Use TS?

Issue : https://scality.atlassian.net/browse/ARSN-426

.eslintrc

@@ -1 +1,6 @@
-{ "extends": "scality" }
+{
+    "extends": "scality",
+    "parserOptions": {
+        "ecmaVersion": 2020
+    }
+}

.github/workflows/codeql.yaml

@@ -0,0 +1,25 @@
+---
+name: codeQL
+
+on:
+  push:
+    branches: [development/*, stabilization/*, hotfix/*]
+  pull_request:
+    branches: [development/*, stabilization/*, hotfix/*]
+  workflow_dispatch:
+
+jobs:
+  analyze:
+    name: Static analysis with CodeQL
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - name: Initialize CodeQL
+        uses: github/codeql-action/init@v3
+        with:
+          languages: javascript, typescript
+
+      - name: Build and analyze
+        uses: github/codeql-action/analyze@v3

.github/workflows/dependency-review.yaml

@@ -0,0 +1,16 @@
+---
+name: dependency review
+
+on:
+  pull_request:
+    branches: [development/*, stabilization/*, hotfix/*]
+
+jobs:
+  dependency-review:
+    runs-on: ubuntu-latest
+    steps:
+      - name: 'Checkout Repository'
+        uses: actions/checkout@v4
+
+      - name: 'Dependency Review'
+        uses: actions/dependency-review-action@v4

.github/workflows/tests.yaml

@@ -25,18 +25,18 @@ jobs:
           - 6379:6379
     steps:
     - name: Checkout
-      uses: actions/checkout@v2
-    - uses: actions/setup-node@v2
+      uses: actions/checkout@v4
+    - uses: actions/setup-node@v4
       with:
         node-version: '16'
         cache: 'yarn'
     - name: install dependencies
-      run: yarn install --frozen-lockfile --prefer-offline
+      run: yarn install --frozen-lockfile --prefer-offline --network-concurrency 1
       continue-on-error: true # TODO ARSN-97 Remove it when no errors in TS
     - name: lint yaml
       run: yarn --silent lint_yml
     - name: lint javascript
-      run: yarn --silent lint -- --max-warnings 0
+      run: yarn --silent lint --max-warnings 0
     - name: lint markdown
       run: yarn --silent lint_md
     - name: add hostname
@@ -46,7 +46,9 @@ jobs:
       run: yarn --silent coverage
     - name: run functional tests
       run: yarn ft_test
-    - uses: codecov/codecov-action@v2
+    - uses: codecov/codecov-action@v4
+      with:
+        token: ${{ secrets.CODECOV_TOKEN }}
     - name: run executables tests
       run: yarn install && yarn test
       working-directory: 'lib/executables/pensieveCreds/'
@@ -57,9 +59,9 @@ jobs:
     runs-on: ubuntu-latest
     steps:
     - name: Checkout
-      uses: actions/checkout@v2
+      uses: actions/checkout@v4
     - name: Install NodeJS
-      uses: actions/setup-node@v2
+      uses: actions/setup-node@v4
       with:
         node-version: '16'
         cache: yarn
@@ -70,7 +72,7 @@ jobs:
       run: yarn build
       continue-on-error: true # TODO ARSN-97 Remove it when no errors in TS
     - name: Upload artifacts
-      uses: scality/action-artifacts@v2
+      uses: scality/action-artifacts@v4
       with:
         url: https://artifacts.scality.net
         user: ${{ secrets.ARTIFACTS_USER }}

.swcrc

@@ -0,0 +1,12 @@
+{
+  "$schema": "https://swc.rs/schema.json",
+  "jsc": {
+    "parser": {
+      "syntax": "typescript"
+    },
+    "target": "es2017"
+  },
+  "module": {
+    "type": "commonjs"
+  }
+}

documentation/BucketInfoModelVersion.md

@@ -178,3 +178,83 @@ this._serverSideEncryption.configuredMasterKeyId = configuredMasterKeyId || unde
 ### Usage
 
 Used to store the users configured KMS key id
+
+## Model version 15
+
+### Properties Added
+
+```javascript
+this._tags = tags || null;
+```
+
+The Tag Set of a bucket is an array of objects with Key and Value:
+
+```javascript
+[
+    {
+        Key: 'something',
+        Value: 'some_data'
+    }
+]
+```
+
+## Model version 16
+
+### Properties Added
+
+```javascript
+this._capabilities = capabilities || undefined;
+```
+
+For capacity-enabled buckets, contains the following data:
+
+```javascript
+{
+    _capabilities: {
+        VeeamSOSApi?: {
+            SystemInfo?: {
+                ProtocolVersion: String,
+                ModelName: String,
+                ProtocolCapabilities: {
+                    CapacityInfo: Boolean,
+                    UploadSessions: Boolean,
+                    IAMSTS: Boolean,
+                },
+                APIEndpoints: {
+                    IAMEndpoint: String,
+                    STSEndpoint: String,
+                },
+                SystemRecommendations?: {
+                    S3ConcurrentTaskLimit: Number,
+                    S3MultiObjectDelete: Number,
+                    StorageCurrentTasksLimit: Number,
+                    KbBlockSize: Number,
+                }
+                LastModified?: String,
+            },
+            CapacityInfo?: {
+                Capacity: Number,
+                Available: Number,
+                Used: Number,
+                LastModified?: String,
+            },
+        }
+    },
+}
+```
+
+### Usage
+
+Used to store bucket tagging
+
+## Model version 17
+
+### Properties Added
+
+```javascript
+this._quotaMax = quotaMax || 0;
+```
+
+### Usage
+
+Used to store bucket quota

documentation/listingAlgos/delimiter.md

@@ -0,0 +1,27 @@
+# Delimiter
+
+The Delimiter class handles raw listings from the database with an
+optional delimiter, and fills in a curated listing with "Contents" and
+"CommonPrefixes" as a result.
+
+## Expected Behavior
+
+- only lists keys belonging to the given **prefix** (if provided)
+
+- groups listed keys that have a common prefix ending with a delimiter
+  inside CommonPrefixes
+
+- can take a **marker** or **continuationToken** to list from a specific key
+
+- can take a **maxKeys** parameter to limit how many keys can be returned
+
+## State Chart
+
+- States with grey background are *Idle* states, which are waiting for
+  a new listing key
+
+- States with blue background are *Processing* states, which are
+  actively processing a new listing key passed by the filter()
+  function
+
+![Delimiter State Chart](./pics/delimiterStateChart.svg)

documentation/listingAlgos/delimiterMaster.md

@@ -0,0 +1,45 @@
+# DelimiterMaster
+
+The DelimiterMaster class handles raw listings from the database of a
+versioned or non-versioned bucket with an optional delimiter, and
+fills in a curated listing with "Contents" and "CommonPrefixes" as a
+result.
+
+## Expected Behavior
+
+- only lists latest versions of versioned buckets
+
+- only lists keys belonging to the given **prefix** (if provided)
+
+- does not list latest versions that are delete markers
+
+- groups listed keys that have a common prefix ending with a delimiter
+  inside CommonPrefixes
+
+- can take a **marker** or **continuationToken** to list from a specific key
+
+- can take a **maxKeys** parameter to limit how many keys can be returned
+
+- reconciles internal PHD keys with the next version (those are
+  created when a specific version that is the latest version is
+  deleted)
+
+- skips internal keys like replay keys
+
+## State Chart
+
+- States with grey background are *Idle* states, which are waiting for
+  a new listing key
+
+- States with blue background are *Processing* states, which are
+  actively processing a new listing key passed by the filter()
+  function
+
+### Bucket Vformat=v0
+
+![DelimiterMaster State Chart for v0 format](./pics/delimiterMasterV0StateChart.svg)
+
+### Bucket Vformat=v1
+
+For buckets in versioning key format **v1**, the algorithm used is the
+one from [Delimiter](delimiter.md).

documentation/listingAlgos/delimiterVersions.md

@@ -0,0 +1,33 @@
+# DelimiterVersions
+
+The DelimiterVersions class handles raw listings from the database of a
+versioned or non-versioned bucket with an optional delimiter, and
+fills in a curated listing with "Versions" and "CommonPrefixes" as a
+result.
+
+## Expected Behavior
+
+- lists individual distinct versions of versioned buckets
+
+- only lists keys belonging to the given **prefix** (if provided)
+
+- groups listed keys that have a common prefix ending with a delimiter
+  inside CommonPrefixes
+
+- can take a **keyMarker** and optionally a **versionIdMarker** to
+  list from a specific key or version
+
+- can take a **maxKeys** parameter to limit how many keys can be returned
+
+- skips internal keys like replay keys
+
+## State Chart
+
+- States with grey background are *Idle* states, which are waiting for
+  a new listing key
+
+- States with blue background are *Processing* states, which are
+  actively processing a new listing key passed by the filter()
+  function
+
+![DelimiterVersions State Chart](./pics/delimiterVersionsStateChart.svg)

documentation/listingAlgos/pics/delimiterMasterV0StateChart.dot

@@ -0,0 +1,45 @@
+digraph {
+  node [shape="box",style="filled,rounded",fontsize=16,fixedsize=true,width=3];
+  edge [fontsize=14];
+  rankdir=TB;
+
+  START [shape="circle",width=0.2,label="",style="filled",fillcolor="black"]
+  END [shape="circle",width=0.2,label="",style="filled",fillcolor="black",peripheries=2]
+
+  node [fillcolor="lightgrey"];
+  "NotSkippingPrefixNorVersions.Idle" [label="NotSkippingPrefixNorVersions",group="NotSkippingPrefixNorVersions",width=4];
+  "SkippingPrefix.Idle" [label="SkippingPrefix",group="SkippingPrefix"];
+  "SkippingVersions.Idle" [label="SkippingVersions",group="SkippingVersions"];
+  "WaitVersionAfterPHD.Idle" [label="WaitVersionAfterPHD",group="WaitVersionAfterPHD"];
+
+  node [fillcolor="lightblue"];
+  "NotSkippingPrefixNorVersions.Processing" [label="NotSkippingPrefixNorVersions",group="NotSkippingPrefixNorVersions",width=4];
+  "SkippingPrefix.Processing" [label="SkippingPrefix",group="SkippingPrefix"];
+  "SkippingVersions.Processing" [label="SkippingVersions",group="SkippingVersions"];
+  "WaitVersionAfterPHD.Processing" [label="WaitVersionAfterPHD",group="WaitVersionAfterPHD"];
+
+  START -> "SkippingVersions.Idle" [label="[marker != undefined]"]
+  START -> "NotSkippingPrefixNorVersions.Idle" [label="[marker == undefined]"]
+
+  "NotSkippingPrefixNorVersions.Idle" -> "NotSkippingPrefixNorVersions.Processing" [label="filter(key, value)"]
+  "SkippingPrefix.Idle" -> "SkippingPrefix.Processing" [label="filter(key, value)"]
+  "SkippingVersions.Idle" -> "SkippingVersions.Processing" [label="filter(key, value)"]
+  "WaitVersionAfterPHD.Idle" -> "WaitVersionAfterPHD.Processing" [label="filter(key, value)"]
+
+
+  "NotSkippingPrefixNorVersions.Processing" -> "SkippingVersions.Idle" [label="[Version.isDeleteMarker(value)]\n-> FILTER_ACCEPT"]
+  "NotSkippingPrefixNorVersions.Processing" -> "WaitVersionAfterPHD.Idle" [label="[Version.isPHD(value)]\n-> FILTER_ACCEPT"]
+  "NotSkippingPrefixNorVersions.Processing" -> "SkippingPrefix.Idle" [label="[key.startsWith(<ReplayPrefix>)]\n/ prefix <- <ReplayPrefix>\n-> FILTER_SKIP"]
+  "NotSkippingPrefixNorVersions.Processing" -> END [label="[isListableKey(key, value) and\nKeys == maxKeys]\n-> FILTER_END"]
+  "NotSkippingPrefixNorVersions.Processing" -> "SkippingPrefix.Idle" [label="[isListableKey(key, value) and\nnKeys < maxKeys and\nhasDelimiter(key)]\n/ prefix <- prefixOf(key)\n/ CommonPrefixes.append(prefixOf(key))\n-> FILTER_ACCEPT"]
+  "NotSkippingPrefixNorVersions.Processing" -> "SkippingVersions.Idle" [label="[isListableKey(key, value) and\nnKeys < maxKeys and\nnot hasDelimiter(key)]\n/ Contents.append(key, value)\n-> FILTER_ACCEPT"]
+
+  "SkippingPrefix.Processing" -> "SkippingPrefix.Idle" [label="[key.startsWith(prefix)]\n-> FILTER_SKIP"]
+  "SkippingPrefix.Processing" -> "NotSkippingPrefixNorVersions.Processing" [label="[not key.startsWith(prefix)]"]
+
+  "SkippingVersions.Processing" -> "SkippingVersions.Idle" [label="[isVersionKey(key)]\n-> FILTER_SKIP"]
+  "SkippingVersions.Processing" -> "NotSkippingPrefixNorVersions.Processing" [label="[not isVersionKey(key)]"]
+
+  "WaitVersionAfterPHD.Processing" -> "NotSkippingPrefixNorVersions.Processing" [label="[isVersionKey(key) and master(key) == PHDkey]\n/ key <- master(key)"]
+  "WaitVersionAfterPHD.Processing" -> "NotSkippingPrefixNorVersions.Processing" [label="[not isVersionKey(key) or master(key) != PHDkey]"]
+}

documentation/listingAlgos/pics/delimiterMasterV0StateChart.svg

@@ -0,0 +1,216 @@
+<?xml version="1.0" encoding="UTF-8" standalone="no"?>
+<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN"
+ "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">
+<!-- Generated by graphviz version 2.43.0 (0)
+ -->
+<!-- Title: %3 Pages: 1 -->
+<svg width="2313pt" height="460pt"
+ viewBox="0.00 0.00 2313.37 460.00" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
+<g id="graph0" class="graph" transform="scale(1 1) rotate(0) translate(4 456)">
+<title>%3</title>
+<polygon fill="white" stroke="transparent" points="-4,4 -4,-456 2309.37,-456 2309.37,4 -4,4"/>
+<!-- START -->
+<g id="node1" class="node">
+<title>START</title>
+<ellipse fill="black" stroke="black" cx="35.37" cy="-445" rx="7" ry="7"/>
+</g>
+<!-- NotSkippingPrefixNorVersions.Idle -->
+<g id="node3" class="node">
+<title>NotSkippingPrefixNorVersions.Idle</title>
+<path fill="lightgrey" stroke="black" d="M925.37,-387C925.37,-387 661.37,-387 661.37,-387 655.37,-387 649.37,-381 649.37,-375 649.37,-375 649.37,-363 649.37,-363 649.37,-357 655.37,-351 661.37,-351 661.37,-351 925.37,-351 925.37,-351 931.37,-351 937.37,-357 937.37,-363 937.37,-363 937.37,-375 937.37,-375 937.37,-381 931.37,-387 925.37,-387"/>
+<text text-anchor="middle" x="793.37" y="-365.2" font-family="Times,serif" font-size="16.00">NotSkippingPrefixNorVersions</text>
+</g>
+<!-- START&#45;&gt;NotSkippingPrefixNorVersions.Idle -->
+<g id="edge2" class="edge">
+<title>START&#45;&gt;NotSkippingPrefixNorVersions.Idle</title>
+<path fill="none" stroke="black" d="M42.39,-443.31C95.3,-438.15 434.98,-404.99 638.94,-385.08"/>
+<polygon fill="black" stroke="black" points="639.54,-388.53 649.15,-384.08 638.86,-381.57 639.54,-388.53"/>
+<text text-anchor="middle" x="497.87" y="-408.8" font-family="Times,serif" font-size="14.00">[marker == undefined]</text>
+</g>
+<!-- SkippingVersions.Idle -->
+<g id="node5" class="node">
+<title>SkippingVersions.Idle</title>
+<path fill="lightgrey" stroke="black" d="M242.37,-138C242.37,-138 50.37,-138 50.37,-138 44.37,-138 38.37,-132 38.37,-126 38.37,-126 38.37,-114 38.37,-114 38.37,-108 44.37,-102 50.37,-102 50.37,-102 242.37,-102 242.37,-102 248.37,-102 254.37,-108 254.37,-114 254.37,-114 254.37,-126 254.37,-126 254.37,-132 248.37,-138 242.37,-138"/>
+<text text-anchor="middle" x="146.37" y="-116.2" font-family="Times,serif" font-size="16.00">SkippingVersions</text>
+</g>
+<!-- START&#45;&gt;SkippingVersions.Idle -->
+<g id="edge1" class="edge">
+<title>START&#45;&gt;SkippingVersions.Idle</title>
+<path fill="none" stroke="black" d="M33.04,-438.14C20.64,-405.9 -34.57,-248.17 33.37,-156 36.76,-151.4 40.74,-147.39 45.16,-143.89"/>
+<polygon fill="black" stroke="black" points="47.27,-146.68 53.53,-138.13 43.3,-140.92 47.27,-146.68"/>
+<text text-anchor="middle" x="85.87" y="-321.8" font-family="Times,serif" font-size="14.00">[marker != undefined]</text>
+</g>
+<!-- END -->
+<g id="node2" class="node">
+<title>END</title>
+<ellipse fill="black" stroke="black" cx="727.37" cy="-120" rx="7" ry="7"/>
+<ellipse fill="none" stroke="black" cx="727.37" cy="-120" rx="11" ry="11"/>
+</g>
+<!-- NotSkippingPrefixNorVersions.Processing -->
+<g id="node7" class="node">
+<title>NotSkippingPrefixNorVersions.Processing</title>
+<path fill="lightblue" stroke="black" d="M925.37,-300C925.37,-300 661.37,-300 661.37,-300 655.37,-300 649.37,-294 649.37,-288 649.37,-288 649.37,-276 649.37,-276 649.37,-270 655.37,-264 661.37,-264 661.37,-264 925.37,-264 925.37,-264 931.37,-264 937.37,-270 937.37,-276 937.37,-276 937.37,-288 937.37,-288 937.37,-294 931.37,-300 925.37,-300"/>
+<text text-anchor="middle" x="793.37" y="-278.2" font-family="Times,serif" font-size="16.00">NotSkippingPrefixNorVersions</text>
+</g>
+<!-- NotSkippingPrefixNorVersions.Idle&#45;&gt;NotSkippingPrefixNorVersions.Processing -->
+<g id="edge3" class="edge">
+<title>NotSkippingPrefixNorVersions.Idle&#45;&gt;NotSkippingPrefixNorVersions.Processing</title>
+<path fill="none" stroke="black" d="M793.37,-350.8C793.37,-339.16 793.37,-323.55 793.37,-310.24"/>
+<polygon fill="black" stroke="black" points="796.87,-310.18 793.37,-300.18 789.87,-310.18 796.87,-310.18"/>
+<text text-anchor="middle" x="851.37" y="-321.8" font-family="Times,serif" font-size="14.00">filter(key, value)</text>
+</g>
+<!-- SkippingPrefix.Idle -->
+<g id="node4" class="node">
+<title>SkippingPrefix.Idle</title>
+<path fill="lightgrey" stroke="black" d="M1209.37,-138C1209.37,-138 1017.37,-138 1017.37,-138 1011.37,-138 1005.37,-132 1005.37,-126 1005.37,-126 1005.37,-114 1005.37,-114 1005.37,-108 1011.37,-102 1017.37,-102 1017.37,-102 1209.37,-102 1209.37,-102 1215.37,-102 1221.37,-108 1221.37,-114 1221.37,-114 1221.37,-126 1221.37,-126 1221.37,-132 1215.37,-138 1209.37,-138"/>
+<text text-anchor="middle" x="1113.37" y="-116.2" font-family="Times,serif" font-size="16.00">SkippingPrefix</text>
+</g>
+<!-- SkippingPrefix.Processing -->
+<g id="node8" class="node">
+<title>SkippingPrefix.Processing</title>
+<path fill="lightblue" stroke="black" d="M1070.37,-36C1070.37,-36 878.37,-36 878.37,-36 872.37,-36 866.37,-30 866.37,-24 866.37,-24 866.37,-12 866.37,-12 866.37,-6 872.37,0 878.37,0 878.37,0 1070.37,0 1070.37,0 1076.37,0 1082.37,-6 1082.37,-12 1082.37,-12 1082.37,-24 1082.37,-24 1082.37,-30 1076.37,-36 1070.37,-36"/>
+<text text-anchor="middle" x="974.37" y="-14.2" font-family="Times,serif" font-size="16.00">SkippingPrefix</text>
+</g>
+<!-- SkippingPrefix.Idle&#45;&gt;SkippingPrefix.Processing -->
+<g id="edge4" class="edge">
+<title>SkippingPrefix.Idle&#45;&gt;SkippingPrefix.Processing</title>
+<path fill="none" stroke="black" d="M1011.89,-101.96C994.96,-97.13 981.04,-91.17 975.37,-84 967.11,-73.56 966.25,-58.93 967.72,-46.2"/>
+<polygon fill="black" stroke="black" points="971.22,-46.52 969.4,-36.09 964.31,-45.38 971.22,-46.52"/>
+<text text-anchor="middle" x="1033.37" y="-65.3" font-family="Times,serif" font-size="14.00">filter(key, value)</text>
+</g>
+<!-- SkippingVersions.Processing -->
+<g id="node9" class="node">
+<title>SkippingVersions.Processing</title>
+<path fill="lightblue" stroke="black" d="M381.37,-36C381.37,-36 189.37,-36 189.37,-36 183.37,-36 177.37,-30 177.37,-24 177.37,-24 177.37,-12 177.37,-12 177.37,-6 183.37,0 189.37,0 189.37,0 381.37,0 381.37,0 387.37,0 393.37,-6 393.37,-12 393.37,-12 393.37,-24 393.37,-24 393.37,-30 387.37,-36 381.37,-36"/>
+<text text-anchor="middle" x="285.37" y="-14.2" font-family="Times,serif" font-size="16.00">SkippingVersions</text>
+</g>
+<!-- SkippingVersions.Idle&#45;&gt;SkippingVersions.Processing -->
+<g id="edge5" class="edge">
+<title>SkippingVersions.Idle&#45;&gt;SkippingVersions.Processing</title>
+<path fill="none" stroke="black" d="M141.4,-101.91C138.35,-87.58 136.8,-67.37 147.37,-54 151.89,-48.28 161.64,-43.34 173.99,-39.12"/>
+<polygon fill="black" stroke="black" points="175.39,-42.36 183.89,-36.04 173.3,-35.67 175.39,-42.36"/>
+<text text-anchor="middle" x="205.37" y="-65.3" font-family="Times,serif" font-size="14.00">filter(key, value)</text>
+</g>
+<!-- WaitVersionAfterPHD.Idle -->
+<g id="node6" class="node">
+<title>WaitVersionAfterPHD.Idle</title>
+<path fill="lightgrey" stroke="black" d="M1534.37,-138C1534.37,-138 1342.37,-138 1342.37,-138 1336.37,-138 1330.37,-132 1330.37,-126 1330.37,-126 1330.37,-114 1330.37,-114 1330.37,-108 1336.37,-102 1342.37,-102 1342.37,-102 1534.37,-102 1534.37,-102 1540.37,-102 1546.37,-108 1546.37,-114 1546.37,-114 1546.37,-126 1546.37,-126 1546.37,-132 1540.37,-138 1534.37,-138"/>
+<text text-anchor="middle" x="1438.37" y="-116.2" font-family="Times,serif" font-size="16.00">WaitVersionAfterPHD</text>
+</g>
+<!-- WaitVersionAfterPHD.Processing -->
+<g id="node10" class="node">
+<title>WaitVersionAfterPHD.Processing</title>
+<path fill="lightblue" stroke="black" d="M1534.37,-36C1534.37,-36 1342.37,-36 1342.37,-36 1336.37,-36 1330.37,-30 1330.37,-24 1330.37,-24 1330.37,-12 1330.37,-12 1330.37,-6 1336.37,0 1342.37,0 1342.37,0 1534.37,0 1534.37,0 1540.37,0 1546.37,-6 1546.37,-12 1546.37,-12 1546.37,-24 1546.37,-24 1546.37,-30 1540.37,-36 1534.37,-36"/>
+<text text-anchor="middle" x="1438.37" y="-14.2" font-family="Times,serif" font-size="16.00">WaitVersionAfterPHD</text>
+</g>
+<!-- WaitVersionAfterPHD.Idle&#45;&gt;WaitVersionAfterPHD.Processing -->
+<g id="edge6" class="edge">
+<title>WaitVersionAfterPHD.Idle&#45;&gt;WaitVersionAfterPHD.Processing</title>
+<path fill="none" stroke="black" d="M1438.37,-101.58C1438.37,-86.38 1438.37,-64.07 1438.37,-46.46"/>
+<polygon fill="black" stroke="black" points="1441.87,-46.22 1438.37,-36.22 1434.87,-46.22 1441.87,-46.22"/>
+<text text-anchor="middle" x="1496.37" y="-65.3" font-family="Times,serif" font-size="14.00">filter(key, value)</text>
+</g>
+<!-- NotSkippingPrefixNorVersions.Processing&#45;&gt;END -->
+<g id="edge10" class="edge">
+<title>NotSkippingPrefixNorVersions.Processing&#45;&gt;END</title>
+<path fill="none" stroke="black" d="M649.15,-273.62C611.7,-268.54 578.44,-260.07 566.37,-246 540.33,-215.64 540,-186.08 566.37,-156 586.46,-133.07 673.88,-148.86 702.37,-138 705.22,-136.91 708.06,-135.44 710.76,-133.82"/>
+<polygon fill="black" stroke="black" points="712.88,-136.61 719.13,-128.05 708.91,-130.84 712.88,-136.61"/>
+<text text-anchor="middle" x="672.87" y="-212.3" font-family="Times,serif" font-size="14.00">[isListableKey(key, value) and</text>
+<text text-anchor="middle" x="672.87" y="-197.3" font-family="Times,serif" font-size="14.00">Keys == maxKeys]</text>
+<text text-anchor="middle" x="672.87" y="-182.3" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_END</text>
+</g>
+<!-- NotSkippingPrefixNorVersions.Processing&#45;&gt;SkippingPrefix.Idle -->
+<g id="edge9" class="edge">
+<title>NotSkippingPrefixNorVersions.Processing&#45;&gt;SkippingPrefix.Idle</title>
+<path fill="none" stroke="black" d="M937.6,-274.31C1018.89,-269.01 1106.69,-260.11 1119.37,-246 1143.16,-219.51 1134.03,-175.72 1124.38,-147.62"/>
+<polygon fill="black" stroke="black" points="1127.6,-146.22 1120.86,-138.04 1121.03,-148.64 1127.6,-146.22"/>
+<text text-anchor="middle" x="1254.37" y="-212.3" font-family="Times,serif" font-size="14.00">[key.startsWith(&lt;ReplayPrefix&gt;)]</text>
+<text text-anchor="middle" x="1254.37" y="-197.3" font-family="Times,serif" font-size="14.00">/ prefix &lt;&#45; &lt;ReplayPrefix&gt;</text>
+<text text-anchor="middle" x="1254.37" y="-182.3" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_SKIP</text>
+</g>
+<!-- NotSkippingPrefixNorVersions.Processing&#45;&gt;SkippingPrefix.Idle -->
+<g id="edge11" class="edge">
+<title>NotSkippingPrefixNorVersions.Processing&#45;&gt;SkippingPrefix.Idle</title>
+<path fill="none" stroke="black" d="M799.18,-263.65C800.96,-258.05 802.85,-251.79 804.37,-246 814.73,-206.45 793.03,-183.41 823.37,-156 851.23,-130.83 954.1,-142.59 991.37,-138 992.65,-137.84 993.94,-137.68 995.24,-137.52"/>
+<polygon fill="black" stroke="black" points="995.81,-140.98 1005.29,-136.25 994.93,-134.03 995.81,-140.98"/>
+<text text-anchor="middle" x="969.37" y="-234.8" font-family="Times,serif" font-size="14.00">[isListableKey(key, value) and</text>
+<text text-anchor="middle" x="969.37" y="-219.8" font-family="Times,serif" font-size="14.00">nKeys &lt; maxKeys and</text>
+<text text-anchor="middle" x="969.37" y="-204.8" font-family="Times,serif" font-size="14.00">hasDelimiter(key)]</text>
+<text text-anchor="middle" x="969.37" y="-189.8" font-family="Times,serif" font-size="14.00">/ prefix &lt;&#45; prefixOf(key)</text>
+<text text-anchor="middle" x="969.37" y="-174.8" font-family="Times,serif" font-size="14.00">/ CommonPrefixes.append(prefixOf(key))</text>
+<text text-anchor="middle" x="969.37" y="-159.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_ACCEPT</text>
+</g>
+<!-- NotSkippingPrefixNorVersions.Processing&#45;&gt;SkippingVersions.Idle -->
+<g id="edge7" class="edge">
+<title>NotSkippingPrefixNorVersions.Processing&#45;&gt;SkippingVersions.Idle</title>
+<path fill="none" stroke="black" d="M649.11,-279.23C439.56,-275.94 73.58,-267.19 53.37,-246 25.76,-217.06 30.6,-188.89 53.37,-156 56.56,-151.39 60.44,-147.39 64.78,-143.91"/>
+<polygon fill="black" stroke="black" points="66.8,-146.76 73.04,-138.2 62.83,-141 66.8,-146.76"/>
+<text text-anchor="middle" x="167.87" y="-204.8" font-family="Times,serif" font-size="14.00">[Version.isDeleteMarker(value)]</text>
+<text text-anchor="middle" x="167.87" y="-189.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_ACCEPT</text>
+</g>
+<!-- NotSkippingPrefixNorVersions.Processing&#45;&gt;SkippingVersions.Idle -->
+<g id="edge12" class="edge">
+<title>NotSkippingPrefixNorVersions.Processing&#45;&gt;SkippingVersions.Idle</title>
+<path fill="none" stroke="black" d="M649.33,-279.1C514.97,-275.99 331.4,-267.75 305.37,-246 273.69,-219.53 311.53,-185.22 282.37,-156 276.73,-150.36 270.32,-145.59 263.42,-141.56"/>
+<polygon fill="black" stroke="black" points="264.92,-138.39 254.44,-136.84 261.67,-144.59 264.92,-138.39"/>
+<text text-anchor="middle" x="411.87" y="-227.3" font-family="Times,serif" font-size="14.00">[isListableKey(key, value) and</text>
+<text text-anchor="middle" x="411.87" y="-212.3" font-family="Times,serif" font-size="14.00">nKeys &lt; maxKeys and</text>
+<text text-anchor="middle" x="411.87" y="-197.3" font-family="Times,serif" font-size="14.00">not hasDelimiter(key)]</text>
+<text text-anchor="middle" x="411.87" y="-182.3" font-family="Times,serif" font-size="14.00">/ Contents.append(key, value)</text>
+<text text-anchor="middle" x="411.87" y="-167.3" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_ACCEPT</text>
+</g>
+<!-- NotSkippingPrefixNorVersions.Processing&#45;&gt;WaitVersionAfterPHD.Idle -->
+<g id="edge8" class="edge">
+<title>NotSkippingPrefixNorVersions.Processing&#45;&gt;WaitVersionAfterPHD.Idle</title>
+<path fill="none" stroke="black" d="M937.38,-280.87C1099.43,-279.42 1344.59,-272.74 1378.37,-246 1411.11,-220.08 1384.48,-192.16 1405.37,-156 1407.38,-152.52 1409.8,-149.11 1412.4,-145.87"/>
+<polygon fill="black" stroke="black" points="1415.16,-148.04 1419.13,-138.21 1409.9,-143.41 1415.16,-148.04"/>
+<text text-anchor="middle" x="1486.87" y="-204.8" font-family="Times,serif" font-size="14.00">[Version.isPHD(value)]</text>
+<text text-anchor="middle" x="1486.87" y="-189.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_ACCEPT</text>
+</g>
+<!-- SkippingPrefix.Processing&#45;&gt;SkippingPrefix.Idle -->
+<g id="edge13" class="edge">
+<title>SkippingPrefix.Processing&#45;&gt;SkippingPrefix.Idle</title>
+<path fill="none" stroke="black" d="M1064.61,-36.08C1074.44,-40.7 1083.66,-46.57 1091.37,-54 1101.65,-63.92 1107.13,-78.81 1110.04,-91.84"/>
+<polygon fill="black" stroke="black" points="1106.62,-92.56 1111.88,-101.76 1113.5,-91.29 1106.62,-92.56"/>
+<text text-anchor="middle" x="1190.37" y="-72.8" font-family="Times,serif" font-size="14.00">[key.startsWith(prefix)]</text>
+<text text-anchor="middle" x="1190.37" y="-57.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_SKIP</text>
+</g>
+<!-- SkippingPrefix.Processing&#45;&gt;NotSkippingPrefixNorVersions.Processing -->
+<g id="edge14" class="edge">
+<title>SkippingPrefix.Processing&#45;&gt;NotSkippingPrefixNorVersions.Processing</title>
+<path fill="none" stroke="black" d="M899.82,-36.01C864.18,-48.2 824.54,-68.57 802.37,-102 771.84,-148.02 779.31,-216.26 786.77,-253.8"/>
+<polygon fill="black" stroke="black" points="783.43,-254.92 788.94,-263.97 790.28,-253.46 783.43,-254.92"/>
+<text text-anchor="middle" x="899.37" y="-116.3" font-family="Times,serif" font-size="14.00">[not key.startsWith(prefix)]</text>
+</g>
+<!-- SkippingVersions.Processing&#45;&gt;SkippingVersions.Idle -->
+<g id="edge15" class="edge">
+<title>SkippingVersions.Processing&#45;&gt;SkippingVersions.Idle</title>
+<path fill="none" stroke="black" d="M283.88,-36.24C281.71,-50.87 276.4,-71.43 263.37,-84 258.07,-89.11 252.06,-93.48 245.62,-97.21"/>
+<polygon fill="black" stroke="black" points="243.85,-94.19 236.61,-101.92 247.09,-100.39 243.85,-94.19"/>
+<text text-anchor="middle" x="349.87" y="-72.8" font-family="Times,serif" font-size="14.00">[isVersionKey(key)]</text>
+<text text-anchor="middle" x="349.87" y="-57.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_SKIP</text>
+</g>
+<!-- SkippingVersions.Processing&#45;&gt;NotSkippingPrefixNorVersions.Processing -->
+<g id="edge16" class="edge">
+<title>SkippingVersions.Processing&#45;&gt;NotSkippingPrefixNorVersions.Processing</title>
+<path fill="none" stroke="black" d="M382.46,-36.08C396.72,-40.7 410.82,-46.57 423.37,-54 476.67,-85.57 487.28,-102.42 518.37,-156 539.39,-192.23 514.46,-218.85 546.37,-246 561.72,-259.06 598.56,-267.25 639.23,-272.39"/>
+<polygon fill="black" stroke="black" points="639.01,-275.89 649.36,-273.59 639.84,-268.93 639.01,-275.89"/>
+<text text-anchor="middle" x="590.37" y="-116.3" font-family="Times,serif" font-size="14.00">[not isVersionKey(key)]</text>
+</g>
+<!-- WaitVersionAfterPHD.Processing&#45;&gt;NotSkippingPrefixNorVersions.Processing -->
+<g id="edge17" class="edge">
+<title>WaitVersionAfterPHD.Processing&#45;&gt;NotSkippingPrefixNorVersions.Processing</title>
+<path fill="none" stroke="black" d="M1536.41,-36.13C1544.73,-40.79 1552.27,-46.65 1558.37,-54 1585.64,-86.89 1597.89,-215.12 1568.37,-246 1547.29,-268.05 1167.71,-276.42 947.74,-279.43"/>
+<polygon fill="black" stroke="black" points="947.67,-275.93 937.71,-279.57 947.76,-282.93 947.67,-275.93"/>
+<text text-anchor="middle" x="1758.37" y="-123.8" font-family="Times,serif" font-size="14.00">[isVersionKey(key) and master(key) == PHDkey]</text>
+<text text-anchor="middle" x="1758.37" y="-108.8" font-family="Times,serif" font-size="14.00">/ key &lt;&#45; master(key)</text>
+</g>
+<!-- WaitVersionAfterPHD.Processing&#45;&gt;NotSkippingPrefixNorVersions.Processing -->
+<g id="edge18" class="edge">
+<title>WaitVersionAfterPHD.Processing&#45;&gt;NotSkippingPrefixNorVersions.Processing</title>
+<path fill="none" stroke="black" d="M1546.51,-21.25C1677.94,-26.54 1888.29,-44.09 1937.37,-102 1947.71,-114.21 1946.85,-125.11 1937.37,-138 1841.62,-268.08 1749.48,-218.23 1590.37,-246 1471.26,-266.79 1143.92,-275.5 947.77,-278.94"/>
+<polygon fill="black" stroke="black" points="947.6,-275.44 937.66,-279.11 947.72,-282.44 947.6,-275.44"/>
+<text text-anchor="middle" x="2124.87" y="-116.3" font-family="Times,serif" font-size="14.00">[not isVersionKey(key) or master(key) != PHDkey]</text>
+</g>
+</g>
+</svg>

documentation/listingAlgos/pics/delimiterStateChart.dot

@@ -0,0 +1,35 @@
+digraph {
+  node [shape="box",style="filled,rounded",fontsize=16,fixedsize=true,width=3];
+  edge [fontsize=14];
+  rankdir=TB;
+
+  START [shape="circle",width=0.2,label="",style="filled",fillcolor="black"]
+  END [shape="circle",width=0.2,label="",style="filled",fillcolor="black",peripheries=2]
+
+  node [fillcolor="lightgrey"];
+  "NotSkipping.Idle" [label="NotSkipping",group="NotSkipping"];
+  "NeverSkipping.Idle" [label="NeverSkipping",group="NeverSkipping"];
+  "NotSkippingPrefix.Idle" [label="NotSkippingPrefix",group="NotSkippingPrefix"];
+  "SkippingPrefix.Idle" [label="SkippingPrefix",group="SkippingPrefix"];
+
+  node [fillcolor="lightblue"];
+  "NeverSkipping.Processing" [label="NeverSkipping",group="NeverSkipping"];
+  "NotSkippingPrefix.Processing" [label="NotSkippingPrefix",group="NotSkippingPrefix"];
+  "SkippingPrefix.Processing" [label="SkippingPrefix",group="SkippingPrefix"];
+
+  START -> "NotSkipping.Idle"
+  "NotSkipping.Idle" -> "NeverSkipping.Idle" [label="[delimiter == undefined]"]
+  "NotSkipping.Idle" -> "NotSkippingPrefix.Idle" [label="[delimiter == '/']"]
+
+  "NeverSkipping.Idle" -> "NeverSkipping.Processing" [label="filter(key, value)"]
+  "NotSkippingPrefix.Idle" -> "NotSkippingPrefix.Processing" [label="filter(key, value)"]
+  "SkippingPrefix.Idle" -> "SkippingPrefix.Processing" [label="filter(key, value)"]
+
+  "NeverSkipping.Processing" -> END [label="[nKeys == maxKeys]\n-> FILTER_END"]
+  "NeverSkipping.Processing" -> "NeverSkipping.Idle" [label="[nKeys < maxKeys]\n/ Contents.append(key, value)\n -> FILTER_ACCEPT"]
+  "NotSkippingPrefix.Processing" -> END [label="[nKeys == maxKeys]\n -> FILTER_END"]
+  "NotSkippingPrefix.Processing" -> "SkippingPrefix.Idle" [label="[nKeys < maxKeys and hasDelimiter(key)]\n/ prefix <- prefixOf(key)\n/ CommonPrefixes.append(prefixOf(key))\n-> FILTER_ACCEPT"]
+  "NotSkippingPrefix.Processing" -> "NotSkippingPrefix.Idle" [label="[nKeys < maxKeys and not hasDelimiter(key)]\n/ Contents.append(key, value)\n -> FILTER_ACCEPT"]
+  "SkippingPrefix.Processing" -> "SkippingPrefix.Idle" [label="[key.startsWith(prefix)]\n-> FILTER_SKIP"]
+  "SkippingPrefix.Processing" -> "NotSkippingPrefix.Processing" [label="[not key.startsWith(prefix)]"]
+}

documentation/listingAlgos/pics/delimiterStateChart.svg

@@ -0,0 +1,166 @@
+<?xml version="1.0" encoding="UTF-8" standalone="no"?>
+<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN"
+ "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">
+<!-- Generated by graphviz version 2.43.0 (0)
+ -->
+<!-- Title: %3 Pages: 1 -->
+<svg width="975pt" height="533pt"
+ viewBox="0.00 0.00 975.00 533.00" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
+<g id="graph0" class="graph" transform="scale(1 1) rotate(0) translate(4 529)">
+<title>%3</title>
+<polygon fill="white" stroke="transparent" points="-4,4 -4,-529 971,-529 971,4 -4,4"/>
+<!-- START -->
+<g id="node1" class="node">
+<title>START</title>
+<ellipse fill="black" stroke="black" cx="283" cy="-518" rx="7" ry="7"/>
+</g>
+<!-- NotSkipping.Idle -->
+<g id="node3" class="node">
+<title>NotSkipping.Idle</title>
+<path fill="lightgrey" stroke="black" d="M379,-474C379,-474 187,-474 187,-474 181,-474 175,-468 175,-462 175,-462 175,-450 175,-450 175,-444 181,-438 187,-438 187,-438 379,-438 379,-438 385,-438 391,-444 391,-450 391,-450 391,-462 391,-462 391,-468 385,-474 379,-474"/>
+<text text-anchor="middle" x="283" y="-452.2" font-family="Times,serif" font-size="16.00">NotSkipping</text>
+</g>
+<!-- START&#45;&gt;NotSkipping.Idle -->
+<g id="edge1" class="edge">
+<title>START&#45;&gt;NotSkipping.Idle</title>
+<path fill="none" stroke="black" d="M283,-510.58C283,-504.23 283,-494.07 283,-484.3"/>
+<polygon fill="black" stroke="black" points="286.5,-484.05 283,-474.05 279.5,-484.05 286.5,-484.05"/>
+</g>
+<!-- END -->
+<g id="node2" class="node">
+<title>END</title>
+<ellipse fill="black" stroke="black" cx="196" cy="-120" rx="7" ry="7"/>
+<ellipse fill="none" stroke="black" cx="196" cy="-120" rx="11" ry="11"/>
+</g>
+<!-- NeverSkipping.Idle -->
+<g id="node4" class="node">
+<title>NeverSkipping.Idle</title>
+<path fill="lightgrey" stroke="black" d="M262,-387C262,-387 70,-387 70,-387 64,-387 58,-381 58,-375 58,-375 58,-363 58,-363 58,-357 64,-351 70,-351 70,-351 262,-351 262,-351 268,-351 274,-357 274,-363 274,-363 274,-375 274,-375 274,-381 268,-387 262,-387"/>
+<text text-anchor="middle" x="166" y="-365.2" font-family="Times,serif" font-size="16.00">NeverSkipping</text>
+</g>
+<!-- NotSkipping.Idle&#45;&gt;NeverSkipping.Idle -->
+<g id="edge2" class="edge">
+<title>NotSkipping.Idle&#45;&gt;NeverSkipping.Idle</title>
+<path fill="none" stroke="black" d="M216.5,-437.82C206.51,-433.18 196.91,-427.34 189,-420 182.25,-413.74 177.33,-405.11 173.81,-396.79"/>
+<polygon fill="black" stroke="black" points="177.05,-395.47 170.3,-387.31 170.49,-397.9 177.05,-395.47"/>
+<text text-anchor="middle" x="279.5" y="-408.8" font-family="Times,serif" font-size="14.00">[delimiter == undefined]</text>
+</g>
+<!-- NotSkippingPrefix.Idle -->
+<g id="node5" class="node">
+<title>NotSkippingPrefix.Idle</title>
+<path fill="lightgrey" stroke="black" d="M496,-387C496,-387 304,-387 304,-387 298,-387 292,-381 292,-375 292,-375 292,-363 292,-363 292,-357 298,-351 304,-351 304,-351 496,-351 496,-351 502,-351 508,-357 508,-363 508,-363 508,-375 508,-375 508,-381 502,-387 496,-387"/>
+<text text-anchor="middle" x="400" y="-365.2" font-family="Times,serif" font-size="16.00">NotSkippingPrefix</text>
+</g>
+<!-- NotSkipping.Idle&#45;&gt;NotSkippingPrefix.Idle -->
+<g id="edge3" class="edge">
+<title>NotSkipping.Idle&#45;&gt;NotSkippingPrefix.Idle</title>
+<path fill="none" stroke="black" d="M340.77,-437.93C351.2,-433.2 361.45,-427.29 370,-420 377.58,-413.53 383.76,-404.65 388.51,-396.16"/>
+<polygon fill="black" stroke="black" points="391.63,-397.74 393.08,-387.24 385.4,-394.54 391.63,-397.74"/>
+<text text-anchor="middle" x="442.5" y="-408.8" font-family="Times,serif" font-size="14.00">[delimiter == &#39;/&#39;]</text>
+</g>
+<!-- NeverSkipping.Processing -->
+<g id="node7" class="node">
+<title>NeverSkipping.Processing</title>
+<path fill="lightblue" stroke="black" d="M204,-270C204,-270 12,-270 12,-270 6,-270 0,-264 0,-258 0,-258 0,-246 0,-246 0,-240 6,-234 12,-234 12,-234 204,-234 204,-234 210,-234 216,-240 216,-246 216,-246 216,-258 216,-258 216,-264 210,-270 204,-270"/>
+<text text-anchor="middle" x="108" y="-248.2" font-family="Times,serif" font-size="16.00">NeverSkipping</text>
+</g>
+<!-- NeverSkipping.Idle&#45;&gt;NeverSkipping.Processing -->
+<g id="edge4" class="edge">
+<title>NeverSkipping.Idle&#45;&gt;NeverSkipping.Processing</title>
+<path fill="none" stroke="black" d="M64.1,-350.93C47.33,-346.11 33.58,-340.17 28,-333 15.72,-317.21 17.05,-304.74 28,-288 30.93,-283.52 34.58,-279.6 38.69,-276.19"/>
+<polygon fill="black" stroke="black" points="40.97,-278.86 47.1,-270.22 36.92,-273.16 40.97,-278.86"/>
+<text text-anchor="middle" x="86" y="-306.8" font-family="Times,serif" font-size="14.00">filter(key, value)</text>
+</g>
+<!-- NotSkippingPrefix.Processing -->
+<g id="node8" class="node">
+<title>NotSkippingPrefix.Processing</title>
+<path fill="lightblue" stroke="black" d="M554,-270C554,-270 362,-270 362,-270 356,-270 350,-264 350,-258 350,-258 350,-246 350,-246 350,-240 356,-234 362,-234 362,-234 554,-234 554,-234 560,-234 566,-240 566,-246 566,-246 566,-258 566,-258 566,-264 560,-270 554,-270"/>
+<text text-anchor="middle" x="458" y="-248.2" font-family="Times,serif" font-size="16.00">NotSkippingPrefix</text>
+</g>
+<!-- NotSkippingPrefix.Idle&#45;&gt;NotSkippingPrefix.Processing -->
+<g id="edge5" class="edge">
+<title>NotSkippingPrefix.Idle&#45;&gt;NotSkippingPrefix.Processing</title>
+<path fill="none" stroke="black" d="M395.69,-350.84C392.38,-333.75 390.03,-307.33 401,-288 403.42,-283.74 406.58,-279.94 410.19,-276.55"/>
+<polygon fill="black" stroke="black" points="412.5,-279.18 418.1,-270.18 408.11,-273.73 412.5,-279.18"/>
+<text text-anchor="middle" x="459" y="-306.8" font-family="Times,serif" font-size="14.00">filter(key, value)</text>
+</g>
+<!-- SkippingPrefix.Idle -->
+<g id="node6" class="node">
+<title>SkippingPrefix.Idle</title>
+<path fill="lightgrey" stroke="black" d="M554,-138C554,-138 362,-138 362,-138 356,-138 350,-132 350,-126 350,-126 350,-114 350,-114 350,-108 356,-102 362,-102 362,-102 554,-102 554,-102 560,-102 566,-108 566,-114 566,-114 566,-126 566,-126 566,-132 560,-138 554,-138"/>
+<text text-anchor="middle" x="458" y="-116.2" font-family="Times,serif" font-size="16.00">SkippingPrefix</text>
+</g>
+<!-- SkippingPrefix.Processing -->
+<g id="node9" class="node">
+<title>SkippingPrefix.Processing</title>
+<path fill="lightblue" stroke="black" d="M691,-36C691,-36 499,-36 499,-36 493,-36 487,-30 487,-24 487,-24 487,-12 487,-12 487,-6 493,0 499,0 499,0 691,0 691,0 697,0 703,-6 703,-12 703,-12 703,-24 703,-24 703,-30 697,-36 691,-36"/>
+<text text-anchor="middle" x="595" y="-14.2" font-family="Times,serif" font-size="16.00">SkippingPrefix</text>
+</g>
+<!-- SkippingPrefix.Idle&#45;&gt;SkippingPrefix.Processing -->
+<g id="edge6" class="edge">
+<title>SkippingPrefix.Idle&#45;&gt;SkippingPrefix.Processing</title>
+<path fill="none" stroke="black" d="M452.35,-101.95C448.76,-87.65 446.54,-67.45 457,-54 461.44,-48.29 471.08,-43.36 483.3,-39.15"/>
+<polygon fill="black" stroke="black" points="484.61,-42.41 493.1,-36.07 482.51,-35.73 484.61,-42.41"/>
+<text text-anchor="middle" x="515" y="-65.3" font-family="Times,serif" font-size="14.00">filter(key, value)</text>
+</g>
+<!-- NeverSkipping.Processing&#45;&gt;END -->
+<g id="edge7" class="edge">
+<title>NeverSkipping.Processing&#45;&gt;END</title>
+<path fill="none" stroke="black" d="M102.91,-233.88C97.93,-213.45 93.18,-179.15 109,-156 123.79,-134.35 154.41,-126.09 175.08,-122.94"/>
+<polygon fill="black" stroke="black" points="175.62,-126.4 185.11,-121.69 174.76,-119.45 175.62,-126.4"/>
+<text text-anchor="middle" x="185" y="-189.8" font-family="Times,serif" font-size="14.00">[nKeys == maxKeys]</text>
+<text text-anchor="middle" x="185" y="-174.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_END</text>
+</g>
+<!-- NeverSkipping.Processing&#45;&gt;NeverSkipping.Idle -->
+<g id="edge8" class="edge">
+<title>NeverSkipping.Processing&#45;&gt;NeverSkipping.Idle</title>
+<path fill="none" stroke="black" d="M129.49,-270.27C134.87,-275.48 140.18,-281.55 144,-288 153.56,-304.17 159.09,-324.63 162.21,-340.81"/>
+<polygon fill="black" stroke="black" points="158.78,-341.49 163.94,-350.74 165.68,-340.29 158.78,-341.49"/>
+<text text-anchor="middle" x="265.5" y="-321.8" font-family="Times,serif" font-size="14.00">[nKeys &lt; maxKeys]</text>
+<text text-anchor="middle" x="265.5" y="-306.8" font-family="Times,serif" font-size="14.00">/ Contents.append(key, value)</text>
+<text text-anchor="middle" x="265.5" y="-291.8" font-family="Times,serif" font-size="14.00"> &#45;&gt; FILTER_ACCEPT</text>
+</g>
+<!-- NotSkippingPrefix.Processing&#45;&gt;END -->
+<g id="edge9" class="edge">
+<title>NotSkippingPrefix.Processing&#45;&gt;END</title>
+<path fill="none" stroke="black" d="M349.96,-237.93C333,-232.81 316.36,-225.74 302,-216 275.27,-197.87 285.01,-177.6 261,-156 247.64,-143.98 229.41,-134.62 215.65,-128.62"/>
+<polygon fill="black" stroke="black" points="216.74,-125.28 206.16,-124.7 214.07,-131.75 216.74,-125.28"/>
+<text text-anchor="middle" x="378" y="-189.8" font-family="Times,serif" font-size="14.00">[nKeys == maxKeys]</text>
+<text text-anchor="middle" x="378" y="-174.8" font-family="Times,serif" font-size="14.00"> &#45;&gt; FILTER_END</text>
+</g>
+<!-- NotSkippingPrefix.Processing&#45;&gt;NotSkippingPrefix.Idle -->
+<g id="edge11" class="edge">
+<title>NotSkippingPrefix.Processing&#45;&gt;NotSkippingPrefix.Idle</title>
+<path fill="none" stroke="black" d="M499.64,-270.11C506.59,-274.86 512.87,-280.76 517,-288 526.9,-305.38 528.94,-316.96 517,-333 513.56,-337.62 509.53,-341.66 505.07,-345.18"/>
+<polygon fill="black" stroke="black" points="502.89,-342.43 496.63,-350.98 506.85,-348.2 502.89,-342.43"/>
+<text text-anchor="middle" x="690.5" y="-321.8" font-family="Times,serif" font-size="14.00">[nKeys &lt; maxKeys and not hasDelimiter(key)]</text>
+<text text-anchor="middle" x="690.5" y="-306.8" font-family="Times,serif" font-size="14.00">/ Contents.append(key, value)</text>
+<text text-anchor="middle" x="690.5" y="-291.8" font-family="Times,serif" font-size="14.00"> &#45;&gt; FILTER_ACCEPT</text>
+</g>
+<!-- NotSkippingPrefix.Processing&#45;&gt;SkippingPrefix.Idle -->
+<g id="edge10" class="edge">
+<title>NotSkippingPrefix.Processing&#45;&gt;SkippingPrefix.Idle</title>
+<path fill="none" stroke="black" d="M458,-233.74C458,-211.98 458,-174.32 458,-148.56"/>
+<polygon fill="black" stroke="black" points="461.5,-148.33 458,-138.33 454.5,-148.33 461.5,-148.33"/>
+<text text-anchor="middle" x="609.5" y="-204.8" font-family="Times,serif" font-size="14.00">[nKeys &lt; maxKeys and hasDelimiter(key)]</text>
+<text text-anchor="middle" x="609.5" y="-189.8" font-family="Times,serif" font-size="14.00">/ prefix &lt;&#45; prefixOf(key)</text>
+<text text-anchor="middle" x="609.5" y="-174.8" font-family="Times,serif" font-size="14.00">/ CommonPrefixes.append(prefixOf(key))</text>
+<text text-anchor="middle" x="609.5" y="-159.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_ACCEPT</text>
+</g>
+<!-- SkippingPrefix.Processing&#45;&gt;SkippingPrefix.Idle -->
+<g id="edge12" class="edge">
+<title>SkippingPrefix.Processing&#45;&gt;SkippingPrefix.Idle</title>
+<path fill="none" stroke="black" d="M593.49,-36.23C591.32,-50.84 586,-71.39 573,-84 567.75,-89.09 561.77,-93.45 555.38,-97.17"/>
+<polygon fill="black" stroke="black" points="553.66,-94.12 546.43,-101.87 556.91,-100.32 553.66,-94.12"/>
+<text text-anchor="middle" x="672" y="-72.8" font-family="Times,serif" font-size="14.00">[key.startsWith(prefix)]</text>
+<text text-anchor="middle" x="672" y="-57.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_SKIP</text>
+</g>
+<!-- SkippingPrefix.Processing&#45;&gt;NotSkippingPrefix.Processing -->
+<g id="edge13" class="edge">
+<title>SkippingPrefix.Processing&#45;&gt;NotSkippingPrefix.Processing</title>
+<path fill="none" stroke="black" d="M703.16,-31.64C728.6,-36.87 750.75,-44.11 759,-54 778.46,-77.34 776.26,-200.01 762,-216 749.37,-230.17 656.13,-239.42 576.2,-244.84"/>
+<polygon fill="black" stroke="black" points="575.77,-241.36 566.03,-245.51 576.24,-248.34 575.77,-241.36"/>
+<text text-anchor="middle" x="870" y="-116.3" font-family="Times,serif" font-size="14.00">[not key.startsWith(prefix)]</text>
+</g>
+</g>
+</svg>

documentation/listingAlgos/pics/delimiterVersionsStateChart.dot

@@ -0,0 +1,50 @@
+digraph {
+  node [shape="box",style="filled,rounded",fontsize=16,fixedsize=true,width=3];
+  edge [fontsize=14];
+  rankdir=TB;
+
+  START [shape="circle",width=0.2,label="",style="filled",fillcolor="black"]
+  END [shape="circle",width=0.2,label="",style="filled",fillcolor="black",peripheries=2]
+
+  node [fillcolor="lightgrey"];
+  "NotSkipping.Idle" [label="NotSkipping",group="NotSkipping",width=4];
+  "SkippingPrefix.Idle" [label="SkippingPrefix",group="SkippingPrefix"];
+  "WaitForNullKey.Idle" [label="WaitForNullKey",group="WaitForNullKey"];
+  "SkippingVersions.Idle" [label="SkippingVersions",group="SkippingVersions"];
+
+  node [fillcolor="lightblue"];
+  "NotSkipping.Processing" [label="NotSkipping",group="NotSkipping",width=4];
+  "NotSkippingV0.Processing" [label="NotSkippingV0",group="NotSkipping",width=4];
+  "NotSkippingV1.Processing" [label="NotSkippingV1",group="NotSkipping",width=4];
+  "NotSkippingCommon.Processing" [label="NotSkippingCommon",group="NotSkipping",width=4];
+  "SkippingPrefix.Processing" [label="SkippingPrefix",group="SkippingPrefix"];
+  "WaitForNullKey.Processing" [label="WaitForNullKey",group="WaitForNullKey"];
+  "SkippingVersions.Processing" [label="SkippingVersions",group="SkippingVersions"];
+
+  START -> "WaitForNullKey.Idle" [label="[versionIdMarker != undefined]"]
+  START -> "NotSkipping.Idle" [label="[versionIdMarker == undefined]"]
+
+  "NotSkipping.Idle" -> "NotSkipping.Processing" [label="filter(key, value)"]
+  "SkippingPrefix.Idle" -> "SkippingPrefix.Processing" [label="filter(key, value)"]
+  "WaitForNullKey.Idle" -> "WaitForNullKey.Processing" [label="filter(key, value)"]
+  "SkippingVersions.Idle" -> "SkippingVersions.Processing" [label="filter(key, value)"]
+
+  "NotSkipping.Processing" -> "NotSkippingV0.Processing" [label="vFormat='v0'"]
+  "NotSkipping.Processing" -> "NotSkippingV1.Processing" [label="vFormat='v1'"]
+
+  "WaitForNullKey.Processing" -> "NotSkipping.Processing" [label="master(key) != keyMarker"]
+  "WaitForNullKey.Processing" -> "SkippingVersions.Processing" [label="master(key) == keyMarker"]
+  "NotSkippingV0.Processing" -> "SkippingPrefix.Idle" [label="[key.startsWith(<ReplayPrefix>)]\n/ prefix <- <ReplayPrefix>\n-> FILTER_SKIP"]
+  "NotSkippingV0.Processing" -> "NotSkipping.Idle" [label="[Version.isPHD(value)]\n-> FILTER_ACCEPT"]
+  "NotSkippingV0.Processing" -> "NotSkippingCommon.Processing" [label="[not key.startsWith(<ReplayPrefix>)\nand not Version.isPHD(value)]"]
+  "NotSkippingV1.Processing" -> "NotSkippingCommon.Processing" [label="[always]"]
+  "NotSkippingCommon.Processing" -> END [label="[isListableKey(key, value) and\nKeys == maxKeys]\n-> FILTER_END"]
+  "NotSkippingCommon.Processing" -> "SkippingPrefix.Idle" [label="[isListableKey(key, value) and\nnKeys < maxKeys and\nhasDelimiter(key)]\n/ prefix <- prefixOf(key)\n/ CommonPrefixes.append(prefixOf(key))\n-> FILTER_ACCEPT"]
+  "NotSkippingCommon.Processing" -> "NotSkipping.Idle" [label="[isListableKey(key, value) and\nnKeys < maxKeys and\nnot hasDelimiter(key)]\n/ Contents.append(key, versionId, value)\n-> FILTER_ACCEPT"]
+
+  "SkippingPrefix.Processing" -> "SkippingPrefix.Idle" [label="[key.startsWith(prefix)]\n-> FILTER_SKIP"]
+  "SkippingPrefix.Processing" -> "NotSkipping.Processing" [label="[not key.startsWith(prefix)]"]
+  "SkippingVersions.Processing" -> "NotSkipping.Processing" [label="master(key) !== keyMarker or \nversionId > versionIdMarker"]
+  "SkippingVersions.Processing" -> "SkippingVersions.Idle" [label="master(key) === keyMarker and \nversionId < versionIdMarker\n-> FILTER_SKIP"]
+  "SkippingVersions.Processing" -> "SkippingVersions.Idle" [label="master(key) === keyMarker and \nversionId == versionIdMarker\n-> FILTER_ACCEPT"]
+}

