Merge pull request #716 from jmt-gh/abs_overdrive

Add support for leveraging chapter data directly from Overdrive mp3s during scanning

.gitignore

@@ -11,6 +11,7 @@ test/
 /client/.nuxt/
 /client/dist/
 /dist/
+library/
 
 sw.*
-.DS_STORE
+.DS_STORE

client/pages/config/index.vue

@@ -103,6 +103,16 @@
         </ui-tooltip>
       </div>
 
+      <div class="flex items-center py-2">
+        <ui-toggle-switch v-model="newServerSettings.scannerPreferOverdriveMediaMarker" :disabled="updatingServerSettings" @input="(val) => updateSettingsKey('scannerPreferOverdriveMediaMarker', val)" />
+        <ui-tooltip :text="tooltips.scannerPreferOverdriveMediaMarker">
+          <p class="pl-4 text-lg">
+            Scanner prefer Overdrive Media Markers for chapters
+            <span class="material-icons icon-text">info_outlined</span>
+          </p>
+        </ui-tooltip>
+      </div>
+
       <div class="flex items-center py-2">
         <ui-toggle-switch v-model="newServerSettings.scannerPreferOpfMetadata" :disabled="updatingServerSettings" @input="(val) => updateSettingsKey('scannerPreferOpfMetadata', val)" />
         <ui-tooltip :text="tooltips.scannerPreferOpfMetadata">
@@ -245,7 +255,8 @@ export default {
         storeCoverWithItem: 'By default covers are stored in /metadata/items, enabling this setting will store covers in your library item folder. Only one file named "cover" will be kept',
         storeMetadataWithItem: 'By default metadata files are stored in /metadata/items, enabling this setting will store metadata files in your library item folders. Uses .abs file extension',
         coverAspectRatio: 'Prefer to use square covers over standard 1.6:1 book covers',
-        enableEReader: 'E-reader is still a work in progress, but use this setting to open it up to all your users (or use the "Experimental Features" toggle below just for you)'
+        enableEReader: 'E-reader is still a work in progress, but use this setting to open it up to all your users (or use the "Experimental Features" toggle below just for you)',
+        scannerPreferOverdriveMediaMarker: 'MP3 files from Overdrive come with chapter timings embedded as custom metadata. Enabling this will use these tags for chapter timings automatically'
       },
       showConfirmPurgeCache: false
     }

server/controllers/PodcastController.js

@@ -103,7 +103,7 @@ class PodcastController {
         Logger.error('Invalid podcast feed request response')
         return res.status(500).send('Bad response from feed request')
       }
-      Logger.debug(`[PdocastController] Podcast feed size ${(data.data.length / 1024 / 1024).toFixed(2)}MB`)
+      Logger.debug(`[PodcastController] Podcast feed size ${(data.data.length / 1024 / 1024).toFixed(2)}MB`)
       var payload = await parsePodcastRssFeedXml(data.data, false, includeRaw)
       if (!payload) {
         return res.status(500).send('Invalid podcast RSS feed')

server/objects/mediaTypes/Book.js

@@ -3,6 +3,7 @@ const Logger = require('../../Logger')
 const BookMetadata = require('../metadata/BookMetadata')
 const { areEquivalent, copyValue } = require('../../utils/index')
 const { parseOpfMetadataXML } = require('../../utils/parsers/parseOpfMetadata')
+const { overdriveMediaMarkersExist, parseOverdriveMediaMarkersAsChapters } = require('../../utils/parsers/parseOverdriveMediaMarkers')
 const abmetadataGenerator = require('../../utils/abmetadataGenerator')
 const { readTextFile } = require('../../utils/fileUtils')
 const AudioFile = require('../files/AudioFile')
@@ -360,10 +361,11 @@ class Book {
     this.rebuildTracks()
   }
 
-  rebuildTracks() {
+  rebuildTracks(preferOverdriveMediaMarker) {
+    Logger.debug(`[Book] Tracks being rebuilt...!`)
     this.audioFiles.sort((a, b) => a.index - b.index)
     this.missingParts = []
-    this.setChapters()
+    this.setChapters(preferOverdriveMediaMarker)
     this.checkUpdateMissingTracks()
   }
 
