Add version information in generated outputs

For mp4 outputs, a metadata box with version information is added to
moov box.

For mpd outputs, a text comment with version information is added in
the beginning of mpd file.

Issue #60

Change-Id: I783ba370781c0a8f77c910ff1172bad2e7edff75

gyp_packager.py

@@ -1,35 +1,34 @@
-#!/usr/bin/env python
+#!/usr/bin/python
 #
 # Copyright 2014 Google Inc. All rights reserved.
 #
 # Use of this source code is governed by a BSD-style
 # license that can be found in the LICENSE file or at
 # https://developers.google.com/open-source/licenses/bsd
-#
+"""This script wraps gyp and sets up build environments.
-# This script is a wrapper for Packager that adds some support for how GYP
+
-# is invoked by Packager.
+Build instructions:
-#
+
-# Build instructions:
+1. Setup gyp: ./gyp_packager.py or use gclient runhooks
-#
+
-# 1. Setup gyp: ./gyp_packager.py
+clang is enabled by default, which can be disabled by overriding
-#
+GYP_DEFINE environment variable, i.e.
-# clang is not enabled by default, which can be enabled by overriding
+"GYP_DEFINES='clang=0' gclient runhooks".
-# GYP_DEFINE environment variable, i.e.
+
-# "GYP_DEFINES='clang=1' ./gyp_packager.py".
+Ninja is the default build system. User can also change to make by
-#
+overriding GYP_GENERATORS to make, i.e.
-# Ninja is the default build system. User can also change to make by
+"GYP_GENERATORS='make' gclient runhooks".
-# overriding GYP_GENERATORS to make, i.e.
+
-# "GYP_GENERATORS='make' ./gyp_packager.py".
+2. The first step generates the make files but does not start the
-#
+build process. Ninja is the default build system. Refer to Ninja
-# 2. The first step generates the make files but does not start the
+manual on how to do the build.
-# build process. Ninja is the default build system. Refer to Ninja
+
-# manual on how to do the build.
+Common syntaxes: ninja -C out/{Debug/Release} [Module]
-#
+Module is optional. If not specified, build everything.
-# Common syntaxes: ninja -C out/{Debug/Release} [Module]
+
-# Module is optional. If not specified, build everything.
+Step 1 is only required if there is any gyp file change. Otherwise, you
-#
+may just run ninja.
-# Step 1 is only required if there is any gyp file change. Otherwise, you
+"""
-# may just run ninja.
 
 import os
 import sys
@@ -37,11 +36,14 @@ import sys
 checkout_dir = os.path.dirname(os.path.realpath(__file__))
 src_dir = os.path.join(checkout_dir, 'packager')
 
+# Workaround the dynamic path.
+# pylint: disable=g-import-not-at-top,g-bad-import-order
+
 sys.path.insert(0, os.path.join(src_dir, 'build'))
-import gyp_helper  # Workaround the dynamic path. pylint: disable-msg=F0401
+import gyp_helper
 
 sys.path.insert(0, os.path.join(src_dir, 'tools', 'gyp', 'pylib'))
-import gyp  # Workaround the dynamic path. pylint: disable-msg=F0401
+import gyp
 
 if __name__ == '__main__':
   args = sys.argv[1:]
@@ -58,12 +60,15 @@ if __name__ == '__main__':
   args.extend(['-I' + os.path.join(src_dir, 'build', 'common.gypi')])
 
   # Set these default GYP_DEFINES if user does not set the value explicitly.
-  _DEFAULT_DEFINES = {'test_isolation_mode': 'noop', 'use_glib': 0,
+  _DEFAULT_DEFINES = {'test_isolation_mode': 'noop',
-                      'use_openssl': 1, 'use_x11': 0,
+                      'use_glib': 0,
-                      'linux_use_gold_binary': 0, 'linux_use_gold_flags': 0}
+                      'use_openssl': 1,
+                      'use_x11': 0,
+                      'linux_use_gold_binary': 0,
+                      'linux_use_gold_flags': 0}
 
-  gyp_defines = (os.environ['GYP_DEFINES'] if os.environ.get('GYP_DEFINES')
+  gyp_defines = (os.environ['GYP_DEFINES'] if os.environ.get('GYP_DEFINES') else
-                 else '')
+                 '')
   for key in _DEFAULT_DEFINES:
     if key not in gyp_defines:
       gyp_defines += ' {0}={1}'.format(key, _DEFAULT_DEFINES[key])

packager/app/mpd_generator.cc

@@ -11,6 +11,7 @@
 #include "packager/base/strings/string_split.h"
 #include "packager/base/strings/stringprintf.h"
 #include "packager/mpd/util/mpd_writer.h"
+#include "packager/version/version.h"
 
 namespace edash_packager {
 namespace {
@@ -88,7 +89,9 @@ int MpdMain(int argc, char** argv) {
 
   ExitStatus status = CheckRequiredFlags();
   if (status != kSuccess) {
-    google::ShowUsageWithFlags(argv[0]);
+    std::string version_string =
+        base::StringPrintf("mpd_generator version %s", kPackagerVersion);
+    google::ShowUsageWithFlags(version_string.c_str());
     return status;
   }
 

packager/app/packager_main.cc

@@ -36,6 +36,7 @@
 #include "packager/mpd/base/media_info.pb.h"
 #include "packager/mpd/base/mpd_builder.h"
 #include "packager/mpd/base/simple_mpd_notifier.h"
+#include "packager/version/version.h"
 
 DEFINE_bool(use_fake_clock_for_muxer,
             false,
@@ -45,37 +46,36 @@ DEFINE_bool(use_fake_clock_for_muxer,
 
 namespace {
 const char kUsage[] =
-    "Packager driver program. Sample Usage:\n"
+    "Packager driver program. Usage:\n\n"
     "%s [flags] <stream_descriptor> ...\n"
     "stream_descriptor consists of comma separated field_name/value pairs:\n"
     "field_name=value,[field_name=value,]...\n"
     "Supported field names are as follows:\n"
-    "  - input (in): Required input/source media file path or network stream "
+    "  - input (in): Required input/source media file path or network stream\n"
-    "URL.\n"
+    "    URL.\n"
-    "  - stream_selector (stream): Required field with value 'audio', 'video', "
+    "  - stream_selector (stream): Required field with value 'audio',\n"
-    "or stream number (zero based).\n"
+    "    'video', or stream number (zero based).\n"
-    "  - output (out): Required output file (single file) or initialization "
+    "  - output (out): Required output file (single file) or initialization\n"
-    "file path (multiple file).\n"
+    "    file path (multiple file).\n"
-    "  - segment_template (segment): Optional value which specifies the "
+    "  - segment_template (segment): Optional value which specifies the\n"
-    "naming  pattern for the segment files, and that the stream should be "
+    "    naming  pattern for the segment files, and that the stream should be\n"
-    "split into multiple files. Its presence should be consistent across "
+    "    split into multiple files. Its presence should be consistent across\n"
-    "streams.\n"
+    "    streams.\n"
-    "  - bandwidth (bw): Optional value which contains a user-specified "
+    "  - bandwidth (bw): Optional value which contains a user-specified\n"
-    "content bit rate for the stream, in bits/sec. If specified, this value is "
+    "    content bit rate for the stream, in bits/sec. If specified, this\n"
-    "propagated to the $Bandwidth$ template parameter for segment names. "
+    "    value is propagated to the $Bandwidth$ template parameter for\n"
-    "If not specified, its value may be estimated.\n"
+    "    segment names. If not specified, its value may be estimated.\n"
-    "  - language (lang): Optional value which contains a user-specified "
+    "  - language (lang): Optional value which contains a user-specified\n"
-    "language tag. If specified, this value overrides any language metadata "
+    "    language tag. If specified, this value overrides any language\n"
-    "in the input track.\n"
+    "    metadata in the input track.\n"
-    "  - output_format (format): Optional value which specifies the format "
+    "  - output_format (format): Optional value which specifies the format\n"
-    "of the output files (MP4 or WebM).  If not specified, it will be "
+    "    of the output files (MP4 or WebM).  If not specified, it will be\n"
-    "derived from the file extension of the output file.\n";
+    "    derived from the file extension of the output file.\n";
 
 const char kMediaInfoSuffix[] = ".media_info";
 
 enum ExitStatus {
   kSuccess = 0,
-  kNoArgument,
   kArgumentValidationFailed,
   kPackagingFailed,
   kInternalError,
@@ -441,8 +441,10 @@ int PackagerMain(int argc, char** argv) {
   google::SetUsageMessage(base::StringPrintf(kUsage, argv[0]));
   google::ParseCommandLineFlags(&argc, &argv, true);
   if (argc < 2) {
-    google::ShowUsageWithFlags(argv[0]);
+    std::string version_string =
-    return kNoArgument;
+        base::StringPrintf("edash-packager version %s", kPackagerVersion);
+    google::ShowUsageWithFlags(version_string.c_str());
+    return kSuccess;
   }
 
   if (!ValidateWidevineCryptoFlags() || !ValidateFixedCryptoFlags())

packager/app/packager_util.cc

@@ -25,6 +25,16 @@
 #include "packager/mpd/base/mpd_builder.h"
 
