil: wait longer before checking purge results
multiple cpu running may be slower than single cpu running, so it may
take longer time to remove files.
Increase from 5ms to 20ms to give it enough time.
2015-06-12 14:36:15 -07:00
Xiang Li
684c721307 Merge pull request #2970 from yichengq/fix-stream-test
rafthttp: use buffered channel as recv/prop chan
2015-06-12 14:34:52 -07:00
Yicheng Qin
dccec11bb4 Merge pull request #2973 from yichengq/fix-recv-log
rafthttp: fix the misformat logging line, and rename internal var for more clarity
2015-06-12 14:27:17 -07:00
Yicheng Qin
36f75cf062 rafthttp: use buffered channel as recv/prop chan
So it ensures that the message will not be discarded because the receive
side has not been ready, which happens easily in multiple core test.

Use log.fatal instead of log.error. The test exits when there is
something wrong because the error may affect following test cases.
2015-06-12 14:25:11 -07:00
Yicheng Qin
2f05b24d6d rafthttp: {from, to} -> {local, remote} in stream
{from, to} nameings are confused when it both dials and receives
messages from the remote. Change it to {local, remote} for better
clarity.
2015-06-12 14:17:30 -07:00
Yicheng Qin
bcc1aadea9 rafthttp: fix the misformat logging line
before:
```
2015/06/12 20:06:19 rafthttp: dropped MsgApp from %!s(uint64=2) since
receiving buffer is full
```

after:
```
2015/06/12 13:51:38 rafthttp: dropped MsgProp from 2 since receiving
buffer is full
```
2015-06-12 14:12:49 -07:00
Yicheng Qin
ae42371ee2 Merge pull request #2965 from yichengq/fix-issue2904
integration: fix TestIssue2904 in multiple cores
2015-06-12 13:53:00 -07:00
Xiang Li
b98aa3a9e0 Merge pull request #2972 from yichengq/test-longer
test: extend integration timeout to 10m
2015-06-12 13:42:54 -07:00
Yicheng Qin
768cb437bc test: extend integration timeout to 10m
We test with `-cpu 1,2,4` now, and it takes longer time.
2015-06-12 13:41:35 -07:00
Yicheng Qin
796d99c390 integration: fix TestIssue2904 when multiple cores
Do not wait for the cluster view of removed member to match with
expected view, since removed member does not apply entries after it is
removed.
2015-06-12 10:20:27 -07:00
Yicheng Qin
ea3c7d1d31 Merge pull request #2960 from yichengq/fix-drop-flood
rafthttp: pretty print message drop info
2015-06-12 09:23:23 -07:00
Yicheng Qin
0de0e4b77c rafthttp: pretty print message drop info 2015-06-12 09:14:53 -07:00
Yicheng Qin
e46fa0a213 Merge pull request #2957 from yichengq/fix-pipeline-test
rafthttp: fix TestStopBlockedPipeline
2015-06-12 08:03:12 -07:00
Yicheng Qin
c21cc5b39b rafthttp: fix TestStopBlockedPipeline
Refactor the fake cancel implementation.

The old one may cancel other in-flight message in random, which leaves
the original target message blocked forever.
2015-06-12 07:55:12 -07:00
Yicheng Qin
29dca49cb5 rafthttp: wait 1ms before enabling cancel
CancelRequest only effects on in-flight request, so we need to wait
for Do(request) called before enabling cancel.
2015-06-12 07:55:06 -07:00
Xiang Li
d8e1950d4e Merge pull request #2963 from xiang90/fix_discovery_error
etcdmain: exit if discovery fails
2015-06-11 16:11:59 -07:00
Xiang Li
6c8b32d316 etcdmain: exit if discovery fails
Fix #2919

If discovery fails, etcd will hang there and does nothing. This
commit fixes the problem.
2015-06-11 15:45:00 -07:00
Xiang Li
3e706c745c Merge pull request #2953 from yichengq/etcdmain-plog
etcdmain: var log -> plog
2015-06-11 15:30:18 -07:00
Xiang Li
1c19eb47b5 Merge pull request #2956 from xiang90/log
all pkgs use leveled log
2015-06-11 15:29:44 -07:00
Xiang Li
2c5ab7ff8b discovery: fix infoln -> info 2015-06-11 14:22:14 -07:00
Xiang Li
8ad7ed321e *:godep log pkg 2015-06-11 14:22:14 -07:00
Xiang Li
2373fd8426 wal: fix the left logging using default log 2015-06-11 14:22:14 -07:00
Xiang Li
2db8b53c4b discovery: use leveled log 2015-06-11 14:22:14 -07:00
Xiang Li
f013a627a4 etcdserver/stats: use leveled log 2015-06-11 14:22:14 -07:00
Xiang Li
cf7cb2b8a9 etcdserver/security: use leveled log 2015-06-11 14:22:14 -07:00
Xiang Li
2f795e42d0 httptypes: use leveled log 2015-06-11 14:19:53 -07:00
Xiang Li
4b5dbeff9b pkg/pbutil: use leveled log 2015-06-11 14:19:53 -07:00
Xiang Li
865a5ffc61 pkg/osutil: use leveled log 2015-06-11 14:19:53 -07:00
Xiang Li
a45f53986f pkg/netutil: use leveled log 2015-06-11 14:19:52 -07:00
Xiang Li
69819d334a pkg/flags: use leveled log 2015-06-11 14:19:52 -07:00
Barak Michener
7bf0479e66 Merge pull request #2882 from barakmich/security_client_new
*: Add security/authorization to etcd/client and etcdctl
2015-06-11 13:40:32 -04:00
Yicheng Qin
1764837783 etcdmain: clean up plog.Printf
Put it into different log levels.
2015-06-11 10:24:02 -07:00
Yicheng Qin
ecdf0a8146 Merge pull request #2959 from yichengq/fix-update-member
rafthttp: fix TestUpdateMember
2015-06-11 10:02:13 -07:00
Yicheng Qin
1af2b4cad7 rafthttp: fix TestUpdateMember
Before this PR, it may error like this:

```
--- FAIL: TestUpdateMember-2 (0.00s)
		server_test.go:950: action =
		[{ApplyConfChange:ConfChangeUpdateNode []}
{ProposeConfChange:ConfChangeUpdateNode []}], want
[{ProposeConfChange:ConfChangeUpdateNode []}
{ApplyConfChange:ConfChangeUpdateNode []}]
```

This fixes the test by recording the proposal event in time.
2015-06-11 09:45:34 -07:00
Yicheng Qin
cd629c9b44 Merge pull request #2939 from yichengq/fix-update-attr
etcdserver: allow to update attributes of removed member
2015-06-10 16:53:39 -07:00
Yicheng Qin
8725e69cf7 etcdserver: allow to update attributes of removed member
There exist the possiblity to update attributes of removed member in
reasonable workflow:
1. start member A
2. leader receives the proposal to remove member A
2. member A sends the proposal of update its attribute to the leader
3. leader commits the two proposals
So etcdserver should allow to update attributes of removed member.
2015-06-10 16:52:18 -07:00
Xiang Li
743ac73b11 Merge pull request #2954 from xiang90/fix_test
proxy: fix test
2015-06-10 16:44:58 -07:00
Yicheng Qin
ed1c5a73d1 Merge pull request #2951 from yichengq/fix-proxy-acurls
etcdmain: fix that advertise-client-urls is required in proxy mode
2015-06-10 16:42:06 -07:00
Xiang Li
612ecbc89d proxy: fix test 2015-06-10 16:31:42 -07:00
Yicheng Qin
cf7c83b304 etcdmain: fix that advertise-client-urls is required in proxy mode
etcd proxy doesn't need to set advertise-client-urls because the flag is
not used.
2015-06-10 16:22:32 -07:00
Yicheng Qin
5a9c2851a7 etcdmain: var log -> plog
So the variable name doesn't mess up with standard package name.
2015-06-10 16:19:06 -07:00
Yicheng Qin
0a3a2720a1 Merge pull request #2923 from yichengq/rafthttp-status
rafthttp: pretty print connection error
2015-06-10 16:17:07 -07:00
Xiang Li
f64a8214f7 Merge pull request #2952 from xiang90/fileutil
fileutil: use leveled logging
2015-06-10 16:01:24 -07:00
Xiang Li
dc87454487 fileutil: return on error and send it to error chan 2015-06-10 15:59:24 -07:00
Xiang Li
e2c2f098bc fileutil: use leveled logging 2015-06-10 15:57:59 -07:00
Yicheng Qin
d92c89516b rafthttp: fix capnslog package name 2015-06-10 15:43:54 -07:00
Yicheng Qin
1dbe72bb74 rafthttp: pretty print connection error
1. print out the status change of connection with peer
2. only print the first error for repeated ones
2015-06-10 15:43:49 -07:00
Yicheng Qin
30db41e031 Procfile: use -listen-client-urls instead of -bind-addr
-bind-addr is etcd 0.4 flag, and we should deprecate it.

Moreover, this makes Procfile fit the workflow we mention in the doc,
which helps ourselves find the problem first.
2015-06-10 15:13:33 -07:00
Xiang Li
37f9534109 Merge pull request #2950 from xiang90/test_cpu
test: run with cpu = 1,2,4
2015-06-10 15:09:45 -07:00
Yicheng Qin
4e79abcfeb Merge pull request #2944 from yichengq/fix-2procs
pkg/testutil: ForceGosched -> WaitSchedule
2015-06-10 14:44:32 -07:00
Yicheng Qin
018fb8e6d9 pkg/testutil: ForceGosched -> WaitSchedule
ForceGosched() performs bad when GOMAXPROCS>1. When GOMAXPROCS=1, it
could promise that other goroutines run long enough
because it always yield the processor to other goroutines. But it cannot
yield processor to goroutine running on other processors. So when
GOMAXPROCS>1, the yield may finish when goroutine on the other
processor just runs for little time.

Here is a test to confirm the case:

```
package main

import (
	"fmt"
	"runtime"
	"testing"
)

func ForceGosched() {
	// possibility enough to sched up to 10 go routines.
	for i := 0; i < 10000; i++ {
		runtime.Gosched()
	}
}

var d int

func loop(c chan struct{}) {
	for {
		select {
		case <-c:
			for i := 0; i < 1000; i++ {
				fmt.Sprintf("come to time %d", i)
			}
			d++
		}
	}
}

func TestLoop(t *testing.T) {
	c := make(chan struct{}, 1)
	go loop(c)
	c <- struct{}{}
	ForceGosched()
	if d != 1 {
		t.Fatal("d is not incremented")
	}
}
```

`go test -v -race` runs well, but `GOMAXPROCS=2 go test -v -race` fails.

Change the functionality to waiting for schedule to happen.
2015-06-10 14:37:41 -07:00
Xiang Li
2d21904cfd test: run with cpu = 1,2,4 2015-06-10 14:26:17 -07:00
Barak Michener
a4d1a5a6e5 *: Add security/auth support to etcdctl and etcd/client
add godep for speakeasy and auth entry parsing
add security_user to client
add role to client
add role commands
add auth support to etcdclient and etcdctl(member/user)
add enable/disable to etcdctl
better error messages, read/write/readwrite
Bump go-etcd to include codec changes, add new dependency
verify the error for revoke/add if nothing changed, remove security-merging prefix
2015-06-10 16:58:10 -04:00
Xiang Li
97709b202d Merge pull request #2930 from xiang90/storage_restore
storage: initial snapshot and restore
2015-06-10 11:38:57 -07:00
Xiang Li
ba9a46aa02 storage: initial snapshot and restore
Snapshot takes an io.Writer and writes the entire backend data to
the given writer. Snapshot writes a consistent view and does not
block other storage operations.

Restore restores the in-memory states (index and book keeping) of
the storage from the backend data.
2015-06-10 11:32:10 -07:00
Yicheng Qin
1403783326 Merge pull request #2911 from yichengq/rafthttp-plog
rafthttp: use leveled logger
2015-06-09 16:16:33 -07:00
Yicheng Qin
f1e995b070 rafthttp: use leveled logger 2015-06-09 16:15:02 -07:00
Xiang Li
19ef3a0982 Merge pull request #2934 from xiang90/etcdserver_log
etcdserver: use leveled logging
2015-06-09 15:53:52 -07:00
Xiang Li
e0f9796653 etcdserver: use leveled logging
Leveled logging for etcdserver pkg.
2015-06-09 13:53:07 -07:00
Yicheng Qin
9fbd2599ad Merge pull request #2940 from yichengq/improve-raft-loop
etcdserver: stop raft loop when receiving stop signal
2015-06-09 11:24:53 -07:00
Yicheng Qin
0814966ca2 etcdserver: stop raft loop when receiving stop signal
When it waits for apply to be done, it should stop the loop if it
receives stop signal.

This helps to print out panic information. Before this PR, if the panic
happens when server loop is applying entries, server loop will wait for
raft loop to stop forever.
2015-06-09 11:11:53 -07:00
Xiang Li
ebb767765e Merge pull request #2941 from bakins/http-log
Simple debug HTTP request logging
2015-06-09 10:52:13 -07:00
Brian Akins
d8a836e618 Simple debug HTTP request logging 2015-06-09 13:40:37 -04:00
Xiang Li
1ff86556b7 Merge pull request #2937 from xiang90/http_log
etcdhttp: use leveled logging
2015-06-09 09:35:17 -07:00
Xiang Li
0adeee2965 etcdhttp: use leveled logging 2015-06-09 09:26:57 -07:00
Xiang Li
3390f38bba Merge pull request #2925 from yichengq/doc-gomaxprocs
docs: document cpu cores deployment
2015-06-08 13:49:28 -07:00
Yicheng Qin
471cf82905 docs: document maximal OS threads 2015-06-08 12:00:33 -07:00
Xiang Li
e0d5116683 Merge pull request #2926 from xiang90/raft_log
raft: make the repeated log message under bad path debug level
2015-06-08 10:57:12 -07:00
Xiang Li
1279e495f0 raft: make the repeated log message under bad path debug level 2015-06-05 17:29:24 -07:00
Xiang Li
05b55d9d75 Merge pull request #2921 from xiang90/fix_watch_cancel
client: fix cancel watch
2015-06-05 15:46:16 -07:00
Xiang Li
15ac4f08f8 client: fix cancel watch
ioutil.ReadAll is a blocking call, we need to wait cancelation
during the call.
2015-06-05 15:40:43 -07:00
Xiang Li
976ac65c86 Merge pull request #2894 from xiang90/refactor_keyIndex
Storage initial compaction
2015-06-05 12:38:11 -07:00
Xiang Li
511f323424 Merge pull request #2916 from luan/build-script-git-fallback
Unexpected dependency in build script
2015-06-05 10:17:15 -07:00
Connor and Luan Santos
17d5381059 build: default git sha to GitNotFound in case git fails 2015-06-05 10:09:50 -07:00
Xiang Li
f47ed4a364 storage: initial compact 2015-06-05 09:22:44 -07:00
Jonathan Boulle
60ca9ebab1 Merge pull request #2915 from jonboulle/master
docs: readme/branch-management cleanup
2015-06-04 16:54:15 -07:00
Jonathan Boulle
048a948eca docs: readme/branch-management cleanup 2015-06-04 16:41:32 -07:00
Xiang Li
75ddf05ca1 Merge pull request #2910 from xiang90/etcdctl
etcdctl: cleanup
2015-06-03 10:53:13 -07:00
Xiang Li
f9c67daee5 Merge pull request #2912 from xiang90/client-curl
client: support printing cURL command
2015-06-03 10:15:00 -07:00
Xiang Li
4f2df84a38 client: support printing cURL command 2015-06-03 10:02:37 -07:00
Yicheng Qin
9e8d589163 Merge pull request #2906 from yichengq/fix-pipeline-stop
rafthttp: fix pipeline.stop may block
2015-06-03 08:47:17 -07:00
Xiang Li
f0edf06b6d etcdctl: minor cleanup 2015-06-02 19:50:37 -07:00
Xiang Li
079e7c10a0 etcdctl: move format to format.go 2015-06-02 19:29:05 -07:00
Xiang Li
26682b663d etcdctl: cleanup etcdctl exit code 2015-06-02 19:01:41 -07:00
Yicheng Qin
7f8925e172 rafthttp: fix pipeline.stop may block
This PR makes pipeline.stop stop quickly. It cancels inflight requests,
and stops sending messages in the buffer.
2015-06-02 17:15:44 -07:00
Xiang Li
627929d2f4 Merge pull request #2909 from xiang90/logger
*: rename logger to plog
2015-06-02 15:03:28 -07:00
Xiang Li
711451ce2d *: rename logger to plog 2015-06-02 14:58:24 -07:00
Xiang Li
28878e34ff Merge pull request #2903 from xiang90/chord_rafthttp
rafhttp: clean up logging messages
2015-06-02 14:44:40 -07:00
Yicheng Qin
b74082c06c Merge pull request #2889 from yichengq/version-runtime-enforce
rafthttp: version enforcement on rafthttp messages
2015-06-02 14:37:38 -07:00
Yicheng Qin
c371d8c65c rafthttp: version enforcement on rafthttp messages
This PR sets etcd version and min cluster version in request header,
and let server check version compatibility. rafthttp server
will reject any message from peer with incompatible version(too low
version or too high version), and print out warning logs.
2015-06-02 13:33:18 -07:00
Xiang Li
2bf64b4adf Merge pull request #2898 from xiang90/raft_log
raft use leveled logger
2015-06-02 13:04:02 -07:00
Xiang Li
1561b85bf3 raft: drop the raft prefix in logging 2015-06-02 12:50:42 -07:00
Xiang Li
3af4a45d7b etcdserver: make raft use leveled logger 2015-06-02 12:50:42 -07:00
Xiang Li
89f6f988cb Godeps: update logger pkg 2015-06-02 12:50:42 -07:00
Xiang Li
46b5eb051e Merge pull request #2896 from xiang90/wal_log
wal: use leveled logger
2015-06-02 11:39:25 -07:00
Xiang Li
59dd1eeaf0 Merge pull request #2897 from xiang90/snapshot_logger
snap: use leveled logger
2015-06-02 11:39:17 -07:00
Xiang Li
a8af787971 Merge pull request #2902 from BlueDragonX/bug-proxyreq-closed
Reuse a bytes buffer as proxy request body.
2015-06-02 10:37:48 -07:00
Ryan Bourgeois
4e85f932e0 proxy: Reuse a bytes buffer as proxy request body.
The call to transport.RoundTrip closes the request body regardless of
the value of request.Closed. This causes subsequent calls to RoundTrip
using the same request body to fail.

Fixes #2895
2015-06-02 10:27:20 -07:00
Xiang Li
2b5f417113 Merge pull request #2901 from xiang90/fix_urlpick
rafthttp: move mu to the top in urlPicker struct
2015-06-01 23:53:57 -07:00
Yicheng Qin
1cd5c7efee Merge pull request #2900 from yichengq/proxy-maxidle
etcdmain: increase maxIdleConnsPerHost in proxy transport
2015-06-01 23:31:35 -07:00
Xiang Li
a7a4233f0b rafhttp: clean up logging messages 2015-06-01 17:18:37 -07:00
Xiang Li
b660ee408f rafthttp: move mu to the top in urlPicker struct
mutex protects all the fields.
2015-06-01 16:40:18 -07:00
Yicheng Qin
0589afe605 etcdmain: increase maxIdleConnsPerHost in proxy transport
This PR set maxIdleConnsPerHost to 128 to let proxy handle 128 concurrent
requests in long term smoothly.
If the number of concurrent requests is bigger than this value,
proxy needs to create one new connection when handling each request in
the delta, which is bad because the creation consumes resource and may
eat up your ephemeral port.
2015-06-01 16:19:36 -07:00
Xiang Li
ae5f7c943b snap: use leveled logger 2015-06-01 14:07:30 -07:00
Xiang Li
185d2bced4 wal: use leveled logger 2015-06-01 13:38:50 -07:00
Xiang Li
8825af47a0 Merge pull request #2893 from eparis/unfuck-godeps
godeps: fix and update dependencies
2015-06-01 10:16:58 -07:00
Eric Paris
af5286c63b Fix godeps to be usable
Godeps should allow me to do
  godep restore
  godep save -r ./...

But that doesn't work. Try it.

This requires update to the following packages:
github.com/prometheus/client_golang/
github.com/prometheus/procfs
github.com/matttproud/golang_protobuf_extensions/

There were 2 major problems.

1. godeps have code.google.com/p/goprotobuf but that repo doesn't exist
2. prometheus/client_golang/_vendor moved to other packages and godep
(with -r) can't handle it.

At the end of this we should be able to use godeps again without tons of
black magic.  uggh.  what a pain in the ass.

The black magic to actually get godeps back in shape was:

```bash
 # remove code.google.com/p/goprotobuf (doesn't exist)
 # remove all _vendor lines from prometheus (we still have other
 # prometheus lines so restore still works)
vi Godeps/Godeps.json

 # remove all the crazy vendoring crud because godep doesn't handle it
 # correctly
find . -name \*.go | xargs sed -i
's|github.com/coreos/etcd/Godeps/_workspace/src/||'

 # ok now, restore as best we can (everything except it wines about
 # goprotobuf
godep restore

 # now update the packages which were using the old (dead) goprotobuf
go get -u github.com/prometheus/client_golang/
go get -u github.com/matttproud/golang_protobuf_extensions/
 # update prometheus procfs because prometheus/client_golang/ has a
 # dependancy on this update
go get -u github.com/prometheus/procfs

 # get rid of Godeps directory entirely
git rm -rf Godeps

 # ok, now, rewrite the Godeps directory and redo the path rewrites
godep  save -r ./...

 # now put Godeps back into git
git add Godeps/

 # commit the new code
git commit -aA

 # And now, you can use godeps!
godep restore
godep save -r ./...
git diff
 # nothing!!
```
2015-05-31 23:52:16 -04:00
Xiang Li
d417b36e5d storage: refactor key_index 2015-05-31 15:24:04 -07:00
Xiang Li
7735501407 Merge pull request #2874 from xiang90/storeAPI
kv api of stroage
2015-05-31 15:17:20 -07:00
Xiang Li
815fe327dd Merge pull request #2890 from xiang90/fix_raft_comment
raft: remove wrong invariant
2015-05-30 13:28:13 -07:00
Xiang Li
0ca6be31f8 raft: remove wrong invariant
The commit > unstable might not true for follower. The leader only need
to ensure the entry is stored on the majority of nodes to commit an
entry. So the minority of the cluster might receive commit > unstable
append request. This is normal.
2015-05-29 18:48:59 -07:00
Xiang Li
871107c65a Merge pull request #2883 from alexaltair/master
etcdmain: use double-dash in message flag
2015-05-28 14:33:14 -07:00
Xiang Li
4e97305df0 Merge pull request #2878 from xiang90/fix_raft_node
raft: fix raft node start bug
2015-05-28 14:31:25 -07:00
Alex Altair
6f8c36c2ab etcdmain: use double-dash in message flag 2015-05-28 13:09:44 -07:00
Barak Michener
ce5e14e713 Merge pull request #2881 from barakmich/go-etcd-update
Godep: update go-etcd version
2015-05-28 14:10:50 -04:00
Barak Michener
f6f7ef6b3a Godep: update go-etcd version 2015-05-28 14:02:14 -04:00
Xiang Li
6c207b9277 stroage: kill todo 2015-05-27 14:46:59 -07:00
Xiang Li
de1c9c08e1 Merge pull request #2842 from SpencerBrown/SpencerBrown-patch-2
docs: add client flags to examples in clustering.md
2015-05-27 14:28:38 -07:00
Xiang Li
69d02410cf stroage: adopt KV interface 2015-05-27 14:24:23 -07:00
Xiang Li
6f0558b999 Merge pull request #2871 from xiang90/cluster_id
rafthttp: print out log when clusterID mismatch instead of exiting
2015-05-27 13:34:27 -07:00
Xiang Li
085447ed85 raft: fix raft node start bug
raft node should set initial prev hard state to empty.
Or it will not send the first hard coded state to application
until the state changes again.

This commit fixs the issue. It introduce a small overhead, that
the same tate might send to application twice when restarting.
But this is fine.
2015-05-27 13:32:04 -07:00
Xiang Li
cbb8b9bb08 stroage: add tnx id 2015-05-27 10:35:51 -07:00
Xiang Li
7ad2b22498 Merge pull request #2876 from xiang90/little_fix
etcdmian: remove main prefix in logging
2015-05-27 10:11:34 -07:00
Yicheng Qin
1d6e9fd387 Merge pull request #2875 from yichengq/verbose-integration
test: run integration tests in verbose mode
2015-05-27 10:09:56 -07:00
Xiang Li
7875de7d2f etcdmian: remove main prefix in logging
We are using new log pkg, which adds the prefix for us.
2015-05-27 10:01:22 -07:00
Xiang Li
9c1aec6877 storage: add rangeKeys func 2015-05-27 09:58:21 -07:00
Yicheng Qin
fde7a7a10c test: run integration tests in verbose mode
Travis doesn't print out the final result of integration tests
sometimes, and verbose mode helps us debug.
2015-05-27 09:57:44 -07:00
Xiang Li
4e0b28f1ca Merge pull request #2872 from bprashanth/log_gomax
etcdmain: explicitly set gomaxprocs and log its value
2015-05-27 09:57:12 -07:00
Prashanth Balasubramanian
1e15b05e4c etcdmain: explicitly set gomaxprocs and log its value 2015-05-27 09:53:05 -07:00
Xiang Li
fb12a4e412 storage: fix a deadlock in batch tx 2015-05-27 09:31:11 -07:00
Xiang Li
93ecf36855 storage: support tnx 2015-05-27 09:31:11 -07:00
Xiang Li
9db360387d storage: support Range 2015-05-27 09:31:11 -07:00
Xiang Li
7bb388ed52 storage: initial kv api 2015-05-27 09:31:11 -07:00
Xiang Li
9be6a7c8fd Merge pull request #2831 from xiang90/index
storage: initial index and key index
2015-05-27 09:29:42 -07:00
Xiang Li
49da7b6556 storage: add boltdb as dependency 2015-05-27 09:24:49 -07:00
Xiang Li
0d3d4c5b01 rafthttp: print out log when clusterID mismatch instead of exiting
We have heard from several users that they do not expect a clusterID
mismatch to kill the cluster.
2015-05-26 16:05:58 -07:00
Yicheng Qin
5d741e4945 Merge pull request #2797 from yichengq/stream-2.0
rafthttp: try stream msgappV1 handler if msgappV2 is unsupported
2015-05-26 15:09:51 -07:00
Yicheng Qin
19fc1a7137 rafthttp: update streamReader term in time
Because etcd 2.1 will build stream to any existing peers and etcd 2.0
requires the remote to provide most updated term, it is
necessary for streamReader to know the latest term.
2015-05-26 14:52:42 -07:00
Yicheng Qin
fad2c09fa8 rafthttp: not log expected timeout as error
The network timeout from stream with etcd 2.0 is expected because etcd
2.0 doesn't heartbeat on idle connections.
2015-05-26 14:52:41 -07:00
Yicheng Qin
38b8e848ac rafthttp: try stream msgappV1 handler if msgappV2 is unsupported
This helps etcd 2.1 connect to msgappV1 handler when the remote member
doesn't support msgappV2. And it doesn't print out unsupported handler
error to make log clean.
2015-05-26 14:52:41 -07:00
Xiang Li
42fe370b35 Merge pull request #2848 from xiang90/metrics
*: use namespace and subsystem in metrics
2015-05-26 14:44:54 -07:00
Yicheng Qin
60c8719d08 Merge pull request #2782 from yichengq/not-close-stream
rafthttp: only close streamMsgApp when updating term
2015-05-26 14:41:22 -07:00
Xiang Li
34ac145b38 *: use namespace and subsystem in metrics
Fix #2841.

From Prometheus developer:
```
the recommended way for etcd as an open source project and under
consideration of its size would be etcd_<subsystem>_<name>.
```

We made the naming change accordingly.
2015-05-26 14:39:04 -07:00
Xiang Li
3028edd7dc Merge pull request #2856 from xiang90/mrefactor
etcdserver: refactore member.go
2015-05-26 14:37:37 -07:00
Yicheng Qin
4d8be39fd1 Merge pull request #2870 from yichengq/enable-travis-govet
travis: stop install tools cover and vet
2015-05-26 11:59:42 -07:00
Xiang Li
c951c22fff Merge pull request #2861 from barakmich/2859
etcdserver: fix go vet. Fixes #2859
2015-05-26 11:06:55 -07:00
Yicheng Qin
90ad78aa46 travis: stop install tools cover and vet
There is no need to install them separately because they have been
downloaded in the default go root directory.
2015-05-26 11:03:53 -07:00
Xiang Li
1be69b1391 Merge pull request #2864 from schmichael/mention-metafora
docs: mention metafora distributed task library
2015-05-22 13:24:20 -07:00
Michael Schurter
e93242967c docs: mention metafora distributed task library
Metafora uses etcd as a task broker, command channel, and state store.
2015-05-22 13:17:05 -07:00
Spencer Brown
0e49a0a3ef docs: add client flags to examples in clustering.md
to make it a complete functional example
2015-05-22 14:18:14 -05:00
Barak Michener
9ef098c5ed etcdserver: fix go vet. Fixes #2859 2015-05-22 13:54:54 -04:00
Xiang Li
58eefda72d Merge pull request #2840 from yichengq/revert-url-equal
Revert "Treat URLs have same IP address as same"
2015-05-21 19:27:19 -07:00
Xiang Li
4a72d3a8bb etcdserver: refactore member.go 2015-05-21 09:19:29 -07:00
Xiang Li
e332e86b5d storage: address barak's comments 2015-05-20 17:47:35 -07:00
Xiang Li
0ad6d7e3ba Merge pull request #2853 from bdarnell/status
raft: MultiNode.Status returns nil for non-existent groups.
2015-05-20 13:07:23 -07:00
Ben Darnell
d58fac453d raft: MultiNode.Status returns nil for non-existent groups.
Previously it would panic if the group did not exist.
2015-05-20 15:45:38 -04:00
Xiang Li
781eccb337 Merge pull request #2852 from bdarnell/hex-node-id
raft: Format node IDs as hex in DescribeMessage.
2015-05-20 12:34:35 -07:00
Ben Darnell
ef721db247 raft: Format node IDs as hex in DescribeMessage.
This is how they are printed in all other log messages.
2015-05-20 15:32:56 -04:00
Xiang Li
260aad5468 Merge pull request #2830 from xiang90/join_checking
checking cluster version compatibility before joining the existing cluster
2015-05-20 12:25:50 -07:00
Xiang Li
aa417ab644 etcdserver: log the per endpoint error in getVersion 2015-05-20 12:10:10 -07:00
Xiang Li
db7db689a6 etcdserver: check cluster version compability when joining 2015-05-19 10:19:41 -07:00
Xiang Li
845cb61213 storage: add kv and event proto 2015-05-18 14:35:10 -07:00
Barak Michener
00ed4fe778 Merge pull request #2764 from barakmich/2755
security: Lazily create the security directories. Fixes #2755.
2015-05-18 17:34:13 -04:00
Barak Michener
a88a53274f security: Lazily create the security directories. Fixes #2755, may find new instances for #2741
revert the kv integration test

fix nits

amend security mention of GUEST
2015-05-18 17:28:04 -04:00
Xiang Li
6ee5cd9105 Merge pull request #2675 from xiang90/v3rfc
doc: v3api rfc
2015-05-18 13:52:54 -07:00
Xiang Li
7c879ee576 doc: v3api rfc 2015-05-18 13:48:16 -07:00
Yicheng Qin
3153e635d5 Revert "Treat URLs have same IP address as same"
This reverts commit f8ce5996b0.

etcd no longer resolves TCP addresses passed in through flags,
so there is no need to compare hostname and IP slices anymore.
(for more details: a3892221ee)

Conflicts:
	etcdserver/cluster.go
	etcdserver/config.go
	pkg/netutil/netutil.go
	pkg/netutil/netutil_test.go
2015-05-16 03:21:10 -07:00
Marcus Stong
b3e6ad136a docs: add node-etcd-config to libs and tools doc 2015-05-16 02:02:44 -07:00
Xiang Li
9575cc4258 storage: add delete example 2015-05-15 19:33:59 -07:00
Yicheng Qin
2e43ac8463 rafthttp: add test for streamReader.updateMsgAppTerm 2015-05-15 11:21:54 -07:00
Yicheng Qin
8637a4bf69 rafthttp: only close streamMsgApp when updating term
In all stream types, streamMsgApp needs to be closed when
updating term because its stream connection can only be used under
a certain term. But there is no need to close other streams, which
may waste time and reduce performance.
2015-05-15 11:21:54 -07:00
Yicheng Qin
9699a501f3 Merge pull request #2833 from yichengq/rename-closer
rafthttp: resetCloser -> close
2015-05-15 11:18:58 -07:00
Yicheng Qin
8e0992a28b rafthttp: resetCloser -> close
name 'close' is shorter and more straightforward.
2015-05-14 22:24:05 -07:00
Yicheng Qin
fc4543a3fd Merge pull request #2628 from yichengq/improve-msgappv2
rafthttp: reduce allocs in msgappv2
2015-05-14 21:18:16 -07:00
Xiang Li
4b0d9f69c7 storage: add a simple backend and kv example 2015-05-14 20:43:32 -07:00
Yicheng Qin
d611904a41 Merge pull request #2828 from yichengq/cluster-health-log
etcdctl/cluster_health: improve output if failed to get leader stats
2015-05-14 19:01:48 -07:00
Yicheng Qin
3d8fe3b3ca etcdctl/cluster_health: improve output if failed to get leader stats
When failing to get leader stats, it said 'cluster is unhealthy' before.
This is confusing when it cannot get stats because advertised client urls
are set wrong and the cluster is healthy.
2015-05-14 18:52:10 -07:00
Xiang Li
9d831e3075 *: godep btree 2015-05-14 17:59:55 -07:00
Xiang Li
660fd5e3e1 storage: add comment around compact 2015-05-14 17:55:54 -07:00
Xiang Li
ee47973199 storage: initial index 2015-05-14 17:53:41 -07:00
Xiang Li
32d44aa3b2 storage: initial key index 2015-05-14 17:35:12 -07:00
Xiang Li
556713739c Merge pull request #2823 from alexwlchan/master
docs: small fixes to spelling and similar
2015-05-14 15:41:14 -07:00
Xiang Li
9f8342dba4 etcdserver: do not get local version via HTTP 2015-05-13 17:19:32 -07:00
Xiang Li
988c30bfba etcdserver: getVersion returns both server and cluster version 2015-05-13 17:04:46 -07:00
Xiang Li
1a9dcd2f72 Merge pull request #2826 from yichengq/fix-wait-test
pkg/wait: fix TestWaitTestStress
2015-05-13 15:56:26 -07:00
Xiang Li
132b12f8db Merge pull request #2827 from xiang90/cluster_v
etcdhttp: version endpoint also returns cluster version.
2015-05-13 15:54:54 -07:00
Xiang Li
6296054ff6 etcdhttp: version endpoint also returns cluster version. 2015-05-13 15:48:10 -07:00
Yicheng Qin
256a7cfe8c pkg/wait: fix TestWaitTestStress
The test may fail if two consequent time.Now() returns the same value.
Sleep 1ns to avoid this situation.
2015-05-13 13:41:34 -07:00
Yicheng Qin
75ee7f4aa1 Merge pull request #2821 from yichengq/private-cluster
etcdserver: stop exposing Cluster struct
2015-05-13 10:26:48 -07:00
Xiang Li
2690535f8a Merge pull request #2820 from xiang90/cap
version capability checking
2015-05-13 10:16:49 -07:00
Xiang Li
d3b1d5c008 etcdhttp: support capability checking
etcdhttp will check the cluster version and update its
capability version periodically.

Any new handler's after 2.0 needs to wrap by capability handler
to ensure it is not accessable until rolling upgrade finished.
2015-05-13 10:11:35 -07:00
Yicheng Qin
a6a649f1c3 etcdserver: stop exposing Cluster struct
After this PR, only cluster's interface Cluster is exposed, which makes
code much cleaner. And it avoids external packages to rely on cluster
struct in the future.
2015-05-13 10:01:25 -07:00
Xiang Li
19ab1cb2a9 Merge pull request #2822 from xiang90/rm_log
etcdserver: remove unnecessary around detect datadir
2015-05-13 09:27:21 -07:00
Alex Chan
0c63e16ae0 docs: small fixes to spelling and similar
This commit is a collection of fixes to spelling, capitalisation
and spacing. No substantial changes.
2015-05-13 11:45:00 +01:00
Xiang Li
f2905f2828 etcdserver: remove unnecessary around detect datadir
The log is super unhelpful. When I have a 2.1.0 etcd, it prints out
`2.0.1 vaild dir`. I have no idea why the data dir of a 2.1.0 etcd is
2.0.1.
2015-05-12 22:06:42 -07:00
Yicheng Qin
f4c51cb5a1 Merge pull request #2766 from yichengq/345
*: extract types.Cluster from etcdserver.Cluster
2015-05-12 15:52:24 -07:00
Yicheng Qin
032db5e396 *: extract types.Cluster from etcdserver.Cluster
The PR extracts types.Cluster from etcdserver.Cluster. types.Cluster
is used for flag parsing and etcdserver config.

There is no need to expose etcdserver.Cluster public, which contains
lots of etcdserver internal details and methods. This is the first step
for it.
2015-05-12 14:53:11 -07:00
Xiang Li
197437316f Merge pull request #2804 from xiang90/vv
etcdserver: support update cluster version through raft
2015-05-12 14:31:27 -07:00
Xiang Li
e866314b94 etcdserver: support update cluster version through raft
1. Persist the cluster version change through raft. When the member is restarted, it can recover
the previous known decided cluster version.

2. When there is a new leader, it is forced to do a version checking immediately. This helps to
update the first cluster version fast.
2015-05-12 11:44:34 -07:00
Yicheng Qin
f1502e970a Merge pull request #2813 from sckott/r-library
Documentation: add the R client etseed to libraries-and-tools.md
2015-05-12 10:43:59 -07:00
Xiang Li
93b610ac8d Merge pull request #2809 from xiang90/fix_discovery_err
discovery: do not return raw error from etcd store
2015-05-12 09:59:56 -07:00
Nick Owens
b764f07e34 Merge pull request #2811 from mischief/plan9-lock
pkg/fileutil: add plan9 lockfile support
2015-05-11 17:42:40 -07:00
Xiang Li
91cbf47a2a etcdmain: better error msg when detected duplicate id in discovery 2015-05-11 17:34:44 -07:00
Scott Chamberlain
5203de5566 Documentation: add the R client etseed to libraries-and-tools.md
etseed is an R client for etcd.
2015-05-11 15:43:31 -07:00
mischief
2e8c932ab0 pkg/fileutil: add plan9 lockfile support 2015-05-11 13:24:01 -07:00
Xiang Li
e9931fb8b1 discovery: do not return error from etcd
We used to return `key not found` directly to the
user due to a bug. We fixed the bug and added a test
case in this commit.
2015-05-11 10:49:57 -07:00
Yicheng Qin
3d242695b3 Merge pull request #2775 from yichengq/proxy-doc
docs: proxy needs accessible advertise client urls
2015-05-10 10:18:40 -07:00
Xiang Li
42783c1faa Merge pull request #2805 from MSamman/more_version_info
version: added more version information
2015-05-08 20:28:10 -07:00
Mohammad Samman
3914defd8a version: added more version information
added more version information output to aid debugging
print etcd Version, Git SHA, Go runtime version, OS
and architecture

Fixes #2560
2015-05-09 03:21:10 +00:00
Yicheng Qin
1abf2636b5 docs: proxy needs accessible advertise client urls
Users cannot use proxy if -advertise-client-urls is set correctly.
Especially mention this in the doc to help them bypass the wrong
settings.
2015-05-07 22:53:42 -07:00
Xiang Li
b24dd8e4e6 Merge pull request #2792 from ecnahc515/client_create_dir
client: Support creating directory through KeysAPI
2015-05-07 11:13:49 -07:00
Chance Zibolski
48e144ae2e client: Support creating directory through KeysAPI
Creating a directory is done using the Set() method and a SetOptions
struct with it's Dir field set to true.
2015-05-07 10:47:18 -07:00
Brian Waldon
eb930c3298 Merge pull request #2787 from bcwaldon/ttldur
client: add Node.TTLDuration()
2015-05-05 15:21:44 -07:00
Brian Waldon
ee9e336fd4 client: add Node.TTLDuration() 2015-05-05 15:03:24 -07:00
Barak Michener
d101568ac9 Merge pull request #2788 from barakmich/roadmap
*: Initial roadmap
2015-05-05 16:43:06 -04:00
Barak Michener
d4bd57229d *: Initial roadmap 2015-05-05 16:05:35 -04:00
Yicheng Qin
0b082b7bd4 Merge pull request #2771 from sorah/close-ongoing-conn
Fix connection leak when client disconnected
2015-04-29 20:46:13 -07:00
Shota Fukumori (sora_h)
a68efe7d1e proxy: Fix connection leak when client disconnect
established connections were leaked when client disconnected before
proxyreq completes. This happens all time for wait=true requests.
2015-04-30 11:41:42 +09:00
Yicheng Qin
0a6f481ca5 Merge pull request #2773 from yichengq/add-flag-help
tools/functional-testing: add help message for flags
2015-04-29 14:18:32 -07:00
Yicheng Qin
e71d43b58e tools/functional-testing: add help message for flags
Help users to understand what these flags are for.
2015-04-29 13:59:55 -07:00
Xiang Li
0fbf90b1e0 Merge pull request #2774 from xiang90/cluster
etcdserver: rename StoreAdminPrefix to StoreClusterPrefix
2015-04-29 12:20:04 -07:00
Xiang Li
94ffd72c7e etcdserver: rename StoreAdminPrefix to StoreClusterPrefix
We store cluster related key in StoreAdminPrefix for some
historical reason. The previous API is called admin. But now,
the admin name is gone and `cluster` is a more clear and correct
name.
2015-04-29 12:05:51 -07:00
Xiang Li
a4e35f4650 Merge pull request #2718 from xiang90/version
support cluster-wide version sync
2015-04-29 11:45:21 -07:00
Xiang Li
6699107f61 *: add cluster version and cluster version detection.
Cluster version is the min major.minor of all members in
the etcd cluster. Cluster version is set to the min version
that a etcd member is compatible with when first bootstrapp.

During a rolling upgrades, the cluster version will be updated
automatically.

For example:

```
Cluster [a:1, b:1 ,c:1] -> clusterVersion 1

update a -> 2, b -> 2

after a detection

Cluster [a:2, b:2 ,c:1] -> clusterVersion 1, since c is still 1

update c -> 2

after a detection

Cluster [a:2, b:2 ,c:2] -> clusterVersion 2
```

The API/raft component can utilize clusterVersion to determine if
it can accept a client request or a raft RPC.

We choose polling rather than pushing since we want to use the same
logic for cluster version detection and (TODO) cluster version checking.

Before a member actually joins a etcd cluster, it should check the version
of the cluster. Push does not work since the other members cannot push
version info to it before it actually joins. Moreover, we do not want our
raft RPC system (which is doing the heartbeat pushing) to coordinate cluster version.
2015-04-29 11:31:59 -07:00
Yicheng Qin
33febb979c Merge pull request #2761 from yichengq/344
etcdmain: advertise-client-urls must be set if listen-client-urls is set
2015-04-29 10:27:10 -07:00
Yicheng Qin
3f90394fbb etcdmain: advertise-client-urls must be set if listen-client-urls is set
Before this PR, people can set listen-client-urls without setting
advertise-client-urls, and leaves advertise-client-urls as default
localhost value. The client libraries which sync the cluster info
fetch wrong advertise-client-urls and cannot connect to the cluster.
This PR avoids this case and provides better UX.

On the other hand, this change is safe because people always want to set
advertise-client-urls if listen-client-urls is set. The default localhost
advertise url cannot be accessed from the outside, and should always be
set except that etcd is bootstrapped with no flag.
2015-04-29 09:52:15 -07:00
Brandon Philips
beb606f066 Merge pull request #2704 from philips/build-aci-port-mountpoint
scripts: build-aci update to have mountPoint and ports
2015-04-29 07:30:44 -07:00
Yicheng Qin
6c77e7a737 Merge pull request #2768 from coreos/docs-formatting
docs: fix code block formatting
2015-04-28 14:02:30 -07:00
Barak Michener
2a50f7a1aa Merge pull request #2770 from barakmich/new_logger
etcdmain: fix logging flag documentation
2015-04-28 16:53:58 -04:00
Barak Michener
ad8e3ea5dc etcdmain: fix logging flag documentation 2015-04-28 16:31:19 -04:00
Barak Michener
2299e35d99 Merge pull request #2769 from barakmich/new_logger
etcdmain: New logger
2015-04-28 16:06:51 -04:00
Barak Michener
b369cf037a etcdmain: New Logging Package
use capnslog

Vendor capnslog and set the flags in etcd main

remove package prefix from etcdmain
2015-04-28 15:42:32 -04:00
Rob Szumski
bfd4a29f67 docs: fix code block formatting 2015-04-28 11:17:13 -07:00
Yicheng Qin
0d6e062b5b Merge pull request #2738 from sublimino/patch-1
docs: fix link to etcd-migrate in README.md
2015-04-27 22:12:57 -07:00
Yicheng Qin
eafdd3b718 Merge pull request #2730 from yichengq/tester-key-param
main: parameterize stress key size and key suffix range
2015-04-27 17:02:36 -07:00
Yicheng Qin
057d21cf79 main: parameterize stress key size and key suffix range
It faciliates tester to adjust the size of each request, the number of
keys in the store and the size of snapshot.
2015-04-27 16:46:56 -07:00
Xiang Li
33f3bb3074 Merge pull request #2754 from xiang90/member_change
integration: add a test case for a full cluster rotation
2015-04-27 15:44:15 -07:00
Xiang Li
077c8397d2 integration: add a test case for a full cluster rotation 2015-04-27 15:38:06 -07:00
Yicheng Qin
d080c33c07 Merge pull request #2762 from yichengq/343
rafthttp: stop etcd if it is found removed when stream dial
2015-04-27 15:10:39 -07:00
Yicheng Qin
1c1cccd236 rafthttp: stop etcd if it is found removed when stream dial
The original process is stopping etcd only when pipeline message finds itself
has been removed. After this PR, stream dial has this functionality too.
It helps fast etcd stop, which doesn't need to wait for stream break to
fall back to pipeline, and wait for election timeout to send out message
to detect self removal.
2015-04-27 15:10:00 -07:00
Xiang Li
be6f49ba32 Merge pull request #2758 from lavagetto/master
docs: clarify the disaster recovery guide
2015-04-25 10:54:17 -07:00
Giuseppe Lavagetto
968f3d9711 docs: clarify the disaster recovery guide
A bit was missing from the documentation on disaster recovery, the reset
of the advertised peer urls for the node recovered from backup. Without
that, any subsequent server joining the cluster would not be able to
speak to the first node.
2015-04-25 18:54:29 +02:00
Xiang Li
f31a57d02e Merge pull request #2757 from yichengq/fix-typo
client: fix test name typo
2015-04-24 18:06:59 -07:00
Yicheng Qin
39dae50e71 client: fix test name typo
This is introduced at d89a862
2015-04-24 18:05:18 -07:00
Xiang Li
f244ae4aa5 Merge pull request #2756 from xiang90/client_gone
client: 410 is a vaild response for member.Remove
2015-04-24 17:17:12 -07:00
Xiang Li
91c45c3243 client: 410 is a vaild response for member.Remove
When removing a member, etcdserver might return 410 that indicates
the member has been removed. To client, 410 is a vaild response since
the client might do internal retry.
2015-04-24 17:01:23 -07:00
Xiang Li
b6aa31a5b6 Merge pull request #2750 from xiang90/member_test
integration: add tests around the membership change issues
2015-04-24 13:22:19 -07:00
Xiang Li
a42b9708ae integration: add tests around the membership change issues 2015-04-24 13:07:43 -07:00
Yicheng Qin
ebecee34e0 Merge pull request #2701 from yichengq/rafthttp-anon
rafthttp: add remotes
2015-04-24 13:04:37 -07:00
Xiang Li
49f4c17767 Merge pull request #2751 from akolb1/solaris_fix3
pkg/fileutil: add filelock support for solaris
2015-04-24 12:50:13 -07:00
Alexander Kolbasov
39c7060d3b pkg/fileutil: add filelock support for solaris 2015-04-24 12:18:08 -07:00
Yicheng Qin
9f19b5660f rafthttp: add AddRemote
Add remotes to rafthttp, who help newly joined members catch up the
progress of the cluster. It supports basic message sending to remote, and
has no stream connection for simplicity. remotes will not be used
after the latest peers have been added into rafthttp.
2015-04-24 11:49:23 -07:00
Xiang Li
41c7b43dc5 Merge pull request #2749 from junxu/master
raft: fix typo in raftlog
2015-04-24 07:47:01 -07:00
xujun
6b7891c643 raft: fix typo in raftlog
fix typo in String() method of raftlog which will misorder
the "committed" and "unstable.offset" output.
2015-04-24 03:28:57 -04:00
Xiang Li
b5d4d9ae9b Merge pull request #2713 from xiaost/etcdserver-skip-empty-entry
etcdserver: apply: skip empty Entry
2015-04-23 21:24:25 -07:00
xiaost
cab1e9a723 etcdserver: skip noop entry in apply 2015-04-24 12:15:51 +08:00
Yicheng Qin
0d25b20fc0 *: bump to v2.1.0-alpha.0+git 2015-04-23 15:02:51 -07:00
Andrew Martin
efb0b6e5c8 Fix link to etcd-migrate in README.md 2015-04-23 17:03:08 +01:00
Yicheng Qin
1811701427 Revert "etcdserver: fix cluster fallback recovery"
This reverts commit cff005777a.

Conflicts:
	etcdserver/server.go
2015-04-19 11:34:33 -07:00
Yicheng Qin
88224f6f4e Revert "etcdserver: not apply stale conf change in cluster and transport"
This reverts commit 40197f0698.
2015-04-19 11:08:03 -07:00
Brandon Philips
00044cd3bd scripts: build-aci update to have mountPoint and ports
Expose the etcd ports and data-dir mountPoint for future releases.
2015-04-17 14:57:15 -04:00
Yicheng Qin
51548acb4f rafthttp: reduce allocs in msgappv2
The patch decreases the allocs when sending one AppEntry in msgappv2
stream from 30 to 9. This helps reduce CPU load when etcd is under
high write load.
2015-04-06 09:45:39 -07:00
Kelsey Hightower
b15806e189 etcdctl: update the ls subcommand help to match behavior
Currently the `etcdctl ls` subcommand help output is a bit misleading.
It mentions that using the `--recursive` flag will output all keys and
values for a given path:

    --recursive  returns all values for key and child keys

This is inaccurate. The `--recursive` will only output the key names
under the given path. Fix the issue by updating the help string for
the `--recursive` flag.

    --recursive  returns all key names recursively for the given path

Fixes #2379.
2015-03-03 06:25:22 -08:00
597 changed files with 113838 additions and 16211 deletions

View File

@@ -4,8 +4,6 @@ go:
- 1.4
install:
- go get golang.org/x/tools/cmd/cover
- go get golang.org/x/tools/cmd/vet
- go get github.com/barakmich/go-nyet
script:

View File

@@ -1,6 +1,6 @@
# How to contribute
etcd is Apache 2.0 licensed and accepts contributions via Github pull requests. This document outlines some of the conventions on commit message formatting, contact points for developers and other resources to make getting your contribution into etcd easier.
etcd is Apache 2.0 licensed and accepts contributions via GitHub pull requests. This document outlines some of the conventions on commit message formatting, contact points for developers and other resources to make getting your contribution into etcd easier.
# Email and chat

