Oops. Datestamp CHANGES file.

git-svn-id: http://viewvc.tigris.org/svn/viewvc/tags/1.1.19@2886 8cb11bc2-c004-0410-86c3-e597b4017df7
Tag the 1.1.19 final release.
2019-04-16 04:14:59 +03:00 · 2013-04-22 19:21:28 +00:00 · 2013-04-22 19:09:17 +00:00 · 2013-04-19 20:41:18 +00:00 · 2013-04-19 20:35:23 +00:00 · 2013-04-18 14:06:58 +00:00
261 changed files with 9645 additions and 11436 deletions
--- a/193
+++ b/193
@@ -1,4 +1,172 @@
-Version 1.1.0 (released ??-???-????)
+Version 1.1.19 (released 22-Apr-2013)
+
+  * improve root lookup performance (issue #523)
+  * new 'max_filesize_kbytes' config option and handling (issue #524)
+  * tarball generation improvements:
+    - preserve Subversion symlinks in generated tarballs (issue #487)
+    - reduce memory usage of tarball generation logic
+    - fix double compression of generated tarballs (issue #525)
+  * file content handling improvements:
+    - expanded support for encoding detection and transcoding (issue #11)
+    - fix tab-to-space conversion bugs in markup, annotate, and diff views
+    - fix handling of trailing whitespace in diff view
+  * add support for timestamp display in ISO8601 format (issue #46)
+
+Version 1.1.18 (released 28-Feb-2013)
+
+  * fix exception raised by BDB-backed SVN repositories (issue #519)
+  * hide revision-less files when rcsparse is in use
+  * include branchpoints in branch views using rcsparse (issue #347)
+  * miscellaneous cvsdb improvements:
+    - add --port option to make-database (issue #521)
+    - explicitly name columns in queries (issue #522)
+    - update MySQL syntax to avoid discontinued "TYPE=" terms
+
+Version 1.1.17 (released 25-Oct-2012)
+
+  * fix exception caused by uninitialized variable usage (issue #516)
+
+Version 1.1.16 (released 24-Oct-2012)
+
+  * security fix: escape "extra" diff info to avoid XSS attack (issue #515)
+  * add 'binary_mime_types' configuration option and handling (issue #510)
+  * fix 'select for diffs' persistence across log pages (issue #512)
+  * remove lock status and filesize check on directories in remote SVN views
+  * fix bogus 'Annotation of' page title for non-annotated view (issue #514)
+
+Version 1.1.15 (released 22-Jun-2012)
+
+  * security fix: complete authz support for remote SVN views (issue #353)
+  * security fix: log msg leak in SVN revision view with unreadable copy source
+  * fix several instances of incorrect information in remote SVN views
+  * increase performance of some revision metadata lookups in remote SVN views
+  * fix RSS feed regression introduced in 1.1.14
+
+Version 1.1.14 (released 12-Jun-2012)
+
+  * fix annotation of svn files with non-URI-safe paths (issue #504)
+  * handle file:/// Subversion rootpaths as local roots (issue #446)
+  * fix bug caused by trying to case-normalize anon usernames (issue #505)
+  * speed up log handling by reusing tokenization results (issue #506)
+  * add support for custom revision log markup rules (issue #246)
+
+Version 1.1.13 (released 23-Jan-2012)
+
+  * fix svndbadmin failure on deleted paths under Subversion 1.7 (issue #499)
+  * fix annotation of files in svn roots with non-URI-safe paths
+  * fix stray annotation warning in markup display of images
+  * more gracefully handle attempts to display binary content (issue #501)
+
+Version 1.1.12 (released 03-Nov-2011)
+
+  * fix path display in patch and certain diff views (issue #485)
+  * fix broken cvsdb glob searching (issue 486)
+  * allow svn revision specifiers to have leading r's (issue #441, #448) 
+  * allow environmental override of configuration location (issue #494)
+  * fix exception HTML-escaping non-string data under WSGI (issue #454)
+  * add links to root logs from roots view (issue #470)
+  * use Pygments lexer-guessing functionality (issue #495)
+
+Version 1.1.11 (released 17-May-2011)
+
+  * security fix: remove user-reachable override of cvsdb row limit
+  * fix broken standalone.py -c and -d options handling
+  * add --help option to standalone.py
+  * fix stack trace when asked to checkout a directory (issue #478)
+  * improve memory usage and speed of revision log markup (issue #477)
+  * fix broken annotation view in CVS keyword-bearing files (issue #479)
+  * warn users when query results are incomplete (issue #433)
+  * avoid parsing errors on RCS newphrases in the admin section (issue #483)
+  * make rlog parsing code more robust in certain error cases (issue #444)
+
+Version 1.1.10 (released 15-Mar-2011)
+
+  * fix stack trace in Subversion revision info logic (issue #475, issue #476)
+
+Version 1.1.9 (released 18-Feb-2011)
+
+  * vcauth universal access determinations (issue #425)
+  * rework svn revision info cache for performance
+  * make revision log "extra pages" count configurable
+  * fix Subversion 1.4.x revision log compatibility code regression
+  * display sanitized error when authzfile is malformed
+  * restore markup of URLs in file contents (issue #455)
+  * optionally display last-committed metadata in roots view (issue #457)
+
+Version 1.1.8 (released 02-Dec-2010)
+
+  * fix slowness triggered by allow_compress=1 configuration (issue #467)
+  * allow use of 'fcrypt' for Windows standalone.py authn support (issue #471)
+  * yield more useful error on directory markup/annotate request (issue #472)
+
+Version 1.1.7 (released 09-Sep-2010)
+
+  * display Subversion revision properties in the revision view (issue #453)
+  * fix exception in 'standalone.py -r REPOS' when run without a config file
+  * fix standalone.py server root deployments (--script-alias='')
+  * add Basic authentication support to standalone.py (Unix only) (issue #49)
+  * fix obscure "unexpected NULL parent pool" Subversion bindings error
+  * enable path info / link display in remote Subversion root revision view
+  * fix vhost name case handling inconsistency (issue #466)
+  * use svn:mime-type property charset param as encoding hint
+  * markup Subversion revision references in log messages (issue #313)
+  * add rudimentary support for FastCGI-based deployments (issue #464)
+  * fix query script WSGI deployment
+  * add configuration to fix query script cross-linking to ViewVC
+
+Version 1.1.6 (released 02-Jun-2010)
+
+  * add rudimentary support for WSGI-based deployments (issue #397)
+  * fix exception caused by trying to HTML-escape non-string data (issue #454)
+  * fix incorrect RSS feed Content-Type header (issue #449)
+  * fix RSS <title> encoding problem (issue #451)
+  * allow 'svndbadmin purge' to work on missing repositories (issue #452)
+
+Version 1.1.5 (released 29-Mar-2010)
+
+  * security fix: escape user-provided search_re input to avoid XSS attack
+
+Version 1.1.4 (released 10-Mar-2010)
+
+  * security fix: escape user-provided query form input to avoid XSS attack
+  * fix standalone.py failure (when per-root options aren't used) (issue #445)
+  * fix annotate failure caused by ignored svn_config_dir (issue #447)
+
+Version 1.1.3 (released 22-Dec-2009)
+
+  * security fix: add root listing support of per-root authz config
+  * security fix: query.py requires 'forbidden' authorizer (or none) in config
+  * fix URL-ification of truncated log messages (issue #3)
+  * fix regexp input validation (issue #426, #427, #440)
+  * add support for configurable tab-to-spaces conversion
+  * fix not-a-sequence error in diff view
+  * allow viewvc-install to work when templates-contrib is absent
+  * minor template improvements/corrections
+  * expose revision metadata in diff view (issue #431)
+  * markup file/directory item property URLs and email addresses (issue #434)
+  * make ViewVC cross copies in Subversion history by default
+  * fix bug that caused standalone.py failure under Python 1.5.2 (issue #442)
+  * fix support for per-vhost overrides of authorizer parameters (issue #411)
+  * fix root name identification in query.py interface
+
+Version 1.1.2 (released 11-Aug-2009)
+
+  * security fix: validate the 'view' parameter to avoid XSS attack
+  * security fix: avoid printing illegal parameter names and values
+  * add optional support for character encoding detection (issue #400)
+  * fix username case handling in svnauthz module (issue #419)
+  * fix cvsdbadmin/svnadmin rebuild error on missing repos (issue #420)
+  * don't drop leading blank lines from colorized file contents (issue #422)
+  * add file.ezt template logic for optionally hiding binary file contents
+
+Version 1.1.1 (released 03-Jun-2009)
+
+  * fix broken query form (missing required template variables) (issue #416)
+  * fix bug in cvsdb which caused rebuild operations to lose data (issue #417)
+  * fix cvsdb purge/rebuild repos lookup to error on missing repos
+  * fix misleading file contents view page title
+
+Version 1.1.0 (released 13-May-2009)

  * add support for full content diffs (issue #153)
  * make many more data dictionary items available to all views
@@ -12,9 +180,7 @@ Version 1.1.0 (released ??-???-????)
  * add support for query by log message (issues #22, #121)
  * fix bug parsing 'svn blame' output with too-long author names (issue #221)
  * fix default standalone.py port to be within private IANA range (issue #234)
-  * add support for integration with GNU source-highlight (issue #285)
-  * add unified configury of allowed views
-  * add support for disabling the checkout view (now the default state)
+  * add unified configury of allowed views; checkout view disabled by default
  * add support for ranges of revisions to svndbadmin (issue #224)
  * make the query handling more forgiving of malformatted subdirs (issue #244)
  * add support for per-root configuration overrides (issue #371)
@@ -36,7 +202,24 @@ Version 1.1.0 (released ??-???-????)
  * show RSS/query links only for roots found in commits database (issue #357)
  * recognize Subversion svn:mime-type property values (issue #364)
  * hide CVS files when viewing tags/branches on which they don't exist
-  * add support for hiding errorful entries from the directory view (issue #105)
+  * allow hiding of errorful entries from the directory view (issue #105)
+  * fix directory view sorting UI
+  * tolerate malformed Accept-Language headers (issue #396)
+  * allow MIME type mapping overrides in ViewVC configuration (issue #401)
+  * fix exception in rev-sorted remote Subversion directory views (issue #409)
+  * allow setting of page sizes for log and dir views individually (issue #402)
+
+Version 1.0.9 (released 11-Aug-2009)
+
+  * security fix: validate the 'view' parameter to avoid XSS attack
+  * security fix: avoid printing illegal parameter names and values
+
+Version 1.0.8 (released 05-May-2009)
+
+  * fix directory view sorting UI
+  * tolerate malformed Accept-Language headers (issue #396)
+  * fix directory log views in revision-less Subversion repositories
+  * fix exception in rev-sorted remote Subversion directory views (issue #409)

 Version 1.0.7 (released 14-Oct-2008)

--- a/160
+++ b/160
@@ -19,7 +19,7 @@ Congratulations on getting this far. :-)

    For CVS Support: 

-      * Python 1.5.2 or later
+      * Python 1.5.2 or later (sorry, no 3.x support yet)
          (http://www.python.org/)
      * RCS, Revision Control System
          (http://www.cs.purdue.edu/homes/trinkle/RCS/)
@@ -30,11 +30,11 @@ Congratulations on getting this far. :-)

    For Subversion Support:

-      * Python 2.0 or later
+      * Python 2.0 or later (sorry, no 3.x support yet)
          (http://www.python.org/)
      * Subversion, Version Control System, 1.3.1 or later
          (binary installation and Python bindings)
-          (http://subversion.tigris.org/)
+          (http://subversion.apache.org/)

    Optional:

@@ -139,7 +139,7 @@ installation instructions.
      default_root
      root_as_url_component
      rcs_dir
-      mime_types_file
+      mime_types_files

   There are some other options that are usually nice to change. See
   viewvc.conf for more information.  ViewVC provides a working,
@@ -168,72 +168,145 @@ checkin database working are below.
 APACHE CONFIGURATION
 --------------------

-1) Find out where the web server configuration file is kept. Typical
-   locations are /etc/httpd/httpd.conf, /etc/httpd/conf/httpd.conf,
-   and /etc/apache/httpd.conf. Depending on how apache was installed,
-   you may also look under /usr/local/etc or /etc/local. Use the vendor
-   documentation or the find utility if in doubt.
+1) Locate your Apache configuration file(s).

-Either METHOD A: 
-2) The ScriptAlias directive is very useful for pointing 
+   Typical locations are /etc/httpd/httpd.conf,
+   /etc/httpd/conf/httpd.conf, and /etc/apache/httpd.conf. Depending
+   on how Apache was installed, you may also look under /usr/local/etc
+   or /etc/local. Use the vendor documentation or the find utility if
+   in doubt.
+
+2) Depending on how your Apache configuration is setup by default, you
+   might need to explicitly allow high-level access to the ViewVC
+   install location.
+
+      <Directory <VIEWVC_INSTALLATION_DIRECTORY>>
+        Order allow,deny
+        Allow from all
+      </Directory>
+
+   For example, if ViewVC is installed in /usr/local/viewvc-1.0 on
+   your system:
+
+      <Directory /usr/local/viewvc-1.0>
+        Order allow,deny
+        Allow from all
+      </Directory>
+
+3) Configure Apache to expose ViewVC to users at the URL of your choice.
+
+   ViewVC provides several different ways to do this.  Choose one of
+   the following methods:
+
+   -----------------------------------
+   METHOD A:  CGI mode via ScriptAlias
+   -----------------------------------
+   The ScriptAlias directive is very useful for pointing 
   directly to the viewvc.cgi script.  Simply insert a line containing

-     ScriptAlias /viewvc <VIEWVC_INSTALLATION_DIRECTORY>/bin/cgi/viewvc.cgi
+      ScriptAlias /viewvc <VIEWVC_INSTALLATION_DIRECTORY>/bin/cgi/viewvc.cgi

   into your httpd.conf file.  Choose the location in httpd.conf where 
   also the other ScriptAlias lines reside.  Some examples:

-     ScriptAlias /viewvc /usr/local/viewvc-1.0/bin/cgi/viewvc.cgi
-     ScriptAlias /query /usr/local/viewvc-1.0/bin/cgi/query.cgi
+      ScriptAlias /viewvc /usr/local/viewvc-1.0/bin/cgi/viewvc.cgi
+      ScriptAlias /query /usr/local/viewvc-1.0/bin/cgi/query.cgi

-   continue with step 3).
-
-or alternatively METHOD B: 
-2) Copy the CGI scripts from 
+   ----------------------------------------
+   METHOD B:  CGI mode in cgi-bin directory
+   ----------------------------------------
+   Copy the CGI scripts from 
   <VIEWVC_INSTALLATION_DIRECTORY>/bin/cgi/*.cgi
   to the /cgi-bin/ directory configured in your httpd.conf file.

-   continue with step 3).
+   You can override configuration file location using:
+
+       SetEnv VIEWVC_CONF_PATHNAME /etc/viewvc.conf
   
-and then there's METHOD C:
-2) Copy the CGI scripts from 
+   ------------------------------------------
+   METHOD C:  CGI mode in ExecCGI'd directory
+   ------------------------------------------
+   Copy the CGI scripts from 
   <VIEWVC_INSTALLATION_DIRECTORY>/bin/cgi/*.cgi
   to the directory of your choosing in the Document Root adding the following
-   apache directives for the directory in httpd.conf or an .htaccess file:
+   Apache directives for the directory in httpd.conf or an .htaccess file:

-     Options +ExecCGI
-     AddHandler cgi-script .cgi
+      Options +ExecCGI
+      AddHandler cgi-script .cgi
     
-   (Note: For this to work mod_cgi has to be loaded. And for the .htaccess file
+   Note: For this to work mod_cgi has to be loaded.  And for the .htaccess file
   to be effective, "AllowOverride All" or "AllowOverride Options FileInfo"
-   need to have been specified for the directory.)
+   needs to have been specified for the directory.
   
-   continue with step 3).
-
-or if you've got Mod_Python installed you can use METHOD D:
-2) Copy the Python scripts and .htaccess file from 
+   ------------------------------------------
+   METHOD D:  Using mod_python (if installed)
+   ------------------------------------------
+   Copy the Python scripts and .htaccess file from 
   <VIEWVC_INSTALLATION_DIRECTORY>/bin/mod_python/
-   to a directory being served by apache.
+   to a directory being served by Apache.
   
   In httpd.conf, make sure that "AllowOverride All" or at least 
   "AllowOverride FileInfo Options" are enabled for the directory
   you copied the files to.
+
+   You can override configuration file location using:
+
+       SetEnv VIEWVC_CONF_PATHNAME /etc/viewvc.conf
   
   Note: If you are using Mod_Python under Apache 1.3 the tarball generation
   feature may not work because it uses multithreading.  This works fine
   under Apache 2.
   
-   continue with step 3).
+   ----------------------------------------
+   METHOD E:  Using mod_wsgi (if installed)
+   ----------------------------------------
+   Copy the Python scripts file from
+   <VIEWVC_INSTALLATION_DIRECTORY>/bin/mod_python/
+   to the directory of your choosing.  Modify httpd.conf with the
+   following directives:

-3) Restart apache.  The commands to do this vary.  "httpd -k restart" and 
-   "apache -k restart" are two common variants.  On RedHat Linux it is
-   done using the command "/sbin/service httpd restart" and on SuSE Linux
-   it is done with "rcapache restart" 
+      WSGIScriptAlias /viewvc <VIEWVC_INSTALLATION_DIRECTORY>/bin/wsgi/viewvc.wsgi
+      WSGIScriptAlias /query <VIEWVC_INSTALLATION_DIRECTORY>/bin/wsgi/query.wsgi

-4) Optional: Add access control.
+   You'll probably also need the following directive because of the
+   not-quite-sanctioned way that ViewVC manipulates Python objects.

-   In your httpd.conf you can control access to certain modules by adding 
-   directives like this:
+      WSGIApplicationGroup %{GLOBAL}
+
+   Note: WSGI support in ViewVC is at this time quite rudimentary,
+   bordering on downright experimental.  Your mileage may vary.
+
+   -----------------------------------------
+   METHOD F:  Using mod_fcgid (if installed)
+   -----------------------------------------
+
+   This uses ViewVC's WSGI support (from above), but supports using FastCGI,
+   and is a somewhat hybrid approach of several of the above methods.
+
+   Especially if fcgi is already being used for other purposes, e.g. PHP,
+   also using fcgi can prevent the need for including additional modules
+   (e.g. mod_python or mod_wsgi) within Apache, which may help lessen Apache's
+   memory usage and/or help improve performance.
+
+   This depends on mod_fcgid:
+
+      http://httpd.apache.org/mod_fcgid/
+
+   as well as the fcgi server from Python's flup package:
+
+      http://pypi.python.org/pypi/flup
+      http://trac.saddi.com/flup
+
+   The following are some example httpd.conf fragments you can use to
+   support this configuration:
+
+      ScriptAlias /viewvc /usr/local/viewvc/bin/wsgi/viewvc.fcgi
+      ScriptAlias /query /usr/local/viewvc/bin/wsgi/query.fcgi
+
+4) [Optional] Add access control.
+
+   In your httpd.conf you can control access to certain modules by
+   adding directives like this:

      <Location "<url to viewvc.cgi>/<modname_you_wish_to_access_ctl>">
        AllowOverride None
@@ -251,7 +324,14 @@ or if you've got Mod_Python installed you can use METHOD D:
      http://<server_name>/viewvc/~checkout~/<module_name>
      http://<server_name>/viewvc/<module_name>.tar.gz?view=tar

-5) Optional: Protect your ViewVC instance from server-whacking webcrawlers.
+5) Restart Apache.
+
+   The commands to do this vary.  "httpd -k restart" and "apache -k
+   restart" are two common variants.  On RedHat Linux it is done using
+   the command "/sbin/service httpd restart" and on SuSE Linux it is
+   done with "rcapache restart".  Other systems use "apachectl restart".
+
+6) Optional: Protect your ViewVC instance from server-whacking webcrawlers.

   As ViewVC is a web-based application which each page containing various
   links to other pages and views, you can expect your server's performance
--- a/LICENSE.html
+++ b/LICENSE.html
@@ -15,7 +15,7 @@

 <blockquote>

-<p><strong>Copyright &copy; 1999-2008 The ViewCVS Group.  All rights
+<p><strong>Copyright &copy; 1999-2013 The ViewCVS Group.  All rights
   reserved.</strong></p>

 <p>By using ViewVC, you agree to the terms and conditions set forth
@@ -59,6 +59,11 @@
  <li>March 17, 2006 &mdash; software renamed from "ViewCVS"</li>
  <li>April 10, 2007 &mdash; copyright years updated</li>
  <li>February 22, 2008 &mdash; copyright years updated</li>
+  <li>March 18, 2009 &mdash; copyright years updated</li>
+  <li>March 29, 2010 &mdash; copyright years updated</li>
+  <li>February 18, 2011 &mdash; copyright years updated</li>
+  <li>January 23, 2012 &mdash; copyright years updated</li>
+  <li>January 04, 2013 &mdash; copyright years updated</li>
 </ul>

 </body>
--- a/1
+++ b/1
@@ -4,4 +4,3 @@ Please read the file INSTALL for more information.

 And see windows/README for more information on running ViewVC on
 Microsoft Windows.
-
--- a/bin/asp/query.asp
+++ b/bin/asp/query.asp
@@ -3,7 +3,7 @@

 # -*-python-*-
 #
-# Copyright (C) 1999-2006 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -54,7 +54,10 @@ import query
 server = sapi.AspServer(Server, Request, Response, Application)
 try:
  cfg = viewvc.load_config(CONF_PATHNAME, server)
-  query.main(server, cfg, "viewvc.asp")
+  viewvc_base_url = cfg.query.viewvc_base_url
+  if viewvc_base_url is None:
+    viewvc_base_url = "viewvc.asp"
+  query.main(server, cfg, viewvc_base_url)
 finally:
  s.close()

--- a/bin/asp/viewvc.asp
+++ b/bin/asp/viewvc.asp
@@ -3,7 +3,7 @@

 # -*-python-*-
 #
-# Copyright (C) 1999-2006 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
--- a/bin/cgi/query.cgi
+++ b/bin/cgi/query.cgi
@@ -1,7 +1,7 @@
 #!/usr/bin/env python
 # -*-python-*-
 #
-# Copyright (C) 1999-2006 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -54,4 +54,7 @@ import query

 server = sapi.CgiServer()
 cfg = viewvc.load_config(CONF_PATHNAME, server)
-query.main(server, cfg, "viewvc.cgi")
+viewvc_base_url = cfg.query.viewvc_base_url
+if viewvc_base_url is None:
+  viewvc_base_url = "viewvc.cgi"
+query.main(server, cfg, viewvc_base_url)
--- a/bin/cgi/viewvc.cgi
+++ b/bin/cgi/viewvc.cgi
@@ -1,7 +1,7 @@
 #!/usr/bin/env python
 # -*-python-*-
 #
-# Copyright (C) 1999-2006 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
--- a/bin/custis-install
+++ b/bin/custis-install
@@ -1,216 +0,0 @@
-#!/bin/sh
-# ViewVC installation script for CustIS
-
-if [ -f custis-install-config ]; then
-    . custis-install-config
-else
-    cat >custis-install-config <<EOF
-CVS_ASYNC=
-SVN_ASYNC=
-CVS_USER=www-data
-CVS_GROUP=cvs.users
-CVSROOTS=
-SVNROOT=
-SVN_FORBIDDENRE=
-
-DB_HOST=
-DB_PORT=
-DB_SOCKET=/var/run/mysqld/mysqld.sock
-DB=
-DB_USER=
-DB_PASSWD=
-
-VIEWVC_DIR=
-VIEWVC_URI=/
-VIEWVC_URI_SLASH=/
-VIEWVC_STATIC_URI=/static
-http_proxy=
-no_proxy=
-EOF
-    echo Empty 'custis-install-config' initialized, please edit it before installation.
-    exit
-fi
-
-if [ ! "$DB" -o ! "$VIEWVC_DIR" ]; then
-    echo Please set up 'custis-install-config' before installation.
-    exit
-fi
-
-################################################################################
-
-export http_proxy
-export no_proxy
-
-DEPS="python libapache2-mod-python rcs diff cvsnt subversion subversion-tools python-setuptools python-subversion python-mysqldb cvsgraph"
-
-echo "*** Installing dependency packages: $DEPS"
-apt-get install $DEPS
-
-echo "*** Installing Pygments for Python"
-wget http://pypi.python.org/packages/source/P/Pygments/Pygments-0.11.1.tar.gz
-tar -zxf Pygments-0.11.1.tar.gz
-cd Pygments-0.11.1
-python setup.py install
-cd ..
-
-echo "*** Installing ViewVC into $VIEWVC_DIR"
-if [ ! -e $VIEWVC_DIR/viewvc.conf ]; then
-../viewvc-install <<EOF
-$VIEWVC_DIR
-
-
-
-
-
-EOF
-fi
-
-echo "*** Writing ViewVC configuration into $VIEWVC_DIR/viewvc.conf"
-for CVSROOT in $CVSROOTS; do
-    if [ "$vccvsroots" ]; then
-	vccvsroots="$vccvsroots, "
-    fi
-    vcrootname=`basename $CVSROOT`
-    vccvsroots="$vccvsroots$vcrootname: $CVSROOT"
-done;
-cat >"$VIEWVC_DIR/viewvc.conf" <<EOF
-[general]
-cvs_roots = $vccvsroots
-root_parents = $SVNROOT : svn
-cvsnt_exe_path = /usr/bin/cvsnt
-mime_types_file = /etc/mime.types
-address = Admin address: stas [gav-gav] custis [ru]
-kv_files =
-languages = en-us, ru-ru
-
-[utilities]
-rcs_dir = /usr/bin
-cvsnt = /usr/bin/cvsnt
-svn = /usr/bin/svn
-diff = /usr/bin/diff
-cvsgraph = /usr/bin/cvsgraph
-
-[options]
-allowed_views = markup, annotate, roots
-authorizer = forbiddenre
-checkout_magic = 0
-cross_copies = 1
-cvsgraph_conf = $VIEWVC_DIR/cvsgraph.conf
-default_file_view = log
-diff_format = h
-docroot = $VIEWVC_STATIC_URI
-enable_syntax_coloration = 1
-generate_etags = 1
-hide_attic = 1
-hide_cvsroot = 1
-hide_errorful_entries = 0
-hr_breakable = 1
-hr_funout = 0
-hr_ignore_keyword_subst = 1
-hr_ignore_white = 1
-hr_intraline = 0
-http_expiration_time = 600
-limit_changes = 100
-log_sort = date
-mangle_email_addresses = 1
-root_as_url_component = 1
-short_log_len = 80
-show_log_in_markup = 1
-show_logs = 1
-show_subdir_lastmod = 0
-sort_by = file
-sort_group_dirs = 1
-svn_config_dir =
-template_dir = templates
-use_cvsgraph = 1
-use_localtime = 1
-use_pagesize = 0
-use_rcsparse = 0
-use_re_search = 0
-
-[templates]
-
-[cvsdb]
-enabled = 1
-host = $DB_HOST
-socket = $DB_SOCKET
-database_name = $DB
-user = $DB_USER
-passwd = $DB_PASSWD
-readonly_user = $DB_USER
-readonly_passwd = $DB_PASSWD
-
-[vhosts]
-
-[authz-forbidden]
-forbidden =
-
-[authz-forbiddenre]
-forbiddenre = $SVN_FORBIDDENRE
-
-[authz-svnauthz]
-authzfile =
-EOF
-
-echo "*** Initializing database: $DB using $DB_USER@$DB_HOST"
-$VIEWVC_DIR/bin/make-database <<EOF
-$DB_HOST
-$DB_USER
-$DB_PASSWD
-$DB
-EOF
-
-echo "*** Configuring Apache"
-cat >/etc/apache2/sites-available/viewvc <<EOF
-<VirtualHost *:80>
-    ServerName viewvc.office.custis.ru
-    ServerAlias viewvc
-    ServerAdmin sysadmins@custis.ru
-    ServerSignature Off
-    ErrorLog /var/log/apache2/viewvc-error.log
-    TransferLog /var/log/apache2/viewvc-access.log
-    LogLevel warn
-    # ViewVC at $VIEWVC_URI installed at $VIEWVC_DIR
-    DocumentRoot $VIEWVC_DIR/bin/mod_python
-    Alias $VIEWVC_STATIC_URI $VIEWVC_DIR/templates/docroot
-    Alias $VIEWVC_URI $VIEWVC_DIR/bin/mod_python/
-    <Location ~ ^${VIEWVC_URI_SLASH}*(\?.*)?$>
-        Options -Indexes
-        RewriteEngine On
-        RewriteRule .* ${VIEWVC_URI_SLASH}viewvc.py%{REQUEST_URI} [R,L]
-    </Location>
-    <Directory $VIEWVC_DIR/bin/mod_python>
-        Options +ExecCGI
-        AddHandler python-program .py
-        PythonHandler handler
-        PythonDebug Off
-    </Directory>
-</VirtualHost>
-EOF
-a2enmod python
-a2enmod mod_python
-a2enmod rewrite
-a2ensite viewvc
-
-echo "*** Restarting Apache"
-apache2ctl stop
-sleep 1
-apache2ctl start
-
-echo "*** Building commit database for CVS"
-for CVSROOT in $CVSROOTS; do
-    $VIEWVC_DIR/bin/cvsdbadmin rebuild $CVSROOT
-done;
-
-echo "*** Building commit database for Subversion repositories"
-for i in `ls $SVNROOT`; do
-    if [ -d "$SVNROOT/$i" ]; then
-        $VIEWVC_DIR/bin/svndbadmin -v rebuild "$SVNROOT/$i"
-    fi
-done;
-
-# setup hooks for CVS
-./setup-cvs-hooks "$CVSROOTS" "$VIEWVC_DIR" "$CVS_USER" "$CVS_GROUP" "$CVS_ASYNC"
-
-# setup hooks for Subversion
-./setup-svn-hooks "$SVNROOT" "$VIEWVC_DIR" "$SVN_ASYNC"
--- a/bin/cvsdbadmin
+++ b/bin/cvsdbadmin
@@ -1,7 +1,7 @@
-#!/usr/bin/python
+#!/usr/bin/env python
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -44,23 +44,17 @@ import string
 import cvsdb
 import viewvc
 import vclib.ccvs
-from stat import *
-  

-def UpdateFile(db, repository, path, update, latest_checkin, quiet_level):
+
+def UpdateFile(db, repository, path, update, quiet_level):
    try:
        if update:
-            mtime = os.stat(repository.rcsfile(path, 1))[ST_MTIME]
-            if mtime < latest_checkin:
-              return
            commit_list = cvsdb.GetUnrecordedCommitList(repository, path, db)
        else:
            commit_list = cvsdb.GetCommitListFromRCSFile(repository, path)
    except cvsdb.error, e:
        print '[ERROR] %s' % (e)
        return
-    except vclib.ItemNotFound, e:
-        return

    file = string.join(path, "/")
    printing = 0
@@ -83,8 +77,7 @@ def UpdateFile(db, repository, path, update, latest_checkin, quiet_level):
        print


-def RecurseUpdate(db, repository, directory, update, latest_checkin,
-                  quiet_level):
+def RecurseUpdate(db, repository, directory, update, quiet_level):
    for entry in repository.listdir(directory, None, {}):
        path = directory + [entry.name]

@@ -92,13 +85,11 @@ def RecurseUpdate(db, repository, directory, update, latest_checkin,
            continue

        if entry.kind is vclib.DIR:
-            RecurseUpdate(db, repository, path, update, latest_checkin,
-                          quiet_level)
+            RecurseUpdate(db, repository, path, update, quiet_level)
            continue

        if entry.kind is vclib.FILE:
-            UpdateFile(db, repository, path, update, latest_checkin,
-                       quiet_level)
+            UpdateFile(db, repository, path, update, quiet_level)

 def RootPath(path, quiet_level):
    """Break os path into cvs root path and other parts"""
@@ -196,11 +187,8 @@ if __name__ == '__main__':
        if command in ('rebuild', 'update'):
            repository = vclib.ccvs.CVSRepository(None, rootpath, None,
                                                  cfg.utilities, 0)
-            latest_checkin = db.GetLatestCheckinTime(repository)
-            if latest_checkin is None:
-                command = 'rebuild'
            RecurseUpdate(db, repository, path_parts,
-                          command == 'update', latest_checkin, quiet_level)
+                          command == 'update', quiet_level)
    except KeyboardInterrupt:
        print
        print '** break **'
--- a/bin/cvsnt-import-cvsacl
+++ b/bin/cvsnt-import-cvsacl
@@ -1,57 +0,0 @@
-#!/usr/bin/perl
-
-use strict;
-my ($type, $path, $branch, $user, $perm);
-
-my $R = ''; # -R for recursive
-my $lvl = {
-    r => 1,
-    t => 2,
-    w => 3,
-    c => 4,
-    a => 4,
-    p => 5,
-};
-
-while(<>)
-{
-    chomp;
-    next if /^\s*#/so;
-    ($type, $path, $branch, $user, $perm) = split /:/, $_;
-    ($user, $perm) = split /!/, $user;
-    next unless $perm;
-    $perm = [ sort { $lvl->{$b} cmp $lvl->{$a} } split //, $perm ];
-    $perm = $perm->[0];
-    if ($perm eq 't')
-    {
-        $perm = 'read,tag,nowrite,nocreate,nocontrol';
-    }
-    elsif ($perm eq 'r')
-    {
-        $perm = 'read,notag,nowrite,nocreate,nocontrol';
-    }
-    elsif ($perm eq 'w')
-    {
-        $perm = 'read,tag,write,nocreate,nocontrol';
-    }
-    elsif ($perm eq 'c' || $perm eq 'a')
-    {
-        $perm = 'read,tag,write,create,nocontrol';
-    }
-    elsif ($perm eq 'p')
-    {
-        $perm = 'read,tag,write,create,control';
-    }
-    print "cvs rchacl$R -a $perm";
-    print " -u '$user'" if $user ne 'ALL';
-    print " -r '$branch'" if $branch ne 'ALL';
-    if ($path ne 'ALL')
-    {
-        print " '$path'";
-    }
-    else
-    {
-        print ' `ls $CVSROOT | grep -v '."'#cvs'`";
-    }
-    print "\n";
-}
--- a/bin/loginfo-handler
+++ b/bin/loginfo-handler
@@ -1,7 +1,7 @@
 #!/usr/bin/env python
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
--- a/bin/make-database
+++ b/bin/make-database
@@ -1,7 +1,7 @@
 #!/usr/bin/env python
 # -*-python-*-
 #
-# Copyright (C) 1999-2009 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -40,7 +40,7 @@ CREATE TABLE branches (
  branch varchar(64) binary DEFAULT '' NOT NULL,
  PRIMARY KEY (id),
  UNIQUE branch (branch)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS checkins;
 CREATE TABLE checkins (
@@ -62,9 +62,8 @@ CREATE TABLE checkins (
  KEY repositoryid_2 (repositoryid),
  KEY dirid (dirid),
  KEY fileid (fileid),
-  KEY branchid (branchid),
-  KEY descid (descid)
-) TYPE=MyISAM;
+  KEY branchid (branchid)
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS descs;
 CREATE TABLE descs (
@@ -72,9 +71,8 @@ CREATE TABLE descs (
  description text,
  hash bigint(20) DEFAULT '0' NOT NULL,
  PRIMARY KEY (id),
-  KEY hash (hash),
-  FULLTEXT KEY description (description)
-) TYPE=MyISAM;
+  KEY hash (hash)
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS dirs;
 CREATE TABLE dirs (
@@ -82,7 +80,7 @@ CREATE TABLE dirs (
  dir varchar(255) binary DEFAULT '' NOT NULL,
  PRIMARY KEY (id),
  UNIQUE dir (dir)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS files;
 CREATE TABLE files (
@@ -90,7 +88,7 @@ CREATE TABLE files (
  file varchar(255) binary DEFAULT '' NOT NULL,
  PRIMARY KEY (id),
  UNIQUE file (file)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS people;
 CREATE TABLE people (
@@ -98,7 +96,7 @@ CREATE TABLE people (
  who varchar(128) binary DEFAULT '' NOT NULL,
  PRIMARY KEY (id),
  UNIQUE who (who)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS repositories;
 CREATE TABLE repositories (
@@ -106,7 +104,7 @@ CREATE TABLE repositories (
  repository varchar(64) binary DEFAULT '' NOT NULL,
  PRIMARY KEY (id),
  UNIQUE repository (repository)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS tags;
 CREATE TABLE tags (
@@ -120,7 +118,7 @@ CREATE TABLE tags (
  KEY dirid (dirid),
  KEY fileid (fileid),
  KEY branchid (branchid)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;
 """

 ## ------------------------------------------------------------------------
@@ -134,7 +132,7 @@ CREATE TABLE branches (
  branch varchar(64) binary DEFAULT '' NOT NULL,
  PRIMARY KEY (id),
  UNIQUE branch (branch)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS commits;
 CREATE TABLE commits (
@@ -158,7 +156,7 @@ CREATE TABLE commits (
  KEY fileid (fileid),
  KEY branchid (branchid),
  KEY descid (descid)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS descs;
 CREATE TABLE descs (
@@ -167,7 +165,7 @@ CREATE TABLE descs (
  hash bigint(20) DEFAULT '0' NOT NULL,
  PRIMARY KEY (id),
  KEY hash (hash)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS dirs;
 CREATE TABLE dirs (
@@ -175,7 +173,7 @@ CREATE TABLE dirs (
  dir varchar(255) binary DEFAULT '' NOT NULL,
  PRIMARY KEY (id),
  UNIQUE dir (dir)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS files;
 CREATE TABLE files (
@@ -183,7 +181,7 @@ CREATE TABLE files (
  file varchar(255) binary DEFAULT '' NOT NULL,
  PRIMARY KEY (id),
  UNIQUE file (file)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS people;
 CREATE TABLE people (
@@ -191,7 +189,7 @@ CREATE TABLE people (
  who varchar(128) binary DEFAULT '' NOT NULL,
  PRIMARY KEY (id),
  UNIQUE who (who)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS repositories;
 CREATE TABLE repositories (
@@ -199,7 +197,7 @@ CREATE TABLE repositories (
  repository varchar(64) binary DEFAULT '' NOT NULL,
  PRIMARY KEY (id),
  UNIQUE repository (repository)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS tags;
 CREATE TABLE tags (
@@ -213,7 +211,7 @@ CREATE TABLE tags (
  KEY dirid (dirid),
  KEY fileid (fileid),
  KEY branchid (branchid)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;

 DROP TABLE IF EXISTS metadata;
 CREATE TABLE metadata (
@@ -221,7 +219,7 @@ CREATE TABLE metadata (
  value text,
  PRIMARY KEY (name),
  UNIQUE name (name)
-) TYPE=MyISAM;
+) ENGINE=MyISAM;
 INSERT INTO metadata (name, value) VALUES ('version', '1');
 """

@@ -247,6 +245,10 @@ options.)  This script will use the 'mysql' program to create the
 database for you.  You will then need to set the appropriate
 parameters in the [cvsdb] section of your viewvc.conf file.

+NOTE: If a hostname or port is supplied at the command line or during
+interactive prompting, this script will pass '--protocol=TCP' to
+'mysql'.
+
 Options:

  --dbname=ARG        Use ARG as the ViewVC database name to create.
@@ -255,7 +257,8 @@ Options:
  --help              Show this usage message.
  
  --hostname=ARG      Use ARG as the hostname for the MySQL connection.
-                      [Default: localhost]
+
+  --port=ARG          Use ARG as the port for the MySQL connection.

  --password=ARG      Use ARG as the password for the MySQL connection.

