X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Futils.py;h=90e0ed9ab7eb36d4c9f30efffdd2868e9cfe6d47;hb=70a1165b32acf253905109e9b4f245295d67af1f;hp=b568288faad1e3a0b09f240edfa1f89301799125;hpb=91410c9bfa9fd8f01fb817474bcc7b0db5cabf95;p=youtube-dl

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index b568288fa..90e0ed9ab 100644
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -75,7 +75,7 @@ def preferredencoding():
     try:
         pref = locale.getpreferredencoding()
         'TEST'.encode(pref)
-    except:
+    except Exception:
         pref = 'UTF-8'
 
     return pref
@@ -127,7 +127,7 @@ def write_json_file(obj, fn):
             except OSError:
                 pass
         os.rename(tf.name, fn)
-    except:
+    except Exception:
         try:
             os.remove(tf.name)
         except OSError:
@@ -252,15 +252,12 @@ def sanitize_open(filename, open_mode):
             raise
 
         # In case of error, try to remove win32 forbidden chars
-        alt_filename = os.path.join(
-            re.sub('[/<>:"\\|\\\\?\\*]', '#', path_part)
-            for path_part in os.path.split(filename)
-        )
+        alt_filename = sanitize_path(filename)
         if alt_filename == filename:
             raise
         else:
             # An exception here should be caught in the caller
-            stream = open(encodeFilename(filename), open_mode)
+            stream = open(encodeFilename(alt_filename), open_mode)
             return (stream, alt_filename)
 
 
@@ -311,6 +308,31 @@ def sanitize_filename(s, restricted=False, is_id=False):
     return result
 
 
+def sanitize_path(s):
+    """Sanitizes and normalizes path on Windows"""
+    if sys.platform != 'win32':
+        return s
+    drive, _ = os.path.splitdrive(s)
+    unc, _ = os.path.splitunc(s)
+    unc_or_drive = unc or drive
+    norm_path = os.path.normpath(remove_start(s, unc_or_drive)).split(os.path.sep)
+    if unc_or_drive:
+        norm_path.pop(0)
+    sanitized_path = [
+        path_part if path_part in ['.', '..'] else re.sub('(?:[/<>:"\\|\\\\?\\*]|\.$)', '#', path_part)
+        for path_part in norm_path]
+    if unc_or_drive:
+        sanitized_path.insert(0, unc_or_drive + os.path.sep)
+    return os.path.join(*sanitized_path)
+
+
+def sanitize_url_path_consecutive_slashes(url):
+    """Collapses consecutive slashes in URLs' path"""
+    parsed_url = list(compat_urlparse.urlparse(url))
+    parsed_url[2] = re.sub(r'/{2,}', '/', parsed_url[2])
+    return compat_urlparse.urlunparse(parsed_url)
+
+
 def orderedSet(iterable):
     """ Remove all duplicates from the input iterable """
     res = []
@@ -326,7 +348,7 @@ def _htmlentity_transform(entity):
     if entity in compat_html_entities.name2codepoint:
         return compat_chr(compat_html_entities.name2codepoint[entity])
 
-    mobj = re.match(r'#(x?[0-9]+)', entity)
+    mobj = re.match(r'#(x[0-9a-fA-F]+|[0-9]+)', entity)
     if mobj is not None:
         numstr = mobj.group(1)
         if numstr.startswith('x'):
@@ -1771,10 +1793,21 @@ def match_filter_func(filter_str):
 
 
 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
+    def __init__(self, proxies=None):
+        # Set default handlers
+        for type in ('http', 'https'):
+            setattr(self, '%s_open' % type,
+                    lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
+                        meth(r, proxy, type))
+        return compat_urllib_request.ProxyHandler.__init__(self, proxies)
+
     def proxy_open(self, req, proxy, type):
-        req_proxy = req.headers.get('Ytdl-Request-Proxy')
+        req_proxy = req.headers.get('Ytdl-request-proxy')
         if req_proxy is not None:
             proxy = req_proxy
-            del req.headers['Ytdl-Request-Proxy']
+            del req.headers['Ytdl-request-proxy']
+
+        if proxy == '__noproxy__':
+            return None  # No Proxy
         return compat_urllib_request.ProxyHandler.proxy_open(
             self, req, proxy, type)