 DEFINE_bool(dump_stream_info, false, "Dump demuxed stream info.");
+DEFINE_bool(override_version_string,
+            false,
+            "Override packager version string in the generated outputs with "
+            "--test_version_string if it is set to true. Should be used for "
+            "testing only.");
+DEFINE_string(test_version_string,
+              "",
+              "Packager version string for testing. Ignored if "
+              "--override_version_string is false. Should be used for testing "
+              "only.");
 
 namespace edash_packager {
 namespace media {
@@ -146,6 +156,8 @@ bool GetMuxerOptions(MuxerOptions* muxer_options) {
   muxer_options->fragment_sap_aligned = FLAGS_fragment_sap_aligned;
   muxer_options->num_subsegments_per_sidx = FLAGS_num_subsegments_per_sidx;
   muxer_options->temp_dir = FLAGS_temp_dir;
+  if (FLAGS_override_version_string)
+    muxer_options->packager_version_string = FLAGS_test_version_string;
   return true;
 }
 
@@ -158,6 +170,8 @@ bool GetMpdOptions(MpdOptions* mpd_options) {
   mpd_options->time_shift_buffer_depth = FLAGS_time_shift_buffer_depth;
   mpd_options->suggested_presentation_delay =
       FLAGS_suggested_presentation_delay;
+  if (FLAGS_override_version_string)
+    mpd_options->packager_version_string = FLAGS_test_version_string;
   return true;
 }
 

packager/app/test/packager_app.py

@@ -31,6 +31,13 @@ class PackagerApp(object):
     cmd = [self.binary, input_str, '--dump_stream_info']
     return subprocess.check_output(cmd)
 
+  def Version(self):
+    output = subprocess.check_output([self.binary])
+    # The output should of the form:
+    #   edash-packager version xxx: Description...
+    # We consider everything before ':' part of version.
+    return output.split(':')[0]
+
   def Package(self, streams, flags=None):
     if flags is None:
       flags = []

packager/app/test/packager_test.py

@@ -38,6 +38,11 @@ class PackagerAppTest(unittest.TestCase):
   def testBuildingCode(self):
     self.assertEqual(0, self.packager.BuildSrc())
 
+  def testVersion(self):
+    self.assertRegexpMatches(
+        self.packager.Version(), '^edash-packager version '
+        '((?P<tag>[\w\.]+)-)?(?P<hash>[a-f\d]+)-(debug|release)$')
+
   def testDumpStreamInfo(self):
     test_file = os.path.join(self.test_data_dir, 'bear-640x360.mp4')
     stream_info = self.packager.DumpStreamInfo(test_file)
@@ -346,6 +351,10 @@ class PackagerAppTest(unittest.TestCase):
     # Use fake clock, so output can be compared.
     if use_fake_clock:
       flags.append('--use_fake_clock_for_muxer')
+
+    # Override packager version string for testing.
+    flags += ['--override_version_string', '--test_version_string',
+              '<tag>-<hash>-<test>']
     return flags
 
   def _CompareWithGold(self, test_output, golden_file_name):

packager/app/test/testdata/bear-640x360-a-cenc-golden.mp4.media_info

@@ -1,4 +1,4 @@
-bandwidth: 128728
+bandwidth: 129127
 audio_info {
   codec: "mp4a.40.2"
   sampling_frequency: 44100
@@ -8,11 +8,11 @@ audio_info {
 }
 init_range {
   begin: 0
-  end: 816
+  end: 954
 }
 index_range {
-  begin: 817
+  begin: 955
-  end: 884
+  end: 1022
 }
 media_file_name: "place_holder"
 media_duration_seconds: 2.7631745

packager/app/test/testdata/bear-640x360-av-cenc-golden.mpd

@@ -1,28 +1,29 @@
 <?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/google/edash-packager version <tag>-<hash>-<test>-->
 <MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xlink="http://www.w3.org/1999/xlink" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" xmlns:cenc="urn:mpeg:cenc:2013" minBufferTime="PT2S" type="static" profiles="urn:mpeg:dash:profile:isoff-on-demand:2011" mediaPresentationDuration="PT2.763174533843994S">
   <Period>
     <AdaptationSet id="0" contentType="video" width="640" height="360" frameRate="30000/1001" subsegmentAlignment="true" par="16:9">
-      <Representation id="0" bandwidth="885151" codecs="avc1.64001e" mimeType="video/mp4" sar="1:1" width="640" height="360" frameRate="30000/1001">
+      <Representation id="0" bandwidth="885555" codecs="avc1.64001e" mimeType="video/mp4" sar="1:1" width="640" height="360" frameRate="30000/1001">
         <ContentProtection value="cenc" schemeIdUri="urn:mpeg:dash:mp4protection:2011" cenc:default_KID="31323334-3536-3738-3930-313233343536"/>
         <ContentProtection schemeIdUri="urn:uuid:edef8ba9-79d6-4ace-a3c8-27dcd51d21ed">
           <cenc:pssh>AAAAMHBzc2gAAAAA7e+LqXnWSs6jyCfc1R0h7QAAABAxMjM0NTY3ODkwMTIzNDU2</cenc:pssh>
         </ContentProtection>
         <BaseURL>output_video.mp4</BaseURL>
-        <SegmentBase indexRange="941-1008" timescale="30000">
+        <SegmentBase indexRange="1079-1146" timescale="30000">
-          <Initialization range="0-940"/>
+          <Initialization range="0-1078"/>
         </SegmentBase>
       </Representation>
     </AdaptationSet>
     <AdaptationSet id="1" contentType="audio" subsegmentAlignment="true">
-      <Representation id="1" bandwidth="128728" codecs="mp4a.40.2" mimeType="audio/mp4" audioSamplingRate="44100">
+      <Representation id="1" bandwidth="129127" codecs="mp4a.40.2" mimeType="audio/mp4" audioSamplingRate="44100">
         <AudioChannelConfiguration schemeIdUri="urn:mpeg:dash:23003:3:audio_channel_configuration:2011" value="2"/>
         <ContentProtection value="cenc" schemeIdUri="urn:mpeg:dash:mp4protection:2011" cenc:default_KID="31323334-3536-3738-3930-313233343536"/>
         <ContentProtection schemeIdUri="urn:uuid:edef8ba9-79d6-4ace-a3c8-27dcd51d21ed">
           <cenc:pssh>AAAAMHBzc2gAAAAA7e+LqXnWSs6jyCfc1R0h7QAAABAxMjM0NTY3ODkwMTIzNDU2</cenc:pssh>
         </ContentProtection>
         <BaseURL>output_audio.mp4</BaseURL>
-        <SegmentBase indexRange="817-884" timescale="44100">
+        <SegmentBase indexRange="955-1022" timescale="44100">
-          <Initialization range="0-816"/>
+          <Initialization range="0-954"/>
         </SegmentBase>
       </Representation>
     </AdaptationSet>

packager/app/test/testdata/bear-640x360-av-cenc-iop-golden.mpd

@@ -1,4 +1,5 @@
 <?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/google/edash-packager version <tag>-<hash>-<test>-->
 <MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xlink="http://www.w3.org/1999/xlink" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" xmlns:cenc="urn:mpeg:cenc:2013" minBufferTime="PT2S" type="static" profiles="urn:mpeg:dash:profile:isoff-on-demand:2011" mediaPresentationDuration="PT2.763174533843994S">
   <Period>
     <AdaptationSet id="0" contentType="video" width="640" height="360" frameRate="30000/1001" subsegmentAlignment="true" par="16:9">
@@ -6,10 +7,10 @@
       <ContentProtection schemeIdUri="urn:uuid:edef8ba9-79d6-4ace-a3c8-27dcd51d21ed">
         <cenc:pssh>AAAAMHBzc2gAAAAA7e+LqXnWSs6jyCfc1R0h7QAAABAxMjM0NTY3ODkwMTIzNDU2</cenc:pssh>
       </ContentProtection>
-      <Representation id="0" bandwidth="885151" codecs="avc1.64001e" mimeType="video/mp4" sar="1:1" width="640" height="360" frameRate="30000/1001">
+      <Representation id="0" bandwidth="885555" codecs="avc1.64001e" mimeType="video/mp4" sar="1:1" width="640" height="360" frameRate="30000/1001">
         <BaseURL>output_video.mp4</BaseURL>
-        <SegmentBase indexRange="941-1008" timescale="30000">
+        <SegmentBase indexRange="1079-1146" timescale="30000">
-          <Initialization range="0-940"/>
+          <Initialization range="0-1078"/>
         </SegmentBase>
       </Representation>
     </AdaptationSet>
@@ -18,11 +19,11 @@
       <ContentProtection schemeIdUri="urn:uuid:edef8ba9-79d6-4ace-a3c8-27dcd51d21ed">
         <cenc:pssh>AAAAMHBzc2gAAAAA7e+LqXnWSs6jyCfc1R0h7QAAABAxMjM0NTY3ODkwMTIzNDU2</cenc:pssh>
       </ContentProtection>
-      <Representation id="1" bandwidth="128728" codecs="mp4a.40.2" mimeType="audio/mp4" audioSamplingRate="44100">
+      <Representation id="1" bandwidth="129127" codecs="mp4a.40.2" mimeType="audio/mp4" audioSamplingRate="44100">
         <AudioChannelConfiguration schemeIdUri="urn:mpeg:dash:23003:3:audio_channel_configuration:2011" value="2"/>
         <BaseURL>output_audio.mp4</BaseURL>
-        <SegmentBase indexRange="817-884" timescale="44100">
+        <SegmentBase indexRange="955-1022" timescale="44100">
-          <Initialization range="0-816"/>
+          <Initialization range="0-954"/>
         </SegmentBase>
       </Representation>
     </AdaptationSet>