@@ -275,10 +278,11 @@ Options:
 if __name__ == "__main__":
  try:
    # Parse the command-line options, if any.
-    dbname = version = hostname = username = password = None
+    dbname = version = hostname = port = username = password = None
    opts, args = getopt.getopt(sys.argv[1:], '', [ 'dbname=',
                                                   'help',
                                                   'hostname=',
+                                                   'port=',
                                                   'password=',
                                                   'username=',
                                                   'version=',
@@ -292,6 +296,8 @@ if __name__ == "__main__":
        dbname = value
      elif name == '--hostname':
        hostname = value
+      elif name == '--port':
+        port = value
      elif name == '--username':
        username = value
      elif name == '--password':
@@ -304,7 +310,9 @@ if __name__ == "__main__":

    # Prompt for information not provided via command-line options.
    if hostname is None:
-      hostname = raw_input("MySQL Hostname [default: localhost]: ") or ""
+      hostname = raw_input("MySQL Hostname (leave blank for default): ")
+    if port is None:
+      port = raw_input("MySQL Port (leave blank for default): ")
    if username is None:
      username = raw_input("MySQL User: ")
    if password is None:
@@ -312,7 +320,7 @@ if __name__ == "__main__":
    if dbname is None:
      dbname = raw_input("ViewVC Database Name [default: ViewVC]: ") or "ViewVC"

-    # Create the database
+    # Create the database.
    dscript = string.replace(DATABASE_SCRIPT_COMMON, "<dbname>", dbname)
    if version == "1.0":
      print BONSAI_COMPAT
@@ -320,16 +328,22 @@ if __name__ == "__main__":
    else:
      dscript = dscript + DATABASE_SCRIPT_VERSION_1

-    host_option = hostname and "--host=%s" % (hostname) or ""
+    # Calculate command arguments.
+    cmd_args = "--user=%s --password=%s" % (username, password)
+    if hostname or port:
+      cmd_args = cmd_args + " --protocol=TCP"
+    if hostname:
+      cmd_args = cmd_args + " --host=%s" % (hostname)
+    if port:
+      cmd_args = cmd_args + " --port=%s" % (port)
+      
    if sys.platform == "win32":
-      cmd = "mysql --user=%s --password=%s %s "\
-            % (username, password, host_option)
+      cmd = "mysql %s" % (cmd_args)
      mysql = os.popen(cmd, "w") # popen2.Popen3 is not provided on windows
      mysql.write(dscript)
      status = mysql.close()
    else:
-      cmd = "{ mysql --user=%s --password=%s %s ; } 2>&1" \
-            % (username, password, host_option)
+      cmd = "{ mysql %s ; } 2>&1" % (cmd_args)
      pipes = popen2.Popen3(cmd)
      pipes.tochild.write(dscript)
      pipes.tochild.close()
--- a/bin/mod_python/.htaccess
+++ b/bin/mod_python/.htaccess
@@ -1,3 +1,3 @@
 AddHandler python-program .py
 PythonHandler handler
-PythonDebug Off
+PythonDebug On
--- a/bin/mod_python/handler.py
+++ b/bin/mod_python/handler.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2006 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
--- a/bin/mod_python/query.py
+++ b/bin/mod_python/query.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -65,7 +65,10 @@ cfg = viewvc.load_config(CONF_PATHNAME)
 def index(req):
  server = sapi.ModPythonServer(req)
  try:
-    query.main(server, cfg, "viewvc.py")
+    viewvc_base_url = cfg.query.viewvc_base_url
+    if viewvc_base_url is None:
+      viewvc_base_url = "viewvc.py"
+    query.main(server, cfg, viewvc_base_url)
  finally:
    server.close()

--- a/bin/mod_python/viewvc.py
+++ b/bin/mod_python/viewvc.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
--- a/bin/setup-cvs-hooks
+++ b/bin/setup-cvs-hooks
@@ -1,57 +0,0 @@
-#!/bin/sh
-
-CVSROOTS=$1
-VIEWVC_DIR=$2
-CVS_USER=$3
-CVS_GROUP=$4
-CVS_ASYNC=$5
-
-if [ ! "$CVSROOTS" -o ! "$VIEWVC_DIR" -o ! "$CVS_USER" -o ! "$CVS_GROUP" ]; then
-    echo "USAGE: $0 <CVS_ROOTS> <VIEWVC_DIR> <CVS_USER> <CVS_GROUP> [CVS_ASYNC]"
-    exit
-fi
-
-echo "*** Setting up commit hooks for CVS repositories (job=$CVS_ASYNC)"
-chgrp $CVS_GROUP $VIEWVC_DIR
-chmod 775 $VIEWVC_DIR
-chmod 1777 .
-for CVSROOT in $CVSROOTS; do
-vcrootname=`basename $CVSROOT`
-if [ "$CVS_ASYNC" ]; then
-    VHOOK="touch $VIEWVC_DIR/.cvs-updated-$vcrootname"
-else
-    VHOOK="$VIEWVC_DIR/bin/cvsdbadmin update $CVSROOT >/dev/null"
-fi;
-su $CVS_USER -s /bin/sh <<EOSH
-CVSROOT=$CVSROOT cvs co CVSROOT
-if [ $? -eq 0 ]; then
-cd CVSROOT
-mv loginfo loginfo.bak
-grep -v '/.cvs-updated' <loginfo.bak | grep -v '/bin/cvsdbadmin update' >loginfo
-cat >>loginfo <<EOF
-ALL $VHOOK
-EOF
-cvs ci -m 'CVS commit hook for ViewVC' loginfo
-cd ..
-rm -Rf CVSROOT
-fi
-EOSH
-if [ "$CVS_ASYNC" ]; then
-    cat >$VIEWVC_DIR/cvshook-$vcrootname <<EOF
-#!/bin/sh
-if [ -e "$VIEWVC_DIR/.cvs-updated-$vcrootname" ]; then
-    rm "$VIEWVC_DIR/.cvs-updated-$vcrootname"
-    $VIEWVC_DIR/bin/cvsdbadmin update $CVSROOT >/dev/null
-fi
-EOF
-    chmod 755 $VIEWVC_DIR/cvshook-$vcrootname
-    cat >/etc/cron.d/viewvc-cvs-$vcrootname <<EOF
-*/5 * * * * root $VIEWVC_DIR/cvshook-$vcrootname
-EOF
-fi
-done;
-
-if [ "$CVS_ASYNC" ]; then
-    echo "*** Reloading cron daemon"
-    /etc/init.d/cron reload
-fi
--- a/bin/setup-svn-hooks
+++ b/bin/setup-svn-hooks
@@ -1,36 +0,0 @@
-#!/bin/sh
-
-SVNROOT=$1
-VIEWVC_DIR=$2
-SVN_ASYNC=$3
-
-if [ ! "$SVNROOT" -o ! "$VIEWVC_DIR" ]; then
-    echo "USAGE: $0 <SVNROOT> <VIEWVC_DIR>"
-    exit
-fi
-
-echo "*** Setting up commit hooks for Subversion"
-for i in `ls $SVNROOT`; do
-    if [ -d "$SVNROOT/$i" ]; then
-        if [ "$SVN_ASYNC" ]; then
-            cat >"$SVNROOT/$i/hooks/post-commit.tmp" <<EOF
-#!/bin/sh
-echo "\$1 \$2" >> $VIEWVC_DIR/.svn-updated
-EOF
-        else
-            cat >"$SVNROOT/$i/hooks/post-commit.tmp" <<EOF
-#!/bin/sh
-$VIEWVC_DIR/bin/svndbadmin update \$1 \$2
-EOF
-        fi
-        chmod 755 "$SVNROOT/$i/hooks/post-commit.tmp"
-        mv "$SVNROOT/$i/hooks/post-commit.tmp" "$SVNROOT/$i/hooks/post-commit"
-    fi
-done;
-
-if [ "$SVN_ASYNC" ]; then
-    cat >/etc/cron.d/viewvc-svn <<EOF
-*/10 * * * * root $VIEWVC_DIR/bin/svnupdate-async.sh "$VIEWVC_DIR"
-EOF
-    /etc/init.d/cron reload
-fi
--- a/bin/standalone.py
+++ b/bin/standalone.py
--- a/bin/svndbadmin
+++ b/bin/svndbadmin
@@ -1,7 +1,7 @@
 #!/usr/bin/env python
 # -*-python-*-
 #
-# Copyright (C) 2004-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 2004-2013 The ViewCVS Group. All Rights Reserved.
 # Copyright (C) 2004-2007 James Henstridge
 #
 # By using this file, you agree to the terms and conditions set forth in
@@ -153,43 +153,46 @@ class SvnRev:
        fsroot = self._get_root_for_rev(rev)
        
        # find changes in the revision
-        editor = svn.repos.RevisionChangeCollector(repo.fs, rev)
+        editor = svn.repos.ChangeCollector(repo.fs, fsroot)
        e_ptr, e_baton = svn.delta.make_editor(editor)
        svn.repos.svn_repos_replay(fsroot, e_ptr, e_baton)

        self.changes = []
-        changes_hash = {}
        for path, change in editor.changes.items():
            # skip non-file changes
            if change.item_kind != svn.core.svn_node_file:
                continue

            # deal with the change types we handle
+            action = None
            base_root = None
+            base_path = change.base_path
            if change.base_path:
                base_root = self._get_root_for_rev(change.base_rev)
-                
-            if not change.path:
+
+            # figure out what kind of change this is, and get a diff
+            # object for it.  note that prior to 1.4 Subversion's
+            # bindings didn't give us change.action, but that's okay
+            # because back then deleted paths always had a change.path
+            # of None.
+            if hasattr(change, 'action') \
+               and change.action == svn.repos.CHANGE_ACTION_DELETE:
+                action = 'remove'
+            elif not change.path:
                action = 'remove'
            elif change.added:
                action = 'add'
            else:
                action = 'change'

-            diffobj = svn.fs.FileDiff(base_root and base_root or None,
-                                      base_root and change.base_path or None,
-                                      change.path and fsroot or None,
-                                      change.path and change.path or None)
+            if action == 'remove':
+                diffobj = svn.fs.FileDiff(base_root, base_path, None, None)
+            else:
+                diffobj = svn.fs.FileDiff(base_root, base_path,
+                                          fsroot, change.path)
+
            diff_fp = diffobj.get_pipe()
            plus, minus = _get_diff_counts(diff_fp)
-
-            # CustIS Bug 50473: a workaround for svnlib behaviour in file movements (FILE1 -> FILE2 + FILE1 -> null)
-            if change.base_path:
-                if not change.path and changes_hash.get(change.base_path, '') != '':
-                    minus = 0
-                elif change.path:
-                    changes_hash[change.base_path] = change.path
-
            self.changes.append((path, action, plus, minus))

    def _get_root_for_rev(self, rev):
@@ -251,6 +254,7 @@ def main(command, repository, revs=[], verbose=0, force=0):
    cfg = viewvc.load_config(CONF_PATHNAME)
    db = cvsdb.ConnectDatabase(cfg)

+    # Purge what must be purged.
    if command in ('rebuild', 'purge'):
        if verbose:
            print "Purging commit info for repository root `%s'" % repository
@@ -261,18 +265,24 @@ def main(command, repository, revs=[], verbose=0, force=0):
                sys.stderr.write("ERROR: " + str(e) + "\n")
                sys.exit(1)

-    repo = SvnRepo(repository)
-    if command == 'rebuild' or (command == 'update' and not revs):
-        for rev in range(repo.rev_max+1):
-            handle_revision(db, command, repo, rev, verbose)
-    elif command == 'update':
-        if revs[0] is None:
-            revs[0] = repo.rev_max
-        if revs[1] is None:
-            revs[1] = repo.rev_max
-        revs.sort()
-        for rev in range(revs[0], revs[1]+1):
-            handle_revision(db, command, repo, rev, verbose, force)
+    # Record what must be recorded.
+    if command in ('rebuild', 'update'):
+        if not os.path.exists(repository):
+            sys.stderr.write('ERROR: could not find repository %s\n'
+                             % (repository))
+            sys.exit(1)
+        repo = SvnRepo(repository)
+        if command == 'rebuild' or (command == 'update' and not revs):
+            for rev in range(repo.rev_max+1):
+                handle_revision(db, command, repo, rev, verbose)
+        elif command == 'update':
+            if revs[0] is None:
+                revs[0] = repo.rev_max
+            if revs[1] is None:
+                revs[1] = repo.rev_max
+            revs.sort()
+            for rev in range(revs[0], revs[1]+1):
+                handle_revision(db, command, repo, rev, verbose, force)

 def _rev2int(r):
    if r == 'HEAD':
@@ -290,7 +300,7 @@ def usage():
 located at REPOS-PATH.

 Usage: 1. %s [-v] rebuild REPOS-PATH
-       2. %s [-v] update REPOS-PATH [REV:[REV2]] [--force]
+       2. %s [-v] update REPOS-PATH [REV[:REV2]] [--force]
       3. %s [-v] purge REPOS-PATH

 1.  Rebuild the commit database information for the repository located
@@ -339,12 +349,6 @@ if __name__ == '__main__':
        sys.stderr.write('ERROR: unknown command %s\n' % command)
        usage()

-    repository = args[2]
-    if not os.path.exists(repository):
-        sys.stderr.write('ERROR: could not find repository %s\n' % args[2])
-        usage()
-    repository = vclib.svn.canonicalize_rootpath(repository)
-
    revs = []
    if len(sys.argv) > 3:
        if command == 'rebuild':
@@ -367,6 +371,7 @@ if __name__ == '__main__':
        rev = None

    try:
+        repository = vclib.svn.canonicalize_rootpath(args[2])
        repository = cvsdb.CleanRepository(os.path.abspath(repository))
        main(command, repository, revs, verbose, force)
    except KeyboardInterrupt:
--- a/bin/svnupdate-async
+++ b/bin/svnupdate-async
@@ -1,50 +0,0 @@
-#!/usr/bin/perl
-# Скрипт для обновления SVN репозиториев svndbadmin-ом
-# Берёт номера ревизий и имена репозиториев из перечисленных файлов или STDIN,
-# группирует их по номерам и выводит список команд, необходимых для обновления
-
-use strict;
-
-# первый аргумент - путь к svndbadmin
-my $svndbadmin = shift @ARGV
-    || die "USAGE: $0 <path_to_svndbadmin> FILES...";
-
-# считываем названия репозиториев и номера ревизий из файла
-my $tou = {};
-my ($repos, $rev);
-while (<>)
-{
-    s/^\s+//so;
-    s/\s+$//so;
-    ($repos, $rev) = split /\s+/, $_;
-    $tou->{$repos}->{$rev} = 1;
-}
-
-# превращаем номера ревизий в диапазоны ревизий
-my ($i, $j, $r, $nr);
-foreach $repos (keys %$tou)
-{
-    $rev = [ sort keys %{$tou->{$repos}} ];
-    $nr = [];
-    $j = 0;
-    for $i (1..@$rev)
-    {
-        if ($i > $#$rev || $rev->[$i]-$rev->[$j] > $i-$j)
-        {
-            $r = $rev->[$j];
-            $r .= ":".$rev->[$i-1] if $i-$j > 1;
-            push @$nr, $r;
-            $j = $i;
-        }
-    }
-    $tou->{$repos} = $nr;
-}
-
-# выводим список команд для выполнения
-foreach $repos (keys %$tou)
-{
-    foreach (@{$tou->{$repos}})
-    {
-        print "$svndbadmin update $repos $_\n";
-    }
-}
--- a/bin/svnupdate-async.sh
+++ b/bin/svnupdate-async.sh
@@ -1,8 +0,0 @@
-#!/bin/sh
-
-VIEWVC_DIR=$1
-
-test -f "$VIEWVC_DIR/.svn-updating" -o ! -f "$VIEWVC_DIR/.svn-updated" && exit 0
-mv "$VIEWVC_DIR/.svn-updated" "$VIEWVC_DIR/.svn-updating"
-"$VIEWVC_DIR/bin/svnupdate-async" "$VIEWVC_DIR/bin/svndbadmin" "$VIEWVC_DIR/.svn-updating" | sh >/dev/null
-rm "$VIEWVC_DIR/.svn-updating"
--- a/bin/wsgi/query.fcgi
+++ b/bin/wsgi/query.fcgi
@@ -0,0 +1,54 @@
+#!/usr/bin/python
+# -*-python-*-
+#
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
+#
+# By using this file, you agree to the terms and conditions set forth in
+# the LICENSE.html file which can be found at the top level of the ViewVC
+# distribution or at http://viewvc.org/license-1.html.
+#
+# For more information, visit http://viewvc.org/
+#
+# -----------------------------------------------------------------------
+#
+# viewvc: View CVS/SVN repositories via a web browser
+#
+# -----------------------------------------------------------------------
+#
+# This is a fcgi entry point for the query ViewVC app. It's appropriate
+# for use with mod_fcgid and flup. It defines a single application function
+# that is a valid fcgi entry point.
+#
+# mod_fcgid: http://httpd.apache.org/mod_fcgid/
+# flup:
+#   http://pypi.python.org/pypi/flup
+#   http://trac.saddi.com/flup
+#
+# -----------------------------------------------------------------------
+
+import sys, os
+
+LIBRARY_DIR = None
+CONF_PATHNAME = None
+
+if LIBRARY_DIR:
+  sys.path.insert(0, LIBRARY_DIR)
+else:
+  sys.path.insert(0, os.path.abspath(os.path.join(sys.argv[0],
+                                                  "../../../lib")))
+
+import sapi
+import viewvc
+import query
+from flup.server import fcgi
+
+def application(environ, start_response):
+  server = sapi.WsgiServer(environ, start_response)
+  cfg = viewvc.load_config(CONF_PATHNAME, server)
+  viewvc_base_url = cfg.query.viewvc_base_url
+  if viewvc_base_url is None:
+    viewvc_base_url = "viewvc.fcgi"
+  query.main(server, cfg, viewvc_base_url)
+  return []
+
+fcgi.WSGIServer(application).run()
--- a/bin/wsgi/query.wsgi
+++ b/bin/wsgi/query.wsgi
@@ -0,0 +1,45 @@
+# -*-python-*-
+#
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
+#
+# By using this file, you agree to the terms and conditions set forth in
+# the LICENSE.html file which can be found at the top level of the ViewVC
+# distribution or at http://viewvc.org/license-1.html.
+#
+# For more information, visit http://viewvc.org/
+#
+# -----------------------------------------------------------------------
+#
+# viewvc: View CVS/SVN repositories via a web browser
+#
+# -----------------------------------------------------------------------
+#
+# This is a wsgi entry point for the query ViewVC app. It's appropriate
+# for use with mod_wsgi. It defines a single application function that
+# is a valid wsgi entry point.
+#
+# -----------------------------------------------------------------------
+
+import sys, os
+
+LIBRARY_DIR = None
+CONF_PATHNAME = None
+
+if LIBRARY_DIR:
+  sys.path.insert(0, LIBRARY_DIR)
+else:
+  sys.path.insert(0, os.path.abspath(os.path.join(sys.argv[0],
+                                                  "../../../lib")))
+
+import sapi
+import viewvc
+import query
+
+def application(environ, start_response):
+  server = sapi.WsgiServer(environ, start_response)
+  cfg = viewvc.load_config(CONF_PATHNAME, server)
+  viewvc_base_url = cfg.query.viewvc_base_url
+  if viewvc_base_url is None:
+    viewvc_base_url = "viewvc.wsgi"
+  query.main(server, cfg, viewvc_base_url)
+  return []
--- a/bin/wsgi/viewvc.fcgi
+++ b/bin/wsgi/viewvc.fcgi
@@ -0,0 +1,50 @@
+#!/usr/bin/python
+# -*-python-*-
+#
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
+#
+# By using this file, you agree to the terms and conditions set forth in
+# the LICENSE.html file which can be found at the top level of the ViewVC
+# distribution or at http://viewvc.org/license-1.html.
+#
+# For more information, visit http://viewvc.org/
+#
+# -----------------------------------------------------------------------
+#
+# viewvc: View CVS/SVN repositories via a web browser
+#
+# -----------------------------------------------------------------------
+#
+# This is a fcgi entry point for the main ViewVC app. It's appropriate
+# for use with mod_fcgid and flup. It defines a single application function
+# that is a valid fcgi entry point.
+#
+# mod_fcgid: http://httpd.apache.org/mod_fcgid/
+# flup:
+#   http://pypi.python.org/pypi/flup
+#   http://trac.saddi.com/flup
+#
+# -----------------------------------------------------------------------
+
+import sys, os
+
+LIBRARY_DIR = None
+CONF_PATHNAME = None
+
+if LIBRARY_DIR:
+  sys.path.insert(0, LIBRARY_DIR)
+else:
+  sys.path.insert(0, os.path.abspath(os.path.join(sys.argv[0],
+                                                  "../../../lib")))
+
+import sapi
+import viewvc
+from flup.server import fcgi
+
+def application(environ, start_response):
+  server = sapi.WsgiServer(environ, start_response)
+  cfg = viewvc.load_config(CONF_PATHNAME, server)
+  viewvc.main(server, cfg)
+  return []
+
+fcgi.WSGIServer(application).run()
--- a/bin/wsgi/viewvc.wsgi
+++ b/bin/wsgi/viewvc.wsgi
@@ -0,0 +1,41 @@
+# -*-python-*-
+#
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
+#
+# By using this file, you agree to the terms and conditions set forth in
+# the LICENSE.html file which can be found at the top level of the ViewVC
+# distribution or at http://viewvc.org/license-1.html.
+#
+# For more information, visit http://viewvc.org/
+#
+# -----------------------------------------------------------------------
+#
+# viewvc: View CVS/SVN repositories via a web browser
+#
+# -----------------------------------------------------------------------
+#
+# This is a wsgi entry point for the main ViewVC app. It's appropriate
+# for use with mod_wsgi. It defines a single application function that
+# is a valid wsgi entry point.
+#
+# -----------------------------------------------------------------------
+
+import sys, os
+
+LIBRARY_DIR = None
+CONF_PATHNAME = None
+
+if LIBRARY_DIR:
+  sys.path.insert(0, LIBRARY_DIR)
+else:
+  sys.path.insert(0, os.path.abspath(os.path.join(sys.argv[0],
+                                                  "../../../lib")))
+
+import sapi
+import viewvc
+
+def application(environ, start_response):
+  server = sapi.WsgiServer(environ, start_response)
+  cfg = viewvc.load_config(CONF_PATHNAME, server)
+  viewvc.main(server, cfg)
+  return []
--- a/conf/viewvc.conf.dist
+++ b/conf/viewvc.conf.dist
--- a/docs/template-authoring-guide.html
+++ b/docs/template-authoring-guide.html
@@ -1,6 +1,6 @@
 <html>
 <head>
-<title>ViewVC 1.0 Template Authoring Guide</title>
+<title>ViewVC 1.1 Template Authoring Guide</title>
 <style>
 body {
  background-color: rgb(180,193,205);
@@ -38,13 +38,13 @@ td {
 </head>
 <body>

-<h1>ViewVC 1.0 Template Authoring Guide</h1>
+<h1>ViewVC 1.1 Template Authoring Guide</h1>

 <div class="h2">
 <h2 id="introduction">Introduction</h2>

 <p>This document represents an (unfinished) attempt at providing
-   documentation for how to customize ViewVC 1.0-dev's HTML output via
+   instructions for how to customize ViewVC's HTML output via
   modification of its templates.</p>

 </div>
@@ -161,12 +161,6 @@ td {
      resource.  Valid only when <var>pathtype</var> is <tt>file</tt>
      or (for Subversion roots) <tt>dir</tt>.</td>
 </tr>
-<tr class="varlevel1">
-  <td class="varname">log_rev_href</td>
-  <td>String</td>
-  <td>Revision number of the file-revision currently being viewed, or
-      None.</td>
-</tr>
 <tr class="varlevel1">
  <td class="varname">nav_path</td>
  <td>List</td>
@@ -321,7 +315,7 @@ td {
 <tr class="varlevel1">
  <td class="varname">pathrev_hidden_values</td>
  <td>List</td>
-  <td>Hidden value name/value pairs for the revision/tag selection form.</td>
+  <td>Hidden field name/value pairs for the revision/tag selection form.</td>
 </tr>
 <tr class="varlevel1">
  <td class="varname">pathrev_clear_action</td>
@@ -331,7 +325,7 @@ td {
 <tr class="varlevel1">
  <td class="varname">pathrev_clear_hidden_values</td>
  <td>List</td>
-  <td>Hidden value name/value pairs for the path revision clear button.</td>
+  <td>Hidden field name/value pairs for the path revision clear button.</td>
 </tr>
 </tbody>
 </table>
@@ -450,10 +444,10 @@ td {
 <tr class="varlevel1">
  <td class="varname">annotation</td>
  <td>String</td>
-  <td>If set, indicates that annotations were requested.  Valid values
-      are "annotated" (annotation was successful), "binary" (file contents
-      are not line-based and human-readable), and "error" (something went
-      wrong during annotation).</td>
+  <td>Valid values are "none" (no annotations were attempted),
+      "annotated" (annotation was successful), "binary" (file contents
+      are not line-based and human-readable), and "error" (something
+      went wrong during annotation).</td>
 </tr>
 <tr class="varlevel1">
  <td class="varname">author</td>
@@ -725,19 +719,29 @@ td {
 <tr class="varlevel1">
  <td class="varname">diff_format_hidden_values</td>
  <td>List</td>
-  <td>Hidden value name/value pairs for the diff format selection form.</td>
+  <td>Hidden field name/value pairs for the diff format selection form.</td>
 </tr>
 <tr class="varlevel1">
  <td class="varname">left</td>
  <td>Container</td>
  <td>Container object for grouping information about the left file.</td>
 </tr>
+<tr class="varlevel2">
+  <td class="varname">left.ago</td>
+  <td>String</td>
+  <td>Text description of the time elapsed since <var>left.date</date>.</td>
+</tr>
 <tr class="varlevel2">
  <td class="varname">left.annotate_href</td>
  <td>String</td>
  <td>URL of the ViewVC annotation view for the left file.
      Valid only when <var>entries.pathtype</var> is <tt>file</tt>.</td>
 </tr>
+<tr class="varlevel2">
+  <td class="varname">left.author</td>
+  <td>String</td>
+  <td>Author of the revision of the left file.</td>
+</tr>
 <tr class="varlevel2">
  <td class="varname">left.date</td>
  <td>String</td>
@@ -755,6 +759,11 @@ td {
  <td>URL to download the HEAD revision of the left file as
      <tt>text/plain</tt>.</td>
 </tr>
+<tr class="varlevel2">
+  <td class="varname">left.log</td>
+  <td>String</td>
+  <td>Log message of the left file revision.</td>
+</tr>
 <tr class="varlevel2">
  <td class="varname">left.path</td>
  <td>String</td>
@@ -778,6 +787,11 @@ td {
      current revision.  Valid only when <var>roottype</var> is
      <tt>svn</tt>.</td>
 </tr>
+<tr class="varlevel2">
+  <td class="varname">left.size</td>
+  <td>String</td>
+  <td>Size of the left file revision, in bytes. Subversion only.</td>
+</tr>
 <tr class="varlevel2">
  <td class="varname">left.tag</td>
  <td>String</td>
@@ -788,6 +802,11 @@ td {
  <td>String</td>
  <td>This is a URL for the markup view of the left file.</td>
 </tr>
+<tr class="varlevel1">
+  <td class="varname">patch_href</td>
+  <td>String</td>
+  <td>URL of the patch view for the file.</td>
+</tr>
 <tr class="varlevel1">
  <td class="varname">raw_diff</td>
  <td>String</td>
@@ -799,12 +818,22 @@ td {
  <td>Container</td>
  <td>Container object for grouping information about the right file.</td>
 </tr>
+<tr class="varlevel2">
+  <td class="varname">right.author</td>
+  <td>String</td>
+  <td>Author of the revision of the right file.</td>
+</tr>
 <tr class="varlevel2">
  <td class="varname">right.annotate_href</td>
  <td>String</td>
  <td>URL of the ViewVC annotation view for the right file.
      Valid only when <var>entries.pathtype</var> is <tt>file</tt>.</td>
 </tr>
+<tr class="varlevel2">
+  <td class="varname">right.author</td>
+  <td>String</td>
+  <td>Author of the revision of the right file.</td>
+</tr>
 <tr class="varlevel2">
  <td class="varname">right.date</td>
  <td>String</td>
@@ -822,6 +851,11 @@ td {
  <td>URL to download the HEAD revision of the right file as
      <tt>text/plain</tt>.</td>
 </tr>
+<tr class="varlevel2">
+  <td class="varname">right.log</td>
+  <td>String</td>
+  <td>Log message of the right file revision.</td>
+</tr>
 <tr class="varlevel2">
  <td class="varname">right.path</td>
  <td>String</td>
@@ -845,6 +879,11 @@ td {
      current revision.  Valid only when <var>roottype</var> is
      <tt>svn</tt>.</td>
 </tr>
+<tr class="varlevel2">
+  <td class="varname">right.size</td>
+  <td>String</td>
+  <td>Size of the right file revision, in bytes. Subversion only.</td>
+</tr>
 <tr class="varlevel2">
  <td class="varname">right.tag</td>
  <td>String</td>
@@ -905,7 +944,7 @@ td {
 <tr class="varlevel1">
  <td class="varname">dir_paging_hidden_values</td>
  <td>List</td>
-  <td>Hidden value name/value pairs for the page selection form.</td>
+  <td>Hidden field name/value pairs for the page selection form.</td>
 </tr>
 <tr class="varlevel1">
  <td class="varname">entries</td>
@@ -1087,23 +1126,16 @@ td {
  <td>String</td>
  <td>Current search expression, if any.</td>
 </tr>
-<tr class="varlevel1">
-  <td class="varname">search_re_form</td>
-  <td>Boolean</td>
-  <td>Indicates whether or not to display the regular expression search
-      form. Value depends on the whether searching is enabled in the
-      configuration and whether or not the current directory is
-      empty.</td>
-</tr>
 <tr class="varlevel1">
  <td class="varname">search_re_action</td>
  <td>String</td>
-  <td>Form action URL for the regular expression search form.</td>
+  <td>Form action URL for the regular expression search form,
+      if searching is available.</td>
 </tr>
 <tr class="varlevel1">
  <td class="varname">search_re_hidden_values</td>
  <td>List</td>
-  <td>Hidden value name/value pairs for the regular expression search form.</td>
+  <td>Hidden field name/value pairs for the regular expression search form.</td>
 </tr>
 <tr class="varlevel1">
  <td class="varname">show_attic_href</td>
@@ -1249,7 +1281,7 @@ td {
 <tr class="varlevel1">
  <td class="varname">diff_select_hidden_values</td>
  <td>List</td>
-  <td>Hidden value name/value pairs for the diff selection form.</td>
+  <td>Hidden field name/value pairs for the diff selection form.</td>
 </tr>
 <tr class="varlevel1">
  <td class="varname">entries</td>
@@ -1530,7 +1562,7 @@ td {
 <tr class="varlevel1">
  <td class="varname">log_paging_hidden_values</td>
  <td>List</td>
-  <td>Hidden value name/value pairs for the page selection form.</td>
+  <td>Hidden field name/value pairs for the page selection form.</td>
 </tr>
 <tr class="varlevel1">
  <td class="varname">logsort</td>
@@ -1546,7 +1578,7 @@ td {
 <tr class="varlevel1">
  <td class="varname">logsort_hidden_values</td>
  <td>List</td>
-  <td>Hidden value name/value pairs for the log sort drop down box</td>
+  <td>Hidden field name/value pairs for the log sort drop down box</td>
 </tr>
 <tr class="varlevel1">
  <td class="varname">mime_type</td>
@@ -1794,6 +1826,14 @@ td {
  <td>Indicates how query results are being sorted. Possible values: 
      <tt>date</tt>, <tt>author</tt>, and <tt>file</tt>.</td>
 </tr>
+<tr class="varlevel1">
+  <td class="varname">row_limit_reached</td>
+  <td>Boolean</td>
+  <td>Indicates whether the internal database row limit threshold (set
+      via the <code>cvsdb.row_limit</code>
+      and <code>cvsdb.rss_row_limit</code> configuration options) was
+      reached by the query.</td>
+</tr>
 <tr class="varlevel1">
  <td class="varname">show_branch</td>
  <td>Boolean</td>
@@ -1910,7 +1950,7 @@ td {
 <tr class="varlevel1">
  <td class="varname">query_hidden_values</td>
  <td>List</td>
-  <td>Hidden value name/value pairs for query form.</td>
+  <td>Hidden field name/value pairs for query form.</td>
 </tr>
 <tr class="varlevel1">
  <td class="varname">querysort</td>
@@ -2050,7 +2090,7 @@ td {
 <tr class="varlevel1">
  <td class="varname">jump_rev_hidden_values</td>
  <td>List</td>
-  <td>Hidden value name/value pairs for revision jump form.</td>
+  <td>Hidden field name/value pairs for revision jump form.</td>
 </tr>
 <tr class="varlevel1">
  <td class="varname">limit_changes</td>
@@ -2073,6 +2113,11 @@ td {
  <td>String</td>
  <td>URL for the current view but with <tt>limit_changes</tt> disabled.</td>
 </tr>
+<tr class="varlevel1">
+  <td class="varname">num_changes</td>
+  <td>String</td>
+  <td>Number of paths changed in this revision.</td>
+</tr>
 <tr class="varlevel1">
  <td class="varname">next_href</td>
  <td>String</td>
@@ -2112,6 +2157,38 @@ td {
  <td>List</td>
  <td>Set of configured viewable repositories.</td>
 </tr>
+<tr class="varlevel2">
+  <td class="varname">roots.ago</td>
+  <td>String</td>
+  <td>Textual description of the time since <var>roots.date</var>.</td>
+</tr>
+<tr class="varlevel2">
+  <td class="varname">roots.author</td>
+  <td>String</td>
+  <td>Username of the last modifier of the root.</td>
+</tr>
+<tr class="varlevel2">
+  <td class="varname">root.date</td>
+  <td>String</td>
+  <td>Date (in UTC if not otherwise configured) of the last
+      modification of the root.</td>
+</tr>
+<tr class="varlevel2">
+  <td class="varname">roots.href</td>
+  <td>String</td>
+  <td>URL of root directory view for a configured repository.</td>
+</tr>
+<tr class="varlevel2">
+  <td class="varname">roots.log</td>
+  <td>String</td>
+  <td>Log message of last modification to the root.</td>
+</tr>
+<tr class="varlevel2">
+  <td class="varname">roots.log_href</td>
+  <td>String</td>
+  <td>URL of log revision view for the top-most (root) directory of
+    the root (repository).</td>
+</tr>
 <tr class="varlevel2">
  <td class="varname">roots.name</td>
  <td>String</td>
@@ -2125,17 +2202,24 @@ td {
      configuration can have negative security implications.  Use this
      token at your own risk.</td>
 </tr>
+<tr class="varlevel2">
+  <td class="varname">roots.rev</td>
+  <td>String</td>
+  <td>Youngest revision of the root.</td>
+</tr>
+<tr class="varlevel2">
+  <td class="varname">roots.short_log</td>
+  <td>String</td>
+  <td>Log message of last modification to the root, truncated to
+      contain no more than the number of characters specified by
+      the <code>short_log_len</code> configuration option.</td>
+</tr>
 <tr class="varlevel2">
  <td class="varname">roots.type</td>
  <td>String</td>
  <td>Version control type of a configured repository.  Valid
      values: <tt>cvs</tt>, <tt>svn</tt>.</td>
 </tr>
-<tr class="varlevel2">
-  <td class="varname">roots.href</td>
-  <td>String</td>
-  <td>URL of root directory view for a configured repository.</td>
-</tr>
 </tbody>
 </table>

--- a/docs/upgrading-howto.html
+++ b/docs/upgrading-howto.html
@@ -137,9 +137,8 @@ td {
       configuration's "general" section.</li>

   <li>Finally, ensure that that the new <code>authorizer</code>
-       option is set to either "forbidden" (which is the default) or
-       "forbiddenre", depending on which of those you were using in
-       ViewVC 1.0.x.</li>
+       option is set to either "forbidden" or "forbiddenre", depending
+       on which of those you were using in ViewVC 1.0.x.</li>

 </ol>

@@ -215,10 +214,9 @@ td {
  <li>options/py2html_path</li>
  <li>options/use_enscript</li>
  <li>options/use_highlight</li>
+  <li>options/use_pagesize</li>
  <li>options/use_php</li>
  <li>options/use_py2html</li>
-  <li>options/use_pygments</li>
-  <li>options/use_source_highlight</li>
 </ul>

 </div>
@@ -226,16 +224,23 @@ td {
 <div class="h3">
 <h3>Checkin Database</h3>

-<p>In ViewVC 1.1, the <code>svndbadmin</code> program's "rebuild"
-   subcommand has had its purpose become more defined.  It no longer
-   accepts a revision argument, and therefore can now only be used to
-   completely rebuild the entirety of the checkin database information
-   for a Subversion repository (instead of being able to only update
-   the information related to single Subversion revision).  For
-   per-revision updating, use <code>svndbadmin update</code> and
+<p>ViewVC 1.1 introduces to the <code>cvsdbadmin</code>
+   and <code>svndbadmin</code> tools a new "purge" operation, which
+   allows you to remove all the information related to a given root
+   from your checkins database (without disturbing the information
+   associated with other roots).  Likewise, the "rebuild" command in
+   those tools now implies a "purge" followed by an update.</p>
+
+<p>As a related change, the <code>svndbadmin</code> program's
+   "rebuild" subcommand has had its purpose become more defined.  It
+   no longer accepts a revision argument, and therefore can now only
+   be used to completely rebuild the entirety of the checkin database
+   information for a Subversion repository (instead of being able to
+   only update the information related to single Subversion revision).
+   For per-revision updating, use <code>svndbadmin update</code> and
   provide a revision (or revision range).  And to get the previous
-   rebuild-a-revision effect, pass the new <code>--force</code>
-   option to <code>svndbadmin update</code>.</p>
+   rebuild-a-revision effect, pass the new <code>--force</code> option
+   to <code>svndbadmin update</code>.</p>

 <p>In other words, where you once did this:</p>

@@ -249,6 +254,19 @@ td {
 </pre>
 </blockquote>

+<p>To enhance the performance of the new "purge" operation, ViewVC 1.1
+   introduces some slight changes to the checkin database schema.  If
+   you use the <code>make-database</code> tool to (re)create your
+   checkins database, it will by default employ the new database
+   schema.  This should cause the database to be virtually unusable by
+   previous versions of ViewVC, and that's by design.  If, however,
+   you need to (re)create your checkins database and you require
+   compatibility with previous versions of ViewVC or ViewCVS, simply
+   pass the "--version=1.0" option to the <code>make-database</code>
+   script.  Note that your "purge" and "rebuild" operations could be
+   abysmally slow, though, as that version of the database schema is
+   not optimized for those operations.</p>
+
 </div>

 <div class="h3">
@@ -296,6 +314,22 @@ td {
  <li>options/allow_tar</li>
 </ul>

+<p>ViewVC now honors the "svn:mime-type" property stored on
+   Subversion-versioned files as the primary source of MIME type
+   determination (before falling back to name-based MIME mappings and
+   such).  However, this can negatively affect the viewability of
+   certain files &mdash; especially images &mdash; whose
+   "svn:mime-type" properties are set incorrectly, such as will happen
+   if Subversion itself merely determines that the file isn't
+   human-readable and uses the "application/octet-stream" MIME type to
+   record this determination.  To make ViewVC <em>not</em> honor the
+   "svn:mime-type" property value, set the <code>svn_ignore_mimetype</code>
+   configuration option.</p>
+
+<p>Speaking of MIME types, the option <code>mime_types_file</code> is
+   now <code>mime_types_files</code>, as it now carries multiple paths
+   to MIME mappings files, ordered by preference.</p>
+
 <p>The <code>use_rcsparse</code> option was moved from the "general"
   section to the "options" section.</p>

@@ -314,7 +348,7 @@ td {
 all = viewvc.*

 [all-options]
-allow_tar = 1
+allowed_views = annotate, diff, markup, tar
 </pre>
 </blockquote>

@@ -324,7 +358,7 @@ allow_tar = 1
 all = viewvc.*

 [vhost-all/options]
-allow_tar = 1
+allowed_views = annotate, diff, markup, tar
 </pre>
 </blockquote>

@@ -631,6 +665,11 @@ allow_tar = 1
  <td>revision.ezt</td>
  <td>now is an iterable list of objects with .name and .value attributes</td>
 </tr>
+<tr class="added">
+  <td class="varname">num_changes</td>
+  <td>revision.ezt</td>
+  <td>added</td>
+</tr>
 </tbody>
 </table>