View File

@@ -61,7 +61,7 @@ After your cluster is up and running, adding or removing members is done via [ru
### Member Migration
When there is a scheduled machine maintenance or retirement, you might want to migrate an etcd member to another machine without losing the data and changing the member ID.
When there is a scheduled machine maintenance or retirement, you might want to migrate an etcd member to another machine without losing the data and changing the member ID.
The data directory contains all the data to recover a member to its point-in-time state. To migrate a member:
@@ -102,7 +102,7 @@ $ sudo systemctl stop etcd
#### Copy the data directory of the now-idle member to the new machine
```
$ tar -cvzf node1.etcd.tar.gz /var/lib/etcd/node1.etcd
$ tar -cvzf node1.etcd.tar.gz /var/lib/etcd/node1.etcd
```
```
@@ -181,11 +181,13 @@ Once you have verified that etcd has started successfully, shut it down and move
#### Restoring the cluster
Now that the node is running successfully, you can add more nodes to the cluster and restore resiliency. See the [runtime configuration](runtime-configuration.md) guide for more details.
Now that the node is running successfully, you should [change its advertised peer URLs](other_apis.md#change-the-peer-urls-of-a-member), as the `--force-new-cluster` has set the peer URL to the default (listening on localhost).
You can then add more nodes to the cluster and restore resiliency. See the [runtime configuration](runtime-configuration.md) guide for more details.
### Client Request Timeout
etcd sets different timeouts for various types of client requests. The timeout value is not tunable now, which will be improved soon(https://github.com/coreos/etcd/issues/2038).
etcd sets different timeouts for various types of client requests. The timeout value is not tunable now, which will be improved soon (https://github.com/coreos/etcd/issues/2038).
#### Get requests
@@ -207,3 +209,11 @@ If the request times out, it indicates two possibilities:
2. the majority of the cluster is not functioning.
If timeout happens several times continuously, administrators should check status of cluster and resolve it as soon as possible.
### Best Practices
#### Maximum OS threads
By default, etcd uses the default configuration of the Go 1.4 runtime, which means that at most one operating system thread will be used to execute code simultaneously. (Note that this default behavior [may change in Go 1.5](https://docs.google.com/document/d/1At2Ls5_fhJQ59kDK2DFVhFu3g5mATSXqqV5QrxinasI/edit)).
When using etcd in heavy-load scenarios on machines with multiple cores it will usually be desirable to increase the number of threads that etcd can utilize. To do this, simply set the environment variable `GOMAXPROCS` to the desired number when starting etcd. For more information on this variable, see the Go [runtime](https://golang.org/pkg/runtime) documentation.

View File

@@ -78,7 +78,7 @@ X-Raft-Index: 5398
X-Raft-Term: 1
```
- `X-Etcd-Index` is the current etcd index as explained above.
- `X-Etcd-Index` is the current etcd index as explained above. When request is a watch on key space, `X-Etcd-Index` is the current etcd index when the watch starts, which means that the watched event may happen after `X-Etcd-Index`.
- `X-Raft-Index` is similar to the etcd index but is for the underlying raft protocol
- `X-Raft-Term` is an integer that will increase whenever an etcd master election happens in the cluster. If this number is increasing rapidly, you may need to tune the election timeout. See the [tuning][tuning] section for details.
@@ -277,7 +277,7 @@ The first terminal should get the notification and return with the same response
However, the watch command can do more than this.
Using the index, we can watch for commands that have happened in the past.
This is useful for ensuring you don't miss events between watch commands.
Typically, we watch again from the (modifiedIndex + 1) of the node we got.
Typically, we watch again from the `modifiedIndex` + 1 of the node we got.
Let's try to watch for the set command of index 7 again:
@@ -287,49 +287,75 @@ curl 'http://127.0.0.1:2379/v2/keys/foo?wait=true&waitIndex=7'
The watch command returns immediately with the same response as previously.
If we were to restart the watch from index 8 with:
```sh
curl 'http://127.0.0.1:2379/v2/keys/foo?wait=true&waitIndex=8'
```
Then even if etcd is on index 9 or 800, the first event to occur to the `/foo`
key between 8 and the current index will be returned.
**Note**: etcd only keeps the responses of the most recent 1000 events across all etcd keys.
It is recommended to send the response to another thread to process immediately
instead of blocking the watch while processing the result.
If we miss all the 1000 events, we need to recover the current state of the
watching key space. First, We do a get and then start to watch from the (etcdIndex + 1).
#### Watch from cleared event index
For example, we set `/foo="bar"` for 2000 times and tries to wait from index 7.
If we miss all the 1000 events, we need to recover the current state of the
watching key space through a get and then start to watch from the
`X-Etcd-Index` + 1.
For example, we set `/other="bar"` for 2000 times and try to wait from index 8.
```sh
curl 'http://127.0.0.1:2379/v2/keys/foo?wait=true&waitIndex=7'
curl 'http://127.0.0.1:2379/v2/keys/foo?wait=true&waitIndex=8'
```
We get the index is outdated response, since we miss the 1000 events kept in etcd.
```
{"errorCode":401,"message":"The event in requested index is outdated and cleared","cause":"the requested history has been cleared [1003/7]","index":2002}
{"errorCode":401,"message":"The event in requested index is outdated and cleared","cause":"the requested history has been cleared [1008/8]","index":2007}
```
To start watch, first we need to fetch the current state of key `/foo` and the etcdIndex.
To start watch, first we need to fetch the current state of key `/foo`:
```sh
curl 'http://127.0.0.1:2379/v2/keys/foo' -vv
```
```
< HTTP/1.1 200 OK
< Content-Type: application/json
< X-Etcd-Cluster-Id: 7e27652122e8b2ae
< X-Etcd-Index: 2002
< X-Etcd-Index: 2007
< X-Raft-Index: 2615
< X-Raft-Term: 2
< Date: Mon, 05 Jan 2015 18:54:43 GMT
< Transfer-Encoding: chunked
<
{"action":"get","node":{"key":"/foo","value":"","modifiedIndex":2002,"createdIndex":2002}}
{"action":"get","node":{"key":"/foo","value":"bar","modifiedIndex":7,"createdIndex":7}}
```
The `X-Etcd-Index` is important. It is the index when we got the value of `/foo`.
So we can watch again from the (`X-Etcd-Index` + 1) without missing an event after the last get.
Unlike watches we use the `X-Etcd-Index` + 1 of the response as a `waitIndex`
instead of the node's `modifiedIndex` + 1 for two reasons:
1. The `X-Etcd-Index` is always greater than or equal to the `modifiedIndex` when
getting a key because `X-Etcd-Index` is the current etcd index, and the `modifiedIndex`
is the index of an event already stored in etcd.
2. None of the events represented by indexes between `modifiedIndex` and
`X-Etcd-Index` will be related to the key being fetched.
Using the `modifiedIndex` + 1 is functionally equivalent for subsequent
watches, but since it is smaller than the `X-Etcd-Index` + 1, we may receive a
`401 EventIndexCleared` error immediately.
So the first watch after the get should be:
```sh
curl 'http://127.0.0.1:2379/v2/keys/foo?wait=true&waitIndex=2003'
curl 'http://127.0.0.1:2379/v2/keys/foo?wait=true&waitIndex=2008'
```
### Atomically Creating In-Order Keys
Using `POST` on a directory, you can create keys with key names that are created in-order.
@@ -870,7 +896,7 @@ Here we see the `/message` key but our hidden `/_message` key is not returned.
### Setting a key from a file
You can also use etcd to store small configuration files, json documents, XML documents, etc directly.
You can also use etcd to store small configuration files, JSON documents, XML documents, etc directly.
For example you can use curl to upload a simple text file and encode it:
```

View File

@@ -12,14 +12,14 @@ There are three types of resources in etcd
#### Users
A user is an identity to be authenticated. Each user can have multiple roles. The user has a capability (such as reading or writing) on the resource if one of the roles has that capability.
A user named `root` is required before security can be enabled, and it always has the ROOT role. The ROOT role can be granted to multiple users, but `root` is required for recovery purposes.
A user named `root` is required before authentication can be enabled, and it always has the ROOT role. The ROOT role can be granted to multiple users, but `root` is required for recovery purposes.
#### Roles
Each role has exact one associated Permission List. An permission list exists for each permission on key-value resources.
The special static ROOT (named `root`) role has a full permissions on all key-value resources, the permission to manage user resources and settings resources. Only the ROOT role has the permission to manage user resources and modify settings resources. The ROOT role is built-in and does not need to be created.
There is also a special GUEST role, named 'guest'. These are the permissions given to unauthenticated requests to etcd. This role will be created when security is enabled, unless it already exists, and by default allows access to the full keyspace due to backward compatability. (etcd did not previously authenticate any actions.). This role can be modified by a ROOT role holder at any time.
There is also a special GUEST role, named 'guest'. These are the permissions given to unauthenticated requests to etcd. This role will be created automatically, and by default allows access to the full keyspace due to backward compatability. (etcd did not previously authenticate any actions.). This role can be modified by a ROOT role holder at any time, to reduce the capabilities of unauthenticated users.
#### Permissions
@@ -35,7 +35,7 @@ A permission on `/foo` is for that exact key or directory, not its children or r
### Settings Resources
Specific settings for the cluster as a whole. This can include adding and removing cluster members, enabling or disabling security, replacing certificates, and any other dynamic configuration by the administrator (holder of the ROOT role).
Specific settings for the cluster as a whole. This can include adding and removing cluster members, enabling or disabling authentication, replacing certificates, and any other dynamic configuration by the administrator (holder of the ROOT role).
## v2 Auth
@@ -43,8 +43,8 @@ Specific settings for the cluster as a whole. This can include adding and removi
We only support [Basic Auth](http://en.wikipedia.org/wiki/Basic_access_authentication) for the first version. Client needs to attach the basic auth to the HTTP Authorization Header.
### Authorization field for operations
Added to requests to /v2/keys, /v2/security
Add code 403 Forbidden to the set of responses from the v2 API
Added to requests to /v2/keys, /v2/auth
Add code 401 Unauthorized to the set of responses from the v2 API
Authorization: Basic {encoded string}
### Future Work
@@ -52,7 +52,7 @@ Other types of auth can be considered for the future (eg, signed certs, public k
### Things out of Scope for etcd Permissions
* Pluggable AUTH backends like LDAP (other Authorization tokens generated by LDAP et al may be a possiblity)
* Pluggable AUTH backends like LDAP (other Authorization tokens generated by LDAP et al may be a possibility)
* Very fine-grained access controls (eg: users modifying keys outside work hours)
@@ -65,34 +65,74 @@ An Error JSON corresponds to:
"description" : "The longer helpful description of the error."
}
#### Enable and Disable Authentication
**Get auth status**
GET /v2/auth/enable
Sent Headers:
Possible Status Codes:
200 OK
200 Body:
{
"enabled": true
}
**Enable auth**
PUT /v2/auth/enable
Sent Headers:
Put Body: (empty)
Possible Status Codes:
200 OK
400 Bad Request (if root user has not been created)
409 Conflict (already enabled)
200 Body: (empty)
**Disable auth**
DELETE /v2/auth/enable
Sent Headers:
Authorization: Basic <RootAuthString>
Possible Status Codes:
200 OK
401 Unauthorized (if not a root user)
409 Conflict (already disabled)
200 Body: (empty)
#### Users
The User JSON object is formed as follows:
```
{
"user": "userName"
"password": "password"
"user": "userName",
"password": "password",
"roles": [
"role1",
"role2"
],
"grant": [],
"revoke": [],
"revoke": []
}
```
Password is only passed when necessary. Last Modified is set by the server and ignored in all client posts.
Password is only passed when necessary.
**Get a list of users**
GET/HEAD /v2/security/user
GET/HEAD /v2/auth/users
Sent Headers:
Authorization: Basic <BasicAuthString>
Possible Status Codes:
200 OK
403 Forbidden
401 Unauthorized
200 Headers:
Content-type: application/json
200 Body:
@@ -102,19 +142,19 @@ GET/HEAD /v2/security/user
**Get User Details**
GET/HEAD /v2/security/users/alice
GET/HEAD /v2/auth/users/alice
Sent Headers:
Authorization: Basic <BasicAuthString>
Possible Status Codes:
200 OK
403 Forbidden
401 Unauthorized
404 Not Found
200 Headers:
Content-type: application/json
200 Body:
{
"user" : "alice"
"user" : "alice",
"roles" : ["fleet", "etcd"]
}
@@ -122,7 +162,7 @@ GET/HEAD /v2/security/users/alice
A user can be created with initial roles, if filled in. However, no roles are required; only the username and password fields
PUT /v2/security/users/charlie
PUT /v2/auth/users/charlie
Sent Headers:
Authorization: Basic <BasicAuthString>
@@ -132,19 +172,26 @@ PUT /v2/security/users/charlie
* Grant/Revoke/Password filled in when updating (to grant roles, revoke roles, or change the password).
Possible Status Codes:
200 OK
403 Forbidden
409 Conflict (if exists)
200 Body: (empty)
201 Created
400 Bad Request
401 Unauthorized
404 Not Found (update non-existent users)
409 Conflict (when granting duplicated roles or revoking non-existent roles)
200 Headers:
Content-type: application/json
200 Body:
JSON state of the user
**Remove A User**
DELETE /v2/security/users/charlie
DELETE /v2/auth/users/charlie
Sent Headers:
Authorization: Basic <BasicAuthString>
Possible Status Codes:
200 OK
403 Forbidden
401 Unauthorized
403 Forbidden (remove root user when auth is enabled)
404 Not Found
200 Headers:
200 Body: (empty)
@@ -156,28 +203,26 @@ A full role structure may look like this. A Permission List structure is used fo
{
"role" : "fleet",
"permissions" : {
"kv" {
"kv" : {
"read" : [ "/fleet/" ],
"write": [ "/fleet/" ],
"write": [ "/fleet/" ]
}
}
},
"grant" : {"kv": {...}},
"revoke": {"kv": {...}},
"members" : ["alice", "bob"]
"revoke": {"kv": {...}}
}
```
**Get a list of Roles**
GET/HEAD /v2/security/roles
GET/HEAD /v2/auth/roles
Sent Headers:
Authorization: Basic <BasicAuthString>
Possible Status Codes:
200 OK
403 Forbidden
401 Unauthorized
200 Headers:
ETag: "<hash of list of roles>"
Content-type: application/json
200 Body:
{
@@ -186,31 +231,30 @@ GET/HEAD /v2/security/roles
**Get Role Details**
GET/HEAD /v2/security/roles/fleet
GET/HEAD /v2/auth/roles/fleet
Sent Headers:
Authorization: Basic <BasicAuthString>
Possible Status Codes:
200 OK
403 Forbidden
401 Unauthorized
404 Not Found
200 Headers:
ETag: "roles/fleet:<lastModified>"
Content-type: application/json
200 Body:
{
"role" : "fleet",
"read": {
"prefixesAllowed": ["/fleet/"],
},
"write": {
"prefixesAllowed": ["/fleet/"],
},
"permissions" : {
"kv" : {
"read": [ "/fleet/" ],
"write": [ "/fleet/" ]
}
}
}
**Create Or Update A Role**
PUT /v2/security/roles/rocket
PUT /v2/auth/roles/rkt
Sent Headers:
Authorization: Basic <BasicAuthString>
@@ -219,107 +263,86 @@ PUT /v2/security/roles/rocket
* Starting permission set if creating
* Granted/Revoked permission set if updating
Possible Status Codes:
200 OK
201 Created
403 Forbidden
404 Not Found
409 Conflict (if exists)
400 Bad Request
401 Unauthorized
404 Not Found (update non-existent roles)
409 Conflict (when granting duplicated permission or revoking non-existent permission)
200 Body:
JSON state of the role
**Remove A Role**
DELETE /v2/security/roles/rocket
DELETE /v2/auth/roles/rkt
Sent Headers:
Authorization: Basic <BasicAuthString>
Possible Status Codes:
200 OK
403 Forbidden
401 Unauthorized
403 Forbidden (remove root)
404 Not Found
200 Headers:
200 Body: (empty)
#### Enable and Disable Security
**Get security status**
GET /v2/security/enable
Sent Headers:
Possible Status Codes:
200 OK
200 Body:
{
"enabled": true
}
**Enable security**
PUT /v2/security/enable
Sent Headers:
Put Body: (empty)
Possible Status Codes:
200 OK
400 Bad Request (if not a root user)
200 Body: (empty)
**Disable security**
DELETE /v2/security/enable
Sent Headers:
Authorization: Basic <RootAuthString>
Possible Status Codes:
200 OK
403 Forbidden (if not a root user)
200 Body: (empty)
## Example Workflow
Let's walk through an example to show two tenants (applications, in our case) using etcd permissions.
### Enable security
### Create root role
```
PUT /v2/security/enable
Headers:
Put Body:
{"user" : "root", "password": "root"}
```
### Change root's password
```
PUT /v2/security/users/root
Headers:
Authorization: Basic <root:root>
PUT /v2/auth/users/root
Put Body:
{"user" : "root", "password": "betterRootPW!"}
```
### Create Roles for the Applications
Create the rocket role fully specified:
### Enable auth
```
PUT /v2/security/roles/rocket
PUT /v2/auth/enable
```
### Modify guest role (revoke write permission)
```
PUT /v2/auth/roles/guest
Headers:
Authorization: Basic <root:betterRootPW!>
Body:
Put Body:
{
"role" : "rocket",
"role" : "guest",
"revoke" : {
"kv" : {
"write": [
"*"
]
}
}
}
```
### Create Roles for the Applications
Create the rkt role fully specified:
```
PUT /v2/auth/roles/rkt
Headers:
Authorization: Basic <root:betterRootPW!>
Body:
{
"role" : "rkt",
"permissions" : {
"kv": {
"read": [
"/rocket/*"
"/rkt/*"
],
"write": [
"/rocket/*"
"/rkt/*"
]
}
}
@@ -329,23 +352,23 @@ PUT /v2/security/roles/rocket
But let's make fleet just a basic role for now:
```
PUT /v2/security/roles/fleet
PUT /v2/auth/roles/fleet
Headers:
Authorization: Basic <root:betterRootPW!>
Body:
Body:
{
"role" : "fleet",
"role" : "fleet"
}
```
### Optional: Add some permissions to the roles
### Optional: Grant some permissions to the roles
Well, we finally figured out where we want fleet to live. Let's fix it.
(Note that we avoided this in the rocket case. So this step is optional.)
(Note that we avoided this in the rkt case. So this step is optional.)
```
PUT /v2/security/roles/fleet
PUT /v2/auth/roles/fleet
Headers:
Authorization: Basic <root:betterRootPW!>
Put Body:
@@ -354,7 +377,7 @@ PUT /v2/security/roles/fleet
"grant" : {
"kv" : {
"read": [
"/rocket/fleet",
"/rkt/fleet",
"/fleet/*"
]
}
@@ -367,15 +390,15 @@ PUT /v2/security/roles/fleet
Same as before, let's use rocket all at once and fleet separately
```
PUT /v2/security/users/rocketuser
PUT /v2/auth/users/rktuser
Headers:
Authorization: Basic <root:betterRootPW!>
Body:
{"user" : "rocketuser", "password" : "rocketpw", "roles" : ["rocket"]}
{"user" : "rktuser", "password" : "rktpw", "roles" : ["rkt"]}
```
```
PUT /v2/security/users/fleetuser
PUT /v2/auth/users/fleetuser
Headers:
Authorization: Basic <root:betterRootPW!>
Body:
@@ -387,23 +410,25 @@ PUT /v2/security/users/fleetuser
Likewise, let's explicitly grant fleetuser access.
```
PUT /v2/security/users/fleetuser
PUT /v2/auth/users/fleetuser
Headers:
Authorization: Basic <root:betterRootPW!>
Body:
{"user": "fleetuser", "grant": ["fleet"]}
{"user": "fleetuser", "grant": ["fleet"]}
```
#### Start to use fleetuser and rocketuser
#### Start to use fleetuser and rktuser
For example:
```
PUT /v2/keys/rocket/RocketData
PUT /v2/keys/rkt/RktData
Headers:
Authorization: Basic <rocketuser:rocketpw>
Authorization: Basic <rktuser:rktpw>
Body:
value=launch
```
Reads and writes outside the prefixes granted will fail with a 403 Forbidden.
Reads and writes outside the prefixes granted will fail with a 401 Unauthorized.

View File

@@ -0,0 +1,179 @@
# Authentication Guide
**NOTE: The authentication feature is considered experimental. We may change workflow without warning in future releases.**
## Overview
Authentication -- having users and roles in etcd -- was added in etcd 2.1. This guide will help you set up basic authentication in etcd.
etcd before 2.1 was a completely open system; anyone with access to the API could change keys. In order to preserve backward compatibility and upgradability, this feature is off by default.
For a full discussion of the RESTful API, see [the authentication API documentation](auth_api.md)
## Special Users and Roles
There is one special user, `root`, and there are two special roles, `root` and `guest`.
### User `root`
User `root` must be created before security can be activated. It has the `root` role and allows for the changing of anything inside etcd. The idea behind the `root` user is for recovery purposes -- a password is generated and stored somewhere -- and the root role is granted to the administrator accounts on the system. In the future, for troubleshooting and recovery, we will need to assume some access to the system, and future documentation will assume this root user (though anyone with the role will suffice).
### Role `root`
Role `root` cannot be modified, but it may be granted to any user. Having access via the root role not only allows global read-write access (as was the case before 2.1) but allows modification of the authentication policy and all administrative things, like modifying the cluster membership.
### Role `guest`
The `guest` role defines the permissions granted to any request that does not provide an authentication. This will be created on security activation (if it doesn't already exist) to have full access to all keys, as was true in etcd 2.0. It may be modified at any time, and cannot be removed.
## Working with users
The `user` subcommand for `etcdctl` handles all things having to do with user accounts.
A listing of users can be found with
```
$ etcdctl user list
```
Creating a user is as easy as
```
$ etcdctl user add myusername
```
And there will be prompt for a new password.
Roles can be granted and revoked for a user with
```
$ etcdctl user grant myusername -roles foo,bar,baz
$ etcdctl user revoke myusername -roles bar,baz
```
We can look at this user with
```
$ etcdctl user get myusername
```
And the password for a user can be changed with
```
$ etcdctl user passwd myusername
```
Which will prompt again for a new password.
To delete an account, there's always
```
$ etcdctl user remove myusername
```
## Working with roles
The `role` subcommand for `etcdctl` handles all things having to do with access controls for particular roles, as were granted to individual users.
A listing of roles can be found with
```
$ etcdctl role list
```
A new role can be created with
```
$ etcdctl role add myrolename
```
A role has no password; we are merely defining a new set of access rights.
Roles are granted access to various parts of the keyspace, a single path at a time.
Reading a path is simple; if the path ends in `*`, that key **and all keys prefixed with it**, are granted to holders of this role. If it does not end in `*`, only that key and that key alone is granted.
Access can be granted as either read, write, or both, as in the following examples:
```
# Give read access to keys under the /foo directory
$ etcdctl role grant myrolename -path '/foo/*' -read
# Give write-only access to the key at /foo/bar
$ etcdctl role grant myrolename -path '/foo/bar' -write
# Give full access to keys under /pub
$ etcdctl role grant myrolename -path '/pub/*' -readwrite
```
Beware that
```
# Give full access to keys under /pub??
$ etcdctl role grant myrolename -path '/pub*' -readwrite
```
Without the slash may include keys under `/publishing`, for example. To do both, grant `/pub` and `/pub/*`
To see what's granted, we can look at the role at any time:
```
$ etcdctl role get myrolename
```
Revocation of permissions is done the same logical way:
```
$ etcdctl role revoke myrolename -path '/foo/bar' -write
```
As is removing a role entirely
```
$ etcdctl role remove myrolename
```
## Enabling authentication
The minimal steps to enabling auth follow. The administrator can set up users and roles before or after enabling authentication, as a matter of preference.
Make sure the root user is created:
```
$ etcdctl user add root
New password:
```
And enable authentication
```
$ etcdctl auth enable
```
After this, etcd is running with authentication enabled. To disable it for any reason, use the reciprocal command:
```
$ etcdctl -u root:rootpw auth disable
```
It would also be good to check what guests (unauthenticated users) are allowed to do:
```
$ etcdctl -u root:rootpw role get guest
```
And modify this role appropriately, depending on your policies.
## Using `etcdctl` to authenticate
`etcdctl` supports a similar flag as `curl` for authentication.
```
$ etcdctl -u user:password get foo
```
or if you prefer to be prompted:
```
$ etcdctl -u user get foo
```
Otherwise, all `etcdctl` commands remain the same. Users and roles can still be created and modified, but require authentication by a user with the root role.

View File

@@ -79,7 +79,7 @@ The consistent read guarantees the sequential consistency within one client that
Each etcd member will proxy the request to leader and only return the result to user after the result is applied on the local member. Thus after the write succeed, the user is guaranteed to see the value on the member it sent the request to.
Reads do not provide linearizability. If you want linearizabilable read, you need to set quorum option to true.
Reads do not provide linearizability. If you want linearizable read, you need to set quorum option to true.
**Previous behavior**
@@ -108,7 +108,7 @@ A size key needs to be provided inside a [discovery token][discoverytoken].
[memberapi]: other_apis.md
## HTTP Key Value API
- The follower can now transparently proxy write equests to the leader. Clients will no longer see 307 redirections to the leader from etcd.
- The follower can now transparently proxy write requests to the leader. Clients will no longer see 307 redirections to the leader from etcd.
- Expiration time is in UTC instead of local time.

View File

@@ -14,7 +14,7 @@ GCE n1-highcpu-2 machine type
## Testing
Bootstrap another machine and use benchmark tool to send requests to etcd cluster.
Bootstrap another machine and use benchmark tool [boom](https://github.com/rakyll/boom) to send requests to each etcd member.
## Performance

View File

@@ -1,24 +1,24 @@
## Branch Managemnt
## Branch Management
### Guide
- New development occurs on the master branch
- New development occurs on the [master branch](https://github.com/coreos/etcd/tree/master)
- Master branch should always have a green build!
- Backwards-compatible bug fixes should target the master branch and ported to stable
- Backwards-compatible bug fixes should target the master branch and subsequently be ported to stable branches
- Once the master branch is ready for release, it will be tagged and become the new stable branch.
The etcd team adopts a rolling release model and support one stable version of etcd going forward.
The etcd team has adopted a _rolling release model_ and supports one stable version of etcd.
### Master branch
The master branch is our development branch. It is where all the new features go into first.
The `master` branch is our development branch. All new features land here first.
If you want to try new features, pull the master branch and play on it. But the branch is not really stable because new features may introduce bugs.
If you want to try new features, pull `master` and play with it. Note that `master` may not be stable because new features may introduce bugs.
Before the release of the next stable version, feature PRs will be frozen. We will focus on the testing, bug-fix and documentation for one to two weeks.
### Stable branches
All branches with prefix 'release-' are stable branches.
All branches with prefix `release-` are considered _stable_ branches.
After a Minor release (http://semver.org/), we will have a new stable branch for that release. We will keep fixing the backwards-compatible bugs for the latest stable release, but not the olders. The bug fixes Patch release will be once every two weeks, given any patches.
After every minor release (http://semver.org/), we will have a new stable branch for that release. We will keep fixing the backwards-compatible bugs for the latest stable release, but not previous releases. The _patch_ release, incorporating any bug fixes, will be once every two weeks, given any patches.

View File

@@ -43,6 +43,8 @@ On each machine you would start etcd with these flags:
```
$ etcd -name infra0 -initial-advertise-peer-urls http://10.0.1.10:2380 \
-listen-peer-urls http://10.0.1.10:2380 \
-listen-client-urls http://10.0.1.10:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.10:2379 \
-initial-cluster-token etcd-cluster-1 \
-initial-cluster infra0=http://10.0.1.10:2380,infra1=http://10.0.1.11:2380,infra2=http://10.0.1.12:2380 \
-initial-cluster-state new
@@ -50,6 +52,8 @@ $ etcd -name infra0 -initial-advertise-peer-urls http://10.0.1.10:2380 \
```
$ etcd -name infra1 -initial-advertise-peer-urls http://10.0.1.11:2380 \
-listen-peer-urls http://10.0.1.11:2380 \
-listen-client-urls http://10.0.1.11:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.11:2379 \
-initial-cluster-token etcd-cluster-1 \
-initial-cluster infra0=http://10.0.1.10:2380,infra1=http://10.0.1.11:2380,infra2=http://10.0.1.12:2380 \
-initial-cluster-state new
@@ -57,6 +61,8 @@ $ etcd -name infra1 -initial-advertise-peer-urls http://10.0.1.11:2380 \
```
$ etcd -name infra2 -initial-advertise-peer-urls http://10.0.1.12:2380 \
-listen-peer-urls http://10.0.1.12:2380 \
-listen-client-urls http://10.0.1.12:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.12:2379 \
-initial-cluster-token etcd-cluster-1 \
-initial-cluster infra0=http://10.0.1.10:2380,infra1=http://10.0.1.11:2380,infra2=http://10.0.1.12:2380 \
-initial-cluster-state new
@@ -71,6 +77,8 @@ In the following example, we have not included our new host in the list of enume
```
$ etcd -name infra1 -initial-advertise-peer-urls http://10.0.1.11:2380 \
-listen-peer-urls https://10.0.1.11:2380 \
-listen-client-urls http://10.0.1.11:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.11:2379 \
-initial-cluster infra0=http://10.0.1.10:2380 \
-initial-cluster-state new
etcd: infra1 not listed in the initial cluster config
@@ -82,6 +90,8 @@ In this example, we are attempting to map a node (infra0) on a different address
```
$ etcd -name infra0 -initial-advertise-peer-urls http://127.0.0.1:2380 \
-listen-peer-urls http://10.0.1.10:2380 \
-listen-client-urls http://10.0.1.10:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.10:2379 \
-initial-cluster infra0=http://10.0.1.10:2380,infra1=http://10.0.1.11:2380,infra2=http://10.0.1.12:2380 \
-initial-cluster-state=new
etcd: error setting up initial cluster: infra0 has different advertised URLs in the cluster and advertised peer URLs list
@@ -93,6 +103,8 @@ If you configure a peer with a different set of configuration and attempt to joi
```
$ etcd -name infra3 -initial-advertise-peer-urls http://10.0.1.13:2380 \
-listen-peer-urls http://10.0.1.13:2380 \
-listen-client-urls http://10.0.1.13:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.13:2379 \
-initial-cluster infra0=http://10.0.1.10:2380,infra1=http://10.0.1.11:2380,infra3=http://10.0.1.13:2380 \
-initial-cluster-state=new
etcd: conflicting cluster ID to the target cluster (c6ab534d07e8fcc4 != bc25ea2a74fb18b0). Exiting.
@@ -137,16 +149,22 @@ Now we start etcd with those relevant flags for each member:
```
$ etcd -name infra0 -initial-advertise-peer-urls http://10.0.1.10:2380 \
-listen-peer-urls http://10.0.1.10:2380 \
-listen-client-urls http://10.0.1.10:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.10:2379 \
-discovery https://myetcd.local/v2/keys/discovery/6c007a14875d53d9bf0ef5a6fc0257c817f0fb83
```
```
$ etcd -name infra1 -initial-advertise-peer-urls http://10.0.1.11:2380 \
-listen-peer-urls http://10.0.1.11:2380 \
-listen-client-urls http://10.0.1.11:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.11:2379 \
-discovery https://myetcd.local/v2/keys/discovery/6c007a14875d53d9bf0ef5a6fc0257c817f0fb83
```
```
$ etcd -name infra2 -initial-advertise-peer-urls http://10.0.1.12:2380 \
-listen-peer-urls http://10.0.1.12:2380 \
-listen-client-urls http://10.0.1.12:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.12:2379 \
-discovery https://myetcd.local/v2/keys/discovery/6c007a14875d53d9bf0ef5a6fc0257c817f0fb83
```
@@ -181,16 +199,22 @@ Now we start etcd with those relevant flags for each member:
```
$ etcd -name infra0 -initial-advertise-peer-urls http://10.0.1.10:2380 \
-listen-peer-urls http://10.0.1.10:2380 \
-listen-client-urls http://10.0.1.10:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.10:2379 \
-discovery https://discovery.etcd.io/3e86b59982e49066c5d813af1c2e2579cbf573de
```
```
$ etcd -name infra1 -initial-advertise-peer-urls http://10.0.1.11:2380 \
-listen-peer-urls http://10.0.1.11:2380 \
-listen-client-urls http://10.0.1.11:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.11:2379 \
-discovery https://discovery.etcd.io/3e86b59982e49066c5d813af1c2e2579cbf573de
```
```
$ etcd -name infra2 -initial-advertise-peer-urls http://10.0.1.12:2380 \
-listen-peer-urls http://10.0.1.12:2380 \
-listen-client-urls http://10.0.1.12:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.12:2379 \
-discovery https://discovery.etcd.io/3e86b59982e49066c5d813af1c2e2579cbf573de
```
@@ -206,6 +230,8 @@ You can use the environment variable `ETCD_DISCOVERY_PROXY` to cause etcd to use
```
$ etcd -name infra0 -initial-advertise-peer-urls http://10.0.1.10:2380 \
-listen-peer-urls http://10.0.1.10:2380 \
-listen-client-urls http://10.0.1.10:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.10:2379 \
-discovery https://discovery.etcd.io/3e86b59982e49066c5d813af1c2e2579cbf573de
etcd: error: the cluster doesnt have a size configuration value in https://discovery.etcd.io/3e86b59982e49066c5d813af1c2e2579cbf573de/_config
exit 1
@@ -218,6 +244,8 @@ This error will occur if the discovery cluster already has the configured number
```
$ etcd -name infra0 -initial-advertise-peer-urls http://10.0.1.10:2380 \
-listen-peer-urls http://10.0.1.10:2380 \
-listen-client-urls http://10.0.1.10:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.10:2379 \
-discovery https://discovery.etcd.io/3e86b59982e49066c5d813af1c2e2579cbf573de \
-discovery-fallback exit
etcd: discovery: cluster is full
@@ -232,6 +260,8 @@ ignored on this machine.
```
$ etcd -name infra0 -initial-advertise-peer-urls http://10.0.1.10:2380 \
-listen-peer-urls http://10.0.1.10:2380 \
-listen-client-urls http://10.0.1.10:2379,http://127.0.0.1:2379 \
-advertise-client-urls http://10.0.1.10:2379 \
-discovery https://discovery.etcd.io/3e86b59982e49066c5d813af1c2e2579cbf573de
etcdserver: discovery token ignored since a cluster has already been initialized. Valid log found at /var/lib/etcd
```
@@ -264,7 +294,7 @@ infra2.example.com. 300 IN A 10.0.1.12
```
#### Bootstrap the etcd cluster using DNS
etcd cluster memebers can listen on domain names or IP address, the bootstrap process will resolve DNS A records.
etcd cluster members can listen on domain names or IP address, the bootstrap process will resolve DNS A records.
```
$ etcd -name infra0 \

View File

@@ -13,6 +13,7 @@ To start etcd automatically using custom settings at startup in Linux, using a [
##### -name
+ Human-readable name for this member.
+ default: "default"
+ This value is referenced as this node's own entries listed in the `-initial-cluster` flag (Ex: `default=http://localhost:2380` or `default=http://localhost:2380,default=http://localhost:7001`). This needs to match the key used in the flag if you're using [static boostrapping](clustering.md#static).
##### -data-dir
+ Path to the data directory.
@@ -66,6 +67,7 @@ To start etcd automatically using custom settings at startup in Linux, using a [
##### -initial-cluster
+ Initial cluster configuration for bootstrapping.
+ default: "default=http://localhost:2380,default=http://localhost:7001"
+ The key is the value of the `-name` flag for each node provided. The default uses `default` for the key because this is the default for the `-name` flag.
##### -initial-cluster-state
+ Initial cluster state ("new" or "existing"). Set to `new` for all members present during initial static or DNS bootstrapping. If this option is set to `existing`, etcd will attempt to join the existing cluster. If the wrong value is set, etcd will attempt to start but fail safely.
@@ -105,6 +107,27 @@ To start etcd automatically using custom settings at startup in Linux, using a [
+ Proxy mode setting ("off", "readonly" or "on").
+ default: "off"
##### -proxy-failure-wait
+ Time (in milliseconds) an endpoint will be held in a failed state before being reconsidered for proxied requests.
+ default: 5000
##### -proxy-refresh-interval
+ Time (in milliseconds) of the endpoints refresh interval.
+ default: 30000
##### -proxy-dial-timeout
+ Time (in milliseconds) for a dial to timeout or 0 to disable the timeout
+ default: 1000
##### -proxy-write-timeout
+ Time (in milliseconds) for a write to timeout or 0 to disable the timeout.
+ default: 5000
##### -proxy-read-timeout
+ Time (in milliseconds) for a read to timeout or 0 to disable the timeout.
+ Don't change this value if you use watches because they are using long polling requests.
+ default: 0
### Security Flags
The security flags help to [build a secure etcd cluster][security].
@@ -149,6 +172,17 @@ The security flags help to [build a secure etcd cluster][security].
+ Path to the peer server TLS trusted CA file.
+ default: none
### Logging Flags
##### -debug
+ Drop the default log level to DEBUG for all subpackages.
+ default: false (INFO for all packages)
##### -log-package-levels
+ Set individual etcd subpackages to specific log levels. An example being `etcdserver=WARNING,security=DEBUG`
+ default: none (INFO for all packages)
### Unsafe Flags
Please be CAUTIOUS when using unsafe flags because it will break the guarantees given by the consensus protocol.

View File

@@ -22,6 +22,10 @@ The node in each member follows raft consensus protocol to replicate logs. Clust
Peer is another member of the same cluster.
### Proposal
A proposal is a request (for example a write request, a configuration change request) that needs to go through raft protocol.
### Client
Client is a caller of the cluster's HTTP API.

View File

@@ -9,6 +9,7 @@
- [etcd-browser](https://github.com/henszey/etcd-browser) - A web-based key/value editor for etcd using AngularJS
- [etcd-lock](https://github.com/datawisesystems/etcd-lock) - Master election & distributed r/w lock implementation using etcd - Supports v2
- [etcd-console](https://github.com/matishsiao/etcd-console) - A web-base key/value editor for etcd using PHP
- [etcd-viewer](https://github.com/nikfoundas/etcd-viewer) - An etcd key-value store editor/viewer written in Java
**Go libraries**
@@ -33,6 +34,7 @@
- [stianeikeland/node-etcd](https://github.com/stianeikeland/node-etcd) - Supports v2 (w Coffeescript)
- [lavagetto/nodejs-etcd](https://github.com/lavagetto/nodejs-etcd) - Supports v2
- [deedubs/node-etcd-config](https://github.com/deedubs/node-etcd-config) - Supports v2
**Ruby libraries**
@@ -68,7 +70,11 @@
**Haskell libraries**
- [wereHamster/etcd-hs](https://github.com/wereHamster/etcd-hs)
**R libraries**
- [ropensci/etseed](https://github.com/ropensci/etseed)
**Tcl libraries**
- [efrecon/etcd-tcl](https://github.com/efrecon/etcd-tcl) - Supports v2, except wait.
@@ -110,3 +116,5 @@ A detailed recap of client functionalities can be found in the [clients compatib
- [skynetservices/skydns](https://github.com/skynetservices/skydns) - RFC compliant DNS server
- [xordataexchange/crypt](https://github.com/xordataexchange/crypt) - Securely store values in etcd using GPG encryption
- [spf13/viper](https://github.com/spf13/viper) - Go configuration library, reads values from ENV, pflags, files, and etcd with optional encryption
- [lytics/metafora](https://github.com/lytics/metafora) - Go distributed task library
- [ryandoyle/nss-etcd](https://github.com/ryandoyle/nss-etcd) - A GNU libc NSS module for resolving names from etcd.

137
Documentation/metrics.md Normal file
View File

@@ -0,0 +1,137 @@
## Metrics
**NOTE: The metrics feature is considered as an experimental. We might add/change/remove metrics without warning in the future releases.**
etcd uses [Prometheus](http://prometheus.io/) for metrics reporting in the server. The metrics can be used for real-time monitoring and debugging.
The simplest way to see the available metrics is to cURL the metrics endpoint `/metrics` of etcd. The format is described [here](http://prometheus.io/docs/instrumenting/exposition_formats/).
You can also follow the doc [here](http://prometheus.io/docs/introduction/getting_started/) to start a Promethus server and monitor etcd metrics.
The naming of metrics follows the suggested [best practice of Promethus](http://prometheus.io/docs/practices/naming/). A metric name has an `etcd` prefix as its namespace and a subsystem prefix (for example `wal` and `etcdserver`).
etcd now exposes the following metrics:
### etcdserver
| Name | Description | Type |
|-----------------------------------------|--------------------------------------------------|---------|
| file_descriptors_used_total | The total number of file descriptors used | Gauge |
| proposal_durations_milliseconds | The latency distributions of committing proposal | Summary |
| pending_proposal_total | The total number of pending proposals | Gauge |
| proposal_failed_total | The total number of failed proposals | Counter |
High file descriptors (`file_descriptors_used_total`) usage (near the file descriptors limitation of the process) indicates a potential out of file descriptors issue. That might cause etcd fails to create new WAL files and panics.
[Proposal](glossary.md#proposal) durations (`proposal_durations_milliseconds`) give you an summary about the proposal commit latency. Latency can be introduced into this process by network and disk IO.
Pending proposal (`pending_proposal_total`) gives you an idea about how many proposal are in the queue and waiting for commit. An increasing pending number indicates a high client load or an unstable cluster.
Failed proposals (`proposal_failed_total`) are normally related to two issues: temporary failures related to a leader election or longer duration downtime caused by a loss of quorum in the cluster.
### store
These metrics describe the accesses into the data store of etcd members that exist in the cluster. They
are useful to count what kind of actions are taken by users. It is also useful to see and whether all etcd members
"see" the same set of data mutations, and whether reads and watches (which are local) are equally distributed.
All these metrics are prefixed with `etcd_store_`.
| Name | Description | Type |
|---------------------------|------------------------------------------------------------------------------------------|--------------------|
| reads_total | Total number of reads from store, should differ among etcd members (local reads). | Counter(action) |
| writes_total | Total number of writes to store, should be same among all etcd members. | Counter(action) |
| reads_failed_total | Number of failed reads from store (e.g. key missing) on local reads. | Counter(action) |
| writes_failed_total | Number of failed writes to store (e.g. failed compare and swap). | Counter(action) |
| expires_total | Total number of expired keys (due to TTL).   | Counter |
| watch_requests_totals | Total number of incoming watch requests to this etcd member (local watches). | Counter |
| watchers | Current count of active watchers on this etcd member. | Gauge |
Both `reads_total` and `writes_total` count both successful and failed requests. `reads_failed_total` and
`writes_failed_total` count failed requests. A lot of failed writes indicate possible contentions on keys (e.g. when
doing `compareAndSet`), and read failures indicate that some clients try to access keys that don't exist.
Example Prometheus queries that may be useful from these metrics (across all etcd members):
* `sum(rate(etcd_store_reads_total{job="etcd"}[1m])) by (action)`
`max(rate(etcd_store_writes_total{job="etcd"}[1m])) by (action)`
Rate of reads and writes by action, across all servers across a time window of `1m`. The reason why `max` is used
for writes as opposed to `sum` for reads is because all of etcd nodes in the cluster apply all writes to their stores.
Shows the rate of successfull readonly/write queries across all servers, across a time window of `1m`.
* `sum(rate(etcd_store_watch_requests_total{job="etcd"}[1m]))`
Shows rate of new watch requests per second. Likely driven by how often watched keys change.
* `sum(etcd_store_watchers{job="etcd"})`
Number of active watchers across all etcd servers.
### wal
| Name | Description | Type |
|------------------------------------|--------------------------------------------------|---------|
| fsync_durations_microseconds | The latency distributions of fsync called by wal | Summary |
| last_index_saved | The index of the last entry saved by wal | Gauge |
Abnormally high fsync duration (`fsync_durations_microseconds`) indicates disk issues and might cause the cluster to be unstable.
### snapshot
| Name | Description | Type |
|--------------------------------------------|------------------------------------------------------------|---------|
| snapshot_save_total_durations_microseconds | The total latency distributions of save called by snapshot | Summary |
Abnormally high snapshot duration (`snapshot_save_total_durations_microseconds`) indicates disk issues and might cause the cluster to be unstable.
### rafthttp
| Name | Description | Type | Labels |
|-----------------------------------|--------------------------------------------|---------|--------------------------------|
| message_sent_latency_microseconds | The latency distributions of messages sent | Summary | sendingType, msgType, remoteID |
| message_sent_failed_total | The total number of failed messages sent | Summary | sendingType, msgType, remoteID |
Abnormally high message duration (`message_sent_latency_microseconds`) indicates network issues and might cause the cluster to be unstable.
An increase in message failures (`message_sent_failed_total`) indicates more severe network issues and might cause the cluster to be unstable.
Label `sendingType` is the connection type to send messages. `message`, `msgapp` and `msgappv2` use HTTP streaming, while `pipeline` does HTTP request for each message.
Label `msgType` is the type of raft message. `MsgApp` is log replication message; `MsgSnap` is snapshot install message; `MsgProp` is proposal forward message; the others are used to maintain raft internal status. If you have a large snapshot, you would expect a long msgSnap sending latency. For other types of messages, you would expect low latency, which is comparable to your ping latency if you have enough network bandwidth.
Label `remoteID` is the member ID of the message destination.
### proxy
etcd members operating in proxy mode do not do store operations. They forward all requests
to cluster instances.
Tracking the rate of requests coming from a proxy allows one to pin down which machine is performing most reads/writes.
All these metrics are prefixed with `etcd_proxy_`
| Name | Description | Type |
|---------------------------|-----------------------------------------------------------------------------------------|--------------------|
| requests_total | Total number of requests by this proxy instance. . | Counter(method) |
| handled_total | Total number of fully handled requests, with responses from etcd members. | Counter(method) |
| dropped_total | Total number of dropped requests due to forwarding errors to etcd members.  | Counter(method,error) |
| handling_duration_seconds | Bucketed handling times by HTTP method, including round trip to member instances. | Histogram(method) |
Example Prometheus queries that may be useful from these metrics (across all etcd servers):
* `sum(rate(etcd_proxy_handled_total{job="etcd"}[1m])) by (method)`
Rate of requests (by HTTP method) handled by all proxies, across a window of `1m`.
* `histogram_quantile(0.9, sum(increase(etcd_proxy_events_handling_time_seconds_bucket{job="etcd",method="GET"}[5m])) by (le))`
`histogram_quantile(0.9, sum(increase(etcd_proxy_events_handling_time_seconds_bucket{job="etcd",method!="GET"}[5m])) by (le))`
Show the 0.90-tile latency (in seconds) of handling of user requestsacross all proxy machines, with a window of `5m`.
* `sum(rate(etcd_proxy_dropped_total{job="etcd"}[1m])) by (proxying_error)`
Number of failed request on the proxy. This should be 0, spikes here indicate connectivity issues to etcd cluster.

View File

@@ -6,6 +6,8 @@ etcd currently supports two proxy modes: `readwrite` and `readonly`. The default
The proxy will shuffle the list of cluster members periodically to avoid sending all connections to a single member.
The member list used by proxy consists of all client URLs advertised within the cluster, as specified in each members' `-advertise-client-urls` flag. If this flag is set incorrectly, requests sent to the proxy are forwarded to wrong addresses and then fail. The fix for this problem is to restart etcd member with correct `-advertise-client-urls` flag. After client URLs list in proxy is recalculated, which happens every 30 seconds, requests will be forwarded correctly.
### Using an etcd proxy
To start etcd in proxy mode, you need to provide three flags: `proxy`, `listen-client-urls`, and `initial-cluster` (or `discovery`).
@@ -16,7 +18,7 @@ The proxy will be listening on `listen-client-urls` and forward requests to the
#### Start an etcd proxy with a static configuration
To start a proxy that will connect to a statically defined etcd cluster, specify the `initial-cluster` flag:
```
etcd -proxy on -listen-client-urls 127.0.0.1:8080 -initial-cluster infra0=http://10.0.1.10:2380,infra1=http://10.0.1.11:2380,infra2=http://10.0.1.12:2380
etcd -proxy on -listen-client-urls http://127.0.0.1:8080 -initial-cluster infra0=http://10.0.1.10:2380,infra1=http://10.0.1.11:2380,infra2=http://10.0.1.12:2380
```
#### Start an etcd proxy with the discovery service
@@ -25,7 +27,7 @@ If you bootstrap an etcd cluster using the [discovery service][discovery-service
To start a proxy using the discovery service, specify the `discovery` flag. The proxy will wait until the etcd cluster defined at the `discovery` url finishes bootstrapping, and then start to forward the requests.
```
etcd -proxy on -listen-client-urls 127.0.0.1:8080 -discovery https://discovery.etcd.io/3e86b59982e49066c5d813af1c2e2579cbf573de
etcd -proxy on -listen-client-urls http://127.0.0.1:8080 -discovery https://discovery.etcd.io/3e86b59982e49066c5d813af1c2e2579cbf573de
```
#### Fallback to proxy mode with discovery service

191
Documentation/rfc/v3api.md Normal file
View File

@@ -0,0 +1,191 @@
## Design
1. Flatten binary key-value space
2. Keep the event history until compaction
- access to old version of keys
- user controlled history compaction
3. Support range query
- Pagination support with limit argument
- Support consistency guarantee across multiple range queries
4. Replace TTL key with Lease
- more efficient/ low cost keep alive
- a logical group of TTL keys
5. Replace CAS/CAD with multi-object Tnx
- MUCH MORE powerful and flexible
6. Support efficient watching with multiple ranges
7. RPC API supports the completed set of APIs.
- more efficient than JSON/HTTP
- additional tnx/lease support
8. HTTP API supports a subset of APIs.
- easy for people to try out etcd
- easy for people to write simple etcd application
## Protobuf Defined API
[protobuf](./v3api.proto)
### Examples
#### Put a key (foo=bar)
```
// A put is always successful
Put( PutRequest { key = foo, value = bar } )
PutResponse {
cluster_id = 0x1000,
member_id = 0x1,
index = 1,
raft_term = 0x1,
}
```
#### Get a key (assume we have foo=bar)
```
Get ( RangeRequest { key = foo } )
RangeResponse {
cluster_id = 0x1000,
member_id = 0x1,
index = 1,
raft_term = 0x1,
kvs = {
{
key = foo,
value = bar,
create_index = 1,
mod_index = 1,
version = 1;
},
},
}
```
#### Range over a key space (assume we have foo0=bar0… foo100=bar100)
```
Range ( RangeRequest { key = foo, end_key = foo80, limit = 30 } )
RangeResponse {
cluster_id = 0x1000,
member_id = 0x1,
index = 100,
raft_term = 0x1,
kvs = {
{
key = foo0,
value = bar0,
create_index = 1,
mod_index = 1,
version = 1;
},
...,
{
key = foo30,
value = bar30,
create_index = 30,
mod_index = 30,
version = 1;
},
},
}
```
#### Finish a tnx (assume we have foo0=bar0, foo1=bar1)
```
Tnx(TnxRequest {
// mod_index of foo0 is equal to 1, mod_index of foo1 is greater than 1
compare = {
{compareType = equal, key = foo0, mod_index = 1},
{compareType = greater, key = foo1, mod_index = 1}}
},
// if the comparison succeeds, put foo2 = bar2
success = {PutRequest { key = foo2, value = success }},
// if the comparison fails, put foo2=fail
failure = {PutRequest { key = foo2, value = failure }},
)
TnxResponse {
cluster_id = 0x1000,
member_id = 0x1,
index = 3,
raft_term = 0x1,
succeeded = true,
responses = {
// response of PUT foo2=success
{
cluster_id = 0x1000,
member_id = 0x1,
index = 3,
raft_term = 0x1,
}
}
}
```
#### Watch on a key/range
```
Watch( WatchRequest{
key = foo,
end_key = fop, // prefix foo
start_index = 20,
end_index = 10000,
// server decided notification frequency
progress_notification = true,
}
… // this can be a watch request stream
)
// put (foo0=bar0) event at 3
WatchResponse {
cluster_id = 0x1000,
member_id = 0x1,
index = 3,
raft_term = 0x1,
event_type = put,
kv = {
key = foo0,
value = bar0,
create_index = 1,
mod_index = 1,
version = 1;
},
}
// a notification at 2000
WatchResponse {
cluster_id = 0x1000,
member_id = 0x1,
index = 2000,
raft_term = 0x1,
// nil event as notification
}
// put (foo0=bar3000) event at 3000
WatchResponse {
cluster_id = 0x1000,
member_id = 0x1,
index = 3000,
raft_term = 0x1,
event_type = put,
kv = {
key = foo0,
value = bar3000,
create_index = 1,
mod_index = 3000,
version = 2;
},
}
```

View File

@@ -0,0 +1,272 @@
syntax = "proto3";
// Interface exported by the server.
service etcd {
// Range gets the keys in the range from the store.
rpc Range(RangeRequest) returns (RangeResponse) {}
// Put puts the given key into the store.
// A put request increases the index of the store,
// and generates one event in the event history.
rpc Put(PutRequest) returns (PutResponse) {}
// Delete deletes the given range from the store.
// A delete request increase the index of the store,
// and generates one event in the event history.
rpc DeleteRange(DeleteRangeRequest) returns (DeleteRangeResponse) {}
// Tnx processes all the requests in one transaction.
// A tnx request increases the index of the store,
// and generates events with the same index in the event history.
rpc Tnx(TnxRequest) returns (TnxResponse) {}
// Watch watches the events happening or happened in etcd. Both input and output
// are stream. One watch rpc can watch for multiple ranges and get a stream of
// events. The whole events history can be watched unless compacted.
rpc WatchRange(stream WatchRangeRequest) returns (stream WatchRangeResponse) {}
// Compact compacts the event history in etcd. User should compact the
// event history periodically, or it will grow infinitely.
rpc Compact(CompactionRequest) returns (CompactionResponse) {}
// LeaseCreate creates a lease. A lease has a TTL. The lease will expire if the
// server does not receive a keepAlive within TTL from the lease holder.
// All keys attached to the lease will be expired and deleted if the lease expires.
// The key expiration generates an event in event history.
rpc LeaseCreate(LeaseCreateRequest) returns (LeaseCreateResponse) {}
// LeaseRevoke revokes a lease. All the key attached to the lease will be expired and deleted.
rpc LeaseRevoke(LeaseRevokeRequest) returns (LeaseRevokeResponse) {}
// LeaseAttach attaches keys with a lease.
rpc LeaseAttach(LeaseAttachRequest) returns (LeaseAttachResponse) {}
// LeaseTnx likes Tnx. It has two addition success and failure LeaseAttachRequest list.
// If the Tnx is successful, then the success list will be executed. Or the failure list
// will be executed.
rpc LeaseTnx(LeaseTnxRequest) returns (LeaseTnxResponse) {}
// KeepAlive keeps the lease alive.
rpc LeaseKeepAlive(stream LeaseKeepAliveRequest) returns (stream LeaseKeepAliveResponse) {}
}
message ResponseHeader {
// an error type message?
optional string error = 1;
optional uint64 cluster_id = 2;
optional uint64 member_id = 3;
// index of the store when the request was applied.
optional int64 index = 4;
// term of raft when the request was applied.
optional uint64 raft_term = 5;
}
message RangeRequest {
// if the range_end is not given, the request returns the key.
optional bytes key = 1;
// if the range_end is given, it gets the keys in range [key, range_end).
optional bytes range_end = 2;
// limit the number of keys returned.
optional int64 limit = 3;
// the response will be consistent with previous request with same token if the token is
// given and is vaild.
optional bytes consistent_token = 4;
}
message RangeResponse {
optional ResponseHeader header = 1;
repeated KeyValue kvs = 2;
optional bytes consistent_token = 3;
}
message PutRequest {
optional bytes key = 1;
optional bytes value = 2;
}
message PutResponse {
optional ResponseHeader header = 1;
}
message DeleteRangeRequest {
// if the range_end is not given, the request deletes the key.
optional bytes key = 1;
// if the range_end is given, it deletes the keys in range [key, range_end).
optional bytes range_end = 2;
}
message DeleteRangeResponse {
optional ResponseHeader header = 1;
}
message RequestUnion {
oneof request {
RangeRequest request_range = 1;
PutRequest request_put = 2;
DeleteRangeRequest request_delete_range = 3;
}
}
message ResponseUnion {
oneof response {
RangeResponse reponse_range = 1;
PutResponse response_put = 2;
DeleteRangeResponse response_delete_range = 3;
}
}
message Compare {
enum CompareType {
EQUAL = 0;
GREATER = 1;
LESS = 2;
}
optional CompareType type = 1;
// key path
optional bytes key = 2;
oneof target {
// version of the given key
int64 version = 3;
// create index of the given key
int64 create_index = 4;
// last modified index of the given key
int64 mod_index = 5;
// value of the given key
bytes value = 6;
}
}
// First all the compare requests are processed.
// If all the compare succeed, all the success
// requests will be processed.
// Or all the failure requests will be processed and
// all the errors in the comparison will be returned.
// From google paxosdb paper:
// Our implementation hinges around a powerful primitive which we call MultiOp. All other database
// operations except for iteration are implemented as a single call to MultiOp. A MultiOp is applied atomically
// and consists of three components:
// 1. A list of tests called guard. Each test in guard checks a single entry in the database. It may check
// for the absence or presence of a value, or compare with a given value. Two different tests in the guard
// may apply to the same or different entries in the database. All tests in the guard are applied and
// MultiOp returns the results. If all tests are true, MultiOp executes t op (see item 2 below), otherwise
// it executes f op (see item 3 below).
// 2. A list of database operations called t op. Each operation in the list is either an insert, delete, or
// lookup operation, and applies to a single database entry. Two different operations in the list may apply
// to the same or different entries in the database. These operations are executed
// if guard evaluates to
// true.
// 3. A list of database operations called f op. Like t op, but executed if guard evaluates to false.
message TnxRequest {
repeated Compare compare = 1;
repeated RequestUnion success = 2;
repeated RequestUnion failure = 3;
}
message TnxResponse {
optional ResponseHeader header = 1;
optional bool succeeded = 2;
repeated ResponseUnion responses = 3;
}
message KeyValue {
optional bytes key = 1;
// mod_index is the last modified index of the key.
optional int64 create_index = 2;
optional int64 mod_index = 3;
// version is the version of the key. A deletion resets
// the version to zero and any modification of the key
// increases its version.
optional int64 version = 4;
optional bytes value = 5;
}
message WatchRangeRequest {
// if the range_end is not given, the request returns the key.
optional bytes key = 1;
// if the range_end is given, it gets the keys in range [key, range_end).
optional bytes range_end = 2;
// start_index is an optional index (including) to watch from. No start_index is "now".
optional int64 start_index = 3;
// end_index is an optional index (excluding) to end watch. No end_index is "forever".
optional int64 end_index = 4;
optional bool progress_notification = 5;
}
message WatchRangeResponse {
optional ResponseHeader header = 1;
repeated Event events = 2;
}
message Event {
enum EventType {
PUT = 0;
DELETE = 1;
EXPIRE = 2;
}
optional EventType event_type = 1;
// a put event contains the current key-value
// a delete/expire event contains the previous
// key-value
optional KeyValue kv = 2;
}
message CompactionRequest {
optional int64 index = 1;
}
message CompactionResponse {
optional ResponseHeader header = 1;
}
message LeaseCreateRequest {
// advisory ttl in seconds
optional int64 ttl = 1;
}
message LeaseCreateResponse {
optional ResponseHeader header = 1;
optional int64 lease_id = 2;
// server decided ttl in second
optional int64 ttl = 3;
optional string error = 4;
}
message LeaseRevokeRequest {
optional int64 lease_id = 1;
}
message LeaseRevokeResponse {
optional ResponseHeader header = 1;
}
message LeaseTnxRequest {
optional TnxRequest request = 1;
repeated LeaseAttachRequest success = 2;
repeated LeaseAttachRequest failure = 3;
}
message LeaseTnxResponse {
optional ResponseHeader header = 1;
optional TnxResponse response = 2;
repeated LeaseAttachResponse attach_responses = 3;
}
message LeaseAttachRequest {
optional int64 lease_id = 1;
optional bytes key = 2;
}
message LeaseAttachResponse {
optional ResponseHeader header = 1;
}
message LeaseKeepAliveRequest {
optional int64 lease_id = 1;
}
message LeaseKeepAliveResponse {
optional ResponseHeader header = 1;
optional int64 lease_id = 2;
optional int64 ttl = 3;
}

View File

@@ -93,7 +93,7 @@ Adding a member is a two step process:
* Add the new member to the cluster via the [members API](other_apis.md#post-v2members) or the `etcdctl member add` command.
* Start the new member with the new cluster configuration, including a list of the updated members (existing members + the new member).
Using `etcdctl` let's add the new member to the cluster by specifing its [name](configuration.md#-name) and [advertised peer URLs](configuration.md#-initial-advertise-peer-urls):
Using `etcdctl` let's add the new member to the cluster by specifying its [name](configuration.md#-name) and [advertised peer URLs](configuration.md#-initial-advertise-peer-urls):
```
$ etcdctl member add infra3 http://10.0.1.13:2380

View File

@@ -147,6 +147,13 @@ The etcd members will form a cluster and all communication between members in th
## Frequently Asked Questions
### My cluster is not working with peer tls configuration?
The internal protocol of etcd v2.0.x uses a lot of short-lived HTTP connections.
So, when enabling TLS you may need to increase the heartbeat interval and election timeouts to reduce internal cluster connection churn.
A reasonable place to start are these values: ` --heartbeat-interval 500 --election-timeout 2500`.
This issues is resolved in the etcd v2.1.x series of releases which uses fewer connections.
### I'm seeing a SSLv3 alert handshake failure when using SSL client authentication?
The `crypto/tls` package of `golang` checks the key usage of the certificate public key before using it.

