X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/3ae74f711947d73bf6627bf312edeec41cec85c3..5e6a74e1f4f6cb3624a350840a6b59faa931e40e:/youtube_dl/utils.py?ds=sidebyside diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index 201802c..057cd20 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -1,17 +1,24 @@ #!/usr/bin/env python # -*- coding: utf-8 -*- +import ctypes import datetime import email.utils import errno import gzip +import itertools import io import json import locale +import math import os +import pipes import platform import re +import ssl import socket +import struct +import subprocess import sys import traceback import zlib @@ -66,6 +73,12 @@ try: except ImportError: # Python 2 from urllib2 import HTTPError as compat_HTTPError +try: + from urllib.request import urlretrieve as compat_urlretrieve +except ImportError: # Python 2 + from urllib import urlretrieve as compat_urlretrieve + + try: from subprocess import DEVNULL compat_subprocess_get_DEVNULL = lambda: DEVNULL @@ -169,7 +182,7 @@ def compat_ord(c): compiled_regex_type = type(re.compile('')) std_headers = { - 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20100101 Firefox/10.0', + 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20100101 Firefox/10.0 (Chrome)', 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7', 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8', 'Accept-Encoding': 'gzip, deflate', @@ -213,7 +226,7 @@ if sys.version_info >= (2,7): def find_xpath_attr(node, xpath, key, val): """ Find the xpath xpath[@key=val] """ assert re.match(r'^[a-zA-Z]+$', key) - assert re.match(r'^[a-zA-Z0-9@\s]*$', val) + assert re.match(r'^[a-zA-Z0-9@\s:._]*$', val) expr = xpath + u"[@%s='%s']" % (key, val) return node.find(expr) else: @@ -223,6 +236,19 @@ else: return f return None +# On python2.6 the xml.etree.ElementTree.Element methods don't support +# the namespace parameter +def xpath_with_ns(path, ns_map): + components = [c.split(':') for c in path.split('/')] + replaced = [] + for c in components: + if len(c) == 1: + replaced.append(c[0]) + else: + ns, tag = c + replaced.append('{%s}%s' % (ns_map[ns], tag)) + return '/'.join(replaced) + def htmlentity_transform(matchobj): """Transforms an HTML entity to a character. @@ -249,7 +275,17 @@ def htmlentity_transform(matchobj): return (u'&%s;' % entity) 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 AttrParser(compat_html_parser.HTMLParser): +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 @@ -257,10 +293,9 @@ class AttrParser(compat_html_parser.HTMLParser): self.result = None self.started = False self.depth = {} - self.html = None self.watch_startpos = False self.error_count = 0 - compat_html_parser.HTMLParser.__init__(self) + BaseHTMLParser.__init__(self) def error(self, message): if self.error_count > 10 or self.started: @@ -269,11 +304,6 @@ class AttrParser(compat_html_parser.HTMLParser): self.error_count += 1 self.goahead(1) - def loads(self, html): - self.html = html - self.feed(html) - self.close() - def handle_starttag(self, tag, attrs): attrs = dict(attrs) if self.started: @@ -334,6 +364,38 @@ def get_element_by_attribute(attribute, value, html): 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') + + def get_result(self): + return self.result + +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() + def clean_html(html): """Clean an HTML snippet into a readable string""" @@ -440,12 +502,13 @@ def unescapeHTML(s): result = re.sub(u'(?u)&(.+?);', htmlentity_transform, s) return result -def encodeFilename(s): + +def encodeFilename(s, for_subprocess=False): """ @param s The name of the file """ - assert type(s) == type(u'') + assert type(s) == compat_str # Python 3 has a Unicode API if sys.version_info >= (3, 0): @@ -455,12 +518,18 @@ def encodeFilename(s): # Pass u'' directly to use Unicode APIs on Windows 2000 and up # (Detecting Windows NT 4 is tricky because 'major >= 4' would # match Windows 9x series as well. Besides, NT 4 is obsolete.) - return s + if not for_subprocess: + return s + else: + # For subprocess calls, encode with locale encoding + # Refer to http://stackoverflow.com/a/9951851/35070 + encoding = preferredencoding() else: encoding = sys.getfilesystemencoding() - if encoding is None: - encoding = 'utf-8' - return