@@ -395,9 +397,16 @@ class Book {
     return wasUpdated
   }
 
-  setChapters() {
+  setChapters(preferOverdriveMediaMarker = false) {
     // If 1 audio file without chapters, then no chapters will be set
     var includedAudioFiles = this.audioFiles.filter(af => !af.exclude)
+
+    // If overdrive media markers are present and preferred, use those instead
+    if (preferOverdriveMediaMarker && overdriveMediaMarkersExist(includedAudioFiles)) {
+      Logger.info('[Book] Overdrive Media Markers and preference found! Using these for chapter definitions')
+      return this.chapters = parseOverdriveMediaMarkersAsChapters(includedAudioFiles)
+    }
+
     if (includedAudioFiles.length === 1) {
       // 1 audio file with chapters
       if (includedAudioFiles[0].chapters) {

server/objects/metadata/AudioMetaTags.js

@@ -20,6 +20,7 @@ class AudioMetaTags {
     this.tagIsbn = null
     this.tagLanguage = null
     this.tagASIN = null
+    this.tagOverdriveMediaMarker = null
 
     if (metadata) {
       this.construct(metadata)
@@ -58,6 +59,7 @@ class AudioMetaTags {
     this.tagIsbn = metadata.tagIsbn || null
     this.tagLanguage = metadata.tagLanguage || null
     this.tagASIN = metadata.tagASIN || null
+    this.tagOverdriveMediaMarker = metadata.tagOverdriveMediaMarker || null
   }
 
   // Data parsed in prober.js
@@ -82,6 +84,7 @@ class AudioMetaTags {
     this.tagIsbn = payload.file_tag_isbn || null
     this.tagLanguage = payload.file_tag_language || null
     this.tagASIN = payload.file_tag_asin || null
+    this.tagOverdriveMediaMarker = payload.file_tag_overdrive_media_marker || null
   }
 
   updateData(payload) {
@@ -105,7 +108,8 @@ class AudioMetaTags {
       tagEncodedBy: payload.file_tag_encodedby || null,
       tagIsbn: payload.file_tag_isbn || null,
       tagLanguage: payload.file_tag_language || null,
-      tagASIN: payload.file_tag_asin || null
+      tagASIN: payload.file_tag_asin || null,
+      tagOverdriveMediaMarker: payload.file_tag_overdrive_media_marker || null,
     }
 
     var hasUpdates = false

server/objects/metadata/BookMetadata.js

@@ -262,6 +262,10 @@ class BookMetadata {
       {
         tag: 'tagASIN',
         key: 'asin'
+      },
+      {
+        tag: 'tagOverdriveMediaMarker',
+        key: 'overdriveMediaMarker'
       }
     ]
 

server/objects/settings/ServerSettings.js

@@ -13,6 +13,7 @@ class ServerSettings {
     this.scannerPreferOpfMetadata = false
     this.scannerPreferMatchedMetadata = false
     this.scannerDisableWatcher = false 
+    this.scannerPreferOverdriveMediaMarker = false
 
     // Metadata - choose to store inside users library item folder
     this.storeCoverWithItem = false
@@ -65,6 +66,7 @@ class ServerSettings {
     this.scannerPreferOpfMetadata = !!settings.scannerPreferOpfMetadata
     this.scannerPreferMatchedMetadata = !!settings.scannerPreferMatchedMetadata
     this.scannerDisableWatcher = !!settings.scannerDisableWatcher
+    this.scannerPreferOverdriveMediaMarker = !!settings.scannerPreferOverdriveMediaMarker
 