View File

@@ -25,8 +25,8 @@ The election timeout should be set based on the heartbeat interval and your netw
Election timeouts should be at least 10 times your ping time so it can account for variance in your network.
For example, if the ping time between your nodes is 10ms then you should have at least a 100ms election timeout.
You should also set your election timeout to at least 4 to 5 times your heartbeat interval to account for variance in leader replication.
For a heartbeat interval of 50ms you should set your election timeout to at least 200ms - 250ms.
You should also set your election timeout to at least 5 to 10 times your heartbeat interval to account for variance in leader replication.
For a heartbeat interval of 50ms you should set your election timeout to at least 250ms - 500ms.
You can override the default values on the command line:
@@ -62,13 +62,3 @@ $ etcd -snapshot-count=5000
# Environment variables:
$ ETCD_SNAPSHOT_COUNT=5000 etcd
```
You can also disable snapshotting by adding the following to your command line:
```sh
# Command line arguments:
$ etcd -snapshot false
# Environment variables:
$ ETCD_SNAPSHOT=false etcd
```

View File

@@ -0,0 +1,112 @@
## Upgrade etcd to 2.1
In the general case, upgrading from etcd 2.0 to 2.1 can be a zero-downtime, rolling upgrade:
- one by one, stop the etcd v2.0 processes and replace them with etcd v2.1 processes
- after you are running all v2.1 processes, new features in v2.1 are available to the cluster
Before [starting an upgrade](#upgrade-procedure), read through the rest of this guide to prepare.
### Upgrade Checklists
#### Upgrade Requirement
To upgrade an existing etcd deployment to 2.1, you must be running 2.0. If youre running a version of etcd before 2.0, you must upgrade to [2.0](https://github.com/coreos/etcd/releases/tag/v2.0.13) before upgrading to 2.1.
Also, to ensure a smooth rolling upgrade, your running cluster must be healthy. You can check the health of the cluster by using `etcdctl cluster-health` command.
#### Preparedness
Before upgrading etcd, always test the services relying on etcd in a staging environment before deploying the upgrade to the production environment.
You might also want to [backup your data directory](admin_guide.md#backing-up-the-datastore) for a potential [downgrade](#downgrade).
etcd 2.1 introduces a new [authentication](auth_api.md) feature, which is disabled by default. If your deployment depends on these, you may want to test the auth features before enabling them in production.
#### Mixed Versions
While upgrading, an etcd cluster supports mixed versions of etcd members. The cluster is only considered upgraded once all its members are upgraded to 2.1.
Internally, etcd members negotiate with each other to determine the overall etcd cluster version, which controls the reported cluster version and the supported features. For example, if you are mid-upgrade, any 2.1 features (such as the the authentication feature mentioned above) wont be available.
#### Limitations
If you encounter any issues during the upgrade, you can attempt to restart the etcd process in trouble using a newer v2.1 binary to solve the problem. One known issue is that etcd v2.0.0 and v2.0.2 may panic during rolling upgrades due to an existing bug, which has been fixed since etcd v2.0.3.
It might take up to 2 minutes for the newly upgraded member to catch up with the existing cluster when the total data size is larger than 50MB (You can check the size of the existing snapshot to know about the rough data size). In other words, it is safest to wait for 2 minutes before upgrading the next member.
If you have even more data, this might take more time. If you have a data size larger than 100MB you should contact us before upgrading, so we can make sure the upgrades work smoothly.
#### Downgrade
If all members have been upgraded to v2.1, the cluster will be upgraded to v2.1, and downgrade is **not possible**. If any member is still v2.0, the cluster will remain in v2.0, and you can go back to use v2.0 binary.
Please [backup your data directory](admin_guide.md#backing-up-the-datastore) of all etcd members if you want to downgrade the cluster, even if it is upgraded.
### Upgrade Procedure
#### 1. Check upgrade requirements.
```
$ etcdctl cluster-health
cluster is healthy
member 6e3bd23ae5f1eae0 is healthy
member 924e2e83e93f2560 is healthy
member a8266ecf031671f3 is healthy
$ curl http://127.0.0.1:4001/version
etcd 2.0.x
```
#### 2. Stop the existing etcd process
You will see similar error logging from other etcd processes in your cluster. This is normal, since you just shut down a member.
```
2015/06/23 15:45:09 sender: error posting to 6e3bd23ae5f1eae0: dial tcp 127.0.0.1:7002: connection refused
2015/06/23 15:45:09 sender: the connection with 6e3bd23ae5f1eae0 became inactive
2015/06/23 15:45:11 rafthttp: encountered error writing to server log stream: write tcp 127.0.0.1:53783: broken pipe
2015/06/23 15:45:11 rafthttp: server streaming to 6e3bd23ae5f1eae0 at term 2 has been stopped
2015/06/23 15:45:11 stream: error sending message: stopped
2015/06/23 15:45:11 stream: stopping the stream server...
```
You could [backup your data directory](https://github.com/coreos/etcd/blob/7f7e2cc79d9c5c342a6eb1e48c386b0223cf934e/Documentation/admin_guide.md#backing-up-the-datastore) for data safety.
```
$ etcdctl backup \
--data-dir /var/lib/etcd \
--backup-dir /tmp/etcd_backup
```
#### 3. Drop-in etcd v2.1 binary and start the new etcd process
You will see the etcd publish its information to the cluster.
```
2015/06/23 15:45:39 etcdserver: published {Name:infra2 ClientURLs:[http://localhost:4002]} to cluster e9c7614f68f35fb2
```
You could verify the cluster becomes healthy.
```
$ etcdctl cluster-health
cluster is healthy
member 6e3bd23ae5f1eae0 is healthy
member 924e2e83e93f2560 is healthy
member a8266ecf031671f3 is healthy
```
#### 4. Repeat step 2 to step 3 for all other members
#### 5. Finish
When all members are upgraded, you will see the cluster is upgraded to 2.1 successfully:
```
2015/06/23 15:46:35 etcdserver: updated the cluster version from 2.0.0 to 2.1.0
```
```
$ curl http://127.0.0.1:4001/version
{"etcdserver":"2.1.x","etcdcluster":"2.1.0"}
```

93
Godeps/Godeps.json generated
View File

@@ -6,9 +6,26 @@
],
"Deps": [
{
"ImportPath": "code.google.com/p/goprotobuf/proto",
"Comment": "go.r60-163",
"Rev": "9352842ae63ee1d7e74e074ce7bb10370c4b6b9e"
"ImportPath": "bitbucket.org/ww/goautoneg",
"Comment": "null-5",
"Rev": "75cd24fc2f2c2a2088577d12123ddee5f54e0675"
},
{
"ImportPath": "github.com/beorn7/perks/quantile",
"Rev": "b965b613227fddccbfffe13eae360ed3fa822f8d"
},
{
"ImportPath": "github.com/bgentry/speakeasy",
"Rev": "5dfe43257d1f86b96484e760f2f0c4e2559089c7"
},
{
"ImportPath": "github.com/boltdb/bolt",
"Comment": "v1.0-71-g71f28ea",
"Rev": "71f28eaecbebd00604d87bb1de0dae8fcfa54bbd"
},
{
"ImportPath": "github.com/bradfitz/http2",
"Rev": "3e36af6d3af0e56fa3da71099f864933dea3d9fb"
},
{
"ImportPath": "github.com/codegangsta/cli",
@@ -17,49 +34,55 @@
},
{
"ImportPath": "github.com/coreos/go-etcd/etcd",
"Comment": "v0.2.0-rc1-130-g6aa2da5",
"Rev": "6aa2da5a7a905609c93036b9307185a04a5a84a5"
"Comment": "v2.0.0-13-g4cceaf7",
"Rev": "4cceaf7283b76f27c4a732b20730dcdb61053bf5"
},
{
"ImportPath": "github.com/coreos/go-semver/semver",
"Rev": "568e959cd89871e61434c1143528d9162da89ef2"
},
{
"ImportPath": "github.com/coreos/pkg/capnslog",
"Rev": "99f6e6b8f8ea30b0f82769c1411691c44a66d015"
},
{
"ImportPath": "github.com/gogo/protobuf/proto",
"Rev": "bc946d07d1016848dfd2507f90f0859c9471681e"
"Rev": "64f27bf06efee53589314a6e5a4af34cdd85adf6"
},
{
"ImportPath": "github.com/golang/glog",
"Rev": "44145f04b68cf362d9c4df2182967c2275eaefed"
},
{
"ImportPath": "github.com/golang/protobuf/proto",
"Rev": "5677a0e3d5e89854c9974e1256839ee23f8233ca"
},
{
"ImportPath": "github.com/google/btree",
"Rev": "cc6329d4279e3f025a53a83c397d2339b5705c45"
},
{
"ImportPath": "github.com/jonboulle/clockwork",
"Rev": "72f9bd7c4e0c2a40055ab3d0f09654f730cce982"
},
{
"ImportPath": "github.com/matttproud/golang_protobuf_extensions/ext",
"Rev": "7a864a042e844af638df17ebbabf8183dace556a"
},
{
"ImportPath": "github.com/prometheus/client_golang/_vendor/goautoneg",
"Comment": "0.1.0-22-g70f5497",
"Rev": "70f54973fb9187e8773d738cb6ef6881333f5f25"
},
{
"ImportPath": "github.com/prometheus/client_golang/_vendor/perks/quantile",
"Comment": "0.1.0-22-g70f5497",
"Rev": "70f54973fb9187e8773d738cb6ef6881333f5f25"
"ImportPath": "github.com/matttproud/golang_protobuf_extensions/pbutil",
"Rev": "fc2b8d3a73c4867e51861bbdd5ae3c1f0869dd6a"
},
{
"ImportPath": "github.com/prometheus/client_golang/model",
"Comment": "0.1.0-22-g70f5497",
"Rev": "70f54973fb9187e8773d738cb6ef6881333f5f25"
"Comment": "0.5.0-10-ga842dc1",
"Rev": "a842dc11e0621c34a71cab634d1d0190a59802a8"
},
{
"ImportPath": "github.com/prometheus/client_golang/prometheus",
"Comment": "0.1.0-22-g70f5497",
"Rev": "70f54973fb9187e8773d738cb6ef6881333f5f25"
"Comment": "0.5.0-10-ga842dc1",
"Rev": "a842dc11e0621c34a71cab634d1d0190a59802a8"
},
{
"ImportPath": "github.com/prometheus/client_golang/text",
"Comment": "0.1.0-22-g70f5497",
"Rev": "70f54973fb9187e8773d738cb6ef6881333f5f25"
"Comment": "0.5.0-10-ga842dc1",
"Rev": "a842dc11e0621c34a71cab634d1d0190a59802a8"
},
{
"ImportPath": "github.com/prometheus/client_model/go",
@@ -68,12 +91,16 @@
},
{
"ImportPath": "github.com/prometheus/procfs",
"Rev": "92faa308558161acab0ada1db048e9996ecec160"
"Rev": "ee2372b58cee877abe07cde670d04d3b3bac5ee6"
},
{
"ImportPath": "github.com/stretchr/testify/assert",
"Rev": "9cc77fa25329013ce07362c7742952ff887361f2"
},
{
"ImportPath": "github.com/ugorji/go/codec",
"Rev": "821cda7e48749cacf7cad2c6ed01e96457ca7e9d"
},
{
"ImportPath": "golang.org/x/crypto/bcrypt",
"Rev": "1351f936d976c60a0a48d728281922cf63eafb8d"
@@ -85,6 +112,22 @@
{
"ImportPath": "golang.org/x/net/context",
"Rev": "7dbad50ab5b31073856416cdcfeb2796d682f844"
},
{
"ImportPath": "golang.org/x/oauth2",
"Rev": "3046bc76d6dfd7d3707f6640f85e42d9c4050f50"
},
{
"ImportPath": "google.golang.org/cloud/compute/metadata",
"Rev": "f20d6dcccb44ed49de45ae3703312cb46e627db1"
},
{
"ImportPath": "google.golang.org/cloud/internal",
"Rev": "f20d6dcccb44ed49de45ae3703312cb46e627db1"
},
{
"ImportPath": "google.golang.org/grpc",
"Rev": "f5ebd86be717593ab029545492c93ddf8914832b"
}
]
}

View File

@@ -1,40 +0,0 @@
# Go support for Protocol Buffers - Google's data interchange format
#
# Copyright 2010 The Go Authors. All rights reserved.
# http://code.google.com/p/goprotobuf/
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are
# met:
#
# * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# * Redistributions in binary form must reproduce the above
# copyright notice, this list of conditions and the following disclaimer
# in the documentation and/or other materials provided with the
# distribution.
# * Neither the name of Google Inc. nor the names of its
# contributors may be used to endorse or promote products derived from
# this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
install:
go install
test: install generate-test-pbs
go test
generate-test-pbs:
make install && cd testdata && make

File diff suppressed because it is too large Load Diff

View File

@@ -1,174 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2011 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
// Protocol buffer deep copy.
// TODO: MessageSet and RawMessage.
package proto
import (
"log"
"reflect"
"strings"
)
// Clone returns a deep copy of a protocol buffer.
func Clone(pb Message) Message {
in := reflect.ValueOf(pb)
if in.IsNil() {
return pb
}
out := reflect.New(in.Type().Elem())
// out is empty so a merge is a deep copy.
mergeStruct(out.Elem(), in.Elem())
return out.Interface().(Message)
}
// Merge merges src into dst.
// Required and optional fields that are set in src will be set to that value in dst.
// Elements of repeated fields will be appended.
// Merge panics if src and dst are not the same type, or if dst is nil.
func Merge(dst, src Message) {
in := reflect.ValueOf(src)
out := reflect.ValueOf(dst)
if out.IsNil() {
panic("proto: nil destination")
}
if in.Type() != out.Type() {
// Explicit test prior to mergeStruct so that mistyped nils will fail
panic("proto: type mismatch")
}
if in.IsNil() {
// Merging nil into non-nil is a quiet no-op
return
}
mergeStruct(out.Elem(), in.Elem())
}
func mergeStruct(out, in reflect.Value) {
for i := 0; i < in.NumField(); i++ {
f := in.Type().Field(i)
if strings.HasPrefix(f.Name, "XXX_") {
continue
}
mergeAny(out.Field(i), in.Field(i))
}
if emIn, ok := in.Addr().Interface().(extendableProto); ok {
emOut := out.Addr().Interface().(extendableProto)
mergeExtension(emOut.ExtensionMap(), emIn.ExtensionMap())
}
uf := in.FieldByName("XXX_unrecognized")
if !uf.IsValid() {
return
}
uin := uf.Bytes()
if len(uin) > 0 {
out.FieldByName("XXX_unrecognized").SetBytes(append([]byte(nil), uin...))
}
}
func mergeAny(out, in reflect.Value) {
if in.Type() == protoMessageType {
if !in.IsNil() {
if out.IsNil() {
out.Set(reflect.ValueOf(Clone(in.Interface().(Message))))
} else {
Merge(out.Interface().(Message), in.Interface().(Message))
}
}
return
}
switch in.Kind() {
case reflect.Bool, reflect.Float32, reflect.Float64, reflect.Int32, reflect.Int64,
reflect.String, reflect.Uint32, reflect.Uint64:
out.Set(in)
case reflect.Ptr:
if in.IsNil() {
return
}
if out.IsNil() {
out.Set(reflect.New(in.Elem().Type()))
}
mergeAny(out.Elem(), in.Elem())
case reflect.Slice:
if in.IsNil() {
return
}
if in.Type().Elem().Kind() == reflect.Uint8 {
// []byte is a scalar bytes field, not a repeated field.
// Make a deep copy.
// Append to []byte{} instead of []byte(nil) so that we never end up
// with a nil result.
out.SetBytes(append([]byte{}, in.Bytes()...))
return
}
n := in.Len()
if out.IsNil() {
out.Set(reflect.MakeSlice(in.Type(), 0, n))
}
switch in.Type().Elem().Kind() {
case reflect.Bool, reflect.Float32, reflect.Float64, reflect.Int32, reflect.Int64,
reflect.String, reflect.Uint32, reflect.Uint64:
out.Set(reflect.AppendSlice(out, in))
default:
for i := 0; i < n; i++ {
x := reflect.Indirect(reflect.New(in.Type().Elem()))
mergeAny(x, in.Index(i))
out.Set(reflect.Append(out, x))
}
}
case reflect.Struct:
mergeStruct(out, in)
default:
// unknown type, so not a protocol buffer
log.Printf("proto: don't know how to copy %v", in)
}
}
func mergeExtension(out, in map[int32]Extension) {
for extNum, eIn := range in {
eOut := Extension{desc: eIn.desc}
if eIn.value != nil {
v := reflect.New(reflect.TypeOf(eIn.value)).Elem()
mergeAny(v, reflect.ValueOf(eIn.value))
eOut.value = v.Interface()
}
if eIn.enc != nil {
eOut.enc = make([]byte, len(eIn.enc))
copy(eOut.enc, eIn.enc)
}
out[extNum] = eOut
}
}

View File

@@ -1,201 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2011 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
package proto_test
import (
"testing"
pb "./testdata"
"github.com/coreos/etcd/Godeps/_workspace/src/code.google.com/p/goprotobuf/proto"
)
var cloneTestMessage = &pb.MyMessage{
Count: proto.Int32(42),
Name: proto.String("Dave"),
Pet: []string{"bunny", "kitty", "horsey"},
Inner: &pb.InnerMessage{
Host: proto.String("niles"),
Port: proto.Int32(9099),
Connected: proto.Bool(true),
},
Others: []*pb.OtherMessage{
{
Value: []byte("some bytes"),
},
},
Somegroup: &pb.MyMessage_SomeGroup{
GroupField: proto.Int32(6),
},
RepBytes: [][]byte{[]byte("sham"), []byte("wow")},
}
func init() {
ext := &pb.Ext{
Data: proto.String("extension"),
}
if err := proto.SetExtension(cloneTestMessage, pb.E_Ext_More, ext); err != nil {
panic("SetExtension: " + err.Error())
}
}
func TestClone(t *testing.T) {
m := proto.Clone(cloneTestMessage).(*pb.MyMessage)
if !proto.Equal(m, cloneTestMessage) {
t.Errorf("Clone(%v) = %v", cloneTestMessage, m)
}
// Verify it was a deep copy.
*m.Inner.Port++
if proto.Equal(m, cloneTestMessage) {
t.Error("Mutating clone changed the original")
}
// Byte fields and repeated fields should be copied.
if &m.Pet[0] == &cloneTestMessage.Pet[0] {
t.Error("Pet: repeated field not copied")
}
if &m.Others[0] == &cloneTestMessage.Others[0] {
t.Error("Others: repeated field not copied")
}
if &m.Others[0].Value[0] == &cloneTestMessage.Others[0].Value[0] {
t.Error("Others[0].Value: bytes field not copied")
}
if &m.RepBytes[0] == &cloneTestMessage.RepBytes[0] {
t.Error("RepBytes: repeated field not copied")
}
if &m.RepBytes[0][0] == &cloneTestMessage.RepBytes[0][0] {
t.Error("RepBytes[0]: bytes field not copied")
}
}
func TestCloneNil(t *testing.T) {
var m *pb.MyMessage
if c := proto.Clone(m); !proto.Equal(m, c) {
t.Errorf("Clone(%v) = %v", m, c)
}
}
var mergeTests = []struct {
src, dst, want proto.Message
}{
{
src: &pb.MyMessage{
Count: proto.Int32(42),
},
dst: &pb.MyMessage{
Name: proto.String("Dave"),
},
want: &pb.MyMessage{
Count: proto.Int32(42),
Name: proto.String("Dave"),
},
},
{
src: &pb.MyMessage{
Inner: &pb.InnerMessage{
Host: proto.String("hey"),
Connected: proto.Bool(true),
},
Pet: []string{"horsey"},
Others: []*pb.OtherMessage{
{
Value: []byte("some bytes"),
},
},
},
dst: &pb.MyMessage{
Inner: &pb.InnerMessage{
Host: proto.String("niles"),
Port: proto.Int32(9099),
},
Pet: []string{"bunny", "kitty"},
Others: []*pb.OtherMessage{
{
Key: proto.Int64(31415926535),
},
{
// Explicitly test a src=nil field
Inner: nil,
},
},
},
want: &pb.MyMessage{
Inner: &pb.InnerMessage{
Host: proto.String("hey"),
Connected: proto.Bool(true),
Port: proto.Int32(9099),
},
Pet: []string{"bunny", "kitty", "horsey"},
Others: []*pb.OtherMessage{
{
Key: proto.Int64(31415926535),
},
{},
{
Value: []byte("some bytes"),
},
},
},
},
{
src: &pb.MyMessage{
RepBytes: [][]byte{[]byte("wow")},
},
dst: &pb.MyMessage{
Somegroup: &pb.MyMessage_SomeGroup{
GroupField: proto.Int32(6),
},
RepBytes: [][]byte{[]byte("sham")},
},
want: &pb.MyMessage{
Somegroup: &pb.MyMessage_SomeGroup{
GroupField: proto.Int32(6),
},
RepBytes: [][]byte{[]byte("sham"), []byte("wow")},
},
},
// Check that a scalar bytes field replaces rather than appends.
{
src: &pb.OtherMessage{Value: []byte("foo")},
dst: &pb.OtherMessage{Value: []byte("bar")},
want: &pb.OtherMessage{Value: []byte("foo")},
},
}
func TestMerge(t *testing.T) {
for _, m := range mergeTests {
got := proto.Clone(m.dst)
proto.Merge(got, m.src)
if !proto.Equal(got, m.want) {
t.Errorf("Merge(%v, %v)\n got %v\nwant %v\n", m.dst, m.src, got, m.want)
}
}
}

View File

@@ -1,721 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2010 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
package proto
/*
* Routines for decoding protocol buffer data to construct in-memory representations.
*/
import (
"errors"
"fmt"
"io"
"os"
"reflect"
)
// errOverflow is returned when an integer is too large to be represented.
var errOverflow = errors.New("proto: integer overflow")
// The fundamental decoders that interpret bytes on the wire.
// Those that take integer types all return uint64 and are
// therefore of type valueDecoder.
// DecodeVarint reads a varint-encoded integer from the slice.
// It returns the integer and the number of bytes consumed, or
// zero if there is not enough.
// This is the format for the
// int32, int64, uint32, uint64, bool, and enum
// protocol buffer types.
func DecodeVarint(buf []byte) (x uint64, n int) {
// x, n already 0
for shift := uint(0); shift < 64; shift += 7 {
if n >= len(buf) {
return 0, 0
}
b := uint64(buf[n])
n++
x |= (b & 0x7F) << shift
if (b & 0x80) == 0 {
return x, n
}
}
// The number is too large to represent in a 64-bit value.
return 0, 0
}
// DecodeVarint reads a varint-encoded integer from the Buffer.
// This is the format for the
// int32, int64, uint32, uint64, bool, and enum
// protocol buffer types.
func (p *Buffer) DecodeVarint() (x uint64, err error) {
// x, err already 0
i := p.index
l := len(p.buf)
for shift := uint(0); shift < 64; shift += 7 {
if i >= l {
err = io.ErrUnexpectedEOF
return
}
b := p.buf[i]
i++
x |= (uint64(b) & 0x7F) << shift
if b < 0x80 {
p.index = i
return
}
}
// The number is too large to represent in a 64-bit value.
err = errOverflow
return
}
// DecodeFixed64 reads a 64-bit integer from the Buffer.
// This is the format for the
// fixed64, sfixed64, and double protocol buffer types.
func (p *Buffer) DecodeFixed64() (x uint64, err error) {
// x, err already 0
i := p.index + 8
if i < 0 || i > len(p.buf) {
err = io.ErrUnexpectedEOF
return
}
p.index = i
x = uint64(p.buf[i-8])
x |= uint64(p.buf[i-7]) << 8
x |= uint64(p.buf[i-6]) << 16
x |= uint64(p.buf[i-5]) << 24
x |= uint64(p.buf[i-4]) << 32
x |= uint64(p.buf[i-3]) << 40
x |= uint64(p.buf[i-2]) << 48
x |= uint64(p.buf[i-1]) << 56
return
}
// DecodeFixed32 reads a 32-bit integer from the Buffer.
// This is the format for the
// fixed32, sfixed32, and float protocol buffer types.
func (p *Buffer) DecodeFixed32() (x uint64, err error) {
// x, err already 0
i := p.index + 4
if i < 0 || i > len(p.buf) {
err = io.ErrUnexpectedEOF
return
}
p.index = i
x = uint64(p.buf[i-4])
x |= uint64(p.buf[i-3]) << 8
x |= uint64(p.buf[i-2]) << 16
x |= uint64(p.buf[i-1]) << 24
return
}
// DecodeZigzag64 reads a zigzag-encoded 64-bit integer
// from the Buffer.
// This is the format used for the sint64 protocol buffer type.
func (p *Buffer) DecodeZigzag64() (x uint64, err error) {
x, err = p.DecodeVarint()
if err != nil {
return
}
x = (x >> 1) ^ uint64((int64(x&1)<<63)>>63)
return
}
// DecodeZigzag32 reads a zigzag-encoded 32-bit integer
// from the Buffer.
// This is the format used for the sint32 protocol buffer type.
func (p *Buffer) DecodeZigzag32() (x uint64, err error) {
x, err = p.DecodeVarint()
if err != nil {
return
}
x = uint64((uint32(x) >> 1) ^ uint32((int32(x&1)<<31)>>31))
return
}
// These are not ValueDecoders: they produce an array of bytes or a string.
// bytes, embedded messages
// DecodeRawBytes reads a count-delimited byte buffer from the Buffer.
// This is the format used for the bytes protocol buffer
// type and for embedded messages.
func (p *Buffer) DecodeRawBytes(alloc bool) (buf []byte, err error) {
n, err := p.DecodeVarint()
if err != nil {
return
}
nb := int(n)
if nb < 0 {
return nil, fmt.Errorf("proto: bad byte length %d", nb)
}
end := p.index + nb
if end < p.index || end > len(p.buf) {
return nil, io.ErrUnexpectedEOF
}
if !alloc {
// todo: check if can get more uses of alloc=false
buf = p.buf[p.index:end]
p.index += nb
return
}
buf = make([]byte, nb)
copy(buf, p.buf[p.index:])
p.index += nb
return
}
// DecodeStringBytes reads an encoded string from the Buffer.
// This is the format used for the proto2 string type.
func (p *Buffer) DecodeStringBytes() (s string, err error) {
buf, err := p.DecodeRawBytes(false)
if err != nil {
return
}
return string(buf), nil
}
// Skip the next item in the buffer. Its wire type is decoded and presented as an argument.
// If the protocol buffer has extensions, and the field matches, add it as an extension.
// Otherwise, if the XXX_unrecognized field exists, append the skipped data there.
func (o *Buffer) skipAndSave(t reflect.Type, tag, wire int, base structPointer, unrecField field) error {
oi := o.index
err := o.skip(t, tag, wire)
if err != nil {
return err
}
if !unrecField.IsValid() {
return nil
}
ptr := structPointer_Bytes(base, unrecField)
// Add the skipped field to struct field
obuf := o.buf
o.buf = *ptr
o.EncodeVarint(uint64(tag<<3 | wire))
*ptr = append(o.buf, obuf[oi:o.index]...)
o.buf = obuf
return nil
}
// Skip the next item in the buffer. Its wire type is decoded and presented as an argument.
func (o *Buffer) skip(t reflect.Type, tag, wire int) error {
var u uint64
var err error
switch wire {
case WireVarint:
_, err = o.DecodeVarint()
case WireFixed64:
_, err = o.DecodeFixed64()
case WireBytes:
_, err = o.DecodeRawBytes(false)
case WireFixed32:
_, err = o.DecodeFixed32()
case WireStartGroup:
for {
u, err = o.DecodeVarint()
if err != nil {
break
}
fwire := int(u & 0x7)
if fwire == WireEndGroup {
break
}
ftag := int(u >> 3)
err = o.skip(t, ftag, fwire)
if err != nil {
break
}
}
default:
err = fmt.Errorf("proto: can't skip unknown wire type %d for %s", wire, t)
}
return err
}
// Unmarshaler is the interface representing objects that can
// unmarshal themselves. The method should reset the receiver before
// decoding starts. The argument points to data that may be
// overwritten, so implementations should not keep references to the
// buffer.
type Unmarshaler interface {
Unmarshal([]byte) error
}
// Unmarshal parses the protocol buffer representation in buf and places the
// decoded result in pb. If the struct underlying pb does not match
// the data in buf, the results can be unpredictable.
//
// Unmarshal resets pb before starting to unmarshal, so any
// existing data in pb is always removed. Use UnmarshalMerge
// to preserve and append to existing data.
func Unmarshal(buf []byte, pb Message) error {
pb.Reset()
return UnmarshalMerge(buf, pb)
}
// UnmarshalMerge parses the protocol buffer representation in buf and
// writes the decoded result to pb. If the struct underlying pb does not match
// the data in buf, the results can be unpredictable.
//
// UnmarshalMerge merges into existing data in pb.
// Most code should use Unmarshal instead.
func UnmarshalMerge(buf []byte, pb Message) error {
// If the object can unmarshal itself, let it.
if u, ok := pb.(Unmarshaler); ok {
return u.Unmarshal(buf)
}
return NewBuffer(buf).Unmarshal(pb)
}
// Unmarshal parses the protocol buffer representation in the
// Buffer and places the decoded result in pb. If the struct
// underlying pb does not match the data in the buffer, the results can be
// unpredictable.
func (p *Buffer) Unmarshal(pb Message) error {
// If the object can unmarshal itself, let it.
if u, ok := pb.(Unmarshaler); ok {
err := u.Unmarshal(p.buf[p.index:])
p.index = len(p.buf)
return err
}
typ, base, err := getbase(pb)
if err != nil {
return err
}
err = p.unmarshalType(typ.Elem(), GetProperties(typ.Elem()), false, base)
if collectStats {
stats.Decode++
}
return err
}
// unmarshalType does the work of unmarshaling a structure.
func (o *Buffer) unmarshalType(st reflect.Type, prop *StructProperties, is_group bool, base structPointer) error {
var state errorState
required, reqFields := prop.reqCount, uint64(0)
var err error
for err == nil && o.index < len(o.buf) {
oi := o.index
var u uint64
u, err = o.DecodeVarint()
if err != nil {
break
}
wire := int(u & 0x7)
if wire == WireEndGroup {
if is_group {
return nil // input is satisfied
}
return fmt.Errorf("proto: %s: wiretype end group for non-group", st)
}
tag := int(u >> 3)
if tag <= 0 {
return fmt.Errorf("proto: %s: illegal tag %d (wire type %d)", st, tag, wire)
}
fieldnum, ok := prop.decoderTags.get(tag)
if !ok {
// Maybe it's an extension?
if prop.extendable {
if e := structPointer_Interface(base, st).(extendableProto); isExtensionField(e, int32(tag)) {
if err = o.skip(st, tag, wire); err == nil {
ext := e.ExtensionMap()[int32(tag)] // may be missing
ext.enc = append(ext.enc, o.buf[oi:o.index]...)
e.ExtensionMap()[int32(tag)] = ext
}
continue
}
}
err = o.skipAndSave(st, tag, wire, base, prop.unrecField)
continue
}
p := prop.Prop[fieldnum]
if p.dec == nil {
fmt.Fprintf(os.Stderr, "proto: no protobuf decoder for %s.%s\n", st, st.Field(fieldnum).Name)
continue
}
dec := p.dec
if wire != WireStartGroup && wire != p.WireType {
if wire == WireBytes && p.packedDec != nil {
// a packable field
dec = p.packedDec
} else {
err = fmt.Errorf("proto: bad wiretype for field %s.%s: got wiretype %d, want %d", st, st.Field(fieldnum).Name, wire, p.WireType)
continue
}
}
decErr := dec(o, p, base)
if decErr != nil && !state.shouldContinue(decErr, p) {
err = decErr
}
if err == nil && p.Required {
// Successfully decoded a required field.
if tag <= 64 {
// use bitmap for fields 1-64 to catch field reuse.
var mask uint64 = 1 << uint64(tag-1)
if reqFields&mask == 0 {
// new required field
reqFields |= mask
required--
}
} else {
// This is imprecise. It can be fooled by a required field
// with a tag > 64 that is encoded twice; that's very rare.
// A fully correct implementation would require allocating
// a data structure, which we would like to avoid.
required--
}
}
}
if err == nil {
if is_group {
return io.ErrUnexpectedEOF
}
if state.err != nil {
return state.err
}
if required > 0 {
// Not enough information to determine the exact field. If we use extra
// CPU, we could determine the field only if the missing required field
// has a tag <= 64 and we check reqFields.
return &RequiredNotSetError{"{Unknown}"}
}
}
return err
}
// Individual type decoders
// For each,
// u is the decoded value,
// v is a pointer to the field (pointer) in the struct
// Sizes of the pools to allocate inside the Buffer.
// The goal is modest amortization and allocation
// on at least 16-byte boundaries.
const (
boolPoolSize = 16
uint32PoolSize = 8
uint64PoolSize = 4
)
// Decode a bool.
func (o *Buffer) dec_bool(p *Properties, base structPointer) error {
u, err := p.valDec(o)
if err != nil {
return err
}
if len(o.bools) == 0 {
o.bools = make([]bool, boolPoolSize)
}
o.bools[0] = u != 0
*structPointer_Bool(base, p.field) = &o.bools[0]
o.bools = o.bools[1:]
return nil
}
// Decode an int32.
func (o *Buffer) dec_int32(p *Properties, base structPointer) error {
u, err := p.valDec(o)
if err != nil {
return err
}
word32_Set(structPointer_Word32(base, p.field), o, uint32(u))
return nil
}
// Decode an int64.
func (o *Buffer) dec_int64(p *Properties, base structPointer) error {
u, err := p.valDec(o)
if err != nil {
return err
}
word64_Set(structPointer_Word64(base, p.field), o, u)
return nil
}
// Decode a string.
func (o *Buffer) dec_string(p *Properties, base structPointer) error {
s, err := o.DecodeStringBytes()
if err != nil {
return err
}
sp := new(string)
*sp = s
*structPointer_String(base, p.field) = sp
return nil
}
// Decode a slice of bytes ([]byte).
func (o *Buffer) dec_slice_byte(p *Properties, base structPointer) error {
b, err := o.DecodeRawBytes(true)
if err != nil {
return err
}
*structPointer_Bytes(base, p.field) = b
return nil
}
// Decode a slice of bools ([]bool).
func (o *Buffer) dec_slice_bool(p *Properties, base structPointer) error {
u, err := p.valDec(o)
if err != nil {
return err
}
v := structPointer_BoolSlice(base, p.field)
*v = append(*v, u != 0)
return nil
}
// Decode a slice of bools ([]bool) in packed format.
func (o *Buffer) dec_slice_packed_bool(p *Properties, base structPointer) error {
v := structPointer_BoolSlice(base, p.field)
nn, err := o.DecodeVarint()
if err != nil {
return err
}
nb := int(nn) // number of bytes of encoded bools
y := *v
for i := 0; i < nb; i++ {
u, err := p.valDec(o)
if err != nil {
return err
}
y = append(y, u != 0)
}
*v = y
return nil
}
// Decode a slice of int32s ([]int32).
func (o *Buffer) dec_slice_int32(p *Properties, base structPointer) error {
u, err := p.valDec(o)
if err != nil {
return err
}
structPointer_Word32Slice(base, p.field).Append(uint32(u))
return nil
}
// Decode a slice of int32s ([]int32) in packed format.
func (o *Buffer) dec_slice_packed_int32(p *Properties, base structPointer) error {
v := structPointer_Word32Slice(base, p.field)
nn, err := o.DecodeVarint()
if err != nil {
return err
}
nb := int(nn) // number of bytes of encoded int32s
fin := o.index + nb
if fin < o.index {
return errOverflow
}
for o.index < fin {
u, err := p.valDec(o)
if err != nil {
return err
}
v.Append(uint32(u))
}
return nil
}
// Decode a slice of int64s ([]int64).
func (o *Buffer) dec_slice_int64(p *Properties, base structPointer) error {
u, err := p.valDec(o)
if err != nil {
return err
}
structPointer_Word64Slice(base, p.field).Append(u)
return nil
}
// Decode a slice of int64s ([]int64) in packed format.
func (o *Buffer) dec_slice_packed_int64(p *Properties, base structPointer) error {
v := structPointer_Word64Slice(base, p.field)
nn, err := o.DecodeVarint()
if err != nil {
return err
}
nb := int(nn) // number of bytes of encoded int64s
fin := o.index + nb
if fin < o.index {
return errOverflow
}
for o.index < fin {
u, err := p.valDec(o)
if err != nil {
return err
}
v.Append(u)
}
return nil
}
// Decode a slice of strings ([]string).
func (o *Buffer) dec_slice_string(p *Properties, base structPointer) error {
s, err := o.DecodeStringBytes()
if err != nil {
return err
}
v := structPointer_StringSlice(base, p.field)
*v = append(*v, s)
return nil
}
// Decode a slice of slice of bytes ([][]byte).
func (o *Buffer) dec_slice_slice_byte(p *Properties, base structPointer) error {
b, err := o.DecodeRawBytes(true)
if err != nil {
return err
}
v := structPointer_BytesSlice(base, p.field)
*v = append(*v, b)
return nil
}
// Decode a group.
func (o *Buffer) dec_struct_group(p *Properties, base structPointer) error {
bas := structPointer_GetStructPointer(base, p.field)
if structPointer_IsNil(bas) {
// allocate new nested message
bas = toStructPointer(reflect.New(p.stype))
structPointer_SetStructPointer(base, p.field, bas)
}
return o.unmarshalType(p.stype, p.sprop, true, bas)
}
// Decode an embedded message.
func (o *Buffer) dec_struct_message(p *Properties, base structPointer) (err error) {
raw, e := o.DecodeRawBytes(false)
if e != nil {
return e
}
bas := structPointer_GetStructPointer(base, p.field)
if structPointer_IsNil(bas) {
// allocate new nested message
bas = toStructPointer(reflect.New(p.stype))
structPointer_SetStructPointer(base, p.field, bas)
}
// If the object can unmarshal itself, let it.
if p.isUnmarshaler {
iv := structPointer_Interface(bas, p.stype)
return iv.(Unmarshaler).Unmarshal(raw)
}
obuf := o.buf
oi := o.index
o.buf = raw
o.index = 0
err = o.unmarshalType(p.stype, p.sprop, false, bas)
o.buf = obuf
o.index = oi
return err
}
// Decode a slice of embedded messages.
func (o *Buffer) dec_slice_struct_message(p *Properties, base structPointer) error {
return o.dec_slice_struct(p, false, base)
}
// Decode a slice of embedded groups.
func (o *Buffer) dec_slice_struct_group(p *Properties, base structPointer) error {
return o.dec_slice_struct(p, true, base)
}
// Decode a slice of structs ([]*struct).
func (o *Buffer) dec_slice_struct(p *Properties, is_group bool, base structPointer) error {
v := reflect.New(p.stype)
bas := toStructPointer(v)
structPointer_StructPointerSlice(base, p.field).Append(bas)
if is_group {
err := o.unmarshalType(p.stype, p.sprop, is_group, bas)
return err
}
raw, err := o.DecodeRawBytes(false)
if err != nil {
return err
}
// If the object can unmarshal itself, let it.
if p.isUnmarshaler {
iv := v.Interface()
return iv.(Unmarshaler).Unmarshal(raw)
}
obuf := o.buf
oi := o.index
o.buf = raw
o.index = 0
err = o.unmarshalType(p.stype, p.sprop, is_group, bas)
o.buf = obuf
o.index = oi
return err
}

File diff suppressed because it is too large Load Diff

View File

@@ -1,241 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2011 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
// Protocol buffer comparison.
// TODO: MessageSet.
package proto
import (
"bytes"
"log"
"reflect"
"strings"
)
/*
Equal returns true iff protocol buffers a and b are equal.
The arguments must both be pointers to protocol buffer structs.
Equality is defined in this way:
- Two messages are equal iff they are the same type,
corresponding fields are equal, unknown field sets
are equal, and extensions sets are equal.
- Two set scalar fields are equal iff their values are equal.
If the fields are of a floating-point type, remember that
NaN != x for all x, including NaN.
- Two repeated fields are equal iff their lengths are the same,
and their corresponding elements are equal (a "bytes" field,
although represented by []byte, is not a repeated field)
- Two unset fields are equal.
- Two unknown field sets are equal if their current
encoded state is equal.
- Two extension sets are equal iff they have corresponding
elements that are pairwise equal.
- Every other combination of things are not equal.
The return value is undefined if a and b are not protocol buffers.
*/
func Equal(a, b Message) bool {
if a == nil || b == nil {
return a == b
}
v1, v2 := reflect.ValueOf(a), reflect.ValueOf(b)
if v1.Type() != v2.Type() {
return false
}
if v1.Kind() == reflect.Ptr {
if v1.IsNil() {
return v2.IsNil()
}
if v2.IsNil() {
return false
}
v1, v2 = v1.Elem(), v2.Elem()
}
if v1.Kind() != reflect.Struct {
return false
}
return equalStruct(v1, v2)
}
// v1 and v2 are known to have the same type.
func equalStruct(v1, v2 reflect.Value) bool {
for i := 0; i < v1.NumField(); i++ {
f := v1.Type().Field(i)
if strings.HasPrefix(f.Name, "XXX_") {
continue
}
f1, f2 := v1.Field(i), v2.Field(i)
if f.Type.Kind() == reflect.Ptr {
if n1, n2 := f1.IsNil(), f2.IsNil(); n1 && n2 {
// both unset
continue
} else if n1 != n2 {
// set/unset mismatch
return false
}
b1, ok := f1.Interface().(raw)
if ok {
b2 := f2.Interface().(raw)
// RawMessage
if !bytes.Equal(b1.Bytes(), b2.Bytes()) {
return false
}
continue
}
f1, f2 = f1.Elem(), f2.Elem()
}
if !equalAny(f1, f2) {
return false
}
}
if em1 := v1.FieldByName("XXX_extensions"); em1.IsValid() {
em2 := v2.FieldByName("XXX_extensions")
if !equalExtensions(v1.Type(), em1.Interface().(map[int32]Extension), em2.Interface().(map[int32]Extension)) {
return false
}
}
uf := v1.FieldByName("XXX_unrecognized")
if !uf.IsValid() {
return true
}
u1 := uf.Bytes()
u2 := v2.FieldByName("XXX_unrecognized").Bytes()
if !bytes.Equal(u1, u2) {
return false
}
return true
}
// v1 and v2 are known to have the same type.
func equalAny(v1, v2 reflect.Value) bool {
if v1.Type() == protoMessageType {
m1, _ := v1.Interface().(Message)
m2, _ := v2.Interface().(Message)
return Equal(m1, m2)
}
switch v1.Kind() {
case reflect.Bool:
return v1.Bool() == v2.Bool()
case reflect.Float32, reflect.Float64:
return v1.Float() == v2.Float()
case reflect.Int32, reflect.Int64:
return v1.Int() == v2.Int()
case reflect.Ptr:
return equalAny(v1.Elem(), v2.Elem())
case reflect.Slice:
if v1.Type().Elem().Kind() == reflect.Uint8 {
// short circuit: []byte
if v1.IsNil() != v2.IsNil() {
return false
}
return bytes.Equal(v1.Interface().([]byte), v2.Interface().([]byte))
}
if v1.Len() != v2.Len() {
return false
}
for i := 0; i < v1.Len(); i++ {
if !equalAny(v1.Index(i), v2.Index(i)) {
return false
}
}
return true
case reflect.String:
return v1.Interface().(string) == v2.Interface().(string)
case reflect.Struct:
return equalStruct(v1, v2)
case reflect.Uint32, reflect.Uint64:
return v1.Uint() == v2.Uint()
}
// unknown type, so not a protocol buffer
log.Printf("proto: don't know how to compare %v", v1)
return false
}
// base is the struct type that the extensions are based on.
// em1 and em2 are extension maps.
func equalExtensions(base reflect.Type, em1, em2 map[int32]Extension) bool {
if len(em1) != len(em2) {
return false
}
for extNum, e1 := range em1 {
e2, ok := em2[extNum]
if !ok {
return false
}
m1, m2 := e1.value, e2.value
if m1 != nil && m2 != nil {
// Both are unencoded.
if !equalAny(reflect.ValueOf(m1), reflect.ValueOf(m2)) {
return false
}
continue
}
// At least one is encoded. To do a semantically correct comparison
// we need to unmarshal them first.
var desc *ExtensionDesc
if m := extensionMaps[base]; m != nil {
desc = m[extNum]
}
if desc == nil {
log.Printf("proto: don't know how to compare extension %d of %v", extNum, base)
continue
}
var err error
if m1 == nil {
m1, err = decodeExtension(e1.enc, desc)
}
if m2 == nil && err == nil {
m2, err = decodeExtension(e2.enc, desc)
}
if err != nil {
// The encoded form is invalid.
log.Printf("proto: badly encoded extension %d of %v: %v", extNum, base, err)
return false
}
if !equalAny(reflect.ValueOf(m1), reflect.ValueOf(m2)) {
return false
}
}
return true
}

View File

@@ -1,166 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2011 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
package proto_test
import (
"testing"
pb "./testdata"
. "github.com/coreos/etcd/Godeps/_workspace/src/code.google.com/p/goprotobuf/proto"
)
// Four identical base messages.
// The init function adds extensions to some of them.
var messageWithoutExtension = &pb.MyMessage{Count: Int32(7)}
var messageWithExtension1a = &pb.MyMessage{Count: Int32(7)}
var messageWithExtension1b = &pb.MyMessage{Count: Int32(7)}
var messageWithExtension2 = &pb.MyMessage{Count: Int32(7)}
// Two messages with non-message extensions.
var messageWithInt32Extension1 = &pb.MyMessage{Count: Int32(8)}
var messageWithInt32Extension2 = &pb.MyMessage{Count: Int32(8)}
func init() {
ext1 := &pb.Ext{Data: String("Kirk")}
ext2 := &pb.Ext{Data: String("Picard")}
// messageWithExtension1a has ext1, but never marshals it.
if err := SetExtension(messageWithExtension1a, pb.E_Ext_More, ext1); err != nil {
panic("SetExtension on 1a failed: " + err.Error())
}
// messageWithExtension1b is the unmarshaled form of messageWithExtension1a.
if err := SetExtension(messageWithExtension1b, pb.E_Ext_More, ext1); err != nil {
panic("SetExtension on 1b failed: " + err.Error())
}
buf, err := Marshal(messageWithExtension1b)
if err != nil {
panic("Marshal of 1b failed: " + err.Error())
}
messageWithExtension1b.Reset()
if err := Unmarshal(buf, messageWithExtension1b); err != nil {
panic("Unmarshal of 1b failed: " + err.Error())
}
// messageWithExtension2 has ext2.
if err := SetExtension(messageWithExtension2, pb.E_Ext_More, ext2); err != nil {
panic("SetExtension on 2 failed: " + err.Error())
}
if err := SetExtension(messageWithInt32Extension1, pb.E_Ext_Number, Int32(23)); err != nil {
panic("SetExtension on Int32-1 failed: " + err.Error())
}
if err := SetExtension(messageWithInt32Extension1, pb.E_Ext_Number, Int32(24)); err != nil {
panic("SetExtension on Int32-2 failed: " + err.Error())
}
}
var EqualTests = []struct {
desc string
a, b Message
exp bool
}{
{"different types", &pb.GoEnum{}, &pb.GoTestField{}, false},
{"equal empty", &pb.GoEnum{}, &pb.GoEnum{}, true},
{"nil vs nil", nil, nil, true},
{"typed nil vs typed nil", (*pb.GoEnum)(nil), (*pb.GoEnum)(nil), true},
{"typed nil vs empty", (*pb.GoEnum)(nil), &pb.GoEnum{}, false},
{"different typed nil", (*pb.GoEnum)(nil), (*pb.GoTestField)(nil), false},
{"one set field, one unset field", &pb.GoTestField{Label: String("foo")}, &pb.GoTestField{}, false},
{"one set field zero, one unset field", &pb.GoTest{Param: Int32(0)}, &pb.GoTest{}, false},
{"different set fields", &pb.GoTestField{Label: String("foo")}, &pb.GoTestField{Label: String("bar")}, false},
{"equal set", &pb.GoTestField{Label: String("foo")}, &pb.GoTestField{Label: String("foo")}, true},
{"repeated, one set", &pb.GoTest{F_Int32Repeated: []int32{2, 3}}, &pb.GoTest{}, false},
{"repeated, different length", &pb.GoTest{F_Int32Repeated: []int32{2, 3}}, &pb.GoTest{F_Int32Repeated: []int32{2}}, false},
{"repeated, different value", &pb.GoTest{F_Int32Repeated: []int32{2}}, &pb.GoTest{F_Int32Repeated: []int32{3}}, false},
{"repeated, equal", &pb.GoTest{F_Int32Repeated: []int32{2, 4}}, &pb.GoTest{F_Int32Repeated: []int32{2, 4}}, true},
{"repeated, nil equal nil", &pb.GoTest{F_Int32Repeated: nil}, &pb.GoTest{F_Int32Repeated: nil}, true},
{"repeated, nil equal empty", &pb.GoTest{F_Int32Repeated: nil}, &pb.GoTest{F_Int32Repeated: []int32{}}, true},
{"repeated, empty equal nil", &pb.GoTest{F_Int32Repeated: []int32{}}, &pb.GoTest{F_Int32Repeated: nil}, true},
{
"nested, different",
&pb.GoTest{RequiredField: &pb.GoTestField{Label: String("foo")}},
&pb.GoTest{RequiredField: &pb.GoTestField{Label: String("bar")}},
false,
},
{
"nested, equal",
&pb.GoTest{RequiredField: &pb.GoTestField{Label: String("wow")}},
&pb.GoTest{RequiredField: &pb.GoTestField{Label: String("wow")}},
true,
},
{"bytes", &pb.OtherMessage{Value: []byte("foo")}, &pb.OtherMessage{Value: []byte("foo")}, true},
{"bytes, empty", &pb.OtherMessage{Value: []byte{}}, &pb.OtherMessage{Value: []byte{}}, true},
{"bytes, empty vs nil", &pb.OtherMessage{Value: []byte{}}, &pb.OtherMessage{Value: nil}, false},
{
"repeated bytes",
&pb.MyMessage{RepBytes: [][]byte{[]byte("sham"), []byte("wow")}},
&pb.MyMessage{RepBytes: [][]byte{[]byte("sham"), []byte("wow")}},
true,
},
{"extension vs. no extension", messageWithoutExtension, messageWithExtension1a, false},
{"extension vs. same extension", messageWithExtension1a, messageWithExtension1b, true},
{"extension vs. different extension", messageWithExtension1a, messageWithExtension2, false},
{"int32 extension vs. itself", messageWithInt32Extension1, messageWithInt32Extension1, true},
{"int32 extension vs. a different int32", messageWithInt32Extension1, messageWithInt32Extension2, false},
{
"message with group",
&pb.MyMessage{
Count: Int32(1),
Somegroup: &pb.MyMessage_SomeGroup{
GroupField: Int32(5),
},
},
&pb.MyMessage{
Count: Int32(1),
Somegroup: &pb.MyMessage_SomeGroup{
GroupField: Int32(5),
},
},
true,
},
}
func TestEqual(t *testing.T) {
for _, tc := range EqualTests {
if res := Equal(tc.a, tc.b); res != tc.exp {
t.Errorf("%v: Equal(%v, %v) = %v, want %v", tc.desc, tc.a, tc.b, res, tc.exp)
}
}
}

View File

@@ -1,353 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2010 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
package proto
/*
* Types and routines for supporting protocol buffer extensions.
*/
import (
"errors"
"reflect"
"strconv"
"sync"
)
// ErrMissingExtension is the error returned by GetExtension if the named extension is not in the message.
var ErrMissingExtension = errors.New("proto: missing extension")
// ExtensionRange represents a range of message extensions for a protocol buffer.
// Used in code generated by the protocol compiler.
type ExtensionRange struct {
Start, End int32 // both inclusive
}
// extendableProto is an interface implemented by any protocol buffer that may be extended.
type extendableProto interface {
Message
ExtensionRangeArray() []ExtensionRange
ExtensionMap() map[int32]Extension
}
var extendableProtoType = reflect.TypeOf((*extendableProto)(nil)).Elem()
// ExtensionDesc represents an extension specification.
// Used in generated code from the protocol compiler.
type ExtensionDesc struct {
ExtendedType Message // nil pointer to the type that is being extended
ExtensionType interface{} // nil pointer to the extension type
Field int32 // field number
Name string // fully-qualified name of extension, for text formatting
Tag string // protobuf tag style
}
func (ed *ExtensionDesc) repeated() bool {
t := reflect.TypeOf(ed.ExtensionType)
return t.Kind() == reflect.Slice && t.Elem().Kind() != reflect.Uint8
}
// Extension represents an extension in a message.
type Extension struct {
// When an extension is stored in a message using SetExtension
// only desc and value are set. When the message is marshaled
// enc will be set to the encoded form of the message.
//
// When a message is unmarshaled and contains extensions, each
// extension will have only enc set. When such an extension is
// accessed using GetExtension (or GetExtensions) desc and value
// will be set.
desc *ExtensionDesc
value interface{}
enc []byte
}
// SetRawExtension is for testing only.
func SetRawExtension(base extendableProto, id int32, b []byte) {
base.ExtensionMap()[id] = Extension{enc: b}
}
// isExtensionField returns true iff the given field number is in an extension range.
func isExtensionField(pb extendableProto, field int32) bool {
for _, er := range pb.ExtensionRangeArray() {
if er.Start <= field && field <= er.End {
return true
}
}
return false
}
// checkExtensionTypes checks that the given extension is valid for pb.
func checkExtensionTypes(pb extendableProto, extension *ExtensionDesc) error {
// Check the extended type.
if a, b := reflect.TypeOf(pb), reflect.TypeOf(extension.ExtendedType); a != b {
return errors.New("proto: bad extended type; " + b.String() + " does not extend " + a.String())
}
// Check the range.
if !isExtensionField(pb, extension.Field) {
return errors.New("proto: bad extension number; not in declared ranges")
}
return nil
}
// extPropKey is sufficient to uniquely identify an extension.
type extPropKey struct {
base reflect.Type
field int32
}
var extProp = struct {
sync.RWMutex
m map[extPropKey]*Properties
}{
m: make(map[extPropKey]*Properties),
}
func extensionProperties(ed *ExtensionDesc) *Properties {
key := extPropKey{base: reflect.TypeOf(ed.ExtendedType), field: ed.Field}
extProp.RLock()
if prop, ok := extProp.m[key]; ok {
extProp.RUnlock()
return prop
}
extProp.RUnlock()
extProp.Lock()
defer extProp.Unlock()
// Check again.
if prop, ok := extProp.m[key]; ok {
return prop
}
prop := new(Properties)
prop.Init(reflect.TypeOf(ed.ExtensionType), "unknown_name", ed.Tag, nil)
extProp.m[key] = prop
return prop
}
// encodeExtensionMap encodes any unmarshaled (unencoded) extensions in m.
func encodeExtensionMap(m map[int32]Extension) error {
for k, e := range m {
if e.value == nil || e.desc == nil {
// Extension is only in its encoded form.
continue
}
// We don't skip extensions that have an encoded form set,
// because the extension value may have been mutated after
// the last time this function was called.
et := reflect.TypeOf(e.desc.ExtensionType)
props := extensionProperties(e.desc)
p := NewBuffer(nil)
// If e.value has type T, the encoder expects a *struct{ X T }.
// Pass a *T with a zero field and hope it all works out.
x := reflect.New(et)
x.Elem().Set(reflect.ValueOf(e.value))
if err := props.enc(p, props, toStructPointer(x)); err != nil {
return err
}
e.enc = p.buf
m[k] = e
}
return nil
}
func sizeExtensionMap(m map[int32]Extension) (n int) {
for _, e := range m {
if e.value == nil || e.desc == nil {
// Extension is only in its encoded form.
n += len(e.enc)
continue
}
// We don't skip extensions that have an encoded form set,
// because the extension value may have been mutated after
// the last time this function was called.
et := reflect.TypeOf(e.desc.ExtensionType)
props := extensionProperties(e.desc)
// If e.value has type T, the encoder expects a *struct{ X T }.
// Pass a *T with a zero field and hope it all works out.
x := reflect.New(et)
x.Elem().Set(reflect.ValueOf(e.value))
n += props.size(props, toStructPointer(x))
}
return
}
// HasExtension returns whether the given extension is present in pb.
func HasExtension(pb extendableProto, extension *ExtensionDesc) bool {
// TODO: Check types, field numbers, etc.?
_, ok := pb.ExtensionMap()[extension.Field]
return ok
}
// ClearExtension removes the given extension from pb.
func ClearExtension(pb extendableProto, extension *ExtensionDesc) {
// TODO: Check types, field numbers, etc.?
delete(pb.ExtensionMap(), extension.Field)
}
// GetExtension parses and returns the given extension of pb.
// If the extension is not present it returns ErrMissingExtension.
func GetExtension(pb extendableProto, extension *ExtensionDesc) (interface{}, error) {
if err := checkExtensionTypes(pb, extension); err != nil {
return nil, err
}
emap := pb.ExtensionMap()
e, ok := emap[extension.Field]
if !ok {
return nil, ErrMissingExtension
}
if e.value != nil {
// Already decoded. Check the descriptor, though.
if e.desc != extension {
// This shouldn't happen. If it does, it means that
// GetExtension was called twice with two different
// descriptors with the same field number.
return nil, errors.New("proto: descriptor conflict")
}
return e.value, nil
}
v, err := decodeExtension(e.enc, extension)
if err != nil {
return nil, err
}
// Remember the decoded version and drop the encoded version.
// That way it is safe to mutate what we return.
e.value = v
e.desc = extension
e.enc = nil
emap[extension.Field] = e
return e.value, nil
}
// decodeExtension decodes an extension encoded in b.
func decodeExtension(b []byte, extension *ExtensionDesc) (interface{}, error) {
o := NewBuffer(b)
t := reflect.TypeOf(extension.ExtensionType)
rep := extension.repeated()
props := extensionProperties(extension)
// t is a pointer to a struct, pointer to basic type or a slice.
// Allocate a "field" to store the pointer/slice itself; the
// pointer/slice will be stored here. We pass
// the address of this field to props.dec.
// This passes a zero field and a *t and lets props.dec
// interpret it as a *struct{ x t }.
value := reflect.New(t).Elem()
for {
// Discard wire type and field number varint. It isn't needed.
if _, err := o.DecodeVarint(); err != nil {
return nil, err
}
if err := props.dec(o, props, toStructPointer(value.Addr())); err != nil {
return nil, err
}
if !rep || o.index >= len(o.buf) {
break
}
}
return value.Interface(), nil
}
// GetExtensions returns a slice of the extensions present in pb that are also listed in es.
// The returned slice has the same length as es; missing extensions will appear as nil elements.
func GetExtensions(pb Message, es []*ExtensionDesc) (extensions []interface{}, err error) {
epb, ok := pb.(extendableProto)
if !ok {
err = errors.New("proto: not an extendable proto")
return
}
extensions = make([]interface{}, len(es))
for i, e := range es {
extensions[i], err = GetExtension(epb, e)
if err == ErrMissingExtension {
err = nil
}
if err != nil {
return
}
}
return
}
// SetExtension sets the specified extension of pb to the specified value.
func SetExtension(pb extendableProto, extension *ExtensionDesc, value interface{}) error {
if err := checkExtensionTypes(pb, extension); err != nil {
return err
}
typ := reflect.TypeOf(extension.ExtensionType)
if typ != reflect.TypeOf(value) {
return errors.New("proto: bad extension value type")
}
pb.ExtensionMap()[extension.Field] = Extension{desc: extension, value: value}
return nil
}
// A global registry of extensions.
// The generated code will register the generated descriptors by calling RegisterExtension.
var extensionMaps = make(map[reflect.Type]map[int32]*ExtensionDesc)
// RegisterExtension is called from the generated code.
func RegisterExtension(desc *ExtensionDesc) {
st := reflect.TypeOf(desc.ExtendedType).Elem()
m := extensionMaps[st]
if m == nil {
m = make(map[int32]*ExtensionDesc)
extensionMaps[st] = m
}
if _, ok := m[desc.Field]; ok {
panic("proto: duplicate extension registered: " + st.String() + " " + strconv.Itoa(int(desc.Field)))
}
m[desc.Field] = desc
}
// RegisteredExtensions returns a map of the registered extensions of a
// protocol buffer struct, indexed by the extension number.
// The argument pb should be a nil pointer to the struct type.
func RegisteredExtensions(pb Message) map[int32]*ExtensionDesc {
return extensionMaps[reflect.TypeOf(pb).Elem()]
}

View File

@@ -1,94 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2014 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
package proto_test
import (
"testing"
pb "./testdata"
"github.com/coreos/etcd/Godeps/_workspace/src/code.google.com/p/goprotobuf/proto"
)
func TestGetExtensionsWithMissingExtensions(t *testing.T) {
msg := &pb.MyMessage{}
ext1 := &pb.Ext{}
if err := proto.SetExtension(msg, pb.E_Ext_More, ext1); err != nil {
t.Fatalf("Could not set ext1: %s", ext1)
}
exts, err := proto.GetExtensions(msg, []*proto.ExtensionDesc{
pb.E_Ext_More,
pb.E_Ext_Text,
})
if err != nil {
t.Fatalf("GetExtensions() failed: %s", err)
}
if exts[0] != ext1 {
t.Errorf("ext1 not in returned extensions: %T %v", exts[0], exts[0])
}
if exts[1] != nil {
t.Errorf("ext2 in returned extensions: %T %v", exts[1], exts[1])
}
}
func TestGetExtensionStability(t *testing.T) {
check := func(m *pb.MyMessage) bool {
ext1, err := proto.GetExtension(m, pb.E_Ext_More)
if err != nil {
t.Fatalf("GetExtension() failed: %s", err)
}
ext2, err := proto.GetExtension(m, pb.E_Ext_More)
if err != nil {
t.Fatalf("GetExtension() failed: %s", err)
}
return ext1 == ext2
}
msg := &pb.MyMessage{Count: proto.Int32(4)}
ext0 := &pb.Ext{}
if err := proto.SetExtension(msg, pb.E_Ext_More, ext0); err != nil {
t.Fatalf("Could not set ext1: %s", ext0)
}
if !check(msg) {
t.Errorf("GetExtension() not stable before marshaling")
}
bb, err := proto.Marshal(msg)
if err != nil {
t.Fatalf("Marshal() failed: %s", err)
}
msg1 := &pb.MyMessage{}
err = proto.Unmarshal(bb, msg1)
if err != nil {
t.Fatalf("Unmarshal() failed: %s", err)
}
if !check(msg1) {
t.Errorf("GetExtension() not stable after unmarshaling")
}
}

View File

@@ -1,740 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2010 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
/*
Package proto converts data structures to and from the wire format of
protocol buffers. It works in concert with the Go source code generated
for .proto files by the protocol compiler.
A summary of the properties of the protocol buffer interface
for a protocol buffer variable v:
- Names are turned from camel_case to CamelCase for export.
- There are no methods on v to set fields; just treat
them as structure fields.
- There are getters that return a field's value if set,
and return the field's default value if unset.
The getters work even if the receiver is a nil message.
- The zero value for a struct is its correct initialization state.
All desired fields must be set before marshaling.
- A Reset() method will restore a protobuf struct to its zero state.
- Non-repeated fields are pointers to the values; nil means unset.
That is, optional or required field int32 f becomes F *int32.
- Repeated fields are slices.
- Helper functions are available to aid the setting of fields.
Helpers for getting values are superseded by the
GetFoo methods and their use is deprecated.
msg.Foo = proto.String("hello") // set field
- Constants are defined to hold the default values of all fields that
have them. They have the form Default_StructName_FieldName.
Because the getter methods handle defaulted values,
direct use of these constants should be rare.
- Enums are given type names and maps from names to values.
Enum values are prefixed with the enum's type name. Enum types have
a String method, and a Enum method to assist in message construction.
- Nested groups and enums have type names prefixed with the name of
the surrounding message type.
- Extensions are given descriptor names that start with E_,
followed by an underscore-delimited list of the nested messages
that contain it (if any) followed by the CamelCased name of the
extension field itself. HasExtension, ClearExtension, GetExtension
and SetExtension are functions for manipulating extensions.
- Marshal and Unmarshal are functions to encode and decode the wire format.
The simplest way to describe this is to see an example.
Given file test.proto, containing
package example;
enum FOO { X = 17; };
message Test {
required string label = 1;
optional int32 type = 2 [default=77];
repeated int64 reps = 3;
optional group OptionalGroup = 4 {
required string RequiredField = 5;
}
}
The resulting file, test.pb.go, is:
package example
import "code.google.com/p/goprotobuf/proto"
type FOO int32
const (
FOO_X FOO = 17
)
var FOO_name = map[int32]string{
17: "X",
}
var FOO_value = map[string]int32{
"X": 17,
}
func (x FOO) Enum() *FOO {
p := new(FOO)
*p = x
return p
}
func (x FOO) String() string {
return proto.EnumName(FOO_name, int32(x))
}
type Test struct {
Label *string `protobuf:"bytes,1,req,name=label" json:"label,omitempty"`
Type *int32 `protobuf:"varint,2,opt,name=type,def=77" json:"type,omitempty"`
Reps []int64 `protobuf:"varint,3,rep,name=reps" json:"reps,omitempty"`
Optionalgroup *Test_OptionalGroup `protobuf:"group,4,opt,name=OptionalGroup" json:"optionalgroup,omitempty"`
XXX_unrecognized []byte `json:"-"`
}
func (this *Test) Reset() { *this = Test{} }
func (this *Test) String() string { return proto.CompactTextString(this) }
const Default_Test_Type int32 = 77
func (this *Test) GetLabel() string {
if this != nil && this.Label != nil {
return *this.Label
}
return ""
}
func (this *Test) GetType() int32 {
if this != nil && this.Type != nil {
return *this.Type
}
return Default_Test_Type
}
func (this *Test) GetOptionalgroup() *Test_OptionalGroup {
if this != nil {
return this.Optionalgroup
}
return nil
}
type Test_OptionalGroup struct {
RequiredField *string `protobuf:"bytes,5,req" json:"RequiredField,omitempty"`
XXX_unrecognized []byte `json:"-"`
}
func (this *Test_OptionalGroup) Reset() { *this = Test_OptionalGroup{} }
func (this *Test_OptionalGroup) String() string { return proto.CompactTextString(this) }
func (this *Test_OptionalGroup) GetRequiredField() string {
if this != nil && this.RequiredField != nil {
return *this.RequiredField
}
return ""
}
func init() {
proto.RegisterEnum("example.FOO", FOO_name, FOO_value)
}
To create and play with a Test object:
package main
import (
"log"
"code.google.com/p/goprotobuf/proto"
"./example.pb"
)
func main() {
test := &example.Test{
Label: proto.String("hello"),
Type: proto.Int32(17),
Optionalgroup: &example.Test_OptionalGroup{
RequiredField: proto.String("good bye"),
},
}
data, err := proto.Marshal(test)
if err != nil {
log.Fatal("marshaling error: ", err)
}
newTest := new(example.Test)
err = proto.Unmarshal(data, newTest)
if err != nil {
log.Fatal("unmarshaling error: ", err)
}
// Now test and newTest contain the same data.
if test.GetLabel() != newTest.GetLabel() {
log.Fatalf("data mismatch %q != %q", test.GetLabel(), newTest.GetLabel())
}
// etc.
}
*/
package proto
import (
"encoding/json"
"fmt"
"log"
"reflect"
"strconv"
"sync"
)
// Message is implemented by generated protocol buffer messages.
type Message interface {
Reset()
String() string
ProtoMessage()
}
// Stats records allocation details about the protocol buffer encoders
// and decoders. Useful for tuning the library itself.
type Stats struct {
Emalloc uint64 // mallocs in encode
Dmalloc uint64 // mallocs in decode
Encode uint64 // number of encodes
Decode uint64 // number of decodes
Chit uint64 // number of cache hits
Cmiss uint64 // number of cache misses
Size uint64 // number of sizes
}
// Set to true to enable stats collection.
const collectStats = false
var stats Stats
// GetStats returns a copy of the global Stats structure.
func GetStats() Stats { return stats }
// A Buffer is a buffer manager for marshaling and unmarshaling
// protocol buffers. It may be reused between invocations to
// reduce memory usage. It is not necessary to use a Buffer;
// the global functions Marshal and Unmarshal create a
// temporary Buffer and are fine for most applications.
type Buffer struct {
buf []byte // encode/decode byte stream
index int // write point
// pools of basic types to amortize allocation.
bools []bool
uint32s []uint32
uint64s []uint64
// extra pools, only used with pointer_reflect.go
int32s []int32
int64s []int64
float32s []float32
float64s []float64
}
// NewBuffer allocates a new Buffer and initializes its internal data to
// the contents of the argument slice.
func NewBuffer(e []byte) *Buffer {
return &Buffer{buf: e}
}
// Reset resets the Buffer, ready for marshaling a new protocol buffer.
func (p *Buffer) Reset() {
p.buf = p.buf[0:0] // for reading/writing
p.index = 0 // for reading
}
// SetBuf replaces the internal buffer with the slice,
// ready for unmarshaling the contents of the slice.
func (p *Buffer) SetBuf(s []byte) {
p.buf = s
p.index = 0
}
// Bytes returns the contents of the Buffer.
func (p *Buffer) Bytes() []byte { return p.buf }
/*
* Helper routines for simplifying the creation of optional fields of basic type.
*/
// Bool is a helper routine that allocates a new bool value
// to store v and returns a pointer to it.
func Bool(v bool) *bool {
return &v
}
// Int32 is a helper routine that allocates a new int32 value
// to store v and returns a pointer to it.
func Int32(v int32) *int32 {
return &v
}
// Int is a helper routine that allocates a new int32 value
// to store v and returns a pointer to it, but unlike Int32
// its argument value is an int.
func Int(v int) *int32 {
p := new(int32)
*p = int32(v)
return p
}
// Int64 is a helper routine that allocates a new int64 value
// to store v and returns a pointer to it.