packager/app/test/testdata/bear-640x360-av-golden.mpd

@@ -1,20 +1,21 @@
 <?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/google/edash-packager version <tag>-<hash>-<test>-->
 <MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xlink="http://www.w3.org/1999/xlink" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" xmlns:cenc="urn:mpeg:cenc:2013" minBufferTime="PT2S" type="static" profiles="urn:mpeg:dash:profile:isoff-on-demand:2011" mediaPresentationDuration="PT2.763174533843994S">
   <Period>
     <AdaptationSet id="0" contentType="video" width="640" height="360" frameRate="30000/1001" subsegmentAlignment="true" par="16:9">
-      <Representation id="0" bandwidth="881637" codecs="avc1.64001e" mimeType="video/mp4" sar="1:1" width="640" height="360" frameRate="30000/1001">
+      <Representation id="0" bandwidth="882040" codecs="avc1.64001e" mimeType="video/mp4" sar="1:1" width="640" height="360" frameRate="30000/1001">
         <BaseURL>output_video.mp4</BaseURL>
-        <SegmentBase indexRange="677-744" timescale="30000">
+        <SegmentBase indexRange="815-882" timescale="30000">
-          <Initialization range="0-676"/>
+          <Initialization range="0-814"/>
         </SegmentBase>
       </Representation>
     </AdaptationSet>
     <AdaptationSet id="1" contentType="audio" subsegmentAlignment="true">
-      <Representation id="1" bandwidth="126087" codecs="mp4a.40.2" mimeType="audio/mp4" audioSamplingRate="44100">
+      <Representation id="1" bandwidth="126487" codecs="mp4a.40.2" mimeType="audio/mp4" audioSamplingRate="44100">
         <AudioChannelConfiguration schemeIdUri="urn:mpeg:dash:23003:3:audio_channel_configuration:2011" value="2"/>
         <BaseURL>output_audio.mp4</BaseURL>
-        <SegmentBase indexRange="611-678" timescale="44100">
+        <SegmentBase indexRange="749-816" timescale="44100">
-          <Initialization range="0-610"/>
+          <Initialization range="0-748"/>
         </SegmentBase>
       </Representation>
     </AdaptationSet>

packager/app/test/testdata/bear-640x360-av-live-cenc-golden.mpd

@@ -1,4 +1,5 @@
 <?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/google/edash-packager version <tag>-<hash>-<test>-->
 <MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xlink="http://www.w3.org/1999/xlink" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" xmlns:cenc="urn:mpeg:cenc:2013" minBufferTime="PT2S" type="dynamic" profiles="urn:mpeg:dash:profile:isoff-live:2011" availabilityStartTime="place_holder" minimumUpdatePeriod="PT5S" timeShiftBufferDepth="PT1800S">
   <Period start="PT0S">
     <AdaptationSet id="0" contentType="video" width="640" height="360" frameRate="30000/1001" segmentAlignment="true" par="16:9">

packager/app/test/testdata/bear-640x360-av-live-cenc-iop-golden.mpd

@@ -1,4 +1,5 @@
 <?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/google/edash-packager version <tag>-<hash>-<test>-->
 <MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xlink="http://www.w3.org/1999/xlink" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" xmlns:cenc="urn:mpeg:cenc:2013" minBufferTime="PT2S" type="dynamic" profiles="urn:mpeg:dash:profile:isoff-live:2011" availabilityStartTime="place_holder" minimumUpdatePeriod="PT5S" timeShiftBufferDepth="PT1800S">
   <Period start="PT0S">
     <AdaptationSet id="0" contentType="video" width="640" height="360" frameRate="30000/1001" segmentAlignment="true" par="16:9">

packager/app/test/testdata/bear-640x360-av-live-cenc-rotation-golden.mpd

@@ -1,4 +1,5 @@
 <?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/google/edash-packager version <tag>-<hash>-<test>-->
 <MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xlink="http://www.w3.org/1999/xlink" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" xmlns:cenc="urn:mpeg:cenc:2013" minBufferTime="PT2S" type="dynamic" profiles="urn:mpeg:dash:profile:isoff-live:2011" availabilityStartTime="place_holder" minimumUpdatePeriod="PT5S" timeShiftBufferDepth="PT1800S">
   <Period start="PT0S">
     <AdaptationSet id="0" contentType="video" width="640" height="360" frameRate="30000/1001" segmentAlignment="true" par="16:9">

packager/app/test/testdata/bear-640x360-av-live-cenc-rotation-iop-golden.mpd

@@ -1,4 +1,5 @@
 <?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/google/edash-packager version <tag>-<hash>-<test>-->
 <MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xlink="http://www.w3.org/1999/xlink" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" xmlns:cenc="urn:mpeg:cenc:2013" minBufferTime="PT2S" type="dynamic" profiles="urn:mpeg:dash:profile:isoff-live:2011" availabilityStartTime="place_holder" minimumUpdatePeriod="PT5S" timeShiftBufferDepth="PT1800S">
   <Period start="PT0S">
     <AdaptationSet id="0" contentType="video" width="640" height="360" frameRate="30000/1001" segmentAlignment="true" par="16:9">

packager/app/test/testdata/bear-640x360-av-live-golden.mpd

@@ -1,4 +1,5 @@
 <?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/google/edash-packager version <tag>-<hash>-<test>-->
 <MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xlink="http://www.w3.org/1999/xlink" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" xmlns:cenc="urn:mpeg:cenc:2013" minBufferTime="PT2S" type="dynamic" profiles="urn:mpeg:dash:profile:isoff-live:2011" availabilityStartTime="place_holder" minimumUpdatePeriod="PT5S" timeShiftBufferDepth="PT1800S">
   <Period start="PT0S">
     <AdaptationSet id="0" contentType="video" width="640" height="360" frameRate="30000/1001" segmentAlignment="true" par="16:9">

packager/app/test/testdata/bear-640x360-avt-golden.mpd

@@ -1,4 +1,5 @@
 <?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/google/edash-packager version <tag>-<hash>-<test>-->
 <MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xlink="http://www.w3.org/1999/xlink" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" xmlns:cenc="urn:mpeg:cenc:2013" minBufferTime="PT2S" type="static" profiles="urn:mpeg:dash:profile:isoff-on-demand:2011" mediaPresentationDuration="PT2.763174533843994S">
   <Period>
     <AdaptationSet id="0" contentType="text">
@@ -7,19 +8,19 @@
       </Representation>
     </AdaptationSet>
     <AdaptationSet id="1" contentType="video" width="640" height="360" frameRate="30000/1001" subsegmentAlignment="true" par="16:9">
-      <Representation id="1" bandwidth="881637" codecs="avc1.64001e" mimeType="video/mp4" sar="1:1" width="640" height="360" frameRate="30000/1001">
+      <Representation id="1" bandwidth="882040" codecs="avc1.64001e" mimeType="video/mp4" sar="1:1" width="640" height="360" frameRate="30000/1001">
         <BaseURL>output_video.mp4</BaseURL>
-        <SegmentBase indexRange="677-744" timescale="30000">
+        <SegmentBase indexRange="815-882" timescale="30000">
-          <Initialization range="0-676"/>
+          <Initialization range="0-814"/>
         </SegmentBase>
       </Representation>
     </AdaptationSet>
     <AdaptationSet id="2" contentType="audio" subsegmentAlignment="true">
-      <Representation id="2" bandwidth="126087" codecs="mp4a.40.2" mimeType="audio/mp4" audioSamplingRate="44100">
+      <Representation id="2" bandwidth="126487" codecs="mp4a.40.2" mimeType="audio/mp4" audioSamplingRate="44100">
         <AudioChannelConfiguration schemeIdUri="urn:mpeg:dash:23003:3:audio_channel_configuration:2011" value="2"/>
         <BaseURL>output_audio.mp4</BaseURL>
-        <SegmentBase indexRange="611-678" timescale="44100">
+        <SegmentBase indexRange="749-816" timescale="44100">
-          <Initialization range="0-610"/>
+          <Initialization range="0-748"/>
         </SegmentBase>
       </Representation>
     </AdaptationSet>

