feat: Add support for ALAC codec (shaka-project#1299)

Co-authored-by: Cosmin Stejerean <[email protected]>

packager/app/test/packager_test.py

@@ -1477,6 +1477,15 @@ def testFlacWithEncryption(self):
     self.assertPackageSuccess(streams, flags)
     self._CheckTestResults('flac-with-encryption', verify_decryption=True)
 
+  def testAlac(self):
+    streams = [
+      self._GetStream('audio', test_file='bear-alac.mp4'),
+    ]
+    flags = self._GetFlags(output_dash=True)
+
+    self.assertPackageSuccess(streams, flags)
+    self._CheckTestResults('audio-alac')
+
   def testAv1Mp4WithEncryption(self):
     self.assertPackageSuccess(
         self._GetStreams(['video'], test_files=['bear-av1.mp4']),

packager/app/test/testdata/audio-alac/output.mpd

@@ -0,0 +1,15 @@
+<?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="PT2.739955S">
+  <Period id="0">
+    <AdaptationSet id="0" contentType="audio" subsegmentAlignment="true">
+      <Representation id="0" bandwidth="1046256" codecs="alac" mimeType="audio/mp4" audioSamplingRate="44100">
+        <AudioChannelConfiguration schemeIdUri="urn:mpeg:dash:23003:3:audio_channel_configuration:2011" value="2"/>
+        <BaseURL>bear-alac-audio.mp4</BaseURL>
+        <SegmentBase indexRange="762-829" timescale="44100">
+          <Initialization range="0-761"/>
+        </SegmentBase>
+      </Representation>
+    </AdaptationSet>
+  </Period>
+</MPD>

packager/media/base/audio_stream_info.cc

@@ -25,6 +25,8 @@ std::string AudioCodecToString(Codec codec) {
       return "AAC";
     case kCodecAC3:
       return "AC3";
+    case kCodecALAC:
+      return "ALAC";
     case kCodecDTSC:
       return "DTSC";
     case kCodecDTSE:
@@ -138,6 +140,8 @@ std::string AudioStreamInfo::GetCodecString(Codec codec,
       return "mp4a.40." + absl::StrFormat("%hhu", audio_object_type);
     case kCodecAC3:
       return "ac-3";
+    case kCodecALAC:
+      return "alac";
     case kCodecDTSC:
       return "dtsc";
     case kCodecDTSE:

packager/media/base/fourccs.h

@@ -23,6 +23,7 @@ enum FourCC : uint32_t {
   FOURCC_ac_3 = 0x61632d33,  // "ac-3"
   FOURCC_ac_4 = 0x61632d34,  // "ac-4"
   FOURCC_ac3d = 0x61633364,
+  FOURCC_alac = 0x616c6163,
   FOURCC_apad = 0x61706164,
   FOURCC_av01 = 0x61763031,
   FOURCC_av1C = 0x61763143,

packager/media/base/stream_info.h

@@ -41,6 +41,7 @@ enum Codec {
   kCodecAAC = kCodecAudio,
   kCodecAC3,
   kCodecAC4,
+  kCodecALAC,
   // TODO(kqyang): Use kCodecDTS and a kDtsStreamFormat for the various DTS
   // streams.
   kCodecDTSC,

packager/media/formats/mp4/box_definitions.cc

@@ -1969,6 +1969,27 @@ size_t FlacSpecific::ComputeSizeInternal() {
   return HeaderSize() + data.size();
 }
 
+ALACSpecific::ALACSpecific() = default;
+ALACSpecific::~ALACSpecific() = default;
+
+FourCC ALACSpecific::BoxType() const {
+  return FOURCC_alac;
+}
+
+bool ALACSpecific::ReadWriteInternal(BoxBuffer* buffer) {
+  RCHECK(ReadWriteHeaderInternal(buffer));
+  size_t size = buffer->Reading() ? buffer->BytesLeft() : data.size();
+  RCHECK(buffer->ReadWriteVector(&data, size));
+  return true;
+}
+
+size_t ALACSpecific::ComputeSizeInternal() {
+  // This box is optional. Skip it if not initialized.
+  if (data.empty())
+    return 0;
+  return HeaderSize() + data.size();
+}
+
 AudioSampleEntry::AudioSampleEntry() = default;
 AudioSampleEntry::~AudioSampleEntry() = default;
 