documentation/listingAlgos/pics/delimiterVersionsStateChart.svg

@@ -0,0 +1,265 @@
+<?xml version="1.0" encoding="UTF-8" standalone="no"?>
+<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN"
+ "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">
+<!-- Generated by graphviz version 2.43.0 (0)
+ -->
+<!-- Title: %3 Pages: 1 -->
+<svg width="1522pt" height="922pt"
+ viewBox="0.00 0.00 1522.26 922.00" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
+<g id="graph0" class="graph" transform="scale(1 1) rotate(0) translate(4 918)">
+<title>%3</title>
+<polygon fill="white" stroke="transparent" points="-4,4 -4,-918 1518.26,-918 1518.26,4 -4,4"/>
+<!-- START -->
+<g id="node1" class="node">
+<title>START</title>
+<ellipse fill="black" stroke="black" cx="393.26" cy="-907" rx="7" ry="7"/>
+</g>
+<!-- NotSkipping.Idle -->
+<g id="node3" class="node">
+<title>NotSkipping.Idle</title>
+<path fill="lightgrey" stroke="black" d="M436.26,-675C436.26,-675 172.26,-675 172.26,-675 166.26,-675 160.26,-669 160.26,-663 160.26,-663 160.26,-651 160.26,-651 160.26,-645 166.26,-639 172.26,-639 172.26,-639 436.26,-639 436.26,-639 442.26,-639 448.26,-645 448.26,-651 448.26,-651 448.26,-663 448.26,-663 448.26,-669 442.26,-675 436.26,-675"/>
+<text text-anchor="middle" x="304.26" y="-653.2" font-family="Times,serif" font-size="16.00">NotSkipping</text>
+</g>
+<!-- START&#45;&gt;NotSkipping.Idle -->
+<g id="edge2" class="edge">
+<title>START&#45;&gt;NotSkipping.Idle</title>
+<path fill="none" stroke="black" d="M391.06,-899.87C380.45,-870.31 334.26,-741.58 313.93,-684.93"/>
+<polygon fill="black" stroke="black" points="317.12,-683.46 310.45,-675.23 310.53,-685.82 317.12,-683.46"/>
+<text text-anchor="middle" x="470.76" y="-783.8" font-family="Times,serif" font-size="14.00">[versionIdMarker == undefined]</text>
+</g>
+<!-- WaitForNullKey.Idle -->
+<g id="node5" class="node">
+<title>WaitForNullKey.Idle</title>
+<path fill="lightgrey" stroke="black" d="M692.26,-849C692.26,-849 500.26,-849 500.26,-849 494.26,-849 488.26,-843 488.26,-837 488.26,-837 488.26,-825 488.26,-825 488.26,-819 494.26,-813 500.26,-813 500.26,-813 692.26,-813 692.26,-813 698.26,-813 704.26,-819 704.26,-825 704.26,-825 704.26,-837 704.26,-837 704.26,-843 698.26,-849 692.26,-849"/>
+<text text-anchor="middle" x="596.26" y="-827.2" font-family="Times,serif" font-size="16.00">WaitForNullKey</text>
+</g>
+<!-- START&#45;&gt;WaitForNullKey.Idle -->
+<g id="edge1" class="edge">
+<title>START&#45;&gt;WaitForNullKey.Idle</title>
+<path fill="none" stroke="black" d="M399.56,-903.7C420.56,-896.05 489.7,-870.85 540.08,-852.48"/>
+<polygon fill="black" stroke="black" points="541.38,-855.73 549.57,-849.02 538.98,-849.16 541.38,-855.73"/>
+<text text-anchor="middle" x="608.76" y="-870.8" font-family="Times,serif" font-size="14.00">[versionIdMarker != undefined]</text>
+</g>
+<!-- END -->
+<g id="node2" class="node">
+<title>END</title>
+<ellipse fill="black" stroke="black" cx="45.26" cy="-120" rx="7" ry="7"/>
+<ellipse fill="none" stroke="black" cx="45.26" cy="-120" rx="11" ry="11"/>
+</g>
+<!-- NotSkipping.Processing -->
+<g id="node7" class="node">
+<title>NotSkipping.Processing</title>
+<path fill="lightblue" stroke="black" d="M761.26,-558C761.26,-558 497.26,-558 497.26,-558 491.26,-558 485.26,-552 485.26,-546 485.26,-546 485.26,-534 485.26,-534 485.26,-528 491.26,-522 497.26,-522 497.26,-522 761.26,-522 761.26,-522 767.26,-522 773.26,-528 773.26,-534 773.26,-534 773.26,-546 773.26,-546 773.26,-552 767.26,-558 761.26,-558"/>
+<text text-anchor="middle" x="629.26" y="-536.2" font-family="Times,serif" font-size="16.00">NotSkipping</text>
+</g>
+<!-- NotSkipping.Idle&#45;&gt;NotSkipping.Processing -->
+<g id="edge3" class="edge">
+<title>NotSkipping.Idle&#45;&gt;NotSkipping.Processing</title>
+<path fill="none" stroke="black" d="M333.17,-638.98C364.86,-620.99 417.68,-592.92 466.26,-576 483.64,-569.95 502.44,-564.74 520.88,-560.34"/>
+<polygon fill="black" stroke="black" points="521.83,-563.71 530.78,-558.04 520.25,-556.89 521.83,-563.71"/>
+<text text-anchor="middle" x="524.26" y="-594.8" font-family="Times,serif" font-size="14.00">filter(key, value)</text>
+</g>
+<!-- SkippingPrefix.Idle -->
+<g id="node4" class="node">
+<title>SkippingPrefix.Idle</title>
+<path fill="lightgrey" stroke="black" d="M662.26,-138C662.26,-138 470.26,-138 470.26,-138 464.26,-138 458.26,-132 458.26,-126 458.26,-126 458.26,-114 458.26,-114 458.26,-108 464.26,-102 470.26,-102 470.26,-102 662.26,-102 662.26,-102 668.26,-102 674.26,-108 674.26,-114 674.26,-114 674.26,-126 674.26,-126 674.26,-132 668.26,-138 662.26,-138"/>
+<text text-anchor="middle" x="566.26" y="-116.2" font-family="Times,serif" font-size="16.00">SkippingPrefix</text>
+</g>
+<!-- SkippingPrefix.Processing -->
+<g id="node11" class="node">
+<title>SkippingPrefix.Processing</title>
+<path fill="lightblue" stroke="black" d="M779.26,-36C779.26,-36 587.26,-36 587.26,-36 581.26,-36 575.26,-30 575.26,-24 575.26,-24 575.26,-12 575.26,-12 575.26,-6 581.26,0 587.26,0 587.26,0 779.26,0 779.26,0 785.26,0 791.26,-6 791.26,-12 791.26,-12 791.26,-24 791.26,-24 791.26,-30 785.26,-36 779.26,-36"/>
+<text text-anchor="middle" x="683.26" y="-14.2" font-family="Times,serif" font-size="16.00">SkippingPrefix</text>
+</g>
+<!-- SkippingPrefix.Idle&#45;&gt;SkippingPrefix.Processing -->
+<g id="edge4" class="edge">
+<title>SkippingPrefix.Idle&#45;&gt;SkippingPrefix.Processing</title>
+<path fill="none" stroke="black" d="M552.64,-101.74C543.31,-87.68 534.41,-67.95 545.26,-54 549.71,-48.29 559.34,-43.36 571.56,-39.15"/>
+<polygon fill="black" stroke="black" points="572.87,-42.41 581.36,-36.07 570.77,-35.73 572.87,-42.41"/>
+<text text-anchor="middle" x="603.26" y="-65.3" font-family="Times,serif" font-size="14.00">filter(key, value)</text>
+</g>
+<!-- WaitForNullKey.Processing -->
+<g id="node12" class="node">
+<title>WaitForNullKey.Processing</title>
+<path fill="lightblue" stroke="black" d="M692.26,-762C692.26,-762 500.26,-762 500.26,-762 494.26,-762 488.26,-756 488.26,-750 488.26,-750 488.26,-738 488.26,-738 488.26,-732 494.26,-726 500.26,-726 500.26,-726 692.26,-726 692.26,-726 698.26,-726 704.26,-732 704.26,-738 704.26,-738 704.26,-750 704.26,-750 704.26,-756 698.26,-762 692.26,-762"/>
+<text text-anchor="middle" x="596.26" y="-740.2" font-family="Times,serif" font-size="16.00">WaitForNullKey</text>
+</g>
+<!-- WaitForNullKey.Idle&#45;&gt;WaitForNullKey.Processing -->
+<g id="edge5" class="edge">
+<title>WaitForNullKey.Idle&#45;&gt;WaitForNullKey.Processing</title>
+<path fill="none" stroke="black" d="M596.26,-812.8C596.26,-801.16 596.26,-785.55 596.26,-772.24"/>
+<polygon fill="black" stroke="black" points="599.76,-772.18 596.26,-762.18 592.76,-772.18 599.76,-772.18"/>
+<text text-anchor="middle" x="654.26" y="-783.8" font-family="Times,serif" font-size="14.00">filter(key, value)</text>
+</g>
+<!-- SkippingVersions.Idle -->
+<g id="node6" class="node">
+<title>SkippingVersions.Idle</title>
+<path fill="lightgrey" stroke="black" d="M1241.26,-558C1241.26,-558 1049.26,-558 1049.26,-558 1043.26,-558 1037.26,-552 1037.26,-546 1037.26,-546 1037.26,-534 1037.26,-534 1037.26,-528 1043.26,-522 1049.26,-522 1049.26,-522 1241.26,-522 1241.26,-522 1247.26,-522 1253.26,-528 1253.26,-534 1253.26,-534 1253.26,-546 1253.26,-546 1253.26,-552 1247.26,-558 1241.26,-558"/>
+<text text-anchor="middle" x="1145.26" y="-536.2" font-family="Times,serif" font-size="16.00">SkippingVersions</text>
+</g>
+<!-- SkippingVersions.Processing -->
+<g id="node13" class="node">
+<title>SkippingVersions.Processing</title>
+<path fill="lightblue" stroke="black" d="M1241.26,-675C1241.26,-675 1049.26,-675 1049.26,-675 1043.26,-675 1037.26,-669 1037.26,-663 1037.26,-663 1037.26,-651 1037.26,-651 1037.26,-645 1043.26,-639 1049.26,-639 1049.26,-639 1241.26,-639 1241.26,-639 1247.26,-639 1253.26,-645 1253.26,-651 1253.26,-651 1253.26,-663 1253.26,-663 1253.26,-669 1247.26,-675 1241.26,-675"/>
+<text text-anchor="middle" x="1145.26" y="-653.2" font-family="Times,serif" font-size="16.00">SkippingVersions</text>
+</g>
+<!-- SkippingVersions.Idle&#45;&gt;SkippingVersions.Processing -->
+<g id="edge6" class="edge">
+<title>SkippingVersions.Idle&#45;&gt;SkippingVersions.Processing</title>
+<path fill="none" stroke="black" d="M1145.26,-558.25C1145.26,-576.77 1145.26,-606.45 1145.26,-628.25"/>
+<polygon fill="black" stroke="black" points="1141.76,-628.53 1145.26,-638.53 1148.76,-628.53 1141.76,-628.53"/>
+<text text-anchor="middle" x="1203.26" y="-594.8" font-family="Times,serif" font-size="14.00">filter(key, value)</text>
+</g>
+<!-- NotSkippingV0.Processing -->
+<g id="node8" class="node">
+<title>NotSkippingV0.Processing</title>
+<path fill="lightblue" stroke="black" d="M436.26,-411C436.26,-411 172.26,-411 172.26,-411 166.26,-411 160.26,-405 160.26,-399 160.26,-399 160.26,-387 160.26,-387 160.26,-381 166.26,-375 172.26,-375 172.26,-375 436.26,-375 436.26,-375 442.26,-375 448.26,-381 448.26,-387 448.26,-387 448.26,-399 448.26,-399 448.26,-405 442.26,-411 436.26,-411"/>
+<text text-anchor="middle" x="304.26" y="-389.2" font-family="Times,serif" font-size="16.00">NotSkippingV0</text>
+</g>
+<!-- NotSkipping.Processing&#45;&gt;NotSkippingV0.Processing -->
+<g id="edge7" class="edge">
+<title>NotSkipping.Processing&#45;&gt;NotSkippingV0.Processing</title>
+<path fill="none" stroke="black" d="M573.96,-521.95C558.07,-516.64 540.84,-510.46 525.26,-504 460.22,-477.02 387.62,-439.36 343.97,-415.84"/>
+<polygon fill="black" stroke="black" points="345.57,-412.72 335.11,-411.04 342.24,-418.88 345.57,-412.72"/>
+<text text-anchor="middle" x="573.76" y="-462.8" font-family="Times,serif" font-size="14.00">vFormat=&#39;v0&#39;</text>
+</g>
+<!-- NotSkippingV1.Processing -->
+<g id="node9" class="node">
+<title>NotSkippingV1.Processing</title>
+<path fill="lightblue" stroke="black" d="M758.26,-411C758.26,-411 494.26,-411 494.26,-411 488.26,-411 482.26,-405 482.26,-399 482.26,-399 482.26,-387 482.26,-387 482.26,-381 488.26,-375 494.26,-375 494.26,-375 758.26,-375 758.26,-375 764.26,-375 770.26,-381 770.26,-387 770.26,-387 770.26,-399 770.26,-399 770.26,-405 764.26,-411 758.26,-411"/>
+<text text-anchor="middle" x="626.26" y="-389.2" font-family="Times,serif" font-size="16.00">NotSkippingV1</text>
+</g>
+<!-- NotSkipping.Processing&#45;&gt;NotSkippingV1.Processing -->
+<g id="edge8" class="edge">
+<title>NotSkipping.Processing&#45;&gt;NotSkippingV1.Processing</title>
+<path fill="none" stroke="black" d="M628.91,-521.8C628.39,-496.94 627.44,-450.74 626.83,-421.23"/>
+<polygon fill="black" stroke="black" points="630.32,-421.11 626.62,-411.18 623.33,-421.25 630.32,-421.11"/>
+<text text-anchor="middle" x="676.76" y="-462.8" font-family="Times,serif" font-size="14.00">vFormat=&#39;v1&#39;</text>
+</g>
+<!-- NotSkippingV0.Processing&#45;&gt;NotSkipping.Idle -->
+<g id="edge12" class="edge">
+<title>NotSkippingV0.Processing&#45;&gt;NotSkipping.Idle</title>
+<path fill="none" stroke="black" d="M304.26,-411.25C304.26,-455.74 304.26,-574.61 304.26,-628.62"/>
+<polygon fill="black" stroke="black" points="300.76,-628.81 304.26,-638.81 307.76,-628.81 300.76,-628.81"/>
+<text text-anchor="middle" x="385.76" y="-543.8" font-family="Times,serif" font-size="14.00">[Version.isPHD(value)]</text>
+<text text-anchor="middle" x="385.76" y="-528.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_ACCEPT</text>
+</g>
+<!-- NotSkippingV0.Processing&#45;&gt;SkippingPrefix.Idle -->
+<g id="edge11" class="edge">
+<title>NotSkippingV0.Processing&#45;&gt;SkippingPrefix.Idle</title>
+<path fill="none" stroke="black" d="M448.41,-376.93C508.52,-369.95 565.63,-362.09 570.26,-357 622.9,-299.12 594.8,-196.31 577.11,-147.78"/>
+<polygon fill="black" stroke="black" points="580.33,-146.4 573.53,-138.28 573.78,-148.87 580.33,-146.4"/>
+<text text-anchor="middle" x="720.26" y="-297.8" font-family="Times,serif" font-size="14.00">[key.startsWith(&lt;ReplayPrefix&gt;)]</text>
+<text text-anchor="middle" x="720.26" y="-282.8" font-family="Times,serif" font-size="14.00">/ prefix &lt;&#45; &lt;ReplayPrefix&gt;</text>
+<text text-anchor="middle" x="720.26" y="-267.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_SKIP</text>
+</g>
+<!-- NotSkippingCommon.Processing -->
+<g id="node10" class="node">
+<title>NotSkippingCommon.Processing</title>
+<path fill="lightblue" stroke="black" d="M436.26,-304.5C436.26,-304.5 172.26,-304.5 172.26,-304.5 166.26,-304.5 160.26,-298.5 160.26,-292.5 160.26,-292.5 160.26,-280.5 160.26,-280.5 160.26,-274.5 166.26,-268.5 172.26,-268.5 172.26,-268.5 436.26,-268.5 436.26,-268.5 442.26,-268.5 448.26,-274.5 448.26,-280.5 448.26,-280.5 448.26,-292.5 448.26,-292.5 448.26,-298.5 442.26,-304.5 436.26,-304.5"/>
+<text text-anchor="middle" x="304.26" y="-282.7" font-family="Times,serif" font-size="16.00">NotSkippingCommon</text>
+</g>
+<!-- NotSkippingV0.Processing&#45;&gt;NotSkippingCommon.Processing -->
+<g id="edge13" class="edge">
+<title>NotSkippingV0.Processing&#45;&gt;NotSkippingCommon.Processing</title>
+<path fill="none" stroke="black" d="M304.26,-374.74C304.26,-358.48 304.26,-333.85 304.26,-314.9"/>
+<polygon fill="black" stroke="black" points="307.76,-314.78 304.26,-304.78 300.76,-314.78 307.76,-314.78"/>
+<text text-anchor="middle" x="435.26" y="-345.8" font-family="Times,serif" font-size="14.00">[not key.startsWith(&lt;ReplayPrefix&gt;)</text>
+<text text-anchor="middle" x="435.26" y="-330.8" font-family="Times,serif" font-size="14.00">and not Version.isPHD(value)]</text>
+</g>
+<!-- NotSkippingV1.Processing&#45;&gt;NotSkippingCommon.Processing -->
+<g id="edge14" class="edge">
+<title>NotSkippingV1.Processing&#45;&gt;NotSkippingCommon.Processing</title>
+<path fill="none" stroke="black" d="M616.43,-374.83C606.75,-359.62 590.48,-338.14 570.26,-327 549.98,-315.83 505.48,-307.38 458.57,-301.23"/>
+<polygon fill="black" stroke="black" points="458.9,-297.74 448.53,-299.95 458.01,-304.69 458.9,-297.74"/>
+<text text-anchor="middle" x="632.26" y="-338.3" font-family="Times,serif" font-size="14.00">[always]</text>
+</g>
+<!-- NotSkippingCommon.Processing&#45;&gt;END -->
+<g id="edge15" class="edge">
+<title>NotSkippingCommon.Processing&#45;&gt;END</title>
+<path fill="none" stroke="black" d="M159.92,-279.56C109.8,-274.24 62.13,-264.33 46.26,-246 20.92,-216.72 30.42,-167.54 38.5,-140.42"/>
+<polygon fill="black" stroke="black" points="41.94,-141.16 41.67,-130.57 35.27,-139.02 41.94,-141.16"/>
+<text text-anchor="middle" x="152.76" y="-212.3" font-family="Times,serif" font-size="14.00">[isListableKey(key, value) and</text>
+<text text-anchor="middle" x="152.76" y="-197.3" font-family="Times,serif" font-size="14.00">Keys == maxKeys]</text>
+<text text-anchor="middle" x="152.76" y="-182.3" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_END</text>
+</g>
+<!-- NotSkippingCommon.Processing&#45;&gt;NotSkipping.Idle -->
+<g id="edge17" class="edge">
+<title>NotSkippingCommon.Processing&#45;&gt;NotSkipping.Idle</title>
+<path fill="none" stroke="black" d="M214.74,-304.54C146.51,-322.73 57.06,-358.99 13.26,-429 -49.27,-528.95 128.43,-602.49 233.32,-635.95"/>
+<polygon fill="black" stroke="black" points="232.34,-639.31 242.93,-638.97 234.43,-632.63 232.34,-639.31"/>
+<text text-anchor="middle" x="156.76" y="-492.8" font-family="Times,serif" font-size="14.00">[isListableKey(key, value) and</text>
+<text text-anchor="middle" x="156.76" y="-477.8" font-family="Times,serif" font-size="14.00">nKeys &lt; maxKeys and</text>
+<text text-anchor="middle" x="156.76" y="-462.8" font-family="Times,serif" font-size="14.00">not hasDelimiter(key)]</text>
+<text text-anchor="middle" x="156.76" y="-447.8" font-family="Times,serif" font-size="14.00">/ Contents.append(key, versionId, value)</text>
+<text text-anchor="middle" x="156.76" y="-432.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_ACCEPT</text>
+</g>
+<!-- NotSkippingCommon.Processing&#45;&gt;SkippingPrefix.Idle -->
+<g id="edge16" class="edge">
+<title>NotSkippingCommon.Processing&#45;&gt;SkippingPrefix.Idle</title>
+<path fill="none" stroke="black" d="M292.14,-268.23C288.18,-261.59 284.27,-253.75 282.26,-246 272.21,-207.28 255.76,-185.96 282.26,-156 293.6,-143.18 374.98,-134.02 447.74,-128.3"/>
+<polygon fill="black" stroke="black" points="448.24,-131.77 457.94,-127.51 447.7,-124.79 448.24,-131.77"/>
+<text text-anchor="middle" x="428.26" y="-234.8" font-family="Times,serif" font-size="14.00">[isListableKey(key, value) and</text>
+<text text-anchor="middle" x="428.26" y="-219.8" font-family="Times,serif" font-size="14.00">nKeys &lt; maxKeys and</text>
+<text text-anchor="middle" x="428.26" y="-204.8" font-family="Times,serif" font-size="14.00">hasDelimiter(key)]</text>
+<text text-anchor="middle" x="428.26" y="-189.8" font-family="Times,serif" font-size="14.00">/ prefix &lt;&#45; prefixOf(key)</text>
+<text text-anchor="middle" x="428.26" y="-174.8" font-family="Times,serif" font-size="14.00">/ CommonPrefixes.append(prefixOf(key))</text>
+<text text-anchor="middle" x="428.26" y="-159.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_ACCEPT</text>
+</g>
+<!-- SkippingPrefix.Processing&#45;&gt;SkippingPrefix.Idle -->
+<g id="edge18" class="edge">
+<title>SkippingPrefix.Processing&#45;&gt;SkippingPrefix.Idle</title>
+<path fill="none" stroke="black" d="M681.57,-36.04C679.28,-50.54 673.9,-71.03 661.26,-84 656.4,-88.99 650.77,-93.28 644.72,-96.95"/>
+<polygon fill="black" stroke="black" points="642.71,-94.06 635.6,-101.92 646.05,-100.21 642.71,-94.06"/>
+<text text-anchor="middle" x="759.26" y="-72.8" font-family="Times,serif" font-size="14.00">[key.startsWith(prefix)]</text>
+<text text-anchor="middle" x="759.26" y="-57.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_SKIP</text>
+</g>
+<!-- SkippingPrefix.Processing&#45;&gt;NotSkipping.Processing -->
+<g id="edge19" class="edge">
+<title>SkippingPrefix.Processing&#45;&gt;NotSkipping.Processing</title>
+<path fill="none" stroke="black" d="M791.46,-33.51C815.84,-38.71 837.21,-45.46 846.26,-54 868.07,-74.57 864.26,-89.02 864.26,-119 864.26,-394 864.26,-394 864.26,-394 864.26,-462.4 791.27,-499.6 726.64,-519.12"/>
+<polygon fill="black" stroke="black" points="725.39,-515.84 716.77,-521.99 727.35,-522.56 725.39,-515.84"/>
+<text text-anchor="middle" x="961.26" y="-282.8" font-family="Times,serif" font-size="14.00">[not key.startsWith(prefix)]</text>
+</g>
+<!-- WaitForNullKey.Processing&#45;&gt;NotSkipping.Processing -->
+<g id="edge9" class="edge">
+<title>WaitForNullKey.Processing&#45;&gt;NotSkipping.Processing</title>
+<path fill="none" stroke="black" d="M599.08,-725.78C604.81,-690.67 617.89,-610.59 624.8,-568.31"/>
+<polygon fill="black" stroke="black" points="628.3,-568.61 626.46,-558.18 621.39,-567.48 628.3,-568.61"/>
+<text text-anchor="middle" x="707.26" y="-653.3" font-family="Times,serif" font-size="14.00">master(key) != keyMarker</text>
+</g>
+<!-- WaitForNullKey.Processing&#45;&gt;SkippingVersions.Processing -->
+<g id="edge10" class="edge">
+<title>WaitForNullKey.Processing&#45;&gt;SkippingVersions.Processing</title>
+<path fill="none" stroke="black" d="M704.4,-726.26C797.32,-711.87 931.09,-691.16 1026.87,-676.33"/>
+<polygon fill="black" stroke="black" points="1027.55,-679.77 1036.89,-674.78 1026.47,-672.85 1027.55,-679.77"/>
+<text text-anchor="middle" x="1001.26" y="-696.8" font-family="Times,serif" font-size="14.00">master(key) == keyMarker</text>
+</g>
+<!-- SkippingVersions.Processing&#45;&gt;SkippingVersions.Idle -->
+<g id="edge21" class="edge">
+<title>SkippingVersions.Processing&#45;&gt;SkippingVersions.Idle</title>
+<path fill="none" stroke="black" d="M1241.89,-638.98C1249.74,-634.29 1256.75,-628.4 1262.26,-621 1274.21,-604.96 1274.21,-592.04 1262.26,-576 1258.82,-571.38 1254.79,-567.34 1250.33,-563.82"/>
+<polygon fill="black" stroke="black" points="1252.11,-560.8 1241.89,-558.02 1248.15,-566.57 1252.11,-560.8"/>
+<text text-anchor="middle" x="1392.26" y="-609.8" font-family="Times,serif" font-size="14.00">master(key) === keyMarker and </text>
+<text text-anchor="middle" x="1392.26" y="-594.8" font-family="Times,serif" font-size="14.00">versionId &lt; versionIdMarker</text>
+<text text-anchor="middle" x="1392.26" y="-579.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_SKIP</text>
+</g>
+<!-- SkippingVersions.Processing&#45;&gt;SkippingVersions.Idle -->
+<g id="edge22" class="edge">
+<title>SkippingVersions.Processing&#45;&gt;SkippingVersions.Idle</title>
+<path fill="none" stroke="black" d="M1036.97,-654.38C978.97,-650.96 915.73,-642.25 897.26,-621 884.15,-605.9 884.15,-591.1 897.26,-576 914.65,-555.99 971.71,-547.1 1026.73,-543.28"/>
+<polygon fill="black" stroke="black" points="1027.21,-546.76 1036.97,-542.62 1026.76,-539.77 1027.21,-546.76"/>
+<text text-anchor="middle" x="1019.26" y="-609.8" font-family="Times,serif" font-size="14.00">master(key) === keyMarker and </text>
+<text text-anchor="middle" x="1019.26" y="-594.8" font-family="Times,serif" font-size="14.00">versionId == versionIdMarker</text>
+<text text-anchor="middle" x="1019.26" y="-579.8" font-family="Times,serif" font-size="14.00">&#45;&gt; FILTER_ACCEPT</text>
+</g>
+<!-- SkippingVersions.Processing&#45;&gt;NotSkipping.Processing -->
+<g id="edge20" class="edge">
+<title>SkippingVersions.Processing&#45;&gt;NotSkipping.Processing</title>
+<path fill="none" stroke="black" d="M1037.02,-651.24C897.84,-644.67 672.13,-632.37 657.26,-621 641.04,-608.6 634.18,-586.13 631.3,-568.16"/>
+<polygon fill="black" stroke="black" points="634.76,-567.68 630.02,-558.21 627.82,-568.57 634.76,-567.68"/>
+<text text-anchor="middle" x="770.26" y="-602.3" font-family="Times,serif" font-size="14.00">master(key) !== keyMarker or </text>
+<text text-anchor="middle" x="770.26" y="-587.3" font-family="Times,serif" font-size="14.00">versionId &gt; versionIdMarker</text>
+</g>
+</g>
+</svg>

index.ts

@@ -1,6 +1,9 @@
 import * as evaluators from './lib/policyEvaluator/evaluator';
 import evaluatePrincipal from './lib/policyEvaluator/principal';
-import RequestContext from './lib/policyEvaluator/RequestContext';
+import RequestContext, {
+    actionNeedQuotaCheck,
+    actionNeedQuotaCheckCopy,
+    actionWithDataDeletion } from './lib/policyEvaluator/RequestContext';
 import * as requestUtils from './lib/policyEvaluator/requestUtils';
 import * as actionMaps from './lib/policyEvaluator/utils/actionMaps';
 import { validateUserPolicy } from './lib/policy/policyValidator'
@@ -25,6 +28,7 @@ import * as objectRestore from './lib/s3middleware/objectRestore';
 import * as lifecycleHelpers from './lib/s3middleware/lifecycleHelpers';
 export { default as errors } from './lib/errors';
 export { default as Clustering } from './lib/Clustering';
+export * as ClusterRPC from './lib/clustering/ClusterRPC';
 export * as ipCheck from './lib/ipCheck';
 export * as auth from './lib/auth/auth';
 export * as constants from './lib/constants';
@@ -48,12 +52,15 @@ export const algorithms = {
         Skip: require('./lib/algos/list/skip'),
     },
     cache: {
+        GapSet: require('./lib/algos/cache/GapSet'),
+        GapCache: require('./lib/algos/cache/GapCache'),
         LRUCache: require('./lib/algos/cache/LRUCache'),
     },
     stream: {
         MergeStream: require('./lib/algos/stream/MergeStream'),
     },
     SortedSet: require('./lib/algos/set/SortedSet'),
+    Heap: require('./lib/algos/heap/Heap'),
 };
 
 export const policies = {
@@ -63,6 +70,9 @@ export const policies = {
     RequestContext,
     requestUtils,
     actionMaps,
+    actionNeedQuotaCheck,
+    actionWithDataDeletion,
+    actionNeedQuotaCheckCopy,
 };
 
 export const testing = {

lib/algos/cache/GapCache.ts

@@ -0,0 +1,363 @@
+import { OrderedSet } from '@js-sdsl/ordered-set';
+import {
+    default as GapSet,
+    GapSetEntry,
+} from './GapSet';
+
+// the API is similar but is not strictly a superset of GapSetInterface
+// so we don't extend from it
+export interface GapCacheInterface {
+    exposureDelayMs: number;
+    maxGapWeight: number;
+    size: number;
+
+    setGap: (firstKey: string, lastKey: string, weight: number) => void;
+    removeOverlappingGaps: (overlappingKeys: string[]) => number;
+    lookupGap: (minKey: string, maxKey?: string) => Promise<GapSetEntry | null>;
+    [Symbol.iterator]: () => Iterator<GapSetEntry>;
+    toArray: () => GapSetEntry[];
+};
+
+class GapCacheUpdateSet {
+    newGaps: GapSet;
+    updatedKeys: OrderedSet<string>;
+
+    constructor(maxGapWeight: number) {
+        this.newGaps = new GapSet(maxGapWeight);
+        this.updatedKeys = new OrderedSet();
+    }
+
+    addUpdateBatch(updatedKeys: OrderedSet<string>): void {
+        this.updatedKeys.union(updatedKeys);
+    }
+};
+
+/**
+ * Cache of listing "gaps" i.e. ranges of keys that can be skipped
+ * over during listing (because they only contain delete markers as
+ * latest versions).
+ *
+ * Typically, a single GapCache instance would be attached to a raft session.
+ *
+ * The API usage is as follows:
+ *
+ * - Initialize a GapCache instance by calling start() (this starts an internal timer)
+ *
+ * - Insert a gap or update an existing one via setGap()
+ *
+ * - Lookup existing gaps via lookupGap()
+ *
+ * - Invalidate gaps that overlap a specific set of keys via removeOverlappingGaps()
+ *
+ * - Shut down a GapCache instance by calling stop() (this stops the internal timer)
+ *
+ * Gaps inserted via setGap() are not exposed immediately to lookupGap(), but only:
+ *
+ * - after a certain delay always larger than 'exposureDelayMs' and usually shorter
+ *   than twice this value (but might be slightly longer in rare cases)
+ *
+ * - and only if they haven't been invalidated by a recent call to removeOverlappingGaps()
+ *
+ * This ensures atomicity between gap creation and invalidation from updates under
+ * the condition that a gap is created from first key to last key within the time defined
+ * by 'exposureDelayMs'.
+ *
+ * The implementation is based on two extra temporary "update sets" on top of the main
+ * exposed gap set, one called "staging" and the other "frozen", each containing a
+ * temporary updated gap set and a list of updated keys to invalidate gaps with (coming
+ * from calls to removeOverlappingGaps()). Every "exposureDelayMs" milliseconds, the frozen
+ * gaps are invalidated by all key updates coming from either of the "staging" or "frozen"
+ * update set, then merged into the exposed gaps set, after which the staging updates become
+ * the frozen updates and won't receive any new gap until the next cycle.
+ */
+export default class GapCache implements GapCacheInterface {
+    _exposureDelayMs: number;
+    maxGaps: number;
+
+    _stagingUpdates: GapCacheUpdateSet;
+    _frozenUpdates: GapCacheUpdateSet;
+    _exposedGaps: GapSet;
+    _exposeFrozenInterval: NodeJS.Timeout | null;
+
+    /**
+     * @constructor
+     *
+     * @param {number} exposureDelayMs - minimum delay between
+     * insertion of a gap via setGap() and its exposure via
+     * lookupGap()
+     * @param {number} maxGaps - maximum number of cached gaps, after
+     * which no new gap can be added by setGap(). (Note: a future
+     * improvement could replace this by an eviction strategy)
+     * @param {number} maxGapWeight - maximum "weight" of individual
+     * cached gaps, which is also the granularity for
+     * invalidation. Individual gaps can be chained together,
+     * which lookupGap() transparently consolidates in the response
+     * into a single large gap.
+     */
+    constructor(exposureDelayMs: number, maxGaps: number, maxGapWeight: number) {
+        this._exposureDelayMs = exposureDelayMs;
+        this.maxGaps = maxGaps;
+
+        this._stagingUpdates = new GapCacheUpdateSet(maxGapWeight);
+        this._frozenUpdates = new GapCacheUpdateSet(maxGapWeight);
+        this._exposedGaps = new GapSet(maxGapWeight);
+        this._exposeFrozenInterval = null;
+    }
+
+    /**
+     * Create a GapCache from an array of exposed gap entries (used in tests)
+     *
+     * @return {GapCache} - a new GapCache instance
+     */
+    static createFromArray(
+        gaps: GapSetEntry[],
+        exposureDelayMs: number,
+        maxGaps: number,
+        maxGapWeight: number
+    ): GapCache {
+        const gapCache = new GapCache(exposureDelayMs, maxGaps, maxGapWeight);
+        gapCache._exposedGaps = GapSet.createFromArray(gaps, maxGapWeight)
+        return gapCache;
+    }
+
+    /**
+     * Internal helper to remove gaps in the staging and frozen sets
+     * overlapping with previously updated keys, right before the
+     * frozen gaps get exposed.
+     *
+     * @return {undefined}
+     */
+    _removeOverlappingGapsBeforeExpose(): void {
+        for (const { updatedKeys } of [this._stagingUpdates, this._frozenUpdates]) {
+            if (updatedKeys.size() === 0) {
+                continue;
+            }
+            for (const { newGaps } of [this._stagingUpdates, this._frozenUpdates]) {
+                if (newGaps.size === 0) {
+                    continue;
+                }
+                newGaps.removeOverlappingGaps(updatedKeys);
+            }
+        }
+    }
+
+    /**
+     * This function is the core mechanism that updates the exposed gaps in the
+     * cache. It is called on a regular interval defined by 'exposureDelayMs'.
+     *
+     * It does the following in order:
+     *
+     * - remove gaps from the frozen set that overlap with any key present in a
+     *   batch passed to removeOverlappingGaps() since the last two triggers of
+     *   _exposeFrozen()
+     *
+     * - merge the remaining gaps from the frozen set to the exposed set, which
+     *   makes them visible from calls to lookupGap()
+     *
+     * - rotate by freezing the currently staging updates and initiating a new
+     *   staging updates set
+     *
+     * @return {undefined}
+     */
+    _exposeFrozen(): void {
+        this._removeOverlappingGapsBeforeExpose();
+        for (const gap of this._frozenUpdates.newGaps) {
+            // Use a trivial strategy to keep the cache size within
+            // limits: refuse to add new gaps when the size is above
+            // the 'maxGaps' threshold. We solely rely on
+            // removeOverlappingGaps() to make space for new gaps.
+            if (this._exposedGaps.size < this.maxGaps) {
+                this._exposedGaps.setGap(gap.firstKey, gap.lastKey, gap.weight);
+            }
+        }
+        this._frozenUpdates = this._stagingUpdates;
+        this._stagingUpdates = new GapCacheUpdateSet(this.maxGapWeight);
+    }
+
+    /**
+     * Start the internal GapCache timer
+     *
+     * @return {undefined}
+     */
+    start(): void {
+        if (this._exposeFrozenInterval) {
+            return;
+        }
+        this._exposeFrozenInterval = setInterval(
+            () => this._exposeFrozen(),
+            this._exposureDelayMs);
+    }
+
+    /**
+     * Stop the internal GapCache timer
+     *
+     * @return {undefined}
+     */
+    stop(): void {
+        if (this._exposeFrozenInterval) {
+            clearInterval(this._exposeFrozenInterval);
+            this._exposeFrozenInterval = null;
+        }
+    }
+
+    /**
+     * Record a gap between two keys, associated with a weight to
+     * limit individual gap's spanning ranges in the cache, for a more
+     * granular invalidation.
+     *
+     * The function handles splitting and merging existing gaps to
+     * maintain an optimal weight of cache entries.
+     *
+     * NOTE 1: the caller must ensure that the full length of the gap
+     * between 'firstKey' and 'lastKey' has been built from a listing
+     * snapshot that is more recent than 'exposureDelayMs' milliseconds,
+     * in order to guarantee that the exposed gap will be fully
+     * covered (and potentially invalidated) from recent calls to
+     * removeOverlappingGaps().
+     *
+     * NOTE 2: a usual pattern when building a large gap from multiple
+     * calls to setGap() is to start the next gap from 'lastKey',
+     * which will be passed as 'firstKey' in the next call, so that
+     * gaps can be chained together and consolidated by lookupGap().
+     *
+     * @param {string} firstKey - first key of the gap
+     * @param {string} lastKey - last key of the gap, must be greater
+     * or equal than 'firstKey'
+     * @param {number} weight - total weight between 'firstKey' and 'lastKey'
+     * @return {undefined}
+     */
+    setGap(firstKey: string, lastKey: string, weight: number): void {
+        this._stagingUpdates.newGaps.setGap(firstKey, lastKey, weight);
+    }
+
+    /**
+     * Remove gaps that overlap with a given set of keys. Used to
+     * invalidate gaps when keys are inserted or deleted.
+     *
+     * @param {OrderedSet<string> | string[]} overlappingKeys - remove gaps that
+     * overlap with any of this set of keys
+     * @return {number} - how many gaps were removed from the exposed
+     * gaps only (overlapping gaps not yet exposed are also invalidated
+     * but are not accounted for in the returned value)
+     */
+    removeOverlappingGaps(overlappingKeys: OrderedSet<string> | string[]): number {
+        let overlappingKeysSet;
+        if (Array.isArray(overlappingKeys)) {
+            overlappingKeysSet = new OrderedSet(overlappingKeys);
+        } else {
+            overlappingKeysSet = overlappingKeys;
+        }
+        this._stagingUpdates.addUpdateBatch(overlappingKeysSet);
+        return this._exposedGaps.removeOverlappingGaps(overlappingKeysSet);
+    }
+
+    /**
+     * Lookup the next exposed gap that overlaps with [minKey, maxKey]. Internally
+     * chained gaps are coalesced in the response into a single contiguous large gap.
+     *
+     * @param {string} minKey - minimum key overlapping with the returned gap
+     * @param {string} [maxKey] - maximum key overlapping with the returned gap
+     * @return {Promise<GapSetEntry | null>} - result of the lookup if a gap
+     *   was found, null otherwise, as a Promise
+     */
+    lookupGap(minKey: string, maxKey?: string): Promise<GapSetEntry | null> {
+        return this._exposedGaps.lookupGap(minKey, maxKey);
+    }
+
+    /**
+     * Get the maximum weight setting for individual gaps.
+     *
+     * @return {number} - maximum weight of individual gaps
+     */
+    get maxGapWeight(): number {
+        return this._exposedGaps.maxWeight;
+    }
+
+    /**
+     * Set the maximum weight setting for individual gaps.
+     *
+     * @param {number} gapWeight - maximum weight of individual gaps
+     */
+    set maxGapWeight(gapWeight: number) {
+        this._exposedGaps.maxWeight = gapWeight;
+        // also update transient gap sets
+        this._stagingUpdates.newGaps.maxWeight = gapWeight;
+        this._frozenUpdates.newGaps.maxWeight = gapWeight;
+    }
+
+    /**
+     * Get the exposure delay in milliseconds, which is the minimum
+     * time after which newly cached gaps will be exposed by
+     * lookupGap().
+     *
+     * @return {number} - exposure delay in milliseconds
+     */
+    get exposureDelayMs(): number {
+        return this._exposureDelayMs;
+    }
+
+    /**
+     * Set the exposure delay in milliseconds, which is the minimum
+     * time after which newly cached gaps will be exposed by
+     * lookupGap(). Setting this attribute automatically updates the
+     * internal state to honor the new value.
+     *
+     * @param {number} - exposure delay in milliseconds
+     */
+    set exposureDelayMs(exposureDelayMs: number) {
+        if (exposureDelayMs !== this._exposureDelayMs) {
+            this._exposureDelayMs = exposureDelayMs;
+            if (this._exposeFrozenInterval) {
+                // invalidate all pending gap updates, as the new interval may not be
+                // safe for them
+                this._stagingUpdates = new GapCacheUpdateSet(this.maxGapWeight);
+                this._frozenUpdates = new GapCacheUpdateSet(this.maxGapWeight);
+
+                // reinitialize the _exposeFrozenInterval timer with the updated delay
+                this.stop();
+                this.start();
+            }
+        }
+    }
+
+    /**
+     * Get the number of exposed gaps
+     *
+     * @return {number} number of exposed gaps
+     */
+    get size(): number {
+        return this._exposedGaps.size;
+    }
+
+    /**
+     * Iterate over exposed gaps
+     *
+     * @return {Iterator<GapSetEntry>} an iterator over exposed gaps
+     */
+    [Symbol.iterator](): Iterator<GapSetEntry> {
+        return this._exposedGaps[Symbol.iterator]();
+    }
+
+    /**
+     * Get an array of all exposed gaps
+     *
+     * @return {GapSetEntry[]} array of exposed gaps
+     */
+    toArray(): GapSetEntry[] {
+        return this._exposedGaps.toArray();
+    }
+
+    /**
+     * Clear all exposed and staging gaps from the cache.
+     *
+     * Note: retains invalidating updates from removeOverlappingGaps()
+     * for correctness of gaps inserted afterwards.
+     *
+     * @return {undefined}
+     */
+    clear(): void {
+        this._stagingUpdates.newGaps = new GapSet(this.maxGapWeight);
+        this._frozenUpdates.newGaps = new GapSet(this.maxGapWeight);
+        this._exposedGaps = new GapSet(this.maxGapWeight);
+    }
+}

lib/algos/cache/GapSet.ts

@@ -0,0 +1,366 @@
+import assert from 'assert';
+import { OrderedSet } from '@js-sdsl/ordered-set';
+
+import errors from '../../errors';
+
+export type GapSetEntry = {
+    firstKey: string,
+    lastKey: string,
+    weight: number,
+};
+
+export interface GapSetInterface {
+    maxWeight: number;
+    size: number;
+
+    setGap: (firstKey: string, lastKey: string, weight: number) => GapSetEntry;
+    removeOverlappingGaps: (overlappingKeys: string[]) => number;
+    lookupGap: (minKey: string, maxKey?: string) => Promise<GapSetEntry | null>;
+    [Symbol.iterator]: () => Iterator<GapSetEntry>;
+    toArray: () => GapSetEntry[];
+};
+
+/**
+ * Specialized data structure to support caching of listing "gaps",
+ * i.e. ranges of keys that can be skipped over during listing
+ * (because they only contain delete markers as latest versions)
+ */
+export default class GapSet implements GapSetInterface, Iterable<GapSetEntry> {
+    _gaps: OrderedSet<GapSetEntry>;
+    _maxWeight: number;
+
+    /**
+     * @constructor
+
+     * @param {number} maxWeight - weight threshold for each cached
+     * gap (unitless). Triggers splitting gaps when reached
+     */
+    constructor(maxWeight: number) {
+        this._gaps = new OrderedSet(
+            [],
+            (left: GapSetEntry, right: GapSetEntry) => (
+                left.firstKey < right.firstKey ? -1 :
+                left.firstKey > right.firstKey ? 1 : 0
+            )
+        );
+        this._maxWeight = maxWeight;
+    }
+
+    /**
+     * Create a GapSet from an array of gap entries (used in tests)
+     */
+    static createFromArray(gaps: GapSetEntry[], maxWeight: number): GapSet {
+        const gapSet = new GapSet(maxWeight);
+        for (const gap of gaps) {
+            gapSet._gaps.insert(gap);
+        }
+        return gapSet;
+    }
+
+    /**
+     * Record a gap between two keys, associated with a weight to limit
+     * individual gap sizes in the cache.
+     *
+     * The function handles splitting and merging existing gaps to
+     * maintain an optimal weight of cache entries.
+     *
+     * @param {string} firstKey - first key of the gap
+     * @param {string} lastKey - last key of the gap, must be greater
+     * or equal than 'firstKey'
+     * @param {number} weight - total weight between 'firstKey' and 'lastKey'
+     * @return {GapSetEntry} - existing or new gap entry
+     */
+    setGap(firstKey: string, lastKey: string, weight: number): GapSetEntry {
+        assert(lastKey >= firstKey);
+
+        // Step 1/4: Find the closest left-overlapping gap, and either re-use it
+        // or chain it with a new gap depending on the weights if it exists (otherwise
+        // just creates a new gap).
+        const curGapIt = this._gaps.reverseLowerBound(<GapSetEntry>{ firstKey });
+        let curGap;
+        if (curGapIt.isAccessible()) {
+            curGap = curGapIt.pointer;
+            if (curGap.lastKey >= lastKey) {
+                // return fully overlapping gap already cached
+                return curGap;
+            }
+        }
+        let remainingWeight = weight;
+        if (!curGap                         // no previous gap
+            || curGap.lastKey < firstKey    // previous gap not overlapping
+            || (curGap.lastKey === firstKey // previous gap overlapping by one key...
+                && curGap.weight + weight > this._maxWeight) // ...but we can't extend it
+           ) {
+            // create a new gap indexed by 'firstKey'
+            curGap = { firstKey, lastKey: firstKey, weight: 0 };
+            this._gaps.insert(curGap);
+        } else if (curGap.lastKey > firstKey && weight > this._maxWeight) {
+            // previous gap is either fully or partially contained in the new gap
+            // and cannot be extended: substract its weight from the total (heuristic
+            // in case the previous gap doesn't start at 'firstKey', which is the
+            // uncommon case)
+            remainingWeight -= curGap.weight;
+
+            // there may be an existing chained gap starting with the previous gap's
+            // 'lastKey': use it if it exists
+            const chainedGapIt = this._gaps.find(<GapSetEntry>{ firstKey: curGap.lastKey });
+            if (chainedGapIt.isAccessible()) {
+                curGap = chainedGapIt.pointer;
+            } else {
+                // no existing chained gap: chain a new gap to the previous gap
+                curGap = {
+                    firstKey: curGap.lastKey,
+                    lastKey: curGap.lastKey,
+                    weight: 0,
+                };
+                this._gaps.insert(curGap);
+            }
+        }
+        // Step 2/4: Cleanup existing gaps fully included in firstKey -> lastKey, and
+        // aggregate their weights in curGap to define the minimum weight up to the
+        // last merged gap.
+        let nextGap;
+        while (true) {
+            const nextGapIt = this._gaps.upperBound(<GapSetEntry>{ firstKey: curGap.firstKey });
+            nextGap = nextGapIt.isAccessible() && nextGapIt.pointer;
+            // stop the cleanup when no more gap or if the next gap is not fully
+            // included in curGap
+            if (!nextGap || nextGap.lastKey > lastKey) {
+                break;
+            }
+            this._gaps.eraseElementByIterator(nextGapIt);
+            curGap.lastKey = nextGap.lastKey;
+            curGap.weight += nextGap.weight;
+        }
+
+        // Step 3/4: Extend curGap to lastKey, adjusting the weight.
+        // At this point, curGap weight is the minimum weight of the finished gap, save it
+        // for step 4.
+        let minMergedWeight = curGap.weight;
+        if (curGap.lastKey === firstKey && firstKey !== lastKey) {
+            // extend the existing gap by the full amount 'firstKey -> lastKey'
+            curGap.lastKey = lastKey;
+            curGap.weight += remainingWeight;
+        } else if (curGap.lastKey <= lastKey) {
+            curGap.lastKey = lastKey;
+            curGap.weight = remainingWeight;
+        }
+
+        // Step 4/4: Find the closest right-overlapping gap, and if it exists, either merge
+        // it or chain it with curGap depending on the weights.
+        if (nextGap && nextGap.firstKey <= lastKey) {
+            // nextGap overlaps with the new gap: check if we can merge it
+            minMergedWeight += nextGap.weight;
+            let mergedWeight;
+            if (lastKey === nextGap.firstKey) {
+                // nextGap is chained with curGap: add the full weight of nextGap
+                mergedWeight = curGap.weight + nextGap.weight;
+            } else {
+                // strict overlap: don't add nextGap's weight unless
+                // it's larger than the sum of merged ranges (as it is
+                // then included in `minMergedWeight`)
+                mergedWeight = Math.max(curGap.weight, minMergedWeight);
+            }
+            if (mergedWeight <= this._maxWeight) {
+                // merge nextGap into curGap
+                curGap.lastKey = nextGap.lastKey;
+                curGap.weight = mergedWeight;
+                this._gaps.eraseElementByKey(nextGap);
+            } else {
+                // adjust the last key to chain with nextGap and substract the next
+                // gap's weight from curGap (heuristic)
+                curGap.lastKey = nextGap.firstKey;
+                curGap.weight = Math.max(mergedWeight - nextGap.weight, 0);
+                curGap = nextGap;
+            }
+        }
+        // return a copy of curGap
+        return Object.assign({}, curGap);
+    }
+
+    /**
+     * Remove gaps that overlap with one or more keys in a given array or
+     * OrderedSet. Used to invalidate gaps when keys are inserted or deleted.
+     *
+     * @param {OrderedSet<string> | string[]} overlappingKeys - remove gaps that overlap
+     * with any of this set of keys
+     * @return {number} - how many gaps were removed
+     */
+    removeOverlappingGaps(overlappingKeys: OrderedSet<string> | string[]): number {
+        // To optimize processing with a large number of keys and/or gaps, this function:
+        //
+        // 1. converts the overlappingKeys array to a OrderedSet (if not already a OrderedSet)
+        // 2. queries both the gaps set and the overlapping keys set in a loop, which allows:
+        //    - skipping ranges of overlapping keys at once when there is no new overlapping gap
+        //    - skipping ranges of gaps at once when there is no overlapping key
+        //
+        // This way, it is efficient when the number of non-overlapping gaps is large
+        // (which is the most common case in practice).
+
+        let overlappingKeysSet;
+        if (Array.isArray(overlappingKeys)) {
+            overlappingKeysSet = new OrderedSet(overlappingKeys);
+        } else {
+            overlappingKeysSet = overlappingKeys;
+        }
+        const firstKeyIt = overlappingKeysSet.begin();
+        let currentKey = firstKeyIt.isAccessible() && firstKeyIt.pointer;
+        let nRemoved = 0;
+        while (currentKey) {
+            const closestGapIt = this._gaps.reverseUpperBound(<GapSetEntry>{ firstKey: currentKey });
+            if (closestGapIt.isAccessible()) {
+                const closestGap = closestGapIt.pointer;
+                if (currentKey <= closestGap.lastKey) {
+                    // currentKey overlaps closestGap: remove the gap
+                    this._gaps.eraseElementByIterator(closestGapIt);
+                    nRemoved += 1;
+                }
+            }
+            const nextGapIt = this._gaps.lowerBound(<GapSetEntry>{ firstKey: currentKey });
+            if (!nextGapIt.isAccessible()) {
+                // no more gap: we're done
+                return nRemoved;
+            }
+            const nextGap = nextGapIt.pointer;
+            // advance to the last key potentially overlapping with nextGap
+            let currentKeyIt = overlappingKeysSet.reverseLowerBound(nextGap.lastKey);
+            if (currentKeyIt.isAccessible()) {
+                currentKey = currentKeyIt.pointer;
+                if (currentKey >= nextGap.firstKey) {
+                    // currentKey overlaps nextGap: remove the gap
+                    this._gaps.eraseElementByIterator(nextGapIt);
+                    nRemoved += 1;
+                }
+            }
+            // advance to the first key potentially overlapping with another gap
+            currentKeyIt = overlappingKeysSet.lowerBound(nextGap.lastKey);
+            currentKey = currentKeyIt.isAccessible() && currentKeyIt.pointer;
+        }
+        return nRemoved;
+    }
+
+    /**
+     * Internal helper to coalesce multiple chained gaps into a single gap.
+     *
+     * It is only used to construct lookupGap() return values and
+     * doesn't modify the GapSet.
+     *
+     * NOTE: The function may take a noticeable amount of time and CPU
+     * to execute if a large number of chained gaps have to be
+     * coalesced, but it should never take more than a few seconds. In
+     * most cases it should take less than a millisecond. It regularly
+     * yields to the nodejs event loop to avoid blocking it during a
+     * long execution.
+     *
+     * @param {GapSetEntry} firstGap - first gap of the chain to coalesce with
+     * the next ones in the chain
+     * @return {Promise<GapSetEntry>} - a new coalesced entry, as a Promise
+     */
+    _coalesceGapChain(firstGap: GapSetEntry): Promise<GapSetEntry> {
+        return new Promise(resolve => {
+            const coalescedGap: GapSetEntry = Object.assign({}, firstGap);
+            const coalesceGapChainIteration = () => {
+                // efficiency trade-off: 100 iterations of log(N) complexity lookups should
+                // not block the event loop for too long
+                for (let opCounter = 0; opCounter < 100; ++opCounter) {
+                    const chainedGapIt = this._gaps.find(
+                            <GapSetEntry>{ firstKey: coalescedGap.lastKey });
+                    if (!chainedGapIt.isAccessible()) {
+                        // chain is complete
+                        return resolve(coalescedGap);
+                    }
+                    const chainedGap = chainedGapIt.pointer;
+                    if (chainedGap.firstKey === chainedGap.lastKey) {
+                        // found a single-key gap: chain is complete
+                        return resolve(coalescedGap);
+                    }
+                    coalescedGap.lastKey = chainedGap.lastKey;
+                    coalescedGap.weight += chainedGap.weight;
+                }
+                // yield to the event loop before continuing the process
+                // of coalescing the gap chain
+                return process.nextTick(coalesceGapChainIteration);
+            };
+            coalesceGapChainIteration();
+        });
+    }
+
+    /**
+     * Lookup the next gap that overlaps with [minKey, maxKey]. Internally chained
+     * gaps are coalesced in the response into a single contiguous large gap.
+     *
+     * @param {string} minKey - minimum key overlapping with the returned gap
+     * @param {string} [maxKey] - maximum key overlapping with the returned gap
+     * @return {Promise<GapSetEntry | null>} - result of the lookup if a gap
+     *   was found, null otherwise, as a Promise
+     */
+    async lookupGap(minKey: string, maxKey?: string): Promise<GapSetEntry | null> {
+        let firstGap: GapSetEntry | null = null;
+        const minGapIt = this._gaps.reverseLowerBound(<GapSetEntry>{ firstKey: minKey });
+        const minGap = minGapIt.isAccessible() && minGapIt.pointer;
+        if (minGap && minGap.lastKey >= minKey) {
+            firstGap = minGap;
+        } else {
+            const maxGapIt = this._gaps.upperBound(<GapSetEntry>{ firstKey: minKey });
+            const maxGap = maxGapIt.isAccessible() && maxGapIt.pointer;
+            if (maxGap && (maxKey === undefined || maxGap.firstKey <= maxKey)) {
+                firstGap = maxGap;
+            }
+        }
+        if (!firstGap) {
+            return null;
+        }
+        return this._coalesceGapChain(firstGap);
+    }
+
+    /**
+     * Get the maximum weight setting for individual gaps.
+     *
+     * @return {number} - maximum weight of individual gaps
+     */
+    get maxWeight(): number {
+        return this._maxWeight;
+    }
+
+    /**
+     * Set the maximum weight setting for individual gaps.
+     *
+     * @param {number} gapWeight - maximum weight of individual gaps
+     */
+    set maxWeight(gapWeight: number) {
+        this._maxWeight = gapWeight;
+    }
+
+    /**
+     * Get the number of gaps stored in this set.
+     *
+     * @return {number} - number of gaps stored in this set
+     */
+    get size(): number {
+        return this._gaps.size();
+    }
+
+    /**
+     * Iterate over each gap of the set, ordered by first key
+     *
+     * @return {Iterator<GapSetEntry>} - an iterator over all gaps
+     *   Example:
+     *     for (const gap of myGapSet) { ... }
+     */
+    [Symbol.iterator](): Iterator<GapSetEntry> {
+        return this._gaps[Symbol.iterator]();
+    }
+
+    /**
+     * Return an array containing all gaps, ordered by first key
+     *
+     * NOTE: there is a toArray() method in the OrderedSet implementation
+     * but it does not scale well and overflows the stack quickly. This is
+     * why we provide an implementation based on an iterator.
+     *
+     * @return {GapSetEntry[]} - an array containing all gaps
+     */
+    toArray(): GapSetEntry[] {
+        return [...this];
+    }
+}

lib/algos/heap/Heap.ts

@@ -0,0 +1,124 @@
+export enum HeapOrder {
+    Min = -1,
+    Max = 1,
+}
+
+export enum CompareResult {
+    LT = -1,
+    EQ = 0,
+    GT = 1,
+}
+
+export type CompareFunction = (x: any, y: any) => CompareResult;
+
+export class Heap {
+    size: number;
+    _maxSize: number;
+    _order: HeapOrder;
+    _heap: any[];
+    _cmpFn: CompareFunction;
+    
+    constructor(size: number, order: HeapOrder, cmpFn: CompareFunction) {
+        this.size = 0;
+        this._maxSize = size;
+        this._order = order;
+        this._cmpFn = cmpFn;
+        this._heap = new Array<any>(this._maxSize);
+    }
+
+    _parent(i: number): number {
+        return Math.floor((i - 1) / 2);
+    }
+
+    _left(i: number): number {
+        return Math.floor((2 * i) + 1);
+    }
+
+    _right(i: number): number {
+        return Math.floor((2 * i) + 2);
+    }
+
+    _shouldSwap(childIdx: number, parentIdx: number): boolean {
+        return this._cmpFn(this._heap[childIdx], this._heap[parentIdx]) as number === this._order as number;
+    }
+
+    _swap(i: number, j: number) {
+        const tmp = this._heap[i];
+        this._heap[i] = this._heap[j];
+        this._heap[j] = tmp;
+    }
+
+    _heapify(i: number) {
+        const l = this._left(i);
+        const r = this._right(i);
+        let c = i;
+
+        if (l < this.size && this._shouldSwap(l, c)) {
+            c = l;
+        }
+
+        if (r < this.size && this._shouldSwap(r, c)) {
+            c = r;
+        }
+
+        if (c != i) {
+            this._swap(c, i);
+            this._heapify(c);
+        }
+    }
+
+    add(item: any): any {
+        if (this.size >= this._maxSize) {
+            return new Error('Max heap size reached');
+        }
+
+        ++this.size;
+        let c = this.size - 1;
+        this._heap[c] = item;
+
+        while (c > 0) {
+            if (!this._shouldSwap(c, this._parent(c))) {
+                return null;
+            }
+
+            this._swap(c, this._parent(c));
+            c = this._parent(c);
+        }
+
+        return null;
+    };
+
+    remove(): any {
+        if (this.size <= 0) {
+            return null;
+        }
+
+        const ret = this._heap[0];
+        this._heap[0] = this._heap[this.size - 1];
+        this._heapify(0);
+        --this.size;
+
+        return ret;
+    };
+
+    peek(): any {
+        if (this.size <= 0) {
+            return null;
+        }
+
+        return this._heap[0];
+    };
+}
+
+export class MinHeap extends Heap {
+    constructor(size: number, cmpFn: CompareFunction) {
+        super(size, HeapOrder.Min, cmpFn);
+    }
+}
+
+export class MaxHeap extends Heap {
+    constructor(size: number, cmpFn: CompareFunction) {
+        super(size, HeapOrder.Max, cmpFn);
+    }
+} 
+

lib/algos/list/Extension.js

@@ -1,6 +1,6 @@
 'use strict'; // eslint-disable-line strict
 
-const { FILTER_SKIP, SKIP_NONE } = require('./tools');
+const { FILTER_ACCEPT, SKIP_NONE } = require('./tools');
 