packager/app/test/testdata/bear-640x360-hevc-v-cenc-golden.mpd

@@ -1,15 +1,16 @@
 <?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/google/edash-packager version <tag>-<hash>-<test>-->
 <MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xlink="http://www.w3.org/1999/xlink" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" xmlns:cenc="urn:mpeg:cenc:2013" minBufferTime="PT2S" type="static" profiles="urn:mpeg:dash:profile:isoff-on-demand:2011" mediaPresentationDuration="PT2.7694332599639893S">
   <Period>
     <AdaptationSet id="0" contentType="video" width="640" height="360" frameRate="30000/1001" subsegmentAlignment="true" par="16:9">
-      <Representation id="0" bandwidth="264624" codecs="hev1.1.6.L63.80" mimeType="video/mp4" sar="1:1" width="640" height="360" frameRate="30000/1001">
+      <Representation id="0" bandwidth="265022" codecs="hev1.1.6.L63.80" mimeType="video/mp4" sar="1:1" width="640" height="360" frameRate="30000/1001">
         <ContentProtection value="cenc" schemeIdUri="urn:mpeg:dash:mp4protection:2011" cenc:default_KID="31323334-3536-3738-3930-313233343536"/>
         <ContentProtection schemeIdUri="urn:uuid:edef8ba9-79d6-4ace-a3c8-27dcd51d21ed">
           <cenc:pssh>AAAAMHBzc2gAAAAA7e+LqXnWSs6jyCfc1R0h7QAAABAxMjM0NTY3ODkwMTIzNDU2</cenc:pssh>
         </ContentProtection>
         <BaseURL>output_video.mp4</BaseURL>
-        <SegmentBase indexRange="2721-2764" timescale="30000">
+        <SegmentBase indexRange="2859-2902" timescale="30000">
-          <Initialization range="0-2720"/>
+          <Initialization range="0-2858"/>
         </SegmentBase>
       </Representation>
     </AdaptationSet>

packager/app/test/testdata/bear-640x360-v-cenc-golden.mp4.media_info

@@ -1,4 +1,4 @@
-bandwidth: 885151
+bandwidth: 885555
 video_info {
   codec: "avc1.64001e"
   width: 640
@@ -11,11 +11,11 @@ video_info {
 }
 init_range {
   begin: 0
-  end: 940
+  end: 1078
 }
 index_range {
-  begin: 941
+  begin: 1079
-  end: 1008
+  end: 1146
 }
 media_file_name: "place_holder"
 media_duration_seconds: 2.7360666

packager/app/test/testdata/bear-640x360-v-golden.mpd

@@ -1,11 +1,12 @@
 <?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/google/edash-packager version <tag>-<hash>-<test>-->
 <MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xlink="http://www.w3.org/1999/xlink" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" xmlns:cenc="urn:mpeg:cenc:2013" minBufferTime="PT2S" type="static" profiles="urn:mpeg:dash:profile:isoff-on-demand:2011" mediaPresentationDuration="PT2.7360665798187256S">
   <Period>
     <AdaptationSet id="0" contentType="video" width="640" height="360" frameRate="30000/1001" subsegmentAlignment="true" par="16:9">
-      <Representation id="0" bandwidth="881637" codecs="avc1.64001e" mimeType="video/mp4" sar="1:1" width="640" height="360" frameRate="30000/1001">
+      <Representation id="0" bandwidth="882040" codecs="avc1.64001e" mimeType="video/mp4" sar="1:1" width="640" height="360" frameRate="30000/1001">
         <BaseURL>output_0.mp4</BaseURL>
-        <SegmentBase indexRange="677-744" timescale="30000">
+        <SegmentBase indexRange="815-882" timescale="30000">
-          <Initialization range="0-676"/>
+          <Initialization range="0-814"/>
         </SegmentBase>
       </Representation>
     </AdaptationSet>

packager/app/test/testdata/subtitle-english-vtt-golden.mpd

@@ -1,4 +1,5 @@
 <?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/google/edash-packager version <tag>-<hash>-<test>-->
 <MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xlink="http://www.w3.org/1999/xlink" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" xmlns:cenc="urn:mpeg:cenc:2013" minBufferTime="PT2S" type="static" profiles="urn:mpeg:dash:profile:isoff-on-demand:2011" mediaPresentationDuration="PT0S">
   <Period>
     <AdaptationSet id="0" contentType="text">

packager/media/base/media_base.gyp

@@ -84,6 +84,7 @@
         '../../third_party/boringssl/boringssl.gyp:boringssl',
         '../../third_party/curl/curl.gyp:libcurl',
         '../../third_party/libxml/libxml.gyp:libxml',
+        '../../version/version.gyp:version',
       ],
     },
     {

packager/media/base/muxer_options.cc

@@ -5,6 +5,7 @@
 // https://developers.google.com/open-source/licenses/bsd
 
 #include "packager/media/base/muxer_options.h"
+#include "packager/version/version.h"
 
 namespace edash_packager {
 namespace media {
@@ -16,7 +17,8 @@ MuxerOptions::MuxerOptions()
       segment_sap_aligned(false),
       fragment_sap_aligned(false),
       num_subsegments_per_sidx(0),
-      bandwidth(0) {}
+      bandwidth(0),
+      packager_version_string(kPackagerVersion) {}
 MuxerOptions::~MuxerOptions() {}
 
 }  // namespace media

packager/media/base/muxer_options.h

@@ -66,6 +66,9 @@ struct MuxerOptions {
   /// User-specified bit rate for the media stream. If zero, the muxer will
   /// attempt to estimate.
   uint32_t bandwidth;
+
+  /// Specify the version string to be embedded in the output files.
+  std::string packager_version_string;
 };
 
 }  // namespace media

packager/media/formats/mp4/box_definitions.cc

@@ -63,6 +63,11 @@ bool IsIvSizeValid(size_t iv_size) {
 // bit(5) Reserved // 0
 const uint8_t kDdtsExtraData[] = {0xe4, 0x7c, 0, 4, 0, 0x0f, 0};
 
+// ID3v2 header: http://id3.org/id3v2.4.0-structure
+const uint32_t kID3v2HeaderSize = 10;
+const char kID3v2Identifier[] = "ID3";
+const uint16_t kID3v2Version = 0x0400;  // id3v2.4.0
+
 // Utility functions to check if the 64bit integers can fit in 32bit integer.
 bool IsFitIn32Bits(uint64_t a) {
   return a <= std::numeric_limits<uint32_t>::max();
@@ -89,6 +94,36 @@ namespace edash_packager {
 namespace media {
 namespace mp4 {
 
+namespace {
+
+TrackType FourCCToTrackType(FourCC fourcc) {
+  switch (fourcc) {
+    case FOURCC_VIDE:
+      return kVideo;
+    case FOURCC_SOUN:
+      return kAudio;
+    case FOURCC_TEXT:
+      return kText;
+    default:
+      return kInvalid;
+  }
+}
+
+FourCC TrackTypeToFourCC(TrackType track_type) {
+  switch (track_type) {
+    case kVideo:
+      return FOURCC_VIDE;
+    case kAudio:
+      return FOURCC_SOUN;
+    case kText:
+      return FOURCC_TEXT;
+    default:
+      return FOURCC_NULL;
+  }
+}
+
+}  // namespace
+
 FileType::FileType() : major_brand(FOURCC_NULL), minor_version(0) {}
 FileType::~FileType() {}
 FourCC FileType::BoxType() const { return FOURCC_FTYP; }
@@ -967,44 +1002,37 @@ uint32_t Edit::ComputeSizeInternal() {
   return HeaderSize() + list.ComputeSize();
 }
 