func Int64(v int64) *int64 {
return &v
}
// Float32 is a helper routine that allocates a new float32 value
// to store v and returns a pointer to it.
func Float32(v float32) *float32 {
return &v
}
// Float64 is a helper routine that allocates a new float64 value
// to store v and returns a pointer to it.
func Float64(v float64) *float64 {
return &v
}
// Uint32 is a helper routine that allocates a new uint32 value
// to store v and returns a pointer to it.
func Uint32(v uint32) *uint32 {
p := new(uint32)
*p = v
return p
}
// Uint64 is a helper routine that allocates a new uint64 value
// to store v and returns a pointer to it.
func Uint64(v uint64) *uint64 {
return &v
}
// String is a helper routine that allocates a new string value
// to store v and returns a pointer to it.
func String(v string) *string {
return &v
}
// EnumName is a helper function to simplify printing protocol buffer enums
// by name. Given an enum map and a value, it returns a useful string.
func EnumName(m map[int32]string, v int32) string {
s, ok := m[v]
if ok {
return s
}
return strconv.Itoa(int(v))
}
// UnmarshalJSONEnum is a helper function to simplify recovering enum int values
// from their JSON-encoded representation. Given a map from the enum's symbolic
// names to its int values, and a byte buffer containing the JSON-encoded
// value, it returns an int32 that can be cast to the enum type by the caller.
//
// The function can deal with both JSON representations, numeric and symbolic.
func UnmarshalJSONEnum(m map[string]int32, data []byte, enumName string) (int32, error) {
if data[0] == '"' {
// New style: enums are strings.
var repr string
if err := json.Unmarshal(data, &repr); err != nil {
return -1, err
}
val, ok := m[repr]
if !ok {
return 0, fmt.Errorf("unrecognized enum %s value %q", enumName, repr)
}
return val, nil
}
// Old style: enums are ints.
var val int32
if err := json.Unmarshal(data, &val); err != nil {
return 0, fmt.Errorf("cannot unmarshal %#q into enum %s", data, enumName)
}
return val, nil
}
// DebugPrint dumps the encoded data in b in a debugging format with a header
// including the string s. Used in testing but made available for general debugging.
func (o *Buffer) DebugPrint(s string, b []byte) {
var u uint64
obuf := o.buf
index := o.index
o.buf = b
o.index = 0
depth := 0
fmt.Printf("\n--- %s ---\n", s)
out:
for {
for i := 0; i < depth; i++ {
fmt.Print(" ")
}
index := o.index
if index == len(o.buf) {
break
}
op, err := o.DecodeVarint()
if err != nil {
fmt.Printf("%3d: fetching op err %v\n", index, err)
break out
}
tag := op >> 3
wire := op & 7
switch wire {
default:
fmt.Printf("%3d: t=%3d unknown wire=%d\n",
index, tag, wire)
break out
case WireBytes:
var r []byte
r, err = o.DecodeRawBytes(false)
if err != nil {
break out
}
fmt.Printf("%3d: t=%3d bytes [%d]", index, tag, len(r))
if len(r) <= 6 {
for i := 0; i < len(r); i++ {
fmt.Printf(" %.2x", r[i])
}
} else {
for i := 0; i < 3; i++ {
fmt.Printf(" %.2x", r[i])
}
fmt.Printf(" ..")
for i := len(r) - 3; i < len(r); i++ {
fmt.Printf(" %.2x", r[i])
}
}
fmt.Printf("\n")
case WireFixed32:
u, err = o.DecodeFixed32()
if err != nil {
fmt.Printf("%3d: t=%3d fix32 err %v\n", index, tag, err)
break out
}
fmt.Printf("%3d: t=%3d fix32 %d\n", index, tag, u)
case WireFixed64:
u, err = o.DecodeFixed64()
if err != nil {
fmt.Printf("%3d: t=%3d fix64 err %v\n", index, tag, err)
break out
}
fmt.Printf("%3d: t=%3d fix64 %d\n", index, tag, u)
break
case WireVarint:
u, err = o.DecodeVarint()
if err != nil {
fmt.Printf("%3d: t=%3d varint err %v\n", index, tag, err)
break out
}
fmt.Printf("%3d: t=%3d varint %d\n", index, tag, u)
case WireStartGroup:
if err != nil {
fmt.Printf("%3d: t=%3d start err %v\n", index, tag, err)
break out
}
fmt.Printf("%3d: t=%3d start\n", index, tag)
depth++
case WireEndGroup:
depth--
if err != nil {
fmt.Printf("%3d: t=%3d end err %v\n", index, tag, err)
break out
}
fmt.Printf("%3d: t=%3d end\n", index, tag)
}
}
if depth != 0 {
fmt.Printf("%3d: start-end not balanced %d\n", o.index, depth)
}
fmt.Printf("\n")
o.buf = obuf
o.index = index
}
// SetDefaults sets unset protocol buffer fields to their default values.
// It only modifies fields that are both unset and have defined defaults.
// It recursively sets default values in any non-nil sub-messages.
func SetDefaults(pb Message) {
setDefaults(reflect.ValueOf(pb), true, false)
}
// v is a pointer to a struct.
func setDefaults(v reflect.Value, recur, zeros bool) {
v = v.Elem()
defaultMu.RLock()
dm, ok := defaults[v.Type()]
defaultMu.RUnlock()
if !ok {
dm = buildDefaultMessage(v.Type())
defaultMu.Lock()
defaults[v.Type()] = dm
defaultMu.Unlock()
}
for _, sf := range dm.scalars {
f := v.Field(sf.index)
if !f.IsNil() {
// field already set
continue
}
dv := sf.value
if dv == nil && !zeros {
// no explicit default, and don't want to set zeros
continue
}
fptr := f.Addr().Interface() // **T
// TODO: Consider batching the allocations we do here.
switch sf.kind {
case reflect.Bool:
b := new(bool)
if dv != nil {
*b = dv.(bool)
}
*(fptr.(**bool)) = b
case reflect.Float32:
f := new(float32)
if dv != nil {
*f = dv.(float32)
}
*(fptr.(**float32)) = f
case reflect.Float64:
f := new(float64)
if dv != nil {
*f = dv.(float64)
}
*(fptr.(**float64)) = f
case reflect.Int32:
// might be an enum
if ft := f.Type(); ft != int32PtrType {
// enum
f.Set(reflect.New(ft.Elem()))
if dv != nil {
f.Elem().SetInt(int64(dv.(int32)))
}
} else {
// int32 field
i := new(int32)
if dv != nil {
*i = dv.(int32)
}
*(fptr.(**int32)) = i
}
case reflect.Int64:
i := new(int64)
if dv != nil {
*i = dv.(int64)
}
*(fptr.(**int64)) = i
case reflect.String:
s := new(string)
if dv != nil {
*s = dv.(string)
}
*(fptr.(**string)) = s
case reflect.Uint8:
// exceptional case: []byte
var b []byte
if dv != nil {
db := dv.([]byte)
b = make([]byte, len(db))
copy(b, db)
} else {
b = []byte{}
}
*(fptr.(*[]byte)) = b
case reflect.Uint32:
u := new(uint32)
if dv != nil {
*u = dv.(uint32)
}
*(fptr.(**uint32)) = u
case reflect.Uint64:
u := new(uint64)
if dv != nil {
*u = dv.(uint64)
}
*(fptr.(**uint64)) = u
default:
log.Printf("proto: can't set default for field %v (sf.kind=%v)", f, sf.kind)
}
}
for _, ni := range dm.nested {
f := v.Field(ni)
if f.IsNil() {
continue
}
// f is *T or []*T
if f.Kind() == reflect.Ptr {
setDefaults(f, recur, zeros)
} else {
for i := 0; i < f.Len(); i++ {
e := f.Index(i)
if e.IsNil() {
continue
}
setDefaults(e, recur, zeros)
}
}
}
}
var (
// defaults maps a protocol buffer struct type to a slice of the fields,
// with its scalar fields set to their proto-declared non-zero default values.
defaultMu sync.RWMutex
defaults = make(map[reflect.Type]defaultMessage)
int32PtrType = reflect.TypeOf((*int32)(nil))
)
// defaultMessage represents information about the default values of a message.
type defaultMessage struct {
scalars []scalarField
nested []int // struct field index of nested messages
}
type scalarField struct {
index int // struct field index
kind reflect.Kind // element type (the T in *T or []T)
value interface{} // the proto-declared default value, or nil
}
func ptrToStruct(t reflect.Type) bool {
return t.Kind() == reflect.Ptr && t.Elem().Kind() == reflect.Struct
}
// t is a struct type.
func buildDefaultMessage(t reflect.Type) (dm defaultMessage) {
sprop := GetProperties(t)
for _, prop := range sprop.Prop {
fi, ok := sprop.decoderTags.get(prop.Tag)
if !ok {
// XXX_unrecognized
continue
}
ft := t.Field(fi).Type
// nested messages
if ptrToStruct(ft) || (ft.Kind() == reflect.Slice && ptrToStruct(ft.Elem())) {
dm.nested = append(dm.nested, fi)
continue
}
sf := scalarField{
index: fi,
kind: ft.Elem().Kind(),
}
// scalar fields without defaults
if !prop.HasDefault {
dm.scalars = append(dm.scalars, sf)
continue
}
// a scalar field: either *T or []byte
switch ft.Elem().Kind() {
case reflect.Bool:
x, err := strconv.ParseBool(prop.Default)
if err != nil {
log.Printf("proto: bad default bool %q: %v", prop.Default, err)
continue
}
sf.value = x
case reflect.Float32:
x, err := strconv.ParseFloat(prop.Default, 32)
if err != nil {
log.Printf("proto: bad default float32 %q: %v", prop.Default, err)
continue
}
sf.value = float32(x)
case reflect.Float64:
x, err := strconv.ParseFloat(prop.Default, 64)
if err != nil {
log.Printf("proto: bad default float64 %q: %v", prop.Default, err)
continue
}
sf.value = x
case reflect.Int32:
x, err := strconv.ParseInt(prop.Default, 10, 32)
if err != nil {
log.Printf("proto: bad default int32 %q: %v", prop.Default, err)
continue
}
sf.value = int32(x)
case reflect.Int64:
x, err := strconv.ParseInt(prop.Default, 10, 64)
if err != nil {
log.Printf("proto: bad default int64 %q: %v", prop.Default, err)
continue
}
sf.value = x
case reflect.String:
sf.value = prop.Default
case reflect.Uint8:
// []byte (not *uint8)
sf.value = []byte(prop.Default)
case reflect.Uint32:
x, err := strconv.ParseUint(prop.Default, 10, 32)
if err != nil {
log.Printf("proto: bad default uint32 %q: %v", prop.Default, err)
continue
}
sf.value = uint32(x)
case reflect.Uint64:
x, err := strconv.ParseUint(prop.Default, 10, 64)
if err != nil {
log.Printf("proto: bad default uint64 %q: %v", prop.Default, err)
continue
}
sf.value = x
default:
log.Printf("proto: unhandled def kind %v", ft.Elem().Kind())
continue
}
dm.scalars = append(dm.scalars, sf)
}
return dm
}

View File

@@ -1,287 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2010 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
package proto
/*
* Support for message sets.
*/
import (
"bytes"
"encoding/json"
"errors"
"fmt"
"reflect"
"sort"
)
// ErrNoMessageTypeId occurs when a protocol buffer does not have a message type ID.
// A message type ID is required for storing a protocol buffer in a message set.
var ErrNoMessageTypeId = errors.New("proto does not have a message type ID")
// The first two types (_MessageSet_Item and MessageSet)
// model what the protocol compiler produces for the following protocol message:
// message MessageSet {
// repeated group Item = 1 {
// required int32 type_id = 2;
// required string message = 3;
// };
// }
// That is the MessageSet wire format. We can't use a proto to generate these
// because that would introduce a circular dependency between it and this package.
//
// When a proto1 proto has a field that looks like:
// optional message<MessageSet> info = 3;
// the protocol compiler produces a field in the generated struct that looks like:
// Info *_proto_.MessageSet `protobuf:"bytes,3,opt,name=info"`
// The package is automatically inserted so there is no need for that proto file to
// import this package.
type _MessageSet_Item struct {
TypeId *int32 `protobuf:"varint,2,req,name=type_id"`
Message []byte `protobuf:"bytes,3,req,name=message"`
}
type MessageSet struct {
Item []*_MessageSet_Item `protobuf:"group,1,rep"`
XXX_unrecognized []byte
// TODO: caching?
}
// Make sure MessageSet is a Message.
var _ Message = (*MessageSet)(nil)
// messageTypeIder is an interface satisfied by a protocol buffer type
// that may be stored in a MessageSet.
type messageTypeIder interface {
MessageTypeId() int32
}
func (ms *MessageSet) find(pb Message) *_MessageSet_Item {
mti, ok := pb.(messageTypeIder)
if !ok {
return nil
}
id := mti.MessageTypeId()
for _, item := range ms.Item {
if *item.TypeId == id {
return item
}
}
return nil
}
func (ms *MessageSet) Has(pb Message) bool {
if ms.find(pb) != nil {
return true
}
return false
}
func (ms *MessageSet) Unmarshal(pb Message) error {
if item := ms.find(pb); item != nil {
return Unmarshal(item.Message, pb)
}
if _, ok := pb.(messageTypeIder); !ok {
return ErrNoMessageTypeId
}
return nil // TODO: return error instead?
}
func (ms *MessageSet) Marshal(pb Message) error {
msg, err := Marshal(pb)
if err != nil {
return err
}
if item := ms.find(pb); item != nil {
// reuse existing item
item.Message = msg
return nil
}
mti, ok := pb.(messageTypeIder)
if !ok {
return ErrNoMessageTypeId
}
mtid := mti.MessageTypeId()
ms.Item = append(ms.Item, &_MessageSet_Item{
TypeId: &mtid,
Message: msg,
})
return nil
}
func (ms *MessageSet) Reset() { *ms = MessageSet{} }
func (ms *MessageSet) String() string { return CompactTextString(ms) }
func (*MessageSet) ProtoMessage() {}
// Support for the message_set_wire_format message option.
func skipVarint(buf []byte) []byte {
i := 0
for ; buf[i]&0x80 != 0; i++ {
}
return buf[i+1:]
}
// MarshalMessageSet encodes the extension map represented by m in the message set wire format.
// It is called by generated Marshal methods on protocol buffer messages with the message_set_wire_format option.
func MarshalMessageSet(m map[int32]Extension) ([]byte, error) {
if err := encodeExtensionMap(m); err != nil {
return nil, err
}
// Sort extension IDs to provide a deterministic encoding.
// See also enc_map in encode.go.
ids := make([]int, 0, len(m))
for id := range m {
ids = append(ids, int(id))
}
sort.Ints(ids)
ms := &MessageSet{Item: make([]*_MessageSet_Item, 0, len(m))}
for _, id := range ids {
e := m[int32(id)]
// Remove the wire type and field number varint, as well as the length varint.
msg := skipVarint(skipVarint(e.enc))
ms.Item = append(ms.Item, &_MessageSet_Item{
TypeId: Int32(int32(id)),
Message: msg,
})
}
return Marshal(ms)
}
// UnmarshalMessageSet decodes the extension map encoded in buf in the message set wire format.
// It is called by generated Unmarshal methods on protocol buffer messages with the message_set_wire_format option.
func UnmarshalMessageSet(buf []byte, m map[int32]Extension) error {
ms := new(MessageSet)
if err := Unmarshal(buf, ms); err != nil {
return err
}
for _, item := range ms.Item {
id := *item.TypeId
msg := item.Message
// Restore wire type and field number varint, plus length varint.
// Be careful to preserve duplicate items.
b := EncodeVarint(uint64(id)<<3 | WireBytes)
if ext, ok := m[id]; ok {
// Existing data; rip off the tag and length varint
// so we join the new data correctly.
// We can assume that ext.enc is set because we are unmarshaling.
o := ext.enc[len(b):] // skip wire type and field number
_, n := DecodeVarint(o) // calculate length of length varint
o = o[n:] // skip length varint
msg = append(o, msg...) // join old data and new data
}
b = append(b, EncodeVarint(uint64(len(msg)))...)
b = append(b, msg...)
m[id] = Extension{enc: b}
}
return nil
}
// MarshalMessageSetJSON encodes the extension map represented by m in JSON format.
// It is called by generated MarshalJSON methods on protocol buffer messages with the message_set_wire_format option.
func MarshalMessageSetJSON(m map[int32]Extension) ([]byte, error) {
var b bytes.Buffer
b.WriteByte('{')
// Process the map in key order for deterministic output.
ids := make([]int32, 0, len(m))
for id := range m {
ids = append(ids, id)
}
sort.Sort(int32Slice(ids)) // int32Slice defined in text.go
for i, id := range ids {
ext := m[id]
if i > 0 {
b.WriteByte(',')
}
msd, ok := messageSetMap[id]
if !ok {
// Unknown type; we can't render it, so skip it.
continue
}
fmt.Fprintf(&b, `"[%s]":`, msd.name)
x := ext.value
if x == nil {
x = reflect.New(msd.t.Elem()).Interface()
if err := Unmarshal(ext.enc, x.(Message)); err != nil {
return nil, err
}
}
d, err := json.Marshal(x)
if err != nil {
return nil, err
}
b.Write(d)
}
b.WriteByte('}')
return b.Bytes(), nil
}
// UnmarshalMessageSetJSON decodes the extension map encoded in buf in JSON format.
// It is called by generated UnmarshalJSON methods on protocol buffer messages with the message_set_wire_format option.
func UnmarshalMessageSetJSON(buf []byte, m map[int32]Extension) error {
// Common-case fast path.
if len(buf) == 0 || bytes.Equal(buf, []byte("{}")) {
return nil
}
// This is fairly tricky, and it's not clear that it is needed.
return errors.New("TODO: UnmarshalMessageSetJSON not yet implemented")
}
// A global registry of types that can be used in a MessageSet.
var messageSetMap = make(map[int32]messageSetDesc)
type messageSetDesc struct {
t reflect.Type // pointer to struct
name string
}
// RegisterMessageSetType is called from the generated code.
func RegisterMessageSetType(m Message, fieldNum int32, name string) {
messageSetMap[fieldNum] = messageSetDesc{
t: reflect.TypeOf(m),
name: name,
}
}

View File

@@ -1,66 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2014 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
package proto
import (
"bytes"
"testing"
)
func TestUnmarshalMessageSetWithDuplicate(t *testing.T) {
// Check that a repeated message set entry will be concatenated.
in := &MessageSet{
Item: []*_MessageSet_Item{
{TypeId: Int32(12345), Message: []byte("hoo")},
{TypeId: Int32(12345), Message: []byte("hah")},
},
}
b, err := Marshal(in)
if err != nil {
t.Fatalf("Marshal: %v", err)
}
t.Logf("Marshaled bytes: %q", b)
m := make(map[int32]Extension)
if err := UnmarshalMessageSet(b, m); err != nil {
t.Fatalf("UnmarshalMessageSet: %v", err)
}
ext, ok := m[12345]
if !ok {
t.Fatalf("Didn't retrieve extension 12345; map is %v", m)
}
// Skip wire type/field number and length varints.
got := skipVarint(skipVarint(ext.enc))
if want := []byte("hoohah"); !bytes.Equal(got, want) {
t.Errorf("Combined extension is %q, want %q", got, want)
}
}

View File

@@ -1,384 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2012 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
// +build appengine,!appenginevm
// This file contains an implementation of proto field accesses using package reflect.
// It is slower than the code in pointer_unsafe.go but it avoids package unsafe and can
// be used on App Engine.
package proto
import (
"math"
"reflect"
)
// A structPointer is a pointer to a struct.
type structPointer struct {
v reflect.Value
}
// toStructPointer returns a structPointer equivalent to the given reflect value.
// The reflect value must itself be a pointer to a struct.
func toStructPointer(v reflect.Value) structPointer {
return structPointer{v}
}
// IsNil reports whether p is nil.
func structPointer_IsNil(p structPointer) bool {
return p.v.IsNil()
}
// Interface returns the struct pointer as an interface value.
func structPointer_Interface(p structPointer, _ reflect.Type) interface{} {
return p.v.Interface()
}
// A field identifies a field in a struct, accessible from a structPointer.
// In this implementation, a field is identified by the sequence of field indices
// passed to reflect's FieldByIndex.
type field []int
// toField returns a field equivalent to the given reflect field.
func toField(f *reflect.StructField) field {
return f.Index
}
// invalidField is an invalid field identifier.
var invalidField = field(nil)
// IsValid reports whether the field identifier is valid.
func (f field) IsValid() bool { return f != nil }
// field returns the given field in the struct as a reflect value.
func structPointer_field(p structPointer, f field) reflect.Value {
// Special case: an extension map entry with a value of type T
// passes a *T to the struct-handling code with a zero field,
// expecting that it will be treated as equivalent to *struct{ X T },
// which has the same memory layout. We have to handle that case
// specially, because reflect will panic if we call FieldByIndex on a
// non-struct.
if f == nil {
return p.v.Elem()
}
return p.v.Elem().FieldByIndex(f)
}
// ifield returns the given field in the struct as an interface value.
func structPointer_ifield(p structPointer, f field) interface{} {
return structPointer_field(p, f).Addr().Interface()
}
// Bytes returns the address of a []byte field in the struct.
func structPointer_Bytes(p structPointer, f field) *[]byte {
return structPointer_ifield(p, f).(*[]byte)
}
// BytesSlice returns the address of a [][]byte field in the struct.
func structPointer_BytesSlice(p structPointer, f field) *[][]byte {
return structPointer_ifield(p, f).(*[][]byte)
}
// Bool returns the address of a *bool field in the struct.
func structPointer_Bool(p structPointer, f field) **bool {
return structPointer_ifield(p, f).(**bool)
}
// BoolSlice returns the address of a []bool field in the struct.
func structPointer_BoolSlice(p structPointer, f field) *[]bool {
return structPointer_ifield(p, f).(*[]bool)
}
// String returns the address of a *string field in the struct.
func structPointer_String(p structPointer, f field) **string {
return structPointer_ifield(p, f).(**string)
}
// StringSlice returns the address of a []string field in the struct.
func structPointer_StringSlice(p structPointer, f field) *[]string {
return structPointer_ifield(p, f).(*[]string)
}
// ExtMap returns the address of an extension map field in the struct.
func structPointer_ExtMap(p structPointer, f field) *map[int32]Extension {
return structPointer_ifield(p, f).(*map[int32]Extension)
}
// SetStructPointer writes a *struct field in the struct.
func structPointer_SetStructPointer(p structPointer, f field, q structPointer) {
structPointer_field(p, f).Set(q.v)
}
// GetStructPointer reads a *struct field in the struct.
func structPointer_GetStructPointer(p structPointer, f field) structPointer {
return structPointer{structPointer_field(p, f)}
}
// StructPointerSlice the address of a []*struct field in the struct.
func structPointer_StructPointerSlice(p structPointer, f field) structPointerSlice {
return structPointerSlice{structPointer_field(p, f)}
}
// A structPointerSlice represents the address of a slice of pointers to structs
// (themselves messages or groups). That is, v.Type() is *[]*struct{...}.
type structPointerSlice struct {
v reflect.Value
}
func (p structPointerSlice) Len() int { return p.v.Len() }
func (p structPointerSlice) Index(i int) structPointer { return structPointer{p.v.Index(i)} }
func (p structPointerSlice) Append(q structPointer) {
p.v.Set(reflect.Append(p.v, q.v))
}
var (
int32Type = reflect.TypeOf(int32(0))
uint32Type = reflect.TypeOf(uint32(0))
float32Type = reflect.TypeOf(float32(0))
int64Type = reflect.TypeOf(int64(0))
uint64Type = reflect.TypeOf(uint64(0))
float64Type = reflect.TypeOf(float64(0))
)
// A word32 represents a field of type *int32, *uint32, *float32, or *enum.
// That is, v.Type() is *int32, *uint32, *float32, or *enum and v is assignable.
type word32 struct {
v reflect.Value
}
// IsNil reports whether p is nil.
func word32_IsNil(p word32) bool {
return p.v.IsNil()
}
// Set sets p to point at a newly allocated word with bits set to x.
func word32_Set(p word32, o *Buffer, x uint32) {
t := p.v.Type().Elem()
switch t {
case int32Type:
if len(o.int32s) == 0 {
o.int32s = make([]int32, uint32PoolSize)
}
o.int32s[0] = int32(x)
p.v.Set(reflect.ValueOf(&o.int32s[0]))
o.int32s = o.int32s[1:]
return
case uint32Type:
if len(o.uint32s) == 0 {
o.uint32s = make([]uint32, uint32PoolSize)
}
o.uint32s[0] = x
p.v.Set(reflect.ValueOf(&o.uint32s[0]))
o.uint32s = o.uint32s[1:]
return
case float32Type:
if len(o.float32s) == 0 {
o.float32s = make([]float32, uint32PoolSize)
}
o.float32s[0] = math.Float32frombits(x)
p.v.Set(reflect.ValueOf(&o.float32s[0]))
o.float32s = o.float32s[1:]
return
}
// must be enum
p.v.Set(reflect.New(t))
p.v.Elem().SetInt(int64(int32(x)))
}
// Get gets the bits pointed at by p, as a uint32.
func word32_Get(p word32) uint32 {
elem := p.v.Elem()
switch elem.Kind() {
case reflect.Int32:
return uint32(elem.Int())
case reflect.Uint32:
return uint32(elem.Uint())
case reflect.Float32:
return math.Float32bits(float32(elem.Float()))
}
panic("unreachable")
}
// Word32 returns a reference to a *int32, *uint32, *float32, or *enum field in the struct.
func structPointer_Word32(p structPointer, f field) word32 {
return word32{structPointer_field(p, f)}
}
// A word32Slice is a slice of 32-bit values.
// That is, v.Type() is []int32, []uint32, []float32, or []enum.
type word32Slice struct {
v reflect.Value
}
func (p word32Slice) Append(x uint32) {
n, m := p.v.Len(), p.v.Cap()
if n < m {
p.v.SetLen(n + 1)
} else {
t := p.v.Type().Elem()
p.v.Set(reflect.Append(p.v, reflect.Zero(t)))
}
elem := p.v.Index(n)
switch elem.Kind() {
case reflect.Int32:
elem.SetInt(int64(int32(x)))
case reflect.Uint32:
elem.SetUint(uint64(x))
case reflect.Float32:
elem.SetFloat(float64(math.Float32frombits(x)))
}
}
func (p word32Slice) Len() int {
return p.v.Len()
}
func (p word32Slice) Index(i int) uint32 {
elem := p.v.Index(i)
switch elem.Kind() {
case reflect.Int32:
return uint32(elem.Int())
case reflect.Uint32:
return uint32(elem.Uint())
case reflect.Float32:
return math.Float32bits(float32(elem.Float()))
}
panic("unreachable")
}
// Word32Slice returns a reference to a []int32, []uint32, []float32, or []enum field in the struct.
func structPointer_Word32Slice(p structPointer, f field) word32Slice {
return word32Slice{structPointer_field(p, f)}
}
// word64 is like word32 but for 64-bit values.
type word64 struct {
v reflect.Value
}
func word64_Set(p word64, o *Buffer, x uint64) {
t := p.v.Type().Elem()
switch t {
case int64Type:
if len(o.int64s) == 0 {
o.int64s = make([]int64, uint64PoolSize)
}
o.int64s[0] = int64(x)
p.v.Set(reflect.ValueOf(&o.int64s[0]))
o.int64s = o.int64s[1:]
return
case uint64Type:
if len(o.uint64s) == 0 {
o.uint64s = make([]uint64, uint64PoolSize)
}
o.uint64s[0] = x
p.v.Set(reflect.ValueOf(&o.uint64s[0]))
o.uint64s = o.uint64s[1:]
return
case float64Type:
if len(o.float64s) == 0 {
o.float64s = make([]float64, uint64PoolSize)
}
o.float64s[0] = math.Float64frombits(x)
p.v.Set(reflect.ValueOf(&o.float64s[0]))
o.float64s = o.float64s[1:]
return
}
panic("unreachable")
}
func word64_IsNil(p word64) bool {
return p.v.IsNil()
}
func word64_Get(p word64) uint64 {
elem := p.v.Elem()
switch elem.Kind() {
case reflect.Int64:
return uint64(elem.Int())
case reflect.Uint64:
return elem.Uint()
case reflect.Float64:
return math.Float64bits(elem.Float())
}
panic("unreachable")
}
func structPointer_Word64(p structPointer, f field) word64 {
return word64{structPointer_field(p, f)}
}
type word64Slice struct {
v reflect.Value
}
func (p word64Slice) Append(x uint64) {
n, m := p.v.Len(), p.v.Cap()
if n < m {
p.v.SetLen(n + 1)
} else {
t := p.v.Type().Elem()
p.v.Set(reflect.Append(p.v, reflect.Zero(t)))
}
elem := p.v.Index(n)
switch elem.Kind() {
case reflect.Int64:
elem.SetInt(int64(int64(x)))
case reflect.Uint64:
elem.SetUint(uint64(x))
case reflect.Float64:
elem.SetFloat(float64(math.Float64frombits(x)))
}
}
func (p word64Slice) Len() int {
return p.v.Len()
}
func (p word64Slice) Index(i int) uint64 {
elem := p.v.Index(i)
switch elem.Kind() {
case reflect.Int64:
return uint64(elem.Int())
case reflect.Uint64:
return uint64(elem.Uint())
case reflect.Float64:
return math.Float64bits(float64(elem.Float()))
}
panic("unreachable")
}
func structPointer_Word64Slice(p structPointer, f field) word64Slice {
return word64Slice{structPointer_field(p, f)}
}

View File

@@ -1,218 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2012 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
// +build !appengine appenginevm
// This file contains the implementation of the proto field accesses using package unsafe.
package proto
import (
"reflect"
"unsafe"
)
// NOTE: These type_Foo functions would more idiomatically be methods,
// but Go does not allow methods on pointer types, and we must preserve
// some pointer type for the garbage collector. We use these
// funcs with clunky names as our poor approximation to methods.
//
// An alternative would be
// type structPointer struct { p unsafe.Pointer }
// but that does not registerize as well.
// A structPointer is a pointer to a struct.
type structPointer unsafe.Pointer
// toStructPointer returns a structPointer equivalent to the given reflect value.
func toStructPointer(v reflect.Value) structPointer {
return structPointer(unsafe.Pointer(v.Pointer()))
}
// IsNil reports whether p is nil.
func structPointer_IsNil(p structPointer) bool {
return p == nil
}
// Interface returns the struct pointer, assumed to have element type t,
// as an interface value.
func structPointer_Interface(p structPointer, t reflect.Type) interface{} {
return reflect.NewAt(t, unsafe.Pointer(p)).Interface()
}
// A field identifies a field in a struct, accessible from a structPointer.
// In this implementation, a field is identified by its byte offset from the start of the struct.
type field uintptr
// toField returns a field equivalent to the given reflect field.
func toField(f *reflect.StructField) field {
return field(f.Offset)
}
// invalidField is an invalid field identifier.
const invalidField = ^field(0)
// IsValid reports whether the field identifier is valid.
func (f field) IsValid() bool {
return f != ^field(0)
}
// Bytes returns the address of a []byte field in the struct.
func structPointer_Bytes(p structPointer, f field) *[]byte {
return (*[]byte)(unsafe.Pointer(uintptr(p) + uintptr(f)))
}
// BytesSlice returns the address of a [][]byte field in the struct.
func structPointer_BytesSlice(p structPointer, f field) *[][]byte {
return (*[][]byte)(unsafe.Pointer(uintptr(p) + uintptr(f)))
}
// Bool returns the address of a *bool field in the struct.
func structPointer_Bool(p structPointer, f field) **bool {
return (**bool)(unsafe.Pointer(uintptr(p) + uintptr(f)))
}
// BoolSlice returns the address of a []bool field in the struct.
func structPointer_BoolSlice(p structPointer, f field) *[]bool {
return (*[]bool)(unsafe.Pointer(uintptr(p) + uintptr(f)))
}
// String returns the address of a *string field in the struct.
func structPointer_String(p structPointer, f field) **string {
return (**string)(unsafe.Pointer(uintptr(p) + uintptr(f)))
}
// StringSlice returns the address of a []string field in the struct.
func structPointer_StringSlice(p structPointer, f field) *[]string {
return (*[]string)(unsafe.Pointer(uintptr(p) + uintptr(f)))
}
// ExtMap returns the address of an extension map field in the struct.
func structPointer_ExtMap(p structPointer, f field) *map[int32]Extension {
return (*map[int32]Extension)(unsafe.Pointer(uintptr(p) + uintptr(f)))
}
// SetStructPointer writes a *struct field in the struct.
func structPointer_SetStructPointer(p structPointer, f field, q structPointer) {
*(*structPointer)(unsafe.Pointer(uintptr(p) + uintptr(f))) = q
}
// GetStructPointer reads a *struct field in the struct.
func structPointer_GetStructPointer(p structPointer, f field) structPointer {
return *(*structPointer)(unsafe.Pointer(uintptr(p) + uintptr(f)))
}
// StructPointerSlice the address of a []*struct field in the struct.
func structPointer_StructPointerSlice(p structPointer, f field) *structPointerSlice {
return (*structPointerSlice)(unsafe.Pointer(uintptr(p) + uintptr(f)))
}
// A structPointerSlice represents a slice of pointers to structs (themselves submessages or groups).
type structPointerSlice []structPointer
func (v *structPointerSlice) Len() int { return len(*v) }
func (v *structPointerSlice) Index(i int) structPointer { return (*v)[i] }
func (v *structPointerSlice) Append(p structPointer) { *v = append(*v, p) }
// A word32 is the address of a "pointer to 32-bit value" field.
type word32 **uint32
// IsNil reports whether *v is nil.
func word32_IsNil(p word32) bool {
return *p == nil
}
// Set sets *v to point at a newly allocated word set to x.
func word32_Set(p word32, o *Buffer, x uint32) {
if len(o.uint32s) == 0 {
o.uint32s = make([]uint32, uint32PoolSize)
}
o.uint32s[0] = x
*p = &o.uint32s[0]
o.uint32s = o.uint32s[1:]
}
// Get gets the value pointed at by *v.
func word32_Get(p word32) uint32 {
return **p
}
// Word32 returns the address of a *int32, *uint32, *float32, or *enum field in the struct.
func structPointer_Word32(p structPointer, f field) word32 {
return word32((**uint32)(unsafe.Pointer(uintptr(p) + uintptr(f))))
}
// A word32Slice is a slice of 32-bit values.
type word32Slice []uint32
func (v *word32Slice) Append(x uint32) { *v = append(*v, x) }
func (v *word32Slice) Len() int { return len(*v) }
func (v *word32Slice) Index(i int) uint32 { return (*v)[i] }
// Word32Slice returns the address of a []int32, []uint32, []float32, or []enum field in the struct.
func structPointer_Word32Slice(p structPointer, f field) *word32Slice {
return (*word32Slice)(unsafe.Pointer(uintptr(p) + uintptr(f)))
}
// word64 is like word32 but for 64-bit values.
type word64 **uint64
func word64_Set(p word64, o *Buffer, x uint64) {
if len(o.uint64s) == 0 {
o.uint64s = make([]uint64, uint64PoolSize)
}
o.uint64s[0] = x
*p = &o.uint64s[0]
o.uint64s = o.uint64s[1:]
}
func word64_IsNil(p word64) bool {
return *p == nil
}
func word64_Get(p word64) uint64 {
return **p
}
func structPointer_Word64(p structPointer, f field) word64 {
return word64((**uint64)(unsafe.Pointer(uintptr(p) + uintptr(f))))
}
// word64Slice is like word32Slice but for 64-bit values.
type word64Slice []uint64
func (v *word64Slice) Append(x uint64) { *v = append(*v, x) }
func (v *word64Slice) Len() int { return len(*v) }
func (v *word64Slice) Index(i int) uint64 { return (*v)[i] }
func structPointer_Word64Slice(p structPointer, f field) *word64Slice {
return (*word64Slice)(unsafe.Pointer(uintptr(p) + uintptr(f)))
}

View File

@@ -1,662 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2010 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
package proto
/*
* Routines for encoding data into the wire format for protocol buffers.
*/
import (
"fmt"
"os"
"reflect"
"sort"
"strconv"
"strings"
"sync"
)
const debug bool = false
// Constants that identify the encoding of a value on the wire.
const (
WireVarint = 0
WireFixed64 = 1
WireBytes = 2
WireStartGroup = 3
WireEndGroup = 4
WireFixed32 = 5
)
const startSize = 10 // initial slice/string sizes
// Encoders are defined in encode.go
// An encoder outputs the full representation of a field, including its
// tag and encoder type.
type encoder func(p *Buffer, prop *Properties, base structPointer) error
// A valueEncoder encodes a single integer in a particular encoding.
type valueEncoder func(o *Buffer, x uint64) error
// Sizers are defined in encode.go
// A sizer returns the encoded size of a field, including its tag and encoder
// type.
type sizer func(prop *Properties, base structPointer) int
// A valueSizer returns the encoded size of a single integer in a particular
// encoding.
type valueSizer func(x uint64) int
// Decoders are defined in decode.go
// A decoder creates a value from its wire representation.
// Unrecognized subelements are saved in unrec.
type decoder func(p *Buffer, prop *Properties, base structPointer) error
// A valueDecoder decodes a single integer in a particular encoding.
type valueDecoder func(o *Buffer) (x uint64, err error)
// tagMap is an optimization over map[int]int for typical protocol buffer
// use-cases. Encoded protocol buffers are often in tag order with small tag
// numbers.
type tagMap struct {
fastTags []int
slowTags map[int]int
}
// tagMapFastLimit is the upper bound on the tag number that will be stored in
// the tagMap slice rather than its map.
const tagMapFastLimit = 1024
func (p *tagMap) get(t int) (int, bool) {
if t > 0 && t < tagMapFastLimit {
if t >= len(p.fastTags) {
return 0, false
}
fi := p.fastTags[t]
return fi, fi >= 0
}
fi, ok := p.slowTags[t]
return fi, ok
}
func (p *tagMap) put(t int, fi int) {
if t > 0 && t < tagMapFastLimit {
for len(p.fastTags) < t+1 {
p.fastTags = append(p.fastTags, -1)
}
p.fastTags[t] = fi
return
}
if p.slowTags == nil {
p.slowTags = make(map[int]int)
}
p.slowTags[t] = fi
}
// StructProperties represents properties for all the fields of a struct.
// decoderTags and decoderOrigNames should only be used by the decoder.
type StructProperties struct {
Prop []*Properties // properties for each field
reqCount int // required count
decoderTags tagMap // map from proto tag to struct field number
decoderOrigNames map[string]int // map from original name to struct field number
order []int // list of struct field numbers in tag order
unrecField field // field id of the XXX_unrecognized []byte field
extendable bool // is this an extendable proto
}
// Implement the sorting interface so we can sort the fields in tag order, as recommended by the spec.
// See encode.go, (*Buffer).enc_struct.
func (sp *StructProperties) Len() int { return len(sp.order) }
func (sp *StructProperties) Less(i, j int) bool {
return sp.Prop[sp.order[i]].Tag < sp.Prop[sp.order[j]].Tag
}
func (sp *StructProperties) Swap(i, j int) { sp.order[i], sp.order[j] = sp.order[j], sp.order[i] }
// Properties represents the protocol-specific behavior of a single struct field.
type Properties struct {
Name string // name of the field, for error messages
OrigName string // original name before protocol compiler (always set)
Wire string
WireType int
Tag int
Required bool
Optional bool
Repeated bool
Packed bool // relevant for repeated primitives only
Enum string // set for enum types only
Default string // default value
HasDefault bool // whether an explicit default was provided
def_uint64 uint64
enc encoder
valEnc valueEncoder // set for bool and numeric types only
field field
tagcode []byte // encoding of EncodeVarint((Tag<<3)|WireType)
tagbuf [8]byte
stype reflect.Type // set for struct types only
sprop *StructProperties // set for struct types only
isMarshaler bool
isUnmarshaler bool
size sizer
valSize valueSizer // set for bool and numeric types only
dec decoder
valDec valueDecoder // set for bool and numeric types only
// If this is a packable field, this will be the decoder for the packed version of the field.
packedDec decoder
}
// String formats the properties in the protobuf struct field tag style.
func (p *Properties) String() string {
s := p.Wire
s = ","
s += strconv.Itoa(p.Tag)
if p.Required {
s += ",req"
}
if p.Optional {
s += ",opt"
}
if p.Repeated {
s += ",rep"
}
if p.Packed {
s += ",packed"
}
if p.OrigName != p.Name {
s += ",name=" + p.OrigName
}
if len(p.Enum) > 0 {
s += ",enum=" + p.Enum
}
if p.HasDefault {
s += ",def=" + p.Default
}
return s
}
// Parse populates p by parsing a string in the protobuf struct field tag style.
func (p *Properties) Parse(s string) {
// "bytes,49,opt,name=foo,def=hello!"
fields := strings.Split(s, ",") // breaks def=, but handled below.
if len(fields) < 2 {
fmt.Fprintf(os.Stderr, "proto: tag has too few fields: %q\n", s)
return
}
p.Wire = fields[0]
switch p.Wire {
case "varint":
p.WireType = WireVarint
p.valEnc = (*Buffer).EncodeVarint
p.valDec = (*Buffer).DecodeVarint
p.valSize = sizeVarint
case "fixed32":
p.WireType = WireFixed32
p.valEnc = (*Buffer).EncodeFixed32
p.valDec = (*Buffer).DecodeFixed32
p.valSize = sizeFixed32
case "fixed64":
p.WireType = WireFixed64
p.valEnc = (*Buffer).EncodeFixed64
p.valDec = (*Buffer).DecodeFixed64
p.valSize = sizeFixed64
case "zigzag32":
p.WireType = WireVarint
p.valEnc = (*Buffer).EncodeZigzag32
p.valDec = (*Buffer).DecodeZigzag32
p.valSize = sizeZigzag32
case "zigzag64":
p.WireType = WireVarint
p.valEnc = (*Buffer).EncodeZigzag64
p.valDec = (*Buffer).DecodeZigzag64
p.valSize = sizeZigzag64
case "bytes", "group":
p.WireType = WireBytes
// no numeric converter for non-numeric types
default:
fmt.Fprintf(os.Stderr, "proto: tag has unknown wire type: %q\n", s)
return
}
var err error
p.Tag, err = strconv.Atoi(fields[1])
if err != nil {
return
}
for i := 2; i < len(fields); i++ {
f := fields[i]
switch {
case f == "req":
p.Required = true
case f == "opt":
p.Optional = true
case f == "rep":
p.Repeated = true
case f == "packed":
p.Packed = true
case strings.HasPrefix(f, "name="):
p.OrigName = f[5:]
case strings.HasPrefix(f, "enum="):
p.Enum = f[5:]
case strings.HasPrefix(f, "def="):
p.HasDefault = true
p.Default = f[4:] // rest of string
if i+1 < len(fields) {
// Commas aren't escaped, and def is always last.
p.Default += "," + strings.Join(fields[i+1:], ",")
break
}
}
}
}
func logNoSliceEnc(t1, t2 reflect.Type) {
fmt.Fprintf(os.Stderr, "proto: no slice oenc for %T = []%T\n", t1, t2)
}
var protoMessageType = reflect.TypeOf((*Message)(nil)).Elem()
// Initialize the fields for encoding and decoding.
func (p *Properties) setEncAndDec(typ reflect.Type, lockGetProp bool) {
p.enc = nil
p.dec = nil
p.size = nil
switch t1 := typ; t1.Kind() {
default:
fmt.Fprintf(os.Stderr, "proto: no coders for %v\n", t1)
case reflect.Ptr:
switch t2 := t1.Elem(); t2.Kind() {
default:
fmt.Fprintf(os.Stderr, "proto: no encoder function for %T -> %T\n", t1, t2)
break
case reflect.Bool:
p.enc = (*Buffer).enc_bool
p.dec = (*Buffer).dec_bool
p.size = size_bool
case reflect.Int32:
p.enc = (*Buffer).enc_int32
p.dec = (*Buffer).dec_int32
p.size = size_int32
case reflect.Uint32:
p.enc = (*Buffer).enc_uint32
p.dec = (*Buffer).dec_int32 // can reuse
p.size = size_uint32
case reflect.Int64, reflect.Uint64:
p.enc = (*Buffer).enc_int64
p.dec = (*Buffer).dec_int64
p.size = size_int64
case reflect.Float32:
p.enc = (*Buffer).enc_uint32 // can just treat them as bits
p.dec = (*Buffer).dec_int32
p.size = size_uint32
case reflect.Float64:
p.enc = (*Buffer).enc_int64 // can just treat them as bits
p.dec = (*Buffer).dec_int64
p.size = size_int64
case reflect.String:
p.enc = (*Buffer).enc_string
p.dec = (*Buffer).dec_string
p.size = size_string
case reflect.Struct:
p.stype = t1.Elem()
p.isMarshaler = isMarshaler(t1)
p.isUnmarshaler = isUnmarshaler(t1)
if p.Wire == "bytes" {
p.enc = (*Buffer).enc_struct_message
p.dec = (*Buffer).dec_struct_message
p.size = size_struct_message
} else {
p.enc = (*Buffer).enc_struct_group
p.dec = (*Buffer).dec_struct_group
p.size = size_struct_group
}
}
case reflect.Slice:
switch t2 := t1.Elem(); t2.Kind() {
default:
logNoSliceEnc(t1, t2)
break
case reflect.Bool:
if p.Packed {
p.enc = (*Buffer).enc_slice_packed_bool
p.size = size_slice_packed_bool
} else {
p.enc = (*Buffer).enc_slice_bool
p.size = size_slice_bool
}
p.dec = (*Buffer).dec_slice_bool
p.packedDec = (*Buffer).dec_slice_packed_bool
case reflect.Int32:
if p.Packed {
p.enc = (*Buffer).enc_slice_packed_int32
p.size = size_slice_packed_int32
} else {
p.enc = (*Buffer).enc_slice_int32
p.size = size_slice_int32
}
p.dec = (*Buffer).dec_slice_int32
p.packedDec = (*Buffer).dec_slice_packed_int32
case reflect.Uint32:
if p.Packed {
p.enc = (*Buffer).enc_slice_packed_uint32
p.size = size_slice_packed_uint32
} else {
p.enc = (*Buffer).enc_slice_uint32
p.size = size_slice_uint32
}
p.dec = (*Buffer).dec_slice_int32
p.packedDec = (*Buffer).dec_slice_packed_int32
case reflect.Int64, reflect.Uint64:
if p.Packed {
p.enc = (*Buffer).enc_slice_packed_int64
p.size = size_slice_packed_int64
} else {
p.enc = (*Buffer).enc_slice_int64
p.size = size_slice_int64
}
p.dec = (*Buffer).dec_slice_int64
p.packedDec = (*Buffer).dec_slice_packed_int64
case reflect.Uint8:
p.enc = (*Buffer).enc_slice_byte
p.dec = (*Buffer).dec_slice_byte
p.size = size_slice_byte
case reflect.Float32, reflect.Float64:
switch t2.Bits() {
case 32:
// can just treat them as bits
if p.Packed {
p.enc = (*Buffer).enc_slice_packed_uint32
p.size = size_slice_packed_uint32
} else {
p.enc = (*Buffer).enc_slice_uint32
p.size = size_slice_uint32
}
p.dec = (*Buffer).dec_slice_int32
p.packedDec = (*Buffer).dec_slice_packed_int32
case 64:
// can just treat them as bits
if p.Packed {
p.enc = (*Buffer).enc_slice_packed_int64
p.size = size_slice_packed_int64
} else {
p.enc = (*Buffer).enc_slice_int64
p.size = size_slice_int64
}
p.dec = (*Buffer).dec_slice_int64
p.packedDec = (*Buffer).dec_slice_packed_int64
default:
logNoSliceEnc(t1, t2)
break
}
case reflect.String:
p.enc = (*Buffer).enc_slice_string
p.dec = (*Buffer).dec_slice_string
p.size = size_slice_string
case reflect.Ptr:
switch t3 := t2.Elem(); t3.Kind() {
default:
fmt.Fprintf(os.Stderr, "proto: no ptr oenc for %T -> %T -> %T\n", t1, t2, t3)
break
case reflect.Struct:
p.stype = t2.Elem()
p.isMarshaler = isMarshaler(t2)
p.isUnmarshaler = isUnmarshaler(t2)
if p.Wire == "bytes" {
p.enc = (*Buffer).enc_slice_struct_message
p.dec = (*Buffer).dec_slice_struct_message
p.size = size_slice_struct_message
} else {
p.enc = (*Buffer).enc_slice_struct_group
p.dec = (*Buffer).dec_slice_struct_group
p.size = size_slice_struct_group
}
}
case reflect.Slice:
switch t2.Elem().Kind() {
default:
fmt.Fprintf(os.Stderr, "proto: no slice elem oenc for %T -> %T -> %T\n", t1, t2, t2.Elem())
break
case reflect.Uint8:
p.enc = (*Buffer).enc_slice_slice_byte
p.dec = (*Buffer).dec_slice_slice_byte
p.size = size_slice_slice_byte
}
}
}
// precalculate tag code
wire := p.WireType
if p.Packed {
wire = WireBytes
}
x := uint32(p.Tag)<<3 | uint32(wire)
i := 0
for i = 0; x > 127; i++ {
p.tagbuf[i] = 0x80 | uint8(x&0x7F)
x >>= 7
}
p.tagbuf[i] = uint8(x)
p.tagcode = p.tagbuf[0 : i+1]
if p.stype != nil {
if lockGetProp {
p.sprop = GetProperties(p.stype)
} else {
p.sprop = getPropertiesLocked(p.stype)
}
}
}
var (
marshalerType = reflect.TypeOf((*Marshaler)(nil)).Elem()
unmarshalerType = reflect.TypeOf((*Unmarshaler)(nil)).Elem()
)
// isMarshaler reports whether type t implements Marshaler.
func isMarshaler(t reflect.Type) bool {
// We're checking for (likely) pointer-receiver methods
// so if t is not a pointer, something is very wrong.
// The calls above only invoke isMarshaler on pointer types.
if t.Kind() != reflect.Ptr {
panic("proto: misuse of isMarshaler")
}
return t.Implements(marshalerType)
}
// isUnmarshaler reports whether type t implements Unmarshaler.
func isUnmarshaler(t reflect.Type) bool {
// We're checking for (likely) pointer-receiver methods
// so if t is not a pointer, something is very wrong.
// The calls above only invoke isUnmarshaler on pointer types.
if t.Kind() != reflect.Ptr {
panic("proto: misuse of isUnmarshaler")
}
return t.Implements(unmarshalerType)
}
// Init populates the properties from a protocol buffer struct tag.
func (p *Properties) Init(typ reflect.Type, name, tag string, f *reflect.StructField) {
p.init(typ, name, tag, f, true)
}
func (p *Properties) init(typ reflect.Type, name, tag string, f *reflect.StructField, lockGetProp bool) {
// "bytes,49,opt,def=hello!"
p.Name = name
p.OrigName = name
if f != nil {
p.field = toField(f)
}
if tag == "" {
return
}
p.Parse(tag)
p.setEncAndDec(typ, lockGetProp)
}
var (
mutex sync.Mutex
propertiesMap = make(map[reflect.Type]*StructProperties)
)
// GetProperties returns the list of properties for the type represented by t.
// t must represent a generated struct type of a protocol message.
func GetProperties(t reflect.Type) *StructProperties {
if t.Kind() != reflect.Struct {
panic("proto: type must have kind struct")
}
mutex.Lock()
sprop := getPropertiesLocked(t)
mutex.Unlock()
return sprop
}
// getPropertiesLocked requires that mutex is held.
func getPropertiesLocked(t reflect.Type) *StructProperties {
if prop, ok := propertiesMap[t]; ok {
if collectStats {
stats.Chit++
}
return prop
}
if collectStats {
stats.Cmiss++
}
prop := new(StructProperties)
// in case of recursive protos, fill this in now.
propertiesMap[t] = prop
// build properties
prop.extendable = reflect.PtrTo(t).Implements(extendableProtoType)
prop.unrecField = invalidField
prop.Prop = make([]*Properties, t.NumField())
prop.order = make([]int, t.NumField())
for i := 0; i < t.NumField(); i++ {
f := t.Field(i)
p := new(Properties)
name := f.Name
p.init(f.Type, name, f.Tag.Get("protobuf"), &f, false)
if f.Name == "XXX_extensions" { // special case
p.enc = (*Buffer).enc_map
p.dec = nil // not needed
p.size = size_map
}
if f.Name == "XXX_unrecognized" { // special case
prop.unrecField = toField(&f)
}
prop.Prop[i] = p
prop.order[i] = i
if debug {
print(i, " ", f.Name, " ", t.String(), " ")
if p.Tag > 0 {
print(p.String())
}
print("\n")
}
if p.enc == nil && !strings.HasPrefix(f.Name, "XXX_") {
fmt.Fprintln(os.Stderr, "proto: no encoder for", f.Name, f.Type.String(), "[GetProperties]")
}
}
// Re-order prop.order.
sort.Sort(prop)
// build required counts
// build tags
reqCount := 0
prop.decoderOrigNames = make(map[string]int)
for i, p := range prop.Prop {
if strings.HasPrefix(p.Name, "XXX_") {
// Internal fields should not appear in tags/origNames maps.
// They are handled specially when encoding and decoding.
continue
}
if p.Required {
reqCount++
}
prop.decoderTags.put(p.Tag, i)
prop.decoderOrigNames[p.OrigName] = i
}
prop.reqCount = reqCount
return prop
}
// Return the Properties object for the x[0]'th field of the structure.
func propByIndex(t reflect.Type, x []int) *Properties {
if len(x) != 1 {
fmt.Fprintf(os.Stderr, "proto: field index dimension %d (not 1) for type %s\n", len(x), t)
return nil
}
prop := GetProperties(t)
return prop.Prop[x[0]]
}
// Get the address and type of a pointer to a struct from an interface.
func getbase(pb Message) (t reflect.Type, b structPointer, err error) {
if pb == nil {
err = ErrNil
return
}
// get the reflect type of the pointer to the struct.
t = reflect.TypeOf(pb)
// get the address of the struct.
value := reflect.ValueOf(pb)
b = toStructPointer(value)
return
}
// A global registry of enum types.
// The generated code will register the generated maps by calling RegisterEnum.
var enumValueMaps = make(map[string]map[string]int32)
// RegisterEnum is called from the generated code to install the enum descriptor
// maps into the global table to aid parsing text format protocol buffers.
func RegisterEnum(typeName string, unusedNameMap map[int32]string, valueMap map[string]int32) {
if _, ok := enumValueMaps[typeName]; ok {
panic("proto: duplicate enum registered: " + typeName)
}
enumValueMaps[typeName] = valueMap
}

View File

@@ -1,120 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2012 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
package proto_test
import (
"log"
"testing"
pb "./testdata"
. "github.com/coreos/etcd/Godeps/_workspace/src/code.google.com/p/goprotobuf/proto"
)
var messageWithExtension1 = &pb.MyMessage{Count: Int32(7)}
// messageWithExtension2 is in equal_test.go.
var messageWithExtension3 = &pb.MyMessage{Count: Int32(8)}
func init() {
if err := SetExtension(messageWithExtension1, pb.E_Ext_More, &pb.Ext{Data: String("Abbott")}); err != nil {
log.Panicf("SetExtension: %v", err)
}
if err := SetExtension(messageWithExtension3, pb.E_Ext_More, &pb.Ext{Data: String("Costello")}); err != nil {
log.Panicf("SetExtension: %v", err)
}
// Force messageWithExtension3 to have the extension encoded.
Marshal(messageWithExtension3)
}
var SizeTests = []struct {
desc string
pb Message
}{
{"empty", &pb.OtherMessage{}},
// Basic types.
{"bool", &pb.Defaults{F_Bool: Bool(true)}},
{"int32", &pb.Defaults{F_Int32: Int32(12)}},
{"negative int32", &pb.Defaults{F_Int32: Int32(-1)}},
{"small int64", &pb.Defaults{F_Int64: Int64(1)}},
{"big int64", &pb.Defaults{F_Int64: Int64(1 << 20)}},
{"negative int64", &pb.Defaults{F_Int64: Int64(-1)}},
{"fixed32", &pb.Defaults{F_Fixed32: Uint32(71)}},
{"fixed64", &pb.Defaults{F_Fixed64: Uint64(72)}},
{"uint32", &pb.Defaults{F_Uint32: Uint32(123)}},
{"uint64", &pb.Defaults{F_Uint64: Uint64(124)}},
{"float", &pb.Defaults{F_Float: Float32(12.6)}},
{"double", &pb.Defaults{F_Double: Float64(13.9)}},
{"string", &pb.Defaults{F_String: String("niles")}},
{"bytes", &pb.Defaults{F_Bytes: []byte("wowsa")}},
{"bytes, empty", &pb.Defaults{F_Bytes: []byte{}}},
{"sint32", &pb.Defaults{F_Sint32: Int32(65)}},
{"sint64", &pb.Defaults{F_Sint64: Int64(67)}},
{"enum", &pb.Defaults{F_Enum: pb.Defaults_BLUE.Enum()}},
// Repeated.
{"empty repeated bool", &pb.MoreRepeated{Bools: []bool{}}},
{"repeated bool", &pb.MoreRepeated{Bools: []bool{false, true, true, false}}},
{"packed repeated bool", &pb.MoreRepeated{BoolsPacked: []bool{false, true, true, false, true, true, true}}},
{"repeated int32", &pb.MoreRepeated{Ints: []int32{1, 12203, 1729, -1}}},
{"repeated int32 packed", &pb.MoreRepeated{IntsPacked: []int32{1, 12203, 1729}}},
{"repeated int64 packed", &pb.MoreRepeated{Int64SPacked: []int64{
// Need enough large numbers to verify that the header is counting the number of bytes
// for the field, not the number of elements.
1 << 62, 1 << 62, 1 << 62, 1 << 62, 1 << 62, 1 << 62, 1 << 62, 1 << 62, 1 << 62, 1 << 62,
1 << 62, 1 << 62, 1 << 62, 1 << 62, 1 << 62, 1 << 62, 1 << 62, 1 << 62, 1 << 62, 1 << 62,
}}},
{"repeated string", &pb.MoreRepeated{Strings: []string{"r", "ken", "gri"}}},
{"repeated fixed", &pb.MoreRepeated{Fixeds: []uint32{1, 2, 3, 4}}},
// Nested.
{"nested", &pb.OldMessage{Nested: &pb.OldMessage_Nested{Name: String("whatever")}}},
{"group", &pb.GroupOld{G: &pb.GroupOld_G{X: Int32(12345)}}},
// Other things.
{"unrecognized", &pb.MoreRepeated{XXX_unrecognized: []byte{13<<3 | 0, 4}}},
{"extension (unencoded)", messageWithExtension1},
{"extension (encoded)", messageWithExtension3},
}
func TestSize(t *testing.T) {
for _, tc := range SizeTests {
size := Size(tc.pb)
b, err := Marshal(tc.pb)
if err != nil {
t.Errorf("%v: Marshal failed: %v", tc.desc, err)
continue
}
if size != len(b) {
t.Errorf("%v: Size(%v) = %d, want %d", tc.desc, tc.pb, size, len(b))
t.Logf("%v: bytes: %#v", tc.desc, b)
}
}
}

View File

@@ -1,50 +0,0 @@
# Go support for Protocol Buffers - Google's data interchange format
#
# Copyright 2010 The Go Authors. All rights reserved.
# http://code.google.com/p/goprotobuf/
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are
# met:
#
# * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# * Redistributions in binary form must reproduce the above
# copyright notice, this list of conditions and the following disclaimer
# in the documentation and/or other materials provided with the
# distribution.
# * Neither the name of Google Inc. nor the names of its
# contributors may be used to endorse or promote products derived from
# this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
include ../../Make.protobuf
all: regenerate
regenerate:
rm -f test.pb.go
make test.pb.go
# The following rules are just aids to development. Not needed for typical testing.
diff: regenerate
hg diff test.pb.go
restore:
cp test.pb.go.golden test.pb.go
preserve:
cp test.pb.go test.pb.go.golden

View File

@@ -1,86 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2012 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
// Verify that the compiler output for test.proto is unchanged.
package testdata
import (
"crypto/sha1"
"fmt"
"io/ioutil"
"os"
"os/exec"
"path/filepath"
"testing"
)
// sum returns in string form (for easy comparison) the SHA-1 hash of the named file.
func sum(t *testing.T, name string) string {
data, err := ioutil.ReadFile(name)
if err != nil {
t.Fatal(err)
}
t.Logf("sum(%q): length is %d", name, len(data))
hash := sha1.New()
_, err = hash.Write(data)
if err != nil {
t.Fatal(err)
}
return fmt.Sprintf("% x", hash.Sum(nil))
}
func run(t *testing.T, name string, args ...string) {
cmd := exec.Command(name, args...)
cmd.Stdin = os.Stdin
cmd.Stdout = os.Stdout
cmd.Stderr = os.Stderr
err := cmd.Run()
if err != nil {
t.Fatal(err)
}
}
func TestGolden(t *testing.T) {
// Compute the original checksum.
goldenSum := sum(t, "test.pb.go")
// Run the proto compiler.
run(t, "protoc", "--go_out="+os.TempDir(), "test.proto")
newFile := filepath.Join(os.TempDir(), "test.pb.go")
defer os.Remove(newFile)
// Compute the new checksum.
newSum := sum(t, newFile)
// Verify
if newSum != goldenSum {
run(t, "diff", "-u", "test.pb.go", newFile)
t.Fatal("Code generated by protoc-gen-go has changed; update test.pb.go")
}
}

File diff suppressed because it is too large Load Diff

View File

@@ -1,428 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2010 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
// A feature-rich test file for the protocol compiler and libraries.
syntax = "proto2";
package testdata;
enum FOO { FOO1 = 1; };
message GoEnum {
required FOO foo = 1;
}
message GoTestField {
required string Label = 1;
required string Type = 2;
}
message GoTest {
// An enum, for completeness.
enum KIND {
VOID = 0;
// Basic types
BOOL = 1;
BYTES = 2;
FINGERPRINT = 3;
FLOAT = 4;
INT = 5;
STRING = 6;
TIME = 7;
// Groupings
TUPLE = 8;
ARRAY = 9;
MAP = 10;
// Table types
TABLE = 11;
// Functions
FUNCTION = 12; // last tag
};
// Some typical parameters
required KIND Kind = 1;
optional string Table = 2;
optional int32 Param = 3;
// Required, repeated and optional foreign fields.
required GoTestField RequiredField = 4;
repeated GoTestField RepeatedField = 5;
optional GoTestField OptionalField = 6;
// Required fields of all basic types
required bool F_Bool_required = 10;
required int32 F_Int32_required = 11;
required int64 F_Int64_required = 12;
required fixed32 F_Fixed32_required = 13;
required fixed64 F_Fixed64_required = 14;
required uint32 F_Uint32_required = 15;
required uint64 F_Uint64_required = 16;
required float F_Float_required = 17;
required double F_Double_required = 18;
required string F_String_required = 19;
required bytes F_Bytes_required = 101;
required sint32 F_Sint32_required = 102;
required sint64 F_Sint64_required = 103;
// Repeated fields of all basic types
repeated bool F_Bool_repeated = 20;
repeated int32 F_Int32_repeated = 21;
repeated int64 F_Int64_repeated = 22;
repeated fixed32 F_Fixed32_repeated = 23;
repeated fixed64 F_Fixed64_repeated = 24;
repeated uint32 F_Uint32_repeated = 25;
repeated uint64 F_Uint64_repeated = 26;
repeated float F_Float_repeated = 27;
repeated double F_Double_repeated = 28;
repeated string F_String_repeated = 29;
repeated bytes F_Bytes_repeated = 201;
repeated sint32 F_Sint32_repeated = 202;
repeated sint64 F_Sint64_repeated = 203;
// Optional fields of all basic types
optional bool F_Bool_optional = 30;
optional int32 F_Int32_optional = 31;
optional int64 F_Int64_optional = 32;
optional fixed32 F_Fixed32_optional = 33;
optional fixed64 F_Fixed64_optional = 34;
optional uint32 F_Uint32_optional = 35;
optional uint64 F_Uint64_optional = 36;
optional float F_Float_optional = 37;
optional double F_Double_optional = 38;
optional string F_String_optional = 39;
optional bytes F_Bytes_optional = 301;
optional sint32 F_Sint32_optional = 302;
optional sint64 F_Sint64_optional = 303;
// Default-valued fields of all basic types
optional bool F_Bool_defaulted = 40 [default=true];
optional int32 F_Int32_defaulted = 41 [default=32];
optional int64 F_Int64_defaulted = 42 [default=64];
optional fixed32 F_Fixed32_defaulted = 43 [default=320];
optional fixed64 F_Fixed64_defaulted = 44 [default=640];
optional uint32 F_Uint32_defaulted = 45 [default=3200];
optional uint64 F_Uint64_defaulted = 46 [default=6400];
optional float F_Float_defaulted = 47 [default=314159.];
optional double F_Double_defaulted = 48 [default=271828.];
optional string F_String_defaulted = 49 [default="hello, \"world!\"\n"];
optional bytes F_Bytes_defaulted = 401 [default="Bignose"];
optional sint32 F_Sint32_defaulted = 402 [default = -32];
optional sint64 F_Sint64_defaulted = 403 [default = -64];
// Packed repeated fields (no string or bytes).
repeated bool F_Bool_repeated_packed = 50 [packed=true];
repeated int32 F_Int32_repeated_packed = 51 [packed=true];
repeated int64 F_Int64_repeated_packed = 52 [packed=true];
repeated fixed32 F_Fixed32_repeated_packed = 53 [packed=true];
repeated fixed64 F_Fixed64_repeated_packed = 54 [packed=true];
repeated uint32 F_Uint32_repeated_packed = 55 [packed=true];
repeated uint64 F_Uint64_repeated_packed = 56 [packed=true];
repeated float F_Float_repeated_packed = 57 [packed=true];
repeated double F_Double_repeated_packed = 58 [packed=true];
repeated sint32 F_Sint32_repeated_packed = 502 [packed=true];
repeated sint64 F_Sint64_repeated_packed = 503 [packed=true];
// Required, repeated, and optional groups.
required group RequiredGroup = 70 {
required string RequiredField = 71;
};
repeated group RepeatedGroup = 80 {
required string RequiredField = 81;
};
optional group OptionalGroup = 90 {
required string RequiredField = 91;
};
}
// For testing skipping of unrecognized fields.
// Numbers are all big, larger than tag numbers in GoTestField,
// the message used in the corresponding test.
message GoSkipTest {
required int32 skip_int32 = 11;
required fixed32 skip_fixed32 = 12;
required fixed64 skip_fixed64 = 13;
required string skip_string = 14;
required group SkipGroup = 15 {
required int32 group_int32 = 16;
required string group_string = 17;
}
}
// For testing packed/non-packed decoder switching.
// A serialized instance of one should be deserializable as the other.
message NonPackedTest {
repeated int32 a = 1;
}
message PackedTest {
repeated int32 b = 1 [packed=true];
}
message MaxTag {
// Maximum possible tag number.
optional string last_field = 536870911;
}
message OldMessage {
message Nested {
optional string name = 1;
}
optional Nested nested = 1;
optional int32 num = 2;
}
// NewMessage is wire compatible with OldMessage;
// imagine it as a future version.
message NewMessage {
message Nested {
optional string name = 1;
optional string food_group = 2;
}
optional Nested nested = 1;
// This is an int32 in OldMessage.
optional int64 num = 2;
}
// Smaller tests for ASCII formatting.
message InnerMessage {
required string host = 1;
optional int32 port = 2 [default=4000];
optional bool connected = 3;
}
message OtherMessage {
optional int64 key = 1;
optional bytes value = 2;
optional float weight = 3;
optional InnerMessage inner = 4;
}
message MyMessage {
required int32 count = 1;
optional string name = 2;
optional string quote = 3;
repeated string pet = 4;
optional InnerMessage inner = 5;
repeated OtherMessage others = 6;
repeated InnerMessage rep_inner = 12;
enum Color {
RED = 0;
GREEN = 1;
BLUE = 2;
};
optional Color bikeshed = 7;
optional group SomeGroup = 8 {
optional int32 group_field = 9;
}
// This field becomes [][]byte in the generated code.
repeated bytes rep_bytes = 10;
optional double bigfloat = 11;
extensions 100 to max;
}
message Ext {
extend MyMessage {
optional Ext more = 103;
optional string text = 104;
optional int32 number = 105;
}
optional string data = 1;
}
extend MyMessage {
repeated string greeting = 106;
}
message MyMessageSet {
option message_set_wire_format = true;
extensions 100 to max;
}
message Empty {
}
extend MyMessageSet {
optional Empty x201 = 201;
optional Empty x202 = 202;
optional Empty x203 = 203;
optional Empty x204 = 204;
optional Empty x205 = 205;
optional Empty x206 = 206;
optional Empty x207 = 207;
optional Empty x208 = 208;
optional Empty x209 = 209;
optional Empty x210 = 210;
optional Empty x211 = 211;
optional Empty x212 = 212;
optional Empty x213 = 213;
optional Empty x214 = 214;
optional Empty x215 = 215;
optional Empty x216 = 216;
optional Empty x217 = 217;
optional Empty x218 = 218;
optional Empty x219 = 219;
optional Empty x220 = 220;
optional Empty x221 = 221;
optional Empty x222 = 222;
optional Empty x223 = 223;
optional Empty x224 = 224;
optional Empty x225 = 225;
optional Empty x226 = 226;
optional Empty x227 = 227;
optional Empty x228 = 228;
optional Empty x229 = 229;
optional Empty x230 = 230;
optional Empty x231 = 231;
optional Empty x232 = 232;
optional Empty x233 = 233;
optional Empty x234 = 234;
optional Empty x235 = 235;
optional Empty x236 = 236;
optional Empty x237 = 237;
optional Empty x238 = 238;
optional Empty x239 = 239;
optional Empty x240 = 240;
optional Empty x241 = 241;
optional Empty x242 = 242;
optional Empty x243 = 243;
optional Empty x244 = 244;
optional Empty x245 = 245;
optional Empty x246 = 246;
optional Empty x247 = 247;
optional Empty x248 = 248;
optional Empty x249 = 249;
optional Empty x250 = 250;
}
message MessageList {
repeated group Message = 1 {
required string name = 2;
required int32 count = 3;
}
}
message Strings {
optional string string_field = 1;
optional bytes bytes_field = 2;
}
message Defaults {
enum Color {
RED = 0;
GREEN = 1;
BLUE = 2;
}
// Default-valued fields of all basic types.
// Same as GoTest, but copied here to make testing easier.
optional bool F_Bool = 1 [default=true];
optional int32 F_Int32 = 2 [default=32];
optional int64 F_Int64 = 3 [default=64];
optional fixed32 F_Fixed32 = 4 [default=320];
optional fixed64 F_Fixed64 = 5 [default=640];