     this.storeCoverWithItem = !!settings.storeCoverWithItem
     if (settings.storeCoverWithBook != undefined) { // storeCoverWithBook was old name of setting < v2
@@ -111,6 +113,7 @@ class ServerSettings {
       scannerPreferOpfMetadata: this.scannerPreferOpfMetadata,
       scannerPreferMatchedMetadata: this.scannerPreferMatchedMetadata,
       scannerDisableWatcher: this.scannerDisableWatcher,
+      scannerPreferOverdriveMediaMarker: this.scannerPreferOverdriveMediaMarker,
       storeCoverWithItem: this.storeCoverWithItem,
       storeMetadataWithItem: this.storeMetadataWithItem,
       rateLimitLoginRequests: this.rateLimitLoginRequests,

server/scanner/LibraryScan.js

@@ -34,6 +34,7 @@ class LibraryScan {
   get forceRescan() { return !!this._scanOptions.forceRescan }
   get preferAudioMetadata() { return !!this._scanOptions.preferAudioMetadata }
   get preferOpfMetadata() { return !!this._scanOptions.preferOpfMetadata }
+  get preferOverdriveMediaMarker() { return !!this._scanOptions.preferOverdriveMediaMarker }
   get findCovers() { return !!this._scanOptions.findCovers }
   get timestamp() {
     return (new Date()).toISOString()

server/scanner/MediaFileScanner.js

@@ -195,7 +195,7 @@ class MediaFileScanner {
     }
   }
 
-  async scanMediaFiles(mediaLibraryFiles, scanData, libraryItem, preferAudioMetadata, libraryScan = null) {
+  async scanMediaFiles(mediaLibraryFiles, scanData, libraryItem, preferAudioMetadata, preferOverdriveMediaMarker, libraryScan = null) {
     var hasUpdated = false
 
     var mediaScanResult = await this.executeMediaFileScans(libraryItem.mediaType, mediaLibraryFiles, scanData)
@@ -208,6 +208,7 @@ class MediaFileScanner {
     } else if (mediaScanResult.audioFiles.length) {
       if (libraryScan) {
         libraryScan.addLog(LogLevel.DEBUG, `Library Item "${scanData.path}" Audio file scan took ${mediaScanResult.elapsed}ms for ${mediaScanResult.audioFiles.length} with average time of ${mediaScanResult.averageScanDuration}ms`)
+        Logger.debug(`Library Item "${scanData.path}" Audio file scan took ${mediaScanResult.elapsed}ms for ${mediaScanResult.audioFiles.length} with average time of ${mediaScanResult.averageScanDuration}ms`)
       }
 
       var totalAudioFilesToInclude = mediaScanResult.audioFiles.length
@@ -247,7 +248,7 @@ class MediaFileScanner {
         }
 
         if (hasUpdated) {
-          libraryItem.media.rebuildTracks()
+          libraryItem.media.rebuildTracks(preferOverdriveMediaMarker)
         }
       } else { // Podcast Media Type
         var existingAudioFiles = mediaScanResult.audioFiles.filter(af => libraryItem.media.findFileWithInode(af.ino))

server/scanner/ScanOptions.js

@@ -9,6 +9,7 @@ class ScanOptions {
     this.preferAudioMetadata = false
     this.preferOpfMetadata = false
     this.preferMatchedMetadata = false
+    this.preferOverdriveMediaMarker = false
 
     if (options) {
       this.construct(options)
@@ -34,7 +35,8 @@ class ScanOptions {
       storeCoverWithItem: this.storeCoverWithItem,
       preferAudioMetadata: this.preferAudioMetadata,
       preferOpfMetadata: this.preferOpfMetadata,
-      preferMatchedMetadata: this.preferMatchedMetadata
+      preferMatchedMetadata: this.preferMatchedMetadata,
+      preferOverdriveMediaMarker: this.preferOverdriveMediaMarker
     }
   }
 