-HandlerReference::HandlerReference() : type(kInvalid) {}
+HandlerReference::HandlerReference() : handler_type(FOURCC_NULL) {}
 HandlerReference::~HandlerReference() {}
 FourCC HandlerReference::BoxType() const { return FOURCC_HDLR; }
 
 bool HandlerReference::ReadWriteInternal(BoxBuffer* buffer) {
-  FourCC hdlr_type = FOURCC_NULL;
   std::vector<uint8_t> handler_name;
   if (!buffer->Reading()) {
-    if (type == kVideo) {
+    switch (handler_type) {
-      hdlr_type = FOURCC_VIDE;
+      case FOURCC_VIDE:
         handler_name.assign(kVideoHandlerName,
                             kVideoHandlerName + arraysize(kVideoHandlerName));
-    } else if (type == kAudio) {
+        break;
-      hdlr_type = FOURCC_SOUN;
+      case FOURCC_SOUN:
         handler_name.assign(kAudioHandlerName,
                             kAudioHandlerName + arraysize(kAudioHandlerName));
-    } else if (type == kText) {
+        break;
-      hdlr_type = FOURCC_TEXT;
+      case FOURCC_TEXT:
         handler_name.assign(kTextHandlerName,
                             kTextHandlerName + arraysize(kTextHandlerName));
-    } else {
+        break;
+      case FOURCC_ID32:
+        break;
+      default:
         NOTIMPLEMENTED();
         return false;
     }
   }
   RCHECK(ReadWriteHeaderInternal(buffer) &&
          buffer->IgnoreBytes(4) &&  // predefined.
-         buffer->ReadWriteFourCC(&hdlr_type));
+         buffer->ReadWriteFourCC(&handler_type));
-  if (buffer->Reading()) {
+  if (!buffer->Reading()) {
-    // Note: for reading, remaining fields in box ignored.
-    if (hdlr_type == FOURCC_VIDE) {
-      type = kVideo;
-    } else if (hdlr_type == FOURCC_SOUN) {
-      type = kAudio;
-    } else {
-      type = kInvalid;
-    }
-  } else {
     RCHECK(buffer->IgnoreBytes(12) &&  // reserved.
            buffer->ReadWriteVector(&handler_name, handler_name.size()));
   }
@@ -1013,16 +1041,161 @@ bool HandlerReference::ReadWriteInternal(BoxBuffer* buffer) {
 
 uint32_t HandlerReference::ComputeSizeInternal() {
   uint32_t box_size = HeaderSize() + kFourCCSize + 16;  // 16 bytes Reserved
-  if (type == kVideo) {
+  switch (handler_type) {
+    case FOURCC_VIDE:
       box_size += sizeof(kVideoHandlerName);
-  } else if (type == kAudio) {
+      break;
+    case FOURCC_SOUN:
       box_size += sizeof(kAudioHandlerName);
-  } else {
+      break;
+    case FOURCC_TEXT:
       box_size += sizeof(kTextHandlerName);
+      break;
+    case FOURCC_ID32:
+      break;
+    default:
+      NOTIMPLEMENTED();
   }
   return box_size;
 }
 
+bool Language::ReadWrite(BoxBuffer* buffer) {
+  if (buffer->Reading()) {
+    // Read language codes into temp first then use BitReader to read the
+    // values. ISO-639-2/T language code: unsigned int(5)[3] language (2 bytes).
+    std::vector<uint8_t> temp;
+    RCHECK(buffer->ReadWriteVector(&temp, 2));
+
+    BitReader bit_reader(&temp[0], 2);
+    bit_reader.SkipBits(1);
+    char language[3];
+    for (int i = 0; i < 3; ++i) {
+      CHECK(bit_reader.ReadBits(5, &language[i]));
+      language[i] += 0x60;
+    }
+    code.assign(language, 3);
+  } else {
+    // Set up default language if it is not set.
+    const char kUndefinedLanguage[] = "und";
+    if (code.empty())
+      code = kUndefinedLanguage;
+    DCHECK_EQ(code.size(), 3u);
+
+    // Lang format: bit(1) pad, unsigned int(5)[3] language.
+    uint16_t lang = 0;
+    for (int i = 0; i < 3; ++i)
+      lang |= (code[i] - 0x60) << ((2 - i) * 5);
+    RCHECK(buffer->ReadWriteUInt16(&lang));
+  }
+  return true;
+}
+
+uint32_t Language::ComputeSize() const {
+  // ISO-639-2/T language code: unsigned int(5)[3] language (2 bytes).
+  return 2;
+}
+
+bool PrivFrame::ReadWrite(BoxBuffer* buffer) {
+  FourCC fourcc = FOURCC_PRIV;
+  RCHECK(buffer->ReadWriteFourCC(&fourcc));
+  if (fourcc != FOURCC_PRIV) {
+    VLOG(1) << "Skip unrecognized id3 frame during read: "
+            << FourCCToString(fourcc);
+    return true;
+  }
+
+  uint32_t frame_size = owner.size() + 1 + value.size();
+  // size should be encoded as synchsafe integer, which is not support here.
+  // We don't expect frame_size to be larger than 0x7F. Synchsafe integers less
+  // than 0x7F is encoded in the same way as normal integer.
+  DCHECK_LT(frame_size, 0x7Fu);
+  uint16_t flags = 0;
+  RCHECK(buffer->ReadWriteUInt32(&frame_size) &&
+         buffer->ReadWriteUInt16(&flags));
+
+  if (buffer->Reading()) {
+    std::string str;
+    RCHECK(buffer->ReadWriteString(&str, frame_size));
+    // |owner| is null terminated.
+    size_t pos = str.find('\0');
+    RCHECK(pos < str.size());
+    owner = str.substr(0, pos);
+    value = str.substr(pos + 1);
+  } else {
+    uint8_t byte = 0;  // Null terminating byte between owner and value.
+    RCHECK(buffer->ReadWriteString(&owner, owner.size()) &&
+           buffer->ReadWriteUInt8(&byte) &&
+           buffer->ReadWriteString(&value, value.size()));
+  }
+  return true;
+}
+
+uint32_t PrivFrame::ComputeSize() const {
+  if (owner.empty() && value.empty())
+    return 0;
+  const uint32_t kFourCCSize = 4;
+  return kFourCCSize + sizeof(uint32_t) + sizeof(uint16_t) + owner.size() + 1 +
+         value.size();
+}
+
+ID3v2::ID3v2() {}
+ID3v2::~ID3v2() {}
+
+FourCC ID3v2::BoxType() const { return FOURCC_ID32; }
+
+bool ID3v2::ReadWriteInternal(BoxBuffer* buffer) {
+  RCHECK(ReadWriteHeaderInternal(buffer) &&
+         language.ReadWrite(buffer));
+
+  // Read/Write ID3v2 header
+  std::string id3v2_identifier = kID3v2Identifier;
+  uint16_t version = kID3v2Version;
+  // We only support PrivateFrame in ID3.
+  uint32_t data_size = private_frame.ComputeSize();
+  // size should be encoded as synchsafe integer, which is not support here.
+  // We don't expect data_size to be larger than 0x7F. Synchsafe integers less
+  // than 0x7F is encoded in the same way as normal integer.
+  DCHECK_LT(data_size, 0x7Fu);
+  uint8_t flags = 0;
+  RCHECK(buffer->ReadWriteString(&id3v2_identifier, id3v2_identifier.size()) &&
+         buffer->ReadWriteUInt16(&version) &&
+         buffer->ReadWriteUInt8(&flags) &&
+         buffer->ReadWriteUInt32(&data_size));
+
+  RCHECK(private_frame.ReadWrite(buffer));
+  return true;
+}
+
+uint32_t ID3v2::ComputeSizeInternal() {
+  uint32_t private_frame_size = private_frame.ComputeSize();
+  // Skip ID3v2 box generation if there is no private frame.
+  return private_frame_size == 0 ? 0 : HeaderSize() + language.ComputeSize() +
+                                           kID3v2HeaderSize +
+                                           private_frame_size;
+}
+
+Metadata::Metadata() {}
+Metadata::~Metadata() {}
+
+FourCC Metadata::BoxType() const {
+  return FOURCC_META;
+}
+
+bool Metadata::ReadWriteInternal(BoxBuffer* buffer) {
+  RCHECK(ReadWriteHeaderInternal(buffer) &&
+         buffer->PrepareChildren() &&
+         buffer->ReadWriteChild(&handler) &&
+         buffer->TryReadWriteChild(&id3v2));
+  return true;
+}
+
+uint32_t Metadata::ComputeSizeInternal() {
+  uint32_t id3v2_size = id3v2.ComputeSize();
+  // Skip metadata box generation if there is no metadata box.
+  return id3v2_size == 0 ? 0
+                         : HeaderSize() + handler.ComputeSize() + id3v2_size;
+}
+
 CodecConfigurationRecord::CodecConfigurationRecord() : box_type(FOURCC_NULL) {}
 CodecConfigurationRecord::~CodecConfigurationRecord() {}
 FourCC CodecConfigurationRecord::BoxType() const {
@@ -1369,9 +1542,7 @@ uint32_t WVTTSampleEntry::ComputeSizeInternal() {
 }
 
 MediaHeader::MediaHeader()
-    : creation_time(0), modification_time(0), timescale(0), duration(0) {
+    : creation_time(0), modification_time(0), timescale(0), duration(0) {}
-  language[0] = 0;
-}
 MediaHeader::~MediaHeader() {}
 FourCC MediaHeader::BoxType() const { return FOURCC_MDHD; }
 
