[utils] Add support for zero years and months in parse_duration

[youtube-dl] / youtube_dl / utils.py

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index c9cbd58426e2c63e95d317ef81b01472d2a166c6..34866a54b6efc122f4d0edb22712c503fa448ec0 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -596,7 +596,7 @@ def unescapeHTML(s):
     assert type(s) == compat_str
 
     return re.sub(
-        r'&([^;]+;)', lambda m: _htmlentity_transform(m.group(1)), s)
+        r'&([^&;]+;)', lambda m: _htmlentity_transform(m.group(1)), s)
 
 
 def get_subprocess_encoding():
@@ -1815,6 +1815,10 @@ def float_or_none(v, scale=1, invscale=1, default=None):
         return default
 
 
+def bool_or_none(v, default=None):
+    return v if isinstance(v, bool) else default
+
+
 def strip_or_none(v):
     return None if v is None else v.strip()
 
@@ -1831,10 +1835,20 @@ def parse_duration(s):
         days, hours, mins, secs, ms = m.groups()
     else:
         m = re.match(
-            r'''(?ix)(?:P?T)?
+            r'''(?ix)(?:P?
+                (?:
+                    [0-9]+\s*y(?:ears?)?\s*
+                )?
+                (?:
+                    [0-9]+\s*m(?:onths?)?\s*
+                )?
+                (?:
+                    [0-9]+\s*w(?:eeks?)?\s*
+                )?
                 (?:
                     (?P<days>[0-9]+)\s*d(?:ays?)?\s*
                 )?
+                T)?
                 (?:
                     (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
                 )?
@@ -1929,7 +1943,7 @@ class PagedList(object):
 
 
 class OnDemandPagedList(PagedList):
-    def __init__(self, pagefunc, pagesize, use_cache=False):
+    def __init__(self, pagefunc, pagesize, use_cache=True):
         self._pagefunc = pagefunc
         self._pagesize = pagesize
         self._use_cache = use_cache
@@ -2568,14 +2582,18 @@ def srt_subtitles_timecode(seconds):
 
 
 def dfxp2srt(dfxp_data):
+    '''
+    @param dfxp_data A bytes-like object containing DFXP data
+    @returns A unicode object containing converted SRT 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',
+        (b'http://www.w3.org/ns/ttml', [
+            b'http://www.w3.org/2004/11/ttaf1',
+            b'http://www.w3.org/2006/04/ttaf1',
+            b'http://www.w3.org/2006/10/ttaf1',
         ]),
-        ('http://www.w3.org/ns/ttml#styling', [
-            'http://www.w3.org/ns/ttml#style',
+        (b'http://www.w3.org/ns/ttml#styling', [
+            b'http://www.w3.org/ns/ttml#style',
         ]),
     )
 
@@ -2670,7 +2688,7 @@ def dfxp2srt(dfxp_data):
         for ns in v:
             dfxp_data = dfxp_data.replace(ns, k)
 
-    dfxp = compat_etree_fromstring(dfxp_data.encode('utf-8'))
+    dfxp = compat_etree_fromstring(dfxp_data)
     out = []
     paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')