Merge branch 'brightcove_in_page_embed' of https://github.com/remitamine/youtube...

[youtube-dl.git] / youtube_dl / utils.py

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index c761ea22a4580216bb2484bd08ec7e8de1d724e4..65556d056a7edfe12b94e82d0f55baa93360aa70 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -179,10 +179,19 @@ def xpath_with_ns(path, ns_map):
 
 
 def xpath_element(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
-    if sys.version_info < (2, 7):  # Crazy 2.6
-        xpath = xpath.encode('ascii')
+    def _find_xpath(xpath):
+        if sys.version_info < (2, 7):  # Crazy 2.6
+            xpath = xpath.encode('ascii')
+        return node.find(xpath)
+
+    if isinstance(xpath, (str, compat_str)):
+        n = _find_xpath(xpath)
+    else:
+        for xp in xpath:
+            n = _find_xpath(xp)
+            if n is not None:
+                break
 
-    n = node.find(xpath)
     if n is None:
         if default is not NO_DEFAULT:
             return default
@@ -250,6 +259,15 @@ def get_element_by_attribute(attribute, value, html):
     return unescapeHTML(res)
 
 
+def extract_attributes(attributes_str, attributes_regex=r'(?s)\s*([^\s=]+)\s*=\s*["\']([^"\']+)["\']'):
+    attributes = re.findall(attributes_regex, attributes_str)
+    attributes_dict = {}
+    if attributes:
+        for (attribute_name, attribute_value) in attributes:
+            attributes_dict[attribute_name] = attribute_value
+    return attributes_dict
+
+
 def clean_html(html):
     """Clean an HTML snippet into a readable string"""
 
@@ -357,7 +375,7 @@ def sanitize_path(s):
     if drive_or_unc:
         norm_path.pop(0)
     sanitized_path = [
-        path_part if path_part in ['.', '..'] else re.sub('(?:[/<>:"\\|\\\\?\\*]|\.$)', '#', path_part)
+        path_part if path_part in ['.', '..'] else re.sub('(?:[/<>:"\\|\\\\?\\*]|[\s.]$)', '#', path_part)
         for path_part in norm_path]
     if drive_or_unc:
         sanitized_path.insert(0, drive_or_unc + os.path.sep)
@@ -815,9 +833,11 @@ def parse_iso8601(date_str, delimiter='T', timezone=None):
     if date_str is None:
         return None
 
+    date_str = re.sub(r'\.[0-9]+', '', date_str)
+
     if timezone is None:
         m = re.search(
-            r'(\.[0-9]+)?(?:Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
+            r'(?:Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
             date_str)
         if not m:
             timezone = datetime.timedelta()
@@ -830,9 +850,12 @@ def parse_iso8601(date_str, delimiter='T', timezone=None):
                 timezone = datetime.timedelta(
                     hours=sign * int(m.group('hours')),
                     minutes=sign * int(m.group('minutes')))
-    date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
-    dt = datetime.datetime.strptime(date_str, date_format) - timezone
-    return calendar.timegm(dt.timetuple())
+    try:
+        date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
+        dt = datetime.datetime.strptime(date_str, date_format) - timezone
+        return calendar.timegm(dt.timetuple())
+    except ValueError:
+        pass
 
 
 def unified_strdate(date_str, day_first=True):
@@ -897,7 +920,8 @@ def unified_strdate(date_str, day_first=True):
         timetuple = email.utils.parsedate_tz(date_str)
         if timetuple:
             upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
-    return upload_date
+    if upload_date is not None:
+        return compat_str(upload_date)
 
 
 def determine_ext(url, default_ext='unknown_video'):