 // Use a heuristic to amortize the cost of JSON
 // serialization/deserialization only on largest metadata where the
@@ -92,21 +92,26 @@ class Extension {
      * @param {object} entry - a listing entry from metadata
      *                         expected format: { key, value }
      * @return {number} - result of filtering the entry:
-     *                    > 0: entry is accepted and included in the result
-     *                    = 0: entry is accepted but not included (skipping)
-     *                    < 0: entry is not accepted, listing should finish
+     *                    FILTER_ACCEPT: entry is accepted and may or not be included
+     *                                   in the result
+     *                    FILTER_SKIP: listing may skip directly (with "gte" param) to
+     *                                 the key returned by the skipping() method
+     *                    FILTER_END: the results are complete, listing can be stopped
      */
-    filter(entry) {
-        return entry ? FILTER_SKIP : FILTER_SKIP;
+    filter(/* entry: { key, value } */) {
+        return FILTER_ACCEPT;
     }
 
     /**
-     * Provides the insight into why filter is skipping an entry. This could be
-     * because it is skipping a range of delimited keys or a range of specific
-     * version when doing master version listing.
+     * Provides the next key at which the listing task is allowed to skip to.
+     * This could allow to skip over:
+     * - a key prefix ending with the delimiter
+     * - all remaining versions of an object when doing a current
+     *   versions listing in v0 format
+     * - a cached "gap" of deleted objects when doing a current
+     *   versions listing in v0 format
      *
-     * @return {string} - the insight: a common prefix or a master key,
-     *                                 or SKIP_NONE if there is no insight
+     * @return {string} - the next key at which the listing task is allowed to skip to
      */
     skipping() {
         return SKIP_NONE;

lib/algos/list/MPU.js

@@ -1,7 +1,7 @@
 'use strict'; // eslint-disable-line strict
 
 const { inc, checkLimit, listingParamsMasterKeysV0ToV1,
-    FILTER_END, FILTER_ACCEPT } = require('./tools');
+    FILTER_END, FILTER_ACCEPT, SKIP_NONE } = require('./tools');
 const DEFAULT_MAX_KEYS = 1000;
 const VSConst = require('../../versioning/constants').VersioningConstants;
 const { DbPrefixes, BucketVersioningKeyFormat } = VSConst;
@@ -163,7 +163,7 @@ class MultipartUploads {
     }
 
     skipping() {
-        return '';
+        return SKIP_NONE;
     }
 
     /**

lib/algos/list/basic.js

@@ -2,7 +2,7 @@
 
 const Extension = require('./Extension').default;
 
-const { checkLimit, FILTER_END, FILTER_ACCEPT, FILTER_SKIP } = require('./tools');
+const { checkLimit, FILTER_END, FILTER_ACCEPT } = require('./tools');
 const DEFAULT_MAX_KEYS = 10000;
 
 /**
@@ -91,7 +91,7 @@ class List extends Extension {
      *                     < 0 : listing done
      */
     filter(elem) {
-        // Check first in case of maxkeys <= 0
+        // Check if the result array is full
         if (this.keys >= this.maxKeys) {
             return FILTER_END;
         }
@@ -99,7 +99,7 @@ class List extends Extension {
              this.filterKeyStartsWith !== undefined) &&
             typeof elem === 'object' &&
             !this.customFilter(elem.value)) {
-            return FILTER_SKIP;
+            return FILTER_ACCEPT;
         }
         if (typeof elem === 'object') {
             this.res.push({

lib/algos/list/delimiter.js

@@ -1,274 +0,0 @@
-'use strict'; // eslint-disable-line strict
-
-const Extension = require('./Extension').default;
-const { inc, listingParamsMasterKeysV0ToV1,
-    FILTER_END, FILTER_ACCEPT, FILTER_SKIP } = require('./tools');
-const VSConst = require('../../versioning/constants').VersioningConstants;
-const { DbPrefixes, BucketVersioningKeyFormat } = VSConst;
-
-/**
- * Find the common prefix in the path
- *
- * @param {String} key            - path of the object
- * @param {String} delimiter      - separator
- * @param {Number} delimiterIndex - 'folder' index in the path
- * @return {String}               - CommonPrefix
- */
-function getCommonPrefix(key, delimiter, delimiterIndex) {
-    return key.substring(0, delimiterIndex + delimiter.length);
-}
-
-/**
- * Handle object listing with parameters
- *
- * @prop {String[]} CommonPrefixes     - 'folders' defined by the delimiter
- * @prop {String[]} Contents           - 'files' to list
- * @prop {Boolean} IsTruncated         - truncated listing flag
- * @prop {String|undefined} NextMarker - marker per amazon format
- * @prop {Number} keys                 - count of listed keys
- * @prop {String|undefined} delimiter  - separator per amazon format
- * @prop {String|undefined} prefix     - prefix per amazon format
- * @prop {Number} maxKeys              - number of keys to list
- */
-class Delimiter extends Extension {
-    /**
-     * Create a new Delimiter instance
-     * @constructor
-     * @param {Object}  parameters                     - listing parameters
-     * @param {String}  [parameters.delimiter]         - delimiter per amazon
-     *                                                   format
-     * @param {String}  [parameters.prefix]            - prefix per amazon
-     *                                                   format
-     * @param {String}  [parameters.marker]            - marker per amazon
-     *                                                   format
-     * @param {Number}  [parameters.maxKeys]           - number of keys to list
-     * @param {Boolean} [parameters.v2]                - indicates whether v2
-     *                                                   format
-     * @param {String}  [parameters.startAfter]        - marker per amazon
-     *                                                   format
-     * @param {String}  [parameters.continuationToken] - obfuscated amazon
-     *                                                   token
-     * @param {Boolean} [parameters.alphabeticalOrder] - Either the result is
-     *                                                   alphabetically ordered
-     *                                                   or not
-     * @param {RequestLogger} logger                   - The logger of the
-     *                                                   request
-     * @param {String} [vFormat]                       - versioning key format
-     */
-    constructor(parameters, logger, vFormat) {
-        super(parameters, logger);
-        // original listing parameters
-        this.delimiter = parameters.delimiter;
-        this.prefix = parameters.prefix;
-        this.marker = parameters.marker;
-        this.maxKeys = parameters.maxKeys || 1000;
-        this.startAfter = parameters.startAfter;
-        this.continuationToken = parameters.continuationToken;
-        this.alphabeticalOrder =
-          typeof parameters.alphabeticalOrder !== 'undefined' ?
-              parameters.alphabeticalOrder : true;
-
-        this.vFormat = vFormat || BucketVersioningKeyFormat.v0;
-        // results
-        this.CommonPrefixes = [];
-        this.Contents = [];
-        this.IsTruncated = false;
-        this.NextMarker = parameters.marker;
-        this.NextContinuationToken =
-            parameters.continuationToken || parameters.startAfter;
-
-        this.startMarker = parameters.v2 ? 'startAfter' : 'marker';
-        this.continueMarker = parameters.v2 ? 'continuationToken' : 'marker';
-        this.nextContinueMarker = parameters.v2 ?
-            'NextContinuationToken' : 'NextMarker';
-
-        if (this.delimiter !== undefined &&
-            this[this.nextContinueMarker] !== undefined &&
-            this[this.nextContinueMarker].startsWith(this.prefix || '')) {
-            const nextDelimiterIndex =
-                this[this.nextContinueMarker].indexOf(this.delimiter,
-                    this.prefix ? this.prefix.length : 0);
-            this[this.nextContinueMarker] =
-                this[this.nextContinueMarker].slice(0, nextDelimiterIndex +
-                                        this.delimiter.length);
-        }
-
-        Object.assign(this, {
-            [BucketVersioningKeyFormat.v0]: {
-                genMDParams: this.genMDParamsV0,
-                getObjectKey: this.getObjectKeyV0,
-                skipping: this.skippingV0,
-            },
-            [BucketVersioningKeyFormat.v1]: {
-                genMDParams: this.genMDParamsV1,
-                getObjectKey: this.getObjectKeyV1,
-                skipping: this.skippingV1,
-            },
-        }[this.vFormat]);
-    }
-
-    genMDParamsV0() {
-        const params = {};
-        if (this.prefix) {
-            params.gte = this.prefix;
-            params.lt = inc(this.prefix);
-        }
-        const startVal = this[this.continueMarker] || this[this.startMarker];
-        if (startVal) {
-            if (params.gte && params.gte > startVal) {
-                return params;
-            }
-            delete params.gte;
-            params.gt = startVal;
-        }
-        return params;
-    }
-
-    genMDParamsV1() {
-        const params = this.genMDParamsV0();
-        return listingParamsMasterKeysV0ToV1(params);
-    }
-
-    /**
-     * check if the max keys count has been reached and set the
-     * final state of the result if it is the case
-     * @return {Boolean} - indicates if the iteration has to stop
-     */
-    _reachedMaxKeys() {
-        if (this.keys >= this.maxKeys) {
-            // In cases of maxKeys <= 0 -> IsTruncated = false
-            this.IsTruncated = this.maxKeys > 0;
-            return true;
-        }
-        return false;
-    }
-
-    /**
-     *  Add a (key, value) tuple to the listing
-     *  Set the NextMarker to the current key
-     *  Increment the keys counter
-     *  @param {String} key   - The key to add
-     *  @param {String} value - The value of the key
-     *  @return {number}      - indicates if iteration should continue
-     */
-    addContents(key, value) {
-        if (this._reachedMaxKeys()) {
-            return FILTER_END;
-        }
-        this.Contents.push({ key, value: this.trimMetadata(value) });
-        this[this.nextContinueMarker] = key;
-        ++this.keys;
-        return FILTER_ACCEPT;
-    }
-
-    getObjectKeyV0(obj) {
-        return obj.key;
-    }
-
-    getObjectKeyV1(obj) {
-        return obj.key.slice(DbPrefixes.Master.length);
-    }
-
-    /**
-     *  Filter to apply on each iteration, based on:
-     *  - prefix
-     *  - delimiter
-     *  - maxKeys
-     *  The marker is being handled directly by levelDB
-     *  @param {Object} obj       - The key and value of the element
-     *  @param {String} obj.key   - The key of the element
-     *  @param {String} obj.value - The value of the element
-     *  @return {number}          - indicates if iteration should continue
-     */
-    filter(obj) {
-        const key = this.getObjectKey(obj);
-        const value = obj.value;
-        if ((this.prefix && !key.startsWith(this.prefix))
-            || (this.alphabeticalOrder
-                && typeof this[this.nextContinueMarker] === 'string'
-                && key <= this[this.nextContinueMarker])) {
-            return FILTER_SKIP;
-        }
-        if (this.delimiter) {
-            const baseIndex = this.prefix ? this.prefix.length : 0;
-            const delimiterIndex = key.indexOf(this.delimiter, baseIndex);
-            if (delimiterIndex === -1) {
-                return this.addContents(key, value);
-            }
-            return this.addCommonPrefix(key, delimiterIndex);
-        }
-        return this.addContents(key, value);
-    }
-
-    /**
-     * Add a Common Prefix in the list
-     * @param {String} key   - object name
-     * @param {Number} index - after prefix starting point
-     * @return {Boolean}     - indicates if iteration should continue
-     */
-    addCommonPrefix(key, index) {
-        const commonPrefix = getCommonPrefix(key, this.delimiter, index);
-        if (this.CommonPrefixes.indexOf(commonPrefix) === -1
-                && this[this.nextContinueMarker] !== commonPrefix) {
-            if (this._reachedMaxKeys()) {
-                return FILTER_END;
-            }
-            this.CommonPrefixes.push(commonPrefix);
-            this[this.nextContinueMarker] = commonPrefix;
-            ++this.keys;
-            return FILTER_ACCEPT;
-        }
-        return FILTER_SKIP;
-    }
-
-    /**
-     * If repd happens to want to skip listing on a bucket in v0
-     * versioning key format, here is an idea.
-     *
-     * @return {string} - the present range (NextMarker) if repd believes
-     *                    that it's enough and should move on
-     */
-    skippingV0() {
-        return this[this.nextContinueMarker];
-    }
-
-    /**
-     * If repd happens to want to skip listing on a bucket in v1
-     * versioning key format, here is an idea.
-     *
-     * @return {string} - the present range (NextMarker) if repd believes
-     *                    that it's enough and should move on
-     */
-    skippingV1() {
-        return DbPrefixes.Master + this[this.nextContinueMarker];
-    }
-
-    /**
-     *  Return an object containing all mandatory fields to use once the
-     *  iteration is done, doesn't show a NextMarker field if the output
-     *  isn't truncated
-     *  @return {Object} - following amazon format
-     */
-    result() {
-        /* NextMarker is only provided when delimiter is used.
-         * specified in v1 listing documentation
-         * http://docs.aws.amazon.com/AmazonS3/latest/API/RESTBucketGET.html
-         */
-        const result = {
-            CommonPrefixes: this.CommonPrefixes,
-            Contents: this.Contents,
-            IsTruncated: this.IsTruncated,
-            Delimiter: this.delimiter,
-        };
-        if (this.parameters.v2) {
-            result.NextContinuationToken = this.IsTruncated
-                ? this.NextContinuationToken : undefined;
-        } else {
-            result.NextMarker = (this.IsTruncated && this.delimiter)
-                ? this.NextMarker : undefined;
-        }
-        return result;
-    }
-}
-
-module.exports = { Delimiter };

lib/algos/list/delimiter.ts

@@ -0,0 +1,356 @@
+'use strict'; // eslint-disable-line strict
+
+const Extension = require('./Extension').default;
+const { inc, listingParamsMasterKeysV0ToV1,
+    FILTER_END, FILTER_ACCEPT, FILTER_SKIP, SKIP_NONE } = require('./tools');
+const VSConst = require('../../versioning/constants').VersioningConstants;
+const { DbPrefixes, BucketVersioningKeyFormat } = VSConst;
+
+export interface FilterState {
+    id: number,
+};
+
+export interface FilterReturnValue {
+    FILTER_ACCEPT,
+    FILTER_SKIP,
+    FILTER_END,
+};
+
+export const enum DelimiterFilterStateId {
+    NotSkipping = 1,
+    SkippingPrefix = 2,
+};
+
+export interface DelimiterFilterState_NotSkipping extends FilterState {
+    id: DelimiterFilterStateId.NotSkipping,
+};
+
+export interface DelimiterFilterState_SkippingPrefix extends FilterState {
+    id: DelimiterFilterStateId.SkippingPrefix,
+    prefix: string;
+};
+
+type KeyHandler = (key: string, value: string) => FilterReturnValue;
+
+export type ResultObject = {
+    CommonPrefixes: string[];
+    Contents: {
+        key: string;
+        value: string;
+    }[];
+    IsTruncated: boolean;
+    Delimiter ?: string;
+    NextMarker ?: string;
+    NextContinuationToken ?: string;
+};
+
+/**
+ * Handle object listing with parameters
+ *
+ * @prop {String[]} CommonPrefixes     - 'folders' defined by the delimiter
+ * @prop {String[]} Contents           - 'files' to list
+ * @prop {Boolean} IsTruncated         - truncated listing flag
+ * @prop {String|undefined} NextMarker - marker per amazon format
+ * @prop {Number} keys                 - count of listed keys
+ * @prop {String|undefined} delimiter  - separator per amazon format
+ * @prop {String|undefined} prefix     - prefix per amazon format
+ * @prop {Number} maxKeys              - number of keys to list
+ */
+export class Delimiter extends Extension {
+
+    state: FilterState;
+    keyHandlers: { [id: number]: KeyHandler };
+
+    /**
+     * Create a new Delimiter instance
+     * @constructor
+     * @param {Object}  parameters                     - listing parameters
+     * @param {String}  [parameters.delimiter]         - delimiter per amazon
+     *                                                   format
+     * @param {String}  [parameters.prefix]            - prefix per amazon
+     *                                                   format
+     * @param {String}  [parameters.marker]            - marker per amazon
+     *                                                   format
+     * @param {Number}  [parameters.maxKeys]           - number of keys to list
+     * @param {Boolean} [parameters.v2]                - indicates whether v2
+     *                                                   format
+     * @param {String}  [parameters.startAfter]        - marker per amazon
+     *                                                   format
+     * @param {String}  [parameters.continuationToken] - obfuscated amazon
+     *                                                   token
+     * @param {RequestLogger} logger                   - The logger of the
+     *                                                   request
+     * @param {String} [vFormat]                       - versioning key format
+     */
+    constructor(parameters, logger, vFormat) {
+        super(parameters, logger);
+        // original listing parameters
+        this.delimiter = parameters.delimiter;
+        this.prefix = parameters.prefix;
+        this.maxKeys = parameters.maxKeys || 1000;
+
+        if (parameters.v2) {
+            this.marker = parameters.continuationToken || parameters.startAfter;
+        } else {
+            this.marker = parameters.marker;
+        }
+        this.nextMarker = this.marker;
+
+        this.vFormat = vFormat || BucketVersioningKeyFormat.v0;
+        // results
+        this.CommonPrefixes = [];
+        this.Contents = [];
+        this.IsTruncated = false;
+        this.keyHandlers = {};
+
+        Object.assign(this, {
+            [BucketVersioningKeyFormat.v0]: {
+                genMDParams: this.genMDParamsV0,
+                getObjectKey: this.getObjectKeyV0,
+                skipping: this.skippingV0,
+            },
+            [BucketVersioningKeyFormat.v1]: {
+                genMDParams: this.genMDParamsV1,
+                getObjectKey: this.getObjectKeyV1,
+                skipping: this.skippingV1,
+            },
+        }[this.vFormat]);
+
+        // if there is a delimiter, we may skip ranges by prefix,
+        // hence using the NotSkippingPrefix flavor that checks the
+        // subprefix up to the delimiter for the NotSkipping state
+        if (this.delimiter) {
+            this.setKeyHandler(
+                DelimiterFilterStateId.NotSkipping,
+                this.keyHandler_NotSkippingPrefix.bind(this));
+        } else {
+            // listing without a delimiter never has to skip over any
+            // prefix -> use NeverSkipping flavor for the NotSkipping
+            // state
+            this.setKeyHandler(
+                DelimiterFilterStateId.NotSkipping,
+                this.keyHandler_NeverSkipping.bind(this));
+        }
+        this.setKeyHandler(
+            DelimiterFilterStateId.SkippingPrefix,
+            this.keyHandler_SkippingPrefix.bind(this));
+
+        this.state = <DelimiterFilterState_NotSkipping> {
+            id: DelimiterFilterStateId.NotSkipping,
+        };
+    }
+
+    genMDParamsV0() {
+        const params: { gt ?: string, gte ?: string, lt ?: string } = {};
+        if (this.prefix) {
+            params.gte = this.prefix;
+            params.lt = inc(this.prefix);
+        }
+        if (this.marker && this.delimiter) {
+            const commonPrefix = this.getCommonPrefix(this.marker);
+            if (commonPrefix) {
+                const afterPrefix = inc(commonPrefix);
+                if (!params.gte || afterPrefix > params.gte) {
+                    params.gte = afterPrefix;
+                }
+            }
+        }
+        if (this.marker && (!params.gte || this.marker >= params.gte)) {
+            delete params.gte;
+            params.gt = this.marker;
+        }
+        return params;
+    }
+
+    genMDParamsV1() {
+        const params = this.genMDParamsV0();
+        return listingParamsMasterKeysV0ToV1(params);
+    }
+
+    /**
+     * check if the max keys count has been reached and set the
+     * final state of the result if it is the case
+     * @return {Boolean} - indicates if the iteration has to stop
+     */
+    _reachedMaxKeys(): boolean {
+        if (this.keys >= this.maxKeys) {
+            // In cases of maxKeys <= 0 -> IsTruncated = false
+            this.IsTruncated = this.maxKeys > 0;
+            return true;
+        }
+        return false;
+    }
+
+    /**
+     *  Add a (key, value) tuple to the listing
+     *  Set the NextMarker to the current key
+     *  Increment the keys counter
+     *  @param {String} key   - The key to add
+     *  @param {String} value - The value of the key
+     *  @return {number}      - indicates if iteration should continue
+     */
+    addContents(key: string, value: string): void {
+        this.Contents.push({ key, value: this.trimMetadata(value) });
+        ++this.keys;
+        this.nextMarker = key;
+    }
+
+    getCommonPrefix(key: string): string | undefined {
+        if (!this.delimiter) {
+            return undefined;
+        }
+        const baseIndex = this.prefix ? this.prefix.length : 0;
+        const delimiterIndex = key.indexOf(this.delimiter, baseIndex);
+        if (delimiterIndex === -1) {
+            return undefined;
+        }
+        return key.substring(0, delimiterIndex + this.delimiter.length);
+    }
+
+    /**
+     * Add a Common Prefix in the list
+     * @param {String} commonPrefix   - common prefix to add
+     * @param {String} key            - full key starting with commonPrefix
+     * @return {Boolean}     - indicates if iteration should continue
+     */
+    addCommonPrefix(commonPrefix: string, key: string): void {
+        // add the new prefix to the list
+        this.CommonPrefixes.push(commonPrefix);
+        ++this.keys;
+        this.nextMarker = commonPrefix;
+    }
+
+    addCommonPrefixOrContents(key: string, value: string): string | undefined {
+        // add the subprefix to the common prefixes if the key has the delimiter
+        const commonPrefix = this.getCommonPrefix(key);
+        if (commonPrefix) {
+            this.addCommonPrefix(commonPrefix, key);
+            return commonPrefix;
+        }
+        this.addContents(key, value);
+        return undefined;
+    }
+
+    getObjectKeyV0(obj: { key: string }): string {
+        return obj.key;
+    }
+
+    getObjectKeyV1(obj: { key: string }): string {
+        return obj.key.slice(DbPrefixes.Master.length);
+    }
+
+    /**
+     *  Filter to apply on each iteration, based on:
+     *  - prefix
+     *  - delimiter
+     *  - maxKeys
+     *  The marker is being handled directly by levelDB
+     *  @param {Object} obj       - The key and value of the element
+     *  @param {String} obj.key   - The key of the element
+     *  @param {String} obj.value - The value of the element
+     *  @return {number}          - indicates if iteration should continue
+     */
+    filter(obj: { key: string, value: string }): FilterReturnValue {
+        const key = this.getObjectKey(obj);
+        const value = obj.value;
+
+        return this.handleKey(key, value);
+    }
+
+    setState(state: FilterState): void {
+        this.state = state;
+    }
+
+    setKeyHandler(stateId: number, keyHandler: KeyHandler): void {
+        this.keyHandlers[stateId] = keyHandler;
+    }
+
+    handleKey(key: string, value: string): FilterReturnValue {
+        return this.keyHandlers[this.state.id](key, value);
+    }
+
+    keyHandler_NeverSkipping(key: string, value: string): FilterReturnValue {
+        if (this._reachedMaxKeys()) {
+            return FILTER_END;
+        }
+        this.addContents(key, value);
+        return FILTER_ACCEPT;
+    }
+
+    keyHandler_NotSkippingPrefix(key: string, value: string): FilterReturnValue {
+        if (this._reachedMaxKeys()) {
+            return FILTER_END;
+        }
+        const commonPrefix = this.addCommonPrefixOrContents(key, value);
+        if (commonPrefix) {
+            // transition into SkippingPrefix state to skip all following keys
+            // while they start with the same prefix
+            this.setState(<DelimiterFilterState_SkippingPrefix> {
+                id: DelimiterFilterStateId.SkippingPrefix,
+                prefix: commonPrefix,
+            });
+        }
+        return FILTER_ACCEPT;
+    }
+
+    keyHandler_SkippingPrefix(key: string, value: string): FilterReturnValue {
+        const { prefix } = <DelimiterFilterState_SkippingPrefix> this.state;
+        if (key.startsWith(prefix)) {
+            return FILTER_SKIP;
+        }
+        this.setState(<DelimiterFilterState_NotSkipping> {
+            id: DelimiterFilterStateId.NotSkipping,
+        });
+        return this.handleKey(key, value);
+    }
+
+    skippingBase(): string | undefined {
+        switch (this.state.id) {
+        case DelimiterFilterStateId.SkippingPrefix:
+            const { prefix } = <DelimiterFilterState_SkippingPrefix> this.state;
+            return inc(prefix);
+
+        default:
+            return SKIP_NONE;
+        }
+    }
+
+    skippingV0() {
+        return this.skippingBase();
+    }
+
+    skippingV1() {
+        const skipTo = this.skippingBase();
+        if (skipTo === SKIP_NONE) {
+            return SKIP_NONE;
+        }
+        return DbPrefixes.Master + skipTo;
+    }
+
+    /**
+     *  Return an object containing all mandatory fields to use once the
+     *  iteration is done, doesn't show a NextMarker field if the output
+     *  isn't truncated
+     *  @return {Object} - following amazon format
+     */
+    result(): ResultObject {
+        /* NextMarker is only provided when delimiter is used.
+         * specified in v1 listing documentation
+         * http://docs.aws.amazon.com/AmazonS3/latest/API/RESTBucketGET.html
+         */
+        const result: ResultObject = {
+            CommonPrefixes: this.CommonPrefixes,
+            Contents: this.Contents,
+            IsTruncated: this.IsTruncated,
+            Delimiter: this.delimiter,
+        };
+        if (this.parameters.v2) {
+            result.NextContinuationToken = this.IsTruncated
+                ? this.nextMarker : undefined;
+        } else {
+            result.NextMarker = (this.IsTruncated && this.delimiter)
+                ? this.nextMarker : undefined;
+        }
+        return result;
+    }
+}

lib/algos/list/delimiterCurrent.ts

@@ -0,0 +1,127 @@
+const { DelimiterMaster } = require('./delimiterMaster');
+const { FILTER_ACCEPT, FILTER_END } = require('./tools');
+
+type ResultObject = {
+    Contents: {
+        key: string;
+        value: string;
+    }[];
+    IsTruncated: boolean;
+    NextMarker ?: string;
+};
+
+/**
+ * Handle object listing with parameters. This extends the base class DelimiterMaster
+ * to return the master/current versions.
+ */
+class DelimiterCurrent extends DelimiterMaster {
+    /**
+     * Delimiter listing of current versions.
+     * @param {Object}  parameters            - listing parameters
+     * @param {String}  parameters.beforeDate - limit the response to keys older than beforeDate
+     * @param {String}  parameters.excludedDataStoreName - excluded datatore name
+     * @param {Number} parameters.maxScannedLifecycleListingEntries - max number of entries to be scanned
+     * @param {RequestLogger} logger          - The logger of the request
+     * @param {String} [vFormat]              - versioning key format
+     */
+    constructor(parameters, logger, vFormat) {
+        super(parameters, logger, vFormat);
+
+        this.beforeDate = parameters.beforeDate;
+        this.excludedDataStoreName = parameters.excludedDataStoreName;
+        this.maxScannedLifecycleListingEntries = parameters.maxScannedLifecycleListingEntries;
+        this.scannedKeys = 0;
+    }
+
+    genMDParamsV0() {
+        const params = super.genMDParamsV0();
+        // lastModified and dataStoreName parameters are used by metadata that enables built-in filtering, 
+        // a feature currently exclusive to MongoDB
+        if (this.beforeDate) {
+            params.lastModified = {
+                lt: this.beforeDate,
+            };
+        }
+
+        if (this.excludedDataStoreName) {
+            params.dataStoreName = {
+                ne: this.excludedDataStoreName,
+            }
+        }
+
+        return params;
+    }
+
+    /**
+     * Parses the stringified entry's value.
+     * @param s - sringified value
+     * @return - undefined if parsing fails, otherwise it contains the parsed value.
+     */
+    _parse(s) {
+        let p;
+        try {
+            p = JSON.parse(s);
+        } catch (e: any) {
+            this.logger.warn(
+                'Could not parse Object Metadata while listing',
+                { err: e.toString() });
+        }
+        return p;
+    }
+
+    /**
+     * check if the max keys count has been reached and set the
+     * final state of the result if it is the case
+     *
+     * specialized implementation on DelimiterCurrent to also check
+     * the number of scanned keys
+     *
+     * @return {Boolean} - indicates if the iteration has to stop
+     */
+    _reachedMaxKeys(): boolean {
+        if (this.maxScannedLifecycleListingEntries && this.scannedKeys >= this.maxScannedLifecycleListingEntries) {
+            this.IsTruncated = true;
+            this.logger.info('listing stopped due to reaching the maximum scanned entries limit',
+                {
+                    maxScannedLifecycleListingEntries: this.maxScannedLifecycleListingEntries,
+                    scannedKeys: this.scannedKeys,
+                });
+            return true;
+        }
+        return super._reachedMaxKeys();
+    }
+
+    addContents(key, value) {
+        ++this.scannedKeys;
+        const parsedValue = this._parse(value);
+        // if parsing fails, skip the key.
+        if (parsedValue) {
+            const lastModified = parsedValue['last-modified'];
+            const dataStoreName = parsedValue.dataStoreName;
+            // We then check if the current version is older than the "beforeDate" and
+            // "excludedDataStoreName" is not specified or if specified and the data store name is different.
+            if ((!this.beforeDate || (lastModified && lastModified < this.beforeDate)) &&
+                (!this.excludedDataStoreName || dataStoreName !== this.excludedDataStoreName)) {
+                super.addContents(key, value);
+            }
+            // In the event of a timeout occurring before any content is added,
+            // NextMarker is updated even if the object is not eligible.
+            // It minimizes the amount of data that the client needs to re-process if the request times out.
+            this.nextMarker = key;
+        }
+    }
+
+    result(): object {
+        const result: ResultObject = {
+            Contents: this.Contents,
+            IsTruncated: this.IsTruncated,
+        };
+
+        if (this.IsTruncated) {
+            result.NextMarker = this.nextMarker;
+        }
+
+        return result;
+    }
+}
+module.exports = { DelimiterCurrent };

lib/algos/list/delimiterMaster.js

@@ -1,196 +0,0 @@
-'use strict'; // eslint-disable-line strict
-
-const Delimiter = require('./delimiter').Delimiter;
-const Version = require('../../versioning/Version').Version;
-const VSConst = require('../../versioning/constants').VersioningConstants;
-const { BucketVersioningKeyFormat } = VSConst;
-const { FILTER_ACCEPT, FILTER_SKIP, SKIP_NONE } = require('./tools');
-
-const VID_SEP = VSConst.VersionId.Separator;
-const { DbPrefixes } = VSConst;
-
-/**
- * Handle object listing with parameters. This extends the base class Delimiter
- * to return the raw master versions of existing objects.
- */
-class DelimiterMaster extends Delimiter {
-    /**
-     * Delimiter listing of master versions.
-     * @param {Object}  parameters            - listing parameters
-     * @param {String}  parameters.delimiter  - delimiter per amazon format
-     * @param {String}  parameters.prefix     - prefix per amazon format
-     * @param {String}  parameters.marker     - marker per amazon format
-     * @param {Number}  parameters.maxKeys    - number of keys to list
-     * @param {Boolean} parameters.v2         - indicates whether v2 format
-     * @param {String}  parameters.startAfter - marker per amazon v2 format
-     * @param {String}  parameters.continuationToken - obfuscated amazon token
-     * @param {RequestLogger} logger          - The logger of the request
-     * @param {String} [vFormat]              - versioning key format
-     */
-    constructor(parameters, logger, vFormat) {
-        super(parameters, logger, vFormat);
-        // non-PHD master version or a version whose master is a PHD version
-        this.prvKey = undefined;
-        this.prvPHDKey = undefined;
-        this.inReplayPrefix = false;
-
-        Object.assign(this, {
-            [BucketVersioningKeyFormat.v0]: {
-                filter: this.filterV0,
-                skipping: this.skippingV0,
-            },
-            [BucketVersioningKeyFormat.v1]: {
-                filter: this.filterV1,
-                skipping: this.skippingV1,
-            },
-        }[this.vFormat]);
-    }
-
-    /**
-     *  Filter to apply on each iteration for buckets in v0 format,
-     *  based on:
-     *  - prefix
-     *  - delimiter
-     *  - maxKeys
-     *  The marker is being handled directly by levelDB
-     *  @param {Object} obj       - The key and value of the element
-     *  @param {String} obj.key   - The key of the element
-     *  @param {String} obj.value - The value of the element
-     *  @return {number}          - indicates if iteration should continue
-     */
-    filterV0(obj) {
-        let key = obj.key;
-        const value = obj.value;
-
-        if (key.startsWith(DbPrefixes.Replay)) {
-            this.inReplayPrefix = true;
-            return FILTER_SKIP;
-        }
-        this.inReplayPrefix = false;
-
-        /* Skip keys not starting with the prefix or not alphabetically
-         * ordered. */
-        if ((this.prefix && !key.startsWith(this.prefix))
-                || (typeof this[this.nextContinueMarker] === 'string' &&
-                    key <= this[this.nextContinueMarker])) {
-            return FILTER_SKIP;
-        }
-
-        /* Skip version keys (<key><versionIdSeparator><version>) if we already
-         * have a master version. */
-        const versionIdIndex = key.indexOf(VID_SEP);
-        if (versionIdIndex >= 0) {
-            key = key.slice(0, versionIdIndex);
-            /* - key === this.prvKey is triggered when a master version has
-             *   been accepted for this key,
-             * - key === this.NextMarker or this.NextContinueToken is triggered
-             *   when a listing page ends on an accepted obj and the next page
-             *   starts with a version of this object.
-             *   In that case prvKey is default set to undefined
-             *   in the constructor and comparing to NextMarker is the only
-             *   way to know  we should not accept this version. This test is
-             *   not redundant with the one at the beginning of this function,
-             *   we are comparing here the key without the version suffix,
-             * - key startsWith the previous NextMarker happens because we set
-             *   NextMarker to the common prefix instead of the whole key
-             *   value. (TODO: remove this test once ZENKO-1048 is fixed)
-             *   */
-            if (key === this.prvKey || key === this[this.nextContinueMarker] ||
-                (this.delimiter &&
-                key.startsWith(this[this.nextContinueMarker]))) {
-                /* master version already filtered */
-                return FILTER_SKIP;
-            }
-        }
-        if (Version.isPHD(value)) {
-            /* master version is a PHD version, we want to wait for the next
-             * one:
-             * - Set the prvKey to undefined to not skip the next version,
-             * - return accept to avoid users to skip the next values in range
-             *   (skip scan mechanism in metadata backend like Metadata or
-             *   MongoClient). */
-            this.prvKey = undefined;
-            this.prvPHDKey = key;
-            return FILTER_ACCEPT;
-        }
-        if (Version.isDeleteMarker(value)) {
-            /* This entry is a deleteMarker which has not been filtered by the
-             * version test. Either :
-             * - it is a deleteMarker on the master version, we want to SKIP
-             *   all the following entries with this key (no master version),
-             * - or a deleteMarker following a PHD (setting prvKey to undefined
-             *   when an entry is a PHD avoids the skip on version for the
-             *   next entry). In that case we expect the master version to
-             *   follow. */
-            if (key === this.prvPHDKey) {
-                this.prvKey = undefined;
-                return FILTER_ACCEPT;
-            }
-            this.prvKey = key;
-            return FILTER_SKIP;
-        }
-
-        this.prvKey = key;
-        if (this.delimiter) {
-            // check if the key has the delimiter
-            const baseIndex = this.prefix ? this.prefix.length : 0;
-            const delimiterIndex = key.indexOf(this.delimiter, baseIndex);
-            if (delimiterIndex >= 0) {
-                // try to add the prefix to the list
-                return this.addCommonPrefix(key, delimiterIndex);
-            }
-        }
-        return this.addContents(key, value);
-    }
-
-    /**
-     *  Filter to apply on each iteration for buckets in v1 format,
-     *  based on:
-     *  - prefix
-     *  - delimiter
-     *  - maxKeys
-     *  The marker is being handled directly by levelDB
-     *  @param {Object} obj       - The key and value of the element
-     *  @param {String} obj.key   - The key of the element
-     *  @param {String} obj.value - The value of the element
-     *  @return {number}          - indicates if iteration should continue
-     */
-    filterV1(obj) {
-        // Filtering master keys in v1 is simply listing the master
-        // keys, as the state of version keys do not change the
-        // result, so we can use Delimiter method directly.
-        return super.filter(obj);
-    }
-
-    skippingBase() {
-        if (this[this.nextContinueMarker]) {
-            // next marker or next continuation token:
-            // - foo/ : skipping foo/
-            // - foo  : skipping foo.
-            const index = this[this.nextContinueMarker].
-                lastIndexOf(this.delimiter);
-            if (index === this[this.nextContinueMarker].length - 1) {
-                return this[this.nextContinueMarker];
-            }
-            return this[this.nextContinueMarker] + VID_SEP;
-        }
-        return SKIP_NONE;
-    }
-
-    skippingV0() {
-        if (this.inReplayPrefix) {
-            return DbPrefixes.Replay;
-        }
-        return this.skippingBase();
-    }
-
-    skippingV1() {
-        const skipTo = this.skippingBase();
-        if (skipTo === SKIP_NONE) {
-            return SKIP_NONE;
-        }
-        return DbPrefixes.Master + skipTo;
-    }
-}
-
-module.exports = { DelimiterMaster };

lib/algos/list/delimiterMaster.ts

@@ -0,0 +1,620 @@
+import {
+    Delimiter,
+    FilterState,
+    FilterReturnValue,
+    DelimiterFilterStateId,
+    DelimiterFilterState_NotSkipping,
+    DelimiterFilterState_SkippingPrefix,
+    ResultObject,
+} from './delimiter';
+const Version = require('../../versioning/Version').Version;
+const VSConst = require('../../versioning/constants').VersioningConstants;
+const { BucketVersioningKeyFormat } = VSConst;
+const { FILTER_ACCEPT, FILTER_SKIP, FILTER_END, SKIP_NONE, inc } = require('./tools');
+
+import { GapSetEntry } from '../cache/GapSet';
+import { GapCacheInterface } from '../cache/GapCache';
+
+const VID_SEP = VSConst.VersionId.Separator;
+const { DbPrefixes } = VSConst;
+
+export const enum DelimiterMasterFilterStateId {
+    SkippingVersionsV0 = 101,
+    WaitVersionAfterPHDV0 = 102,
+    SkippingGapV0 = 103,
+};
+
+interface DelimiterMasterFilterState_SkippingVersionsV0 extends FilterState {
+    id: DelimiterMasterFilterStateId.SkippingVersionsV0,
+    masterKey: string,
+};
+
+interface DelimiterMasterFilterState_WaitVersionAfterPHDV0 extends FilterState {
+    id: DelimiterMasterFilterStateId.WaitVersionAfterPHDV0,
+    masterKey: string,
+};
+
+interface DelimiterMasterFilterState_SkippingGapV0 extends FilterState {
+    id: DelimiterMasterFilterStateId.SkippingGapV0,
+};
+
+export const enum GapCachingState {
+    NoGapCache = 0, // there is no gap cache
+    UnknownGap = 1, // waiting for a cache lookup
+    GapLookupInProgress = 2, // asynchronous gap lookup in progress
+    GapCached = 3, // an upcoming or already skippable gap is cached
+    NoMoreGap = 4, // the cache doesn't have any more gaps inside the listed range
+};
+
+type GapCachingInfo_NoGapCache = {
+    state: GapCachingState.NoGapCache;
+};
+
+type GapCachingInfo_NoCachedGap = {
+    state: GapCachingState.UnknownGap
+        | GapCachingState.GapLookupInProgress
+    gapCache: GapCacheInterface;
+};
+
+type GapCachingInfo_GapCached = {
+    state: GapCachingState.GapCached;
+    gapCache: GapCacheInterface;
+    gapCached: GapSetEntry;
+};
+
+type GapCachingInfo_NoMoreGap = {
+    state: GapCachingState.NoMoreGap;
+};
+
+type GapCachingInfo = GapCachingInfo_NoGapCache
+    | GapCachingInfo_NoCachedGap
+    | GapCachingInfo_GapCached
+    | GapCachingInfo_NoMoreGap;
+
+
+export const enum GapBuildingState {
+    Disabled = 0, // no gap cache or no gap building needed (e.g. in V1 versioning format)
+    NotBuilding = 1, // not currently building a gap (i.e. not listing within a gap)
+    Building = 2, // currently building a gap (i.e. listing within a gap)
+    Expired = 3, // not allowed to build due to exposure delay timeout
+};
+
+type GapBuildingInfo_NothingToBuild = {
+    state: GapBuildingState.Disabled | GapBuildingState.Expired;
+};
+
+type GapBuildingParams = {
+    /**
+     * minimum weight for a gap to be created in the cache
+     */
+    minGapWeight: number;
+    /**
+     * trigger a cache setGap() call every N skippable keys
+     */
+    triggerSaveGapWeight: number;
+    /**
+     * timestamp to assess whether we're still inside the validity period to
+     * be allowed to build gaps
+     */
+    initTimestamp: number;
+};
+
+type GapBuildingInfo_NotBuilding = {
+    state: GapBuildingState.NotBuilding;
+    gapCache: GapCacheInterface;
+    params: GapBuildingParams;
+};
+
+type GapBuildingInfo_Building = {
+    state: GapBuildingState.Building;
+    gapCache: GapCacheInterface;
+    params: GapBuildingParams;
+    /**
+     * Gap currently being created
+     */
+    gap: GapSetEntry;
+    /**
+     * total current weight of the gap being created
+     */
+    gapWeight: number;
+};
+
+type GapBuildingInfo = GapBuildingInfo_NothingToBuild
+    | GapBuildingInfo_NotBuilding
+    | GapBuildingInfo_Building;
+
+/**
+ * Handle object listing with parameters. This extends the base class Delimiter
+ * to return the raw master versions of existing objects.
+ */
+export class DelimiterMaster extends Delimiter {
+
+    _gapCaching: GapCachingInfo;
+    _gapBuilding: GapBuildingInfo;
+    _refreshedBuildingParams: GapBuildingParams | null;
+
+    /**
+     * Delimiter listing of master versions.
+     * @param {Object}  parameters            - listing parameters
+     * @param {String}  [parameters.delimiter]  - delimiter per amazon format
+     * @param {String}  [parameters.prefix]     - prefix per amazon format
+     * @param {String}  [parameters.marker]     - marker per amazon format
+     * @param {Number}  [parameters.maxKeys]    - number of keys to list
+     * @param {Boolean} [parameters.v2]         - indicates whether v2 format
+     * @param {String}  [parameters.startAfter] - marker per amazon v2 format
+     * @param {String}  [parameters.continuationToken] - obfuscated amazon token
+     * @param {RequestLogger} logger            - The logger of the request
+     * @param {String}  [vFormat="v0"]          - versioning key format
+     */
+    constructor(parameters, logger, vFormat?: string) {
+        super(parameters, logger, vFormat);
+
+        if (this.vFormat === BucketVersioningKeyFormat.v0) {
+            // override Delimiter's implementation of NotSkipping for
+            // DelimiterMaster logic (skipping versions and special
+            // handling of delete markers and PHDs)
+            this.setKeyHandler(
+                DelimiterFilterStateId.NotSkipping,
+                this.keyHandler_NotSkippingPrefixNorVersionsV0.bind(this));
+
+            // add extra state handlers specific to DelimiterMaster with v0 format
+            this.setKeyHandler(
+                DelimiterMasterFilterStateId.SkippingVersionsV0,
+                this.keyHandler_SkippingVersionsV0.bind(this));
+
+            this.setKeyHandler(
+                DelimiterMasterFilterStateId.WaitVersionAfterPHDV0,
+                this.keyHandler_WaitVersionAfterPHDV0.bind(this));
+
+            this.setKeyHandler(
+                DelimiterMasterFilterStateId.SkippingGapV0,
+                this.keyHandler_SkippingGapV0.bind(this));
+
+            if (this.marker) {
+                // distinct initial state to include some special logic
+                // before the first master key is found that does not have
+                // to be checked afterwards
+                this.state = <DelimiterMasterFilterState_SkippingVersionsV0> {
+                    id: DelimiterMasterFilterStateId.SkippingVersionsV0,
+                    masterKey: this.marker,
+                };
+            } else {
+                this.state = <DelimiterFilterState_NotSkipping> {
+                    id: DelimiterFilterStateId.NotSkipping,
+                };
+            }
+        } else {
+            // save base implementation of the `NotSkipping` state in
+            // Delimiter before overriding it with ours, to be able to call it from there
+            this.keyHandler_NotSkipping_Delimiter = this.keyHandlers[DelimiterFilterStateId.NotSkipping];
+            this.setKeyHandler(
+                DelimiterFilterStateId.NotSkipping,
+                this.keyHandler_NotSkippingPrefixNorVersionsV1.bind(this));
+        }
+        // in v1, we can directly use Delimiter's implementation,
+        // which is already set to the proper state
+
+        // default initialization of the gap cache and building states, can be
+        // set by refreshGapCache()
+        this._gapCaching = {
+            state: GapCachingState.NoGapCache,
+        };
+        this._gapBuilding = {
+            state: GapBuildingState.Disabled,
+        };
+        this._refreshedBuildingParams = null;
+    }
+
+    /**
+     * Get the validity period left before a refresh of the gap cache is needed
+     * to continue building new gaps.
+     *
+     * @return {number|null} one of:
+     * - the remaining time in milliseconds in which gaps can be added to the
+     *   cache before a call to refreshGapCache() is required
+     * - or 0 if there is no time left and a call to refreshGapCache() is required
+     *   to resume caching gaps
+     * - or null if refreshing the cache is never needed (because the gap cache
+     *   is either not available or not used)
+     */
+    getGapBuildingValidityPeriodMs(): number | null {
+        let gapBuilding;
+        switch (this._gapBuilding.state) {
+        case GapBuildingState.Disabled:
+            return null;
+        case GapBuildingState.Expired:
+            return 0;
+        case GapBuildingState.NotBuilding:
+            gapBuilding = <GapBuildingInfo_NotBuilding> this._gapBuilding;
+            break;
+        case GapBuildingState.Building:
+            gapBuilding = <GapBuildingInfo_Building> this._gapBuilding;
+            break;
+        }
+        const { gapCache, params } = gapBuilding;
+        const elapsedTime = Date.now() - params.initTimestamp;
+        return Math.max(gapCache.exposureDelayMs - elapsedTime, 0);
+    }
+
+    /**
+     * Refresh the gaps caching logic (gaps are series of current delete markers
+     * in V0 bucket metadata format). It has two effects:
+     *
+     * - starts exposing existing and future gaps from the cache to efficiently
+     *   skip over series of current delete markers that have been seen and cached
+     *   earlier
+     *
+     * - enables building and caching new gaps (or extend existing ones), for a
+     *   limited time period defined by the `gapCacheProxy.exposureDelayMs` value
+     *   in milliseconds. To refresh the validity period and resume building and
+     *   caching new gaps, one must restart a new listing from the database (starting
+     *   at the current listing key, included), then call refreshGapCache() again.
+     *
+     * @param {GapCacheInterface} gapCacheProxy - API proxy to the gaps cache
+     *   (the proxy should handle prefixing object keys with the bucket name)
+     * @param {number} [minGapWeight=100] - minimum weight of a gap for it to be
+     *   added in the cache
+     * @param {number} [triggerSaveGapWeight] - cumulative weight to wait for
+     *   before saving the current building gap. Cannot be greater than
+     *   `gapCacheProxy.maxGapWeight` (the value is thresholded to `maxGapWeight`
+     *   otherwise). Defaults to `gapCacheProxy.maxGapWeight / 2`.
+     * @return {undefined}
+     */
+    refreshGapCache(
+        gapCacheProxy: GapCacheInterface,
+        minGapWeight?: number,
+        triggerSaveGapWeight?: number
+    ): void {
+        if (this.vFormat !== BucketVersioningKeyFormat.v0) {
+            return;
+        }
+        if (this._gapCaching.state === GapCachingState.NoGapCache) {
+            this._gapCaching = {
+                state: GapCachingState.UnknownGap,
+                gapCache: gapCacheProxy,
+            };
+        }
+        const refreshedBuildingParams: GapBuildingParams = {
+            minGapWeight: minGapWeight || 100,
+            triggerSaveGapWeight: triggerSaveGapWeight
+                || Math.trunc(gapCacheProxy.maxGapWeight / 2),
+            initTimestamp: Date.now(),
+        };
+        if (this._gapBuilding.state === GapBuildingState.Building) {
+            // refreshed params will be applied as soon as the current building gap is saved
+            this._refreshedBuildingParams = refreshedBuildingParams;
+        } else {
+            this._gapBuilding = {
+                state: GapBuildingState.NotBuilding,
+                gapCache: gapCacheProxy,
+                params: refreshedBuildingParams,
+            };
+        }
+    }
+
+    /**
+     * Trigger a lookup of the closest upcoming or already skippable gap.
+     *
+     * @param {string} fromKey - lookup a gap not before 'fromKey'
+     * @return {undefined} - the lookup is asynchronous and its
+     * response is handled inside this function
+     */
+    _triggerGapLookup(gapCaching: GapCachingInfo_NoCachedGap, fromKey: string): void {
+        this._gapCaching = {
+            state: GapCachingState.GapLookupInProgress,
+            gapCache: gapCaching.gapCache,
+        };
+        const maxKey = this.prefix ? inc(this.prefix) : undefined;
+        gapCaching.gapCache.lookupGap(fromKey, maxKey).then(_gap => {
+            const gap = <GapSetEntry | null> _gap;
+            if (gap) {
+                this._gapCaching = {
+                    state: GapCachingState.GapCached,
+                    gapCache: gapCaching.gapCache,
+                    gapCached: gap,
+                };
+            } else {
+                this._gapCaching = {
+                    state: GapCachingState.NoMoreGap,
+                };
+            }
+        });
+    }
+
+    _checkGapOnMasterDeleteMarker(key: string): FilterReturnValue {
+        switch (this._gapBuilding.state) {
+        case GapBuildingState.Disabled:
+        case GapBuildingState.Expired:
+            break;
+        case GapBuildingState.NotBuilding:
+            this._createBuildingGap(key, 1);
+            break;
+        case GapBuildingState.Building:
+            this._updateBuildingGap(key);
+            break;
+        }
+        if (this._gapCaching.state === GapCachingState.GapCached) {
+            const { gapCached } = this._gapCaching;
+            if (key >= gapCached.firstKey) {
+                if (key <= gapCached.lastKey) {
+                    // we are inside the last looked up cached gap: transition to
+                    // 'SkippingGapV0' state
+                    this.setState(<DelimiterMasterFilterState_SkippingGapV0> {
+                        id: DelimiterMasterFilterStateId.SkippingGapV0,
+                    });
+                    // cut the current gap before skipping, it will be merged or
+                    // chained with the existing one (depending on its weight)
+                    if (this._gapBuilding.state === GapBuildingState.Building) {
+                        // substract 1 from the weight because we are going to chain this gap,
+                        // which has an overlap of one key.
+                        this._gapBuilding.gap.weight -= 1;
+                        this._cutBuildingGap();
+                    }
+                    return FILTER_SKIP;
+                }
+                // as we are past the cached gap, we will need another lookup
+                this._gapCaching = {
+                    state: GapCachingState.UnknownGap,
+                    gapCache: this._gapCaching.gapCache,
+                };
+            }
+        }
+        if (this._gapCaching.state === GapCachingState.UnknownGap) {
+            this._triggerGapLookup(this._gapCaching, key);
+        }
+        return FILTER_ACCEPT;
+    }
+
+    filter_onNewMasterKeyV0(key: string, value: string): FilterReturnValue {
+        // if this master key is a delete marker, accept it without
+        // adding the version to the contents
+        if (Version.isDeleteMarker(value)) {
+            // update the state to start skipping versions of the new master key
+            this.setState(<DelimiterMasterFilterState_SkippingVersionsV0> {
+                id: DelimiterMasterFilterStateId.SkippingVersionsV0,
+                masterKey: key,
+            });
+            return this._checkGapOnMasterDeleteMarker(key);
+        }
+        if (Version.isPHD(value)) {
+            // master version is a PHD version: wait for the first
+            // following version that will be considered as the actual
+            // master key
+            this.setState(<DelimiterMasterFilterState_WaitVersionAfterPHDV0> {
+                id: DelimiterMasterFilterStateId.WaitVersionAfterPHDV0,
+                masterKey: key,
+            });
+            return FILTER_ACCEPT;
+        }
+        // cut the current gap as soon as a non-deleted entry is seen
+        this._cutBuildingGap();
+
+        if (key.startsWith(DbPrefixes.Replay)) {
+            // skip internal replay prefix entirely
+            this.setState(<DelimiterFilterState_SkippingPrefix> {
+                id: DelimiterFilterStateId.SkippingPrefix,
+                prefix: DbPrefixes.Replay,
+            });
+            return FILTER_SKIP;
+        }
+        if (this._reachedMaxKeys()) {
+            return FILTER_END;
+        }
+
+        const commonPrefix = this.addCommonPrefixOrContents(key, value);
+        if (commonPrefix) {
+            // transition into SkippingPrefix state to skip all following keys
+            // while they start with the same prefix
+            this.setState(<DelimiterFilterState_SkippingPrefix> {
+                id: DelimiterFilterStateId.SkippingPrefix,
+                prefix: commonPrefix,
+            });
+            return FILTER_ACCEPT;
+        }
+        // update the state to start skipping versions of the new master key
+        this.setState(<DelimiterMasterFilterState_SkippingVersionsV0> {
+            id: DelimiterMasterFilterStateId.SkippingVersionsV0,
+            masterKey: key,
+        });
+        return FILTER_ACCEPT;
+    }
+
+    keyHandler_NotSkippingPrefixNorVersionsV0(key: string, value: string): FilterReturnValue {
+        return this.filter_onNewMasterKeyV0(key, value);
+    }
+
+    filter_onNewMasterKeyV1(key: string, value: string): FilterReturnValue {
+        // if this master key is a delete marker, accept it without
+        // adding the version to the contents
+        if (Version.isDeleteMarker(value)) {
+            return FILTER_ACCEPT;
+        }
+        // use base Delimiter's implementation
+        return this.keyHandler_NotSkipping_Delimiter(key, value);
+    }
+
+    keyHandler_NotSkippingPrefixNorVersionsV1(key: string, value: string): FilterReturnValue {
+        return this.filter_onNewMasterKeyV1(key, value);
+    }
+
+    keyHandler_SkippingVersionsV0(key: string, value: string): FilterReturnValue {
+        /* In the SkippingVersionsV0 state, skip all version keys
+         * (<key><versionIdSeparator><version>) */
+        const versionIdIndex = key.indexOf(VID_SEP);
+        if (versionIdIndex !== -1) {
+            // version keys count in the building gap weight because they must
+            // also be listed until skipped
+            if (this._gapBuilding.state === GapBuildingState.Building) {
+                this._updateBuildingGap(key);
+            }
+            return FILTER_SKIP;
+        }
+        return this.filter_onNewMasterKeyV0(key, value);
+    }
+
+    keyHandler_WaitVersionAfterPHDV0(key: string, value: string): FilterReturnValue {
+        // After a PHD key is encountered, the next version key of the
+        // same object if it exists is the new master key, hence
+        // consider it as such and call 'onNewMasterKeyV0' (the test
+        // 'masterKey == phdKey' is probably redundant when we already
+        // know we have a versioned key, since all objects in v0 have
+        // a master key, but keeping it in doubt)
+        const { masterKey: phdKey } = <DelimiterMasterFilterState_WaitVersionAfterPHDV0> this.state;
+        const versionIdIndex = key.indexOf(VID_SEP);
+        if (versionIdIndex !== -1) {
+            const masterKey = key.slice(0, versionIdIndex);
+            if (masterKey === phdKey) {
+                return this.filter_onNewMasterKeyV0(masterKey, value);
+            }
+        }
+        return this.filter_onNewMasterKeyV0(key, value);
+    }
+
+    keyHandler_SkippingGapV0(key: string, value: string): FilterReturnValue {
+        const { gapCache, gapCached } = <GapCachingInfo_GapCached> this._gapCaching;
+        if (key <= gapCached.lastKey) {
+            return FILTER_SKIP;
+        }
+        this._gapCaching = {
+            state: GapCachingState.UnknownGap,
+            gapCache,
+        };
+        this.setState(<DelimiterMasterFilterState_SkippingVersionsV0> {
+            id: DelimiterMasterFilterStateId.SkippingVersionsV0,
+        });
+        // Start a gap with weight=0 from the latest skippable key. This will
+        // allow to extend the gap just skipped with a chained gap in case
+        // other delete markers are seen after the existing gap is skipped.
+        this._createBuildingGap(gapCached.lastKey, 0, gapCached.weight);
+
+        return this.handleKey(key, value);
+    }
+
+    skippingBase(): string | undefined {
+        switch (this.state.id) {
+        case DelimiterMasterFilterStateId.SkippingVersionsV0:
+            const { masterKey } = <DelimiterMasterFilterState_SkippingVersionsV0> this.state;
+            return masterKey + inc(VID_SEP);
+
+        case DelimiterMasterFilterStateId.SkippingGapV0:
+            const { gapCached } = <GapCachingInfo_GapCached> this._gapCaching;
+            return gapCached.lastKey;
+
+        default:
+            return super.skippingBase();
+        }
+    }
+
+    result(): ResultObject {
+        this._cutBuildingGap();
+        return super.result();
+    }
+
+    _checkRefreshedBuildingParams(params: GapBuildingParams): GapBuildingParams {
+        if (this._refreshedBuildingParams) {
+            const newParams = this._refreshedBuildingParams;
+            this._refreshedBuildingParams = null;
+            return newParams;
+        }
+        return params;
+    }
+
+    /**
+     * Save the gap being built if allowed (i.e. still within the
+     * allocated exposure time window).
+     *
+     * @return {boolean} - true if the gap was saved, false if we are
+     * outside the allocated exposure time window.
+     */
+    _saveBuildingGap(): boolean {
+        const { gapCache, params, gap, gapWeight } =
+              <GapBuildingInfo_Building> this._gapBuilding;
+        const totalElapsed = Date.now() - params.initTimestamp;
+        if (totalElapsed >= gapCache.exposureDelayMs) {
+            this._gapBuilding = {
+                state: GapBuildingState.Expired,
+            };
+            this._refreshedBuildingParams = null;
+            return false;
+        }
+        const { firstKey, lastKey, weight } = gap;
+        gapCache.setGap(firstKey, lastKey, weight);
+        this._gapBuilding = {
+            state: GapBuildingState.Building,
+            gapCache,
+            params: this._checkRefreshedBuildingParams(params),
+            gap: {
+                firstKey: gap.lastKey,
+                lastKey: gap.lastKey,
+                weight: 0,
+            },
+            gapWeight,
+        };
+        return true;
+    }
+
+    /**
+     * Create a new gap to be extended afterwards
+     *
+     * @param {string} newKey - gap's first key
+     * @param {number} startWeight - initial weight of the building gap (usually 0 or 1)
+     * @param {number} [cachedWeight] - if continuing a cached gap, weight of the existing
+     *   cached portion
+     * @return {undefined}
+     */
+    _createBuildingGap(newKey: string, startWeight: number, cachedWeight?: number): void {
+        if (this._gapBuilding.state === GapBuildingState.NotBuilding) {
+            const { gapCache, params } = <GapBuildingInfo_NotBuilding> this._gapBuilding;
+            this._gapBuilding = {
+                state: GapBuildingState.Building,
+                gapCache,
+                params: this._checkRefreshedBuildingParams(params),
+                gap: {
+                    firstKey: newKey,
+                    lastKey: newKey,
+                    weight: startWeight,
+                },
+                gapWeight: (cachedWeight || 0) + startWeight,
+            };
+        }
+    }
+
+    _updateBuildingGap(newKey: string): void {
+        const gapBuilding = <GapBuildingInfo_Building> this._gapBuilding;
+        const { params, gap } = gapBuilding;
+        gap.lastKey = newKey;
+        gap.weight += 1;
+        gapBuilding.gapWeight += 1;
+        // the GapCache API requires updating a gap regularly because it can only split
+        // it once per update, by the known last key. In practice the default behavior
+        // is to trigger an update after a number of keys that is half the maximum weight.
+        // It is also useful for other listings to benefit from the cache sooner.
+        if (gapBuilding.gapWeight >= params.minGapWeight &&
+            gap.weight >= params.triggerSaveGapWeight) {
+            this._saveBuildingGap();
+        }
+    }
+
+    _cutBuildingGap(): void {
+        if (this._gapBuilding.state === GapBuildingState.Building) {
+            let gapBuilding = <GapBuildingInfo_Building> this._gapBuilding;
+            let { gapCache, params, gap, gapWeight } = gapBuilding;
+            // only set gaps that are significant enough in weight and
+            // with a non-empty extension
+            if (gapWeight >= params.minGapWeight && gap.weight > 0) {
+                // we're done if we were not allowed to save the gap
+                if (!this._saveBuildingGap()) {
+                    return;
+                }
+                // params may have been refreshed, reload them
+                gapBuilding = <GapBuildingInfo_Building> this._gapBuilding;
+                params = gapBuilding.params;
+            }
+            this._gapBuilding = {
+                state: GapBuildingState.NotBuilding,
+                gapCache,
+                params,
+            };
+        }
+    }
+}