optional uint32 F_Uint32 = 6 [default=3200];
optional uint64 F_Uint64 = 7 [default=6400];
optional float F_Float = 8 [default=314159.];
optional double F_Double = 9 [default=271828.];
optional string F_String = 10 [default="hello, \"world!\"\n"];
optional bytes F_Bytes = 11 [default="Bignose"];
optional sint32 F_Sint32 = 12 [default=-32];
optional sint64 F_Sint64 = 13 [default=-64];
optional Color F_Enum = 14 [default=GREEN];
// More fields with crazy defaults.
optional float F_Pinf = 15 [default=inf];
optional float F_Ninf = 16 [default=-inf];
optional float F_Nan = 17 [default=nan];
// Sub-message.
optional SubDefaults sub = 18;
// Redundant but explicit defaults.
optional string str_zero = 19 [default=""];
}
message SubDefaults {
optional int64 n = 1 [default=7];
}
message RepeatedEnum {
enum Color {
RED = 1;
}
repeated Color color = 1;
}
message MoreRepeated {
repeated bool bools = 1;
repeated bool bools_packed = 2 [packed=true];
repeated int32 ints = 3;
repeated int32 ints_packed = 4 [packed=true];
repeated int64 int64s_packed = 7 [packed=true];
repeated string strings = 5;
repeated fixed32 fixeds = 6;
}
// GroupOld and GroupNew have the same wire format.
// GroupNew has a new field inside a group.
message GroupOld {
optional group G = 101 {
optional int32 x = 2;
}
}
message GroupNew {
optional group G = 101 {
optional int32 x = 2;
optional int32 y = 3;
}
}
message FloatingPoint {
required double f = 1;
}

View File

@@ -1,695 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2010 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
package proto
// Functions for writing the text protocol buffer format.
import (
"bufio"
"bytes"
"encoding"
"fmt"
"io"
"log"
"math"
"os"
"reflect"
"sort"
"strings"
)
var (
newline = []byte("\n")
spaces = []byte(" ")
gtNewline = []byte(">\n")
endBraceNewline = []byte("}\n")
backslashN = []byte{'\\', 'n'}
backslashR = []byte{'\\', 'r'}
backslashT = []byte{'\\', 't'}
backslashDQ = []byte{'\\', '"'}
backslashBS = []byte{'\\', '\\'}
posInf = []byte("inf")
negInf = []byte("-inf")
nan = []byte("nan")
)
type writer interface {
io.Writer
WriteByte(byte) error
}
// textWriter is an io.Writer that tracks its indentation level.
type textWriter struct {
ind int
complete bool // if the current position is a complete line
compact bool // whether to write out as a one-liner
w writer
}
func (w *textWriter) WriteString(s string) (n int, err error) {
if !strings.Contains(s, "\n") {
if !w.compact && w.complete {
w.writeIndent()
}
w.complete = false
return io.WriteString(w.w, s)
}
// WriteString is typically called without newlines, so this
// codepath and its copy are rare. We copy to avoid
// duplicating all of Write's logic here.
return w.Write([]byte(s))
}
func (w *textWriter) Write(p []byte) (n int, err error) {
newlines := bytes.Count(p, newline)
if newlines == 0 {
if !w.compact && w.complete {
w.writeIndent()
}
n, err = w.w.Write(p)
w.complete = false
return n, err
}
frags := bytes.SplitN(p, newline, newlines+1)
if w.compact {
for i, frag := range frags {
if i > 0 {
if err := w.w.WriteByte(' '); err != nil {
return n, err
}
n++
}
nn, err := w.w.Write(frag)
n += nn
if err != nil {
return n, err
}
}
return n, nil
}
for i, frag := range frags {
if w.complete {
w.writeIndent()
}
nn, err := w.w.Write(frag)
n += nn
if err != nil {
return n, err
}
if i+1 < len(frags) {
if err := w.w.WriteByte('\n'); err != nil {
return n, err
}
n++
}
}
w.complete = len(frags[len(frags)-1]) == 0
return n, nil
}
func (w *textWriter) WriteByte(c byte) error {
if w.compact && c == '\n' {
c = ' '
}
if !w.compact && w.complete {
w.writeIndent()
}
err := w.w.WriteByte(c)
w.complete = c == '\n'
return err
}
func (w *textWriter) indent() { w.ind++ }
func (w *textWriter) unindent() {
if w.ind == 0 {
log.Printf("proto: textWriter unindented too far")
return
}
w.ind--
}
func writeName(w *textWriter, props *Properties) error {
if _, err := w.WriteString(props.OrigName); err != nil {
return err
}
if props.Wire != "group" {
return w.WriteByte(':')
}
return nil
}
var (
messageSetType = reflect.TypeOf((*MessageSet)(nil)).Elem()
)
// raw is the interface satisfied by RawMessage.
type raw interface {
Bytes() []byte
}
func writeStruct(w *textWriter, sv reflect.Value) error {
if sv.Type() == messageSetType {
return writeMessageSet(w, sv.Addr().Interface().(*MessageSet))
}
st := sv.Type()
sprops := GetProperties(st)
for i := 0; i < sv.NumField(); i++ {
fv := sv.Field(i)
props := sprops.Prop[i]
name := st.Field(i).Name
if strings.HasPrefix(name, "XXX_") {
// There are two XXX_ fields:
// XXX_unrecognized []byte
// XXX_extensions map[int32]proto.Extension
// The first is handled here;
// the second is handled at the bottom of this function.
if name == "XXX_unrecognized" && !fv.IsNil() {
if err := writeUnknownStruct(w, fv.Interface().([]byte)); err != nil {
return err
}
}
continue
}
if fv.Kind() == reflect.Ptr && fv.IsNil() {
// Field not filled in. This could be an optional field or
// a required field that wasn't filled in. Either way, there
// isn't anything we can show for it.
continue
}
if fv.Kind() == reflect.Slice && fv.IsNil() {
// Repeated field that is empty, or a bytes field that is unused.
continue
}
if props.Repeated && fv.Kind() == reflect.Slice {
// Repeated field.
for j := 0; j < fv.Len(); j++ {
if err := writeName(w, props); err != nil {
return err
}
if !w.compact {
if err := w.WriteByte(' '); err != nil {
return err
}
}
v := fv.Index(j)
if v.Kind() == reflect.Ptr && v.IsNil() {
// A nil message in a repeated field is not valid,
// but we can handle that more gracefully than panicking.
if _, err := w.Write([]byte("<nil>\n")); err != nil {
return err
}
continue
}
if err := writeAny(w, v, props); err != nil {
return err
}
if err := w.WriteByte('\n'); err != nil {
return err
}
}
continue
}
if err := writeName(w, props); err != nil {
return err
}
if !w.compact {
if err := w.WriteByte(' '); err != nil {
return err
}
}
if b, ok := fv.Interface().(raw); ok {
if err := writeRaw(w, b.Bytes()); err != nil {
return err
}
continue
}
// Enums have a String method, so writeAny will work fine.
if err := writeAny(w, fv, props); err != nil {
return err
}
if err := w.WriteByte('\n'); err != nil {
return err
}
}
// Extensions (the XXX_extensions field).
pv := sv.Addr()
if pv.Type().Implements(extendableProtoType) {
if err := writeExtensions(w, pv); err != nil {
return err
}
}
return nil
}
// writeRaw writes an uninterpreted raw message.
func writeRaw(w *textWriter, b []byte) error {
if err := w.WriteByte('<'); err != nil {
return err
}
if !w.compact {
if err := w.WriteByte('\n'); err != nil {
return err
}
}
w.indent()
if err := writeUnknownStruct(w, b); err != nil {
return err
}
w.unindent()
if err := w.WriteByte('>'); err != nil {
return err
}
return nil
}
// writeAny writes an arbitrary field.
func writeAny(w *textWriter, v reflect.Value, props *Properties) error {
v = reflect.Indirect(v)
// Floats have special cases.
if v.Kind() == reflect.Float32 || v.Kind() == reflect.Float64 {
x := v.Float()
var b []byte
switch {
case math.IsInf(x, 1):
b = posInf
case math.IsInf(x, -1):
b = negInf
case math.IsNaN(x):
b = nan
}
if b != nil {
_, err := w.Write(b)
return err
}
// Other values are handled below.
}
// We don't attempt to serialise every possible value type; only those
// that can occur in protocol buffers.
switch v.Kind() {
case reflect.Slice:
// Should only be a []byte; repeated fields are handled in writeStruct.
if err := writeString(w, string(v.Interface().([]byte))); err != nil {
return err
}
case reflect.String:
if err := writeString(w, v.String()); err != nil {
return err
}
case reflect.Struct:
// Required/optional group/message.
var bra, ket byte = '<', '>'
if props != nil && props.Wire == "group" {
bra, ket = '{', '}'
}
if err := w.WriteByte(bra); err != nil {
return err
}
if !w.compact {
if err := w.WriteByte('\n'); err != nil {
return err
}
}
w.indent()
if tm, ok := v.Interface().(encoding.TextMarshaler); ok {
text, err := tm.MarshalText()
if err != nil {
return err
}
if _, err = w.Write(text); err != nil {
return err
}
} else if err := writeStruct(w, v); err != nil {
return err
}
w.unindent()
if err := w.WriteByte(ket); err != nil {
return err
}
default:
_, err := fmt.Fprint(w, v.Interface())
return err
}
return nil
}
// equivalent to C's isprint.
func isprint(c byte) bool {
return c >= 0x20 && c < 0x7f
}
// writeString writes a string in the protocol buffer text format.
// It is similar to strconv.Quote except we don't use Go escape sequences,
// we treat the string as a byte sequence, and we use octal escapes.
// These differences are to maintain interoperability with the other
// languages' implementations of the text format.
func writeString(w *textWriter, s string) error {
// use WriteByte here to get any needed indent
if err := w.WriteByte('"'); err != nil {
return err
}
// Loop over the bytes, not the runes.
for i := 0; i < len(s); i++ {
var err error
// Divergence from C++: we don't escape apostrophes.
// There's no need to escape them, and the C++ parser
// copes with a naked apostrophe.
switch c := s[i]; c {
case '\n':
_, err = w.w.Write(backslashN)
case '\r':
_, err = w.w.Write(backslashR)
case '\t':
_, err = w.w.Write(backslashT)
case '"':
_, err = w.w.Write(backslashDQ)
case '\\':
_, err = w.w.Write(backslashBS)
default:
if isprint(c) {
err = w.w.WriteByte(c)
} else {
_, err = fmt.Fprintf(w.w, "\\%03o", c)
}
}
if err != nil {
return err
}
}
return w.WriteByte('"')
}
func writeMessageSet(w *textWriter, ms *MessageSet) error {
for _, item := range ms.Item {
id := *item.TypeId
if msd, ok := messageSetMap[id]; ok {
// Known message set type.
if _, err := fmt.Fprintf(w, "[%s]: <\n", msd.name); err != nil {
return err
}
w.indent()
pb := reflect.New(msd.t.Elem())
if err := Unmarshal(item.Message, pb.Interface().(Message)); err != nil {
if _, err := fmt.Fprintf(w, "/* bad message: %v */\n", err); err != nil {
return err
}
} else {
if err := writeStruct(w, pb.Elem()); err != nil {
return err
}
}
} else {
// Unknown type.
if _, err := fmt.Fprintf(w, "[%d]: <\n", id); err != nil {
return err
}
w.indent()
if err := writeUnknownStruct(w, item.Message); err != nil {
return err
}
}
w.unindent()
if _, err := w.Write(gtNewline); err != nil {
return err
}
}
return nil
}
func writeUnknownStruct(w *textWriter, data []byte) (err error) {
if !w.compact {
if _, err := fmt.Fprintf(w, "/* %d unknown bytes */\n", len(data)); err != nil {
return err
}
}
b := NewBuffer(data)
for b.index < len(b.buf) {
x, err := b.DecodeVarint()
if err != nil {
_, err := fmt.Fprintf(w, "/* %v */\n", err)
return err
}
wire, tag := x&7, x>>3
if wire == WireEndGroup {
w.unindent()
if _, err := w.Write(endBraceNewline); err != nil {
return err
}
continue
}
if _, err := fmt.Fprint(w, tag); err != nil {
return err
}
if wire != WireStartGroup {
if err := w.WriteByte(':'); err != nil {
return err
}
}
if !w.compact || wire == WireStartGroup {
if err := w.WriteByte(' '); err != nil {
return err
}
}
switch wire {
case WireBytes:
buf, e := b.DecodeRawBytes(false)
if e == nil {
_, err = fmt.Fprintf(w, "%q", buf)
} else {
_, err = fmt.Fprintf(w, "/* %v */", e)
}
case WireFixed32:
x, err = b.DecodeFixed32()
err = writeUnknownInt(w, x, err)
case WireFixed64:
x, err = b.DecodeFixed64()
err = writeUnknownInt(w, x, err)
case WireStartGroup:
err = w.WriteByte('{')
w.indent()
case WireVarint:
x, err = b.DecodeVarint()
err = writeUnknownInt(w, x, err)
default:
_, err = fmt.Fprintf(w, "/* unknown wire type %d */", wire)
}
if err != nil {
return err
}
if err = w.WriteByte('\n'); err != nil {
return err
}
}
return nil
}
func writeUnknownInt(w *textWriter, x uint64, err error) error {
if err == nil {
_, err = fmt.Fprint(w, x)
} else {
_, err = fmt.Fprintf(w, "/* %v */", err)
}
return err
}
type int32Slice []int32
func (s int32Slice) Len() int { return len(s) }
func (s int32Slice) Less(i, j int) bool { return s[i] < s[j] }
func (s int32Slice) Swap(i, j int) { s[i], s[j] = s[j], s[i] }
// writeExtensions writes all the extensions in pv.
// pv is assumed to be a pointer to a protocol message struct that is extendable.
func writeExtensions(w *textWriter, pv reflect.Value) error {
emap := extensionMaps[pv.Type().Elem()]
ep := pv.Interface().(extendableProto)
// Order the extensions by ID.
// This isn't strictly necessary, but it will give us
// canonical output, which will also make testing easier.
m := ep.ExtensionMap()
ids := make([]int32, 0, len(m))
for id := range m {
ids = append(ids, id)
}
sort.Sort(int32Slice(ids))
for _, extNum := range ids {
ext := m[extNum]
var desc *ExtensionDesc
if emap != nil {
desc = emap[extNum]
}
if desc == nil {
// Unknown extension.
if err := writeUnknownStruct(w, ext.enc); err != nil {
return err
}
continue
}
pb, err := GetExtension(ep, desc)
if err != nil {
if _, err := fmt.Fprintln(os.Stderr, "proto: failed getting extension: ", err); err != nil {
return err
}
continue
}
// Repeated extensions will appear as a slice.
if !desc.repeated() {
if err := writeExtension(w, desc.Name, pb); err != nil {
return err
}
} else {
v := reflect.ValueOf(pb)
for i := 0; i < v.Len(); i++ {
if err := writeExtension(w, desc.Name, v.Index(i).Interface()); err != nil {
return err
}
}
}
}
return nil
}
func writeExtension(w *textWriter, name string, pb interface{}) error {
if _, err := fmt.Fprintf(w, "[%s]:", name); err != nil {
return err
}
if !w.compact {
if err := w.WriteByte(' '); err != nil {
return err
}
}
if err := writeAny(w, reflect.ValueOf(pb), nil); err != nil {
return err
}
if err := w.WriteByte('\n'); err != nil {
return err
}
return nil
}
func (w *textWriter) writeIndent() {
if !w.complete {
return
}
remain := w.ind * 2
for remain > 0 {
n := remain
if n > len(spaces) {
n = len(spaces)
}
w.w.Write(spaces[:n])
remain -= n
}
w.complete = false
}
func marshalText(w io.Writer, pb Message, compact bool) error {
val := reflect.ValueOf(pb)
if pb == nil || val.IsNil() {
w.Write([]byte("<nil>"))
return nil
}
var bw *bufio.Writer
ww, ok := w.(writer)
if !ok {
bw = bufio.NewWriter(w)
ww = bw
}
aw := &textWriter{
w: ww,
complete: true,
compact: compact,
}
if tm, ok := pb.(encoding.TextMarshaler); ok {
text, err := tm.MarshalText()
if err != nil {
return err
}
if _, err = aw.Write(text); err != nil {
return err
}
if bw != nil {
return bw.Flush()
}
return nil
}
// Dereference the received pointer so we don't have outer < and >.
v := reflect.Indirect(val)
if err := writeStruct(aw, v); err != nil {
return err
}
if bw != nil {
return bw.Flush()
}
return nil
}
// MarshalText writes a given protocol buffer in text format.
// The only errors returned are from w.
func MarshalText(w io.Writer, pb Message) error {
return marshalText(w, pb, false)
}
// MarshalTextString is the same as MarshalText, but returns the string directly.
func MarshalTextString(pb Message) string {
var buf bytes.Buffer
marshalText(&buf, pb, false)
return buf.String()
}
// CompactText writes a given protocol buffer in compact text format (one line).
func CompactText(w io.Writer, pb Message) error { return marshalText(w, pb, true) }
// CompactTextString is the same as CompactText, but returns the string directly.
func CompactTextString(pb Message) string {
var buf bytes.Buffer
marshalText(&buf, pb, true)
return buf.String()
}

View File

@@ -1,687 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2010 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
package proto
// Functions for parsing the Text protocol buffer format.
// TODO: message sets.
import (
"encoding"
"errors"
"fmt"
"reflect"
"strconv"
"strings"
"unicode/utf8"
)
type ParseError struct {
Message string
Line int // 1-based line number
Offset int // 0-based byte offset from start of input
}
func (p *ParseError) Error() string {
if p.Line == 1 {
// show offset only for first line
return fmt.Sprintf("line 1.%d: %v", p.Offset, p.Message)
}
return fmt.Sprintf("line %d: %v", p.Line, p.Message)
}
type token struct {
value string
err *ParseError
line int // line number
offset int // byte number from start of input, not start of line
unquoted string // the unquoted version of value, if it was a quoted string
}
func (t *token) String() string {
if t.err == nil {
return fmt.Sprintf("%q (line=%d, offset=%d)", t.value, t.line, t.offset)
}
return fmt.Sprintf("parse error: %v", t.err)
}
type textParser struct {
s string // remaining input
done bool // whether the parsing is finished (success or error)
backed bool // whether back() was called
offset, line int
cur token
}
func newTextParser(s string) *textParser {
p := new(textParser)
p.s = s
p.line = 1
p.cur.line = 1
return p
}
func (p *textParser) errorf(format string, a ...interface{}) *ParseError {
pe := &ParseError{fmt.Sprintf(format, a...), p.cur.line, p.cur.offset}
p.cur.err = pe
p.done = true
return pe
}
// Numbers and identifiers are matched by [-+._A-Za-z0-9]
func isIdentOrNumberChar(c byte) bool {
switch {
case 'A' <= c && c <= 'Z', 'a' <= c && c <= 'z':
return true
case '0' <= c && c <= '9':
return true
}
switch c {
case '-', '+', '.', '_':
return true
}
return false
}
func isWhitespace(c byte) bool {
switch c {
case ' ', '\t', '\n', '\r':
return true
}
return false
}
func (p *textParser) skipWhitespace() {
i := 0
for i < len(p.s) && (isWhitespace(p.s[i]) || p.s[i] == '#') {
if p.s[i] == '#' {
// comment; skip to end of line or input
for i < len(p.s) && p.s[i] != '\n' {
i++
}
if i == len(p.s) {
break
}
}
if p.s[i] == '\n' {
p.line++
}
i++
}
p.offset += i
p.s = p.s[i:len(p.s)]
if len(p.s) == 0 {
p.done = true
}
}
func (p *textParser) advance() {
// Skip whitespace
p.skipWhitespace()
if p.done {
return
}
// Start of non-whitespace
p.cur.err = nil
p.cur.offset, p.cur.line = p.offset, p.line
p.cur.unquoted = ""
switch p.s[0] {
case '<', '>', '{', '}', ':', '[', ']', ';', ',':
// Single symbol
p.cur.value, p.s = p.s[0:1], p.s[1:len(p.s)]
case '"', '\'':
// Quoted string
i := 1
for i < len(p.s) && p.s[i] != p.s[0] && p.s[i] != '\n' {
if p.s[i] == '\\' && i+1 < len(p.s) {
// skip escaped char
i++
}
i++
}
if i >= len(p.s) || p.s[i] != p.s[0] {
p.errorf("unmatched quote")
return
}
unq, err := unquoteC(p.s[1:i], rune(p.s[0]))
if err != nil {
p.errorf("invalid quoted string %v", p.s[0:i+1])
return
}
p.cur.value, p.s = p.s[0:i+1], p.s[i+1:len(p.s)]
p.cur.unquoted = unq
default:
i := 0
for i < len(p.s) && isIdentOrNumberChar(p.s[i]) {
i++
}
if i == 0 {
p.errorf("unexpected byte %#x", p.s[0])
return
}
p.cur.value, p.s = p.s[0:i], p.s[i:len(p.s)]
}
p.offset += len(p.cur.value)
}
var (
errBadUTF8 = errors.New("proto: bad UTF-8")
errBadHex = errors.New("proto: bad hexadecimal")
)
func unquoteC(s string, quote rune) (string, error) {
// This is based on C++'s tokenizer.cc.
// Despite its name, this is *not* parsing C syntax.
// For instance, "\0" is an invalid quoted string.
// Avoid allocation in trivial cases.
simple := true
for _, r := range s {
if r == '\\' || r == quote {
simple = false
break
}
}
if simple {
return s, nil
}
buf := make([]byte, 0, 3*len(s)/2)
for len(s) > 0 {
r, n := utf8.DecodeRuneInString(s)
if r == utf8.RuneError && n == 1 {
return "", errBadUTF8
}
s = s[n:]
if r != '\\' {
if r < utf8.RuneSelf {
buf = append(buf, byte(r))
} else {
buf = append(buf, string(r)...)
}
continue
}
ch, tail, err := unescape(s)
if err != nil {
return "", err
}
buf = append(buf, ch...)
s = tail
}
return string(buf), nil
}
func unescape(s string) (ch string, tail string, err error) {
r, n := utf8.DecodeRuneInString(s)
if r == utf8.RuneError && n == 1 {
return "", "", errBadUTF8
}
s = s[n:]
switch r {
case 'a':
return "\a", s, nil
case 'b':
return "\b", s, nil
case 'f':
return "\f", s, nil
case 'n':
return "\n", s, nil
case 'r':
return "\r", s, nil
case 't':
return "\t", s, nil
case 'v':
return "\v", s, nil
case '?':
return "?", s, nil // trigraph workaround
case '\'', '"', '\\':
return string(r), s, nil
case '0', '1', '2', '3', '4', '5', '6', '7', 'x', 'X':
if len(s) < 2 {
return "", "", fmt.Errorf(`\%c requires 2 following digits`, r)
}
base := 8
ss := s[:2]
s = s[2:]
if r == 'x' || r == 'X' {
base = 16
} else {
ss = string(r) + ss
}
i, err := strconv.ParseUint(ss, base, 8)
if err != nil {
return "", "", err
}
return string([]byte{byte(i)}), s, nil
case 'u', 'U':
n := 4
if r == 'U' {
n = 8
}
if len(s) < n {
return "", "", fmt.Errorf(`\%c requires %d digits`, r, n)
}
bs := make([]byte, n/2)
for i := 0; i < n; i += 2 {
a, ok1 := unhex(s[i])
b, ok2 := unhex(s[i+1])
if !ok1 || !ok2 {
return "", "", errBadHex
}
bs[i/2] = a<<4 | b
}
s = s[n:]
return string(bs), s, nil
}
return "", "", fmt.Errorf(`unknown escape \%c`, r)
}
// Adapted from src/pkg/strconv/quote.go.
func unhex(b byte) (v byte, ok bool) {
switch {
case '0' <= b && b <= '9':
return b - '0', true
case 'a' <= b && b <= 'f':
return b - 'a' + 10, true
case 'A' <= b && b <= 'F':
return b - 'A' + 10, true
}
return 0, false
}
// Back off the parser by one token. Can only be done between calls to next().
// It makes the next advance() a no-op.
func (p *textParser) back() { p.backed = true }
// Advances the parser and returns the new current token.
func (p *textParser) next() *token {
if p.backed || p.done {
p.backed = false
return &p.cur
}
p.advance()
if p.done {
p.cur.value = ""
} else if len(p.cur.value) > 0 && p.cur.value[0] == '"' {
// Look for multiple quoted strings separated by whitespace,
// and concatenate them.
cat := p.cur
for {
p.skipWhitespace()
if p.done || p.s[0] != '"' {
break
}
p.advance()
if p.cur.err != nil {
return &p.cur
}
cat.value += " " + p.cur.value
cat.unquoted += p.cur.unquoted
}
p.done = false // parser may have seen EOF, but we want to return cat
p.cur = cat
}
return &p.cur
}
// Return a RequiredNotSetError indicating which required field was not set.
func (p *textParser) missingRequiredFieldError(sv reflect.Value) *RequiredNotSetError {
st := sv.Type()
sprops := GetProperties(st)
for i := 0; i < st.NumField(); i++ {
if !isNil(sv.Field(i)) {
continue
}
props := sprops.Prop[i]
if props.Required {
return &RequiredNotSetError{fmt.Sprintf("%v.%v", st, props.OrigName)}
}
}
return &RequiredNotSetError{fmt.Sprintf("%v.<unknown field name>", st)} // should not happen
}
// Returns the index in the struct for the named field, as well as the parsed tag properties.
func structFieldByName(st reflect.Type, name string) (int, *Properties, bool) {
sprops := GetProperties(st)
i, ok := sprops.decoderOrigNames[name]
if ok {
return i, sprops.Prop[i], true
}
return -1, nil, false
}
// Consume a ':' from the input stream (if the next token is a colon),
// returning an error if a colon is needed but not present.
func (p *textParser) checkForColon(props *Properties, typ reflect.Type) *ParseError {
tok := p.next()
if tok.err != nil {
return tok.err
}
if tok.value != ":" {
// Colon is optional when the field is a group or message.
needColon := true
switch props.Wire {
case "group":
needColon = false
case "bytes":
// A "bytes" field is either a message, a string, or a repeated field;
// those three become *T, *string and []T respectively, so we can check for
// this field being a pointer to a non-string.
if typ.Kind() == reflect.Ptr {
// *T or *string
if typ.Elem().Kind() == reflect.String {
break
}
} else if typ.Kind() == reflect.Slice {
// []T or []*T
if typ.Elem().Kind() != reflect.Ptr {
break
}
}
needColon = false
}
if needColon {
return p.errorf("expected ':', found %q", tok.value)
}
p.back()
}
return nil
}
func (p *textParser) readStruct(sv reflect.Value, terminator string) error {
st := sv.Type()
reqCount := GetProperties(st).reqCount
var reqFieldErr error
fieldSet := make(map[string]bool)
// A struct is a sequence of "name: value", terminated by one of
// '>' or '}', or the end of the input. A name may also be
// "[extension]".
for {
tok := p.next()
if tok.err != nil {
return tok.err
}
if tok.value == terminator {
break
}
if tok.value == "[" {
// Looks like an extension.
//
// TODO: Check whether we need to handle
// namespace rooted names (e.g. ".something.Foo").
tok = p.next()
if tok.err != nil {
return tok.err
}
var desc *ExtensionDesc
// This could be faster, but it's functional.
// TODO: Do something smarter than a linear scan.
for _, d := range RegisteredExtensions(reflect.New(st).Interface().(Message)) {
if d.Name == tok.value {
desc = d
break
}
}
if desc == nil {
return p.errorf("unrecognized extension %q", tok.value)
}
// Check the extension terminator.
tok = p.next()
if tok.err != nil {
return tok.err
}
if tok.value != "]" {
return p.errorf("unrecognized extension terminator %q", tok.value)
}
props := &Properties{}
props.Parse(desc.Tag)
typ := reflect.TypeOf(desc.ExtensionType)
if err := p.checkForColon(props, typ); err != nil {
return err
}
rep := desc.repeated()
// Read the extension structure, and set it in
// the value we're constructing.
var ext reflect.Value
if !rep {
ext = reflect.New(typ).Elem()
} else {
ext = reflect.New(typ.Elem()).Elem()
}
if err := p.readAny(ext, props); err != nil {
if _, ok := err.(*RequiredNotSetError); !ok {
return err
}
reqFieldErr = err
}
ep := sv.Addr().Interface().(extendableProto)
if !rep {
SetExtension(ep, desc, ext.Interface())
} else {
old, err := GetExtension(ep, desc)
var sl reflect.Value
if err == nil {
sl = reflect.ValueOf(old) // existing slice
} else {
sl = reflect.MakeSlice(typ, 0, 1)
}
sl = reflect.Append(sl, ext)
SetExtension(ep, desc, sl.Interface())
}
} else {
// This is a normal, non-extension field.
name := tok.value
fi, props, ok := structFieldByName(st, name)
if !ok {
return p.errorf("unknown field name %q in %v", name, st)
}
dst := sv.Field(fi)
// Check that it's not already set if it's not a repeated field.
if !props.Repeated && fieldSet[name] {
return p.errorf("non-repeated field %q was repeated", name)
}
if err := p.checkForColon(props, st.Field(fi).Type); err != nil {
return err
}
// Parse into the field.
fieldSet[name] = true
if err := p.readAny(dst, props); err != nil {
if _, ok := err.(*RequiredNotSetError); !ok {
return err
}
reqFieldErr = err
} else if props.Required {
reqCount--
}
}
// For backward compatibility, permit a semicolon or comma after a field.
tok = p.next()
if tok.err != nil {
return tok.err
}
if tok.value != ";" && tok.value != "," {
p.back()
}
}
if reqCount > 0 {
return p.missingRequiredFieldError(sv)
}
return reqFieldErr
}
func (p *textParser) readAny(v reflect.Value, props *Properties) error {
tok := p.next()
if tok.err != nil {
return tok.err
}
if tok.value == "" {
return p.errorf("unexpected EOF")
}
switch fv := v; fv.Kind() {
case reflect.Slice:
at := v.Type()
if at.Elem().Kind() == reflect.Uint8 {
// Special case for []byte
if tok.value[0] != '"' && tok.value[0] != '\'' {
// Deliberately written out here, as the error after
// this switch statement would write "invalid []byte: ...",
// which is not as user-friendly.
return p.errorf("invalid string: %v", tok.value)
}
bytes := []byte(tok.unquoted)
fv.Set(reflect.ValueOf(bytes))
return nil
}
// Repeated field. May already exist.
flen := fv.Len()
if flen == fv.Cap() {
nav := reflect.MakeSlice(at, flen, 2*flen+1)
reflect.Copy(nav, fv)
fv.Set(nav)
}
fv.SetLen(flen + 1)
// Read one.
p.back()
return p.readAny(fv.Index(flen), props)
case reflect.Bool:
// Either "true", "false", 1 or 0.
switch tok.value {
case "true", "1":
fv.SetBool(true)
return nil
case "false", "0":
fv.SetBool(false)
return nil
}
case reflect.Float32, reflect.Float64:
v := tok.value
// Ignore 'f' for compatibility with output generated by C++, but don't
// remove 'f' when the value is "-inf" or "inf".
if strings.HasSuffix(v, "f") && tok.value != "-inf" && tok.value != "inf" {
v = v[:len(v)-1]
}
if f, err := strconv.ParseFloat(v, fv.Type().Bits()); err == nil {
fv.SetFloat(f)
return nil
}
case reflect.Int32:
if x, err := strconv.ParseInt(tok.value, 0, 32); err == nil {
fv.SetInt(x)
return nil
}
if len(props.Enum) == 0 {
break
}
m, ok := enumValueMaps[props.Enum]
if !ok {
break
}
x, ok := m[tok.value]
if !ok {
break
}
fv.SetInt(int64(x))
return nil
case reflect.Int64:
if x, err := strconv.ParseInt(tok.value, 0, 64); err == nil {
fv.SetInt(x)
return nil
}
case reflect.Ptr:
// A basic field (indirected through pointer), or a repeated message/group
p.back()
fv.Set(reflect.New(fv.Type().Elem()))
return p.readAny(fv.Elem(), props)
case reflect.String:
if tok.value[0] == '"' || tok.value[0] == '\'' {
fv.SetString(tok.unquoted)
return nil
}
case reflect.Struct:
var terminator string
switch tok.value {
case "{":
terminator = "}"
case "<":
terminator = ">"
default:
return p.errorf("expected '{' or '<', found %q", tok.value)
}
// TODO: Handle nested messages which implement encoding.TextUnmarshaler.
return p.readStruct(fv, terminator)
case reflect.Uint32:
if x, err := strconv.ParseUint(tok.value, 0, 32); err == nil {
fv.SetUint(uint64(x))
return nil
}
case reflect.Uint64:
if x, err := strconv.ParseUint(tok.value, 0, 64); err == nil {
fv.SetUint(x)
return nil
}
}
return p.errorf("invalid %v: %v", v.Type(), tok.value)
}
// UnmarshalText reads a protocol buffer in Text format. UnmarshalText resets pb
// before starting to unmarshal, so any existing data in pb is always removed.
// If a required field is not set and no other error occurs,
// UnmarshalText returns *RequiredNotSetError.
func UnmarshalText(s string, pb Message) error {
if um, ok := pb.(encoding.TextUnmarshaler); ok {
err := um.UnmarshalText([]byte(s))
return err
}
pb.Reset()
v := reflect.ValueOf(pb)
if pe := newTextParser(s).readStruct(v.Elem(), ""); pe != nil {
return pe
}
return nil
}

View File

@@ -1,468 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2010 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
package proto_test
import (
"math"
"reflect"
"testing"
. "./testdata"
. "github.com/coreos/etcd/Godeps/_workspace/src/code.google.com/p/goprotobuf/proto"
)
type UnmarshalTextTest struct {
in string
err string // if "", no error expected
out *MyMessage
}
func buildExtStructTest(text string) UnmarshalTextTest {
msg := &MyMessage{
Count: Int32(42),
}
SetExtension(msg, E_Ext_More, &Ext{
Data: String("Hello, world!"),
})
return UnmarshalTextTest{in: text, out: msg}
}
func buildExtDataTest(text string) UnmarshalTextTest {
msg := &MyMessage{
Count: Int32(42),
}
SetExtension(msg, E_Ext_Text, String("Hello, world!"))
SetExtension(msg, E_Ext_Number, Int32(1729))
return UnmarshalTextTest{in: text, out: msg}
}
func buildExtRepStringTest(text string) UnmarshalTextTest {
msg := &MyMessage{
Count: Int32(42),
}
if err := SetExtension(msg, E_Greeting, []string{"bula", "hola"}); err != nil {
panic(err)
}
return UnmarshalTextTest{in: text, out: msg}
}
var unMarshalTextTests = []UnmarshalTextTest{
// Basic
{
in: " count:42\n name:\"Dave\" ",
out: &MyMessage{
Count: Int32(42),
Name: String("Dave"),
},
},
// Empty quoted string
{
in: `count:42 name:""`,
out: &MyMessage{
Count: Int32(42),
Name: String(""),
},
},
// Quoted string concatenation
{
in: `count:42 name: "My name is "` + "\n" + `"elsewhere"`,
out: &MyMessage{
Count: Int32(42),
Name: String("My name is elsewhere"),
},
},
// Quoted string with escaped apostrophe
{
in: `count:42 name: "HOLIDAY - New Year\'s Day"`,
out: &MyMessage{
Count: Int32(42),
Name: String("HOLIDAY - New Year's Day"),
},
},
// Quoted string with single quote
{
in: `count:42 name: 'Roger "The Ramster" Ramjet'`,
out: &MyMessage{
Count: Int32(42),
Name: String(`Roger "The Ramster" Ramjet`),
},
},
// Quoted string with all the accepted special characters from the C++ test
{
in: `count:42 name: ` + "\"\\\"A string with \\' characters \\n and \\r newlines and \\t tabs and \\001 slashes \\\\ and multiple spaces\"",
out: &MyMessage{
Count: Int32(42),
Name: String("\"A string with ' characters \n and \r newlines and \t tabs and \001 slashes \\ and multiple spaces"),
},
},
// Quoted string with quoted backslash
{
in: `count:42 name: "\\'xyz"`,
out: &MyMessage{
Count: Int32(42),
Name: String(`\'xyz`),
},
},
// Quoted string with UTF-8 bytes.
{
in: "count:42 name: '\303\277\302\201\xAB'",
out: &MyMessage{
Count: Int32(42),
Name: String("\303\277\302\201\xAB"),
},
},
// Bad quoted string
{
in: `inner: < host: "\0" >` + "\n",
err: `line 1.15: invalid quoted string "\0"`,
},
// Number too large for int64
{
in: "count: 1 others { key: 123456789012345678901 }",
err: "line 1.23: invalid int64: 123456789012345678901",
},
// Number too large for int32
{
in: "count: 1234567890123",
err: "line 1.7: invalid int32: 1234567890123",
},
// Number in hexadecimal
{
in: "count: 0x2beef",
out: &MyMessage{
Count: Int32(0x2beef),
},
},
// Number in octal
{
in: "count: 024601",
out: &MyMessage{
Count: Int32(024601),
},
},
// Floating point number with "f" suffix
{
in: "count: 4 others:< weight: 17.0f >",
out: &MyMessage{
Count: Int32(4),
Others: []*OtherMessage{
{
Weight: Float32(17),
},
},
},
},
// Floating point positive infinity
{
in: "count: 4 bigfloat: inf",
out: &MyMessage{
Count: Int32(4),
Bigfloat: Float64(math.Inf(1)),
},
},
// Floating point negative infinity
{
in: "count: 4 bigfloat: -inf",
out: &MyMessage{
Count: Int32(4),
Bigfloat: Float64(math.Inf(-1)),
},
},
// Number too large for float32
{
in: "others:< weight: 12345678901234567890123456789012345678901234567890 >",
err: "line 1.17: invalid float32: 12345678901234567890123456789012345678901234567890",
},
// Number posing as a quoted string
{
in: `inner: < host: 12 >` + "\n",
err: `line 1.15: invalid string: 12`,
},
// Quoted string posing as int32
{
in: `count: "12"`,
err: `line 1.7: invalid int32: "12"`,
},
// Quoted string posing a float32
{
in: `others:< weight: "17.4" >`,
err: `line 1.17: invalid float32: "17.4"`,
},
// Enum
{
in: `count:42 bikeshed: BLUE`,
out: &MyMessage{
Count: Int32(42),
Bikeshed: MyMessage_BLUE.Enum(),
},
},
// Repeated field
{
in: `count:42 pet: "horsey" pet:"bunny"`,
out: &MyMessage{
Count: Int32(42),
Pet: []string{"horsey", "bunny"},
},
},
// Repeated message with/without colon and <>/{}
{
in: `count:42 others:{} others{} others:<> others:{}`,
out: &MyMessage{
Count: Int32(42),
Others: []*OtherMessage{
{},
{},
{},
{},
},
},
},
// Missing colon for inner message
{
in: `count:42 inner < host: "cauchy.syd" >`,
out: &MyMessage{
Count: Int32(42),
Inner: &InnerMessage{
Host: String("cauchy.syd"),
},
},
},
// Missing colon for string field
{
in: `name "Dave"`,
err: `line 1.5: expected ':', found "\"Dave\""`,
},
// Missing colon for int32 field
{
in: `count 42`,
err: `line 1.6: expected ':', found "42"`,
},
// Missing required field
{
in: `name: "Pawel"`,
err: `proto: required field "testdata.MyMessage.count" not set`,
out: &MyMessage{
Name: String("Pawel"),
},
},
// Repeated non-repeated field
{
in: `name: "Rob" name: "Russ"`,
err: `line 1.12: non-repeated field "name" was repeated`,
},
// Group
{
in: `count: 17 SomeGroup { group_field: 12 }`,
out: &MyMessage{
Count: Int32(17),
Somegroup: &MyMessage_SomeGroup{
GroupField: Int32(12),
},
},
},
// Semicolon between fields
{
in: `count:3;name:"Calvin"`,
out: &MyMessage{
Count: Int32(3),
Name: String("Calvin"),
},
},
// Comma between fields
{
in: `count:4,name:"Ezekiel"`,
out: &MyMessage{
Count: Int32(4),
Name: String("Ezekiel"),
},
},
// Extension
buildExtStructTest(`count: 42 [testdata.Ext.more]:<data:"Hello, world!" >`),
buildExtStructTest(`count: 42 [testdata.Ext.more] {data:"Hello, world!"}`),
buildExtDataTest(`count: 42 [testdata.Ext.text]:"Hello, world!" [testdata.Ext.number]:1729`),
buildExtRepStringTest(`count: 42 [testdata.greeting]:"bula" [testdata.greeting]:"hola"`),
// Big all-in-one
{
in: "count:42 # Meaning\n" +
`name:"Dave" ` +
`quote:"\"I didn't want to go.\"" ` +
`pet:"bunny" ` +
`pet:"kitty" ` +
`pet:"horsey" ` +
`inner:<` +
` host:"footrest.syd" ` +
` port:7001 ` +
` connected:true ` +
`> ` +
`others:<` +
` key:3735928559 ` +
` value:"\x01A\a\f" ` +
`> ` +
`others:<` +
" weight:58.9 # Atomic weight of Co\n" +
` inner:<` +
` host:"lesha.mtv" ` +
` port:8002 ` +
` >` +
`>`,
out: &MyMessage{
Count: Int32(42),
Name: String("Dave"),
Quote: String(`"I didn't want to go."`),
Pet: []string{"bunny", "kitty", "horsey"},
Inner: &InnerMessage{
Host: String("footrest.syd"),
Port: Int32(7001),
Connected: Bool(true),
},
Others: []*OtherMessage{
{
Key: Int64(3735928559),
Value: []byte{0x1, 'A', '\a', '\f'},
},
{
Weight: Float32(58.9),
Inner: &InnerMessage{
Host: String("lesha.mtv"),
Port: Int32(8002),
},
},
},
},
},
}
func TestUnmarshalText(t *testing.T) {
for i, test := range unMarshalTextTests {
pb := new(MyMessage)
err := UnmarshalText(test.in, pb)
if test.err == "" {
// We don't expect failure.
if err != nil {
t.Errorf("Test %d: Unexpected error: %v", i, err)
} else if !reflect.DeepEqual(pb, test.out) {
t.Errorf("Test %d: Incorrect populated \nHave: %v\nWant: %v",
i, pb, test.out)
}
} else {
// We do expect failure.
if err == nil {
t.Errorf("Test %d: Didn't get expected error: %v", i, test.err)
} else if err.Error() != test.err {
t.Errorf("Test %d: Incorrect error.\nHave: %v\nWant: %v",
i, err.Error(), test.err)
} else if _, ok := err.(*RequiredNotSetError); ok && test.out != nil && !reflect.DeepEqual(pb, test.out) {
t.Errorf("Test %d: Incorrect populated \nHave: %v\nWant: %v",
i, pb, test.out)
}
}
}
}
func TestUnmarshalTextCustomMessage(t *testing.T) {
msg := &textMessage{}
if err := UnmarshalText("custom", msg); err != nil {
t.Errorf("Unexpected error from custom unmarshal: %v", err)
}
if UnmarshalText("not custom", msg) == nil {
t.Errorf("Didn't get expected error from custom unmarshal")
}
}
// Regression test; this caused a panic.
func TestRepeatedEnum(t *testing.T) {
pb := new(RepeatedEnum)
if err := UnmarshalText("color: RED", pb); err != nil {
t.Fatal(err)
}
exp := &RepeatedEnum{
Color: []RepeatedEnum_Color{RepeatedEnum_RED},
}
if !Equal(pb, exp) {
t.Errorf("Incorrect populated \nHave: %v\nWant: %v", pb, exp)
}
}
var benchInput string
func init() {
benchInput = "count: 4\n"
for i := 0; i < 1000; i++ {
benchInput += "pet: \"fido\"\n"
}
// Check it is valid input.
pb := new(MyMessage)
err := UnmarshalText(benchInput, pb)
if err != nil {
panic("Bad benchmark input: " + err.Error())
}
}
func BenchmarkUnmarshalText(b *testing.B) {
pb := new(MyMessage)
for i := 0; i < b.N; i++ {
UnmarshalText(benchInput, pb)
}
b.SetBytes(int64(len(benchInput)))
}

View File

@@ -1,407 +0,0 @@
// Go support for Protocol Buffers - Google's data interchange format
//
// Copyright 2010 The Go Authors. All rights reserved.
// http://code.google.com/p/goprotobuf/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
package proto_test
import (
"bytes"
"errors"
"io/ioutil"
"math"
"strings"
"testing"
pb "./testdata"
"github.com/coreos/etcd/Godeps/_workspace/src/code.google.com/p/goprotobuf/proto"
)
// textMessage implements the methods that allow it to marshal and unmarshal
// itself as text.
type textMessage struct {
}
func (*textMessage) MarshalText() ([]byte, error) {
return []byte("custom"), nil
}
func (*textMessage) UnmarshalText(bytes []byte) error {
if string(bytes) != "custom" {
return errors.New("expected 'custom'")
}
return nil
}
func (*textMessage) Reset() {}
func (*textMessage) String() string { return "" }
func (*textMessage) ProtoMessage() {}
func newTestMessage() *pb.MyMessage {
msg := &pb.MyMessage{
Count: proto.Int32(42),
Name: proto.String("Dave"),
Quote: proto.String(`"I didn't want to go."`),
Pet: []string{"bunny", "kitty", "horsey"},
Inner: &pb.InnerMessage{
Host: proto.String("footrest.syd"),
Port: proto.Int32(7001),
Connected: proto.Bool(true),
},
Others: []*pb.OtherMessage{
{
Key: proto.Int64(0xdeadbeef),
Value: []byte{1, 65, 7, 12},
},
{
Weight: proto.Float32(6.022),
Inner: &pb.InnerMessage{
Host: proto.String("lesha.mtv"),
Port: proto.Int32(8002),
},
},
},
Bikeshed: pb.MyMessage_BLUE.Enum(),
Somegroup: &pb.MyMessage_SomeGroup{
GroupField: proto.Int32(8),
},
// One normally wouldn't do this.
// This is an undeclared tag 13, as a varint (wire type 0) with value 4.
XXX_unrecognized: []byte{13<<3 | 0, 4},
}
ext := &pb.Ext{
Data: proto.String("Big gobs for big rats"),
}
if err := proto.SetExtension(msg, pb.E_Ext_More, ext); err != nil {
panic(err)
}
greetings := []string{"adg", "easy", "cow"}
if err := proto.SetExtension(msg, pb.E_Greeting, greetings); err != nil {
panic(err)
}
// Add an unknown extension. We marshal a pb.Ext, and fake the ID.
b, err := proto.Marshal(&pb.Ext{Data: proto.String("3G skiing")})
if err != nil {
panic(err)
}
b = append(proto.EncodeVarint(201<<3|proto.WireBytes), b...)
proto.SetRawExtension(msg, 201, b)
// Extensions can be plain fields, too, so let's test that.
b = append(proto.EncodeVarint(202<<3|proto.WireVarint), 19)
proto.SetRawExtension(msg, 202, b)
return msg
}
const text = `count: 42
name: "Dave"
quote: "\"I didn't want to go.\""
pet: "bunny"
pet: "kitty"
pet: "horsey"
inner: <
host: "footrest.syd"
port: 7001
connected: true
>
others: <
key: 3735928559
value: "\001A\007\014"
>
others: <
weight: 6.022
inner: <
host: "lesha.mtv"
port: 8002
>
>
bikeshed: BLUE
SomeGroup {
group_field: 8
}
/* 2 unknown bytes */
13: 4
[testdata.Ext.more]: <
data: "Big gobs for big rats"
>
[testdata.greeting]: "adg"
[testdata.greeting]: "easy"
[testdata.greeting]: "cow"
/* 13 unknown bytes */
201: "\t3G skiing"
/* 3 unknown bytes */
202: 19
`
func TestMarshalText(t *testing.T) {
buf := new(bytes.Buffer)
if err := proto.MarshalText(buf, newTestMessage()); err != nil {
t.Fatalf("proto.MarshalText: %v", err)
}
s := buf.String()
if s != text {
t.Errorf("Got:\n===\n%v===\nExpected:\n===\n%v===\n", s, text)
}
}
func TestMarshalTextCustomMessage(t *testing.T) {
buf := new(bytes.Buffer)
if err := proto.MarshalText(buf, &textMessage{}); err != nil {
t.Fatalf("proto.MarshalText: %v", err)
}
s := buf.String()
if s != "custom" {
t.Errorf("Got %q, expected %q", s, "custom")
}
}
func TestMarshalTextNil(t *testing.T) {
want := "<nil>"
tests := []proto.Message{nil, (*pb.MyMessage)(nil)}
for i, test := range tests {
buf := new(bytes.Buffer)
if err := proto.MarshalText(buf, test); err != nil {
t.Fatal(err)
}
if got := buf.String(); got != want {
t.Errorf("%d: got %q want %q", i, got, want)
}
}
}
func TestMarshalTextUnknownEnum(t *testing.T) {
// The Color enum only specifies values 0-2.
m := &pb.MyMessage{Bikeshed: pb.MyMessage_Color(3).Enum()}
got := m.String()
const want = `bikeshed:3 `
if got != want {
t.Errorf("\n got %q\nwant %q", got, want)
}
}
func BenchmarkMarshalTextBuffered(b *testing.B) {
buf := new(bytes.Buffer)
m := newTestMessage()
for i := 0; i < b.N; i++ {
buf.Reset()
proto.MarshalText(buf, m)
}
}
func BenchmarkMarshalTextUnbuffered(b *testing.B) {
w := ioutil.Discard
m := newTestMessage()
for i := 0; i < b.N; i++ {
proto.MarshalText(w, m)
}
}
func compact(src string) string {
// s/[ \n]+/ /g; s/ $//;
dst := make([]byte, len(src))
space, comment := false, false
j := 0
for i := 0; i < len(src); i++ {
if strings.HasPrefix(src[i:], "/*") {
comment = true
i++
continue
}
if comment && strings.HasPrefix(src[i:], "*/") {
comment = false
i++
continue
}
if comment {
continue
}
c := src[i]
if c == ' ' || c == '\n' {
space = true
continue
}
if j > 0 && (dst[j-1] == ':' || dst[j-1] == '<' || dst[j-1] == '{') {
space = false
}
if c == '{' {
space = false
}
if space {
dst[j] = ' '
j++
space = false
}
dst[j] = c
j++
}
if space {
dst[j] = ' '
j++
}
return string(dst[0:j])
}
var compactText = compact(text)
func TestCompactText(t *testing.T) {
s := proto.CompactTextString(newTestMessage())
if s != compactText {
t.Errorf("Got:\n===\n%v===\nExpected:\n===\n%v\n===\n", s, compactText)
}
}
func TestStringEscaping(t *testing.T) {
testCases := []struct {
in *pb.Strings
out string
}{
{
// Test data from C++ test (TextFormatTest.StringEscape).
// Single divergence: we don't escape apostrophes.
&pb.Strings{StringField: proto.String("\"A string with ' characters \n and \r newlines and \t tabs and \001 slashes \\ and multiple spaces")},
"string_field: \"\\\"A string with ' characters \\n and \\r newlines and \\t tabs and \\001 slashes \\\\ and multiple spaces\"\n",
},
{
// Test data from the same C++ test.
&pb.Strings{StringField: proto.String("\350\260\267\346\255\214")},
"string_field: \"\\350\\260\\267\\346\\255\\214\"\n",
},
{
// Some UTF-8.
&pb.Strings{StringField: proto.String("\x00\x01\xff\x81")},
`string_field: "\000\001\377\201"` + "\n",
},
}
for i, tc := range testCases {
var buf bytes.Buffer
if err := proto.MarshalText(&buf, tc.in); err != nil {
t.Errorf("proto.MarsalText: %v", err)
continue
}
s := buf.String()
if s != tc.out {
t.Errorf("#%d: Got:\n%s\nExpected:\n%s\n", i, s, tc.out)
continue
}
// Check round-trip.
pb := new(pb.Strings)
if err := proto.UnmarshalText(s, pb); err != nil {
t.Errorf("#%d: UnmarshalText: %v", i, err)
continue
}
if !proto.Equal(pb, tc.in) {
t.Errorf("#%d: Round-trip failed:\nstart: %v\n end: %v", i, tc.in, pb)
}
}
}
// A limitedWriter accepts some output before it fails.
// This is a proxy for something like a nearly-full or imminently-failing disk,
// or a network connection that is about to die.
type limitedWriter struct {
b bytes.Buffer
limit int
}
var outOfSpace = errors.New("proto: insufficient space")
func (w *limitedWriter) Write(p []byte) (n int, err error) {
var avail = w.limit - w.b.Len()
if avail <= 0 {
return 0, outOfSpace
}
if len(p) <= avail {
return w.b.Write(p)
}
n, _ = w.b.Write(p[:avail])
return n, outOfSpace
}
func TestMarshalTextFailing(t *testing.T) {
// Try lots of different sizes to exercise more error code-paths.
for lim := 0; lim < len(text); lim++ {
buf := new(limitedWriter)
buf.limit = lim
err := proto.MarshalText(buf, newTestMessage())
// We expect a certain error, but also some partial results in the buffer.
if err != outOfSpace {
t.Errorf("Got:\n===\n%v===\nExpected:\n===\n%v===\n", err, outOfSpace)
}
s := buf.b.String()
x := text[:buf.limit]
if s != x {
t.Errorf("Got:\n===\n%v===\nExpected:\n===\n%v===\n", s, x)
}
}
}
func TestFloats(t *testing.T) {
tests := []struct {
f float64
want string
}{
{0, "0"},
{4.7, "4.7"},
{math.Inf(1), "inf"},
{math.Inf(-1), "-inf"},
{math.NaN(), "nan"},
}
for _, test := range tests {
msg := &pb.FloatingPoint{F: &test.f}
got := strings.TrimSpace(msg.String())
want := `f:` + test.want
if got != want {
t.Errorf("f=%f: got %q, want %q", test.f, got, want)
}
}
}
func TestRepeatedNilText(t *testing.T) {
m := &pb.MessageList{
Message: []*pb.MessageList_Message{
nil,
&pb.MessageList_Message{
Name: proto.String("Horse"),
},
nil,
},
}
want := `Message <nil>
Message {
name: "Horse"
}
Message <nil>
`
if s := proto.MarshalTextString(m); s != want {
t.Errorf(" got: %s\nwant: %s", s, want)
}
}

View File

@@ -5,11 +5,12 @@ package quantile_test
import (
"bufio"
"fmt"
"github.com/bmizerany/perks/quantile"
"log"
"os"
"strconv"
"time"
"github.com/coreos/etcd/Godeps/_workspace/src/github.com/beorn7/perks/quantile"
)
func Example_simple() {
@@ -17,7 +18,11 @@ func Example_simple() {
go sendFloats(ch)
// Compute the 50th, 90th, and 99th percentile.
q := quantile.NewTargeted(0.50, 0.90, 0.99)
q := quantile.NewTargeted(map[float64]float64{
0.50: 0.005,
0.90: 0.001,
0.99: 0.0001,
})
for v := range ch {
q.Insert(v)
}
@@ -28,8 +33,8 @@ func Example_simple() {
fmt.Println("count:", q.Count())
// Output:
// perc50: 5
// perc90: 14
// perc99: 40
// perc90: 16
// perc99: 223
// count: 2388
}
@@ -52,7 +57,7 @@ func Example_mergeMultipleStreams() {
// even if we do not plan to query them all here.
ch := make(chan quantile.Samples)
getDBQuerySamples(ch)
q := quantile.NewTargeted(0.90)
q := quantile.NewTargeted(map[float64]float64{0.90: 0.001})
for samples := range ch {
q.Merge(samples)
}
@@ -67,7 +72,11 @@ func Example_window() {
go sendStreamValues(ch)
tick := time.NewTicker(1 * time.Minute)
q := quantile.NewTargeted(0.90, 0.95, 0.99)
q := quantile.NewTargeted(map[float64]float64{
0.90: 0.001,
0.95: 0.0005,
0.99: 0.0001,
})
for {
select {
case t := <-tick.C:

View File

@@ -113,7 +113,8 @@ func TestHighBiasedQuery(t *testing.T) {
verifyHighPercsWithRelativeEpsilon(t, a, s)
}
func TestTargetedMerge(t *testing.T) {
// BrokenTestTargetedMerge is broken, see Merge doc comment.
func BrokenTestTargetedMerge(t *testing.T) {
rand.Seed(42)
s1 := NewTargeted(Targets)
s2 := NewTargeted(Targets)
@@ -123,7 +124,8 @@ func TestTargetedMerge(t *testing.T) {
verifyPercsWithAbsoluteEpsilon(t, a, s1)
}
func TestLowBiasedMerge(t *testing.T) {
// BrokenTestLowBiasedMerge is broken, see Merge doc comment.
func BrokenTestLowBiasedMerge(t *testing.T) {
rand.Seed(42)
s1 := NewLowBiased(RelativeEpsilon)
s2 := NewLowBiased(RelativeEpsilon)
@@ -133,7 +135,8 @@ func TestLowBiasedMerge(t *testing.T) {
verifyLowPercsWithRelativeEpsilon(t, a, s2)
}
func TestHighBiasedMerge(t *testing.T) {
// BrokenTestHighBiasedMerge is broken, see Merge doc comment.
func BrokenTestHighBiasedMerge(t *testing.T) {
rand.Seed(42)
s1 := NewHighBiased(RelativeEpsilon)
s2 := NewHighBiased(RelativeEpsilon)

View File

@@ -0,0 +1,2 @@
example/example
example/example.exe

View File

@@ -0,0 +1,201 @@
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work
(an example is provided in the Appendix below).
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding those notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf
of any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
END OF TERMS AND CONDITIONS
APPENDIX: How to apply the Apache License to your work.
To apply the Apache License to your work, attach the following
boilerplate notice, with the fields enclosed by brackets "[]"
replaced with your own identifying information. (Don't include
the brackets!) The text should be enclosed in the appropriate
comment syntax for the file format. We also recommend that a
file or class name and description of purpose be included on the
same "printed page" as the copyright notice for easier
identification within third-party archives.
Copyright [2013] [the CloudFoundry Authors]
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

View File

@@ -0,0 +1,30 @@
# Speakeasy
This package provides cross-platform Go (#golang) helpers for taking user input
from the terminal while not echoing the input back (similar to `getpasswd`). The
package uses syscalls to avoid any dependence on cgo, and is therefore
compatible with cross-compiling.
[![GoDoc](https://godoc.org/github.com/bgentry/speakeasy?status.png)][godoc]
## Unicode
Multi-byte unicode characters work successfully on Mac OS X. On Windows,
however, this may be problematic (as is UTF in general on Windows). Other
platforms have not been tested.
## License
The code herein was not written by me, but was compiled from two separate open
source packages. Unix portions were imported from [gopass][gopass], while
Windows portions were imported from the [CloudFoundry Go CLI][cf-cli]'s
[Windows terminal helpers][cf-ui-windows].
The [license for the windows portion](./LICENSE_WINDOWS) has been copied exactly
from the source (though I attempted to fill in the correct owner in the
boilerplate copyright notice).
[cf-cli]: https://github.com/cloudfoundry/cli "CloudFoundry Go CLI"
[cf-ui-windows]: https://github.com/cloudfoundry/cli/blob/master/src/cf/terminal/ui_windows.go "CloudFoundry Go CLI Windows input helpers"
[godoc]: https://godoc.org/github.com/bgentry/speakeasy "speakeasy on Godoc.org"
[gopass]: https://code.google.com/p/gopass "gopass"

View File

@@ -0,0 +1,18 @@
package main
import (
"fmt"
"os"
"github.com/coreos/etcd/Godeps/_workspace/src/github.com/bgentry/speakeasy"
)
func main() {
password, err := speakeasy.Ask("Please enter a password: ")
if err != nil {
fmt.Println(err)
os.Exit(1)
}
fmt.Printf("Password result: %q\n", password)
fmt.Printf("Password len: %d\n", len(password))
}

View File

@@ -0,0 +1,47 @@
package speakeasy
import (
"fmt"
"io"
"os"
"strings"
)
// Ask the user to enter a password with input hidden. prompt is a string to
// display before the user's input. Returns the provided password, or an error
// if the command failed.
func Ask(prompt string) (password string, err error) {
return FAsk(os.Stdout, prompt)
}
// Same as the Ask function, except it is possible to specify the file to write
// the prompt to.
func FAsk(file *os.File, prompt string) (password string, err error) {
if prompt != "" {
fmt.Fprint(file, prompt) // Display the prompt.
}
password, err = getPassword()
// Carriage return after the user input.
fmt.Fprintln(file, "")
return
}
func readline() (value string, err error) {
var valb []byte
var n int
b := make([]byte, 1)
for {
// read one byte at a time so we don't accidentally read extra bytes
n, err = os.Stdin.Read(b)
if err != nil && err != io.EOF {
return "", err
}
if n == 0 || b[0] == '\n' {
break
}
valb = append(valb, b[0])
}
return strings.TrimSuffix(string(valb), "\r"), nil
}

View File

@@ -0,0 +1,93 @@
// based on https://code.google.com/p/gopass
// Author: johnsiilver@gmail.com (John Doak)
//
// Original code is based on code by RogerV in the golang-nuts thread:
// https://groups.google.com/group/golang-nuts/browse_thread/thread/40cc41e9d9fc9247
// +build darwin freebsd linux netbsd openbsd
package speakeasy
import (
"fmt"
"os"
"os/signal"
"strings"
"syscall"
)
const sttyArg0 = "/bin/stty"
var (
sttyArgvEOff []string = []string{"stty", "-echo"}
sttyArgvEOn []string = []string{"stty", "echo"}
ws syscall.WaitStatus = 0
)
// getPassword gets input hidden from the terminal from a user. This is
// accomplished by turning off terminal echo, reading input from the user and
// finally turning on terminal echo.
func getPassword() (password string, err error) {
sig := make(chan os.Signal, 10)
brk := make(chan bool)
// File descriptors for stdin, stdout, and stderr.
fd := []uintptr{os.Stdin.Fd(), os.Stdout.Fd(), os.Stderr.Fd()}
// Setup notifications of termination signals to channel sig, create a process to
// watch for these signals so we can turn back on echo if need be.
signal.Notify(sig, syscall.SIGHUP, syscall.SIGINT, syscall.SIGKILL, syscall.SIGQUIT,
syscall.SIGTERM)
go catchSignal(fd, sig, brk)
// Turn off the terminal echo.
pid, err := echoOff(fd)
if err != nil {
return "", err
}
// Turn on the terminal echo and stop listening for signals.
defer close(brk)
defer echoOn(fd)
syscall.Wait4(pid, &ws, 0, nil)
line, err := readline()
if err == nil {
password = strings.TrimSpace(line)
} else {
err = fmt.Errorf("failed during password entry: %s", err)
}
return password, err
}
// echoOff turns off the terminal echo.
func echoOff(fd []uintptr) (int, error) {
pid, err := syscall.ForkExec(sttyArg0, sttyArgvEOff, &syscall.ProcAttr{Dir: "", Files: fd})
if err != nil {
return 0, fmt.Errorf("failed turning off console echo for password entry:\n\t%s", err)
}
return pid, nil
}
// echoOn turns back on the terminal echo.
func echoOn(fd []uintptr) {
// Turn on the terminal echo.
pid, e := syscall.ForkExec(sttyArg0, sttyArgvEOn, &syscall.ProcAttr{Dir: "", Files: fd})
if e == nil {
syscall.Wait4(pid, &ws, 0, nil)
}
}
// catchSignal tries to catch SIGKILL, SIGQUIT and SIGINT so that we can turn
// terminal echo back on before the program ends. Otherwise the user is left
// with echo off on their terminal.
func catchSignal(fd []uintptr, sig chan os.Signal, brk chan bool) {
select {
case <-sig:
echoOn(fd)
os.Exit(-1)
case <-brk:
}
}

View File

@@ -0,0 +1,43 @@
// +build windows
package speakeasy
import (
"os"
"syscall"
)
// SetConsoleMode function can be used to change value of ENABLE_ECHO_INPUT:
// http://msdn.microsoft.com/en-us/library/windows/desktop/ms686033(v=vs.85).aspx
const ENABLE_ECHO_INPUT = 0x0004
func getPassword() (password string, err error) {
hStdin := syscall.Handle(os.Stdin.Fd())
var oldMode uint32
err = syscall.GetConsoleMode(hStdin, &oldMode)
if err != nil {
return
}
var newMode uint32 = (oldMode &^ ENABLE_ECHO_INPUT)
err = setConsoleMode(hStdin, newMode)
defer setConsoleMode(hStdin, oldMode)
if err != nil {
return
}
return readline()
}
func setConsoleMode(console syscall.Handle, mode uint32) (err error) {
dll := syscall.MustLoadDLL("kernel32")
proc := dll.MustFindProc("SetConsoleMode")
r, _, err := proc.Call(uintptr(console), uintptr(mode))
if r == 0 {
return err
}
return nil
}

View File

@@ -0,0 +1,3 @@
*.prof
*.test
/bin/

20
Godeps/_workspace/src/github.com/boltdb/bolt/LICENSE generated vendored Normal file
View File

@@ -0,0 +1,20 @@
The MIT License (MIT)
Copyright (c) 2013 Ben Johnson
Permission is hereby granted, free of charge, to any person obtaining a copy of
this software and associated documentation files (the "Software"), to deal in
the Software without restriction, including without limitation the rights to
use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of
the Software, and to permit persons to whom the Software is furnished to do so,
subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS
FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR
COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER
IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

54
Godeps/_workspace/src/github.com/boltdb/bolt/Makefile generated vendored Normal file
View File

@@ -0,0 +1,54 @@
TEST=.
BENCH=.
COVERPROFILE=/tmp/c.out
BRANCH=`git rev-parse --abbrev-ref HEAD`
COMMIT=`git rev-parse --short HEAD`
GOLDFLAGS="-X main.branch $(BRANCH) -X main.commit $(COMMIT)"
default: build
bench:
go test -v -test.run=NOTHINCONTAINSTHIS -test.bench=$(BENCH)
# http://cloc.sourceforge.net/
cloc:
@cloc --not-match-f='Makefile|_test.go' .
cover: fmt
go test -coverprofile=$(COVERPROFILE) -test.run=$(TEST) $(COVERFLAG) .
go tool cover -html=$(COVERPROFILE)
rm $(COVERPROFILE)
cpuprofile: fmt
@go test -c
@./bolt.test -test.v -test.run=$(TEST) -test.cpuprofile cpu.prof
# go get github.com/kisielk/errcheck
errcheck:
@echo "=== errcheck ==="
@errcheck github.com/boltdb/bolt
fmt:
@go fmt ./...
get:
@go get -d ./...
build: get
@mkdir -p bin
@go build -ldflags=$(GOLDFLAGS) -a -o bin/bolt ./cmd/bolt
test: fmt
@go get github.com/stretchr/testify/assert
@echo "=== TESTS ==="
@go test -v -cover -test.run=$(TEST)
@echo ""
@echo ""
@echo "=== CLI ==="
@go test -v -test.run=$(TEST) ./cmd/bolt
@echo ""
@echo ""
@echo "=== RACE DETECTOR ==="
@go test -v -race -test.run="TestSimulate_(100op|1000op)"
.PHONY: bench cloc cover cpuprofile fmt memprofile test

591
Godeps/_workspace/src/github.com/boltdb/bolt/README.md generated vendored Normal file
View File

@@ -0,0 +1,591 @@
Bolt [![Build Status](https://drone.io/github.com/boltdb/bolt/status.png)](https://drone.io/github.com/boltdb/bolt/latest) [![Coverage Status](https://coveralls.io/repos/boltdb/bolt/badge.png?branch=master)](https://coveralls.io/r/boltdb/bolt?branch=master) [![GoDoc](https://godoc.org/github.com/boltdb/bolt?status.png)](https://godoc.org/github.com/boltdb/bolt) ![Version](http://img.shields.io/badge/version-1.0-green.png)
====
Bolt is a pure Go key/value store inspired by [Howard Chu's][hyc_symas] and
the [LMDB project][lmdb]. The goal of the project is to provide a simple,
fast, and reliable database for projects that don't require a full database
server such as Postgres or MySQL.
Since Bolt is meant to be used as such a low-level piece of functionality,
simplicity is key. The API will be small and only focus on getting values
and setting values. That's it.
[hyc_symas]: https://twitter.com/hyc_symas
[lmdb]: http://symas.com/mdb/
## Project Status
Bolt is stable and the API is fixed. Full unit test coverage and randomized
black box testing are used to ensure database consistency and thread safety.
Bolt is currently in high-load production environments serving databases as
large as 1TB. Many companies such as Shopify and Heroku use Bolt-backed
services every day.
## Getting Started
### Installing
To start using Bolt, install Go and run `go get`:
```sh
$ go get github.com/boltdb/bolt/...
```
This will retrieve the library and install the `bolt` command line utility into
your `$GOBIN` path.
### Opening a database
The top-level object in Bolt is a `DB`. It is represented as a single file on
your disk and represents a consistent snapshot of your data.
To open your database, simply use the `bolt.Open()` function:
```go
package main
import (
"log"
"github.com/boltdb/bolt"
)
func main() {
// Open the my.db data file in your current directory.
// It will be created if it doesn't exist.
db, err := bolt.Open("my.db", 0600, nil)
if err != nil {
log.Fatal(err)
}
defer db.Close()
...
}
```
Please note that Bolt obtains a file lock on the data file so multiple processes
cannot open the same database at the same time. Opening an already open Bolt
database will cause it to hang until the other process closes it. To prevent
an indefinite wait you can pass a timeout option to the `Open()` function:
```go
db, err := bolt.Open("my.db", 0600, &bolt.Options{Timeout: 1 * time.Second})
```
### Transactions
Bolt allows only one read-write transaction at a time but allows as many
read-only transactions as you want at a time. Each transaction has a consistent
view of the data as it existed when the transaction started.
Individual transactions and all objects created from them (e.g. buckets, keys)
are not thread safe. To work with data in multiple goroutines you must start
a transaction for each one or use locking to ensure only one goroutine accesses
a transaction at a time. Creating transaction from the `DB` is thread safe.
#### Read-write transactions
To start a read-write transaction, you can use the `DB.Update()` function:
```go
err := db.Update(func(tx *bolt.Tx) error {
...
return nil
})
```
Inside the closure, you have a consistent view of the database. You commit the
transaction by returning `nil` at the end. You can also rollback the transaction
at any point by returning an error. All database operations are allowed inside
a read-write transaction.
Always check the return error as it will report any disk failures that can cause
your transaction to not complete. If you return an error within your closure
it will be passed through.
#### Read-only transactions
To start a read-only transaction, you can use the `DB.View()` function:
```go
err := db.View(func(tx *bolt.Tx) error {
...
return nil
})
```
You also get a consistent view of the database within this closure, however,
no mutating operations are allowed within a read-only transaction. You can only
retrieve buckets, retrieve values, and copy the database within a read-only
transaction.
#### Batch read-write transactions
Each `DB.Update()` waits for disk to commit the writes. This overhead
can be minimized by combining multiple updates with the `DB.Batch()`
function:
```go
err := db.Batch(func(tx *bolt.Tx) error {
...
return nil
})
```
Concurrent Batch calls are opportunistically combined into larger
transactions. Batch is only useful when there are multiple goroutines
calling it.
The trade-off is that `Batch` can call the given
function multiple times, if parts of the transaction fail. The
function must be idempotent and side effects must take effect only
after a successful return from `DB.Batch()`.
For example: don't display messages from inside the function, instead
set variables in the enclosing scope:
```go
var id uint64
err := db.Batch(func(tx *bolt.Tx) error {
// Find last key in bucket, decode as bigendian uint64, increment
// by one, encode back to []byte, and add new key.
...
id = newValue
return nil
})
if err != nil {
return ...
}
fmt.Println("Allocated ID %d", id)
```
#### Managing transactions manually
The `DB.View()` and `DB.Update()` functions are wrappers around the `DB.Begin()`
function. These helper functions will start the transaction, execute a function,
and then safely close your transaction if an error is returned. This is the
recommended way to use Bolt transactions.
However, sometimes you may want to manually start and end your transactions.
You can use the `Tx.Begin()` function directly but _please_ be sure to close the
