feat: HLS / DASH support forced subtitle

docs/source/options/dash_stream_descriptors.rst

@@ -12,5 +12,5 @@ DASH specific stream descriptor fields
 
     Optional semicolon separated list of values for DASH Role element. The
     value should be one of: **caption**, **subtitle**, **main**, **alternate**,
-    **supplementary**, **commentary**, **description** and **dub**. See 
-    DASH (ISO/IEC 23009-1) specification for details.
+    **supplementary**, **commentary**, **description**, **dub** and **forced-subtitle** .
+    See DASH (ISO/IEC 23009-1) specification for details.

docs/source/options/stream_descriptors.rst

@@ -73,6 +73,15 @@ These are the available fields:
     CEA allows specifying up to 4 streams within a single video stream. If not
     specified, all subtitles will be merged together.
 
+:forced_subtitle:
+
+    Optional boolean value (0|1). If set to 1 indicates that this stream is a
+    Forced Narrative subtitle that should be displayed when subtitles are otherwise
+    off, for example used to caption short portions of the audio that might be in
+    a foreign language. For DASH this will set role to **forced_subtitle**, for HLS
+    it will set FORCED=YES and AUTOSELECT=YES. Only valid for subtitles.
+
+
 .. include:: /options/drm_stream_descriptors.rst
 .. include:: /options/dash_stream_descriptors.rst
 .. include:: /options/hls_stream_descriptors.rst

include/packager/packager.h

@@ -152,6 +152,9 @@ struct StreamDescriptor {
   /// Set to true to indicate that the stream is for hls only.
   bool hls_only = false;
 
+  /// Optional, indicates if this is a Forced Narrative subtitle stream.
+  bool forced_subtitle = false;
+
   /// Optional for DASH output. It defines the Label element in Adaptation Set.
   std::string dash_label;
 };

packager/app/packager_main.cc

@@ -120,8 +120,16 @@ const char kUsage[] =
     "    in the format: scheme_id_uri=value.\n"
     "  - dash_roles (roles): Optional semicolon separated list of values for\n"
     "    DASH Role elements. The value should be one of: caption, subtitle,\n"
-    "    main, alternate, supplementary, commentary, description and dub. See\n"
-    "    DASH (ISO/IEC 23009-1) specification for details.\n";
+    "    forced-subtitle, main, alternate, supplementary, commentary, \n"
+    "    description and dub. See DASH\n"
+    "    (ISO/IEC 23009-1) specification for details.\n"
+    "  - forced_subtitle: Optional boolean value (0|1). If set to 1 \n"
+    "    indicates that this stream is a Forced Narrative subtitle that \n"
+    "    should be displayed when subtitles are otherwise off, for example \n"
+    "    used to caption short portions of the audio that might be in a \n"
+    "    foreign language. For DASH this will set role to forced_subtitle, \n"
+    "    for HLS it will set FORCED=YES and AUTOSELECT=YES. \n"
+    "    Only valid for subtitles.";
 
 // Labels for parameters in RawKey key info.
 const char kDrmLabelLabel[] = "label";

packager/app/stream_descriptor.cc

@@ -40,6 +40,7 @@ enum FieldType {
   kDashOnlyField,
   kHlsOnlyField,
   kDashLabelField,
+  kForcedSubtitleField,
 };
 
 struct FieldNameToTypeMapping {
@@ -88,6 +89,7 @@ const FieldNameToTypeMapping kFieldNameTypeMappings[] = {
     {"dash_only", kDashOnlyField},
     {"hls_only", kHlsOnlyField},
     {"dash_label", kDashLabelField},
+    {"forced_subtitle", kForcedSubtitleField},
 };
 
 FieldType GetFieldType(const std::string& field_name) {
@@ -255,12 +257,35 @@ std::optional<StreamDescriptor> ParseStreamDescriptor(
       case kDashLabelField:
         descriptor.dash_label = pair.second;
         break;
+      case kForcedSubtitleField:
+        unsigned forced_subtitle_value;
+        if (!absl::SimpleAtoi(pair.second, &forced_subtitle_value)) {
+          LOG(ERROR) << "Non-numeric option for forced field "
+                        "specified ("
+                     << pair.second << ").";
+          return std::nullopt;
+        }
+        if (forced_subtitle_value > 1) {
+          LOG(ERROR) << "forced should be either 0 or 1.";
+          return std::nullopt;
+        }
+        descriptor.forced_subtitle = forced_subtitle_value > 0;
+        break;
       default:
         LOG(ERROR) << "Unknown field in stream descriptor (\"" << pair.first
                    << "\").";
         return std::nullopt;
     }
   }
+
+  if (descriptor.forced_subtitle) {
+    auto itr = std::find(descriptor.dash_roles.begin(),
+                         descriptor.dash_roles.end(), "forced-subtitle");
+    if (itr == descriptor.dash_roles.end()) {
+      descriptor.dash_roles.push_back("forced-subtitle");
+    }
+  }
+
   return descriptor;
 }
 

