[utils] Improve xpath_text

[youtube-dl] / youtube_dl / utils.py

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index e265c75742b6d22647965e6a0261f24bb4a73a00..cc792be6ab2fb077d0f530396cd8fdc9072d8df4 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -141,7 +141,7 @@ def write_json_file(obj, fn):
 if sys.version_info >= (2, 7):
     def find_xpath_attr(node, xpath, key, val=None):
         """ Find the xpath xpath[@key=val] """
-        assert re.match(r'^[a-zA-Z-]+$', key)
+        assert re.match(r'^[a-zA-Z_-]+$', key)
         if val:
             assert re.match(r'^[a-zA-Z0-9@\s:._-]*$', val)
         expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
@@ -176,12 +176,12 @@ def xpath_with_ns(path, ns_map):
     return '/'.join(replaced)
 
 
-def xpath_text(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
+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')
 
     n = node.find(xpath)
-    if n is None or n.text is None:
+    if n is None:
         if default is not NO_DEFAULT:
             return default
         elif fatal:
@@ -189,9 +189,37 @@ def xpath_text(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
             raise ExtractorError('Could not find XML element %s' % name)
         else:
             return None
+    return n
+
+
+def xpath_text(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
+    n = xpath_element(node, xpath, name, fatal=fatal, default=default)
+    if n is None or n == default:
+        return n
+    if n.text is None:
+        if default is not NO_DEFAULT:
+            return default
+        elif fatal:
+            name = xpath if name is None else name
+            raise ExtractorError('Could not find XML element\'s text %s' % name)
+        else:
+            return None
     return n.text
 
 
+def xpath_attr(node, xpath, key, name=None, fatal=False, default=NO_DEFAULT):
+    n = find_xpath_attr(node, xpath, key)
+    if n is None:
+        if default is not NO_DEFAULT:
+            return default
+        elif fatal:
+            name = '%s[@%s]' % (xpath, key) if name is None else name
+            raise ExtractorError('Could not find XML attribute %s' % name)
+        else:
+            return None
+    return n.attrib[key]
+
+
 def get_element_by_id(id, html):
     """Return the content of the tag with the specified ID in the passed HTML document"""
     return get_element_by_attribute("id", id, html)
@@ -587,6 +615,11 @@ class ContentTooShortError(Exception):
 
 
 def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs):
+    # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
+    # expected HTTP responses to meet HTTP/1.0 or later (see also
+    # https://github.com/rg3/youtube-dl/issues/6727)
+    if sys.version_info < (3, 0):
+        kwargs['strict'] = True
     hc = http_class(*args, **kwargs)
     source_address = ydl_handler._params.get('source_address')
     if source_address is not None: