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

[youtube-dl] / youtube_dl / utils.py

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 558c9c7d5a21c646a11720221fd1648c137ac68e..65556d056a7edfe12b94e82d0f55baa93360aa70 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -36,6 +36,7 @@ import zlib
 from .compat import (
     compat_basestring,
     compat_chr,
+    compat_etree_fromstring,
     compat_html_entities,
     compat_http_client,
     compat_kwargs,
@@ -178,10 +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
@@ -249,6 +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"""
 
@@ -356,7 +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)
@@ -901,7 +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'):
@@ -1656,29 +1676,6 @@ def encode_dict(d, encoding='utf-8'):
     return dict((k.encode(encoding), v.encode(encoding)) for k, v in d.items())
 
 
-try:
-    etree_iter = xml.etree.ElementTree.Element.iter
-except AttributeError:  # Python <=2.6
-    etree_iter = lambda n: n.findall('.//*')
-
-
-def parse_xml(s):
-    class TreeBuilder(xml.etree.ElementTree.TreeBuilder):
-        def doctype(self, name, pubid, system):
-            pass  # Ignore doctypes
-
-    parser = xml.etree.ElementTree.XMLParser(target=TreeBuilder())
-    kwargs = {'parser': parser} if sys.version_info >= (2, 7) else {}
-    tree = xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs)
-    # Fix up XML parser in Python 2.x
-    if sys.version_info < (3, 0):
-        for n in etree_iter(tree):
-            if n.text is not None:
-                if not isinstance(n.text, compat_str):
-                    n.text = n.text.decode('utf-8')
-    return tree
-
-
 US_RATINGS = {
     'G': 0,
     'PG': 10,
@@ -1979,7 +1976,7 @@ def dfxp2srt(dfxp_data):
 
         return out
 
-    dfxp = xml.etree.ElementTree.fromstring(dfxp_data.encode('utf-8'))
+    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('.//p')