packager/app/test/packager_test.py

@@ -310,7 +310,8 @@ def _GetStream(self,
                  skip_encryption=None,
                  bandwidth=None,
                  split_content_on_ad_cues=False,
-                 test_file=None):
+                 test_file=None,
+                 forced_subtitle=None):
     """Get a stream descriptor as a string.
 
 
@@ -347,8 +348,9 @@ def _GetStream(self,
           into multiple files, with a total of NumAdCues + 1 files.
       test_file: The input file to use. If the input file is not specified, a
           default file will be used.
-
-
+      forced_subtitle: If set to true, it marks this as a Forced Narrative
+          subtitle, marked in DASH using forced-subtitle role and
+          in HLS using FORCED=YES.
     Returns:
       A string that makes up a single stream descriptor for input to the
       packager.
@@ -402,6 +404,9 @@ def _GetStream(self,
     if dash_only:
       stream.Append('dash_only', 1)
 
+    if forced_subtitle:
+      stream.Append('forced_subtitle', 1)
+
     if dash_label:
       stream.Append('dash_label', dash_label)
 
@@ -799,6 +804,21 @@ def testDashLabel(self):
     self.assertPackageSuccess(streams, self._GetFlags(output_dash=True))
     self._CheckTestResults('dash-label')
 
+  def testForcedSubtitle(self):
+    streams = [
+      self._GetStream('audio', hls=True),
+      self._GetStream('video', hls=True),
+    ]
+
+    streams += self._GetStreams(
+        ['text'],
+        test_files=['bear-english.vtt'],
+        forced_subtitle=True)
+
+    self.assertPackageSuccess(streams, self._GetFlags(output_dash=True,
+                                                      output_hls=True))
+    self._CheckTestResults('forced-subtitle')
+
   def testAudioVideoWithLanguageOverride(self):
     self.assertPackageSuccess(
         self._GetStreams(['audio', 'video'], language='por', hls=True),

packager/app/test/testdata/forced-subtitle/bear-640x360-audio.m3u8

@@ -0,0 +1,16 @@
+#EXTM3U
+#EXT-X-VERSION:6
+## Generated with https://github.com/shaka-project/shaka-packager version <tag>-<hash>-<test>
+#EXT-X-TARGETDURATION:5
+#EXT-X-PLAYLIST-TYPE:VOD
+#EXT-X-MAP:URI="bear-640x360-audio.mp4",BYTERANGE="804@0"
+#EXTINF:1.022,
+#EXT-X-BYTERANGE:17028@872
+bear-640x360-audio.mp4
+#EXTINF:0.998,
+#EXT-X-BYTERANGE:16285
+bear-640x360-audio.mp4
+#EXTINF:0.720,
+#EXT-X-BYTERANGE:9558
+bear-640x360-audio.mp4
+#EXT-X-ENDLIST

packager/app/test/testdata/forced-subtitle/bear-640x360-video-iframe.m3u8

@@ -0,0 +1,17 @@
+#EXTM3U
+#EXT-X-VERSION:6
+## Generated with https://github.com/shaka-project/shaka-packager version <tag>-<hash>-<test>
+#EXT-X-TARGETDURATION:5
+#EXT-X-PLAYLIST-TYPE:VOD
+#EXT-X-I-FRAMES-ONLY
+#EXT-X-MAP:URI="bear-640x360-video.mp4",BYTERANGE="870@0"
+#EXTINF:1.001,
+#EXT-X-BYTERANGE:15581@938
+bear-640x360-video.mp4
+#EXTINF:1.001,
+#EXT-X-BYTERANGE:18221@100251
+bear-640x360-video.mp4
+#EXTINF:0.734,
+#EXT-X-BYTERANGE:19663@222058
+bear-640x360-video.mp4
+#EXT-X-ENDLIST