transaction.
```go
// Start a writable transaction.
tx, err := db.Begin(true)
if err != nil {
return err
}
defer tx.Rollback()
// Use the transaction...
_, err := tx.CreateBucket([]byte("MyBucket"))
if err != nil {
return err
}
// Commit the transaction and check for error.
if err := tx.Commit(); err != nil {
return err
}
```
The first argument to `DB.Begin()` is a boolean stating if the transaction
should be writable.
### Using buckets
Buckets are collections of key/value pairs within the database. All keys in a
bucket must be unique. You can create a bucket using the `DB.CreateBucket()`
function:
```go
db.Update(func(tx *bolt.Tx) error {
b, err := tx.CreateBucket([]byte("MyBucket"))
if err != nil {
return fmt.Errorf("create bucket: %s", err)
}
return nil
})
```
You can also create a bucket only if it doesn't exist by using the
`Tx.CreateBucketIfNotExists()` function. It's a common pattern to call this
function for all your top-level buckets after you open your database so you can
guarantee that they exist for future transactions.
To delete a bucket, simply call the `Tx.DeleteBucket()` function.
### Using key/value pairs
To save a key/value pair to a bucket, use the `Bucket.Put()` function:
```go
db.Update(func(tx *bolt.Tx) error {
b := tx.Bucket([]byte("MyBucket"))
err := b.Put([]byte("answer"), []byte("42"))
return err
})
```
This will set the value of the `"answer"` key to `"42"` in the `MyBucket`
bucket. To retrieve this value, we can use the `Bucket.Get()` function:
```go
db.View(func(tx *bolt.Tx) error {
b := tx.Bucket([]byte("MyBucket"))
v := b.Get([]byte("answer"))
fmt.Printf("The answer is: %s\n", v)
return nil
})
```
The `Get()` function does not return an error because its operation is
guarenteed to work (unless there is some kind of system failure). If the key
exists then it will return its byte slice value. If it doesn't exist then it
will return `nil`. It's important to note that you can have a zero-length value
set to a key which is different than the key not existing.
Use the `Bucket.Delete()` function to delete a key from the bucket.
Please note that values returned from `Get()` are only valid while the
transaction is open. If you need to use a value outside of the transaction
then you must use `copy()` to copy it to another byte slice.
### Iterating over keys
Bolt stores its keys in byte-sorted order within a bucket. This makes sequential
iteration over these keys extremely fast. To iterate over keys we'll use a
`Cursor`:
```go
db.View(func(tx *bolt.Tx) error {
b := tx.Bucket([]byte("MyBucket"))
c := b.Cursor()
for k, v := c.First(); k != nil; k, v = c.Next() {
fmt.Printf("key=%s, value=%s\n", k, v)
}
return nil
})
```
The cursor allows you to move to a specific point in the list of keys and move
forward or backward through the keys one at a time.
The following functions are available on the cursor:
```
First() Move to the first key.
Last() Move to the last key.
Seek() Move to a specific key.
Next() Move to the next key.
Prev() Move to the previous key.
```
When you have iterated to the end of the cursor then `Next()` will return `nil`.
You must seek to a position using `First()`, `Last()`, or `Seek()` before
calling `Next()` or `Prev()`. If you do not seek to a position then these
functions will return `nil`.
#### Prefix scans
To iterate over a key prefix, you can combine `Seek()` and `bytes.HasPrefix()`:
```go
db.View(func(tx *bolt.Tx) error {
c := tx.Bucket([]byte("MyBucket")).Cursor()
prefix := []byte("1234")
for k, v := c.Seek(prefix); bytes.HasPrefix(k, prefix); k, v = c.Next() {
fmt.Printf("key=%s, value=%s\n", k, v)
}
return nil
})
```
#### Range scans
Another common use case is scanning over a range such as a time range. If you
use a sortable time encoding such as RFC3339 then you can query a specific
date range like this:
```go
db.View(func(tx *bolt.Tx) error {
// Assume our events bucket has RFC3339 encoded time keys.
c := tx.Bucket([]byte("Events")).Cursor()
// Our time range spans the 90's decade.
min := []byte("1990-01-01T00:00:00Z")
max := []byte("2000-01-01T00:00:00Z")
// Iterate over the 90's.
for k, v := c.Seek(min); k != nil && bytes.Compare(k, max) <= 0; k, v = c.Next() {
fmt.Printf("%s: %s\n", k, v)
}
return nil
})
```
#### ForEach()
You can also use the function `ForEach()` if you know you'll be iterating over
all the keys in a bucket:
```go
db.View(func(tx *bolt.Tx) error {
b := tx.Bucket([]byte("MyBucket"))
b.ForEach(func(k, v []byte) error {
fmt.Printf("key=%s, value=%s\n", k, v)
return nil
})
return nil
})
```
### Nested buckets
You can also store a bucket in a key to create nested buckets. The API is the
same as the bucket management API on the `DB` object:
```go
func (*Bucket) CreateBucket(key []byte) (*Bucket, error)
func (*Bucket) CreateBucketIfNotExists(key []byte) (*Bucket, error)
func (*Bucket) DeleteBucket(key []byte) error
```
### Database backups
Bolt is a single file so it's easy to backup. You can use the `Tx.WriteTo()`
function to write a consistent view of the database to a writer. If you call
this from a read-only transaction, it will perform a hot backup and not block
your other database reads and writes. It will also use `O_DIRECT` when available
to prevent page cache trashing.
One common use case is to backup over HTTP so you can use tools like `cURL` to
do database backups:
```go
func BackupHandleFunc(w http.ResponseWriter, req *http.Request) {
err := db.View(func(tx *bolt.Tx) error {
w.Header().Set("Content-Type", "application/octet-stream")
w.Header().Set("Content-Disposition", `attachment; filename="my.db"`)
w.Header().Set("Content-Length", strconv.Itoa(int(tx.Size())))
_, err := tx.WriteTo(w)
return err
})
if err != nil {
http.Error(w, err.Error(), http.StatusInternalServerError)
}
}
```
Then you can backup using this command:
```sh
$ curl http://localhost/backup > my.db
```
Or you can open your browser to `http://localhost/backup` and it will download
automatically.
If you want to backup to another file you can use the `Tx.CopyFile()` helper
function.
### Statistics
The database keeps a running count of many of the internal operations it
performs so you can better understand what's going on. By grabbing a snapshot
of these stats at two points in time we can see what operations were performed
in that time range.
For example, we could start a goroutine to log stats every 10 seconds:
```go
go func() {
// Grab the initial stats.
prev := db.Stats()
for {
// Wait for 10s.
time.Sleep(10 * time.Second)
// Grab the current stats and diff them.
stats := db.Stats()
diff := stats.Sub(&prev)
// Encode stats to JSON and print to STDERR.
json.NewEncoder(os.Stderr).Encode(diff)
// Save stats for the next loop.
prev = stats
}
}()
```
It's also useful to pipe these stats to a service such as statsd for monitoring
or to provide an HTTP endpoint that will perform a fixed-length sample.
## Resources
For more information on getting started with Bolt, check out the following articles:
* [Intro to BoltDB: Painless Performant Persistence](http://npf.io/2014/07/intro-to-boltdb-painless-performant-persistence/) by [Nate Finch](https://github.com/natefinch).
* [Bolt -- an embedded key/value database for Go](https://www.progville.com/go/bolt-embedded-db-golang/) by Progville
## Comparison with other databases
### Postgres, MySQL, & other relational databases
Relational databases structure data into rows and are only accessible through
the use of SQL. This approach provides flexibility in how you store and query
your data but also incurs overhead in parsing and planning SQL statements. Bolt
accesses all data by a byte slice key. This makes Bolt fast to read and write
data by key but provides no built-in support for joining values together.
Most relational databases (with the exception of SQLite) are standalone servers
that run separately from your application. This gives your systems
flexibility to connect multiple application servers to a single database
server but also adds overhead in serializing and transporting data over the
network. Bolt runs as a library included in your application so all data access
has to go through your application's process. This brings data closer to your
application but limits multi-process access to the data.
### LevelDB, RocksDB
LevelDB and its derivatives (RocksDB, HyperLevelDB) are similar to Bolt in that
they are libraries bundled into the application, however, their underlying
structure is a log-structured merge-tree (LSM tree). An LSM tree optimizes
random writes by using a write ahead log and multi-tiered, sorted files called
SSTables. Bolt uses a B+tree internally and only a single file. Both approaches
have trade offs.
If you require a high random write throughput (>10,000 w/sec) or you need to use
spinning disks then LevelDB could be a good choice. If your application is
read-heavy or does a lot of range scans then Bolt could be a good choice.
One other important consideration is that LevelDB does not have transactions.
It supports batch writing of key/values pairs and it supports read snapshots
but it will not give you the ability to do a compare-and-swap operation safely.
Bolt supports fully serializable ACID transactions.
### LMDB
Bolt was originally a port of LMDB so it is architecturally similar. Both use
a B+tree, have ACID semantics with fully serializable transactions, and support
lock-free MVCC using a single writer and multiple readers.
The two projects have somewhat diverged. LMDB heavily focuses on raw performance
while Bolt has focused on simplicity and ease of use. For example, LMDB allows
several unsafe actions such as direct writes for the sake of performance. Bolt
opts to disallow actions which can leave the database in a corrupted state. The
only exception to this in Bolt is `DB.NoSync`.
There are also a few differences in API. LMDB requires a maximum mmap size when
opening an `mdb_env` whereas Bolt will handle incremental mmap resizing
automatically. LMDB overloads the getter and setter functions with multiple
flags whereas Bolt splits these specialized cases into their own functions.
## Caveats & Limitations
It's important to pick the right tool for the job and Bolt is no exception.
Here are a few things to note when evaluating and using Bolt:
* Bolt is good for read intensive workloads. Sequential write performance is
also fast but random writes can be slow. You can add a write-ahead log or
[transaction coalescer](https://github.com/boltdb/coalescer) in front of Bolt
to mitigate this issue.
* Bolt uses a B+tree internally so there can be a lot of random page access.
SSDs provide a significant performance boost over spinning disks.
* Try to avoid long running read transactions. Bolt uses copy-on-write so
old pages cannot be reclaimed while an old transaction is using them.
* Byte slices returned from Bolt are only valid during a transaction. Once the
transaction has been committed or rolled back then the memory they point to
can be reused by a new page or can be unmapped from virtual memory and you'll
see an `unexpected fault address` panic when accessing it.
* Be careful when using `Bucket.FillPercent`. Setting a high fill percent for
buckets that have random inserts will cause your database to have very poor
page utilization.
* Use larger buckets in general. Smaller buckets causes poor page utilization
once they become larger than the page size (typically 4KB).
* Bulk loading a lot of random writes into a new bucket can be slow as the
page will not split until the transaction is committed. Randomly inserting
more than 100,000 key/value pairs into a single new bucket in a single
transaction is not advised.
* Bolt uses a memory-mapped file so the underlying operating system handles the
caching of the data. Typically, the OS will cache as much of the file as it
can in memory and will release memory as needed to other processes. This means
that Bolt can show very high memory usage when working with large databases.
However, this is expected and the OS will release memory as needed. Bolt can
handle databases much larger than the available physical RAM.
* Because of the way pages are laid out on disk, Bolt cannot truncate data files
and return free pages back to the disk. Instead, Bolt maintains a free list
of unused pages within its data file. These free pages can be reused by later
transactions. This works well for many use cases as databases generally tend
to grow. However, it's important to note that deleting large chunks of data
will not allow you to reclaim that space on disk.
For more information on page allocation, [see this comment][page-allocation].
[page-allocation]: https://github.com/boltdb/bolt/issues/308#issuecomment-74811638
## Other Projects Using Bolt
Below is a list of public, open source projects that use Bolt:
* [Operation Go: A Routine Mission](http://gocode.io) - An online programming game for Golang using Bolt for user accounts and a leaderboard.
* [Bazil](https://github.com/bazillion/bazil) - A file system that lets your data reside where it is most convenient for it to reside.
* [DVID](https://github.com/janelia-flyem/dvid) - Added Bolt as optional storage engine and testing it against Basho-tuned leveldb.
* [Skybox Analytics](https://github.com/skybox/skybox) - A standalone funnel analysis tool for web analytics.
* [Scuttlebutt](https://github.com/benbjohnson/scuttlebutt) - Uses Bolt to store and process all Twitter mentions of GitHub projects.
* [Wiki](https://github.com/peterhellberg/wiki) - A tiny wiki using Goji, BoltDB and Blackfriday.
* [ChainStore](https://github.com/nulayer/chainstore) - Simple key-value interface to a variety of storage engines organized as a chain of operations.
* [MetricBase](https://github.com/msiebuhr/MetricBase) - Single-binary version of Graphite.
* [Gitchain](https://github.com/gitchain/gitchain) - Decentralized, peer-to-peer Git repositories aka "Git meets Bitcoin".
* [event-shuttle](https://github.com/sclasen/event-shuttle) - A Unix system service to collect and reliably deliver messages to Kafka.
* [ipxed](https://github.com/kelseyhightower/ipxed) - Web interface and api for ipxed.
* [BoltStore](https://github.com/yosssi/boltstore) - Session store using Bolt.
* [photosite/session](http://godoc.org/bitbucket.org/kardianos/photosite/session) - Sessions for a photo viewing site.
* [LedisDB](https://github.com/siddontang/ledisdb) - A high performance NoSQL, using Bolt as optional storage.
* [ipLocator](https://github.com/AndreasBriese/ipLocator) - A fast ip-geo-location-server using bolt with bloom filters.
* [cayley](https://github.com/google/cayley) - Cayley is an open-source graph database using Bolt as optional backend.
* [bleve](http://www.blevesearch.com/) - A pure Go search engine similar to ElasticSearch that uses Bolt as the default storage backend.
* [tentacool](https://github.com/optiflows/tentacool) - REST api server to manage system stuff (IP, DNS, Gateway...) on a linux server.
* [SkyDB](https://github.com/skydb/sky) - Behavioral analytics database.
* [Seaweed File System](https://github.com/chrislusf/weed-fs) - Highly scalable distributed key~file system with O(1) disk read.
* [InfluxDB](http://influxdb.com) - Scalable datastore for metrics, events, and real-time analytics.
If you are using Bolt in a project please send a pull request to add it to the list.

135
Godeps/_workspace/src/github.com/boltdb/bolt/batch.go generated vendored Normal file
View File

@@ -0,0 +1,135 @@
package bolt
import (
"errors"
"fmt"
"sync"
"time"
)
// Batch calls fn as part of a batch. It behaves similar to Update,
// except:
//
// 1. concurrent Batch calls can be combined into a single Bolt
// transaction.
//
// 2. the function passed to Batch may be called multiple times,
// regardless of whether it returns error or not.
//
// This means that Batch function side effects must be idempotent and
// take permanent effect only after a successful return is seen in
// caller.
//
// Batch is only useful when there are multiple goroutines calling it.
func (db *DB) Batch(fn func(*Tx) error) error {
errCh := make(chan error, 1)
db.batchMu.Lock()
if (db.batch == nil) || (db.batch != nil && len(db.batch.calls) >= db.MaxBatchSize) {
// There is no existing batch, or the existing batch is full; start a new one.
db.batch = &batch{
db: db,
}
db.batch.timer = time.AfterFunc(db.MaxBatchDelay, db.batch.trigger)
}
db.batch.calls = append(db.batch.calls, call{fn: fn, err: errCh})
if len(db.batch.calls) >= db.MaxBatchSize {
// wake up batch, it's ready to run
go db.batch.trigger()
}
db.batchMu.Unlock()
err := <-errCh
if err == trySolo {
err = db.Update(fn)
}
return err
}
type call struct {
fn func(*Tx) error
err chan<- error
}
type batch struct {
db *DB
timer *time.Timer
start sync.Once
calls []call
}
// trigger runs the batch if it hasn't already been run.
func (b *batch) trigger() {
b.start.Do(b.run)
}
// run performs the transactions in the batch and communicates results
// back to DB.Batch.
func (b *batch) run() {
b.db.batchMu.Lock()
b.timer.Stop()
// Make sure no new work is added to this batch, but don't break
// other batches.
if b.db.batch == b {
b.db.batch = nil
}
b.db.batchMu.Unlock()
retry:
for len(b.calls) > 0 {
var failIdx = -1
err := b.db.Update(func(tx *Tx) error {
for i, c := range b.calls {
if err := safelyCall(c.fn, tx); err != nil {
failIdx = i
return err
}
}
return nil
})
if failIdx >= 0 {
// take the failing transaction out of the batch. it's
// safe to shorten b.calls here because db.batch no longer
// points to us, and we hold the mutex anyway.
c := b.calls[failIdx]
b.calls[failIdx], b.calls = b.calls[len(b.calls)-1], b.calls[:len(b.calls)-1]
// tell the submitter re-run it solo, continue with the rest of the batch
c.err <- trySolo
continue retry
}
// pass success, or bolt internal errors, to all callers
for _, c := range b.calls {
if c.err != nil {
c.err <- err
}
}
break retry
}
}
// trySolo is a special sentinel error value used for signaling that a
// transaction function should be re-run. It should never be seen by
// callers.
var trySolo = errors.New("batch function returned an error and should be re-run solo")
type panicked struct {
reason interface{}
}
func (p panicked) Error() string {
if err, ok := p.reason.(error); ok {
return err.Error()
}
return fmt.Sprintf("panic: %v", p.reason)
}
func safelyCall(fn func(*Tx) error, tx *Tx) (err error) {
defer func() {
if p := recover(); p != nil {
err = panicked{p}
}
}()
return fn(tx)
}

View File

@@ -0,0 +1,170 @@
package bolt_test
import (
"bytes"
"encoding/binary"
"errors"
"hash/fnv"
"sync"
"testing"
"github.com/coreos/etcd/Godeps/_workspace/src/github.com/boltdb/bolt"
)
func validateBatchBench(b *testing.B, db *TestDB) {
var rollback = errors.New("sentinel error to cause rollback")
validate := func(tx *bolt.Tx) error {
bucket := tx.Bucket([]byte("bench"))
h := fnv.New32a()
buf := make([]byte, 4)
for id := uint32(0); id < 1000; id++ {
binary.LittleEndian.PutUint32(buf, id)
h.Reset()
h.Write(buf[:])
k := h.Sum(nil)
v := bucket.Get(k)
if v == nil {
b.Errorf("not found id=%d key=%x", id, k)
continue
}
if g, e := v, []byte("filler"); !bytes.Equal(g, e) {
b.Errorf("bad value for id=%d key=%x: %s != %q", id, k, g, e)
}
if err := bucket.Delete(k); err != nil {
return err
}
}
// should be empty now
c := bucket.Cursor()
for k, v := c.First(); k != nil; k, v = c.Next() {
b.Errorf("unexpected key: %x = %q", k, v)
}
return rollback
}
if err := db.Update(validate); err != nil && err != rollback {
b.Error(err)
}
}
func BenchmarkDBBatchAutomatic(b *testing.B) {
db := NewTestDB()
defer db.Close()
db.MustCreateBucket([]byte("bench"))
b.ResetTimer()
for i := 0; i < b.N; i++ {
start := make(chan struct{})
var wg sync.WaitGroup
for round := 0; round < 1000; round++ {
wg.Add(1)
go func(id uint32) {
defer wg.Done()
<-start
h := fnv.New32a()
buf := make([]byte, 4)
binary.LittleEndian.PutUint32(buf, id)
h.Write(buf[:])
k := h.Sum(nil)
insert := func(tx *bolt.Tx) error {
b := tx.Bucket([]byte("bench"))
return b.Put(k, []byte("filler"))
}
if err := db.Batch(insert); err != nil {
b.Error(err)
return
}
}(uint32(round))
}
close(start)
wg.Wait()
}
b.StopTimer()
validateBatchBench(b, db)
}
func BenchmarkDBBatchSingle(b *testing.B) {
db := NewTestDB()
defer db.Close()
db.MustCreateBucket([]byte("bench"))
b.ResetTimer()
for i := 0; i < b.N; i++ {
start := make(chan struct{})
var wg sync.WaitGroup
for round := 0; round < 1000; round++ {
wg.Add(1)
go func(id uint32) {
defer wg.Done()
<-start
h := fnv.New32a()
buf := make([]byte, 4)
binary.LittleEndian.PutUint32(buf, id)
h.Write(buf[:])
k := h.Sum(nil)
insert := func(tx *bolt.Tx) error {
b := tx.Bucket([]byte("bench"))
return b.Put(k, []byte("filler"))
}
if err := db.Update(insert); err != nil {
b.Error(err)
return
}
}(uint32(round))
}
close(start)
wg.Wait()
}
b.StopTimer()
validateBatchBench(b, db)
}
func BenchmarkDBBatchManual10x100(b *testing.B) {
db := NewTestDB()
defer db.Close()
db.MustCreateBucket([]byte("bench"))
b.ResetTimer()
for i := 0; i < b.N; i++ {
start := make(chan struct{})
var wg sync.WaitGroup
for major := 0; major < 10; major++ {
wg.Add(1)
go func(id uint32) {
defer wg.Done()
<-start
insert100 := func(tx *bolt.Tx) error {
h := fnv.New32a()
buf := make([]byte, 4)
for minor := uint32(0); minor < 100; minor++ {
binary.LittleEndian.PutUint32(buf, uint32(id*100+minor))
h.Reset()
h.Write(buf[:])
k := h.Sum(nil)
b := tx.Bucket([]byte("bench"))
if err := b.Put(k, []byte("filler")); err != nil {
return err
}
}
return nil
}
if err := db.Update(insert100); err != nil {
b.Fatal(err)
}
}(uint32(major))
}
close(start)
wg.Wait()
}
b.StopTimer()
validateBatchBench(b, db)
}

View File

@@ -0,0 +1,148 @@
package bolt_test
import (
"encoding/binary"
"fmt"
"io/ioutil"
"log"
"math/rand"
"net/http"
"net/http/httptest"
"os"
"github.com/coreos/etcd/Godeps/_workspace/src/github.com/boltdb/bolt"
)
// Set this to see how the counts are actually updated.
const verbose = false
// Counter updates a counter in Bolt for every URL path requested.
type counter struct {
db *bolt.DB
}
func (c counter) ServeHTTP(rw http.ResponseWriter, req *http.Request) {
// Communicates the new count from a successful database
// transaction.
var result uint64
increment := func(tx *bolt.Tx) error {
b, err := tx.CreateBucketIfNotExists([]byte("hits"))
if err != nil {
return err
}
key := []byte(req.URL.String())
// Decode handles key not found for us.
count := decode(b.Get(key)) + 1
b.Put(key, encode(count))
// All good, communicate new count.
result = count
return nil
}
if err := c.db.Batch(increment); err != nil {
http.Error(rw, err.Error(), 500)
return
}
if verbose {
log.Printf("server: %s: %d", req.URL.String(), result)
}
rw.Header().Set("Content-Type", "application/octet-stream")
fmt.Fprintf(rw, "%d\n", result)
}
func client(id int, base string, paths []string) error {
// Process paths in random order.
rng := rand.New(rand.NewSource(int64(id)))
permutation := rng.Perm(len(paths))
for i := range paths {
path := paths[permutation[i]]
resp, err := http.Get(base + path)
if err != nil {
return err
}
defer resp.Body.Close()
buf, err := ioutil.ReadAll(resp.Body)
if err != nil {
return err
}
if verbose {
log.Printf("client: %s: %s", path, buf)
}
}
return nil
}
func ExampleDB_Batch() {
// Open the database.
db, _ := bolt.Open(tempfile(), 0666, nil)
defer os.Remove(db.Path())
defer db.Close()
// Start our web server
count := counter{db}
srv := httptest.NewServer(count)
defer srv.Close()
// Decrease the batch size to make things more interesting.
db.MaxBatchSize = 3
// Get every path multiple times concurrently.
const clients = 10
paths := []string{
"/foo",
"/bar",
"/baz",
"/quux",
"/thud",
"/xyzzy",
}
errors := make(chan error, clients)
for i := 0; i < clients; i++ {
go func(id int) {
errors <- client(id, srv.URL, paths)
}(i)
}
// Check all responses to make sure there's no error.
for i := 0; i < clients; i++ {
if err := <-errors; err != nil {
fmt.Printf("client error: %v", err)
return
}
}
// Check the final result
db.View(func(tx *bolt.Tx) error {
b := tx.Bucket([]byte("hits"))
c := b.Cursor()
for k, v := c.First(); k != nil; k, v = c.Next() {
fmt.Printf("hits to %s: %d\n", k, decode(v))
}
return nil
})
// Output:
// hits to /bar: 10
// hits to /baz: 10
// hits to /foo: 10
// hits to /quux: 10
// hits to /thud: 10
// hits to /xyzzy: 10
}
// encode marshals a counter.
func encode(n uint64) []byte {
buf := make([]byte, 8)
binary.BigEndian.PutUint64(buf, n)
return buf
}
// decode unmarshals a counter. Nil buffers are decoded as 0.
func decode(buf []byte) uint64 {
if buf == nil {
return 0
}
return binary.BigEndian.Uint64(buf)
}

View File

@@ -0,0 +1,167 @@
package bolt_test
import (
"testing"
"time"
"github.com/coreos/etcd/Godeps/_workspace/src/github.com/boltdb/bolt"
)
// Ensure two functions can perform updates in a single batch.
func TestDB_Batch(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.MustCreateBucket([]byte("widgets"))
// Iterate over multiple updates in separate goroutines.
n := 2
ch := make(chan error)
for i := 0; i < n; i++ {
go func(i int) {
ch <- db.Batch(func(tx *bolt.Tx) error {
return tx.Bucket([]byte("widgets")).Put(u64tob(uint64(i)), []byte{})
})
}(i)
}
// Check all responses to make sure there's no error.
for i := 0; i < n; i++ {
if err := <-ch; err != nil {
t.Fatal(err)
}
}
// Ensure data is correct.
db.MustView(func(tx *bolt.Tx) error {
b := tx.Bucket([]byte("widgets"))
for i := 0; i < n; i++ {
if v := b.Get(u64tob(uint64(i))); v == nil {
t.Errorf("key not found: %d", i)
}
}
return nil
})
}
func TestDB_Batch_Panic(t *testing.T) {
db := NewTestDB()
defer db.Close()
var sentinel int
var bork = &sentinel
var problem interface{}
var err error
// Execute a function inside a batch that panics.
func() {
defer func() {
if p := recover(); p != nil {
problem = p
}
}()
err = db.Batch(func(tx *bolt.Tx) error {
panic(bork)
})
}()
// Verify there is no error.
if g, e := err, error(nil); g != e {
t.Fatalf("wrong error: %v != %v", g, e)
}
// Verify the panic was captured.
if g, e := problem, bork; g != e {
t.Fatalf("wrong error: %v != %v", g, e)
}
}
func TestDB_BatchFull(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.MustCreateBucket([]byte("widgets"))
const size = 3
// buffered so we never leak goroutines
ch := make(chan error, size)
put := func(i int) {
ch <- db.Batch(func(tx *bolt.Tx) error {
return tx.Bucket([]byte("widgets")).Put(u64tob(uint64(i)), []byte{})
})
}
db.MaxBatchSize = size
// high enough to never trigger here
db.MaxBatchDelay = 1 * time.Hour
go put(1)
go put(2)
// Give the batch a chance to exhibit bugs.
time.Sleep(10 * time.Millisecond)
// not triggered yet
select {
case <-ch:
t.Fatalf("batch triggered too early")
default:
}
go put(3)
// Check all responses to make sure there's no error.
for i := 0; i < size; i++ {
if err := <-ch; err != nil {
t.Fatal(err)
}
}
// Ensure data is correct.
db.MustView(func(tx *bolt.Tx) error {
b := tx.Bucket([]byte("widgets"))
for i := 1; i <= size; i++ {
if v := b.Get(u64tob(uint64(i))); v == nil {
t.Errorf("key not found: %d", i)
}
}
return nil
})
}
func TestDB_BatchTime(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.MustCreateBucket([]byte("widgets"))
const size = 1
// buffered so we never leak goroutines
ch := make(chan error, size)
put := func(i int) {
ch <- db.Batch(func(tx *bolt.Tx) error {
return tx.Bucket([]byte("widgets")).Put(u64tob(uint64(i)), []byte{})
})
}
db.MaxBatchSize = 1000
db.MaxBatchDelay = 0
go put(1)
// Batch must trigger by time alone.
// Check all responses to make sure there's no error.
for i := 0; i < size; i++ {
if err := <-ch; err != nil {
t.Fatal(err)
}
}
// Ensure data is correct.
db.MustView(func(tx *bolt.Tx) error {
b := tx.Bucket([]byte("widgets"))
for i := 1; i <= size; i++ {
if v := b.Get(u64tob(uint64(i))); v == nil {
t.Errorf("key not found: %d", i)
}
}
return nil
})
}

View File

@@ -0,0 +1,7 @@
package bolt
// maxMapSize represents the largest mmap size supported by Bolt.
const maxMapSize = 0x7FFFFFFF // 2GB
// maxAllocSize is the size used when creating array pointers.
const maxAllocSize = 0xFFFFFFF

View File

@@ -0,0 +1,7 @@
package bolt
// maxMapSize represents the largest mmap size supported by Bolt.
const maxMapSize = 0xFFFFFFFFFFFF // 256TB
// maxAllocSize is the size used when creating array pointers.
const maxAllocSize = 0x7FFFFFFF

View File

@@ -0,0 +1,7 @@
package bolt
// maxMapSize represents the largest mmap size supported by Bolt.
const maxMapSize = 0x7FFFFFFF // 2GB
// maxAllocSize is the size used when creating array pointers.
const maxAllocSize = 0xFFFFFFF

View File

@@ -0,0 +1,12 @@
package bolt
import (
"syscall"
)
var odirect = syscall.O_DIRECT
// fdatasync flushes written data to a file descriptor.
func fdatasync(db *DB) error {
return syscall.Fdatasync(int(db.file.Fd()))
}

View File

@@ -0,0 +1,29 @@
package bolt
import (
"syscall"
"unsafe"
)
const (
msAsync = 1 << iota // perform asynchronous writes
msSync // perform synchronous writes
msInvalidate // invalidate cached data
)
var odirect int
func msync(db *DB) error {
_, _, errno := syscall.Syscall(syscall.SYS_MSYNC, uintptr(unsafe.Pointer(db.data)), uintptr(db.datasz), msInvalidate)
if errno != 0 {
return errno
}
return nil
}
func fdatasync(db *DB) error {
if db.data != nil {
return msync(db)
}
return db.file.Sync()
}

View File

@@ -0,0 +1,36 @@
package bolt_test
import (
"fmt"
"path/filepath"
"reflect"
"runtime"
"testing"
)
// assert fails the test if the condition is false.
func assert(tb testing.TB, condition bool, msg string, v ...interface{}) {
if !condition {
_, file, line, _ := runtime.Caller(1)
fmt.Printf("\033[31m%s:%d: "+msg+"\033[39m\n\n", append([]interface{}{filepath.Base(file), line}, v...)...)
tb.FailNow()
}
}
// ok fails the test if an err is not nil.
func ok(tb testing.TB, err error) {
if err != nil {
_, file, line, _ := runtime.Caller(1)
fmt.Printf("\033[31m%s:%d: unexpected error: %s\033[39m\n\n", filepath.Base(file), line, err.Error())
tb.FailNow()
}
}
// equals fails the test if exp is not equal to act.
func equals(tb testing.TB, exp, act interface{}) {
if !reflect.DeepEqual(exp, act) {
_, file, line, _ := runtime.Caller(1)
fmt.Printf("\033[31m%s:%d:\n\n\texp: %#v\n\n\tgot: %#v\033[39m\n\n", filepath.Base(file), line, exp, act)
tb.FailNow()
}
}

View File

@@ -0,0 +1,80 @@
// +build !windows,!plan9
package bolt
import (
"fmt"
"os"
"syscall"
"time"
"unsafe"
)
// flock acquires an advisory lock on a file descriptor.
func flock(f *os.File, timeout time.Duration) error {
var t time.Time
for {
// If we're beyond our timeout then return an error.
// This can only occur after we've attempted a flock once.
if t.IsZero() {
t = time.Now()
} else if timeout > 0 && time.Since(t) > timeout {
return ErrTimeout
}
// Otherwise attempt to obtain an exclusive lock.
err := syscall.Flock(int(f.Fd()), syscall.LOCK_EX|syscall.LOCK_NB)
if err == nil {
return nil
} else if err != syscall.EWOULDBLOCK {
return err
}
// Wait for a bit and try again.
time.Sleep(50 * time.Millisecond)
}
}
// funlock releases an advisory lock on a file descriptor.
func funlock(f *os.File) error {
return syscall.Flock(int(f.Fd()), syscall.LOCK_UN)
}
// mmap memory maps a DB's data file.
func mmap(db *DB, sz int) error {
// Truncate and fsync to ensure file size metadata is flushed.
// https://github.com/boltdb/bolt/issues/284
if err := db.file.Truncate(int64(sz)); err != nil {
return fmt.Errorf("file resize error: %s", err)
}
if err := db.file.Sync(); err != nil {
return fmt.Errorf("file sync error: %s", err)
}
// Map the data file to memory.
b, err := syscall.Mmap(int(db.file.Fd()), 0, sz, syscall.PROT_READ, syscall.MAP_SHARED)
if err != nil {
return err
}
// Save the original byte slice and convert to a byte array pointer.
db.dataref = b
db.data = (*[maxMapSize]byte)(unsafe.Pointer(&b[0]))
db.datasz = sz
return nil
}
// munmap unmaps a DB's data file from memory.
func munmap(db *DB) error {
// Ignore the unmap if we have no mapped data.
if db.dataref == nil {
return nil
}
// Unmap using the original byte slice.
err := syscall.Munmap(db.dataref)
db.dataref = nil
db.data = nil
db.datasz = 0
return err
}

View File

@@ -0,0 +1,74 @@
package bolt
import (
"fmt"
"os"
"syscall"
"time"
"unsafe"
)
var odirect int
// fdatasync flushes written data to a file descriptor.
func fdatasync(db *DB) error {
return db.file.Sync()
}
// flock acquires an advisory lock on a file descriptor.
func flock(f *os.File, _ time.Duration) error {
return nil
}
// funlock releases an advisory lock on a file descriptor.
func funlock(f *os.File) error {
return nil
}
// mmap memory maps a DB's data file.
// Based on: https://github.com/edsrzf/mmap-go
func mmap(db *DB, sz int) error {
// Truncate the database to the size of the mmap.
if err := db.file.Truncate(int64(sz)); err != nil {
return fmt.Errorf("truncate: %s", err)
}
// Open a file mapping handle.
sizelo := uint32(sz >> 32)
sizehi := uint32(sz) & 0xffffffff
h, errno := syscall.CreateFileMapping(syscall.Handle(db.file.Fd()), nil, syscall.PAGE_READONLY, sizelo, sizehi, nil)
if h == 0 {
return os.NewSyscallError("CreateFileMapping", errno)
}
// Create the memory map.
addr, errno := syscall.MapViewOfFile(h, syscall.FILE_MAP_READ, 0, 0, uintptr(sz))
if addr == 0 {
return os.NewSyscallError("MapViewOfFile", errno)
}
// Close mapping handle.
if err := syscall.CloseHandle(syscall.Handle(h)); err != nil {
return os.NewSyscallError("CloseHandle", err)
}
// Convert to a byte array.
db.data = ((*[maxMapSize]byte)(unsafe.Pointer(addr)))
db.datasz = sz
return nil
}
// munmap unmaps a pointer from a file.
// Based on: https://github.com/edsrzf/mmap-go
func munmap(db *DB) error {
if db.data == nil {
return nil
}
addr := (uintptr)(unsafe.Pointer(&db.data[0]))
if err := syscall.UnmapViewOfFile(addr); err != nil {
return os.NewSyscallError("UnmapViewOfFile", err)
}
return nil
}

View File

@@ -0,0 +1,10 @@
// +build !windows,!plan9,!linux,!openbsd
package bolt
var odirect int
// fdatasync flushes written data to a file descriptor.
func fdatasync(db *DB) error {
return db.file.Sync()
}

743
Godeps/_workspace/src/github.com/boltdb/bolt/bucket.go generated vendored Normal file
View File

@@ -0,0 +1,743 @@
package bolt
import (
"bytes"
"fmt"
"unsafe"
)
const (
// MaxKeySize is the maximum length of a key, in bytes.
MaxKeySize = 32768
// MaxValueSize is the maximum length of a value, in bytes.
MaxValueSize = 4294967295
)
const (
maxUint = ^uint(0)
minUint = 0
maxInt = int(^uint(0) >> 1)
minInt = -maxInt - 1
)
const bucketHeaderSize = int(unsafe.Sizeof(bucket{}))
const (
minFillPercent = 0.1
maxFillPercent = 1.0
)
// DefaultFillPercent is the percentage that split pages are filled.
// This value can be changed by setting Bucket.FillPercent.
const DefaultFillPercent = 0.5
// Bucket represents a collection of key/value pairs inside the database.
type Bucket struct {
*bucket
tx *Tx // the associated transaction
buckets map[string]*Bucket // subbucket cache
page *page // inline page reference
rootNode *node // materialized node for the root page.
nodes map[pgid]*node // node cache
// Sets the threshold for filling nodes when they split. By default,
// the bucket will fill to 50% but it can be useful to increase this
// amount if you know that your write workloads are mostly append-only.
//
// This is non-persisted across transactions so it must be set in every Tx.
FillPercent float64
}
// bucket represents the on-file representation of a bucket.
// This is stored as the "value" of a bucket key. If the bucket is small enough,
// then its root page can be stored inline in the "value", after the bucket
// header. In the case of inline buckets, the "root" will be 0.
type bucket struct {
root pgid // page id of the bucket's root-level page
sequence uint64 // monotonically incrementing, used by NextSequence()
}
// newBucket returns a new bucket associated with a transaction.
func newBucket(tx *Tx) Bucket {
var b = Bucket{tx: tx, FillPercent: DefaultFillPercent}
if tx.writable {
b.buckets = make(map[string]*Bucket)
b.nodes = make(map[pgid]*node)
}
return b
}
// Tx returns the tx of the bucket.
func (b *Bucket) Tx() *Tx {
return b.tx
}
// Root returns the root of the bucket.
func (b *Bucket) Root() pgid {
return b.root
}
// Writable returns whether the bucket is writable.
func (b *Bucket) Writable() bool {
return b.tx.writable
}
// Cursor creates a cursor associated with the bucket.
// The cursor is only valid as long as the transaction is open.
// Do not use a cursor after the transaction is closed.
func (b *Bucket) Cursor() *Cursor {
// Update transaction statistics.
b.tx.stats.CursorCount++
// Allocate and return a cursor.
return &Cursor{
bucket: b,
stack: make([]elemRef, 0),
}
}
// Bucket retrieves a nested bucket by name.
// Returns nil if the bucket does not exist.
func (b *Bucket) Bucket(name []byte) *Bucket {
if b.buckets != nil {
if child := b.buckets[string(name)]; child != nil {
return child
}
}
// Move cursor to key.
c := b.Cursor()
k, v, flags := c.seek(name)
// Return nil if the key doesn't exist or it is not a bucket.
if !bytes.Equal(name, k) || (flags&bucketLeafFlag) == 0 {
return nil
}
// Otherwise create a bucket and cache it.
var child = b.openBucket(v)
if b.buckets != nil {
b.buckets[string(name)] = child
}
return child
}
// Helper method that re-interprets a sub-bucket value
// from a parent into a Bucket
func (b *Bucket) openBucket(value []byte) *Bucket {
var child = newBucket(b.tx)
// If this is a writable transaction then we need to copy the bucket entry.
// Read-only transactions can point directly at the mmap entry.
if b.tx.writable {
child.bucket = &bucket{}
*child.bucket = *(*bucket)(unsafe.Pointer(&value[0]))
} else {
child.bucket = (*bucket)(unsafe.Pointer(&value[0]))
}
// Save a reference to the inline page if the bucket is inline.
if child.root == 0 {
child.page = (*page)(unsafe.Pointer(&value[bucketHeaderSize]))
}
return &child
}
// CreateBucket creates a new bucket at the given key and returns the new bucket.
// Returns an error if the key already exists, if the bucket name is blank, or if the bucket name is too long.
func (b *Bucket) CreateBucket(key []byte) (*Bucket, error) {
if b.tx.db == nil {
return nil, ErrTxClosed
} else if !b.tx.writable {
return nil, ErrTxNotWritable
} else if len(key) == 0 {
return nil, ErrBucketNameRequired
}
// Move cursor to correct position.
c := b.Cursor()
k, _, flags := c.seek(key)
// Return an error if there is an existing key.
if bytes.Equal(key, k) {
if (flags & bucketLeafFlag) != 0 {
return nil, ErrBucketExists
} else {
return nil, ErrIncompatibleValue
}
}
// Create empty, inline bucket.
var bucket = Bucket{
bucket: &bucket{},
rootNode: &node{isLeaf: true},
FillPercent: DefaultFillPercent,
}
var value = bucket.write()
// Insert into node.
key = cloneBytes(key)
c.node().put(key, key, value, 0, bucketLeafFlag)
// Since subbuckets are not allowed on inline buckets, we need to
// dereference the inline page, if it exists. This will cause the bucket
// to be treated as a regular, non-inline bucket for the rest of the tx.
b.page = nil
return b.Bucket(key), nil
}
// CreateBucketIfNotExists creates a new bucket if it doesn't already exist and returns a reference to it.
// Returns an error if the bucket name is blank, or if the bucket name is too long.
func (b *Bucket) CreateBucketIfNotExists(key []byte) (*Bucket, error) {
child, err := b.CreateBucket(key)
if err == ErrBucketExists {
return b.Bucket(key), nil
} else if err != nil {
return nil, err
}
return child, nil
}
// DeleteBucket deletes a bucket at the given key.
// Returns an error if the bucket does not exists, or if the key represents a non-bucket value.
func (b *Bucket) DeleteBucket(key []byte) error {
if b.tx.db == nil {
return ErrTxClosed
} else if !b.Writable() {
return ErrTxNotWritable
}
// Move cursor to correct position.
c := b.Cursor()
k, _, flags := c.seek(key)
// Return an error if bucket doesn't exist or is not a bucket.
if !bytes.Equal(key, k) {
return ErrBucketNotFound
} else if (flags & bucketLeafFlag) == 0 {
return ErrIncompatibleValue
}
// Recursively delete all child buckets.
child := b.Bucket(key)
err := child.ForEach(func(k, v []byte) error {
if v == nil {
if err := child.DeleteBucket(k); err != nil {
return fmt.Errorf("delete bucket: %s", err)
}
}
return nil
})
if err != nil {
return err
}
// Remove cached copy.
delete(b.buckets, string(key))
// Release all bucket pages to freelist.
child.nodes = nil
child.rootNode = nil
child.free()
// Delete the node if we have a matching key.
c.node().del(key)
return nil
}
// Get retrieves the value for a key in the bucket.
// Returns a nil value if the key does not exist or if the key is a nested bucket.
// The returned value is only valid for the life of the transaction.
func (b *Bucket) Get(key []byte) []byte {
k, v, flags := b.Cursor().seek(key)
// Return nil if this is a bucket.
if (flags & bucketLeafFlag) != 0 {
return nil
}
// If our target node isn't the same key as what's passed in then return nil.
if !bytes.Equal(key, k) {
return nil
}
return v
}
// Put sets the value for a key in the bucket.
// If the key exist then its previous value will be overwritten.
// Returns an error if the bucket was created from a read-only transaction, if the key is blank, if the key is too large, or if the value is too large.
func (b *Bucket) Put(key []byte, value []byte) error {
if b.tx.db == nil {
return ErrTxClosed
} else if !b.Writable() {
return ErrTxNotWritable
} else if len(key) == 0 {
return ErrKeyRequired
} else if len(key) > MaxKeySize {
return ErrKeyTooLarge
} else if int64(len(value)) > MaxValueSize {
return ErrValueTooLarge
}
// Move cursor to correct position.
c := b.Cursor()
k, _, flags := c.seek(key)
// Return an error if there is an existing key with a bucket value.
if bytes.Equal(key, k) && (flags&bucketLeafFlag) != 0 {
return ErrIncompatibleValue
}
// Insert into node.
key = cloneBytes(key)
c.node().put(key, key, value, 0, 0)
return nil
}
// Delete removes a key from the bucket.
// If the key does not exist then nothing is done and a nil error is returned.
// Returns an error if the bucket was created from a read-only transaction.
func (b *Bucket) Delete(key []byte) error {
if b.tx.db == nil {
return ErrTxClosed
} else if !b.Writable() {
return ErrTxNotWritable
}
// Move cursor to correct position.
c := b.Cursor()
_, _, flags := c.seek(key)
// Return an error if there is already existing bucket value.
if (flags & bucketLeafFlag) != 0 {
return ErrIncompatibleValue
}
// Delete the node if we have a matching key.
c.node().del(key)
return nil
}
// NextSequence returns an autoincrementing integer for the bucket.
func (b *Bucket) NextSequence() (uint64, error) {
if b.tx.db == nil {
return 0, ErrTxClosed
} else if !b.Writable() {
return 0, ErrTxNotWritable
}
// Materialize the root node if it hasn't been already so that the
// bucket will be saved during commit.
if b.rootNode == nil {
_ = b.node(b.root, nil)
}
// Increment and return the sequence.
b.bucket.sequence++
return b.bucket.sequence, nil
}
// ForEach executes a function for each key/value pair in a bucket.
// If the provided function returns an error then the iteration is stopped and
// the error is returned to the caller.
func (b *Bucket) ForEach(fn func(k, v []byte) error) error {
if b.tx.db == nil {
return ErrTxClosed
}
c := b.Cursor()
for k, v := c.First(); k != nil; k, v = c.Next() {
if err := fn(k, v); err != nil {
return err
}
}
return nil
}
// Stat returns stats on a bucket.
func (b *Bucket) Stats() BucketStats {
var s, subStats BucketStats
pageSize := b.tx.db.pageSize
s.BucketN += 1
if b.root == 0 {
s.InlineBucketN += 1
}
b.forEachPage(func(p *page, depth int) {
if (p.flags & leafPageFlag) != 0 {
s.KeyN += int(p.count)
// used totals the used bytes for the page
used := pageHeaderSize
if p.count != 0 {
// If page has any elements, add all element headers.
used += leafPageElementSize * int(p.count-1)
// Add all element key, value sizes.
// The computation takes advantage of the fact that the position
// of the last element's key/value equals to the total of the sizes
// of all previous elements' keys and values.
// It also includes the last element's header.
lastElement := p.leafPageElement(p.count - 1)
used += int(lastElement.pos + lastElement.ksize + lastElement.vsize)
}
if b.root == 0 {
// For inlined bucket just update the inline stats
s.InlineBucketInuse += used
} else {
// For non-inlined bucket update all the leaf stats
s.LeafPageN++
s.LeafInuse += used
s.LeafOverflowN += int(p.overflow)
// Collect stats from sub-buckets.
// Do that by iterating over all element headers
// looking for the ones with the bucketLeafFlag.
for i := uint16(0); i < p.count; i++ {
e := p.leafPageElement(i)
if (e.flags & bucketLeafFlag) != 0 {
// For any bucket element, open the element value
// and recursively call Stats on the contained bucket.
subStats.Add(b.openBucket(e.value()).Stats())
}
}
}
} else if (p.flags & branchPageFlag) != 0 {
s.BranchPageN++
lastElement := p.branchPageElement(p.count - 1)
// used totals the used bytes for the page
// Add header and all element headers.
used := pageHeaderSize + (branchPageElementSize * int(p.count-1))
// Add size of all keys and values.
// Again, use the fact that last element's position equals to
// the total of key, value sizes of all previous elements.
used += int(lastElement.pos + lastElement.ksize)
s.BranchInuse += used
s.BranchOverflowN += int(p.overflow)
}
// Keep track of maximum page depth.
if depth+1 > s.Depth {
s.Depth = (depth + 1)
}
})
// Alloc stats can be computed from page counts and pageSize.
s.BranchAlloc = (s.BranchPageN + s.BranchOverflowN) * pageSize
s.LeafAlloc = (s.LeafPageN + s.LeafOverflowN) * pageSize
// Add the max depth of sub-buckets to get total nested depth.
s.Depth += subStats.Depth
// Add the stats for all sub-buckets
s.Add(subStats)
return s
}
// forEachPage iterates over every page in a bucket, including inline pages.
func (b *Bucket) forEachPage(fn func(*page, int)) {
// If we have an inline page then just use that.
if b.page != nil {
fn(b.page, 0)
return
}
// Otherwise traverse the page hierarchy.
b.tx.forEachPage(b.root, 0, fn)
}
// forEachPageNode iterates over every page (or node) in a bucket.
// This also includes inline pages.
func (b *Bucket) forEachPageNode(fn func(*page, *node, int)) {
// If we have an inline page or root node then just use that.
if b.page != nil {
fn(b.page, nil, 0)
return
}
b._forEachPageNode(b.root, 0, fn)
}
func (b *Bucket) _forEachPageNode(pgid pgid, depth int, fn func(*page, *node, int)) {
var p, n = b.pageNode(pgid)
// Execute function.
fn(p, n, depth)
// Recursively loop over children.
if p != nil {
if (p.flags & branchPageFlag) != 0 {
for i := 0; i < int(p.count); i++ {
elem := p.branchPageElement(uint16(i))
b._forEachPageNode(elem.pgid, depth+1, fn)
}
}
} else {
if !n.isLeaf {
for _, inode := range n.inodes {
b._forEachPageNode(inode.pgid, depth+1, fn)
}
}
}
}
// spill writes all the nodes for this bucket to dirty pages.
func (b *Bucket) spill() error {
// Spill all child buckets first.
for name, child := range b.buckets {
// If the child bucket is small enough and it has no child buckets then
// write it inline into the parent bucket's page. Otherwise spill it
// like a normal bucket and make the parent value a pointer to the page.
var value []byte
if child.inlineable() {
child.free()
value = child.write()
} else {
if err := child.spill(); err != nil {
return err
}
// Update the child bucket header in this bucket.
value = make([]byte, unsafe.Sizeof(bucket{}))
var bucket = (*bucket)(unsafe.Pointer(&value[0]))
*bucket = *child.bucket
}
// Skip writing the bucket if there are no materialized nodes.
if child.rootNode == nil {
continue
}
// Update parent node.
var c = b.Cursor()
k, _, flags := c.seek([]byte(name))
if !bytes.Equal([]byte(name), k) {
panic(fmt.Sprintf("misplaced bucket header: %x -> %x", []byte(name), k))
}
if flags&bucketLeafFlag == 0 {
panic(fmt.Sprintf("unexpected bucket header flag: %x", flags))
}
c.node().put([]byte(name), []byte(name), value, 0, bucketLeafFlag)
}
// Ignore if there's not a materialized root node.
if b.rootNode == nil {
return nil
}
// Spill nodes.
if err := b.rootNode.spill(); err != nil {
return err
}
b.rootNode = b.rootNode.root()
// Update the root node for this bucket.
if b.rootNode.pgid >= b.tx.meta.pgid {
panic(fmt.Sprintf("pgid (%d) above high water mark (%d)", b.rootNode.pgid, b.tx.meta.pgid))
}
b.root = b.rootNode.pgid
return nil
}
// inlineable returns true if a bucket is small enough to be written inline
// and if it contains no subbuckets. Otherwise returns false.
func (b *Bucket) inlineable() bool {
var n = b.rootNode
// Bucket must only contain a single leaf node.
if n == nil || !n.isLeaf {
return false
}
// Bucket is not inlineable if it contains subbuckets or if it goes beyond
// our threshold for inline bucket size.
var size = pageHeaderSize
for _, inode := range n.inodes {
size += leafPageElementSize + len(inode.key) + len(inode.value)
if inode.flags&bucketLeafFlag != 0 {
return false
} else if size > b.maxInlineBucketSize() {
return false
}
}
return true
}
// Returns the maximum total size of a bucket to make it a candidate for inlining.
func (b *Bucket) maxInlineBucketSize() int {
return b.tx.db.pageSize / 4
}
// write allocates and writes a bucket to a byte slice.
func (b *Bucket) write() []byte {
// Allocate the appropriate size.
var n = b.rootNode
var value = make([]byte, bucketHeaderSize+n.size())
// Write a bucket header.
var bucket = (*bucket)(unsafe.Pointer(&value[0]))
*bucket = *b.bucket
// Convert byte slice to a fake page and write the root node.
var p = (*page)(unsafe.Pointer(&value[bucketHeaderSize]))
n.write(p)
return value
}
// rebalance attempts to balance all nodes.
func (b *Bucket) rebalance() {
for _, n := range b.nodes {
n.rebalance()
}
for _, child := range b.buckets {
child.rebalance()
}
}
// node creates a node from a page and associates it with a given parent.
func (b *Bucket) node(pgid pgid, parent *node) *node {
_assert(b.nodes != nil, "nodes map expected")
// Retrieve node if it's already been created.
if n := b.nodes[pgid]; n != nil {
return n
}
// Otherwise create a node and cache it.
n := &node{bucket: b, parent: parent}
if parent == nil {
b.rootNode = n
} else {
parent.children = append(parent.children, n)
}
// Use the inline page if this is an inline bucket.
var p = b.page
if p == nil {
p = b.tx.page(pgid)
}
// Read the page into the node and cache it.
n.read(p)
b.nodes[pgid] = n
// Update statistics.
b.tx.stats.NodeCount++
return n
}
// free recursively frees all pages in the bucket.
func (b *Bucket) free() {
if b.root == 0 {
return
}
var tx = b.tx
b.forEachPageNode(func(p *page, n *node, _ int) {
if p != nil {
tx.db.freelist.free(tx.meta.txid, p)
} else {
n.free()
}
})
b.root = 0
}
// dereference removes all references to the old mmap.
func (b *Bucket) dereference() {
if b.rootNode != nil {
b.rootNode.root().dereference()
}
for _, child := range b.buckets {
child.dereference()
}
}
// pageNode returns the in-memory node, if it exists.
// Otherwise returns the underlying page.
func (b *Bucket) pageNode(id pgid) (*page, *node) {
// Inline buckets have a fake page embedded in their value so treat them
// differently. We'll return the rootNode (if available) or the fake page.
if b.root == 0 {
if id != 0 {
panic(fmt.Sprintf("inline bucket non-zero page access(2): %d != 0", id))
}
if b.rootNode != nil {
return nil, b.rootNode
}
return b.page, nil
}
// Check the node cache for non-inline buckets.
if b.nodes != nil {
if n := b.nodes[id]; n != nil {
return nil, n
}
}
// Finally lookup the page from the transaction if no node is materialized.
return b.tx.page(id), nil
}
// BucketStats records statistics about resources used by a bucket.
type BucketStats struct {
// Page count statistics.
BranchPageN int // number of logical branch pages
BranchOverflowN int // number of physical branch overflow pages
LeafPageN int // number of logical leaf pages
LeafOverflowN int // number of physical leaf overflow pages
// Tree statistics.
KeyN int // number of keys/value pairs
Depth int // number of levels in B+tree
// Page size utilization.
BranchAlloc int // bytes allocated for physical branch pages
BranchInuse int // bytes actually used for branch data
LeafAlloc int // bytes allocated for physical leaf pages
LeafInuse int // bytes actually used for leaf data
// Bucket statistics
BucketN int // total number of buckets including the top bucket
InlineBucketN int // total number on inlined buckets
InlineBucketInuse int // bytes used for inlined buckets (also accounted for in LeafInuse)
}
func (s *BucketStats) Add(other BucketStats) {
s.BranchPageN += other.BranchPageN
s.BranchOverflowN += other.BranchOverflowN
s.LeafPageN += other.LeafPageN
s.LeafOverflowN += other.LeafOverflowN
s.KeyN += other.KeyN
if s.Depth < other.Depth {
s.Depth = other.Depth
}
s.BranchAlloc += other.BranchAlloc
s.BranchInuse += other.BranchInuse
s.LeafAlloc += other.LeafAlloc
s.LeafInuse += other.LeafInuse
s.BucketN += other.BucketN
s.InlineBucketN += other.InlineBucketN
s.InlineBucketInuse += other.InlineBucketInuse
}
// cloneBytes returns a copy of a given slice.
func cloneBytes(v []byte) []byte {
var clone = make([]byte, len(v))
copy(clone, v)
return clone
}

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

View File

@@ -0,0 +1,145 @@
package main_test
import (
"bytes"
"io/ioutil"
"os"
"strconv"
"testing"
"github.com/coreos/etcd/Godeps/_workspace/src/github.com/boltdb/bolt"
"github.com/coreos/etcd/Godeps/_workspace/src/github.com/boltdb/bolt/cmd/bolt"
)
// Ensure the "info" command can print information about a database.
func TestInfoCommand_Run(t *testing.T) {
db := MustOpen(0666, nil)
db.DB.Close()
defer db.Close()
// Run the info command.
m := NewMain()
if err := m.Run("info", db.Path); err != nil {
t.Fatal(err)
}
}
// Ensure the "stats" command can execute correctly.
func TestStatsCommand_Run(t *testing.T) {
// Ignore
if os.Getpagesize() != 4096 {
t.Skip("system does not use 4KB page size")
}
db := MustOpen(0666, nil)
defer db.Close()
if err := db.Update(func(tx *bolt.Tx) error {
// Create "foo" bucket.
b, err := tx.CreateBucket([]byte("foo"))
if err != nil {
return err
}
for i := 0; i < 10; i++ {
if err := b.Put([]byte(strconv.Itoa(i)), []byte(strconv.Itoa(i))); err != nil {
return err
}
}
// Create "bar" bucket.
b, err = tx.CreateBucket([]byte("bar"))
if err != nil {
return err
}
for i := 0; i < 100; i++ {
if err := b.Put([]byte(strconv.Itoa(i)), []byte(strconv.Itoa(i))); err != nil {
return err
}
}
// Create "baz" bucket.
b, err = tx.CreateBucket([]byte("baz"))
if err != nil {
return err
}
if err := b.Put([]byte("key"), []byte("value")); err != nil {
return err
}
return nil
}); err != nil {
t.Fatal(err)
}
db.DB.Close()
// Generate expected result.
exp := "Aggregate statistics for 3 buckets\n\n" +
"Page count statistics\n" +
"\tNumber of logical branch pages: 0\n" +
"\tNumber of physical branch overflow pages: 0\n" +
"\tNumber of logical leaf pages: 1\n" +
"\tNumber of physical leaf overflow pages: 0\n" +
"Tree statistics\n" +
"\tNumber of keys/value pairs: 111\n" +
"\tNumber of levels in B+tree: 1\n" +
"Page size utilization\n" +
"\tBytes allocated for physical branch pages: 0\n" +
"\tBytes actually used for branch data: 0 (0%)\n" +
"\tBytes allocated for physical leaf pages: 4096\n" +
"\tBytes actually used for leaf data: 1996 (48%)\n" +
"Bucket statistics\n" +
"\tTotal number of buckets: 3\n" +
"\tTotal number on inlined buckets: 2 (66%)\n" +
"\tBytes used for inlined buckets: 236 (11%)\n"
// Run the command.
m := NewMain()
if err := m.Run("stats", db.Path); err != nil {
t.Fatal(err)
} else if m.Stdout.String() != exp {
t.Fatalf("unexpected stdout:\n\n%s", m.Stdout.String())
}
}
// Main represents a test wrapper for main.Main that records output.
type Main struct {
*main.Main
Stdin bytes.Buffer
Stdout bytes.Buffer
Stderr bytes.Buffer
}
// NewMain returns a new instance of Main.
func NewMain() *Main {
m := &Main{Main: main.NewMain()}
m.Main.Stdin = &m.Stdin
m.Main.Stdout = &m.Stdout
m.Main.Stderr = &m.Stderr
return m
}
// MustOpen creates a Bolt database in a temporary location.
func MustOpen(mode os.FileMode, options *bolt.Options) *DB {
// Create temporary path.
f, _ := ioutil.TempFile("", "bolt-")
f.Close()
os.Remove(f.Name())
db, err := bolt.Open(f.Name(), mode, options)
if err != nil {
panic(err.Error())
}
return &DB{DB: db, Path: f.Name()}
}
// DB is a test wrapper for bolt.DB.
type DB struct {
*bolt.DB
Path string
}
// Close closes and removes the database.
func (db *DB) Close() error {
defer os.Remove(db.Path)
return db.DB.Close()
}

384
Godeps/_workspace/src/github.com/boltdb/bolt/cursor.go generated vendored Normal file
View File

@@ -0,0 +1,384 @@
package bolt
import (
"bytes"
"fmt"
"sort"
)
// Cursor represents an iterator that can traverse over all key/value pairs in a bucket in sorted order.
// Cursors see nested buckets with value == nil.
// Cursors can be obtained from a transaction and are valid as long as the transaction is open.
//
// Keys and values returned from the cursor are only valid for the life of the transaction.
//
// Changing data while traversing with a cursor may cause it to be invalidated
// and return unexpected keys and/or values. You must reposition your cursor
// after mutating data.
type Cursor struct {
bucket *Bucket
stack []elemRef
}
// Bucket returns the bucket that this cursor was created from.
func (c *Cursor) Bucket() *Bucket {
return c.bucket
}
// First moves the cursor to the first item in the bucket and returns its key and value.
// If the bucket is empty then a nil key and value are returned.
// The returned key and value are only valid for the life of the transaction.
func (c *Cursor) First() (key []byte, value []byte) {
_assert(c.bucket.tx.db != nil, "tx closed")
c.stack = c.stack[:0]
p, n := c.bucket.pageNode(c.bucket.root)
c.stack = append(c.stack, elemRef{page: p, node: n, index: 0})
c.first()
k, v, flags := c.keyValue()
if (flags & uint32(bucketLeafFlag)) != 0 {
return k, nil
}
return k, v
}
// Last moves the cursor to the last item in the bucket and returns its key and value.
// If the bucket is empty then a nil key and value are returned.
// The returned key and value are only valid for the life of the transaction.
func (c *Cursor) Last() (key []byte, value []byte) {
_assert(c.bucket.tx.db != nil, "tx closed")
c.stack = c.stack[:0]
p, n := c.bucket.pageNode(c.bucket.root)
ref := elemRef{page: p, node: n}
ref.index = ref.count() - 1
c.stack = append(c.stack, ref)
c.last()
k, v, flags := c.keyValue()
if (flags & uint32(bucketLeafFlag)) != 0 {
return k, nil
}
return k, v
}
// Next moves the cursor to the next item in the bucket and returns its key and value.
// If the cursor is at the end of the bucket then a nil key and value are returned.
// The returned key and value are only valid for the life of the transaction.
func (c *Cursor) Next() (key []byte, value []byte) {
_assert(c.bucket.tx.db != nil, "tx closed")
k, v, flags := c.next()
if (flags & uint32(bucketLeafFlag)) != 0 {
return k, nil
}
return k, v
}
// Prev moves the cursor to the previous item in the bucket and returns its key and value.
// If the cursor is at the beginning of the bucket then a nil key and value are returned.
// The returned key and value are only valid for the life of the transaction.
func (c *Cursor) Prev() (key []byte, value []byte) {
_assert(c.bucket.tx.db != nil, "tx closed")
// Attempt to move back one element until we're successful.
// Move up the stack as we hit the beginning of each page in our stack.
for i := len(c.stack) - 1; i >= 0; i-- {
elem := &c.stack[i]
if elem.index > 0 {
elem.index--
break
}
c.stack = c.stack[:i]
}
// If we've hit the end then return nil.
if len(c.stack) == 0 {
return nil, nil
}
// Move down the stack to find the last element of the last leaf under this branch.
c.last()
k, v, flags := c.keyValue()
if (flags & uint32(bucketLeafFlag)) != 0 {
return k, nil
}
return k, v
}
// Seek moves the cursor to a given key and returns it.
// If the key does not exist then the next key is used. If no keys
// follow, a nil key is returned.
// The returned key and value are only valid for the life of the transaction.
func (c *Cursor) Seek(seek []byte) (key []byte, value []byte) {
k, v, flags := c.seek(seek)
// If we ended up after the last element of a page then move to the next one.
if ref := &c.stack[len(c.stack)-1]; ref.index >= ref.count() {
k, v, flags = c.next()
}
if k == nil {
return nil, nil
} else if (flags & uint32(bucketLeafFlag)) != 0 {
return k, nil
}
return k, v
}
// Delete removes the current key/value under the cursor from the bucket.
// Delete fails if current key/value is a bucket or if the transaction is not writable.
func (c *Cursor) Delete() error {
if c.bucket.tx.db == nil {
return ErrTxClosed
} else if !c.bucket.Writable() {
return ErrTxNotWritable
}
key, _, flags := c.keyValue()
// Return an error if current value is a bucket.
if (flags & bucketLeafFlag) != 0 {
return ErrIncompatibleValue
}
c.node().del(key)
return nil
}
// seek moves the cursor to a given key and returns it.
// If the key does not exist then the next key is used.
func (c *Cursor) seek(seek []byte) (key []byte, value []byte, flags uint32) {
_assert(c.bucket.tx.db != nil, "tx closed")
// Start from root page/node and traverse to correct page.
c.stack = c.stack[:0]
c.search(seek, c.bucket.root)
ref := &c.stack[len(c.stack)-1]
// If the cursor is pointing to the end of page/node then return nil.
if ref.index >= ref.count() {
return nil, nil, 0
}
// If this is a bucket then return a nil value.
return c.keyValue()
}
// first moves the cursor to the first leaf element under the last page in the stack.
func (c *Cursor) first() {
for {
// Exit when we hit a leaf page.
var ref = &c.stack[len(c.stack)-1]
if ref.isLeaf() {
break
}
// Keep adding pages pointing to the first element to the stack.
var pgid pgid
if ref.node != nil {
pgid = ref.node.inodes[ref.index].pgid
} else {
pgid = ref.page.branchPageElement(uint16(ref.index)).pgid
}
p, n := c.bucket.pageNode(pgid)
c.stack = append(c.stack, elemRef{page: p, node: n, index: 0})
}
}
// last moves the cursor to the last leaf element under the last page in the stack.
func (c *Cursor) last() {
for {
// Exit when we hit a leaf page.
ref := &c.stack[len(c.stack)-1]
if ref.isLeaf() {
break
}
// Keep adding pages pointing to the last element in the stack.
var pgid pgid
if ref.node != nil {
pgid = ref.node.inodes[ref.index].pgid
} else {
pgid = ref.page.branchPageElement(uint16(ref.index)).pgid
}
p, n := c.bucket.pageNode(pgid)
var nextRef = elemRef{page: p, node: n}
nextRef.index = nextRef.count() - 1
c.stack = append(c.stack, nextRef)
}
}
// next moves to the next leaf element and returns the key and value.
// If the cursor is at the last leaf element then it stays there and returns nil.
func (c *Cursor) next() (key []byte, value []byte, flags uint32) {
// Attempt to move over one element until we're successful.
// Move up the stack as we hit the end of each page in our stack.
var i int
for i = len(c.stack) - 1; i >= 0; i-- {
elem := &c.stack[i]
if elem.index < elem.count()-1 {
elem.index++
break
}
}
// If we've hit the root page then stop and return. This will leave the
// cursor on the last element of the last page.
if i == -1 {
return nil, nil, 0
}
// Otherwise start from where we left off in the stack and find the
// first element of the first leaf page.
c.stack = c.stack[:i+1]
c.first()