lib/algos/list/delimiterNonCurrent.js

@@ -0,0 +1,202 @@
+const { DelimiterVersions } = require('./delimiterVersions');
+const { FILTER_END, FILTER_SKIP } = require('./tools');
+
+const TRIM_METADATA_MIN_BLOB_SIZE = 10000;
+
+/**
+ * Handle object listing with parameters. This extends the base class DelimiterVersions
+ * to return the raw non-current versions objects.
+ */
+class DelimiterNonCurrent extends DelimiterVersions {
+    /**
+     * Delimiter listing of non-current versions.
+     * @param {Object}  parameters                  - listing parameters
+     * @param {String}  parameters.keyMarker        - key marker
+     * @param {String}  parameters.versionIdMarker  - version id marker
+     * @param {String}  parameters.beforeDate       - limit the response to keys with stale date older than beforeDate.
+     * “stale date” is the date on when a version becomes non-current.
+     * @param {Number} parameters.maxScannedLifecycleListingEntries - max number of entries to be scanned
+     * @param {String} parameters.excludedDataStoreName - exclude dataStoreName matches from the versions
+     * @param {RequestLogger} logger                - The logger of the request
+     * @param {String} [vFormat]                    - versioning key format
+     */
+    constructor(parameters, logger, vFormat) {
+        super(parameters, logger, vFormat);
+
+        this.beforeDate = parameters.beforeDate;
+        this.excludedDataStoreName = parameters.excludedDataStoreName;
+        this.maxScannedLifecycleListingEntries = parameters.maxScannedLifecycleListingEntries;
+
+        // internal state
+        this.prevKey = null;
+        this.staleDate = null;
+
+        this.scannedKeys = 0;
+    }
+
+    getLastModified(value) {
+        let lastModified;
+        try {
+            const v = JSON.parse(value);
+            lastModified = v['last-modified'];
+        } catch (e) {
+            this.logger.warn('could not parse Object Metadata while listing',
+                {
+                    method: 'getLastModified',
+                    err: e.toString(),
+                });
+        }
+        return lastModified;
+    }
+
+    // Overwrite keyHandler_SkippingVersions to include the last version from the previous listing.
+    // The creation (last-modified) date of this version will be the stale date for the following version.
+    // eslint-disable-next-line camelcase
+    keyHandler_SkippingVersions(key, versionId, value) {
+        if (key === this.keyMarker) {
+            // since the nonversioned key equals the marker, there is
+            // necessarily a versionId in this key
+            const _versionId = versionId;
+            if (_versionId < this.versionIdMarker) {
+                // skip all versions until marker
+                return FILTER_SKIP;
+            }
+        }
+        this.setState({
+            id: 1 /* NotSkipping */,
+        });
+        return this.handleKey(key, versionId, value);
+    }
+
+    filter(obj) {
+        if (this.maxScannedLifecycleListingEntries && this.scannedKeys >= this.maxScannedLifecycleListingEntries) {
+            this.IsTruncated = true;
+            this.logger.info('listing stopped due to reaching the maximum scanned entries limit',
+                {
+                    maxScannedLifecycleListingEntries: this.maxScannedLifecycleListingEntries,
+                    scannedKeys: this.scannedKeys,
+                });
+            return FILTER_END;
+        }
+        ++this.scannedKeys;
+        return super.filter(obj);
+    }
+
+    /**
+     * NOTE: Each version of a specific key is sorted from the latest to the oldest
+     * thanks to the way version ids are generated.
+     * DESCRIPTION: Skip the version if it represents the master key, but keep its last-modified date in memory,
+     * which will be the stale date of the following version.
+     * The following version is pushed only:
+     * - if the "stale date" (picked up from the previous version) is available (JSON.parse has not failed),
+     * - if "beforeDate" is not specified or if specified and the "stale date" is older.
+     * - if "excludedDataStoreName" is not specified or if specified and the data store name is different
+     * The in-memory "stale date" is then updated with the version's last-modified date to be used for
+     * the following version.
+     * The process stops and returns the available results if either:
+     * - no more metadata key is left to be processed
+     * - the listing reaches the maximum number of key to be returned
+     * - the internal timeout is reached
+     *  @param {String} key   - The key to add
+     *  @param {String} versionId - The version id
+     *  @param {String} value - The value of the key
+     *  @return {undefined}
+     */
+    addVersion(key, versionId, value) {
+        this.nextKeyMarker = key;
+        this.nextVersionIdMarker = versionId;
+
+        // Skip the version if it represents the non-current version, but keep its last-modified date,
+        // which will be the stale date of the following version.
+        const isCurrentVersion = key !== this.prevKey;
+        if (isCurrentVersion) {
+            this.staleDate = this.getLastModified(value);
+            this.prevKey = key;
+            return;
+        }
+
+        // The following version is pushed only:
+        // - if the "stale date" (picked up from the previous version) is available (JSON.parse has not failed),
+        // - if "beforeDate" is not specified or if specified and the "stale date" is older.
+        // - if "excludedDataStoreName" is not specified or if specified and the data store name is different
+        let lastModified;
+        if (this.staleDate && (!this.beforeDate || this.staleDate < this.beforeDate)) {
+            const parsedValue = this._parse(value);
+            // if parsing fails, skip the key.
+            if (parsedValue) {
+                const dataStoreName = parsedValue.dataStoreName;
+                lastModified = parsedValue['last-modified'];
+                if (!this.excludedDataStoreName || dataStoreName !== this.excludedDataStoreName) {
+                    const s = this._stringify(parsedValue, this.staleDate);
+                    // check that _stringify succeeds to only push objects with a defined staleDate.
+                    if (s) {
+                        this.Versions.push({ key, value: s });
+                        ++this.keys;
+                    }
+                }
+            }
+        }
+
+        // The in-memory "stale date" is then updated with the version's last-modified date to be used for
+        // the following version.
+        this.staleDate = lastModified || this.getLastModified(value);
+
+        return;
+    }
+
+    /**
+     * Parses the stringified entry's value and remove the location property if too large.
+     * @param {string} s - sringified value
+     * @return {object} p - undefined if parsing fails, otherwise it contains the parsed value.
+     */
+    _parse(s) {
+        let p;
+        try {
+            p = JSON.parse(s);
+            if (s.length >= TRIM_METADATA_MIN_BLOB_SIZE) {
+                delete p.location;
+            }
+        } catch (e) {
+            this.logger.warn('Could not parse Object Metadata while listing', {
+                method: 'DelimiterNonCurrent._parse',
+                err: e.toString(),
+            });
+        }
+        return p;
+    }
+
+    _stringify(parsedMD, staleDate) {
+        const p = parsedMD;
+        let s = undefined;
+        p.staleDate = staleDate;
+        try {
+            s = JSON.stringify(p);
+        } catch (e) {
+            this.logger.warn('could not stringify Object Metadata while listing', {
+                method: 'DelimiterNonCurrent._stringify',
+                err: e.toString(),
+            });
+        }
+        return s;
+    }
+
+    result() {
+        const { Versions, IsTruncated, NextKeyMarker, NextVersionIdMarker } = super.result();
+
+        const result = {
+            Contents: Versions,
+            IsTruncated,
+        };
+
+        if (NextKeyMarker) {
+            result.NextKeyMarker = NextKeyMarker;
+        }
+
+        if (NextVersionIdMarker) {
+            result.NextVersionIdMarker = NextVersionIdMarker;
+        }
+
+        return result;
+    }
+}
+module.exports = { DelimiterNonCurrent };

lib/algos/list/delimiterOrphanDeleteMarker.js

@@ -0,0 +1,204 @@
+const DelimiterVersions = require('./delimiterVersions').DelimiterVersions;
+const { FILTER_END } = require('./tools');
+const TRIM_METADATA_MIN_BLOB_SIZE = 10000;
+/**
+ * Handle object listing with parameters. This extends the base class DelimiterVersions
+ * to return the orphan delete markers. Orphan delete markers are also
+ * refered as expired object delete marker.
+ * They are delete marker with zero noncurrent versions.
+ */
+class DelimiterOrphanDeleteMarker extends DelimiterVersions {
+    /**
+     * Delimiter listing of orphan delete markers.
+     * @param {Object}  parameters            - listing parameters
+     * @param {String}  parameters.beforeDate - limit the response to keys older than beforeDate
+     * @param {Number} parameters.maxScannedLifecycleListingEntries - max number of entries to be scanned
+     * @param {RequestLogger} logger          - The logger of the request
+     * @param {String} [vFormat]              - versioning key format
+     */
+    constructor(parameters, logger, vFormat) {
+        const {
+            marker,
+            maxKeys,
+            prefix,
+            beforeDate,
+            maxScannedLifecycleListingEntries,
+        } = parameters;
+
+        const versionParams = {
+            // The orphan delete marker logic uses the term 'marker' instead of 'keyMarker',
+            // as the latter could suggest the presence of a 'versionIdMarker'.
+            keyMarker: marker,
+            maxKeys,
+            prefix,
+        };
+        super(versionParams, logger, vFormat);
+
+        this.maxScannedLifecycleListingEntries = maxScannedLifecycleListingEntries;
+        this.beforeDate = beforeDate;
+        // this.prevKeyName is used as a marker for the next listing when the current one reaches its entry limit.
+        // We cannot rely on this.keyName, as it contains the name of the current key.
+        // In the event of a listing interruption due to reaching the maximum scanned entries,
+        // relying on this.keyName would cause the next listing to skip the current key because S3 starts
+        // listing after the marker.
+        this.prevKeyName = null;
+        this.keyName = null;
+        this.value = null;
+        this.scannedKeys = 0;
+    }
+
+    _reachedMaxKeys() {
+        if (this.keys >= this.maxKeys) {
+            return true;
+        }
+        return false;
+    }
+
+    _addOrphan() {
+        const parsedValue = this._parse(this.value);
+        // if parsing fails, skip the key.
+        if (parsedValue) {
+            const lastModified = parsedValue['last-modified'];
+            const isDeleteMarker = parsedValue.isDeleteMarker;
+            // We then check if the orphan version is a delete marker and if it is older than the "beforeDate"
+            if ((!this.beforeDate || (lastModified && lastModified < this.beforeDate)) && isDeleteMarker) {
+                // Prefer returning an untrimmed data rather than stopping the service in case of parsing failure.
+                const s = this._stringify(parsedValue) || this.value;
+                this.Versions.push({ key: this.keyName, value: s });
+                this.nextKeyMarker = this.keyName;
+                ++this.keys;
+            }
+        }
+    }
+
+    /**
+     * Parses the stringified entry's value and remove the location property if too large.
+     * @param {string} s - sringified value
+     * @return {object} p - undefined if parsing fails, otherwise it contains the parsed value.
+     */
+    _parse(s) {
+        let p;
+        try {
+            p = JSON.parse(s);
+            if (s.length >= TRIM_METADATA_MIN_BLOB_SIZE) {
+                delete p.location;
+            }
+        } catch (e) {
+            this.logger.warn('Could not parse Object Metadata while listing', {
+                method: 'DelimiterOrphanDeleteMarker._parse',
+                err: e.toString(),
+            });
+        }
+        return p;
+    }
+
+    _stringify(value) {
+        const p = value;
+        let s = undefined;
+        try {
+            s = JSON.stringify(p);
+        } catch (e) {
+            this.logger.warn('could not stringify Object Metadata while listing',
+                {
+                    method: 'DelimiterOrphanDeleteMarker._stringify',
+                    err: e.toString(),
+                });
+        }
+        return s;
+    }
+    /**
+     * The purpose of _isMaxScannedEntriesReached is to restrict the number of scanned entries,
+     * thus controlling resource overhead (CPU...).
+     * @return {boolean} isMaxScannedEntriesReached - true if the maximum limit on the number
+     * of entries scanned has been reached, false otherwise.
+    */
+    _isMaxScannedEntriesReached() {
+        return this.maxScannedLifecycleListingEntries && this.scannedKeys >= this.maxScannedLifecycleListingEntries;
+    }
+
+    filter(obj) {
+        if (this._isMaxScannedEntriesReached()) {
+            this.nextKeyMarker = this.prevKeyName;
+            this.IsTruncated = true;
+            this.logger.info('listing stopped due to reaching the maximum scanned entries limit',
+                {
+                    maxScannedLifecycleListingEntries: this.maxScannedLifecycleListingEntries,
+                    scannedKeys: this.scannedKeys,
+                });
+            return FILTER_END;
+        }
+        ++this.scannedKeys;
+        return super.filter(obj);
+    }
+
+    /**
+     * NOTE: Each version of a specific key is sorted from the latest to the oldest
+     * thanks to the way version ids are generated.
+     * DESCRIPTION: For a given key, the latest version is kept in memory since it is the current version.
+     * If the following version reference a new key, it means that the previous one was an orphan version.
+     * We then check if the orphan version is a delete marker and if it is older than the "beforeDate"
+     * The process stops and returns the available results if either:
+     * - no more metadata key is left to be processed
+     * - the listing reaches the maximum number of key to be returned
+     * - the internal timeout is reached
+     * NOTE: we cannot leverage MongoDB to list keys older than "beforeDate"
+     * because then we will not be able to assess its orphanage.
+     *  @param {String} key   - The object key.
+     * @param {String} versionId   - The object version id.
+     *  @param {String} value - The value of the key
+     *  @return {undefined}
+     */
+    addVersion(key, versionId, value) {
+        // For a given key, the youngest version is kept in memory since it represents the current version.
+        if (key !== this.keyName) {
+            // If this.value is defined, it means that <this.keyName, this.value> pair is "allowed" to be an orphan.
+            if (this.value) {
+                this._addOrphan();
+            }
+            this.prevKeyName = this.keyName;
+            this.keyName = key;
+            this.value = value;
+
+            return;
+        }
+
+        // If the key is not the current version, we can skip it in the next listing
+        // in the case where the current listing is interrupted due to reaching the maximum scanned entries.
+        this.prevKeyName = key;
+        this.keyName = key;
+        this.value = null;
+
+        return;
+    }
+
+    result() {
+        // Only check for remaining last orphan delete marker if the listing is not interrupted.
+        // This will help avoid false positives.
+        if (!this._isMaxScannedEntriesReached()) {
+            // The following check makes sure the last orphan delete marker is not forgotten.
+            if (this.keys < this.maxKeys) {
+                if (this.value) {
+                    this._addOrphan();
+                }
+            // The following make sure that if makeKeys is reached, isTruncated is set to true.
+            // We moved the "isTruncated" from _reachedMaxKeys to make sure we take into account the last entity
+            // if listing is truncated right before the last entity and the last entity is a orphan delete marker.
+            } else {
+                this.IsTruncated = this.maxKeys > 0;
+            }
+        }
+
+        const result = {
+            Contents: this.Versions,
+            IsTruncated: this.IsTruncated,
+        };
+
+        if (this.IsTruncated) {
+            result.NextMarker = this.nextKeyMarker;
+        }
+
+        return result;
+    }
+}
+
+module.exports = { DelimiterOrphanDeleteMarker };

lib/algos/list/delimiterVersions.js

@@ -1,283 +0,0 @@
-'use strict'; // eslint-disable-line strict
-
-const Delimiter = require('./delimiter').Delimiter;
-const Version = require('../../versioning/Version').Version;
-const VSConst = require('../../versioning/constants').VersioningConstants;
-const { inc, FILTER_END, FILTER_ACCEPT, FILTER_SKIP, SKIP_NONE } =
-    require('./tools');
-
-const VID_SEP = VSConst.VersionId.Separator;
-const { DbPrefixes, BucketVersioningKeyFormat } = VSConst;
-
-/**
- * Handle object listing with parameters
- *
- * @prop {String[]} CommonPrefixes     - 'folders' defined by the delimiter
- * @prop {String[]} Contents           - 'files' to list
- * @prop {Boolean} IsTruncated         - truncated listing flag
- * @prop {String|undefined} NextMarker - marker per amazon format
- * @prop {Number} keys                 - count of listed keys
- * @prop {String|undefined} delimiter  - separator per amazon format
- * @prop {String|undefined} prefix     - prefix per amazon format
- * @prop {Number} maxKeys              - number of keys to list
- */
-class DelimiterVersions extends Delimiter {
-    constructor(parameters, logger, vFormat) {
-        super(parameters, logger, vFormat);
-        // specific to version listing
-        this.keyMarker = parameters.keyMarker;
-        this.versionIdMarker = parameters.versionIdMarker;
-        // internal state
-        this.masterKey = undefined;
-        this.masterVersionId = undefined;
-        // listing results
-        this.NextMarker = parameters.keyMarker;
-        this.NextVersionIdMarker = undefined;
-        this.inReplayPrefix = false;
-
-        Object.assign(this, {
-            [BucketVersioningKeyFormat.v0]: {
-                genMDParams: this.genMDParamsV0,
-                filter: this.filterV0,
-                skipping: this.skippingV0,
-            },
-            [BucketVersioningKeyFormat.v1]: {
-                genMDParams: this.genMDParamsV1,
-                filter: this.filterV1,
-                skipping: this.skippingV1,
-            },
-        }[this.vFormat]);
-    }
-
-    genMDParamsV0() {
-        const params = {};
-        if (this.parameters.prefix) {
-            params.gte = this.parameters.prefix;
-            params.lt = inc(this.parameters.prefix);
-        }
-        if (this.parameters.keyMarker) {
-            if (params.gte && params.gte > this.parameters.keyMarker) {
-                return params;
-            }
-            delete params.gte;
-            if (this.parameters.versionIdMarker) {
-                // versionIdMarker should always come with keyMarker
-                // but may not be the other way around
-                params.gt = this.parameters.keyMarker
-                    + VID_SEP
-                    + this.parameters.versionIdMarker;
-            } else {
-                params.gt = inc(this.parameters.keyMarker + VID_SEP);
-            }
-        }
-        return params;
-    }
-
-    genMDParamsV1() {
-        // return an array of two listing params sets to ask for
-        // synchronized listing of M and V ranges
-        const params = [{}, {}];
-        if (this.parameters.prefix) {
-            params[0].gte = DbPrefixes.Master + this.parameters.prefix;
-            params[0].lt = DbPrefixes.Master + inc(this.parameters.prefix);
-            params[1].gte = DbPrefixes.Version + this.parameters.prefix;
-            params[1].lt = DbPrefixes.Version + inc(this.parameters.prefix);
-        } else {
-            params[0].gte = DbPrefixes.Master;
-            params[0].lt = inc(DbPrefixes.Master); // stop after the last master key
-            params[1].gte = DbPrefixes.Version;
-            params[1].lt = inc(DbPrefixes.Version); // stop after the last version key
-        }
-        if (this.parameters.keyMarker) {
-            if (params[1].gte <= DbPrefixes.Version + this.parameters.keyMarker) {
-                delete params[0].gte;
-                delete params[1].gte;
-                params[0].gt = DbPrefixes.Master + inc(this.parameters.keyMarker + VID_SEP);
-                if (this.parameters.versionIdMarker) {
-                    // versionIdMarker should always come with keyMarker
-                    // but may not be the other way around
-                    params[1].gt = DbPrefixes.Version
-                        + this.parameters.keyMarker
-                        + VID_SEP
-                        + this.parameters.versionIdMarker;
-                } else {
-                    params[1].gt = DbPrefixes.Version
-                        + inc(this.parameters.keyMarker + VID_SEP);
-                }
-            }
-        }
-        return params;
-    }
-
-    /**
-     * Used to synchronize listing of M and V prefixes by object key
-     *
-     * @param {object} masterObj object listed from first range
-     * returned by genMDParamsV1() (the master keys range)
-     * @param {object} versionObj object listed from second range
-     * returned by genMDParamsV1() (the version keys range)
-     * @return {number} comparison result:
-     *   * -1 if master key < version key
-     *   * 1 if master key > version key
-     */
-    compareObjects(masterObj, versionObj) {
-        const masterKey = masterObj.key.slice(DbPrefixes.Master.length);
-        const versionKey = versionObj.key.slice(DbPrefixes.Version.length);
-        return masterKey < versionKey ? -1 : 1;
-    }
-
-    /**
-     *  Add a (key, versionId, value) tuple to the listing.
-     *  Set the NextMarker to the current key
-     *  Increment the keys counter
-     *  @param {object} obj             - the entry to add to the listing result
-     *  @param {String} obj.key         - The key to add
-     *  @param {String} obj.versionId   - versionId
-     *  @param {String} obj.value       - The value of the key
-     *  @return {Boolean} - indicates if iteration should continue
-     */
-    addContents(obj) {
-        if (this._reachedMaxKeys()) {
-            return FILTER_END;
-        }
-        this.Contents.push({
-            key: obj.key,
-            value: this.trimMetadata(obj.value),
-            versionId: obj.versionId,
-        });
-        this.NextMarker = obj.key;
-        this.NextVersionIdMarker = obj.versionId;
-        ++this.keys;
-        return FILTER_ACCEPT;
-    }
-
-    /**
-     *  Filter to apply on each iteration if bucket is in v0
-     *  versioning key format, based on:
-     *  - prefix
-     *  - delimiter
-     *  - maxKeys
-     *  The marker is being handled directly by levelDB
-     *  @param {Object} obj       - The key and value of the element
-     *  @param {String} obj.key   - The key of the element
-     *  @param {String} obj.value - The value of the element
-     *  @return {number}          - indicates if iteration should continue
-     */
-    filterV0(obj) {
-        if (obj.key.startsWith(DbPrefixes.Replay)) {
-            this.inReplayPrefix = true;
-            return FILTER_SKIP;
-        }
-        this.inReplayPrefix = false;
-
-        if (Version.isPHD(obj.value)) {
-            // return accept to avoid skipping the next values in range
-            return FILTER_ACCEPT;
-        }
-        return this.filterCommon(obj.key, obj.value);
-    }
-
-    /**
-     *  Filter to apply on each iteration if bucket is in v1
-     *  versioning key format, based on:
-     *  - prefix
-     *  - delimiter
-     *  - maxKeys
-     *  The marker is being handled directly by levelDB
-     *  @param {Object} obj       - The key and value of the element
-     *  @param {String} obj.key   - The key of the element
-     *  @param {String} obj.value - The value of the element
-     *  @return {number}          - indicates if iteration should continue
-     */
-    filterV1(obj) {
-        if (Version.isPHD(obj.value)) {
-            // return accept to avoid skipping the next values in range
-            return FILTER_ACCEPT;
-        }
-        // this function receives both M and V keys, but their prefix
-        // length is the same so we can remove their prefix without
-        // looking at the type of key
-        return this.filterCommon(obj.key.slice(DbPrefixes.Master.length),
-            obj.value);
-    }
-
-    filterCommon(key, value) {
-        if (this.prefix && !key.startsWith(this.prefix)) {
-            return FILTER_SKIP;
-        }
-        let nonversionedKey;
-        let versionId = undefined;
-        const versionIdIndex = key.indexOf(VID_SEP);
-        if (versionIdIndex < 0) {
-            nonversionedKey = key;
-            this.masterKey = key;
-            this.masterVersionId =
-                Version.from(value).getVersionId() || 'null';
-            versionId = this.masterVersionId;
-        } else {
-            nonversionedKey = key.slice(0, versionIdIndex);
-            versionId = key.slice(versionIdIndex + 1);
-            // skip a version key if it is the master version
-            if (this.masterKey === nonversionedKey && this.masterVersionId === versionId) {
-                return FILTER_SKIP;
-            }
-            this.masterKey = undefined;
-            this.masterVersionId = undefined;
-        }
-        if (this.delimiter) {
-            const baseIndex = this.prefix ? this.prefix.length : 0;
-            const delimiterIndex = nonversionedKey.indexOf(this.delimiter, baseIndex);
-            if (delimiterIndex >= 0) {
-                return this.addCommonPrefix(nonversionedKey, delimiterIndex);
-            }
-        }
-        return this.addContents({ key: nonversionedKey, value, versionId });
-    }
-
-    skippingV0() {
-        if (this.inReplayPrefix) {
-            return DbPrefixes.Replay;
-        }
-        if (this.NextMarker) {
-            const index = this.NextMarker.lastIndexOf(this.delimiter);
-            if (index === this.NextMarker.length - 1) {
-                return this.NextMarker;
-            }
-        }
-        return SKIP_NONE;
-    }
-
-    skippingV1() {
-        const skipV0 = this.skippingV0();
-        if (skipV0 === SKIP_NONE) {
-            return SKIP_NONE;
-        }
-        // skip to the same object key in both M and V range listings
-        return [DbPrefixes.Master + skipV0,
-            DbPrefixes.Version + skipV0];
-    }
-
-    /**
-     *  Return an object containing all mandatory fields to use once the
-     *  iteration is done, doesn't show a NextMarker field if the output
-     *  isn't truncated
-     *  @return {Object} - following amazon format
-     */
-    result() {
-        /* NextMarker is only provided when delimiter is used.
-         * specified in v1 listing documentation
-         * http://docs.aws.amazon.com/AmazonS3/latest/API/RESTBucketGET.html
-         */
-        return {
-            CommonPrefixes: this.CommonPrefixes,
-            Versions: this.Contents,
-            IsTruncated: this.IsTruncated,
-            NextKeyMarker: this.IsTruncated ? this.NextMarker : undefined,
-            NextVersionIdMarker: this.IsTruncated ?
-                this.NextVersionIdMarker : undefined,
-            Delimiter: this.delimiter,
-        };
-    }
-}
-
-module.exports = { DelimiterVersions };

lib/algos/list/delimiterVersions.ts

@@ -0,0 +1,535 @@
+'use strict'; // eslint-disable-line strict
+
+const Extension = require('./Extension').default;
+
+import {
+    FilterState,
+    FilterReturnValue,
+} from './delimiter';
+
+const Version = require('../../versioning/Version').Version;
+const VSConst = require('../../versioning/constants').VersioningConstants;
+const { inc, FILTER_END, FILTER_ACCEPT, FILTER_SKIP, SKIP_NONE } =
+    require('./tools');
+
+const VID_SEP = VSConst.VersionId.Separator;
+const { DbPrefixes, BucketVersioningKeyFormat } = VSConst;
+
+export const enum DelimiterVersionsFilterStateId {
+    NotSkipping = 1,
+    SkippingPrefix = 2,
+    SkippingVersions = 3,
+};
+
+export interface DelimiterVersionsFilterState_NotSkipping extends FilterState {
+    id: DelimiterVersionsFilterStateId.NotSkipping,
+};
+
+export interface DelimiterVersionsFilterState_SkippingPrefix extends FilterState {
+    id: DelimiterVersionsFilterStateId.SkippingPrefix,
+    prefix: string;
+};
+
+export interface DelimiterVersionsFilterState_SkippingVersions extends FilterState {
+    id: DelimiterVersionsFilterStateId.SkippingVersions,
+    gt: string;
+};
+
+type KeyHandler = (key: string, versionId: string | undefined, value: string) => FilterReturnValue;
+
+type ResultObject = {
+    CommonPrefixes: string[],
+    Versions: {
+        key: string;
+        value: string;
+        versionId: string;
+    }[];
+    IsTruncated: boolean;
+    Delimiter ?: string;
+    NextKeyMarker ?: string;
+    NextVersionIdMarker ?: string;
+};
+
+type GenMDParamsItem = {
+    gt ?: string,
+    gte ?: string,
+    lt ?: string,
+};
+
+/**
+ * Handle object listing with parameters
+ *
+ * @prop {String[]} CommonPrefixes     - 'folders' defined by the delimiter
+ * @prop {String[]} Contents           - 'files' to list
+ * @prop {Boolean} IsTruncated         - truncated listing flag
+ * @prop {String|undefined} NextMarker - marker per amazon format
+ * @prop {Number} keys                 - count of listed keys
+ * @prop {String|undefined} delimiter  - separator per amazon format
+ * @prop {String|undefined} prefix     - prefix per amazon format
+ * @prop {Number} maxKeys              - number of keys to list
+ */
+export class DelimiterVersions extends Extension {
+
+    state: FilterState;
+    keyHandlers: { [id: number]: KeyHandler };
+
+    constructor(parameters, logger, vFormat) {
+        super(parameters, logger);
+        // original listing parameters
+        this.delimiter = parameters.delimiter;
+        this.prefix = parameters.prefix;
+        this.maxKeys = parameters.maxKeys || 1000;
+        // specific to version listing
+        this.keyMarker = parameters.keyMarker;
+        this.versionIdMarker = parameters.versionIdMarker;
+        // internal state
+        this.masterKey = undefined;
+        this.masterVersionId = undefined;
+        this.nullKey = null;
+        this.vFormat = vFormat || BucketVersioningKeyFormat.v0;
+        // listing results
+        this.CommonPrefixes = [];
+        this.Versions = [];
+        this.IsTruncated = false;
+        this.nextKeyMarker = parameters.keyMarker;
+        this.nextVersionIdMarker = undefined;
+
+        this.keyHandlers = {};
+
+        Object.assign(this, {
+            [BucketVersioningKeyFormat.v0]: {
+                genMDParams: this.genMDParamsV0,
+                getObjectKey: this.getObjectKeyV0,
+                skipping: this.skippingV0,
+            },
+            [BucketVersioningKeyFormat.v1]: {
+                genMDParams: this.genMDParamsV1,
+                getObjectKey: this.getObjectKeyV1,
+                skipping: this.skippingV1,
+            },
+        }[this.vFormat]);
+
+        if (this.vFormat === BucketVersioningKeyFormat.v0) {
+            this.setKeyHandler(
+                DelimiterVersionsFilterStateId.NotSkipping,
+                this.keyHandler_NotSkippingV0.bind(this));
+        } else {
+            this.setKeyHandler(
+                DelimiterVersionsFilterStateId.NotSkipping,
+                this.keyHandler_NotSkippingV1.bind(this));
+        }
+        this.setKeyHandler(
+            DelimiterVersionsFilterStateId.SkippingPrefix,
+            this.keyHandler_SkippingPrefix.bind(this));
+
+        this.setKeyHandler(
+            DelimiterVersionsFilterStateId.SkippingVersions,
+            this.keyHandler_SkippingVersions.bind(this));
+
+        if (this.versionIdMarker) {
+            this.state = <DelimiterVersionsFilterState_SkippingVersions> {
+                id: DelimiterVersionsFilterStateId.SkippingVersions,
+                gt: `${this.keyMarker}${VID_SEP}${this.versionIdMarker}`,
+            };
+        } else {
+            this.state = <DelimiterVersionsFilterState_NotSkipping> {
+                id: DelimiterVersionsFilterStateId.NotSkipping,
+            };
+        }
+    }
+
+    genMDParamsV0() {
+        const params: GenMDParamsItem = {};
+        if (this.prefix) {
+            params.gte = this.prefix;
+            params.lt = inc(this.prefix);
+        }
+        if (this.keyMarker && this.delimiter) {
+            const commonPrefix = this.getCommonPrefix(this.keyMarker);
+            if (commonPrefix) {
+                const afterPrefix = inc(commonPrefix);
+                if (!params.gte || afterPrefix > params.gte) {
+                    params.gte = afterPrefix;
+                }
+            }
+        }
+        if (this.keyMarker && (!params.gte || this.keyMarker >= params.gte)) {
+            delete params.gte;
+            if (this.versionIdMarker) {
+                // start from the beginning of versions so we can
+                // check if there's a null key and fetch it
+                // (afterwards, we can skip the rest of versions until
+                // we reach versionIdMarker)
+                params.gte = `${this.keyMarker}${VID_SEP}`;
+            } else {
+                params.gt = `${this.keyMarker}${inc(VID_SEP)}`;
+            }
+        }
+        return params;
+    }
+
+    genMDParamsV1() {
+        // return an array of two listing params sets to ask for
+        // synchronized listing of M and V ranges
+        const v0Params: GenMDParamsItem = this.genMDParamsV0();
+        const mParams: GenMDParamsItem = {};
+        const vParams: GenMDParamsItem = {};
+        if (v0Params.gt) {
+            mParams.gt = `${DbPrefixes.Master}${v0Params.gt}`;
+            vParams.gt = `${DbPrefixes.Version}${v0Params.gt}`;
+        } else if (v0Params.gte) {
+            mParams.gte = `${DbPrefixes.Master}${v0Params.gte}`;
+            vParams.gte = `${DbPrefixes.Version}${v0Params.gte}`;
+        } else {
+            mParams.gte = DbPrefixes.Master;
+            vParams.gte = DbPrefixes.Version;
+        }
+        if (v0Params.lt) {
+            mParams.lt = `${DbPrefixes.Master}${v0Params.lt}`;
+            vParams.lt = `${DbPrefixes.Version}${v0Params.lt}`;
+        } else {
+            mParams.lt = inc(DbPrefixes.Master);
+            vParams.lt = inc(DbPrefixes.Version);
+        }
+        return [mParams, vParams];
+    }
+
+    /**
+     * check if the max keys count has been reached and set the
+     * final state of the result if it is the case
+     * @return {Boolean} - indicates if the iteration has to stop
+     */
+    _reachedMaxKeys(): boolean {
+        if (this.keys >= this.maxKeys) {
+            // In cases of maxKeys <= 0 -> IsTruncated = false
+            this.IsTruncated = this.maxKeys > 0;
+            return true;
+        }
+        return false;
+    }
+
+    /**
+     * Used to synchronize listing of M and V prefixes by object key
+     *
+     * @param {object} masterObj object listed from first range
+     * returned by genMDParamsV1() (the master keys range)
+     * @param {object} versionObj object listed from second range
+     * returned by genMDParamsV1() (the version keys range)
+     * @return {number} comparison result:
+     *   * -1 if master key < version key
+     *   * 1 if master key > version key
+     */
+    compareObjects(masterObj, versionObj) {
+        const masterKey = masterObj.key.slice(DbPrefixes.Master.length);
+        const versionKey = versionObj.key.slice(DbPrefixes.Version.length);
+        return masterKey < versionKey ? -1 : 1;
+    }
+
+    /**
+     * Parse a listing key into its nonversioned key and version ID components
+     *
+     * @param {string} key - full listing key
+     * @return {object} obj
+     * @return {string} obj.key - nonversioned part of key
+     * @return {string} [obj.versionId] - version ID in the key
+     */
+    parseKey(fullKey: string): { key: string, versionId ?: string } {
+        const versionIdIndex = fullKey.indexOf(VID_SEP);
+        if (versionIdIndex === -1) {
+            return { key: fullKey };
+        }
+        const nonversionedKey: string = fullKey.slice(0, versionIdIndex);
+        let versionId: string = fullKey.slice(versionIdIndex + 1);
+        return { key: nonversionedKey, versionId };
+    }
+
+    /**
+     * Include a key in the listing output, in the Versions or CommonPrefix result
+     *
+     * @param {string} key - key (without version ID)
+     * @param {string} versionId - version ID
+     * @param {string} value - metadata value
+     * @return {undefined}
+     */
+    addKey(key: string, versionId: string, value: string) {
+        // add the subprefix to the common prefixes if the key has the delimiter
+        const commonPrefix = this.getCommonPrefix(key);
+        if (commonPrefix) {
+            this.addCommonPrefix(commonPrefix);
+            // transition into SkippingPrefix state to skip all following keys
+            // while they start with the same prefix
+            this.setState(<DelimiterVersionsFilterState_SkippingPrefix> {
+                id: DelimiterVersionsFilterStateId.SkippingPrefix,
+                prefix: commonPrefix,
+            });
+        } else {
+            this.addVersion(key, versionId, value);
+        }
+    }
+
+    /**
+     *  Add a (key, versionId, value) tuple to the listing.
+     *  Set the NextMarker to the current key
+     *  Increment the keys counter
+     *  @param {String} key         - The key to add
+     *  @param {String} versionId   - versionId
+     *  @param {String} value       - The value of the key
+     *  @return {undefined}
+     */
+    addVersion(key: string, versionId: string, value: string) {
+        this.Versions.push({
+            key,
+            versionId,
+            value: this.trimMetadata(value),
+        });
+        this.nextKeyMarker = key;
+        this.nextVersionIdMarker = versionId;
+        ++this.keys;
+    }
+
+    getCommonPrefix(key: string): string | undefined {
+        if (!this.delimiter) {
+            return undefined;
+        }
+        const baseIndex = this.prefix ? this.prefix.length : 0;
+        const delimiterIndex = key.indexOf(this.delimiter, baseIndex);
+        if (delimiterIndex === -1) {
+            return undefined;
+        }
+        return key.substring(0, delimiterIndex + this.delimiter.length);
+    }
+
+    /**
+     * Add a Common Prefix in the list
+     * @param {String} commonPrefix   - common prefix to add
+     * @return {undefined}
+     */
+    addCommonPrefix(commonPrefix: string): void {
+        // add the new prefix to the list
+        this.CommonPrefixes.push(commonPrefix);
+        ++this.keys;
+        this.nextKeyMarker = commonPrefix;
+        this.nextVersionIdMarker = undefined;
+    }
+
+    /**
+     * Cache the current null key, to save it for outputting it later at
+     * the correct position
+     *
+     * @param {String} key        - nonversioned key of the null key
+     * @param {String} versionId  - real version ID of the null key
+     * @param {String} value      - value of the null key
+     * @return {undefined}
+     */
+    cacheNullKey(key: string, versionId: string, value: string): void {
+        this.nullKey = { key, versionId, value };
+    }
+
+    getObjectKeyV0(obj: { key: string }): string {
+        return obj.key;
+    }
+
+    getObjectKeyV1(obj: { key: string }): string {
+        return obj.key.slice(DbPrefixes.Master.length);
+    }
+
+    /**
+     *  Filter to apply on each iteration, based on:
+     *  - prefix
+     *  - delimiter
+     *  - maxKeys
+     *  The marker is being handled directly by levelDB
+     *  @param {Object} obj       - The key and value of the element
+     *  @param {String} obj.key   - The key of the element
+     *  @param {String} obj.value - The value of the element
+     *  @return {number}          - indicates if iteration should continue
+     */
+    filter(obj: { key: string, value: string }): FilterReturnValue {
+        const key = this.getObjectKey(obj);
+        const value = obj.value;
+
+        const { key: nonversionedKey, versionId: keyVersionId } = this.parseKey(key);
+        if (this.nullKey) {
+            if (this.nullKey.key !== nonversionedKey
+                || this.nullKey.versionId < <string> keyVersionId) {
+                this.handleKey(
+                    this.nullKey.key, this.nullKey.versionId, this.nullKey.value);
+                this.nullKey = null;
+            }
+        }
+        if (keyVersionId === '') {
+            // null key
+            this.cacheNullKey(nonversionedKey, Version.from(value).getVersionId(), value);
+            if (this.state.id === DelimiterVersionsFilterStateId.SkippingVersions) {
+                return FILTER_SKIP;
+            }
+            return FILTER_ACCEPT;
+        }
+        return this.handleKey(nonversionedKey, keyVersionId, value);
+    }
+
+    setState(state: FilterState): void {
+        this.state = state;
+    }
+
+    setKeyHandler(stateId: number, keyHandler: KeyHandler): void {
+        this.keyHandlers[stateId] = keyHandler;
+    }
+
+    handleKey(key: string, versionId: string | undefined, value: string): FilterReturnValue {
+        return this.keyHandlers[this.state.id](key, versionId, value);
+    }
+
+    keyHandler_NotSkippingV0(key: string, versionId: string | undefined, value: string): FilterReturnValue {
+        if (key.startsWith(DbPrefixes.Replay)) {
+            // skip internal replay prefix entirely
+            this.setState(<DelimiterVersionsFilterState_SkippingPrefix> {
+                id: DelimiterVersionsFilterStateId.SkippingPrefix,
+                prefix: DbPrefixes.Replay,
+            });
+            return FILTER_SKIP;
+        }
+        if (Version.isPHD(value)) {
+            return FILTER_ACCEPT;
+        }
+        return this.filter_onNewKey(key, versionId, value);
+    }
+
+    keyHandler_NotSkippingV1(key: string, versionId: string | undefined, value: string): FilterReturnValue {
+        // NOTE: this check on PHD is only useful for Artesca, S3C
+        // does not use PHDs in V1 format
+        if (Version.isPHD(value)) {
+            return FILTER_ACCEPT;
+        }
+        return this.filter_onNewKey(key, versionId, value);
+    }
+
+    filter_onNewKey(key: string, versionId: string | undefined, value: string): FilterReturnValue {
+        if (this._reachedMaxKeys()) {
+            return FILTER_END;
+        }
+        if (versionId === undefined) {
+            this.masterKey = key;
+            this.masterVersionId = Version.from(value).getVersionId() || 'null';
+            this.addKey(this.masterKey, this.masterVersionId, value);
+        } else {
+            if (this.masterKey === key && this.masterVersionId === versionId) {
+                // do not add a version key if it is the master version
+                return FILTER_ACCEPT;
+            }
+            this.addKey(key, versionId, value);
+        }
+        return FILTER_ACCEPT;
+    }
+
+    keyHandler_SkippingPrefix(key: string, versionId: string | undefined, value: string): FilterReturnValue {
+        const { prefix } = <DelimiterVersionsFilterState_SkippingPrefix> this.state;
+        if (key.startsWith(prefix)) {
+            return FILTER_SKIP;
+        }
+        this.setState(<DelimiterVersionsFilterState_NotSkipping> {
+            id: DelimiterVersionsFilterStateId.NotSkipping,
+        });
+        return this.handleKey(key, versionId, value);
+    }
+
+    keyHandler_SkippingVersions(key: string, versionId: string | undefined, value: string): FilterReturnValue {
+        if (key === this.keyMarker) {
+            // since the nonversioned key equals the marker, there is
+            // necessarily a versionId in this key
+            const _versionId = <string> versionId;
+            if (_versionId < this.versionIdMarker) {
+                // skip all versions until marker
+                return FILTER_SKIP;
+            }
+            if (_versionId === this.versionIdMarker) {
+                // nothing left to skip, so return ACCEPT, but don't add this version
+                return FILTER_ACCEPT;
+            }
+        }
+        this.setState(<DelimiterVersionsFilterState_NotSkipping> {
+            id: DelimiterVersionsFilterStateId.NotSkipping,
+        });
+        return this.handleKey(key, versionId, value);
+    }
+
+    skippingBase(): string | undefined {
+        switch (this.state.id) {
+        case DelimiterVersionsFilterStateId.SkippingPrefix:
+            const { prefix } = <DelimiterVersionsFilterState_SkippingPrefix> this.state;
+            return inc(prefix);
+
+        case DelimiterVersionsFilterStateId.SkippingVersions:
+            const { gt } = <DelimiterVersionsFilterState_SkippingVersions> this.state;
+            // the contract of skipping() is to return the first key
+            // that can be skipped to, so adding a null byte to skip
+            // over the existing versioned key set in 'gt'
+            return `${gt}\0`;
+
+        default:
+            return SKIP_NONE;
+        }
+    }
+
+    skippingV0() {
+        return this.skippingBase();
+    }
+
+    skippingV1() {
+        const skipTo = this.skippingBase();
+        if (skipTo === SKIP_NONE) {
+            return SKIP_NONE;
+        }
+        // skip to the same object key in both M and V range listings
+        return [
+            `${DbPrefixes.Master}${skipTo}`,
+            `${DbPrefixes.Version}${skipTo}`,
+        ];
+    }
+
+    /**
+     *  Return an object containing all mandatory fields to use once the
+     *  iteration is done, doesn't show a NextMarker field if the output
+     *  isn't truncated
+     *  @return {Object} - following amazon format
+     */
+    result() {
+        // Add the last null key if still in cache (when it is the
+        // last version of the last key)
+        //
+        // NOTE: _reachedMaxKeys sets IsTruncated to true when it
+        // returns true. Here we want this because either:
+        //
+        // - we did not reach the max keys yet so the result is not
+        // - truncated, and there is still room for the null key in
+        // - the results
+        //
+        // - OR we reached it already while having to process a new
+        //   key (so the result is truncated even without the null key)
+        //
+        // - OR we are *just* below the limit but the null key to add
+        //   does not fit, so we know the result is now truncated
+        //   because there remains the null key to be output.
+        //
+        if (this.nullKey) {
+            this.handleKey(this.nullKey.key, this.nullKey.versionId, this.nullKey.value);
+        }
+        const result: ResultObject = {
+            CommonPrefixes: this.CommonPrefixes,
+            Versions: this.Versions,
+            IsTruncated: this.IsTruncated,
+        };
+        if (this.delimiter) {
+            result.Delimiter = this.delimiter;
+        }
+        if (this.IsTruncated) {
+            result.NextKeyMarker = this.nextKeyMarker;
+            if (this.nextVersionIdMarker) {
+                result.NextVersionIdMarker = this.nextVersionIdMarker;
+            }
+        };
+        return result;
+    }
+}
+
+module.exports = { DelimiterVersions };

lib/algos/list/exportAlgos.js

@@ -6,4 +6,7 @@ module.exports = {
     DelimiterMaster: require('./delimiterMaster')
         .DelimiterMaster,
     MPU: require('./MPU').MultipartUploads,
+    DelimiterCurrent: require('./delimiterCurrent').DelimiterCurrent,
+    DelimiterNonCurrent: require('./delimiterNonCurrent').DelimiterNonCurrent,
+    DelimiterOrphanDeleteMarker: require('./delimiterOrphanDeleteMarker').DelimiterOrphanDeleteMarker,
 };

lib/algos/list/skip.js

@@ -52,21 +52,21 @@ class Skip {
         assert(this.skipRangeCb);
 
         const filteringResult = this.extension.filter(entry);
-        const skippingRange = this.extension.skipping();
+        const skipTo = this.extension.skipping();
 
         if (filteringResult === FILTER_END) {
             this.listingEndCb();
         } else if (filteringResult === FILTER_SKIP
-                   && skippingRange !== SKIP_NONE) {
+                   && skipTo !== SKIP_NONE) {
             if (++this.streakLength >= MAX_STREAK_LENGTH) {
                 let newRange;
-                if (Array.isArray(skippingRange)) {
+                if (Array.isArray(skipTo)) {
                     newRange = [];
-                    for (let i = 0; i < skippingRange.length; ++i) {
-                        newRange.push(this._inc(skippingRange[i]));
+                    for (let i = 0; i < skipTo.length; ++i) {
+                        newRange.push(skipTo[i]);
                     }
                 } else {
-                    newRange = this._inc(skippingRange);
+                    newRange = skipTo;
                 }
                 /* Avoid to loop on the same range again and again. */
                 if (newRange === this.gteParams) {
@@ -79,16 +79,6 @@ class Skip {
             this.streakLength = 0;
         }
     }
-
-    _inc(str) {
-        if (!str) {
-            return str;
-        }
-        const lastCharValue = str.charCodeAt(str.length - 1);
-        const lastCharNewValue = String.fromCharCode(lastCharValue + 1);
-
-        return `${str.slice(0, str.length - 1)}${lastCharNewValue}`;
-    }
 }
 
 

lib/auth/Vault.ts

@@ -14,7 +14,7 @@ function vaultSignatureCb(
     err: Error | null,
     authInfo: { message: { body: any } },
     log: Logger,
-    callback: (err: Error | null, data?: any, results?: any, params?: any) => void,
+    callback: (err: Error | null, data?: any, results?: any, params?: any, infos?: any) => void,
     streamingV4Params?: any
 ) {
     // vaultclient API guarantees that it returns:
@@ -38,7 +38,9 @@ function vaultSignatureCb(
     }
     // @ts-ignore
     log.addDefaultFields(auditLog);
-    return callback(null, userInfo, authorizationResults, streamingV4Params);
+    return callback(null, userInfo, authorizationResults, streamingV4Params, {
+        accountQuota: info.accountQuota || {},
+    });
 }
 
 export type AuthV4RequestParams = {
@@ -384,4 +386,19 @@ export default class Vault {
             return callback(null, respBody);
         });
     }
+
+    report(log: Logger, callback: (err: Error | null, data?: any) => void) {
+        // call the report function of the client
+        if (!this.client.report) {
+            return callback(null, {});
+        }
+        // @ts-ignore
+        return this.client.report(log.getSerializedUids(), (err: Error | null, obj?: any) => {
+            if (err) {
+                log.debug(`error from ${this.implName}`, { error: err });
+                return callback(err);
+            }
+            return callback(null, obj);
+        });
+    }
 }

lib/auth/auth.ts

@@ -163,6 +163,20 @@ function doAuth(
     return cb(errors.InternalError);
 }
 
+/**
+ * This function will generate a version 4 content-md5 header
+ * It looks at the request path to determine what kind of header encoding is required
+ *
+ * @param path - the request path
+ * @param payload - the request payload to hash
+ */
+function generateContentMD5Header(
+    path: string,
+    payload: string,
+) {
+    const encoding = path && path.startsWith('/_/backbeat/') ? 'hex' : 'base64';
+    return crypto.createHash('md5').update(payload, 'binary').digest(encoding);
+}
 /**
  * This function will generate a version 4 header
  *
@@ -175,6 +189,7 @@ function doAuth(
  * @param [proxyPath] - path that gets proxied by reverse proxy
  * @param [sessionToken] - security token if the access/secret keys
  *                                are temporary credentials from STS
+ * @param [payload] - body of the request if any
  */
 function generateV4Headers(
     request: any,
@@ -182,8 +197,9 @@ function generateV4Headers(
     accessKey: string,
     secretKeyValue: string,
     awsService: string,
-    proxyPath: string,
-    sessionToken: string
+    proxyPath?: string,
+    sessionToken?: string,
+    payload?: string,
 ) {
     Object.assign(request, { headers: {} });
     const amzDate = convertUTCtoISO8601(Date.now());
@@ -196,7 +212,7 @@ function generateV4Headers(
     const timestamp = amzDate;
     const algorithm = 'AWS4-HMAC-SHA256';
 
-    let payload = '';
+    payload = payload || '';
     if (request.method === 'POST') {
         payload = queryString.stringify(data, undefined, undefined, {
             encodeURIComponent,
@@ -207,6 +223,7 @@ function generateV4Headers(
     request.setHeader('host', request._headers.host);
     request.setHeader('x-amz-date', amzDate);
     request.setHeader('x-amz-content-sha256', payloadChecksum);
+    request.setHeader('content-md5', generateContentMD5Header(request.path, payload));
 
     if (sessionToken) {
         request.setHeader('x-amz-security-token', sessionToken);
@@ -217,6 +234,7 @@ function generateV4Headers(
         .filter(headerName =>
             headerName.startsWith('x-amz-')
             || headerName.startsWith('x-scal-')
+            || headerName === 'content-md5'
             || headerName === 'host',
         ).sort().join(';');
     const params = { request, signedHeaders, payloadChecksum,

lib/auth/backends/ChainBackend.ts

@@ -133,23 +133,37 @@ export default class ChainBackend extends BaseBackend {
                 return;
             }
 
