[ContentCacheManager] use PDF.js Xref table instead of stream detection (#242)

* make the content cache manager tests configurable

* extend stream content in unit tests

* [ContentCacheManagerTests] prepare for full object caching

* filesystem stream for pdfjs

* working??

* cleaning up

* handle overflow

* [misc] install pdfjs-dist

* [misc] move pdfjs code into app/lib/ and scripts/, also use CamelCase

* [misc] abstract the file loading and parsing of xRef tables into helper

* [misc] pdfjsTests: add snapshot based tests for the Xref table parser

* [misc] FSStream: throw proper error and drop commented code

* [misc] FSStream: integrate throwing of MissingDataException into getter

* [misc] pdfjs: fix eslint errors

* [misc] pdfjs: run format_fix

* [misc] pdfjs: allocate very small non empty dummy buffers explicitly

* [misc] install @overleaf/o-error

* [ContentCacheManager] use PDF.js Xref table instead of stream detection

Co-Authored-By: Brian Gough <brian.gough@overleaf.com>

* [pdfjs] parseXrefTable: handle empty PDF files gracefully

Co-authored-by: Brian Gough <brian.gough@overleaf.com>

services/clsi/app/js/ContentCacheManager.js

@@ -7,45 +7,97 @@ const fs = require('fs')
 const crypto = require('crypto')
 const Path = require('path')
 const Settings = require('settings-sharelatex')
+const OError = require('@overleaf/o-error')
 const pLimit = require('p-limit')
-
-const MIN_CHUNK_SIZE = Settings.pdfCachingMinChunkSize
-
-const START_OF_STREAM_MARKER = 'stream'
-const END_OF_STREAM_MARKER = 'endstream'
-const START_OF_STREAM_MARKER_LENGTH = START_OF_STREAM_MARKER.length
-const END_OF_STREAM_MARKER_LENGTH = END_OF_STREAM_MARKER.length
+const { parseXrefTable } = require('../lib/pdfjs/parseXrefTable')
 
 /**
  *
  * @param {String} contentDir path to directory where content hash files are cached
  * @param {String} filePath the pdf file to scan for streams
+ * @param {number} size the pdf size
  */
-async function update(contentDir, filePath) {
-  const stream = fs.createReadStream(filePath)
-  const extractor = new PdfStreamsExtractor()
+async function update(contentDir, filePath, size) {
   const ranges = []
   const newRanges = []
   // keep track of hashes expire old ones when they reach a generation > N.
   const tracker = await HashFileTracker.from(contentDir)
   tracker.updateAge()
 
-  for await (const chunk of stream) {
-    const pdfStreams = extractor.consume(chunk)
-    for (const pdfStream of pdfStreams) {
-      if (pdfStream.end - pdfStream.start < MIN_CHUNK_SIZE) continue
-      const hash = pdfStreamHash(pdfStream.buffers)
+  const rawTable = await parseXrefTable(filePath, size)
+  rawTable.sort((a, b) => {
+    return a.offset - b.offset
+  })
+  rawTable.forEach((obj, idx) => {
+    obj.idx = idx
+  })
+
+  const uncompressedObjects = []
+  for (const object of rawTable) {
+    if (!object.uncompressed) {
+      continue
+    }
+    const nextObject = rawTable[object.idx + 1]
+    if (!nextObject) {
+      // Ignore this possible edge case.
+      // The last object should be part of the xRef table.
+      continue
+    } else {
+      object.endOffset = nextObject.offset
+    }
+    const size = object.endOffset - object.offset
+    object.size = size
+    if (size < Settings.pdfCachingMinChunkSize) {
+      continue
+    }
+    uncompressedObjects.push(object)
+  }
 
-      const range = { start: pdfStream.start, end: pdfStream.end, hash }
+  const handle = await fs.promises.open(filePath)
+  try {
+    for (const object of uncompressedObjects) {
+      let buffer = Buffer.alloc(object.size, 0)
+      const { bytesRead } = await handle.read(
+        buffer,
+        0,
+        object.size,
+        object.offset
+      )
+      if (bytesRead !== object.size) {
+        throw new OError('could not read full chunk', {
+          object,
+          bytesRead
+        })
+      }
+      const idxObj = buffer.indexOf('obj')
+      if (idxObj > 100) {
+        throw new OError('objectId is too large', {
+          object,
+          idxObj
+        })
+      }
+      const objectIdRaw = buffer.subarray(0, idxObj)
+      buffer = buffer.subarray(objectIdRaw.byteLength)
+
+      const hash = pdfStreamHash(buffer)
+      const range = {
+        objectId: objectIdRaw.toString(),
+        start: object.offset + objectIdRaw.byteLength,
+        end: object.endOffset,
+        hash
+      }
       ranges.push(range)
 
       // Optimization: Skip writing of duplicate streams.
       if (tracker.track(range)) continue
 
-      await writePdfStream(contentDir, hash, pdfStream.buffers)
+      await writePdfStream(contentDir, hash, buffer)
       newRanges.push(range)
     }
+  } finally {
+    await handle.close()
   }
+
   const reclaimedSpace = await tracker.deleteStaleHashes(5)
   await tracker.flush()
   return [ranges, newRanges, reclaimedSpace]
@@ -142,94 +194,21 @@ class HashFileTracker {
   }
 }
 
-class PdfStreamsExtractor {
-  constructor() {
-    this.fileIndex = 0
-    this.inStream = false
-    this.streamStartIndex = 0
-    this.buffers = []
-    this.lastChunk = Buffer.alloc(0)
-  }
-
-  consume(chunk) {
-    let chunkIndex = 0
-    const pdfStreams = []
-    chunk = Buffer.concat([this.lastChunk, chunk])
-    while (true) {
-      if (!this.inStream) {
-        // Not in a stream, look for stream start
-        const index = chunk.indexOf(START_OF_STREAM_MARKER, chunkIndex)
-        if (index === -1) {
-          // Couldn't find stream start
-          break
-        }
-        // Found stream start, start a stream
-        this.inStream = true
-        this.streamStartIndex = this.fileIndex + index
-        chunkIndex = index
-      } else {
-        // In a stream, look for stream end
-        const index = chunk.indexOf(END_OF_STREAM_MARKER, chunkIndex)
-        if (index === -1) {
-          break
-        }
-        // add "endstream" part
-        const endIndex = index + END_OF_STREAM_MARKER_LENGTH
-        this.buffers.push(chunk.slice(chunkIndex, endIndex))
-        pdfStreams.push({
-          start: this.streamStartIndex,
-          end: this.fileIndex + endIndex,
-          buffers: this.buffers
-        })
-        this.inStream = false
-        this.buffers = []
-        chunkIndex = endIndex
-      }
-    }
-
-    const remaining = chunk.length - chunkIndex
-    const nextMarkerLength = this.inStream
-      ? END_OF_STREAM_MARKER_LENGTH
-      : START_OF_STREAM_MARKER_LENGTH
-    if (remaining > nextMarkerLength) {
-      const retainMarkerSection = chunk.length - nextMarkerLength
-      if (this.inStream) {
-        this.buffers.push(chunk.slice(chunkIndex, retainMarkerSection))
-      }
-      this.lastChunk = chunk.slice(retainMarkerSection)
-      this.fileIndex += retainMarkerSection
-    } else {
-      this.lastChunk = chunk.slice(chunkIndex)
-      this.fileIndex += chunkIndex
-    }
-    return pdfStreams
-  }
-}
-
-function pdfStreamHash(buffers) {
+function pdfStreamHash(buffer) {
   const hash = crypto.createHash('sha256')
-  for (const buffer of buffers) {
-    hash.update(buffer)
-  }
+  hash.update(buffer)
   return hash.digest('hex')
 }
 