@@ -659,7 +698,7 @@ allow_tar = 1
 <h3>Checkin Database</h3>
    
 <p>ViewVC 1.0 reads and writes commit times in the MySQL database in
-   UTC time rather than local time. This can cause times displayed on
+   UTC time rather than local time.  This can cause times displayed on
   the query page to be a few hours off if an old database is being
   used with a new version of ViewVC.  The best way to fix this is to
   rebuild the database with the new version of cvsdbadmin, but it
@@ -702,7 +741,8 @@ allow_tar = 1
  <li>options/root_as_url_component</li>
  <li>options/default_file_view</li>
  <li>options/sort_group_dirs</li>
-  <li>options/use_pagesize</li>
+  <li>options/dir_pagesize</li>
+  <li>options/log_pagesize</li>
  <li>options/limit_changes</li>
  <li>options/use_localtime</li>
  <li>options/cross_copies</li>
--- a/docs/url-reference.html
+++ b/docs/url-reference.html
@@ -996,7 +996,7 @@ th.caption {
    <td>file query string</td>
  </tr>
  <tr>
-    <td><code>file_match=FILE_MATCH</code></td>
+    <td><code>file_match=<var>FILE_MATCH</var></code></td>
    <td>optional</td>
    <td>"exact" "like" "glob" "regex" or "notregex" determining type
        of file match</td>
@@ -1007,7 +1007,7 @@ th.caption {
    <td>author query string</td>
  </tr>
  <tr>
-    <td><code>who_match=WHO_MATCH</code></td>
+    <td><code>who_match=<var>WHO_MATCH</var></code></td>
    <td>optional</td>
    <td>"exact" "like" "glob" "regex" or "notregex" determining type
        of author match</td>
@@ -1024,36 +1024,36 @@ th.caption {
        of log message match</td>
  </tr>
  <tr>
-    <td><code>querysort=SORT</code></td>
+    <td><code>querysort=<var>SORT</var></code></td>
    <td>optional</td>
    <td>"date" "author" or "file" determining order of query results</td>
  </tr>
  <tr>
-    <td><code>date=DATE</code></td>
+    <td><code>date=<var>DATE</var></code></td>
    <td>optional</td>
    <td>"hours" "day" "week" "month" "all" or "explicit" to filter
        query results by date</td>
  </tr>
  <tr>
-    <td><code>hours=HOURS</code></td>
+    <td><code>hours=<var>HOURS</var></code></td>
    <td>optional</td>
    <td>number of hours back to include results from when
        <code><var>DATE</var></code> is "hours"</td>
  </tr>
  <tr>
-    <td><code>mindate=MINDATE</code></td>
+    <td><code>mindate=<var>MINDATE</var></code></td>
    <td>optional</td>
    <td>earliest date to include results from when
        <code><var>DATE</var></code> is "explicit"</td>
  </tr>
  <tr>
-    <td><code>maxdate=MAXDATE</code></td>
+    <td><code>maxdate=<var>MAXDATE</var></code></td>
    <td>optional</td>
    <td>latest date to include results from when
        <code><var>DATE</var></code> is "explicit"</td>
  </tr>
  <tr>
-    <td><code>limit_changes=LIMIT_CHANGES</code></td>
+    <td><code>limit_changes=<var>LIMIT_CHANGES</var></code></td>
    <td>optional</td>
    <td>maximum number of files to list per commit in query
        results. Default is value of <code>limit_changes</code>
@@ -1113,7 +1113,7 @@ th.caption {
    <td>branch query string</td>
  </tr>
  <tr>
-    <td><code>branch_match=BRANCH_MATCH</code></td>
+    <td><code>branch_match=<var>BRANCH_MATCH</var></code></td>
    <td>optional</td>
    <td>"exact" "like" "glob" "regex" or "notregex" determining type
        of branch match</td>
@@ -1129,7 +1129,7 @@ th.caption {
    <td>file query string</td>
  </tr>
  <tr>
-    <td><code>file_match=FILE_MATCH</code></td>
+    <td><code>file_match=<var>FILE_MATCH</var></code></td>
    <td>optional</td>
    <td>"exact" "like" "glob" "regex" or "notregex" determining type
        of file match</td>
@@ -1140,7 +1140,7 @@ th.caption {
    <td>author query string</td>
  </tr>
  <tr>
-    <td><code>who_match=WHO_MATCH</code></td>
+    <td><code>who_match=<var>WHO_MATCH</var></code></td>
    <td>optional</td>
    <td>"exact" "like" "glob" "regex" or "notregex" determining type
        of author match</td>
@@ -1157,50 +1157,43 @@ th.caption {
        of log message match</td>
  </tr>
  <tr>
-    <td><code>querysort=SORT</code></td>
+    <td><code>querysort=<var>SORT</var></code></td>
    <td>optional</td>
    <td>"date" "author" or "file" determining order of query results</td>
  </tr>
  <tr>
-    <td><code>date=DATE</code></td>
+    <td><code>date=<var>DATE</var></code></td>
    <td>optional</td>
    <td>"hours" "day" "week" "month" "all" or "explicit" to filter
        query results by date</td>
  </tr>
  <tr>
-    <td><code>hours=HOURS</code></td>
+    <td><code>hours=<var>HOURS</var></code></td>
    <td>optional</td>
    <td>number of hours back to include results from when
        <code><var>DATE</var></code> is "hours"</td>
  </tr>
  <tr>
-    <td><code>mindate=MINDATE</code></td>
+    <td><code>mindate=<var>MINDATE</var></code></td>
    <td>optional</td>
    <td>earliest date to include results from when
        <code><var>DATE</var></code> is "explicit"</td>
  </tr>
  <tr>
-    <td><code>maxdate=MAXDATE</code></td>
+    <td><code>maxdate=<var>MAXDATE</var></code></td>
    <td>optional</td>
    <td>latest date to include results from when
        <code><var>DATE</var></code> is "explicit"</td>
  </tr>
  <tr>
-    <td><code>format=FORMAT</code></td>
+    <td><code>format=<var>FORMAT</var></code></td>
    <td>optional</td>
    <td>"rss" or "backout" values to generate an rss feed or list of
        commands to back out changes instead showing a normal query result
        page</td>
  </tr>
  <tr>
-    <td><code>limit=LIMIT</code></td>
-    <td>optional</td>
-    <td>maximum number of file-revisions to process during a
-        query. Default is value of <code>row_limit</code> configuration
-        option</td>
-  </tr>
-  <tr>
-    <td><code>limit_changes=LIMIT_CHANGES</code></td>
+    <td><code>limit_changes=<var>LIMIT_CHANGES</var></code></td>
    <td>optional</td>
    <td>maximum number of files to list per commit in query
        results. Default is value of <code>limit_changes</code>
@@ -1254,7 +1247,7 @@ th.caption {
    <td><a href="#revision-param"><code>revision</code> parameter</a></td>
  </tr>
  <tr>
-    <td><code>limit_changes=LIMIT_CHANGES</code></td>
+    <td><code>limit_changes=<var>LIMIT_CHANGES</var></code></td>
    <td>optional</td>
    <td>maximum number of files to list per commit. Default is value
        of <code>limit_changes</code> configuration option</td>
--- a/elemx/Makefile
+++ b/elemx/Makefile
@@ -0,0 +1,54 @@
+TARGETS = python/elx-python java/elx-java
+
+all : $(TARGETS)
+
+CFLAGS = -g -Wpointer-arith -Wwrite-strings -Wshadow -Wall
+CPPFLAGS = -Ipython -Ijava -I.
+
+# the scanner depends on tokens generated from python.y
+python/scanner.c : python/python.c
+
+# the keywords also need the tokens in python.h
+python/py_keywords.c : python/python.c
+
+# we need the scanner tokens in py_scan.h and keywords in py_keywords.h
+python/elx-python.o : python/elx-python.c python/py_keywords.c
+
+# we need java.[ch] generated first to get the tokens in java.h
+java/j_scan.c : java/java.c java/j_keywords.c
+
+# the keywords also need the tokens in java.h
+java/j_keywords.c : java/java.c
+
+# we need the scanner tokens in j_scan.h and keywords in j_keywords.h
+java/elx-java.o : java/elx-java.c java/j_keywords.c java/java.c
+
+
+python/elx-python : python/elx-python.o python/scanner.o python/python.o \
+			python/py_keywords.o elx-common.o
+	$(CC) -o $@ $^
+
+java/elx-java : java/elx-java.o java/j_scan.o java/java.o java/j_keywords.o \
+			elx-common.o
+	$(CC) -o $@ $^
+
+clean :
+	rm -f *.o
+	rm -f python/*.o python/python.[ch] python/py_keywords.[ch]
+	rm -f java/*.o java/java.[ch] java/j_keywords.[ch] java/j_scan.[ch]
+	rm -f python/*.output java/*.output
+	rm -f $(TARGETS)
+
+.SUFFIXES:
+.SUFFIXES: .c .y .shilka .o
+
+%.c : %.y
+	@d="`echo $@ | sed 's/\.c//'`" ; \
+	echo msta -d -enum -v -o $$d $< ; \
+	msta -d -enum -v -o $$d $<
+
+%.c : %.shilka
+	@d="`echo $@ | sed 's,/[^/]*$$,,'`" ; \
+	f="`echo $< | sed 's,.*/,,'`" ; \
+	echo shilka -length -no-definitions -interface $$f ; \
+	(cd $$d && shilka -length -no-definitions -interface $$f)
--- a/elemx/elx-common.c
+++ b/elemx/elx-common.c
@@ -0,0 +1,110 @@
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <errno.h>
+
+#include "elx.h"
+
+#define ELX_ELEMS_EXT   ".elx"
+#define ELX_SYMBOLS_EXT ".els"
+
+
+static void usage(const char *progname)
+{
+    fprintf(stderr, "USAGE: %s FILENAME\n", progname);
+    exit(1);
+}
+
+static const char * build_one(const char *base, int len, const char *suffix)
+{
+    int slen = strlen(suffix);
+    char *fn;
+
+    fn = malloc(len + slen + 1);
+    memcpy(fn, base, len);
+    memcpy(fn + len, suffix, slen);
+    fn[len + slen] = '\0';
+
+    return fn;
+}
+
+elx_context_t *elx_process_args(int argc, const char **argv)
+{
+    elx_context_t *ec;
+    const char *input_fn;
+    const char *p;
+    int len;
+
+    /* ### in the future, we can expand this for more options */
+
+    if (argc != 2)
+    {
+        usage(argv[0]);
+        /* NOTREACHED */
+    }
+
+    input_fn = argv[1];
+
+    p = strrchr(input_fn, '.');
+    if (p == NULL)
+        len = strlen(input_fn);
+    else
+        len = p - argv[1];
+
+    ec = malloc(sizeof(*ec));
+    ec->input_fn = input_fn;
+    ec->elx_fn = build_one(input_fn, len, ELX_ELEMS_EXT);
+    ec->sym_fn = build_one(input_fn, len, ELX_SYMBOLS_EXT);
+
+    return ec;
+}
+
+void elx_open_files(elx_context_t *ec)
+{
+    const char *fn;
+    const char *op;
+
+    if ((ec->input_fp = fopen(ec->input_fn, "r")) == NULL)
+    {
+        fn = ec->input_fn;
+        op = "reading";
+        goto error;
+    }
+    if ((ec->elx_fp = fopen(ec->elx_fn, "w")) == NULL)
+    {
+        fn = ec->elx_fn;
+        op = "writing";
+        goto error;
+    }
+    if ((ec->sym_fp = fopen(ec->sym_fn, "w")) == NULL)
+    {
+        fn = ec->sym_fn;
+        op = "writing";
+        goto error;
+    }
+    return;
+
+  error:
+    fprintf(stderr, "ERROR: file \"%s\" could not be opened for %s.\n       %s\n",
+            fn, op, strerror(errno));
+    exit(2);
+}
+
+void elx_close_files(elx_context_t *ec)
+{
+    fclose(ec->input_fp);
+    fclose(ec->elx_fp);
+    fclose(ec->sym_fp);
+}
+
+void elx_issue_token(elx_context_t *ec,
+                     char which, int start, int len,
+                     const char *symbol)
+{
+    fprintf(ec->elx_fp, "%c %d %d\n", which, start, len);
+
+    if (ELX_DEFINES_SYM(which))
+    {
+        fprintf(ec->sym_fp, "%s %d %s\n", symbol, start, ec->input_fn);
+    }
+}
--- a/elemx/elx.h
+++ b/elemx/elx.h
@@ -0,0 +1,54 @@
+#ifndef ELX_H
+#define ELX_H
+
+#include <stdio.h>
+
+#ifdef __cplusplus
+extern "C" {
+#endif /* __cplusplus */
+
+
+#define ELX_COMMENT     'C'     /* a comment */
+#define ELX_STRING      'S'     /* a string constant */
+#define ELX_KEYWORD     'K'     /* a language keyword */
+#define ELX_GLOBAL_FDEF 'F'     /* function defn in global (visible) scope */
+#define ELX_LOCAL_FDEF  'L'     /* function defn in local (hidden) scope */
+#define ELX_METHOD_DEF  'M'     /* method definition */
+#define ELX_FUNC_REF    'R'     /* function reference / call */
+
+#define ELX_DEFINES_SYM(c) ((c) == ELX_GLOBAL_FDEF || (c) == ELX_LOCAL_FDEF \
+                            || (c) == ELX_METHOD_DEF)
+
+
+typedef struct
+{
+    /* input filename */
+    const char *input_fn;
+
+    /* output filenames: element extractions, and symbols */
+    const char *elx_fn;
+    const char *sym_fn;
+
+    /* file pointers for each of the input/output files */
+    FILE *input_fp;
+    FILE *elx_fp;
+    FILE *sym_fp;
+
+} elx_context_t;
+
+elx_context_t *elx_process_args(int argc, const char **argv);
+
+void elx_open_files(elx_context_t *ec);
+void elx_close_files(elx_context_t *ec);
+
+
+void elx_issue_token(elx_context_t *ec,
+                     char which, int start, int len,
+                     const char *symbol);
+
+
+#ifdef __cplusplus
+}
+#endif /* __cplusplus */
+
+#endif /* ELX_H */
--- a/elemx/elx_html.py
+++ b/elemx/elx_html.py
@@ -0,0 +1,121 @@
+#!/usr/bin/env python
+#
+# generate HTML given an input file and an element file
+#
+
+import re
+import string
+import cgi
+import struct
+
+_re_elem = re.compile('([a-zA-Z]) ([0-9]+) ([0-9]+)\n')
+
+CHUNK_SIZE = 98304 # 4096*24
+
+
+class ElemParser:
+  "Parse an elements file, extracting the token type, start, and length."
+
+  def __init__(self, efile):
+    self.efile = efile
+
+  def get(self):
+    line = self.efile.readline()
+    if not line:
+      return None, None, None
+    t, s, e = string.split(line)
+    return t, int(s)-1, int(e)
+
+  def unused_get(self):
+    record = self.efile.read(9)
+    if not record:
+      return None, None, None
+    return struct.unpack('>cii', record)
+
+
+class Writer:
+  "Generate output, including copying from another input."
+
+  def __init__(self, ifile, ofile):
+    self.ifile = ifile
+    self.ofile = ofile
+
+    self.buf = ifile.read(CHUNK_SIZE)
+    self.offset = 0
+
+  def write(self, data):
+    self.ofile.write(data)
+
+  def copy(self, pos, amt):
+    "Copy 'amt' bytes from position 'pos' of input to output."
+    idx = pos - self.offset
+    self.ofile.write(cgi.escape(buffer(self.buf, idx, amt)))
+    amt = amt - (len(self.buf) - idx)
+    while amt > 0:
+      self._more()
+      self.ofile.write(cgi.escape(buffer(self.buf, 0, amt)))
+      amt = amt - len(self.buf)
+
+  def flush(self, pos):
+    "Flush the rest of the input to the output."
+    idx = pos - self.offset
+    self.ofile.write(cgi.escape(buffer(self.buf, idx)))
+    while 1:
+      buf = self.ifile.read(CHUNK_SIZE)
+      if not buf:
+        break
+      self.ofile.write(cgi.escape(buf))
+
+  def _more(self):
+    self.offset = self.offset + len(self.buf)
+    self.buf = self.ifile.read(CHUNK_SIZE)
+
+
+def generate(input, elems, output, genpage=0):
+  ep = ElemParser(elems)
+  w = Writer(input, output)
+  cur = 0
+  if genpage:
+    w.write('''\
+<html><head><title>ELX Output Page</title>
+<style type="text/css">
+  .elx_C { color: firebrick; font-style: italic; }
+  .elx_S { color: #bc8f8f; font-weight: bold; }
+  .elx_K { color: purple; font-weight: bold }
+  .elx_F { color: blue; font-weight: bold; }
+  .elx_L { color: blue; font-weight: bold; }
+  .elx_M { color: blue; font-weight: bold; }
+  .elx_R { color: blue; font-weight: bold; }
+</style>
+</head>
+<body>
+''')
+  w.write('<pre>')
+  while 1:
+    type, start, length = ep.get()
+    if type is None:
+      break
+    if cur < start:
+      # print out some plain text up to 'cur'
+      w.copy(cur, start - cur)
+
+    # wrap a bit o' formatting here
+    w.write('<span class="elx_%s">' % type)
+
+    # copy over the token
+    w.copy(start, length)
+
+    # and close up the formatting
+    w.write('</span>')
+
+    cur = start + length
+
+  # all done.
+  w.flush(cur)
+  w.write('</pre>')
+  if genpage:
+    w.write('</body></html>\n')
+
+if __name__ == '__main__':
+  import sys
+  generate(open(sys.argv[1]), open(sys.argv[2]), sys.stdout, 1)
--- a/elemx/elx_page.sh
+++ b/elemx/elx_page.sh
@@ -0,0 +1,26 @@
+#!/bin/sh
+
+if test "$#" != 2; then
+  echo "USAGE: $0 SOURCE-FILE ELX-FILE"
+  exit 1
+fi
+
+cat <<EOF
+<html><head><title>ELX Output Page</title>
+<style type="text/css">
+  .elx_C { color: firebrick; font-style: italic; }
+  .elx_S { color: #bc8f8f; font-weight: bold; }
+  .elx_K { color: purple; font-weight: bold }
+  .elx_F { color: blue; font-weight: bold; }
+  .elx_L { color: blue; font-weight: bold; }
+  .elx_M { color: blue; font-weight: bold; }
+  .elx_R { color: blue; font-weight: bold; }
+</style>
+</head>
+<body>
+EOF
+
+dirname="`dirname $0`"
+python2 $dirname/elx_html.py $1 $2
+
+echo "</body></html>"
--- a/elemx/java/elx-java.c
+++ b/elemx/java/elx-java.c
@@ -0,0 +1,148 @@
+#include <stdio.h>
+#include <stdlib.h>
+
+#include "java.h"
+#include "j_keywords.h"
+#include "elx.h"
+
+/* from j_scan.c */
+extern int yylex(void);
+extern void yylex_start(int *error_flag);
+extern void yylex_finish(void);
+extern const char *get_identifier(void);
+
+static const char *fname;
+static int saw_error = 0;
+
+static int lineno = 1;
+static int hpos = 1;
+static int fpos = 0;
+
+static int token_start = 0;
+static int start_lineno;
+static int start_hpos;
+
+static elx_context_t *ectx;
+
+
+//#define DEBUG_SCANNER
+
+/* if we're debugging, then the scanner looks for this var */
+int yysdebug = 0;
+
+/* and the parser looks for this */
+int yydebug = 1;
+
+
+void yyserror(const char *msg)
+{
+    fprintf(stderr, "%s:%d:%d: lex error: %s\n",
+            fname, start_lineno, start_hpos, msg);
+    saw_error = 1;
+}
+
+void yyerror(const char *msg)
+{
+    fprintf(stderr, "%s:%d:%d: parse error: %s\n",
+            fname, start_lineno, start_hpos, msg);
+    saw_error = 1;
+}
+
+int yyslex(void)
+{
+    int c = fgetc(ectx->input_fp);
+
+    if (c == EOF)
+        return -1;      /* tell lexer we're done */
+
+    ++fpos;
+    if (c == '\n')
+    {
+        hpos = 1;
+        ++lineno;
+    }
+    else
+        ++hpos;
+
+//    printf("char: '%c'\n", c);
+
+    return c;
+}
+
+void issue_token(char which)
+{
+    const char *ident = NULL;
+
+    if (ELX_DEFINES_SYM(which))
+        ident = get_identifier();
+    else
+        ident = NULL;
+
+    elx_issue_token(ectx, which, token_start, fpos - token_start + 1, ident);
+}
+
+void mark_token_start(void)
+{
+    token_start = fpos;
+    start_lineno = lineno;
+    start_hpos = hpos;
+}
+
+#ifdef DEBUG_SCANNER
+
+void gen_scan_tokens(void)
+{
+    while (1)
+    {
+        int v = yylex();
+
+        if (v == TK_IDENTIFIER)
+            printf("%d-%d: %d '%s'\n",
+                   token_start, fpos-1, v, get_identifier());
+        else
+            printf("%d-%d: %d\n", token_start, fpos-1, v);
+
+        /* end of parse? */
+        if (v <= 0)
+            break;
+    }
+}
+
+#else /* DEBUG_SCANNER */
+
+static void gen_elx_tokens(void)
+{
+    /* ### what to do with the result? should have seen/set saw_error */
+    (void) yyparse();
+}
+
+#endif /* DEBUG_SCANNER */
+
+int main(int argc, const char **argv)
+{
+    int errcode;
+
+    ectx = elx_process_args(argc, argv);
+
+    yylex_start(&errcode);
+    if (errcode)
+    {
+        fprintf(stderr, "error: yylex_start: %d\n", errcode);
+        return EXIT_FAILURE;
+    }
+
+    elx_open_files(ectx);
+
+#ifdef DEBUG_SCANNER
+    gen_scan_tokens();
+#else
+    gen_elx_tokens();
+#endif
+
+    yylex_finish();
+    elx_close_files(ectx);
+
+    if (saw_error)
+        return EXIT_FAILURE;
+    return EXIT_SUCCESS;
+}
--- a/elemx/java/j_keywords.shilka
+++ b/elemx/java/j_keywords.shilka
@@ -0,0 +1,65 @@
+%local {
+/* get the KR_* values */
+#include "java.h"
+}
+
+%%
+abstract
+boolean
+break
+byte
+/* byvalue */
+case
+/* cast */
+catch
+char
+class
+/* const */
+continue
+default
+do
+double
+else
+extends
+false
+final
+finally
+float
+for
+/* future */
+/* generic */
+/* goto */
+if
+implements
+import
+/* inner */
+instanceof
+int
+interface
+long
+native
+new
+null
+/* operator */
+/* outer */
+package
+private
+protected
+public
+/* rest */
+return
+short
+static
+super
+switch
+synchronized
+this
+throw
+throws
+transient
+true
+try
+/* var */
+void
+volatile
+while
--- a/elemx/java/j_scan.y
+++ b/elemx/java/j_scan.y
@@ -0,0 +1,135 @@
+%start token
+%scanner
+
+%local {
+#include "elx.h"
+
+/* from elx-java.c */
+void yyserror(const char *msg);
+int yyslex(void);
+
+/* for the TK_ symbols, generated from java.y */
+#include "java.h"
+
+/* for keyword recognition */
+#include "j_keywords.h"
+
+extern void issue_token(char which);
+extern void mark_token_start(void);
+
+#define MAX_IDENT 200
+static int idlen;
+static char identifier[MAX_IDENT+1];
+#define INIT_IDENT(c) (identifier[0] = (c), idlen = 1)
+#define ADD_IDENT(c) if (idlen == MAX_IDENT) return E_IDENT_TOO_LONG; \
+                     else identifier[idlen++] = (c)
+
+/* ### is there a better place? */
+#define E_IDENT_TOO_LONG  (-100)
+
+static int lookup(void);
+}
+
+
+%%
+
+token : pure_ws* { mark_token_start(); } slash_op
+
+slash_op : "/=" { return TK_OPERATOR; }
+	 | comment token
+	 | '/' { return TK_OPERATOR; }
+	 | one_token
+	 |
+	 ;
+
+one_token : t_identifier { return lookup(); }
+	  | t_literal { return TK_LITERAL; }
+	  | t_operator { return TK_OPERATOR; }
+	  | t_chars { return yysprev_char; }
+	  | t_inc_dec { return TK_INC_DEC; }
+	  | t_bracket
+          ;
+
+t_identifier : alpha { INIT_IDENT(yysprev_char); }
+	       ( alphanum { ADD_IDENT(yysprev_char); } )*
+
+alpha : 'a' - 'z' | 'A' - 'Z' | '_' | '$'
+alphanum : alpha | digit
+
+digit : '0' - '9'
+hexdigit : digit | 'a' - 'f' | 'A' - 'F'
+octal : '0' - '7'
+
+t_literal : number | string | char_constant
+
+number : ('1' - '9') digit* decimal_suffix
+       | '.' digit+ [exponent] [float_suffix]
+       | '0' (('x' | 'X') hexdigit+ | octal+) decimal_suffix
+       ;
+decimal_suffix : ('.' digit* [exponent] [float_suffix])
+	       | 'l' | 'L'
+	       | /* nothing */
+	       ;
+exponent : ('e' | 'E') ['+' | '-'] digit+
+float_suffix : 'f' | 'F' | 'd' | 'D'
+
+string : '"' string_char* '"' { issue_token(ELX_STRING); }
+string_char : '\1' -> '"' | '"' <-> '\\' | '\\' <- '\377' | '\\' '\1' - '\377'
+
+char_constant : '\'' one_char '\''
+one_char : '\1' -> '\'' | '\'' <-> '\\' | '\\' <- '\377' | '\\' '\1' - '\377'
+
+comment : ( "//" line_comment_char* '\n'
+	  | "/*" (block_comment_char | '*' block_non_term_char)* "*/"
+	  ) { issue_token(ELX_COMMENT); }
+	;
+line_comment_char : '\1' -> '\n' | '\n' <- '\377'
+block_comment_char : '\1' -> '*' | '*' <- '\377'
+block_non_term_char : '\1' -> '/' | '/' <- '\377'
+
+t_operator : "<<" | ">>" | ">>>"
+           | ">=" | "<=" | "==" | "!=" | "&&" | "||"
+	   | "*=" | "%=" | "+=" | "-=" | "<<=" | ">>="
+	   | ">>>=" | "&=" | "^=" | "|="
+	   | '<' | '>' | '%' | '^' | '&' | '|'
+	   ;
+t_inc_dec : "++" | "--"
+
+/* note: could not use ws* ; the '[' form would only reduce on $end
+   rather than "any" character. that meant we could not recognize '['
+   within the program text. separating out the cases Does The Right
+   Thing */
+t_bracket : '[' { return '['; }
+	  | '[' ']' { return TK_DIM; }
+	  | '[' ws+ ']' { return TK_DIM; }
+	  ;
+
+t_chars : ',' | ';' | '.' | '{' | '}' | '=' | '(' | ')' | ':'
+        | ']' | '!' | '~' | '+' | '-' | '*' | '?'
+	;
+
+ws : pure_ws | comment
+
+pure_ws : ' ' | '\t' | '\n' | '\f'
+
+%%
+
+static int lookup(void)
+{
+    int kw = KR_find_keyword(identifier, idlen);
+
+    if (kw == KR__not_found)
+    {
+        /* terminate so user can grab an identifier string */
+        identifier[idlen] = '\0';
+        return TK_IDENTIFIER;
+    }
+    
+    issue_token(ELX_KEYWORD);
+    return kw;
+}
+
+const char *get_identifier(void)
+{
+    return identifier;
+}
--- a/elemx/java/java.y
+++ b/elemx/java/java.y
@@ -0,0 +1,458 @@
+%token KR_abstract
+%token KR_boolean KR_break KR_byte /* KR_byvalue */
+%token KR_case /* KR_cast */ KR_catch KR_char KR_class /* KR_const */ KR_continue
+%token KR_default KR_do KR_double
+%token KR_else KR_extends
+%token KR_false KR_final KR_finally KR_float KR_for /* KR_future */
+/* %token KR_generic KR_goto */
+%token KR_if KR_implements KR_import /* KR_inner */ KR_instanceof KR_int KR_interface
+%token KR_long
+%token KR_native KR_new KR_null
+/* %token KR_operator KR_outer */
+%token KR_package KR_private KR_protected KR_public
+%token /* KR_rest */ KR_return
+%token KR_short KR_static KR_super KR_switch KR_synchronized
+%token KR_this KR_throw KR_throws KR_transient KR_true KR_try
+%token /* KR_var */ KR_void KR_volatile
+%token KR_while
+
+%token TK_OP_ASSIGN TK_OPERATOR TK_IDENTIFIER TK_LITERAL
+%token TK_DIM TK_INC_DEC
+
+%start CompilationUnit
+
+/* the standard if/then/else conflict */
+/* %expect 1 */
+
+%{
+#include "elx.h"
+
+void yyerror(const char *msg);
+int yylex(void);
+
+/* ### should come from an elx-python.h or something */
+void issue_token(char which);
+%}
+
+%export {
+/* the main parsing function */
+int yyparse(void);
+
+/* need to define the 'not found' in addition to the regular keywords */
+#define KR__not_found 0
+}
+
+%%
+
+TypeSpecifier
+	: TypeName
+	| TypeNameDims
+	;
+
+TypeNameDims
+	: TypeName TK_DIM+
+	;
+
+TypeNameDot
+	: NamePeriod
+	| PrimitiveType '.'
+	;
+
+TypeName
+	: PrimitiveType
+	| NamePeriod TK_IDENTIFIER
+	| TK_IDENTIFIER
+	;
+
+NamePeriod
+	: TK_IDENTIFIER '.'
+	| NamePeriod TK_IDENTIFIER '.'
+	;
+
+TypeNameList
+	: TypeName / ','
+	;
+
+PrimitiveType
+	: KR_boolean
+	| KR_byte
+	| KR_char
+	| KR_double
+	| KR_float
+	| KR_int
+	| KR_long
+	| KR_short
+	| KR_void
+	;
+
+CompilationUnit
+	: PackageStatement [ImportStatements] [TypeDeclarations]
+	|                   ImportStatements  [TypeDeclarations]
+	|                                      TypeDeclarations
+	;
+
+PackageStatement
+	: KR_package (TK_IDENTIFIER / '.') ';'
+	;
+
+TypeDeclarations
+	: TypeDeclaration+
+	;
+
+TypeDeclaration
+	: ClassDeclaration
+	| InterfaceDeclaration
+	;
+
+ImportStatements
+	: ImportStatement+
+	;
+
+ImportStatement
+	: KR_import TK_IDENTIFIER ('.' TK_IDENTIFIER)* [".*"] ';'
+	;
+/*
+QualifiedName
+	: TK_IDENTIFIER / '.'
+	;
+*/
+ClassDeclaration
+        : [Modifiers] KR_class TK_IDENTIFIER [Super] [Interfaces] ClassBody
+	;
+
+Modifiers
+	: Modifier+
+	;
+
+Modifier
+	: KR_abstract
+	| KR_final
+	| KR_public
+	| KR_protected
+	| KR_private
+	| KR_static
+	| KR_transient
+	| KR_volatile
+	| KR_native
+	| KR_synchronized
+	;
+
+Super
+	: KR_extends TypeNameList
+	;
+
+Interfaces
+	: KR_implements TypeNameList
+	;
+
+ClassBody
+	: '{' FieldDeclaration* '}'
+	;
+
+FieldDeclaration
+	: FieldVariableDeclaration
+	| MethodDeclaration
+	| ConstructorDeclaration
+	| StaticInitializer
+	;
+
+FieldVariableDeclaration
+	: [Modifiers] TypeSpecifier VariableDeclarators ';'
+	;
+
+VariableDeclarators
+	: VariableDeclarator / ','
+	;
+
+VariableDeclarator
+	: DeclaratorName ['=' VariableInitializer]
+	;
+
+VariableInitializer
+	: Expression
+        | '{' [ArrayInitializers] '}'
+        ;
+
+ArrayInitializers
+	: VariableInitializer ( ',' [VariableInitializer] )*
+	;
+
+MethodDeclaration
+	: [Modifiers] TypeSpecifier MethodDeclarator [Throws] MethodBody
+	;
+
+MethodDeclarator
+	: DeclaratorName '(' [ParameterList] ')' TK_DIM*
+	;
+
+ParameterList
+	: Parameter / ','
+	;
+
+Parameter
+	: TypeSpecifier DeclaratorName
+	;
+
+DeclaratorName
+	: TK_IDENTIFIER TK_DIM*
+        ;
+
+Throws
+	: KR_throws TypeNameList
+	;
+
+MethodBody
+	: Block
+	| ';'
+	;
+
+ConstructorDeclaration
+	: [Modifiers] ConstructorDeclarator [Throws] Block
+	;
+
+ConstructorDeclarator
+	: TypeName '(' [ParameterList] ')'
+	;
+
+StaticInitializer
+	: KR_static Block
+	;
+
+InterfaceDeclaration
+	: [Modifiers] KR_interface TK_IDENTIFIER [ExtendsInterfaces] InterfaceBody
+	;
+
+ExtendsInterfaces
+	: KR_extends TypeNameList
+        ;
+
+InterfaceBody
+	: '{' FieldDeclaration+ '}'
+	;
+
+Block
+	: '{' LocalVariableDeclarationOrStatement* '}'
+        ;
+
+LocalVariableDeclarationOrStatement
+	: LocalVariableDeclarationStatement
+	| Statement
+	;
+
+LocalVariableDeclarationStatement
+	: TypeSpecifier VariableDeclarators ';'
+	;
+
+Statement
+	: EmptyStatement
+	| LabeledStatement
+	| ExpressionStatement ';'
+        | SelectionStatement
+        | IterationStatement
+	| JumpStatement
+	| GuardingStatement
+	| Block
+	;
+
+EmptyStatement
+	: ';'
+        ;
+
+LabeledStatement
+	: TK_IDENTIFIER ':' LocalVariableDeclarationOrStatement
+        | KR_case ConstantExpression ':' LocalVariableDeclarationOrStatement
+	| KR_default ':' LocalVariableDeclarationOrStatement
+        ;
+
+ExpressionStatement
+	: Expression
+	;
+
+SelectionStatement
+	: KR_if '(' Expression ')' Statement [KR_else Statement]
+        | KR_switch '(' Expression ')' Block
+        ;
+
+IterationStatement
+	: KR_while '(' Expression ')' Statement
+	| KR_do Statement KR_while '(' Expression ')' ';'
+	| KR_for '(' ForInit ForExpr [ForIncr] ')' Statement
+	;
+
+ForInit
+	: ExpressionStatements ';'
+	| LocalVariableDeclarationStatement
+	| ';'
+	;
+
+ForExpr
+	: [Expression] ';'
+	;
+
+ForIncr
+	: ExpressionStatements
+	;
+
+ExpressionStatements
+	: ExpressionStatement / ','
+	;
+
+JumpStatement
+	: KR_break [TK_IDENTIFIER] ';'
+        | KR_continue [TK_IDENTIFIER] ';'
+	| KR_return [Expression] ';'
+	| KR_throw Expression ';'
+	;
+
+GuardingStatement
+	: KR_synchronized '(' Expression ')' Statement
+	| KR_try Block Finally
+	| KR_try Block Catches
+	| KR_try Block Catches Finally
+	;
+
+Catches
+	: Catch+
+	;
+
+Catch
+	: KR_catch '(' TypeSpecifier [TK_IDENTIFIER] ')' Block
+	;
+
+Finally
+	: KR_finally Block
+	;
+
+ArgumentList
+	: Expression / ','
+	;
+
+PrimaryExpression
+	: TK_LITERAL
+	| KR_true | KR_false
+	| KR_this
+	| KR_null
+	| KR_super
+	| '(' Expression ')'
+	;
+
+PostfixExpression
+	: PrimaryExpression Trailers
+
+	| TypeName AltTrailers
+	| TypeNameDot FollowsPeriod Trailers
+	| TypeNameDot DimAllocation TypeTrailers
+	| TypeNameDims TypeTrailers
+
+	| KR_new TypeName AltTrailers
+	| KR_new TypeNameDims TypeTrailers
+        ;
+
+DimAllocation
+	: KR_new TypeNameDims
+	;
+
+PostfixDims
+	: TK_DIM+ '.' KR_class
+	;
+
+FollowsPeriod
+	: KR_this
+	| KR_class
+	| KR_super
+	| KR_new TypeName NoPeriodsTrailer
+	;
+
+NoPeriodsTrailer
+	: '[' Expression ']'
+	| '(' [ArgumentList] ')'
+	;
+
+NoDimTrailer
+	: NoPeriodsTrailer
+	| '.' (FollowsPeriod | TK_IDENTIFIER)
+	;
+
+AnyTrailer
+	: NoDimTrailer
+	| PostfixDims
+	;
+
+Trailers
+	: (AnyTrailer | DimAllocation NoDimTrailer)* [DimAllocation]
+	;
+
+AltTrailers
+	: NoPeriodsTrailer Trailers
+	|
+	;
+
+TypeTrailers
+	: NoDimTrailer Trailers
+	|
+	;
+
+CastablePrefixExpression
+	: PostfixExpression [TK_INC_DEC]
+	| LogicalUnaryOperator CastExpression
+	;
+
+LogicalUnaryOperator
+	: '~'
+	| '!'
+	;
+
+UnaryOperator
+	: '+'
+	| '-'
+	| TK_INC_DEC
+	;
+
+/* note: we don't actually have grammar for a cast. we just rely on:
+   (expr) (argument)
+   as our parse match */
+CastExpression
+	: /* '(' PrimitiveType ')' CastExpression
+	| '(' NamePeriod TK_IDENTIFIER TK_DIM TK_DIM* ')' CastablePrefixExpression
+	| '(' NamePeriod TK_IDENTIFIER ')' CastablePrefixExpression
+	| '(' TK_IDENTIFIER TK_DIM TK_DIM* ')' CastablePrefixExpression
+	| '(' TK_IDENTIFIER ')' CastablePrefixExpression
+	| */ UnaryOperator CastExpression
+	| CastablePrefixExpression
+	;
+
+BinaryExpression
+	: CastExpression
+	| BinaryExpression TK_BINARY CastExpression
+	| BinaryExpression KR_instanceof TypeSpecifier
+	;
+
+ConditionalExpression
+	: BinaryExpression
+	| BinaryExpression '?' Expression ':' ConditionalExpression
+	;
+
+AssignmentExpression
+	: ConditionalExpression [AssignmentOperator AssignmentExpression]
+	;
+
+AssignmentOperator
+	: '='
+	| TK_OP_ASSIGN
+	;
+
+Expression
+	: AssignmentExpression
+        ;
+
+ConstantExpression
+	: ConditionalExpression
+	;
+
+/*
+TK_OPERATOR : OP_LOR | OP_LAND
+	    | OP_EQ | OP_NE | OP_LE | OP_GE
+	    | OP_SHL | OP_SHR | OP_SHRR
+	    ;
+*/
+TK_BINARY : TK_OPERATOR | '+' | '-' | '*'
--- a/elemx/python/elx-python.c
+++ b/elemx/python/elx-python.c
@@ -0,0 +1,150 @@
+#include <stdio.h>
+#include <stdlib.h>
+
+#include "scanner.h"
+#include "python.h"
+#include "py_keywords.h"
+#include "elx.h"
+
+extern int yylex(void);
+
+static const char *fname;
+static int saw_error = 0;
+static void *scan_ctx;
+static elx_context_t *ectx;
+
+void yyerror(const char *msg)
+{
+    int sl, sc, el, ec;
+
+    scanner_token_linecol(scan_ctx, &sl, &sc, &el, &ec);
+    fprintf(stderr, "%s:%d:%d: parse error: %s\n", fname, sl, sc, msg);
+    saw_error = 1;
+}
+
+int reader(void *user_ctx)
+{
+    FILE *inf = user_ctx;
+    int c = fgetc(inf);
+
+    if (c == EOF)
+        return SCANNER_EOF;
+
+//    printf("char: '%c'\n", c);
+
+    return c;
+}
+
+void issue_token(char which)
+{
+    int start;
+    int end;
+    const char *ident = NULL;
+
+    scanner_token_range(scan_ctx, &start, &end);
+
+    if (ELX_DEFINES_SYM(which))
+    {
+        int length;
+
+        scanner_identifier(scan_ctx, &ident, &length);
+    }
+
+    elx_issue_token(ectx, which, start, end - start + 1, ident);
+}
+
+int yylex(void)
+{
+    int v;
+
+    do {
+        v = scanner_get_token(scan_ctx);
+
+        if (v == TK_COMMENT)
+        {
+            issue_token(ELX_COMMENT);
+        }
+    } while (v == TK_COMMENT);
+
+    /* is this identifier a keyword? */
+    if (v == TK_IDENTIFIER)
+    {
+        const char *ident;
+        int length;
+        int kw;
+
+        scanner_identifier(scan_ctx, &ident, &length);
+#if 0
+        printf("id=%s\n", ident);
+#endif
+
+        kw = KR_find_keyword(ident, length);
+        if (kw != KR__not_found)
+        {
+            v = kw;
+            issue_token(ELX_KEYWORD);
+        }
+    }
+    else if (v == TK_STRING)
+    {
+        issue_token(ELX_STRING);
+    }
+
+//    printf("token=%d\n", v);
+
+    return v;
+}
+
+#ifdef DEBUG_SCANNER
+
+void gen_scan_tokens(void)
+{
+    while (1)
+    {
+        int v = scanner_get_token(scan_ctx);
+        int sl, sc, el, ec;
+
+        scanner_token_linecol(scan_ctx, &sl, &sc, &el, &ec);
+        if (v == TK_NEWLINE)
+            printf("%d,%d: NEWLINE\n", sl, sc);
+        else if (v == TK_INDENT)
+            printf("%d,%d: INDENT\n", el, ec);
+        else if (v == TK_DEDENT)
+            printf("%d,%d: DEDENT\n", el, ec);
+        else
+            printf("%d,%d-%d,%d: %d\n", sl, sc, el, ec, v);
+
+        /* end of parse? */
+        if (v <= 0)
+            break;
+    }
+}
+
+#endif /* DEBUG_SCANNER */
+
+static void gen_elx_tokens(void)
+{
+    /* ### what to do with the result? should have seen/set saw_error */
+    (void) yyparse();
+}
+
+int main(int argc, const char **argv)
+{
+    ectx = elx_process_args(argc, argv);
+    elx_open_files(ectx);
+
+    scan_ctx = scanner_begin(reader, ectx->input_fp);
+
+#ifdef DEBUG_SCANNER
+    gen_scan_tokens();
+#else
+    gen_elx_tokens();
+#endif
+
+    scanner_end(scan_ctx);
+    elx_close_files(ectx);
+
+    if (saw_error)
+        return EXIT_FAILURE;
+    return EXIT_SUCCESS;
+}
--- a/elemx/python/py_keywords.shilka
+++ b/elemx/python/py_keywords.shilka
@@ -0,0 +1,36 @@
+%local {
+/* get the KR_* values */
+#include "python.h"
+}
+
+%%
+and
+/* as */
+assert
+break
+class
+continue
+def
+del
+elif
+else
+except
+exec
+finally
+for
+from
+global
+if
+import
+in
+is
+lambda
+not
+or
+pass
+print
+raise
+return
+try
+while
+yield
--- a/elemx/python/python.y
+++ b/elemx/python/python.y
@@ -0,0 +1,135 @@
+
+%token TK_COMMENT TK_IDENTIFIER TK_NUMBER
+%token TK_OPERATOR TK_STRING
+%token TK_INDENT TK_DEDENT TK_NEWLINE
+
+%token KR_and KR_assert KR_break KR_class KR_continue KR_def
+%token KR_del KR_elif KR_else KR_except KR_exec KR_finally
+%token KR_for KR_from KR_global KR_if KR_import KR_in KR_is
+%token KR_lambda KR_not KR_or KR_pass KR_print KR_raise
+%token KR_return KR_try KR_while KR_yield
+
+%start file_input
+
+%{
+#include "elx.h"
+
+void yyerror(const char *msg);
+int yylex(void);
+
+/* ### should come from an elx-python.h or something */
+void issue_token(char which);
+%}
+
+%export {
+/* the main parsing function */
+int yyparse(void);
+
+/* need to define the 'not found' in addition to the regular keywords */
+#define KR__not_found 0
+}
+
+%%
+
+file_input: (TK_NEWLINE | stmt)*
+
+NAME: TK_IDENTIFIER
+
+funcdef: KR_def NAME { issue_token(ELX_LOCAL_FDEF); } parameters ':' suite
+parameters: '(' [varargslist] ')'
+varargslist: paramdef (',' paramdef)* [',' [varargsdef]]
+           | varargsdef
+	   ;
+/* the TK_OPERATOR represents '*' or '**' */
+varargsdef: TK_OPERATOR NAME [',' TK_OPERATOR NAME]
+paramdef: fpdef [TK_OPERATOR test]
+fpdef: NAME | '(' fplist ')'
+fplist: fpdef (',' fpdef)* [',']
+
+stmt: simple_stmt | compound_stmt
+simple_stmt: small_stmt (';' small_stmt)* [';'] TK_NEWLINE
+small_stmt: expr_stmt | print_stmt | raise_stmt
+	  | import_stmt | global_stmt | exec_stmt | assert_stmt
+          | KR_del exprlist
+	  | KR_pass
+	  | KR_break
+	  | KR_continue
+	  | KR_return [testlist]
+	  | KR_yield testlist
+	  ;
+
+/* expr_stmt is normally assignment, which we get thru TK_OPERATOR in 'expr' */
+expr_stmt: testlist
+
+/* a print normally allows '>> test'; since that is a TK_OPERATOR, we
+   get it as part of 'factor'. this rule also allows for a trailing
+   comma in '>> test,' which the normal print doesn't */
+print_stmt: KR_print [test (',' test)* [',']]
+
+raise_stmt: KR_raise [test [',' test [',' test]]]
+
+/* the TK_OPERATOR represents '*' */
+import_stmt: KR_import dotted_as_name (',' dotted_as_name)*
+           | KR_from dotted_name KR_import (TK_OPERATOR | import_as_name (',' import_as_name)*)
+import_as_name: NAME [NAME NAME]
+dotted_as_name: dotted_name [NAME NAME]
+dotted_name: NAME ('.' NAME)*
+global_stmt: KR_global NAME (',' NAME)*
+exec_stmt: KR_exec expr [KR_in test [',' test]]
+assert_stmt: KR_assert test [',' test]
+
+compound_stmt: if_stmt | while_stmt | for_stmt | try_stmt | funcdef | classdef
+if_stmt: KR_if test ':' suite (KR_elif test ':' suite)* [KR_else ':' suite]
+while_stmt: KR_while test ':' suite [KR_else ':' suite]
+for_stmt: KR_for exprlist KR_in testlist ':' suite [KR_else ':' suite]
+try_stmt: KR_try ':' suite (except_clause ':' suite)+
+           [KR_else ':' suite] | KR_try ':' suite KR_finally ':' suite
+/* NB compile.c makes sure that the default except clause is last */
+except_clause: KR_except [test [',' test]]
+suite: simple_stmt | TK_NEWLINE TK_INDENT stmt+ TK_DEDENT
+
+test: test_factor (test_op test_factor | KR_is [KR_not] factor)*
+      [TK_OPERATOR lambdef] | lambdef
+test_op: bin_op | KR_in
+test_factor: KR_not* factor
+
+expr: factor (expr_op factor)*
+expr_op: bin_op | KR_is [KR_not]
+
+factor: TK_OPERATOR* atom trailer*
+
+bin_op: TK_OPERATOR | KR_or | KR_and | KR_not KR_in
+
+atom: '(' [testlist] ')' | '[' [listmaker] ']' | '{' [dictmaker] '}'
+    | '`' testlist_no_trailing '`' | TK_IDENTIFIER | TK_NUMBER | TK_STRING+
+listmaker: test ( list_for | (',' test)* [','] )
+lambdef: KR_lambda [varargslist] ':' test
+trailer: '(' [arglist] ')' | '[' subscriptlist ']' | '.' NAME
+subscriptlist: subscript (',' subscript)* [',']
+subscript: '.' '.' '.' | test | [test] ':' [test] [sliceop]
+sliceop: ':' [test]
+exprlist: expr (',' expr)* [',']
+testlist: test (',' test)* [',']
+testlist_no_trailing: test (',' test)*
+testlist_safe: test [(',' test)+ [',']] /* doesn't match: test, */
+dictmaker: test ':' test (',' test ':' test)* [',']
+
+classdef: KR_class NAME ['(' testlist ')'] ':' suite
+
+/* arguments are normally 'keyword = test; since '=' is TK_OPERATOR, we
+   match keyword arguments as part of 'test' (in 'expr').
+   
+   the vararg portion is normally '* test' or '** test'; since '*' and
+   '**' are TK_OPERATOR, we match varargs as part of 'test' (in
+   'factor')
+   
+   thus, all argument forms are simply 'test'
+   
+   varargs does not normally allow a trailing comma, but we can
+   simplify things and allow a match
+*/
+arglist: test (',' test)* [',']
+
+list_iter: list_for | list_if
+list_for: KR_for exprlist KR_in testlist_safe [list_iter]
+list_if: KR_if test [list_iter]
--- a/elemx/python/scanner.c
+++ b/elemx/python/scanner.c
@@ -0,0 +1,523 @@
+
+#include <stdlib.h>
+#include <ctype.h>
+#include <assert.h>
+#include <string.h>
+
+#include "python.h"     /* get the TK_ values */
+#include "scanner.h"
+
+#define SCANNER_EMPTY (SCANNER_EOF - 1) /* -2 */
+#define SCANNER_TABSIZE 8
+#define SCANNER_MAXINDENT 100
+#define SCANNER_MAXIDLEN  200
+
+typedef struct
+{
+    get_char_t getfunc;
+    void *user_ctx;
+
+    char saved;
+    int was_newline;    /* was previous character a newline? */
+
+    int start;          /* start position of last token returned */
+    int start_col;
+    int start_line;
+
+    int fpos;           /* file position */
+    int lineno;         /* file line number */
+    int line_pos;       /* file position of current line's first char */
+
+    int nesting_level;
+
+    int indent;                         /* which indent */
+    int indents[SCANNER_MAXINDENT];     /* the set of indents */
+
+    int dedent_count;                   /* how many DEDENTs to issue */
+
+    int skip_newline;   /* skip the newline after a blank_line + comment */
+
+    int idlen;
+    char identifier[SCANNER_MAXIDLEN];  /* accumulated identifier */
+
+} scanner_ctx;
+
+
+static int next_char(scanner_ctx *ctx)
+{
+    int c;
+
+    ++ctx->fpos;
+
+    if (ctx->saved == SCANNER_EMPTY)
+    {
+        return (*ctx->getfunc)(ctx->user_ctx);
+    }
+
+    c = ctx->saved;
+    ctx->saved = SCANNER_EMPTY;
+    return c;
+}
+
+static void backup_char(scanner_ctx *ctx, int c)
+{
+    assert(ctx->saved == SCANNER_EMPTY);
+    ctx->saved = c;
+    ctx->was_newline = 0;       /* we may have put it back */
+    --ctx->fpos;
+}
+
+/* called to note that we've moved on to another line */
+static void on_next_line(scanner_ctx *ctx)
+{
+    ctx->line_pos = ctx->fpos;
+    ++ctx->lineno;
+}
+
+void *scanner_begin(get_char_t getfunc, void *user_ctx)
+{
+    scanner_ctx *ctx = malloc(sizeof(*ctx));
+
+    memset(ctx, 0, sizeof(*ctx));
+    ctx->getfunc = getfunc;
+    ctx->user_ctx = user_ctx;
+    ctx->saved = SCANNER_EMPTY;
+    ctx->lineno = 1;
+
+    return ctx;
+}
+
+int scanner_get_token(void *opaque_ctx)
+{
+    scanner_ctx *ctx = opaque_ctx;
+    int c;
+    int c2;
+    int blank_line;
+
+    if (ctx->dedent_count)
+    {
+        --ctx->dedent_count;
+        return TK_DEDENT;
+    }
+
+  nextline:
+    blank_line = 0;
+    /* if we're at the start of the line, then get the indentation level */
+    if (ctx->fpos == ctx->line_pos)
+    {
+        int col = 0;
+
+        while (1)
+        {
+            c = next_char(ctx);
+            if (c == ' ')
+                ++col;
+            else if (c == '\t')
+                col = (col / SCANNER_TABSIZE + 1) * SCANNER_TABSIZE;
+            else if (c == '\f')         /* ^L / formfeed */
+                col = 0;
+            else
+                break;
+        }
+        backup_char(ctx, c);
+
+        if (c == '#' || c == '\n')
+        {
+            /* this is a "blank" line and doesn't count towards indentation,
+               and it doesn't produce NEWLINE tokens */
+            blank_line = 1;
+        }
+
+        /* if it isn't blank, and we aren't inside nesting expressions, then
+           we need to handle INDENT/DEDENT */
+        if (!blank_line && ctx->nesting_level == 0)
+        {
+            int last_indent = ctx->indents[ctx->indent];
+
+            if (col == last_indent)
+            {
+                /* no change */
+            }
+            else if (col > last_indent)
+            {
+                if (ctx->indent == SCANNER_MAXINDENT - 1)
+                {
+                    /* oops. too deep. */
+                    return E_TOO_MANY_INDENTS;
+                }
+                ctx->indents[++ctx->indent] = col;
+                return TK_INDENT;
+            }
+            else /* col < last_indent */
+            {
+                /* find the previous indentation that matches this one */
+                while (ctx->indent > 0
+                       && col < ctx->indents[ctx->indent])
+                {
+                    ++ctx->dedent_count;
+                    --ctx->indent;
+                }
+                if (col != ctx->indents[ctx->indent])
+                {
+                    /* oops. dedent doesn't match any indent. */
+                    return E_DEDENT_MISMATCH;
+                }
+
+                /* deliver one dedent now */
+                --ctx->dedent_count;
+                return TK_DEDENT;
+            }
+        } /* !blank_line ... */
+    } /* start of line */
+
+    /* start here if we see a line continuation */
+ read_more:
+
+    do {
+        c = next_char(ctx);
+    } while (c == ' ' || c == '\t' || c == '\f');
+
+    /* here is where the token starts */
+    ctx->start = ctx->fpos;
+    ctx->start_line = ctx->lineno;
+    ctx->start_col = ctx->fpos - ctx->line_pos;
+
+    /* comment? */
+    if (c == '#')
+    {
+        do {
+            c = next_char(ctx);
+        } while (c != SCANNER_EOF && c != '\n');
+
+        /* if we are suppressing newlines because this is a blank line, then
+           leave a marker to skip the newline, next time through. */
+        if (blank_line && c == '\n')
+            ctx->skip_newline = 1;
+
+        /* put back whatever we sucked up */
+        backup_char(ctx, c);
+
+        return TK_COMMENT;
+    }
+
+    /* Look for an identifier */
+    if (isalpha(c) || c == '_')
+    {
+        ctx->idlen = 0;
+
+        /* is this actually a string? */
+        if (c == 'r' || c == 'R')
+        {
+            ctx->identifier[ctx->idlen++] = c;
+            c = next_char(ctx);
+            if (c == '"' || c == '\'')
+                goto parse_string;
+        }
+        else if (c == 'u' || c == 'U')
+        {
+            ctx->identifier[ctx->idlen++] = c;
+            c = next_char(ctx);
+            if (c == 'r' || c == 'R')
+            {
+                ctx->identifier[ctx->idlen++] = c;
+                c = next_char(ctx);
+            }
+            if (c == '"' || c == '\'')
+                goto parse_string;
+        }
+
+        while (isalnum(c) || c == '_') {
+            /* store the character if there is room for it, and room left
+               for a null-terminator. */
+            if (ctx->idlen < SCANNER_MAXIDLEN-1)
+                ctx->identifier[ctx->idlen++] = c;
+            c = next_char(ctx);
+        }
+        backup_char(ctx, c);
+
+        /* ### check for a keyword */
+        return TK_IDENTIFIER;
+    }
+
+    if (c == '\n')
+    {
+        on_next_line(ctx);
+
+        /* don't report NEWLINE tokens for blank lines or nested exprs */
+        if (blank_line || ctx->nesting_level > 0 || ctx->skip_newline)
+        {
+            ctx->skip_newline = 0;
+            goto nextline;
+        }
+
+        return TK_NEWLINE;
+    }
+
+    if (c == '.')
+    {
+        c = next_char(ctx);
+        if (isdigit(c))
+            goto parse_fraction;
+        backup_char(ctx, c);
+        return '.';
+    }
+
+    if (isdigit(c))
+    {
+        if (c == '0')
+        {
+            c = next_char(ctx);
+            if (c == 'x' || c == 'X')
+            {
+                do {
+                    c = next_char(ctx);
+                } while (isxdigit(c));
+                goto skip_fp;
+            }
+            else if (isdigit(c))
+            {
+                do {
+                    c = next_char(ctx);
+                } while (isdigit(c));
+            }
+            if (c == '.')
+                goto parse_fraction;
+            if (c == 'e' || c == 'E')
+                goto parse_exponent;
+            if (c == 'j' || c == 'J')
+                goto parse_imaginary;
+        skip_fp:
+            /* this point: parsed an octal, decimal, or hexadecimal */
+
+            if (c == 'l' || c == 'L')
+            {
+                /* we consumed just enough. stop and return a NUMBER */
+                return TK_NUMBER;
+            }
+
+            /* consumed too much. backup and return a NUMBER */
+            backup_char(ctx, c);
+            return TK_NUMBER;
+        }
+
+        /* decimal number */
+        do {
+            c = next_char(ctx);
+        } while (isdigit(c));
+
+        if (c == 'l' || c == 'L')
+        {
+            /* we consumed just enogh. stop and return a NUMBER */
+            return TK_NUMBER;
+        }
+
+        if (c == '.')
+        {
+        parse_fraction:
+            do {
+                c = next_char(ctx);
+            } while (isdigit(c));
+        }
+
+        if (c == 'e' || c == 'E')
+        {
+        parse_exponent:
+            c = next_char(ctx);
+            if (c == '+' || c == '-')
+                c = next_char(ctx);
+            if (!isdigit(c))
+            {
+                backup_char(ctx, c);
+                return E_BAD_NUMBER;
+            }
+            do {
+                c = next_char(ctx);
+            } while (isdigit(c));
+        }
+
+        if (c == 'j' || c == 'J')
+        {
+        parse_imaginary:
+            c = next_char(ctx);
+        }
+
+        /* one too far. backup and return a NUMBER */
+        backup_char(ctx, c);
+        return TK_NUMBER;
+
+    } /* isdigit */
+
+parse_string:
+    if (c == '\'' || c == '"')
+    {
+        int second_quote_pos = ctx->fpos + 1;
+        int which_quote = c;
+        int is_triple = 0;
+        int quote_count = 0;
+
+        while (1)
+        {
+            c = next_char(ctx);
+            if (c == '\n')
+            {
+                on_next_line(ctx);
+
+                if (!is_triple)
+                    return E_UNTERM_STRING;
+                quote_count = 0;
+            }
+            else if (c == SCANNER_EOF)
+            {
+                return E_UNTERM_STRING;
+            }
+            else if (c == which_quote)
+            {
+                ++quote_count;
+                if (ctx->fpos == second_quote_pos)
+                {
+                    c = next_char(ctx);
+                    if (c == which_quote)
+                    {
+                        is_triple = 1;
+                        quote_count = 0;
+                        continue;
+                    }
+                    /* we just read one past the empty string. back up. */
+                    backup_char(ctx, c);
+                }
+
+                /* this quote may have terminated the string */
+                if (!is_triple || quote_count == 3)
+                    return TK_STRING;
+            }
+            else if (c == '\\')
+            {
+                c = next_char(ctx);
+                if (c == SCANNER_EOF)
+                    return E_UNTERM_STRING;
+                if (c == '\n')
+                    on_next_line(ctx);
+                quote_count = 0;
+            }
+            else
+            {
+                quote_count = 0;
+            }
+        }
+
+        /* NOTREACHED */
+    }
+
+    /* line continuation */
+    if (c == '\\')
+    {
+        c = next_char(ctx);
+        if (c != '\n')
+            return E_BAD_CONTINUATION;
+
+        on_next_line(ctx);
+        goto read_more;
+    }
+
+    /* look for operators */
+
+    /* the nesting operators */
+    if (c == '(' || c == '[' || c == '{')
+    {
+        ++ctx->nesting_level;
+        return c;
+    }
+    if (c == ')' || c == ']' || c == '}')
+    {
+        --ctx->nesting_level;
+        return c;
+    }
+
+    /* look for up-to-3-char ops */
+    if (c == '<' || c == '>' || c == '*' || c == '/')
+    {
+        c2 = next_char(ctx);
+        if (c == c2)
+        {
+            c2 = next_char(ctx);
+            if (c2 != '=')
+            {
+                /* oops. one too far. */
+                backup_char(ctx, c2);
+            }
+            return TK_OPERATOR;
+        }
+
+        if (c == '<' && c2 == '>')
+            return TK_OPERATOR;
+
+        if (c2 != '=')
+        {
+            /* one char too far. */
+            backup_char(ctx, c2);
+        }
+        return TK_OPERATOR;
+    }
+
+    /* look for 2-char ops */
+    if (c == '=' || c == '!' || c == '+' || c == '-'
+        || c == '|' || c == '%' || c == '&' || c == '^')
+    {
+        c2 = next_char(ctx);
+        if (c2 == '=')
+            return TK_OPERATOR;
+
+        /* oops. too far. */
+        backup_char(ctx, c2);
+        return TK_OPERATOR;
+    }
+
+    /* ### should all of these return 'c' ? */
+    if (c == ':' || c == ',' || c == ';' || c == '`')
+        return c;
+
+    /* as a unary operator, this must be a TK_OPERATOR */
+    if (c == '~')
+        return TK_OPERATOR;
+
+    /* if we have an EOF, then just return it */
+    if (c == SCANNER_EOF)
+        return SCANNER_EOF;
+
+    /* unknown input */
+    return E_UNKNOWN_TOKEN;
+}
+
+void scanner_identifier(void *opaque_ctx, const char **ident, int *len)
+{
+    scanner_ctx *ctx = opaque_ctx;
+
+    ctx->identifier[ctx->idlen] = '\0';
+    *ident = ctx->identifier;
+    *len = ctx->idlen;
+}
+
+void scanner_token_range(void *opaque_ctx, int *start, int *end)
+{
+    scanner_ctx *ctx = opaque_ctx;
+
+    *start = ctx->start;
+    *end = ctx->fpos;
+}
+
+void scanner_token_linecol(void *opaque_ctx,
+                           int *sline, int *scol, int *eline, int *ecol)
+{
+    scanner_ctx *ctx = opaque_ctx;
+
+    *sline = ctx->start_line;
+    *scol = ctx->start_col;
+
+    *eline = ctx->lineno;
+    *ecol = ctx->fpos - ctx->line_pos;
+}
+
+void scanner_end(void *ctx)
+{
+    free(ctx);
+}
--- a/elemx/python/scanner.h
+++ b/elemx/python/scanner.h
@@ -0,0 +1,42 @@
+#ifndef SCANNER_H
+#define SCANNER_H
+
+
+#ifdef __cplusplus
+extern "C" {
+#endif /* __cplusplus */
+
+
+/* constants and errors returned by the scanner */
+enum
+{
+    SCANNER_EOF = -1,           /* returned by get_char_t and
+                                   scanner_get_token to symbolize EOF */
+
+    E_TOO_MANY_INDENTS = -100,  /* too many indents */
+    E_DEDENT_MISMATCH,          /* no matching indent */
+    E_BAD_CONTINUATION,         /* character occurred after \ */
+    E_BAD_NUMBER,               /* parse error in a number */
+    E_UNKNOWN_TOKEN,            /* dunno what we found */
+    E_UNTERM_STRING             /* unterminated string constant */
+};
+
+typedef int (*get_char_t)(void *user_ctx);
+
+void *scanner_begin(get_char_t getfunc, void *user_ctx);
+
+int scanner_get_token(void *ctx);
+
+void scanner_identifier(void *ctx, const char **ident, int *len);
+void scanner_token_range(void *ctx, int *start, int *end);
+void scanner_token_linecol(void *ctx,
+                           int *sline, int *scol, int *eline, int *ecol);
+
+void scanner_end(void *ctx);
+
+
+#ifdef __cplusplus
+}
+#endif /* __cplusplus */
+
+#endif /* SCANNER_H */
--- a/lib/accept.py
+++ b/lib/accept.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2009 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
--- a/lib/blame.py
+++ b/lib/blame.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 # Copyright (C) 2000 Curt Hagenlocher <curt@hagenlocher.org>
 #
 # By using this file, you agree to the terms and conditions set forth in
@@ -32,9 +32,8 @@ import os
 import re
 import time
 import math
-import cgi
 import vclib
-
+import sapi