@@ -1382,42 +1553,16 @@ bool MediaHeader::ReadWriteInternal(BoxBuffer* buffer) {
   RCHECK(buffer->ReadWriteUInt64NBytes(&creation_time, num_bytes) &&
          buffer->ReadWriteUInt64NBytes(&modification_time, num_bytes) &&
          buffer->ReadWriteUInt32(&timescale) &&
-         buffer->ReadWriteUInt64NBytes(&duration, num_bytes));
+         buffer->ReadWriteUInt64NBytes(&duration, num_bytes) &&
-
+         language.ReadWrite(buffer) &&
-  if (buffer->Reading()) {
+         buffer->IgnoreBytes(2));  // predefined.
-    // Read language codes into temp first then use BitReader to read the
-    // values. ISO-639-2/T language code: unsigned int(5)[3] language (2 bytes).
-    std::vector<uint8_t> temp;
-    RCHECK(buffer->ReadWriteVector(&temp, 2));
-
-    BitReader bit_reader(&temp[0], 2);
-    bit_reader.SkipBits(1);
-    for (int i = 0; i < 3; ++i) {
-      CHECK(bit_reader.ReadBits(5, &language[i]));
-      language[i] += 0x60;
-    }
-    language[3] = '\0';
-  } else {
-    // Set up default language if it is not set.
-    const char kUndefinedLanguage[] = "und";
-    if (language[0] == 0)
-      strcpy(language, kUndefinedLanguage);
-
-    // Lang format: bit(1) pad, unsigned int(5)[3] language.
-    uint16_t lang = 0;
-    for (int i = 0; i < 3; ++i)
-      lang |= (language[i] - 0x60) << ((2 - i) * 5);
-    RCHECK(buffer->ReadWriteUInt16(&lang));
-  }
-
-  RCHECK(buffer->IgnoreBytes(2));  // predefined.
   return true;
 }
 
 uint32_t MediaHeader::ComputeSizeInternal() {
   version = IsFitIn32Bits(creation_time, modification_time, duration) ? 0 : 1;
   return HeaderSize() + sizeof(timescale) +
-         sizeof(uint32_t) * (1 + version) * 3 + 2 +  // 2 bytes language.
+         sizeof(uint32_t) * (1 + version) * 3 + language.ComputeSize() +
          2;  // 2 bytes predefined.
 }
 
@@ -1580,24 +1725,30 @@ FourCC Media::BoxType() const { return FOURCC_MDIA; }
 bool Media::ReadWriteInternal(BoxBuffer* buffer) {
   RCHECK(ReadWriteHeaderInternal(buffer) &&
          buffer->PrepareChildren() &&
-         buffer->ReadWriteChild(&header) &&
+         buffer->ReadWriteChild(&header));
-         buffer->ReadWriteChild(&handler));
   if (buffer->Reading()) {
+    RCHECK(buffer->ReadWriteChild(&handler));
     // Maddeningly, the HandlerReference box specifies how to parse the
     // SampleDescription box, making the latter the only box (of those that we
     // support) which cannot be parsed correctly on its own (or even with
     // information from its strict ancestor tree). We thus copy the handler type
     // to the sample description box *before* parsing it to provide this
     // information while parsing.
-    information.sample_table.description.type = handler.type;
+    information.sample_table.description.type =
+        FourCCToTrackType(handler.handler_type);
   } else {
-    DCHECK_EQ(information.sample_table.description.type, handler.type);
+    handler.handler_type =
+        TrackTypeToFourCC(information.sample_table.description.type);
+    RCHECK(handler.handler_type != FOURCC_NULL);
+    RCHECK(buffer->ReadWriteChild(&handler));
   }
   RCHECK(buffer->ReadWriteChild(&information));
   return true;
 }
 
 uint32_t Media::ComputeSizeInternal() {
+  handler.handler_type =
+      TrackTypeToFourCC(information.sample_table.description.type);
   return HeaderSize() + header.ComputeSize() + handler.ComputeSize() +
          information.ComputeSize();
 }
@@ -1702,6 +1853,7 @@ bool Movie::ReadWriteInternal(BoxBuffer* buffer) {
   RCHECK(ReadWriteHeaderInternal(buffer) &&
          buffer->PrepareChildren() &&
          buffer->ReadWriteChild(&header) &&
+         buffer->TryReadWriteChild(&metadata) &&
          buffer->TryReadWriteChild(&extends));
   if (buffer->Reading()) {
     BoxReader* reader = buffer->reader();
@@ -1718,8 +1870,8 @@ bool Movie::ReadWriteInternal(BoxBuffer* buffer) {
 }
 
 uint32_t Movie::ComputeSizeInternal() {
-  uint32_t box_size =
+  uint32_t box_size = HeaderSize() + header.ComputeSize() +
-      HeaderSize() + header.ComputeSize() + extends.ComputeSize();
+                      metadata.ComputeSize() + extends.ComputeSize();
   for (uint32_t i = 0; i < tracks.size(); ++i)
     box_size += tracks[i].ComputeSize();
   for (uint32_t i = 0; i < pssh.size(); ++i)

packager/media/formats/mp4/box_definitions.h

@@ -222,7 +222,41 @@ struct Edit : Box {
 struct HandlerReference : FullBox {
   DECLARE_BOX_METHODS(HandlerReference);
 
-  TrackType type;
+  FourCC handler_type;
+};
+
+struct Language {
+  bool ReadWrite(BoxBuffer* buffer);
+  uint32_t ComputeSize() const;
+
+  std::string code;
+};
+
+/// Implemented per http://id3.org/id3v2.4.0-frames.
+struct PrivFrame {
+  bool ReadWrite(BoxBuffer* buffer);
+  uint32_t ComputeSize() const;
+
+  std::string owner;
+  std::string value;
+};
+
+/// Implemented per http://mp4ra.org/specs.html#id3v2 and
+///                 http://id3.org/id3v2.4.0-structure.
+struct ID3v2 : FullBox {
+  DECLARE_BOX_METHODS(ID3v2);
+
+  Language language;
+
+  /// We only support PrivateFrame in ID3. Other frames are ignored.
+  PrivFrame private_frame;
+};
+
+struct Metadata : FullBox {
+  DECLARE_BOX_METHODS(Metadata);
+
+  HandlerReference handler;
+  ID3v2 id3v2;
 };
 
 struct CodecConfigurationRecord : Box {
@@ -422,8 +456,7 @@ struct MediaHeader : FullBox {
   uint64_t modification_time;
   uint32_t timescale;
   uint64_t duration;
-  // 3-char language code + 1 null terminating char.
+  Language language;
-  char language[4];
 };
 
 struct VideoMediaHeader : FullBox {
@@ -519,6 +552,7 @@ struct Movie : Box {
   DECLARE_BOX_METHODS(Movie);
 
   MovieHeader header;
+  Metadata metadata;  // Used to hold version information.
   MovieExtends extends;
   std::vector<Track> tracks;
   std::vector<ProtectionSystemSpecificHeader> pssh;

packager/media/formats/mp4/box_definitions_comparison.h

@@ -180,7 +180,24 @@ inline bool operator==(const Edit& lhs, const Edit& rhs) {
 
 inline bool operator==(const HandlerReference& lhs,
                        const HandlerReference& rhs) {
-  return lhs.type == rhs.type;
+  return lhs.handler_type == rhs.handler_type;
+}
+
+inline bool operator==(const Language& lhs,
+                       const Language& rhs) {
+  return lhs.code == rhs.code;
+}
+
+inline bool operator==(const PrivFrame& lhs, const PrivFrame& rhs) {
+  return lhs.owner == rhs.owner && lhs.value == rhs.value;
+}
+
+inline bool operator==(const ID3v2& lhs, const ID3v2& rhs) {
+  return lhs.language == rhs.language && lhs.private_frame == rhs.private_frame;
+}
+
+inline bool operator==(const Metadata& lhs, const Metadata& rhs) {
+  return lhs.handler == rhs.handler && lhs.id3v2 == rhs.id3v2;
 }
 
 inline bool operator==(const CodecConfigurationRecord& lhs,
@@ -252,7 +269,7 @@ inline bool operator==(const MediaHeader& lhs, const MediaHeader& rhs) {
   return lhs.creation_time == rhs.creation_time &&
          lhs.modification_time == rhs.modification_time &&
          lhs.timescale == rhs.timescale && lhs.duration == rhs.duration &&
-         strcmp(lhs.language, rhs.language) == 0;
+         lhs.language == rhs.language;
 }
 
 inline bool operator==(const VideoMediaHeader& lhs,

packager/media/formats/mp4/box_definitions_unittest.cc

@@ -286,11 +286,29 @@ class BoxDefinitionsTestGeneral : public testing::Test {
 
   void Modify(Edit* edts) { Modify(&edts->list); }
 
-  void Fill(HandlerReference* hdlr) {
+  void Fill(HandlerReference* hdlr) { hdlr->handler_type = FOURCC_VIDE; }
-    hdlr->type = kSampleDescriptionTrackType;
+
+  void Modify(HandlerReference* hdlr) { hdlr->handler_type = FOURCC_SOUN; }
+
+  void Fill(ID3v2* id3v2) {
+    id3v2->language.code = "eng";
+    id3v2->private_frame.owner = "edash-packager";
+    id3v2->private_frame.value = "version 1.2.0-debug";
   }
 