-async function writePdfStream(dir, hash, buffers) {
+async function writePdfStream(dir, hash, buffer) {
   const filename = Path.join(dir, hash)
   const atomicWriteFilename = filename + '~'
-  const file = await fs.promises.open(atomicWriteFilename, 'w')
   if (Settings.enablePdfCachingDark) {
     // Write an empty file in dark mode.
-    buffers = []
+    buffer = Buffer.alloc(0)
   }
   try {
-    try {
-      for (const buffer of buffers) {
-        await file.write(buffer)
-      }
-    } finally {
-      await file.close()
-    }
+    await fs.promises.writeFile(atomicWriteFilename, buffer)
     await fs.promises.rename(atomicWriteFilename, filename)
   } catch (err) {
     try {
@@ -247,5 +226,8 @@ function promiseMapWithLimit(concurrency, array, fn) {
 
 module.exports = {
   HASH_REGEX: /^[0-9a-f]{64}$/,
-  update: callbackify(update)
+  update: callbackify(update),
+  promises: {
+    update
+  }
 }

services/clsi/app/js/OutputCacheManager.js

@@ -275,36 +275,39 @@ module.exports = OutputCacheManager = {
           outputDir,
           OutputCacheManager.path(outputFile.build, outputFile.path)
         )
+        const pdfSize = outputFile.size
         const timer = new Metrics.Timer('compute-pdf-ranges')
-        ContentCacheManager.update(contentDir, outputFilePath, function (
-          err,
-          result
-        ) {
-          if (err) return callback(err, outputFiles)
-          const [contentRanges, newContentRanges, reclaimedSpace] = result
+        ContentCacheManager.update(
+          contentDir,
+          outputFilePath,
+          pdfSize,
+          function (err, result) {
+            if (err) return callback(err, outputFiles)
+            const [contentRanges, newContentRanges, reclaimedSpace] = result
 
-          if (Settings.enablePdfCachingDark) {
-            // In dark mode we are doing the computation only and do not emit
-            //  any ranges to the frontend.
-          } else {
-            outputFile.contentId = Path.basename(contentDir)
-            outputFile.ranges = contentRanges
-          }
+            if (Settings.enablePdfCachingDark) {
+              // In dark mode we are doing the computation only and do not emit
+              //  any ranges to the frontend.
+            } else {
+              outputFile.contentId = Path.basename(contentDir)
+              outputFile.ranges = contentRanges
+            }
 
-          timings['compute-pdf-caching'] = timer.done()
-          stats['pdf-caching-n-ranges'] = contentRanges.length
-          stats['pdf-caching-total-ranges-size'] = contentRanges.reduce(
-            (sum, next) => sum + (next.end - next.start),
-            0
-          )
-          stats['pdf-caching-n-new-ranges'] = newContentRanges.length
-          stats['pdf-caching-new-ranges-size'] = newContentRanges.reduce(
-            (sum, next) => sum + (next.end - next.start),
-            0
-          )
-          stats['pdf-caching-reclaimed-space'] = reclaimedSpace
-          callback(null, outputFiles)
-        })
+            timings['compute-pdf-caching'] = timer.done()
+            stats['pdf-caching-n-ranges'] = contentRanges.length
+            stats['pdf-caching-total-ranges-size'] = contentRanges.reduce(
+              (sum, next) => sum + (next.end - next.start),
+              0
+            )
+            stats['pdf-caching-n-new-ranges'] = newContentRanges.length
+            stats['pdf-caching-new-ranges-size'] = newContentRanges.reduce(
+              (sum, next) => sum + (next.end - next.start),
+              0
+            )
+            stats['pdf-caching-reclaimed-space'] = reclaimedSpace
+            callback(null, outputFiles)
+          }
+        )
       } else {
         callback(null, outputFiles)
       }

services/clsi/app/lib/pdfjs/FSPdfManager.js

@@ -0,0 +1,43 @@
+const { PDFDocument } = require('pdfjs-dist/lib/core/document')
+const { LocalPdfManager } = require('pdfjs-dist/lib/core/pdf_manager')
+const { MissingDataException } = require('pdfjs-dist/lib/core/core_utils')
+const { FSStream } = require('./FSStream')
+
+class FSPdfManager extends LocalPdfManager {
+  constructor(docId, options) {
+    const nonEmptyDummyBuffer = Buffer.alloc(1, 0)
+    super(docId, nonEmptyDummyBuffer)
+    this.stream = new FSStream(options.fh, 0, options.size)
+    this.pdfDocument = new PDFDocument(this, this.stream)
+  }
+
+  async ensure(obj, prop, args) {
+    try {
+      const value = obj[prop]
+      if (typeof value === 'function') {
+        return value.apply(obj, args)
+      }
+      return value
+    } catch (ex) {
+      if (!(ex instanceof MissingDataException)) {
+        throw ex
+      }
+      await this.requestRange(ex.begin, ex.end)
+      return this.ensure(obj, prop, args)
+    }
+  }
+
+  requestRange(begin, end) {
+    return this.stream.requestRange(begin, end)
+  }
+
+  requestLoadedStream() {}
+
+  onLoadedStream() {}
+
+  terminate(reason) {}
+}
+
+module.exports = {
+  FSPdfManager
+}

services/clsi/app/lib/pdfjs/FSStream.js

@@ -0,0 +1,138 @@
+const { Stream } = require('pdfjs-dist/lib/core/stream')
+const { MissingDataException } = require('pdfjs-dist/lib/core/core_utils')
+
+const BUF_SIZE = 1024 // read from the file in 1024 byte pages
+
+class FSStream extends Stream {
+  constructor(fh, start, length, dict, cachedBytes) {
+    const nonEmptyDummyBuffer = Buffer.alloc(1, 0)
+    super(nonEmptyDummyBuffer, start, length, dict)
+    delete this.bytes
+    this.fh = fh
+    this.cachedBytes = cachedBytes || []
+  }
+
+  get length() {
+    return this.end - this.start
+  }
+
+  get isEmpty() {
+    return this.length === 0
+  }
+
+  // Manage cached reads from the file
+
+  requestRange(begin, end) {
+    // expand small ranges to read a larger amount
+    if (end - begin < BUF_SIZE) {
+      end = begin + BUF_SIZE
+    }
+    end = Math.min(end, this.length)
+    // keep a cache of previous reads with {begin,end,buffer} values
+    const result = {
+      begin: begin,
+      end: end,
+      buffer: Buffer.alloc(end - begin, 0)
+    }
+    this.cachedBytes.push(result)
+    return this.fh.read(result.buffer, 0, end - begin, begin)
+  }
+
+  _ensureGetPos(pos) {
+    const found = this.cachedBytes.find((x) => {
+      return x.begin <= pos && pos < x.end
+    })
+    if (!found) {
+      throw new MissingDataException(pos, pos + 1)
+    }
+    return found
+  }
+
+  _ensureGetRange(begin, end) {
+    end = Math.min(end, this.length) // BG: handle overflow case
+    const found = this.cachedBytes.find((x) => {
+      return x.begin <= begin && end <= x.end
+    })
+    if (!found) {
+      throw new MissingDataException(begin, end)
+    }
+    return found
+  }
+
+  _readByte(found, pos) {
+    return found.buffer[pos - found.begin]
+  }
+
+  _readBytes(found, pos, end) {
+    return found.buffer.subarray(pos - found.begin, end - found.begin)
+  }
+
+  // handle accesses to the bytes
+
+  ensureByte(pos) {
+    this._ensureGetPos(pos) // may throw a MissingDataException
+  }
+
+  getByte() {
+    const pos = this.pos
+    if (this.pos >= this.end) {
+      return -1
+    }
+    const found = this._ensureGetPos(pos)
+    return this._readByte(found, this.pos++)
+  }
+
+  // BG: for a range, end is not included (see Buffer.subarray for example)
+
+  ensureBytes(length, forceClamped = false) {
+    const pos = this.pos
+    this._ensureGetRange(pos, pos + length)
+  }
+
+  getBytes(length, forceClamped = false) {
+    const pos = this.pos
+    const strEnd = this.end
+
+    const found = this._ensureGetRange(pos, pos + length)
+    if (!length) {
+      const subarray = this._readBytes(found, pos, strEnd)
+      // `this.bytes` is always a `Uint8Array` here.
+      return forceClamped ? new Uint8ClampedArray(subarray) : subarray
+    }
+    let end = pos + length
+    if (end > strEnd) {
+      end = strEnd
+    }
+    this.pos = end
+    const subarray = this._readBytes(found, pos, end)
+    // `this.bytes` is always a `Uint8Array` here.
+    return forceClamped ? new Uint8ClampedArray(subarray) : subarray
+  }
+
+  getByteRange() {
+    // BG: this isn't needed as far as I can tell
+    throw new Error('not implemented')
+  }
+
+  reset() {
+    this.pos = this.start
+  }
+
+  moveStart() {
+    this.start = this.pos
+  }
+
+  makeSubStream(start, length, dict = null) {
+    // BG: had to add this check for null length, it is being called with only
+    // the start value at one point in the xref decoding. The intent is clear
+    // enough
+    // - a null length means "to the end of the file" -- not sure how it is
+    //   working in the existing pdfjs code without this.
+    if (!length) {
+      length = this.end - start
+    }
+    return new FSStream(this.fh, start, length, dict, this.cachedBytes)
+  }
+}
+
+module.exports = { FSStream }