 re_includes = re.compile('\\#(\\s*)include(\\s*)"(.*?)"')

@@ -75,14 +74,15 @@ class HTMLBlameSource:
    self.path_parts = path_parts
    self.diff_url = diff_url
    self.include_url = include_url
-    self.annotation, self.revision = self.repos.annotate(path_parts, opt_rev)
+    self.annotation, self.revision = self.repos.annotate(path_parts, opt_rev,
+                                                         True)

  def __getitem__(self, idx):
    item = self.annotation.__getitem__(idx)
    diff_url = None
    if item.prev_rev:
      diff_url = '%sr1=%s&amp;r2=%s' % (self.diff_url, item.prev_rev, item.rev)
-    thisline = link_includes(cgi.escape(item.text), self.repos,
+    thisline = link_includes(sapi.escape(item.text), self.repos,
                             self.path_parts, self.include_url)
    return _item(text=thisline, line_number=item.line_number,
                 rev=item.rev, prev_rev=item.prev_rev,
--- a/lib/compat.py
+++ b/lib/compat.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2007 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
--- a/lib/config.py
+++ b/lib/config.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2009 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -19,82 +19,155 @@ import os
 import string
 import ConfigParser
 import fnmatch
-import vclib
-import vclib.ccvs
-import vclib.svn
-import cvsdb
+

 #########################################################################
 #
 # CONFIGURATION
+# -------------
 #
 # There are three forms of configuration:
 #
-#       1) edit the viewvc.conf created by the viewvc-install(er)
-#       2) as (1), but delete all unchanged entries from viewvc.conf
-#       3) do not use viewvc.conf and just edit the defaults in this file
+#    1. edit the viewvc.conf created by the viewvc-install(er)
+#    2. as (1), but delete all unchanged entries from viewvc.conf
+#    3. do not use viewvc.conf and just edit the defaults in this file
 #
 # Most users will want to use (1), but there are slight speed advantages
 # to the other two options. Note that viewvc.conf values are a bit easier
 # to work with since it is raw text, rather than python literal values.
 #
+#
+# A WORD ABOUT OPTION LAYERING/OVERRIDES
+# --------------------------------------
+#
+# ViewVC has three "layers" of configuration options:
+#
+#    1. base configuration options - very basic configuration bits
+#       found in sections like 'general', 'options', etc.
+#    2. vhost overrides - these options overlay/override the base
+#       configuration on a per-vhost basis.
+#    3. root overrides - these options overlay/override the base
+#       configuration and vhost overrides on a per-root basis.
+#
+# Here's a diagram of the valid overlays/overrides:
+#
+#         PER-ROOT          PER-VHOST            BASE
+#       
+#                         ,-----------.     ,-----------.
+#                         | vhost-*/  |     |           |
+#                         |  general  | --> |  general  |
+#                         |           |     |           |
+#                         `-----------'     `-----------'
+#       ,-----------.     ,-----------.     ,-----------.
+#       |  root-*/  |     | vhost-*/  |     |           |
+#       |  options  | --> |  options  | --> |  options  |
+#       |           |     |           |     |           |
+#       `-----------'     `-----------'     `-----------'
+#       ,-----------.     ,-----------.     ,-----------.
+#       |  root-*/  |     | vhost-*/  |     |           |
+#       | templates | --> | templates | --> | templates |
+#       |           |     |           |     |           |
+#       `-----------'     `-----------'     `-----------'
+#       ,-----------.     ,-----------.     ,-----------.
+#       |  root-*/  |     | vhost-*/  |     |           |
+#       | utilities | --> | utilities | --> | utilities |
+#       |           |     |           |     |           |
+#       `-----------'     `-----------'     `-----------'
+#                         ,-----------.     ,-----------.
+#                         | vhost-*/  |     |           |
+#                         |   cvsdb   | --> |   cvsdb   |
+#                         |           |     |           |
+#                         `-----------'     `-----------'
+#       ,-----------.     ,-----------.     ,-----------.
+#       |  root-*/  |     | vhost-*/  |     |           |
+#       |  authz-*  | --> |  authz-*  | --> |  authz-*  |
+#       |           |     |           |     |           |
+#       `-----------'     `-----------'     `-----------'
+#                                           ,-----------.
+#                                           |           |
+#                                           |  vhosts   |
+#                                           |           |
+#                                           `-----------'
+#                                           ,-----------.
+#                                           |           |
+#                                           |   query   |
+#                                           |           |
+#                                           `-----------'
+#
+# ### TODO:  Figure out what this all means for the 'kv' stuff.
+#
 #########################################################################

 class Config:
-  _sections = ('general', 'utilities', 'options', 'cvsdb', 'templates')
-  _force_multi_value = ('cvs_roots', 'svn_roots', 'languages', 'kv_files',
-                        'root_parents', 'allowed_views', 'mime_types_files')
+  _base_sections = (
+    # Base configuration sections.
+    'authz-*',
+    'cvsdb',
+    'general',
+    'options',
+    'query',
+    'templates',
+    'utilities',
+    )
+  _force_multi_value = (
+    # Configuration values with multiple, comma-separated values.
+    'allowed_views',
+    'binary_mime_types',
+    'custom_log_formatting',
+    'cvs_roots',
+    'kv_files',
+    'languages',
+    'mime_types_files',
+    'root_parents',
+    'svn_roots',
+    )
+  _allowed_overrides = {
+    # Mapping of override types to allowed overridable sections.
+    'vhost' : ('authz-*',
+               'cvsdb',
+               'general',
+               'options',
+               'templates',
+               'utilities',
+               ),
+    'root'  : ('authz-*',
+               'options',
+               'templates',
+               'utilities',
+               )
+    }

  def __init__(self):
-    for section in self._sections:
+    self.root_options_overlayed = 0
+    for section in self._base_sections:
+      if section[-1] == '*':
+        continue
      setattr(self, section, _sub_config())

-  def load_config(self, pathname, vhost=None, rootname=None):
+  def load_config(self, pathname, vhost=None):
+    """Load the configuration file at PATHNAME, applying configuration
+    settings there as overrides to the built-in default values.  If
+    VHOST is provided, also process the configuration overrides
+    specific to that virtual host."""
+    
    self.conf_path = os.path.isfile(pathname) and pathname or None
    self.base = os.path.dirname(pathname)
    self.parser = ConfigParser.ConfigParser()
+    self.parser.optionxform = lambda x: x # don't case-normalize option names.
    self.parser.read(self.conf_path or [])
-
-    for section in self._sections:
-      if self.parser.has_section(section):
+    
+    for section in self.parser.sections():
+      if self._is_allowed_section(section, self._base_sections):
        self._process_section(self.parser, section, section)

    if vhost and self.parser.has_section('vhosts'):
      self._process_vhost(self.parser, vhost)

-    if rootname:
-      self._process_root_options(self.parser, rootname)
-    self.expand_root_parents()
-    cvsdb.setencs(self.options.encodings.split(':'))
-
-  def expand_root_parents(self):
-    """Expand the configured root parents into individual roots."""
-
-    # Each item in root_parents is a "directory : repo_type" string.
-    for pp in self.general.root_parents:
-      pos = string.rfind(pp, ':')
-      if pos < 0:
-        raise debug.ViewVCException(
-          "The path '%s' in 'root_parents' does not include a "
-          "repository type." % (pp))
-
-      repo_type = string.strip(pp[pos+1:])
-      pp = os.path.normpath(string.strip(pp[:pos]))
-
-      if repo_type == 'cvs':
-        roots = vclib.ccvs.expand_root_parent(pp)
-        if self.options.hide_cvsroot and roots.has_key('CVSROOT'):
-          del roots['CVSROOT']
-        self.general.cvs_roots.update(roots)
-      elif repo_type == 'svn':
-        roots = vclib.svn.expand_root_parent(pp)
-        self.general.svn_roots.update(roots)
-      else:
-        raise debug.ViewVCException(
-          "The path '%s' in 'root_parents' has an unrecognized "
-          "repository type." % (pp))
-
  def load_kv_files(self, language):
+    """Process the key/value (kv) files specified in the
+    configuration, merging their values into the configuration as
+    dotted heirarchical items."""
+    
    kv = _sub_config()

    for fname in self.general.kv_files:
@@ -107,6 +180,7 @@ class Config:
      fname = string.replace(fname, '%lang%', language)

      parser = ConfigParser.ConfigParser()
+      parser.optionxform = lambda x: x # don't case-normalize option names.
      parser.read(os.path.join(self.base, fname))
      for section in parser.sections():
        for option in parser.options(section):
@@ -124,10 +198,12 @@ class Config:
    return kv

  def path(self, path):
-    """Return path relative to the config file directory"""
+    """Return PATH relative to the config file directory."""
    return os.path.join(self.base, path)

  def _process_section(self, parser, section, subcfg_name):
+    if not hasattr(self, subcfg_name):
+      setattr(self, subcfg_name, _sub_config())
    sc = getattr(self, subcfg_name)

    for opt in parser.options(section):
@@ -140,25 +216,54 @@ class Config:
        except ValueError:
          pass

+      ### FIXME: This feels like unnecessary depth of knowledge for a
+      ### semi-generic configuration object.
      if opt == 'cvs_roots' or opt == 'svn_roots':
        value = _parse_roots(opt, value)

      setattr(sc, opt, value)

+  def _is_allowed_section(self, section, allowed_sections):
+    """Return 1 iff SECTION is an allowed section, defined as being
+    explicitly present in the ALLOWED_SECTIONS list or present in the
+    form 'someprefix-*' in that list."""
+    
+    for allowed_section in allowed_sections:
+      if allowed_section[-1] == '*':
+        if _startswith(section, allowed_section[:-1]):
+          return 1
+      elif allowed_section == section:
+        return 1
+    return 0
+
+  def _is_allowed_override(self, sectype, secspec, section):
+    """Test if SECTION is an allowed override section for sections of
+    type SECTYPE ('vhosts' or 'root', currently) and type-specifier
+    SECSPEC (a rootname or vhostname, currently).  If it is, return
+    the overridden base section name.  If it's not an override section
+    at all, return None.  And if it's an override section but not an
+    allowed one, raise IllegalOverrideSection."""
+
+    cv = '%s-%s/' % (sectype, secspec)
+    lcv = len(cv)
+    if section[:lcv] != cv:
+      return None
+    base_section = section[lcv:]
+    if self._is_allowed_section(base_section,
+                                self._allowed_overrides[sectype]):
+      return base_section
+    raise IllegalOverrideSection(sectype, section)
+
  def _process_vhost(self, parser, vhost):
-    # find a vhost name for this vhost, if any (if not, we've nothing to do)
+    # Find a vhost name for this VHOST, if any (else, we've nothing to do).
    canon_vhost = self._find_canon_vhost(parser, vhost)
    if not canon_vhost:
      return

-    # overlay any option sections associated with this vhost name
-    cv = 'vhost-%s/' % (canon_vhost)
-    lcv = len(cv)
+    # Overlay any option sections associated with this vhost name.
    for section in parser.sections():
-      if section[:lcv] == cv:
-        base_section = section[lcv:]
-        if base_section not in self._sections:
-          raise IllegalOverrideSection('vhost', section)
+      base_section = self._is_allowed_override('vhost', canon_vhost, section)
+      if base_section:
        self._process_section(parser, section, base_section)

  def _find_canon_vhost(self, parser, vhost):
@@ -173,26 +278,29 @@ class Config:

    return None

-  def _process_root_options(self, parser, rootname):
-    rn = 'root-%s/' % (rootname)
-    lrn = len(rn)
-    for section in parser.sections():
-      if section[:lrn] == rn:
-        base_section = section[lrn:]
-        if base_section in self._sections:
-          if base_section == 'general':
-            raise IllegalOverrideSection('root', section)
-          self._process_section(parser, section, base_section)
-        elif _startswith(base_section, 'authz-'):
-          pass
-        else:
-          raise IllegalOverrideSection('root', section)
-          
  def overlay_root_options(self, rootname):
-    "Overly per-root options atop the existing option set."
+    """Overlay per-root options for ROOTNAME atop the existing option
+    set.  This is a destructive change to the configuration."""
+
+    did_overlay = 0
+    
    if not self.conf_path:
      return
-    self._process_root_options(self.parser, rootname)
+
+    for section in self.parser.sections():
+      base_section = self._is_allowed_override('root', rootname, section)
+      if base_section:
+        # We can currently only deal with root overlays happening
+        # once, so check that we've not yet done any overlaying of
+        # per-root options.
+        assert(self.root_options_overlayed == 0)
+        self._process_section(self.parser, section, base_section)
+        did_overlay = 1
+
+    # If we actually did any overlaying, remember this fact so we
+    # don't do it again later.
+    if did_overlay:
+      self.root_options_overlayed = 1

  def _get_parser_items(self, parser, section):
    """Basically implement ConfigParser.items() for pre-Python-2.3 versions."""
@@ -203,24 +311,67 @@ class Config:
      for option in parser.options(section):
        d[option] = parser.get(section, option)
      return d.items()
-    
-  def get_authorizer_params(self, authorizer, rootname=None):
-    if not self.conf_path:
-      return {}

+  def get_authorizer_and_params_hack(self, rootname):
+    """Return a 2-tuple containing the name and parameters of the
+    authorizer configured for use with ROOTNAME.
+
+    ### FIXME: This whole thing is a hack caused by our not being able
+    ### to non-destructively overlay root options when trying to do
+    ### something like a root listing (which might need to get
+    ### different authorizer bits for each and every root in the list).
+    ### Until we have a good way to do that, we expose this function,
+    ### which assumes that base and per-vhost configuration has been
+    ### absorbed into this object and that per-root options have *not*
+    ### been overlayed.  See issue #371."""
+
+    # We assume that per-root options have *not* been overlayed.
+    assert(self.root_options_overlayed == 0)
+
+    if not self.conf_path:
+      return None, {}
+
+    # Figure out the authorizer by searching first for a per-root
+    # override, then falling back to the base/vhost configuration.
+    authorizer = None
+    root_options_section = 'root-%s/options' % (rootname)
+    if self.parser.has_section(root_options_section) \
+       and self.parser.has_option(root_options_section, 'authorizer'):
+      authorizer = self.parser.get(root_options_section, 'authorizer')
+    if not authorizer:
+      authorizer = self.options.authorizer
+
+    # No authorizer?  Get outta here.
+    if not authorizer:
+      return None, {}
+
+    # Dig up the parameters for the authorizer, starting with the
+    # base/vhost items, then overlaying any root-specific ones we find.
    params = {}
    authz_section = 'authz-%s' % (authorizer)
+    if hasattr(self, authz_section):
+      sub_config = getattr(self, authz_section)
+      for attr in dir(sub_config):
+        params[attr] = getattr(sub_config, attr)
+    root_authz_section = 'root-%s/authz-%s' % (rootname, authorizer)
    for section in self.parser.sections():
-      if section == authz_section:
+      if section == root_authz_section:
        for key, value in self._get_parser_items(self.parser, section):
          params[key] = value
-    if rootname:
-      root_authz_section = 'root-%s/authz-%s' % (rootname, authorizer)
-      for section in self.parser.sections():
-        if section == root_authz_section:
-          for key, value in self._get_parser_items(self.parser, section):
-            params[key] = value
-    params['__config'] = self
+    return authorizer, params
+
+  def get_authorizer_params(self, authorizer=None):
+    """Return a dictionary of parameter names and values which belong
+    to the configured authorizer (or AUTHORIZER, if provided)."""
+    params = {}
+    if authorizer is None:
+      authorizer = self.options.authorizer
+    if authorizer:
+      authz_section = 'authz-%s' % (self.options.authorizer)
+      if hasattr(self, authz_section):
+        sub_config = getattr(self, authz_section)
+        for attr in dir(sub_config):
+          params[attr] = getattr(sub_config, attr)
    return params
  
  def set_defaults(self):
@@ -249,11 +400,14 @@ class Config:
    self.options.allowed_views = ['annotate', 'diff', 'markup', 'roots']
    self.options.authorizer = None
    self.options.mangle_email_addresses = 0
+    self.options.custom_log_formatting = []
    self.options.default_file_view = "log"
+    self.options.binary_mime_types = []
    self.options.http_expiration_time = 600
    self.options.generate_etags = 1
    self.options.svn_ignore_mimetype = 0
    self.options.svn_config_dir = None
+    self.options.max_filesize_kbytes = 512
    self.options.use_rcsparse = 0
    self.options.sort_by = 'file'
    self.options.sort_group_dirs = 1
@@ -271,23 +425,23 @@ class Config:
    self.options.template_dir = "templates"
    self.options.docroot = None
    self.options.show_subdir_lastmod = 0
+    self.options.show_roots_lastmod = 0
    self.options.show_logs = 1
    self.options.show_log_in_markup = 1
-    self.options.cross_copies = 0
+    self.options.cross_copies = 1
    self.options.use_localtime = 0
+    self.options.iso8601_timestamps = 0
    self.options.short_log_len = 80
    self.options.enable_syntax_coloration = 1
+    self.options.tabsize = 8
    self.options.detect_encoding = 0
    self.options.use_cvsgraph = 0
    self.options.cvsgraph_conf = "cvsgraph.conf"
-    self.options.allowed_cvsgraph_useropts = []
    self.options.use_re_search = 0
    self.options.dir_pagesize = 0
    self.options.log_pagesize = 0
+    self.options.log_pagesextra = 3
    self.options.limit_changes = 100
-    self.options.cvs_ondisk_charset = 'cp1251'
-    self.options.binary_mime_re = '^(?!text/|.*\Wxml)'
-    self.options.encodings = 'utf-8:cp1251:iso-8859-1'

    self.templates.diff = None
    self.templates.directory = None
@@ -303,7 +457,6 @@ class Config:
    self.cvsdb.enabled = 0
    self.cvsdb.host = ''
    self.cvsdb.port = 3306
-    self.cvsdb.socket = ''
    self.cvsdb.database_name = ''
    self.cvsdb.user = ''
    self.cvsdb.passwd = ''
@@ -312,8 +465,9 @@ class Config:
    self.cvsdb.row_limit = 1000
    self.cvsdb.rss_row_limit = 100
    self.cvsdb.check_database_for_root = 0
-    self.cvsdb.fulltext_min_relevance = 0.2

+    self.query.viewvc_base_url = None
+    
 def _startswith(somestr, substr):
  return somestr[:len(substr)] == substr

--- a/lib/cvsdb.py
+++ b/lib/cvsdb.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2009 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -37,32 +37,14 @@ error = "cvsdb error"
 ## defined to actually be complete; it should run well off of any DBI 2.0
 ## complient database interface

-encs = [ "utf-8", "cp1251", "iso-8859-1" ]
-
-def utf8string(value):
-    for e in encs:
-        try:
-            value = value.decode(e)
-            break
-        except: pass
-    return value.encode("utf-8")
-
-def setencs(e):
-    global encs
-    encs = e
-
 class CheckinDatabase:
-    def __init__(self, host, port, socket, user, passwd, database, row_limit, min_relevance, authorizer = None):
+    def __init__(self, host, port, user, passwd, database):
        self._host = host
        self._port = port
-        self._socket = socket
        self._user = user
        self._passwd = passwd
        self._database = database
-        self._row_limit = row_limit
        self._version = None
-        self._min_relevance = min_relevance
-        self.authorizer = authorizer

        ## database lookup caches
        self._get_cache = {}
@@ -71,7 +53,7 @@ class CheckinDatabase:

    def Connect(self):
        self.db = dbi.connect(
-            self._host, self._port, self._socket, self._user, self._passwd, self._database)
+            self._host, self._port, self._user, self._passwd, self._database)
        cursor = self.db.cursor()
        cursor.execute("SET AUTOCOMMIT=1")
        table_list = self.GetTableList()
@@ -84,16 +66,14 @@ class CheckinDatabase:
        else:
            self._version = 0
        if self._version > CURRENT_SCHEMA_VERSION:
-           raise DatabaseVersionError("Database version %d is newer than the "
-                                      "last version supported by this "
-                                      "software." % (self._version))
+            raise DatabaseVersionError("Database version %d is newer than the "
+                                       "last version supported by this "
+                                       "software." % (self._version))

    def sql_get_id(self, table, column, value, auto_set):
-        value = utf8string(value)
-
        sql = "SELECT id FROM %s WHERE %s=%%s" % (table, column)
        sql_args = (value, )
-
+        
        cursor = self.db.cursor()
        cursor.execute(sql, sql_args)
        try:
@@ -103,7 +83,7 @@ class CheckinDatabase:
                return None
        else:
            return str(int(id))
-
+   
        ## insert the new identifier
        sql = "INSERT INTO %s(%s) VALUES(%%s)" % (table, column)
        sql_args = (value, )
@@ -188,6 +168,9 @@ class CheckinDatabase:

        return list

+    def GetCommitsTable(self):
+        return self._version >= 1 and 'commits' or 'checkins'
+        
    def GetTableList(self):
        sql = "SHOW TABLES"
        cursor = self.db.cursor()
@@ -254,17 +237,13 @@ class CheckinDatabase:
    def GetRepository(self, id):
        return self.get("repositories", "repository", id)

-    def GetRepositoryList(self):
-        return self.get_list("repositories", repository)
-
    def SQLGetDescriptionID(self, description, auto_set = 1):
-        description = utf8string(description)
        ## lame string hash, blame Netscape -JMP
        hash = len(description)

        sql = "SELECT id FROM descs WHERE hash=%s AND description=%s"
        sql_args = (hash, description)
-
+        
        cursor = self.db.cursor()
        cursor.execute(sql, sql_args)
        try:
@@ -278,7 +257,7 @@ class CheckinDatabase:
        sql = "INSERT INTO descs (hash,description) values (%s,%s)"
        sql_args = (hash, description)
        cursor.execute(sql, sql_args)
-
+        
        return self.GetDescriptionID(description, 0)

    def GetDescriptionID(self, description, auto_set = 1):
@@ -314,26 +293,6 @@ class CheckinDatabase:
    def GetAuthorList(self):
        return self.get_list("people", 1)

-    def GetLatestCheckinTime(self, repository):
-        repository_id = self.GetRepositoryID(repository.rootpath, 0)
-        if repository_id is None:
-            return None
-
-        commits_table = self._version >= 1 and 'commits' or 'checkins'
-        sql = "SELECT ci_when FROM %s WHERE "\
-              "repositoryid = %%s ORDER BY ci_when DESC LIMIT 1" % (commits_table)
-        sql_args = (repository_id)
-
-        cursor = self.db.cursor()
-        cursor.execute(sql, sql_args)
-        ci_when = None
-        try:
-            ci_when = cursor.fetchone()[0]
-        except TypeError:
-            return None
-        
-        return dbi.TicksFromDateTime(ci_when)
-
    def AddCommitList(self, commit_list):
        for commit in commit_list:
            self.AddCommit(commit)
@@ -352,8 +311,7 @@ class CheckinDatabase:
        minus_count = commit.GetMinusCount() or '0'
        description_id = self.GetDescriptionID(commit.GetDescription())

-        commits_table = self._version >= 1 and 'commits' or 'checkins'
-        sql = "REPLACE INTO %s" % (commits_table)
+        sql = "REPLACE INTO %s" % (self.GetCommitsTable())
        sql = sql + \
              "  (type,ci_when,whoid,repositoryid,dirid,fileid,revision,"\
              "   stickytag,branchid,addedlines,removedlines,descid)"\
@@ -394,91 +352,91 @@ class CheckinDatabase:
                match = " LIKE "
            elif query_entry.match == "glob":
                match = " REGEXP "
-                # use fnmatch to translate the glob into a regexp
+                # Use fnmatch to translate the glob into a regular
+                # expression.  Sadly, we have to account for the fact
+                # that in Python 2.6, fnmatch.translate() started
+                # sticking '\Z(?ms)' at the end of the regular
+                # expression instead of just '$', and doesn't prepend
+                # the '^'.
                data = fnmatch.translate(data)
-                if data[0] != '^': data = '^' + data
+                if data[0] != '^':
+                    data = '^' + data
+                if data[-7:] == '\Z(?ms)':
+                    data = data[:-7] + '$'
            elif query_entry.match == "regex":
                match = " REGEXP "
            elif query_entry.match == "notregex":
                match = " NOT REGEXP "
+
            sqlList.append("%s%s%s" % (field, match, self.db.literal(data)))

        return "(%s)" % (string.join(sqlList, " OR "))

-    def CreateSQLQueryString(self, query):
-        commits_table = self._version >= 1 and 'commits' or 'checkins'
-        fields = [
-            commits_table+".*",
-            "repositories.repository AS repository_name",
-            "dirs.dir AS dir_name",
-            "files.file AS file_name"]
-        tableList = [
-            (commits_table, None),
-            ("repositories","(%s.repositoryid=repositories.id)" % (commits_table)),
-            ("dirs", "(%s.dirid=dirs.id)" % (commits_table)),
-            ("files", "(%s.fileid=files.id)" % (commits_table))]
+    def CreateSQLQueryString(self, query, detect_leftover=0):
+        commits_table = self.GetCommitsTable()
+        tableList = [(commits_table, None)]
        condList = []
-        
-        if len(query.text_query):
-            tableList.append(("descs", "(descs.id=%s.descid)" % (commits_table)))
-            temp = "MATCH (descs.description) AGAINST (%s" % (self.db.literal(query.text_query))
-            condList.append("%s IN BOOLEAN MODE) > %s" % (temp, self._min_relevance))
-            fields.append("%s) AS relevance" % temp)
-        else:
-            fields.append("'' AS relevance")

        if len(query.repository_list):
+            tableList.append(("repositories",
+                              "(%s.repositoryid=repositories.id)"
+                              % (commits_table)))
            temp = self.SQLQueryListString("repositories.repository",
                                           query.repository_list)
            condList.append(temp)

        if len(query.branch_list):
-            tableList.append(("branches", "(%s.branchid=branches.id)" % (commits_table)))
+            tableList.append(("branches",
+                              "(%s.branchid=branches.id)" % (commits_table)))
            temp = self.SQLQueryListString("branches.branch",
                                           query.branch_list)
            condList.append(temp)

        if len(query.directory_list):
+            tableList.append(("dirs",
+                              "(%s.dirid=dirs.id)" % (commits_table)))
            temp = self.SQLQueryListString("dirs.dir", query.directory_list)
            condList.append(temp)
            
        if len(query.file_list):
-            tableList.append(("files", "(%s.fileid=files.id)" % (commits_table)))
+            tableList.append(("files",
+                              "(%s.fileid=files.id)" % (commits_table)))
            temp = self.SQLQueryListString("files.file", query.file_list)
            condList.append(temp)
            
-        if len(query.revision_list):
-            condList.append("(%s.revision IN (" % (commits_table) + ','.join(map(lambda s: self.db.literal(s), query.revision_list)) + "))")
-            
        if len(query.author_list):
-            tableList.append(("people", "(%s.whoid=people.id)" % (commits_table)))
+            tableList.append(("people",
+                              "(%s.whoid=people.id)" % (commits_table)))
            temp = self.SQLQueryListString("people.who", query.author_list)
            condList.append(temp)
            
        if len(query.comment_list):
-            tableList.append(("descs", "(%s.descid=descs.id)" % (commits_table)))
+            tableList.append(("descs",
+                              "(%s.descid=descs.id)" % (commits_table)))
            temp = self.SQLQueryListString("descs.description",
                                           query.comment_list)
            condList.append(temp)
            
        if query.from_date:
-            temp = "(%s.ci_when>=\"%s\")" % (commits_table, str(query.from_date))
+            temp = "(%s.ci_when>=\"%s\")" \
+                   % (commits_table, str(query.from_date))
            condList.append(temp)

        if query.to_date:
-            temp = "(%s.ci_when<=\"%s\")" % (commits_table, str(query.to_date))
+            temp = "(%s.ci_when<=\"%s\")" \
+                   % (commits_table, str(query.to_date))
            condList.append(temp)

        if query.sort == "date":
            order_by = "ORDER BY %s.ci_when DESC,descid" % (commits_table)
        elif query.sort == "author":
-            tableList.append(("people", "(%s.whoid=people.id)" % (commits_table)))
+            tableList.append(("people",
+                              "(%s.whoid=people.id)" % (commits_table)))
            order_by = "ORDER BY people.who,descid"
        elif query.sort == "file":
-            tableList.append(("files", "(%s.fileid=files.id)" % (commits_table)))
+            tableList.append(("files",
+                              "(%s.fileid=files.id)" % (commits_table)))
            order_by = "ORDER BY files.file,descid"
-        elif query.sort == "relevance" and len(query.text_query):
-            order_by = "ORDER BY relevance DESC,%s.ci_when DESC,descid" % (commits_table)

        ## exclude duplicates from the table list, and split out join
        ## conditions from table names.  In future, the join conditions
@@ -491,59 +449,51 @@ class CheckinDatabase:
                tables.append(table)
                if cond is not None: joinConds.append(cond)

-        fields = string.join(fields, ",")
        tables = string.join(tables, ",")
        conditions = string.join(joinConds + condList, " AND ")
        conditions = conditions and "WHERE %s" % conditions