return c.keyValue()
}
// search recursively performs a binary search against a given page/node until it finds a given key.
func (c *Cursor) search(key []byte, pgid pgid) {
p, n := c.bucket.pageNode(pgid)
if p != nil && (p.flags&(branchPageFlag|leafPageFlag)) == 0 {
panic(fmt.Sprintf("invalid page type: %d: %x", p.id, p.flags))
}
e := elemRef{page: p, node: n}
c.stack = append(c.stack, e)
// If we're on a leaf page/node then find the specific node.
if e.isLeaf() {
c.nsearch(key)
return
}
if n != nil {
c.searchNode(key, n)
return
}
c.searchPage(key, p)
}
func (c *Cursor) searchNode(key []byte, n *node) {
var exact bool
index := sort.Search(len(n.inodes), func(i int) bool {
// TODO(benbjohnson): Optimize this range search. It's a bit hacky right now.
// sort.Search() finds the lowest index where f() != -1 but we need the highest index.
ret := bytes.Compare(n.inodes[i].key, key)
if ret == 0 {
exact = true
}
return ret != -1
})
if !exact && index > 0 {
index--
}
c.stack[len(c.stack)-1].index = index
// Recursively search to the next page.
c.search(key, n.inodes[index].pgid)
}
func (c *Cursor) searchPage(key []byte, p *page) {
// Binary search for the correct range.
inodes := p.branchPageElements()
var exact bool
index := sort.Search(int(p.count), func(i int) bool {
// TODO(benbjohnson): Optimize this range search. It's a bit hacky right now.
// sort.Search() finds the lowest index where f() != -1 but we need the highest index.
ret := bytes.Compare(inodes[i].key(), key)
if ret == 0 {
exact = true
}
return ret != -1
})
if !exact && index > 0 {
index--
}
c.stack[len(c.stack)-1].index = index
// Recursively search to the next page.
c.search(key, inodes[index].pgid)
}
// nsearch searches the leaf node on the top of the stack for a key.
func (c *Cursor) nsearch(key []byte) {
e := &c.stack[len(c.stack)-1]
p, n := e.page, e.node
// If we have a node then search its inodes.
if n != nil {
index := sort.Search(len(n.inodes), func(i int) bool {
return bytes.Compare(n.inodes[i].key, key) != -1
})
e.index = index
return
}
// If we have a page then search its leaf elements.
inodes := p.leafPageElements()
index := sort.Search(int(p.count), func(i int) bool {
return bytes.Compare(inodes[i].key(), key) != -1
})
e.index = index
}
// keyValue returns the key and value of the current leaf element.
func (c *Cursor) keyValue() ([]byte, []byte, uint32) {
ref := &c.stack[len(c.stack)-1]
if ref.count() == 0 || ref.index >= ref.count() {
return nil, nil, 0
}
// Retrieve value from node.
if ref.node != nil {
inode := &ref.node.inodes[ref.index]
return inode.key, inode.value, inode.flags
}
// Or retrieve value from page.
elem := ref.page.leafPageElement(uint16(ref.index))
return elem.key(), elem.value(), elem.flags
}
// node returns the node that the cursor is currently positioned on.
func (c *Cursor) node() *node {
_assert(len(c.stack) > 0, "accessing a node with a zero-length cursor stack")
// If the top of the stack is a leaf node then just return it.
if ref := &c.stack[len(c.stack)-1]; ref.node != nil && ref.isLeaf() {
return ref.node
}
// Start from root and traverse down the hierarchy.
var n = c.stack[0].node
if n == nil {
n = c.bucket.node(c.stack[0].page.id, nil)
}
for _, ref := range c.stack[:len(c.stack)-1] {
_assert(!n.isLeaf, "expected branch node")
n = n.childAt(int(ref.index))
}
_assert(n.isLeaf, "expected leaf node")
return n
}
// elemRef represents a reference to an element on a given page/node.
type elemRef struct {
page *page
node *node
index int
}
// isLeaf returns whether the ref is pointing at a leaf page/node.
func (r *elemRef) isLeaf() bool {
if r.node != nil {
return r.node.isLeaf
}
return (r.page.flags & leafPageFlag) != 0
}
// count returns the number of inodes or page elements.
func (r *elemRef) count() int {
if r.node != nil {
return len(r.node.inodes)
}
return int(r.page.count)
}

View File

@@ -0,0 +1,511 @@
package bolt_test
import (
"bytes"
"encoding/binary"
"fmt"
"os"
"sort"
"testing"
"testing/quick"
"github.com/coreos/etcd/Godeps/_workspace/src/github.com/boltdb/bolt"
)
// Ensure that a cursor can return a reference to the bucket that created it.
func TestCursor_Bucket(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
b, _ := tx.CreateBucket([]byte("widgets"))
c := b.Cursor()
equals(t, b, c.Bucket())
return nil
})
}
// Ensure that a Tx cursor can seek to the appropriate keys.
func TestCursor_Seek(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
b, err := tx.CreateBucket([]byte("widgets"))
ok(t, err)
ok(t, b.Put([]byte("foo"), []byte("0001")))
ok(t, b.Put([]byte("bar"), []byte("0002")))
ok(t, b.Put([]byte("baz"), []byte("0003")))
_, err = b.CreateBucket([]byte("bkt"))
ok(t, err)
return nil
})
db.View(func(tx *bolt.Tx) error {
c := tx.Bucket([]byte("widgets")).Cursor()
// Exact match should go to the key.
k, v := c.Seek([]byte("bar"))
equals(t, []byte("bar"), k)
equals(t, []byte("0002"), v)
// Inexact match should go to the next key.
k, v = c.Seek([]byte("bas"))
equals(t, []byte("baz"), k)
equals(t, []byte("0003"), v)
// Low key should go to the first key.
k, v = c.Seek([]byte(""))
equals(t, []byte("bar"), k)
equals(t, []byte("0002"), v)
// High key should return no key.
k, v = c.Seek([]byte("zzz"))
assert(t, k == nil, "")
assert(t, v == nil, "")
// Buckets should return their key but no value.
k, v = c.Seek([]byte("bkt"))
equals(t, []byte("bkt"), k)
assert(t, v == nil, "")
return nil
})
}
func TestCursor_Delete(t *testing.T) {
db := NewTestDB()
defer db.Close()
var count = 1000
// Insert every other key between 0 and $count.
db.Update(func(tx *bolt.Tx) error {
b, _ := tx.CreateBucket([]byte("widgets"))
for i := 0; i < count; i += 1 {
k := make([]byte, 8)
binary.BigEndian.PutUint64(k, uint64(i))
b.Put(k, make([]byte, 100))
}
b.CreateBucket([]byte("sub"))
return nil
})
db.Update(func(tx *bolt.Tx) error {
c := tx.Bucket([]byte("widgets")).Cursor()
bound := make([]byte, 8)
binary.BigEndian.PutUint64(bound, uint64(count/2))
for key, _ := c.First(); bytes.Compare(key, bound) < 0; key, _ = c.Next() {
if err := c.Delete(); err != nil {
return err
}
}
c.Seek([]byte("sub"))
err := c.Delete()
equals(t, err, bolt.ErrIncompatibleValue)
return nil
})
db.View(func(tx *bolt.Tx) error {
b := tx.Bucket([]byte("widgets"))
equals(t, b.Stats().KeyN, count/2+1)
return nil
})
}
// Ensure that a Tx cursor can seek to the appropriate keys when there are a
// large number of keys. This test also checks that seek will always move
// forward to the next key.
//
// Related: https://github.com/boltdb/bolt/pull/187
func TestCursor_Seek_Large(t *testing.T) {
db := NewTestDB()
defer db.Close()
var count = 10000
// Insert every other key between 0 and $count.
db.Update(func(tx *bolt.Tx) error {
b, _ := tx.CreateBucket([]byte("widgets"))
for i := 0; i < count; i += 100 {
for j := i; j < i+100; j += 2 {
k := make([]byte, 8)
binary.BigEndian.PutUint64(k, uint64(j))
b.Put(k, make([]byte, 100))
}
}
return nil
})
db.View(func(tx *bolt.Tx) error {
c := tx.Bucket([]byte("widgets")).Cursor()
for i := 0; i < count; i++ {
seek := make([]byte, 8)
binary.BigEndian.PutUint64(seek, uint64(i))
k, _ := c.Seek(seek)
// The last seek is beyond the end of the the range so
// it should return nil.
if i == count-1 {
assert(t, k == nil, "")
continue
}
// Otherwise we should seek to the exact key or the next key.
num := binary.BigEndian.Uint64(k)
if i%2 == 0 {
equals(t, uint64(i), num)
} else {
equals(t, uint64(i+1), num)
}
}
return nil
})
}
// Ensure that a cursor can iterate over an empty bucket without error.
func TestCursor_EmptyBucket(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
_, err := tx.CreateBucket([]byte("widgets"))
return err
})
db.View(func(tx *bolt.Tx) error {
c := tx.Bucket([]byte("widgets")).Cursor()
k, v := c.First()
assert(t, k == nil, "")
assert(t, v == nil, "")
return nil
})
}
// Ensure that a Tx cursor can reverse iterate over an empty bucket without error.
func TestCursor_EmptyBucketReverse(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
_, err := tx.CreateBucket([]byte("widgets"))
return err
})
db.View(func(tx *bolt.Tx) error {
c := tx.Bucket([]byte("widgets")).Cursor()
k, v := c.Last()
assert(t, k == nil, "")
assert(t, v == nil, "")
return nil
})
}
// Ensure that a Tx cursor can iterate over a single root with a couple elements.
func TestCursor_Iterate_Leaf(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
tx.Bucket([]byte("widgets")).Put([]byte("baz"), []byte{})
tx.Bucket([]byte("widgets")).Put([]byte("foo"), []byte{0})
tx.Bucket([]byte("widgets")).Put([]byte("bar"), []byte{1})
return nil
})
tx, _ := db.Begin(false)
c := tx.Bucket([]byte("widgets")).Cursor()
k, v := c.First()
equals(t, string(k), "bar")
equals(t, v, []byte{1})
k, v = c.Next()
equals(t, string(k), "baz")
equals(t, v, []byte{})
k, v = c.Next()
equals(t, string(k), "foo")
equals(t, v, []byte{0})
k, v = c.Next()
assert(t, k == nil, "")
assert(t, v == nil, "")
k, v = c.Next()
assert(t, k == nil, "")
assert(t, v == nil, "")
tx.Rollback()
}
// Ensure that a Tx cursor can iterate in reverse over a single root with a couple elements.
func TestCursor_LeafRootReverse(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
tx.Bucket([]byte("widgets")).Put([]byte("baz"), []byte{})
tx.Bucket([]byte("widgets")).Put([]byte("foo"), []byte{0})
tx.Bucket([]byte("widgets")).Put([]byte("bar"), []byte{1})
return nil
})
tx, _ := db.Begin(false)
c := tx.Bucket([]byte("widgets")).Cursor()
k, v := c.Last()
equals(t, string(k), "foo")
equals(t, v, []byte{0})
k, v = c.Prev()
equals(t, string(k), "baz")
equals(t, v, []byte{})
k, v = c.Prev()
equals(t, string(k), "bar")
equals(t, v, []byte{1})
k, v = c.Prev()
assert(t, k == nil, "")
assert(t, v == nil, "")
k, v = c.Prev()
assert(t, k == nil, "")
assert(t, v == nil, "")
tx.Rollback()
}
// Ensure that a Tx cursor can restart from the beginning.
func TestCursor_Restart(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
tx.Bucket([]byte("widgets")).Put([]byte("bar"), []byte{})
tx.Bucket([]byte("widgets")).Put([]byte("foo"), []byte{})
return nil
})
tx, _ := db.Begin(false)
c := tx.Bucket([]byte("widgets")).Cursor()
k, _ := c.First()
equals(t, string(k), "bar")
k, _ = c.Next()
equals(t, string(k), "foo")
k, _ = c.First()
equals(t, string(k), "bar")
k, _ = c.Next()
equals(t, string(k), "foo")
tx.Rollback()
}
// Ensure that a Tx can iterate over all elements in a bucket.
func TestCursor_QuickCheck(t *testing.T) {
f := func(items testdata) bool {
db := NewTestDB()
defer db.Close()
// Bulk insert all values.
tx, _ := db.Begin(true)
tx.CreateBucket([]byte("widgets"))
b := tx.Bucket([]byte("widgets"))
for _, item := range items {
ok(t, b.Put(item.Key, item.Value))
}
ok(t, tx.Commit())
// Sort test data.
sort.Sort(items)
// Iterate over all items and check consistency.
var index = 0
tx, _ = db.Begin(false)
c := tx.Bucket([]byte("widgets")).Cursor()
for k, v := c.First(); k != nil && index < len(items); k, v = c.Next() {
equals(t, k, items[index].Key)
equals(t, v, items[index].Value)
index++
}
equals(t, len(items), index)
tx.Rollback()
return true
}
if err := quick.Check(f, qconfig()); err != nil {
t.Error(err)
}
}
// Ensure that a transaction can iterate over all elements in a bucket in reverse.
func TestCursor_QuickCheck_Reverse(t *testing.T) {
f := func(items testdata) bool {
db := NewTestDB()
defer db.Close()
// Bulk insert all values.
tx, _ := db.Begin(true)
tx.CreateBucket([]byte("widgets"))
b := tx.Bucket([]byte("widgets"))
for _, item := range items {
ok(t, b.Put(item.Key, item.Value))
}
ok(t, tx.Commit())
// Sort test data.
sort.Sort(revtestdata(items))
// Iterate over all items and check consistency.
var index = 0
tx, _ = db.Begin(false)
c := tx.Bucket([]byte("widgets")).Cursor()
for k, v := c.Last(); k != nil && index < len(items); k, v = c.Prev() {
equals(t, k, items[index].Key)
equals(t, v, items[index].Value)
index++
}
equals(t, len(items), index)
tx.Rollback()
return true
}
if err := quick.Check(f, qconfig()); err != nil {
t.Error(err)
}
}
// Ensure that a Tx cursor can iterate over subbuckets.
func TestCursor_QuickCheck_BucketsOnly(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
b, err := tx.CreateBucket([]byte("widgets"))
ok(t, err)
_, err = b.CreateBucket([]byte("foo"))
ok(t, err)
_, err = b.CreateBucket([]byte("bar"))
ok(t, err)
_, err = b.CreateBucket([]byte("baz"))
ok(t, err)
return nil
})
db.View(func(tx *bolt.Tx) error {
var names []string
c := tx.Bucket([]byte("widgets")).Cursor()
for k, v := c.First(); k != nil; k, v = c.Next() {
names = append(names, string(k))
assert(t, v == nil, "")
}
equals(t, names, []string{"bar", "baz", "foo"})
return nil
})
}
// Ensure that a Tx cursor can reverse iterate over subbuckets.
func TestCursor_QuickCheck_BucketsOnly_Reverse(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
b, err := tx.CreateBucket([]byte("widgets"))
ok(t, err)
_, err = b.CreateBucket([]byte("foo"))
ok(t, err)
_, err = b.CreateBucket([]byte("bar"))
ok(t, err)
_, err = b.CreateBucket([]byte("baz"))
ok(t, err)
return nil
})
db.View(func(tx *bolt.Tx) error {
var names []string
c := tx.Bucket([]byte("widgets")).Cursor()
for k, v := c.Last(); k != nil; k, v = c.Prev() {
names = append(names, string(k))
assert(t, v == nil, "")
}
equals(t, names, []string{"foo", "baz", "bar"})
return nil
})
}
func ExampleCursor() {
// Open the database.
db, _ := bolt.Open(tempfile(), 0666, nil)
defer os.Remove(db.Path())
defer db.Close()
// Start a read-write transaction.
db.Update(func(tx *bolt.Tx) error {
// Create a new bucket.
tx.CreateBucket([]byte("animals"))
// Insert data into a bucket.
b := tx.Bucket([]byte("animals"))
b.Put([]byte("dog"), []byte("fun"))
b.Put([]byte("cat"), []byte("lame"))
b.Put([]byte("liger"), []byte("awesome"))
// Create a cursor for iteration.
c := b.Cursor()
// Iterate over items in sorted key order. This starts from the
// first key/value pair and updates the k/v variables to the
// next key/value on each iteration.
//
// The loop finishes at the end of the cursor when a nil key is returned.
for k, v := c.First(); k != nil; k, v = c.Next() {
fmt.Printf("A %s is %s.\n", k, v)
}
return nil
})
// Output:
// A cat is lame.
// A dog is fun.
// A liger is awesome.
}
func ExampleCursor_reverse() {
// Open the database.
db, _ := bolt.Open(tempfile(), 0666, nil)
defer os.Remove(db.Path())
defer db.Close()
// Start a read-write transaction.
db.Update(func(tx *bolt.Tx) error {
// Create a new bucket.
tx.CreateBucket([]byte("animals"))
// Insert data into a bucket.
b := tx.Bucket([]byte("animals"))
b.Put([]byte("dog"), []byte("fun"))
b.Put([]byte("cat"), []byte("lame"))
b.Put([]byte("liger"), []byte("awesome"))
// Create a cursor for iteration.
c := b.Cursor()
// Iterate over items in reverse sorted key order. This starts
// from the last key/value pair and updates the k/v variables to
// the previous key/value on each iteration.
//
// The loop finishes at the beginning of the cursor when a nil key
// is returned.
for k, v := c.Last(); k != nil; k, v = c.Prev() {
fmt.Printf("A %s is %s.\n", k, v)
}
return nil
})
// Output:
// A liger is awesome.
// A dog is fun.
// A cat is lame.
}

732
Godeps/_workspace/src/github.com/boltdb/bolt/db.go generated vendored Normal file
View File

@@ -0,0 +1,732 @@
package bolt
import (
"fmt"
"hash/fnv"
"os"
"runtime"
"runtime/debug"
"strings"
"sync"
"time"
"unsafe"
)
// The largest step that can be taken when remapping the mmap.
const maxMmapStep = 1 << 30 // 1GB
// The data file format version.
const version = 2
// Represents a marker value to indicate that a file is a Bolt DB.
const magic uint32 = 0xED0CDAED
// IgnoreNoSync specifies whether the NoSync field of a DB is ignored when
// syncing changes to a file. This is required as some operating systems,
// such as OpenBSD, do not have a unified buffer cache (UBC) and writes
// must be synchronzied using the msync(2) syscall.
const IgnoreNoSync = runtime.GOOS == "openbsd"
// Default values if not set in a DB instance.
const (
DefaultMaxBatchSize int = 1000
DefaultMaxBatchDelay = 10 * time.Millisecond
)
// DB represents a collection of buckets persisted to a file on disk.
// All data access is performed through transactions which can be obtained through the DB.
// All the functions on DB will return a ErrDatabaseNotOpen if accessed before Open() is called.
type DB struct {
// When enabled, the database will perform a Check() after every commit.
// A panic is issued if the database is in an inconsistent state. This
// flag has a large performance impact so it should only be used for
// debugging purposes.
StrictMode bool
// Setting the NoSync flag will cause the database to skip fsync()
// calls after each commit. This can be useful when bulk loading data
// into a database and you can restart the bulk load in the event of
// a system failure or database corruption. Do not set this flag for
// normal use.
//
// If the package global IgnoreNoSync constant is true, this value is
// ignored. See the comment on that constant for more details.
//
// THIS IS UNSAFE. PLEASE USE WITH CAUTION.
NoSync bool
// MaxBatchSize is the maximum size of a batch. Default value is
// copied from DefaultMaxBatchSize in Open.
//
// If <=0, disables batching.
//
// Do not change concurrently with calls to Batch.
MaxBatchSize int
// MaxBatchDelay is the maximum delay before a batch starts.
// Default value is copied from DefaultMaxBatchDelay in Open.
//
// If <=0, effectively disables batching.
//
// Do not change concurrently with calls to Batch.
MaxBatchDelay time.Duration
path string
file *os.File
dataref []byte // mmap'ed readonly, write throws SEGV
data *[maxMapSize]byte
datasz int
meta0 *meta
meta1 *meta
pageSize int
opened bool
rwtx *Tx
txs []*Tx
freelist *freelist
stats Stats
batchMu sync.Mutex
batch *batch
rwlock sync.Mutex // Allows only one writer at a time.
metalock sync.Mutex // Protects meta page access.
mmaplock sync.RWMutex // Protects mmap access during remapping.
statlock sync.RWMutex // Protects stats access.
ops struct {
writeAt func(b []byte, off int64) (n int, err error)
}
}
// Path returns the path to currently open database file.
func (db *DB) Path() string {
return db.path
}
// GoString returns the Go string representation of the database.
func (db *DB) GoString() string {
return fmt.Sprintf("bolt.DB{path:%q}", db.path)
}
// String returns the string representation of the database.
func (db *DB) String() string {
return fmt.Sprintf("DB<%q>", db.path)
}
// Open creates and opens a database at the given path.
// If the file does not exist then it will be created automatically.
// Passing in nil options will cause Bolt to open the database with the default options.
func Open(path string, mode os.FileMode, options *Options) (*DB, error) {
var db = &DB{opened: true}
// Set default options if no options are provided.
if options == nil {
options = DefaultOptions
}
// Set default values for later DB operations.
db.MaxBatchSize = DefaultMaxBatchSize
db.MaxBatchDelay = DefaultMaxBatchDelay
// Open data file and separate sync handler for metadata writes.
db.path = path
var err error
if db.file, err = os.OpenFile(db.path, os.O_RDWR|os.O_CREATE, mode); err != nil {
_ = db.close()
return nil, err
}
// Lock file so that other processes using Bolt cannot use the database
// at the same time. This would cause corruption since the two processes
// would write meta pages and free pages separately.
if err := flock(db.file, options.Timeout); err != nil {
_ = db.close()
return nil, err
}
// Default values for test hooks
db.ops.writeAt = db.file.WriteAt
// Initialize the database if it doesn't exist.
if info, err := db.file.Stat(); err != nil {
return nil, fmt.Errorf("stat error: %s", err)
} else if info.Size() == 0 {
// Initialize new files with meta pages.
if err := db.init(); err != nil {
return nil, err
}
} else {
// Read the first meta page to determine the page size.
var buf [0x1000]byte
if _, err := db.file.ReadAt(buf[:], 0); err == nil {
m := db.pageInBuffer(buf[:], 0).meta()
if err := m.validate(); err != nil {
return nil, fmt.Errorf("meta0 error: %s", err)
}
db.pageSize = int(m.pageSize)
}
}
// Memory map the data file.
if err := db.mmap(0); err != nil {
_ = db.close()
return nil, err
}
// Read in the freelist.
db.freelist = newFreelist()
db.freelist.read(db.page(db.meta().freelist))
// Mark the database as opened and return.
return db, nil
}
// mmap opens the underlying memory-mapped file and initializes the meta references.
// minsz is the minimum size that the new mmap can be.
func (db *DB) mmap(minsz int) error {
db.mmaplock.Lock()
defer db.mmaplock.Unlock()
info, err := db.file.Stat()
if err != nil {
return fmt.Errorf("mmap stat error: %s", err)
} else if int(info.Size()) < db.pageSize*2 {
return fmt.Errorf("file size too small")
}
// Ensure the size is at least the minimum size.
var size = int(info.Size())
if size < minsz {
size = minsz
}
size, err = db.mmapSize(size)
if err != nil {
return err
}
// Dereference all mmap references before unmapping.
if db.rwtx != nil {
db.rwtx.root.dereference()
}
// Unmap existing data before continuing.
if err := db.munmap(); err != nil {
return err
}
// Memory-map the data file as a byte slice.
if err := mmap(db, size); err != nil {
return err
}
// Save references to the meta pages.
db.meta0 = db.page(0).meta()
db.meta1 = db.page(1).meta()
// Validate the meta pages.
if err := db.meta0.validate(); err != nil {
return fmt.Errorf("meta0 error: %s", err)
}
if err := db.meta1.validate(); err != nil {
return fmt.Errorf("meta1 error: %s", err)
}
return nil
}
// munmap unmaps the data file from memory.
func (db *DB) munmap() error {
if err := munmap(db); err != nil {
return fmt.Errorf("unmap error: " + err.Error())
}
return nil
}
// mmapSize determines the appropriate size for the mmap given the current size
// of the database. The minimum size is 1MB and doubles until it reaches 1GB.
// Returns an error if the new mmap size is greater than the max allowed.
func (db *DB) mmapSize(size int) (int, error) {
// Double the size from 1MB until 1GB.
for i := uint(20); i <= 30; i++ {
if size <= 1<<i {
return 1 << i, nil
}
}
// Verify the requested size is not above the maximum allowed.
if size > maxMapSize {
return 0, fmt.Errorf("mmap too large")
}
// If larger than 1GB then grow by 1GB at a time.
sz := int64(size)
if remainder := sz % int64(maxMmapStep); remainder > 0 {
sz += int64(maxMmapStep) - remainder
}
// Ensure that the mmap size is a multiple of the page size.
// This should always be true since we're incrementing in MBs.
pageSize := int64(db.pageSize)
if (sz % pageSize) != 0 {
sz = ((sz / pageSize) + 1) * pageSize
}
// If we've exceeded the max size then only grow up to the max size.
if sz > maxMapSize {
sz = maxMapSize
}
return int(sz), nil
}
// init creates a new database file and initializes its meta pages.
func (db *DB) init() error {
// Set the page size to the OS page size.
db.pageSize = os.Getpagesize()
// Create two meta pages on a buffer.
buf := make([]byte, db.pageSize*4)
for i := 0; i < 2; i++ {
p := db.pageInBuffer(buf[:], pgid(i))
p.id = pgid(i)
p.flags = metaPageFlag
// Initialize the meta page.
m := p.meta()
m.magic = magic
m.version = version
m.pageSize = uint32(db.pageSize)
m.freelist = 2
m.root = bucket{root: 3}
m.pgid = 4
m.txid = txid(i)
}
// Write an empty freelist at page 3.
p := db.pageInBuffer(buf[:], pgid(2))
p.id = pgid(2)
p.flags = freelistPageFlag
p.count = 0
// Write an empty leaf page at page 4.
p = db.pageInBuffer(buf[:], pgid(3))
p.id = pgid(3)
p.flags = leafPageFlag
p.count = 0
// Write the buffer to our data file.
if _, err := db.ops.writeAt(buf, 0); err != nil {
return err
}
if err := fdatasync(db); err != nil {
return err
}
return nil
}
// Close releases all database resources.
// All transactions must be closed before closing the database.
func (db *DB) Close() error {
db.metalock.Lock()
defer db.metalock.Unlock()
return db.close()
}
func (db *DB) close() error {
db.opened = false
db.freelist = nil
db.path = ""
// Clear ops.
db.ops.writeAt = nil
// Close the mmap.
if err := db.munmap(); err != nil {
return err
}
// Close file handles.
if db.file != nil {
// Unlock the file.
_ = funlock(db.file)
// Close the file descriptor.
if err := db.file.Close(); err != nil {
return fmt.Errorf("db file close: %s", err)
}
db.file = nil
}
return nil
}
// Begin starts a new transaction.
// Multiple read-only transactions can be used concurrently but only one
// write transaction can be used at a time. Starting multiple write transactions
// will cause the calls to block and be serialized until the current write
// transaction finishes.
//
// IMPORTANT: You must close read-only transactions after you are finished or
// else the database will not reclaim old pages.
func (db *DB) Begin(writable bool) (*Tx, error) {
if writable {
return db.beginRWTx()
}
return db.beginTx()
}
func (db *DB) beginTx() (*Tx, error) {
// Lock the meta pages while we initialize the transaction. We obtain
// the meta lock before the mmap lock because that's the order that the
// write transaction will obtain them.
db.metalock.Lock()
// Obtain a read-only lock on the mmap. When the mmap is remapped it will
// obtain a write lock so all transactions must finish before it can be
// remapped.
db.mmaplock.RLock()
// Exit if the database is not open yet.
if !db.opened {
db.mmaplock.RUnlock()
db.metalock.Unlock()
return nil, ErrDatabaseNotOpen
}
// Create a transaction associated with the database.
t := &Tx{}
t.init(db)
// Keep track of transaction until it closes.
db.txs = append(db.txs, t)
n := len(db.txs)
// Unlock the meta pages.
db.metalock.Unlock()
// Update the transaction stats.
db.statlock.Lock()
db.stats.TxN++
db.stats.OpenTxN = n
db.statlock.Unlock()
return t, nil
}
func (db *DB) beginRWTx() (*Tx, error) {
// Obtain writer lock. This is released by the transaction when it closes.
// This enforces only one writer transaction at a time.
db.rwlock.Lock()
// Once we have the writer lock then we can lock the meta pages so that
// we can set up the transaction.
db.metalock.Lock()
defer db.metalock.Unlock()
// Exit if the database is not open yet.
if !db.opened {
db.rwlock.Unlock()
return nil, ErrDatabaseNotOpen
}
// Create a transaction associated with the database.
t := &Tx{writable: true}
t.init(db)
db.rwtx = t
// Free any pages associated with closed read-only transactions.
var minid txid = 0xFFFFFFFFFFFFFFFF
for _, t := range db.txs {
if t.meta.txid < minid {
minid = t.meta.txid
}
}
if minid > 0 {
db.freelist.release(minid - 1)
}
return t, nil
}
// removeTx removes a transaction from the database.
func (db *DB) removeTx(tx *Tx) {
// Release the read lock on the mmap.
db.mmaplock.RUnlock()
// Use the meta lock to restrict access to the DB object.
db.metalock.Lock()
// Remove the transaction.
for i, t := range db.txs {
if t == tx {
db.txs = append(db.txs[:i], db.txs[i+1:]...)
break
}
}
n := len(db.txs)
// Unlock the meta pages.
db.metalock.Unlock()
// Merge statistics.
db.statlock.Lock()
db.stats.OpenTxN = n
db.stats.TxStats.add(&tx.stats)
db.statlock.Unlock()
}
// Update executes a function within the context of a read-write managed transaction.
// If no error is returned from the function then the transaction is committed.
// If an error is returned then the entire transaction is rolled back.
// Any error that is returned from the function or returned from the commit is
// returned from the Update() method.
//
// Attempting to manually commit or rollback within the function will cause a panic.
func (db *DB) Update(fn func(*Tx) error) error {
t, err := db.Begin(true)
if err != nil {
return err
}
// Make sure the transaction rolls back in the event of a panic.
defer func() {
if t.db != nil {
t.rollback()
}
}()
// Mark as a managed tx so that the inner function cannot manually commit.
t.managed = true
// If an error is returned from the function then rollback and return error.
err = fn(t)
t.managed = false
if err != nil {
_ = t.Rollback()
return err
}
return t.Commit()
}
// View executes a function within the context of a managed read-only transaction.
// Any error that is returned from the function is returned from the View() method.
//
// Attempting to manually rollback within the function will cause a panic.
func (db *DB) View(fn func(*Tx) error) error {
t, err := db.Begin(false)
if err != nil {
return err
}
// Make sure the transaction rolls back in the event of a panic.
defer func() {
if t.db != nil {
t.rollback()
}
}()
// Mark as a managed tx so that the inner function cannot manually rollback.
t.managed = true
// If an error is returned from the function then pass it through.
err = fn(t)
t.managed = false
if err != nil {
_ = t.Rollback()
return err
}
if err := t.Rollback(); err != nil {
return err
}
return nil
}
// Stats retrieves ongoing performance stats for the database.
// This is only updated when a transaction closes.
func (db *DB) Stats() Stats {
db.statlock.RLock()
defer db.statlock.RUnlock()
return db.stats
}
// This is for internal access to the raw data bytes from the C cursor, use
// carefully, or not at all.
func (db *DB) Info() *Info {
return &Info{uintptr(unsafe.Pointer(&db.data[0])), db.pageSize}
}
// page retrieves a page reference from the mmap based on the current page size.
func (db *DB) page(id pgid) *page {
pos := id * pgid(db.pageSize)
return (*page)(unsafe.Pointer(&db.data[pos]))
}
// pageInBuffer retrieves a page reference from a given byte array based on the current page size.
func (db *DB) pageInBuffer(b []byte, id pgid) *page {
return (*page)(unsafe.Pointer(&b[id*pgid(db.pageSize)]))
}
// meta retrieves the current meta page reference.
func (db *DB) meta() *meta {
if db.meta0.txid > db.meta1.txid {
return db.meta0
}
return db.meta1
}
// allocate returns a contiguous block of memory starting at a given page.
func (db *DB) allocate(count int) (*page, error) {
// Allocate a temporary buffer for the page.
buf := make([]byte, count*db.pageSize)
p := (*page)(unsafe.Pointer(&buf[0]))
p.overflow = uint32(count - 1)
// Use pages from the freelist if they are available.
if p.id = db.freelist.allocate(count); p.id != 0 {
return p, nil
}
// Resize mmap() if we're at the end.
p.id = db.rwtx.meta.pgid
var minsz = int((p.id+pgid(count))+1) * db.pageSize
if minsz >= db.datasz {
if err := db.mmap(minsz); err != nil {
return nil, fmt.Errorf("mmap allocate error: %s", err)
}
}
// Move the page id high water mark.
db.rwtx.meta.pgid += pgid(count)
return p, nil
}
// Options represents the options that can be set when opening a database.
type Options struct {
// Timeout is the amount of time to wait to obtain a file lock.
// When set to zero it will wait indefinitely. This option is only
// available on Darwin and Linux.
Timeout time.Duration
}
// DefaultOptions represent the options used if nil options are passed into Open().
// No timeout is used which will cause Bolt to wait indefinitely for a lock.
var DefaultOptions = &Options{
Timeout: 0,
}
// Stats represents statistics about the database.
type Stats struct {
// Freelist stats
FreePageN int // total number of free pages on the freelist
PendingPageN int // total number of pending pages on the freelist
FreeAlloc int // total bytes allocated in free pages
FreelistInuse int // total bytes used by the freelist
// Transaction stats
TxN int // total number of started read transactions
OpenTxN int // number of currently open read transactions
TxStats TxStats // global, ongoing stats.
}
// Sub calculates and returns the difference between two sets of database stats.
// This is useful when obtaining stats at two different points and time and
// you need the performance counters that occurred within that time span.
func (s *Stats) Sub(other *Stats) Stats {
if other == nil {
return *s
}
var diff Stats
diff.FreePageN = s.FreePageN
diff.PendingPageN = s.PendingPageN
diff.FreeAlloc = s.FreeAlloc
diff.FreelistInuse = s.FreelistInuse
diff.TxN = other.TxN - s.TxN
diff.TxStats = s.TxStats.Sub(&other.TxStats)
return diff
}
func (s *Stats) add(other *Stats) {
s.TxStats.add(&other.TxStats)
}
type Info struct {
Data uintptr
PageSize int
}
type meta struct {
magic uint32
version uint32
pageSize uint32
flags uint32
root bucket
freelist pgid
pgid pgid
txid txid
checksum uint64
}
// validate checks the marker bytes and version of the meta page to ensure it matches this binary.
func (m *meta) validate() error {
if m.checksum != 0 && m.checksum != m.sum64() {
return ErrChecksum
} else if m.magic != magic {
return ErrInvalid
} else if m.version != version {
return ErrVersionMismatch
}
return nil
}
// copy copies one meta object to another.
func (m *meta) copy(dest *meta) {
*dest = *m
}
// write writes the meta onto a page.
func (m *meta) write(p *page) {
if m.root.root >= m.pgid {
panic(fmt.Sprintf("root bucket pgid (%d) above high water mark (%d)", m.root.root, m.pgid))
} else if m.freelist >= m.pgid {
panic(fmt.Sprintf("freelist pgid (%d) above high water mark (%d)", m.freelist, m.pgid))
}
// Page id is either going to be 0 or 1 which we can determine by the transaction ID.
p.id = pgid(m.txid % 2)
p.flags |= metaPageFlag
// Calculate the checksum.
m.checksum = m.sum64()
m.copy(p.meta())
}
// generates the checksum for the meta.
func (m *meta) sum64() uint64 {
var h = fnv.New64a()
_, _ = h.Write((*[unsafe.Offsetof(meta{}.checksum)]byte)(unsafe.Pointer(m))[:])
return h.Sum64()
}
// _assert will panic with a given formatted message if the given condition is false.
func _assert(condition bool, msg string, v ...interface{}) {
if !condition {
panic(fmt.Sprintf("assertion failed: "+msg, v...))
}
}
func warn(v ...interface{}) { fmt.Fprintln(os.Stderr, v...) }
func warnf(msg string, v ...interface{}) { fmt.Fprintf(os.Stderr, msg+"\n", v...) }
func printstack() {
stack := strings.Join(strings.Split(string(debug.Stack()), "\n")[2:], "\n")
fmt.Fprintln(os.Stderr, stack)
}

790
Godeps/_workspace/src/github.com/boltdb/bolt/db_test.go generated vendored Normal file
View File

@@ -0,0 +1,790 @@
package bolt_test
import (
"encoding/binary"
"errors"
"flag"
"fmt"
"io/ioutil"
"os"
"regexp"
"runtime"
"sort"
"strings"
"testing"
"time"
"github.com/coreos/etcd/Godeps/_workspace/src/github.com/boltdb/bolt"
)
var statsFlag = flag.Bool("stats", false, "show performance stats")
// Ensure that opening a database with a bad path returns an error.
func TestOpen_BadPath(t *testing.T) {
db, err := bolt.Open("", 0666, nil)
assert(t, err != nil, "err: %s", err)
assert(t, db == nil, "")
}
// Ensure that a database can be opened without error.
func TestOpen(t *testing.T) {
path := tempfile()
defer os.Remove(path)
db, err := bolt.Open(path, 0666, nil)
assert(t, db != nil, "")
ok(t, err)
equals(t, db.Path(), path)
ok(t, db.Close())
}
// Ensure that opening an already open database file will timeout.
func TestOpen_Timeout(t *testing.T) {
if runtime.GOOS == "windows" {
t.Skip("timeout not supported on windows")
}
path := tempfile()
defer os.Remove(path)
// Open a data file.
db0, err := bolt.Open(path, 0666, nil)
assert(t, db0 != nil, "")
ok(t, err)
// Attempt to open the database again.
start := time.Now()
db1, err := bolt.Open(path, 0666, &bolt.Options{Timeout: 100 * time.Millisecond})
assert(t, db1 == nil, "")
equals(t, bolt.ErrTimeout, err)
assert(t, time.Since(start) > 100*time.Millisecond, "")
db0.Close()
}
// Ensure that opening an already open database file will wait until its closed.
func TestOpen_Wait(t *testing.T) {
if runtime.GOOS == "windows" {
t.Skip("timeout not supported on windows")
}
path := tempfile()
defer os.Remove(path)
// Open a data file.
db0, err := bolt.Open(path, 0666, nil)
assert(t, db0 != nil, "")
ok(t, err)
// Close it in just a bit.
time.AfterFunc(100*time.Millisecond, func() { db0.Close() })
// Attempt to open the database again.
start := time.Now()
db1, err := bolt.Open(path, 0666, &bolt.Options{Timeout: 200 * time.Millisecond})
assert(t, db1 != nil, "")
ok(t, err)
assert(t, time.Since(start) > 100*time.Millisecond, "")
}
// Ensure that opening a database does not increase its size.
// https://github.com/boltdb/bolt/issues/291
func TestOpen_Size(t *testing.T) {
// Open a data file.
db := NewTestDB()
path := db.Path()
defer db.Close()
// Insert until we get above the minimum 4MB size.
ok(t, db.Update(func(tx *bolt.Tx) error {
b, _ := tx.CreateBucketIfNotExists([]byte("data"))
for i := 0; i < 10000; i++ {
ok(t, b.Put([]byte(fmt.Sprintf("%04d", i)), make([]byte, 1000)))
}
return nil
}))
// Close database and grab the size.
db.DB.Close()
sz := fileSize(path)
if sz == 0 {
t.Fatalf("unexpected new file size: %d", sz)
}
// Reopen database, update, and check size again.
db0, err := bolt.Open(path, 0666, nil)
ok(t, err)
ok(t, db0.Update(func(tx *bolt.Tx) error { return tx.Bucket([]byte("data")).Put([]byte{0}, []byte{0}) }))
ok(t, db0.Close())
newSz := fileSize(path)
if newSz == 0 {
t.Fatalf("unexpected new file size: %d", newSz)
}
// Compare the original size with the new size.
if sz != newSz {
t.Fatalf("unexpected file growth: %d => %d", sz, newSz)
}
}
// Ensure that opening a database beyond the max step size does not increase its size.
// https://github.com/boltdb/bolt/issues/303
func TestOpen_Size_Large(t *testing.T) {
if testing.Short() {
t.Skip("short mode")
}
// Open a data file.
db := NewTestDB()
path := db.Path()
defer db.Close()
// Insert until we get above the minimum 4MB size.
var index uint64
for i := 0; i < 10000; i++ {
ok(t, db.Update(func(tx *bolt.Tx) error {
b, _ := tx.CreateBucketIfNotExists([]byte("data"))
for j := 0; j < 1000; j++ {
ok(t, b.Put(u64tob(index), make([]byte, 50)))
index++
}
return nil
}))
}
// Close database and grab the size.
db.DB.Close()
sz := fileSize(path)
if sz == 0 {
t.Fatalf("unexpected new file size: %d", sz)
} else if sz < (1 << 30) {
t.Fatalf("expected larger initial size: %d", sz)
}
// Reopen database, update, and check size again.
db0, err := bolt.Open(path, 0666, nil)
ok(t, err)
ok(t, db0.Update(func(tx *bolt.Tx) error { return tx.Bucket([]byte("data")).Put([]byte{0}, []byte{0}) }))
ok(t, db0.Close())
newSz := fileSize(path)
if newSz == 0 {
t.Fatalf("unexpected new file size: %d", newSz)
}
// Compare the original size with the new size.
if sz != newSz {
t.Fatalf("unexpected file growth: %d => %d", sz, newSz)
}
}
// Ensure that a re-opened database is consistent.
func TestOpen_Check(t *testing.T) {
path := tempfile()
defer os.Remove(path)
db, err := bolt.Open(path, 0666, nil)
ok(t, err)
ok(t, db.View(func(tx *bolt.Tx) error { return <-tx.Check() }))
db.Close()
db, err = bolt.Open(path, 0666, nil)
ok(t, err)
ok(t, db.View(func(tx *bolt.Tx) error { return <-tx.Check() }))
db.Close()
}
// Ensure that the database returns an error if the file handle cannot be open.
func TestDB_Open_FileError(t *testing.T) {
path := tempfile()
defer os.Remove(path)
_, err := bolt.Open(path+"/youre-not-my-real-parent", 0666, nil)
assert(t, err.(*os.PathError) != nil, "")
equals(t, path+"/youre-not-my-real-parent", err.(*os.PathError).Path)
equals(t, "open", err.(*os.PathError).Op)
}
// Ensure that write errors to the meta file handler during initialization are returned.
func TestDB_Open_MetaInitWriteError(t *testing.T) {
t.Skip("pending")
}
// Ensure that a database that is too small returns an error.
func TestDB_Open_FileTooSmall(t *testing.T) {
path := tempfile()
defer os.Remove(path)
db, err := bolt.Open(path, 0666, nil)
ok(t, err)
db.Close()
// corrupt the database
ok(t, os.Truncate(path, int64(os.Getpagesize())))
db, err = bolt.Open(path, 0666, nil)
equals(t, errors.New("file size too small"), err)
}
// TODO(benbjohnson): Test corruption at every byte of the first two pages.
// Ensure that a database cannot open a transaction when it's not open.
func TestDB_Begin_DatabaseNotOpen(t *testing.T) {
var db bolt.DB
tx, err := db.Begin(false)
assert(t, tx == nil, "")
equals(t, err, bolt.ErrDatabaseNotOpen)
}
// Ensure that a read-write transaction can be retrieved.
func TestDB_BeginRW(t *testing.T) {
db := NewTestDB()
defer db.Close()
tx, err := db.Begin(true)
assert(t, tx != nil, "")
ok(t, err)
assert(t, tx.DB() == db.DB, "")
equals(t, tx.Writable(), true)
ok(t, tx.Commit())
}
// Ensure that opening a transaction while the DB is closed returns an error.
func TestDB_BeginRW_Closed(t *testing.T) {
var db bolt.DB
tx, err := db.Begin(true)
equals(t, err, bolt.ErrDatabaseNotOpen)
assert(t, tx == nil, "")
}
// Ensure a database can provide a transactional block.
func TestDB_Update(t *testing.T) {
db := NewTestDB()
defer db.Close()
err := db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
b := tx.Bucket([]byte("widgets"))
b.Put([]byte("foo"), []byte("bar"))
b.Put([]byte("baz"), []byte("bat"))
b.Delete([]byte("foo"))
return nil
})
ok(t, err)
err = db.View(func(tx *bolt.Tx) error {
assert(t, tx.Bucket([]byte("widgets")).Get([]byte("foo")) == nil, "")
equals(t, []byte("bat"), tx.Bucket([]byte("widgets")).Get([]byte("baz")))
return nil
})
ok(t, err)
}
// Ensure a closed database returns an error while running a transaction block
func TestDB_Update_Closed(t *testing.T) {
var db bolt.DB
err := db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
return nil
})
equals(t, err, bolt.ErrDatabaseNotOpen)
}
// Ensure a panic occurs while trying to commit a managed transaction.
func TestDB_Update_ManualCommit(t *testing.T) {
db := NewTestDB()
defer db.Close()
var ok bool
db.Update(func(tx *bolt.Tx) error {
func() {
defer func() {
if r := recover(); r != nil {
ok = true
}
}()
tx.Commit()
}()
return nil
})
assert(t, ok, "expected panic")
}
// Ensure a panic occurs while trying to rollback a managed transaction.
func TestDB_Update_ManualRollback(t *testing.T) {
db := NewTestDB()
defer db.Close()
var ok bool
db.Update(func(tx *bolt.Tx) error {
func() {
defer func() {
if r := recover(); r != nil {
ok = true
}
}()
tx.Rollback()
}()
return nil
})
assert(t, ok, "expected panic")
}
// Ensure a panic occurs while trying to commit a managed transaction.
func TestDB_View_ManualCommit(t *testing.T) {
db := NewTestDB()
defer db.Close()
var ok bool
db.Update(func(tx *bolt.Tx) error {
func() {
defer func() {
if r := recover(); r != nil {
ok = true
}
}()
tx.Commit()
}()
return nil
})
assert(t, ok, "expected panic")
}
// Ensure a panic occurs while trying to rollback a managed transaction.
func TestDB_View_ManualRollback(t *testing.T) {
db := NewTestDB()
defer db.Close()
var ok bool
db.Update(func(tx *bolt.Tx) error {
func() {
defer func() {
if r := recover(); r != nil {
ok = true
}
}()
tx.Rollback()
}()
return nil
})
assert(t, ok, "expected panic")
}
// Ensure a write transaction that panics does not hold open locks.
func TestDB_Update_Panic(t *testing.T) {
db := NewTestDB()
defer db.Close()
func() {
defer func() {
if r := recover(); r != nil {
t.Log("recover: update", r)
}
}()
db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
panic("omg")
})
}()
// Verify we can update again.
err := db.Update(func(tx *bolt.Tx) error {
_, err := tx.CreateBucket([]byte("widgets"))
return err
})
ok(t, err)
// Verify that our change persisted.
err = db.Update(func(tx *bolt.Tx) error {
assert(t, tx.Bucket([]byte("widgets")) != nil, "")
return nil
})
}
// Ensure a database can return an error through a read-only transactional block.
func TestDB_View_Error(t *testing.T) {
db := NewTestDB()
defer db.Close()
err := db.View(func(tx *bolt.Tx) error {
return errors.New("xxx")
})
equals(t, errors.New("xxx"), err)
}
// Ensure a read transaction that panics does not hold open locks.
func TestDB_View_Panic(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
return nil
})
func() {
defer func() {
if r := recover(); r != nil {
t.Log("recover: view", r)
}
}()
db.View(func(tx *bolt.Tx) error {
assert(t, tx.Bucket([]byte("widgets")) != nil, "")
panic("omg")
})
}()
// Verify that we can still use read transactions.
db.View(func(tx *bolt.Tx) error {
assert(t, tx.Bucket([]byte("widgets")) != nil, "")
return nil
})
}
// Ensure that an error is returned when a database write fails.
func TestDB_Commit_WriteFail(t *testing.T) {
t.Skip("pending") // TODO(benbjohnson)
}
// Ensure that DB stats can be returned.
func TestDB_Stats(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
_, err := tx.CreateBucket([]byte("widgets"))
return err
})
stats := db.Stats()
equals(t, 2, stats.TxStats.PageCount)
equals(t, 0, stats.FreePageN)
equals(t, 2, stats.PendingPageN)
}
// Ensure that database pages are in expected order and type.
func TestDB_Consistency(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
_, err := tx.CreateBucket([]byte("widgets"))
return err
})
for i := 0; i < 10; i++ {
db.Update(func(tx *bolt.Tx) error {
ok(t, tx.Bucket([]byte("widgets")).Put([]byte("foo"), []byte("bar")))
return nil
})
}
db.Update(func(tx *bolt.Tx) error {
p, _ := tx.Page(0)
assert(t, p != nil, "")
equals(t, "meta", p.Type)
p, _ = tx.Page(1)
assert(t, p != nil, "")
equals(t, "meta", p.Type)
p, _ = tx.Page(2)
assert(t, p != nil, "")
equals(t, "free", p.Type)
p, _ = tx.Page(3)
assert(t, p != nil, "")
equals(t, "free", p.Type)
p, _ = tx.Page(4)
assert(t, p != nil, "")
equals(t, "leaf", p.Type)
p, _ = tx.Page(5)
assert(t, p != nil, "")
equals(t, "freelist", p.Type)
p, _ = tx.Page(6)
assert(t, p == nil, "")
return nil
})
}
// Ensure that DB stats can be substracted from one another.
func TestDBStats_Sub(t *testing.T) {
var a, b bolt.Stats
a.TxStats.PageCount = 3
a.FreePageN = 4
b.TxStats.PageCount = 10
b.FreePageN = 14
diff := b.Sub(&a)
equals(t, 7, diff.TxStats.PageCount)
// free page stats are copied from the receiver and not subtracted
equals(t, 14, diff.FreePageN)
}
func ExampleDB_Update() {
// Open the database.
db, _ := bolt.Open(tempfile(), 0666, nil)
defer os.Remove(db.Path())
defer db.Close()
// Execute several commands within a write transaction.
err := db.Update(func(tx *bolt.Tx) error {
b, err := tx.CreateBucket([]byte("widgets"))
if err != nil {
return err
}
if err := b.Put([]byte("foo"), []byte("bar")); err != nil {
return err
}
return nil
})
// If our transactional block didn't return an error then our data is saved.
if err == nil {
db.View(func(tx *bolt.Tx) error {
value := tx.Bucket([]byte("widgets")).Get([]byte("foo"))
fmt.Printf("The value of 'foo' is: %s\n", value)
return nil
})
}
// Output:
// The value of 'foo' is: bar
}
func ExampleDB_View() {
// Open the database.
db, _ := bolt.Open(tempfile(), 0666, nil)
defer os.Remove(db.Path())
defer db.Close()
// Insert data into a bucket.
db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("people"))
b := tx.Bucket([]byte("people"))
b.Put([]byte("john"), []byte("doe"))
b.Put([]byte("susy"), []byte("que"))
return nil
})
// Access data from within a read-only transactional block.
db.View(func(tx *bolt.Tx) error {
v := tx.Bucket([]byte("people")).Get([]byte("john"))
fmt.Printf("John's last name is %s.\n", v)
return nil
})
// Output:
// John's last name is doe.
}
func ExampleDB_Begin_ReadOnly() {
// Open the database.
db, _ := bolt.Open(tempfile(), 0666, nil)
defer os.Remove(db.Path())
defer db.Close()
// Create a bucket.
db.Update(func(tx *bolt.Tx) error {
_, err := tx.CreateBucket([]byte("widgets"))
return err
})
// Create several keys in a transaction.
tx, _ := db.Begin(true)
b := tx.Bucket([]byte("widgets"))
b.Put([]byte("john"), []byte("blue"))
b.Put([]byte("abby"), []byte("red"))
b.Put([]byte("zephyr"), []byte("purple"))
tx.Commit()
// Iterate over the values in sorted key order.
tx, _ = db.Begin(false)
c := tx.Bucket([]byte("widgets")).Cursor()
for k, v := c.First(); k != nil; k, v = c.Next() {
fmt.Printf("%s likes %s\n", k, v)
}
tx.Rollback()
// Output:
// abby likes red
// john likes blue
// zephyr likes purple
}
// TestDB represents a wrapper around a Bolt DB to handle temporary file
// creation and automatic cleanup on close.
type TestDB struct {
*bolt.DB
}
// NewTestDB returns a new instance of TestDB.
func NewTestDB() *TestDB {
db, err := bolt.Open(tempfile(), 0666, nil)
if err != nil {
panic("cannot open db: " + err.Error())
}
return &TestDB{db}
}
// MustView executes a read-only function. Panic on error.
func (db *TestDB) MustView(fn func(tx *bolt.Tx) error) {
if err := db.DB.View(func(tx *bolt.Tx) error {
return fn(tx)
}); err != nil {
panic(err.Error())
}
}
// MustUpdate executes a read-write function. Panic on error.
func (db *TestDB) MustUpdate(fn func(tx *bolt.Tx) error) {
if err := db.DB.View(func(tx *bolt.Tx) error {
return fn(tx)
}); err != nil {
panic(err.Error())
}
}
// MustCreateBucket creates a new bucket. Panic on error.
func (db *TestDB) MustCreateBucket(name []byte) {
if err := db.Update(func(tx *bolt.Tx) error {
_, err := tx.CreateBucket([]byte(name))
return err
}); err != nil {
panic(err.Error())
}
}
// Close closes the database and deletes the underlying file.
func (db *TestDB) Close() {
// Log statistics.
if *statsFlag {
db.PrintStats()
}
// Check database consistency after every test.
db.MustCheck()
// Close database and remove file.
defer os.Remove(db.Path())
db.DB.Close()
}
// PrintStats prints the database stats
func (db *TestDB) PrintStats() {
var stats = db.Stats()
fmt.Printf("[db] %-20s %-20s %-20s\n",
fmt.Sprintf("pg(%d/%d)", stats.TxStats.PageCount, stats.TxStats.PageAlloc),
fmt.Sprintf("cur(%d)", stats.TxStats.CursorCount),
fmt.Sprintf("node(%d/%d)", stats.TxStats.NodeCount, stats.TxStats.NodeDeref),
)
fmt.Printf(" %-20s %-20s %-20s\n",
fmt.Sprintf("rebal(%d/%v)", stats.TxStats.Rebalance, truncDuration(stats.TxStats.RebalanceTime)),
fmt.Sprintf("spill(%d/%v)", stats.TxStats.Spill, truncDuration(stats.TxStats.SpillTime)),
fmt.Sprintf("w(%d/%v)", stats.TxStats.Write, truncDuration(stats.TxStats.WriteTime)),
)
}
// MustCheck runs a consistency check on the database and panics if any errors are found.
func (db *TestDB) MustCheck() {
db.View(func(tx *bolt.Tx) error {
// Collect all the errors.
var errors []error
for err := range tx.Check() {
errors = append(errors, err)
if len(errors) > 10 {
break
}
}
// If errors occurred, copy the DB and print the errors.
if len(errors) > 0 {
var path = tempfile()
tx.CopyFile(path, 0600)
// Print errors.
fmt.Print("\n\n")
fmt.Printf("consistency check failed (%d errors)\n", len(errors))
for _, err := range errors {
fmt.Println(err)
}
fmt.Println("")
fmt.Println("db saved to:")
fmt.Println(path)
fmt.Print("\n\n")
os.Exit(-1)
}
return nil
})
}
// CopyTempFile copies a database to a temporary file.
func (db *TestDB) CopyTempFile() {
path := tempfile()
db.View(func(tx *bolt.Tx) error { return tx.CopyFile(path, 0600) })
fmt.Println("db copied to: ", path)
}
// tempfile returns a temporary file path.
func tempfile() string {
f, _ := ioutil.TempFile("", "bolt-")
f.Close()
os.Remove(f.Name())
return f.Name()
}
// mustContainKeys checks that a bucket contains a given set of keys.
func mustContainKeys(b *bolt.Bucket, m map[string]string) {
found := make(map[string]string)
b.ForEach(func(k, _ []byte) error {
found[string(k)] = ""
return nil
})
// Check for keys found in bucket that shouldn't be there.
var keys []string
for k, _ := range found {
if _, ok := m[string(k)]; !ok {
keys = append(keys, k)
}
}
if len(keys) > 0 {
sort.Strings(keys)
panic(fmt.Sprintf("keys found(%d): %s", len(keys), strings.Join(keys, ",")))
}
// Check for keys not found in bucket that should be there.
for k, _ := range m {
if _, ok := found[string(k)]; !ok {
keys = append(keys, k)
}
}
if len(keys) > 0 {
sort.Strings(keys)
panic(fmt.Sprintf("keys not found(%d): %s", len(keys), strings.Join(keys, ",")))
}
}
func trunc(b []byte, length int) []byte {
if length < len(b) {
return b[:length]
}
return b
}
func truncDuration(d time.Duration) string {
return regexp.MustCompile(`^(\d+)(\.\d+)`).ReplaceAllString(d.String(), "$1")
}
func fileSize(path string) int64 {
fi, err := os.Stat(path)
if err != nil {
return 0
}
return fi.Size()
}
func warn(v ...interface{}) { fmt.Fprintln(os.Stderr, v...) }
func warnf(msg string, v ...interface{}) { fmt.Fprintf(os.Stderr, msg+"\n", v...) }
// u64tob converts a uint64 into an 8-byte slice.
func u64tob(v uint64) []byte {
b := make([]byte, 8)
binary.BigEndian.PutUint64(b, v)
return b
}
// btou64 converts an 8-byte slice into an uint64.
func btou64(b []byte) uint64 { return binary.BigEndian.Uint64(b) }

44
Godeps/_workspace/src/github.com/boltdb/bolt/doc.go generated vendored Normal file
View File

@@ -0,0 +1,44 @@
/*
Package bolt implements a low-level key/value store in pure Go. It supports
fully serializable transactions, ACID semantics, and lock-free MVCC with
multiple readers and a single writer. Bolt can be used for projects that
want a simple data store without the need to add large dependencies such as
Postgres or MySQL.
Bolt is a single-level, zero-copy, B+tree data store. This means that Bolt is
optimized for fast read access and does not require recovery in the event of a
system crash. Transactions which have not finished committing will simply be
rolled back in the event of a crash.
The design of Bolt is based on Howard Chu's LMDB database project.
Bolt currently works on Windows, Mac OS X, and Linux.
Basics
There are only a few types in Bolt: DB, Bucket, Tx, and Cursor. The DB is
a collection of buckets and is represented by a single file on disk. A bucket is
a collection of unique keys that are associated with values.
Transactions provide either read-only or read-write access to the database.
Read-only transactions can retrieve key/value pairs and can use Cursors to
iterate over the dataset sequentially. Read-write transactions can create and
delete buckets and can insert and remove keys. Only one read-write transaction
is allowed at a time.
Caveats
The database uses a read-only, memory-mapped data file to ensure that
applications cannot corrupt the database, however, this means that keys and
values returned from Bolt cannot be changed. Writing to a read-only byte slice
will cause Go to panic.
Keys and values retrieved from the database are only valid for the life of
the transaction. When used outside the transaction, these byte slices can
point to different data or can point to invalid memory which will cause a panic.
*/
package bolt

66
Godeps/_workspace/src/github.com/boltdb/bolt/errors.go generated vendored Normal file
View File

@@ -0,0 +1,66 @@
package bolt
import "errors"
// These errors can be returned when opening or calling methods on a DB.
var (
// ErrDatabaseNotOpen is returned when a DB instance is accessed before it
// is opened or after it is closed.
ErrDatabaseNotOpen = errors.New("database not open")
// ErrDatabaseOpen is returned when opening a database that is
// already open.
ErrDatabaseOpen = errors.New("database already open")
// ErrInvalid is returned when a data file is not a Bolt-formatted database.
ErrInvalid = errors.New("invalid database")
// ErrVersionMismatch is returned when the data file was created with a
// different version of Bolt.
ErrVersionMismatch = errors.New("version mismatch")
// ErrChecksum is returned when either meta page checksum does not match.
ErrChecksum = errors.New("checksum error")
// ErrTimeout is returned when a database cannot obtain an exclusive lock
// on the data file after the timeout passed to Open().
ErrTimeout = errors.New("timeout")
)
// These errors can occur when beginning or committing a Tx.
var (
// ErrTxNotWritable is returned when performing a write operation on a
// read-only transaction.
ErrTxNotWritable = errors.New("tx not writable")
// ErrTxClosed is returned when committing or rolling back a transaction
// that has already been committed or rolled back.
ErrTxClosed = errors.New("tx closed")
)
// These errors can occur when putting or deleting a value or a bucket.
var (
// ErrBucketNotFound is returned when trying to access a bucket that has
// not been created yet.
ErrBucketNotFound = errors.New("bucket not found")
// ErrBucketExists is returned when creating a bucket that already exists.
ErrBucketExists = errors.New("bucket already exists")
// ErrBucketNameRequired is returned when creating a bucket with a blank name.
ErrBucketNameRequired = errors.New("bucket name required")
// ErrKeyRequired is returned when inserting a zero-length key.
ErrKeyRequired = errors.New("key required")
// ErrKeyTooLarge is returned when inserting a key that is larger than MaxKeySize.
ErrKeyTooLarge = errors.New("key too large")
// ErrValueTooLarge is returned when inserting a value that is larger than MaxValueSize.
ErrValueTooLarge = errors.New("value too large")
// ErrIncompatibleValue is returned when trying create or delete a bucket
// on an existing non-bucket key or when trying to create or delete a
// non-bucket key on an existing bucket key.
ErrIncompatibleValue = errors.New("incompatible value")
)

View File

@@ -0,0 +1,241 @@
package bolt
import (
"fmt"
"sort"
"unsafe"
)
// freelist represents a list of all pages that are available for allocation.
// It also tracks pages that have been freed but are still in use by open transactions.
type freelist struct {
ids []pgid // all free and available free page ids.
pending map[txid][]pgid // mapping of soon-to-be free page ids by tx.
cache map[pgid]bool // fast lookup of all free and pending page ids.
}
// newFreelist returns an empty, initialized freelist.
func newFreelist() *freelist {
return &freelist{
pending: make(map[txid][]pgid),
cache: make(map[pgid]bool),
}
}
// size returns the size of the page after serialization.
func (f *freelist) size() int {
return pageHeaderSize + (int(unsafe.Sizeof(pgid(0))) * f.count())
}
// count returns count of pages on the freelist
func (f *freelist) count() int {
return f.free_count() + f.pending_count()
}
// free_count returns count of free pages
func (f *freelist) free_count() int {
return len(f.ids)
}
// pending_count returns count of pending pages
func (f *freelist) pending_count() int {
var count int
for _, list := range f.pending {
count += len(list)
}
return count
}
// all returns a list of all free ids and all pending ids in one sorted list.
func (f *freelist) all() []pgid {
ids := make([]pgid, len(f.ids))
copy(ids, f.ids)
for _, list := range f.pending {
ids = append(ids, list...)
}
sort.Sort(pgids(ids))
return ids
}
// allocate returns the starting page id of a contiguous list of pages of a given size.
// If a contiguous block cannot be found then 0 is returned.
func (f *freelist) allocate(n int) pgid {
if len(f.ids) == 0 {
return 0
}
var initial, previd pgid
for i, id := range f.ids {
if id <= 1 {
panic(fmt.Sprintf("invalid page allocation: %d", id))
}
// Reset initial page if this is not contiguous.
if previd == 0 || id-previd != 1 {
initial = id
}
// If we found a contiguous block then remove it and return it.
if (id-initial)+1 == pgid(n) {
// If we're allocating off the beginning then take the fast path
// and just adjust the existing slice. This will use extra memory
// temporarily but the append() in free() will realloc the slice
// as is necessary.
if (i + 1) == n {
f.ids = f.ids[i+1:]
} else {
copy(f.ids[i-n+1:], f.ids[i+1:])
f.ids = f.ids[:len(f.ids)-n]
}
// Remove from the free cache.
for i := pgid(0); i < pgid(n); i++ {
delete(f.cache, initial+i)
}
return initial
}
previd = id
}
return 0
}
// free releases a page and its overflow for a given transaction id.
// If the page is already free then a panic will occur.
func (f *freelist) free(txid txid, p *page) {
if p.id <= 1 {
panic(fmt.Sprintf("cannot free page 0 or 1: %d", p.id))
}
// Free page and all its overflow pages.
var ids = f.pending[txid]
for id := p.id; id <= p.id+pgid(p.overflow); id++ {
// Verify that page is not already free.
if f.cache[id] {
panic(fmt.Sprintf("page %d already freed", id))
}
// Add to the freelist and cache.
ids = append(ids, id)
f.cache[id] = true
}
f.pending[txid] = ids
}
// release moves all page ids for a transaction id (or older) to the freelist.
func (f *freelist) release(txid txid) {
for tid, ids := range f.pending {
if tid <= txid {
// Move transaction's pending pages to the available freelist.
// Don't remove from the cache since the page is still free.
f.ids = append(f.ids, ids...)
delete(f.pending, tid)
}
}
sort.Sort(pgids(f.ids))
}
// rollback removes the pages from a given pending tx.
func (f *freelist) rollback(txid txid) {
// Remove page ids from cache.
for _, id := range f.pending[txid] {
delete(f.cache, id)
}
// Remove pages from pending list.
delete(f.pending, txid)
}
// freed returns whether a given page is in the free list.
func (f *freelist) freed(pgid pgid) bool {
return f.cache[pgid]
}
// read initializes the freelist from a freelist page.
func (f *freelist) read(p *page) {
// If the page.count is at the max uint16 value (64k) then it's considered
// an overflow and the size of the freelist is stored as the first element.
idx, count := 0, int(p.count)
if count == 0xFFFF {
idx = 1
count = int(((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[0])
}
// Copy the list of page ids from the freelist.
ids := ((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[idx:count]
f.ids = make([]pgid, len(ids))
copy(f.ids, ids)
// Make sure they're sorted.
sort.Sort(pgids(f.ids))
// Rebuild the page cache.
f.reindex()
}
// write writes the page ids onto a freelist page. All free and pending ids are
// saved to disk since in the event of a program crash, all pending ids will
// become free.
func (f *freelist) write(p *page) error {
// Combine the old free pgids and pgids waiting on an open transaction.
ids := f.all()
// Update the header flag.
p.flags |= freelistPageFlag
// The page.count can only hold up to 64k elements so if we overflow that
// number then we handle it by putting the size in the first element.
if len(ids) < 0xFFFF {
p.count = uint16(len(ids))
copy(((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[:], ids)
} else {
p.count = 0xFFFF
((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[0] = pgid(len(ids))
copy(((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[1:], ids)
}
return nil
}
// reload reads the freelist from a page and filters out pending items.
func (f *freelist) reload(p *page) {
f.read(p)
// Build a cache of only pending pages.
pcache := make(map[pgid]bool)
for _, pendingIDs := range f.pending {
for _, pendingID := range pendingIDs {
pcache[pendingID] = true
}
}
// Check each page in the freelist and build a new available freelist
// with any pages not in the pending lists.
var a []pgid
for _, id := range f.ids {
if !pcache[id] {
a = append(a, id)
}
}
f.ids = a
// Once the available list is rebuilt then rebuild the free cache so that
// it includes the available and pending free pages.
f.reindex()
}
// reindex rebuilds the free cache based on available and pending free lists.
func (f *freelist) reindex() {
f.cache = make(map[pgid]bool)
for _, id := range f.ids {
f.cache[id] = true
}
for _, pendingIDs := range f.pending {
for _, pendingID := range pendingIDs {
f.cache[pendingID] = true
}
}
}

View File

@@ -0,0 +1,129 @@
package bolt
import (
"reflect"
"testing"
"unsafe"
)
// Ensure that a page is added to a transaction's freelist.
func TestFreelist_free(t *testing.T) {
f := newFreelist()
f.free(100, &page{id: 12})
if !reflect.DeepEqual([]pgid{12}, f.pending[100]) {
t.Fatalf("exp=%v; got=%v", []pgid{12}, f.pending[100])
}
}
// Ensure that a page and its overflow is added to a transaction's freelist.
func TestFreelist_free_overflow(t *testing.T) {
f := newFreelist()
f.free(100, &page{id: 12, overflow: 3})
if exp := []pgid{12, 13, 14, 15}; !reflect.DeepEqual(exp, f.pending[100]) {
t.Fatalf("exp=%v; got=%v", exp, f.pending[100])
}
}
// Ensure that a transaction's free pages can be released.
func TestFreelist_release(t *testing.T) {
f := newFreelist()
f.free(100, &page{id: 12, overflow: 1})
f.free(100, &page{id: 9})