services/clsi/app/lib/pdfjs/parseXrefTable.js

@@ -0,0 +1,24 @@
+const fs = require('fs')
+const { FSPdfManager } = require('./FSPdfManager')
+
+async function parseXrefTable(path, size) {
+  if (size === 0) {
+    return []
+  }
+
+  const file = await fs.promises.open(path)
+  try {
+    const manager = new FSPdfManager(0, { fh: file, size })
+
+    await manager.ensureDoc('checkHeader')
+    await manager.ensureDoc('parseStartXRef')
+    await manager.ensureDoc('parse')
+    return manager.pdfDocument.catalog.xref.entries
+  } finally {
+    file.close()
+  }
+}
+
+module.exports = {
+  parseXrefTable
+}

services/clsi/package-lock.json

@@ -1008,9 +1008,19 @@
       }
     },
     "@overleaf/o-error": {
-      "version": "3.0.0",
-      "resolved": "https://registry.npmjs.org/@overleaf/o-error/-/o-error-3.0.0.tgz",
-      "integrity": "sha512-LsM2s6Iy9G97ktPo0ys4VxtI/m3ahc1ZHwjo5XnhXtjeIkkkVAehsrcRRoV/yWepPjymB0oZonhcfojpjYR/tg=="
+      "version": "3.3.1",
+      "resolved": "https://registry.npmjs.org/@overleaf/o-error/-/o-error-3.3.1.tgz",
+      "integrity": "sha512-1FRBYZO0lbJ0U+FRGZVS8ou6RhEw3e2B86WW/NbtBw554g0h5iC8ESf+juIfPMU/WDf/JDIFbg3eB/LnP2RSow==",
+      "requires": {
+        "core-js": "^3.8.3"
+      },
+      "dependencies": {
+        "core-js": {
+          "version": "3.13.0",
+          "resolved": "https://registry.npmjs.org/core-js/-/core-js-3.13.0.tgz",
+          "integrity": "sha512-iWDbiyha1M5vFwPFmQnvRv+tJzGbFAm6XimJUT0NgHYW3xZEs1SkCAcasWSVFxpI2Xb/V1DDJckq3v90+bQnog=="
+        }
+      }
     },
     "@protobufjs/aspromise": {
       "version": "1.1.2",
@@ -5594,6 +5604,11 @@
       "integrity": "sha1-uULm1L3mUwBe9rcTYd74cn0GReA=",
       "dev": true
     },
+    "pdfjs-dist": {
+      "version": "2.7.570",
+      "resolved": "https://registry.npmjs.org/pdfjs-dist/-/pdfjs-dist-2.7.570.tgz",
+      "integrity": "sha512-/ZkA1FwkEOyDaq11JhMLazdwQAA0F9uwrP7h/1L9Akt9KWh1G5/tkzS+bPuUELq2s2GDFnaT+kooN/aSjT7DXQ=="
+    },
     "performance-now": {
       "version": "2.1.0",
       "resolved": "https://registry.npmjs.org/performance-now/-/performance-now-2.1.0.tgz",

services/clsi/package.json

@@ -10,7 +10,7 @@
     "start": "node $NODE_APP_OPTIONS app.js",
     "test:acceptance:_run": "mocha --recursive --reporter spec --timeout 15000 --exit $@ test/acceptance/js",
     "test:acceptance": "npm run test:acceptance:_run -- --grep=$MOCHA_GREP",
-    "test:unit:_run": "mocha --recursive --reporter spec $@ test/unit/js",
+    "test:unit:_run": "mocha --recursive --reporter spec $@ test/unit/js test/unit/lib",
     "test:unit": "npm run test:unit:_run -- --grep=$MOCHA_GREP",
     "nodemon": "nodemon --config nodemon.json",
     "lint": "node_modules/.bin/eslint --max-warnings 0 .",
@@ -20,6 +20,7 @@
   "author": "James Allen <james@sharelatex.com>",
   "dependencies": {
     "@overleaf/metrics": "^3.5.1",
+    "@overleaf/o-error": "^3.3.1",
     "async": "3.2.0",
     "body-parser": "^1.19.0",
     "diskusage": "^1.1.3",
@@ -32,6 +33,7 @@
     "logger-sharelatex": "^2.2.0",
     "mysql": "^2.18.1",
     "p-limit": "^3.1.0",
+    "pdfjs-dist": "^2.7.570",
     "request": "^2.88.2",
     "sequelize": "^5.21.5",
     "settings-sharelatex": "^1.1.0",

services/clsi/scripts/demo-pdfjs-Xref.js

@@ -0,0 +1,12 @@
+const fs = require('fs')
+const { parseXrefTable } = require('../app/lib/pdfjs/parseXrefTable')
+
+const pdfPath = process.argv[2]
+
+async function main() {
+  const size = (await fs.promises.stat(pdfPath)).size
+  const xRefEntries = await parseXrefTable(pdfPath, size)
+  console.log('Xref entries', xRefEntries)
+}
+
+main().catch(console.error)

services/clsi/test/unit/js/ContentCacheManagerTests.js

@@ -1,201 +1,123 @@
+const fs = require('fs')
 const Path = require('path')
-const crypto = require('crypto')
-const { Readable } = require('stream')
-const SandboxedModule = require('sandboxed-module')
-const sinon = require('sinon')
 const { expect } = require('chai')
 
 const MODULE_PATH = '../../../app/js/ContentCacheManager'
 
-class FakeFile {
-  constructor() {
-    this.closed = false
-    this.contents = []
-  }
-
-  async write(blob) {
-    this.contents.push(blob)
-    return this
-  }
-
-  async close() {
-    this.closed = true
-    return this
-  }
-
-  toJSON() {
-    return {
-      contents: Buffer.concat(this.contents).toString(),
-      closed: this.closed
-    }
-  }
-}
-
-function hash(blob) {
-  const hash = crypto.createHash('sha256')
-  hash.update(blob)
-  return hash.digest('hex')
-}
-
 describe('ContentCacheManager', function () {
   let contentDir, pdfPath
-  let ContentCacheManager, fs, files, Settings
-  function load() {
-    ContentCacheManager = SandboxedModule.require(MODULE_PATH, {
-      requires: {
-        fs,
-        'settings-sharelatex': Settings
-      }
-    })
-  }
+  let ContentCacheManager, files, Settings
+  before(function () {
+    Settings = require('settings-sharelatex')
+    ContentCacheManager = require(MODULE_PATH)
+  })
   let contentRanges, newContentRanges, reclaimed
-  function run(filePath, done) {
-    ContentCacheManager.update(contentDir, filePath, (err, ranges) => {
-      if (err) return done(err)
-      let newlyReclaimed
-      ;[contentRanges, newContentRanges, newlyReclaimed] = ranges
-      reclaimed += newlyReclaimed
-      done()
-    })
+  async function run(filePath, size) {
+    const result = await ContentCacheManager.promises.update(
+      contentDir,
+      filePath,
+      size
+    )
+    let newlyReclaimed
+    ;[contentRanges, newContentRanges, newlyReclaimed] = result
+    reclaimed += newlyReclaimed
+
+    const fileNames = await fs.promises.readdir(contentDir)
+    files = {}
+    for (const fileName of fileNames) {
+      const path = Path.join(contentDir, fileName)
+      files[path] = await fs.promises.readFile(path)
+    }
   }
-
-  beforeEach(function () {
-    reclaimed = 0
+  before(function () {
     contentDir =
       '/app/output/602cee6f6460fca0ba7921e6/content/1797a7f48f9-5abc1998509dea1f'
     pdfPath =
       '/app/output/602cee6f6460fca0ba7921e6/generated-files/1797a7f48ea-8ac6805139f43351/output.pdf'
-    Settings = {
-      pdfCachingMinChunkSize: 1024,
-      enablePdfCachingDark: false
-    }
-    files = {}
-    fs = {
-      createReadStream: sinon.stub().returns(Readable.from([])),
-      promises: {
-        async writeFile(name, blob) {
-          const file = new FakeFile()
-          await file.write(Buffer.from(blob))
-          await file.close()
-          files[name] = file
-        },
-        async readFile(name) {
-          if (!files[name]) {
-            throw new Error()
-          }
-          return files[name].toJSON().contents
-        },
-        async open(name) {
-          files[name] = new FakeFile()
-          return files[name]
-        },
-        async stat(name) {
-          if (!files[name]) {
-            throw new Error()
-          }
-        },
-        async rename(oldName, newName) {
-          if (!files[oldName]) {
-            throw new Error()
-          }
-          files[newName] = files[oldName]
-          delete files[oldName]
-        },
-        async unlink(name) {
-          if (!files[name]) {
-            throw new Error()
-          }
-          delete files[name]
-        }
-      }
-    }
+
+    reclaimed = 0
+    Settings.pdfCachingMinChunkSize = 1024
+  })
+
+  before(async function () {
+    await fs.promises.rmdir(contentDir, { recursive: true })
+    await fs.promises.mkdir(contentDir, { recursive: true })
+    await fs.promises.mkdir(Path.dirname(pdfPath), { recursive: true })
   })
 