s.encode(encoding, 'ignore') + if encoding is None: + encoding = 'utf-8' + return s.encode(encoding, 'ignore') + def decodeOption(optval): if optval is None: @@ -479,19 +548,40 @@ def formatSeconds(secs): else: return '%d' % secs -def make_HTTPS_handler(opts): - if sys.version_info < (3,2): - # Python's 2.x handler is very simplistic - return compat_urllib_request.HTTPSHandler() + +def make_HTTPS_handler(opts_no_check_certificate, **kwargs): + 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_SSLv3) + except ssl.SSLError: + self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_SSLv23) + + class HTTPSHandlerV3(compat_urllib_request.HTTPSHandler): + def https_open(self, req): + return self.do_open(HTTPSConnectionV3, req) + return HTTPSHandlerV3(**kwargs) else: - import ssl - context = ssl.SSLContext(ssl.PROTOCOL_SSLv23) - context.set_default_verify_paths() - + context = ssl.SSLContext(ssl.PROTOCOL_SSLv3) context.verify_mode = (ssl.CERT_NONE - if opts.no_check_certificate + if opts_no_check_certificate else ssl.CERT_REQUIRED) - return compat_urllib_request.HTTPSHandler(context=context) + context.set_default_verify_paths() + try: + context.load_default_certs() + except AttributeError: + pass # Python < 3.4 + return compat_urllib_request.HTTPSHandler(context=context, **kwargs) class ExtractorError(Exception): """Error during info extraction.""" @@ -516,6 +606,11 @@ class ExtractorError(Exception): return u''.join(traceback.format_tb(self.traceback)) +class RegexNotFoundError(ExtractorError): + """Error when a regex didn't match""" + pass + + class DownloadError(Exception): """Download Error exception. @@ -657,19 +752,39 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler): https_request = http_request https_response = http_response + def unified_strdate(date_str): """Return a string with the date in the format YYYYMMDD""" upload_date = None #Replace commas - date_str = date_str.replace(',',' ') + date_str = date_str.replace(',', ' ') # %z (UTC offset) is only supported in python>=3.2 - date_str = re.sub(r' (\+|-)[\d]*$', '', date_str) - format_expressions = ['%d %B %Y', '%B %d %Y', '%b %d %Y', '%Y-%m-%d', '%d/%m/%Y', '%Y/%m/%d %H:%M:%S', '%d.%m.%Y %H:%M'] + date_str = re.sub(r' ?(\+|-)[0-9]{2}:?[0-9]{2}$', '', date_str) + format_expressions = [ + '%d %B %Y', + '%d %b %Y', + '%B %d %Y', + '%b %d %Y', + '%Y-%m-%d', + '%d/%m/%Y', + '%Y/%m/%d %H:%M:%S', + '%Y-%m-%d %H:%M:%S', + '%d.%m.%Y %H:%M', + '%Y-%m-%dT%H:%M:%SZ', + '%Y-%m-%dT%H:%M:%S.%fZ', + '%Y-%m-%dT%H:%M:%S.%f0Z', + '%Y-%m-%dT%H:%M:%S', + '%Y-%m-%dT%H:%M', + ] for expression in format_expressions: try: upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d') - except: + except ValueError: pass + if upload_date is None: + timetuple = email.utils.parsedate_tz(date_str) + if timetuple: + upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d') return upload_date def determine_ext(url, default_ext=u'unknown_video'): @@ -708,6 +823,15 @@ def date_from_str(date_str): return today + delta return datetime.datetime.strptime(date_str, "%Y%m%d").date() +def hyphenate_date(date_str): + """ + Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format""" + match = re.match(r'^(\d\d\d\d)(\d\d)(\d\d)$', date_str) + if match is not None: + return '-'.join(match.groups()) + else: + return date_str + class DateRange(object): """Represents a time interval between two dates""" def __init__(self, start=None, end=None): @@ -745,6 +869,28 @@ def platform_name(): return res +def write_string(s, out=None): + if out is None: + out = sys.stderr + assert type(s) == compat_str + + if ('b' in getattr(out, 'mode', '') or + sys.version_info[0] < 3): # Python 2 lies about mode of sys.stderr + s = s.encode(preferredencoding(), 'ignore') + try: + out.write(s) + except UnicodeEncodeError: + # In Windows shells, this can fail even when the codec is just charmap!? + # See https://wiki.python.org/moin/PrintFails#Issue + if sys.platform == 'win32' and hasattr(out, 'encoding'): + s = s.encode(out.encoding, 'ignore').decode(out.encoding) + out.write(s) + else: + raise + + out.flush() + + def bytes_to_intlist(bs): if not bs: return [] @@ -761,3 +907,335 @@ def intlist_to_bytes(xs): return ''.join([chr(x) for x in xs]) else: return bytes(xs) + + +def