X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Futils.py;h=6d27b80c02912b01b7f64ef85f20097f8216abe4;hb=48dce58ca907921f5013367a7b22235b3a3a05df;hp=22a39a0ab7af191095242e6b2a3f580e59e17c37;hpb=25cb05bda9c8609c5e2f329f199735837b4d8c60;p=youtube-dl diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index 22a39a0ab..6d27b80c0 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -35,6 +35,7 @@ import xml.etree.ElementTree import zlib from .compat import ( + compat_HTMLParser, compat_basestring, compat_chr, compat_etree_fromstring, @@ -46,9 +47,11 @@ from .compat import ( compat_str, compat_urllib_error, compat_urllib_parse, + compat_urllib_parse_urlencode, compat_urllib_parse_urlparse, compat_urllib_request, compat_urlparse, + compat_xpath, shlex_quote, ) @@ -164,12 +167,7 @@ if sys.version_info >= (2, 7): return node.find(expr) else: def find_xpath_attr(node, xpath, key, val=None): - # Here comes the crazy part: In 2.6, if the xpath is a unicode, - # .//node does not match if a node is a direct child of . ! - if isinstance(xpath, compat_str): - xpath = xpath.encode('ascii') - - for f in node.findall(xpath): + for f in node.findall(compat_xpath(xpath)): if key not in f.attrib: continue if val is None or f.attrib.get(key) == val: @@ -194,9 +192,7 @@ def xpath_with_ns(path, ns_map): def xpath_element(node, xpath, name=None, fatal=False, default=NO_DEFAULT): def _find_xpath(xpath): - if sys.version_info < (2, 7): # Crazy 2.6 - xpath = xpath.encode('ascii') - return node.find(xpath) + return node.find(compat_xpath(xpath)) if isinstance(xpath, (str, compat_str)): n = _find_xpath(xpath) @@ -273,6 +269,38 @@ def get_element_by_attribute(attribute, value, html): return unescapeHTML(res) +class HTMLAttributeParser(compat_HTMLParser): + """Trivial HTML parser to gather the attributes for a single element""" + def __init__(self): + self.attrs = {} + compat_HTMLParser.__init__(self) + + def handle_starttag(self, tag, attrs): + self.attrs = dict(attrs) + + +def extract_attributes(html_element): + """Given a string for an HTML element such as + + Decode and return a dictionary of attributes. + { + 'a': 'foo', 'b': 'bar', c: 'baz', d: 'boz', + 'empty': '', 'noval': None, 'entity': '&', + 'sq': '"', 'dq': '\'' + }. + NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions, + but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5. + """ + parser = HTMLAttributeParser() + parser.feed(html_element) + parser.close() + return parser.attrs + + def clean_html(html): """Clean an HTML snippet into a readable string""" @@ -389,9 +417,12 @@ def sanitize_path(s): # Prepend protocol-less URLs with `http:` scheme in order to mitigate the number of # unwanted failures due to missing protocol +def sanitize_url(url): + return 'http:%s' % url if url.startswith('//') else url + + def sanitized_Request(url, *args, **kwargs): - return compat_urllib_request.Request( - 'http:%s' % url if url.startswith('//') else url, *args, **kwargs) + return compat_urllib_request.Request(sanitize_url(url), *args, **kwargs) def orderedSet(iterable): @@ -1288,7 +1319,7 @@ def shell_quote(args): def smuggle_url(url, data): """ Pass additional data in a URL for internal use. """ - sdata = compat_urllib_parse.urlencode( + sdata = compat_urllib_parse_urlencode( {'__youtubedl_smuggle': json.dumps(data)}) return url + '#' + sdata @@ -1316,6 +1347,17 @@ def format_bytes(bytes): return '%.2f%s' % (converted, suffix) +def lookup_unit_table(unit_table, s): + units_re = '|'.join(re.escape(u) for u in unit_table) + m = re.match( + r'(?P[0-9]+(?:[,.][0-9]*)?)\s*(?P%s)\b' % units_re, s) + if not m: + return None + num_str = m.group('num').replace(',', '.') + mult = unit_table[m.group('unit')] + return int(float(num_str) * mult) + + def parse_filesize(s): if s is None: return None @@ -1359,15 +1401,28 @@ def parse_filesize(s): 'Yb': 1000 ** 8, } - units_re = '|'.join(re.escape(u) for u in _UNIT_TABLE) - m = re.match( - r'(?P[0-9]+(?:[,.][0-9]*)?)\s*(?P%s)' % units_re, s) - if not m: + return lookup_unit_table(_UNIT_TABLE, s) + + +def parse_count(s): + if s is None: return None - num_str = m.group('num').replace(',', '.') - mult = _UNIT_TABLE[m.group('unit')] - return int(float(num_str) * mult) + s = s.strip() + + if re.match(r'^[\d,.]+$', s): + return str_to_int(s) + + _UNIT_TABLE = { + 'k': 1000, + 'K': 1000, + 'm': 1000 ** 2, + 'M': 1000 ** 2, + 'kk': 1000 ** 2, + 'KK': 1000 ** 2, + } + + return lookup_unit_table(_UNIT_TABLE, s) def month_by_name(name): @@ -1695,6 +1750,7 @@ def escape_url(url): """Escape URL as suggested by RFC 3986""" url_parsed = compat_urllib_parse_urlparse(url) return url_parsed._replace( + netloc=url_parsed.netloc.encode('idna').decode('ascii'), path=escape_rfc3986(url_parsed.path), params=escape_rfc3986(url_parsed.params), query=escape_rfc3986(url_parsed.query), @@ -1704,7 +1760,8 @@ def escape_url(url): try: struct.pack('!I', 0) except TypeError: - # In Python 2.6 (and some 2.7 versions), struct requires a bytes argument + # In Python 2.6 and 2.7.x < 2.7.7, struct requires a bytes argument + # See https://bugs.python.org/issue19099 def struct_pack(spec, *args): if isinstance(spec, compat_str): spec = spec.encode('ascii') @@ -1736,22 +1793,15 @@ def read_batch_urls(batch_fd): def urlencode_postdata(*args, **kargs): - return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii') + return compat_urllib_parse_urlencode(*args, **kargs).encode('ascii') def update_url_query(url, query): parsed_url = compat_urlparse.urlparse(url) qs = compat_parse_qs(parsed_url.query) qs.update(query) - qs = encode_dict(qs) return compat_urlparse.urlunparse(parsed_url._replace( - query=compat_urllib_parse.urlencode(qs, True))) - - -def encode_dict(d, encoding='utf-8'): - def encode(v): - return v.encode(encoding) if isinstance(v, compat_basestring) else v - return dict((encode(k), encode(v)) for k, v in d.items()) + query=compat_urllib_parse_urlencode(qs, True))) def dict_get(d, key_or_keys, default=None, skip_false_values=True):