-  describe('with a small minChunkSize', function () {
-    beforeEach(function () {
-      Settings.pdfCachingMinChunkSize = 1
-      load()
+  describe('minimal', function () {
+    const PATH_MINIMAL = 'test/acceptance/fixtures/minimal.pdf'
+    const OBJECT_ID_1 = '9 0 '
+    const HASH_LARGE =
+      'd7cfc73ad2fba4578a437517923e3714927bbf35e63ea88bd93c7a8076cf1fcd'
+    const OBJECT_ID_2 = '10 0 '
+    const HASH_SMALL =
+      '896749b8343851b0dc385f71616916a7ba0434fcfb56d1fc7e27cd139eaa2f71'
+    function getChunkPath(hash) {
+      return Path.join('test/unit/js/snapshots/minimalCompile/chunks', hash)
+    }
+    let MINIMAL_SIZE, RANGE_1, RANGE_2, h1, h2, START_1, START_2, END_1, END_2
+    before(async function () {
+      await fs.promises.copyFile(PATH_MINIMAL, pdfPath)
+      const MINIMAL = await fs.promises.readFile(PATH_MINIMAL)
+      MINIMAL_SIZE = (await fs.promises.stat(PATH_MINIMAL)).size
+      RANGE_1 = await fs.promises.readFile(getChunkPath(HASH_LARGE))
+      RANGE_2 = await fs.promises.readFile(getChunkPath(HASH_SMALL))
+      h1 = HASH_LARGE
+      h2 = HASH_SMALL
+      START_1 = MINIMAL.indexOf(RANGE_1)
+      END_1 = START_1 + RANGE_1.byteLength
+      START_2 = MINIMAL.indexOf(RANGE_2)
+      END_2 = START_2 + RANGE_2.byteLength
     })
+    async function runWithMinimal() {
+      await run(pdfPath, MINIMAL_SIZE)
+    }
 
-    describe('when the ranges are split across chunks', function () {
-      const RANGE_1 = 'stream123endstream'
-      const RANGE_2 = 'stream(||)endstream'
-      const RANGE_3 = 'stream!$%/=endstream'
-      const h1 = hash(RANGE_1)
-      const h2 = hash(RANGE_2)
-      const h3 = hash(RANGE_3)
-      function runWithSplitStream(done) {
-        fs.createReadStream
-          .withArgs(pdfPath)
-          .returns(
-            Readable.from([
-              Buffer.from('abcstr'),
-              Buffer.from('eam123endstreamABC'),
-              Buffer.from('str'),
-              Buffer.from('eam(||'),
-              Buffer.from(')end'),
-              Buffer.from('stream-_~stream!$%/=endstream')
-            ])
-          )
-        run(pdfPath, done)
-      }
-      beforeEach(function (done) {
-        runWithSplitStream(done)
+    describe('with two ranges qualifying', function () {
+      before(function () {
+        Settings.pdfCachingMinChunkSize = 500
       })
-
-      it('should produce three ranges', function () {
-        expect(contentRanges).to.have.length(3)
+      before(async function () {
+        await runWithMinimal()
+      })
+      it('should produce two ranges', function () {
+        expect(contentRanges).to.have.length(2)
       })
 
       it('should find the correct offsets', function () {
         expect(contentRanges).to.deep.equal([
           {
-            start: 3,
-            end: 21,
-            hash: hash(RANGE_1)
-          },
-          {
-            start: 24,
-            end: 43,
-            hash: hash(RANGE_2)
+            objectId: OBJECT_ID_1,
+            start: START_1,
+            end: END_1,
+            hash: h1
           },
           {
-            start: 46,
-            end: 66,
-            hash: hash(RANGE_3)
+            objectId: OBJECT_ID_2,
+            start: START_2,
+            end: END_2,
+            hash: h2
           }
         ])
       })
 
       it('should store the contents', function () {
-        expect(JSON.parse(JSON.stringify(files))).to.deep.equal({
-          [Path.join(contentDir, h1)]: {
-            contents: RANGE_1,
-            closed: true
-          },
-          [Path.join(contentDir, h2)]: {
-            contents: RANGE_2,
-            closed: true
-          },
-          [Path.join(contentDir, h3)]: {
-            contents: RANGE_3,
-            closed: true
-          },
-          [Path.join(contentDir, '.state.v0.json')]: {
-            contents: JSON.stringify({
+        expect(files).to.deep.equal({
+          [Path.join(contentDir, h1)]: RANGE_1,
+          [Path.join(contentDir, h2)]: RANGE_2,
+          [Path.join(contentDir, '.state.v0.json')]: Buffer.from(
+            JSON.stringify({
               hashAge: [
                 [h1, 0],
-                [h2, 0],
-                [h3, 0]
+                [h2, 0]
               ],
               hashSize: [
-                [h1, 18],
-                [h2, 19],
-                [h3, 20]
+                [h1, RANGE_1.byteLength],
+                [h2, RANGE_2.byteLength]
               ]
-            }),
-            closed: true
-          }
+            })
+          )
         })
       })
 
@@ -203,71 +125,46 @@ describe('ContentCacheManager', function () {
         expect(contentRanges).to.deep.equal(newContentRanges)
       })
 
-      describe('when re-running with one stream removed', function () {
-        function runWithOneSplitStreamRemoved(done) {
-          fs.createReadStream
-            .withArgs(pdfPath)
-            .returns(
-              Readable.from([
-                Buffer.from('abcstr'),
-                Buffer.from('eam123endstreamABC'),
-                Buffer.from('stream!$%/=endstream')
-              ])
-            )
-          run(pdfPath, done)
-        }
-        beforeEach(function (done) {
-          runWithOneSplitStreamRemoved(done)
+      describe('when re-running with one range too small', function () {
+        before(function () {
+          Settings.pdfCachingMinChunkSize = 1024
+        })
+
+        before(async function () {
+          await runWithMinimal()
         })
 
-        it('should produce two ranges', function () {
-          expect(contentRanges).to.have.length(2)
+        it('should produce one range', function () {
+          expect(contentRanges).to.have.length(1)
         })
 
         it('should find the correct offsets', function () {
           expect(contentRanges).to.deep.equal([
             {
-              start: 3,
-              end: 21,
-              hash: hash(RANGE_1)
-            },
-            {
-              start: 24,
-              end: 44,
-              hash: hash(RANGE_3)
+              objectId: OBJECT_ID_1,
+              start: START_1,
+              end: END_1,
+              hash: h1
             }
           ])
         })
 
         it('should update the age of the 2nd range', function () {
-          expect(JSON.parse(JSON.stringify(files))).to.deep.equal({
-            [Path.join(contentDir, h1)]: {
-              contents: RANGE_1,
-              closed: true
-            },
-            [Path.join(contentDir, h2)]: {
-              contents: RANGE_2,
-              closed: true
-            },
-            [Path.join(contentDir, h3)]: {
-              contents: RANGE_3,
-              closed: true
-            },
-            [Path.join(contentDir, '.state.v0.json')]: {
-              contents: JSON.stringify({
+          expect(files).to.deep.equal({
+            [Path.join(contentDir, h1)]: RANGE_1,
+            [Path.join(contentDir, h2)]: RANGE_2,
+            [Path.join(contentDir, '.state.v0.json')]: Buffer.from(
+              JSON.stringify({
                 hashAge: [
                   [h1, 0],
-                  [h2, 1],
-                  [h3, 0]
+                  [h2, 1]
                 ],
                 hashSize: [
-                  [h1, 18],
-                  [h2, 19],
-                  [h3, 20]
+                  [h1, RANGE_1.byteLength],
+                  [h2, RANGE_2.byteLength]
                 ]
-              }),
-              closed: true
-            }
+              })
+            )
           })
         })
 