@@ -2011,6 +2032,7 @@ bool AudioSampleEntry::ReadWriteInternal(BoxBuffer* buffer) {
   RCHECK(buffer->TryReadWriteChild(&dops));
   RCHECK(buffer->TryReadWriteChild(&dfla));
   RCHECK(buffer->TryReadWriteChild(&mhac));
+  RCHECK(buffer->TryReadWriteChild(&alac));
 
   // Somehow Edge does not support having sinf box before codec_configuration,
   // box, so just do it in the end of AudioSampleEntry. See
@@ -2036,6 +2058,7 @@ size_t AudioSampleEntry::ComputeSizeInternal() {
          sizeof(samplesize) + sizeof(samplerate) + sinf.ComputeSize() +
          esds.ComputeSize() + ddts.ComputeSize() + dac3.ComputeSize() +
          dec3.ComputeSize() + dops.ComputeSize() + dfla.ComputeSize() +
+         dac4.ComputeSize() + mhac.ComputeSize() + alac.ComputeSize() +
          dac4.ComputeSize() + mhac.ComputeSize() + udts.ComputeSize() +
          // Reserved and predefined bytes.
          6 + 8 +  // 6 + 8 bytes reserved.

packager/media/formats/mp4/box_definitions.h

@@ -382,6 +382,15 @@ struct FlacSpecific : FullBox {
   std::vector<uint8_t> data;
 };
 
+// ALAC specific decoder configuration box:
+//   https://wiki.multimedia.cx/index.php/Apple_Lossless_Audio_Coding
+// We do not care about the actual data inside, which is simply copied over.
+struct ALACSpecific : FullBox {
+  DECLARE_BOX_METHODS(ALACSpecific);
+
+  std::vector<uint8_t> data;
+};
+
 struct AudioSampleEntry : Box {
   DECLARE_BOX_METHODS(AudioSampleEntry);
 
@@ -409,6 +418,7 @@ struct AudioSampleEntry : Box {
   OpusSpecific dops;
   FlacSpecific dfla;
   MHAConfiguration mhac;
+  ALACSpecific alac;
 };
 
 struct WebVTTConfigurationBox : Box {

packager/media/formats/mp4/mp4_media_parser.cc

@@ -102,6 +102,8 @@ Codec FourCCToCodec(FourCC fourcc) {
       return kCodecEAC3;
     case FOURCC_ac_4:
       return kCodecAC4;
+    case FOURCC_alac:
+      return kCodecALAC;
     case FOURCC_fLaC:
       return kCodecFlac;
     case FOURCC_mha1:
@@ -514,6 +516,9 @@ bool MP4MediaParser::ParseMoov(BoxReader* reader) {
             return false;
           }
           break;
+        case FOURCC_alac:
+          codec_config = entry.alac.data;
+          break;
         case FOURCC_fLaC:
           codec_config = entry.dfla.data;
           break;

packager/media/formats/mp4/mp4_muxer.cc

@@ -76,6 +76,8 @@ FourCC CodecToFourCC(Codec codec, H26xStreamFormat h26x_stream_format) {
       return FOURCC_mp4a;
     case kCodecAC3:
       return FOURCC_ac_3;
+    case kCodecALAC:
+      return FOURCC_alac;
     case kCodecDTSC:
       return FOURCC_dtsc;
     case kCodecDTSH:
@@ -512,6 +514,9 @@ bool MP4Muxer::GenerateAudioTrak(const AudioStreamInfo* audio_info,
     case kCodecAC4:
       audio.dac4.data = audio_info->codec_config();
       break;
+    case kCodecALAC:
+      audio.alac.data = audio_info->codec_config();
+      break;
     case kCodecFlac:
       audio.dfla.data = audio_info->codec_config();
       break;