X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Futils.py;h=d22b0313460f4dc42453bcd7772bd59f5dcfe074;hb=3fcfb8e9faf3cf1dcadedd6fecc5158a86d07065;hp=2546fa45d21599e89a71ee7843cd443585c4d1f9;hpb=be5f2c192cde200b6fcc8d754fde8b5e479a300c;p=youtube-dl

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 2546fa45d..d22b03134 100644
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -411,25 +411,9 @@ def make_HTTPS_handler(params, **kwargs):
             pass
 
     if sys.version_info < (3, 2):
-        import httplib
-
-        class HTTPSConnectionV3(httplib.HTTPSConnection):
-            def __init__(self, *args, **kwargs):
-                httplib.HTTPSConnection.__init__(self, *args, **kwargs)
-
-            def connect(self):
-                sock = socket.create_connection((self.host, self.port), self.timeout)
-                if getattr(self, '_tunnel_host', False):
-                    self.sock = sock
-                    self._tunnel()
-                try:
-                    self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_TLSv1)
-                except ssl.SSLError:
-                    self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_SSLv23)
-
-        return YoutubeDLHTTPSHandler(params, https_conn_class=HTTPSConnectionV3, **kwargs)
+        return YoutubeDLHTTPSHandler(params, **kwargs)
     else:  # Python < 3.4
-        context = ssl.SSLContext(ssl.PROTOCOL_SSLv23)
+        context = ssl.SSLContext(ssl.PROTOCOL_TLSv1)
         context.verify_mode = (ssl.CERT_NONE
                                if opts_no_check_certificate
                                else ssl.CERT_REQUIRED)
@@ -560,7 +544,9 @@ def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs):
                 sock = compat_socket_create_connection(
                     (self.host, self.port), self.timeout, sa)
                 if is_https:
-                    self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file)
+                    self.sock = ssl.wrap_socket(
+                        sock, self.key_file, self.cert_file,
+                        ssl_version=ssl.PROTOCOL_TLSv1)
                 else:
                     self.sock = sock
             hc.connect = functools.partial(_hc_connect, hc)
@@ -612,17 +598,14 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
 
     def http_request(self, req):
         for h, v in std_headers.items():
-            if h not in req.headers:
+            # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
+            # The dict keys are capitalized because of this bug by urllib
+            if h.capitalize() not in req.headers:
                 req.add_header(h, v)
         if 'Youtubedl-no-compression' in req.headers:
             if 'Accept-encoding' in req.headers:
                 del req.headers['Accept-encoding']
             del req.headers['Youtubedl-no-compression']
-        if 'Youtubedl-user-agent' in req.headers:
-            if 'User-agent' in req.headers:
-                del req.headers['User-agent']
-            req.headers['User-agent'] = req.headers['Youtubedl-user-agent']
-            del req.headers['Youtubedl-user-agent']
 
         if sys.version_info < (2, 7) and '#' in req.get_full_url():
             # Python 2.6 is brain-dead when it comes to fragments
@@ -875,6 +858,9 @@ def _windows_write_string(s, out):
     except AttributeError:
         # If the output stream doesn't have a fileno, it's virtual
         return False
+    except io.UnsupportedOperation:
+        # Some strange Windows pseudo files?
+        return False
     if fileno not in WIN_OUTPUT_IDS:
         return False
 
@@ -1277,7 +1263,7 @@ def parse_duration(s):
     s = s.strip()
 
     m = re.match(
-        r'''(?ix)T?
+        r'''(?ix)(?:P?T)?
         (?:
             (?P<only_mins>[0-9.]+)\s*(?:mins?|minutes?)\s*|
             (?P<only_hours>[0-9.]+)\s*(?:hours?)|
@@ -1612,6 +1598,14 @@ def urlhandle_detect_ext(url_handle):
     except AttributeError:  # Python < 3
         getheader = url_handle.info().getheader
 
+    cd = getheader('Content-Disposition')
+    if cd:
+        m = re.match(r'attachment;\s*filename="(?P<filename>[^"]+)"', cd)
+        if m:
+            e = determine_ext(m.group('filename'), default_ext=None)
+            if e:
+                return e
+
     return getheader('Content-Type').split("/")[1]
 
 
@@ -1623,3 +1617,45 @@ def age_restricted(content_limit, age_limit):
     if content_limit is None:
         return False  # Content available for everyone
     return age_limit < content_limit
+
+
+def is_html(first_bytes):
+    """ Detect whether a file contains HTML by examining its first bytes. """
+
+    BOMS = [
+        (b'\xef\xbb\xbf', 'utf-8'),
+        (b'\x00\x00\xfe\xff', 'utf-32-be'),
+        (b'\xff\xfe\x00\x00', 'utf-32-le'),
+        (b'\xff\xfe', 'utf-16-le'),
+        (b'\xfe\xff', 'utf-16-be'),
+    ]
+    for bom, enc in BOMS:
+        if first_bytes.startswith(bom):
+            s = first_bytes[len(bom):].decode(enc, 'replace')
+            break
+    else:
+        s = first_bytes.decode('utf-8', 'replace')
+
+    return re.match(r'^\s*<', s)
+
+
+def determine_protocol(info_dict):
+    protocol = info_dict.get('protocol')
+    if protocol is not None:
+        return protocol
+
+    url = info_dict['url']
+    if url.startswith('rtmp'):
+        return 'rtmp'
+    elif url.startswith('mms'):
+        return 'mms'
+    elif url.startswith('rtsp'):
+        return 'rtsp'
+
+    ext = determine_ext(url)
+    if ext == 'm3u8':
+        return 'm3u8'
+    elif ext == 'f4m':
+        return 'f4m'
+
+    return compat_urllib_parse_urlparse(url).scheme