@@ -277,53 +174,35 @@ describe('ContentCacheManager', function () {
 
         describe('when re-running 5 more times', function () {
           for (let i = 0; i < 5; i++) {
-            beforeEach(function (done) {
-              runWithOneSplitStreamRemoved(done)
+            before(async function () {
+              await runWithMinimal()
             })
           }
 
-          it('should still produce two ranges', function () {
-            expect(contentRanges).to.have.length(2)
+          it('should still produce one range', function () {
+            expect(contentRanges).to.have.length(1)
           })
 
           it('should still find the correct offsets', function () {
             expect(contentRanges).to.deep.equal([
               {
-                start: 3,
-                end: 21,
-                hash: hash(RANGE_1)
-              },
-              {
-                start: 24,
-                end: 44,
-                hash: hash(RANGE_3)
+                objectId: OBJECT_ID_1,
+                start: START_1,
+                end: END_1,
+                hash: h1
               }
             ])
           })
 
           it('should delete the 2nd range', function () {
-            expect(JSON.parse(JSON.stringify(files))).to.deep.equal({
-              [Path.join(contentDir, h1)]: {
-                contents: RANGE_1,
-                closed: true
-              },
-              [Path.join(contentDir, h3)]: {
-                contents: RANGE_3,
-                closed: true
-              },
-              [Path.join(contentDir, '.state.v0.json')]: {
-                contents: JSON.stringify({
-                  hashAge: [
-                    [h1, 0],
-                    [h3, 0]
-                  ],
-                  hashSize: [
-                    [h1, 18],
-                    [h3, 20]
-                  ]
-                }),
-                closed: true
-              }
+            expect(files).to.deep.equal({
+              [Path.join(contentDir, h1)]: RANGE_1,
+              [Path.join(contentDir, '.state.v0.json')]: Buffer.from(
+                JSON.stringify({
+                  hashAge: [[h1, 0]],
+                  hashSize: [[h1, RANGE_1.byteLength]]
+                })
+              )
             })
           })
 
@@ -332,7 +211,7 @@ describe('ContentCacheManager', function () {
           })
 
           it('should yield the reclaimed space', function () {
-            expect(reclaimed).to.equal(RANGE_2.length)
+            expect(reclaimed).to.equal(RANGE_2.byteLength)
           })
         })
       })

services/clsi/test/unit/js/snapshots/minimalCompile/chunks/896749b8343851b0dc385f71616916a7ba0434fcfb56d1fc7e27cd139eaa2f71

@@ -0,0 +1,7 @@
+obj
+<< /Type /ObjStm /Length 447 /Filter /FlateDecode /N 5 /First 32 >>
+stream
+xœ…RQkÛ0~߯¸Ç�K'É’¥Ð$„vkGIJ[(yð‘Ò8Ø*ëþýî$¯IûR°déîû>ÝI"H@¥¤Ï9@J!`	Vú/gg fí>BZx†J˜Ãù9ˆÛ®]-B„'ZNg îÂk„%¥i\÷!ÅÍõüêÇôûäfŽ’4š®�“mÝ�©²ô4ô«®9ĶãCY]\ב@‹—ßñï!Ð	4cþ
U‰‡f·=�ËJgOÚÊòg³î©>—’ËÿÌÛzzÆ>A)å§C9Wøw³K³qªPÃœ#®”à¤/48•¢VX/¶TŒð¦p	§-%2"³­*¡B;ä‡XâêÚ2£,9’õG¥z;ž†E–Î¥øØJj/‰c
+ón­%¯�ˆÌ¢ê¨áµ¦f3æˆÎ]!µ¢«¤ç=µyšÍ
,Ós]ô«@ïe+COW.C³ÙÒÖkÚ’	Øc_Å¥X‹ v¢í>ˆNüù‚�ÃÌ2u¶«7=˜ìŒñ¸}¥—	#•®HVér–9?kv
ªì6ÞÿªŸÃG«^Åz׬.ö‚ÒvÃó=UyjÀïŠÇ¡p£‘ÍõzìË2
+endstream
+endobj

services/clsi/test/unit/lib/pdfjsTests.js

@@ -0,0 +1,78 @@
+const fs = require('fs')
+const Path = require('path')
+const { expect } = require('chai')
+const { parseXrefTable } = require('../../../app/lib/pdfjs/parseXrefTable')
+const PATH_EXAMPLES = 'test/acceptance/fixtures/examples/'
+const PATH_SNAPSHOTS = 'test/unit/lib/snapshots/'
+const EXAMPLES = fs.readdirSync(PATH_EXAMPLES)
+
+function snapshotPath(example) {
+  return Path.join(PATH_SNAPSHOTS, example, 'XrefTable.json')
+}
+
+function pdfPath(example) {
+  return Path.join(PATH_EXAMPLES, example, 'output.pdf')
+}
+
+async function loadContext(example) {
+  const size = (await fs.promises.stat(pdfPath(example))).size
+
+  let blob
+  try {
+    blob = await fs.promises.readFile(snapshotPath(example))
+  } catch (e) {
+    if (e.code !== 'ENOENT') {
+      throw e
+    }
+  }
+  const snapshot = blob ? JSON.parse(blob) : null
+  return {
+    size,
+    snapshot
+  }
+}
+
+async function backFillSnapshot(example, size) {
+  const table = await parseXrefTable(pdfPath(example), size)
+  await fs.promises.mkdir(Path.dirname(snapshotPath(example)), {
+    recursive: true
+  })
+  await fs.promises.writeFile(
+    snapshotPath(example),
+    JSON.stringify(table, null, 2)
+  )
+  return table
+}
+
+describe('pdfjs', function () {
+  describe('when the pdf is an empty file', function () {
+    it('should yield no entries', async function () {
+      const path = 'does/not/matter.pdf'
+      const table = await parseXrefTable(path, 0)
+      expect(table).to.deep.equal([])
+    })
+  })
+
+  for (const example of EXAMPLES) {
+    describe(example, function () {
+      let size, snapshot
+      before('load snapshot', async function () {
+        const ctx = await loadContext(example)
+        size = ctx.size
+        snapshot = ctx.snapshot
+      })
+
+      before('back fill new snapshot', async function () {
+        if (snapshot === null) {
+          console.error('back filling snapshot for', example)
+          snapshot = await backFillSnapshot(example, size)
+        }
+      })
+
+      it('should produce the expected xRef table', async function () {
+        const table = await parseXrefTable(pdfPath(example), size)
+        expect(table).to.deep.equal(snapshot)
+      })
+    })
+  }
+})

