X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Futils.py;h=475fad3c903f9a2923def9f186c746d067807a68;hb=8fb3ac3649ca7df6f328971f58afa84dd9d05cc6;hp=4be3239268fd66e11bce4c2efd1cb26565e60883;hpb=d79323136fabc2cd72afc7c124e17797e32df514;p=youtube-dl diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index 4be323926..475fad3c9 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -17,6 +17,7 @@ import io import json import locale import math +import operator import os import pipes import platform @@ -32,6 +33,7 @@ import xml.etree.ElementTree import zlib from .compat import ( + compat_basestring, compat_chr, compat_getenv, compat_html_entities, @@ -60,6 +62,11 @@ std_headers = { } +ENGLISH_MONTH_NAMES = [ + 'January', 'February', 'March', 'April', 'May', 'June', + 'July', 'August', 'September', 'October', 'November', 'December'] + + def preferredencoding(): """Get preferred encoding. @@ -140,7 +147,7 @@ else: def find_xpath_attr(node, xpath, key, val): # 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, unicode): + if isinstance(xpath, compat_str): xpath = xpath.encode('ascii') for f in node.findall(xpath): @@ -606,11 +613,6 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler): 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 @@ -659,31 +661,37 @@ class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler): self._params = params def https_open(self, req): + kwargs = {} + if hasattr(self, '_context'): # python > 2.6 + kwargs['context'] = self._context + if hasattr(self, '_check_hostname'): # python 3.x + kwargs['check_hostname'] = self._check_hostname return self.do_open(functools.partial( _create_http_connection, self, self._https_conn_class, True), - req) + req, **kwargs) -def parse_iso8601(date_str, delimiter='T'): +def parse_iso8601(date_str, delimiter='T', timezone=None): """ Return a UNIX timestamp from the given date """ if date_str is None: return None - m = re.search( - r'(\.[0-9]+)?(?:Z$| ?(?P\+|-)(?P[0-9]{2}):?(?P[0-9]{2})$)', - date_str) - if not m: - timezone = datetime.timedelta() - else: - date_str = date_str[:-len(m.group(0))] - if not m.group('sign'): + if timezone is None: + m = re.search( + r'(\.[0-9]+)?(?:Z$| ?(?P\+|-)(?P[0-9]{2}):?(?P[0-9]{2})$)', + date_str) + if not m: timezone = datetime.timedelta() else: - sign = 1 if m.group('sign') == '+' else -1 - timezone = datetime.timedelta( - hours=sign * int(m.group('hours')), - minutes=sign * int(m.group('minutes'))) + date_str = date_str[:-len(m.group(0))] + if not m.group('sign'): + timezone = datetime.timedelta() + else: + sign = 1 if m.group('sign') == '+' else -1 + timezone = datetime.timedelta( + hours=sign * int(m.group('hours')), + minutes=sign * int(m.group('minutes'))) date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter) dt = datetime.datetime.strptime(date_str, date_format) - timezone return calendar.timegm(dt.timetuple()) @@ -700,7 +708,7 @@ def unified_strdate(date_str, day_first=True): # %z (UTC offset) is only supported in python>=3.2 date_str = re.sub(r' ?(\+|-)[0-9]{2}:?[0-9]{2}$', '', date_str) # Remove AM/PM + timezone - date_str = re.sub(r'(?i)\s*(?:AM|PM)\s+[A-Z]+', '', date_str) + date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str) format_expressions = [ '%d %B %Y', @@ -863,6 +871,9 @@ 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 @@ -889,8 +900,8 @@ def _windows_write_string(s, out): def not_a_console(handle): if handle == INVALID_HANDLE_VALUE or handle is None: return True - return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR - or GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0) + return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR or + GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0) if not_a_console(h): return False @@ -1179,11 +1190,18 @@ def get_term_width(): def month_by_name(name): """ Return the number of a month by (locale-independently) English name """ - ENGLISH_NAMES = [ - 'January', 'February', 'March', 'April', 'May', 'June', - 'July', 'August', 'September', 'October', 'November', 'December'] try: - return ENGLISH_NAMES.index(name) + 1 + return ENGLISH_MONTH_NAMES.index(name) + 1 + except ValueError: + return None + + +def month_by_abbreviation(abbrev): + """ Return the number of a month by (locale-independently) English + abbreviations """ + + try: + return [s[:3] for s in ENGLISH_MONTH_NAMES].index(abbrev) + 1 except ValueError: return None @@ -1259,7 +1277,7 @@ def float_or_none(v, scale=1, invscale=1, default=None): def parse_duration(s): - if not isinstance(s, basestring if sys.version_info < (3, 0) else compat_str): + if not isinstance(s, compat_basestring): return None s = s.strip() @@ -1271,7 +1289,10 @@ def parse_duration(s): (?P[0-9.]