-        ## limit the number of rows requested or we could really slam
-        ## a server with a large database
+        ## apply the query's row limit, if any (so we avoid really
+        ## slamming a server with a large database)
        limit = ""
        if query.limit:
-            limit = "LIMIT %s" % (str(query.limit))
-        elif self._row_limit:
-            limit = "LIMIT %s" % (str(self._row_limit))
+            if detect_leftover:
+                limit = "LIMIT %s" % (str(query.limit + 1))
+            else:
+                limit = "LIMIT %s" % (str(query.limit))

-        sql = "SELECT %s FROM %s %s %s %s" % (
-            fields, tables, conditions, order_by, limit)
+        sql = "SELECT %s.* FROM %s %s %s %s" \
+              % (commits_table, tables, conditions, order_by, limit)

        return sql
    
-    def check_commit_access(self, repos, dir, file, rev):
-        if self.authorizer:
-            rootname = repos.split('/')
-            rootname = rootname.pop()
-            path_parts = dir.split('/')
-            path_parts.append(file)
-            return self.authorizer.check_path_access(rootname, path_parts, vclib.FILE, rev)
-        return True
-
    def RunQuery(self, query):
-        sql = self.CreateSQLQueryString(query)
+        sql = self.CreateSQLQueryString(query, 1)
        cursor = self.db.cursor()
        cursor.execute(sql)
-
+        query.SetExecuted()
+        row_count = 0
+        
        while 1:
            row = cursor.fetchone()
            if not row:
                break
-
+            row_count = row_count + 1
+            if query.limit and (row_count > query.limit):
+                query.SetLimitReached()
+                break
+            
            (dbType, dbCI_When, dbAuthorID, dbRepositoryID, dbDirID,
             dbFileID, dbRevision, dbStickyTag, dbBranchID, dbAddedLines,
-             dbRemovedLines, dbDescID, dbRepositoryName, dbDirName,
-             dbFileName, dbRelevance) = row
-
-            if not self.check_commit_access(dbRepositoryName, dbDirName, dbFileName, dbRevision):
-                continue
+             dbRemovedLines, dbDescID) = row

            commit = LazyCommit(self)
            if dbType == 'Add':
-                commit.SetTypeAdd()
+              commit.SetTypeAdd()
            elif dbType == 'Remove':
-                commit.SetTypeRemove()
+              commit.SetTypeRemove()
            else:
-                commit.SetTypeChange()
-
+              commit.SetTypeChange()
            commit.SetTime(dbi.TicksFromDateTime(dbCI_When))
            commit.SetFileID(dbFileID)
            commit.SetDirectoryID(dbDirID)
@@ -554,7 +504,6 @@ class CheckinDatabase:
            commit.SetPlusCount(dbAddedLines)
            commit.SetMinusCount(dbRemovedLines)
            commit.SetDescriptionID(dbDescID)
-            commit.SetRelevance(dbRelevance)

            query.AddCommit(commit)

@@ -571,13 +520,15 @@ class CheckinDatabase:
        if file_id == None:
            return None

-        commits_table = self._version >= 1 and 'commits' or 'checkins'
-        sql = "SELECT * FROM %s WHERE "\
+        sql = "SELECT type, ci_when, whoid, repositoryid, dirid, fileid, " \
+              "revision, stickytag, branchid, addedlines, removedlines, " \
+              "descid "\
+              "  FROM %s WHERE "\
              "  repositoryid=%%s "\
              "  AND dirid=%%s"\
              "  AND fileid=%%s"\
              "  AND revision=%%s"\
-              % (commits_table)
+              % (self.GetCommitsTable())
        sql_args = (repository_id, dir_id, file_id, commit.GetRevision())

        cursor = self.db.cursor()
@@ -594,14 +545,13 @@ class CheckinDatabase:
    def sql_delete(self, table, key, value, keep_fkey = None):
        sql = "DELETE FROM %s WHERE %s=%%s" % (table, key)
        sql_args = (value, )
-        commits_table = self._version >= 1 and 'commits' or 'checkins'
        if keep_fkey:
            sql += " AND %s NOT IN (SELECT %s FROM %s WHERE %s = %%s)" \
-                   % (key, keep_fkey, commits_table, keep_fkey)
+                   % (key, keep_fkey, self.GetCommitsTable(), keep_fkey)
            sql_args = (value, value)
        cursor = self.db.cursor()
        cursor.execute(sql, sql_args)
-
+        
    def sql_purge(self, table, key, fkey, ftable):
        sql = "DELETE FROM %s WHERE %s NOT IN (SELECT %s FROM %s)" \
              % (table, key, fkey, ftable)
@@ -623,7 +573,10 @@ class CheckinDatabase:
            self.sql_purge('descs', 'id', 'descid', 'commits')
            self.sql_purge('people', 'id', 'whoid', 'commits')
        else:
-            sql = "SELECT * FROM checkins WHERE repositoryid=%s"
+            sql = "SELECT type, ci_when, whoid, repositoryid, dirid, " \
+                  "fileid, revision, stickytag, branchid, addedlines, " \
+                  "removedlines, descid "\
+                  "  FROM checkins WHERE repositoryid=%s"
            sql_args = (rep_id, )
            cursor = self.db.cursor()
            cursor.execute(sql, sql_args)
@@ -680,7 +633,6 @@ class Commit:
        self.__pluscount = ''
        self.__minuscount = ''
        self.__description = ''
-        self.__relevance = ''
        self.__gmt_time = 0.0
        self.__type = Commit.CHANGE

@@ -749,12 +701,6 @@ class Commit:
    def GetDescription(self):
        return self.__description

-    def SetRelevance(self, relevance):
-        self.__relevance = relevance
-
-    def GetRelevance(self):
-        return self.__relevance
-
    def SetTypeChange(self):
        self.__type = Commit.CHANGE

@@ -843,8 +789,9 @@ class QueryEntry:
        self.data = data
        self.match = match

-## CheckinDatabaseQueryData is a object which contains the search parameters
-## for a query to the CheckinDatabase
+## CheckinDatabaseQuery is an object which contains the search
+## parameters for a query to the Checkin Database and -- after the
+## query is executed -- the data returned by the query.
 class CheckinDatabaseQuery:
    def __init__(self):
        ## sorting
@@ -855,10 +802,8 @@ class CheckinDatabaseQuery:
        self.branch_list = []
        self.directory_list = []
        self.file_list = []
-        self.revision_list = []
        self.author_list = []
        self.comment_list = []
-        self.text_query = ""

        ## date range in DBI 2.0 timedate objects
        self.from_date = None
@@ -866,7 +811,8 @@ class CheckinDatabaseQuery:

        ## limit on number of rows to return
        self.limit = None
-
+        self.limit_reached = 0
+        
        ## list of commits -- filled in by CVS query
        self.commit_list = []

@@ -874,8 +820,8 @@ class CheckinDatabaseQuery:
        ## are added
        self.commit_cb = None

-    def SetTextQuery(self, query):
-        self.text_query = query
+        ## has this query been run?
+        self.executed = 0

    def SetRepository(self, repository, match = "exact"):
        self.repository_list.append(QueryEntry(repository, match))
@@ -889,15 +835,10 @@ class CheckinDatabaseQuery:
    def SetFile(self, file, match = "exact"):
        self.file_list.append(QueryEntry(file, match))

-    def SetRevision(self, revision):
-        r = re.compile('\s*[,;]+\s*')
-        for i in r.split(revision):
-            self.revision_list.append(i)
-
    def SetAuthor(self, author, match = "exact"):
        self.author_list.append(QueryEntry(author, match))

-    def SetComment(self, comment, match = "fulltext"):
+    def SetComment(self, comment, match = "exact"):
        self.comment_list.append(QueryEntry(comment, match))

    def SetSortMethod(self, sort):
@@ -927,6 +868,20 @@ class CheckinDatabaseQuery:
    def AddCommit(self, commit):
        self.commit_list.append(commit)

+    def SetExecuted(self):
+        self.executed = 1
+
+    def SetLimitReached(self):
+        self.limit_reached = 1
+
+    def GetLimitReached(self):
+        assert self.executed
+        return self.limit_reached
+
+    def GetCommitList(self):
+        assert self.executed
+        return self.commit_list
+        

 ##
 ## entrypoints
@@ -937,21 +892,20 @@ def CreateCommit():
 def CreateCheckinQuery():
    return CheckinDatabaseQuery()

-def ConnectDatabase(cfg, authorizer=None, readonly=0):
+def ConnectDatabase(cfg, readonly=0):
    if readonly:
        user = cfg.cvsdb.readonly_user
        passwd = cfg.cvsdb.readonly_passwd
    else:
        user = cfg.cvsdb.user
        passwd = cfg.cvsdb.passwd
-    db = CheckinDatabase(cfg.cvsdb.host, cfg.cvsdb.port, cfg.cvsdb.socket, user, passwd,
-                         cfg.cvsdb.database_name, cfg.cvsdb.row_limit, cfg.cvsdb.fulltext_min_relevance,
-                         authorizer)
+    db = CheckinDatabase(cfg.cvsdb.host, cfg.cvsdb.port, user, passwd,
+                         cfg.cvsdb.database_name)
    db.Connect()
    return db

-def ConnectDatabaseReadOnly(cfg, authorizer):
-    return ConnectDatabase(cfg, authorizer, 1)
+def ConnectDatabaseReadOnly(cfg):
+    return ConnectDatabase(cfg, 1)

 def GetCommitListFromRCSFile(repository, path_parts, revision=None):
    commit_list = []
--- a/lib/dbi.py
+++ b/lib/dbi.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2006 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -59,13 +59,5 @@ def TicksFromDateTime(datetime):
  else:
    return time.mktime(t[:8] + (-1,))
    
-def connect(host, port, socket, user, passwd, db, charset = 'utf8'):
-    return MySQLdb.connect(
-        host = host,
-        port = port,
-        unix_socket = socket,
-        user = user,
-        passwd = passwd,
-        db = db,
-        charset = charset,
-        use_unicode = charset == 'utf8')
+def connect(host, port, user, passwd, db):
+    return MySQLdb.connect(host=host, port=port, user=user, passwd=passwd, db=db)
--- a/lib/debug.py
+++ b/lib/debug.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
--- a/lib/ezt.py
+++ b/lib/ezt.py
@@ -347,7 +347,7 @@ class Template:
      for_names = [ ]

    if base_format:
-      program.append((self._cmd_format, _printers[base_format]))
+      program.append((self._cmd_format, _formatters[base_format]))

    for i in range(len(parts)):
      piece = parts[i]
@@ -405,13 +405,13 @@ class Template:
          elif cmd == 'format':
            if args[1][0]:
              # argument is a variable reference
-              printer = args[1]
+              formatter = args[1]
            else:
-              # argument is a string constant referring to built-in printer
-              printer = _printers.get(args[1][1])
-              if not printer:
+              # argument is a string constant referring to built-in formatter
+              formatter = _formatters.get(args[1][1])
+              if not formatter:
                raise UnknownFormatConstantError(str(args[1:]))
-            program.append((self._cmd_format, printer))
+            program.append((self._cmd_format, formatter))

          # remember the cmd, current pos, args, and a section placeholder
          stack.append([cmd, len(program), args[1:], None])
@@ -465,13 +465,13 @@ class Template:
    except TypeError:
      raise Exception("Unprintable value type for '%s'" % (str(valrefs[0][0])))

-  def _cmd_format(self, printer, ctx):
-    if type(printer) is TupleType:
-      printer = _get_value(printer, ctx)
-    ctx.printers.append(printer)
+  def _cmd_format(self, formatter, ctx):
+    if type(formatter) is TupleType:
+      formatter = _get_value(formatter, ctx)
+    ctx.formatters.append(formatter)

  def _cmd_end_format(self, valref, ctx):
-    ctx.printers.pop()
+    ctx.formatters.pop()

  def _cmd_include(self, (valref, reader), ctx):
    fname = _get_value(valref, ctx)
@@ -637,14 +637,23 @@ def _get_value((refname, start, rest), ctx):
  # string or a sequence
  return ob

+def _print_formatted(formatters, ctx, chunk):
+  # print chunk to ctx.fp after running it sequentially through formatters
+  for formatter in formatters:
+    chunk = formatter(chunk)
+  ctx.fp.write(chunk)
+  
 def _write_value(value, args, ctx):
  # value is a callback function, generates its own output
  if callable(value):
    apply(value, [ctx] + list(args))
    return

-  # pop printer in case it recursively calls _write_value
-  printer = ctx.printers.pop()
+  # squirrel away formatters in case one of them recursively calls
+  # _write_value() -- we'll use them (in reverse order) to format our
+  # output.
+  formatters = ctx.formatters[:]
+  formatters.reverse()

  try:
    # if the value has a 'read' attribute, then it is a stream: copy it
@@ -653,7 +662,7 @@ def _write_value(value, args, ctx):
        chunk = value.read(16384)
        if not chunk:
          break
-        printer(ctx, chunk)
+        _print_formatted(formatters, ctx, chunk)

    # value is a substitution pattern
    elif args:
@@ -666,14 +675,16 @@ def _write_value(value, args, ctx):
            piece = args[idx]
          else:
            piece = '<undef>'
-        printer(ctx, piece)
+        _print_formatted(formatters, ctx, piece)

    # plain old value, write to output
    else:
-      printer(ctx, value)
+      _print_formatted(formatters, ctx, value)

  finally:
-    ctx.printers.append(printer)
+    # restore our formatters
+    formatters.reverse()
+    ctx.formatters = formatters


 class TemplateData:
@@ -715,7 +726,7 @@ class Context:
  """A container for the execution context"""
  def __init__(self, fp):
    self.fp = fp
-    self.printers = []
+    self.formatters = []
  def write(self, value, args=()):
    _write_value(value, args, self)

@@ -828,26 +839,26 @@ class BaseUnavailableError(EZTException):
 class UnknownFormatConstantError(EZTException):
  """The format specifier is an unknown value."""

-def _raw_printer(ctx, s):
-  try: s = s.encode('utf-8')
-  except: pass
-  ctx.fp.write(s)
+def _raw_formatter(s):
+  return s

-def _html_printer(ctx, s):
-  try: s = s.encode('utf-8')
-  except: pass
-  ctx.fp.write(cgi.escape(s))
+def _html_formatter(s):
+  return cgi.escape(s)

-def _uri_printer(ctx, s):
-  try: s = s.encode('utf-8')
-  except: pass
-  ctx.fp.write(urllib.quote(s))
+def _xml_formatter(s):
+  s = s.replace('&', '&#x26;')
+  s = s.replace('<', '&#x3C;')
+  s = s.replace('>', '&#x3E;')
+  return s

-_printers = {
-  FORMAT_RAW  : _raw_printer,
-  FORMAT_HTML : _html_printer,
-  FORMAT_XML  : _html_printer,
-  FORMAT_URI  : _uri_printer,
+def _uri_formatter(s):
+  return urllib.quote(s)
+
+_formatters = {
+  FORMAT_RAW  : _raw_formatter,
+  FORMAT_HTML : _html_formatter,
+  FORMAT_XML  : _xml_formatter,
+  FORMAT_URI  : _uri_formatter,
 }

 # --- standard test environment ---
@@ -865,7 +876,7 @@ def test_parse():
         ['', '["a \\"b[foo]" c.d f]', None, '']

 def _test(argv):
-  import doctest, ezt
+  import doctest, ezt           
  verbose = "-v" in argv
  return doctest.testmod(ezt, verbose=verbose)

--- a/lib/idiff.py
+++ b/lib/idiff.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2006 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -20,7 +20,7 @@ import difflib
 import sys
 import re
 import ezt
-import cgi
+import sapi

 def sidebyside(fromlines, tolines, context):
  """Generate side by side diff"""
@@ -49,18 +49,18 @@ def _mdiff_split(flag, (line_number, text)):
  while True:
    m = _re_mdiff.search(text, pos)
    if not m:
-      segments.append(_item(text=cgi.escape(text[pos:]), type=None))
+      segments.append(_item(text=sapi.escape(text[pos:]), type=None))
      break

    if m.start() > pos:
-      segments.append(_item(text=cgi.escape(text[pos:m.start()]), type=None))
+      segments.append(_item(text=sapi.escape(text[pos:m.start()]), type=None))

    if m.group(1) == "+":
-      segments.append(_item(text=cgi.escape(m.group(2)), type="add"))
+      segments.append(_item(text=sapi.escape(m.group(2)), type="add"))
    elif m.group(1) == "-":
-      segments.append(_item(text=cgi.escape(m.group(2)), type="remove"))
+      segments.append(_item(text=sapi.escape(m.group(2)), type="remove"))
    elif m.group(1) == "^":
-      segments.append(_item(text=cgi.escape(m.group(2)), type="change"))
+      segments.append(_item(text=sapi.escape(m.group(2)), type="change"))

    pos = m.end()

@@ -166,12 +166,12 @@ def _differ_split(row, guide):

    for m in _re_differ.finditer(guide, pos):
      if m.start() > pos:
-        segments.append(_item(text=cgi.escape(line[pos:m.start()]), type=None))
-      segments.append(_item(text=cgi.escape(line[m.start():m.end()]),
+        segments.append(_item(text=sapi.escape(line[pos:m.start()]), type=None))
+      segments.append(_item(text=sapi.escape(line[m.start():m.end()]),
                            type="change"))
      pos = m.end()

-  segments.append(_item(text=cgi.escape(line[pos:]), type=None))
+  segments.append(_item(text=sapi.escape(line[pos:]), type=None))

  return _item(gap=ezt.boolean(gap), type=type, segments=segments,
               left_number=left_number, right_number=right_number)
--- a/lib/popen.py
+++ b/lib/popen.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2009 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
--- a/lib/query.py
+++ b/lib/query.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python
 # -*-python-*-
 #
-# Copyright (C) 1999-2009 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -40,22 +40,12 @@ class FormData:
        self.file = ""
        self.who = ""
        self.sortby = ""
-        self.textquery = ""
        self.date = ""
        self.hours = 0

        self.decode_thyself(form)
        
    def decode_thyself(self, form):
-        try:
-            self.textquery = string.strip(form["textquery"].value)
-        except KeyError:
-            pass
-        except TypeError:
-            pass
-        else:
-            self.valid = 1
-        
        try:
            self.repository = string.strip(form["repository"].value)
        except KeyError:
@@ -227,8 +217,9 @@ def decode_command(cmd):
    else:
        return "exact"

-def form_to_cvsdb_query(form_data):
+def form_to_cvsdb_query(cfg, form_data):
    query = cvsdb.CreateCheckinQuery()
+    query.SetLimit(cfg.cvsdb.row_limit)

    if form_data.repository:
        for cmd, str in listparse_string(form_data.repository):
@@ -255,15 +246,10 @@ def form_to_cvsdb_query(form_data):
            cmd = decode_command(cmd)
            query.SetAuthor(str, cmd)

-    if form_data.textquery:
-        query.SetTextQuery(form_data.textquery)
-
    if form_data.sortby == "author":
        query.SetSortMethod("author")
    elif form_data.sortby == "file":
        query.SetSortMethod("file")
-    elif form_data.sortby == "relevance" and form_data.textquery:
-        query.SetSortMethod("relevance")
    else:
        query.SetSortMethod("date")

@@ -290,8 +276,30 @@ def prev_rev(rev):
    return string.join(r, '.')

 def is_forbidden(cfg, cvsroot_name, module):
-    auth_params = cfg.get_authorizer_params('forbidden', cvsroot_name)
-    forbidden = auth_params.get('forbidden', '')
+    '''Return 1 if MODULE in CVSROOT_NAME is forbidden; return 0 otherwise.'''
+
+    # CVSROOT_NAME might be None here if the data comes from an
+    # unconfigured root.  This interfaces doesn't care that the root
+    # isn't configured, but if that's the case, it will consult only
+    # the base and per-vhost configuration for authorizer and
+    # authorizer parameters.
+    if cvsroot_name:
+        authorizer, params = cfg.get_authorizer_and_params_hack(cvsroot_name)
+    else:
+        authorizer = cfg.options.authorizer
+        params = cfg.get_authorizer_params()
+        
+    # If CVSROOT_NAME isn't configured to use an authorizer, nothing
+    # is forbidden.  If it's configured to use something other than
+    # the 'forbidden' authorizer, complain.  Otherwise, check for
+    # forbiddenness per the PARAMS as expected.
+    if not authorizer:
+        return 0
+    if authorizer != 'forbidden':    
+        raise Exception("The 'forbidden' authorizer is the only one supported "
+                        "by this interface.  The '%s' root is configured to "
+                        "use a different one." % (cvsroot_name))
+    forbidden = params.get('forbidden', '')
    forbidden = map(string.strip, filter(None, string.split(forbidden, ',')))
    default = 0
    for pat in forbidden:
@@ -304,12 +312,8 @@ def is_forbidden(cfg, cvsroot_name, module):
    return default
    
 def build_commit(server, cfg, desc, files, cvsroots, viewvc_link):
-    ob = _item(num_files=len(files), files=[], plus=0, minus=0)
-    
-    if desc:
-        ob.log = string.replace(server.escape(desc), '\n', '<br />')
-    else:
-        ob.log = '&nbsp;'
+    ob = _item(num_files=len(files), files=[])
+    ob.log = desc and string.replace(server.escape(desc), '\n', '<br />') or ''

    for commit in files:
        repository = commit.GetRepository()
@@ -343,9 +347,10 @@ def build_commit(server, cfg, desc, files, cvsroots, viewvc_link):
        except:
            raise Exception, str([directory, commit.GetFile()])

-        ## if we couldn't find the cvsroot path configured in the 
-        ## viewvc.conf file, then don't make the link
-        if cvsroot_name:
+        ## If we couldn't find the cvsroot path configured in the
+        ## viewvc.conf file, or we don't have a VIEWVC_LINK, then
+        ## don't make the link.
+        if cvsroot_name and viewvc_link:
            flink = '[%s] <a href="%s/%s?root=%s">%s</a>' % (
                    cvsroot_name, viewvc_link, urllib.quote(file),
                    cvsroot_name, file)
@@ -359,9 +364,6 @@ def build_commit(server, cfg, desc, files, cvsroots, viewvc_link):
            flink = '[%s] %s' % (repository, file)
            dlink = None

-        ob.relevance = commit.GetRelevance()
-        ob.plus += int(commit.GetPlusCount())
-        ob.minus += int(commit.GetMinusCount())
        ob.files.append(_item(date=ctime,
                              author=commit.GetAuthor(),
                              link=flink,
@@ -375,23 +377,28 @@ def build_commit(server, cfg, desc, files, cvsroots, viewvc_link):

    return ob

-def run_query(server, cfg, db, form_data, viewvc_link):
-    query = form_to_cvsdb_query(form_data)
+def run_query(server, cfg, form_data, viewvc_link):
+    query = form_to_cvsdb_query(cfg, form_data)
+    db = cvsdb.ConnectDatabaseReadOnly(cfg)
    db.RunQuery(query)

-    if not query.commit_list:
-        return [ ]
+    commit_list = query.GetCommitList()
+    if not commit_list:
+        return [ ], 0
+
+    row_limit_reached = query.GetLimitReached()

    commits = [ ]
    files = [ ]

    cvsroots = {}
+    viewvc.expand_root_parents(cfg)
    rootitems = cfg.general.svn_roots.items() + cfg.general.cvs_roots.items()
    for key, value in rootitems:
        cvsroots[cvsdb.CleanRepository(value)] = key

-    current_desc = query.commit_list[0].GetDescription()
-    for commit in query.commit_list:
+    current_desc = commit_list[0].GetDescription()
+    for commit in commit_list:
        desc = commit.GetDescription()
        if current_desc == desc:
            files.append(commit)
@@ -414,7 +421,7 @@ def run_query(server, cfg, db, form_data, viewvc_link):
        return len(commit.files) > 0
    commits = filter(_only_with_files, commits)
  
-    return commits
+    return commits, row_limit_reached

 def main(server, cfg, viewvc_link):
  try:
@@ -422,37 +429,36 @@ def main(server, cfg, viewvc_link):
    form = server.FieldStorage()
    form_data = FormData(form)

-    db = cvsdb.ConnectDatabaseReadOnly(cfg, None)
    if form_data.valid:
-        commits = run_query(server, cfg, db, form_data, viewvc_link)
+        commits, row_limit_reached = run_query(server, cfg,
+                                               form_data, viewvc_link)
        query = None
    else:
        commits = [ ]
+        row_limit_reached = 0
        query = 'skipped'

+    docroot = cfg.options.docroot
+    if docroot is None and viewvc_link:
+        docroot = viewvc_link + '/' + viewvc.docroot_magic_path
+        
    data = ezt.TemplateData({
      'cfg' : cfg,
      'address' : cfg.general.address,
      'vsn' : viewvc.__version__,
-
-      'textquery' : server.escape(form_data.textquery, 1),
-      'repository' : server.escape(form_data.repository, 1),
-      'branch' : server.escape(form_data.branch, 1),
-      'directory' : server.escape(form_data.directory, 1),
-      'file' : server.escape(form_data.file, 1),
-      'who' : server.escape(form_data.who, 1),
-      'docroot' : cfg.options.docroot is None \
-                  and viewvc_link + '/' + viewvc.docroot_magic_path \
-                  or cfg.options.docroot,
-
+      'repository' : server.escape(form_data.repository),
+      'branch' : server.escape(form_data.branch),
+      'directory' : server.escape(form_data.directory),
+      'file' : server.escape(form_data.file),
+      'who' : server.escape(form_data.who),
+      'docroot' : docroot,
      'sortby' : form_data.sortby,
      'date' : form_data.date,
-
      'query' : query,
+      'row_limit_reached' : ezt.boolean(row_limit_reached),
      'commits' : commits,
      'num_commits' : len(commits),
      'rss_href' : None,
-      'repositories' : db.GetRepositoryList(),
      'hours' : form_data.hours and form_data.hours or 2,
      })

--- a/lib/sapi.py
+++ b/lib/sapi.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2006 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -20,13 +20,27 @@ import string
 import os
 import sys
 import re
+import cgi


-# global server object. It will be either a CgiServer or a proxy to
-# an AspServer or ModPythonServer object.
+# global server object. It will be either a CgiServer, a WsgiServer,
+# or a proxy to an AspServer or ModPythonServer object.
 server = None


+# Simple HTML string escaping.  Note that we always escape the
+# double-quote character -- ViewVC shouldn't ever need to preserve
+# that character as-is, and sometimes needs to embed escaped values
+# into HTML attributes.
+def escape(s):
+  s = str(s)
+  s = string.replace(s, '&', '&amp;')
+  s = string.replace(s, '>', '&gt;')
+  s = string.replace(s, '<', '&lt;')
+  s = string.replace(s, '"', "&quot;")
+  return s
+
+  
 class Server:
  def __init__(self):
    self.pageGlobals = {}
@@ -34,6 +48,9 @@ class Server:
  def self(self):
    return self

+  def escape(self, s):
+    return escape(s)
+    
  def close(self):
    pass

@@ -129,9 +146,6 @@ class CgiServer(Server):
    global server
    server = self

-    global cgi
-    import cgi
-
  def addheader(self, name, value):
    self.headers.append((name, value))

@@ -160,9 +174,6 @@ class CgiServer(Server):
    self.header(status='301 Moved')
    sys.stdout.write('This document is located <a href="%s">here</a>.\n' % url)

-  def escape(self, s, quote = None):
-    return cgi.escape(s, quote)
-
  def getenv(self, name, value=None):
    ret = os.environ.get(name, value)
    if self.iis and name == 'PATH_INFO' and ret:
@@ -175,7 +186,7 @@ class CgiServer(Server):
  def FieldStorage(fp=None, headers=None, outerboundary="",
                 environ=os.environ, keep_blank_values=0, strict_parsing=0):
    return cgi.FieldStorage(fp, headers, outerboundary, environ,
-      keep_blank_values, strict_parsing)
+                            keep_blank_values, strict_parsing)

  def write(self, s):
    sys.stdout.write(s)
@@ -187,6 +198,63 @@ class CgiServer(Server):
    return sys.stdout


+class WsgiServer(Server):
+  def __init__(self, environ, start_response):
+    Server.__init__(self)
+
+    self._environ = environ
+    self._start_response = start_response;
+    self._headers = []
+    self._wsgi_write = None
+    self.headerSent = False
+
+    global server
+    server = self
+
+    global cgi
+    import cgi
+
+  def addheader(self, name, value):
+    self._headers.append((name, value))
+
+  def header(self, content_type='text/html; charset=UTF-8', status=None):
+    if not status:
+      status = "200 OK"
+    if not self.headerSent:
+      self.headerSent = True
+      self._headers.insert(0, ("Content-Type", content_type),)
+      self._wsgi_write = self._start_response("%s" % status, self._headers)
+
+  def redirect(self, url):
+    """Redirect client to url. This discards any data that has been queued
+    to be sent to the user. But there should never by any anyway.
+    """
+    self.addheader('Location', url)
+    self.header(status='301 Moved')
+    self._wsgi_write('This document is located <a href="%s">here</a>.' % url)
+
+  def getenv(self, name, value=None):
+    return self._environ.get(name, value)
+
+  def params(self):
+    return cgi.parse(environ=self._environ, fp=self._environ["wsgi.input"])
+
+  def FieldStorage(self, fp=None, headers=None, outerboundary="",
+                   environ=os.environ, keep_blank_values=0, strict_parsing=0):
+    return cgi.FieldStorage(self._environ["wsgi.input"], headers,
+                            outerboundary, self._environ, keep_blank_values,
+                            strict_parsing)
+
+  def write(self, s):
+    self._wsgi_write(s)
+
+  def flush(self):
+    pass
+
+  def file(self):
+    return File(self)
+
+
 class AspServer(ThreadedServer):
  def __init__(self, Server, Request, Response, Application):
    ThreadedServer.__init__(self)
@@ -219,9 +287,6 @@ class AspServer(ThreadedServer):
  def redirect(self, url):
    self.response.Redirect(url)

-  def escape(self, s, quote = None):
-    return self.server.HTMLEncode(str(s))
-
  def getenv(self, name, value = None):
    ret = self.request.ServerVariables(name)()
    if not type(ret) is types.UnicodeType:
@@ -283,9 +348,6 @@ class ModPythonServer(ThreadedServer):
    self.request = request
    self.headerSent = 0
    
-    global cgi
-    import cgi
-
  def addheader(self, name, value):
    self.request.headers_out.add(name, value)

@@ -308,9 +370,6 @@ class ModPythonServer(ThreadedServer):
    self.request.write("You are being redirected to <a href=\"%s\">%s</a>"
                       % (url, url))

-  def escape(self, s, quote = None):
-    return cgi.escape(s, quote)
-
  def getenv(self, name, value = None):
    try:
      return self.request.subprocess_env[name]
--- a/lib/vcauth/init.py
+++ b/lib/vcauth/init.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 2006-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 2006-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -29,7 +29,15 @@ class GenericViewVCAuthorizer:
  def check_root_access(self, rootname):
    """Return 1 iff the associated username is permitted to read ROOTNAME."""
    pass
-  
+
+  def check_universal_access(self, rootname):
+    """Return 1 if the associated username is permitted to read every
+    path in the repository at every revision, 0 if the associated
+    username is prohibited from reading any path in the repository, or
+    None if no such determination can be made (perhaps because the
+    cost of making it is too great)."""
+    pass
+    
  def check_path_access(self, rootname, path_parts, pathtype, rev=None):
    """Return 1 iff the associated username is permitted to read
    revision REV of the path PATH_PARTS (of type PATHTYPE) in
@@ -44,6 +52,9 @@ class ViewVCAuthorizer(GenericViewVCAuthorizer):
  """The uber-permissive authorizer."""
  def check_root_access(self, rootname):
    return 1
+
+  def check_universal_access(self, rootname):
+    return 1
    
  def check_path_access(self, rootname, path_parts, pathtype, rev=None):
    return 1
--- a/lib/vcauth/cvsntacl/init.py
+++ b/lib/vcauth/cvsntacl/init.py
@@ -1,100 +0,0 @@
-# -*-python-*-
-#
-# Copyright (C) 2009 Vitaliy Filippov.
-#
-# By using this file, you agree to the terms and conditions set forth in
-# the LICENSE.html file which can be found at the top level of the ViewVC
-# distribution or at http://viewvc.org/license-1.html.
-#
-# For more information, visit http://viewvc.org/
-#
-# -----------------------------------------------------------------------
-
-import vcauth
-import vclib
-import string
-from xml.dom.ext.reader import Sax2
-from xml import xpath
-
-class ViewVCAuthorizer(vcauth.GenericViewVCAuthorizer):
-  """An authorizer making use of CVSnt access control lists (which are in form
-     of XML files in CVS/ subdirectories in the repository)."""
-
-  def __init__(self, username, params={}):
-    self.username = username
-    self.params = params
-    self.cfg = params['__config']
-    self.default = params.get('default', 0)
-    self.cached = {}
-    self.xmlcache = {}
-
-  def checkr(self, element, paths):
-    r = None
-    for p in paths:
-      nodes = xpath.Evaluate(p, element)
-      if nodes and len(nodes):
-        for c in nodes:
-          if c.nodeName == 'read' and r is None:
-            r = True
-            if c.attributes and len(c.attributes):
-              for a in c.attributes:
-                if a.nodeName == 'deny':
-                  r = not a.value
-      if r is not None:
-        break
-    return r
-
-  def check(self, rootname, path_parts, filename):
-    d = self.cfg.general.cvs_roots.get(rootname,None)
-    if not d:
-      return self.default
-    i = len(path_parts)
-    r = None
-    while i >= 0:
-      try:
-        xml = d
-        if len(path_parts):
-          xml = xml + '/' + string.join(path_parts, '/')
-        xml = xml + '/CVS/fileattr.xml'
-        if self.cached.get(xml, None) is not None:
-          return self.cached.get(xml, None)
-	doc = self.xmlcache.get(xml, None)
-	if doc is None:
-          fp = open(xml, 'rb')
-          doc = Sax2.Reader().fromStream(fp)
-          fp.close()
-	  self.xmlcache[xml] = doc
-        if filename:
-          r = self.checkr(doc.documentElement, [
-            '/fileattr/file[@name=\'%s\']/acl[@user=\'%s\' and not(@branch)]/read' % (filename, self.username),
-            '/fileattr/file[@name=\'%s\']/acl[not(@user) and not(@branch)]/read' % filename,
-            '/fileattr/directory/acl[@user=\'%s\' and not(@branch)]/read' % self.username,
-            '/fileattr/directory/acl[not(@user) and not(@branch)]/read'
-          ] )
-        else:
-          r = self.checkr(doc.documentElement, [
-            '/fileattr/directory/acl[@user=\'%s\' and not(@branch)]/read' % self.username,
-            '/fileattr/directory/acl[not(@user) and not(@branch)]/read'
-          ] )
-        if r is not None:
-          self.cached[xml] = r
-          return r
-        raise Exception(None)
-      except:
-        if len(path_parts) > 0:
-          path_parts = path_parts[:-1]
-        filename = ''
-        i = i-1
-    return self.default
-
-  def check_root_access(self, rootname):
-    return self.check(rootname, [], '')
-
-  def check_path_access(self, rootname, path_parts, pathtype, rev=None):
-    if not path_parts:
-      return 1
-    if pathtype == vclib.DIR:
-      return self.check(rootname, path_parts, '')
-    f = path_parts[-1]
-    path_parts = path_parts[:-1]
-    return self.check(rootname, path_parts, f)
--- a/lib/vcauth/forbidden/init.py
+++ b/lib/vcauth/forbidden/init.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 2006-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 2006-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -23,7 +23,14 @@ class ViewVCAuthorizer(vcauth.GenericViewVCAuthorizer):

  def check_root_access(self, rootname):
    return 1
-  
+
+  def check_universal_access(self, rootname):
+    # If there aren't any forbidden paths, we can grant universal read
+    # access.  Otherwise, we make no claim.
+    if not self.forbidden:
+      return 1
+    return None
+    
  def check_path_access(self, rootname, path_parts, pathtype, rev=None):
    # No path?  No problem.
    if not path_parts:
--- a/lib/vcauth/forbiddenre/init.py
+++ b/lib/vcauth/forbiddenre/init.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 2008-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -46,6 +46,13 @@ class ViewVCAuthorizer(vcauth.GenericViewVCAuthorizer):
  def check_root_access(self, rootname):
    return self._check_root_path_access(rootname)
  
+  def check_universal_access(self, rootname):
+    # If there aren't any forbidden regexps, we can grant universal
+    # read access.  Otherwise, we make no claim.
+    if not self.forbidden:
+      return 1
+    return None
+    
  def check_path_access(self, rootname, path_parts, pathtype, rev=None):
    root_path = rootname
    if path_parts:
--- a/lib/vcauth/grp/init.py
+++ b/lib/vcauth/grp/init.py
@@ -1,57 +0,0 @@
-# -*-python-*-
-#
-# Copyright (C) 2009 Vitaliy Filippov.
-#
-# By using this file, you agree to the terms and conditions set forth in
-# the LICENSE.html file which can be found at the top level of the ViewVC
-# distribution or at http://viewvc.org/license-1.html.
-#
-# For more information, visit http://viewvc.org/
-#
-# -----------------------------------------------------------------------
-
-import vcauth
-import vclib
-import string
-import grp
-import re
-
-class ViewVCAuthorizer(vcauth.GenericViewVCAuthorizer):
-  """A simple authorizer checking system group membership for every
-     repository and every user."""
-
-  def __init__(self, username, params={}):
-    self.username = username
-    self.params = params
-    self.cfg = params['__config']
-    self.cached = {}
-    self.grp = {}
-    self.byroot = {}
-    self.fmt = map(lambda l: l.strip(), params.get('group_name_format', 'svn.%s|svn.%s.ro').split('|'))
-    byr = params.get('by_root', '')
-    for i in byr.split(','):
-      if i.find(':') < 0:
-        continue
-      (root, auth) = i.split(':', 2)
-      self.byroot[root.strip()] = map(lambda l: l.strip(), auth.split('|'))
-
-  def check_root_access(self, rootname):
-    r = self.cached.get(rootname, None)
-    if r is not None:
-      return r
-    try:
-      grent = self.grp.get(rootname, None)
-      if grent is None:
-        grent = map(lambda grn: grp.getgrnam(grn.replace('%s', re.sub('[^\w\.\-]+', '', rootname))), self.byroot.get(rootname, self.fmt))
-        self.grp[rootname] = grent
-      for i in grent:
-        if i.gr_mem and len(i.gr_mem) and self.username in i.gr_mem:
-          r = 1
-          break
-    except:
-      r = 0
-    self.cached[rootname] = r
-    return r
-
-  def check_path_access(self, rootname, path_parts, pathtype, rev=None):
-    return self.check_root_access(rootname)
--- a/lib/vcauth/svnauthz/init.py
+++ b/lib/vcauth/svnauthz/init.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 2006-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 2006-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -34,9 +34,9 @@ class ViewVCAuthorizer(vcauth.GenericViewVCAuthorizer):
    # See if the admin wants us to do case normalization of usernames.
    self.force_username_case = params.get('force_username_case')
    if self.force_username_case == "upper":
-      self.username = username.upper()
+      self.username = username and string.upper(username) or username
    elif self.force_username_case == "lower":
-      self.username = username.lower()
+      self.username = username and string.lower(username) or username
    elif not self.force_username_case:
      self.username = username
    else:
@@ -54,7 +54,10 @@ class ViewVCAuthorizer(vcauth.GenericViewVCAuthorizer):
    # option names.
    cp = ConfigParser()
    cp.optionxform = lambda x: x
-    cp.read(self.authz_file)
+    try:
+      cp.read(self.authz_file)
+    except:
+      raise debug.ViewVCException("Unable to parse configured authzfile file")

    # Figure out if there are any aliases for the current username
    aliases = []
@@ -221,6 +224,36 @@ class ViewVCAuthorizer(vcauth.GenericViewVCAuthorizer):
    paths = self._get_paths_for_root(rootname)
    return (paths is not None) and 1 or 0
  
+  def check_universal_access(self, rootname):
+    paths = self._get_paths_for_root(rootname)
+    if not paths: # None or empty.
+      return 0
+
+    # Search the access determinations.  If there's a mix, we can't
+    # claim a universal access determination.
+    found_allow = 0
+    found_deny = 0
+    for access in paths.values():
+      if access:
+        found_allow = 1
+      else:
+        found_deny = 1
+      if found_allow and found_deny:
+        return None
+
+    # We didn't find both allowances and denials, so we must have
+    # found one or the other.  Denials only is a universal denial.
+    if found_deny:
+      return 0
+
+    # ... but allowances only is only a universal allowance if read
+    # access is granted to the root directory.
+    if found_allow and paths.has_key('/'):
+      return 1
+
+    # Anything else is indeterminable.
+    return None
+    
  def check_path_access(self, rootname, path_parts, pathtype, rev=None):
    # Crawl upward from the path represented by PATH_PARTS toward to
    # the root of the repository, looking for an explicitly grant or
--- a/lib/vcauth/union/init.py
+++ b/lib/vcauth/union/init.py
@@ -1,69 +0,0 @@
-# -*-python-*-
-#
-# Copyright (C) 2009 Vitaliy Filippov.
-#
-# By using this file, you agree to the terms and conditions set forth in
-# the LICENSE.html file which can be found at the top level of the ViewVC
-# distribution or at http://viewvc.org/license-1.html.
-#
-# For more information, visit http://viewvc.org/
-#
-# -----------------------------------------------------------------------
-
-import vcauth
-import vclib
-import string
-
-class ViewVCAuthorizer(vcauth.GenericViewVCAuthorizer):
-  """A 'union' authorizer: it makes possible to use one authorizer for
-     one root and other authorizer for other roots."""
-
-  def __init__(self, username, params={}):
-    self.username = username
-    self.params = params
-    self.cfg = params['__config']
-    self.default = params.get('default', '')
-    self.byroot = {}
-    self.authz = {}
-    union = params.get('union', '')
-    for i in union.split(','):
-      if i.find(':') < 0:
-        continue
-      (root, auth) = i.split(':', 2)
-      self.byroot[root.strip()] = auth.strip()
-
-  def create_authz(self, rootname):
-    aname = self.byroot.get(rootname, '') or self.default
-    if not aname:
-      return None
-    if self.authz.get(aname, None):
-      return self.authz[aname]
-    import imp
-    fp = None
-    try:
-      try:
-        fp, path, desc = imp.find_module(aname, vcauth.__path__)
-        my_auth = imp.load_module('viewvc', fp, path, desc)
-      except ImportError:
-        raise debug.ViewVCException(
-          'Invalid authorizer (%s) specified for root "%s"' \
-          % (self.cfg.options.authorizer, rootname),
-          '500 Internal Server Error')
-    finally:
-      if fp:
-        fp.close()
-    params = self.cfg.get_authorizer_params(aname, rootname)
-    self.authz[aname] = my_auth.ViewVCAuthorizer(self.username, params)
-    return self.authz[aname]
-
-  def check_root_access(self, rootname):
-    a = self.create_authz(rootname)
-    if a:
-      return a.check_root_access(rootname)
-    return None
-
-  def check_path_access(self, rootname, path_parts, pathtype, rev=None):
-    a = self.create_authz(rootname)
-    if a:
-      return a.check_path_access(rootname, path_parts, pathtype, rev)
-    return None
--- a/lib/vclib/init.py
+++ b/lib/vclib/init.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -76,7 +76,7 @@ class Repository:
    """
    pass

-  def openfile(self, path_parts, rev):
+  def openfile(self, path_parts, rev, options):
    """Open a file object to read file contents at a given path and revision.