services/clsi/test/unit/lib/snapshots/asymptote/XrefTable.json

@@ -0,0 +1,356 @@
+[
+  {
+    "offset": 0,
+    "gen": 0,
+    "free": true
+  },
+  {
+    "offset": 123086,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 123405,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 216,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1084,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1244,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 4001,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 4155,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 4297,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 4932,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5307,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5495,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 30246,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 31466,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 38398,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 39039,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 40158,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 40897,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 65550,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 74691,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 81693,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 97169,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 104103,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 111180,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 118555,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 0
+  },
+  {
+    "offset": 6,
+    "gen": 1
+  },
+  {
+    "offset": 6,
+    "gen": 2
+  },
+  {
+    "offset": 6,
+    "gen": 3
+  },
+  {
+    "offset": 6,
+    "gen": 4
+  },
+  {
+    "offset": 6,
+    "gen": 5
+  },
+  {
+    "offset": 6,
+    "gen": 6
+  },
+  {
+    "offset": 6,
+    "gen": 7
+  },
+  {
+    "offset": 6,
+    "gen": 8
+  },
+  {
+    "offset": 6,
+    "gen": 9
+  },
+  {
+    "offset": 6,
+    "gen": 10
+  },
+  {
+    "offset": 6,
+    "gen": 11
+  },
+  {
+    "offset": 6,
+    "gen": 12
+  },
+  {
+    "offset": 6,
+    "gen": 13
+  },
+  {
+    "offset": 6,
+    "gen": 14
+  },
+  {
+    "offset": 6,
+    "gen": 15
+  },
+  {
+    "offset": 6,
+    "gen": 16
+  },
+  {
+    "offset": 6,
+    "gen": 17
+  },
+  {
+    "offset": 6,
+    "gen": 18
+  },
+  {
+    "offset": 6,
+    "gen": 19
+  },
+  {
+    "offset": 6,
+    "gen": 20
+  },
+  {
+    "offset": 6,
+    "gen": 21
+  },
+  {
+    "offset": 6,
+    "gen": 22
+  },
+  {
+    "offset": 6,
+    "gen": 23
+  },
+  {
+    "offset": 6,
+    "gen": 24
+  },
+  {
+    "offset": 6,
+    "gen": 25
+  },
+  {
+    "offset": 6,
+    "gen": 26
+  },
+  {
+    "offset": 6,
+    "gen": 27
+  },
+  {
+    "offset": 6,
+    "gen": 28
+  },
+  {
+    "offset": 6,
+    "gen": 29
+  },
+  {
+    "offset": 6,
+    "gen": 30
+  },
+  {
+    "offset": 6,
+    "gen": 31
+  },
+  {
+    "offset": 6,
+    "gen": 32
+  },
+  {
+    "offset": 6,
+    "gen": 33
+  },
+  {
+    "offset": 6,
+    "gen": 34
+  },
+  {
+    "offset": 6,
+    "gen": 35
+  },
+  {
+    "offset": 6,
+    "gen": 36
+  },
+  {
+    "offset": 6,
+    "gen": 37
+  },
+  {
+    "offset": 6,
+    "gen": 38
+  },
+  {
+    "offset": 6,
+    "gen": 39
+  },
+  {
+    "offset": 6,
+    "gen": 40
+  },
+  {
+    "offset": 6,
+    "gen": 41
+  },
+  {
+    "offset": 6,
+    "gen": 42
+  },
+  {
+    "offset": 6,
+    "gen": 43
+  },
+  {
+    "offset": 6,
+    "gen": 44
+  },
+  {
+    "offset": 6,
+    "gen": 45
+  },
+  {
+    "offset": 6,
+    "gen": 46
+  },
+  {
+    "offset": 6,
+    "gen": 47
+  },
+  {
+    "offset": 6,
+    "gen": 48
+  },
+  {
+    "offset": 6,
+    "gen": 49
+  },
+  {
+    "offset": 6,
+    "gen": 50
+  },
+  {
+    "offset": 6,
+    "gen": 51
+  },
+  {
+    "offset": 6,
+    "gen": 52
+  },
+  {
+    "offset": 6,
+    "gen": 53
+  },
+  {
+    "offset": 6,
+    "gen": 54
+  },
+  {
+    "offset": 6,
+    "gen": 55
+  }
+]

services/clsi/test/unit/lib/snapshots/biber_bibliography/XrefTable.json

@@ -0,0 +1,128 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 9,
+    "gen": 1
+  },
+  {
+    "offset": 9,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 9,
+    "gen": 14
+  },
+  {
+    "offset": 9,
+    "gen": 12
+  },
+  {
+    "offset": 9,
+    "gen": 15
+  },
+  {
+    "offset": 9,
+    "gen": 13
+  },
+  {
+    "offset": 9,
+    "gen": 16
+  },
+  {
+    "offset": 57274,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 9,
+    "gen": 17
+  },
+  {
+    "offset": 9,
+    "gen": 2
+  },
+  {
+    "offset": 9,
+    "gen": 3
+  },
+  {
+    "offset": 9,
+    "gen": 4
+  },
+  {
+    "offset": 9,
+    "gen": 5
+  },
+  {
+    "offset": 9,
+    "gen": 6
+  },
+  {
+    "offset": 522,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 9,
+    "gen": 7
+  },
+  {
+    "offset": 8788,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 9,
+    "gen": 8
+  },
+  {
+    "offset": 17289,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 9,
+    "gen": 9
+  },
+  {
+    "offset": 32619,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 9,
+    "gen": 10
+  },
+  {
+    "offset": 44596,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 9,
+    "gen": 11
+  },
+  {
+    "offset": 9,
+    "gen": 18
+  },
+  {
+    "offset": 57027,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 58655,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/epstopdf/XrefTable.json

@@ -0,0 +1,125 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 208,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 11
+  },
+  {
+    "offset": 6,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 14
+  },
+  {
+    "offset": 29580,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 15
+  },
+  {
+    "offset": 6,
+    "gen": 1
+  },
+  {
+    "offset": 6,
+    "gen": 2
+  },
+  {
+    "offset": 6,
+    "gen": 3
+  },
+  {
+    "offset": 6,
+    "gen": 4
+  },
+  {
+    "offset": 6,
+    "gen": 5
+  },
+  {
+    "offset": 6,
+    "gen": 6
+  },
+  {
+    "offset": 6,
+    "gen": 7
+  },
+  {
+    "offset": 6,
+    "gen": 8
+  },
+  {
+    "offset": 18310,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 9
+  },
+  {
+    "offset": 6,
+    "gen": 10
+  },
+  {
+    "offset": 18554,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 18945,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 19674,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 12
+  },
+  {
+    "offset": 22318,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 13
+  },
+  {
+    "offset": 6,
+    "gen": 16
+  },
+  {
+    "offset": 29321,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 30697,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/feynmf/XrefTable.json

@@ -0,0 +1,115 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 8,
+    "gen": 1
+  },
+  {
+    "offset": 8,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 5
+  },
+  {
+    "offset": 8,
+    "gen": 12
+  },
+  {
+    "offset": 8,
+    "gen": 14
+  },
+  {
+    "offset": 8,
+    "gen": 13
+  },
+  {
+    "offset": 25628,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 15
+  },
+  {
+    "offset": 8,
+    "gen": 2
+  },
+  {
+    "offset": 8,
+    "gen": 3
+  },
+  {
+    "offset": 8,
+    "gen": 4
+  },
+  {
+    "offset": 250,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 6
+  },
+  {
+    "offset": 8,
+    "gen": 7
+  },
+  {
+    "offset": 8,
+    "gen": 8
+  },
+  {
+    "offset": 3854,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 9
+  },
+  {
+    "offset": 11227,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 10
+  },
+  {
+    "offset": 18230,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 11
+  },
+  {
+    "offset": 8,
+    "gen": 16
+  },
+  {
+    "offset": 25381,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 26423,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/feynmp/XrefTable.json

@@ -0,0 +1,102 @@
+[
+  {
+    "offset": 0,
+    "gen": 65535,
+    "free": true
+  },
+  {
+    "offset": 1094,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5512,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1035,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 875,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 856,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1159,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1593,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 3149,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1436,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 2412,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1282,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1768,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1200,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1230,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 2006,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 2625,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 3381,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 4069,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/fontawesome/XrefTable.json

@@ -0,0 +1,93 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 7,
+    "gen": 1
+  },
+  {
+    "offset": 7,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 11
+  },
+  {
+    "offset": 7,
+    "gen": 10
+  },
+  {
+    "offset": 7,
+    "gen": 9
+  },
+  {
+    "offset": 29476,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 12
+  },
+  {
+    "offset": 7,
+    "gen": 7
+  },
+  {
+    "offset": 7,
+    "gen": 2
+  },
+  {
+    "offset": 7,
+    "gen": 8
+  },
+  {
+    "offset": 7,
+    "gen": 3
+  },
+  {
+    "offset": 7,
+    "gen": 4
+  },
+  {
+    "offset": 255,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 5
+  },
+  {
+    "offset": 17910,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 6
+  },
+  {
+    "offset": 7,
+    "gen": 13
+  },
+  {
+    "offset": 29228,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 30448,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/fontawesome_xelatex/XrefTable.json