-            resp.message.body.forEach(policy => {
-                const key = (policy.arn || '') + (policy.versionId || '');
+            const check = (policy) => {
+                const key = (policy.arn || '') + (policy.versionId || '') + (policy.action || '');
                 if (!policyMap[key] || !policyMap[key].isAllowed) {
                     policyMap[key] = policy;
                 }
                 // else is duplicate policy
+            };
+
+            resp.message.body.forEach(policy => {
+                if (Array.isArray(policy)) {
+                    policy.forEach(authResult => check(authResult));
+                } else {
+                    check(policy);
+                }
             });
         });
 
         return Object.keys(policyMap).map(key => {
-            const policyRes: any = { isAllowed: policyMap[key].isAllowed };
+            const policyRes: any = { isAllowed: policyMap[key].isAllowed };
             if (policyMap[key].arn !== '') {
                 policyRes.arn = policyMap[key].arn;
             }
             if (policyMap[key].versionId) {
                 policyRes.versionId = policyMap[key].versionId;
             }
+            if (policyMap[key].isImplicit !== undefined) {
+                policyRes.isImplicit = policyMap[key].isImplicit;
+            }
+            if (policyMap[key].action) {
+                policyRes.action = policyMap[key].action;
+            }
             return policyRes;
         });
     }
@@ -198,4 +212,22 @@ export default class ChainBackend extends BaseBackend {
             return callback(null, res);
         });
     }
+
+    report(reqUid: string, callback: any) {
+        this._forEachClient((client, done) =>
+            client.report(reqUid, done),
+            (err, res) => {
+                if (err) {
+                    return callback(err);
+                }
+                const mergedRes = res.reduce((acc, val) => {
+                    Object.keys(val).forEach(k => {
+                        acc[k] = val[k];
+                    });
+                    return acc;
+                }, {});
+
+                return callback(null, mergedRes);
+            });
+    }
 }

lib/auth/backends/in_memory/Backend.ts

@@ -161,6 +161,10 @@ class InMemoryBackend extends BaseBackend {
         };
         return cb(null, vaultReturnObject);
     }
+
+    report(log: Logger, callback: any) {
+        return callback(null, {});
+    }
 }
 
 

lib/auth/v4/awsURIencode.ts

@@ -35,15 +35,16 @@ export default function awsURIencode(
     encodeSlash?: boolean,
     noEncodeStar?: boolean
 ) {
-    const encSlash = encodeSlash === undefined ? true : encodeSlash;
-    let encoded = '';
     /**
      * Duplicate query params are not suppported by AWS S3 APIs. These params
      * are parsed as Arrays by Node.js HTTP parser which breaks this method
      */
      if (typeof input !== 'string') {
-        return encoded;
+        return '';
     }
+    let encoded = "";
+    const slash = encodeSlash === undefined || encodeSlash ? '%2F' : '/';
+    const star = noEncodeStar !== undefined && noEncodeStar ? '*' : '%2A';
     for (let i = 0; i < input.length; i++) {
         let ch = input.charAt(i);
         if ((ch >= 'A' && ch <= 'Z') ||
@@ -55,9 +56,9 @@ export default function awsURIencode(
         } else if (ch === ' ') {
             encoded = encoded.concat('%20');
         } else if (ch === '/') {
-            encoded = encoded.concat(encSlash ? '%2F' : ch);
+            encoded = encoded.concat(slash);
         } else if (ch === '*') {
-            encoded = encoded.concat(noEncodeStar ? '*' : '%2A');
+            encoded = encoded.concat(star);
         } else {
             if (ch >= '\uD800' && ch <= '\uDBFF') {
                 // If this character is a high surrogate peek the next character

lib/clustering/ClusterRPC.ts

@@ -0,0 +1,569 @@
+import cluster, { Worker } from 'cluster';
+import * as werelogs from 'werelogs';
+
+import { default as errors } from '../../lib/errors';
+
+const rpcLogger = new werelogs.Logger('ClusterRPC');
+
+/**
+ * Remote procedure calls support between cluster workers.
+ *
+ * When using the cluster module, new processes are forked and are
+ * dispatched workloads, usually HTTP requests. The ClusterRPC module
+ * implements a RPC system to send commands to all cluster worker
+ * processes at once from any particular worker, and retrieve their
+ * individual command results, like a distributed map operation.
+ *
+ * The existing nodejs cluster IPC channel is setup from the primary
+ * to each worker, but not between workers, so there has to be a hop
+ * by the primary.
+ *
+ * How a command is treated:
+ *
+ * - a worker sends a command message to the primary
+ *
+ * - the primary then forwards that command to each existing worker
+ *   (including the requestor)
+ *
+ * - each worker then executes the command and returns a result or an
+ *   error
+ *
+ * - the primary gathers all workers results into an array
+ *
+ * - finally, the primary dispatches the results array to the original
+ *   requesting worker
+ *
+ *
+ * Limitations:
+ *
+ * - The command payload must be serializable, which means that:
+ *   - it should not contain circular references
+ *   - it should be of a reasonable size to be sent in a single RPC message
+ *
+ * - The "toWorkers" parameter of value "*" targets the set of workers
+ *   that are available at the time the command is dispatched. Any new
+ *   worker spawned after the command has been dispatched for
+ *   processing, but before the command completes, don't execute
+ *   the command and hence are not part of the results array.
+ *
+ *
+ * To set it up:
+ *
+ * - On the primary:
+ *     if (cluster.isPrimary) {
+ *         setupRPCPrimary();
+ *     }
+ *
+ * - On the workers:
+ *     if (!cluster.isPrimary) {
+ *         setupRPCWorker({
+ *             handler1: (payload: object, uids: string, callback: HandlerCallback) => void,
+ *             handler2: ...
+ *         });
+ *     }
+ *     Handler functions will be passed the command payload, request
+ *     serialized uids, and must call the callback when the worker is done
+ *     processing the command:
+ *         callback(error: Error | null | undefined, result?: any)
+ *
+ * When this setup is done, any worker can start sending commands by calling
+ * the async function sendWorkerCommand().
+ */
+
+// exported types
+
+export type ResultObject = {
+    error: Error | null;
+    result: any;
+};
+
+/**
+ * saved Promise for sendWorkerCommand
+ */
+export type CommandPromise = {
+    resolve: (results?: ResultObject[]) => void;
+    reject: (error: Error) => void;
+    timeout: NodeJS.Timeout | null;
+};
+export type HandlerCallback = (error: (Error & { code?: number }) | null | undefined, result?: any) => void;
+export type HandlerFunction = (payload: object, uids: string, callback: HandlerCallback) => void;
+export type HandlersMap = {
+    [index: string]: HandlerFunction;
+};
+export type PrimaryHandlerFunction = (worker: Worker, payload: object, uids: string, callback: HandlerCallback) => void;
+export type PrimaryHandlersMap = Record<string, PrimaryHandlerFunction>;
+
+// private types
+
+type RPCMessage<T extends string, P> = {
+    type: T;
+    uids: string;
+    payload: P;
+};
+
+type RPCCommandMessage = RPCMessage<'cluster-rpc:command', any> & {
+    toWorkers: string;
+    toHandler: string;
+};
+
+type MarshalledResultObject = {
+    error: string | null;
+    errorCode?: number;
+    result: any;
+};
+
+type RPCCommandResultMessage = RPCMessage<'cluster-rpc:commandResult', MarshalledResultObject>;
+
+type RPCCommandResultsMessage = RPCMessage<'cluster-rpc:commandResults', {
+    results: MarshalledResultObject[];
+}>;
+
+type RPCCommandErrorMessage = RPCMessage<'cluster-rpc:commandError', {
+    error: string;
+}>;
+
+interface RPCSetupOptions {
+    /**
+     * As werelogs is not a peerDependency, arsenal and a parent project
+     * might have their own separate versions duplicated in dependencies.
+     * The config are therefore not shared.
+     * Use this to propagate werelogs config to arsenal's ClusterRPC.
+     */
+    werelogsConfig?: Parameters<typeof werelogs.configure>[0];
+};
+
+/**
+ * In primary: store worker IDs that are waiting to be dispatched
+ * their command's results, as a mapping.
+ */
+const uidsToWorkerId: {
+    [index: string]: number;
+} = {};
+
+
+/**
+ * In primary: store worker responses for commands in progress as a
+ * mapping.
+ *
+ * Result objects are 'null' while the worker is still processing the
+ * command. When a worker finishes processing it stores the result as:
+ *    {
+ *        error: string | null,
+ *        result: any
+ *    }
+ */
+const uidsToCommandResults: {
+    [index: string]: {
+        [index: number]: MarshalledResultObject | null;
+    };
+} = {};
+
+/**
+ * In workers: store promise callbacks for commands waiting to be
+ * dispatched, as a mapping.
+ */
+const uidsToCommandPromise: {
+    [index: string]: CommandPromise;
+} = {};
+
+
+function _isRpcMessage(message) {
+    return (message !== null &&
+            typeof message === 'object' &&
+            typeof message.type === 'string' &&
+            message.type.startsWith('cluster-rpc:'));
+}
+
+/**
+ * Setup cluster RPC system on the primary
+ *
+ * @param {object} [handlers] - mapping of handler names to handler functions
+ *     handler function:
+ *         `handler({Worker} worker, {object} payload, {string} uids, {function} callback)`
+ *     handler callback must be called when worker is done with the command:
+ *         `callback({Error|null} error, {any} [result])`
+ * @return {undefined}
+ */
+export function setupRPCPrimary(handlers?: PrimaryHandlersMap, options?: RPCSetupOptions) {
+    if (options?.werelogsConfig) {
+        werelogs.configure(options.werelogsConfig);
+    }
+    cluster.on('message', (worker, message) => {
+        if (_isRpcMessage(message)) {
+            _handlePrimaryMessage(worker, message, handlers);
+        }
+    });
+}
+
+/**
+ * Setup RPCs on a cluster worker process
+ *
+ * @param {object} handlers - mapping of handler names to handler functions
+ *     handler function:
+ *         handler({object} payload, {string} uids, {function} callback)
+ *     handler callback must be called when worker is done with the command:
+ *         callback({Error|null} error, {any} [result])
+ * @return {undefined}
+ * }
+ */
+export function setupRPCWorker(handlers: HandlersMap, options?: RPCSetupOptions) {
+    if (!process.send) {
+        throw new Error('fatal: cannot setup cluster RPC: "process.send" is not available');
+    }
+    if (options?.werelogsConfig) {
+        werelogs.configure(options.werelogsConfig);
+    }
+    process.on('message', (message: RPCCommandMessage | RPCCommandResultsMessage) => {
+        if (_isRpcMessage(message)) {
+            _handleWorkerMessage(message, handlers);
+        }
+    });
+}
+
+/**
+ * Send a command for workers to execute in parallel, and wait for results
+ *
+ * @param {string} toWorkers - which workers should execute the command
+ *     Currently the supported values are:
+ *         - "*", meaning all workers will execute the command
+ *         - "PRIMARY", meaning primary process will execute the command
+ * @param {string} toHandler - name of handler that will execute the
+ * command in workers, as declared in setupRPCWorker() parameter object
+ * @param {string} uids - unique identifier of the command, must be
+ * unique across all commands in progress
+ * @param {object} payload - message payload, sent as-is to the handler
+ * @param {number} [timeoutMs=60000] - timeout the command with a
+ * "RequestTimeout" error after this number of milliseconds - set to 0
+ * to disable timeouts (the command may then hang forever)
+ * @returns {Promise}
+ */
+export async function sendWorkerCommand(
+    toWorkers: string,
+    toHandler: string,
+    uids: string,
+    payload: object,
+    timeoutMs: number = 60000
+) {
+    if (typeof uids !== 'string') {
+        rpcLogger.error('missing or invalid "uids" field', { uids });
+        throw errors.MissingParameter;
+    }
+    if (uidsToCommandPromise[uids] !== undefined) {
+        rpcLogger.error('a command is already in progress with same uids', { uids });
+        throw errors.OperationAborted;
+    }
+    rpcLogger.info('sending command', { toWorkers, toHandler, uids, payload });
+    return new Promise((resolve, reject) => {
+        let timeout: NodeJS.Timeout | null = null;
+        if (timeoutMs) {
+            timeout = setTimeout(() => {
+                delete uidsToCommandPromise[uids];
+                reject(errors.RequestTimeout);
+            }, timeoutMs);
+        }
+        uidsToCommandPromise[uids] = { resolve, reject, timeout };
+        const message: RPCCommandMessage = {
+            type: 'cluster-rpc:command',
+            toWorkers,
+            toHandler,
+            uids,
+            payload,
+        };
+        return process.send?.(message);
+    });
+}
+
+/**
+ * Get the number of commands in flight
+ * @returns {number}
+ */
+export function getPendingCommandsCount() {
+    return Object.keys(uidsToCommandPromise).length;
+}
+
+
+function _dispatchCommandResultsToWorker(
+    worker: Worker,
+    uids: string,
+    resultsArray: MarshalledResultObject[]
+): void {
+    const message: RPCCommandResultsMessage = {
+        type: 'cluster-rpc:commandResults',
+        uids,
+        payload: {
+            results: resultsArray,
+        },
+    };
+    worker.send(message);
+}
+
+function _dispatchCommandErrorToWorker(
+    worker: Worker,
+    uids: string,
+    error: Error,
+): void {
+    const message: RPCCommandErrorMessage = {
+        type: 'cluster-rpc:commandError',
+        uids,
+        payload: {
+            error: error.message,
+        },
+    };
+    worker.send(message);
+}
+
+function _sendPrimaryCommandResult(
+    worker: Worker,
+    uids: string,
+    error: (Error & { code?: number }) | null | undefined,
+    result?: any
+): void {
+    const message: RPCCommandResultsMessage = {
+        type: 'cluster-rpc:commandResults',
+        uids,
+        payload: {
+            results: [{ error: error?.message || null, errorCode: error?.code, result }],
+        },
+    };
+    worker.send?.(message);
+}
+
+function _handlePrimaryCommandMessage(
+    fromWorker: Worker,
+    logger: any,
+    message: RPCCommandMessage,
+    handlers?: PrimaryHandlersMap
+): void {
+    const { toWorkers, toHandler, uids, payload } = message;
+    if (toWorkers === '*') {
+        if (uidsToWorkerId[uids] !== undefined) {
+            logger.warn('new command already has a waiting worker with same uids', {
+                uids, workerId: uidsToWorkerId[uids],
+            });
+            return undefined;
+        }
+        const commandResults = {};
+        for (const workerId of Object.keys(cluster.workers || {})) {
+            commandResults[workerId] = null;
+        }
+        uidsToWorkerId[uids] = fromWorker?.id;
+        uidsToCommandResults[uids] = commandResults;
+
+        for (const [workerId, worker] of Object.entries(cluster.workers || {})) {
+            logger.debug('sending command message to worker', {
+                workerId, toHandler, payload,
+            });
+            if (worker) {
+                worker.send(message);
+            }
+        }
+    } else if (toWorkers === 'PRIMARY') {
+        const { toHandler, uids, payload } = message;
+        const cb: HandlerCallback = (err, result) => _sendPrimaryCommandResult(fromWorker, uids, err, result);
+
+        if (toHandler in (handlers || {})) {
+            return handlers![toHandler](fromWorker, payload, uids, cb);
+        }
+        logger.error('no such handler in "toHandler" field from worker command message', {
+            toHandler,
+        });
+        return cb(errors.NotImplemented);
+    } else {
+        logger.error('unsupported "toWorkers" field from worker command message', {
+            toWorkers,
+        });
+        if (fromWorker) {
+            _dispatchCommandErrorToWorker(fromWorker, uids, errors.NotImplemented);
+        }
+    }
+}
+
+function _handlePrimaryCommandResultMessage(
+    fromWorkerId: number,
+    logger: any,
+    message: RPCCommandResultMessage
+): void {
+    const { uids, payload } = message;
+    const commandResults = uidsToCommandResults[uids];
+    if (!commandResults) {
+        logger.warn('received command response message from worker for command not in flight', {
+            workerId: fromWorkerId,
+            uids,
+        });
+        return undefined;
+    }
+    if (commandResults[fromWorkerId] === undefined) {
+        logger.warn('received command response message with unexpected worker ID', {
+            workerId: fromWorkerId,
+            uids,
+        });
+        return undefined;
+    }
+    if (commandResults[fromWorkerId] !== null) {
+        logger.warn('ignoring duplicate command response from worker', {
+            workerId: fromWorkerId,
+            uids,
+        });
+        return undefined;
+    }
+    commandResults[fromWorkerId] = payload;
+    const commandResultsArray = Object.values(commandResults);
+    if (commandResultsArray.every(response => response !== null)) {
+        logger.debug('all workers responded to command', { uids });
+        const completeCommandResultsArray = <MarshalledResultObject[]> commandResultsArray;
+        const toWorkerId = uidsToWorkerId[uids];
+        const toWorker = cluster.workers?.[toWorkerId];
+
+        delete uidsToCommandResults[uids];
+        delete uidsToWorkerId[uids];
+
+        if (!toWorker) {
+            logger.warn('worker shut down while its command was executing', {
+                workerId: toWorkerId, uids,
+            });
+            return undefined;
+        }
+        // send back response to original worker
+        _dispatchCommandResultsToWorker(toWorker, uids, completeCommandResultsArray);
+    }
+}
+
+function _handlePrimaryMessage(
+    fromWorker: Worker,
+    message: RPCCommandMessage | RPCCommandResultMessage,
+    handlers?: PrimaryHandlersMap
+): void {
+    const { type: messageType, uids } = message;
+    const logger = rpcLogger.newRequestLoggerFromSerializedUids(uids);
+    logger.debug('primary received message from worker', {
+        workerId: fromWorker?.id, rpcMessage: message,
+    });
+    if (messageType === 'cluster-rpc:command') {
+        return _handlePrimaryCommandMessage(fromWorker, logger, message, handlers);
+    }
+    if (messageType === 'cluster-rpc:commandResult') {
+        return _handlePrimaryCommandResultMessage(fromWorker?.id, logger, message);
+    }
+    logger.error('unsupported message type', {
+        workerId: fromWorker?.id, messageType, uids,
+    });
+    return undefined;
+}
+
+function _sendWorkerCommandResult(
+    uids: string,
+    error: Error | null | undefined,
+    result?: any
+): void {
+    const message: RPCCommandResultMessage = {
+        type: 'cluster-rpc:commandResult',
+        uids,
+        payload: {
+            error: error ? error.message : null,
+            result,
+        },
+    };
+    process.send?.(message);
+}
+
+function _handleWorkerCommandMessage(
+    logger: any,
+    message: RPCCommandMessage,
+    handlers: HandlersMap
+): void {
+    const { toHandler, uids, payload } = message;
+    const cb: HandlerCallback = (err, result) => _sendWorkerCommandResult(uids, err, result);
+
+    if (toHandler in handlers) {
+        return handlers[toHandler](payload, uids, cb);
+    }
+    logger.error('no such handler in "toHandler" field from worker command message', {
+        toHandler,
+    });
+    return cb(errors.NotImplemented);
+}
+
+function _handleWorkerCommandResultsMessage(
+    logger: any,
+    message: RPCCommandResultsMessage,
+): void {
+    const { uids, payload } = message;
+    const { results } = payload;
+    const commandPromise: CommandPromise = uidsToCommandPromise[uids];
+    if (commandPromise === undefined) {
+        logger.error('missing promise for command results', { uids, payload });
+        return undefined;
+    }
+    if (commandPromise.timeout) {
+        clearTimeout(commandPromise.timeout);
+    }
+    delete uidsToCommandPromise[uids];
+    const unmarshalledResults = results.map(workerResult => {
+        let workerError: Error | null = null;
+        if (workerResult.error) {
+            if (workerResult.error in errors) {
+                workerError = errors[workerResult.error];
+            } else {
+                workerError = new Error(workerResult.error);
+            }
+        }
+        if (workerError && workerResult.errorCode) {
+            (workerError as Error & { code: number }).code = workerResult.errorCode;
+        }
+        const unmarshalledResult: ResultObject = {
+            error: workerError,
+            result: workerResult.result,
+        };
+        return unmarshalledResult;
+    });
+    return commandPromise.resolve(unmarshalledResults);
+}
+
+function _handleWorkerCommandErrorMessage(
+    logger: any,
+    message: RPCCommandErrorMessage,
+): void {
+    const { uids, payload } = message;
+    const { error } = payload;
+    const commandPromise: CommandPromise = uidsToCommandPromise[uids];
+    if (commandPromise === undefined) {
+        logger.error('missing promise for command results', { uids, payload });
+        return undefined;
+    }
+    if (commandPromise.timeout) {
+        clearTimeout(commandPromise.timeout);
+    }
+    delete uidsToCommandPromise[uids];
+    let commandError: Error | null = null;
+    if (error in errors) {
+        commandError = errors[error];
+    } else {
+        commandError = new Error(error);
+    }
+    return commandPromise.reject(<Error> commandError);
+}
+
+function _handleWorkerMessage(
+    message: RPCCommandMessage | RPCCommandResultsMessage | RPCCommandErrorMessage,
+    handlers: HandlersMap
+): void {
+    const { type: messageType, uids } = message;
+    const workerId = cluster.worker?.id;
+    const logger = rpcLogger.newRequestLoggerFromSerializedUids(uids);
+    logger.debug('worker received message from primary', {
+        workerId, rpcMessage: message,
+    });
+    if (messageType === 'cluster-rpc:command') {
+        return _handleWorkerCommandMessage(logger, message, handlers);
+    }
+    if (messageType === 'cluster-rpc:commandResults') {
+        return _handleWorkerCommandResultsMessage(logger, message);
+    }
+    if (messageType === 'cluster-rpc:commandError') {
+        return _handleWorkerCommandErrorMessage(logger, message);
+    }
+    logger.error('unsupported message type', {
+        workerId, messageType,
+    });
+    return undefined;
+}

lib/constants.ts

@@ -35,7 +35,13 @@ export const emptyFileMd5 = 'd41d8cd98f00b204e9800998ecf8427e';
 // Version 4 add the Creation-Time and Content-Language attributes,
 //     and add support for x-ms-meta-* headers in UserMetadata
 // Version 5 adds the azureInfo structure
-export const mdModelVersion = 5;
+// Version 6 adds a "deleted" flag that is updated to true before
+// the object gets deleted. This is done to keep object metadata in the
+// oplog when deleting the object, as oplog deletion events don't contain
+// any metadata of the object.
+// version 6 also adds the "isPHD" flag that is used to indicate that the master
+// object is a placeholder and is not up to date.
+export const mdModelVersion = 6;
 /*
  * Splitter is used to build the object name for the overview of a
  * multipart upload and to build the object names for each part of a
@@ -131,6 +137,14 @@ export const supportedNotificationEvents = new Set([
     's3:ObjectTagging:Put',
     's3:ObjectTagging:Delete',
     's3:ObjectAcl:Put',
+    's3:ObjectRestore:*',
+    's3:ObjectRestore:Post',
+    's3:ObjectRestore:Completed',
+    's3:ObjectRestore:Delete',
+    's3:LifecycleTransition',
+    's3:LifecycleExpiration:*',
+    's3:LifecycleExpiration:DeleteMarkerCreated',
+    's3:LifecycleExpiration:Delete',
 ]);
 export const notificationArnPrefix = 'arn:scality:bucketnotif';
 // HTTP server keep-alive timeout is set to a higher value than
@@ -157,3 +171,7 @@ export const maxCachedBuckets = process.env.METADATA_MAX_CACHED_BUCKETS ?
     Number(process.env.METADATA_MAX_CACHED_BUCKETS) : 1000;
 
 export const validRestoreObjectTiers = new Set(['Expedited', 'Standard', 'Bulk']);
+export const maxBatchingConcurrentOperations = 5;
+
+/** For policy resource arn check we allow empty account ID to not break compatibility */
+export const policyArnAllowedEmptyAccountId = ['utapi', 'scuba'];

lib/db.ts

@@ -148,7 +148,7 @@ export class IndexTransaction {
                 'missing condition for conditional put'
             );
         }
-        if (typeof condition.notExists !== 'string') {
+        if (typeof condition.notExists !== 'string' && typeof condition.exists !== 'string') {
             throw propError(
                 'unsupportedConditionalOperation',
                 'missing key or supported condition'

lib/errors/arsenalErrors.ts

@@ -42,7 +42,7 @@ export const BucketAlreadyOwnedByYou: ErrorFormat = {
     code: 409,
 
     description:
-        'Your previous request to create the named bucket succeeded and you already own it. You get this error in all AWS regions except US Standard, us-east-1. In us-east-1 region, you will get 200 OK, but it is no-op (if bucket exists S3 will not do anything).',
+        'A bucket with this name exists and is already owned by you',
 };
 
 export const BucketNotEmpty: ErrorFormat = {
@@ -365,6 +365,11 @@ export const NoSuchWebsiteConfiguration: ErrorFormat = {
     description: 'The specified bucket does not have a website configuration',
 };
 
+export const NoSuchTagSet: ErrorFormat = {
+    code: 404,
+    description: 'The TagSet does not exist',
+};
+
 export const NoSuchUpload: ErrorFormat = {
     code: 404,
     description:
@@ -685,6 +690,11 @@ export const ReportNotPresent: ErrorFormat = {
         'The request was rejected because the credential report does not exist. To generate a credential report, use GenerateCredentialReport.',
 };
 
+export const Found: ErrorFormat = {
+    code: 302,
+    description: 'Resource Found'
+};
+
 // ------------- Special non-AWS S3 errors -------------
 
 export const MPUinProgress: ErrorFormat = {
@@ -1032,3 +1042,15 @@ export const AuthMethodNotImplemented: ErrorFormat = {
     description: 'AuthMethodNotImplemented',
     code: 501,
 };
+
+// --------------------- quotaErros ---------------------
+
+export const NoSuchQuota: ErrorFormat = {
+    code: 404,
+    description: 'The specified resource does not have a quota.',
+};
+
+export const QuotaExceeded: ErrorFormat = {
+    code: 429,
+    description: 'The quota set for the resource is exceeded.',
+};

lib/metrics/ZenkoMetrics.ts

@@ -1,26 +1,19 @@
 import promClient from 'prom-client';
 
-const collectDefaultMetricsIntervalMs =
-    process.env.COLLECT_DEFAULT_METRICS_INTERVAL_MS !== undefined ?
-        Number.parseInt(process.env.COLLECT_DEFAULT_METRICS_INTERVAL_MS, 10) :
-        10000;
-
-promClient.collectDefaultMetrics({ timeout: collectDefaultMetricsIntervalMs });
-
 export default class ZenkoMetrics {
-    static createCounter(params: promClient.CounterConfiguration) {
+    static createCounter(params: promClient.CounterConfiguration<string>) {
         return new promClient.Counter(params);
     }
 
-    static createGauge(params: promClient.GaugeConfiguration) {
+    static createGauge(params: promClient.GaugeConfiguration<string>) {
         return new promClient.Gauge(params);
     }
 
-    static createHistogram(params: promClient.HistogramConfiguration) {
+    static createHistogram(params: promClient.HistogramConfiguration<string>) {
         return new promClient.Histogram(params);
     }
 
-    static createSummary(params: promClient.SummaryConfiguration) {
+    static createSummary(params: promClient.SummaryConfiguration<string>) {
         return new promClient.Summary(params);
     }
 
@@ -28,11 +21,15 @@ export default class ZenkoMetrics {
         return promClient.register.getSingleMetric(name);
     }
 
-    static asPrometheus() {
+    static async asPrometheus() {
         return promClient.register.metrics();
     }
 
     static asPrometheusContentType() {
         return promClient.register.contentType;
     }
+
+    static collectDefaultMetrics() {
+        return promClient.collectDefaultMetrics();
+    }
 }

lib/models/BackendInfo.ts

@@ -1,3 +1,5 @@
+import { RequestLogger } from 'werelogs';
+
 import { legacyLocations } from '../constants';
 import escapeForXml from '../s3middleware/escapeForXml';
 

lib/models/BucketInfo.ts

@@ -8,11 +8,12 @@ import ObjectLockConfiguration from './ObjectLockConfiguration';
 import BucketPolicy from './BucketPolicy';
 import NotificationConfiguration from './NotificationConfiguration';
 import { ACL as OACL } from './ObjectMD';
+import { areTagsValid, BucketTag } from '../s3middleware/tagging';
 
 // WHEN UPDATING THIS NUMBER, UPDATE BucketInfoModelVersion.md CHANGELOG
 // BucketInfoModelVersion.md can be found in documentation/ at the root
 // of this repository
-const modelVersion = 14;
+const modelVersion = 16;
 
 export type CORS = {
     id: string;
@@ -36,6 +37,41 @@ export type VersioningConfiguration = {
     MfaDelete: any;
 };
 
+export type VeeamSOSApi = {
+    SystemInfo?: {
+        ProtocolVersion: string,
+        ModelName: string,
+        ProtocolCapabilities: {
+            CapacityInfo: boolean,
+            UploadSessions: boolean,
+            IAMSTS?: boolean,
+        },
+        APIEndpoints?: {
+            IAMEndpoint: string,
+            STSEndpoint: string,
+        },
+        SystemRecommendations?: {
+            S3ConcurrentTaskLimit: number,
+            S3MultiObjectDelete: number,
+            StorageCurrentTasksLimit: number,
+            KbBlockSize: number,
+        }
+        LastModified?: string,
+    },
+    CapacityInfo?: {
+        Capacity: number,
+        Available: number,
+        Used: number,
+        LastModified?: string,
+    },
+};
+
+// Capabilities contains all specifics from external products supported by
+// our S3 implementation, at bucket level
+export type Capabilities = {
+    VeeamSOSApi?: VeeamSOSApi,
+};
+
 export type ACL = OACL & { WRITE: string[] }
 
 export default class BucketInfo {
@@ -59,11 +95,13 @@ export default class BucketInfo {
     _objectLockEnabled?: boolean;
     _objectLockConfiguration?: any;
     _notificationConfiguration?: any;
-    _tags?: { key: string; value: string }[] | null;
+    _tags?: Array<BucketTag>;
     _readLocationConstraint: string | null;
     _isNFS: boolean | null;
     _azureInfo: any | null;
     _ingestion: { status: 'enabled' | 'disabled' } | null;
+    _capabilities?: Capabilities;
+    _quotaMax: number | 0;
 
     /**
     * Represents all bucket information.
@@ -118,6 +156,9 @@ export default class BucketInfo {
     * @param [objectLockEnabled] - true when object lock enabled
     * @param [objectLockConfiguration] - object lock configuration
     * @param [notificationConfiguration] - bucket notification configuration
+    * @param [tags] - bucket tag set
+    * @param [capabilities] - capabilities for the bucket
+    * @param quotaMax - bucket quota
     */
     constructor(
         name: string,
@@ -144,6 +185,9 @@ export default class BucketInfo {
         objectLockEnabled?: boolean,
         objectLockConfiguration?: any,
         notificationConfiguration?: any,
+        tags?: Array<BucketTag> | [],
+        capabilities?: Capabilities,
+        quotaMax?: number | 0,
     ) {
         assert.strictEqual(typeof name, 'string');
         assert.strictEqual(typeof owner, 'string');
@@ -240,6 +284,15 @@ export default class BucketInfo {
             READ_ACP: [],
         };
 
+        if (tags === undefined) {
+            tags = [] as BucketTag[];
+        }
+        assert.strictEqual(areTagsValid(tags), true);
+        if (quotaMax) {
+            assert.strictEqual(typeof quotaMax, 'number');
+            assert(quotaMax >= 0, 'Quota cannot be negative');
+        }
+
         // IF UPDATING PROPERTIES, INCREMENT MODELVERSION NUMBER ABOVE
         this._acl = aclInstance;
         this._name = name;
@@ -265,6 +318,9 @@ export default class BucketInfo {
         this._objectLockEnabled = objectLockEnabled || false;
         this._objectLockConfiguration = objectLockConfiguration || null;
         this._notificationConfiguration = notificationConfiguration || null;
+        this._tags = tags;
+        this._capabilities = capabilities || undefined;
+        this._quotaMax = quotaMax || 0;
         return this;
     }
 
@@ -298,6 +354,9 @@ export default class BucketInfo {
             objectLockEnabled: this._objectLockEnabled,
             objectLockConfiguration: this._objectLockConfiguration,
             notificationConfiguration: this._notificationConfiguration,
+            tags: this._tags,
+            capabilities: this._capabilities,
+            quotaMax: this._quotaMax,
         };
         const final = this._websiteConfiguration
             ? {
@@ -323,7 +382,8 @@ export default class BucketInfo {
             obj.cors, obj.replicationConfiguration, obj.lifecycleConfiguration,
             obj.bucketPolicy, obj.uid, obj.readLocationConstraint, obj.isNFS,
             obj.ingestion, obj.azureInfo, obj.objectLockEnabled,
-            obj.objectLockConfiguration, obj.notificationConfiguration);
+            obj.objectLockConfiguration, obj.notificationConfiguration, obj.tags,
+            obj.capabilities, obj.quotaMax);
     }
 
     /**
@@ -350,7 +410,8 @@ export default class BucketInfo {
             data._bucketPolicy, data._uid, data._readLocationConstraint,
             data._isNFS, data._ingestion, data._azureInfo,
             data._objectLockEnabled, data._objectLockConfiguration,
-            data._notificationConfiguration);
+            data._notificationConfiguration, data._tags, data._capabilities,
+            data._quotaMax);
     }
 
     /**
@@ -841,4 +902,69 @@ export default class BucketInfo {
         this._objectLockEnabled = enabled;
         return this;
     }
+
+    /**
+     * Get the value of bucket tags
+     * @return - Array of bucket tags
+     */
+    getTags() {
+        return this._tags;
+    }
+    
+    /**
+     * Set bucket tags
+     * @return - bucket info instance
+     */
+    setTags(tags: Array<BucketTag>) {
+        this._tags = tags;
+        return this;
+    }
+
+    /**
+     * Get the value of bucket capabilities
+     * @return - capabilities of the bucket
+     */
+    getCapabilities() {
+        return this._capabilities;
+    }
+
+    /**
+     * Get a specific bucket capability
+     * 
+     * @param capability? - if provided, will return a specific capacity
+     * @return - capability of the bucket
+     */
+    getCapability(capability: string) : VeeamSOSApi | undefined {
+        if (capability && this._capabilities && this._capabilities[capability]) {
+            return this._capabilities[capability];
+        }
+        return undefined;
+    }
+    
+    /**
+     * Set bucket capabilities
+     * @return - bucket info instance
+     */
+    setCapabilities(capabilities: Capabilities) {
+        this._capabilities = capabilities;
+        return this;
+    }
+
+    /**
+     * Get the bucket quota information
+     * @return quotaMax
+     */
+    getQuota() {
+        return this._quotaMax;
+    }
+    
+    /**
+     * Set bucket quota
+     * @param quota - quota to be set
+     * @return - bucket quota info
+     */
+    setQuota(quota: number) {
+        this._quotaMax = quota || 0;
+        return this;
+    }
 }

lib/models/LifecycleConfiguration.ts

@@ -666,13 +666,38 @@ export default class LifecycleConfiguration {
      * @return Returns an error object or `null`
      */
     _checkDate(date: string) {
-        const isoRegex = new RegExp('^(-?(?:[1-9][0-9]*)?[0-9]{4})-' +
-            '(1[0-2]|0[1-9])-(3[01]|0[1-9]|[12][0-9])T(2[0-3]|[01][0-9])' +
-            ':([0-5][0-9]):([0-5][0-9])(.[0-9]+)?(Z)?$');
-        if (!isoRegex.test(date)) {
+        const isoRegex = new RegExp(
+            "^(-?(?:[1-9][0-9]*)?[0-9]{4})" + // Year
+            "-(1[0-2]|0[1-9])" + // Month
+            "-(3[01]|0[1-9]|[12][0-9])" + // Day
+            "T(2[0-3]|[01][0-9])" + // Hour
+            ":([0-5][0-9])" + // Minute
+            ":([0-5][0-9])" + // Second
+            "(\\.[0-9]+)?" + // Fractional second
+            "(Z|[+-][01][0-9]:[0-5][0-9])?$", // Timezone
+            "g"
+        );
+        const matches = [...date.matchAll(isoRegex)];
+        if (matches.length !== 1) {
             const msg = 'Date must be in ISO 8601 format';
             return errors.InvalidArgument.customizeDescription(msg);
         }
+        // Check for a timezone in the last match group. If none, add a Z to indicate UTC.
+        if (!matches[0][matches[0].length-1]) {
+            date += 'Z';
+        }
+        const dateObj = new Date(date);
+        if (Number.isNaN(dateObj.getTime())) {
+            const msg = 'Date is not a valid date';
+            return errors.InvalidArgument.customizeDescription(msg);
+        }
+        if (dateObj.getUTCHours() !== 0
+            || dateObj.getUTCMinutes() !== 0
+            || dateObj.getUTCSeconds() !== 0
+            || dateObj.getUTCMilliseconds() !== 0) {
+            const msg = '\'Date\' must be at midnight GMT';
+            return errors.InvalidArgument.customizeDescription(msg);
+        }
         return null;
     }
 
@@ -844,6 +869,7 @@ export default class LifecycleConfiguration {
      *              days: <value>,
      *              date: <value>,
      *              deleteMarker: <value>
+     *              newerNoncurrentVersions: <value>,
      *          },
      *      ],
      * }
@@ -856,7 +882,8 @@ export default class LifecycleConfiguration {
                 actionName: string;
                 days?: number;
                 date?: number;
-                deleteMarker?: boolean
+                deleteMarker?: boolean;
+                newerNoncurrentVersions?: number
             }[];
         } = {
             propName: 'actions',
@@ -885,8 +912,14 @@ export default class LifecycleConfiguration {
             if (action.error) {
                 actionsObj.error = action.error;
             } else {
-                const actionTimes = ['days', 'date', 'deleteMarker',
-                    'transition', 'nonCurrentVersionTransition'];
+                const actionTimes = [
+                    'days',
+                    'date',
+                    'deleteMarker',
+                    'transition',
+                    'nonCurrentVersionTransition',
+                    'newerNoncurrentVersions'
+                ];
                 actionTimes.forEach(t => {
                     if (action[t]) {
                         // eslint-disable-next-line no-param-reassign
@@ -1032,6 +1065,7 @@ export default class LifecycleConfiguration {
      * nvExpObj = {
      *      error: <error>,
      *      days: <value>,
+     *      newerNoncurrentVersions: <value>,
      * }
      */
     _parseNoncurrentVersionExpiration(rule: any) {
@@ -1042,14 +1076,41 @@ export default class LifecycleConfiguration {
                 'NoncurrentDays');
             return { error };
         }
+
+        const actionParams: {
+            error?: ArsenalError;
+            days: number;
+            newerNoncurrentVersions: number;
+        } = {
+            days: 0,
+            newerNoncurrentVersions: 0,
+        };
+
         const daysInt = parseInt(subNVExp.NoncurrentDays[0], 10);
         if (daysInt < 1) {
             const msg = 'NoncurrentDays is not a positive integer';
             const error = errors.InvalidArgument.customizeDescription(msg);
             return { error };
         } else {
-            return { days: daysInt };
+            actionParams.days = daysInt;
         }
+
+        if (subNVExp.NewerNoncurrentVersions) {
+            const newerVersionsInt = parseInt(subNVExp.NewerNoncurrentVersions[0], 10);
+
+            if (Number.isNaN(newerVersionsInt) || newerVersionsInt < 1) {
+                const msg = 'NewerNoncurrentVersions is not a positive integer';
+                const error = errors.InvalidArgument.customizeDescription(msg);
+                return { error };
+            }
+
+            actionParams.newerNoncurrentVersions = newerVersionsInt;
+
+        } else {
+            actionParams.newerNoncurrentVersions = 0;
+        }
+
+        return actionParams;
     }
 
     /**
@@ -1112,6 +1173,10 @@ export default class LifecycleConfiguration {
                         assert.strictEqual(typeof t.storageClass, 'string');
                     });
                 }
+
+                if (a.newerNoncurrentVersions) {
+                    assert.strictEqual(typeof a.newerNoncurrentVersions, 'number');
+                }
             });
         });
     }
@@ -1161,15 +1226,24 @@ export default class LifecycleConfiguration {
             }
 
             const Actions = actions.map(action => {
-                const { actionName, days, date, deleteMarker,
-                    nonCurrentVersionTransition, transition } = action;
+                const {
+                    actionName,
+                    days,
+                    date,
+                    deleteMarker,
+                    nonCurrentVersionTransition,
+                    transition,
+                    newerNoncurrentVersions,
+                } = action;
                 let Action: any;
                 if (actionName === 'AbortIncompleteMultipartUpload') {
                     Action = `<${actionName}><DaysAfterInitiation>${days}` +
                         `</DaysAfterInitiation></${actionName}>`;
                 } else if (actionName === 'NoncurrentVersionExpiration') {
-                    Action = `<${actionName}><NoncurrentDays>${days}` +
-                        `</NoncurrentDays></${actionName}>`;
+                    const Days = `<NoncurrentDays>${days}</NoncurrentDays>`;
+                    const NewerVersions = newerNoncurrentVersions ?
+                        `<NewerNoncurrentVersions>${newerNoncurrentVersions}</NewerNoncurrentVersions>` : '';
+                    Action = `<${actionName}>${Days}${NewerVersions}</${actionName}>`;
                 } else if (actionName === 'Expiration') {
                     const Days = days ? `<Days>${days}</Days>` : '';
                     const Date = date ? `<Date>${date}</Date>` : '';
@@ -1246,13 +1320,18 @@ export default class LifecycleConfiguration {
             }
 
             actions.forEach(action => {
-                const { actionName, days, date, deleteMarker } = action;
+                const { actionName, days, date, deleteMarker, newerNoncurrentVersions } = action;
                 if (actionName === 'AbortIncompleteMultipartUpload') {
                     entry.addAbortMPU(days!);
                     return;
                 }
                 if (actionName === 'NoncurrentVersionExpiration') {
-                    entry.addNCVExpiration(days!);
+                    entry.addNCVExpiration('NoncurrentDays', days!);
+
+                    if (newerNoncurrentVersions) {
+                        entry.addNCVExpiration('NewerNoncurrentVersions', newerNoncurrentVersions!);
+                    }
+
                     return;
                 }
                 if (actionName === 'Expiration') {
@@ -1289,6 +1368,7 @@ export type Rule = {
         days?: number;
         date?: number;
         deleteMarker?: boolean;
+        newerNoncurrentVersions?: number;
         nonCurrentVersionTransition?: {
             noncurrentDays: number;
             storageClass: string;

lib/models/LifecycleRule.ts

@@ -10,6 +10,10 @@ export type Expiration = {
     Date?: number | boolean;
     Days?: number | boolean;
 };
+export type NoncurrentExpiration = {
+    NoncurrentDays: number | null;
+    NewerNoncurrentVersions: number | null;
+};
 
 /**
  * @class LifecycleRule
@@ -21,9 +25,10 @@ export default class LifecycleRule {
     status: Status;
     tags: Tags;
     expiration?: Expiration;
-    ncvExpiration?: { NoncurrentDays: number };
+    ncvExpiration?: NoncurrentExpiration;
     abortMPU?: { DaysAfterInitiation: number };
     transitions?: any[];
+    ncvTransitions?: any[];
     prefix?: string;
 
     constructor(id: string, status: Status) {
@@ -38,9 +43,10 @@ export default class LifecycleRule {
             ID: string;
             Status: Status;
             Expiration?: Expiration;
-            NoncurrentVersionExpiration?: { NoncurrentDays: number };
+            NoncurrentVersionExpiration?: NoncurrentExpiration;
             AbortIncompleteMultipartUpload?: { DaysAfterInitiation: number };
             Transitions?: any[];
+            NoncurrentVersionTransitions?: any[];
             Filter?: Filter;
             Prefix?: '';
         } = { ID: this.id, Status: this.status };
@@ -49,7 +55,7 @@ export default class LifecycleRule {
             rule.Expiration = this.expiration;
         }
         if (this.ncvExpiration) {
-            rule.NoncurrentVersionExpiration = this.ncvExpiration;
+            rule.NoncurrentVersionExpiration = this.ncvExpiration
         }
         if (this.abortMPU) {
             rule.AbortIncompleteMultipartUpload = this.abortMPU;
@@ -57,6 +63,9 @@ export default class LifecycleRule {
         if (this.transitions) {
             rule.Transitions = this.transitions;
         }
+        if (this.ncvTransitions) {
+            rule.NoncurrentVersionTransitions = this.ncvTransitions;
+        }
 
         const filter = this.buildFilter();
 
@@ -136,15 +145,24 @@ export default class LifecycleRule {
 
     /**
      * NoncurrentVersionExpiration
-     * @param days - NoncurrentDays
+     * @param prop - Property must be defined in `validProps`
+     * @param value - integer for `NoncurrentDays` and `NewerNoncurrentVersions` 
      */
-    addNCVExpiration(days: number) {
-        this.ncvExpiration = { NoncurrentDays: days };
+    addNCVExpiration(prop: 'NoncurrentDays' | 'NewerNoncurrentVersions', value: number): this;
+    addNCVExpiration(prop: string, value: number) {
+        const validProps = ['NoncurrentDays', 'NewerNoncurrentVersions'];
+        if (validProps.includes(prop)) {
+            this.ncvExpiration = this.ncvExpiration || {
+                NoncurrentDays: null,
+                NewerNoncurrentVersions: null,
+            };
+            this.ncvExpiration[prop] = value;
+        }
         return this;
     }
 
     /**
-     * AbortIncompleteMultipartUpload
+     * abortincompletemultipartupload
      * @param days - DaysAfterInitiation
      */
     addAbortMPU(days: number) {
@@ -160,4 +178,13 @@ export default class LifecycleRule {
         this.transitions = transitions;
         return this;
     }
+
+    /**
+     * NonCurrentVersionTransitions
+     * @param nvcTransitions - NonCurrentVersionTransitions 
+     */
+    addNCVTransitions(nvcTransitions) {
+        this.ncvTransitions = nvcTransitions;
+        return this;
+    }
 }

lib/models/ObjectMD.ts

@@ -1,6 +1,7 @@
 import * as crypto from 'crypto';
 import * as constants from '../constants';
 import * as VersionIDUtils from '../versioning/VersionID';
+import { VersioningConstants } from '../versioning/constants';
 import ObjectMDLocation, {
     ObjectMDLocationData,
     Location,
@@ -56,13 +57,16 @@ export type ObjectMDData = {
     'x-amz-server-side-encryption-aws-kms-key-id': string;
     'x-amz-server-side-encryption-customer-algorithm': string;
     'x-amz-website-redirect-location': string;
+    'x-amz-scal-transition-in-progress'?: boolean;
+    'x-amz-scal-transition-time'?: string;
     azureInfo?: any;
     acl: ACL;
     key: string;
     location: null | Location[];
-    // versionId, isNull, nullVersionId and isDeleteMarker
+    // versionId, isNull, isNull2, nullVersionId and isDeleteMarker
     // should be undefined when not set explicitly
     isNull?: boolean;
+    isNull2?: boolean;
     nullVersionId?: string;
     nullUploadId?: string;
     isDeleteMarker?: boolean;
@@ -76,6 +80,16 @@ export type ObjectMDData = {
     dataStoreName: string;
     originOp: string;
     microVersionId?: string;
+    // Deletion flag
+    // Used for keeping object metadata in the oplog event
+    // In case of a deletion the flag is first updated before
+    // deleting the object
+    deleted: boolean;
+    // PHD flag indicates whether the object is a temporary placeholder.
+    // This is the case when the latest version of an object gets deleted
+    // the master is set as a placeholder and gets updated with the new latest
+    // version data after a certain amount of time.
+    isPHD: boolean;
 };
 
 /**
@@ -183,6 +197,7 @@ export default class ObjectMD {
             'x-amz-server-side-encryption-aws-kms-key-id': '',
             'x-amz-server-side-encryption-customer-algorithm': '',
             'x-amz-website-redirect-location': '',
+            'x-amz-scal-transition-in-progress': false,
             acl: {
                 Canned: 'private',
                 FULL_CONTROL: [],
@@ -196,6 +211,7 @@ export default class ObjectMD {
             // versionId, isNull, nullVersionId and isDeleteMarker
             // should be undefined when not set explicitly
             isNull: undefined,
+            isNull2: undefined,
             nullVersionId: undefined,
             nullUploadId: undefined,
             isDeleteMarker: undefined,
@@ -215,6 +231,8 @@ export default class ObjectMD {
             },
             dataStoreName: '',
             originOp: '',
+            deleted: false,
+            isPHD: false,
         };
     }
 
@@ -628,6 +646,48 @@ export default class ObjectMD {
         return this._data['x-amz-website-redirect-location'];
     }
 
+    /**
+     * Set metadata transition in progress value
+     *
+     * @param inProgress - True if transition is in progress, false otherwise
+     * @param transitionTime - Date when the transition started
+     * @return itself
+     */
+    setTransitionInProgress(inProgress: false): this
+    setTransitionInProgress(inProgress: true, transitionTime: Date|string|number): this
+    setTransitionInProgress(inProgress: boolean, transitionTime?: Date|string|number) {
+        this._data['x-amz-scal-transition-in-progress'] = inProgress;
+        if (!inProgress || !transitionTime) {
+            delete this._data['x-amz-scal-transition-time'];
+        } else {
+            if (typeof transitionTime === 'number') {
+                transitionTime = new Date(transitionTime);
+            }
+            if (transitionTime instanceof Date) {
+                transitionTime = transitionTime.toISOString();
+            }
+            this._data['x-amz-scal-transition-time'] = transitionTime;
+        }
+        return this;
+    }
+
+    /**
+     * Get metadata transition in progress value
+     *
+     * @return True if transition is in progress, false otherwise
+     */
+    getTransitionInProgress() {
+        return this._data['x-amz-scal-transition-in-progress'];
+    }
+
+    /**
+     * Gets the transition time of the object.
+     * @returns The transition time of the object.
+     */
+    getTransitionTime() {
+        return this._data['x-amz-scal-transition-time'];
+    }
+
     /**
      * Set access control list
      *
@@ -776,6 +836,31 @@ export default class ObjectMD {
         return this._data.isNull || false;
     }
 
+    /**
+     * Set metadata isNull2 value
+     *
+     * @param isNull2 - Whether new version is null or not AND has
+     * been put with a Cloudserver handling null keys (i.e. supporting
+     * S3C-7352)
+
+     * @return itself
+     */
+    setIsNull2(isNull2: boolean) {
+        this._data.isNull2 = isNull2;
+        return this;
+    }
+
+    /**
+     * Get metadata isNull2 value
+     *
+     * @return isNull2 - Whether new version is null or not AND has
+     * been put with a Cloudserver handling null keys (i.e. supporting
+     * S3C-7352)
+     */
+    getIsNull2() {
+        return this._data.isNull2 || false;
+    }
+
     /**
      * Set metadata nullVersionId value
      *
@@ -867,6 +952,9 @@ export default class ObjectMD {
      * @return The object versionId
      */
     getVersionId() {
+        if (this.getIsNull()) {
+            return VersioningConstants.ExternalNullVersionId;
+        }
         return this._data.versionId;
     }
 
@@ -874,13 +962,16 @@ export default class ObjectMD {
      * Get metadata versionId value in encoded form (the one visible
      * to the S3 API user)
      *
-     * @return The encoded object versionId
+     * @return {undefined|string} The encoded object versionId
      */
     getEncodedVersionId() {
         const versionId = this.getVersionId();
-        if (versionId) {
+        if (versionId === VersioningConstants.ExternalNullVersionId) {
+            return versionId;
+        } else if (versionId) {
             return VersionIDUtils.encode(versionId);
         }
+        return undefined;
     }
 
     /**
@@ -1354,4 +1445,40 @@ export default class ObjectMD {
         }
         return this;
     }
+
+    /**
+    * Set deleted flag
+    * @param {Boolean} value deleted object
+    * @return {ObjectMD}
+    */
+     setDeleted(value) {
+        this._data.deleted = value;
+        return this;
+    }
+    
+    /**
+    * Get deleted flag
+    * @return {Boolean}
+    */
+    getDeleted() {
+        return this._data.deleted;
+    }
+
+    /**
+    * Set isPHD flag
+    * @param {Boolean} value isPHD value
+    * @return {ObjectMD}
+    */
+    setIsPHD(value) {
+        this._data.isPHD = value;
+        return this;
+    }
+
+    /**
+    * Get isPHD flag
+    * @return {Boolean}
+    */
+    getIsPHD() {
+        return this._data.isPHD;
+    }
 }

lib/models/ReplicationConfiguration.ts

@@ -1,6 +1,8 @@
 import assert from 'assert';
 import UUID from 'uuid';
 
+import { RequestLogger } from 'werelogs';
+
 import escapeForXml from '../s3middleware/escapeForXml';
 import errors from '../errors';
 import { isValidBucketName } from '../s3routes/routesUtils';
@@ -359,6 +361,11 @@ export default class ReplicationConfiguration {
                 (endpoint: any) => endpoint.site === storageClass
             );
             if (endpoint) {
+                // We do not support replication to cold location. 
+                // Only transition to cold location is supported.
+                if (endpoint.site && this._config.locationConstraints[endpoint.site]?.isCold) {
+                    return false;
+                }
                 // If this._hasScalityDestination was not set to true in any
                 // previous iteration or by a prior rule's storage class, then
                 // check if the current endpoint is a Scality destination.

lib/network/http/server.ts

@@ -1,5 +1,6 @@
 import * as http from 'http';
 import * as https from 'https';
+import { https as HttpsAgent } from 'httpagent';
 import * as tls from 'tls';
 import * as net from 'net';
 import assert from 'assert';
@@ -409,7 +410,11 @@ export default class Server {
                     method: 'arsenal.network.Server.start',
                     port: this._port,
                 });
-                this._https.agent = new https.Agent(this._https);
+                this._https.agent = new HttpsAgent.Agent(this._https, {
+                    // Do not enforce the maximum number of sockets for the
+                    // main server, as it might be able to serve more clients.
+                    maxSockets: false,
+                });
                 this._server = https.createServer(this._https,
                     (req, res) => this._onRequest(req, res));
             } else {
@@ -430,7 +435,6 @@ export default class Server {
             this._server.on('connection', sock => {
                 // Setting no delay of the socket to the value configured
                 // TODO fix this
-                // @ts-expect-errors
                 sock.setNoDelay(this.isNoDelay());
                 sock.on('error', err => this._logger.info(
                     'socket error - request rejected', { error: err }));

lib/network/http/utils.ts

@@ -77,10 +77,11 @@ export function getByteRangeFromSpec(
             objectSize - 1] };
     }
     if (rangeSpec.start < objectSize) {
-        // test is false if end is undefined
-        return { range: [rangeSpec.start,
-            ((rangeSpec.end && (rangeSpec.end < objectSize)) ?
-            rangeSpec.end : objectSize - 1)] };
+        // test is false if end is undefined or end is greater than objectSize
+        const end: number = rangeSpec.end !== undefined && rangeSpec.end < objectSize
+            ? rangeSpec.end
+            : objectSize - 1;
+        return { range: [rangeSpec.start, end] };
     }
     return { error: errors.InvalidRange };
 }

lib/network/kmip/codec/ttlv.ts

@@ -20,7 +20,7 @@ function _ttlvPadVector(vec: any[]) {
     return vec;
 }
 
-function _throwError(logger: werelogs.Logger, msg: string, data?: LogDictionnary) {
+function _throwError(logger: werelogs.Logger, msg: string, data?: LogDictionary) {
     logger.error(msg, data);
     throw Error(msg);
 }

lib/network/probe/HealthProbeServer.ts

@@ -62,7 +62,7 @@ export default class HealthProbeServer extends httpServer {
     _onLiveness(
         _req: http.IncomingMessage,
         res: http.ServerResponse,
-        log: RequestLogger,
+        log: werelogs.RequestLogger,
     ) {
         if (this._livenessCheck(log)) {
             sendSuccess(res, log);
@@ -74,7 +74,7 @@ export default class HealthProbeServer extends httpServer {
     _onReadiness(
         _req: http.IncomingMessage,
         res: http.ServerResponse,
-        log: RequestLogger,
+        log: werelogs.RequestLogger,
     ) {
         if (this._readinessCheck(log)) {
             sendSuccess(res, log);
@@ -84,10 +84,11 @@ export default class HealthProbeServer extends httpServer {
     }
 
     // expose metrics to Prometheus
-    _onMetrics(_req: http.IncomingMessage, res: http.ServerResponse) {
+    async _onMetrics(_req: http.IncomingMessage, res: http.ServerResponse) {
+        const metrics = await ZenkoMetrics.asPrometheus();
         res.writeHead(200, {
             'Content-Type': ZenkoMetrics.asPrometheusContentType(),
         });
-        res.end(ZenkoMetrics.asPrometheus());
+        res.end(metrics);
     }
 }

lib/network/probe/ProbeServer.ts

@@ -16,7 +16,7 @@ export const DEFAULT_METRICS_ROUTE = '/metrics';
  * @param log - Werelogs instance for logging if you choose to
  */
 
-export type ProbeDelegate = (res: http.ServerResponse, log: RequestLogger) => string | void
+export type ProbeDelegate = (res: http.ServerResponse, log: werelogs.RequestLogger) => string | void
 
 export type ProbeServerParams = {
     port: number;

lib/network/probe/Utils.ts

@@ -1,4 +1,7 @@
 import * as http from 'http';
+
+import { RequestLogger } from 'werelogs';
+
 import { ArsenalError } from '../../errors';
 
 /**

lib/network/rest/RESTClient.ts

@@ -4,7 +4,7 @@ import * as werelogs from 'werelogs';
 import * as constants from '../../constants';
 import * as utils from './utils';
 import errors, { ArsenalError } from '../../errors';
-import HttpAgent from 'agentkeepalive';
+import { http as HttpAgent } from 'httpagent';
 import * as stream from 'stream';
 
 function setRequestUids(reqHeaders: http.IncomingHttpHeaders, reqUids: string) {
@@ -71,7 +71,7 @@ function makeErrorFromHTTPResponse(response: http.IncomingMessage) {
 export default class RESTClient {
     host: string;
     port: number;
-    httpAgent: HttpAgent;
+    httpAgent: http.Agent;
     logging: werelogs.Logger;
     isPassthrough: boolean;
 
@@ -98,10 +98,10 @@ export default class RESTClient {
         this.port = params.port;
         this.isPassthrough = params.isPassthrough || false;
         this.logging = new (params.logApi || werelogs).Logger('DataFileRESTClient');
-        this.httpAgent = new HttpAgent({
+        this.httpAgent = new HttpAgent.Agent({
             keepAlive: true,
             freeSocketTimeout: constants.httpClientFreeSocketTimeout,
-        });
+        }) as http.Agent;
     }
 
     /** Destroy the HTTP agent, forcing a close of the remaining open connections */
@@ -119,7 +119,7 @@ export default class RESTClient {
         method: string,
         headers: http.OutgoingHttpHeaders | null,
         key: string | null,
-        log: RequestLogger,
+        log: werelogs.RequestLogger,
         responseCb: (res: http.IncomingMessage) => void,
     ) {
         const reqHeaders = headers || {};

lib/network/rest/RESTServer.ts

@@ -25,7 +25,7 @@ function setContentRange(
 
 function sendError(
     res: http.ServerResponse,
-    log: RequestLogger,
+    log: werelogs.RequestLogger,
     error: ArsenalError,
     optMessage?: string,
 ) {
@@ -68,7 +68,6 @@ export default class RESTServer extends httpServer {
     }) {
         assert(params.port);
 
