# Newline vs <br />
html = html.replace('\n', ' ')
- html = re.sub(r'\s*<\s*br\s*/?\s*>\s*', '\n', html)
- html = re.sub(r'<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
+ html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
+ html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
# Strip html tags
html = re.sub('<.*?>', '', html)
# Replace html entities
def try_get(src, getter, expected_type=None):
- try:
- v = getter(src)
- except (AttributeError, KeyError, TypeError, IndexError):
- pass
- else:
- if expected_type is None or isinstance(v, expected_type):
- return v
+ if not isinstance(getter, (list, tuple)):
+ getter = [getter]
+ for get in getter:
+ try:
+ v = get(src)
+ except (AttributeError, KeyError, TypeError, IndexError):
+ pass
+ else:
+ if expected_type is None or isinstance(v, expected_type):
+ return v
def encode_compat_str(string, encoding=preferredencoding(), errors='strict'):
return {
'3gpp': '3gp',
'smptett+xml': 'tt',
- 'srt': 'srt',
'ttaf+xml': 'dfxp',
'ttml+xml': 'ttml',
- 'vtt': 'vtt',
'x-flv': 'flv',
'x-mp4-fragmented': 'mp4',
'x-ms-wmv': 'wmv',
'x-mpegurl': 'm3u8',
'vnd.apple.mpegurl': 'm3u8',
'dash+xml': 'mpd',
- 'f4m': 'f4m',
'f4m+xml': 'f4m',
'hds+xml': 'f4m',
'vnd.ms-sstr+xml': 'ism',
'quicktime': 'mov',
+ 'mp2t': 'ts',
}.get(res, res)
def dfxp2srt(dfxp_data):
+ LEGACY_NAMESPACES = (
+ ('http://www.w3.org/ns/ttml', [
+ 'http://www.w3.org/2004/11/ttaf1',
+ 'http://www.w3.org/2006/04/ttaf1',
+ 'http://www.w3.org/2006/10/ttaf1',
+ ]),
+ ('http://www.w3.org/ns/ttml#styling', [
+ 'http://www.w3.org/ns/ttml#style',
+ ]),
+ )
+
+ SUPPORTED_STYLING = [
+ 'color',
+ 'fontFamily',
+ 'fontSize',
+ 'fontStyle',
+ 'fontWeight',
+ 'textDecoration'
+ ]
+
_x = functools.partial(xpath_with_ns, ns_map={
'ttml': 'http://www.w3.org/ns/ttml',
- 'ttaf1': 'http://www.w3.org/2006/10/ttaf1',
- 'ttaf1_0604': 'http://www.w3.org/2006/04/ttaf1',
+ 'tts': 'http://www.w3.org/ns/ttml#styling',
})
+ styles = {}
+ default_style = {}
+
class TTMLPElementParser(object):
- out = ''
+ _out = ''
+ _unclosed_elements = []
+ _applied_styles = []
def start(self, tag, attrib):
- if tag in (_x('ttml:br'), _x('ttaf1:br'), 'br'):
- self.out += '\n'
+ if tag in (_x('ttml:br'), 'br'):
+ self._out += '\n'
+ else:
+ unclosed_elements = []
+ style = {}
+ element_style_id = attrib.get('style')
+ if default_style:
+ style.update(default_style)
+ if element_style_id:
+ style.update(styles.get(element_style_id, {}))
+ for prop in SUPPORTED_STYLING:
+ prop_val = attrib.get(_x('tts:' + prop))
+ if prop_val:
+ style[prop] = prop_val
+ if style:
+ font = ''
+ for k, v in sorted(style.items()):
+ if self._applied_styles and self._applied_styles[-1].get(k) == v:
+ continue
+ if k == 'color':
+ font += ' color="%s"' % v
+ elif k == 'fontSize':
+ font += ' size="%s"' % v
+ elif k == 'fontFamily':
+ font += ' face="%s"' % v
+ elif k == 'fontWeight' and v == 'bold':
+ self._out += '<b>'
+ unclosed_elements.append('b')
+ elif k == 'fontStyle' and v == 'italic':
+ self._out += '<i>'
+ unclosed_elements.append('i')
+ elif k == 'textDecoration' and v == 'underline':
+ self._out += '<u>'
+ unclosed_elements.append('u')
+ if font:
+ self._out += '<font' + font + '>'
+ unclosed_elements.append('font')
+ applied_style = {}
+ if self._applied_styles:
+ applied_style.update(self._applied_styles[-1])
+ applied_style.update(style)
+ self._applied_styles.append(applied_style)
+ self._unclosed_elements.append(unclosed_elements)
def end(self, tag):
- pass
+ if tag not in (_x('ttml:br'), 'br'):
+ unclosed_elements = self._unclosed_elements.pop()
+ for element in reversed(unclosed_elements):
+ self._out += '</%s>' % element
+ if unclosed_elements and self._applied_styles:
+ self._applied_styles.pop()
def data(self, data):
- self.out += data
+ self._out += data
def close(self):
- return self.out.strip()
+ return self._out.strip()
def parse_node(node):
target = TTMLPElementParser()
parser.feed(xml.etree.ElementTree.tostring(node))
return parser.close()
+ for k, v in LEGACY_NAMESPACES:
+ for ns in v:
+ dfxp_data = dfxp_data.replace(ns, k)
+
dfxp = compat_etree_fromstring(dfxp_data.encode('utf-8'))
out = []
- paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall(_x('.//ttaf1:p')) or dfxp.findall(_x('.//ttaf1_0604:p')) or dfxp.findall('.//p')
+ paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
if not paras:
raise ValueError('Invalid dfxp/TTML subtitle')
+ repeat = False
+ while True:
+ for style in dfxp.findall(_x('.//ttml:style')):
+ style_id = style.get('id')
+ parent_style_id = style.get('style')
+ if parent_style_id:
+ if parent_style_id not in styles:
+ repeat = True
+ continue
+ styles[style_id] = styles[parent_style_id].copy()
+ for prop in SUPPORTED_STYLING:
+ prop_val = style.get(_x('tts:' + prop))
+ if prop_val:
+ styles.setdefault(style_id, {})[prop] = prop_val
+ if repeat:
+ repeat = False
+ else:
+ break
+
+ for p in ('body', 'div'):
+ ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
+ if ele is None:
+ continue
+ style = styles.get(ele.get('style'))
+ if not style:
+ continue
+ default_style.update(style)
+
for para, index in zip(paras, itertools.count(1)):
begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
end_time = parse_dfxp_time_expr(para.attrib.get('end'))