@@ -0,0 +1,126 @@
+[
+  {
+    "offset": 0,
+    "gen": 0,
+    "free": true
+  },
+  {
+    "offset": 6338,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 216,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 707,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 757,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 888,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 991,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1257,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1678,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 2050,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 4246,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 4339,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5382,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5475,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5513,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 15,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 1
+  },
+  {
+    "offset": 15,
+    "gen": 2
+  },
+  {
+    "offset": 15,
+    "gen": 3
+  },
+  {
+    "offset": 15,
+    "gen": 4
+  },
+  {
+    "offset": 15,
+    "gen": 5
+  },
+  {
+    "offset": 15,
+    "gen": 6
+  },
+  {
+    "offset": 15,
+    "gen": 7
+  },
+  {
+    "offset": 15,
+    "gen": 8
+  },
+  {
+    "offset": 15,
+    "gen": 9
+  },
+  {
+    "offset": 15,
+    "gen": 10
+  }
+]

services/clsi/test/unit/lib/snapshots/glossaries/XrefTable.json

@@ -0,0 +1,94 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 7,
+    "gen": 1
+  },
+  {
+    "offset": 7,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 10
+  },
+  {
+    "offset": 7,
+    "gen": 9
+  },
+  {
+    "offset": 7,
+    "gen": 8
+  },
+  {
+    "offset": 33085,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 11
+  },
+  {
+    "offset": 7,
+    "gen": 2
+  },
+  {
+    "offset": 7,
+    "gen": 3
+  },
+  {
+    "offset": 7,
+    "gen": 4
+  },
+  {
+    "offset": 445,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 5
+  },
+  {
+    "offset": 10048,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 6
+  },
+  {
+    "offset": 18151,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 7
+  },
+  {
+    "offset": 7,
+    "gen": 12
+  },
+  {
+    "offset": 32838,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 34157,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/gnuplot/XrefTable.json

@@ -0,0 +1,89 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 4,
+    "gen": 0
+  },
+  {
+    "offset": 4,
+    "gen": 1
+  },
+  {
+    "offset": 4,
+    "gen": 2
+  },
+  {
+    "offset": 22047,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 4,
+    "gen": 4
+  },
+  {
+    "offset": 4,
+    "gen": 3
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 4,
+    "gen": 10
+  },
+  {
+    "offset": 4,
+    "gen": 9
+  },
+  {
+    "offset": 4,
+    "gen": 11
+  },
+  {
+    "offset": 4,
+    "gen": 5
+  },
+  {
+    "offset": 4,
+    "gen": 6
+  },
+  {
+    "offset": 6451,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 4,
+    "gen": 7
+  },
+  {
+    "offset": 14825,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 4,
+    "gen": 8
+  },
+  {
+    "offset": 4,
+    "gen": 12
+  },
+  {
+    "offset": 21800,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 22696,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/hebrew/XrefTable.json

@@ -0,0 +1,81 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 6,
+    "gen": 1
+  },
+  {
+    "offset": 6,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 8
+  },
+  {
+    "offset": 6,
+    "gen": 7
+  },
+  {
+    "offset": 22744,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 9
+  },
+  {
+    "offset": 6,
+    "gen": 2
+  },
+  {
+    "offset": 6,
+    "gen": 6
+  },
+  {
+    "offset": 6,
+    "gen": 3
+  },
+  {
+    "offset": 364,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 4
+  },
+  {
+    "offset": 12163,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 5
+  },
+  {
+    "offset": 6,
+    "gen": 10
+  },
+  {
+    "offset": 22496,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 23856,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/knitr/XrefTable.json

@@ -0,0 +1,145 @@
+[
+  {
+    "offset": 0,
+    "gen": 0,
+    "free": true
+  },
+  {
+    "offset": 43543,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 43792,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 216,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 734,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 784,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 912,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1020,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1544,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5791,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 12911,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 23655,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 30651,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 42597,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 14,
+    "gen": 0
+  },
+  {
+    "offset": 14,
+    "gen": 1
+  },
+  {
+    "offset": 14,
+    "gen": 2
+  },
+  {
+    "offset": 14,
+    "gen": 3
+  },
+  {
+    "offset": 14,
+    "gen": 4
+  },
+  {
+    "offset": 14,
+    "gen": 5
+  },
+  {
+    "offset": 14,
+    "gen": 6
+  },
+  {
+    "offset": 14,
+    "gen": 7
+  },
+  {
+    "offset": 14,
+    "gen": 8
+  },
+  {
+    "offset": 14,
+    "gen": 9
+  },
+  {
+    "offset": 14,
+    "gen": 10
+  },
+  {
+    "offset": 14,
+    "gen": 11
+  },
+  {
+    "offset": 14,
+    "gen": 12
+  },
+  {
+    "offset": 14,
+    "gen": 13
+  },
+  {
+    "offset": 14,
+    "gen": 14
+  },
+  {
+    "offset": 14,
+    "gen": 15
+  },
+  {
+    "offset": 14,
+    "gen": 16
+  }
+]

services/clsi/test/unit/lib/snapshots/knitr_utf8/XrefTable.json

@@ -0,0 +1,179 @@
+[
+  {
+    "offset": 0,
+    "gen": 0,
+    "free": true
+  },
+  {
+    "offset": 75291,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 75540,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 216,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 790,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 840,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 975,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 1083,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 2127,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 13797,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 23679,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 31863,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 36111,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 50346,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 61562,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 73508,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 16,
+    "gen": 0
+  },
+  {
+    "offset": 16,
+    "gen": 1
+  },
+  {
+    "offset": 16,
+    "gen": 2
+  },
+  {
+    "offset": 16,
+    "gen": 3
+  },
+  {
+    "offset": 16,
+    "gen": 4
+  },
+  {
+    "offset": 16,
+    "gen": 5
+  },
+  {
+    "offset": 16,
+    "gen": 6
+  },
+  {
+    "offset": 16,
+    "gen": 7
+  },
+  {
+    "offset": 16,
+    "gen": 8
+  },
+  {
+    "offset": 16,
+    "gen": 9
+  },
+  {
+    "offset": 16,
+    "gen": 10
+  },
+  {
+    "offset": 16,
+    "gen": 11
+  },
+  {
+    "offset": 16,
+    "gen": 12
+  },
+  {
+    "offset": 16,
+    "gen": 13
+  },
+  {
+    "offset": 16,
+    "gen": 14
+  },
+  {
+    "offset": 16,
+    "gen": 15
+  },
+  {
+    "offset": 16,
+    "gen": 16
+  },
+  {
+    "offset": 16,
+    "gen": 17
+  },
+  {
+    "offset": 16,
+    "gen": 18
+  },
+  {
+    "offset": 16,
+    "gen": 19
+  },
+  {
+    "offset": 16,
+    "gen": 20
+  },
+  {
+    "offset": 16,
+    "gen": 21
+  },
+  {
+    "offset": 16,
+    "gen": 22
+  }
+]

services/clsi/test/unit/lib/snapshots/latex_compiler/XrefTable.json

@@ -0,0 +1,132 @@
+[
+  {
+    "offset": 0,
+    "gen": 65535,
+    "free": true
+  },
+  {
+    "offset": 17349,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 25448,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 17290,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 17130,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 17109,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 17414,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 17548,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 18912,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 18381,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 22406,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 17897,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 21796,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 18758,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 23361,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 17455,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 17485,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 19218,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 22021,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 22638,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 23599,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 18051,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 18142,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 18657,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 24005,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/lualatex_compiler/XrefTable.json

@@ -0,0 +1,74 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 6,
+    "gen": 1
+  },
+  {
+    "offset": 6,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 4
+  },
+  {
+    "offset": 6,
+    "gen": 6
+  },
+  {
+    "offset": 2320,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 2
+  },
+  {
+    "offset": 202,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 298,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 3
+  },
+  {
+    "offset": 1642,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 5
+  },
+  {
+    "offset": 6,
+    "gen": 7
+  },
+  {
+    "offset": 2120,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 2892,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/makeindex-custom-style/XrefTable.json

@@ -0,0 +1,107 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 5,
+    "gen": 1
+  },
+  {
+    "offset": 5,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 12
+  },
+  {
+    "offset": 30470,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 13
+  },
+  {
+    "offset": 5,
+    "gen": 3
+  },
+  {
+    "offset": 5,
+    "gen": 2
+  },
+  {
+    "offset": 366,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 11
+  },
+  {
+    "offset": 5,
+    "gen": 10
+  },
+  {
+    "offset": 5,
+    "gen": 4
+  },
+  {
+    "offset": 5,
+    "gen": 5
+  },
+  {
+    "offset": 5,
+    "gen": 6
+  },
+  {
+    "offset": 638,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 7
+  },
+  {
+    "offset": 7838,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 8
+  },
+  {
+    "offset": 15674,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 9
+  },
+  {
+    "offset": 5,
+    "gen": 14
+  },
+  {
+    "offset": 30223,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 31457,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/makeindex/XrefTable.json

