Add a stream descriptor field for language.

If a stream has no language metadata, or has the wrong metadata, this
field allows the user to override this on the command-line.

This also maps all specified languages on the command-line to ISO-639-2
tags as required by the MP4 muxer, so that ISO-639-1 tags specified on
the command-line do not cause a DCHECK to fail.

b/18613148

Change-Id: I473baeecbb3d388db5e06d080179ec6a332b4794

packager/app/packager_main.cc

@@ -48,7 +48,10 @@ const char kUsage[] =
     "  - bandwidth (bw): Optional value which contains a user-specified "
     "content bit rate for the stream, in bits/sec. If specified, this value is "
     "propagated to the $Bandwidth$ template parameter for segment names. "
-    "If not specified, its value may be estimated.\n";
+    "If not specified, its value may be estimated.\n"
+    "  - language (lang): Optional value which contains a user-specified "
+    "language tag. If specified, this value overrides any language metadata "
+    "in the input track.\n";
 
 enum ExitStatus {
   kSuccess = 0,
@@ -161,11 +164,11 @@ bool CreateRemuxJobs(const StreamDescriptorList& stream_descriptors,
     scoped_ptr<MuxerListener> muxer_listener;
     DCHECK(!(FLAGS_output_media_info && mpd_notifier));
     if (FLAGS_output_media_info) {
-      const std::string output_mpd_file_name =
+      const std::string output_media_info_file_name =
           stream_muxer_options.output_file_name + ".media_info";
       scoped_ptr<VodMediaInfoDumpMuxerListener>
           vod_media_info_dump_muxer_listener(
-              new VodMediaInfoDumpMuxerListener(output_mpd_file_name));
+              new VodMediaInfoDumpMuxerListener(output_media_info_file_name));
       vod_media_info_dump_muxer_listener->SetContentProtectionSchemeIdUri(
           FLAGS_scheme_id_uri);
       muxer_listener = vod_media_info_dump_muxer_listener.Pass();
@@ -185,6 +188,7 @@ bool CreateRemuxJobs(const StreamDescriptorList& stream_descriptors,
 
     if (!AddStreamToMuxer(remux_jobs->back()->demuxer()->streams(),
                           stream_iter->stream_selector,
+                          stream_iter->language,
                           muxer.get()))
       return false;
     remux_jobs->back()->AddMuxer(muxer.Pass());

packager/app/packager_util.cc

@@ -179,6 +179,7 @@ MediaStream* FindFirstAudioStream(const std::vector<MediaStream*>& streams) {
 
 bool AddStreamToMuxer(const std::vector<MediaStream*>& streams,
                       const std::string& stream_selector,
+                      const std::string& language_override,
                       Muxer* muxer) {
   DCHECK(muxer);
 
@@ -207,6 +208,11 @@ bool AddStreamToMuxer(const std::vector<MediaStream*>& streams,
     LOG(ERROR) << "No " << stream_selector << " stream found in the input.";
     return false;
   }
+
+  if (!language_override.empty()) {
+    stream->info()->set_language(language_override);
+  }
+
   muxer->AddStream(stream);
   return true;
 }

packager/app/packager_util.h

@@ -59,9 +59,12 @@ bool GetMpdOptions(edash_packager::MpdOptions* mpd_options);
 ///        "audio" to select the first audio track, "video" to select the first
 ///        video track, or a decimal number indicating which track number to
 ///        select (start at "1").
+/// @param language_override is a string which, if non-empty, overrides the
+///        stream's language metadata.
 /// @return true if successful, false otherwise.
 bool AddStreamToMuxer(const std::vector<MediaStream*>& streams,
                       const std::string& stream_selector,
+                      const std::string& language_override,
                       Muxer* muxer);
 
 }  // namespace media

packager/app/stream_descriptor.cc

@@ -10,6 +10,7 @@
 #include "packager/base/logging.h"
 #include "packager/base/strings/string_number_conversions.h"
 #include "packager/base/strings/string_split.h"
+#include "packager/mpd/base/language_utils.h"
 
 namespace edash_packager {
 namespace media {
@@ -23,6 +24,7 @@ enum FieldType {
   kOutputField,
   kSegmentTemplateField,
   kBandwidthField,
+  kLanguageField,
 };
 
 struct FieldNameToTypeMapping {
@@ -43,6 +45,8 @@ const FieldNameToTypeMapping kFieldNameTypeMappings[] = {
   { "bandwidth", kBandwidthField },
   { "bw", kBandwidthField },
   { "bitrate", kBandwidthField },
+  { "language", kLanguageField },
+  { "lang", kLanguageField },
 };
 
 FieldType GetFieldType(const std::string& field_name) {
@@ -96,6 +100,16 @@ bool InsertStreamDescriptor(const std::string& descriptor_string,
         descriptor.bandwidth = bw;
         break;
       }
+      case kLanguageField: {
+        std::string language = LanguageToISO_639_2(iter->second);
+        if (language == "und") {
+          LOG(ERROR) << "Unknown/invalid language specified: " << iter->second;
+          return false;
+        }
+        DCHECK_EQ(3, language.size());
+        descriptor.language = language;
+        break;
+      }
       default:
         LOG(ERROR) << "Unknown field in stream descriptor (\"" << iter->first
                    << "\").";

packager/app/stream_descriptor.h

@@ -26,6 +26,7 @@ struct StreamDescriptor {
   std::string output;
   std::string segment_template;
   uint32_t bandwidth;
+  std::string language;
 };
 
 class StreamDescriptorCompareFn {

packager/media/base/stream_info.h

@@ -60,6 +60,8 @@ class StreamInfo : public base::RefCountedThreadSafe<StreamInfo> {
     codec_string_ = codec_string;
   }
 
+  void set_language(const std::string& language) { language_ = language; }
+
  protected:
   friend class base::RefCountedThreadSafe<StreamInfo>;
   virtual ~StreamInfo();

packager/media/formats/mp4/mp4_muxer.cc

@@ -141,10 +141,14 @@ void MP4Muxer::InitializeTrak(const StreamInfo* info, Track* trak) {
   trak->media.header.timescale = info->time_scale();
   trak->media.header.duration = 0;
   if (!info->language().empty()) {
-    DCHECK_EQ(info->language().size(),
-              arraysize(trak->media.header.language) - 1);
-    strcpy(trak->media.header.language, info->language().c_str());
-    trak->media.header.language[info->language().size()] = '\0';
+    const size_t language_size = arraysize(trak->media.header.language) - 1;
+    if (info->language().size() != language_size) {
+      LOG(WARNING) << "'" << info->language() << "' is not a valid ISO-639-2 "
+                   << "language code, ignoring.";
+    } else {
+      memcpy(trak->media.header.language, info->language().c_str(),
+             language_size + 1);
+    }
   }
 }