schmittlauch/youtube-dl
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

[utils] Support TTML without default namespace

Browse source 
In a strict sense such TTML is invalid, but Yahoo uses it.
This commit is contained in:
Yen Chi Hsuan 2015-05-19 00:45:01 +08:00
parent 2aa64b89b3
commit 1b0427e6c4
2 changed files with 21 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

9
youtube_dl/utils.py
View file

@ -1848,9 +1848,9 @@ def dfxp2srt(dfxp_data):
out = str_or_empty(node.text)
for child in node:
if child.tag == _x('ttml:br'):
if child.tag in (_x('ttml:br'), 'br'):
out += '\n' + str_or_empty(child.tail)
elif child.tag == _x('ttml:span'):
elif child.tag in (_x('ttml:span'), 'span'):
out += str_or_empty(parse_node(child))
else:
out += str_or_empty(xml.etree.ElementTree.tostring(child))
@ -1859,7 +1859,10 @@ def dfxp2srt(dfxp_data):
dfxp = xml.etree.ElementTree.fromstring(dfxp_data.encode('utf-8'))
out = []
paras = dfxp.findall(_x('.//ttml:p'))
paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
if not paras:
raise ValueError('Invalid dfxp/TTML subtitle')
for para, index in zip(paras, itertools.count(1)):
begin_time = parse_dfxp_time_expr(para.attrib['begin'])