-  void Modify(HandlerReference* hdlr) { hdlr->type = kAudio; }
+  void Modify(ID3v2* id3v2) {
+    id3v2->language.code = "fre";
+    id3v2->private_frame.value = "version 1.3.1-release";
+  }
+
+  void Fill(Metadata* metadata) {
+    metadata->handler.handler_type = FOURCC_ID32;
+    Fill(&metadata->id3v2);
+  }
+
+  void Modify(Metadata* metadata) {
+    Modify(&metadata->id3v2);
+  }
 
   void Fill(PixelAspectRatio* pasp) {
     pasp->h_spacing = 5;
@@ -517,14 +535,14 @@ class BoxDefinitionsTestGeneral : public testing::Test {
         static_cast<uint64_t>(std::numeric_limits<uint32_t>::max()) + 1;
     mdhd->timescale = 50000;
     mdhd->duration = 250000;
-    strcpy(mdhd->language, "abc");
+    mdhd->language.code = "abc";
     mdhd->version = 1;
   }
 
   void Modify(MediaHeader* mdhd) {
     mdhd->creation_time = 2;
     mdhd->modification_time = std::numeric_limits<uint32_t>::max();
-    strcpy(mdhd->language, "und");
+    mdhd->language.code = "und";
     mdhd->version = 0;
   }
 
@@ -633,6 +651,7 @@ class BoxDefinitionsTestGeneral : public testing::Test {
 
   void Fill(Movie* moov) {
     Fill(&moov->header);
+    Fill(&moov->metadata);
     Fill(&moov->extends);
     moov->tracks.resize(2);
     Fill(&moov->tracks[0]);
@@ -854,6 +873,8 @@ class BoxDefinitionsTestGeneral : public testing::Test {
     Modify(&vttc->cue_payload);
   }
 
+  bool IsOptional(const ID3v2* box) { return true; }
+  bool IsOptional(const Metadata* box) { return true; }
   bool IsOptional(const SampleAuxiliaryInformationOffset* box) { return true; }
   bool IsOptional(const SampleAuxiliaryInformationSize* box) { return true; }
   bool IsOptional(const SampleEncryption* box) { return true; }
@@ -881,8 +902,9 @@ class BoxDefinitionsTestGeneral : public testing::Test {
   scoped_ptr<BufferWriter> buffer_;
 };
 
-typedef testing::Types<
+// GTEST support a maximum of 50 types in the template list, so we have to
-    FileType,
+// break it into two groups.
+typedef testing::Types<FileType,
                        SegmentType,
                        ProtectionSystemSpecificHeader,
                        SampleAuxiliaryInformationOffset,
@@ -897,10 +919,13 @@ typedef testing::Types<
                        EditList,
                        Edit,
                        HandlerReference,
+                       ID3v2,
+                       Metadata,
                        CodecConfigurationRecord,
                        PixelAspectRatio,
                        VideoSampleEntry,
                        ElementaryStreamDescriptor,
+                       DTSSpecific,
                        AudioSampleEntry,
                        WebVTTConfigurationBox,
                        WebVTTSourceLabelBox,
@@ -914,8 +939,9 @@ typedef testing::Types<
                        ChunkLargeOffset,
                        ChunkOffset,
                        SyncSample,
-    SampleTable,
+                       SampleTable>
-    MediaHeader,
+    Boxes;
+typedef testing::Types<MediaHeader,
                        VideoMediaHeader,
                        SoundMediaHeader,
                        SubtitleMediaHeader,
@@ -931,11 +957,7 @@ typedef testing::Types<
                        Movie,
                        TrackFragmentDecodeTime,
                        MovieFragmentHeader,
-    TrackFragmentHeader> Boxes;
+                       TrackFragmentHeader,
-
-// GTEST support a maximum of 50 types in the template list, so we have to
-// break it into two groups.
-typedef testing::Types<
                        TrackFragmentRun,
                        TrackFragment,
                        MovieFragment,
@@ -949,8 +971,8 @@ typedef testing::Types<
                        CuePayloadBox,
                        VTTEmptyCueBox,
                        VTTAdditionalTextBox,
-    VTTCueBox,
+                       VTTCueBox>
-    DTSSpecific> Boxes2;
+    Boxes2;
 
 TYPED_TEST_CASE_P(BoxDefinitionsTestGeneral);
 
@@ -1021,6 +1043,20 @@ INSTANTIATE_TYPED_TEST_CASE_P(BoxDefinitionTypedTests2,
 // Test other cases of box input.
 class BoxDefinitionsTest : public BoxDefinitionsTestGeneral<Box> {};
 
+TEST_F(BoxDefinitionsTest, MediaHandlerType) {
+  Media media;
+  Fill(&media);
+  // Clear handler type. When this box is written, it will derive handler type
+  // from sample table description.
+  media.handler.handler_type = FOURCC_NULL;
+  media.information.sample_table.description.type = kVideo;
+  media.Write(this->buffer_.get());
+
+  Media media_readback;
+  ASSERT_TRUE(ReadBack(&media_readback));
+  ASSERT_EQ(FOURCC_VIDE, media_readback.handler.handler_type);
+}
+
 TEST_F(BoxDefinitionsTest, DTSSampleEntry) {
   AudioSampleEntry entry;
   entry.format = FOURCC_DTSE;

packager/media/formats/mp4/fourccs.h

@@ -14,6 +14,8 @@ namespace mp4 {
 // TODO(rkuroiwa): Make these case sensitive. e.g. FOURCC_avc1.
 enum FourCC {
   FOURCC_NULL = 0,
+  FOURCC_ID32 = 0x49443332,
+  FOURCC_PRIV = 0x50524956,
   FOURCC_AVC1 = 0x61766331,
   FOURCC_AVCC = 0x61766343,
   FOURCC_BLOC = 0x626C6F63,

packager/media/formats/mp4/mp4_media_parser.cc

@@ -327,7 +327,7 @@ bool MP4MediaParser::ParseMoov(BoxReader* reader) {
     RCHECK(desc_idx > 0);
     desc_idx -= 1;  // BMFF descriptor index is one-based
 
-    if (track->media.handler.type == kAudio) {
+    if (samp_descr.type == kAudio) {
       RCHECK(!samp_descr.audio_entries.empty());
 
       // It is not uncommon to find otherwise-valid files with incorrect sample
@@ -428,7 +428,7 @@ bool MP4MediaParser::ParseMoov(BoxReader* reader) {
           duration,
           codec,
           AudioStreamInfo::GetCodecString(codec, audio_object_type),
-          track->media.header.language,
+          track->media.header.language.code,
           entry.samplesize,
           num_channels,
           sampling_frequency,
@@ -439,7 +439,7 @@ bool MP4MediaParser::ParseMoov(BoxReader* reader) {
           is_encrypted));
     }
 
-    if (track->media.handler.type == kVideo) {
+    if (samp_descr.type == kVideo) {
       RCHECK(!samp_descr.video_entries.empty());
       if (desc_idx >= samp_descr.video_entries.size())
         desc_idx = 0;
@@ -527,8 +527,8 @@ bool MP4MediaParser::ParseMoov(BoxReader* reader) {
       DVLOG(1) << "is_video_track_encrypted_: " << is_encrypted;
       streams.push_back(new VideoStreamInfo(
           track->header.track_id, timescale, duration, video_codec,
-          codec_string, track->media.header.language, coded_width, coded_height,
+          codec_string, track->media.header.language.code, coded_width,
-          pixel_width, pixel_height,
+          coded_height, pixel_width, pixel_height,
           0,  // trick_play_rate
           nalu_length_size, vector_as_array(&entry.codec_config_record.data),
           entry.codec_config_record.data.size(), is_encrypted));

packager/media/formats/mp4/mp4_muxer.cc

@@ -187,13 +187,12 @@ void MP4Muxer::InitializeTrak(const StreamInfo* info, Track* trak) {
   trak->media.header.timescale = info->time_scale();
   trak->media.header.duration = 0;
   if (!info->language().empty()) {
-    const size_t language_size = arraysize(trak->media.header.language) - 1;
+    // ISO-639-2/T language code should be 3 characters..
-    if (info->language().size() != language_size) {
+    if (info->language().size() != 3) {
       LOG(WARNING) << "'" << info->language() << "' is not a valid ISO-639-2 "
                    << "language code, ignoring.";
     } else {
-      memcpy(trak->media.header.language, info->language().c_str(),
+      trak->media.header.language.code = info->language();
-             language_size + 1);
     }
   }
 }
@@ -217,8 +216,6 @@ void MP4Muxer::GenerateVideoTrak(const VideoStreamInfo* video_info,
   trak->header.width = video_info->width() * sample_aspect_ratio * 0x10000;
   trak->header.height = video_info->height() * 0x10000;
 