    The return value is a 2-tuple of containg the file object and revision
@@ -86,6 +86,8 @@ class Repository:
    of the repository. e.g. ["subdir1", "subdir2", "filename"]

    rev is the revision of the file to check out
+
+    options is a dictionary of implementation specific options
    """

  def listdir(self, path_parts, rev, options):
@@ -168,20 +170,29 @@ class Repository:
    Return value is a python file object
    """

-  def annotate(self, path_parts, rev):
-    """Return a list of annotate file content lines and a revision.
+  def annotate(self, path_parts, rev, include_text=False):
+    """Return a list of Annotation object, sorted by their
+    "line_number" components, which describe the lines of given
+    version of a file.

-    The result is a list of Annotation objects, sorted by their
-    line_number components.
-    """
+    The file path is specified as a list of components, relative to
+    the root of the repository. e.g. ["subdir1", "subdir2", "filename"]
+
+    rev is the revision of the item to return information about.
+    
+    If include_text is true, populate the Annotation objects' "text"
+    members with the corresponding line of file content; otherwise,
+    leave that member set to None."""

  def revinfo(self, rev):
    """Return information about a global revision

    rev is the revision of the item to return information about
    
-    Return value is a 4-tuple containing the date, author, log
-    message, and a list of ChangedPath items representing paths changed
+    Return value is a 5-tuple containing: the date, author, log
+    message, a list of ChangedPath items representing paths changed,
+    and a dictionary mapping property names to property values for
+    properties stored on an item.

    Raise vclib.UnsupportedFeature if the version control system
    doesn't support a global revision concept.
@@ -196,7 +207,21 @@ class Repository:

    rev is the revision of the item to return information about
    """
+
+  def filesize(self, path_parts, rev):
+    """Return the size of a versioned file's contents if it can be
+    obtained without a brute force measurement; -1 otherwise.
+
+    NOTE: Callers that require a filesize answer when this function
+    returns -1 may obtain it by measuring the data returned via
+    openfile().
    
+    The path is specified as a list of components, relative to the root
+    of the repository. e.g. ["subdir1", "subdir2", "filename"]
+
+    rev is the revision of the item to return information about
+    """
+
    
 # ======================================================================
 class DirEntry:
@@ -328,7 +353,7 @@ def _diff_args(type, options):
  if type == CONTEXT:
    if options.has_key('context'):
      if options['context'] is None:
-        args.append('--context=%i' % 0x01ffffff)
+        args.append('--context=-1')
      else:
        args.append('--context=%i' % options['context'])
    else:
@@ -336,7 +361,7 @@ def _diff_args(type, options):
  elif type == UNIFIED:
    if options.has_key('context'):
      if options['context'] is None:
-        args.append('--unified=%i' % 0x01ffffff)
+        args.append('--unified=-1')
      else:
        args.append('--unified=%i' % options['context'])
    else:
@@ -381,13 +406,13 @@ class _diff_fp:
        self.fp = None
    finally:
      try:
-        if self.temp1 and self.temp1 != '/dev/null':
+        if self.temp1:
          os.remove(self.temp1)
-        self.temp1 = None
+          self.temp1 = None
      finally:
-        if self.temp2 and self.temp2 != '/dev/null':
+        if self.temp2:
          os.remove(self.temp2)
-        self.temp2 = None
+          self.temp2 = None

  def __del__(self):
    self.close()
--- a/lib/vclib/ccvs/init.py
+++ b/lib/vclib/ccvs/init.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -14,6 +14,7 @@ import os.path


 def canonicalize_rootpath(rootpath):
+  assert os.path.isabs(rootpath)
  return os.path.normpath(rootpath)


@@ -22,6 +23,7 @@ def expand_root_parent(parent_path):
  # "CVSROOT/config" is added the set of returned roots.  Or, if the
  # PARENT_PATH itself contains a child "CVSROOT/config", then all its
  # subdirectories are returned as roots.
+  assert os.path.isabs(parent_path)
  roots = {}
  subpaths = os.listdir(parent_path)
  cvsroot = os.path.exists(os.path.join(parent_path, "CVSROOT", "config"))
@@ -33,6 +35,23 @@ def expand_root_parent(parent_path):
  return roots


+def find_root_in_parent(parent_path, rootname):
+  """Search PARENT_PATH for a root named ROOTNAME, returning the
+  canonicalized ROOTPATH of the root if found; return None if no such
+  root is found."""
+
+  assert os.path.isabs(parent_path)
+  # Is PARENT_PATH itself a CVS repository?  If so, we allow ROOTNAME
+  # to be any subdir within it.  Otherwise, we expect
+  # PARENT_PATH/ROOTNAME to be a CVS repository.
+  rootpath = os.path.join(parent_path, rootname)
+  if os.path.exists(os.path.join(parent_path, "CVSROOT", "config")):
+    return canonicalize_rootpath(rootpath)
+  if os.path.exists(os.path.join(rootpath, "CVSROOT", "config")):
+    return canonicalize_rootpath(rootpath)
+  return None
+
+
 def CVSRepository(name, rootpath, authorizer, utilities, use_rcsparse):
  rootpath = canonicalize_rootpath(rootpath)
  if use_rcsparse:
--- a/lib/vclib/ccvs/bincvs.py
+++ b/lib/vclib/ccvs/bincvs.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -21,7 +21,6 @@ import stat
 import string
 import re
 import time
-import cvsdb

 # ViewVC libs
 import compat
@@ -41,6 +40,11 @@ class BaseCVSRepository(vclib.Repository):
    if not vclib.check_root_access(self):
      raise vclib.ReposNotFound(name)

+  def open(self):
+    # See if a universal read access determination can be made.
+    if self.auth and self.auth.check_universal_access(self.name) == 1:
+      self.auth = None
+
  def rootname(self):
    return self.name

@@ -145,6 +149,11 @@ class BaseCVSRepository(vclib.Repository):
                        % (string.join(path_parts, "/")))
    rcsfile = self.rcsfile(path_parts, 1)
    return os.access(rcsfile, os.X_OK)
+  
+  def filesize(self, path_parts, rev):
+    if self.itemtype(path_parts, rev) != vclib.FILE:  # does auth-check
+      raise vclib.Error("Path '%s' is not a file." % (_path_join(path_parts)))
+    return -1


 class BinCVSRepository(BaseCVSRepository):
@@ -163,7 +172,14 @@ class BinCVSRepository(BaseCVSRepository):
      return revs[-1]
    return None

-  def openfile(self, path_parts, rev):
+  def openfile(self, path_parts, rev, options):
+    """see vclib.Repository.openfile docstring
+
+    Option values recognized by this implementation:
+
+      cvs_oldkeywords
+        boolean. true to use the original keyword substitution values.
+    """
    if self.itemtype(path_parts, rev) != vclib.FILE:  # does auth-check
      raise vclib.Error("Path '%s' is not a file."
                        % (string.join(path_parts, "/")))
@@ -171,12 +187,14 @@ class BinCVSRepository(BaseCVSRepository):
      rev_flag = '-p'
    else:
      rev_flag = '-p' + rev
+    if options.get('cvs_oldkeywords', 0):
+      kv_flag = '-ko'
+    else:
+      kv_flag = '-kkv'
    full_name = self.rcsfile(path_parts, root=1, v=0)
-
    used_rlog = 0
    tip_rev = None  # used only if we have to fallback to using rlog
-
-    fp = self.rcs_popen('co', (rev_flag, full_name), 'rb') 
+    fp = self.rcs_popen('co', (kv_flag, rev_flag, full_name), 'rb') 
    try:
      filename, revision = _parse_co_header(fp)
    except COMissingRevision:
@@ -319,13 +337,13 @@ class BinCVSRepository(BaseCVSRepository):
      args = rcs_args
    return popen.popen(cmd, args, mode, capture_err)

-  def annotate(self, path_parts, rev=None):
+  def annotate(self, path_parts, rev=None, include_text=False):
    if self.itemtype(path_parts, rev) != vclib.FILE:  # does auth-check
      raise vclib.Error("Path '%s' is not a file."
                        % (string.join(path_parts, "/")))
-
+                        
    from vclib.ccvs import blame
-    source = blame.BlameSource(self.rcsfile(path_parts, 1), rev)
+    source = blame.BlameSource(self.rcsfile(path_parts, 1), rev, include_text)
    return source, source.revision

  def revinfo(self, rev):
@@ -338,19 +356,13 @@ class BinCVSRepository(BaseCVSRepository):

      ignore_keyword_subst - boolean, ignore keyword substitution
    """
-    if not path_parts1:
-      path_parts1 = path_parts2
-      rev1 = '1.0'
-    if not path_parts2:
-      path_parts2 = path_parts1
-      rev2 = '1.0'
    if self.itemtype(path_parts1, rev1) != vclib.FILE:  # does auth-check
      raise vclib.Error("Path '%s' is not a file."
                        % (string.join(path_parts1, "/")))
    if self.itemtype(path_parts2, rev2) != vclib.FILE:  # does auth-check
      raise vclib.Error("Path '%s' is not a file."
                        % (string.join(path_parts2, "/")))
-
+    
    args = vclib._diff_args(type, options)
    if options.get('ignore_keyword_subst', 0):
      args.append('-kk')
@@ -358,7 +370,8 @@ class BinCVSRepository(BaseCVSRepository):
    rcsfile = self.rcsfile(path_parts1, 1)
    if path_parts1 != path_parts2:
      raise NotImplementedError, "cannot diff across paths in cvs"
-    args.extend(['-N', '-r' + rev1, '-r' + rev2, rcsfile])
+    args.extend(['-r' + rev1, '-r' + rev2, rcsfile])
+    
    fp = self.rcs_popen('rcsdiff', args, 'rt')

    # Eat up the non-GNU-diff-y headers.
@@ -367,6 +380,7 @@ class BinCVSRepository(BaseCVSRepository):
      if not line or line[0:5] == 'diff ':
        break
    return fp
+  

 class CVSDirEntry(vclib.DirEntry):
  def __init__(self, name, kind, errors, in_attic, absent=0):
@@ -838,8 +852,6 @@ def _parse_log_entry(fp):
      raise ValueError, 'invalid year'
  date = compat.timegm(tm)

-  log = cvsdb.utf8string(log)
-
  return Revision(rev, date,
                  # author, state, lines changed
                  match.group(2), match.group(3) == "dead", match.group(5),
@@ -1029,16 +1041,16 @@ def _get_logs(repos, dir_path_parts, entries, view_tag, get_dirs):
        file.errors.append("rlog error: %s" % msg)
        continue

+      tag = None
      if view_tag == 'MAIN' or view_tag == 'HEAD':
        tag = Tag(None, default_branch)
      elif taginfo.has_key(view_tag):
        tag = Tag(None, taginfo[view_tag])
-      elif view_tag:
-        # the tag wasn't found, so skip this file
+      elif view_tag and (eof != _EOF_FILE):
+        # the tag wasn't found, so skip this file (unless we already
+        # know there's nothing left of it to read)
        _skip_file(rlog)
-        eof = 1
-      else:
-        tag = None
+        eof = _EOF_FILE

      # we don't care about the specific values -- just the keys and whether
      # the values point to branches or revisions. this the fastest way to 
--- a/lib/vclib/ccvs/blame.py
+++ b/lib/vclib/ccvs/blame.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 # Copyright (C) 2000 Curt Hagenlocher <curt@hagenlocher.org>
 #
 # By using this file, you agree to the terms and conditions set forth in
@@ -32,7 +32,6 @@ import time
 import math
 import rcsparse
 import vclib
-import cvsdb

 class CVSParser(rcsparse.Sink):
  # Precompiled regular expressions
@@ -415,7 +414,7 @@ class CVSParser(rcsparse.Sink):


 class BlameSource:
-  def __init__(self, rcs_file, opt_rev=None):
+  def __init__(self, rcs_file, opt_rev=None, include_text=False):
    # Parse the CVS file
    parser = CVSParser()
    revision = parser.parse_cvs_file(rcs_file, opt_rev)
@@ -429,6 +428,7 @@ class BlameSource:
    self.lines = lines
    self.num_lines = count
    self.parser = parser
+    self.include_text = include_text

    # keep track of where we are during an iteration
    self.idx = -1
@@ -447,7 +447,9 @@ class BlameSource:
    prev_rev = self.parser.prev_revision.get(rev)
    line_number = idx + 1
    author = self.parser.revision_author[rev]
-    thisline = cvsdb.utf8string(self.lines[idx])
+    thisline = self.lines[idx]
+    if not self.include_text:
+      thisline = None
    ### TODO:  Put a real date in here.
    item = vclib.Annotation(thisline, line_number, rev, prev_rev, author, None)
    self.last = item
--- a/lib/vclib/ccvs/ccvs.py
+++ b/lib/vclib/ccvs/ccvs.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -19,7 +19,6 @@ import tempfile
 import vclib
 import rcsparse
 import blame
-import cvsdb

 ### The functionality shared with bincvs should probably be moved to a
 ### separate module
@@ -67,7 +66,6 @@ class CCVSRepository(BaseCVSRepository):
        entry.path = path
        try:
          rcsparse.parse(open(path, 'rb'), InfoSink(entry, rev, alltags))
-          entry.log = cvsdb.utf8string(entry.log)
        except IOError, e:
          entry.errors.append("rcsparse error: %s" % e)
        except RuntimeError, e:
@@ -121,49 +119,40 @@ class CCVSRepository(BaseCVSRepository):
    return filtered_revs

  def rawdiff(self, path_parts1, rev1, path_parts2, rev2, type, options={}):
-    if path_parts1 and self.itemtype(path_parts1, rev1) != vclib.FILE:  # does auth-check
+    if self.itemtype(path_parts1, rev1) != vclib.FILE:  # does auth-check
      raise vclib.Error("Path '%s' is not a file."
                        % (string.join(path_parts1, "/")))
-    if path_parts2 and self.itemtype(path_parts2, rev2) != vclib.FILE:  # does auth-check
+    if self.itemtype(path_parts2, rev2) != vclib.FILE:  # does auth-check
      raise vclib.Error("Path '%s' is not a file."
                        % (string.join(path_parts2, "/")))
-    if not path_parts1 and not path_parts2:
-      raise vclib.Error("Nothing to diff.")
+    
+    temp1 = tempfile.mktemp()
+    open(temp1, 'wb').write(self.openfile(path_parts1, rev1, {})[0].getvalue())
+    temp2 = tempfile.mktemp()
+    open(temp2, 'wb').write(self.openfile(path_parts2, rev2, {})[0].getvalue())

-    if path_parts1:
-      temp1 = tempfile.mktemp()
-      open(temp1, 'wb').write(self.openfile(path_parts1, rev1)[0].getvalue())
-      r1 = self.itemlog(path_parts1, rev1, vclib.SORTBY_DEFAULT, 0, 0, {})[-1]
-      info1 = (self.rcsfile(path_parts1, root=1, v=0), r1.date, r1.string)
-    else:
-      temp1 = '/dev/null'
-      info1 = ('/dev/null', '', '')
+    r1 = self.itemlog(path_parts1, rev1, vclib.SORTBY_DEFAULT, 0, 0, {})[-1]
+    r2 = self.itemlog(path_parts2, rev2, vclib.SORTBY_DEFAULT, 0, 0, {})[-1]

-    if path_parts2:
-      temp2 = tempfile.mktemp()
-      open(temp2, 'wb').write(self.openfile(path_parts2, rev2)[0].getvalue())
-      r2 = self.itemlog(path_parts2, rev2, vclib.SORTBY_DEFAULT, 0, 0, {})[-1]
-      info2 = (self.rcsfile(path_parts2, root=1, v=0), r2.date, r2.string)
-    else:
-      temp2 = '/dev/null'
-      info2 = ('/dev/null', '', '')
+    info1 = (self.rcsfile(path_parts1, root=1, v=0), r1.date, r1.string)
+    info2 = (self.rcsfile(path_parts2, root=1, v=0), r2.date, r2.string)

    diff_args = vclib._diff_args(type, options)

    return vclib._diff_fp(temp1, temp2, info1, info2,
                          self.utilities.diff or 'diff', diff_args)

-  def annotate(self, path_parts, rev=None):
+  def annotate(self, path_parts, rev=None, include_text=False):
    if self.itemtype(path_parts, rev) != vclib.FILE:  # does auth-check
      raise vclib.Error("Path '%s' is not a file."
                        % (string.join(path_parts, "/")))
-    source = blame.BlameSource(self.rcsfile(path_parts, 1), rev)
+    source = blame.BlameSource(self.rcsfile(path_parts, 1), rev, include_text)
    return source, source.revision

  def revinfo(self, rev):
    raise vclib.UnsupportedFeature

-  def openfile(self, path_parts, rev=None):
+  def openfile(self, path_parts, rev, options):
    if self.itemtype(path_parts, rev) != vclib.FILE:  # does auth-check
      raise vclib.Error("Path '%s' is not a file."
                        % (string.join(path_parts, "/")))
@@ -211,6 +200,7 @@ class InfoSink(MatchingSink):
    self.matching_rev = None
    self.perfect_match = 0
    self.lockinfo = { }
+    self.saw_revision = False

  def define_tag(self, name, revision):
    MatchingSink.define_tag(self, name, revision)
@@ -224,10 +214,17 @@ class InfoSink(MatchingSink):
        self.entry.absent = 1
      raise rcsparse.RCSStopParser

+  def parse_completed(self):
+    if not self.saw_revision:
+      #self.entry.errors.append("No revisions exist on %s" % (view_tag or "MAIN"))
+      self.entry.absent = 1
+    
  def set_locker(self, rev, locker):
    self.lockinfo[rev] = locker
    
  def define_revision(self, revision, date, author, state, branches, next):
+    self.saw_revision = True
+    
    if self.perfect_match:
      return

@@ -235,14 +232,18 @@ class InfoSink(MatchingSink):
    rev = Revision(revision, date, author, state == "dead")
    rev.lockinfo = self.lockinfo.get(revision)
    
-    # perfect match if revision number matches tag number or if revision is on
-    # trunk and tag points to trunk. imperfect match if tag refers to a branch
-    # and this revision is the highest revision so far found on that branch
+    # perfect match if revision number matches tag number or if
+    # revision is on trunk and tag points to trunk.  imperfect match
+    # if tag refers to a branch and either a) this revision is the
+    # highest revision so far found on that branch, or b) this
+    # revision is the branchpoint.
    perfect = ((rev.number == tag.number) or
               (not tag.number and len(rev.number) == 2))
-    if perfect or (tag.is_branch and tag.number == rev.number[:-1] and
-                   (not self.matching_rev or
-                    rev.number > self.matching_rev.number)):
+    if perfect or (tag.is_branch and \
+                   ((tag.number == rev.number[:-1] and
+                     (not self.matching_rev or
+                      rev.number > self.matching_rev.number)) or
+                    (rev.number == tag.number[:-1]))):
      self.matching_rev = rev
      self.perfect_match = perfect

--- a/lib/vclib/ccvs/rcsparse/init.py
+++ b/lib/vclib/ccvs/rcsparse/init.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2006 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
--- a/lib/vclib/ccvs/rcsparse/common.py
+++ b/lib/vclib/ccvs/rcsparse/common.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -194,7 +194,8 @@ class _Parser:
        else:
          # Chew up "newphrase"
          # warn("Unexpected RCS token: $token\n")
-          pass
+          while self.ts.get() != ';':
+            pass
      else:
        if f is None:
          self.ts.unget(token)
--- a/lib/vclib/ccvs/rcsparse/debug.py
+++ b/lib/vclib/ccvs/rcsparse/debug.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2006 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
--- a/lib/vclib/ccvs/rcsparse/default.py
+++ b/lib/vclib/ccvs/rcsparse/default.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -19,7 +19,11 @@ import string
 import common

 class _TokenStream:
-  token_term = string.whitespace + ';:'
+  token_term = string.whitespace + ";:"
+  try:
+    token_term = frozenset(token_term)
+  except NameError:
+    pass

  # the algorithm is about the same speed for any CHUNK_SIZE chosen.
  # grab a good-sized chunk, but not too large to overwhelm memory.
@@ -44,15 +48,17 @@ class _TokenStream:
    # out more complex solutions.

    buf = self.buf
+    lbuf = len(buf)
    idx = self.idx

    while 1:
-      if idx == len(buf):
+      if idx == lbuf:
        buf = self.rcsfile.read(self.CHUNK_SIZE)
        if buf == '':
          # signal EOF by returning None as the token
          del self.buf   # so we fail if get() is called again
          return None
+        lbuf = len(buf)
        idx = 0

      if buf[idx] not in string.whitespace:
@@ -60,7 +66,7 @@ class _TokenStream:

      idx = idx + 1

-    if buf[idx] == ';' or buf[idx] == ':':
+    if buf[idx] in ';:':
      self.buf = buf
      self.idx = idx + 1
      return buf[idx]
@@ -70,17 +76,18 @@ class _TokenStream:
      token = ''
      while 1:
        # find token characters in the current buffer
-        while end < len(buf) and buf[end] not in self.token_term:
+        while end < lbuf and buf[end] not in self.token_term:
          end = end + 1
        token = token + buf[idx:end]

-        if end < len(buf):
+        if end < lbuf:
          # we stopped before the end, so we have a full token
          idx = end
          break

        # we stopped at the end of the buffer, so we may have a partial token
        buf = self.rcsfile.read(self.CHUNK_SIZE)
+        lbuf = len(buf)
        idx = end = 0

      self.buf = buf
@@ -94,22 +101,24 @@ class _TokenStream:
    chunks = [ ]

    while 1:
-      if idx == len(buf):
+      if idx == lbuf:
        idx = 0
        buf = self.rcsfile.read(self.CHUNK_SIZE)
        if buf == '':
          raise RuntimeError, 'EOF'
+        lbuf = len(buf)
      i = string.find(buf, '@', idx)
      if i == -1:
        chunks.append(buf[idx:])
-        idx = len(buf)
+        idx = lbuf
        continue
-      if i == len(buf) - 1:
+      if i == lbuf - 1:
        chunks.append(buf[idx:i])
        idx = 0
        buf = '@' + self.rcsfile.read(self.CHUNK_SIZE)
        if buf == '@':
          raise RuntimeError, 'EOF'
+        lbuf = len(buf)
        continue
      if buf[i + 1] == '@':
        chunks.append(buf[idx:i+1])
--- a/lib/vclib/ccvs/rcsparse/texttools.py
+++ b/lib/vclib/ccvs/rcsparse/texttools.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
--- a/lib/vclib/svn/init.py
+++ b/lib/vclib/svn/init.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2008 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -15,17 +15,50 @@
 import os
 import os.path
 import re
+import urllib

 _re_url = re.compile('^(http|https|file|svn|svn\+[^:]+)://')

-def canonicalize_rootpath(rootpath):
+def _canonicalize_path(path):
+  import svn.core
  try:
-    import svn.core
-    return svn.core.svn_path_canonicalize(rootpath)
-  except:
-    if re.search(_re_url, rootpath):
-      return rootpath[-1] == '/' and rootpath[:-1] or rootpath
-    return os.path.normpath(rootpath)
+    return svn.core.svn_path_canonicalize(path)
+  except AttributeError: # svn_path_canonicalize() appeared in 1.4.0 bindings
+    # There's so much more that we *could* do here, but if we're
+    # here at all its because there's a really old Subversion in
+    # place, and those older Subversion versions cared quite a bit
+    # less about the specifics of path canonicalization.
+    if re.search(_re_url, path):
+      return path.rstrip('/')
+    else:
+      return os.path.normpath(path)
+
+
+def canonicalize_rootpath(rootpath):
+  # Try to canonicalize the rootpath using Subversion semantics.
+  rootpath = _canonicalize_path(rootpath)
+  
+  # ViewVC's support for local repositories is more complete and more
+  # performant than its support for remote ones, so if we're on a
+  # Unix-y system and we have a file:/// URL, convert it to a local
+  # path instead.
+  if os.name == 'posix':
+    rootpath_lower = rootpath.lower()
+    if rootpath_lower in ['file://localhost',
+                          'file://localhost/',
+                          'file://',
+                          'file:///'
+                          ]:
+      return '/'
+    if rootpath_lower.startswith('file://localhost/'):
+      rootpath = os.path.normpath(urllib.unquote(rootpath[16:]))
+    elif rootpath_lower.startswith('file:///'):
+      rootpath = os.path.normpath(urllib.unquote(rootpath[7:]))
+
+  # Ensure that we have an absolute path (or URL), and return.
+  if not re.search(_re_url, rootpath):
+    assert os.path.isabs(rootpath)
+  return rootpath


 def expand_root_parent(parent_path):
@@ -35,6 +68,7 @@ def expand_root_parent(parent_path):
  else:
    # Any subdirectories of PARENT_PATH which themselves have a child
    # "format" are returned as roots.
+    assert os.path.isabs(parent_path)
    subpaths = os.listdir(parent_path)
    for rootname in subpaths:
      rootpath = os.path.join(parent_path, rootname)
@@ -43,6 +77,20 @@ def expand_root_parent(parent_path):
  return roots


+def find_root_in_parent(parent_path, rootname):
+  """Search PARENT_PATH for a root named ROOTNAME, returning the
+  canonicalized ROOTPATH of the root if found; return None if no such
+  root is found."""
+  
+  if not re.search(_re_url, parent_path):
+    assert os.path.isabs(parent_path)
+    rootpath = os.path.join(parent_path, rootname)
+    format_path = os.path.join(rootpath, "format")
+    if os.path.exists(format_path):
+      return canonicalize_rootpath(rootpath)
+  return None
+
+
 def SubversionRepository(name, rootpath, authorizer, utilities, config_dir):
  rootpath = canonicalize_rootpath(rootpath)
  if re.search(_re_url, rootpath):
--- a/lib/vclib/svn/svn_ra.py
+++ b/lib/vclib/svn/svn_ra.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2009 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -18,10 +18,12 @@ import os
 import string
 import re
 import tempfile
-import popen2
 import time
 import urllib
-from svn_repos import Revision, SVNChangedPath, _datestr_to_date, _compare_paths, _path_parts, _cleanup_path, _rev2optrev, _fix_subversion_exception
+from svn_repos import Revision, SVNChangedPath, _datestr_to_date, \
+                      _compare_paths, _path_parts, _cleanup_path, \
+                      _rev2optrev, _fix_subversion_exception, \
+                      _split_revprops, _canonicalize_path
 from svn import core, delta, client, wc, ra


@@ -52,13 +54,71 @@ def get_directory_props(ra_session, path, rev):
    props = ra.svn_ra_get_dir(ra_session, path, rev)
  return props

+def client_log(url, start_rev, end_rev, log_limit, include_changes,
+               cross_copies, cb_func, ctx):
+  include_changes = include_changes and 1 or 0
+  cross_copies = cross_copies and 1 or 0
+  try:
+    client.svn_client_log4([url], start_rev, start_rev, end_rev,
+                           log_limit, include_changes, not cross_copies,
+                           0, None, cb_func, ctx)
+  except AttributeError:
+    # Wrap old svn_log_message_receiver_t interface with a
+    # svn_log_entry_t one.
+    def cb_convert(paths, revision, author, date, message, pool):
+      class svn_log_entry_t:
+        pass
+      log_entry = svn_log_entry_t()
+      log_entry.changed_paths = paths
+      log_entry.revision = revision
+      log_entry.revprops = { core.SVN_PROP_REVISION_LOG : message,
+                             core.SVN_PROP_REVISION_AUTHOR : author,
+                             core.SVN_PROP_REVISION_DATE : date,
+                             }
+      cb_func(log_entry, pool)
+    client.svn_client_log2([url], start_rev, end_rev, log_limit,
+                           include_changes, not cross_copies, cb_convert, ctx)
+
+
+def setup_client_ctx(config_dir):
+  # Ensure that the configuration directory exists.
+  core.svn_config_ensure(config_dir)
+
+  # Fetch the configuration (and 'config' bit thereof).
+  cfg = core.svn_config_get_config(config_dir)
+  config = cfg.get(core.SVN_CONFIG_CATEGORY_CONFIG)
+
+  # Here's the compat-sensitive part: try to use
+  # svn_cmdline_create_auth_baton(), and fall back to making our own
+  # if that fails.
+  try:
+    auth_baton = core.svn_cmdline_create_auth_baton(1, None, None, config_dir,
+                                                    1, 1, config, None)
+  except AttributeError:
+    auth_baton = core.svn_auth_open([
+      client.svn_client_get_simple_provider(),
+      client.svn_client_get_username_provider(),
+      client.svn_client_get_ssl_server_trust_file_provider(),
+      client.svn_client_get_ssl_client_cert_file_provider(),
+      client.svn_client_get_ssl_client_cert_pw_file_provider(),
+      ])
+    if config_dir is not None:
+      core.svn_auth_set_parameter(auth_baton,
+                                  core.SVN_AUTH_PARAM_CONFIG_DIR,
+                                  config_dir)
+
+  # Create, setup, and return the client context baton.
+  ctx = client.svn_client_create_context()
+  ctx.config = cfg
+  ctx.auth_baton = auth_baton
+  return ctx
+
 ### END COMPATABILITY CODE ###


 class LogCollector:
-  ### TODO: Make this thing authz-aware
  
-  def __init__(self, path, show_all_logs, lockinfo):
+  def __init__(self, path, show_all_logs, lockinfo, access_check_func):
    # This class uses leading slashes for paths internally
    if not path:
      self.path = '/'
@@ -67,8 +127,16 @@ class LogCollector:
    self.logs = []
    self.show_all_logs = show_all_logs
    self.lockinfo = lockinfo
+    self.access_check_func = access_check_func
+    self.done = False
+    
+  def add_log(self, log_entry, pool):
+    if self.done:
+      return
+    paths = log_entry.changed_paths
+    revision = log_entry.revision
+    msg, author, date, revprops = _split_revprops(log_entry.revprops)
    
-  def add_log(self, paths, revision, author, date, message, pool):
    # Changed paths have leading slashes
    changed_paths = paths.keys()
    changed_paths.sort(lambda a, b: _compare_paths(a, b))
@@ -88,13 +156,17 @@ class LogCollector:
          if change.copyfrom_path:
            this_path = change.copyfrom_path + self.path[len(changed_path):]
    if self.show_all_logs or this_path:
-      entry = Revision(revision, _datestr_to_date(date), author, message, None,
-                       self.lockinfo, self.path[1:], None, None)
-      self.logs.append(entry)
+      if self.access_check_func is None \
+         or self.access_check_func(self.path[1:], revision):
+        entry = Revision(revision, date, author, msg, None, self.lockinfo,
+                         self.path[1:], None, None)
+        self.logs.append(entry)
+      else:
+        self.done = True
    if this_path:
      self.path = this_path
    
-def temp_checkout(svnrepos, path, rev):
+def cat_to_tempfile(svnrepos, path, rev):
  """Check out file revision to temporary file"""
  temp = tempfile.mktemp()
  stream = core.svn_stream_from_aprfile(temp)
@@ -155,21 +227,8 @@ class RemoteSubversionRepository(vclib.Repository):

  def open(self):
    # Setup the client context baton, complete with non-prompting authstuffs.
-    # TODO: svn_cmdline_setup_auth_baton() is mo' better (when available)
-    core.svn_config_ensure(self.config_dir)
-    self.ctx = client.svn_client_ctx_t()
-    self.ctx.auth_baton = core.svn_auth_open([
-      client.svn_client_get_simple_provider(),
-      client.svn_client_get_username_provider(),
-      client.svn_client_get_ssl_server_trust_file_provider(),
-      client.svn_client_get_ssl_client_cert_file_provider(),
-      client.svn_client_get_ssl_client_cert_pw_file_provider(),
-      ])
-    self.ctx.config = core.svn_config_get_config(self.config_dir)
-    if self.config_dir is not None:
-      core.svn_auth_set_parameter(self.ctx.auth_baton,
-                                  core.SVN_AUTH_PARAM_CONFIG_DIR,
-                                  self.config_dir)
+    self.ctx = setup_client_ctx(self.config_dir)
+    
    ra_callbacks = ra.svn_ra_callbacks_t()
    ra_callbacks.auth_baton = self.ctx.auth_baton
    self.ra_session = ra.svn_ra_open(self.rootpath, ra_callbacks, None,
@@ -177,6 +236,10 @@ class RemoteSubversionRepository(vclib.Repository):
    self.youngest = ra.svn_ra_get_latest_revnum(self.ra_session)
    self._dirent_cache = { }
    self._revinfo_cache = { }
+
+    # See if a universal read access determination can be made.
+    if self.auth and self.auth.check_universal_access(self.name) == 1:
+      self.auth = None
    
  def rootname(self):
    return self.name
@@ -211,25 +274,23 @@ class RemoteSubversionRepository(vclib.Repository):
      raise vclib.ItemNotFound(path_parts)
    return pathtype

-  def openfile(self, path_parts, rev):
+  def openfile(self, path_parts, rev, options):
    path = self._getpath(path_parts)
    if self.itemtype(path_parts, rev) != vclib.FILE:  # does auth-check
      raise vclib.Error("Path '%s' is not a file." % path)
    rev = self._getrev(rev)
    url = self._geturl(path)
-    tmp_file = tempfile.mktemp()
-    stream = core.svn_stream_from_aprfile(tmp_file)
    ### rev here should be the last history revision of the URL
-    client.svn_client_cat(core.Stream(stream), url, _rev2optrev(rev), self.ctx)
-    core.svn_stream_close(stream)
-    return SelfCleanFP(tmp_file), self._get_last_history_rev(path_parts, rev)
+    fp = SelfCleanFP(cat_to_tempfile(self, path, rev))
+    lh_rev, c_rev = self._get_last_history_rev(path_parts, rev)
+    return fp, lh_rev

  def listdir(self, path_parts, rev, options):
    path = self._getpath(path_parts)
    if self.itemtype(path_parts, rev) != vclib.DIR:  # does auth-check
      raise vclib.Error("Path '%s' is not a directory." % path)
    rev = self._getrev(rev)
-    entries = [ ]
+    entries = []
    dirents, locks = self._get_dirents(path, rev)
    for name in dirents.keys():
      entry = dirents[name]
@@ -237,8 +298,9 @@ class RemoteSubversionRepository(vclib.Repository):
        kind = vclib.DIR
      elif entry.kind == core.svn_node_file:
        kind = vclib.FILE
-      if vclib.check_path_access(self, path_parts + [name], kind, rev):
-        entries.append(vclib.DirEntry(name, kind))
+      else:
+        kind = None
+      entries.append(vclib.DirEntry(name, kind))
    return entries

  def dirlogs(self, path_parts, rev, entries, options):
@@ -249,11 +311,13 @@ class RemoteSubversionRepository(vclib.Repository):
    dirents, locks = self._get_dirents(path, rev)
    for entry in entries:
      entry_path_parts = path_parts + [entry.name]
-      if not vclib.check_path_access(self, entry_path_parts, entry.kind, rev):
+      dirent = dirents.get(entry.name, None)
+      # dirents is authz-sanitized, so ensure the entry is found therein.
+      if dirent is None:
        continue
-      dirent = dirents[entry.name]
-      entry.date, entry.author, entry.log, changes = \
-                  self.revinfo(dirent.created_rev)
+      # Get authz-sanitized revision metadata.
+      entry.date, entry.author, entry.log, revprops, changes = \
+                  self._revinfo(dirent.created_rev)
      entry.rev = str(dirent.created_rev)
      entry.size = dirent.size
      entry.lockinfo = None
@@ -267,29 +331,51 @@ class RemoteSubversionRepository(vclib.Repository):
    rev = self._getrev(rev)
    url = self._geturl(path)

-    # Use ls3 to fetch the lock status for this item.
-    lockinfo = None
-    basename = path_parts and path_parts[-1] or ""
-    dirents, locks = list_directory(url, _rev2optrev(rev),
-                                    _rev2optrev(rev), 0, self.ctx)
-    if locks.has_key(basename):
-      lockinfo = locks[basename].owner
+    # If this is a file, fetch the lock status and size (as of REV)
+    # for this item.
+    lockinfo = size_in_rev = None
+    if path_type == vclib.FILE:
+      basename = path_parts[-1]
+      list_url = self._geturl(self._getpath(path_parts[:-1]))
+      dirents, locks = list_directory(list_url, _rev2optrev(rev),
+                                      _rev2optrev(rev), 0, self.ctx)
+      if locks.has_key(basename):
+        lockinfo = locks[basename].owner
+      if dirents.has_key(basename):
+        size_in_rev = dirents[basename].size
+    
+    # Special handling for the 'svn_latest_log' scenario.
+    ### FIXME: Don't like this hack.  We should just introduce
+    ### something more direct in the vclib API.
+    if options.get('svn_latest_log', 0):
+      dir_lh_rev, dir_c_rev = self._get_last_history_rev(path_parts, rev)
+      date, author, log, revprops, changes = self._revinfo(dir_lh_rev)
+      return [vclib.Revision(dir_lh_rev, str(dir_lh_rev), date, author,
+                             None, log, size_in_rev, lockinfo)]

+    def _access_checker(check_path, check_rev):
+      return vclib.check_path_access(self, _path_parts(check_path),
+                                     path_type, check_rev)
+      
    # It's okay if we're told to not show all logs on a file -- all
    # the revisions should match correctly anyway.
-    lc = LogCollector(path, options.get('svn_show_all_dir_logs', 0), lockinfo)
+    lc = LogCollector(path, options.get('svn_show_all_dir_logs', 0),
+                      lockinfo, _access_checker)

    cross_copies = options.get('svn_cross_copies', 0)
    log_limit = 0
    if limit:
      log_limit = first + limit
-    client.svn_client_log2([url], _rev2optrev(rev), _rev2optrev(1),
-                           log_limit, 1, not cross_copies,
-                           lc.add_log, self.ctx)
+    client_log(url, _rev2optrev(rev), _rev2optrev(1), log_limit, 1,
+               cross_copies, lc.add_log, self.ctx)
    revs = lc.logs
    revs.sort()
    prev = None
    for rev in revs:
+      # Swap out revision info with stuff from the cache (which is
+      # authz-sanitized).
+      rev.date, rev.author, rev.log, revprops, changes \
+                = self._revinfo(rev.number)
      rev.prev = prev
      prev = rev
    revs.reverse()
@@ -309,13 +395,25 @@ class RemoteSubversionRepository(vclib.Repository):
                                        _rev2optrev(rev), 0, self.ctx)
    return pairs and pairs[0][1] or {}
  
-  def annotate(self, path_parts, rev):
+  def annotate(self, path_parts, rev, include_text=False):
    path = self._getpath(path_parts)
    if self.itemtype(path_parts, rev) != vclib.FILE:  # does auth-check
      raise vclib.Error("Path '%s' is not a file." % path)
    rev = self._getrev(rev)
    url = self._geturl(path)

+    # Examine logs for the file to determine the oldest revision we are
+    # permitted to see.
+    log_options = {
+      'svn_cross_copies' : 1,
+      'svn_show_all_dir_logs' : 1,
+      }
+    revs = self.itemlog(path_parts, rev, vclib.SORTBY_REV, 0, 0, log_options)
+    oldest_rev = revs[-1].number
+
+    # Now calculate the annotation data.  Note that we'll not
+    # inherently trust the provided author and date, because authz
+    # rules might necessitate that we strip that information out.
    blame_data = []

    def _blame_cb(line_no, revision, author, date,
@@ -323,69 +421,69 @@ class RemoteSubversionRepository(vclib.Repository):
      prev_rev = None
      if revision > 1:
        prev_rev = revision - 1
-      blame_data.append(vclib.Annotation(line, line_no+1, revision, prev_rev,
-                                         author, None))
-      
-    client.svn_client_blame(url, _rev2optrev(1), _rev2optrev(rev),
-                            _blame_cb, self.ctx)

+      # If we have an invalid revision, clear the date and author
+      # values.  Otherwise, if we have authz filtering to do, use the
+      # revinfo cache to do so.
+      if revision < 0:
+        date = author = None
+      elif self.auth:
+        date, author, msg, revprops, changes = self._revinfo(revision)
+
+      # Strip text if the caller doesn't want it.
+      if not include_text:
+        line = None
+      blame_data.append(vclib.Annotation(line, line_no + 1, revision, prev_rev,
+                                         author, date))
+      
+    client.blame2(url, _rev2optrev(rev), _rev2optrev(oldest_rev),
+                  _rev2optrev(rev), _blame_cb, self.ctx)
    return blame_data, rev

  def revinfo(self, rev):
-    rev = self._getrev(rev)
-    cached_info = self._revinfo_cache.get(rev)
-    if not cached_info:
-      cached_info = self._revinfo_raw(rev)
-      self._revinfo_cache[rev] = cached_info
-    return cached_info[0], cached_info[1], cached_info[2], cached_info[3]
+    return self._revinfo(rev, 1)
    
  def rawdiff(self, path_parts1, rev1, path_parts2, rev2, type, options={}):
-
-    if path_parts1 is not None:
-      p1 = self._getpath(path_parts1)
-      r1 = self._getrev(rev1)
-      if not vclib.check_path_access(self, path_parts1, vclib.FILE, rev1):
-        raise vclib.ItemNotFound(path_parts1)
-    else:
-      p1 = None
-
-    if path_parts2 is not None:
-      if not p1:
-        raise vclib.ItemNotFound(parh_parts2)
-      p2 = self._getpath(path_parts2)
-      r2 = self._getrev(rev2)
-      if not vclib.check_path_access(self, path_parts2, vclib.FILE, rev2):
-        raise vclib.ItemNotFound(path_parts2)
-    else:
-      p2 = None
+    p1 = self._getpath(path_parts1)
+    p2 = self._getpath(path_parts2)
+    r1 = self._getrev(rev1)
+    r2 = self._getrev(rev2)
+    if not vclib.check_path_access(self, path_parts1, vclib.FILE, rev1):
+      raise vclib.ItemNotFound(path_parts1)
+    if not vclib.check_path_access(self, path_parts2, vclib.FILE, rev2):
+      raise vclib.ItemNotFound(path_parts2)

    args = vclib._diff_args(type, options)

    def _date_from_rev(rev):
-      date, author, msg, changes = self.revinfo(rev)
+      date, author, msg, revprops, changes = self._revinfo(rev)
      return date
-
+    
    try:
+      temp1 = cat_to_tempfile(self, p1, r1)
+      temp2 = cat_to_tempfile(self, p2, r2)
      info1 = p1, _date_from_rev(r1), r1
      info2 = p2, _date_from_rev(r2), r2
-      if p1:
-        temp1 = temp_checkout(self, p1, r1)
-      else:
-        temp1 = '/dev/null'
-      if p2:
-        temp2 = temp_checkout(self, p2, r2)
-      else:
-        temp2 = '/dev/null'
      return vclib._diff_fp(temp1, temp2, info1, info2, self.diff_cmd, args)
    except core.SubversionException, e:
      _fix_subversion_exception(e)
-      if e.apr_err == core.SVN_ERR_FS_NOT_FOUND:
+      if e.apr_err == vclib.svn.core.SVN_ERR_FS_NOT_FOUND:
        raise vclib.InvalidRevision
      raise
+
  def isexecutable(self, path_parts, rev):
    props = self.itemprops(path_parts, rev) # does authz-check
    return props.has_key(core.SVN_PROP_EXECUTABLE)
  
+  def filesize(self, path_parts, rev):
+    path = self._getpath(path_parts)
+    if self.itemtype(path_parts, rev) != vclib.FILE:  # does auth-check
+      raise vclib.Error("Path '%s' is not a file." % path)
+    rev = self._getrev(rev)
+    dirents, locks = self._get_dirents(self._getpath(path_parts[:-1]), rev)
+    dirent = dirents.get(path_parts[-1], None)
+    return dirent.size
+    
  def _getpath(self, path_parts):
    return string.join(path_parts, '/')

