X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Futils.py;h=67be4a9ae43710fb86bf2087ab3bae4421688f3d;hb=ec5f601670dfb6c39d3a4669898284bb2782dd0c;hp=b0255c943008b50b4af428d218be322b5aa68248;hpb=8c25f81beea169c9d6540eea1a6f71dc045da6ed;p=youtube-dl

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index b0255c943..67be4a9ae 100644
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -1,6 +1,8 @@
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
 
+from __future__ import unicode_literals
+
 import calendar
 import codecs
 import contextlib
@@ -8,7 +10,6 @@ import ctypes
 import datetime
 import email.utils
 import errno
-import getpass
 import gzip
 import itertools
 import io
@@ -33,7 +34,6 @@ from .compat import (
     compat_chr,
     compat_getenv,
     compat_html_entities,
-    compat_html_parser,
     compat_parse_qs,
     compat_str,
     compat_urllib_error,
@@ -73,10 +73,22 @@ def preferredencoding():
 def write_json_file(obj, fn):
     """ Encode obj as JSON and write it to fn, atomically """
 
+    if sys.version_info < (3, 0):
+        encoding = get_filesystem_encoding()
+        # os.path.basename returns a bytes object, but NamedTemporaryFile
+        # will fail if the filename contains non ascii characters unless we
+        # use a unicode object
+        path_basename = lambda f: os.path.basename(fn).decode(encoding)
+        # the same for os.path.dirname
+        path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
+    else:
+        path_basename = os.path.basename
+        path_dirname = os.path.dirname
+
     args = {
         'suffix': '.tmp',
-        'prefix': os.path.basename(fn) + '.',
-        'dir': os.path.dirname(fn),
+        'prefix': path_basename(fn) + '.',
+        'dir': path_dirname(fn),
         'delete': False,
     }
 
@@ -151,127 +163,32 @@ def xpath_text(node, xpath, name=None, fatal=False):
     return n.text
 
 
-compat_html_parser.locatestarttagend = re.compile(r"""<[a-zA-Z][-.a-zA-Z0-9:_]*(?:\s+(?:(?<=['"\s])[^\s/>][^\s/=>]*(?:\s*=+\s*(?:'[^']*'|"[^"]*"|(?!['"])[^>\s]*))?\s*)*)?\s*""", re.VERBOSE) # backport bugfix
-class BaseHTMLParser(compat_html_parser.HTMLParser):
-    def __init(self):
-        compat_html_parser.HTMLParser.__init__(self)
-        self.html = None
-
-    def loads(self, html):
-        self.html = html
-        self.feed(html)
-        self.close()
-
-class AttrParser(BaseHTMLParser):
-    """Modified HTMLParser that isolates a tag with the specified attribute"""
-    def __init__(self, attribute, value):
-        self.attribute = attribute
-        self.value = value
-        self.result = None
-        self.started = False
-        self.depth = {}
-        self.watch_startpos = False
-        self.error_count = 0
-        BaseHTMLParser.__init__(self)
-
-    def error(self, message):
-        if self.error_count > 10 or self.started:
-            raise compat_html_parser.HTMLParseError(message, self.getpos())
-        self.rawdata = '\n'.join(self.html.split('\n')[self.getpos()[0]:]) # skip one line
-        self.error_count += 1
-        self.goahead(1)
-
-    def handle_starttag(self, tag, attrs):
-        attrs = dict(attrs)
-        if self.started:
-            self.find_startpos(None)
-        if self.attribute in attrs and attrs[self.attribute] == self.value:
-            self.result = [tag]
-            self.started = True
-            self.watch_startpos = True
-        if self.started:
-            if not tag in self.depth: self.depth[tag] = 0
-            self.depth[tag] += 1
-
-    def handle_endtag(self, tag):
-        if self.started:
-            if tag in self.depth: self.depth[tag] -= 1
-            if self.depth[self.result[0]] == 0:
-                self.started = False
-                self.result.append(self.getpos())
-
-    def find_startpos(self, x):
-        """Needed to put the start position of the result (self.result[1])
-        after the opening tag with the requested id"""
-        if self.watch_startpos:
-            self.watch_startpos = False
-            self.result.append(self.getpos())
-    handle_entityref = handle_charref = handle_data = handle_comment = \
-    handle_decl = handle_pi = unknown_decl = find_startpos
-
-    def get_result(self):
-        if self.result is None:
-            return None
-        if len(self.result) != 3:
-            return None
-        lines = self.html.split('\n')
-        lines = lines[self.result[1][0]-1:self.result[2][0]]
-        lines[0] = lines[0][self.result[1][1]:]
-        if len(lines) == 1:
-            lines[-1] = lines[-1][:self.result[2][1]-self.result[1][1]]
-        lines[-1] = lines[-1][:self.result[2][1]]
-        return '\n'.join(lines).strip()
-# Hack for https://github.com/rg3/youtube-dl/issues/662
-if sys.version_info < (2, 7, 3):
-    AttrParser.parse_endtag = (lambda self, i:
-        i + len("</scr'+'ipt>")
-        if self.rawdata[i:].startswith("</scr'+'ipt>")
-        else compat_html_parser.HTMLParser.parse_endtag(self, i))
-
 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)
 