@@ -47,6 +49,7 @@ class ScanOptions {
     this.preferAudioMetadata = serverSettings.scannerPreferAudioMetadata
     this.preferOpfMetadata = serverSettings.scannerPreferOpfMetadata
     this.scannerPreferMatchedMetadata = serverSettings.scannerPreferMatchedMetadata
+    this.preferOverdriveMediaMarker = serverSettings.scannerPreferOverdriveMediaMarker
   }
 }
 module.exports = ScanOptions

server/scanner/Scanner.js

@@ -80,7 +80,7 @@ class Scanner {
     // Scan all audio files
     if (libraryItem.hasAudioFiles) {
       var libraryAudioFiles = libraryItem.libraryFiles.filter(lf => lf.fileType === 'audio')
-      if (await MediaFileScanner.scanMediaFiles(libraryAudioFiles, libraryItemData, libraryItem, this.db.serverSettings.scannerPreferAudioMetadata)) {
+      if (await MediaFileScanner.scanMediaFiles(libraryAudioFiles, libraryItemData, libraryItem, this.db.serverSettings.scannerPreferAudioMetadata, this.db.serverSettings.scannerPreferOverdriveMediaMarker)) {
         hasUpdated = true
       }
 
@@ -310,7 +310,7 @@ class Scanner {
 
   async scanNewLibraryItemDataChunk(newLibraryItemsData, libraryScan) {
     var newLibraryItems = await Promise.all(newLibraryItemsData.map((lid) => {
-      return this.scanNewLibraryItem(lid, libraryScan.libraryMediaType, libraryScan.preferAudioMetadata, libraryScan.preferOpfMetadata, libraryScan.findCovers, libraryScan)
+      return this.scanNewLibraryItem(lid, libraryScan.libraryMediaType, libraryScan.preferAudioMetadata, libraryScan.preferOpfMetadata, libraryScan.findCovers, libraryScan.preferOverdriveMediaMarker, libraryScan)
     }))
     newLibraryItems = newLibraryItems.filter(li => li) // Filter out nulls
 
@@ -337,7 +337,7 @@ class Scanner {
     // forceRescan all existing audio files - will probe and update ID3 tag metadata
     var existingAudioFiles = existingLibraryFiles.filter(lf => lf.fileType === 'audio')
     if (libraryScan.scanOptions.forceRescan && existingAudioFiles.length) {
-      if (await MediaFileScanner.scanMediaFiles(existingAudioFiles, scanData, libraryItem, libraryScan.preferAudioMetadata, libraryScan)) {
+      if (await MediaFileScanner.scanMediaFiles(existingAudioFiles, scanData, libraryItem, libraryScan.preferAudioMetadata, libraryScan.preferOverdriveMediaMarker, libraryScan)) {
         hasUpdated = true
       }
     }
@@ -345,7 +345,7 @@ class Scanner {
     var newAudioFiles = newLibraryFiles.filter(lf => lf.fileType === 'audio')
     var removedAudioFiles = filesRemoved.filter(lf => lf.fileType === 'audio')
     if (newAudioFiles.length || removedAudioFiles.length) {
-      if (await MediaFileScanner.scanMediaFiles(newAudioFiles, scanData, libraryItem, libraryScan.preferAudioMetadata, libraryScan)) {
+      if (await MediaFileScanner.scanMediaFiles(newAudioFiles, scanData, libraryItem, libraryScan.preferAudioMetadata, libraryScan.preferOverdriveMediaMarker, libraryScan)) {
         hasUpdated = true
       }
     }
@@ -379,7 +379,7 @@ class Scanner {
     return hasUpdated ? libraryItem : null
   }
 
-  async scanNewLibraryItem(libraryItemData, libraryMediaType, preferAudioMetadata, preferOpfMetadata, findCovers, libraryScan = null) {
+  async scanNewLibraryItem(libraryItemData, libraryMediaType, preferAudioMetadata, preferOpfMetadata, findCovers, preferOverdriveMediaMarker, libraryScan = null) {
     if (libraryScan) libraryScan.addLog(LogLevel.DEBUG, `Scanning new library item "${libraryItemData.path}"`)
     else Logger.debug(`[Scanner] Scanning new item "${libraryItemData.path}"`)
 