-  trak->media.handler.type = kVideo;
-
   VideoSampleEntry video;
   video.format = VideoCodecToFourCC(video_info->codec());
   video.width = video_info->width();
@@ -241,7 +238,6 @@ void MP4Muxer::GenerateAudioTrak(const AudioStreamInfo* audio_info,
   InitializeTrak(audio_info, trak);
 
   trak->header.volume = 0x100;
-  trak->media.handler.type = kAudio;
 
   AudioSampleEntry audio;
   audio.format = AudioCodecToFourCC(audio_info->codec());

packager/media/formats/mp4/segmenter.cc

@@ -238,6 +238,13 @@ Status Segmenter::Initialize(const std::vector<MediaStream*>& streams,
   // Use the reference stream's time scale as movie time scale.
   moov_->header.timescale = sidx_->timescale;
   moof_->header.sequence_number = 1;
+
+  // Fill in version information.
+  moov_->metadata.handler.handler_type = FOURCC_ID32;
+  moov_->metadata.id3v2.language.code = "eng";
+  moov_->metadata.id3v2.private_frame.owner =
+      "https://github.com/google/edash-packager";
+  moov_->metadata.id3v2.private_frame.value = options_.packager_version_string;
   return DoInitialize();
 }
 

packager/media/formats/mp4/segmenter.h

@@ -13,6 +13,7 @@
 #include "packager/base/memory/ref_counted.h"
 #include "packager/base/memory/scoped_ptr.h"
 #include "packager/media/base/status.h"
+#include "packager/media/formats/mp4/box_definitions.h"
 
 namespace edash_packager {
 namespace media {
@@ -30,11 +31,6 @@ namespace mp4 {
 
 class Fragmenter;
 
-struct FileType;
-struct Movie;
-struct MovieFragment;
-struct SegmentIndex;
-
 /// This class defines the Segmenter which is responsible for organizing
 /// fragments into segments/subsegments and package them into a MP4 file.
 /// Inherited by MultiSegmentSegmenter and SingleSegmentSegmenter.

packager/mpd/base/mpd_builder.cc

@@ -494,7 +494,13 @@ xmlDocPtr MpdBuilder::GenerateMpd() {
   }
 
   DCHECK(doc);
-  xmlDocSetRootElement(doc.get(), mpd.Release());
+  std::string version_string =
+      "Generated with https://github.com/google/edash-packager version " +
+      mpd_options_.packager_version_string;
+  xml::scoped_xml_ptr<xmlNode> comment(
+      xmlNewDocComment(doc.get(), BAD_CAST version_string.c_str()));
+  xmlDocSetRootElement(doc.get(), comment.get());
+  xmlAddSibling(comment.release(), mpd.Release());
   return doc.release();
 }
 

packager/mpd/base/mpd_builder_unittest.cc

@@ -174,6 +174,8 @@ class DynamicMpdBuilderTest : public MpdBuilderTest<MpdBuilder::kDynamic> {
   // current time.
   void SetUp() override {
     mpd_.availability_start_time_ = "2011-12-25T12:30:00";
+    // Override packager version string for testing.
+    mpd_.mpd_options_.packager_version_string = "<tag>-<hash>-<test>";
   }
 
   MpdOptions* mutable_mpd_options() { return &mpd_.mpd_options_; }
@@ -1776,6 +1778,8 @@ TEST_F(StaticMpdBuilderTest, Text) {
 TEST_F(DynamicMpdBuilderTest, CheckMpdAttributes) {
   static const char kExpectedOutput[] =
       "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n"
+      "<!--Generated with https://github.com/google/edash-packager "
+      "version <tag>-<hash>-<test>-->\n"
       "<MPD xmlns=\"urn:mpeg:dash:schema:mpd:2011\" "
       "xmlns:xsi=\"http://www.w3.org/2001/XMLSchema-instance\" "
       "xmlns:xlink=\"http://www.w3.org/1999/xlink\" "

packager/mpd/base/mpd_options.h

@@ -7,6 +7,10 @@
 #ifndef MPD_BASE_MPD_OPTIONS_H_
 #define MPD_BASE_MPD_OPTIONS_H_
 
+#include <string>
+
+#include "packager/version/version.h"
+
 namespace edash_packager {
 
 /// Defines Mpd Options.
@@ -17,7 +21,8 @@ struct MpdOptions {
         // TODO(tinskip): Set min_buffer_time in unit tests rather than here.
         min_buffer_time(2.0),
         time_shift_buffer_depth(0),
-        suggested_presentation_delay(0) {}
+        suggested_presentation_delay(0),
+        packager_version_string(kPackagerVersion) {}
 
   ~MpdOptions() {};
 
@@ -26,6 +31,7 @@ struct MpdOptions {
   double min_buffer_time;
   double time_shift_buffer_depth;
   double suggested_presentation_delay;
+  std::string packager_version_string;
 };
 
 }  // namespace edash_packager

packager/mpd/mpd.gyp

@@ -60,6 +60,7 @@
         '../base/base.gyp:base',
         '../media/file/file.gyp:file',
         '../third_party/libxml/libxml.gyp:libxml',
+        '../version/version.gyp:version',
         'media_info_proto',
       ],
       'export_dependent_settings': [

packager/version/generate_version_string.py

@@ -0,0 +1,40 @@
+#!/usr/bin/python
+#
+# Copyright 2015 Google Inc. All rights reserved.
+#
+# Use of this source code is governed by a BSD-style
+# license that can be found in the LICENSE file or at
+# https://developers.google.com/open-source/licenses/bsd
+"""This script is used to generate version string for packager."""
+
+import subprocess
+
+# To support python version before 2.7, which does not have
+# subprocess.check_output.
+if 'check_output' not in dir(subprocess):
+
+  def check_output_implementation(*popenargs, **kwargs):
+    """Implement check_output if it is not available."""
+    if 'stdout' in kwargs:
+      raise ValueError('stdout argument not allowed, it will be overridden.')
+    process = subprocess.Popen(stdout=subprocess.PIPE, *popenargs, **kwargs)
+    output, unused_err = process.communicate()
+    retcode = process.poll()
+    if retcode:
+      cmd = kwargs.get('args')
+      if cmd is None:
+        cmd = popenargs[0]
+      raise subprocess.CalledProcessError(retcode, cmd)
+    return output
+
+  subprocess.check_output = check_output_implementation
+
+if __name__ == '__main__':
+  version_tag = subprocess.check_output(['git', 'tag', '--points-at', 'HEAD'
+                                        ]).rstrip()
+  version_hash = subprocess.check_output(['git', 'rev-parse', '--short', 'HEAD'
+                                         ]).rstrip()
+  if version_tag:
+    print '{0}-{1}'.format(version_tag, version_hash)
+  else:
+    print version_hash

packager/version/version.cc

@@ -0,0 +1,23 @@
+// Copyright 2015 Google Inc. All rights reserved.
+//
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file or at
+// https://developers.google.com/open-source/licenses/bsd
+
+#include "packager/version/version.h"
+
+namespace edash_packager {
+
+#if defined(PACKAGER_VERSION)
+// PACKAGER_VERSION is generated in gyp file using script
+// generate_version_string.py.
+#if defined(NDEBUG)
+const char kPackagerVersion[] = PACKAGER_VERSION "-release";
+#else
+const char kPackagerVersion[] = PACKAGER_VERSION "-debug";
+#endif  // #if defined(NDEBUG)
+#else
+const char kPackagerVersion[] = "";
+#endif  // #if defined(PACKAGER_VERSION)
+
+}  // namespace edash_packager

packager/version/version.gyp

@@ -0,0 +1,24 @@
+# Copyright 2015 Google Inc. All rights reserved.
+#
+# Use of this source code is governed by a BSD-style
+# license that can be found in the LICENSE file or at
+# https://developers.google.com/open-source/licenses/bsd
+
+{
+  'includes': [
+    '../common.gypi',
+  ],
+  'targets': [
+    {
+      'target_name': 'version',
+      'type': '<(component)',
+      'defines': [
+        'PACKAGER_VERSION="<!(python generate_version_string.py)"',
+      ],
+      'sources': [
+        'version.cc',
+        'version.h',
+      ],
+    },
+  ],
+}

packager/version/version.h

@@ -0,0 +1,11 @@
+// Copyright 2015 Google Inc. All rights reserved.
+//
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file or at
+// https://developers.google.com/open-source/licenses/bsd
+
+namespace edash_packager {
+
+extern const char kPackagerVersion[];
+
+}  // namespace edash_packager