Add postprocessor for converting subtitles (closes #4954)

4 files changed, 52 insertions, 0 deletions

diff --git a/youtube_dl/__init__.py b/youtube_dl/__init__.py
index 5ce201800..49f382695 100644
--- a/youtube_dl/__init__.py
+++ b/youtube_dl/__init__.py
@@ -170,6 +170,9 @@ def _real_main(argv=None):
     if opts.recodevideo is not None:
         if opts.recodevideo not in ['mp4', 'flv', 'webm', 'ogg', 'mkv']:
             parser.error('invalid video recode format specified')
+    if opts.convertsubtitles is not None:
+        if opts.convertsubtitles not in ['srt', 'vtt', 'ass']:
+            parser.error('invalid subtitle format specified')
 
     if opts.date is not None:
         date = DateRange.day(opts.date)
@@ -223,6 +226,11 @@ def _real_main(argv=None):
             'key': 'FFmpegVideoConvertor',
             'preferedformat': opts.recodevideo,
         })
+    if opts.convertsubtitles:
+        postprocessors.append({
+            'key': 'FFmpegSubtitlesConvertor',
+            'format': opts.convertsubtitles,
+        })
     if opts.embedsubtitles:
         postprocessors.append({
             'key': 'FFmpegEmbedSubtitle',
diff --git a/youtube_dl/options.py b/youtube_dl/options.py
index 886ce9613..58f811162 100644
--- a/youtube_dl/options.py
+++ b/youtube_dl/options.py
@@ -751,6 +751,10 @@ def parseOpts(overrideArguments=None):
         '--exec',
         metavar='CMD', dest='exec_cmd',
         help='Execute a command on the file after downloading, similar to find\'s -exec syntax. Example: --exec \'adb push {} /sdcard/Music/ && rm {}\'')
+    postproc.add_option(
+        '--convert-subtitles', '--convert-subs',
+        metavar='FORMAT', dest='convertsubtitles', default=None,
+        help='Convert the subtitles to other format (currently supported: srt|ass|vtt)')
 
     parser.add_option_group(general)
     parser.add_option_group(network)
diff --git a/youtube_dl/postprocessor/__init__.py b/youtube_dl/postprocessor/__init__.py
index 0ffbca258..708df3dd4 100644
--- a/youtube_dl/postprocessor/__init__.py
+++ b/youtube_dl/postprocessor/__init__.py
@@ -11,6 +11,7 @@ from .ffmpeg import (
     FFmpegMergerPP,
     FFmpegMetadataPP,
     FFmpegVideoConvertorPP,
+    FFmpegSubtitlesConvertorPP,
 )
 from .xattrpp import XAttrMetadataPP
 from .execafterdownload import ExecAfterDownloadPP
@@ -31,6 +32,7 @@ __all__ = [
     'FFmpegMergerPP',
     'FFmpegMetadataPP',
     'FFmpegPostProcessor',
+    'FFmpegSubtitlesConvertorPP',
     'FFmpegVideoConvertorPP',
     'XAttrMetadataPP',
 ]
diff --git a/youtube_dl/postprocessor/ffmpeg.py b/youtube_dl/postprocessor/ffmpeg.py
index 398fe050e..30094c2f3 100644
--- a/youtube_dl/postprocessor/ffmpeg.py
+++ b/youtube_dl/postprocessor/ffmpeg.py
@@ -1,5 +1,6 @@
 from __future__ import unicode_literals
 
+import io
 import os
 import subprocess
 import sys
@@ -635,3 +636,40 @@ class FFmpegFixupM4aPP(FFmpegPostProcessor):
         os.rename(encodeFilename(temp_filename), encodeFilename(filename))
 
         return True, info
+
+
+class FFmpegSubtitlesConvertorPP(FFmpegPostProcessor):
+    def __init__(self, downloader=None, format=None):
+        super(FFmpegSubtitlesConvertorPP, self).__init__(downloader)
+        self.format = format
+
+    def run(self, info):
+        subs = info.get('requested_subtitles')
+        filename = info['filepath']
+        new_ext = self.format
+        new_format = new_ext
+        if new_format == 'vtt':
+            new_format = 'webvtt'
+        if subs is None:
+            self._downloader.to_screen('[ffmpeg] There aren\'t any subtitles to convert')
+            return True, info
+        self._downloader.to_screen('[ffmpeg] Converting subtitles')
+        for lang, sub in subs.items():
+            ext = sub['ext']
+            if ext == new_ext:
+                self._downloader.to_screen(
+                    '[ffmpeg] Subtitle file for %s is already in the requested'
+                    'format' % new_ext)
+                continue
+            new_file = subtitles_filename(filename, lang, new_ext)
+            self.run_ffmpeg(
+                subtitles_filename(filename, lang, ext),
+                new_file, ['-f', new_format])
+
+            with io.open(new_file, 'rt', encoding='utf-8') as f:
+                subs[lang] = {
+                    'ext': ext,
+                    'data': f.read(),
+                }
+
+        return True, info