-        // @ts-expect-error
         werelogs.configure({
             level: params.log.logLevel,
             dump: params.log.dumpLevel,
@@ -142,7 +141,7 @@ export default class RESTServer extends httpServer {
     _onPut(
         req: http.IncomingMessage,
         res: http.ServerResponse,
-        log: RequestLogger,
+        log: werelogs.RequestLogger,
     ) {
         let size: number;
         try {
@@ -184,7 +183,7 @@ export default class RESTServer extends httpServer {
     _onGet(
         req: http.IncomingMessage,
         res: http.ServerResponse,
-        log: RequestLogger,
+        log: werelogs.RequestLogger,
     ) {
         let pathInfo: ReturnType<typeof parseURL>;
         let rangeSpec: ReturnType<typeof httpUtils.parseRangeSpec> | undefined =
@@ -267,7 +266,7 @@ export default class RESTServer extends httpServer {
     _onDelete(
         req: http.IncomingMessage,
         res: http.ServerResponse,
-        log: RequestLogger,
+        log: werelogs.RequestLogger,
     ) {
         let pathInfo: ReturnType<typeof parseURL>;
         try {

lib/network/rpc/rpc.ts

@@ -1,6 +1,6 @@
 import ioClient from 'socket.io-client';
 import * as http from 'http';
-import io from 'socket.io';
+import { Server as IOServer } from 'socket.io';
 import * as sioStream from './sio-stream';
 import async from 'async';
 import assert from 'assert';
@@ -497,7 +497,7 @@ export function RPCServer(params: {
     assert(params.logger);
 
     const httpServer = http.createServer();
-    const server = io(httpServer);
+    const server = new IOServer(httpServer, { maxHttpBufferSize: 1e8 });
     const log = params.logger;
 
     /**
@@ -508,7 +508,7 @@ export function RPCServer(params: {
      *
      * @param {BaseService} serviceList - list of services to register
      */
-    server.registerServices = function registerServices(...serviceList: any[]) {
+    (server as any).registerServices = function registerServices(...serviceList: any[]) {
         serviceList.forEach(service => {
             const sock = this.of(service.namespace);
             sock.on('connection', conn => {
@@ -536,7 +536,7 @@ export function RPCServer(params: {
         });
     };
 
-    server.listen = function listen(port, bindAddress = undefined) {
+    (server as any).listen = function listen(port, bindAddress = undefined) {
         httpServer.listen(port, bindAddress);
     };
 

lib/policy/resourcePolicySchema.json

@@ -38,7 +38,7 @@
         },
         "principalAWSUserArn": {
             "type": "string",
-            "pattern": "^arn:aws:iam::[0-9]{12}:user/(?!\\*)[\\w+=,.@ -/]{1,64}$"
+            "pattern": "^arn:aws:iam::[0-9]{12}:user/(?!\\*)[\\w+=,.@ -/]{1,2017}$"
         },
         "principalAWSRoleArn": {
             "type": "string",
@@ -360,6 +360,9 @@
             "type": "string",
             "const": "2012-10-17"
         },
+        "Id": {
+            "type": "string"
+        },
         "Statement": {
             "oneOf": [
                 {

lib/policy/userPolicySchema.json

@@ -28,7 +28,7 @@
         },
         "principalAWSUserArn": {
             "type": "string",
-            "pattern": "^arn:aws:iam::[0-9]{12}:user/(?!\\*)[\\w+=,.@ -/]{1,64}$"
+            "pattern": "^arn:aws:iam::[0-9]{12}:user/(?!\\*)[\\w+=,.@ -/]{1,2017}$"
         },
         "principalAWSRoleArn": {
             "type": "string",

lib/policyEvaluator/RequestContext.ts

@@ -12,13 +12,39 @@ import {
     actionMapSSO,
     actionMapSTS,
     actionMapMetadata,
+    actionMapScuba,
 } from './utils/actionMaps';
 
-const _actionNeedQuotaCheck = {
+export const actionNeedQuotaCheck = {
     objectPut: true,
+    objectPutVersion: true,
     objectPutPart: true,
+    objectRestore: true,
 };
 
+/**
+ * This variable describes APIs that change the bytes
+ * stored, requiring quota updates
+ */
+export const actionWithDataDeletion = {
+    objectDelete: true,
+    objectDeleteVersion: true,
+    multipartDelete: true,
+    multiObjectDelete: true,
+};
+
+/**
+ * The function returns true if the current API call is a copy object
+ * and the action requires a quota evaluation logic, post retrieval
+ * of the object metadata.
+ * @param {string} action - the action being performed
+ * @param {string} currentApi - the current API being called
+ * @return {boolean} - whether the action requires a quota check
+ */
+export function actionNeedQuotaCheckCopy(action: string, currentApi: string) {
+    return action === 'objectGet' && (currentApi === 'objectCopy' || currentApi === 'objectPutCopyPart');
+}
+
 function _findAction(service: string, method: string) {
     switch (service) {
         case 's3':
@@ -36,6 +62,8 @@ function _findAction(service: string, method: string) {
             return actionMapSTS[method];
         case 'metadata':
             return actionMapMetadata[method];
+        case 'scuba':
+            return actionMapScuba[method];
         default:
             return undefined;
     }
@@ -105,6 +133,10 @@ function _buildArn(
             return `arn:scality:metadata::${requesterInfo!.accountid}:` +
             `${generalResource}/`;
         }
+        case 'scuba': {
+            return `arn:scality:scuba::${requesterInfo!.accountid}:` +
+            `${generalResource}${specificResource ? '/' + specificResource : ''}`;
+        }
         default:
             return undefined;
     }
@@ -168,12 +200,12 @@ export default class RequestContext {
     _policyArn: string;
     _action?: string;
     _needQuota: boolean;
-    _postXml?: string;
     _requestObjTags: string | null;
     _existingObjTag: string | null;
     _needTagEval: boolean;
     _foundAction?: string;
     _foundResource?: string;
+    _objectLockRetentionDays?: number | null;
 
     constructor(
         headers: { [key: string]: string | string[] },
@@ -192,7 +224,10 @@ export default class RequestContext {
         securityToken: string,
         policyArn: string,
         action?: string,
-        postXml?: string,
+        requestObjTags?: string,
+        existingObjTag?: string,
+        needTagEval?: false,
+        objectLockRetentionDays?: number,
     ) {
         this._headers = headers;
         this._query = query;
@@ -221,11 +256,12 @@ export default class RequestContext {
         this._securityToken = securityToken;
         this._policyArn = policyArn;
         this._action = action;
-        this._needQuota = _actionNeedQuotaCheck[apiMethod] === true;
-        this._postXml = postXml;
-        this._requestObjTags = null;
-        this._existingObjTag = null;
-        this._needTagEval = false;
+        this._needQuota = actionNeedQuotaCheck[apiMethod] === true
+            || actionWithDataDeletion[apiMethod] === true;
+        this._requestObjTags = requestObjTags || null;
+        this._existingObjTag = existingObjTag || null;
+        this._needTagEval = needTagEval || false;
+        this._objectLockRetentionDays = objectLockRetentionDays || null;
         return this;
     }
 
@@ -238,7 +274,7 @@ export default class RequestContext {
             apiMethod: this._apiMethod,
             headers: this._headers,
             query: this._query,
-            requersterInfo: this._requesterInfo,
+            requesterInfo: this._requesterInfo,
             requesterIp: this._requesterIp,
             sslEnabled: this._sslEnabled,
             awsService: this._awsService,
@@ -254,10 +290,10 @@ export default class RequestContext {
             securityToken: this._securityToken,
             policyArn: this._policyArn,
             action: this._action,
-            postXml: this._postXml,
             requestObjTags: this._requestObjTags,
             existingObjTag: this._existingObjTag,
             needTagEval: this._needTagEval,
+            objectLockRetentionDays: this._objectLockRetentionDays,
         };
         return JSON.stringify(requestInfo);
     }
@@ -278,12 +314,28 @@ export default class RequestContext {
         if (resource) {
             obj.specificResource = resource;
         }
-        return new RequestContext(obj.headers, obj.query, obj.generalResource,
-            obj.specificResource, obj.requesterIp, obj.sslEnabled,
-            obj.apiMethod, obj.awsService, obj.locationConstraint,
-            obj.requesterInfo, obj.signatureVersion,
-            obj.authType, obj.signatureAge, obj.securityToken, obj.policyArn,
-            obj.action, obj.postXml);
+        return new RequestContext(
+            obj.headers,
+            obj.query,
+            obj.generalResource,
+            obj.specificResource,
+            obj.requesterIp,
+            obj.sslEnabled,
+            obj.apiMethod,
+            obj.awsService,
+            obj.locationConstraint,
+            obj.requesterInfo,
+            obj.signatureVersion,
+            obj.authType,
+            obj.signatureAge,
+            obj.securityToken,
+            obj.policyArn,
+            obj.action,
+            obj.requestObjTags,
+            obj.existingObjTag,
+            obj.needTagEval,
+            obj.objectLockRetentionDays,
+        );
     }
 
     /**
@@ -627,26 +679,6 @@ export default class RequestContext {
         return this._needQuota;
     }
 
-    /**
-     * Set request post
-     *
-     * @param postXml - request post
-     * @return itself
-     */
-    setPostXml(postXml: string) {
-        this._postXml = postXml;
-        return this;
-    }
-
-    /**
-     * Get request post
-     *
-     * @return request post
-     */
-    getPostXml() {
-        return this._postXml;
-    }
-
     /**
      * Set request object tags
      *
@@ -706,4 +738,24 @@ export default class RequestContext {
     getNeedTagEval() {
         return this._needTagEval;
     }
+
+    /**
+     * Get object lock retention days
+     *
+     * @returns objectLockRetentionDays - object lock retention days 
+     */
+    getObjectLockRetentionDays() {
+        return this._objectLockRetentionDays;
+    }
+
+    /**
+     * Set object lock retention days
+     *
+     * @param objectLockRetentionDays - object lock retention days
+     * @returns itself
+     */
+    setObjectLockRetentionDays(objectLockRetentionDays: number) {
+        this._objectLockRetentionDays = objectLockRetentionDays;
+        return this;
+    }
 }

lib/policyEvaluator/evaluator.ts

@@ -13,7 +13,11 @@ const operatorsWithVariables = ['StringEquals', 'StringNotEquals',
 const operatorsWithNegation = ['StringNotEquals',
     'StringNotEqualsIgnoreCase', 'StringNotLike', 'ArnNotEquals',
     'ArnNotLike', 'NumericNotEquals'];
-const tagConditions = new Set(['s3:ExistingObjectTag', 's3:RequestObjectTagKey', 's3:RequestObjectTagKeys']);
+const tagConditions = new Set([
+    's3:ExistingObjectTag',
+    's3:RequestObjectTagKey',
+    's3:RequestObjectTagKeys',
+]);
 
 
 /**
@@ -24,11 +28,11 @@ const tagConditions = new Set(['s3:ExistingObjectTag', 's3:RequestObjectTagKey',
  * @param log - logger
  * @return true if applicable, false if not
  */
-export const isResourceApplicable = (
+export function isResourceApplicable(
     requestContext: RequestContext,
     statementResource: string | string[],
     log: Logger,
-): boolean => {
+): boolean {
     const resource = requestContext.getResource();
     if (!Array.isArray(statementResource)) {
         // eslint-disable-next-line no-param-reassign
@@ -59,7 +63,7 @@ export const isResourceApplicable = (
         { requestResource: resource });
     // If no match found, no resource is applicable
     return false;
-};
+}
 
 /**
  * Check whether action in policy statement applies to request
@@ -69,11 +73,11 @@ export const isResourceApplicable = (
  * @param log - logger
  * @return true if applicable, false if not
  */
-export const isActionApplicable = (
+export function isActionApplicable(
     requestAction: string,
     statementAction: string | string[],
     log: Logger,
-): boolean => {
+): boolean {
     if (!Array.isArray(statementAction)) {
         // eslint-disable-next-line no-param-reassign
         statementAction = [statementAction];
@@ -95,32 +99,33 @@ export const isActionApplicable = (
         { requestAction });
     // If no match found, return false
     return false;
-};
+}
 
 /**
  * Check whether request meets policy conditions
- * @param requestContext - info about request
- * @param statementCondition - Condition statement from policy
- * @param log - logger
- * @return contains whether conditions are allowed and whether they
- * contain any tag condition keys
+ * @param {RequestContext} requestContext - info about request
+ * @param {object} statementCondition - Condition statement from policy
+ * @param {Logger} log - logger
+ * @return {boolean|null} a condition evaluation result, one of:
+ * - true: condition is met
+ * - false: condition is not met
+ * - null: condition evaluation requires additional info to be
+ *   provided (namely, for tag conditions, request tags and/or object
+ *   tags have to be provided to evaluate the condition)
  */
-export const meetConditions = (
+export function meetConditions(
     requestContext: RequestContext,
     statementCondition: any,
     log: Logger,
-) => {
+): boolean | null {
+    let hasTagConditions = false;
     // The Condition portion of a policy is an object with different
     // operators as keys
-    const conditionEval = {};
-    const operators = Object.keys(statementCondition);
-    const length = operators.length;
-    for (let i = 0; i < length; i++) {
-        const operator = operators[i];
+    for (const operator of Object.keys(statementCondition)) {
         const hasPrefix = operator.includes(':');
         const hasIfExistsCondition = operator.endsWith('IfExists');
         // If has "IfExists" added to operator name, or operator has "ForAnyValue" or
-        // "For All Values" prefix, find operator name without "IfExists" or prefix
+        // "ForAllValues" prefix, find operator name without "IfExists" or prefix
         let bareOperator = hasIfExistsCondition ? operator.slice(0, -8) :
             operator;
         let prefix: string | undefined;
@@ -135,10 +140,6 @@ export const meetConditions = (
         // Note: this should be the actual operator name, not the bareOperator
         const conditionsWithSameOperator = statementCondition[operator];
         const conditionKeys = Object.keys(conditionsWithSameOperator);
-        if (conditionKeys.some(key => tagConditions.has(key)) && !requestContext.getNeedTagEval()) {
-            // @ts-expect-error
-            conditionEval.tagConditions = true;
-        }
         const conditionKeysLength = conditionKeys.length;
         for (let j = 0; j < conditionKeysLength; j++) {
             const key = conditionKeys[j];
@@ -155,6 +156,10 @@ export const meetConditions = (
             // tag key is included in condition key and needs to be
             // moved to value for evaluation, otherwise key/value are unchanged
             const [transformedKey, transformedValue] = transformTagKeyValue(key, value);
+            if (tagConditions.has(transformedKey) && !requestContext.getNeedTagEval()) {
+                hasTagConditions = true;
+                continue;
+            }
             // Pull key using requestContext
             // TODO: If applicable to S3, handle policy set operations
             // where a keyBasedOnRequestContext returns multiple values and
@@ -180,11 +185,10 @@ export const meetConditions = (
                 log.trace('condition not satisfied due to ' +
                 'missing info', { operator,
                     conditionKey: transformedKey, policyValue: transformedValue });
-                return { allow: false };
+                return false;
             }
             // If condition operator prefix is included, the key should be an array
             if (prefix && !Array.isArray(keyBasedOnRequestContext)) {
-                // @ts-expect-error
                 keyBasedOnRequestContext = [keyBasedOnRequestContext];
             }
             // Transalate operator into function using bareOperator
@@ -196,14 +200,16 @@ export const meetConditions = (
             if (!operatorFunction(keyBasedOnRequestContext, transformedValue, prefix)) {
                 log.trace('did not satisfy condition', { operator: bareOperator,
                     keyBasedOnRequestContext, policyValue: transformedValue });
-                return { allow: false };
+                return false;
             }
         }
     }
-    // @ts-expect-error
-    conditionEval.allow = true;
-    return conditionEval;
-};
+    // one or more conditions required tag info to be evaluated
+    if (hasTagConditions) {
+        return null;
+    }
+    return true;
+}
 
 /**
  * Evaluate whether a request is permitted under a policy.
@@ -216,13 +222,15 @@ export const meetConditions = (
  * @return Allow if permitted, Deny if not permitted or Neutral
  * if not applicable
  */
-export const evaluatePolicy = (
+export function evaluatePolicy(
     requestContext: RequestContext,
     policy: any,
     log: Logger,
-): string => {
+): string {
     // TODO: For bucket policies need to add Principal evaluation
-    let verdict = 'Neutral';
+    let allow = false;
+    let allowWithTagCondition = false;
+    let denyWithTagCondition = false;
 
     if (!Array.isArray(policy.Statement)) {
         // eslint-disable-next-line no-param-reassign
@@ -259,10 +267,18 @@ export const evaluatePolicy = (
         }
         const conditionEval = currentStatement.Condition ?
             meetConditions(requestContext, currentStatement.Condition, log) :
-            null;
+            true;
         // If do not meet conditions move on to next statement
-        // @ts-expect-error
-        if (conditionEval && !conditionEval.allow) {
+        if (conditionEval === false) {
+            continue;
+        }
+        // If condition needs tag info to be evaluated, mark and move on to next statement
+        if (conditionEval === null) {
+            if (currentStatement.Effect === 'Deny') {
+                denyWithTagCondition = true;
+            } else {
+                allowWithTagCondition = true;
+            }
             continue;
         }
         if (currentStatement.Effect === 'Deny') {
@@ -271,19 +287,30 @@ export const evaluatePolicy = (
             return 'Deny';
         }
         log.trace('Allow statement applies');
-        // If statement is applicable, conditions are met and Effect is
-        // to Allow, set verdict to Allow
-        verdict = 'Allow';
-        // @ts-expect-error
-        if (conditionEval && conditionEval.tagConditions) {
-            verdict = 'NeedTagConditionEval';
+        // statement is applicable, conditions are met and Effect is
+        // to Allow
+        allow = true;
     }
+    let verdict;
+    if (denyWithTagCondition) {
+        // priority is on checking tags to potentially deny
+        verdict = 'DenyWithTagCondition';
+    } else if (allow) {
+        // at least one statement is an allow
+        verdict = 'Allow';
+    } else if (allowWithTagCondition) {
+        // all allow statements need tag checks
+        verdict = 'AllowWithTagCondition';
+    } else {
+        // no statement matched to allow or deny
+        verdict = 'Neutral';
     }
     log.trace('result of evaluating single policy', { verdict });
     return verdict;
-};
+}
 
 /**
+ * @deprecated Upgrade to standardEvaluateAllPolicies
  * Evaluate whether a request is permitted under a policy.
  * @param requestContext - Info necessary to
  * evaluate permission
@@ -294,24 +321,58 @@ export const evaluatePolicy = (
  * @return Allow if permitted, Deny if not permitted.
  * Default is to Deny. Deny overrides an Allow
  */
-export const evaluateAllPolicies = (
+export function evaluateAllPolicies(
     requestContext: RequestContext,
     allPolicies: any[],
     log: Logger,
-): string => {
+): string {
+    return standardEvaluateAllPolicies(requestContext, allPolicies, log).verdict;
+}
+export function standardEvaluateAllPolicies(
+    requestContext: RequestContext,
+    allPolicies: any[],
+    log: Logger,
+): {
+    verdict: string;
+    isImplicit: boolean;
+} {
     log.trace('evaluating all policies');
-    let verdict = 'Deny';
+    let allow = false;
+    let allowWithTagCondition = false;
+    let denyWithTagCondition = false;
     for (let i = 0; i < allPolicies.length; i++) {
-        const singlePolicyVerdict =
-            evaluatePolicy(requestContext, allPolicies[i], log);
+        const singlePolicyVerdict = evaluatePolicy(requestContext, allPolicies[i], log);
         // If there is any Deny, just return Deny
         if (singlePolicyVerdict === 'Deny') {
-            return 'Deny';
+            return {
+                verdict: 'Deny',
+                isImplicit: false,
+            };
         }
         if (singlePolicyVerdict === 'Allow') {
+            allow = true;
+        } else if (singlePolicyVerdict === 'AllowWithTagCondition') {
+            allowWithTagCondition = true;
+        } else if (singlePolicyVerdict === 'DenyWithTagCondition') {
+            denyWithTagCondition = true;
+        } // else 'Neutral'
+    }
+    let verdict;
+    let isImplicit = false;
+    if (allow) {
+        if (denyWithTagCondition) {
+            verdict = 'NeedTagConditionEval';
+        } else {
             verdict = 'Allow';
         }
+    } else {
+        if (allowWithTagCondition) {
+            verdict = 'NeedTagConditionEval';
+        } else {
+            verdict = 'Deny';
+            isImplicit = true;
         }
-    log.trace('result of evaluating all pollicies', { verdict });
-    return verdict;
-};
+    }
+    log.trace('result of evaluating all policies', { verdict, isImplicit });
+    return { verdict, isImplicit };
+}

lib/policyEvaluator/principal.ts

@@ -23,15 +23,22 @@ export default class Principal {
      * @param statement - Statement policy field
      * @return True if meet conditions
      */
-    static _evaluateCondition(
+    static _evaluateStatement(
         params: Params,
         statement: Statement,
-        // TODO Fix return type
-    ): any {
-        if (statement.Condition) {
-            return meetConditions(params.rc, statement.Condition, params.log);
+    ): 'Neutral' | 'Allow' | 'Deny' {
+        const reverse = !!statement.NotPrincipal;
+        if (reverse) {
+            // In case of anonymous NotPrincipal, this will neutral everyone
+            return 'Neutral';
         }
-        return true;
+        if (statement.Condition) {
+            const conditionEval = meetConditions(params.rc, statement.Condition, params.log);
+            if (conditionEval === false || conditionEval === null) {
+                return 'Neutral';
+            }
+        }
+        return statement.Effect;
     }
 
     /**
@@ -48,19 +55,12 @@ export default class Principal {
         statement: Statement,
         valids: Valid,
     ): 'Neutral' | 'Allow' | 'Deny' {
-        const reverse = !!statement.NotPrincipal;
         const principal = (statement.Principal || statement.NotPrincipal)!;
-        if (typeof principal === 'string' && principal === '*') {
-            if (reverse) {
-                // In case of anonymous NotPrincipal, this will neutral everyone
-                return 'Neutral';
+        const reverse = !!statement.NotPrincipal;
+        if (typeof principal === 'string') {
+            if (principal === '*') {
+                return Principal._evaluateStatement(params, statement);
             }
-            const conditionEval = Principal._evaluateCondition(params, statement);
-            if (!conditionEval || conditionEval.allow === false) {
-                return 'Neutral';
-            }
-            return statement.Effect;
-        } else if (typeof principal === 'string') {
             return 'Deny';
         }
         let ref = [];
@@ -82,28 +82,8 @@ export default class Principal {
         }
         toCheck = Array.isArray(toCheck) ? toCheck : [toCheck];
         ref = Array.isArray(ref) ? ref : [ref];
-        if (toCheck.indexOf('*') !== -1) {
-            if (reverse) {
-                return 'Neutral';
-            }
-            const conditionEval = Principal._evaluateCondition(params, statement);
-            if (!conditionEval || conditionEval.allow === false) {
-                return 'Neutral';
-            }
-            return statement.Effect;
-        }
-        const len = ref.length;
-        for (let i = 0; i < len; ++i) {
-            if (toCheck.indexOf(ref[i]) !== -1) {
-                if (reverse) {
-                    return 'Neutral';
-                }
-                const conditionEval = Principal._evaluateCondition(params, statement);
-                if (!conditionEval || conditionEval.allow === false) {
-                    return 'Neutral';
-                }
-                return statement.Effect;
-            }
+        if (toCheck.includes('*') || ref.some(r => toCheck.includes(r))) {
+            return Principal._evaluateStatement(params, statement);
         }
         if (reverse) {
             return statement.Effect;

lib/policyEvaluator/utils/actionMaps.ts

@@ -4,14 +4,14 @@ const sharedActionMap = {
     bucketDeleteEncryption: 's3:PutEncryptionConfiguration',
     bucketDeletePolicy: 's3:DeleteBucketPolicy',
     bucketDeleteWebsite: 's3:DeleteBucketWebsite',
-    bucketDeleteTagging: 's3:DeleteBucketTagging',
+    bucketDeleteTagging: 's3:PutBucketTagging',
     bucketGet: 's3:ListBucket',
     bucketGetACL: 's3:GetBucketAcl',
     bucketGetCors: 's3:GetBucketCORS',
     bucketGetEncryption: 's3:GetEncryptionConfiguration',
     bucketGetLifecycle: 's3:GetLifecycleConfiguration',
     bucketGetLocation: 's3:GetBucketLocation',
-    bucketGetNotification: 's3:GetBucketNotificationConfiguration',
+    bucketGetNotification: 's3:GetBucketNotification',
     bucketGetObjectLock: 's3:GetBucketObjectLockConfiguration',
     bucketGetPolicy: 's3:GetBucketPolicy',
     bucketGetReplication: 's3:GetReplicationConfiguration',
@@ -23,7 +23,7 @@ const sharedActionMap = {
     bucketPutCors: 's3:PutBucketCORS',
     bucketPutEncryption: 's3:PutEncryptionConfiguration',
     bucketPutLifecycle: 's3:PutLifecycleConfiguration',
-    bucketPutNotification: 's3:PutBucketNotificationConfiguration',
+    bucketPutNotification: 's3:PutBucketNotification',
     bucketPutObjectLock: 's3:PutBucketObjectLockConfiguration',
     bucketPutPolicy: 's3:PutBucketPolicy',
     bucketPutReplication: 's3:PutReplicationConfiguration',
@@ -42,6 +42,7 @@ const sharedActionMap = {
     objectGetLegalHold: 's3:GetObjectLegalHold',
     objectGetRetention: 's3:GetObjectRetention',
     objectGetTagging: 's3:GetObjectTagging',
+    objectHead: 's3:GetObject',
     objectPut: 's3:PutObject',
     objectPutACL: 's3:PutObjectAcl',
     objectPutLegalHold: 's3:PutObjectLegalHold',
@@ -51,6 +52,12 @@ const sharedActionMap = {
     objectPutVersion: 's3:PutObjectVersion',
 };
 
+const actionMapBucketQuotas = {
+    bucketGetQuota: 'scality:GetBucketQuota',
+    bucketUpdateQuota: 'scality:UpdateBucketQuota',
+    bucketDeleteQuota: 'scality:DeleteBucketQuota',
+};
+
 // action map used for request context
 const actionMapRQ = {
     bucketPut: 's3:CreateBucket',
@@ -58,36 +65,35 @@ const actionMapRQ = {
     // see http://docs.aws.amazon.com/AmazonS3/latest/API/
     // RESTBucketDELETEcors.html
     bucketDeleteCors: 's3:PutBucketCORS',
-    bucketDeleteReplication: 's3:DeleteReplicationConfiguration',
-    bucketDeleteLifecycle: 's3:DeleteLifecycleConfiguration',
+    bucketDeleteReplication: 's3:PutReplicationConfiguration',
+    bucketDeleteLifecycle: 's3:PutLifecycleConfiguration',
     completeMultipartUpload: 's3:PutObject',
     initiateMultipartUpload: 's3:PutObject',
     objectDeleteVersion: 's3:DeleteObjectVersion',
     objectDeleteTaggingVersion: 's3:DeleteObjectVersionTagging',
+    objectGetArchiveInfo: 'scality:GetObjectArchiveInfo',
     objectGetVersion: 's3:GetObjectVersion',
     objectGetACLVersion: 's3:GetObjectVersionAcl',
     objectGetTaggingVersion: 's3:GetObjectVersionTagging',
-    objectHead: 's3:GetObject',
     objectPutACLVersion: 's3:PutObjectVersionAcl',
     objectPutPart: 's3:PutObject',
     objectPutTaggingVersion: 's3:PutObjectVersionTagging',
     serviceGet: 's3:ListAllMyBuckets',
     objectReplicate: 's3:ReplicateObject',
-    objectPutRetentionVersion: 's3:PutObjectVersionRetention',
-    objectPutLegalHoldVersion: 's3:PutObjectVersionLegalHold',
+    objectGetRetentionVersion: 's3:GetObjectRetention',
+    objectPutRetentionVersion: 's3:PutObjectRetention',
+    objectGetLegalHoldVersion: 's3:GetObjectLegalHold',
+    objectPutLegalHoldVersion: 's3:PutObjectLegalHold',
+    listObjectVersions: 's3:ListBucketVersions',
     ...sharedActionMap,
+    ...actionMapBucketQuotas,
 };
 
 // action map used for bucket policies
-const actionMapBP = { ...sharedActionMap };
+const actionMapBP = actionMapRQ;
 
 // action map for all relevant s3 actions
 const actionMapS3 = {
-    // TODO
-    // @ts-ignore
-    bucketGetNotification: 's3:GetBucketNotification',
-    // @ts-ignore
-    bucketPutNotification: 's3:PutBucketNotification',
     ...sharedActionMap,
     ...actionMapRQ,
     ...actionMapBP,
@@ -107,7 +113,7 @@ const actionMonitoringMapS3 = {
     bucketGetCors: 'GetBucketCors',
     bucketGetLifecycle: 'GetBucketLifecycleConfiguration',
     bucketGetLocation: 'GetBucketLocation',
-    bucketGetNotification: 'GetBucketNotificationConfiguration',
+    bucketGetNotification: 'GetBucketNotification',
     bucketGetObjectLock: 'GetObjectLockConfiguration',
     bucketGetPolicy: 'GetBucketPolicy',
     bucketGetReplication: 'GetBucketReplication',
@@ -120,7 +126,7 @@ const actionMonitoringMapS3 = {
     bucketPutACL: 'PutBucketAcl',
     bucketPutCors: 'PutBucketCors',
     bucketPutLifecycle: 'PutBucketLifecycleConfiguration',
-    bucketPutNotification: 'PutBucketNotificationConfiguration',
+    bucketPutNotification: 'PutBucketNotification',
     bucketPutObjectLock: 'PutObjectLockConfiguration',
     bucketPutPolicy: 'PutBucketPolicy',
     bucketPutReplication: 'PutBucketReplication',
@@ -153,6 +159,15 @@ const actionMonitoringMapS3 = {
     objectPutTagging: 'PutObjectTagging',
     objectRestore: 'RestoreObject',
     serviceGet: 'ListBuckets',
+    bucketGetQuota: 'GetBucketQuota',
+    bucketUpdateQuota: 'UpdateBucketQuota',
+    bucketDeleteQuota: 'DeleteBucketQuota', 
+};
+
+const actionMapAccountQuotas = {
+    UpdateAccountQuota : 'scality:UpdateAccountQuota',
+    DeleteAccountQuota : 'scality:DeleteAccountQuota',
+    GetAccountQuota : 'scality:GetAccountQuota',
 };
 
 const actionMapIAM = {
@@ -188,10 +203,15 @@ const actionMapIAM = {
     removeUserFromGroup: 'iam:RemoveUserFromGroup',
     updateAccessKey: 'iam:UpdateAccessKey',
     updateGroup: 'iam:UpdateGroup',
+    updateRole: 'iam:UpdateRole',
     updateUser: 'iam:UpdateUser',
     getAccessKeyLastUsed: 'iam:GetAccessKeyLastUsed',
     generateCredentialReport: 'iam:GenerateCredentialReport',
     getCredentialReport: 'iam:GetCredentialReport',
+    tagUser: 'iam:TagUser',
+    unTagUser: 'iam:UntagUser',
+    listUserTags: 'iam:ListUserTags',
+    ...actionMapAccountQuotas,
 };
 
 const actionMapSSO = {
@@ -207,6 +227,14 @@ const actionMapMetadata = {
     default: 'metadata:bucketd',
 };
 
+const actionMapScuba = {
+    GetMetrics: 'scuba:GetMetrics',
+    AdminStartIngest: 'scuba:AdminStartIngest',
+    AdminStopIngest: 'scuba:AdminStopIngest',
+    AdminReadRaftCseq: 'scuba:AdminReadRaftCseq',
+    AdminTriggerRepair: 'scuba:AdminTriggerRepair',
+};
+
 export {
     actionMapRQ,
     actionMapBP,
@@ -216,4 +244,5 @@ export {
     actionMapSSO,
     actionMapSTS,
     actionMapMetadata,
+    actionMapScuba,
 };

lib/policyEvaluator/utils/checkArnMatch.ts

@@ -1,5 +1,5 @@
 import { handleWildcardInResource } from './wildcards';
-
+import { policyArnAllowedEmptyAccountId } from '../../constants';
 /**
  * Checks whether an ARN from a request matches an ARN in a policy
  * to compare against each portion of the ARN from the request
@@ -38,9 +38,10 @@ export default function checkArnMatch(
         const requestSegment = caseSensitive ? requestArnArr[j] :
             requestArnArr[j].toLowerCase();
         const policyArnArr = policyArn.split(':');
-        // We want to allow an empty account ID for utapi service ARNs to not
+        // We want to allow an empty account ID for utapi and scuba service ARNs to not
         // break compatibility.
-        if (j === 4 && policyArnArr[2] === 'utapi' && policyArnArr[4] === '') {
+        if (j === 4 && policyArnAllowedEmptyAccountId.includes(policyArnArr[2])
+            && policyArnArr[4] === '') {
             continue;
         } else if (!segmentRegEx.test(requestSegment)) {
             return false;

lib/policyEvaluator/utils/conditions.ts

@@ -11,31 +11,30 @@ import ipaddr from 'ipaddr.js';
  * @param requestContext - info sent with request
  * @return condition key value
  */
-export const findConditionKey = (
+export function findConditionKey(
     key: string,
     requestContext: RequestContext,
-): string => {
+): any {
     // TODO: Consider combining with findVariable function if no benefit
     // to keeping separate
     const headers = requestContext.getHeaders();
     const query = requestContext.getQuery();
     const requesterInfo = requestContext.getRequesterInfo();
 
-    const map = new Map();
     // Possible AWS Condition keys (http://docs.aws.amazon.com/IAM/latest/
     // UserGuide/reference_policies_elements.html#AvailableKeys)
-
+    switch (key) {
     // aws:CurrentTime – Used for date/time conditions
     // (see Date Condition Operators).
-    map.set('aws:CurrentTime', new Date().toISOString());
+    case 'aws:CurrentTime': return new Date().toISOString();
     // aws:EpochTime – Used for date/time conditions
     // (see Date Condition Operators).
-    map.set('aws:EpochTime', Date.now().toString());
+    case 'aws:EpochTime': return Date.now().toString();
     // aws:TokenIssueTime – Date/time that temporary security
     // credentials were issued (see Date Condition Operators).
     // Only present in requests that are signed using temporary security
     // credentials.
-    map.set('aws:TokenIssueTime', requestContext.getTokenIssueTime());
+    case 'aws:TokenIssueTime': return requestContext.getTokenIssueTime();
     // aws:MultiFactorAuthPresent – Used to check whether MFA was used
     // (see Boolean Condition Operators).
     // Note: This key is only present if MFA was used. So, the following
@@ -45,133 +44,137 @@ export const findConditionKey = (
     // Instead use:
     //     "Condition" :
     //     { "Null" : { "aws:MultiFactorAuthPresent" : true } }
-    map.set('aws:MultiFactorAuthPresent',
-        requestContext.getMultiFactorAuthPresent());
+    case 'aws:MultiFactorAuthPresent': return requestContext.getMultiFactorAuthPresent();
     // aws:MultiFactorAuthAge – Used to check how many seconds since
     // MFA credentials were issued. If MFA was not used,
     // this key is not present
-    map.set('aws:MultiFactorAuthAge', requestContext.getMultiFactorAuthAge());
+    case 'aws:MultiFactorAuthAge': return requestContext.getMultiFactorAuthAge();
     // aws:principaltype states whether the principal is an account,
     // user, federated, or assumed role
     // Note: Docs for conditions have "PrincipalType" but simulator
     // and docs for variables have lowercase
-    map.set('aws:principaltype', requesterInfo.principaltype);
+    case 'aws:principaltype': return requesterInfo.principaltype;
     // aws:Referer – Used to check who referred the client browser to
     // the address the request is being sent to. Only supported by some
     // services, such as S3. Value comes from the referer header in the
     // HTTPS request made to AWS.
-    map.set('aws:referer', headers.referer);
+    case 'aws:referer': return headers.referer;
     // aws:SecureTransport – Used to check whether the request was sent
     // using SSL (see Boolean Condition Operators).
-    map.set('aws:SecureTransport',
-        requestContext.getSslEnabled() ? 'true' : 'false');
+    case 'aws:SecureTransport': return requestContext.getSslEnabled() ? 'true' : 'false';
     // aws:SourceArn – Used check the source of the request,
     // using the ARN of the source. N/A here.
-    map.set('aws:SourceArn', undefined);
+    case 'aws:SourceArn': return undefined;
     // aws:SourceIp – Used to check the requester's IP address
     // (see IP Address Condition Operators)
-    map.set('aws:SourceIp', requestContext.getRequesterIp());
+    case 'aws:SourceIp': return requestContext.getRequesterIp();
     // aws:SourceVpc – Used to restrict access to a specific
     // AWS Virtual Private Cloud. N/A here.
-    map.set('aws:SourceVpc', undefined);
+    case 'aws:SourceVpc': return undefined;
     // aws:SourceVpce – Used to limit access to a specific VPC endpoint
     // N/A here
-    map.set('aws:SourceVpce', undefined);
+    case 'aws:SourceVpce': return undefined;
     // aws:UserAgent – Used to check the requester's client app.
     // (see String Condition Operators)
-    map.set('aws:UserAgent', headers['user-agent']);
+    case 'aws:UserAgent': return headers['user-agent'];
     // aws:userid – Used to check the requester's unique user ID.
     // (see String Condition Operators)
-    map.set('aws:userid', requesterInfo.userid);
+    case 'aws:userid': return requesterInfo.userid;
     // aws:username – Used to check the requester's friendly user name.
     // (see String Condition Operators)
-    map.set('aws:username', requesterInfo.username);
+    case 'aws:username': return requesterInfo.username;
     // Possible condition keys for S3:
     // s3:x-amz-acl is acl request for bucket or object put request
-    map.set('s3:x-amz-acl', headers['x-amz-acl']);
+    case 's3:x-amz-acl': return headers['x-amz-acl'];
     // s3:x-amz-grant-PERMISSION (where permission can be:
     // read, write, read-acp, write-acp or full-control)
     // Value is the value of that header (ex. id of grantee)
-    map.set('s3:x-amz-grant-read', headers['x-amz-grant-read']);
-    map.set('s3:x-amz-grant-write', headers['x-amz-grant-write']);
-    map.set('s3:x-amz-grant-read-acp', headers['x-amz-grant-read-acp']);
-    map.set('s3:x-amz-grant-write-acp', headers['x-amz-grant-write-acp']);
-    map.set('s3:x-amz-grant-full-control', headers['x-amz-grant-full-control']);
+    case 's3:x-amz-grant-read': return headers['x-amz-grant-read'];
+    case 's3:x-amz-grant-write': return headers['x-amz-grant-write'];
+    case 's3:x-amz-grant-read-acp': return headers['x-amz-grant-read-acp'];
+    case 's3:x-amz-grant-write-acp': return headers['x-amz-grant-write-acp'];
+    case 's3:x-amz-grant-full-control': return headers['x-amz-grant-full-control'];
     // s3:x-amz-copy-source is x-amz-copy-source header if applicable on
     // a put object
-    map.set('s3:x-amz-copy-source', headers['x-amz-copy-source']);
+    case 's3:x-amz-copy-source': return headers['x-amz-copy-source'];
     // s3:x-amz-metadata-directive is x-amz-metadata-directive header if
     // applicable on a put object copy.  Determines whether metadata will
     // be copied from original object or replaced. Values or "COPY" or
     // "REPLACE".  Default is "COPY"
-    map.set('s3:x-amz-metadata-directive', headers['metadata-directive']);
+    case 's3:x-amz-metadata-directive': return headers['metadata-directive'];
     // s3:x-amz-server-side-encryption -- Used to require that object put
     // use server side encryption.  Value is the encryption algo such as
     // "AES256"
-    map.set('s3:x-amz-server-side-encryption',
-        headers['x-amz-server-side-encryption']);
+    case 's3:x-amz-server-side-encryption': return headers['x-amz-server-side-encryption'];
     // s3:x-amz-storage-class -- x-amz-storage-class header value
     // (STANDARD, etc.)
-    map.set('s3:x-amz-storage-class', headers['x-amz-storage-class']);
+    case 's3:x-amz-storage-class': return headers['x-amz-storage-class'];
     // s3:VersionId -- version id of object
-    map.set('s3:VersionId', query.versionId);
+    case 's3:VersionId': return query.versionId;
     // s3:LocationConstraint -- Used to restrict creation of bucket
     // in certain region.  Only applicable for CreateBucket
-    map.set('s3:LocationConstraint', requestContext.getLocationConstraint());
+    case 's3:LocationConstraint': return requestContext.getLocationConstraint();
     // s3:delimiter is delimiter for listing request
-    map.set('s3:delimiter', query.delimiter);
+    case 's3:delimiter': return query.delimiter;
     // s3:max-keys is max-keys for listing request
-    map.set('s3:max-keys', query['max-keys']);
+    case 's3:max-keys': return query['max-keys'];
     // s3:prefix is prefix for listing request
-    map.set('s3:prefix', query.prefix);
+    case 's3:prefix': return query.prefix;
     // s3 auth v4 additional condition keys
     // (See http://docs.aws.amazon.com/AmazonS3/latest/API/
     // bucket-policy-s3-sigv4-conditions.html)
     // s3:signatureversion -- Either "AWS" for v2 or
     // "AWS4-HMAC-SHA256" for v4
-    map.set('s3:signatureversion', requestContext.getSignatureVersion());
+    case 's3:signatureversion': return requestContext.getSignatureVersion();
     // s3:authType -- Method of authentication: either "REST-HEADER",
     // "REST-QUERY-STRING" or "POST"
-    map.set('s3:authType', requestContext.getAuthType());
+    case 's3:authType': return requestContext.getAuthType();
     // s3:signatureAge is the length of time, in milliseconds,
     // that a signature is valid in an authenticated request.  So,
     // can use this to limit the age to less than 7 days
-    map.set('s3:signatureAge', requestContext.getSignatureAge());
+    case 's3:signatureAge': return requestContext.getSignatureAge();
     // s3:x-amz-content-sha256 - Valid value is "UNSIGNED-PAYLOAD"
     // so can use this in a deny policy to deny any requests that do not
     // have a signed payload
-    map.set('s3:x-amz-content-sha256', headers['x-amz-content-sha256']);
+    case 's3:x-amz-content-sha256': return headers['x-amz-content-sha256'];
     // s3:ObjLocationConstraint is the location constraint set for an
     // object on a PUT request using the "x-amz-meta-scal-location-constraint"
     // header
-    map.set('s3:ObjLocationConstraint',
-        headers['x-amz-meta-scal-location-constraint']);
-    map.set('sts:ExternalId', requestContext.getRequesterExternalId());
-    map.set('keycloak:groups', requesterInfo.keycloakGroup);
-    map.set('keycloak:roles', requesterInfo.keycloakRole);
-    map.set('iam:PolicyArn', requestContext.getPolicyArn());
+    case 's3:ObjLocationConstraint': return headers['x-amz-meta-scal-location-constraint'];
+    case 'sts:ExternalId': return requestContext.getRequesterExternalId();
+    case 'keycloak:groups': return requesterInfo.keycloakGroup;
+    case 'keycloak:roles': return requesterInfo.keycloakRole;
+    case 'iam:PolicyArn': return requestContext.getPolicyArn();
     // s3:ExistingObjectTag - Used to check that existing object tag has
     // specific tag key and value. Extraction of correct tag key is done in CloudServer.
     // On first pass of policy evaluation, CloudServer information will not be included,
     // so evaluation should be skipped
-    map.set('s3:ExistingObjectTag', requestContext.getNeedTagEval() ? requestContext.getExistingObjTag() : undefined);
+    case 's3:ExistingObjectTag':
+        return requestContext.getNeedTagEval()
+            ? requestContext.getExistingObjTag() : undefined;
     // s3:RequestObjectTag - Used to limit putting object tags to specific
     // tag key and value. N/A here.
     // Requires information from CloudServer
     // On first pass of policy evaluation, CloudServer information will not be included,
     // so evaluation should be skipped
-    map.set('s3:RequestObjectTagKey', requestContext.getNeedTagEval() ? requestContext.getRequestObjTags() : undefined);
+    case 's3:RequestObjectTagKey':
+        return requestContext.getNeedTagEval()
+            ? requestContext.getRequestObjTags() : undefined;
     // s3:RequestObjectTagKeys - Used to limit putting object tags specific tag keys.
     // Requires information from CloudServer.
     // On first pass of policy evaluation, CloudServer information will not be included,
     // so evaluation should be skipped
-    map.set('s3:RequestObjectTagKeys',
-        requestContext.getNeedTagEval() && requestContext.getRequestObjTags()
+    case 's3:RequestObjectTagKeys':
+        return requestContext.getNeedTagEval() && requestContext.getRequestObjTags()
             ? getTagKeys(requestContext.getRequestObjTags()!)
-        : undefined,
-    );
-    return map.get(key);
-};
+            : undefined;
+    // The maximum retention period is 100 years.
+    case 's3:object-lock-remaining-retention-days':
+        return requestContext.getObjectLockRetentionDays() || undefined;
+    default:
+        return undefined;
+    }
+}
 
 
 // Wildcards are allowed in certain string comparison and arn comparisons
@@ -231,7 +234,7 @@ function convertToEpochTime(time: string | string[]) {
  * reference_policies_elements.html)
  * @return true if condition passes and false if not
  */
-export const convertConditionOperator = (operator: string): boolean => {
+export function convertConditionOperator(operator: string): boolean {
     // Policy Validator checks that the condition operator
     // is only one of these strings so should not have undefined
     // or security issue with object assignment
@@ -446,4 +449,4 @@ export const convertConditionOperator = (operator: string): boolean => {
         },
     };
     return operatorMap[operator];
-};
+}

lib/s3middleware/azureHelpers/ResultsCollector.ts

@@ -30,7 +30,7 @@ export default class ResultsCollector extends EventEmitter {
      * @emits ResultCollector#done
      * @emits ResultCollector#error
      */
-    pushResult(err: Error | undefined, subPartIndex: number) {
+    pushResult(err: Error | null | undefined, subPartIndex: number) {
         this._results.push({
             error: err,
             subPartIndex,

lib/s3middleware/azureHelpers/mpuUtils.ts

@@ -1,11 +1,15 @@
 import assert from 'assert';
 import * as crypto from 'crypto';
 import * as stream from 'stream';
+import azure from '@azure/storage-blob';
+
+import { RequestLogger } from 'werelogs';
+
 import ResultsCollector from './ResultsCollector';
 import SubStreamInterface from './SubStreamInterface';
 import * as objectUtils from '../objectUtils';
 import MD5Sum from '../MD5Sum';
-import errors from '../../errors';
+import errors, { ArsenalError } from '../../errors';
 
 export const splitter = '|';
 export const overviewMpuKey = 'azure_mpu';
@@ -61,7 +65,7 @@ export const getBlockId = (
     const paddedSubPart = padString(subPartIndex, 'subPart');
     const blockId = `${uploadId}${splitter}partNumber${paddedPartNumber}` +
         `${splitter}subPart${paddedSubPart}${splitter}`;
-    return padString(blockId, 'part');
+    return Buffer.from(padString(blockId, 'part')).toString('base64');
 };
 
 export const getSummaryPartId = (partNumber: number, eTag: string, size: number) => {
@@ -100,10 +104,17 @@ export const getSubPartIds = (
 ) => [...Array(part.numberSubParts).keys()].map(subPartIndex =>
         getBlockId(uploadId, part.partNumber, subPartIndex));
 
-// TODO Better type this
+type ErrorWrapperFn = (
+    s3Method: string,
+    azureMethod: string,
+    command: (client: azure.ContainerClient) => Promise<any>,
+    log: RequestLogger,
+    cb: (err: ArsenalError | null | undefined) => void,
+) => void
+
 export const putSinglePart = (
-    errorWrapperFn: (first: string, second: string, third: any, log: any, cb: any) => void,
-    request: any,
+    errorWrapperFn: ErrorWrapperFn,
+    request: stream.Readable,
     params: {
         bucketName: string;
         partNumber: number;
@@ -114,44 +125,44 @@ export const putSinglePart = (
     },
     dataStoreName: string,
     log: RequestLogger,
-    cb: any,
+    cb: (err: ArsenalError | null | undefined, dataStoreETag?: string, size?: number) => void,
 ) => {
     const { bucketName, partNumber, size, objectKey, contentMD5, uploadId }
         = params;
     const blockId = getBlockId(uploadId, partNumber, 0);
     const passThrough = new stream.PassThrough();
     const options = contentMD5
-        ? { useTransactionalMD5: true, transactionalContentMD5: contentMD5 }
+        ? { transactionalContentMD5: objectUtils.getMD5Buffer(contentMD5) }
         : {};
     request.pipe(passThrough);
-    return errorWrapperFn('uploadPart', 'createBlockFromStream',
-        [blockId, bucketName, objectKey, passThrough, size, options,
-            (err: any | null, result: any) => {
-                if (err) {
+    return errorWrapperFn('uploadPart', 'createBlockFromStream', async client => {
+        try {
+            const result = await client.getBlockBlobClient(objectKey)
+                .stageBlock(blockId, () => passThrough, size, options);
+            const md5 = result.contentMD5 || '';
+            const eTag = objectUtils.getHexMD5(md5);
+            return eTag
+        } catch (err: any) {
             log.error('Error from Azure data backend uploadPart',
                 { error: err.message, dataStoreName });
             if (err.code === 'ContainerNotFound') {
-                        return cb(errors.NoSuchBucket);
+                throw errors.NoSuchBucket;
             }
             if (err.code === 'InvalidMd5') {
-                        return cb(errors.InvalidDigest);
+                throw errors.InvalidDigest;
             }
             if (err.code === 'Md5Mismatch') {
-                        return cb(errors.BadDigest);
+                throw errors.BadDigest;
             }
-                    return cb(errors.InternalError.customizeDescription(
-                        `Error returned from Azure: ${err.message}`),
+            throw errors.InternalError.customizeDescription(
+                `Error returned from Azure: ${err.message}`
             );
         }    
-                const md5 = result.headers['content-md5'] || '';
-                const eTag = objectUtils.getHexMD5(md5);
-                return cb(null, eTag, size);
-            }], log, cb);
+    }, log, cb);
 };
 
-// TODO type this
-export const putNextSubPart = (
-    errorWrapperFn: any,
+const putNextSubPart = (
+    errorWrapperFn: ErrorWrapperFn,
     partParams: {
         uploadId: string;
         partNumber: number;
@@ -159,11 +170,10 @@ export const putNextSubPart = (
         objectKey: string;
     },
     subPartInfo: { lastPartIndex: number; lastPartSize: number },
-    subPartStream: any,
+    subPartStream: stream.Readable,
     subPartIndex: number,
     resultsCollector: ResultsCollector,
     log: RequestLogger,
-    cb: any,
 ) => {
     const { uploadId, partNumber, bucketName, objectKey } = partParams;
     const subPartSize = getSubPartSize(
@@ -171,14 +181,20 @@ export const putNextSubPart = (
     const subPartId = getBlockId(uploadId, partNumber,
         subPartIndex);
     resultsCollector.pushOp();
-    errorWrapperFn('uploadPart', 'createBlockFromStream',
-        [subPartId, bucketName, objectKey, subPartStream, subPartSize,
-            {}, err => resultsCollector.pushResult(err, subPartIndex)], log, cb);
+    errorWrapperFn('uploadPart', 'createBlockFromStream', async client => {
+        try {
+            const result = await client.getBlockBlobClient(objectKey)
+                .stageBlock(subPartId, () => subPartStream, subPartSize, {});
+            resultsCollector.pushResult(null, subPartIndex);
+        } catch (err: any) {
+            resultsCollector.pushResult(err, subPartIndex);
+        }
+    }, log, () => {});
 };
 
 export const putSubParts = (
-    errorWrapperFn: any,
-    request: any,
+    errorWrapperFn: ErrorWrapperFn,
+    request: stream.Readable,
     params: {
         uploadId: string;
         partNumber: number;
@@ -188,7 +204,7 @@ export const putSubParts = (
     },
     dataStoreName: string,
     log: RequestLogger,
-    cb: any,
+    cb: (err: ArsenalError | null | undefined, dataStoreETag?: string) => void,
 ) => {
     const subPartInfo = getSubPartInfo(params.size);
     const resultsCollector = new ResultsCollector();
@@ -227,14 +243,13 @@ export const putSubParts = (
         const totalLength = streamInterface.getTotalBytesStreamed();
         log.trace('successfully put subparts to Azure',
             { numberSubParts, totalLength });
-        hashedStream.on('hashed', () => cb(null, hashedStream.completedHash,
-            totalLength));
+        hashedStream.on('hashed', () => cb(null, hashedStream.completedHash));
 
         // in case the hashed event was already emitted before the
         // event handler was registered:
         if (hashedStream.completedHash) {
             hashedStream.removeAllListeners('hashed');
-            return cb(null, hashedStream.completedHash, totalLength);
+            return cb(null, hashedStream.completedHash);
         }
         return undefined;
     });
@@ -242,7 +257,7 @@ export const putSubParts = (
     const currentStream = streamInterface.getCurrentStream();
     // start first put to Azure before we start streaming the data
     putNextSubPart(errorWrapperFn, params, subPartInfo,
-        currentStream, 0, resultsCollector, log, cb);
+        currentStream, 0, resultsCollector, log);
 
     request.pipe(hashedStream);
     hashedStream.on('end', () => {
@@ -262,8 +277,8 @@ export const putSubParts = (
             }
             const { nextStream, subPartIndex } =
                 streamInterface.transitionToNextStream();
-            putNextSubPart(errorWrapperFn, params, subPartInfo,
-                nextStream, subPartIndex, resultsCollector, log, cb);
+            putNextSubPart(errorWrapperFn, params, subPartInfo, nextStream,
+                subPartIndex, resultsCollector, log);
             streamInterface.write(firstChunk);
         } else {
             streamInterface.write(data);

lib/s3middleware/lifecycleHelpers/LifecycleDateTime.ts

@@ -1,19 +1,25 @@
-const oneDay = 24 * 60 * 60 * 1000; // Milliseconds in a day.
+import { scaleMsPerDay } from '../objectUtils';
+const msInOneDay = 24 * 60 * 60 * 1000; // Milliseconds in a day.
 
 export default class LifecycleDateTime {
     _transitionOneDayEarlier?: boolean;
     _expireOneDayEarlier?: boolean;
+    _timeProgressionFactor?: number;
+    _scaledMsPerDay: number;
 
     constructor(params?: {
         transitionOneDayEarlier: boolean;
         expireOneDayEarlier: boolean;
+        timeProgressionFactor: number;
     }) {
         this._transitionOneDayEarlier = params?.transitionOneDayEarlier;
         this._expireOneDayEarlier = params?.expireOneDayEarlier;
+        this._timeProgressionFactor = params?.timeProgressionFactor || 1;
+        this._scaledMsPerDay = scaleMsPerDay(this._timeProgressionFactor);
     }
 
     getCurrentDate() {
-        const timeTravel = this._expireOneDayEarlier ? oneDay : 0;
+        const timeTravel = this._expireOneDayEarlier ? msInOneDay : 0;
         return Date.now() + timeTravel;
     }
 
@@ -25,7 +31,7 @@ export default class LifecycleDateTime {
     findDaysSince(date: Date) {
         const now = this.getCurrentDate();
         const diff = now - date.getTime();
-        return Math.floor(diff / (1000 * 60 * 60 * 24));
+        return Math.floor(diff / this._scaledMsPerDay);
     }
 
     /**
@@ -52,8 +58,25 @@ export default class LifecycleDateTime {
         }
         if (transition.Days !== undefined) {
             const lastModifiedTime = this.getTimestamp(lastModified);
-            const timeTravel = this._transitionOneDayEarlier ? -oneDay : 0;
-            return lastModifiedTime + (transition.Days * oneDay) + timeTravel;
+            const timeTravel = this._transitionOneDayEarlier ? -msInOneDay : 0;
+            return lastModifiedTime + (transition.Days * this._scaledMsPerDay) + timeTravel;
+        }
+    }
+
+    /**
+     * Find the Unix time at which the non-current version transition should occur.
+     * @param transition - A non-current version transition from the lifecycle non-current version transitions
+     * @param lastModified - The object's last modified date
+     * @return - The normalized transition timestamp
+     */
+    getNCVTransitionTimestamp(
+        transition: { NoncurrentDays?: number },
+        lastModified: string,
+    ) {
+        if (transition.NoncurrentDays !== undefined) {
+            const lastModifiedTime = this.getTimestamp(lastModified);
+            const timeTravel = this._transitionOneDayEarlier ? -msInOneDay : 0;
+            return lastModifiedTime + (transition.NoncurrentDays * this._scaledMsPerDay) + timeTravel;
         }
     }
 }

lib/s3middleware/lifecycleHelpers/LifecycleUtils.ts

@@ -61,6 +61,47 @@ export default class LifecycleUtils {
         return trans1 > trans2 ? transition1 : transition2;
     }
 
+    /**
+    * Compare two non-current version transition rules and return the one that is most recent.
+    * @param params - The function parameters
+    * @param params.transition1 - A non-current version transition from the current rule
+    * @param params.transition2 - A non-current version transition from the previous rule
+    * @param params.lastModified - The object's last modified
+    * date
+    * @return The most applicable transition rule
+    */
+    compareNCVTransitions(params: {
+        lastModified: string;
+        transition1: any;
+        transition2?: any;
+    }): number | undefined;
+    compareNCVTransitions(params: {
+        lastModified: string;
+        transition1?: any;
+        transition2: any;
+    }): number | undefined;
+    compareNCVTransitions(params: {
+        lastModified: string;
+        transition1: any;
+        transition2: any;
+    }): number | undefined;
+    compareNCVTransitions(params: {
+        lastModified: string;
+        transition1?: any;
+        transition2?: any;
+    }) {
+        const { transition1, transition2, lastModified } = params;
+        if (transition1 === undefined) {
+            return transition2;
+        }
+        if (transition2 === undefined) {
+            return transition1;
+        }
+        const trans1 = this._datetime.getNCVTransitionTimestamp(transition1!, lastModified)!;
+        const trans2 = this._datetime.getNCVTransitionTimestamp(transition2!, lastModified)!;
+        return trans1 > trans2 ? transition1 : transition2;
+    }
+
     // TODO Fix This
     /**
     * Find the most relevant trantition rule for the given transitions array
@@ -98,6 +139,42 @@ export default class LifecycleUtils {
         });
     }
 