f.free(102, &page{id: 39})
f.release(100)
f.release(101)
if exp := []pgid{9, 12, 13}; !reflect.DeepEqual(exp, f.ids) {
t.Fatalf("exp=%v; got=%v", exp, f.ids)
}
f.release(102)
if exp := []pgid{9, 12, 13, 39}; !reflect.DeepEqual(exp, f.ids) {
t.Fatalf("exp=%v; got=%v", exp, f.ids)
}
}
// Ensure that a freelist can find contiguous blocks of pages.
func TestFreelist_allocate(t *testing.T) {
f := &freelist{ids: []pgid{3, 4, 5, 6, 7, 9, 12, 13, 18}}
if id := int(f.allocate(3)); id != 3 {
t.Fatalf("exp=3; got=%v", id)
}
if id := int(f.allocate(1)); id != 6 {
t.Fatalf("exp=6; got=%v", id)
}
if id := int(f.allocate(3)); id != 0 {
t.Fatalf("exp=0; got=%v", id)
}
if id := int(f.allocate(2)); id != 12 {
t.Fatalf("exp=12; got=%v", id)
}
if id := int(f.allocate(1)); id != 7 {
t.Fatalf("exp=7; got=%v", id)
}
if id := int(f.allocate(0)); id != 0 {
t.Fatalf("exp=0; got=%v", id)
}
if id := int(f.allocate(0)); id != 0 {
t.Fatalf("exp=0; got=%v", id)
}
if exp := []pgid{9, 18}; !reflect.DeepEqual(exp, f.ids) {
t.Fatalf("exp=%v; got=%v", exp, f.ids)
}
if id := int(f.allocate(1)); id != 9 {
t.Fatalf("exp=9; got=%v", id)
}
if id := int(f.allocate(1)); id != 18 {
t.Fatalf("exp=18; got=%v", id)
}
if id := int(f.allocate(1)); id != 0 {
t.Fatalf("exp=0; got=%v", id)
}
if exp := []pgid{}; !reflect.DeepEqual(exp, f.ids) {
t.Fatalf("exp=%v; got=%v", exp, f.ids)
}
}
// Ensure that a freelist can deserialize from a freelist page.
func TestFreelist_read(t *testing.T) {
// Create a page.
var buf [4096]byte
page := (*page)(unsafe.Pointer(&buf[0]))
page.flags = freelistPageFlag
page.count = 2
// Insert 2 page ids.
ids := (*[3]pgid)(unsafe.Pointer(&page.ptr))
ids[0] = 23
ids[1] = 50
// Deserialize page into a freelist.
f := newFreelist()
f.read(page)
// Ensure that there are two page ids in the freelist.
if exp := []pgid{23, 50}; !reflect.DeepEqual(exp, f.ids) {
t.Fatalf("exp=%v; got=%v", exp, f.ids)
}
}
// Ensure that a freelist can serialize into a freelist page.
func TestFreelist_write(t *testing.T) {
// Create a freelist and write it to a page.
var buf [4096]byte
f := &freelist{ids: []pgid{12, 39}, pending: make(map[txid][]pgid)}
f.pending[100] = []pgid{28, 11}
f.pending[101] = []pgid{3}
p := (*page)(unsafe.Pointer(&buf[0]))
f.write(p)
// Read the page back out.
f2 := newFreelist()
f2.read(p)
// Ensure that the freelist is correct.
// All pages should be present and in reverse order.
if exp := []pgid{3, 11, 12, 28, 39}; !reflect.DeepEqual(exp, f2.ids) {
t.Fatalf("exp=%v; got=%v", exp, f2.ids)
}
}

627
Godeps/_workspace/src/github.com/boltdb/bolt/node.go generated vendored Normal file
View File

@@ -0,0 +1,627 @@
package bolt
import (
"bytes"
"fmt"
"sort"
"unsafe"
)
// node represents an in-memory, deserialized page.
type node struct {
bucket *Bucket
isLeaf bool
unbalanced bool
spilled bool
key []byte
pgid pgid
parent *node
children nodes
inodes inodes
}
// root returns the top-level node this node is attached to.
func (n *node) root() *node {
if n.parent == nil {
return n
}
return n.parent.root()
}
// minKeys returns the minimum number of inodes this node should have.
func (n *node) minKeys() int {
if n.isLeaf {
return 1
}
return 2
}
// size returns the size of the node after serialization.
func (n *node) size() int {
sz, elsz := pageHeaderSize, n.pageElementSize()
for i := 0; i < len(n.inodes); i++ {
item := &n.inodes[i]
sz += elsz + len(item.key) + len(item.value)
}
return sz
}
// sizeLessThan returns true if the node is less than a given size.
// This is an optimization to avoid calculating a large node when we only need
// to know if it fits inside a certain page size.
func (n *node) sizeLessThan(v int) bool {
sz, elsz := pageHeaderSize, n.pageElementSize()
for i := 0; i < len(n.inodes); i++ {
item := &n.inodes[i]
sz += elsz + len(item.key) + len(item.value)
if sz >= v {
return false
}
}
return true
}
// pageElementSize returns the size of each page element based on the type of node.
func (n *node) pageElementSize() int {
if n.isLeaf {
return leafPageElementSize
}
return branchPageElementSize
}
// childAt returns the child node at a given index.
func (n *node) childAt(index int) *node {
if n.isLeaf {
panic(fmt.Sprintf("invalid childAt(%d) on a leaf node", index))
}
return n.bucket.node(n.inodes[index].pgid, n)
}
// childIndex returns the index of a given child node.
func (n *node) childIndex(child *node) int {
index := sort.Search(len(n.inodes), func(i int) bool { return bytes.Compare(n.inodes[i].key, child.key) != -1 })
return index
}
// numChildren returns the number of children.
func (n *node) numChildren() int {
return len(n.inodes)
}
// nextSibling returns the next node with the same parent.
func (n *node) nextSibling() *node {
if n.parent == nil {
return nil
}
index := n.parent.childIndex(n)
if index >= n.parent.numChildren()-1 {
return nil
}
return n.parent.childAt(index + 1)
}
// prevSibling returns the previous node with the same parent.
func (n *node) prevSibling() *node {
if n.parent == nil {
return nil
}
index := n.parent.childIndex(n)
if index == 0 {
return nil
}
return n.parent.childAt(index - 1)
}
// put inserts a key/value.
func (n *node) put(oldKey, newKey, value []byte, pgid pgid, flags uint32) {
if pgid >= n.bucket.tx.meta.pgid {
panic(fmt.Sprintf("pgid (%d) above high water mark (%d)", pgid, n.bucket.tx.meta.pgid))
} else if len(oldKey) <= 0 {
panic("put: zero-length old key")
} else if len(newKey) <= 0 {
panic("put: zero-length new key")
}
// Find insertion index.
index := sort.Search(len(n.inodes), func(i int) bool { return bytes.Compare(n.inodes[i].key, oldKey) != -1 })
// Add capacity and shift nodes if we don't have an exact match and need to insert.
exact := (len(n.inodes) > 0 && index < len(n.inodes) && bytes.Equal(n.inodes[index].key, oldKey))
if !exact {
n.inodes = append(n.inodes, inode{})
copy(n.inodes[index+1:], n.inodes[index:])
}
inode := &n.inodes[index]
inode.flags = flags
inode.key = newKey
inode.value = value
inode.pgid = pgid
_assert(len(inode.key) > 0, "put: zero-length inode key")
}
// del removes a key from the node.
func (n *node) del(key []byte) {
// Find index of key.
index := sort.Search(len(n.inodes), func(i int) bool { return bytes.Compare(n.inodes[i].key, key) != -1 })
// Exit if the key isn't found.
if index >= len(n.inodes) || !bytes.Equal(n.inodes[index].key, key) {
return
}
// Delete inode from the node.
n.inodes = append(n.inodes[:index], n.inodes[index+1:]...)
// Mark the node as needing rebalancing.
n.unbalanced = true
}
// read initializes the node from a page.
func (n *node) read(p *page) {
n.pgid = p.id
n.isLeaf = ((p.flags & leafPageFlag) != 0)
n.inodes = make(inodes, int(p.count))
for i := 0; i < int(p.count); i++ {
inode := &n.inodes[i]
if n.isLeaf {
elem := p.leafPageElement(uint16(i))
inode.flags = elem.flags
inode.key = elem.key()
inode.value = elem.value()
} else {
elem := p.branchPageElement(uint16(i))
inode.pgid = elem.pgid
inode.key = elem.key()
}
_assert(len(inode.key) > 0, "read: zero-length inode key")
}
// Save first key so we can find the node in the parent when we spill.
if len(n.inodes) > 0 {
n.key = n.inodes[0].key
_assert(len(n.key) > 0, "read: zero-length node key")
} else {
n.key = nil
}
}
// write writes the items onto one or more pages.
func (n *node) write(p *page) {
// Initialize page.
if n.isLeaf {
p.flags |= leafPageFlag
} else {
p.flags |= branchPageFlag
}
if len(n.inodes) >= 0xFFFF {
panic(fmt.Sprintf("inode overflow: %d (pgid=%d)", len(n.inodes), p.id))
}
p.count = uint16(len(n.inodes))
// Loop over each item and write it to the page.
b := (*[maxAllocSize]byte)(unsafe.Pointer(&p.ptr))[n.pageElementSize()*len(n.inodes):]
for i, item := range n.inodes {
_assert(len(item.key) > 0, "write: zero-length inode key")
// Write the page element.
if n.isLeaf {
elem := p.leafPageElement(uint16(i))
elem.pos = uint32(uintptr(unsafe.Pointer(&b[0])) - uintptr(unsafe.Pointer(elem)))
elem.flags = item.flags
elem.ksize = uint32(len(item.key))
elem.vsize = uint32(len(item.value))
} else {
elem := p.branchPageElement(uint16(i))
elem.pos = uint32(uintptr(unsafe.Pointer(&b[0])) - uintptr(unsafe.Pointer(elem)))
elem.ksize = uint32(len(item.key))
elem.pgid = item.pgid
_assert(elem.pgid != p.id, "write: circular dependency occurred")
}
// Write data for the element to the end of the page.
copy(b[0:], item.key)
b = b[len(item.key):]
copy(b[0:], item.value)
b = b[len(item.value):]
}
// DEBUG ONLY: n.dump()
}
// split breaks up a node into multiple smaller nodes, if appropriate.
// This should only be called from the spill() function.
func (n *node) split(pageSize int) []*node {
var nodes []*node
node := n
for {
// Split node into two.
a, b := node.splitTwo(pageSize)
nodes = append(nodes, a)
// If we can't split then exit the loop.
if b == nil {
break
}
// Set node to b so it gets split on the next iteration.
node = b
}
return nodes
}
// splitTwo breaks up a node into two smaller nodes, if appropriate.
// This should only be called from the split() function.
func (n *node) splitTwo(pageSize int) (*node, *node) {
// Ignore the split if the page doesn't have at least enough nodes for
// two pages or if the nodes can fit in a single page.
if len(n.inodes) <= (minKeysPerPage*2) || n.sizeLessThan(pageSize) {
return n, nil
}
// Determine the threshold before starting a new node.
var fillPercent = n.bucket.FillPercent
if fillPercent < minFillPercent {
fillPercent = minFillPercent
} else if fillPercent > maxFillPercent {
fillPercent = maxFillPercent
}
threshold := int(float64(pageSize) * fillPercent)
// Determine split position and sizes of the two pages.
splitIndex, _ := n.splitIndex(threshold)
// Split node into two separate nodes.
// If there's no parent then we'll need to create one.
if n.parent == nil {
n.parent = &node{bucket: n.bucket, children: []*node{n}}
}
// Create a new node and add it to the parent.
next := &node{bucket: n.bucket, isLeaf: n.isLeaf, parent: n.parent}
n.parent.children = append(n.parent.children, next)
// Split inodes across two nodes.
next.inodes = n.inodes[splitIndex:]
n.inodes = n.inodes[:splitIndex]
// Update the statistics.
n.bucket.tx.stats.Split++
return n, next
}
// splitIndex finds the position where a page will fill a given threshold.
// It returns the index as well as the size of the first page.
// This is only be called from split().
func (n *node) splitIndex(threshold int) (index, sz int) {
sz = pageHeaderSize
// Loop until we only have the minimum number of keys required for the second page.
for i := 0; i < len(n.inodes)-minKeysPerPage; i++ {
index = i
inode := n.inodes[i]
elsize := n.pageElementSize() + len(inode.key) + len(inode.value)
// If we have at least the minimum number of keys and adding another
// node would put us over the threshold then exit and return.
if i >= minKeysPerPage && sz+elsize > threshold {
break
}
// Add the element size to the total size.
sz += elsize
}
return
}
// spill writes the nodes to dirty pages and splits nodes as it goes.
// Returns an error if dirty pages cannot be allocated.
func (n *node) spill() error {
var tx = n.bucket.tx
if n.spilled {
return nil
}
// Spill child nodes first. Child nodes can materialize sibling nodes in
// the case of split-merge so we cannot use a range loop. We have to check
// the children size on every loop iteration.
sort.Sort(n.children)
for i := 0; i < len(n.children); i++ {
if err := n.children[i].spill(); err != nil {
return err
}
}
// We no longer need the child list because it's only used for spill tracking.
n.children = nil
// Split nodes into appropriate sizes. The first node will always be n.
var nodes = n.split(tx.db.pageSize)
for _, node := range nodes {
// Add node's page to the freelist if it's not new.
if node.pgid > 0 {
tx.db.freelist.free(tx.meta.txid, tx.page(node.pgid))
node.pgid = 0
}
// Allocate contiguous space for the node.
p, err := tx.allocate((node.size() / tx.db.pageSize) + 1)
if err != nil {
return err
}
// Write the node.
if p.id >= tx.meta.pgid {
panic(fmt.Sprintf("pgid (%d) above high water mark (%d)", p.id, tx.meta.pgid))
}
node.pgid = p.id
node.write(p)
node.spilled = true
// Insert into parent inodes.
if node.parent != nil {
var key = node.key
if key == nil {
key = node.inodes[0].key
}
node.parent.put(key, node.inodes[0].key, nil, node.pgid, 0)
node.key = node.inodes[0].key
_assert(len(node.key) > 0, "spill: zero-length node key")
}
// Update the statistics.
tx.stats.Spill++
}
// If the root node split and created a new root then we need to spill that
// as well. We'll clear out the children to make sure it doesn't try to respill.
if n.parent != nil && n.parent.pgid == 0 {
n.children = nil
return n.parent.spill()
}
return nil
}
// rebalance attempts to combine the node with sibling nodes if the node fill
// size is below a threshold or if there are not enough keys.
func (n *node) rebalance() {
if !n.unbalanced {
return
}
n.unbalanced = false
// Update statistics.
n.bucket.tx.stats.Rebalance++
// Ignore if node is above threshold (25%) and has enough keys.
var threshold = n.bucket.tx.db.pageSize / 4
if n.size() > threshold && len(n.inodes) > n.minKeys() {
return
}
// Root node has special handling.
if n.parent == nil {
// If root node is a branch and only has one node then collapse it.
if !n.isLeaf && len(n.inodes) == 1 {
// Move root's child up.
child := n.bucket.node(n.inodes[0].pgid, n)
n.isLeaf = child.isLeaf
n.inodes = child.inodes[:]
n.children = child.children
// Reparent all child nodes being moved.
for _, inode := range n.inodes {
if child, ok := n.bucket.nodes[inode.pgid]; ok {
child.parent = n
}
}
// Remove old child.
child.parent = nil
delete(n.bucket.nodes, child.pgid)
child.free()
}
return
}
// If node has no keys then just remove it.
if n.numChildren() == 0 {
n.parent.del(n.key)
n.parent.removeChild(n)
delete(n.bucket.nodes, n.pgid)
n.free()
n.parent.rebalance()
return
}
_assert(n.parent.numChildren() > 1, "parent must have at least 2 children")
// Destination node is right sibling if idx == 0, otherwise left sibling.
var target *node
var useNextSibling = (n.parent.childIndex(n) == 0)
if useNextSibling {
target = n.nextSibling()
} else {
target = n.prevSibling()
}
// If target node has extra nodes then just move one over.
if target.numChildren() > target.minKeys() {
if useNextSibling {
// Reparent and move node.
if child, ok := n.bucket.nodes[target.inodes[0].pgid]; ok {
child.parent.removeChild(child)
child.parent = n
child.parent.children = append(child.parent.children, child)
}
n.inodes = append(n.inodes, target.inodes[0])
target.inodes = target.inodes[1:]
// Update target key on parent.
target.parent.put(target.key, target.inodes[0].key, nil, target.pgid, 0)
target.key = target.inodes[0].key
_assert(len(target.key) > 0, "rebalance(1): zero-length node key")
} else {
// Reparent and move node.
if child, ok := n.bucket.nodes[target.inodes[len(target.inodes)-1].pgid]; ok {
child.parent.removeChild(child)
child.parent = n
child.parent.children = append(child.parent.children, child)
}
n.inodes = append(n.inodes, inode{})
copy(n.inodes[1:], n.inodes)
n.inodes[0] = target.inodes[len(target.inodes)-1]
target.inodes = target.inodes[:len(target.inodes)-1]
}
// Update parent key for node.
n.parent.put(n.key, n.inodes[0].key, nil, n.pgid, 0)
n.key = n.inodes[0].key
_assert(len(n.key) > 0, "rebalance(2): zero-length node key")
return
}
// If both this node and the target node are too small then merge them.
if useNextSibling {
// Reparent all child nodes being moved.
for _, inode := range target.inodes {
if child, ok := n.bucket.nodes[inode.pgid]; ok {
child.parent.removeChild(child)
child.parent = n
child.parent.children = append(child.parent.children, child)
}
}
// Copy over inodes from target and remove target.
n.inodes = append(n.inodes, target.inodes...)
n.parent.del(target.key)
n.parent.removeChild(target)
delete(n.bucket.nodes, target.pgid)
target.free()
} else {
// Reparent all child nodes being moved.
for _, inode := range n.inodes {
if child, ok := n.bucket.nodes[inode.pgid]; ok {
child.parent.removeChild(child)
child.parent = target
child.parent.children = append(child.parent.children, child)
}
}
// Copy over inodes to target and remove node.
target.inodes = append(target.inodes, n.inodes...)
n.parent.del(n.key)
n.parent.removeChild(n)
delete(n.bucket.nodes, n.pgid)
n.free()
}
// Either this node or the target node was deleted from the parent so rebalance it.
n.parent.rebalance()
}
// removes a node from the list of in-memory children.
// This does not affect the inodes.
func (n *node) removeChild(target *node) {
for i, child := range n.children {
if child == target {
n.children = append(n.children[:i], n.children[i+1:]...)
return
}
}
}
// dereference causes the node to copy all its inode key/value references to heap memory.
// This is required when the mmap is reallocated so inodes are not pointing to stale data.
func (n *node) dereference() {
if n.key != nil {
key := make([]byte, len(n.key))
copy(key, n.key)
n.key = key
_assert(n.pgid == 0 || len(n.key) > 0, "dereference: zero-length node key on existing node")
}
for i := range n.inodes {
inode := &n.inodes[i]
key := make([]byte, len(inode.key))
copy(key, inode.key)
inode.key = key
_assert(len(inode.key) > 0, "dereference: zero-length inode key")
value := make([]byte, len(inode.value))
copy(value, inode.value)
inode.value = value
}
// Recursively dereference children.
for _, child := range n.children {
child.dereference()
}
// Update statistics.
n.bucket.tx.stats.NodeDeref++
}
// free adds the node's underlying page to the freelist.
func (n *node) free() {
if n.pgid != 0 {
n.bucket.tx.db.freelist.free(n.bucket.tx.meta.txid, n.bucket.tx.page(n.pgid))
n.pgid = 0
}
}
// dump writes the contents of the node to STDERR for debugging purposes.
/*
func (n *node) dump() {
// Write node header.
var typ = "branch"
if n.isLeaf {
typ = "leaf"
}
warnf("[NODE %d {type=%s count=%d}]", n.pgid, typ, len(n.inodes))
// Write out abbreviated version of each item.
for _, item := range n.inodes {
if n.isLeaf {
if item.flags&bucketLeafFlag != 0 {
bucket := (*bucket)(unsafe.Pointer(&item.value[0]))
warnf("+L %08x -> (bucket root=%d)", trunc(item.key, 4), bucket.root)
} else {
warnf("+L %08x -> %08x", trunc(item.key, 4), trunc(item.value, 4))
}
} else {
warnf("+B %08x -> pgid=%d", trunc(item.key, 4), item.pgid)
}
}
warn("")
}
*/
type nodes []*node
func (s nodes) Len() int { return len(s) }
func (s nodes) Swap(i, j int) { s[i], s[j] = s[j], s[i] }
func (s nodes) Less(i, j int) bool { return bytes.Compare(s[i].inodes[0].key, s[j].inodes[0].key) == -1 }
// inode represents an internal node inside of a node.
// It can be used to point to elements in a page or point
// to an element which hasn't been added to a page yet.
type inode struct {
flags uint32
pgid pgid
key []byte
value []byte
}
type inodes []inode

View File

@@ -0,0 +1,156 @@
package bolt
import (
"testing"
"unsafe"
)
// Ensure that a node can insert a key/value.
func TestNode_put(t *testing.T) {
n := &node{inodes: make(inodes, 0), bucket: &Bucket{tx: &Tx{meta: &meta{pgid: 1}}}}
n.put([]byte("baz"), []byte("baz"), []byte("2"), 0, 0)
n.put([]byte("foo"), []byte("foo"), []byte("0"), 0, 0)
n.put([]byte("bar"), []byte("bar"), []byte("1"), 0, 0)
n.put([]byte("foo"), []byte("foo"), []byte("3"), 0, leafPageFlag)
if len(n.inodes) != 3 {
t.Fatalf("exp=3; got=%d", len(n.inodes))
}
if k, v := n.inodes[0].key, n.inodes[0].value; string(k) != "bar" || string(v) != "1" {
t.Fatalf("exp=<bar,1>; got=<%s,%s>", k, v)
}
if k, v := n.inodes[1].key, n.inodes[1].value; string(k) != "baz" || string(v) != "2" {
t.Fatalf("exp=<baz,2>; got=<%s,%s>", k, v)
}
if k, v := n.inodes[2].key, n.inodes[2].value; string(k) != "foo" || string(v) != "3" {
t.Fatalf("exp=<foo,3>; got=<%s,%s>", k, v)
}
if n.inodes[2].flags != uint32(leafPageFlag) {
t.Fatalf("not a leaf: %d", n.inodes[2].flags)
}
}
// Ensure that a node can deserialize from a leaf page.
func TestNode_read_LeafPage(t *testing.T) {
// Create a page.
var buf [4096]byte
page := (*page)(unsafe.Pointer(&buf[0]))
page.flags = leafPageFlag
page.count = 2
// Insert 2 elements at the beginning. sizeof(leafPageElement) == 16
nodes := (*[3]leafPageElement)(unsafe.Pointer(&page.ptr))
nodes[0] = leafPageElement{flags: 0, pos: 32, ksize: 3, vsize: 4} // pos = sizeof(leafPageElement) * 2
nodes[1] = leafPageElement{flags: 0, pos: 23, ksize: 10, vsize: 3} // pos = sizeof(leafPageElement) + 3 + 4
// Write data for the nodes at the end.
data := (*[4096]byte)(unsafe.Pointer(&nodes[2]))
copy(data[:], []byte("barfooz"))
copy(data[7:], []byte("helloworldbye"))
// Deserialize page into a leaf.
n := &node{}
n.read(page)
// Check that there are two inodes with correct data.
if !n.isLeaf {
t.Fatal("expected leaf")
}
if len(n.inodes) != 2 {
t.Fatalf("exp=2; got=%d", len(n.inodes))
}
if k, v := n.inodes[0].key, n.inodes[0].value; string(k) != "bar" || string(v) != "fooz" {
t.Fatalf("exp=<bar,fooz>; got=<%s,%s>", k, v)
}
if k, v := n.inodes[1].key, n.inodes[1].value; string(k) != "helloworld" || string(v) != "bye" {
t.Fatalf("exp=<helloworld,bye>; got=<%s,%s>", k, v)
}
}
// Ensure that a node can serialize into a leaf page.
func TestNode_write_LeafPage(t *testing.T) {
// Create a node.
n := &node{isLeaf: true, inodes: make(inodes, 0), bucket: &Bucket{tx: &Tx{db: &DB{}, meta: &meta{pgid: 1}}}}
n.put([]byte("susy"), []byte("susy"), []byte("que"), 0, 0)
n.put([]byte("ricki"), []byte("ricki"), []byte("lake"), 0, 0)
n.put([]byte("john"), []byte("john"), []byte("johnson"), 0, 0)
// Write it to a page.
var buf [4096]byte
p := (*page)(unsafe.Pointer(&buf[0]))
n.write(p)
// Read the page back in.
n2 := &node{}
n2.read(p)
// Check that the two pages are the same.
if len(n2.inodes) != 3 {
t.Fatalf("exp=3; got=%d", len(n2.inodes))
}
if k, v := n2.inodes[0].key, n2.inodes[0].value; string(k) != "john" || string(v) != "johnson" {
t.Fatalf("exp=<john,johnson>; got=<%s,%s>", k, v)
}
if k, v := n2.inodes[1].key, n2.inodes[1].value; string(k) != "ricki" || string(v) != "lake" {
t.Fatalf("exp=<ricki,lake>; got=<%s,%s>", k, v)
}
if k, v := n2.inodes[2].key, n2.inodes[2].value; string(k) != "susy" || string(v) != "que" {
t.Fatalf("exp=<susy,que>; got=<%s,%s>", k, v)
}
}
// Ensure that a node can split into appropriate subgroups.
func TestNode_split(t *testing.T) {
// Create a node.
n := &node{inodes: make(inodes, 0), bucket: &Bucket{tx: &Tx{db: &DB{}, meta: &meta{pgid: 1}}}}
n.put([]byte("00000001"), []byte("00000001"), []byte("0123456701234567"), 0, 0)
n.put([]byte("00000002"), []byte("00000002"), []byte("0123456701234567"), 0, 0)
n.put([]byte("00000003"), []byte("00000003"), []byte("0123456701234567"), 0, 0)
n.put([]byte("00000004"), []byte("00000004"), []byte("0123456701234567"), 0, 0)
n.put([]byte("00000005"), []byte("00000005"), []byte("0123456701234567"), 0, 0)
// Split between 2 & 3.
n.split(100)
var parent = n.parent
if len(parent.children) != 2 {
t.Fatalf("exp=2; got=%d", len(parent.children))
}
if len(parent.children[0].inodes) != 2 {
t.Fatalf("exp=2; got=%d", len(parent.children[0].inodes))
}
if len(parent.children[1].inodes) != 3 {
t.Fatalf("exp=3; got=%d", len(parent.children[1].inodes))
}
}
// Ensure that a page with the minimum number of inodes just returns a single node.
func TestNode_split_MinKeys(t *testing.T) {
// Create a node.
n := &node{inodes: make(inodes, 0), bucket: &Bucket{tx: &Tx{db: &DB{}, meta: &meta{pgid: 1}}}}
n.put([]byte("00000001"), []byte("00000001"), []byte("0123456701234567"), 0, 0)
n.put([]byte("00000002"), []byte("00000002"), []byte("0123456701234567"), 0, 0)
// Split.
n.split(20)
if n.parent != nil {
t.Fatalf("expected nil parent")
}
}
// Ensure that a node that has keys that all fit on a page just returns one leaf.
func TestNode_split_SinglePage(t *testing.T) {
// Create a node.
n := &node{inodes: make(inodes, 0), bucket: &Bucket{tx: &Tx{db: &DB{}, meta: &meta{pgid: 1}}}}
n.put([]byte("00000001"), []byte("00000001"), []byte("0123456701234567"), 0, 0)
n.put([]byte("00000002"), []byte("00000002"), []byte("0123456701234567"), 0, 0)
n.put([]byte("00000003"), []byte("00000003"), []byte("0123456701234567"), 0, 0)
n.put([]byte("00000004"), []byte("00000004"), []byte("0123456701234567"), 0, 0)
n.put([]byte("00000005"), []byte("00000005"), []byte("0123456701234567"), 0, 0)
// Split.
n.split(4096)
if n.parent != nil {
t.Fatalf("expected nil parent")
}
}

134
Godeps/_workspace/src/github.com/boltdb/bolt/page.go generated vendored Normal file
View File

@@ -0,0 +1,134 @@
package bolt
import (
"fmt"
"os"
"unsafe"
)
const pageHeaderSize = int(unsafe.Offsetof(((*page)(nil)).ptr))
const minKeysPerPage = 2
const branchPageElementSize = int(unsafe.Sizeof(branchPageElement{}))
const leafPageElementSize = int(unsafe.Sizeof(leafPageElement{}))
const (
branchPageFlag = 0x01
leafPageFlag = 0x02
metaPageFlag = 0x04
freelistPageFlag = 0x10
)
const (
bucketLeafFlag = 0x01
)
type pgid uint64
type page struct {
id pgid
flags uint16
count uint16
overflow uint32
ptr uintptr
}
// typ returns a human readable page type string used for debugging.
func (p *page) typ() string {
if (p.flags & branchPageFlag) != 0 {
return "branch"
} else if (p.flags & leafPageFlag) != 0 {
return "leaf"
} else if (p.flags & metaPageFlag) != 0 {
return "meta"
} else if (p.flags & freelistPageFlag) != 0 {
return "freelist"
}
return fmt.Sprintf("unknown<%02x>", p.flags)
}
// meta returns a pointer to the metadata section of the page.
func (p *page) meta() *meta {
return (*meta)(unsafe.Pointer(&p.ptr))
}
// leafPageElement retrieves the leaf node by index
func (p *page) leafPageElement(index uint16) *leafPageElement {
n := &((*[0x7FFFFFF]leafPageElement)(unsafe.Pointer(&p.ptr)))[index]
return n
}
// leafPageElements retrieves a list of leaf nodes.
func (p *page) leafPageElements() []leafPageElement {
return ((*[0x7FFFFFF]leafPageElement)(unsafe.Pointer(&p.ptr)))[:]
}
// branchPageElement retrieves the branch node by index
func (p *page) branchPageElement(index uint16) *branchPageElement {
return &((*[0x7FFFFFF]branchPageElement)(unsafe.Pointer(&p.ptr)))[index]
}
// branchPageElements retrieves a list of branch nodes.
func (p *page) branchPageElements() []branchPageElement {
return ((*[0x7FFFFFF]branchPageElement)(unsafe.Pointer(&p.ptr)))[:]
}
// dump writes n bytes of the page to STDERR as hex output.
func (p *page) hexdump(n int) {
buf := (*[maxAllocSize]byte)(unsafe.Pointer(p))[:n]
fmt.Fprintf(os.Stderr, "%x\n", buf)
}
type pages []*page
func (s pages) Len() int { return len(s) }
func (s pages) Swap(i, j int) { s[i], s[j] = s[j], s[i] }
func (s pages) Less(i, j int) bool { return s[i].id < s[j].id }
// branchPageElement represents a node on a branch page.
type branchPageElement struct {
pos uint32
ksize uint32
pgid pgid
}
// key returns a byte slice of the node key.
func (n *branchPageElement) key() []byte {
buf := (*[maxAllocSize]byte)(unsafe.Pointer(n))
return buf[n.pos : n.pos+n.ksize]
}
// leafPageElement represents a node on a leaf page.
type leafPageElement struct {
flags uint32
pos uint32
ksize uint32
vsize uint32
}
// key returns a byte slice of the node key.
func (n *leafPageElement) key() []byte {
buf := (*[maxAllocSize]byte)(unsafe.Pointer(n))
return buf[n.pos : n.pos+n.ksize]
}
// value returns a byte slice of the node value.
func (n *leafPageElement) value() []byte {
buf := (*[maxAllocSize]byte)(unsafe.Pointer(n))
return buf[n.pos+n.ksize : n.pos+n.ksize+n.vsize]
}
// PageInfo represents human readable information about a page.
type PageInfo struct {
ID int
Type string
Count int
OverflowCount int
}
type pgids []pgid
func (s pgids) Len() int { return len(s) }
func (s pgids) Swap(i, j int) { s[i], s[j] = s[j], s[i] }
func (s pgids) Less(i, j int) bool { return s[i] < s[j] }

View File

@@ -0,0 +1,29 @@
package bolt
import (
"testing"
)
// Ensure that the page type can be returned in human readable format.
func TestPage_typ(t *testing.T) {
if typ := (&page{flags: branchPageFlag}).typ(); typ != "branch" {
t.Fatalf("exp=branch; got=%v", typ)
}
if typ := (&page{flags: leafPageFlag}).typ(); typ != "leaf" {
t.Fatalf("exp=leaf; got=%v", typ)
}
if typ := (&page{flags: metaPageFlag}).typ(); typ != "meta" {
t.Fatalf("exp=meta; got=%v", typ)
}
if typ := (&page{flags: freelistPageFlag}).typ(); typ != "freelist" {
t.Fatalf("exp=freelist; got=%v", typ)
}
if typ := (&page{flags: 20000}).typ(); typ != "unknown<4e20>" {
t.Fatalf("exp=unknown<4e20>; got=%v", typ)
}
}
// Ensure that the hexdump debugging function doesn't blow up.
func TestPage_dump(t *testing.T) {
(&page{id: 256}).hexdump(16)
}

View File

@@ -0,0 +1,79 @@
package bolt_test
import (
"bytes"
"flag"
"fmt"
"math/rand"
"os"
"reflect"
"testing/quick"
"time"
)
// testing/quick defaults to 5 iterations and a random seed.
// You can override these settings from the command line:
//
// -quick.count The number of iterations to perform.
// -quick.seed The seed to use for randomizing.
// -quick.maxitems The maximum number of items to insert into a DB.
// -quick.maxksize The maximum size of a key.
// -quick.maxvsize The maximum size of a value.
//
var qcount, qseed, qmaxitems, qmaxksize, qmaxvsize int
func init() {
flag.IntVar(&qcount, "quick.count", 5, "")
flag.IntVar(&qseed, "quick.seed", int(time.Now().UnixNano())%100000, "")
flag.IntVar(&qmaxitems, "quick.maxitems", 1000, "")
flag.IntVar(&qmaxksize, "quick.maxksize", 1024, "")
flag.IntVar(&qmaxvsize, "quick.maxvsize", 1024, "")
flag.Parse()
fmt.Fprintln(os.Stderr, "seed:", qseed)
fmt.Fprintf(os.Stderr, "quick settings: count=%v, items=%v, ksize=%v, vsize=%v\n", qcount, qmaxitems, qmaxksize, qmaxvsize)
}
func qconfig() *quick.Config {
return &quick.Config{
MaxCount: qcount,
Rand: rand.New(rand.NewSource(int64(qseed))),
}
}
type testdata []testdataitem
func (t testdata) Len() int { return len(t) }
func (t testdata) Swap(i, j int) { t[i], t[j] = t[j], t[i] }
func (t testdata) Less(i, j int) bool { return bytes.Compare(t[i].Key, t[j].Key) == -1 }
func (t testdata) Generate(rand *rand.Rand, size int) reflect.Value {
n := rand.Intn(qmaxitems-1) + 1
items := make(testdata, n)
for i := 0; i < n; i++ {
item := &items[i]
item.Key = randByteSlice(rand, 1, qmaxksize)
item.Value = randByteSlice(rand, 0, qmaxvsize)
}
return reflect.ValueOf(items)
}
type revtestdata []testdataitem
func (t revtestdata) Len() int { return len(t) }
func (t revtestdata) Swap(i, j int) { t[i], t[j] = t[j], t[i] }
func (t revtestdata) Less(i, j int) bool { return bytes.Compare(t[i].Key, t[j].Key) == 1 }
type testdataitem struct {
Key []byte
Value []byte
}
func randByteSlice(rand *rand.Rand, minSize, maxSize int) []byte {
n := rand.Intn(maxSize-minSize) + minSize
b := make([]byte, n)
for i := 0; i < n; i++ {
b[i] = byte(rand.Intn(255))
}
return b
}

View File

@@ -0,0 +1,327 @@
package bolt_test
import (
"bytes"
"fmt"
"math/rand"
"sync"
"testing"
"github.com/coreos/etcd/Godeps/_workspace/src/github.com/boltdb/bolt"
)
func TestSimulate_1op_1p(t *testing.T) { testSimulate(t, 100, 1) }
func TestSimulate_10op_1p(t *testing.T) { testSimulate(t, 10, 1) }
func TestSimulate_100op_1p(t *testing.T) { testSimulate(t, 100, 1) }
func TestSimulate_1000op_1p(t *testing.T) { testSimulate(t, 1000, 1) }
func TestSimulate_10000op_1p(t *testing.T) { testSimulate(t, 10000, 1) }
func TestSimulate_10op_10p(t *testing.T) { testSimulate(t, 10, 10) }
func TestSimulate_100op_10p(t *testing.T) { testSimulate(t, 100, 10) }
func TestSimulate_1000op_10p(t *testing.T) { testSimulate(t, 1000, 10) }
func TestSimulate_10000op_10p(t *testing.T) { testSimulate(t, 10000, 10) }
func TestSimulate_100op_100p(t *testing.T) { testSimulate(t, 100, 100) }
func TestSimulate_1000op_100p(t *testing.T) { testSimulate(t, 1000, 100) }
func TestSimulate_10000op_100p(t *testing.T) { testSimulate(t, 10000, 100) }
func TestSimulate_10000op_1000p(t *testing.T) { testSimulate(t, 10000, 1000) }
// Randomly generate operations on a given database with multiple clients to ensure consistency and thread safety.
func testSimulate(t *testing.T, threadCount, parallelism int) {
if testing.Short() {
t.Skip("skipping test in short mode.")
}
rand.Seed(int64(qseed))
// A list of operations that readers and writers can perform.
var readerHandlers = []simulateHandler{simulateGetHandler}
var writerHandlers = []simulateHandler{simulateGetHandler, simulatePutHandler}
var versions = make(map[int]*QuickDB)
versions[1] = NewQuickDB()
db := NewTestDB()
defer db.Close()
var mutex sync.Mutex
// Run n threads in parallel, each with their own operation.
var wg sync.WaitGroup
var threads = make(chan bool, parallelism)
var i int
for {
threads <- true
wg.Add(1)
writable := ((rand.Int() % 100) < 20) // 20% writers
// Choose an operation to execute.
var handler simulateHandler
if writable {
handler = writerHandlers[rand.Intn(len(writerHandlers))]
} else {
handler = readerHandlers[rand.Intn(len(readerHandlers))]
}
// Execute a thread for the given operation.
go func(writable bool, handler simulateHandler) {
defer wg.Done()
// Start transaction.
tx, err := db.Begin(writable)
if err != nil {
t.Fatal("tx begin: ", err)
}
// Obtain current state of the dataset.
mutex.Lock()
var qdb = versions[tx.ID()]
if writable {
qdb = versions[tx.ID()-1].Copy()
}
mutex.Unlock()
// Make sure we commit/rollback the tx at the end and update the state.
if writable {
defer func() {
mutex.Lock()
versions[tx.ID()] = qdb
mutex.Unlock()
ok(t, tx.Commit())
}()
} else {
defer tx.Rollback()
}
// Ignore operation if we don't have data yet.
if qdb == nil {
return
}
// Execute handler.
handler(tx, qdb)
// Release a thread back to the scheduling loop.
<-threads
}(writable, handler)
i++
if i > threadCount {
break
}
}
// Wait until all threads are done.
wg.Wait()
}
type simulateHandler func(tx *bolt.Tx, qdb *QuickDB)
// Retrieves a key from the database and verifies that it is what is expected.
func simulateGetHandler(tx *bolt.Tx, qdb *QuickDB) {
// Randomly retrieve an existing exist.
keys := qdb.Rand()
if len(keys) == 0 {
return
}
// Retrieve root bucket.
b := tx.Bucket(keys[0])
if b == nil {
panic(fmt.Sprintf("bucket[0] expected: %08x\n", trunc(keys[0], 4)))
}
// Drill into nested buckets.
for _, key := range keys[1 : len(keys)-1] {
b = b.Bucket(key)
if b == nil {
panic(fmt.Sprintf("bucket[n] expected: %v -> %v\n", keys, key))
}
}
// Verify key/value on the final bucket.
expected := qdb.Get(keys)
actual := b.Get(keys[len(keys)-1])
if !bytes.Equal(actual, expected) {
fmt.Println("=== EXPECTED ===")
fmt.Println(expected)
fmt.Println("=== ACTUAL ===")
fmt.Println(actual)
fmt.Println("=== END ===")
panic("value mismatch")
}
}
// Inserts a key into the database.
func simulatePutHandler(tx *bolt.Tx, qdb *QuickDB) {
var err error
keys, value := randKeys(), randValue()
// Retrieve root bucket.
b := tx.Bucket(keys[0])
if b == nil {
b, err = tx.CreateBucket(keys[0])
if err != nil {
panic("create bucket: " + err.Error())
}
}
// Create nested buckets, if necessary.
for _, key := range keys[1 : len(keys)-1] {
child := b.Bucket(key)
if child != nil {
b = child
} else {
b, err = b.CreateBucket(key)
if err != nil {
panic("create bucket: " + err.Error())
}
}
}
// Insert into database.
if err := b.Put(keys[len(keys)-1], value); err != nil {
panic("put: " + err.Error())
}
// Insert into in-memory database.
qdb.Put(keys, value)
}
// QuickDB is an in-memory database that replicates the functionality of the
// Bolt DB type except that it is entirely in-memory. It is meant for testing
// that the Bolt database is consistent.
type QuickDB struct {
sync.RWMutex
m map[string]interface{}
}
// NewQuickDB returns an instance of QuickDB.
func NewQuickDB() *QuickDB {
return &QuickDB{m: make(map[string]interface{})}
}
// Get retrieves the value at a key path.
func (db *QuickDB) Get(keys [][]byte) []byte {
db.RLock()
defer db.RUnlock()
m := db.m
for _, key := range keys[:len(keys)-1] {
value := m[string(key)]
if value == nil {
return nil
}
switch value := value.(type) {
case map[string]interface{}:
m = value
case []byte:
return nil
}
}
// Only return if it's a simple value.
if value, ok := m[string(keys[len(keys)-1])].([]byte); ok {
return value
}
return nil
}
// Put inserts a value into a key path.
func (db *QuickDB) Put(keys [][]byte, value []byte) {
db.Lock()
defer db.Unlock()
// Build buckets all the way down the key path.
m := db.m
for _, key := range keys[:len(keys)-1] {
if _, ok := m[string(key)].([]byte); ok {
return // Keypath intersects with a simple value. Do nothing.
}
if m[string(key)] == nil {
m[string(key)] = make(map[string]interface{})
}
m = m[string(key)].(map[string]interface{})
}
// Insert value into the last key.
m[string(keys[len(keys)-1])] = value
}
// Rand returns a random key path that points to a simple value.
func (db *QuickDB) Rand() [][]byte {
db.RLock()
defer db.RUnlock()
if len(db.m) == 0 {
return nil
}
var keys [][]byte
db.rand(db.m, &keys)
return keys
}
func (db *QuickDB) rand(m map[string]interface{}, keys *[][]byte) {
i, index := 0, rand.Intn(len(m))
for k, v := range m {
if i == index {
*keys = append(*keys, []byte(k))
if v, ok := v.(map[string]interface{}); ok {
db.rand(v, keys)
}
return
}
i++
}
panic("quickdb rand: out-of-range")
}
// Copy copies the entire database.
func (db *QuickDB) Copy() *QuickDB {
db.RLock()
defer db.RUnlock()
return &QuickDB{m: db.copy(db.m)}
}
func (db *QuickDB) copy(m map[string]interface{}) map[string]interface{} {
clone := make(map[string]interface{}, len(m))
for k, v := range m {
switch v := v.(type) {
case map[string]interface{}:
clone[k] = db.copy(v)
default:
clone[k] = v
}
}
return clone
}
func randKey() []byte {
var min, max = 1, 1024
n := rand.Intn(max-min) + min
b := make([]byte, n)
for i := 0; i < n; i++ {
b[i] = byte(rand.Intn(255))
}
return b
}
func randKeys() [][]byte {
var keys [][]byte
var count = rand.Intn(2) + 2
for i := 0; i < count; i++ {
keys = append(keys, randKey())
}
return keys
}
func randValue() []byte {
n := rand.Intn(8192)
b := make([]byte, n)
for i := 0; i < n; i++ {
b[i] = byte(rand.Intn(255))
}
return b
}

585
Godeps/_workspace/src/github.com/boltdb/bolt/tx.go generated vendored Normal file
View File

@@ -0,0 +1,585 @@
package bolt
import (
"fmt"
"io"
"os"
"sort"
"time"
"unsafe"
)
// txid represents the internal transaction identifier.
type txid uint64
// Tx represents a read-only or read/write transaction on the database.
// Read-only transactions can be used for retrieving values for keys and creating cursors.
// Read/write transactions can create and remove buckets and create and remove keys.
//
// IMPORTANT: You must commit or rollback transactions when you are done with
// them. Pages can not be reclaimed by the writer until no more transactions
// are using them. A long running read transaction can cause the database to
// quickly grow.
type Tx struct {
writable bool
managed bool
db *DB
meta *meta
root Bucket
pages map[pgid]*page
stats TxStats
commitHandlers []func()
}
// init initializes the transaction.
func (tx *Tx) init(db *DB) {
tx.db = db
tx.pages = nil
// Copy the meta page since it can be changed by the writer.
tx.meta = &meta{}
db.meta().copy(tx.meta)
// Copy over the root bucket.
tx.root = newBucket(tx)
tx.root.bucket = &bucket{}
*tx.root.bucket = tx.meta.root
// Increment the transaction id and add a page cache for writable transactions.
if tx.writable {
tx.pages = make(map[pgid]*page)
tx.meta.txid += txid(1)
}
}
// ID returns the transaction id.
func (tx *Tx) ID() int {
return int(tx.meta.txid)
}
// DB returns a reference to the database that created the transaction.
func (tx *Tx) DB() *DB {
return tx.db
}
// Size returns current database size in bytes as seen by this transaction.
func (tx *Tx) Size() int64 {
return int64(tx.meta.pgid) * int64(tx.db.pageSize)
}
// Writable returns whether the transaction can perform write operations.
func (tx *Tx) Writable() bool {
return tx.writable
}
// Cursor creates a cursor associated with the root bucket.
// All items in the cursor will return a nil value because all root bucket keys point to buckets.
// The cursor is only valid as long as the transaction is open.
// Do not use a cursor after the transaction is closed.
func (tx *Tx) Cursor() *Cursor {
return tx.root.Cursor()
}
// Stats retrieves a copy of the current transaction statistics.
func (tx *Tx) Stats() TxStats {
return tx.stats
}
// Bucket retrieves a bucket by name.
// Returns nil if the bucket does not exist.
func (tx *Tx) Bucket(name []byte) *Bucket {
return tx.root.Bucket(name)
}
// CreateBucket creates a new bucket.
// Returns an error if the bucket already exists, if the bucket name is blank, or if the bucket name is too long.
func (tx *Tx) CreateBucket(name []byte) (*Bucket, error) {
return tx.root.CreateBucket(name)
}
// CreateBucketIfNotExists creates a new bucket if it doesn't already exist.
// Returns an error if the bucket name is blank, or if the bucket name is too long.
func (tx *Tx) CreateBucketIfNotExists(name []byte) (*Bucket, error) {
return tx.root.CreateBucketIfNotExists(name)
}
// DeleteBucket deletes a bucket.
// Returns an error if the bucket cannot be found or if the key represents a non-bucket value.
func (tx *Tx) DeleteBucket(name []byte) error {
return tx.root.DeleteBucket(name)
}
// ForEach executes a function for each bucket in the root.
// If the provided function returns an error then the iteration is stopped and
// the error is returned to the caller.
func (tx *Tx) ForEach(fn func(name []byte, b *Bucket) error) error {
return tx.root.ForEach(func(k, v []byte) error {
if err := fn(k, tx.root.Bucket(k)); err != nil {
return err
}
return nil
})
}
// OnCommit adds a handler function to be executed after the transaction successfully commits.
func (tx *Tx) OnCommit(fn func()) {
tx.commitHandlers = append(tx.commitHandlers, fn)
}
// Commit writes all changes to disk and updates the meta page.
// Returns an error if a disk write error occurs.
func (tx *Tx) Commit() error {
_assert(!tx.managed, "managed tx commit not allowed")
if tx.db == nil {
return ErrTxClosed
} else if !tx.writable {
return ErrTxNotWritable
}
// TODO(benbjohnson): Use vectorized I/O to write out dirty pages.
// Rebalance nodes which have had deletions.
var startTime = time.Now()
tx.root.rebalance()
if tx.stats.Rebalance > 0 {
tx.stats.RebalanceTime += time.Since(startTime)
}
// spill data onto dirty pages.
startTime = time.Now()
if err := tx.root.spill(); err != nil {
tx.rollback()
return err
}
tx.stats.SpillTime += time.Since(startTime)
// Free the old root bucket.
tx.meta.root.root = tx.root.root
// Free the freelist and allocate new pages for it. This will overestimate
// the size of the freelist but not underestimate the size (which would be bad).
tx.db.freelist.free(tx.meta.txid, tx.db.page(tx.meta.freelist))
p, err := tx.allocate((tx.db.freelist.size() / tx.db.pageSize) + 1)
if err != nil {
tx.rollback()
return err
}
if err := tx.db.freelist.write(p); err != nil {
tx.rollback()
return err
}
tx.meta.freelist = p.id
// Write dirty pages to disk.
startTime = time.Now()
if err := tx.write(); err != nil {
tx.rollback()
return err
}
// If strict mode is enabled then perform a consistency check.
// Only the first consistency error is reported in the panic.
if tx.db.StrictMode {
if err, ok := <-tx.Check(); ok {
panic("check fail: " + err.Error())
}
}
// Write meta to disk.
if err := tx.writeMeta(); err != nil {
tx.rollback()
return err
}
tx.stats.WriteTime += time.Since(startTime)
// Finalize the transaction.
tx.close()
// Execute commit handlers now that the locks have been removed.
for _, fn := range tx.commitHandlers {
fn()
}
return nil
}
// Rollback closes the transaction and ignores all previous updates.
func (tx *Tx) Rollback() error {
_assert(!tx.managed, "managed tx rollback not allowed")
if tx.db == nil {
return ErrTxClosed
}
tx.rollback()
return nil
}
func (tx *Tx) rollback() {
if tx.db == nil {
return
}
if tx.writable {
tx.db.freelist.rollback(tx.meta.txid)
tx.db.freelist.reload(tx.db.page(tx.db.meta().freelist))
}
tx.close()
}
func (tx *Tx) close() {
if tx.db == nil {
return
}
if tx.writable {
// Grab freelist stats.
var freelistFreeN = tx.db.freelist.free_count()
var freelistPendingN = tx.db.freelist.pending_count()
var freelistAlloc = tx.db.freelist.size()
// Remove writer lock.
tx.db.rwlock.Unlock()
// Merge statistics.
tx.db.statlock.Lock()
tx.db.stats.FreePageN = freelistFreeN
tx.db.stats.PendingPageN = freelistPendingN
tx.db.stats.FreeAlloc = (freelistFreeN + freelistPendingN) * tx.db.pageSize
tx.db.stats.FreelistInuse = freelistAlloc
tx.db.stats.TxStats.add(&tx.stats)
tx.db.statlock.Unlock()
} else {
tx.db.removeTx(tx)
}
tx.db = nil
}
// Copy writes the entire database to a writer.
// This function exists for backwards compatibility. Use WriteTo() in
func (tx *Tx) Copy(w io.Writer) error {
_, err := tx.WriteTo(w)
return err
}
// WriteTo writes the entire database to a writer.
// If err == nil then exactly tx.Size() bytes will be written into the writer.
func (tx *Tx) WriteTo(w io.Writer) (n int64, err error) {
// Attempt to open reader directly.
var f *os.File
if f, err = os.OpenFile(tx.db.path, os.O_RDONLY|odirect, 0); err != nil {
// Fallback to a regular open if that doesn't work.
if f, err = os.OpenFile(tx.db.path, os.O_RDONLY, 0); err != nil {
return 0, err
}
}
// Copy the meta pages.
tx.db.metalock.Lock()
n, err = io.CopyN(w, f, int64(tx.db.pageSize*2))
tx.db.metalock.Unlock()
if err != nil {
_ = f.Close()
return n, fmt.Errorf("meta copy: %s", err)
}
// Copy data pages.
wn, err := io.CopyN(w, f, tx.Size()-int64(tx.db.pageSize*2))
n += wn
if err != nil {
_ = f.Close()
return n, err
}
return n, f.Close()
}
// CopyFile copies the entire database to file at the given path.
// A reader transaction is maintained during the copy so it is safe to continue
// using the database while a copy is in progress.
func (tx *Tx) CopyFile(path string, mode os.FileMode) error {
f, err := os.OpenFile(path, os.O_RDWR|os.O_CREATE|os.O_TRUNC, mode)
if err != nil {
return err
}
err = tx.Copy(f)
if err != nil {
_ = f.Close()
return err
}
return f.Close()
}
// Check performs several consistency checks on the database for this transaction.
// An error is returned if any inconsistency is found.
//
// It can be safely run concurrently on a writable transaction. However, this
// incurs a high cost for large databases and databases with a lot of subbuckets
// because of caching. This overhead can be removed if running on a read-only
// transaction, however, it is not safe to execute other writer transactions at
// the same time.
func (tx *Tx) Check() <-chan error {
ch := make(chan error)
go tx.check(ch)
return ch
}
func (tx *Tx) check(ch chan error) {
// Check if any pages are double freed.
freed := make(map[pgid]bool)
for _, id := range tx.db.freelist.all() {
if freed[id] {
ch <- fmt.Errorf("page %d: already freed", id)
}
freed[id] = true
}
// Track every reachable page.
reachable := make(map[pgid]*page)
reachable[0] = tx.page(0) // meta0
reachable[1] = tx.page(1) // meta1
for i := uint32(0); i <= tx.page(tx.meta.freelist).overflow; i++ {
reachable[tx.meta.freelist+pgid(i)] = tx.page(tx.meta.freelist)
}
// Recursively check buckets.
tx.checkBucket(&tx.root, reachable, freed, ch)
// Ensure all pages below high water mark are either reachable or freed.
for i := pgid(0); i < tx.meta.pgid; i++ {
_, isReachable := reachable[i]
if !isReachable && !freed[i] {
ch <- fmt.Errorf("page %d: unreachable unfreed", int(i))
}
}
// Close the channel to signal completion.
close(ch)
}
func (tx *Tx) checkBucket(b *Bucket, reachable map[pgid]*page, freed map[pgid]bool, ch chan error) {
// Ignore inline buckets.
if b.root == 0 {
return
}
// Check every page used by this bucket.
b.tx.forEachPage(b.root, 0, func(p *page, _ int) {
if p.id > tx.meta.pgid {
ch <- fmt.Errorf("page %d: out of bounds: %d", int(p.id), int(b.tx.meta.pgid))
}
// Ensure each page is only referenced once.
for i := pgid(0); i <= pgid(p.overflow); i++ {
var id = p.id + i
if _, ok := reachable[id]; ok {
ch <- fmt.Errorf("page %d: multiple references", int(id))
}
reachable[id] = p
}
// We should only encounter un-freed leaf and branch pages.
if freed[p.id] {
ch <- fmt.Errorf("page %d: reachable freed", int(p.id))
} else if (p.flags&branchPageFlag) == 0 && (p.flags&leafPageFlag) == 0 {
ch <- fmt.Errorf("page %d: invalid type: %s", int(p.id), p.typ())
}
})
// Check each bucket within this bucket.
_ = b.ForEach(func(k, v []byte) error {
if child := b.Bucket(k); child != nil {
tx.checkBucket(child, reachable, freed, ch)
}
return nil
})
}
// allocate returns a contiguous block of memory starting at a given page.
func (tx *Tx) allocate(count int) (*page, error) {
p, err := tx.db.allocate(count)
if err != nil {
return nil, err
}
// Save to our page cache.
tx.pages[p.id] = p
// Update statistics.
tx.stats.PageCount++
tx.stats.PageAlloc += count * tx.db.pageSize
return p, nil
}
// write writes any dirty pages to disk.
func (tx *Tx) write() error {
// Sort pages by id.
pages := make(pages, 0, len(tx.pages))
for _, p := range tx.pages {
pages = append(pages, p)
}
sort.Sort(pages)
// Write pages to disk in order.
for _, p := range pages {
size := (int(p.overflow) + 1) * tx.db.pageSize
buf := (*[maxAllocSize]byte)(unsafe.Pointer(p))[:size]
offset := int64(p.id) * int64(tx.db.pageSize)
if _, err := tx.db.ops.writeAt(buf, offset); err != nil {
return err
}
// Update statistics.
tx.stats.Write++
}
if !tx.db.NoSync || IgnoreNoSync {
if err := fdatasync(tx.db); err != nil {
return err
}
}
// Clear out page cache.
tx.pages = make(map[pgid]*page)
return nil
}
// writeMeta writes the meta to the disk.
func (tx *Tx) writeMeta() error {
// Create a temporary buffer for the meta page.
buf := make([]byte, tx.db.pageSize)
p := tx.db.pageInBuffer(buf, 0)
tx.meta.write(p)
// Write the meta page to file.
if _, err := tx.db.ops.writeAt(buf, int64(p.id)*int64(tx.db.pageSize)); err != nil {
return err
}
if !tx.db.NoSync || IgnoreNoSync {
if err := fdatasync(tx.db); err != nil {
return err
}
}
// Update statistics.
tx.stats.Write++
return nil
}
// page returns a reference to the page with a given id.
// If page has been written to then a temporary bufferred page is returned.
func (tx *Tx) page(id pgid) *page {
// Check the dirty pages first.
if tx.pages != nil {
if p, ok := tx.pages[id]; ok {
return p
}
}
// Otherwise return directly from the mmap.
return tx.db.page(id)
}
// forEachPage iterates over every page within a given page and executes a function.
func (tx *Tx) forEachPage(pgid pgid, depth int, fn func(*page, int)) {
p := tx.page(pgid)
// Execute function.
fn(p, depth)
// Recursively loop over children.
if (p.flags & branchPageFlag) != 0 {
for i := 0; i < int(p.count); i++ {
elem := p.branchPageElement(uint16(i))
tx.forEachPage(elem.pgid, depth+1, fn)
}
}
}
// Page returns page information for a given page number.
// This is only safe for concurrent use when used by a writable transaction.
func (tx *Tx) Page(id int) (*PageInfo, error) {
if tx.db == nil {
return nil, ErrTxClosed
} else if pgid(id) >= tx.meta.pgid {
return nil, nil
}
// Build the page info.
p := tx.db.page(pgid(id))
info := &PageInfo{
ID: id,
Count: int(p.count),
OverflowCount: int(p.overflow),
}
// Determine the type (or if it's free).
if tx.db.freelist.freed(pgid(id)) {
info.Type = "free"
} else {
info.Type = p.typ()
}
return info, nil
}
// TxStats represents statistics about the actions performed by the transaction.
type TxStats struct {
// Page statistics.
PageCount int // number of page allocations
PageAlloc int // total bytes allocated
// Cursor statistics.
CursorCount int // number of cursors created
// Node statistics
NodeCount int // number of node allocations
NodeDeref int // number of node dereferences
// Rebalance statistics.
Rebalance int // number of node rebalances
RebalanceTime time.Duration // total time spent rebalancing
// Split/Spill statistics.
Split int // number of nodes split
Spill int // number of nodes spilled
SpillTime time.Duration // total time spent spilling
// Write statistics.
Write int // number of writes performed
WriteTime time.Duration // total time spent writing to disk
}
func (s *TxStats) add(other *TxStats) {
s.PageCount += other.PageCount
s.PageAlloc += other.PageAlloc
s.CursorCount += other.CursorCount
s.NodeCount += other.NodeCount
s.NodeDeref += other.NodeDeref
s.Rebalance += other.Rebalance
s.RebalanceTime += other.RebalanceTime
s.Split += other.Split
s.Spill += other.Spill
s.SpillTime += other.SpillTime
s.Write += other.Write
s.WriteTime += other.WriteTime
}
// Sub calculates and returns the difference between two sets of transaction stats.
// This is useful when obtaining stats at two different points and time and
// you need the performance counters that occurred within that time span.
func (s *TxStats) Sub(other *TxStats) TxStats {
var diff TxStats
diff.PageCount = s.PageCount - other.PageCount
diff.PageAlloc = s.PageAlloc - other.PageAlloc
diff.CursorCount = s.CursorCount - other.CursorCount
diff.NodeCount = s.NodeCount - other.NodeCount
diff.NodeDeref = s.NodeDeref - other.NodeDeref
diff.Rebalance = s.Rebalance - other.Rebalance
diff.RebalanceTime = s.RebalanceTime - other.RebalanceTime
diff.Split = s.Split - other.Split
diff.Spill = s.Spill - other.Spill
diff.SpillTime = s.SpillTime - other.SpillTime
diff.Write = s.Write - other.Write
diff.WriteTime = s.WriteTime - other.WriteTime
return diff
}

424
Godeps/_workspace/src/github.com/boltdb/bolt/tx_test.go generated vendored Normal file
View File

@@ -0,0 +1,424 @@
package bolt_test
import (
"errors"
"fmt"
"os"
"testing"
"github.com/coreos/etcd/Godeps/_workspace/src/github.com/boltdb/bolt"
)
// Ensure that committing a closed transaction returns an error.
func TestTx_Commit_Closed(t *testing.T) {
db := NewTestDB()
defer db.Close()
tx, _ := db.Begin(true)
tx.CreateBucket([]byte("foo"))
ok(t, tx.Commit())
equals(t, tx.Commit(), bolt.ErrTxClosed)
}
// Ensure that rolling back a closed transaction returns an error.
func TestTx_Rollback_Closed(t *testing.T) {
db := NewTestDB()
defer db.Close()
tx, _ := db.Begin(true)
ok(t, tx.Rollback())
equals(t, tx.Rollback(), bolt.ErrTxClosed)
}
// Ensure that committing a read-only transaction returns an error.
func TestTx_Commit_ReadOnly(t *testing.T) {
db := NewTestDB()
defer db.Close()
tx, _ := db.Begin(false)
equals(t, tx.Commit(), bolt.ErrTxNotWritable)
}
// Ensure that a transaction can retrieve a cursor on the root bucket.
func TestTx_Cursor(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
tx.CreateBucket([]byte("woojits"))
c := tx.Cursor()
k, v := c.First()
equals(t, "widgets", string(k))
assert(t, v == nil, "")
k, v = c.Next()
equals(t, "woojits", string(k))
assert(t, v == nil, "")
k, v = c.Next()
assert(t, k == nil, "")
assert(t, v == nil, "")
return nil
})
}
// Ensure that creating a bucket with a read-only transaction returns an error.
func TestTx_CreateBucket_ReadOnly(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.View(func(tx *bolt.Tx) error {
b, err := tx.CreateBucket([]byte("foo"))
assert(t, b == nil, "")
equals(t, bolt.ErrTxNotWritable, err)
return nil
})
}
// Ensure that creating a bucket on a closed transaction returns an error.
func TestTx_CreateBucket_Closed(t *testing.T) {
db := NewTestDB()
defer db.Close()
tx, _ := db.Begin(true)
tx.Commit()
b, err := tx.CreateBucket([]byte("foo"))
assert(t, b == nil, "")
equals(t, bolt.ErrTxClosed, err)
}
// Ensure that a Tx can retrieve a bucket.
func TestTx_Bucket(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
b := tx.Bucket([]byte("widgets"))
assert(t, b != nil, "")
return nil
})
}
// Ensure that a Tx retrieving a non-existent key returns nil.
func TestTx_Get_Missing(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
tx.Bucket([]byte("widgets")).Put([]byte("foo"), []byte("bar"))
value := tx.Bucket([]byte("widgets")).Get([]byte("no_such_key"))
assert(t, value == nil, "")
return nil
})
}
// Ensure that a bucket can be created and retrieved.
func TestTx_CreateBucket(t *testing.T) {
db := NewTestDB()
defer db.Close()
// Create a bucket.
db.Update(func(tx *bolt.Tx) error {
b, err := tx.CreateBucket([]byte("widgets"))
assert(t, b != nil, "")
ok(t, err)
return nil
})
// Read the bucket through a separate transaction.
db.View(func(tx *bolt.Tx) error {
b := tx.Bucket([]byte("widgets"))
assert(t, b != nil, "")
return nil
})
}
// Ensure that a bucket can be created if it doesn't already exist.
func TestTx_CreateBucketIfNotExists(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
b, err := tx.CreateBucketIfNotExists([]byte("widgets"))
assert(t, b != nil, "")
ok(t, err)
b, err = tx.CreateBucketIfNotExists([]byte("widgets"))
assert(t, b != nil, "")
ok(t, err)
b, err = tx.CreateBucketIfNotExists([]byte{})
assert(t, b == nil, "")
equals(t, bolt.ErrBucketNameRequired, err)
b, err = tx.CreateBucketIfNotExists(nil)
assert(t, b == nil, "")
equals(t, bolt.ErrBucketNameRequired, err)
return nil
})
// Read the bucket through a separate transaction.
db.View(func(tx *bolt.Tx) error {
b := tx.Bucket([]byte("widgets"))
assert(t, b != nil, "")
return nil
})
}
// Ensure that a bucket cannot be created twice.
func TestTx_CreateBucket_Exists(t *testing.T) {
db := NewTestDB()
defer db.Close()
// Create a bucket.
db.Update(func(tx *bolt.Tx) error {
b, err := tx.CreateBucket([]byte("widgets"))
assert(t, b != nil, "")
ok(t, err)
return nil
})
// Create the same bucket again.
db.Update(func(tx *bolt.Tx) error {
b, err := tx.CreateBucket([]byte("widgets"))
assert(t, b == nil, "")
equals(t, bolt.ErrBucketExists, err)
return nil
})
}
// Ensure that a bucket is created with a non-blank name.
func TestTx_CreateBucket_NameRequired(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
b, err := tx.CreateBucket(nil)
assert(t, b == nil, "")
equals(t, bolt.ErrBucketNameRequired, err)
return nil
})
}
// Ensure that a bucket can be deleted.
func TestTx_DeleteBucket(t *testing.T) {
db := NewTestDB()
defer db.Close()
// Create a bucket and add a value.
db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
tx.Bucket([]byte("widgets")).Put([]byte("foo"), []byte("bar"))
return nil
})
// Delete the bucket and make sure we can't get the value.
db.Update(func(tx *bolt.Tx) error {
ok(t, tx.DeleteBucket([]byte("widgets")))
assert(t, tx.Bucket([]byte("widgets")) == nil, "")
return nil
})
db.Update(func(tx *bolt.Tx) error {
// Create the bucket again and make sure there's not a phantom value.
b, err := tx.CreateBucket([]byte("widgets"))
assert(t, b != nil, "")
ok(t, err)
assert(t, tx.Bucket([]byte("widgets")).Get([]byte("foo")) == nil, "")
return nil
})
}
// Ensure that deleting a bucket on a closed transaction returns an error.
func TestTx_DeleteBucket_Closed(t *testing.T) {
db := NewTestDB()
defer db.Close()
tx, _ := db.Begin(true)
tx.Commit()
equals(t, tx.DeleteBucket([]byte("foo")), bolt.ErrTxClosed)
}
// Ensure that deleting a bucket with a read-only transaction returns an error.
func TestTx_DeleteBucket_ReadOnly(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.View(func(tx *bolt.Tx) error {
equals(t, tx.DeleteBucket([]byte("foo")), bolt.ErrTxNotWritable)
return nil
})
}
// Ensure that nothing happens when deleting a bucket that doesn't exist.
func TestTx_DeleteBucket_NotFound(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
equals(t, bolt.ErrBucketNotFound, tx.DeleteBucket([]byte("widgets")))
return nil
})
}
// Ensure that Tx commit handlers are called after a transaction successfully commits.
func TestTx_OnCommit(t *testing.T) {
var x int
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
tx.OnCommit(func() { x += 1 })
tx.OnCommit(func() { x += 2 })
_, err := tx.CreateBucket([]byte("widgets"))
return err
})
equals(t, 3, x)
}
// Ensure that Tx commit handlers are NOT called after a transaction rolls back.
func TestTx_OnCommit_Rollback(t *testing.T) {
var x int
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
tx.OnCommit(func() { x += 1 })
tx.OnCommit(func() { x += 2 })
tx.CreateBucket([]byte("widgets"))
return errors.New("rollback this commit")
})
equals(t, 0, x)
}
// Ensure that the database can be copied to a file path.
func TestTx_CopyFile(t *testing.T) {
db := NewTestDB()
defer db.Close()
var dest = tempfile()
db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
tx.Bucket([]byte("widgets")).Put([]byte("foo"), []byte("bar"))
tx.Bucket([]byte("widgets")).Put([]byte("baz"), []byte("bat"))
return nil
})
ok(t, db.View(func(tx *bolt.Tx) error { return tx.CopyFile(dest, 0600) }))
db2, err := bolt.Open(dest, 0600, nil)
ok(t, err)
defer db2.Close()
db2.View(func(tx *bolt.Tx) error {
equals(t, []byte("bar"), tx.Bucket([]byte("widgets")).Get([]byte("foo")))
equals(t, []byte("bat"), tx.Bucket([]byte("widgets")).Get([]byte("baz")))
return nil
})
}
type failWriterError struct{}
func (failWriterError) Error() string {
return "error injected for tests"
}
type failWriter struct {
// fail after this many bytes
After int
}
func (f *failWriter) Write(p []byte) (n int, err error) {
n = len(p)
if n > f.After {
n = f.After
err = failWriterError{}
}
f.After -= n
return n, err
}
// Ensure that Copy handles write errors right.
func TestTx_CopyFile_Error_Meta(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
tx.Bucket([]byte("widgets")).Put([]byte("foo"), []byte("bar"))
tx.Bucket([]byte("widgets")).Put([]byte("baz"), []byte("bat"))
return nil
})
err := db.View(func(tx *bolt.Tx) error { return tx.Copy(&failWriter{}) })
equals(t, err.Error(), "meta copy: error injected for tests")
}
// Ensure that Copy handles write errors right.
func TestTx_CopyFile_Error_Normal(t *testing.T) {
db := NewTestDB()
defer db.Close()
db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
tx.Bucket([]byte("widgets")).Put([]byte("foo"), []byte("bar"))
tx.Bucket([]byte("widgets")).Put([]byte("baz"), []byte("bat"))
return nil
})
err := db.View(func(tx *bolt.Tx) error { return tx.Copy(&failWriter{3 * db.Info().PageSize}) })
equals(t, err.Error(), "error injected for tests")
}
func ExampleTx_Rollback() {
// Open the database.
db, _ := bolt.Open(tempfile(), 0666, nil)
defer os.Remove(db.Path())
defer db.Close()
// Create a bucket.
db.Update(func(tx *bolt.Tx) error {
_, err := tx.CreateBucket([]byte("widgets"))
return err
})
// Set a value for a key.
db.Update(func(tx *bolt.Tx) error {
return tx.Bucket([]byte("widgets")).Put([]byte("foo"), []byte("bar"))
})
// Update the key but rollback the transaction so it never saves.
tx, _ := db.Begin(true)
b := tx.Bucket([]byte("widgets"))
b.Put([]byte("foo"), []byte("baz"))
tx.Rollback()
// Ensure that our original value is still set.
db.View(func(tx *bolt.Tx) error {
value := tx.Bucket([]byte("widgets")).Get([]byte("foo"))
fmt.Printf("The value for 'foo' is still: %s\n", value)
return nil
})
// Output:
// The value for 'foo' is still: bar
}
func ExampleTx_CopyFile() {
// Open the database.
db, _ := bolt.Open(tempfile(), 0666, nil)
defer os.Remove(db.Path())
defer db.Close()
// Create a bucket and a key.
db.Update(func(tx *bolt.Tx) error {
tx.CreateBucket([]byte("widgets"))
tx.Bucket([]byte("widgets")).Put([]byte("foo"), []byte("bar"))
return nil
})
// Copy the database to another file.
toFile := tempfile()
db.View(func(tx *bolt.Tx) error { return tx.CopyFile(toFile, 0666) })
defer os.Remove(toFile)
// Open the cloned database.
db2, _ := bolt.Open(toFile, 0666, nil)
defer db2.Close()
// Ensure that the key exists in the copy.
db2.View(func(tx *bolt.Tx) error {
value := tx.Bucket([]byte("widgets")).Get([]byte("foo"))
fmt.Printf("The value for 'foo' in the clone is: %s\n", value)
return nil
})
// Output:
// The value for 'foo' in the clone is: bar
}

Some files were not shown because too many files have changed in this diff Show More