added youtube closed captions .srt support (see #90)
authorFilippo Valsorda <filippo.valsorda@gmail.com>
Tue, 13 Mar 2012 22:49:33 +0000 (23:49 +0100)
committerFilippo Valsorda <filippo.valsorda@gmail.com>
Tue, 13 Mar 2012 22:49:33 +0000 (23:49 +0100)
youtube-dl

index 1b381d7b75307008b3686e036417edf155d896e7..5c81973cd4ea33c1c470fa44d4940dce2f7ef90c 100755 (executable)
@@ -490,6 +490,7 @@ class FileDownloader(object):
        updatetime:       Use the Last-modified header to set output file timestamps.
        writedescription: Write the video description to a .description file
        writeinfojson:    Write the video description to a .info.json file
+       writesubtitles:   Write the video subtitles to a .srt file
        """
 
        params = None
@@ -681,6 +682,10 @@ class FileDownloader(object):
                """ Report that the description file is being written """
                self.to_screen(u'[info] Writing video description to: ' + descfn)
 
+       def report_writesubtitles(self, srtfn):
+               """ Report that the subtitles file is being written """
+               self.to_screen(u'[info] Writing video subtitles to: ' + srtfn)
+
        def report_writeinfojson(self, infofn):
                """ Report that the metadata file has been written """
                self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn)
@@ -808,6 +813,21 @@ class FileDownloader(object):
                        except (OSError, IOError):
                                self.trouble(u'ERROR: Cannot write description file ' + descfn)
                                return
+                               
+               if self.params.get('writesubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']:
+                       # subtitles download errors are already managed as troubles in relevant IE
+                       # that way it will silently go on when used with unsupporting IE 
+                       try:
+                               srtfn = filename.rsplit('.', 1)[0] + u'.srt'
+                               self.report_writesubtitles(srtfn)
+                               srtfile = open(_encodeFilename(srtfn), 'wb')
+                               try:
+                                       srtfile.write(info_dict['subtitles'].encode('utf-8'))
+                               finally:
+                                       srtfile.close()
+                       except (OSError, IOError):
+                               self.trouble(u'ERROR: Cannot write subtitles file ' + descfn)
+                               return
 
                if self.params.get('writeinfojson', False):
                        infofn = filename + u'.info.json'
@@ -1206,6 +1226,10 @@ class YoutubeIE(InfoExtractor):
                """Report attempt to download video info webpage."""
                self._downloader.to_screen(u'[youtube] %s: Downloading video info webpage' % video_id)
 
+       def report_video_subtitles_download(self, video_id):
+               """Report attempt to download video info webpage."""
+               self._downloader.to_screen(u'[youtube] %s: Downloading video subtitles' % video_id)
+
        def report_information_extraction(self, video_id):
                """Report attempt to extract video information."""
                self._downloader.to_screen(u'[youtube] %s: Extracting video information' % video_id)
@@ -1218,6 +1242,23 @@ class YoutubeIE(InfoExtractor):
                """Indicate the download will use the RTMP protocol."""
                self._downloader.to_screen(u'[youtube] RTMP download detected')
 
+       def _closed_captions_xml_to_srt(self, xml_string):
+               srt = ''
+               texts = re.findall(r'<text start="([\d\.]+)"( dur="([\d\.]+)")?>([^<]+)</text>', xml_string, re.MULTILINE)
+               # TODO parse xml instead of regex
+               for n, (start, dur_tag, dur, caption) in enumerate(texts):
+                       if not dur: dur = '4'
+                       start = float(start)
+                       end = start + float(dur)
+                       start = "%02i:%02i:%02i,%03i" %(start/(60*60), start/60%60, start%60, start%1*1000)
+                       end = "%02i:%02i:%02i,%03i" %(end/(60*60), end/60%60, end%60, end%1*1000)
+                       caption = re.sub(ur'(?u)&(.+?);', htmlentity_transform, caption)
+                       caption = re.sub(ur'(?u)&(.+?);', htmlentity_transform, caption) # double cycle, inentional
+                       srt += str(n) + '\n'
+                       srt += start + ' --> ' + end + '\n'
+                       srt += caption + '\n\n'
+               return srt
+
        def _print_formats(self, formats):
                print 'Available formats:'
                for x in formats:
@@ -1389,6 +1430,30 @@ class YoutubeIE(InfoExtractor):
                        vwebpage_doc = lxml.etree.parse(StringIO.StringIO(video_webpage), html_parser)
                        video_description = u''.join(vwebpage_doc.xpath('id("eow-description")//text()'))
                        # TODO use another parser
+                       
+               # closed captions
+               video_subtitles = None
+               if self._downloader.params.get('writesubtitles', False):
+                       self.report_video_subtitles_download(video_id)
+                       request = urllib2.Request('http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id)
+                       try:
+                               srt_list = urllib2.urlopen(request).read()
+                       except (urllib2.URLError, httplib.HTTPException, socket.error), err:
+                               self._downloader.trouble(u'WARNING: unable to download video subtitles: %s' % str(err))
+                       else:
+                               srt_lang_list = re.findall(r'lang_code="([\w\-]+)"', srt_list)
+                               if srt_lang_list:
+                                       if 'en' in srt_lang_list: srt_lang = 'en'
+                                       else: srt_lang = srt_lang_list[0] # TODO choose better and provide an override
+                                       request = urllib2.Request('http://video.google.com/timedtext?hl=en&lang=%s&v=%s' % (srt_lang, video_id))
+                                       try:
+                                               srt_xml = urllib2.urlopen(request).read()
+                                       except (urllib2.URLError, httplib.HTTPException, socket.error), err:
+                                               self._downloader.trouble(u'WARNING: unable to download video subtitles: %s' % str(err))
+                                       else:
+                                               video_subtitles = self._closed_captions_xml_to_srt(srt_xml.decode('utf-8'))
+                               else:
+                                       self._downloader.trouble(u'WARNING: video has no subtitles')
 
                # token
                video_token = urllib.unquote_plus(video_info['token'][0])
@@ -1461,6 +1526,7 @@ class YoutubeIE(InfoExtractor):
                                        'thumbnail':    video_thumbnail.decode('utf-8'),
                                        'description':  video_description,
                                        'player_url':   player_url,
+                                       'subtitles':    video_subtitles
                                })
                        except UnavailableVideoError, err:
                                self._downloader.trouble(u'\nERROR: unable to download video')
@@ -4388,6 +4454,9 @@ def parseOpts():
        filesystem.add_option('--write-info-json',
                        action='store_true', dest='writeinfojson',
                        help='write video metadata to a .info.json file', default=False)
+       filesystem.add_option('--write-srt',
+                       action='store_true', dest='writesubtitles',
+                       help='write video subtitles to a .srt file', default=False)
 
 
        postproc.add_option('--extract-audio', action='store_true', dest='extractaudio', default=False,
@@ -4588,6 +4657,7 @@ def _real_main():
                'updatetime': opts.updatetime,
                'writedescription': opts.writedescription,
                'writeinfojson': opts.writeinfojson,
+               'writesubtitles': opts.writesubtitles,
                'matchtitle': opts.matchtitle,
                'rejecttitle': opts.rejecttitle,
                'max_downloads': opts.max_downloads,