+    /**
+    * Find the most relevant non-current version transition rule for the given transitions array
+    * and any previously stored non-current version transition from another rule.
+    * @param params - The function parameters
+    * @param params.transitions - Array of lifecycle non-current version transitions
+    * @param params.lastModified - The object's last modified
+    * date
+    * @return The most applicable non-current version transition rule
+    */
+    getApplicableNCVTransition(params: {
+        store: any;
+        currentDate: Date;
+        transitions: any[];
+        lastModified: string;
+    }) {
+        const { transitions, store, lastModified, currentDate } = params;
+        const transition = transitions.reduce((result, transition) => {
+            const isApplicable = // Is the transition time in the past?
+            this._datetime.getTimestamp(currentDate) >=
+                this._datetime.getNCVTransitionTimestamp(transition, lastModified)!;
+            if (!isApplicable) {
+                return result;
+            }
+            return this.compareNCVTransitions({
+                transition1: transition,
+                transition2: result,
+                lastModified,
+            });
+        }, undefined);
+        return this.compareNCVTransitions({
+            transition1: transition,
+            transition2: store.NoncurrentVersionTransition,
+            lastModified,
+        });
+    }
+
     // TODO
     /**
     * Filter out all rules based on `Status` and `Filter` (Prefix and Tags)
@@ -207,6 +284,7 @@ export default class LifecycleUtils {
                 // Names are long, so obscuring a bit
                 const ncve = 'NoncurrentVersionExpiration';
                 const ncd = 'NoncurrentDays';
+                const nncv = 'NewerNoncurrentVersions';
 
                 if (!store[ncve]) {
                     store[ncve] = {};
@@ -214,6 +292,7 @@ export default class LifecycleUtils {
                 if (!store[ncve][ncd] || rule[ncve][ncd] < store[ncve][ncd]) {
                     store[ncve].ID = rule.ID;
                     store[ncve][ncd] = rule[ncve][ncd];
+                    store[ncve][nncv] = rule[ncve][nncv];
                 }
             }
             if (rule.AbortIncompleteMultipartUpload
@@ -239,7 +318,17 @@ export default class LifecycleUtils {
                     currentDate,
                 });
             }
-            // TODO: Add support for NoncurrentVersionTransitions.
+
+            const ncvt = 'NoncurrentVersionTransitions';
+            const hasNoncurrentVersionTransitions = Array.isArray(rule[ncvt]) && rule[ncvt].length > 0;
+            if (hasNoncurrentVersionTransitions && this._supportedRules.includes('noncurrentVersionTransition')) {
+                store.NoncurrentVersionTransition = this.getApplicableNCVTransition({
+                    transitions: rule.NoncurrentVersionTransitions,
+                    lastModified: metadata.LastModified,
+                    store,
+                    currentDate,
+                });
+            }
             return store;
         }, {});
         // Do not transition to a location where the object is already stored.
@@ -247,6 +336,12 @@ export default class LifecycleUtils {
             && applicableRules.Transition.StorageClass === metadata.StorageClass) {
             applicableRules.Transition = undefined;
         }
+
+        if (applicableRules.NoncurrentVersionTransition
+            && applicableRules.NoncurrentVersionTransition.StorageClass === metadata.StorageClass) {
+            applicableRules.NoncurrentVersionTransition = undefined;
+        }
+
         return applicableRules;
         /* eslint-enable no-param-reassign */
     }

lib/s3middleware/objectUtils.ts

@@ -1,5 +1,21 @@
-export const getHexMD5 = (base64MD5: WithImplicitCoercion<string>) =>
-    Buffer.from(base64MD5, 'base64').toString('hex');
+const msInOneDay = 24 * 60 * 60 * 1000; // Milliseconds in a day.
+
+export const getMD5Buffer = (base64MD5: WithImplicitCoercion<string> | Uint8Array) =>
+    base64MD5 instanceof Uint8Array ? base64MD5 : Buffer.from(base64MD5, 'base64')
+
+export const getHexMD5 = (base64MD5: WithImplicitCoercion<string> | Uint8Array) =>
+    getMD5Buffer(base64MD5).toString('hex');
 
 export const getBase64MD5 = (hexMD5: WithImplicitCoercion<string>) =>
     Buffer.from(hexMD5, 'hex').toString('base64');
+
+
+/**
+ * Calculates the number of scaled milliseconds per day based on the given time progression factor.
+ * This function is intended for testing and simulation purposes only.
+ * @param {number} timeProgressionFactor - The desired time progression factor for scaling.
+ * @returns {number} The number of scaled milliseconds per day.
+ * If the result is 0, the minimum value of 1 millisecond is returned.
+ */
+export const scaleMsPerDay = (timeProgressionFactor: number): number =>
+    Math.round(msInOneDay / (timeProgressionFactor || 1)) || 1;

lib/s3middleware/tagging.ts

@@ -3,6 +3,11 @@ import * as werelogs from 'werelogs';
 import errors, { ArsenalError } from '../errors';
 import escapeForXml from './escapeForXml';
 
+export interface BucketTag {
+    Key: string;
+    Value: string;
+};
+
 const errorInvalidArgument = () => errors.InvalidArgument
     .customizeDescription('The header \'x-amz-tagging\' shall be ' +
   'encoded as UTF-8 then URLEncoded URL query parameters without ' +
@@ -32,6 +37,15 @@ export const _validator = {
         && tag.Key[0] !== undefined && tag.Value[0] !== undefined
         && typeof tag.Key[0] === 'string' && typeof tag.Value[0] === 'string',
 
+    // Allowed characters are letters, whitespace, and numbers, plus
+    // the following special characters: + - = . _ : /
+    // Maximum key length: 128 Unicode characters
+    // Maximum value length: 256 Unicode characters
+    validateTagObjectStructure: (tag: BucketTag) => tag
+        && Object.keys(tag).length === 2
+        && typeof tag.Key === 'string' && typeof tag.Value === 'string'
+        && tag.Key.length >= 1 && tag.Value.length >= 1,
+
     validateXMLStructure: (result: any) =>
         result && Object.keys(result).length === 1 &&
         result.Tagging &&
@@ -100,12 +114,47 @@ function _validateTags(tags: Array<{ Key: string[], Value: string[] }>) {
     }
     // not repeating keys
     if (tags.length > Object.keys(tagsResult).length) {
-        return errors.InvalidTag.customizeDescription('Cannot provide ' +
-        'multiple Tags with the same key');
+        return errors.InvalidTag.customizeDescription(
+            'Cannot provide multiple Tags with the same key'
+        );
     }
     return tagsResult;
 }
 
+/** areTagsValid - Validate bucket tags
+* @param tags - tags parsed from xml to be validated
+* @return result - true if the tags are valide, false otherwise
+*/
+export function areTagsValid(tags: Array<BucketTag>) {
+    if (tags.length === 0) {
+        return true;
+    }
+    // Maximum number of tags per resource: 50
+    if (tags.length > 50) {
+        return false;
+    }
+
+    const tagsResult = {};
+    for (const tag of tags) {
+        if (!_validator.validateTagObjectStructure(tag)) {
+            return false;
+        }
+        const { Key: key, Value: value } = tag;
+
+        const result = _validator.validateKeyValue(key, value);
+        if (result instanceof Error) {
+            return false;
+        }
+
+        tagsResult[key] = value;
+    }
+    // not repeating keys
+    if (tags.length > Object.keys(tagsResult).length) {
+        return false;
+    }
+    return true;
+}
+
 /** parseTagXml - Parse and validate xml body, returning callback with object
 * tags : { key: value}
 * @param xml - xml body to parse and validate

lib/s3routes/routes.ts

@@ -1,4 +1,7 @@
 import assert from 'assert';
+
+import { RequestLogger } from 'werelogs';
+
 import errors from '../errors';
 import routeGET from './routes/routeGET';
 import routePUT from './routes/routePUT';

lib/s3routes/routes/routeDELETE.ts

@@ -1,3 +1,5 @@
+import { RequestLogger } from 'werelogs';
+
 import * as routesUtils from '../routesUtils';
 import errors from '../../errors';
 import StatsClient from '../../metrics/StatsClient';
@@ -41,6 +43,8 @@ export default function routeDELETE(
             return call('bucketDeleteEncryption');
         } else if (query?.tagging !== undefined) {
             return call('bucketDeleteTagging');
+        } else if (query?.quota !== undefined) {
+            return call('bucketDeleteQuota');
         }
         call('bucketDelete');
     } else {

lib/s3routes/routes/routeGET.ts

@@ -1,3 +1,5 @@
+import { RequestLogger } from 'werelogs';
+
 import * as routesUtils from '../routesUtils';
 import errors from '../../errors';
 import * as http from 'http';
@@ -58,6 +60,8 @@ export default function routerGET(
             call('bucketGetEncryption');
         } else if (query.search !== undefined) {
             call('metadataSearch')
+        } else if (query.quota !== undefined) {
+            call('bucketGetQuota');
         } else {
             // GET bucket
             call('bucketGet');

lib/s3routes/routes/routeHEAD.ts

@@ -1,3 +1,5 @@
+import { RequestLogger } from 'werelogs';
+
 import * as routesUtils from '../routesUtils';
 import errors from '../../errors';
 import StatsClient from '../../metrics/StatsClient';

lib/s3routes/routes/routeOPTIONS.ts

@@ -1,3 +1,5 @@
+import { RequestLogger } from 'werelogs';
+
 import * as routesUtils from '../routesUtils';
 import errors from '../../errors';
 import * as http from 'http';

lib/s3routes/routes/routePOST.ts

@@ -1,3 +1,5 @@
+import { RequestLogger } from 'werelogs';
+
 import * as routesUtils from '../routesUtils';
 import errors from '../../errors';
 import * as http from 'http';

lib/s3routes/routes/routePUT.ts

@@ -1,3 +1,5 @@
+import { RequestLogger } from 'werelogs';
+
 import * as routesUtils from '../routesUtils';
 import errors from '../../errors';
 import * as http from 'http';
@@ -103,6 +105,13 @@ export default function routePUT(
                     return routesUtils.responseNoBody(err, corsHeaders,
                         response, 200, log);
                 });
+        } else if (query.quota !== undefined) {
+            api.callApiMethod('bucketUpdateQuota', request, response,
+                log, (err, resHeaders) => {
+                    routesUtils.statsReport500(err, statsClient);
+                    return routesUtils.responseNoBody(err, resHeaders, response,
+                        200, log);
+                });
         } else {
             // PUT bucket
             return api.callApiMethod('bucketPut', request, response, log,

lib/s3routes/routes/routeWebsite.ts

@@ -1,3 +1,5 @@
+import { RequestLogger } from 'werelogs';
+
 import * as routesUtils from '../routesUtils';
 import errors from '../../errors';
 import * as http from 'http';
@@ -27,6 +29,11 @@ export default function routerWebsite(
                 routesUtils.statsReport500(err, statsClient);
                 // request being redirected
                 if (redirectInfo) {
+                    if (err && redirectInfo.withError) {
+                        return routesUtils.redirectRequestOnError(err,
+                            'GET', redirectInfo, dataGetInfo, dataRetrievalParams,
+                            response, resMetaHeaders, log)
+                    }
                     // note that key might have been modified in websiteGet
                     // api to add index document
                     return routesUtils.redirectRequest(redirectInfo,
@@ -57,6 +64,11 @@ export default function routerWebsite(
             (err, resMetaHeaders, redirectInfo, key) => {
                 routesUtils.statsReport500(err, statsClient);
                 if (redirectInfo) {
+                    if (err && redirectInfo.withError) {
+                        return routesUtils.redirectRequestOnError(err,
+                            'HEAD', redirectInfo, null, dataRetrievalParams,
+                            response, resMetaHeaders, log)
+                    }
                     return routesUtils.redirectRequest(redirectInfo,
                         // TODO ARSN-217 encrypted does not exists in request.connection
                         // @ts-ignore

lib/s3routes/routesUtils.ts

@@ -1,12 +1,16 @@
 import * as url from 'url';
+import * as http from 'http';
+import { eachSeries } from 'async';
+
+import { RequestLogger } from 'werelogs';
+
 import * as ipCheck from '../ipCheck';
 import errors, { ArsenalError } from '../errors';
 import * as constants from '../constants';
-import { eachSeries } from 'async';
 import DataWrapper from '../storage/data/DataWrapper';
-import * as http from 'http';
 import StatsClient from '../metrics/StatsClient';
 import { objectKeyByteLimit } from '../constants';
+const jsutil = require('../jsutil');
 
 export type CallApiMethod = (
     methodName: string,
@@ -409,6 +413,7 @@ function retrieveData(
     return eachSeries(locations,
         (current, next) => data.get(current, response, log,
             (err: any, readable: http.IncomingMessage) => {
+                const cbOnce = jsutil.once(next);
                 // NB: readable is of IncomingMessage type
                 if (err) {
                     log.error('failed to get object', {
@@ -416,7 +421,7 @@ function retrieveData(
                         method: 'retrieveData',
                     });
                     _destroyResponse();
-                    return next(err);
+                    return cbOnce(err);
                 }
                 // response.isclosed is set by the S3 server. Might happen if
                 // the S3-client closes the connection before the first request
@@ -430,19 +435,19 @@ function retrieveData(
                     // @ts-ignore
                     responseErr.code = 'ResponseError';
                     responseErr.message = 'response closed by client request before all data sent';
-                    return next(responseErr);
+                    return cbOnce(responseErr);
                 }
                 // readable stream successfully consumed
                 readable.on('end', () => {
                     currentStream = null;
                     log.debug('readable stream end reached');
-                    return next();
+                    return cbOnce();
                 });
                 // errors on server side with readable stream
                 readable.on('error', err => {
                     log.error('error piping data from source');
                     _destroyResponse();
-                    return next(err);
+                    return cbOnce(err);
                 });
                 currentStream = readable;
                 return readable.pipe(response, { end: false });
@@ -689,6 +694,8 @@ export function streamUserErrorPage(
     log: RequestLogger,
 ) {
     setCommonResponseHeaders(corsHeaders, response, log);
+    response.setHeader('x-amz-error-code', err.message);
+    response.setHeader('x-amz-error-message', err.description);
     response.writeHead(err.code, { 'Content-type': 'text/html' });
     response.on('finish', () => {
         // TODO ARSN-216 Fix logger
@@ -871,7 +878,7 @@ export function redirectRequest(
     }
     let redirectLocation = justPath ? `/${redirectKey}` :
         `${redirectProtocol}://${redirectHostName}/${redirectKey}`;
-    if (!redirectKey && redirectLocationHeader) {
+    if (!redirectKey && redirectLocationHeader && redirectLocation !== '/') {
         // remove hanging slash
         redirectLocation = redirectLocation.slice(0, -1);
     }
@@ -888,6 +895,52 @@ export function redirectRequest(
     return undefined;
 }
 
+/**
+ * redirectRequestOnError - redirect with an error body
+ * @param err - arsenal error object
+ * @param method - HTTP method
+ * @param routingInfo - info for routing
+ * @param [routingInfo.withError] - flag to differentiate from routing rules
+ * @param [routingInfo.location] - location header
+ * @param dataLocations --
+ *   - array of locations to get streams from backend
+ * @param retrieveDataParams - params to create instance of
+ * data retrieval function
+ * @param response - response object
+ * @param corsHeaders - CORS-related response headers
+ * @param log - Werelogs instance
+ */
+export function redirectRequestOnError(
+    err: ArsenalError,
+    method: 'HEAD' | 'GET',
+    routingInfo: {
+        withError: true;
+        location: string;
+    },
+    dataLocations: { size: string | number }[] | null,
+    retrieveDataParams: any,
+    response: http.ServerResponse,
+    corsHeaders: { [key: string]: string },
+    log: RequestLogger,
+) {
+    response.setHeader('Location', routingInfo.location);
+
+    if (!dataLocations && err.is.Found) {
+        if (method === 'HEAD') {
+            return errorHeaderResponse(err, response, corsHeaders, log);
+        }
+        response.setHeader('x-amz-error-code', err.message);
+        response.setHeader('x-amz-error-message', err.description);
+        return errorHtmlResponse(err, false, '', response, corsHeaders, log);
+    }
+
+    // This is reached only for website error document (GET only)
+    const overrideErrorCode = err.flatten();
+    overrideErrorCode.code = 301;
+    return streamUserErrorPage(ArsenalError.unflatten(overrideErrorCode)!,
+        dataLocations || [], retrieveDataParams, response, corsHeaders, log);
+}
+
 /**
  * Get bucket name and object name from the request
  * @param request - http request object

lib/storage/data/DataWrapper.js

@@ -2,6 +2,8 @@ const async = require('async');
 const PassThrough = require('stream').PassThrough;
 const assert = require('assert');
 
+const { Logger } = require('werelogs');
+
 const errors = require('../../errors').default;
 const MD5Sum = require('../../s3middleware/MD5Sum').default;
 const NullStream = require('../../s3middleware/nullStream').default;
@@ -27,6 +29,7 @@ class DataWrapper {
         this.metadata = metadata;
         this.locStorageCheckFn = locStorageCheckFn;
         this.vault = vault;
+        this.logger = new Logger('DataWrapper');
     }
 
     put(cipherBundle, value, valueSize, keyContext, backendInfo, log, cb) {
@@ -127,7 +130,7 @@ class DataWrapper {
     }
 
     delete(objectGetInfo, log, cb) {
-        const callback = cb || log.end;
+        const callback = cb || (() => {});
         const isMdModelVersion2 = typeof(objectGetInfo) === 'string';
         const isRequiredStringKey =
             constants.clientsRequireStringKey[this.implName];
@@ -176,7 +179,9 @@ class DataWrapper {
             newObjDataStoreName)) {
             return process.nextTick(cb);
         }
-        log.trace('initiating batch delete', {
+        const delLog = this.logger.newRequestLoggerFromSerializedUids(
+            log.getSerializedUids());
+        delLog.trace('initiating batch delete', {
             keys: locations,
             implName: this.implName,
             method: 'batchDelete',
@@ -202,21 +207,21 @@ class DataWrapper {
             return false;
         });
         if (shouldBatchDelete && keys.length > 1) {
-            return this.client.batchDelete(backendName, { keys }, log, cb);
+            return this.client.batchDelete(backendName, { keys }, delLog, cb);
         }
         return async.eachLimit(locations, 5, (loc, next) => {
-            process.nextTick(() => this.delete(loc, log, next));
+            process.nextTick(() => this.delete(loc, delLog, next));
         },
         err => {
             if (err) {
-                log.end().error('batch delete failed', { error: err });
+                delLog.end().error('batch delete failed', { error: err });
                 // deletion of non-existing objects result in 204
                 if (err.code === 404) {
                     return cb();
                 }
                 return cb(err);
             }
-            log.end().trace('batch delete successfully completed');
+            delLog.end().trace('batch delete successfully completed');
             return cb();
         });
     }
@@ -984,13 +989,14 @@ class DataWrapper {
         return this.client.delete(objectGetInfo, log.getSerializedUids(),
             err => {
                 if (err) {
-                    if (err.is.ObjNotFound) {
+                    // TODO: sproxydclient and hdclient does not return standard Arsenal error yet.
+                    if (err.code === 404) {
                         log.info('no such key in datastore', {
                             objectGetInfo,
                             implName: this.implName,
                             moreRetries: 'no',
                         });
-                        return cb(err);
+                        return cb(errors.ObjNotFound);
                     }
                     log.error('delete error from datastore', {
                         error: err,

lib/storage/data/LocationConstraintParser.js

@@ -1,11 +1,10 @@
-const https = require('https');
-const http = require('http');
+const { http, https } = require('httpagent');
 const url = require('url');
 const AWS = require('aws-sdk');
-const Sproxy = require('sproxydclient');
-const Hyperdrive = require('hdclient');
 const HttpsProxyAgent = require('https-proxy-agent');
 
+require("aws-sdk/lib/maintenance_mode_message").suppress = true;
+
 const constants = require('../../constants');
 const DataFileBackend = require('./file/DataFileInterface');
 const inMemory = require('./in_memory/datastore').backend;
@@ -26,8 +25,13 @@ function parseLC(config, vault) {
         if (locationObj.type === 'file') {
             clients[location] = new DataFileBackend(config);
         }
+        if (locationObj.type === 'vitastor') {
+            const VitastorBackend = require('./vitastor/VitastorBackend');
+            clients[location] = new VitastorBackend(location, locationObj.details);
+        }
         if (locationObj.type === 'scality') {
             if (locationObj.details.connector.sproxyd) {
+                const Sproxy = require('sproxydclient');
                 clients[location] = new Sproxy({
                     bootstrap: locationObj.details.connector
                         .sproxyd.bootstrap,
@@ -42,6 +46,7 @@ function parseLC(config, vault) {
                 });
                 clients[location].clientType = 'scality';
             } else if (locationObj.details.connector.hdclient) {
+                const Hyperdrive = require('hdclient');
                 clients[location] = new Hyperdrive.hdcontroller.HDProxydClient(
                     locationObj.details.connector.hdclient);
                 clients[location].clientType = 'scality';
@@ -77,8 +82,8 @@ function parseLC(config, vault) {
                 connectionAgent = new HttpsProxyAgent(options);
             } else {
                 connectionAgent = sslEnabled ?
-                    new https.Agent(httpAgentConfig) :
-                    new http.Agent(httpAgentConfig);
+                    new https.Agent(httpAgentConfig, { maxSockets: false }) :
+                    new http.Agent(httpAgentConfig, { maxSockets: false });
             }
             const httpOptions = { agent: connectionAgent, timeout: 0 };
             const s3Params = {

lib/storage/data/MultipleBackendGateway.js

@@ -5,6 +5,7 @@ const { parseTagFromQuery } = require('../../s3middleware/tagging');
 const { externalBackendHealthCheckInterval } = require('../../constants');
 const DataFileBackend = require('./file/DataFileInterface');
 const { createLogger, checkExternalBackend } = require('./external/utils');
+const jsutil = require('../../jsutil');
 
 class MultipleBackendGateway {
     constructor(clients, metadata, locStorageCheckFn) {
@@ -199,11 +200,12 @@ class MultipleBackendGateway {
     uploadPart(request, streamingV4Params, stream, size, location, key,
         uploadId, partNumber, bucketName, log, cb) {
         const client = this.clients[location];
+        const cbOnce = jsutil.once(cb);
 
         if (client.uploadPart) {
             return this.locStorageCheckFn(location, size, log, err => {
                 if (err) {
-                    return cb(err);
+                    return cbOnce(err);
                 }
                 return client.uploadPart(request, streamingV4Params, stream,
                     size, key, uploadId, partNumber, bucketName, log,
@@ -217,14 +219,14 @@ class MultipleBackendGateway {
                                     'metric following object PUT failure',
                                         { error: error.message });
                                     }
-                                    return cb(err);
+                                    return cbOnce(err);
                                 });
                         }
-                        return cb(null, partInfo);
+                        return cbOnce(null, partInfo);
                     });
             });
         }
-        return cb();
+        return cbOnce();
     }
 
     listParts(key, uploadId, location, bucketName, partNumberMarker, maxParts,

lib/storage/data/external/AwsClient.js

@@ -8,6 +8,7 @@ const getMetaHeaders =
 const { prepareStream } = require('../../../s3middleware/prepareStream');
 const { createLogger, logHelper, removeQuotes, trimXMetaPrefix } =
     require('./utils');
+const jsutil = require('../../../jsutil');
 
 const missingVerIdInternalError = errors.InternalError.customizeDescription(
     'Invalid state. Please ensure versioning is enabled ' +
@@ -317,9 +318,11 @@ class AwsClient {
     uploadPart(request, streamingV4Params, stream, size, key, uploadId,
         partNumber, bucketName, log, callback) {
         let hashedStream = stream;
+        const cbOnce = jsutil.once(callback);
+
         if (request) {
             const partStream = prepareStream(request, streamingV4Params,
-                this._vault, log, callback);
+                this._vault, log, cbOnce);
             hashedStream = new MD5Sum();
             partStream.pipe(hashedStream);
         }
@@ -333,7 +336,7 @@ class AwsClient {
             if (err) {
                 logHelper(log, 'error', 'err from data backend ' +
                     'on uploadPart', err, this._dataStoreName, this.clientType);
-                return callback(errors.ServiceUnavailable
+                return cbOnce(errors.ServiceUnavailable
                     .customizeDescription('Error returned from ' +
                     `${this.type}: ${err.message}`),
                 );
@@ -347,7 +350,7 @@ class AwsClient {
                 dataStoreName: this._dataStoreName,
                 dataStoreETag: noQuotesETag,
             };
-            return callback(null, dataRetrievalInfo);
+            return cbOnce(null, dataRetrievalInfo);
         });
     }
 

lib/storage/data/external/AzureClient.js

@@ -1,62 +1,109 @@
-const url = require('url');
-
-const azure = require('azure-storage');
+const { BlobServiceClient, StorageSharedKeyCredential, AnonymousCredential } = require('@azure/storage-blob');
+const { ClientSecretCredential } = require('@azure/identity');
 const errors = require('../../../errors').default;
-const azureMpuUtils = require('../../../s3middleware/azureHelpers/mpuUtils').default;
+const azureMpuUtils = require('../../../s3middleware/azureHelpers/mpuUtils');
 const { validateAndFilterMpuParts } =
     require('../../../s3middleware/processMpuParts');
 
 const { createLogger, logHelper, translateAzureMetaHeaders } =
     require('./utils');
+const objectUtils = require('../../../s3middleware/objectUtils');
 
 const constants = require('../../../constants');
 const packageVersion = require('../../../../package.json').version;
 
-azure.Constants.USER_AGENT_PRODUCT_NAME = constants.productName;
-azure.Constants.USER_AGENT_PRODUCT_VERSION = packageVersion;
-
 class AzureClient {
+    static addQueryParams(endpoint, token) {
+        const url = new URL(endpoint);
+        const query = token.startsWith('?') ? token.slice(1) : token;
+        if (!url.search) {
+            url.search = `?${query}`;
+        } else if (url.search === '?') {
+            url.search += query;
+        } else {
+            url.search += `&${query}`;
+        }
+        return url.toString();
+    }
+
     constructor(config) {
         this._azureStorageEndpoint = config.azureStorageEndpoint;
         this._azureStorageCredentials = config.azureStorageCredentials;
         this._azureContainerName = config.azureContainerName;
-        this._client = azure.createBlobService(
-            this._azureStorageCredentials.storageAccountName,
-            this._azureStorageCredentials.storageAccessKey,
-            this._azureStorageEndpoint);
-        this._client.enableGlobalHttpAgent = true;
+        const cred = (credentialsConfig => {
+            switch (credentialsConfig.authMethod) {
+            case 'client-secret':
+                return new ClientSecretCredential(
+                    credentialsConfig.tenantId,
+                    credentialsConfig.clientId,
+                    credentialsConfig.clientKey,
+                );
+
+            case 'shared-access-signature':
+                this._azureStorageEndpoint = AzureClient.addQueryParams(
+                    this._azureStorageEndpoint, credentialsConfig.sasToken);
+                return new AnonymousCredential();
+
+            case 'shared-key':
+            default:
+                return new StorageSharedKeyCredential(
+                    credentialsConfig.storageAccountName,
+                    credentialsConfig.storageAccessKey,
+                );
+            }
+        })(this._azureStorageCredentials);
+        const proxyOptions = (() => {
+            if (!config.proxy || !config.proxy.url) {
+                return undefined;
+            }
+            // NOTE: config.proxy.certs is not supported
+            const parsedUrl = new URL(config.proxy.url);
+            return {
+                host: parsedUrl.host,
+                port: parsedUrl.port || 80,
+                username: parsedUrl.username || undefined,
+                password: parsedUrl.password || undefined,
+            };
+        })();
+        this._client = new BlobServiceClient(this._azureStorageEndpoint, cred, {
+            keepAliveOptions: {
+                enable: false, // Enable use of global HTTP agent
+            },
+            proxyOptions,
+            userAgentOptions: {
+                userAgentPrefix: `${constants.productName}/${packageVersion} `,
+            },
+        }).getContainerClient(this._azureContainerName);
         this._dataStoreName = config.dataStoreName;
         this._bucketMatch = config.bucketMatch;
-        if (config.proxy && config.proxy.url) {
-            const parsedUrl = url.parse(config.proxy.url);
-            if (!parsedUrl.port) {
-                parsedUrl.port = 80;
-            }
-            const proxyParams = parsedUrl;
-            if (config.proxy.certs) {
-                Object.assign(proxyParams, config.proxy.certs);
-            }
-            this._client.setProxy(proxyParams);
-        }
     }
 
-    _errorWrapper(s3Method, azureMethod, args, log, cb) {
+    /**
+     * Run azure method call.
+     * @param {string} [s3Method] S3 method name
+     * @param {string} [azureMethod] Azure method name
+     * @param {ErrorWrapper~Command} [command] Actual command to run
+     * @param {RequestLogger} [log] Logger
+     * @param {ErrorWrapper~Cb} [cb] The final callback
+     * @returns {void}
+     *
+     * @callback ErrorWrapper~Command
+     * @param {azure.ContainerClient} [client] Azure client to use
+     * @returns {Promise<any>}
+     *
+     * @callback ErrorWrapper~Cb
+     * @param {azure.ArsenalError} [arsenalErr] Error returned by the command
+     * @param {any} [result] Result of Azure SDK command
+     * @returns {void}
+     */
+    _errorWrapper(s3Method, azureMethod, command, log, cb) {
         if (log) {
-            log.info(`calling azure ${azureMethod}`);
-        }
-        try {
-            this._client[azureMethod].apply(this._client, args);
-        } catch (err) {
-            const error = errors.ServiceUnavailable;
-            if (log) {
-                log.error('error thrown by Azure Storage Client Library',
-                    { error: err.message, stack: err.stack, s3Method,
-                        azureMethod, dataStoreName: this._dataStoreName });
-            }
-            cb(error.customizeDescription('Error from Azure ' +
-                `method: ${azureMethod} on ${s3Method} S3 call: ` +
-                `${err.message}`));
+            log.info(`calling azure ${azureMethod} in ${s3Method}`);
         }
+        command(this._client).then(
+            result => cb(null, result),
+            cb,
+        );
     }
 
     _createAzureKey(requestBucketName, requestObjectKey,
@@ -119,6 +166,32 @@ class AzureClient {
         };
     }
 
+    /**
+     * Build Azure HTTP headers for content settings
+     * @param {object} [properties]                    The blob properties to set.
+     * @param {string} [properties.contentType]        The MIME content type of the blob.
+     *                                                 The default type is application/octet-stream.
+     * @param {string} [properties.contentEncoding]    The content encodings that have been applied
+     *                                                 to the blob.
+     * @param {string} [properties.contentLanguage]    The natural languages used by this resource.
+     * @param {string} [properties.cacheControl]       The blob's cache control.
+     * @param {string} [properties.contentDisposition] The blob's content disposition.
+     * @param {string} [properties.contentMD5]         The blob's MD5 hash.
+     * @returns {BlobHTTPHeaders} The headers
+     */
+    _getAzureContentSettingsHeaders(properties) {
+        return {
+            blobContentMD5: properties.contentMD5
+                ? objectUtils.getMD5Buffer(properties.contentMD5)
+                : undefined,
+            blobContentType: properties.contentType || undefined,
+            blobCacheControl: properties.cacheControl || undefined,
+            blobContentDisposition: properties.contentDisposition || undefined,
+            blobContentEncoding: properties.contentEncoding || undefined,
+            blobContentLanguage: properties.blobContentLanguage || undefined,
+        };
+    }
+
     put(stream, size, keyContext, reqUids, callback, skey, metadata) {
         const log = createLogger(reqUids);
         // before blob is put, make sure there is no ongoing MPU with same key
@@ -134,50 +207,59 @@ class AzureClient {
                 const options = {
                     metadata: translateAzureMetaHeaders(keyContext.metaHeaders,
                         keyContext.tagging),
-                    contentSettings: {
-                        contentType: keyContext.contentType || undefined,
-                        cacheControl: keyContext.cacheControl || undefined,
-                        contentDisposition: keyContext.contentDisposition ||
-                        undefined,
-                        contentEncoding: keyContext.contentEncoding || undefined,
-                    },
+                    blobHTTPHeaders: this._getAzureContentSettingsHeaders(
+                        keyContext || {}),
                 };
                 if (size === 0) {
-                    return this._errorWrapper('put', 'createBlockBlobFromText',
-                        [this._azureContainerName, azureKey, '', options,
-                            err => {
-                                if (err) {
-                                    logHelper(log, 'error', 'err from Azure PUT data ' +
-                                'backend', err, this._dataStoreName);
-                                    return callback(errors.ServiceUnavailable
-                                        .customizeDescription('Error returned from ' +
-                                `Azure: ${err.message}`));
+                    return this._errorWrapper('put', 'uploadData', async client => {
+                        try {
+                            await client.getBlockBlobClient(azureKey).upload('', 0, options);
+                            return azureKey;
+                        } catch (err) {
+                            logHelper(log, 'error', 'err from Azure PUT data backend',
+                                err, this._dataStoreName);
+                            throw errors.ServiceUnavailable.customizeDescription(
+                                `Error returned from Azure: ${err.message}`);
                         }
-                                return callback(null, azureKey);
-                            }], log, callback);
+                    }, log, callback);
                 }
-                return this._errorWrapper('put', 'createBlockBlobFromStream',
-                    [this._azureContainerName, azureKey, stream, size, options,
-                        err => {
-                            if (err) {
-                                logHelper(log, 'error', 'err from Azure PUT data ' +
-                            'backend', err, this._dataStoreName);
-                                return callback(errors.ServiceUnavailable
-                                    .customizeDescription('Error returned from ' +
-                            `Azure: ${err.message}`));
+                return this._errorWrapper('put', 'createBlockBlobFromStream', async client => {
+                    try {
+                        await client.getBlockBlobClient(azureKey).upload(() => stream, size, options);
+                        return azureKey;
+                    } catch (err) {
+                        logHelper(log, 'error', 'err from Azure PUT data backend',
+                            err, this._dataStoreName);
+                        throw errors.ServiceUnavailable.customizeDescription(
+                            `Error returned from Azure: ${err.message}`);
                     }
-                            return callback(null, azureKey);
-                        }], log, callback);
+                }, log, callback);
             });
     }
 
+    /**
+     * Build BlobRequestConditions from azureStreamingOptions
+     * @param {object} [objectGetInfoOptions]   Azure streaming options
+     * @param {object} [objectGetInfoOptions.accessConditions] Access conditions
+     * @param {Date}   [objectGetInfoOptions.accessConditions.DateUnModifiedSince] Filter objects not
+     *                                  modified since that date.
+     * @returns {BlobRequestConditions} Request conditions
+     */
+    _getAzureConditions(objectGetInfoOptions) {
+        const accessConditions = objectGetInfoOptions.accessConditions || {};
+        return {
+            ifUnmodifiedSince: accessConditions.DateUnModifiedSince || undefined,
+        };
+    }
+
     head(objectGetInfo, reqUids, callback) {
         const log = createLogger(reqUids);
-        const { key, azureStreamingOptions } = objectGetInfo;
-        return this._errorWrapper('head', 'getBlobProperties',
-            [this._azureContainerName, key, azureStreamingOptions,
-                (err, data) => {
-                    if (err) {
+        const { key } = objectGetInfo;
+        return this._errorWrapper('head', 'getBlobProperties', async client => {
+            try {
+                const data = await client.getBlockBlobClient(key).getProperties();
+                return data;
+            } catch (err) {
                 let logLevel;
                 let retError;
                 if (err.code === 'NotFound') {
@@ -185,42 +267,46 @@ class AzureClient {
                     retError = errors.LocationNotFound;
                 } else {
                     logLevel = 'error';
-                            retError = errors.ServiceUnavailable
-                                .customizeDescription(
+                    retError = errors.ServiceUnavailable.customizeDescription(
                         `Error returned from Azure: ${err.message}`);
                 }
                 logHelper(log, logLevel, 'err from Azure HEAD data backend',
                     err, this._dataStoreName);
-                        return callback(retError);
+                throw retError;
             }
-                    return callback(null, data);
-                }], log, callback);
+        }, log, callback);
     }
 
     get(objectGetInfo, range, reqUids, callback) {
         const log = createLogger(reqUids);
-        // for backwards compatibility
         const { key, response, azureStreamingOptions } = objectGetInfo;
-        let streamingOptions;
+        let rangeStart = 0;
+        let rangeEnd = undefined;
         if (azureStreamingOptions) {
             // option coming from api.get()
-            streamingOptions = azureStreamingOptions;
+            rangeStart = (typeof azureStreamingOptions.rangeStart === 'string')
+                ? parseInt(azureStreamingOptions.rangeStart, 10)
+                : azureStreamingOptions.rangeStart;
+            rangeEnd = (typeof azureStreamingOptions.rangeEnd === 'string')
+                ? parseInt(azureStreamingOptions.rangeEnd, 10)
+                : azureStreamingOptions.rangeEnd;
         } else if (range) {
             // option coming from multipleBackend.upload()
-            const rangeStart = (typeof range[0] === 'number') ? range[0].toString() : undefined;
-            const rangeEnd = range[1] ? range[1].toString() : undefined;
-            streamingOptions = { rangeStart, rangeEnd };
+            rangeStart = (typeof range[0] === 'number') ? range[0] : 0;
+            rangeEnd = range[1] || undefined;
         }
-        this._errorWrapper('get', 'getBlobToStream',
-            [this._azureContainerName, key, response, streamingOptions,
-                err => {
-                    if (err) {
+        this._errorWrapper('get', 'getBlobToStream', async client => {
+            try {
+                const rsp = await client.getBlockBlobClient(key)
+                    .download(rangeStart, rangeEnd - rangeStart + 1 || undefined);
+                rsp.readableStreamBody.pipe(response);
+                return response;
+            } catch (err) {
                 logHelper(log, 'error', 'err from Azure GET data backend',
                     err, this._dataStoreName);
-                        return callback(errors.ServiceUnavailable);
+                throw errors.ServiceUnavailable;
             }
-                    return callback(null, response);
-                }], log, callback);
+        }, log, callback);
     }
 
     delete(objectGetInfo, reqUids, callback) {
@@ -230,44 +316,46 @@ class AzureClient {
             objectGetInfo.key;
         let options;
         if (typeof objectGetInfo === 'object') {
-            options = objectGetInfo.options;
+            options = {
+                conditions: this._getAzureConditions(objectGetInfo.options || {}),
+            };
         }
-        return this._errorWrapper('delete', 'deleteBlobIfExists',
-            [this._azureContainerName, key, options,
-                err => {
-                    if (err && err.statusCode === 412) {
-                        return callback(errors.PreconditionFailed);
+        return this._errorWrapper('delete', 'deleteBlobIfExists', async client => {
+            try {
+                await client.getBlockBlobClient(key).deleteIfExists(options);
+            } catch (err) {
+                if (err.statusCode === 412) {
+                    throw errors.PreconditionFailed;
                 }
-                    if (err) {
                 const log = createLogger(reqUids);
-                        logHelper(log, 'error', 'error deleting object from ' +
-                        'Azure datastore', err, this._dataStoreName);
-                        return callback(errors.ServiceUnavailable
-                            .customizeDescription('Error returned from ' +
-                        `Azure: ${err.message}`));
+                logHelper(log, 'error', 'error deleting object from Azure datastore',
+                    err, this._dataStoreName);
+                throw errors.ServiceUnavailable.customizeDescription(
+                    `Error returned from Azure: ${err.message}`);
             }
-                    return callback();
-                }], log, callback);
+        }, log, callback);
     }
 
     healthcheck(location, callback, flightCheckOnStartUp) {
         const azureResp = {};
-        const healthCheckAction = flightCheckOnStartUp ?
-            'createContainerIfNotExists' : 'doesContainerExist';
-        this._errorWrapper('checkAzureHealth', healthCheckAction,
-            [this._azureContainerName, err => {
-                /* eslint-disable no-param-reassign */
-                if (err) {
-                    azureResp[location] = { error: err.message,
-                        external: true };
-                    return callback(null, azureResp);
+        this._errorWrapper('healthcheck', 'checkAzureHealth', async client => {
+            try {
+                if (flightCheckOnStartUp) {
+                    await client.createIfNotExists();
+                } else {
+                    await client.exists();
                 }
                 azureResp[location] = {
-                    message:
-                    'Congrats! You can access the Azure storage account',
+                    message: 'Congrats! You can access the Azure storage account',
                 };
-                return callback(null, azureResp);
-            }], null, callback);
+            } catch (err) {
+                azureResp[location] = {
+                    error: err.message,
+                    external: true,
+                };
+            }
+            return azureResp;
+        }, null, callback);
     }
 
     uploadPart(request, streamingV4Params, partStream, size, key, uploadId,
@@ -321,9 +409,7 @@ class AzureClient {
     completeMPU(jsonList, mdInfo, key, uploadId, bucket, metaHeaders,
         contentSettings, tagging, log, callback) {
         const azureKey = this._createAzureKey(bucket, key, this._bucketMatch);
-        const commitList = {
-            UncommittedBlocks: jsonList.uncommittedBlocks || [],
-        };
+        const commitList = jsonList.uncommittedBlocks || [];
         let filteredPartsObj;
         if (!jsonList.uncommittedBlocks) {
             const { storedParts, mpuOverviewKey, splitter } = mdInfo;
@@ -336,60 +422,56 @@ class AzureClient {
                 // part.locations is always array of 1, which contains data info
                 const subPartIds =
                     azureMpuUtils.getSubPartIds(part.locations[0], uploadId);
-                commitList.UncommittedBlocks.push(...subPartIds);
+                commitList.push(...subPartIds);
             });
         }
         const options = {
-            contentSettings,
+            blobHTTPHeaders: this._getAzureContentSettingsHeaders(contentSettings || {}),
             metadata: translateAzureMetaHeaders(metaHeaders || {}, tagging),
         };
-        return this._errorWrapper('completeMPU', 'commitBlocks',
-            [this._azureContainerName, azureKey, commitList, options,
-                err => {
-                    if (err) {
-                        logHelper(log, 'error', 'err completing MPU on Azure ' +
-                        'datastore', err, this._dataStoreName);
-                        return callback(errors.ServiceUnavailable
-                            .customizeDescription('Error returned from ' +
-                        `Azure: ${err.message}`));
-                    }
-                    const completeObjData = {
+        return this._errorWrapper('completeMPU', 'commitBlocks', async client => {
+            try {
+                await client.getBlockBlobClient(azureKey).commitBlockList(commitList, options);
+                return {
                     key: azureKey,
                     filteredPartsObj,
                 };
-                    return callback(null, completeObjData);
-                }], log, callback);
+            } catch (err) {
+                logHelper(log, 'error', 'err completing MPU on Azure datastore',
+                    err, this._dataStoreName);
+                throw errors.ServiceUnavailable.customizeDescription(
+                    `Error returned from Azure: ${err.message}`);
+            }
+        }, log, callback);
     }
 
     objectPutTagging(key, bucket, objectMD, log, callback) {
         const azureKey = this._createAzureKey(bucket, key, this._bucketMatch);
         const azureMD = this._getMetaHeaders(objectMD);
         azureMD.tags = JSON.stringify(objectMD.tags);
-        this._errorWrapper('objectPutTagging', 'setBlobMetadata',
-            [this._azureContainerName, azureKey, azureMD,
-                err => {
-                    if (err) {
-                        logHelper(log, 'error', 'err putting object tags to ' +
-                        'Azure backend', err, this._dataStoreName);
-                        return callback(errors.ServiceUnavailable);
+        this._errorWrapper('objectPutTagging', 'setBlobMetadata', async client => {
+            try {
+                await client.getBlockBlobClient(azureKey).setMetadata(azureMD);
+            } catch (err) {
+                logHelper(log, 'error', 'err putting object tags to Azure backend',
+                    err, this._dataStoreName);
+                throw errors.ServiceUnavailable;
             }
-                    return callback();
-                }], log, callback);
+        }, log, callback);
     }
 
     objectDeleteTagging(key, bucketName, objectMD, log, callback) {
         const azureKey = this._createAzureKey(bucketName, key, this._bucketMatch);
         const azureMD = this._getMetaHeaders(objectMD);
-        this._errorWrapper('objectDeleteTagging', 'setBlobMetadata',
-            [this._azureContainerName, azureKey, azureMD,
-                err => {
-                    if (err) {
-                        logHelper(log, 'error', 'err putting object tags to ' +
-                        'Azure backend', err, this._dataStoreName);
-                        return callback(errors.ServiceUnavailable);
+        this._errorWrapper('objectDeleteTagging', 'setBlobMetadata', async client => {
+            try {
+                await client.getBlockBlobClient(azureKey).setMetadata(azureMD);
+            } catch (err) {
+                logHelper(log, 'error', 'err putting object tags to Azure backend',
+                    err, this._dataStoreName);
+                throw errors.ServiceUnavailable;
             }
-                    return callback();
-                }], log, callback);
+        }, log, callback);
     }
 
     copyObject(request, destLocationConstraintName, sourceKey,
@@ -406,54 +488,50 @@ class AzureClient {
 
         let options;
         if (storeMetadataParams.metaHeaders) {
-            options = { metadata:
-            translateAzureMetaHeaders(storeMetadataParams.metaHeaders) };
+            options = {
+                metadata: translateAzureMetaHeaders(storeMetadataParams.metaHeaders),
+            };
         }
 
-        this._errorWrapper('copyObject', 'startCopyBlob',
-            [`${this._azureStorageEndpoint}` +
-                `${sourceContainerName}/${sourceKey}`,
-            this._azureContainerName, destAzureKey, options,
-            (err, res) => {
-                if (err) {
-                    if (err.code === 'CannotVerifyCopySource') {
-                        logHelper(log, 'error', 'Unable to access ' +
-                        `${sourceContainerName} Azure Container`, err,
-                        this._dataStoreName);
-                        return callback(errors.AccessDenied
-                            .customizeDescription('Error: Unable to access ' +
-                        `${sourceContainerName} Azure Container`),
+        // TODO: should we use syncCopyBlob() instead? or use poller.pollUntilDone() to wait until complete?
+        this._errorWrapper('copyObject', 'startCopyBlob', async client => {
+            let res;
+            try {
+                const poller = await client.getBlockBlobClient(destAzureKey).beginCopyFromURL(
+                    `${this._azureStorageEndpoint}${sourceContainerName}/${sourceKey}`,
+                    options,
                 );
+
+                res = poller.getOperationState().result;
+                if (res.copyProgress !== 'pending') {
+                    return destAzureKey;
                 }
-                    logHelper(log, 'error', 'error from data backend on ' +
-                    'copyObject', err, this._dataStoreName);
-                    return callback(errors.ServiceUnavailable
-                        .customizeDescription('Error returned from ' +
-                    `AWS: ${err.message}`),
-                    );
-                }
-                if (res.copy.status === 'pending') {
-                    logHelper(log, 'error', 'Azure copy status is pending',
+            } catch (err) {
+                if (err.code === 'CannotVerifyCopySource') { // TOOD: may use a constant (or type) from SDK ??
+                    logHelper(log, 'error',
+                        `Unable to access ${sourceContainerName} Azure Container`,
                         err, this._dataStoreName);
-                    const copyId = res.copy.id;
-                    this._client.abortCopyBlob(this._azureContainerName,
-                        destAzureKey, copyId, err => {
-                            if (err) {
-                                logHelper(log, 'error', 'error from data backend ' +
-                            'on abortCopyBlob', err, this._dataStoreName);
-                                return callback(errors.ServiceUnavailable
-                                    .customizeDescription('Error returned from ' +
-                            `AWS on abortCopyBlob: ${err.message}`),
-                                );
+                    throw errors.AccessDenied.customizeDescription(
+                        `Error: Unable to access ${sourceContainerName} Azure Container`);
                 }
-                            return callback(errors.InvalidObjectState
-                                .customizeDescription('Error: Azure copy status was ' +
-                        'pending. It has been aborted successfully'),
-                            );
-                        });
+                logHelper(log, 'error', 'error from data backend on copyObject',
+                    err, this._dataStoreName);
+                throw errors.ServiceUnavailable.customizeDescription(
+                    `Error returned from AWS: ${err.message}`);
             }
-                return callback(null, destAzureKey);
-            }], log, callback);
+
+            logHelper(log, 'error', 'Azure copy status is pending', {}, this._dataStoreName);
+            try {
+                await client.getBlockBlobClient(destAzureKey).abortCopyFromURL(res.copyId);
+            } catch (err) {
+                logHelper(log, 'error', 'error from data backend on abortCopyBlob',
+                    err, this._dataStoreName);
+                throw errors.ServiceUnavailable.customizeDescription(
+                    `Error returned from AWS on abortCopyBlob: ${err.message}`);
+            }
+            throw errors.InvalidObjectState.customizeDescription(
+                'Error: Azure copy status was pending. It has been aborted successfully');
+        }, log, callback);
     }
 }
 

