R-yan/l1ving_youtube-dl
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

dfxp2srt: Fix disappearing words after <span>s ( '<span>...</span> this-goes-missing' ). Ensure trailing whitespace/newlines are not added.

Browse Source
This commit is contained in:
fnord 2015-07-15 04:29:10 -05:00
parent 056c797523
commit 7db0a3d1bc
1 changed files with 7 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
youtube_dl/utils.py
View File

@ -1850,6 +1850,12 @@ def dfxp2srt(dfxp_data):
'ttaf1': 'http://www.w3.org/2006/10/ttaf1', 'ttaf1': 'http://www.w3.org/2006/10/ttaf1',
}) })
def text_or_empty(v):
str = str_or_none(v, '')
return '' if not re.search(r'[^\s]',str,re.DOTALL) else str
def parse_node(node): def parse_node(node):
str_or_empty = functools.partial(str_or_none, default='') str_or_empty = functools.partial(str_or_none, default='')
@ -1859,7 +1865,7 @@ def dfxp2srt(dfxp_data):
if child.tag in (_x('ttml:br'), _x('ttaf1:br'), 'br'): if child.tag in (_x('ttml:br'), _x('ttaf1:br'), 'br'):
out += '\n' + str_or_empty(child.tail) out += '\n' + str_or_empty(child.tail)
elif child.tag in (_x('ttml:span'), _x('ttaf1:span'), 'span'): elif child.tag in (_x('ttml:span'), _x('ttaf1:span'), 'span'):
out += str_or_empty(parse_node(child)) out += str_or_empty(parse_node(child)) + text_or_empty(child.tail)
else: else:
out += str_or_empty(xml.etree.ElementTree.tostring(child)) out += str_or_empty(xml.etree.ElementTree.tostring(child))