From 1b0427e6c433c0b6db5e210db6e3173e19e702ed Mon Sep 17 00:00:00 2001 From: Yen Chi Hsuan Date: Tue, 19 May 2015 00:45:01 +0800 Subject: [PATCH] [utils] Support TTML without default namespace In a strict sense such TTML is invalid, but Yahoo uses it. --- test/test_utils.py | 15 +++++++++++++++ youtube_dl/utils.py | 9 ++++++--- 2 files changed, 21 insertions(+), 3 deletions(-) diff --git a/test/test_utils.py b/test/test_utils.py index b401070371..e13e11b59f 100644 --- a/test/test_utils.py +++ b/test/test_utils.py @@ -621,6 +621,21 @@ def test_dfxp2srt(self): ''' self.assertEqual(dfxp2srt(dfxp_data), srt_data) + dfxp_data_no_default_namespace = ''' + + +
+

The first line

+
+ +
''' + srt_data = '''1 +00:00:00,000 --> 00:00:01,000 +The first line + +''' + self.assertEqual(dfxp2srt(dfxp_data_no_default_namespace), srt_data) + if __name__ == '__main__': unittest.main() diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index ed9ed9ed63..507f073834 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -1848,9 +1848,9 @@ def parse_node(node): 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 parse_node(node): 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'])