feat(DASH): Add Label element. (#1175)

Add ability to set `Label` tag in MPD, see page 35 of DASH-IF IOP 4.3 https://dashif.org/docs/DASH-IF-IOP-v4.3.pdf Implements #881 --------- Co-authored-by: Cosmin Stejerean <cstejerean@meta.com>
shaka-project · Feb 14, 2024 · b1c5a74 · b1c5a74
1 parent aad2a12
commit b1c5a74
Show file tree

Hide file tree

Showing 18 changed files with 86 additions and 1 deletion.
diff --git a/docs/source/options/dash_options.rst b/docs/source/options/dash_options.rst
@@ -104,4 +104,10 @@ DASH options
 --force_cl_index
 
     True forces the muxer to order streams in the order given 
-    on the command-line. False uses the previous unordered behavior.
+    on the command-line. False uses the previous unordered behavior.
+
+--dash_label <label_name>
+
+    Optional. Will add Label tag to adapation set and will be taken into
+    consideration along with codecs, language, media type (audio, video etc)
+    and container type to create different adaptation sets.
diff --git a/include/packager/packager.h b/include/packager/packager.h
@@ -151,6 +151,9 @@ struct StreamDescriptor {
   bool dash_only = false;
   /// Set to true to indicate that the stream is for hls only.
   bool hls_only = false;
+
+  /// Optional for DASH output. It defines the Label element in Adaptation Set.
+  std::string dash_label;
 };
 
 class SHAKA_EXPORT Packager {

diff --git a/packager/app/stream_descriptor.cc b/packager/app/stream_descriptor.cc
@@ -39,6 +39,7 @@ enum FieldType {
   kDashRolesField,
   kDashOnlyField,
   kHlsOnlyField,
+  kDashLabelField,
 };
 
 struct FieldNameToTypeMapping {
@@ -86,6 +87,7 @@ const FieldNameToTypeMapping kFieldNameTypeMappings[] = {
     {"role", kDashRolesField},
     {"dash_only", kDashOnlyField},
     {"hls_only", kHlsOnlyField},
+    {"dash_label", kDashLabelField},
 };
 
 FieldType GetFieldType(const std::string& field_name) {
@@ -250,6 +252,9 @@ std::optional<StreamDescriptor> ParseStreamDescriptor(
         }
         descriptor.hls_only = hls_only_value > 0;
         break;
+      case kDashLabelField:
+        descriptor.dash_label = pair.second;
+        break;
       default:
         LOG(ERROR) << "Unknown field in stream descriptor (\"" << pair.first
                    << "\").";

diff --git a/packager/app/test/packager_test.py b/packager/app/test/packager_test.py
@@ -304,6 +304,7 @@ def _GetStream(self,
                  dash_accessibilities=None,
                  dash_roles=None,
                  dash_only=None,
+                 dash_label=None,
                  trick_play_factor=None,
                  drm_label=None,
                  skip_encryption=None,
@@ -334,6 +335,7 @@ def _GetStream(self,
       dash_accessibilities: Accessibility element for the DASH stream.
       dash_roles: Role element for the DASH stream.
       dash_only: If set to true, will indicate that the stream is for DASH only.
+      dash_label: Label element for the DASH stream.
       trick_play_factor: Signals the stream is to be used for a trick play
           stream and which key frames to use. A trick play factor of 0 is the
           same as not specifying a trick play factor.
@@ -400,6 +402,9 @@ def _GetStream(self,
     if dash_only:
       stream.Append('dash_only', 1)
 
+    if dash_label:
+      stream.Append('dash_label', dash_label)
+
     requires_init_segment = segmented and base_ext not in [
         'aac', 'ac3', 'ec3', 'ts', 'vtt', 'ttml',
     ]
@@ -786,6 +791,14 @@ def testDashOnlyAndHlsOnly(self):
         self._GetFlags(output_dash=True, output_hls=True))
     self._CheckTestResults('hls-only-dash-only')
 
+  def testDashLabel(self):
+    streams = [
+        self._GetStream('video', dash_label='Main'),
+        self._GetStream('audio', dash_label='English'),
+    ]
+    self.assertPackageSuccess(streams, self._GetFlags(output_dash=True))
+    self._CheckTestResults('dash-label')
+
   def testAudioVideoWithLanguageOverride(self):
     self.assertPackageSuccess(
         self._GetStreams(['audio', 'video'], language='por', hls=True),

diff --git a/packager/app/test/testdata/dash-label/bear-640x360-audio.mp4 b/packager/app/test/testdata/dash-label/bear-640x360-audio.mp4
diff --git a/packager/app/test/testdata/dash-label/bear-640x360-video.mp4 b/packager/app/test/testdata/dash-label/bear-640x360-video.mp4
diff --git a/packager/app/test/testdata/dash-label/output.mpd b/packager/app/test/testdata/dash-label/output.mpd
@@ -0,0 +1,25 @@
+<?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="video" width="640" height="360" frameRate="30000/1001" subsegmentAlignment="true" par="16:9">
+      <Label>Main</Label>
+      <Representation id="0" 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="1" contentType="audio" subsegmentAlignment="true">
+      <Label>English</Label>
+      <Representation id="1" 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>
+  </Period>
+</MPD>
diff --git a/packager/media/event/mpd_notify_muxer_listener.cc b/packager/media/event/mpd_notify_muxer_listener.cc
@@ -88,6 +88,9 @@ void MpdNotifyMuxerListener::OnMediaStart(const MuxerOptions& muxer_options,
   if (index_.has_value())
     media_info->set_index(index_.value());
 