@@ -0,0 +1,90 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 5,
+    "gen": 1
+  },
+  {
+    "offset": 5,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 9
+  },
+  {
+    "offset": 23352,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 10
+  },
+  {
+    "offset": 5,
+    "gen": 3
+  },
+  {
+    "offset": 5,
+    "gen": 2
+  },
+  {
+    "offset": 366,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 8
+  },
+  {
+    "offset": 5,
+    "gen": 4
+  },
+  {
+    "offset": 5,
+    "gen": 5
+  },
+  {
+    "offset": 589,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 6
+  },
+  {
+    "offset": 8425,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 7
+  },
+  {
+    "offset": 5,
+    "gen": 11
+  },
+  {
+    "offset": 23105,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 24245,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/minted/XrefTable.json

@@ -0,0 +1,77 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 6,
+    "gen": 1
+  },
+  {
+    "offset": 6,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 7
+  },
+  {
+    "offset": 6,
+    "gen": 6
+  },
+  {
+    "offset": 19462,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 8
+  },
+  {
+    "offset": 6,
+    "gen": 2
+  },
+  {
+    "offset": 6,
+    "gen": 3
+  },
+  {
+    "offset": 340,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 4
+  },
+  {
+    "offset": 7343,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 5
+  },
+  {
+    "offset": 6,
+    "gen": 9
+  },
+  {
+    "offset": 19215,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 20089,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/multibib_bibliography/XrefTable.json

@@ -0,0 +1,133 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 6,
+    "gen": 1
+  },
+  {
+    "offset": 6,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 14
+  },
+  {
+    "offset": 6,
+    "gen": 15
+  },
+  {
+    "offset": 40591,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 17
+  },
+  {
+    "offset": 6,
+    "gen": 3
+  },
+  {
+    "offset": 6,
+    "gen": 2
+  },
+  {
+    "offset": 290,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 16
+  },
+  {
+    "offset": 6,
+    "gen": 5
+  },
+  {
+    "offset": 6,
+    "gen": 4
+  },
+  {
+    "offset": 595,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 7
+  },
+  {
+    "offset": 6,
+    "gen": 6
+  },
+  {
+    "offset": 844,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 8
+  },
+  {
+    "offset": 6,
+    "gen": 9
+  },
+  {
+    "offset": 6,
+    "gen": 10
+  },
+  {
+    "offset": 1107,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 11
+  },
+  {
+    "offset": 11816,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 12
+  },
+  {
+    "offset": 28180,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6,
+    "gen": 13
+  },
+  {
+    "offset": 6,
+    "gen": 18
+  },
+  {
+    "offset": 40344,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 41851,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/nomenclature/XrefTable.json

@@ -0,0 +1,94 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 7,
+    "gen": 1
+  },
+  {
+    "offset": 7,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 8
+  },
+  {
+    "offset": 7,
+    "gen": 9
+  },
+  {
+    "offset": 7,
+    "gen": 10
+  },
+  {
+    "offset": 32363,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 11
+  },
+  {
+    "offset": 7,
+    "gen": 2
+  },
+  {
+    "offset": 7,
+    "gen": 3
+  },
+  {
+    "offset": 7,
+    "gen": 4
+  },
+  {
+    "offset": 565,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 5
+  },
+  {
+    "offset": 10031,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 6
+  },
+  {
+    "offset": 18203,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 7
+  },
+  {
+    "offset": 7,
+    "gen": 12
+  },
+  {
+    "offset": 32116,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 33492,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/references_in_include/XrefTable.json

@@ -0,0 +1,90 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 5,
+    "gen": 1
+  },
+  {
+    "offset": 5,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 9
+  },
+  {
+    "offset": 15855,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 10
+  },
+  {
+    "offset": 5,
+    "gen": 3
+  },
+  {
+    "offset": 5,
+    "gen": 2
+  },
+  {
+    "offset": 167,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 8
+  },
+  {
+    "offset": 5,
+    "gen": 4
+  },
+  {
+    "offset": 5,
+    "gen": 5
+  },
+  {
+    "offset": 341,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 6
+  },
+  {
+    "offset": 7911,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 5,
+    "gen": 7
+  },
+  {
+    "offset": 5,
+    "gen": 11
+  },
+  {
+    "offset": 15608,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 16597,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/simple_bibliography/XrefTable.json

@@ -0,0 +1,94 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 7,
+    "gen": 1
+  },
+  {
+    "offset": 7,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 9
+  },
+  {
+    "offset": 7,
+    "gen": 8
+  },
+  {
+    "offset": 7,
+    "gen": 10
+  },
+  {
+    "offset": 35573,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 11
+  },
+  {
+    "offset": 7,
+    "gen": 2
+  },
+  {
+    "offset": 7,
+    "gen": 3
+  },
+  {
+    "offset": 7,
+    "gen": 4
+  },
+  {
+    "offset": 373,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 5
+  },
+  {
+    "offset": 8639,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 6
+  },
+  {
+    "offset": 23349,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7,
+    "gen": 7
+  },
+  {
+    "offset": 7,
+    "gen": 12
+  },
+  {
+    "offset": 35326,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 36668,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/subdirectories/XrefTable.json

@@ -0,0 +1,108 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 548,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 9,
+    "gen": 1
+  },
+  {
+    "offset": 9,
+    "gen": 2
+  },
+  {
+    "offset": 9,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 9,
+    "gen": 10
+  },
+  {
+    "offset": 9,
+    "gen": 9
+  },
+  {
+    "offset": 9,
+    "gen": 11
+  },
+  {
+    "offset": 46398,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 9,
+    "gen": 12
+  },
+  {
+    "offset": 8100,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 9,
+    "gen": 3
+  },
+  {
+    "offset": 9,
+    "gen": 4
+  },
+  {
+    "offset": 9,
+    "gen": 5
+  },
+  {
+    "offset": 9693,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 9,
+    "gen": 6
+  },
+  {
+    "offset": 17959,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 9,
+    "gen": 7
+  },
+  {
+    "offset": 34174,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 9,
+    "gen": 8
+  },
+  {
+    "offset": 9,
+    "gen": 13
+  },
+  {
+    "offset": 46151,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 47562,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/tikz_feynman/XrefTable.json

@@ -0,0 +1,145 @@
+[
+  {
+    "offset": 0,
+    "gen": 255,
+    "free": true
+  },
+  {
+    "offset": 8,
+    "gen": 8
+  },
+  {
+    "offset": 8,
+    "gen": 9
+  },
+  {
+    "offset": 8,
+    "gen": 10
+  },
+  {
+    "offset": 8,
+    "gen": 1
+  },
+  {
+    "offset": 8,
+    "gen": 0
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 20
+  },
+  {
+    "offset": 33577,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 3
+  },
+  {
+    "offset": 8,
+    "gen": 2
+  },
+  {
+    "offset": 1151,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 17
+  },
+  {
+    "offset": 8,
+    "gen": 19
+  },
+  {
+    "offset": 8,
+    "gen": 18
+  },
+  {
+    "offset": 8,
+    "gen": 5
+  },
+  {
+    "offset": 8,
+    "gen": 4
+  },
+  {
+    "offset": 2721,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 7
+  },
+  {
+    "offset": 8,
+    "gen": 6
+  },
+  {
+    "offset": 5757,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 11
+  },
+  {
+    "offset": 8,
+    "gen": 12
+  },
+  {
+    "offset": 8,
+    "gen": 13
+  },
+  {
+    "offset": 9558,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 14
+  },
+  {
+    "offset": 18967,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 15
+  },
+  {
+    "offset": 26388,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 16
+  },
+  {
+    "offset": 8,
+    "gen": 21
+  },
+  {
+    "offset": 33354,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 34451,
+    "gen": 0,
+    "uncompressed": true
+  }
+]

services/clsi/test/unit/lib/snapshots/xelatex_compiler/XrefTable.json

@@ -0,0 +1,73 @@
+[
+  {
+    "offset": 0,
+    "gen": 65535,
+    "free": true
+  },
+  {
+    "offset": 8,
+    "gen": 4
+  },
+  {
+    "offset": 8,
+    "gen": 3
+  },
+  {
+    "offset": 8,
+    "gen": 1
+  },
+  {
+    "offset": 8,
+    "gen": 6
+  },
+  {
+    "offset": 8,
+    "gen": 8
+  },
+  {
+    "offset": 15,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 0
+  },
+  {
+    "offset": 6857,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 2
+  },
+  {
+    "offset": 265,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 8,
+    "gen": 7
+  },
+  {
+    "offset": 8,
+    "gen": 5
+  },
+  {
+    "offset": 701,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 6750,
+    "gen": 0,
+    "uncompressed": true
+  },
+  {
+    "offset": 7655,
+    "gen": 0,
+    "uncompressed": true
+  }
+]