@@ -388,7 +388,7 @@ class Scanner {
 
     var mediaFiles = libraryItemData.libraryFiles.filter(lf => lf.fileType === 'audio' || lf.fileType === 'video')
     if (mediaFiles.length) {
-      await MediaFileScanner.scanMediaFiles(mediaFiles, libraryItemData, libraryItem, preferAudioMetadata, libraryScan)
+      await MediaFileScanner.scanMediaFiles(mediaFiles, libraryItemData, libraryItem, preferAudioMetadata, preferOverdriveMediaMarker, libraryScan)
     }
 
     await libraryItem.syncFiles(preferOpfMetadata)
@@ -608,7 +608,7 @@ class Scanner {
     var libraryItemData = await getLibraryItemFileData(libraryMediaType, folder, fullPath, isSingleMediaItem, this.db.serverSettings)
     if (!libraryItemData) return null
     var serverSettings = this.db.serverSettings
-    return this.scanNewLibraryItem(libraryItemData, libraryMediaType, serverSettings.scannerPreferAudioMetadata, serverSettings.scannerPreferOpfMetadata, serverSettings.scannerFindCovers)
+    return this.scanNewLibraryItem(libraryItemData, libraryMediaType, serverSettings.scannerPreferAudioMetadata, serverSettings.scannerPreferOpfMetadata, serverSettings.scannerFindCovers, serverSettings.scannerPreferOverdriveMediaMarker)
   }
 
