[crunchycroll] Fix building of ass subtitles (reported in #4019)

Parse the xml document instead of using regexes, otherwise unicode characters are left unescaped.
This commit is contained in:
Jaime Marquínez Ferrándiz 2014-10-29 21:19:20 +01:00
parent ac645ac7d0
commit d65d628613

View File

@ -109,19 +109,17 @@ class CrunchyrollIE(SubtitlesInfoExtractor):
decrypted_data = intlist_to_bytes(aes_cbc_decrypt(data, key, iv)) decrypted_data = intlist_to_bytes(aes_cbc_decrypt(data, key, iv))
return zlib.decompress(decrypted_data) return zlib.decompress(decrypted_data)
def _convert_subtitles_to_srt(self, subtitles): def _convert_subtitles_to_srt(self, sub_root):
output = '' output = ''
for i, (start, end, text) in enumerate(re.findall(r'<event [^>]*?start="([^"]+)" [^>]*?end="([^"]+)" [^>]*?text="([^"]+)"[^>]*?>', subtitles), 1):
start = start.replace('.', ',') for i, event in enumerate(sub_root.findall('./events/event'), 1):
end = end.replace('.', ',') start = event.attrib['start'].replace('.', ',')
text = clean_html(text) end = event.attrib['end'].replace('.', ',')
text = text.replace('\\N', '\n') text = event.attrib['text'].replace('\\N', '\n')
if not text:
continue
output += '%d\n%s --> %s\n%s\n\n' % (i, start, end, text) output += '%d\n%s --> %s\n%s\n\n' % (i, start, end, text)
return output return output
def _convert_subtitles_to_ass(self, subtitles): def _convert_subtitles_to_ass(self, sub_root):
output = '' output = ''
def ass_bool(strvalue): def ass_bool(strvalue):
@ -130,10 +128,6 @@ class CrunchyrollIE(SubtitlesInfoExtractor):
assvalue = '-1' assvalue = '-1'
return assvalue return assvalue
sub_root = xml.etree.ElementTree.fromstring(subtitles)
if not sub_root:
return output
output = '[Script Info]\n' output = '[Script Info]\n'
output += 'Title: %s\n' % sub_root.attrib["title"] output += 'Title: %s\n' % sub_root.attrib["title"]
output += 'ScriptType: v4.00+\n' output += 'ScriptType: v4.00+\n'
@ -270,10 +264,13 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False) lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False)
if not lang_code: if not lang_code:
continue continue
sub_root = xml.etree.ElementTree.fromstring(subtitle)
if not sub_root:
subtitles[lang_code] = ''
if sub_format == 'ass': if sub_format == 'ass':
subtitles[lang_code] = self._convert_subtitles_to_ass(subtitle) subtitles[lang_code] = self._convert_subtitles_to_ass(sub_root)
else: else:
subtitles[lang_code] = self._convert_subtitles_to_srt(subtitle) subtitles[lang_code] = self._convert_subtitles_to_srt(sub_root)
if self._downloader.params.get('listsubtitles', False): if self._downloader.params.get('listsubtitles', False):
self._list_available_subtitles(video_id, subtitles) self._list_available_subtitles(video_id, subtitles)