lib/storage/data/vitastor/VitastorBackend.ts

@@ -0,0 +1,696 @@
+// Zenko CloudServer Vitastor data storage backend adapter
+// Copyright (c) Vitaliy Filippov, 2019+
+// License: VNPL-1.1 (see README.md for details)
+
+const stream = require('stream');
+
+const vitastor = require('vitastor');
+
+const VOLUME_MAGIC = 'VstS3Vol';
+const OBJECT_MAGIC = 'VstS3Obj';
+const FLAG_DELETED = 2n;
+
+type Volume = {
+    id: number,
+    partial_sectors: {
+        [key: string]: {
+            buffer: Buffer,
+            refs: number,
+        },
+    },
+    header: {
+        location: string,
+        bucket: string,
+        max_size: number,
+        create_ts: number,
+        used_ts: number,
+        size: number,
+        objects: number,
+        removed_objects: number,
+        object_bytes: number,
+        removed_bytes: number,
+    },
+};
+
+type ObjectHeader = {
+    size: number,
+    key: string,
+    part_num?: number,
+};
+
+class VitastorBackend
+{
+    locationName: string;
+    config: {
+        pool_id: number,
+        metadata_image: string,
+        metadata_pool_id: number,
+        metadata_inode_num: number,
+        size_buckets: number[],
+        size_bucket_mul: number,
+        id_batch_size: number,
+        sector_size: number,
+        write_chunk_size: number,
+        read_chunk_size: number,
+        pack_objects: boolean,
+        // and also other parameters for vitastor itself
+    };
+    next_id: number;
+    alloc_id: number;
+    opened: boolean;
+    on_open: ((...args: any[]) => void)[] | null;
+    open_error: Error | null;
+    cli: any;
+    kv: any;
+    volumes: {
+        [bucket: string]: {
+            [max_size: string]: Volume,
+        },
+    };
+    volumes_by_id: {
+        [id: string]: Volume,
+    };
+    volume_delete_stats: {
+        [id: string]: {
+            count: number,
+            bytes: number,
+        },
+    };
+
+    constructor(locationName, config)
+    {
+        this.locationName = locationName;
+        this.config = config;
+        // validate config
+        this.config.pool_id = Number(this.config.pool_id) || 0;
+        if (!this.config.pool_id)
+            throw new Error('pool_id is required for Vitastor');
+        if (!this.config.metadata_image && (!this.config.metadata_pool_id || !this.config.metadata_inode_num))
+            throw new Error('metadata_image or metadata_inode is required for Vitastor');
+        if (!this.config.size_buckets || !this.config.size_buckets.length)
+            this.config.size_buckets = [ 32*1024, 128*1024, 512*1024, 2*1024, 8*1024 ];
+        this.config.size_bucket_mul = Number(this.config.size_bucket_mul) || 2;
+        this.config.id_batch_size = Number(this.config.id_batch_size) || 100;
+        this.config.sector_size = Number(this.config.sector_size) || 0;
+        if (this.config.sector_size < 4096)
+            this.config.sector_size = 4096;
+        this.config.write_chunk_size = Number(this.config.write_chunk_size) || 0;
+        if (this.config.write_chunk_size < this.config.sector_size)
+            this.config.write_chunk_size = 4*1024*1024; // 4 MB
+        this.config.read_chunk_size = Number(this.config.read_chunk_size) || 0;
+        if (this.config.read_chunk_size < this.config.sector_size)
+            this.config.read_chunk_size = 4*1024*1024; // 4 MB
+        this.config.pack_objects = !!this.config.pack_objects;
+        // state
+        this.next_id = 1;
+        this.alloc_id = 0;
+        this.opened = false;
+        this.on_open = null;
+        this.open_error = null;
+        this.cli = new vitastor.Client(config);
+        this.kv = new vitastor.KV(this.cli);
+        // we group objects into volumes by bucket and size
+        this.volumes = {};
+        this.volumes_by_id = {};
+        this.volume_delete_stats = {};
+    }
+
+    async _makeVolumeId()
+    {
+        if (this.next_id <= this.alloc_id)
+        {
+            return this.next_id++;
+        }
+        const id_key = 'id'+this.config.pool_id;
+        const [ err, prev ] = await new Promise<[ any, string ]>(ok => this.kv.get(id_key, (err, value) => ok([ err, value ])));
+        if (err && err != vitastor.ENOENT)
+        {
+            throw new Error(err);
+        }
+        const new_id = (parseInt(prev) || 0) + 1;
+        this.next_id = new_id;
+        this.alloc_id = this.next_id + this.config.id_batch_size - 1;
+        await new Promise((ok, no) => this.kv.set(id_key, this.alloc_id, err => (err ? no(new Error(err)) : ok(null)), cas_old => cas_old === prev));
+        return this.next_id;
+    }
+
+    async _getVolume(bucketName, size)
+    {
+        if (!this.opened)
+        {
+            if (this.on_open)
+            {
+                await new Promise(ok => this.on_open!.push(ok));
+            }
+            else
+            {
+                this.on_open = [];
+                if (this.config.metadata_image)
+                {
+                    const img = new vitastor.Image(this.cli, this.config.metadata_image);
+                    const info = await new Promise<{ pool_id: number, inode_num: number }>(ok => img.get_info(ok));
+                    this.config.metadata_pool_id = info.pool_id;
+                    this.config.metadata_inode_num = info.inode_num;
+                }
+                const kv_config = {};
+                for (const key in this.config)
+                {
+                    if (key.substr(0, 3) === 'kv_')
+                        kv_config[key] = this.config[key];
+                }
+                this.open_error = await new Promise(ok => this.kv.open(
+                    this.config.metadata_pool_id, this.config.metadata_inode_num,
+                    kv_config, err => ok(err ? new Error(err) : null)
+                ));
+                this.opened = true;
+                this.on_open.map(cb => setImmediate(cb));
+                this.on_open = null;
+            }
+        }
+        if (this.open_error)
+        {
+            throw this.open_error;
+        }
+        let i;
+        for (i = 0; i < this.config.size_buckets.length && size >= this.config.size_buckets[i]; i++) {}
+        let s;
+        if (i < this.config.size_buckets.length)
+            s = this.config.size_buckets[i];
+        else if (this.config.size_bucket_mul > 1)
+        {
+            while (size >= s)
+                s = Math.floor(this.config.size_bucket_mul * s);
+        }
+        if (!this.volumes[bucketName])
+        {
+            this.volumes[bucketName] = {};
+        }
+        if (this.volumes[bucketName][s])
+        {
+            return this.volumes[bucketName][s];
+        }
+        const new_id = await this._makeVolumeId();
+        const new_vol = this.volumes[bucketName][s] = {
+            id: new_id,
+            // FIXME: partial_sectors should be written with CAS because otherwise we may lose quick deletes
+            partial_sectors: {},
+            header: {
+                location: this.locationName,
+                bucket: bucketName,
+                max_size: s,
+                create_ts: Date.now(),
+                used_ts: Date.now(),
+                size: this.config.sector_size, // initial position is right after header
+                objects: 0,
+                removed_objects: 0,
+                object_bytes: 0,
+                removed_bytes: 0,
+            },
+        };
+        this.volumes_by_id[new_id] = new_vol;
+        const header_text = JSON.stringify(this.volumes[bucketName][s].header);
+        const buf = Buffer.alloc(this.config.sector_size);
+        buf.write(VOLUME_MAGIC + header_text, 0);
+        await new Promise((ok, no) => this.cli.write(
+            this.config.pool_id, new_id, 0, buf, err => (err ? no(new Error(err)) : ok(null))
+        ));
+        await new Promise((ok, no) => this.kv.set(
+            'vol_'+this.config.pool_id+'_'+new_id, header_text, err => (err ? no(new Error(err)) : ok(null)), cas_old => !cas_old
+        ));
+        return new_vol;
+    }
+
+    toObjectGetInfo(objectKey, bucketName, storageLocation)
+    {
+        return null;
+    }
+
+    _bufferStart(vol, cur_pos, cur_size, cur_chunks, sector_refs)
+    {
+        if ((cur_pos % this.config.sector_size) ||
+            Math.floor((cur_pos + cur_size) / this.config.sector_size) == Math.floor(cur_pos / this.config.sector_size))
+        {
+            const sect_pos = Math.floor(cur_pos / this.config.sector_size) * this.config.sector_size;
+            const sect = vol.partial_sectors[sect_pos]
+                ? vol.partial_sectors[sect_pos].buffer
+                : Buffer.alloc(this.config.sector_size);
+            if (this.config.pack_objects)
+            {
+                // Save only if <pack_objects>
+                if (!vol.partial_sectors[sect_pos])
+                    vol.partial_sectors[sect_pos] = { buffer: sect, refs: 0 };
+                vol.partial_sectors[sect_pos].refs++;
+                sector_refs.push(sect_pos);
+            }
+            let off = cur_pos % this.config.sector_size;
+            let i = 0;
+            for (; i < cur_chunks.length; i++)
+            {
+                let copy_len = this.config.sector_size - off;
+                copy_len = copy_len > cur_chunks[i].length ? cur_chunks[i].length : copy_len;
+                cur_chunks[i].copy(sect, off, 0, copy_len);
+                off += copy_len;
+                if (copy_len < cur_chunks[i].length)
+                {
+                    cur_chunks[i] = cur_chunks[i].slice(copy_len);
+                    cur_size -= copy_len;
+                    break;
+                }
+                else
+                    cur_size -= cur_chunks[i].length;
+            }
+            cur_chunks.splice(0, i, sect);
+            cur_size += this.config.sector_size;
+            cur_pos = sect_pos;
+        }
+        return [ cur_pos, cur_size ];
+    }
+
+    _bufferEnd(vol, cur_pos, cur_size, cur_chunks, sector_refs, write_all)
+    {
+        const write_pos = cur_pos;
+        const write_chunks = cur_chunks;
+        let write_size = cur_size;
+        cur_chunks = [];
+        cur_pos += cur_size;
+        cur_size = 0;
+        let remain = (cur_pos % this.config.sector_size);
+        if (remain > 0)
+        {
+            cur_pos -= remain;
+            let last_sect = null;
+            if (write_all)
+            {
+                last_sect = vol.partial_sectors[cur_pos]
+                    ? vol.partial_sectors[cur_pos].buffer
+                    : Buffer.alloc(this.config.sector_size);
+                if (this.config.pack_objects)
+                {
+                    // Save only if <pack_objects>
+                    if (!vol.partial_sectors[cur_pos])
+                        vol.partial_sectors[cur_pos] = { buffer: last_sect, refs: 0 };
+                    vol.partial_sectors[cur_pos].refs++;
+                    sector_refs.push(cur_pos);
+                }
+            }
+            write_size -= remain;
+            if (write_size < 0)
+                write_size = 0;
+            for (let i = write_chunks.length-1; i >= 0 && remain > 0; i--)
+            {
+                if (write_chunks[i].length <= remain)
+                {
+                    remain -= write_chunks[i].length;
+                    if (write_all)
+                        write_chunks[i].copy(last_sect, remain);
+                    else
+                        cur_chunks.unshift(write_chunks[i]);
+                    write_chunks.pop();
+                }
+                else
+                {
+                    if (write_all)
+                        write_chunks[i].copy(last_sect, 0, write_chunks[i].length - remain);
+                    else
+                        cur_chunks.unshift(write_chunks[i].slice(write_chunks[i].length - remain));
+                    write_chunks[i] = write_chunks[i].slice(0, write_chunks[i].length - remain);
+                    remain = 0;
+                    i++;
+                }
+            }
+            if (write_all)
+            {
+                write_chunks.push(last_sect);
+                write_size += this.config.sector_size;
+            }
+        }
+        for (const chunk of cur_chunks)
+        {
+            cur_size += chunk.length;
+        }
+        return [ write_pos, write_chunks, write_size, cur_pos, cur_size, cur_chunks ];
+    }
+
+    /**
+     * reqUids: string, // request-ids for log, usually joined by ':'
+     * keyContext: {
+     *     // a lot of shit, basically all metadata
+     *     bucketName,
+     *     objectKey,
+     *     owner?,
+     *     namespace?,
+     *     partNumber?,
+     *     uploadId?,
+     *     metaHeaders?,
+     *     isDeleteMarker?,
+     *     tagging?,
+     *     contentType?,
+     *     cacheControl?,
+     *     contentDisposition?,
+     *     contentEncoding?,
+     * },
+     * callback: (error, objectGetInfo: any) => void,
+     */
+    put(stream, size, keyContext, reqUids, callback)
+    {
+        callback = once(callback);
+        this._getVolume(keyContext.bucketName, size)
+            .then(vol => this._put(vol, stream, size, keyContext, reqUids, callback))
+            .catch(callback);
+    }
+
+    _put(vol, stream, size, keyContext, reqUids, callback)
+    {
+        const object_header: ObjectHeader = {
+            size,
+            key: keyContext.objectKey,
+        };
+        if (keyContext.partNumber)
+        {
+            object_header.part_num = keyContext.partNumber;
+        }
+        // header is: <8 bytes magic> <8 bytes flags> <8 bytes json length> <json>
+        const hdr_begin_buf = Buffer.alloc(24);
+        const hdr_json_buf = Buffer.from(JSON.stringify(object_header), 'utf-8');
+        hdr_begin_buf.write(OBJECT_MAGIC);
+        hdr_begin_buf.writeBigInt64LE(BigInt(hdr_json_buf.length), 16);
+        const object_header_buf = Buffer.concat([ hdr_begin_buf, hdr_json_buf ]);
+        const object_pos = vol.header.size;
+        const object_get_info = { volume: vol.id, offset: object_pos, hdrlen: object_header_buf.length, size };
+        let cur_pos = object_pos;
+        let cur_chunks = [ object_header_buf ];
+        let cur_size = object_header_buf.length;
+        let err: Error|null = null;
+        let waiting = 1; // 1 for end or error, 1 for each write request
+        vol.header.size += object_header_buf.length + size;
+        if (!this.config.pack_objects && (vol.header.size % this.config.sector_size))
+        {
+            vol.header.size += this.config.sector_size - (vol.header.size % this.config.sector_size);
+        }
+        const writeChunk = (last) =>
+        {
+            const sector_refs = [];
+            // Handle partial beginning
+            [ cur_pos, cur_size ] = this._bufferStart(vol, cur_pos, cur_size, cur_chunks, sector_refs);
+            // Handle partial end
+            let write_pos, write_chunks, write_size;
+            [ write_pos, write_chunks, write_size, cur_pos, cur_size, cur_chunks ] = this._bufferEnd(vol, cur_pos, cur_size, cur_chunks, sector_refs, last);
+            waiting++;
+            // FIXME: pool_id: maybe it should be stored in volume metadata to allow to migrate volumes?
+            this.cli.write(this.config.pool_id, vol.id, write_pos, write_chunks, (res) =>
+            {
+                for (const sect of sector_refs)
+                {
+                    vol.partial_sectors[sect].refs--;
+                    if (!vol.partial_sectors[sect].refs &&
+                        vol.header.size >= sect+this.config.sector_size)
+                    {
+                        // Forget partial data when it's not needed anymore
+                        delete(vol.partial_sectors[sect]);
+                    }
+                }
+                waiting--;
+                if (res)
+                {
+                    err = new Error(res);
+                    waiting--;
+                }
+                if (!waiting)
+                {
+                    callback(err, err ? null : object_get_info);
+                }
+            });
+        };
+        // Stream data
+        stream.on('error', (e) =>
+        {
+            err = e;
+            waiting--;
+            if (!waiting)
+            {
+                callback(err, null);
+            }
+        });
+        stream.on('end', () =>
+        {
+            if (err)
+            {
+                return;
+            }
+            waiting--;
+            if (cur_size)
+            {
+                // write last chunk
+                writeChunk(true);
+            }
+            if (!waiting)
+            {
+                callback(null, object_get_info);
+            }
+        });
+        stream.on('data', (chunk) =>
+        {
+            if (err)
+            {
+                return;
+            }
+            cur_chunks.push(chunk);
+            cur_size += chunk.length;
+            if (cur_size >= this.config.write_chunk_size)
+            {
+                // got a complete chunk, write it out
+                writeChunk(false);
+            }
+        });
+    }
+
+    /**
+     * objectGetInfo: {
+     *   key: { volume, offset, hdrlen, size }, // from put
+     *   size,
+     *   start,
+     *   dataStoreName,
+     *   dataStoreETag,
+     *   range,
+     *   response: ServerResponse,
+     * },
+     * range?: [ start, end ], // like in HTTP - first byte index, last byte index
+     * callback: (error, readStream) => void,
+     */
+    get(objectGetInfo, range, reqUids, callback)
+    {
+        if (!(objectGetInfo instanceof Object) || !objectGetInfo.key ||
+            !(objectGetInfo.key instanceof Object) || !objectGetInfo.key.volume ||
+            !objectGetInfo.key.offset || !objectGetInfo.key.hdrlen || !objectGetInfo.key.size)
+        {
+            throw new Error('objectGetInfo must be { key: { volume, offset, hdrlen, size } }, but is '+JSON.stringify(objectGetInfo));
+        }
+        const [ start, end ] = range || [];
+        if (start < 0 || end < 0 || end != null && start != null && end < start || start >= objectGetInfo.key.size)
+        {
+            throw new Error('Invalid range: '+start+'-'+end);
+        }
+        let offset = objectGetInfo.key.offset + objectGetInfo.key.hdrlen + (start || 0);
+        let len = objectGetInfo.key.size - (start || 0);
+        if (end)
+        {
+            const len2 = end - (start || 0) + 1;
+            if (len2 < len)
+                len = len2;
+        }
+        callback(null, new VitastorReadStream(this.cli, objectGetInfo.key.volume, offset, len, this.config));
+    }
+
+    /**
+     * objectGetInfo: {
+     *   key: { volume, offset, hdrlen, size }, // from put
+     *   size,
+     *   start,
+     *   dataStoreName,
+     *   dataStoreETag,
+     *   range,
+     *   response: ServerResponse,
+     * },
+     * callback: (error) => void,
+     */
+    delete(objectGetInfo, reqUids, callback)
+    {
+        callback = once(callback);
+        this._delete(objectGetInfo, reqUids)
+            .then(callback)
+            .catch(callback);
+    }
+
+    async _delete(objectGetInfo, reqUids)
+    {
+        if (!(objectGetInfo instanceof Object) || !objectGetInfo.key ||
+            !(objectGetInfo.key instanceof Object) || !objectGetInfo.key.volume ||
+            !objectGetInfo.key.offset || !objectGetInfo.key.hdrlen || !objectGetInfo.key.size)
+        {
+            throw new Error('objectGetInfo must be { key: { volume, offset, hdrlen, size } }, but is '+JSON.stringify(objectGetInfo));
+        }
+        const in_sect_pos = (objectGetInfo.key.offset % this.config.sector_size);
+        const sect_pos = objectGetInfo.key.offset - in_sect_pos;
+        const vol = this.volumes_by_id[objectGetInfo.key.volume];
+        if (vol && vol.partial_sectors[sect_pos])
+        {
+            // The sector may still be written to in corner cases
+            const sect = vol.partial_sectors[sect_pos];
+            const flags = sect.buffer.readBigInt64LE(in_sect_pos + 8);
+            if (!(flags & FLAG_DELETED))
+            {
+                const del_stat = this.volume_delete_stats[vol.id] = (this.volume_delete_stats[vol.id] || { count: 0, bytes: 0 });
+                del_stat.count++;
+                del_stat.bytes += objectGetInfo.key.size;
+                sect.buffer.writeBigInt64LE(flags | FLAG_DELETED, in_sect_pos + 8);
+                sect.refs++;
+                const err = await new Promise<any>(ok => this.cli.write(this.config.pool_id, objectGetInfo.key.volume, sect_pos, sect.buffer, ok));
+                sect.refs--;
+                if (err)
+                {
+                    sect.buffer.writeBigInt64LE(0n, in_sect_pos + 8);
+                    throw new Error(err);
+                }
+            }
+        }
+        else
+        {
+            // RMW with CAS
+            const [ err, buf, version ] = await new Promise<[ any, Buffer, bigint ]>(ok => this.cli.read(
+                this.config.pool_id, objectGetInfo.key.volume, sect_pos, this.config.sector_size,
+                (err, buf, version) => ok([ err, buf, version ])
+            ));
+            if (err)
+            {
+                throw new Error(err);
+            }
+            // FIXME What if JSON crosses sector boundary? Prevent it if we want to pack objects
+            const magic = buf.slice(in_sect_pos, in_sect_pos+8).toString();
+            const flags = buf.readBigInt64LE(in_sect_pos+8);
+            const json_len = Number(buf.readBigInt64LE(in_sect_pos+16));
+            let json_hdr;
+            if (in_sect_pos+24+json_len <= buf.length)
+            {
+                try
+                {
+                    json_hdr = JSON.parse(buf.slice(in_sect_pos+24, in_sect_pos+24+json_len).toString());
+                }
+                catch (e)
+                {
+                }
+            }
+            if (magic !== OBJECT_MAGIC || !json_hdr || json_hdr.size !== objectGetInfo.key.size)
+            {
+                throw new Error(
+                    'header of object with size '+objectGetInfo.key.size+
+                    ' bytes not found in volume '+objectGetInfo.key.volume+' at '+objectGetInfo.key.offset
+                );
+            }
+            else if (!(flags & FLAG_DELETED))
+            {
+                buf.writeBigInt64LE(flags | FLAG_DELETED, in_sect_pos + 8);
+                const err = await new Promise<any>(ok => this.cli.write(this.config.pool_id, objectGetInfo.key.volume, sect_pos, buf, { version: version+1n }, ok));
+                if (err == vitastor.EINTR)
+                {
+                    // Retry
+                    await this._delete(objectGetInfo, reqUids);
+                }
+                else if (err)
+                {
+                    throw new Error(err);
+                }
+                else
+                {
+                    // FIXME: Write deletion statistics to volumes
+                    // FIXME: Implement defragmentation
+                    const del_stat = this.volume_delete_stats[objectGetInfo.key.volume] = (this.volume_delete_stats[objectGetInfo.key.volume] || { count: 0, bytes: 0 });
+                    del_stat.count++;
+                    del_stat.bytes += objectGetInfo.key.size;
+                }
+            }
+        }
+    }
+
+    /**
+     * config: full zenko server config,
+     * callback: (error, stats) => void, // stats is the returned statistics in arbitrary format
+     */
+    getDiskUsage(config, reqUids, callback)
+    {
+        // FIXME: Iterate all volumes and return its sizes and deletion statistics, or maybe just sizes
+        callback(null, {});
+    }
+}
+
+class VitastorReadStream extends stream.Readable
+{
+    constructor(cli, volume_id, offset, len, config, options = undefined)
+    {
+        super(options);
+        this.cli = cli;
+        this.volume_id = volume_id;
+        this.offset = offset;
+        this.end = offset + len;
+        this.pos = offset;
+        this.config = config;
+        this._reading = false;
+    }
+
+    _read(n)
+    {
+        if (this._reading)
+        {
+            return;
+        }
+        // FIXME: Validate object header
+        const chunk_size = n && this.config.read_chunk_size < n ? n : this.config.read_chunk_size;
+        const read_offset = this.pos;
+        const round_offset = read_offset - (read_offset % this.config.sector_size);
+        let read_end = this.end <= read_offset+chunk_size ? this.end : read_offset+chunk_size;
+        const round_end = (read_end % this.config.sector_size)
+            ? read_end + this.config.sector_size - (read_end % this.config.sector_size)
+            : read_end;
+        if (round_end <= this.end)
+            read_end = round_end;
+        this.pos = read_end;
+        if (read_end <= read_offset)
+        {
+            // EOF
+            this.push(null);
+            return;
+        }
+        this._reading = true;
+        this.cli.read(this.config.pool_id, this.volume_id, round_offset, round_end-round_offset, (err, buf, version) =>
+        {
+            this._reading = false;
+            if (err)
+            {
+                this.destroy(new Error(err));
+                return;
+            }
+            if (read_offset != round_offset || round_end != read_end)
+            {
+                buf = buf.subarray(read_offset-round_offset, buf.length-(round_end-read_end));
+            }
+            if (this.push(buf))
+            {
+                this._read(n);
+            }
+        });
+    }
+}
+
+function once(callback)
+{
+    let called = false;
+    return function()
+    {
+        if (!called)
+        {
+            called = true;
+            callback.apply(null, arguments);
+        }
+    };
+}
+
+module.exports = VitastorBackend;

lib/storage/metadata/MetadataWrapper.js

@@ -51,6 +51,36 @@ function _parseListEntries(entries) {
     });
 }
 
+/** _parseLifecycleListEntries - parse the values returned in a lifeycle listing by metadata
+ * @param {object[]} entries - Version or Content entries in a metadata listing
+ * @param {string} entries[].key - metadata key
+ * @param {string} entries[].value - stringified object metadata
+ * @return {object} - mapped array with parsed value or JSON parsing err
+ */
+function _parseLifecycleListEntries(entries) {
+    return entries.map(entry => {
+        const tmp = JSON.parse(entry.value);
+        return {
+            key: entry.key,
+            value: {
+                Size: tmp['content-length'],
+                ETag: tmp['content-md5'],
+                VersionId: tmp.versionId,
+                IsNull: tmp.isNull,
+                LastModified: tmp['last-modified'],
+                Owner: {
+                    DisplayName: tmp['owner-display-name'],
+                    ID: tmp['owner-id'],
+                },
+                StorageClass: tmp['x-amz-storage-class'],
+                tags: tmp.tags,
+                staleDate: tmp.staleDate,
+                dataStoreName: tmp.dataStoreName,
+            },
+        };
+    });
+}
+
 /** parseListEntries - parse the values returned in a listing by metadata
  * @param {object[]} entries - Version or Content entries in a metadata listing
  * @param {string} entries[].key - metadata key
@@ -147,6 +177,42 @@ class MetadataWrapper {
         });
     }
 
+    updateBucketCapabilities(bucketName, bucketMD, capabilityName, capacityField, capability, log, cb) {
+        log.debug('updating bucket capabilities in metadata');
+        // When concurrency update is not supported, we update the whole bucket metadata
+        if (!this.client.putBucketAttributesCapabilities) {
+            return this.updateBucket(bucketName, bucketMD, log, cb);
+        }
+        return this.client.putBucketAttributesCapabilities(bucketName, capabilityName, capacityField, capability,
+            log, err => {
+                if (err) {
+                    log.debug('error from metadata', { implName: this.implName,
+                        error: err });
+                    return cb(err);
+                }
+                log.trace('bucket capabilities updated in metadata');
+                return cb(err);
+            });
+    }
+
+    deleteBucketCapabilities(bucketName, bucketMD, capabilityName, capacityField, log, cb) {
+        log.debug('deleting bucket capabilities in metadata');
+        // When concurrency update is not supported, we update the whole bucket metadata
+        if (!this.client.deleteBucketAttributesCapability) {
+            return this.updateBucket(bucketName, bucketMD, log, cb);
+        }
+        return this.client.deleteBucketAttributesCapability(bucketName, capabilityName, capacityField,
+            log, err => {
+                if (err) {
+                    log.debug('error from metadata', { implName: this.implName,
+                        error: err });
+                    return cb(err);
+                }
+                log.trace('bucket capabilities deleted in metadata');
+                return cb(err);
+            });
+    }
+
     getBucket(bucketName, log, cb) {
         log.debug('getting bucket from metadata');
         this.client.getBucketAttributes(bucketName, log, (err, data) => {
@@ -160,6 +226,19 @@ class MetadataWrapper {
         });
     }
 
+    getBucketQuota(bucketName, log, cb) {
+        log.debug('getting bucket quota from metadata');
+        this.client.getBucketAttributes(bucketName, log, (err, data) => {
+            if (err) {
+                log.debug('error from metadata', { implName: this.implName,
+                    error: err });
+                return cb(err);
+            }
+            const bucketInfo = BucketInfo.fromObj(data);
+            return cb(err, { quota: bucketInfo.getQuota() });
+        });
+    }
+
     deleteBucket(bucketName, log, cb) {
         log.debug('deleting bucket from metadata');
         this.client.deleteBucket(bucketName, log, err => {
@@ -213,6 +292,25 @@ class MetadataWrapper {
             });
     }
 
+    getObjectsMD(bucketName, objNamesWithParams, log, cb) {
+        if (typeof this.client.getObjects !== 'function') {
+            log.debug('backend does not support get object metadata with batching', {
+                implName: this.implName,
+            });
+            return cb(errors.NotImplemented);
+        }
+        log.debug('getting objects from metadata', { objects: objNamesWithParams });
+        return this.client.getObjects(bucketName, objNamesWithParams, log, (err, data) => {
+            if (err) {
+                log.debug('error getting objects from metadata', { implName: this.implName, objects: objNamesWithParams,
+                    err });
+                return cb(err);
+            }
+            log.debug('objects retrieved from metadata', { objects: objNamesWithParams });
+            return cb(err, data);
+        });
+    }
+
     getObjectMD(bucketName, objName, params, log, cb) {
         log.debug('getting object from metadata');
         this.client.getObject(bucketName, objName, params, log, (err, data) => {
@@ -226,7 +324,7 @@ class MetadataWrapper {
         });
     }
 
-    deleteObjectMD(bucketName, objName, params, log, cb) {
+    deleteObjectMD(bucketName, objName, params, log, cb, originOp = 's3:ObjectRemoved:Delete') {
         log.debug('deleting object from metadata');
         this.client.deleteObject(bucketName, objName, params, log, err => {
             if (err) {
@@ -236,7 +334,7 @@ class MetadataWrapper {
             }
             log.debug('object deleted from metadata');
             return cb(err);
-        });
+        }, originOp);
     }
 
     listObject(bucketName, listingParams, log, cb) {
@@ -279,6 +377,29 @@ class MetadataWrapper {
         });
     }
 
+    listLifecycleObject(bucketName, listingParams, log, cb) {
+        log.debug('getting object listing for lifecycle from metadata');
+        this.client.listLifecycleObject(bucketName, listingParams, log, (err, data) => {
+            if (err) {
+                log.error('error from metadata', { implName: this.implName,
+                    err });
+                return cb(err);
+            }
+            log.debug('object listing for lifecycle retrieved from metadata');
+            // eslint-disable-next-line no-param-reassign
+            data.Contents = parseListEntries(data.Contents, _parseLifecycleListEntries);
+            if (data.Contents instanceof Error) {
+                log.error('error parsing metadata listing for lifecycle', {
+                    error: data.Contents,
+                    listingType: listingParams.listingType,
+                    method: 'listLifecycleObject',
+                });
+                return cb(errors.InternalError);
+            }
+            return cb(null, data);
+        });
+    }
+
     listMultipartUploads(bucketName, listingParams, log, cb) {
         this.client.listMultipartUploads(bucketName, listingParams, log,
             (err, data) => {
@@ -427,6 +548,139 @@ class MetadataWrapper {
                 return cb();
             });
     }
+
+    /**
+     * Put bucket indexes
+     *
+     * indexSpec format:
+     * [
+     *      { key:[ { key: "", order: 1 } ... ], name: <id 1>, ... , < backend options> },
+     *      ...
+     *      { key:[ { key: "", order: 1 } ... ], name: <id n>, ... },
+     * ]
+     *
+     *
+     * @param {String} bucketName bucket name
+     * @param {Array<Object>} indexSpecs index specification
+     * @param {Object} log logger
+     * @param {Function} cb callback
+     * @return {undefined}
+     */
+    putBucketIndexes(bucketName, indexSpecs, log, cb) {
+        log.debug('put bucket indexes');
+
+        if (typeof this.client.putBucketIndexes !== 'function') {
+            log.error('error from metadata', {
+                method: 'putBucketIndexes',
+                error: errors.NotImplemented,
+                implName: this.implName,
+            });
+            return cb(errors.NotImplemented);
+        }
+
+        return this.client.putBucketIndexes(bucketName, indexSpecs, log, err => {
+            if (err) {
+                log.debug('error from metadata', {
+                    method: 'putBucketIndexes',
+                    error: err,
+                    implName: this.implName,
+                });
+                return cb(err);
+            }
+            return cb(null);
+        });
+    }
+
+
+    /**
+     * Delete bucket indexes
+     *
+     * indexSpec format:
+     * [
+     *      { key:[ { key: "", order: 1 } ... ], name: <id 1>, ... , < backend options> },
+     *      ...
+     *      { key:[ { key: "", order: 1 } ... ], name: <id n>, ...  },
+     * ]
+     *
+     *
+     * @param {String} bucketName bucket name
+     * @param {Array<Object>} indexSpecs index specification
+     * @param {Object} log logger
+     * @param {Function} cb callback
+     * @return {undefined}
+     */
+    deleteBucketIndexes(bucketName, indexSpecs, log, cb) {
+        log.debug('delete bucket indexes');
+
+        if (typeof this.client.deleteBucketIndexes !== 'function') {
+            log.error('error from metadata', {
+                method: 'deleteBucketIndexes',
+                error: errors.NotImplemented,
+                implName: this.implName,
+            });
+            return cb(errors.NotImplemented);
+        }
+
+        return this.client.deleteBucketIndexes(bucketName, indexSpecs, log, err => {
+            if (err) {
+                log.error('error from metadata', {
+                    method: 'deleteBucketIndexes',
+                    error: err,
+                    implName: this.implName,
+                });
+                return cb(err);
+            }
+            return cb(null);
+        });
+    }
+
+    getBucketIndexes(bucketName, log, cb) {
+        log.debug('get bucket indexes');
+
+        if (typeof this.client.getBucketIndexes !== 'function') {
+            log.debug('error from metadata', {
+                method: 'getBucketIndexes',
+                error: errors.NotImplemented,
+                implName: this.implName,
+            });
+            return cb(errors.NotImplemented);
+        }
+
+        return this.client.getBucketIndexes(bucketName, log, (err, res) => {
+            if (err) {
+                log.debug('error from metadata', {
+                    method: 'getBucketIndexes',
+                    error: err,
+                    implName: this.implName,
+                });
+                return cb(err);
+            }
+            return cb(null, res);
+        });
+    }
+
+    getIndexingJobs(log, cb) {
+        if (typeof this.client.getIndexingJobs !== 'function') {
+            log.debug('error from metadata', {
+                method: 'getIndexingJobs',
+                error: errors.NotImplemented,
+                implName: this.implName,
+            });
+            return cb(errors.NotImplemented);
+        }
+
+        return this.client.getIndexingJobs(log, (err, res) => {
+            if (err) {
+                log.debug('error from metadata', {
+                    method: 'getBucketIndexes',
+                    error: err,
+                    implName: this.implName,
+                });
+                return cb(err);
+            }
+            return cb(null, res);
+        });
+    }
 }
 
 module.exports = MetadataWrapper;

lib/storage/metadata/bucketclient/BucketClientInterface.js

@@ -110,6 +110,17 @@ class BucketClientInterface {
         return null;
     }
 
+    listLifecycleObject(bucketName, params, log, cb) {
+        this.client.listObject(bucketName, log.getSerializedUids(), params,
+            (err, data) => {
+                if (err) {
+                    return cb(err);
+                }
+                return cb(null, JSON.parse(data));
+            });
+        return null;
+    }
+
     listMultipartUploads(bucketName, params, log, cb) {
         this.client.listObject(bucketName, log.getSerializedUids(), params,
             (err, data) => {

lib/storage/metadata/file/BucketFileInterface.js

@@ -325,6 +325,10 @@ class BucketFileInterface {
         return this.internalListObject(bucketName, params, log, cb);
     }
 
+    listLifecycleObject(bucketName, params, log, cb) {
+        return this.internalListObject(bucketName, params, log, cb);
+    }
+
     listMultipartUploads(bucketName, params, log, cb) {
         return this.internalListObject(bucketName, params, log, cb);
     }

lib/storage/metadata/in_memory/metastore.js

@@ -318,6 +318,10 @@ const metastore = {
         });
     },
 
+    listLifecycleObject(bucketName, params, log, cb) {
+        return process.nextTick(cb, errors.NotImplemented);
+    },
+
     listMultipartUploads(bucketName, listingParams, log, cb) {
         process.nextTick(() => {
             metastore.getBucketAttributes(bucketName, log, (err, bucket) => {

lib/storage/metadata/mongoclient/ListRecordStream.js

@@ -108,9 +108,26 @@ class ListRecordStream extends stream.Readable {
             if (value && value.tags) {
                 value.tags = unescape(value.tags);
             }
-            entry = {
-                type: 'put', // updates overwrite the whole metadata,
+            // updates overwrite the whole metadata,
             // so they are considered as puts
+            let type = 'put';
+            // When the object metadata contain the "deleted"
+            // flag, it means that the operation is the update
+            // we perform before the deletion of an object. We
+            // perform the update to keep all the metadata in the
+            // oplog. This update is what will be used by backbeat
+            // as the delete operation so we put the type of operation
+            // for this event to a delete.
+            // Backbeat still receives the actual delete operations
+            // but they are ignored as they don't contain any metadata.
+            // The delete operations are kept in case we want to listen
+            // to delete events comming from special collections other
+            // than "bucket" collections.
+            if (value && value.deleted) {
+                type = 'delete';
+            }
+            entry = {
+                type,
                 key: itemObj.o2._id,
                 // updated value may be either stored directly in 'o'
                 // attribute or in '$set' attribute (supposedly when

lib/storage/metadata/mongoclient/Mongoclient.md

@@ -143,13 +143,13 @@ on top of the delete marker or if the delete marker is deleted.
 ##### Format of master keys
 
 ```
-\x7fM{{key}
+\x7fM{{key}}
 ```
 
 ##### Format of version keys
 
 ```
-\x7fV{{key}\x00{versionId}
+\x7fV{{key}}\x00{{versionId}}
 ```
 
 ##### Sizing considerations
@@ -272,7 +272,61 @@ The mongoclient backend implements a readable key/value stream called
 in Arsenal/lib/algos listing algorithms. Note it does not require any
 LevelDB package.
 
+A versioned object in MongoDB is stored using two types of documents:
+a version document for each version of the object, and a master document
+containing the data of the latest version of the object.
+
+When deleting the latest version of an object, the master document is replaced
+by a temporary placeholder document (marked with `isPHD: true` in its metadata)
+that gets resolved later with the data
+of the new latest version of the object.
+
+Listing a bucket’s objects (not versions) is done in two steps.
+First, all master documents are listed internally
+(these can contain the placeholder documents).
+Next, the placeholder documents are resolved by listing
+the object’s versions and returning the latest one’s data.
+
 #### Generating the UUID
 
 To avoid race conditions we always (try to) generate a new UUID and we
 condition the insertion to the non-existence of the document.
+
+#### Oplog Events
+
+The oplog (operations log) is a special capped collection that keeps a
+rolling record of all operations that modify the data stored in the database.
+
+The oplog is read and used by multiple backbeat extensions such as replication
+and bucket notification to react to database changes.
+
+oplog events have different types such as insertion, update and deletion.
+Within an oplog event we find the changes that occured at object level or
+collection level.
+
+At object level if an object was inserted for example we'll find all of the object's
+data in the event. Update events contain the change that occured to an object, and
+delete events don't contain anything (only the object's _id).
+
+To keep the latest object's metadata before it got deleted in the oplog, we update
+the full object setting a deletion flag before deleting the object. This adds an
+update event containing all of the object's metadata.
+
+Note: updating the object with itself without change creates a no-op event that gets
+ignored.
+
+**Special Cases:**
+When deleting the last version of an object in a versioned bucket, the master object
+goes into a PHD state (temporary non updated state) that gets resolved later. There
+is one case where a master can be non existent, it's where the last version is a
+delete marker. In this case the created PHD master is empty only containing the
+isPHD flag and the versionId.
+
+When versioning is enabled in a bucket, only version events should be processed as
+they already contain all of the information needed.
+
+In non versioned buckets, master object events are the ones to be processed.
+
+In versioning suspended buckets, both master and version events should be processed,
+as the master object itself is considered a null version. No special case is present
+here as the master object is always present.

lib/storage/metadata/mongoclient/readStream.js

@@ -55,15 +55,38 @@ class MongoReadStream extends Readable {
             }
         }
 
+        if (options.lastModified) {
+            query['value.last-modified'] = {};
+
+            if (options.lastModified.lt) {
+                query['value.last-modified'].$lt = options.lastModified.lt;
+            }
+        }
+
+        if (options.dataStoreName) {
+            query['value.dataStoreName'] = {};
+
+            if (options.dataStoreName.ne) {
+                query['value.dataStoreName'].$ne = options.dataStoreName.ne;
+            }
+        }
+
         if (!Object.keys(query._id).length) {
             delete query._id;
         }
 
+        // filtering out objects flagged for deletion
+        query.$or = [
+            { 'value.deleted': { $exists: false } },
+            { 'value.deleted': { $eq: false } },
+        ];
+
         if (searchOptions) {
             Object.assign(query, searchOptions);
         }
 
-        this._cursor = c.find(query).sort({
+        const projection = { 'value.location': 0 };
+        this._cursor = c.find(query, { projection }).sort({
             _id: options.reverse ? -1 : 1,
         });
         if (options.limit && options.limit !== -1) {
@@ -79,15 +102,10 @@ class MongoReadStream extends Readable {
             return;
         }
 
-        this._cursor.next((err, doc) => {
+        this._cursor.next().then(doc => {
             if (this._destroyed) {
                 return;
             }
-            if (err) {
-                this.emit('error', err);
-                return;
-            }
-
             let key = undefined;
             let value = undefined;
 
@@ -111,6 +129,12 @@ class MongoReadStream extends Readable {
                     value,
                 });
             }
+        }).catch(err => {
+            if (this._destroyed) {
+                return;
+            }
+            this.emit('error', err);
+            return;
         });
     }
 
@@ -120,7 +144,7 @@ class MongoReadStream extends Readable {
         }
         this._destroyed = true;
 
-        this._cursor.close(err => {
+        this._cursor.close().catch(err => {
             if (err) {
                 this.emit('error', err);
                 return;

lib/storage/metadata/mongoclient/utils.js

@@ -185,6 +185,48 @@ function formatVersionKey(key, versionId, vFormat) {
     return formatVersionKeyV0(key, versionId);
 }
 
+function indexFormatMongoArrayToObject(mongoIndexArray) {
+    const indexObj = [];
+
+    for (const idx of mongoIndexArray) {
+        const keys = [];
+        let entries = [];
+
+        if (idx.key instanceof Map) {
+            entries = idx.key.entries();
+        } else {
+            entries = Object.entries(idx.key);
+        }
+
+        for (const k of entries) {
+            keys.push({ key: k[0], order: k[1] });
+        }
+
+        indexObj.push({ name: idx.name, keys });
+    }
+
+    return indexObj;
+}
+
+
+function indexFormatObjectToMongoArray(indexObj) {
+    const mongoIndexArray = [];
+
+    for (const idx of indexObj) {
+        const key = new Map();
+
+        for (const k of idx.keys) {
+            key.set(k.key, k.order);
+        }
+
+        // copy all field except keys from idx
+        // eslint-disable-next-line
+        const { keys: _, ...toCopy } = idx;
+        mongoIndexArray.push(Object.assign(toCopy, { name: idx.name, key }));
+    }
+
+    return mongoIndexArray;
+}
 
 module.exports = {
     credPrefix,
@@ -195,4 +237,6 @@ module.exports = {
     translateConditions,
     formatMasterKey,
     formatVersionKey,
+    indexFormatMongoArrayToObject,
+    indexFormatObjectToMongoArray,
 };

lib/storage/utils.js

@@ -10,21 +10,21 @@ function trySetDirSyncFlag(path) {
 
     const GETFLAGS = 2148034049;
     const SETFLAGS = 1074292226;
-    const FS_DIRSYNC_FL = 65536;
+    const FS_DIRSYNC_FL = 65536n;
     const buffer = Buffer.alloc(8, 0);
     const pathFD = fs.openSync(path, 'r');
     const status = ioctl(pathFD, GETFLAGS, buffer);
     assert.strictEqual(status, 0);
-    const currentFlags = buffer.readUIntLE(0, 8);
+    const currentFlags = buffer.readBigInt64LE(0);
     const flags = currentFlags | FS_DIRSYNC_FL;
-    buffer.writeUIntLE(flags, 0, 8);
+    buffer.writeBigInt64LE(flags, 0);
     const status2 = ioctl(pathFD, SETFLAGS, buffer);
     assert.strictEqual(status2, 0);
     fs.closeSync(pathFD);
     const pathFD2 = fs.openSync(path, 'r');
     const confirmBuffer = Buffer.alloc(8, 0);
     ioctl(pathFD2, GETFLAGS, confirmBuffer);
-    assert.strictEqual(confirmBuffer.readUIntLE(0, 8),
+    assert.strictEqual(confirmBuffer.readBigInt64LE(0),
         currentFlags | FS_DIRSYNC_FL, 'FS_DIRSYNC_FL not set');
     fs.closeSync(pathFD2);
 }

lib/versioning/Version.ts

@@ -3,7 +3,7 @@ import { VersioningConstants } from './constants';
 const VID_SEP = VersioningConstants.VersionId.Separator;
 /**
  * Class for manipulating an object version.
- * The format of a version: { isNull, isDeleteMarker, versionId, otherInfo }
+ * The format of a version: { isNull, isNull2, isDeleteMarker, versionId, otherInfo }
  *
  * @note Some of these functions are optimized based on string search
  * prior to a full JSON parse/stringify. (Vinh: 18K op/s are achieved
@@ -13,24 +13,31 @@ const VID_SEP = VersioningConstants.VersionId.Separator;
 export class Version {
     version: {
         isNull?: boolean;
+        isNull2?: boolean;
         isDeleteMarker?: boolean;
         versionId?: string;
         isPHD?: boolean;
+        nullVersionId?: string;
     };
 
     /**
      * Create a new version instantiation from its data object.
      * @param version - the data object to instantiate
      * @param version.isNull - is a null version
+     * @param version.isNull2 - Whether new version is null or not AND has
+     * been put with a Cloudserver handling null keys (i.e. supporting
+     * S3C-7352)
      * @param version.isDeleteMarker - is a delete marker
      * @param version.versionId - the version id
      * @constructor
      */
     constructor(version?: {
         isNull?: boolean;
+        isNull2?: boolean;
         isDeleteMarker?: boolean;
         versionId?: string;
         isPHD?: boolean;
+        nullVersionId?: string;
     }) {
         this.version = version || {};
     }
@@ -83,6 +90,33 @@ export class Version {
         return `{ "isPHD": true, "versionId": "${versionId}" }`;
     }
 
+    /**
+     * Appends a key-value pair to a JSON object represented as a string. It adds
+     * a comma if the object is not empty (i.e., not just '{}'). It assumes the input
+     * string is formatted as a JSON object.
+     *
+     * @param {string} stringifiedObject The JSON object as a string to which the key-value pair will be appended.
+     * @param {string} key The key to append to the JSON object.
+     * @param {string} value The value associated with the key to append to the JSON object.
+     * @returns {string} The updated JSON object as a string with the new key-value pair appended.
+     * @example
+     * _jsonAppend('{"existingKey":"existingValue"}', 'newKey', 'newValue');
+     * // returns '{"existingKey":"existingValue","newKey":"newValue"}'
+     */
+    static _jsonAppend(stringifiedObject: string, key: string, value: string): string {
+        // stringifiedObject value has the format of '{...}'
+        let index = stringifiedObject.length - 2;
+        while (stringifiedObject.charAt(index) === ' ') {
+            index -= 1;
+        }
+        const needComma = stringifiedObject.charAt(index) !== '{';
+        return (
+            `${stringifiedObject.slice(0, stringifiedObject.length - 1)}` +
+            (needComma ? ',' : '') +
+            `"${key}":"${value}"}`
+        );
+    }
+
     /**
      * Put versionId into an object in the (cheap) way of string manipulation,
      * instead of the more expensive alternative parsing and stringification.
@@ -93,14 +127,32 @@ export class Version {
      */
     static appendVersionId(value: string, versionId: string): string {
         // assuming value has the format of '{...}'
-        let index = value.length - 2;
-        while (value.charAt(index--) === ' ');
-        const comma = value.charAt(index + 1) !== '{';
-        return (
-            `${value.slice(0, value.length - 1)}` + // eslint-disable-line
-            (comma ? ',' : '') +
-            `"versionId":"${versionId}"}`
-        );
+        return Version._jsonAppend(value, 'versionId', versionId);
+    }
+
+    /**
+    * Updates or appends a `nullVersionId` property to a JSON-formatted string.
+    * This function first checks if the `nullVersionId` property already exists within the input string.
+    * If it exists, the function updates the `nullVersionId` with the new value provided.
+    * If it does not exist, the function appends a `nullVersionId` property with the provided value.
+    *
+    * @static
+    * @param {string} value - The JSON-formatted string that may already contain a `nullVersionId` property.
+    * @param {string} nullVersionId - The new value for the `nullVersionId` property to be updated or appended.
+    * @returns {string} The updated JSON-formatted string with the new `nullVersionId` value.
+    */
+    static updateOrAppendNullVersionId(value: string, nullVersionId: string): string {
+        // Check if "nullVersionId" already exists in the string
+        const nullVersionIdPattern = /"nullVersionId":"[^"]*"/;
+        const nullVersionIdExists = nullVersionIdPattern.test(value);
+    
+        if (nullVersionIdExists) {
+            // Replace the existing nullVersionId with the new one
+            return value.replace(nullVersionIdPattern, `"nullVersionId":"${nullVersionId}"`);
+        } else {
+            // Append nullVersionId
+            return Version._jsonAppend(value, 'nullVersionId', nullVersionId);
+        }
     }
 
     /**
@@ -121,6 +173,19 @@ export class Version {
         return this.version.isNull ?? false;
     }
 
+    /**
+     * Check if a version is a null version and has
+     * been put with a Cloudserver handling null keys (i.e. supporting
+     * S3C-7352).
+     *
+     * @return - stating if the value is a null version and has
+     * been put with a Cloudserver handling null keys (i.e. supporting
+     * S3C-7352).
+     */
+    isNull2Version(): boolean {
+        return this.version.isNull2 ?? false;
+    }
+
     /**
      * Check if a stringified object is a delete marker.
      *
@@ -190,6 +255,19 @@ export class Version {
         return this;
     }
 
+    /**
+     * Mark that the null version has been put with a Cloudserver handling null keys (i.e. supporting S3C-7352)
+     * 
+     * If `isNull2` is set, `isNull` is also set to maintain consistency.
+     * Explicitly setting both avoids misunderstandings and mistakes in future updates or fixes.
+     * @return - the updated version
+     */
+    setNull2Version() {
+        this.version.isNull2 = true;
+        this.version.isNull = true;
+        return this;
+    }
+
     /**
      * Serialize the version.
      *

lib/versioning/VersioningRequestProcessor.ts

@@ -1,6 +1,8 @@
+import { RequestLogger } from 'werelogs';
+
 import errors, { ArsenalError } from '../errors';
 import { Version } from './Version';
-import { generateVersionId as genVID } from './VersionID';
+import { generateVersionId as genVID, getInfVid } from './VersionID';
 import WriteCache from './WriteCache';
 import WriteGatheringManager from './WriteGatheringManager';
 
@@ -22,11 +24,11 @@ function getPrefixUpperBoundary(prefix: string): string {
     return prefix;
 }
 
-function formatVersionKey(key: string, versionId: string) {
+function formatVersionKey(key: string, versionId: string): string {
     return `${key}${VID_SEP}${versionId}`;
 }
 
-function formatCacheKey(db: string, key: string) {
+function formatCacheKey(db: string, key: string): string {
     // using double VID_SEP to make sure the cache key is unique
     return `${db}${VID_SEP}${VID_SEP}${key}`;
 }
@@ -89,8 +91,10 @@ export default class VersioningRequestProcessor {
         callback: (error: ArsenalError | null, data?: any) => void,
     ) {
         const { db, key, options } = request;
+        logger.addDefaultFields({ bucket: db, key, options });
         if (options && options.versionId) {
-            const versionKey = formatVersionKey(key, options.versionId);
+            const keyVersionId = options.versionId === 'null' ? '' : options.versionId;
+            const versionKey = formatVersionKey(key, keyVersionId);
             return this.wgm.get({ db, key: versionKey }, logger, callback);
         }
         return this.wgm.get(request, logger, (err, data) => {
@@ -101,13 +105,82 @@ export default class VersioningRequestProcessor {
             if (!Version.isPHD(data)) {
                 return callback(null, data);
             }
-            logger.debug('master version is a PHD, getting the latest version',
-                { db, key });
+            logger.debug('master version is a PHD, getting the latest version');
             // otherwise, need to search for the latest version
             return this.getByListing(request, logger, callback);
         });
     }
 
+    /**
+     * Helper that lists version keys for a certain object key,
+     * sorted by version ID. If a null key exists for this object, it is
+     * sorted at the appropriate position by its internal version ID and
+     * its key will be appended its internal version ID.
+     *
+     * @param {string} db - bucket name
+     * @param {string} key - object key
+     * @param {object} [options] - options object
+     * @param {number} [options.limit] - max version keys returned
+     * (returns all object version keys if not specified)
+     * @param {object} logger - logger of the request
+     * @param {function} callback - callback(err, {object|null} master, {array} versions)
+     *                              master: { key, value }
+     *                              versions: [{ key, value }, ...]
+     * @return {undefined}
+     */
+    listVersionKeys(db, key, options, logger, callback) {
+        const { limit } = options || {};
+        const listingParams: any = {};
+        let nullKeyLength;
+        // include master key in v0 listing
+        listingParams.gte = key;
+        listingParams.lt = `${key}${VID_SEPPLUS}`;
+        if (limit !== undefined) {
+            // may have to skip master + null key, so 2 extra to list in the worst case
+            listingParams.limit = limit + 2;
+        }
+        nullKeyLength = key.length + 1;
+        return this.wgm.list({
+            db,
+            params: listingParams,
+        }, logger, (err, rawVersions) => {
+            if (err) {
+                return callback(err);
+            }
+            if (rawVersions.length === 0) {
+                // object does not have any version key
+                return callback(null, null, []);
+            }
+            let versions = rawVersions;
+            let master;
+            // in v0 there is always a master key before versions
+            master = versions.shift();
+            if (versions.length === 0) {
+                return callback(null, master, []);
+            }
+            const firstItem = versions[0];
+            if (firstItem.key.length === nullKeyLength) {
+                // first version is the null key
+                const nullVersion = Version.from(firstItem.value);
+                const nullVersionKey = formatVersionKey(key, <string> nullVersion.getVersionId());
+                // find null key's natural versioning order in the list
+                let nullPos = versions.findIndex(item => item.key > nullVersionKey);
+                if (nullPos === -1) {
+                    nullPos = versions.length;
+                }
+                // move null key at the correct position and append its real version ID to the key
+                versions = versions.slice(1, nullPos)
+                    .concat([{ key: nullVersionKey, value: firstItem.value, isNullKey: true }])
+                    .concat(versions.slice(nullPos));
+            }
+            if (limit !== undefined) {
+                // truncate versions to 'limit' entries
+                versions.splice(limit);
+            }
+            return callback(null, master, versions);
+        });
+    }
+
     /**
      * Get the latest version of an object when the master version is a place
      * holder for deletion. For any given pair of db and key, only a
@@ -132,39 +205,39 @@ export default class VersioningRequestProcessor {
         if (!this.enqueueGet(request, logger, callback)) {
             return null;
         }
-        logger.info('start listing latest versions', { request });
+        logger.info('start listing latest versions');
         // otherwise, search for the latest version
         const cacheKey = formatCacheKey(request.db, request.key);
         clearTimeout(this.repairing[cacheKey]);
         delete this.repairing[cacheKey];
-        const req = { db: request.db, params: {
-            gte: request.key, lt: `${request.key}${VID_SEPPLUS}`, limit: 2 } };
-        return this.wgm.list(req, logger, (err, list) => {
-            logger.info('listing latest versions done', { err, list });
+        return this.listVersionKeys(request.db, request.key, {
+            limit: 1,
+        }, logger, (err, master, versions) => {
+            logger.info('listing latest versions done', { err, master, versions });
             if (err) {
                 return this.dequeueGet(request, err);
             }
-            // the complete list of versions is always: mst, v1, v2, ...
-            if (list.length === 0) {
+            if (!master) {
                 return this.dequeueGet(request, errors.ObjNotFound);
             }
-            if (!Version.isPHD(list[0].value)) {
-                return this.dequeueGet(request, null, list[0].value);
+            if (!Version.isPHD(master.value)) {
+                return this.dequeueGet(request, null, master.value);
             }
-            if (list.length === 1) {
-                logger.info('no other versions', { request });
+            if (versions.length === 0) {
+                logger.info('no other versions');
                 this.dequeueGet(request, errors.ObjNotFound);
                 return this.repairMaster(request, logger,
-                    { type: 'del',
-                        value: list[0].value });
+                    { type: 'del', value: master.value });
             }
             // need repair
-            logger.info('update master by the latest version', { request });
-            const nextValue = list[1].value;
-            this.dequeueGet(request, null, nextValue);
+            logger.info('update master by the latest version');
+            const next = {
+                value: versions[0].value,
+                isNullKey: versions[0].isNullKey,
+            };
+            this.dequeueGet(request, null, next.value);
             return this.repairMaster(request, logger,
-                { type: 'put', value: list[0].value,
-                    nextValue });
+                { type: 'put', value: master.value, next });
         });
     }
 
@@ -227,42 +300,60 @@ export default class VersioningRequestProcessor {
      *                           RepdConnection format { db, key
      *                           [, value][, type], method, options }
      * @param logger - logger
-     * @param hints - storing reparing hints
-     * @param hints.type - type of repair operation ('put' or 'del')
-     * @param hints.value - existing value of the master version (PHD)
-     * @param hints.nextValue - the suggested latest version
-         (for 'put')
+     * @param {object} data - storing reparing hints
+     * @param {string} data.value - existing value of the master version (PHD)
+     * @param {object} data.next - the suggested latest version
+     * @param {string} data.next.value - the suggested latest version value
+     * @param {boolean} data.next.isNullKey - whether the suggested
+     * latest version is a null key
      * @return - to finish the call
      */
-    repairMaster(request: any, logger: RequestLogger, hints: {
+    repairMaster(request: any, logger: RequestLogger, data: {
         type: 'put' | 'del';
         value: string;
-        nextValue?: string;
+        next?: {
+            value: string;
+            isNullKey: boolean;
+        };
     }) {
         const { db, key } = request;
-        logger.info('start repair process', { request });
+        logger.info('start repair process');
         this.writeCache.get({ db, key }, logger, (err, value) => {
             // error or the new version is not a place holder for deletion
             if (err) {
-                return logger.info('error repairing', { request, error: err });
+                if (err.is.ObjNotFound) {
+                    return logger.debug('did not repair master: PHD was deleted');
+                } else {
+                    return logger.error('error repairing', { error: err });
+                }
             }
             if (!Version.isPHD(value)) {
-                return logger.debug('master is updated already', { request });
+                return logger.debug('master is updated already');
             }
             // the latest version is the same place holder for deletion
-            if (hints.value === value) {
+            if (data.value === value) {
                 // update the latest version with the next version
+                const ops: any = [];
+                if (data.next) {
+                    ops.push({ key, value: data.next.value });
+                    // cleanup the null key if it is the new master
+                    if (data.next.isNullKey) {
+                        ops.push({ key: formatVersionKey(key, ''), type: 'del' });
+                    }
+                } else {
+                    ops.push({ key, type: 'del' });
+                }
                 const repairRequest = {
                     db,
-                    array: [
-                        { type: hints.type, key, value: hints.nextValue },
-                    ] };
+                    array: ops,
+                };
                 logger.info('replicate repair request', { repairRequest });
                 return this.writeCache.batch(repairRequest, logger, () => {});
             }
             // The latest version is an updated place holder for deletion,
             // repeat the repair process from listing for latest versions.
             // The queue will ensure single repair process at any moment.
+            logger.info('latest version is an updated PHD');
             return this.getByListing(request, logger, () => {});
         });
     }
@@ -284,6 +375,7 @@ export default class VersioningRequestProcessor {
         callback: (error: ArsenalError | null, data?: any) => void,
     ) {
         const { db, key, value, options } = request;
+        logger.addDefaultFields({ bucket: db, key, options });
         // valid combinations of versioning options:
         // - !versioning && !versionId: normal non-versioning put
         // - versioning && !versionId: create a new version
@@ -337,6 +429,7 @@ export default class VersioningRequestProcessor {
             versionId: string,
         ) => void,
     ) {
+        logger.info('process new version put');
         // making a new versionId and a new version key
         const versionId = this.generateVersionId();
         const versionKey = formatVersionKey(request.key, versionId);
@@ -365,12 +458,22 @@ export default class VersioningRequestProcessor {
         logger: RequestLogger,
         callback: (err: ArsenalError | null, data?: any, versionId?: string) => void,
     ) {
+        logger.info('process version specific put');
         const { db, key } = request;
         // versionId is empty: update the master version
         if (request.options.versionId === '') {
             const versionId = this.generateVersionId();
             const value = Version.appendVersionId(request.value, versionId);
-            return callback(null, [{ key, value }], versionId);
+            const ops: any = [{ key, value }];
+            if (request.options.deleteNullKey) {
+                const nullKey = formatVersionKey(key, '');
+                ops.push({ key: nullKey, type: 'del' });
+            }
+            return callback(null, ops, versionId);
+        }
+        if (request.options.versionId === 'null') {
+            const nullKey = formatVersionKey(key, '');
+            return callback(null, [{ key: nullKey, value: request.value }], 'null');
         }
         // need to get the master version to check if this is the master version
         this.writeCache.get({ db, key }, logger, (err, data) => {
@@ -378,14 +481,115 @@ export default class VersioningRequestProcessor {
                 return callback(err);
             }
             const versionId = request.options.versionId;
-            const versionKey = formatVersionKey(request.key, versionId);
-            const ops = [{ key: versionKey, value: request.value }];
-            if (data === undefined ||
-                (Version.from(data).getVersionId() ?? '') >= versionId) {
-                // master does not exist or is not newer than put
-                // version and needs to be updated as well.
-                // Note that older versions have a greater version ID.
-                ops.push({ key: request.key, value: request.value });
+            const versionKey = formatVersionKey(key, versionId);
+            const ops: any = [];
+            const masterVersion = data !== undefined &&
+                  Version.from(data);
+            // push a version key if we're not updating the null
+            // version (or in legacy Cloudservers not sending the
+            // 'isNull' parameter, but this has an issue, see S3C-7526)
+            if (request.options.isNull !== true) {
+                const versionOp = { key: versionKey, value: request.value };
+                ops.push(versionOp);
+            }
+            if (masterVersion) {
+                // master key exists
+                // note that older versions have a greater version ID
+                const versionIdFromMaster = masterVersion.getVersionId();
+                if (versionIdFromMaster === undefined ||
+                    versionIdFromMaster >= versionId) {
+                    let value = request.value;
+                    logger.debug('version to put is not older than master');
+                    // Delete the deprecated, null key for backward compatibility
+                    // to avoid storing both deprecated and new null keys.
+                    // If master null version was put with an older Cloudserver (or in compat mode),
+                    // there is a possibility that it also has a null versioned key
+                    // associated, so we need to delete it as we write the null key.
+                    // Deprecated null key gets deleted when the new CloudServer:
+                    // - updates metadata of a null master (options.isNull=true)
+                    // - puts metadata on top of a master null key (options.isNull=false)
+                    if (request.options.isNull !== undefined && // new null key behavior when isNull is defined.
+                        masterVersion.isNullVersion() && // master is null
+                        !masterVersion.isNull2Version()) { // master does not support the new null key behavior yet.
+                        const masterNullVersionId = masterVersion.getVersionId();
+                        // The deprecated null key is referenced in the "versionId" property of the master key.
+                        if (masterNullVersionId) {
+                            const oldNullVersionKey = formatVersionKey(key, masterNullVersionId);
+                            ops.push({ key: oldNullVersionKey, type: 'del' });
+                        }
+                    }
+                    // new behavior when isNull is defined is to only
+                    // update the master key if it is the latest
+                    // version, old behavior needs to copy master to
+                    // the null version because older Cloudservers
+                    // rely on version-specific PUT to copy master
+                    // contents to a new null version key (newer ones
+                    // use special versionId="null" requests for this
+                    // purpose).
+                    if (versionIdFromMaster !== versionId ||
+                        request.options.isNull === undefined) {
+                        // master key is strictly older than the put version
+                        let masterVersionId;
+                        if (masterVersion.isNullVersion() && versionIdFromMaster) {
+                            logger.debug('master key is a null version');
+                            masterVersionId = versionIdFromMaster;
+                        } else if (versionIdFromMaster === undefined) {
+                            logger.debug('master key is nonversioned');
+                            // master key does not have a versionID
+                            // => create one with the "infinite" version ID
+                            masterVersionId = getInfVid(this.replicationGroupId);
+                            masterVersion.setVersionId(masterVersionId);
+                        } else {
+                            logger.debug('master key is a regular version');
+                        }
+                        if (request.options.isNull === true) {
+                            if (!masterVersionId) {
+                                // master is a regular version: delete the null key that
+                                // may exist (older null version)
+                                logger.debug('delete null key');
+                                const nullKey = formatVersionKey(key, '');
+                                ops.push({ key: nullKey, type: 'del' });
+                            }
+                        } else if (masterVersionId) {
+                            logger.debug('create version key from master version');
+                            // isNull === false means Cloudserver supports null keys,
+                            // so create a null key in this case, and a version key otherwise
+                            const masterKeyVersionId = request.options.isNull === false ?
+                                  '' : masterVersionId;
+                            const masterVersionKey = formatVersionKey(key, masterKeyVersionId);
+                            masterVersion.setNullVersion();
+                            // isNull === false means Cloudserver supports null keys,
+                            // so create a null key with the isNull2 flag
+                            if (request.options.isNull === false) {
+                                masterVersion.setNull2Version();
+                            // else isNull === undefined means Cloudserver does not support null keys,
+                            // and versionIdFromMaster !== versionId means that a version is PUT on top of a null version
+                            // hence set/update the new master nullVersionId for backward compatibility
+                            } else if (versionIdFromMaster !== versionId) {
+                                // => set the nullVersionId to the master version if put version on top of null version.
+                                value = Version.updateOrAppendNullVersionId(request.value, masterVersionId);
+                            }
+                            ops.push({ key: masterVersionKey,
+                                       value: masterVersion.toString() });
+                        }
+                    } else {
+                        logger.debug('version to put is the master');
+                    }
+                    ops.push({ key, value: value });
+                } else {
+                    logger.debug('version to put is older than master');
+                    if (request.options.isNull === true && !masterVersion.isNullVersion()) {
+                        logger.debug('create or update null key');
+                        const nullKey = formatVersionKey(key, '');
+                        const nullKeyOp = { key: nullKey, value: request.value };
+                        ops.push(nullKeyOp);
+                        // for backward compatibility: remove null version key
+                        ops.push({ key: versionKey, type: 'del' });
+                    }
+                }
+            } else {
+                // master key does not exist: create it
+                ops.push({ key, value: request.value });
             }
             return callback(null, ops, versionId);
         });
@@ -399,8 +603,10 @@ export default class VersioningRequestProcessor {
         callback: (err: ArsenalError | null, data?: any) => void,
     ) {
         const { db, key, options } = request;
+        logger.addDefaultFields({ bucket: db, key, options });
         // no versioning or versioning configuration off
         if (!(options && options.versionId)) {
+            logger.info('process non-versioned delete');
             return this.writeCache.batch({ db,
                 array: [{ key, type: 'del' }] },
             logger, callback);
@@ -438,7 +644,12 @@ export default class VersioningRequestProcessor {
             versionId?: string,
         ) => void,
     ) {
+        logger.info('process version specific delete');
         const { db, key, options } = request;
+        if (options.versionId === 'null') {
+            const nullKey = formatVersionKey(key, '');
+            return callback(null, [{ key: nullKey, type: 'del' }], 'null');
+        }
         // deleting a specific version
         this.writeCache.get({ db, key }, logger, (err, data) => {
             if (err && !err.is.ObjNotFound) {
@@ -446,7 +657,8 @@ export default class VersioningRequestProcessor {
             }
             // delete the specific version
             const versionId = options.versionId;
-            const versionKey = formatVersionKey(key, versionId);
+            const keyVersionId = options.isNull ? '' : versionId;
+            const versionKey = formatVersionKey(key, keyVersionId);
             const ops: any = [{ key: versionKey, type: 'del' }];
             // update the master version as PHD if it is the deleting version
             if (Version.isPHD(data) ||

lib/versioning/WriteCache.ts

@@ -1,3 +1,5 @@
+import { RequestLogger } from 'werelogs';
+
 import errors, { ArsenalError } from '../errors';
 import WriteGatheringManager from './WriteGatheringManager';