get_cachedir(params={}): + cache_root = os.environ.get('XDG_CACHE_HOME', + os.path.expanduser('~/.cache')) + return params.get('cachedir', os.path.join(cache_root, 'youtube-dl')) + + +# Cross-platform file locking +if sys.platform == 'win32': + import ctypes.wintypes + import msvcrt + + class OVERLAPPED(ctypes.Structure): + _fields_ = [ + ('Internal', ctypes.wintypes.LPVOID), + ('InternalHigh', ctypes.wintypes.LPVOID), + ('Offset', ctypes.wintypes.DWORD), + ('OffsetHigh', ctypes.wintypes.DWORD), + ('hEvent', ctypes.wintypes.HANDLE), + ] + + kernel32 = ctypes.windll.kernel32 + LockFileEx = kernel32.LockFileEx + LockFileEx.argtypes = [ + ctypes.wintypes.HANDLE, # hFile + ctypes.wintypes.DWORD, # dwFlags + ctypes.wintypes.DWORD, # dwReserved + ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow + ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh + ctypes.POINTER(OVERLAPPED) # Overlapped + ] + LockFileEx.restype = ctypes.wintypes.BOOL + UnlockFileEx = kernel32.UnlockFileEx + UnlockFileEx.argtypes = [ + ctypes.wintypes.HANDLE, # hFile + ctypes.wintypes.DWORD, # dwReserved + ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow + ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh + ctypes.POINTER(OVERLAPPED) # Overlapped + ] + UnlockFileEx.restype = ctypes.wintypes.BOOL + whole_low = 0xffffffff + whole_high = 0x7fffffff + + def _lock_file(f, exclusive): + overlapped = OVERLAPPED() + overlapped.Offset = 0 + overlapped.OffsetHigh = 0 + overlapped.hEvent = 0 + f._lock_file_overlapped_p = ctypes.pointer(overlapped) + handle = msvcrt.get_osfhandle(f.fileno()) + if not LockFileEx(handle, 0x2 if exclusive else 0x0, 0, + whole_low, whole_high, f._lock_file_overlapped_p): + raise OSError('Locking file failed: %r' % ctypes.FormatError()) + + def _unlock_file(f): + assert f._lock_file_overlapped_p + handle = msvcrt.get_osfhandle(f.fileno()) + if not UnlockFileEx(handle, 0, + whole_low, whole_high, f._lock_file_overlapped_p): + raise OSError('Unlocking file failed: %r' % ctypes.FormatError()) + +else: + import fcntl + + def _lock_file(f, exclusive): + fcntl.lockf(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH) + + def _unlock_file(f): + fcntl.lockf(f, fcntl.LOCK_UN) + + +class locked_file(object): + def __init__(self, filename, mode, encoding=None): + assert mode in ['r', 'a', 'w'] + self.f = io.open(filename, mode, encoding=encoding) + self.mode = mode + + def __enter__(self): + exclusive = self.mode != 'r' + try: + _lock_file(self.f, exclusive) + except IOError: + self.f.close() + raise + return self + + def __exit__(self, etype, value, traceback): + try: + _unlock_file(self.f) + finally: + self.f.close() + + def __iter__(self): + return iter(self.f) + + def write(self, *args): + return self.f.write(*args) + + def read(self, *args): + return self.f.read(*args) + + +def shell_quote(args): + quoted_args = [] + encoding = sys.getfilesystemencoding() + if encoding is None: + encoding = 'utf-8' + for a in args: + if isinstance(a, bytes): + # We may get a filename encoded with 'encodeFilename' + a = a.decode(encoding) + quoted_args.append(pipes.quote(a)) + return u' '.join(quoted_args) + + +def takewhile_inclusive(pred, seq): + """ Like itertools.takewhile, but include the latest evaluated element + (the first element so that Not pred(e)) """ + for e in seq: + yield e + if not pred(e): + return + + +def smuggle_url(url, data): + """ Pass additional data in a URL for internal use. """ + + sdata = compat_urllib_parse.urlencode( + {u'__youtubedl_smuggle': json.dumps(data)}) + return url + u'#' + sdata + + +def unsmuggle_url(smug_url, default=None): + if not '#__youtubedl_smuggle' in smug_url: + return smug_url, default + url, _, sdata = smug_url.rpartition(u'#') + jsond = compat_parse_qs(sdata)[u'__youtubedl_smuggle'][0] + data = json.loads(jsond) + return url, data + + +def format_bytes(bytes): + if bytes is None: + return u'N/A' + if type(bytes) is str: + bytes = float(bytes) + if bytes == 0.0: + exponent = 0 + else: + exponent = int(math.log(bytes, 1024.0)) + suffix = [u'B', u'KiB', u'MiB', u'GiB', u'TiB', u'PiB', u'EiB', u'ZiB', u'YiB'][exponent] + converted = float(bytes) / float(1024 ** exponent) + return u'%.2f%s' % (converted, suffix) + + +def str_to_int(int_str): + int_str = re.sub(r'[,\.]', u'', int_str) + return