@@ -393,8 +491,11 @@ class RemoteSubversionRepository(vclib.Repository):
    if rev is None or rev == 'HEAD':
      return self.youngest
    try:
+      if type(rev) == type(''):
+        while rev[0] == 'r':
+          rev = rev[1:]
      rev = int(rev)
-    except ValueError:
+    except:
      raise vclib.InvalidRevision(rev)
    if (rev < 0) or (rev > self.youngest):
      raise vclib.InvalidRevision(rev)
@@ -403,58 +504,139 @@ class RemoteSubversionRepository(vclib.Repository):
  def _geturl(self, path=None):
    if not path:
      return self.rootpath
-    return self.rootpath + '/' + urllib.quote(path, "/*~")
+    path = self.rootpath + '/' + urllib.quote(path)
+    return _canonicalize_path(path)

  def _get_dirents(self, path, rev):
    """Return a 2-type of dirents and locks, possibly reading/writing
-    from a local cache of that information."""
+    from a local cache of that information.  This functions performs
+    authz checks, stripping out unreadable dirents."""

    dir_url = self._geturl(path)
+    path_parts = _path_parts(path)    
    if path:
      key = str(rev) + '/' + path
    else:
      key = str(rev)
+
+    # Ensure that the cache gets filled...
    dirents_locks = self._dirent_cache.get(key)
    if not dirents_locks:
-      dirents, locks = list_directory(dir_url, _rev2optrev(rev),
-                                      _rev2optrev(rev), 0, self.ctx)
+      tmp_dirents, locks = list_directory(dir_url, _rev2optrev(rev),
+                                          _rev2optrev(rev), 0, self.ctx)
+      dirents = {}
+      for name, dirent in tmp_dirents.items():
+        dirent_parts = path_parts + [name]
+        kind = dirent.kind 
+        if (kind == core.svn_node_dir or kind == core.svn_node_file) \
+           and vclib.check_path_access(self, dirent_parts,
+                                       kind == core.svn_node_dir \
+                                         and vclib.DIR or vclib.FILE, rev):
+          lh_rev, c_rev = self._get_last_history_rev(dirent_parts, rev)
+          dirent.created_rev = lh_rev
+          dirents[name] = dirent
      dirents_locks = [dirents, locks]
      self._dirent_cache[key] = dirents_locks
+
+    # ...then return the goodies from the cache.
    return dirents_locks[0], dirents_locks[1]

  def _get_last_history_rev(self, path_parts, rev):
+    """Return the a 2-tuple which contains:
+         - the last interesting revision equal to or older than REV in
+           the history of PATH_PARTS.
+         - the created_rev of of PATH_PARTS as of REV."""
+    
+    path = self._getpath(path_parts)
    url = self._geturl(self._getpath(path_parts))
    optrev = _rev2optrev(rev)
+
+    # Get the last-changed-rev.
    revisions = []
    def _info_cb(path, info, pool, retval=revisions):
      revisions.append(info.last_changed_rev)
    client.svn_client_info(url, optrev, optrev, _info_cb, 0, self.ctx)
-    return revisions[0]
-    
-  def _revinfo_raw(self, rev):
-    # return 5-tuple (date, author, message, changes)
-    optrev = _rev2optrev(rev)
-    revs = []
+    last_changed_rev = revisions[0]

-    def _log_cb(changed_paths, revision, author,
-                datestr, message, pool, retval=revs):
-      date = _datestr_to_date(datestr)
+    # Now, this object might not have been directly edited since the
+    # last-changed-rev, but it might have been the child of a copy.
+    # To determine this, we'll run a potentially no-op log between
+    # LAST_CHANGED_REV and REV.
+    lc = LogCollector(path, 1, None, None)
+    client_log(url, optrev, _rev2optrev(last_changed_rev), 1, 1, 0,
+               lc.add_log, self.ctx)
+    revs = lc.logs
+    if revs:
+      revs.sort()
+      return revs[0].number, last_changed_rev
+    else:
+      return last_changed_rev, last_changed_rev
+    
+  def _revinfo_fetch(self, rev, include_changed_paths=0):
+    need_changes = include_changed_paths or self.auth
+    revs = []
+    
+    def _log_cb(log_entry, pool, retval=revs):
+      # If Subversion happens to call us more than once, we choose not
+      # to care.
+      if retval:
+        return
+      
+      revision = log_entry.revision
+      msg, author, date, revprops = _split_revprops(log_entry.revprops)
      action_map = { 'D' : vclib.DELETED,
                     'A' : vclib.ADDED,
                     'R' : vclib.REPLACED,
                     'M' : vclib.MODIFIED,
                     }
-      paths = (changed_paths or {}).keys()
+
+      # Easy out: if we won't use the changed-path info, just return a
+      # changes-less tuple.
+      if not need_changes:
+        return revs.append([date, author, msg, revprops, None])
+
+      # Subversion 1.5 and earlier didn't offer the 'changed_paths2'
+      # hash, and in Subversion 1.6, it's offered but broken.
+      try: 
+        changed_paths = log_entry.changed_paths2
+        paths = (changed_paths or {}).keys()
+      except:
+        changed_paths = log_entry.changed_paths
+        paths = (changed_paths or {}).keys()
      paths.sort(lambda a, b: _compare_paths(a, b))
+
+      # If we get this far, our caller needs changed-paths, or we need
+      # them for authz-related sanitization.
      changes = []
      found_readable = found_unreadable = 0
      for path in paths:
-        pathtype = None
        change = changed_paths[path]
+
+        # svn_log_changed_path_t (which we might get instead of the
+        # svn_log_changed_path2_t we'd prefer) doesn't have the
+        # 'node_kind' member.        
+        pathtype = None
+        if hasattr(change, 'node_kind'):
+          if change.node_kind == core.svn_node_dir:
+            pathtype = vclib.DIR
+          elif change.node_kind == core.svn_node_file:
+            pathtype = vclib.FILE
+            
+        # svn_log_changed_path2_t only has the 'text_modified' and
+        # 'props_modified' bits in Subversion 1.7 and beyond.  And
+        # svn_log_changed_path_t is without.
+        text_modified = props_modified = 0
+        if hasattr(change, 'text_modified'):
+          if change.text_modified == core.svn_tristate_true:
+            text_modified = 1
+        if hasattr(change, 'props_modified'):
+          if change.props_modified == core.svn_tristate_true:
+            props_modified = 1
+            
+        # Wrong, diddily wrong wrong wrong.  Can you say,
+        # "Manufacturing data left and right because it hurts to
+        # figure out the right stuff?"
        action = action_map.get(change.action, vclib.MODIFIED)
-        ### Wrong, diddily wrong wrong wrong.  Can you say,
-        ### "Manufacturing data left and right because it hurts to
-        ### figure out the right stuff?"
        if change.copyfrom_path and change.copyfrom_rev:
          is_copy = 1
          base_path = change.copyfrom_path
@@ -467,31 +649,61 @@ class RemoteSubversionRepository(vclib.Repository):
          base_path = path
          base_rev = revision - 1

-        ### Check authz rules (we lie about the path type)
+        # Check authz rules (sadly, we have to lie about the path type)
        parts = _path_parts(path)
        if vclib.check_path_access(self, parts, vclib.FILE, revision):
          if is_copy and base_path and (base_path != path):
            parts = _path_parts(base_path)
-            if vclib.check_path_access(self, parts, vclib.FILE, base_rev):
+            if not vclib.check_path_access(self, parts, vclib.FILE, base_rev):
              is_copy = 0
              base_path = None
              base_rev = None
+              found_unreadable = 1
          changes.append(SVNChangedPath(path, revision, pathtype, base_path,
-                                        base_rev, action, is_copy, 0, 0))
+                                        base_rev, action, is_copy,
+                                        text_modified, props_modified))
          found_readable = 1
        else:
          found_unreadable = 1

+        # If our caller doesn't want changed-path stuff, and we have
+        # the info we need to make an authz determination already,
+        # quit this loop and get on with it.
+        if (not include_changed_paths) and found_unreadable and found_readable:
+          break
+
+      # Filter unreadable information.
      if found_unreadable:
-        message = None
+        msg = None
        if not found_readable:
          author = None
          date = None
-      revs.append([date, author, message, changes])

-    client.svn_client_log([self.rootpath], optrev, optrev,
-                          1, 0, _log_cb, self.ctx)
-    return revs[0][0], revs[0][1], revs[0][2], revs[0][3]
+      # Drop unrequested changes.
+      if not include_changed_paths:
+        changes = None
+
+      # Add this revision information to the "return" array.
+      retval.append([date, author, msg, revprops, changes])
+
+    optrev = _rev2optrev(rev)
+    client_log(self.rootpath, optrev, optrev, 1, need_changes, 0,
+               _log_cb, self.ctx)
+    return tuple(revs[0])
+
+  def _revinfo(self, rev, include_changed_paths=0):
+    """Internal-use, cache-friendly revision information harvester."""
+
+    # Consult the revinfo cache first.  If we don't have cached info,
+    # or our caller wants changed paths and we don't have those for
+    # this revision, go do the real work.
+    rev = self._getrev(rev)
+    cached_info = self._revinfo_cache.get(rev)
+    if not cached_info \
+       or (include_changed_paths and cached_info[4] is None):
+      cached_info = self._revinfo_fetch(rev, include_changed_paths)
+      self._revinfo_cache[rev] = cached_info
+    return cached_info

  ##--- custom --##

@@ -510,23 +722,16 @@ class RemoteSubversionRepository(vclib.Repository):
      old_path = results[old_rev]
    except KeyError:
      raise vclib.ItemNotFound(path)
-  
-    return _cleanup_path(old_path)
+    old_path = _cleanup_path(old_path)
+    old_path_parts = _path_parts(old_path)
+    # Check access (lying about path types)
+    if not vclib.check_path_access(self, old_path_parts, vclib.FILE, old_rev):
+      raise vclib.ItemNotFound(path)
+    return old_path
  
  def created_rev(self, path, rev):
-    # NOTE: We can't use svn_client_propget here because the
-    # interfaces in that layer strip out the properties not meant for
-    # human consumption (such as svn:entry:committed-rev, which we are
-    # using here to get the created revision of PATH@REV).
-    kind = ra.svn_ra_check_path(self.ra_session, path, rev)
-    if kind == core.svn_node_none:
-      raise vclib.ItemNotFound(_path_parts(path))
-    elif kind == core.svn_node_dir:
-      props = get_directory_props(self.ra_session, path, rev)
-    elif kind == core.svn_node_file:
-      fetched_rev, props = ra.svn_ra_get_file(self.ra_session, path, rev, None)
-    return int(props.get(core.SVN_PROP_ENTRY_COMMITTED_REV,
-                         SVN_INVALID_REVNUM))
+    lh_rev, c_rev = self._get_last_history_rev(_path_parts(path), rev)
+    return lh_rev

  def last_rev(self, path, peg_revision, limit_revision=None):
    """Given PATH, known to exist in PEG_REVISION, find the youngest
@@ -561,3 +766,33 @@ class RemoteSubversionRepository(vclib.Repository):
        else:
          peg_revision = mid
      return peg_revision, path
+
+  def get_symlink_target(self, path_parts, rev):
+    """Return the target of the symbolic link versioned at PATH_PARTS
+    in REV, or None if that object is not a symlink."""
+
+    path = self._getpath(path_parts)
+    path_type = self.itemtype(path_parts, rev) # does auth-check
+    rev = self._getrev(rev)
+    url = self._geturl(path)
+
+    # Symlinks must be files with the svn:special property set on them
+    # and with file contents which read "link SOME_PATH".
+    if path_type != vclib.FILE:
+      return None
+    pairs = client.svn_client_proplist2(url, _rev2optrev(rev),
+                                        _rev2optrev(rev), 0, self.ctx)
+    props = pairs and pairs[0][1] or {}
+    if not props.has_key(core.SVN_PROP_SPECIAL):
+      return None
+    pathspec = ''
+    ### FIXME: We're being a touch sloppy here, first by grabbing the
+    ### whole file and then by checking only the first line
+    ### of it.
+    fp = SelfCleanFP(cat_to_tempfile(self, path, rev))
+    pathspec = fp.readline()
+    fp.close()
+    if pathspec[:5] != 'link ':
+      return None
+    return pathspec[5:]
+
--- a/lib/vclib/svn/svn_repos.py
+++ b/lib/vclib/svn/svn_repos.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2009 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
@@ -15,15 +15,14 @@
 import vclib
 import os
 import os.path
-import stat
 import string
 import cStringIO
 import signal
-import shutil
 import time
 import tempfile
 import popen
 import re
+import urllib
 from svn import fs, repos, core, client, delta


@@ -31,13 +30,27 @@ from svn import fs, repos, core, client, delta
 if (core.SVN_VER_MAJOR, core.SVN_VER_MINOR, core.SVN_VER_PATCH) < (1, 3, 1):
  raise Exception, "Version requirement not met (needs 1.3.1 or better)"

+
+### Pre-1.5 Subversion doesn't have SVN_ERR_CEASE_INVOCATION
+try:
+  _SVN_ERR_CEASE_INVOCATION = core.SVN_ERR_CEASE_INVOCATION
+except:
+  _SVN_ERR_CEASE_INVOCATION = core.SVN_ERR_CANCELLED
+
 ### Pre-1.5 SubversionException's might not have the .msg and .apr_err members
 def _fix_subversion_exception(e):
  if not hasattr(e, 'apr_err'):
    e.apr_err = e[1]
  if not hasattr(e, 'message'):
    e.message = e[0]
-  
+
+### Pre-1.4 Subversion doesn't have svn_path_canonicalize()
+def _canonicalize_path(path):
+  try:
+    return core.svn_path_canonicalize(path)
+  except AttributeError:
+    return path
+
 def _allow_all(root, path, pool):
  """Generic authz_read_func that permits access to all paths"""
  return 1
@@ -103,11 +116,37 @@ def _rev2optrev(rev):

 def _rootpath2url(rootpath, path):
  rootpath = os.path.abspath(rootpath)
-  if rootpath and rootpath[0] != '/':
-    rootpath = '/' + rootpath
+  drive, rootpath = os.path.splitdrive(rootpath)
  if os.sep != '/':
    rootpath = string.replace(rootpath, os.sep, '/')
-  return 'file://' + string.join([rootpath, path], "/")
+  rootpath = urllib.quote(rootpath)
+  path = urllib.quote(path)
+  if drive:
+    url = 'file:///' + drive + rootpath + '/' + path
+  else:
+    url = 'file://' + rootpath + '/' + path
+  return _canonicalize_path(url)
+
+
+# Given a dictionary REVPROPS of revision properties, pull special
+# ones out of them and return a 4-tuple containing the log message,
+# the author, the date (converted from the date string property), and
+# a dictionary of any/all other revprops.
+def _split_revprops(revprops):
+  if not revprops:
+    return None, None, None, {}
+  special_props = []
+  for prop in core.SVN_PROP_REVISION_LOG, \
+              core.SVN_PROP_REVISION_AUTHOR, \
+              core.SVN_PROP_REVISION_DATE:
+    if revprops.has_key(prop):
+      special_props.append(revprops[prop])
+      del(revprops[prop])
+    else:
+      special_props.append(None)
+  msg, author, datestr = tuple(special_props)
+  date = _datestr_to_date(datestr)
+  return msg, author, date, revprops  


 def _datestr_to_date(datestr):
@@ -173,64 +212,11 @@ class NodeHistory:
          return
    self.histories.append([revision, _cleanup_path(path)])
    if self.limit and len(self.histories) == self.limit:
-      raise core.SubversionException("", core.SVN_ERR_CEASE_INVOCATION)
+      raise core.SubversionException("", _SVN_ERR_CEASE_INVOCATION)

  def __getitem__(self, idx):
    return self.histories[idx]

-
-def _get_history(svnrepos, path, rev, path_type, limit=0, options={}):
-  if svnrepos.youngest == 0:
-    return []
-
-  rev_paths = []
-  fsroot = svnrepos._getroot(rev)
-  show_all_logs = options.get('svn_show_all_dir_logs', 0)
-  if not show_all_logs:
-    # See if the path is a file or directory.
-    kind = fs.check_path(fsroot, path)
-    if kind is core.svn_node_file:
-      show_all_logs = 1
-      
-  # Instantiate a NodeHistory collector object, and use it to collect
-  # history items for PATH@REV.
-  history = NodeHistory(svnrepos.fs_ptr, show_all_logs, limit)
-  try:
-    repos.svn_repos_history(svnrepos.fs_ptr, path, history.add_history,
-                            1, rev, options.get('svn_cross_copies', 0))
-  except core.SubversionException, e:
-    _fix_subversion_exception(e)
-    if e.apr_err != core.SVN_ERR_CEASE_INVOCATION:
-      raise
-
-  # Now, iterate over those history items, checking for changes of
-  # location, pruning as necessitated by authz rules.
-  for hist_rev, hist_path in history:
-    path_parts = _path_parts(hist_path)
-    if not vclib.check_path_access(svnrepos, path_parts, path_type, hist_rev):
-      break
-    rev_paths.append([hist_rev, hist_path])
-  return rev_paths
-
-
-def _log_helper(svnrepos, path, rev, lockinfo):
-  rev_root = fs.revision_root(svnrepos.fs_ptr, rev)
-
-  # Was this path@rev the target of a copy?
-  copyfrom_rev, copyfrom_path = fs.copied_from(rev_root, path)
-
-  # Assemble our LogEntry
-  date, author, msg, changes = svnrepos.revinfo(rev)
-  if fs.is_file(rev_root, path):
-    size = fs.file_length(rev_root, path)
-  else:
-    size = None
-  entry = Revision(rev, date, author, msg, size, lockinfo, path,
-                   copyfrom_path and _cleanup_path(copyfrom_path),
-                   copyfrom_rev)
-  return entry
-  
-
 def _get_last_history_rev(fsroot, path):
  history = fs.node_history(fsroot, path)
  history = fs.history_prev(history, 0)
@@ -309,14 +295,15 @@ class FileContentsPipe:


 class BlameSource:
-  def __init__(self, local_url, rev, first_rev):
+  def __init__(self, local_url, rev, first_rev, include_text, config_dir):
    self.idx = -1
    self.first_rev = first_rev
    self.blame_data = []
+    self.include_text = include_text

-    ctx = client.ctx_t()
-    core.svn_config_ensure(None)
-    ctx.config = core.svn_config_get_config(None)
+    ctx = client.svn_client_create_context()
+    core.svn_config_ensure(config_dir)
+    ctx.config = core.svn_config_get_config(config_dir)
    ctx.auth_baton = core.svn_auth_open([])
    try:
      ### TODO: Is this use of FIRST_REV always what we want?  Should we
@@ -333,6 +320,8 @@ class BlameSource:
    prev_rev = None
    if rev > self.first_rev:
      prev_rev = rev - 1
+    if not self.include_text:
+      text = None
    self.blame_data.append(vclib.Annotation(text, line_no + 1, rev,
                                            prev_rev, author, None))

@@ -371,7 +360,7 @@ class LocalSubversionRepository(vclib.Repository):
    self.auth = authorizer
    self.svn_client_path = utilities.svn or 'svn'
    self.diff_cmd = utilities.diff or 'diff'
-    self.config_dir = config_dir
+    self.config_dir = config_dir or None

    # See if this repository is even viewable, authz-wise.
    if not vclib.check_root_access(self):
@@ -401,6 +390,10 @@ class LocalSubversionRepository(vclib.Repository):
    self._fsroots = {}
    self._revinfo_cache = {}

+    # See if a universal read access determination can be made.
+    if self.auth and self.auth.check_universal_access(self.name) == 1:
+      self.auth = None
+
  def rootname(self):
    return self.name

@@ -416,19 +409,14 @@ class LocalSubversionRepository(vclib.Repository):
  def itemtype(self, path_parts, rev):
    rev = self._getrev(rev)
    basepath = self._getpath(path_parts)
-    kind = fs.check_path(self._getroot(rev), basepath)
-    pathtype = None
-    if kind == core.svn_node_dir:
-      pathtype = vclib.DIR
-    elif kind == core.svn_node_file:
-      pathtype = vclib.FILE
-    else:
+    pathtype = self._gettype(basepath, rev)
+    if pathtype is None:
      raise vclib.ItemNotFound(path_parts)
    if not vclib.check_path_access(self, path_parts, pathtype, rev):
      raise vclib.ItemNotFound(path_parts)
    return pathtype

-  def openfile(self, path_parts, rev):
+  def openfile(self, path_parts, rev, options):
    path = self._getpath(path_parts)
    if self.itemtype(path_parts, rev) != vclib.FILE:  # does auth-check
      raise vclib.Error("Path '%s' is not a file." % path)
@@ -467,7 +455,7 @@ class LocalSubversionRepository(vclib.Repository):
        continue
      path = self._getpath(entry_path_parts)
      entry_rev = _get_last_history_rev(fsroot, path)
-      date, author, msg, changes = self.revinfo(entry_rev)
+      date, author, msg, revprops, changes = self._revinfo(entry_rev)
      entry.rev = str(entry_rev)
      entry.date = date
      entry.author = author
@@ -516,20 +504,19 @@ class LocalSubversionRepository(vclib.Repository):
    # 'limit' parameter here as numeric cut-off for the depth of our
    # history search.
    if options.get('svn_latest_log', 0):
-      revision = _log_helper(self, path, rev, lockinfo)
+      revision = self._log_helper(path, rev, lockinfo)
      if revision:
        revision.prev = None
        revs.append(revision)
    else:
-      history = _get_history(self, path, rev, path_type,
-                             first + limit, options)
+      history = self._get_history(path, rev, path_type, first + limit, options)
      if len(history) < first:
        history = []
      if limit:
        history = history[first:first+limit]

      for hist_rev, hist_path in history:
-        revision = _log_helper(self, hist_path, hist_rev, lockinfo)
+        revision = self._log_helper(hist_path, hist_rev, lockinfo)
        if revision:
          # If we have unreadable copyfrom data, obscure it.
          if revision.copy_path is not None:
@@ -550,153 +537,45 @@ class LocalSubversionRepository(vclib.Repository):
    fsroot = self._getroot(rev)
    return fs.node_proplist(fsroot, path)
  
-  def annotate(self, path_parts, rev):
+  def annotate(self, path_parts, rev, include_text=False):
    path = self._getpath(path_parts)
    path_type = self.itemtype(path_parts, rev)  # does auth-check
    if path_type != vclib.FILE:
      raise vclib.Error("Path '%s' is not a file." % path)
    rev = self._getrev(rev)
    fsroot = self._getroot(rev)
-    history = _get_history(self, path, rev, path_type, 0,
-                           {'svn_cross_copies': 1})
+    history = self._get_history(path, rev, path_type, 0,
+                                {'svn_cross_copies': 1})
    youngest_rev, youngest_path = history[0]
    oldest_rev, oldest_path = history[-1]
-    source = BlameSource(_rootpath2url(self.rootpath, path),
-                         youngest_rev, oldest_rev)
+    source = BlameSource(_rootpath2url(self.rootpath, path), youngest_rev,
+                         oldest_rev, include_text, self.config_dir)
    return source, youngest_rev

-  def _revinfo_raw(self, rev):
-    fsroot = self._getroot(rev)
-
-    # Get the changes for the revision
-    editor = repos.ChangeCollector(self.fs_ptr, fsroot)
-    e_ptr, e_baton = delta.make_editor(editor)
-    repos.svn_repos_replay(fsroot, e_ptr, e_baton)
-    changes = editor.get_changes()
-    changedpaths = {}
-    
-    # Now get the revision property info.  Would use
-    # editor.get_root_props(), but something is broken there...
-    revprops = fs.revision_proplist(self.fs_ptr, rev)
-    msg = revprops.get(core.SVN_PROP_REVISION_LOG)
-    author = revprops.get(core.SVN_PROP_REVISION_AUTHOR)
-    datestr = revprops.get(core.SVN_PROP_REVISION_DATE)
-
-    # Copy the Subversion changes into a new hash, converting them into
-    # ChangedPath objects.
-    found_readable = found_unreadable = 0
-    for path in changes.keys():
-      change = changes[path]
-      if change.path:
-        change.path = _cleanup_path(change.path)
-      if change.base_path:
-        change.base_path = _cleanup_path(change.base_path)
-      is_copy = 0
-      if not hasattr(change, 'action'): # new to subversion 1.4.0
-        action = vclib.MODIFIED
-        if not change.path:
-          action = vclib.DELETED
-        elif change.added:
-          action = vclib.ADDED
-          replace_check_path = path
-          if change.base_path and change.base_rev:
-            replace_check_path = change.base_path
-          if changedpaths.has_key(replace_check_path) \
-             and changedpaths[replace_check_path].action == vclib.DELETED:
-            action = vclib.REPLACED
-      else:
-        if change.action == repos.CHANGE_ACTION_ADD:
-          action = vclib.ADDED
-        elif change.action == repos.CHANGE_ACTION_DELETE:
-          action = vclib.DELETED
-        elif change.action == repos.CHANGE_ACTION_REPLACE:
-          action = vclib.REPLACED
-        else:
-          action = vclib.MODIFIED
-      if (action == vclib.ADDED or action == vclib.REPLACED) \
-         and change.base_path \
-         and change.base_rev:
-        is_copy = 1
-      if change.item_kind == core.svn_node_dir:
-        pathtype = vclib.DIR
-      elif change.item_kind == core.svn_node_file:
-        pathtype = vclib.FILE
-      else:
-        pathtype = None
-
-      parts = _path_parts(path)
-      if vclib.check_path_access(self, parts, pathtype, rev):
-        if is_copy and change.base_path and (change.base_path != path):
-          parts = _path_parts(change.base_path)
-          if not vclib.check_path_access(self, parts, pathtype, change.base_rev):
-            is_copy = 0
-            change.base_path = None
-            change.base_rev = None
-        changedpaths[path] = SVNChangedPath(path, rev, pathtype,
-                                            change.base_path,
-                                            change.base_rev, action,
-                                            is_copy, change.text_changed,
-                                            change.prop_changes)
-        found_readable = 1
-      else:
-        found_unreadable = 1
-
-    # Return our tuple, auth-filtered: date, author, msg, changes
-    if found_unreadable:
-      msg = None
-      if not found_readable:
-        author = None
-        datestr = None
-
-    date = _datestr_to_date(datestr)
-    return date, author, msg, changedpaths.values()
-
  def revinfo(self, rev):
-    rev = self._getrev(rev)
-    cached_info = self._revinfo_cache.get(rev)
-    if not cached_info:
-      cached_info = self._revinfo_raw(rev)
-      self._revinfo_cache[rev] = cached_info
-    return cached_info[0], cached_info[1], cached_info[2], cached_info[3]
+    return self._revinfo(rev, 1) 
  
  def rawdiff(self, path_parts1, rev1, path_parts2, rev2, type, options={}):
-    if path_parts1:
-      p1 = self._getpath(path_parts1)
-      r1 = self._getrev(rev1)
-      if not vclib.check_path_access(self, path_parts1, vclib.FILE, rev1):
-        raise vclib.ItemNotFound(path_parts1)
-    else:
-      p1 = None
-
-    if path_parts2:
-      p2 = self._getpath(path_parts2)
-      r2 = self._getrev(rev2)
-      if not vclib.check_path_access(self, path_parts2, vclib.FILE, rev2):
-        raise vclib.ItemNotFound(path_parts2)
-    else:
-      if not p1:
-        raise vclib.ItemNotFound(path_parts2)
-      p2 = None
-
+    p1 = self._getpath(path_parts1)
+    p2 = self._getpath(path_parts2)
+    r1 = self._getrev(rev1)
+    r2 = self._getrev(rev2)
+    if not vclib.check_path_access(self, path_parts1, vclib.FILE, rev1):
+      raise vclib.ItemNotFound(path_parts1)
+    if not vclib.check_path_access(self, path_parts2, vclib.FILE, rev2):
+      raise vclib.ItemNotFound(path_parts2)
+    
    args = vclib._diff_args(type, options)

    def _date_from_rev(rev):
-      date, author, msg, changes = self.revinfo(rev)
+      date, author, msg, revprops, changes = self._revinfo(rev)
      return date

    try:
-      if p1:
-        temp1 = temp_checkout(self, p1, r1)
-        info1 = p1, _date_from_rev(r1), r1
-      else:
-        temp1 = '/dev/null'
-        info1 = '/dev/null', _date_from_rev(rev1), rev1
-      if p2:
-        temp2 = temp_checkout(self, p2, r2)
-        info2 = p2, _date_from_rev(r2), r2
-      else:
-        temp2 = '/dev/null'
-        info2 = '/dev/null', _date_from_rev(rev2), rev2
+      temp1 = temp_checkout(self, p1, r1)
+      temp2 = temp_checkout(self, p2, r2)
+      info1 = p1, _date_from_rev(r1), r1
+      info2 = p2, _date_from_rev(r2), r2
      return vclib._diff_fp(temp1, temp2, info1, info2, self.diff_cmd, args)
    except core.SubversionException, e:
      _fix_subversion_exception(e)
@@ -707,6 +586,253 @@ class LocalSubversionRepository(vclib.Repository):
  def isexecutable(self, path_parts, rev):
    props = self.itemprops(path_parts, rev) # does authz-check
    return props.has_key(core.SVN_PROP_EXECUTABLE)
+
+  def filesize(self, path_parts, rev):
+    path = self._getpath(path_parts)
+    if self.itemtype(path_parts, rev) != vclib.FILE:  # does auth-check
+      raise vclib.Error("Path '%s' is not a file." % path)
+    fsroot = self._getroot(self._getrev(rev))
+    return fs.file_length(fsroot, path)   
+
+  ##--- helpers ---##
+
+  def _revinfo(self, rev, include_changed_paths=0):
+    """Internal-use, cache-friendly revision information harvester."""
+
+    def _get_changed_paths(fsroot):
+      """Return a 3-tuple: found_readable, found_unreadable, changed_paths."""
+      editor = repos.ChangeCollector(self.fs_ptr, fsroot)
+      e_ptr, e_baton = delta.make_editor(editor)
+      repos.svn_repos_replay(fsroot, e_ptr, e_baton)
+      changedpaths = {}
+      changes = editor.get_changes()
+    
+      # Copy the Subversion changes into a new hash, checking
+      # authorization and converting them into ChangedPath objects.
+      found_readable = found_unreadable = 0
+      for path in changes.keys():
+        change = changes[path]
+        if change.path:
+          change.path = _cleanup_path(change.path)
+        if change.base_path:
+          change.base_path = _cleanup_path(change.base_path)
+        is_copy = 0
+        if not hasattr(change, 'action'): # new to subversion 1.4.0
+          action = vclib.MODIFIED
+          if not change.path:
+            action = vclib.DELETED
+          elif change.added:
+            action = vclib.ADDED
+            replace_check_path = path
+            if change.base_path and change.base_rev:
+              replace_check_path = change.base_path
+            if changedpaths.has_key(replace_check_path) \
+               and changedpaths[replace_check_path].action == vclib.DELETED:
+              action = vclib.REPLACED
+        else:
+          if change.action == repos.CHANGE_ACTION_ADD:
+            action = vclib.ADDED
+          elif change.action == repos.CHANGE_ACTION_DELETE:
+            action = vclib.DELETED
+          elif change.action == repos.CHANGE_ACTION_REPLACE:
+            action = vclib.REPLACED
+          else:
+            action = vclib.MODIFIED
+        if (action == vclib.ADDED or action == vclib.REPLACED) \
+           and change.base_path \
+           and change.base_rev:
+          is_copy = 1
+        if change.item_kind == core.svn_node_dir:
+          pathtype = vclib.DIR
+        elif change.item_kind == core.svn_node_file:
+          pathtype = vclib.FILE
+        else:
+          pathtype = None
+  
+        parts = _path_parts(path)
+        if vclib.check_path_access(self, parts, pathtype, rev):
+          if is_copy and change.base_path and (change.base_path != path):
+            parts = _path_parts(change.base_path)
+            if not vclib.check_path_access(self, parts, pathtype,
+                                           change.base_rev):
+              is_copy = 0
+              change.base_path = None
+              change.base_rev = None
+              found_unreadable = 1
+          changedpaths[path] = SVNChangedPath(path, rev, pathtype,
+                                              change.base_path,
+                                              change.base_rev, action,
+                                              is_copy, change.text_changed,
+                                              change.prop_changes)
+          found_readable = 1
+        else:
+          found_unreadable = 1
+      return found_readable, found_unreadable, changedpaths.values()
+
+    def _get_change_copyinfo(fsroot, path, change):
+      # If we know the copyfrom info, return it...
+      if hasattr(change, 'copyfrom_known') and change.copyfrom_known:
+        copyfrom_path = change.copyfrom_path
+        copyfrom_rev = change.copyfrom_rev
+      # ...otherwise, if this change could be a copy (that is, it
+      # contains an add action), query the copyfrom info ...
+      elif (change.change_kind == fs.path_change_add or
+            change.change_kind == fs.path_change_replace):
+        copyfrom_rev, copyfrom_path = fs.copied_from(fsroot, path)
+      # ...else, there's no copyfrom info.
+      else:
+        copyfrom_rev = core.SVN_INVALID_REVNUM
+        copyfrom_path = None
+      return copyfrom_path, copyfrom_rev
+      
+    def _simple_auth_check(fsroot):
+      """Return a 2-tuple: found_readable, found_unreadable."""
+      found_unreadable = found_readable = 0
+      if hasattr(fs, 'paths_changed2'):
+        changes = fs.paths_changed2(fsroot)
+      else:
+        changes = fs.paths_changed(fsroot)
+      paths = changes.keys()
+      for path in paths:
+        change = changes[path]
+        pathtype = None
+        if hasattr(change, 'node_kind'):
+          if change.node_kind == core.svn_node_file:
+            pathtype = vclib.FILE
+          elif change.node_kind == core.svn_node_dir:
+            pathtype = vclib.DIR
+        parts = _path_parts(path)
+        if pathtype is None:
+          # Figure out the pathtype so we can query the authz subsystem.
+          if change.change_kind == fs.path_change_delete:
+            # Deletions are annoying, because they might be underneath
+            # copies (make their previous location non-trivial).
+            prev_parts = parts
+            prev_rev = rev - 1
+            parent_parts = parts[:-1]
+            while parent_parts:
+              parent_path = '/' + self._getpath(parent_parts)
+              parent_change = changes.get(parent_path)
+              if not (parent_change and \
+                      (parent_change.change_kind == fs.path_change_add or
+                       parent_change.change_kind == fs.path_change_replace)):
+                del(parent_parts[-1])
+                continue
+              copyfrom_path, copyfrom_rev = \
+                _get_change_copyinfo(fsroot, parent_path, parent_change)
+              if copyfrom_path:
+                prev_rev = copyfrom_rev
+                prev_parts = _path_parts(copyfrom_path) + \
+                             parts[len(parent_parts):]
+                break
+              del(parent_parts[-1])
+            pathtype = self._gettype(self._getpath(prev_parts), prev_rev)
+          else:
+            pathtype = self._gettype(self._getpath(parts), rev)
+        if vclib.check_path_access(self, parts, pathtype, rev):
+          found_readable = 1
+          copyfrom_path, copyfrom_rev = \
+            _get_change_copyinfo(fsroot, path, change)
+          if copyfrom_path and copyfrom_path != path:
+            parts = _path_parts(copyfrom_path)
+            if not vclib.check_path_access(self, parts, pathtype,
+                                           copyfrom_rev):
+              found_unreadable = 1
+        else:
+          found_unreadable = 1
+        if found_readable and found_unreadable:
+          break
+      return found_readable, found_unreadable
+      
+    def _revinfo_helper(rev, include_changed_paths):
+      # Get the revision property info.  (Would use
+      # editor.get_root_props(), but something is broken there...)
+      revprops = fs.revision_proplist(self.fs_ptr, rev)
+      msg, author, date, revprops = _split_revprops(revprops)
+  
+      # Optimization: If our caller doesn't care about the changed
+      # paths, and we don't need them to do authz determinations, let's
+      # get outta here.
+      if self.auth is None and not include_changed_paths:
+        return date, author, msg, revprops, None
+  
+      # If we get here, then we either need the changed paths because we
+      # were asked for them, or we need them to do authorization checks.
+      #
+      # If we only need them for authorization checks, though, we
+      # won't bother generating fully populated ChangedPath items (the
+      # cost is too great).
+      fsroot = self._getroot(rev)
+      if include_changed_paths:
+        found_readable, found_unreadable, changedpaths = \
+          _get_changed_paths(fsroot)
+      else:
+        changedpaths = None
+        found_readable, found_unreadable = _simple_auth_check(fsroot)
+        
+      # Filter our metadata where necessary, and return the requested data.
+      if found_unreadable:
+        msg = None
+        if not found_readable:
+          author = None
+          date = None
+      return date, author, msg, revprops, changedpaths
+
+    # Consult the revinfo cache first.  If we don't have cached info,
+    # or our caller wants changed paths and we don't have those for
+    # this revision, go do the real work.
+    rev = self._getrev(rev)
+    cached_info = self._revinfo_cache.get(rev)
+    if not cached_info \
+       or (include_changed_paths and cached_info[4] is None):
+      cached_info = _revinfo_helper(rev, include_changed_paths)
+      self._revinfo_cache[rev] = cached_info
+    return tuple(cached_info)
+  
+  def _log_helper(self, path, rev, lockinfo):
+    rev_root = fs.revision_root(self.fs_ptr, rev)
+    copyfrom_rev, copyfrom_path = fs.copied_from(rev_root, path)
+    date, author, msg, revprops, changes = self._revinfo(rev)
+    if fs.is_file(rev_root, path):
+      size = fs.file_length(rev_root, path)
+    else:
+      size = None
+    return Revision(rev, date, author, msg, size, lockinfo, path,
+                    copyfrom_path and _cleanup_path(copyfrom_path),
+                    copyfrom_rev)
+
+  def _get_history(self, path, rev, path_type, limit=0, options={}):
+    if self.youngest == 0:
+      return []
+
+    rev_paths = []
+    fsroot = self._getroot(rev)
+    show_all_logs = options.get('svn_show_all_dir_logs', 0)
+    if not show_all_logs:
+      # See if the path is a file or directory.
+      kind = fs.check_path(fsroot, path)
+      if kind is core.svn_node_file:
+        show_all_logs = 1
+      
+    # Instantiate a NodeHistory collector object, and use it to collect
+    # history items for PATH@REV.
+    history = NodeHistory(self.fs_ptr, show_all_logs, limit)
+    try:
+      repos.svn_repos_history(self.fs_ptr, path, history.add_history,
+                              1, rev, options.get('svn_cross_copies', 0))
+    except core.SubversionException, e:
+      _fix_subversion_exception(e)
+      if e.apr_err != _SVN_ERR_CEASE_INVOCATION:
+        raise
+
+    # Now, iterate over those history items, checking for changes of
+    # location, pruning as necessitated by authz rules.
+    for hist_rev, hist_path in history:
+      path_parts = _path_parts(hist_path)
+      if not vclib.check_path_access(self, path_parts, path_type, hist_rev):
+        break
+      rev_paths.append([hist_rev, hist_path])
+    return rev_paths
  
  def _getpath(self, path_parts):
    return string.join(path_parts, '/')
@@ -715,8 +841,11 @@ class LocalSubversionRepository(vclib.Repository):
    if rev is None or rev == 'HEAD':
      return self.youngest
    try:
+      if type(rev) == type(''):
+        while rev[0] == 'r':
+          rev = rev[1:]
      rev = int(rev)
-    except ValueError:
+    except:
      raise vclib.InvalidRevision(rev)
    if (rev < 0) or (rev > self.youngest):
      raise vclib.InvalidRevision(rev)
@@ -729,7 +858,20 @@ class LocalSubversionRepository(vclib.Repository):
      r = self._fsroots[rev] = fs.revision_root(self.fs_ptr, rev)
      return r