+)\s*(?:hours?)| (?: - (?:(?P[0-9]+)\s*(?:[:h]|hours?)\s*)? + (?: + (?:(?P[0-9]+)\s*(?:[:d]|days?)\s*)? + (?P[0-9]+)\s*(?:[:h]|hours?)\s* + )? (?P[0-9]+)\s*(?:[:m]|mins?|minutes?)\s* )? (?P[0-9]+)(?P\.[0-9]+)?\s*(?:s|secs?|seconds?)? @@ -1289,6 +1310,8 @@ def parse_duration(s): res += int(m.group('mins')) * 60 if m.group('hours'): res += int(m.group('hours')) * 60 * 60 + if m.group('days'): + res += int(m.group('days')) * 24 * 60 * 60 if m.group('ms'): res += float(m.group('ms')) return res @@ -1423,7 +1446,7 @@ def uppercase_escape(s): def escape_rfc3986(s): """Escape non-ASCII characters as suggested by RFC 3986""" - if sys.version_info < (3, 0) and isinstance(s, unicode): + if sys.version_info < (3, 0) and isinstance(s, compat_str): s = s.encode('utf-8') return compat_urllib_parse.quote(s, b"%/;:@&=+$,!~*'()?#[]") @@ -1537,9 +1560,9 @@ def js_to_json(code): return '"%s"' % v res = re.sub(r'''(?x) - "(?:[^"\\]*(?:\\\\|\\")?)*"| - '(?:[^'\\]*(?:\\\\|\\')?)*'| - [a-zA-Z_][a-zA-Z_0-9]* + "(?:[^"\\]*(?:\\\\|\\['"nu]))*[^"\\]*"| + '(?:[^'\\]*(?:\\\\|\\['"nu]))*[^'\\]*'| + [a-zA-Z_][.a-zA-Z_0-9]* ''', fix_kv, code) res = re.sub(r',(\s*\])', lambda m: m.group(1), res) return res @@ -1593,6 +1616,15 @@ def args_to_str(args): return ' '.join(shlex_quote(a) for a in args) +def mimetype2ext(mt): + _, _, res = mt.rpartition('/') + + return { + 'x-ms-wmv': 'wmv', + 'x-mp4-fragmented': 'mp4', + }.get(res, res) + + def urlhandle_detect_ext(url_handle): try: url_handle.headers @@ -1608,7 +1640,7 @@ def urlhandle_detect_ext(url_handle): if e: return e - return getheader('Content-Type').split("/")[1] + return mimetype2ext(getheader('Content-Type')) def age_restricted(content_limit, age_limit): @@ -1639,3 +1671,109 @@ def is_html(first_bytes): 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 + + +def render_table(header_row, data): + """ Render a list of rows, each as a list of values """ + table = [header_row] + data + max_lens = [max(len(compat_str(v)) for v in col) for col in zip(*table)] + format_str = ' '.join('%-' + compat_str(ml + 1) + 's' for ml in max_lens[:-1]) + '%s' + return '\n'.join(format_str % tuple(row) for row in table) + + +def _match_one(filter_part, dct): + COMPARISON_OPERATORS = { + '<': operator.lt, + '<=': operator.le, + '>': operator.gt, + '>=': operator.ge, + '=': operator.eq, + '!=': operator.ne, + } + operator_rex = re.compile(r'''(?x)\s* + (?P[a-z_]+) + \s*(?P%s)(?P\s*\?)?\s* + (?: + (?P[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)| + (?P(?![0-9.])[a-z0-9A-Z]*) + ) + \s*$ + ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys()))) + m = operator_rex.search(filter_part) + if m: + op = COMPARISON_OPERATORS[m.group('op')] + if m.group('strval') is not None: + if m.group('op') not in ('=', '!='): + raise ValueError( + 'Operator %s does not support string values!' % m.group('op')) + comparison_value = m.group('strval') + else: + try: + comparison_value = int(m.group('intval')) + except ValueError: + comparison_value = parse_filesize(m.group('intval')) + if comparison_value is None: + comparison_value = parse_filesize(m.group('intval') + 'B') + if comparison_value is None: + raise ValueError( + 'Invalid integer value %r in filter part %r' % ( + m.group('intval'), filter_part)) + actual_value = dct.get(m.group('key')) + if actual_value is None: + return m.group('none_inclusive') + return op(actual_value, comparison_value) + + UNARY_OPERATORS = { + '': lambda v: v is not None, + '!': lambda v: v is None, + } + operator_rex = re.compile(r'''(?x)\s* + (?P%s)\s*(?P[a-z_]+) + \s*$ + ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys()))) + m = operator_rex.search(filter_part) + if m: + op = UNARY_OPERATORS[m.group('op')] + actual_value = dct.get(m.group('key')) + return op(actual_value) + + raise ValueError('Invalid filter part %r' % filter_part) + + +def match_str(filter_str, dct): + """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false """ + + return all( + _match_one(filter_part, dct) for filter_part in filter_str.split('&')) + + +def match_filter_func(filter_str): + def _match_func(info_dict): + if match_str(filter_str, info_dict): + return None + else: + video_title = info_dict.get('title', info_dict.get('id', 'video')) + return '%s does not pass filter %s, skipping ..' % (video_title, filter_str) + return _match_func