int(int_str) + + +def get_term_width(): + columns = os.environ.get('COLUMNS', None) + if columns: + return int(columns) + + try: + sp = subprocess.Popen( + ['stty', 'size'], + stdout=subprocess.PIPE, stderr=subprocess.PIPE) + out, err = sp.communicate() + return int(out.split()[1]) + except: + pass + return None + + +def month_by_name(name): + """ Return the number of a month by (locale-independently) English name """ + + ENGLISH_NAMES = [ + u'January', u'February', u'March', u'April', u'May', u'June', + u'July', u'August', u'September', u'October', u'November', u'December'] + try: + return ENGLISH_NAMES.index(name) + 1 + except ValueError: + return None + + +def fix_xml_ampersands(xml_str): + """Replace all the '&' by '&' in XML""" + return re.sub( + r'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)', + u'&', + xml_str) + + +def setproctitle(title): + assert isinstance(title, compat_str) + try: + libc = ctypes.cdll.LoadLibrary("libc.so.6") + except OSError: + return + title = title + buf = ctypes.create_string_buffer(len(title) + 1) + buf.value = title.encode('utf-8') + try: + libc.prctl(15, ctypes.byref(buf), 0, 0, 0) + except AttributeError: + return # Strange libc, just skip this + + +def remove_start(s, start): + if s.startswith(start): + return s[len(start):] + return s + + +def url_basename(url): + path = compat_urlparse.urlparse(url).path + return path.strip(u'/').split(u'/')[-1] + + +class HEADRequest(compat_urllib_request.Request): + def get_method(self): + return "HEAD" + + +def int_or_none(v, scale=1): + return v if v is None else (int(v) // scale) + + +def parse_duration(s): + if s is None: + return None + + m = re.match( + r'(?:(?:(?P[0-9]+)[:h])?(?P[0-9]+)[:m])?(?P[0-9]+)s?$', s) + if not m: + return None + res = int(m.group('secs')) + if m.group('mins'): + res += int(m.group('mins')) * 60 + if m.group('hours'): + res += int(m.group('hours')) * 60 * 60 + return res + + +def prepend_extension(filename, ext): + name, real_ext = os.path.splitext(filename) + return u'{0}.{1}{2}'.format(name, ext, real_ext) + + +def check_executable(exe, args=[]): + """ Checks if the given binary is installed somewhere in PATH, and returns its name. + args can be a list of arguments for a short output (like -version) """ + try: + subprocess.Popen([exe] + args, stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate() + except OSError: + return False + return exe + + +class PagedList(object): + def __init__(self, pagefunc, pagesize): + self._pagefunc = pagefunc + self._pagesize = pagesize + + def __len__(self): + # This is only useful for tests + return len(self.getslice()) + + def getslice(self, start=0, end=None): + res = [] + for pagenum in itertools.count(start // self._pagesize): + firstid = pagenum * self._pagesize + nextfirstid = pagenum * self._pagesize + self._pagesize + if start >= nextfirstid: + continue + + page_results = list(self._pagefunc(pagenum)) + + startv = ( + start % self._pagesize + if firstid <= start < nextfirstid + else 0) + + endv = ( + ((end - 1) % self._pagesize) + 1 + if (end is not None and firstid <= end <= nextfirstid) + else None) + + if startv != 0 or endv is not None: + page_results = page_results[startv:endv] + res.extend(page_results) + + # A little optimization - if current page is not "full", ie. does + # not contain page_size videos then we can assume that this page + # is the last one - there are no more ids on further pages - + # i.e. no need to query again. + if len(page_results) + startv < self._pagesize: + break + + # If we got the whole page, but the next page is not interesting, + # break out early as well + if end == nextfirstid: + break + return res + + +def uppercase_escape(s): + return re.sub( + r'\\U([0-9a-fA-F]{8})', + lambda m: compat_chr(int(m.group(1), base=16)), s) + +try: + struct.pack(u'!I', 0) +except TypeError: + # In Python 2.6 (and some 2.7 versions), struct requires a bytes argument + def struct_pack(spec, *args): + if isinstance(spec, compat_str): + spec = spec.encode('ascii') + return struct.pack(spec, *args) + + def struct_unpack(spec, *args): + if isinstance(spec, compat_str): + spec = spec.encode('ascii') + return struct.unpack(spec, *args) +else: + struct_pack = struct.pack + struct_unpack = struct.unpack