+  if (!dash_label_.empty())
+    media_info->set_dash_label(dash_label_);
+
   if (is_encrypted_) {
     internal::SetContentProtectionFields(protection_scheme_, default_key_id_,
                                          key_system_info_, media_info.get());

diff --git a/packager/media/event/mpd_notify_muxer_listener.h b/packager/media/event/mpd_notify_muxer_listener.h
@@ -69,6 +69,8 @@ class MpdNotifyMuxerListener : public MuxerListener {
 
   void set_index(std::optional<uint32_t> idx) { index_ = idx; }
 
+  void set_dash_label(std::string label) { dash_label_ = label; }
+
  private:
   MpdNotifyMuxerListener(const MpdNotifyMuxerListener&) = delete;
   MpdNotifyMuxerListener& operator=(const MpdNotifyMuxerListener&) = delete;
@@ -81,6 +83,7 @@ class MpdNotifyMuxerListener : public MuxerListener {
 
   std::vector<std::string> accessibilities_;
   std::vector<std::string> roles_;
+  std::string dash_label_;
 
   std::optional<uint32_t> index_ = 0;
 

diff --git a/packager/media/event/muxer_listener_factory.cc b/packager/media/event/muxer_listener_factory.cc
@@ -45,6 +45,7 @@ std::unique_ptr<MuxerListener> CreateMpdListenerInternal(
   listener->set_accessibilities(stream.dash_accessiblities);
   listener->set_roles(stream.dash_roles);
   listener->set_index(stream.index);
+  listener->set_dash_label(stream.dash_label);
   return listener;
 }
 

diff --git a/packager/media/event/muxer_listener_factory.h b/packager/media/event/muxer_listener_factory.h
@@ -55,6 +55,7 @@ class MuxerListenerFactory {
     std::vector<std::string> dash_roles;
     bool dash_only = false;
     std::optional<uint32_t> index;
+    std::string dash_label;
   };
 
   /// Create a new muxer listener.

diff --git a/packager/mpd/base/adaptation_set.cc b/packager/mpd/base/adaptation_set.cc
@@ -382,6 +382,9 @@ std::optional<xml::XmlNode> AdaptationSet::GetXml() {
     }
   }
 
+  if (!label_.empty() && !adaptation_set.AddLabelElement(label_))
+    return std::nullopt;
+
   for (const auto& representation_pair : representation_map_) {
     const auto& representation = representation_pair.second;
     if (suppress_representation_width)
@@ -474,6 +477,9 @@ void AdaptationSet::UpdateFromMediaInfo(const MediaInfo& media_info) {
     }
   }
 
+  if (media_info.has_dash_label())
+    label_ = media_info.dash_label();
+
   if (media_info.has_video_info()) {
     content_type_ = "video";
   } else if (media_info.has_audio_info()) {

diff --git a/packager/mpd/base/adaptation_set.h b/packager/mpd/base/adaptation_set.h
@@ -330,6 +330,9 @@ class AdaptationSet {
 
   // the command-line index for this AdaptationSet
   std::optional<uint32_t> index_;
+
+  // The label of this AdaptationSet.
+  std::string label_;
 };
 
 }  // namespace shaka

diff --git a/packager/mpd/base/media_info.proto b/packager/mpd/base/media_info.proto
@@ -215,4 +215,7 @@ message MediaInfo {
 
   // stream index for consistent ordering of streams
   optional uint32 index = 28;
+
+  // DASH only. Label element.
+  optional string dash_label = 29;
 }
diff --git a/packager/mpd/base/mpd_utils.cc b/packager/mpd/base/mpd_utils.cc
@@ -161,6 +161,9 @@ std::string GetAdaptationSetKey(const MediaInfo& media_info,
     key.append("unknown:");
   }
 
+  if (media_info.has_dash_label())
+    key.append(media_info.dash_label() + ":");
+
   key.append(MediaInfo_ContainerType_Name(media_info.container_type()));
   if (!ignore_codec) {
     key.append(":");

diff --git a/packager/mpd/base/xml/xml_node.cc b/packager/mpd/base/xml/xml_node.cc
@@ -337,6 +337,12 @@ bool AdaptationSetXmlNode::AddRoleElement(const std::string& scheme_id_uri,
   return AddDescriptor("Role", scheme_id_uri, value);
 }
 
+bool AdaptationSetXmlNode::AddLabelElement(const std::string& value) {
+  XmlNode descriptor("Label");
+  descriptor.SetContent(value);
+  return AddChild(std::move(descriptor));
+}
+
 RepresentationXmlNode::RepresentationXmlNode()
     : RepresentationBaseXmlNode("Representation") {}
 RepresentationXmlNode::~RepresentationXmlNode() {}

diff --git a/packager/mpd/base/xml/xml_node.h b/packager/mpd/base/xml/xml_node.h
@@ -171,6 +171,9 @@ class AdaptationSetXmlNode : public RepresentationBaseXmlNode {
   [[nodiscard]] bool AddRoleElement(const std::string& scheme_id_uri,
                                     const std::string& value);
 
+  /// @param value is element's content.
+  [[nodiscard]] bool AddLabelElement(const std::string& value);
+
  private:
   DISALLOW_COPY_AND_ASSIGN(AdaptationSetXmlNode);
 };

diff --git a/packager/packager.cc b/packager/packager.cc
@@ -75,6 +75,7 @@ MuxerListenerFactory::StreamData ToMuxerListenerData(
   data.dash_roles = stream.dash_roles;
   data.dash_only = stream.dash_only;
   data.index = stream.index;
+  data.dash_label = stream.dash_label;
   return data;
 };