-  ##--- custom --##
+  def _gettype(self, path, rev):
+    # Similar to itemtype(), but without the authz check.  Returns
+    # None for missing paths.
+    try:
+      kind = fs.check_path(self._getroot(rev), path)
+    except:
+      return None
+    if kind == core.svn_node_dir:
+      return vclib.DIR
+    if kind == core.svn_node_file:
+      return vclib.FILE
+    return None
+  
+  ##--- custom ---##

  def get_youngest_revision(self):
    return self.youngest
@@ -807,3 +949,32 @@ class LocalSubversionRepository(vclib.Repository):
        return peg_revision, path
    finally:
      pass
+
+  def get_symlink_target(self, path_parts, rev):
+    """Return the target of the symbolic link versioned at PATH_PARTS
+    in REV, or None if that object is not a symlink."""
+
+    path = self._getpath(path_parts)
+    rev = self._getrev(rev)
+    path_type = self.itemtype(path_parts, rev)  # does auth-check
+    fsroot = self._getroot(rev)
+
+    # Symlinks must be files with the svn:special property set on them
+    # and with file contents which read "link SOME_PATH".
+    if path_type != vclib.FILE:
+      return None
+    props = fs.node_proplist(fsroot, path)
+    if not props.has_key(core.SVN_PROP_SPECIAL):
+      return None
+    pathspec = ''
+    ### FIXME: We're being a touch sloppy here, only checking the first line
+    ### of the file.
+    stream = fs.file_contents(fsroot, path)
+    try:
+      pathspec, eof = core.svn_stream_readline(stream, '\n')
+    finally:
+      core.svn_stream_close(stream)
+    if pathspec[:5] != 'link ':
+      return None
+    return pathspec[5:]
+
--- a/lib/viewvc.py
+++ b/lib/viewvc.py
--- a/lib/win32popen.py
+++ b/lib/win32popen.py
@@ -1,6 +1,6 @@
 # -*-python-*-
 #
-# Copyright (C) 1999-2007 The ViewCVS Group. All Rights Reserved.
+# Copyright (C) 1999-2013 The ViewCVS Group. All Rights Reserved.
 #
 # By using this file, you agree to the terms and conditions set forth in
 # the LICENSE.html file which can be found at the top level of the ViewVC
--- a/notes/TODO
+++ b/notes/TODO
@@ -0,0 +1,53 @@
+PREFACE
+-------
+This file will go away soon after release 0.8.  Please use the SourceForge 
+tracker to resubmit any of the items listed below, if you think, it is 
+still an issue:  
+	http://sourceforge.net/tracker/?group_id=18760
+Before reporting please check, whether someone else has already done this.
+Working patches increase the chance to be included into the next release.
+                                                     -- PeFu / October 2001
+
+TODO ITEMS
+----------
+*) add Tamminen Eero's comments on how to make Linux directly execute
+   the Python script. From email on Feb 19.
+   [ add other examples, such as my /bin/sh hack or the teeny CGI stub 
+     importing the bulk hack ]
+
+*) insert rcs_path into PATH before calling "rcsdiff". rcsdiff might
+   use "co" and needs to find it on the path.
+
+*) show the "locked" flag (attach it to the LogEntry objects).
+   Idea from Russell Gordon <russell@hoopscotch.dhs.org>
+
+*) committing with a specific revision number:
+   http://mailman.lyra.org/pipermail/viewcvs/2000q1/000008.html
+
+*) add capability similar to cvs2cl.pl:
+     http://mailman.lyra.org/pipermail/viewcvs/2000q2/000050.html
+   suggestion from Chris Meyer <cmeyer@gatan.com>.
+
+*) add a tree view of the directory structure (and files?)
+
+*) include a ConfigParser.py to help older Python installations
+
+*) add a check for the rcs programs/paths to viewvc-install. clarify the
+   dependency on RCS in the docs.
+
+*) have a "check" mode that verifies binaries are available on rcs_path
+
+   -> alternately (probably?): use rcsparse rather than external tools
+
+KNOWN BUGS
+----------
+*) time.timezone seems to not be available on some 1.5.2 installs.
+   I was unable to verify this.  On RedHat and SuSE Linux this bug
+   is non existant.
+
+*) With old repositories containing many branches, tags or thousands
+   or revisions, the cvsgraph feature becomes unusable (see INSTALL).
+   ViewVC can't do much about this, but it might be possible to
+   investigate the number of branches, tags and revision in advance
+   and disable the cvsgraph links, if the numbers exceed a certain
+   treshold.
--- a/notes/authz-dev-TODO
+++ b/notes/authz-dev-TODO
@@ -0,0 +1,82 @@
+Here lie TODO items for the pluggable authz system:
+
+  * Subversion uses path privelege to determine visibility of revision
+    metadata.  That logic is pretty Subversion-specific, so it feels like it
+    belongs outside the vcauth library as just a helper function in viewvc.py
+    or something.  The algorithm is something like this (culled from the
+    CollabNet implementation, and not expected to work as edited):
+
+      # Subversion revision access levels
+      REVISION_ACCESS_NONE = 0
+      REVISION_ACCESS_PARTIAL = 1
+      REVISION_ACCESS_FULL = 2
+      
+      def check_svn_revision_access(request, rev):
+        # Check our revision access cache first.
+        if request.rev_access_cache.has_key(rev):
+          return request.rev_access_cache[rev]
+
+        # Check our cached answer to the question "Does the user have 
+        # an all-access or a not-at-all-access pass?"
+        if request.full_access is not None:
+          return request.full_access \
+                   and REVISION_ACCESS_FULL or REVISION_ACCESS_NONE
+ 
+        # Get a list of paths changed in REV.
+        ### FIXME:  There outta be a vclib-complaint way to do this,
+        ### as this won't work for vclib.svn_ra.
+        import svn.fs
+        rev_root = svn.fs.revision_root(self.repos.fs_ptr, rev)
+        changes = svn.fs.paths_changed(rev_root)
+      
+        # Loop over the list of changed paths, asking the access question
+        # for each one.  We'll track whether we've found any readable paths
+        # as well as any un-readable (non-authorized) paths, and quit
+        # checking as soon as we know our revision access level.
+        found_readable = 0
+        found_unreadable = 0
+        for path in changes.keys():
+          parts = _path_parts(path)
+          kind = request.repos.itemtype(parts, rev)
+          if kind == vclib.DIR:
+            access = request.auth.check_dir_access(parts, rev)
+          elif:
+            access = request.auth.check_file_access(parts, rev)
+          if access:
+            found_readable = 1
+          else:
+            found_unreadable = 1
+          # Optimization: if we've found at least one readable, and one
+          # unreadable, we needn't ask about any more paths.
+          if found_readable and found_unreadable:
+            break
+          
+        # If there are paths but we can't read any of them, no access is
+        # granted.
+        if len(changes) and not found_readable:
+          request.rev_access_cache[rev] = REVISION_ACCESS_NONE
+        # If we found at least one unreadable path, partial access is
+        # granted.
+        elif found_unreadable:
+          request.rev_access_cache[rev] = REVISION_ACCESS_PARTIAL
+        # Finally, if there were no paths at all, or none of the existing
+        # ones were unreadable, grant full access.
+        else:
+          request.rev_access_cache[rev] = REVISION_ACCESS_FULL
+        return request.rev_access_cache[rev]
+
+    The problems are:  where does one hang the revision access cache
+    so that it doesn't survive a given request?  On the request, as
+    shown in the edited code above?
+
+    Can we actually get a good interface into the vcauth layer for
+    asking the all-access / no-access question?  Obviously each vcauth
+    provider can cache that value for itself and use it as it deems
+    necessary, but ideally the revision access level question will
+    want to know if said auth provider was able to answer the question
+    and, if so, what the answer was.
+
+    Another off-the-wall idea -- let's just teach Subversion to do
+    this calculation as part of a libsvn_repos API.
+
+ 
--- a/notes/logo/viewvc-logo.odg
+++ b/notes/logo/viewvc-logo.odg
--- a/notes/logo/viewvc-logo.pdf
+++ b/notes/logo/viewvc-logo.pdf
--- a/notes/logo/viewvc-logo.svg
+++ b/notes/logo/viewvc-logo.svg
--- a/notes/releases.txt
+++ b/notes/releases.txt
@@ -0,0 +1,125 @@
+                        RELEASE MANAGEMENT
+
+ViewVC rolls releases from release branches associate with each minor
+version of the software.  For example, the 1.1.0 is rolled from the
+1.1.x branch.  The same is true for the 1.1.1, 1.1.2, ... releases.
+
+There is a script, `tools/make-release', which creates a release
+directory and the various archive files that we distribute.  All other
+steps required to get a ViewVC release out of the door require manual
+execution (currently by C. Michael Pilato).  Those steps are as
+follows:
+
+Checkout a working copy of the release branch for the release you
+intend to roll, and in that working copy, perform the following steps
+(X, Y, and Z below represent integral major, minor, and patch version
+numbers, and not literal):
+
+1.  Review any open bug reports:
+
+       http://viewvc.tigris.org/servlets/ProjectIssues
+
+2.  Add a new subsection to the file 'docs/upgrading.html' describing
+    all user visible changes for users of previous releases of ViewVC.
+    Commit any modifications.  NOTE: This step should not be necessary
+    for patch releases.
+
+3.  Verify that copyright years are correct in both the license-1.html
+    file and the source code.
+
+4.  Update and commit the 'CHANGES' file, using any available crystal
+    balls or other forward-looking devices to take a stab at the
+    release date.
+
+5.  Test, test, test!  There is no automatic testsuite available.  So
+    just run with permuting different `viewvc.conf' settings... and
+    pray.  Fix what needs fixin', keeping the CHANGES file in sync
+    with the branch.
+
+6.  At this point, the source code committed to the release branch
+    should exactly reflect what you wish to distribute and dub "the
+    release".
+
+7.  Update your release branch working copy to HEAD.
+
+       svn up
+
+8.  Edit the file 'lib/viewvc.py' and remove the "-dev" suffix from
+    __version__.  The remainder should be of the form "X.Y.Z", where X,
+    Y, and Z are positive integers.
+
+    *** Do NOT commit this change. ***
+
+9.  "Peg" the contributed templates externals definition to the
+    current HEAD revision:
+
+       svn pedit svn:externals .
+ 
+    (squeeze "-rBASE_REV", where BASE_REV is the current HEAD revision
+    number, between 'templates-contrib' and the target URL).
+
+    *** Do NOT commit this change. ***
+
+10. Tag the release:
+
+       svn cp -m "Tag the X.Y.Z final release." . ^/tags/X.Y.Z
+
+    This will create a copy of the release branch, plus your local
+    modifications to the svn:externals property and lib/viewvc.py
+    file, to the tag location.
+
+11. Revert the changes in your working copy.
+
+       svn revert -R .
+
+12. Go into an empty directory and run the 'make-release' script:
+
+       tools/make-release viewvc-X.Y.Z tags/X.Y.Z
+
+13. Verify the archive files:
+
+       - do they have a LICENSE.html file?
+       - do they have necessary include documentation?
+       - do they *not* have unnecessary stuff?
+       - do they install and work correctly?
+
+14. Upload the created archive files (tar.gz and zip) into the Files
+    and Documents section of the Tigris.org project, and modify the
+    CHECKSUMS document there accordingly:
+
+       http://viewvc.tigris.org/servlets/ProjectDocumentList?folderID=6004
+
+    Also, drop a copy of the archive files into the root directory of
+    the viewvc.org website (unversioned).
+
+15. Update the Tigris.org website (^/trunk/www/index.html) to refer to
+    the new release files and commit.
+
+       svn ci -m "Bump latest advertised release."
+
+16. Back on the release branch, edit the file 'lib/viewvc.py' again,
+    incrementing the patch number assigned to the __version__
+    variable.  Add a new empty block in the branch's CHANGES file.
+    Commit your changes:
+
+       svn ci -m "Begin a new release cycle."
+
+17. Edit the Issue Tracker configuration options, adding a new Version
+    for the just-released one, and a new Milestone for the next patch
+    (and possibly, minor or major) release.  (For the Milestone sort
+    key, use a packed integer XXYYZZ:  1.0.3 == 10003, 2.11.4 == 21104.)
+
+       http://viewvc.tigris.org/issues/editversions.cgi?component=viewvc&action=add
+       http://viewvc.tigris.org/issues/editmilestones.cgi?component=viewvc&action=add
+
+18. Send to the announce@ list a message explaining all the cool new
+    features.
+
+       http://viewvc.tigris.org/ds/viewForumSummary.do?dsForumId=4253
+
+19. Post a new release notification at Freecode.
+
+       https://freecode.com/projects/viewvc/releases/new
+
+20. Merge CHANGES for this release into the CHANGES file for newer
+    release lines and commit.
--- a/notes/vclib-enhancements.txt
+++ b/notes/vclib-enhancements.txt
@@ -0,0 +1,78 @@
+The following is an email from a developer who was integrating bzr
+into ViewVC in which he shares some thoughts on how to further
+abstract the version control system interactions into first-class APIs
+in the vclib module.
+
+   Subject: Re: [ViewCVS-dev] difflib module
+   Date: Wed, 1 Jun 2005 16:59:10 -0800
+   From: "Johan Rydberg" <jrydberg@gnu.org>
+   To: "Michael Pilato" <cmpilato@collab.net>
+   Cc: <viewcvs-dev@lyra.org>
+   
+   "C. Michael Pilato" <cmpilato@collab.net> writes:
+   
+   >> I've tried to minimize the changes to the viewcvs.py, but of course
+   >> there are a few places where some things has to be altered.
+   >
+   > Well, if along the way, you have ideas about how to further abstract
+   > stuff into the vclib/ modules, please post.
+   
+   I came up with a few as off now;
+   
+    * Generalize revision counting; svn starts from 0, bzr starts from 1.
+      Can be done by a constant; request.repos.MIN_REVNO.  For CVS I'm not
+      sure exactly what should be done.  Right now this is only used in
+      view_revision_svn, so it is not a problem in the short term.
+   
+    * Generalize view_diff;
+   
+      * Have a repo-method diff(file1, rev1, file2, rev2, args) that returns
+        (date1, date2, fp).  Means human_readbale_diff and raw_diff does not
+        have to parse dates.  Good for VCS that does not have the date in
+        the diff.   [### DONE ###]
+   
+      * I'm not sure you should require GNU diff.  Some VCS may use own
+        diff mechanisms (bzr uses difflib, _or_ GNU diff when needed.
+        Monotone uses its own, IIRC.)
+   
+    * Generalize view_revision ;
+   
+      * Have a method, revision_info(), which returns (date, author, msg,
+        changes) much like vclib.svn.get_revision_info.  The CVS version
+        can raise a ViewCVSException.  [### DONE ###]
+   
+      * Establish a convention for renamed/copied files; current should
+        work good enough (change.base_path, change.base_rev) but action
+        string must be same for both svn and others.
+   
+    * request.repos.rev (or .revision) should give the current revision
+      number of the repo.  No need for this (from view_directory):
+   
+       if request.roottype == 'svn': 
+         revision = str(vclib.svn.created_rev(request.repos, request.where))
+   
+      If svn needs the full name of the repo, why not give it when the
+      repo is created?
+   
+    * request.repos.youngest vs vclib.svn.get_youngest_revision(REPO)
+   
+    * More object oriented;
+   
+      * The subversion backend is not really object oriented.  viewcfg.py uses
+        a lot function from vclib.svn, which could instead be methods of the
+        Repository class.  Example:
+   
+          diffobj = vclib.svn.do_diff(request.repos, p1, int(rev1),
+                                      p2, int(rev2), args)  
+   
+        This should be a method of the repository;
+   
+          diffobj = request.repos.do_diff(p1, rev1, ...)
+   
+        I have identified the following functions;
+   
+         - vclib.svn.created_rev
+         - vclib.svn.get_youngest_revision
+         - vclib.svn.date_from_rev
+         - vclib.svn.do_diff
+         - vclib.svn.get_revision_info
--- a/templates-contrib/README
+++ b/templates-contrib/README
@@ -1,5 +0,0 @@
-This directory contains ViewVC template sets contributed by their
-respective authors and expected to work against ViewVC 1.0.x.  They
-are not necessarily supported by the ViewVC development community, and
-do not necessarily carry the same license or copyright as ViewVC
-itself.
--- a/templates-contrib/newvc/README
+++ b/templates-contrib/newvc/README
@@ -1,7 +0,0 @@
-Template Set:  newvc
-Author(s):  C. Michael Pilato <cmpilato@red-bean.com>
-Compatibility:  ViewVC 1.1
-
-The "newvc" template set uses top navigation tabs to flip between
-various views of a file or directory, and aims for a clean-yet-modern
-look and feel.
--- a/templates-contrib/newvc/templates/diff.ezt
+++ b/templates-contrib/newvc/templates/diff.ezt
@@ -1,128 +0,0 @@
-[# Setup page definitions]
-  [define page_title]Diff of:[end]
-  [define help_href][docroot]/help_rootview.html[end]
-[# end]
-[include "include/header.ezt" "diff"]
-
-<form method="get" action="[diff_format_action]" style="display: inline;">
-  <div>
-    [for diff_format_hidden_values]<input type="hidden" name="[diff_format_hidden_values.name]" value="[diff_format_hidden_values.value]"/>[end]
-    <select name="diff_format" onchange="submit()">
-      <option value="h" [is diff_format "h"]selected="selected"[end]>Colored Diff</option>
-      <option value="l" [is diff_format "l"]selected="selected"[end]>Long Colored Diff</option>
-      <option value="f" [is diff_format "f"]selected="selected"[end]>Full Colored Diff</option>
-      <option value="u" [is diff_format "u"]selected="selected"[end]>Unidiff</option>
-      <option value="c" [is diff_format "c"]selected="selected"[end]>Context Diff</option>
-      <option value="s" [is diff_format "s"]selected="selected"[end]>Side by Side</option>
-    </select>
-    <input type="submit" value="Show" />
-    (<a href="[patch_href]">Generate patch</a>)
-  </div>
-</form>
-
-<div id="vc_main_body">
-<!-- ************************************************************** -->
-
-[if-any raw_diff]
-<pre>[raw_diff]</pre>
-[else]
-
-[define change_right][end]
-[define last_change_type][end]
-
-[# these should live in stylesheets]
-
-<table cellpadding="0" cellspacing="0" style="width: 100%;">
-[for changes]
- [is changes.type "change"][else]
-  [if-any change_right][change_right][define change_right][end][end]
- [end]
- [is changes.type "header"]
-  <tr>
-    <th class="vc_header" style="width:6%;"><strong>#</strong></th>
-    <th colspan="2" class="vc_header">
-      <strong>Line [changes.line_info_left]</strong> |
-      <strong>Line [changes.line_info_right]</strong> 
-    </th>
-  </tr>
- [else]
- [is changes.type "add"]
-  <tr>
-    <td id="l[changes.line_number]">[if-any right.annotate_href]<a href="[right.annotate_href]#l[changes.line_number]">[changes.line_number]</a>[else][changes.line_number][end]</td>
-    <td class="vc_diff_plusminus"><strong style="color: green;">+</strong></td>
-    <td class="vc_diff_add">[changes.right]</td>
-  </tr>
- [else]
- [is changes.type "remove"]
-  <tr>
-    <td style="text-decoration: line-through">[changes.line_number]</td>
-    <td class="vc_diff_plusminus"><strong style="color: red;">&ndash;</strong></td>
-    <td class="vc_diff_remove">[changes.left]</td>
-  </tr>
- [else]
- [is changes.type "change"]
-  [if-any changes.have_left]
-  <tr>
-    <td style="text-decoration: line-through">[changes.line_number]</td>
-    <td class="vc_diff_plusminus"><strong style="color: yellow;">&lt;</strong></td>
-    <td class="vc_diff_changes1">[changes.left]</td>
-  </tr>
-  [end]
-  [define change_right][change_right]
-  [if-any changes.have_right]
-  <tr>
-    <td id="l[changes.line_number]">[if-any right.annotate_href]<a href="[right.annotate_href]#l[changes.line_number]">[changes.line_number]</a>[else][changes.line_number][end]</td>
-    <td class="vc_diff_plusminus"><strong style="color: yellow;">&gt;</strong></td>
-    <td class="vc_diff_changes2">[changes.right]</td>
-  </tr>[end]
-  [end]
- [else]
- [is changes.type "no-changes"]
-  <tr><td colspan="3" style="vc_diff_nochange"><strong>- No changes -</strong></td></tr>
- [else]
- [is changes.type "binary-diff"]
-  <tr><td colspan="3" class="vc_diff_binary"><strong>- Binary file revisions differ -</strong></td></tr>
- [else]
- [is changes.type "error"]
-  <tr><td colspan="3" class="vc_diff_error"><strong>- ViewVC depends on rcsdiff and GNU diff
-      to create this page.  ViewVC cannot find GNU diff. Even if you
-      have GNU diff installed, the rcsdiff program must be configured
-      and compiled with the GNU diff location.  -</strong></td></tr>
- [else][# a line of context]
-  <tr>
-    <td id="l[changes.line_number]">[if-any right.annotate_href]<a href="[right.annotate_href]#l[changes.line_number]">[changes.line_number]</a>[else][changes.line_number][end]</td>
-    <td class="vc_diff_plusminus">&nbsp;</td>
-    <td style="font-family: monospace; white-space: pre;">[changes.right]</td>
-  </tr>
- [end][end][end][end][end][end][end]
- [define last_change_type][changes.type][end]
-[end]
-[if-any change_right][change_right][end]
-</table>
-
-<h3>Diff Legend</h3>
-<table class="auto" cellspacing="0">
-<tr>
- <td class="vc_diff_plusminus"><strong style="color: red;">&ndash;</strong></td>
- <td class="vc_diff_remove">Removed lines</td>
-</tr>
-<tr>
- <td class="vc_diff_plusminus"><strong style="color: green;">+</strong></td>
- <td class="vc_diff_add">Added lines</td>
-</tr>
-<tr>
- <td class="vc_diff_plusminus"><strong style="color: yellow;">&lt;</strong></td>
- <td class="vc_diff_changes1">Changed lines</td>
-</tr>
-<tr>
- <td class="vc_diff_plusminus"><strong style="color: yellow;">&gt;</strong></td>
- <td class="vc_diff_changes2">Changed lines</td>
-</tr>
-</table>
-
-[end]
-
-<!-- ************************************************************** -->
-</div>
-
-[include "include/footer.ezt"]
--- a/templates-contrib/newvc/templates/directory.ezt
+++ b/templates-contrib/newvc/templates/directory.ezt
@@ -1,139 +0,0 @@
-[# setup page definitions]
-  [define page_title]Index of:[end]
-  [define help_href][docroot]/help_[if-any where]dir[else]root[end]view.html[end]
-[# end]
-[include "include/header.ezt" "directory"]
-
-[if-any where][else]
-  <!-- you may insert repository access instructions here -->
-[end]
-
-<table class="auto">
-[is cfg.options.use_pagesize "0"][else][is picklist_len "1"][else]
-<tr>
-  <td>Jump to page:</td>
-  <td><form method="get" action="[dir_paging_action]">
-        [for dir_paging_hidden_values]<input type="hidden" name="[dir_paging_hidden_values.name]" value="[dir_paging_hidden_values.value]"/>[end]
-        <select name="dir_pagestart"  onchange="submit()">
-          [for picklist]
-            <option [is picklist.count dir_pagestart]selected[end] value="[picklist.count]">Page [picklist.page]: [picklist.start] to [picklist.end]</option>
-          [end]
-        </select>
-        <input type="submit" value="Go" />
-      </form>
-  </td>
-</tr>
-[end][end]
-</table>
-
-<div id="vc_main_body">
-<!-- ************************************************************** -->
-
-<div id="vc_togglables">
-[is roottype "svn"]
-[if-any rev]r<a href="[revision_href]">[rev]</a>[end]
-[else]
-[is num_dead "0"]
-[else]
-  [if-any attic_showing]
-    <a href="[hide_attic_href]">Hide
-  [else]
-    <a href="[show_attic_href]">Show
-  [end]
-  dead files</a>
-[end]
-[end]
-</div>
-
-<table cellspacing="2" class="fixed" id="dirlist">
-<thead>
-<tr>
-  <th style="width: 200px" class="vc_header[is sortby "file"]_sort[end]">
-    <a href="[sortby_file_href]#dirlist">File
-    [is sortby "file"]
-      <img class="vc_sortarrow" alt="[is sortdir "down"](rev)[end]"
-        width="13" height="13"
-        src="[docroot]/images/[is sortdir "up"]up[else]down[end].png" />
-    [end]
-    </a>
-  </th>
-  <th class="vc_header[is sortby "rev"]_sort[end]">
-    <a href="[sortby_rev_href]#dirlist">Last Change
-    [is sortby "rev"]
-      <img class="vc_sortarrow" alt="[is sortdir "down"](rev)[end]"
-        width="13" height="13"
-        src="[docroot]/images/[is sortdir "up"]up[else]down[end].png" />
-    [end]
-    </a>
-  </th>
-</tr>
-</thead>
-
-<tbody>
-[if-any up_href]
-  <tr class="vc_row_odd">
-    <td colspan="2">
-      <a href="[up_href]">
-      <img src="[docroot]/images/back_small.png" alt="" class="vc_icon"
-      />&nbsp;../</a>
-    </td>
-  </tr>
- [end]
-[for entries]
-  [define click_href][is entries.pathtype "dir"][entries.view_href][else][if-any entries.prefer_markup][entries.view_href][else][entries.download_href][end][end][end]
-
-  <tr class="vc_row_[if-index entries even]even[else]odd[end]">
-    <td style="width: 200px" onclick="jumpTo('[click_href]')">
-    <a name="[entries.anchor]" href="[click_href]" title="[is entries.pathtype "dir"]View Directory Contents[else][if-any entries.prefer_markup]View[else]Download[end] File Contents[end]">
-       <img src="[docroot]/images/[is entries.pathtype "dir"]dir[else][is entries.state "dead"]broken[else]text[end][end].png" alt="" class="vc_icon" />
-       [entries.name][is entries.pathtype "dir"]/[end]</a>
-       [is entries.state "dead"](dead)[end]
-    </td>
-
-    <td [if-any entries.log_href]onclick="jumpTo('[entries.log_href]')"[end]>
-      [if-any entries.rev]
-        <strong>[if-any entries.log_href]<a href="[entries.log_href]" title="Revision [entries.rev]">[entries.rev]</a>[else][entries.rev][end]</strong>
-        ([entries.ago] ago)
-        by <em>[entries.author]</em>:
-        [entries.log]
-        [is entries.pathtype "dir"][is roottype "cvs"]
-          <em>(from [entries.log_file]/[entries.log_rev])</em>
-        [end][end]
-      [end]
-    </td>
-  </tr>
-[end]
-</tbody>
-
-</table>
-
-<div id="vc_view_summary">
-[if-any search_re_form]
-<form class="inline" method="get" action="[search_re_action]">
-  <div class="inline">
-    [for search_re_hidden_values]<input type="hidden" name="[search_re_hidden_values.name]" value="[search_re_hidden_values.value]"/>[end]
-    <input type="text" name="search" value="[search_re]" />
-    <input type="submit" value="Search Files" />
-  </div>
-</form>
-[if-any search_re]
-<form class="inline" method="get" action="[search_re_action]">
-  <div class="inline">
-    [for search_re_hidden_values]<input type="hidden" name="[search_re_hidden_values.name]" value="[search_re_hidden_values.value]"/>[end]
-    <input type="submit" value="Show all files" />
-  </div>
-</form>
-[end]
-&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
-[end]
-[include "include/pathrev_form.ezt"]
-&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
-[files_shown] file[is files_shown "1"][else]s[end] shown
-</div>
-
-[include "include/props.ezt"]
-
-<!-- ************************************************************** -->
-</div>
-
-[include "include/footer.ezt"]
--- a/templates-contrib/newvc/templates/docroot/help.css
+++ b/templates-contrib/newvc/templates/docroot/help.css
@@ -1,8 +0,0 @@
-/************************************/
-/***  ViewVC Help CSS Stylesheet ***/
-/************************************/
-body { margin: 0.5em; }
-img { border: none; }
-table { width: 100%; }
-td { vertical-align: top; }
-col.menu { width:12em; }
--- a/templates-contrib/newvc/templates/docroot/help_dirview.html
+++ b/templates-contrib/newvc/templates/docroot/help_dirview.html
@@ -1,126 +0,0 @@
-<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" 
-    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
-<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
-<head>
-  <title>ViewVC Help: Directory View</title>
-  <link rel="stylesheet" href="help.css" type="text/css" />
-</head>
-<body>
-  <table>
-    <col class="menu" />
-    <col />
-    <tr>
-      <td colspan="2">
-	<h1>ViewVC Help: Directory View</h1>
-      </td>
-    </tr>
-    <tr><td>
-       <h3>Help</h3>
-       <a href="help_rootview.html">General</a><br />
-       <strong>Directory&nbsp;View</strong><br />
-       <a href="help_log.html">Log&nbsp;View</a><br />
-
-       <h3>Internet</h3>
-       <a href="http://viewvc.org/index.html">Home</a><br />
-       <a href="http://viewvc.org/upgrading.html">Upgrading</a><br />
-       <a href="http://viewvc.org/contributing.html">Contributing</a><br />
-       <a href="http://viewvc.org/license-1.html">License</a><br />
-    </td><td colspan="2">
-
-    <p>The directory listing view should be a familiar sight to any
-    computer user. It shows the path of the current directory being viewed
-    at the top of the page. Below that is a table summarizing the
-    directory contents, and then comes actual contents, a sortable list of
-    all files and subdirectories inside the current directory.</p>
-
-    <p><a name="summary"></a>The summary table is made up of some or all
-    of the following rows:</p>
-    <ul>
-      <li><a name="summary-files-shown"><strong>Files Shown</strong></a>
-      - Number of files shown in the directory listing. This might be less
-      than the actual number of files in the directory if a
-      <a href="#option-search">regular expression search</a> is in place,
-      hiding files which don't meet the search criteria. In CVS directory
-      listings, this row will also have a link to toggle display of
-      <a href="help_rootview.html#dead-files">dead files</a>, if any are
-      present.</li>
-
-      <li><a name="summary-revision"><strong>Directory
-      Revision</strong></a> - For Subversion directories only.
-      Shown as "# of #" where the first number is the most recent
-      repository revision where the directory (or a path underneath it)
-      was modified. The second number is just the latest repository
-      revision. Both numbers are links to
-      <a href="help_rootview.html#view-rev">revision views</a></li>
-
-     <li><a name="summary-sticky-revision-tag"><strong>Sticky
-     Revision/Tag</strong></a> - shows the current
-     <a href="help_rootview.html#sticky-revision-tag">sticky revision or
-     tag</a> and contains form fields to set or clear it.</li>
-
-     <li><a name="summary-search"><strong>Current Search</strong></a> -
-     If a <a href="#option-search">regular expression search</a> is in place,
-     shows the search string.</li>
-
-     <li><a name="summary-query"><strong>Query</strong></a> - Provides
-     a link to a <a href="help_rootview.html#view-query">query form</a>
-     for the directory</li>
-   </ul>
-
-  <p><a name="list"></a>The actual directory list is a table with
-  filenames and directory names in one column and information about the
-  most recent revisions where each file or directory was modified in the
-  other columns. Column headers can be clicked to sort the directory
-  entries in order by a column, and clicked again to reverse the sort
-  order.</p>
-
-  <p>
-  <!-- If using directory.ezt template -->
-  File names are links to <a href="help_log.html">log views</a>
-  showing a list of revisions where a file was modified. Revision
-  numbers are links to either
-  <a href="help_rootview.html#view-markup">view</a>
-  or <a href="help_rootview.html#view-checkout">download</a> a file
-  (depending on its file type). The links are reversed for directories.
-  Directory revision numbers are links to <a href="help_log.html">log
-  views</a>, while directory names are links showing the contents of those
-  directories.
-
-  <!-- If using dir_alt.ezt template -->
-  <!--
-  File and directory names are links to retrieve their contents.
-  File links may be either
-  <a href="help_rootview.html#view-markup">view</a>
-  or <a href="help_rootview.html#view-download">download</a> links
-  depending on the file type. Directory links go to directory
-  listings. Revision numbers are links to <a href="help_log.html">log
-  views</a> showing lists of revisions where a file or directory was
-  modified.
-  -->
-
-  Also, in CVS repositories with the <a
-  href="help_rootview.html#view-graph">graph view</a> enabled, there
-  will be small icons next to file names which are links to revision
-  graphs.</p>
-
-  <p>Depending on how ViewVC is configured, there may be more options
-  at the bottom of directory pages:</p>
-
-  <ul>
-    <li><a name="option-search"><strong>Regular expression
-    search</strong></a> - If enabled, will show a form field accepting
-    a search string (a
-    <a href="http://doc.python.org/lib/re-syntax.html">python regular
-    expression</a>). Once submitted, only files that have at least
-    one occurance of the expression will show up in directory listings.
-    </li>
-    <li><a name="option-tarball"><strong>Tarball download</strong></a> -
-    If enabled, will show a link to download a gzipped tar archive of
-    the directory contents.</li>
-  </ul>
-
-  </td></tr></table>
-  <hr />
-  <address><a href="mailto:users@viewvc.tigris.org">ViewVC Users Mailinglist</a></address>
-  </body>
-</html>
--- a/templates-contrib/newvc/templates/docroot/help_log.html
+++ b/templates-contrib/newvc/templates/docroot/help_log.html
@@ -1,71 +0,0 @@
-<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" 
-    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
-<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
-<head>
-  <title>ViewVC Help: Log View</title>
-  <link rel="stylesheet" href="help.css" type="text/css" />
-  <meta http-equiv="Content-Type" content="text/html;charset=utf-8" />
-</head>
-<body>
-  <table>
-    <col class="menu" />
-    <col />
-    <tr>
-      <td colspan="2">
-	<h1>ViewVC Help: Log View</h1>
-      </td>
-    </tr>
-    <tr><td>
-       <h3>Help</h3>
-       <a href="help_rootview.html">General</a><br />
-       <a href="help_dirview.html">Directory&nbsp;View</a><br />
-       <strong>Log&nbsp;View</strong><br />
-
-       <h3>Internet</h3>
-       <a href="http://viewvc.org/index.html">Home</a><br />
-       <a href="http://viewvc.org/upgrading.html">Upgrading</a><br />
-       <a href="http://viewvc.org/contributing.html">Contributing</a><br />
-       <a href="http://viewvc.org/license-1.html">License</a><br />
-    </td><td colspan="2">
-    <p>
-      The log view displays the revision history of the selected source
-      file or directory. For each revision the following information is
-      displayed:
-
-      <ul>
-        <li>The revision number. In Subversion repositories, this is a
-            link to the <a href="help_rootview.html#view-rev">revision
-            view</a></li>
-        <li>For files, links to
-        <a href="help_rootview.html#view-markup">view</a>,
-        <a href="help_rootview.html#view-checkout">download</a>, and
-        <a href="help_rootview.html#view-annotate">annotate</a> the
-          revision. For directories, a link to
-        <a href="help_dirview.html">list directory contents</a></li>
-        <li>A link to select the revision for diffs (see below)</li>
-        <li>The date and age of the change</li>
-        <li>The author of the modification</li>
-        <li>The CVS branch (usually <em>MAIN</em>, if not on a branch)</li>
-        <li>Possibly a list of CVS tags bound to the revision (if any)</li>
-        <li>The size of the change measured in added and removed lines of
-            code. (CVS only)</li>
-        <li>The size of the file in bytes at the time of the revision
-            (Subversion only)</li>
-        <li>Links to view diffs to the previous revision or possibly to
-            an arbitrary selected revision (if any, see above)</li>
-        <li>If the revision is the result of a copy, the path and revision
-            copied from</li>
-        <li>If the revision precedes a copy or rename, the path at the
-            time of the revision</li>
-        <li>And last but not least, the commit log message which should tell
-            about the reason for the change.</li>
-      </ul>
-    <p>
-      At the bottom of the page you will find a form which allows
-      to request diffs between arbitrary revisions.
-    </p>
-  </td></tr></table>
-  <hr />
-  <address><a href="mailto:users@viewvc.tigris.org">ViewVC Users Mailinglist</a></address>
-  </body>
-</html>
--- a/templates-contrib/newvc/templates/docroot/help_query.html
+++ b/templates-contrib/newvc/templates/docroot/help_query.html
@@ -1,66 +0,0 @@
-<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" 
-    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
-<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
-<head>
-  <title>ViewVC Help: Query The Commit Database</title>
-  <link rel="stylesheet" href="help.css" type="text/css" />
-</head>
-<body>
-<table>
-  <col class="menu" />
-  <col />
-  <tr>
-    <td colspan="2">
-      <h1>ViewVC Help: Query The Commit Database</h1>
-    </td>
-  </tr>
-  <tr><td>
-     <h3>Other&nbsp;Help:</h3>
-     <a href="help_rootview.html">General</a><br />
-     <a href="help_dirview.html">Directory&nbsp;View</a><br />
-     <a href="help_log.html">Classic&nbsp;Log&nbsp;View</a><br />
-     <a href="help_logtable.html">Alternative&nbsp;Log&nbsp;View</a><br />
-     <strong>Query&nbsp;Database</strong>
-
-     <h3>Internet</h3>
-     <a href="http://viewvc.org/index.html">Home</a><br />
-     <a href="http://viewvc.org/upgrading.html">Upgrading</a><br />
-     <a href="http://viewvc.org/contributing.html">Contributing</a><br />
-     <a href="http://viewvc.org/license-1.html">License</a><br />
-  </td><td colspan="2">
-
-  <p> 
-    Select your parameters for querying the CVS commit database in the
-    form at the top of the page.  You
-    can search for multiple matches by typing a comma-seperated list
-    into the text fields.  Regular expressions, and wildcards are also
-    supported.  Blank text input fields are treated as wildcards.
-  </p>
-  <p>
-    Any of the text entry fields can take a comma-seperated list of
-    search arguments.  For example, to search for all commits from
-    authors <em>jpaint</em> and <em>gstein</em>, just type: <code>jpaint,
-    gstein</code> in the <em>Author</em> input box.  If you are searching
-    for items containing spaces or quotes, you will need to quote your
-    request.  For example, the same search above with quotes is:
-    <code>"jpaint", "gstein"</code>.
-  </p>
-  <p>                           
-    Wildcard and regular expression searches are entered in a similar
-    way to the quoted requests.  You must quote any wildcard or
-    regular expression request, and a command character preceeds the
-    first quote.  The command character <code>l</code>(lowercase L) is for wildcard
-    searches, and the wildcard character is a percent (<code>%</code>).  The
-    command character for regular expressions is <code>r</code>, and is
-    passed directly to MySQL, so you'll need to refer to the MySQL
-    manual for the exact regex syntax.  It is very similar to Perl.  A
-    wildard search for all files with a <em>.py</em> extention is:
-    <code>l"%.py"</code> in the <em>File</em> input box.  The same search done
-    with a regular expression is: <code>r".*\.py"</code>.
-  </p>
-  <p>                  
-    All search types can be mixed, as long as they are seperated by
-    commas.
-  </p>                                                    
-  </td></tr></table>
-</body></html>
--- a/templates-contrib/newvc/templates/docroot/help_rootview.html
+++ b/templates-contrib/newvc/templates/docroot/help_rootview.html
@@ -1,166 +0,0 @@
-<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" 
-    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
-<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
-<head>
-  <title>ViewVC Help: General</title>
-  <link rel="stylesheet" href="help.css" type="text/css" />
-</head>
-<body>
-  <table>
-    <col class="menu" />
-    <col />
-    <tr>
-      <td colspan="2">
-	<h1>ViewVC Help: General</h1>
-      </td>
-    </tr>
-    <tr><td>
-       <h3>Help</h3>
-       <strong>General</strong><br />
-       <a href="help_dirview.html">Directory&nbsp;View</a><br />
-       <a href="help_log.html">Log&nbsp;View</a><br />
-
-       <h3>Internet</h3>
-       <a href="http://viewvc.org/index.html">Home</a><br />
-       <a href="http://viewvc.org/upgrading.html">Upgrading</a><br />
-       <a href="http://viewvc.org/contributing.html">Contributing</a><br />
-       <a href="http://viewvc.org/license-1.html">License</a><br />
-    </td><td colspan="2">
-
-  <p><em>ViewVC</em> is a WWW interface for CVS and Subversion
-  repositories. It allows you to browse the files and directories in a
-  repository while showing you metadata from the repository history: log
-  messages, modification dates, author names, revision numbers, copy
-  history, and so on. It provides several different views of repository
-  data to help you find the information you are looking for:</p>
-
-  <ul>
-    <li><a name="view-dir" href="help_dirview.html"><strong>Directory
-    View</strong></a> - Shows a list of files and subdirectories in a
-    directory of the repository, along with metadata like author names and
-    log entries.</li>
-
-    <li><a name="view-log" href="help_log.html"><strong>Log
-    View</strong></a> - Shows a revision by revision list of all the
-    changes that have made to a file or directory in the repository, with
-    metadata and links to views of each revision.</li>
-
-    <li><a name="view-markup"><strong>File Contents View (Markup
-    View)</strong></a> - Shows the contents of a file at a particular
-    revision, with revision information at the top of the page. File
-    revisions which are GIF, PNG, or JPEG images are displayed inline on
-    the page. Other file types are displayed as marked up text. The markup
-    may be limited to turning URLs and email addresses into links, or
-    configured to show colorized source code.</li>
-
-    <li><a name="view-checkout"><strong>File Download (Checkout
-    View)</strong></a> - Retrieves the unaltered contents of a file
-    revision. Browsers may try to display the file, or just save it to
-    disk.</li>
-
-    <li><a name="view-annotate"><strong>File Annotate View</strong></a> -
-    Shows the contents of a file revision and breaks it down line by line,
-    showing the revision number where each one was last modified, along
-    with links and other information. <em>This view is disabled in some
-    ViewVC configurations</em></li>
-
-    <li><a name="view-diff"><strong>File Diff View</strong></a> - Shows
-    the changes made between two revisions of a file</li>
-
-    <li><a name="view-tarball"><strong>Directory Tarball View</strong> -
-    Retrieves a gzipped tar archive containing the contents of a
-    directory.<em>This view is disabled in the default ViewVC
-    configuration.</em></li>
-
-    <li><a name="view-query"><strong>Directory Query View</strong></a> -
-    Shows information about changes made to all subdirectories and files
-    under a parent directory, sorted and filtered by criteria you specify.
-    <em>This view is disabled in the default ViewVC configuration.</em>
-    </li>
-
-    <li><a name="view-rev"><strong>Revision View</strong> - Shows
-    information about a revision including log message, author, and a list
-    of changed paths. <em>For Subversion repositories only.</em></li>
-
-    <li><a name="view-graph"><strong>Graph View</strong></a> - Shows a
-    graphical representation of a file's revisions and branches complete
-    with tag and author names and links to markup and diff pages.
-    <em>For CVS repositories only, and disabled in the default
-    configuration.</em></li>
-  </ul>
-
-  <h3><a name="multiple-repositories">Multiple Repositories</a></h3>
-
-  <p>A single installation of ViewVC is often used to provide access to
-  more than one repository. In these installations, ViewVC shows a
-  <em>Project Root</em> drop down box in the top right corner of every
-  generated page to allow for quick access to any repository.</p>
-
-  <h3><a name="sticky-revision-tag">Sticky Revision and Tag</a></h3>
-
-  <p>By default, ViewVC will show the files and directories and revisions
-  that currently exist in the repository. But it's also possible to browse
-  the contents of a repository at a point in its past history by choosing
-  a "sticky tag" (in CVS) or a "sticky revision" (in Subversion) from the
-  forms at the top of directory and log pages. They're called sticky
-  because once they're chosen, they stick around when you navigate to
-  other pages, until you reset them. When they're set, directory and log
-  pages only show revisions preceding the specified point in history. In
-  CVS, when a tag refers to a branch or a revision on a branch, only
-  revisions from the branch history are shown, including branch points and
-  their preceding revisions.</p>
-
-  <h3><a name="dead-files">Dead Files</a></h3>
-
-  <p>In CVS directory listings, ViewVC can optionally display dead files.
-  Dead files are files which used to be in a directory but are currently
-  deleted, or files which just don't exist in the currently selected
-  <a href="#sticky-revision-tag">sticky tag</a>. Dead files cannot be
-  shown in Subversion repositories. The only way to see a deleted file in
-  a Subversion directory is to navigate to a sticky revision where the
-  file previously existed.</p>
-
-  <h3><a name="artificial-tags">Artificial Tags</a></h3>
-
-  <p>In CVS Repositories, ViewVC adds artificial tags <em>HEAD</em> and
-  <em>MAIN</em> to tag listings and accepts them in place of revision
-  numbers and real tag names in all URLs. <em>MAIN</em> acts like a branch
-  tag pointing at the default branch, while <em>HEAD</em> acts like a
-  revision tag pointing to the latest revision on the default branch. The
-  default branch is usually just the trunk, but may be set to other
-  branches inside individual repository files. CVS will always check out
-  revisions from a file's default branch when no other branch is specified
-  on the command line.</p>
-
-  <h3><a name="more-information">More Information</a></h3>
-
-  <p>More information about <em>ViewVC</em> is available from
-  <a href="http://viewvc.org/">viewvc.org</a>.
-  See the links below for guides to CVS and Subversion</p>
-
-  <h4>Documentation about CVS</h4>
-  <blockquote>
-    <p>
-      <a href="http://cvsbook.red-bean.com/"><em>Open Source
-      Development with CVS</em></a><br />
-      <a href="http://www.loria.fr/~molli/cvs/doc/cvs_toc.html">CVS
-      User's Guide</a><br />
-      <a href="http://cellworks.washington.edu/pub/docs/cvs/tutorial/cvs_tutorial_1.html">Another CVS tutorial</a><br />
-      <a href="http://www.csc.calpoly.edu/~dbutler/tutorials/winter96/cvs/">Yet another CVS tutorial (a little old, but nice)</a><br />
-      <a href="http://www.cs.utah.edu/dept/old/texinfo/cvs/FAQ.txt">An old but very useful FAQ about CVS</a>
-    </p>
-  </blockquote>
-
-  <h4>Documentation about Subversion</h3>
-  <blockquote>
-    <p>
-      <a href="http://svnbook.red-bean.com/"><em>Version Control with
-      Subversion</em></a><br />
-    </p>
-  </blockquote>
-
-  </td></tr></table>
-  <hr />
-  <address><a href="mailto:users@viewvc.tigris.org">ViewVC Users Mailinglist</a></address>
-  </body>
-</html>
--- a/templates-contrib/newvc/templates/docroot/images/back.png
+++ b/templates-contrib/newvc/templates/docroot/images/back.png
--- a/templates-contrib/newvc/templates/docroot/images/back_small.png
+++ b/templates-contrib/newvc/templates/docroot/images/back_small.png
--- a/templates-contrib/newvc/templates/docroot/images/broken.png
+++ b/templates-contrib/newvc/templates/docroot/images/broken.png
--- a/templates-contrib/newvc/templates/docroot/images/cvs-logo.png
+++ b/templates-contrib/newvc/templates/docroot/images/cvs-logo.png
--- a/templates-contrib/newvc/templates/docroot/images/dir.png
+++ b/templates-contrib/newvc/templates/docroot/images/dir.png
--- a/Show More
+++ b/Show More