packager/app/test/testdata/forced-subtitle/bear-640x360-video.m3u8

@@ -0,0 +1,16 @@
+#EXTM3U
+#EXT-X-VERSION:6
+## Generated with https://github.com/shaka-project/shaka-packager version <tag>-<hash>-<test>
+#EXT-X-TARGETDURATION:5
+#EXT-X-PLAYLIST-TYPE:VOD
+#EXT-X-MAP:URI="bear-640x360-video.mp4",BYTERANGE="870@0"
+#EXTINF:1.001,
+#EXT-X-BYTERANGE:99313@938
+bear-640x360-video.mp4
+#EXTINF:1.001,
+#EXT-X-BYTERANGE:121807
+bear-640x360-video.mp4
+#EXTINF:0.734,
+#EXT-X-BYTERANGE:79662
+bear-640x360-video.mp4
+#EXT-X-ENDLIST

packager/app/test/testdata/forced-subtitle/bear-english-text.vtt

@@ -0,0 +1,11 @@
+WEBVTT
+
+STYLE
+::cue { color:lime }
+
+00:00:00.000 --> 00:00:00.800 align:center
+Yup, that's a bear, eh.
+
+00:00:01.000 --> 00:00:04.700 align:center
+He 's... um... doing bear-like stuff.
+

packager/app/test/testdata/forced-subtitle/output.m3u8

@@ -0,0 +1,13 @@
+#EXTM3U
+## Generated with https://github.com/shaka-project/shaka-packager version <tag>-<hash>-<test>
+
+#EXT-X-INDEPENDENT-SEGMENTS
+
+#EXT-X-MEDIA:TYPE=AUDIO,URI="bear-640x360-audio.m3u8",GROUP-ID="default-audio-group",NAME="stream_0",DEFAULT=NO,AUTOSELECT=YES,CHANNELS="2"
+
+#EXT-X-MEDIA:TYPE=SUBTITLES,URI="stream_2.m3u8",GROUP-ID="default-text-group",NAME="stream_2",DEFAULT=NO,AUTOSELECT=YES,FORCED=YES
+
+#EXT-X-STREAM-INF:BANDWIDTH=1106817,AVERAGE-BANDWIDTH=1004632,CODECS="avc1.64001e,mp4a.40.2",RESOLUTION=640x360,FRAME-RATE=29.970,AUDIO="default-audio-group",SUBTITLES="default-text-group",CLOSED-CAPTIONS=NONE
+bear-640x360-video.m3u8
+
+#EXT-X-I-FRAME-STREAM-INF:BANDWIDTH=214292,AVERAGE-BANDWIDTH=156327,CODECS="avc1.64001e",RESOLUTION=640x360,CLOSED-CAPTIONS=NONE,URI="bear-640x360-video-iframe.m3u8"

packager/app/test/testdata/forced-subtitle/output.mpd

@@ -0,0 +1,29 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/shaka-project/shaka-packager version <tag>-<hash>-<test>-->
+<MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" profiles="urn:mpeg:dash:profile:isoff-on-demand:2011" minBufferTime="PT2S" type="static" mediaPresentationDuration="PT2.736067S">
+  <Period id="0">
+    <AdaptationSet id="0" contentType="audio" subsegmentAlignment="true">
+      <Representation id="0" bandwidth="133334" codecs="mp4a.40.2" mimeType="audio/mp4" audioSamplingRate="44100">
+        <AudioChannelConfiguration schemeIdUri="urn:mpeg:dash:23003:3:audio_channel_configuration:2011" value="2"/>
+        <BaseURL>bear-640x360-audio.mp4</BaseURL>
+        <SegmentBase indexRange="804-871" timescale="44100">
+          <Initialization range="0-803"/>
+        </SegmentBase>
+      </Representation>
+    </AdaptationSet>
+    <AdaptationSet id="1" contentType="video" width="640" height="360" frameRate="30000/1001" subsegmentAlignment="true" par="16:9">
+      <Representation id="1" bandwidth="973483" codecs="avc1.64001e" mimeType="video/mp4" sar="1:1">
+        <BaseURL>bear-640x360-video.mp4</BaseURL>
+        <SegmentBase indexRange="870-937" timescale="30000">
+          <Initialization range="0-869"/>
+        </SegmentBase>
+      </Representation>
+    </AdaptationSet>
+    <AdaptationSet id="2" contentType="text" subsegmentAlignment="true">
+      <Role schemeIdUri="urn:mpeg:dash:role:2011" value="forced-subtitle"/>
+      <Representation id="2" bandwidth="317" mimeType="text/vtt">
+        <BaseURL>bear-english-text.vtt</BaseURL>
+      </Representation>
+    </AdaptationSet>
+  </Period>
+</MPD>

packager/app/test/testdata/forced-subtitle/stream_2.m3u8

@@ -0,0 +1,8 @@
+#EXTM3U
+#EXT-X-VERSION:6
+## Generated with https://github.com/shaka-project/shaka-packager version <tag>-<hash>-<test>
+#EXT-X-TARGETDURATION:5
+#EXT-X-PLAYLIST-TYPE:VOD
+#EXTINF:4.700,
+bear-english-text.vtt
+#EXT-X-ENDLIST

packager/hls/base/master_playlist.cc

@@ -318,11 +318,16 @@ void BuildMediaTag(const MediaPlaylist& playlist,
   } else {
     tag.AddString("DEFAULT", "NO");
   }
-
   if (is_autoselect) {
     tag.AddString("AUTOSELECT", "YES");
   }
 