   async searchForCover(libraryItem, libraryScan = null) {

server/utils/parsers/parseOverdriveMediaMarkers.js

@@ -0,0 +1,148 @@
+const Logger = require('../../Logger')
+
+// given a list of audio files, extract all of the Overdrive Media Markers metaTags, and return an array of them as XML
+function extractOverdriveMediaMarkers(includedAudioFiles) {
+    Logger.debug('[parseOverdriveMediaMarkers] Extracting overdrive media markers')
+    var markers = includedAudioFiles.map((af) => af.metaTags.tagOverdriveMediaMarker).filter(notUndefined => notUndefined !== undefined).filter(elem => { return elem !== null }) || []
+
+    return markers
+}
+
+// given the array of Overdrive Media Markers from generateOverdriveMediaMarkers()
+//  parse and clean them in to something a bit more usable
+function cleanOverdriveMediaMarkers(overdriveMediaMarkers) {
+    Logger.debug('[parseOverdriveMediaMarkers] Cleaning up overdrive media markers')
+    /*
+    returns an array of arrays of objects. Each inner array corresponds to an audio track, with it's objects being a chapter:
+    [
+      [
+       {
+        "Name": "Chapter 1",
+        "Time": "0:00.000"
+      },
+      {
+        "Name": "Chapter 2",
+        "Time": "15:51.000"
+      },
+      { etc }
+      ]
+    ]
+    */
+
+    var parseString = require('xml2js').parseString; // function to convert xml to JSON
+    var parsedOverdriveMediaMarkers = []
+
+    overdriveMediaMarkers.forEach(function (item, index) {
+        var parsed_result
+        parseString(item, function (err, result) {
+            /*  
+            result.Markers.Marker is the result of parsing the XML for the MediaMarker tags for the MP3 file (Part##.mp3)
+            it is shaped like this, and needs further cleaning below:
+            [
+                {
+                    "Name": [
+                        "Chapter 1:  "
+                    ],
+                    "Time": [
+                        "0:00.000"
+                    ]
+                },
+                { 
+                    ANOTHER CHAPTER
+                },
+            ]
+            */
+
+            // The values for Name and Time in results.Markers.Marker are returned as Arrays from parseString and should be strings
+            parsed_result = objectValuesArrayToString(result.Markers.Marker)
+        })
+
+        parsedOverdriveMediaMarkers.push(parsed_result)
+    })
+
+    return removeExtraChapters(parsedOverdriveMediaMarkers)
+}
+
+// given an array of objects, convert any values that are arrays to strings
+function objectValuesArrayToString(arrayOfObjects) {
+    Logger.debug('[parseOverdriveMediaMarkers] Converting Marker object values from arrays to strings')
+    arrayOfObjects.forEach((item) => {
+        Object.keys(item).forEach(key => {
+            item[key] = item[key].toString()
+        })
+    })
+
+    return arrayOfObjects
+}
+
+// Overdrive sometimes has weird chapters and subchapters defined
+//  These aren't necessary, so lets remove them
+function removeExtraChapters(parsedOverdriveMediaMarkers) {
+    Logger.debug('[parseOverdriveMediaMarkers] Removing any unnecessary chapters')
+    const weirdChapterFilterRegex = /([(]\d|[cC]ontinued)/
+    var cleaned = []
+    parsedOverdriveMediaMarkers.forEach(function (item) {
+        cleaned.push(item.filter(chapter => !weirdChapterFilterRegex.test(chapter.Name)))
+    })
+
+    return cleaned
+}
+
+// Given a set of chapters from generateParsedChapters, add the end time to each one
+function addChapterEndTimes(chapters, totalAudioDuration) {
+    Logger.debug('[parseOverdriveMediaMarkers] Adding chapter end times')
+    chapters.forEach((chapter, chapter_index) => {
+        if (chapter_index < chapters.length - 1) {
+            chapter.end = chapters[chapter_index + 1].start
+        } else {
+            chapter.end = totalAudioDuration
+        }
+    })
+
+    return chapters
+}
+
+// The function that actually generates the Chapters object that we update ABS with
+function generateParsedChapters(includedAudioFiles, cleanedOverdriveMediaMarkers) {
+    Logger.debug('[parseOverdriveMediaMarkers] Generating new chapters for ABS')
+    // logic ported over from benonymity's OverdriveChapterizer:
+    //    https://github.com/benonymity/OverdriveChapterizer/blob/main/chapters.py
+    var parsedChapters = []
+    var length = 0.0
+    var index = 0
+    var time = 0.0
+
+    // cleanedOverdriveMediaMarkers is an array of array of objects, where the inner array matches to the included audio files tracks
+    //     this allows us to leverage the individual track durations when calculating the start times of chapters in tracks after the first (using length)
+    includedAudioFiles.forEach((track, track_index) => {
+        cleanedOverdriveMediaMarkers[track_index].forEach((chapter) => {
+            time = chapter.Time.split(":")
+            time = length + parseFloat(time[0]) * 60 + parseFloat(time[1])
+            var newChapterData = {
+                id: index++,
+                start: time,
+                title: chapter.Name
+            }
+            parsedChapters.push(newChapterData)
+        })
+        length += track.duration
+    })
+
+    parsedChapters = addChapterEndTimes(parsedChapters, length) // we need all the start times sorted out before we can add the end times
+
+    return parsedChapters
+}
+
+module.exports.overdriveMediaMarkersExist = (includedAudioFiles) => {
+    return extractOverdriveMediaMarkers(includedAudioFiles).length > 1
+}
+
+module.exports.parseOverdriveMediaMarkersAsChapters = (includedAudioFiles) => {
+    Logger.info('[parseOverdriveMediaMarkers] Parsing of Overdrive Media Markers started')
+
+    var overdriveMediaMarkers = extractOverdriveMediaMarkers(includedAudioFiles)
+    var cleanedOverdriveMediaMarkers = cleanOverdriveMediaMarkers(overdriveMediaMarkers)
+    var parsedChapters = generateParsedChapters(includedAudioFiles, cleanedOverdriveMediaMarkers)
+
+    return parsedChapters
+}

server/utils/prober.js

@@ -192,6 +192,7 @@ function parseTags(format, verbose) {
     file_tag_movement: tryGrabTags(format, 'movement', 'mvin'),
     file_tag_genre1: tryGrabTags(format, 'tmp_genre1', 'genre1'),
     file_tag_genre2: tryGrabTags(format, 'tmp_genre2', 'genre2'),
+    file_tag_overdrive_media_marker: tryGrabTags(format, 'OverDrive MediaMarkers'),
   }
   for (const key in tags) {
     if (!tags[key]) {