+
 def get_element_by_attribute(attribute, value, html):
     """Return the content of the tag with the specified attribute in the passed HTML document"""
-    parser = AttrParser(attribute, value)
-    try:
-        parser.loads(html)
-    except compat_html_parser.HTMLParseError:
-        pass
-    return parser.get_result()
 
-class MetaParser(BaseHTMLParser):
-    """
-    Modified HTMLParser that isolates a meta tag with the specified name 
-    attribute.
-    """
-    def __init__(self, name):
-        BaseHTMLParser.__init__(self)
-        self.name = name
-        self.content = None
-        self.result = None
-
-    def handle_starttag(self, tag, attrs):
-        if tag != 'meta':
-            return
-        attrs = dict(attrs)
-        if attrs.get('name') == self.name:
-            self.result = attrs.get('content')
+    m = re.search(r'''(?xs)
+        <([a-zA-Z0-9:._-]+)
+         (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
+         \s+%s=['"]?%s['"]?
+         (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
+        \s*>
+        (?P<content>.*?)
+        </\1>
+    ''' % (re.escape(attribute), re.escape(value)), html)
 
-    def get_result(self):
-        return self.result
+    if not m:
+        return None
+    res = m.group('content')
 
-def get_meta_content(name, html):
-    """
-    Return the content attribute from the meta tag with the given name attribute.
-    """
-    parser = MetaParser(name)
-    try:
-        parser.loads(html)
-    except compat_html_parser.HTMLParseError:
-        pass
-    return parser.get_result()
+    if res.startswith('"') or res.startswith("'"):
+        res = res[1:-1]
+
+    return unescapeHTML(res)
 
 
 def clean_html(html):
@@ -938,10 +855,7 @@ def bytes_to_intlist(bs):
 def intlist_to_bytes(xs):
     if not xs:
         return b''
-    if isinstance(chr(0), bytes):  # Python 2
-        return ''.join([chr(x) for x in xs])
-    else:
-        return bytes(xs)
+    return struct_pack('%dB' % len(xs), *xs)
 
 
 # Cross-platform file locking
@@ -1338,7 +1252,7 @@ def escape_rfc3986(s):
     """Escape non-ASCII characters as suggested by RFC 3986"""
     if sys.version_info < (3, 0) and isinstance(s, unicode):
         s = s.encode('utf-8')
-    return compat_urllib_parse.quote(s, "%/;:@&=+$,!~*'()?#[]")
+    return compat_urllib_parse.quote(s, b"%/;:@&=+$,!~*'()?#[]")
 
 
 def escape_url(url):
@@ -1429,7 +1343,8 @@ def parse_age_limit(s):
 
 
 def strip_jsonp(code):
-    return re.sub(r'(?s)^[a-zA-Z0-9_]+\s*\(\s*(.*)\);?\s*?\s*$', r'\1', code)
+    return re.sub(
+        r'(?s)^[a-zA-Z0-9_]+\s*\(\s*(.*)\);?\s*?(?://[^\n]*)*$', r'\1', code)
 
 
 def js_to_json(code):