+  if (playlist.stream_type() ==
+          MediaPlaylist::MediaPlaylistStreamType::kSubtitle &&
+      playlist.forced_subtitle()) {
+    tag.AddString("FORCED", "YES");
+  }
+
   const std::vector<std::string>& characteristics = playlist.characteristics();
   if (!characteristics.empty()) {
     tag.AddQuotedString("CHARACTERISTICS", absl::StrJoin(characteristics, ","));
@@ -401,6 +406,12 @@ void BuildMediaTags(
         }
       }
 
+      if (playlist->stream_type() ==
+              MediaPlaylist::MediaPlaylistStreamType::kSubtitle &&
+          playlist->forced_subtitle()) {
+        is_autoselect = true;
+      }
+
       BuildMediaTag(*playlist, group_id, is_default, is_autoselect, base_url,
                     out);
     }

packager/hls/base/media_playlist.cc

@@ -373,6 +373,10 @@ void MediaPlaylist::SetCharacteristicsForTesting(
   characteristics_ = characteristics;
 }
 
+void MediaPlaylist::SetForcedSubtitleForTesting(const bool forced_subtitle) {
+  forced_subtitle_ = forced_subtitle;
+}
+
 bool MediaPlaylist::SetMediaInfo(const MediaInfo& media_info) {
   const int32_t time_scale = GetTimeScale(media_info);
   if (time_scale == 0) {
@@ -400,6 +404,8 @@ bool MediaPlaylist::SetMediaInfo(const MediaInfo& media_info) {
       std::vector<std::string>(media_info_.hls_characteristics().begin(),
                                media_info_.hls_characteristics().end());
 
+  forced_subtitle_ = media_info_.forced_subtitle();
+
   return true;
 }
 

packager/hls/base/media_playlist.h

@@ -90,6 +90,9 @@ class MediaPlaylist {
   /// For testing only.
   void SetLanguageForTesting(const std::string& language);
 
+  /// For testing only.
+  void SetForcedSubtitleForTesting(const bool forced_subtitle);
+
   /// For testing only.
   void SetCharacteristicsForTesting(
       const std::vector<std::string>& characteristics);
@@ -223,6 +226,8 @@ class MediaPlaylist {
     return characteristics_;
   }
 
+  bool forced_subtitle() const { return forced_subtitle_; }
+
   bool is_dvs() const {
     // HLS Authoring Specification for Apple Devices
     // https://developer.apple.com/documentation/http_live_streaming/hls_authoring_specification_for_apple_devices#overview
@@ -262,6 +267,7 @@ class MediaPlaylist {
   std::string codec_;
   std::string language_;
   std::vector<std::string> characteristics_;
+  bool forced_subtitle_ = false;
   uint32_t media_sequence_number_ = 0;
   bool inserted_discontinuity_tag_ = false;
   int discontinuity_sequence_number_ = 0;

packager/media/event/hls_notify_muxer_listener.cc

@@ -26,13 +26,15 @@ HlsNotifyMuxerListener::HlsNotifyMuxerListener(
     const std::string& ext_x_media_name,
     const std::string& ext_x_media_group_id,
     const std::vector<std::string>& characteristics,
+    bool forced_subtitle,
     hls::HlsNotifier* hls_notifier,
     std::optional<uint32_t> index)
     : playlist_name_(playlist_name),
       iframes_only_(iframes_only),
       ext_x_media_name_(ext_x_media_name),
       ext_x_media_group_id_(ext_x_media_group_id),
       characteristics_(characteristics),
+      forced_subtitle_(forced_subtitle),
       hls_notifier_(hls_notifier),
       index_(index) {
   DCHECK(hls_notifier);
@@ -103,6 +105,9 @@ void HlsNotifyMuxerListener::OnMediaStart(const MuxerOptions& muxer_options,
     for (const std::string& characteristic : characteristics_)
       media_info->add_hls_characteristics(characteristic);
   }
+  if (forced_subtitle_) {
+    media_info->set_forced_subtitle(forced_subtitle_);
+  }
   if (index_.has_value())
     media_info->set_index(index_.value());