X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/e8cd8c4bd832446f1971215b9fedc4531555dc1a..252064fa082f84e06d1251185428b44c75c59109:/youtube_dl/utils.py?ds=sidebyside diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index 73fe1ad..64a9618 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -1,11 +1,16 @@ #!/usr/bin/env python # -*- coding: utf-8 -*- +import calendar +import codecs +import contextlib import ctypes import datetime import email.utils import errno +import getpass import gzip +import itertools import io import json import locale @@ -16,9 +21,11 @@ import platform import re import ssl import socket +import struct import subprocess import sys import traceback +import xml.etree.ElementTree import zlib try: @@ -172,6 +179,11 @@ try: except NameError: compat_chr = chr +try: + from xml.etree.ElementTree import ParseError as compat_xml_parse_error +except ImportError: # Python 2.6 + from xml.parsers.expat import ExpatError as compat_xml_parse_error + def compat_ord(c): if type(c) is int: return c else: return ord(c) @@ -224,7 +236,7 @@ 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: @@ -491,13 +503,13 @@ def orderedSet(iterable): res.append(el) return res + def unescapeHTML(s): - """ - @param s a string - """ - assert type(s) == type(u'') + if s is None: + return None + assert type(s) == compat_str - result = re.sub(u'(?u)&(.+?);', htmlentity_transform, s) + result = re.sub(r'(?u)&(.+?);', htmlentity_transform, s) return result @@ -529,6 +541,15 @@ def encodeFilename(s, for_subprocess=False): return s.encode(encoding, 'ignore') +def encodeArgument(s): + if not isinstance(s, compat_str): + # Legacy code that uses byte strings + # Uncomment the following line after fixing all post processors + #assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s)) + s = s.decode('ascii') + return encodeFilename(s, True) + + def decodeOption(optval): if optval is None: return optval @@ -583,13 +604,15 @@ def make_HTTPS_handler(opts_no_check_certificate, **kwargs): class ExtractorError(Exception): """Error during info extraction.""" - def __init__(self, msg, tb=None, expected=False, cause=None): + def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None): """ tb, if given, is the original traceback (so that it can be printed out). If expected is set, this is a normal error message and most likely not a bug in youtube-dl. """ if sys.exc_info()[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError): expected = True + if video_id is not None: + msg = video_id + ': ' + msg if not expected: msg = msg + u'; please report this issue on https://yt-dl.org/bug . Be sure to call youtube-dl with the --verbose flag and include its complete output. Make sure you are using the latest version; type youtube-dl -U to update.' super(ExtractorError, self).__init__(msg) @@ -597,6 +620,7 @@ class ExtractorError(Exception): self.traceback = tb self.exc_info = sys.exc_info() # preserve original exception self.cause = cause + self.video_id = video_id def format_traceback(self): if self.traceback is None: @@ -750,31 +774,69 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler): https_request = http_request https_response = http_response + +def parse_iso8601(date_str, delimiter='T'): + """ Return a UNIX timestamp from the given date """ + + if date_str is None: + return None + + m = re.search( + r'Z$| ?(?P\+|-)(?P[0-9]{2}):?(?P[0-9]{2})$', + date_str) + if not m: + timezone = datetime.timedelta() + else: + date_str = date_str[:-len(m.group(0))] + if not m.group('sign'): + timezone = datetime.timedelta() + else: + sign = 1 if m.group('sign') == '+' else -1 + timezone = datetime.timedelta( + hours=sign * int(m.group('hours')), + minutes=sign * int(m.group('minutes'))) + date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter) + dt = datetime.datetime.strptime(date_str, date_format) - timezone + return calendar.timegm(dt.timetuple()) + + def unified_strdate(date_str): """Return a string with the date in the format YYYYMMDD""" + + if date_str is None: + return None + upload_date = None #Replace commas - date_str = date_str.replace(',',' ') + date_str = date_str.replace(',', ' ') # %z (UTC offset) is only supported in python>=3.2 - date_str = re.sub(r' (\+|-)[\d]*$', '', date_str) + 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', + '%b %dst %Y %I:%M%p', + '%b %dnd %Y %I:%M%p', + '%b %dth %Y %I:%M%p', '%Y-%m-%d', + '%d.%m.%Y', '%d/%m/%Y', '%Y/%m/%d %H:%M:%S', '%Y-%m-%d %H:%M:%S', '%d.%m.%Y %H:%M', + '%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:%S.%f', + '%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) @@ -864,25 +926,97 @@ def platform_name(): return res -def write_string(s, out=None): +def _windows_write_string(s, out): + """ Returns True if the string was written using special methods, + False if it has yet to be written out.""" + # Adapted from http://stackoverflow.com/a/3259271/35070 + + import ctypes + import ctypes.wintypes + + WIN_OUTPUT_IDS = { + 1: -11, + 2: -12, + } + + try: + fileno = out.fileno() + except AttributeError: + # If the output stream doesn't have a fileno, it's virtual + return False + if fileno not in WIN_OUTPUT_IDS: + return False + + GetStdHandle = ctypes.WINFUNCTYPE( + ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)( + ("GetStdHandle", ctypes.windll.kernel32)) + h = GetStdHandle(WIN_OUTPUT_IDS[fileno]) + + WriteConsoleW = ctypes.WINFUNCTYPE( + ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR, + ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD), + ctypes.wintypes.LPVOID)(("WriteConsoleW", ctypes.windll.kernel32)) + written = ctypes.wintypes.DWORD(0) + + GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)(("GetFileType", ctypes.windll.kernel32)) + FILE_TYPE_CHAR = 0x0002 + FILE_TYPE_REMOTE = 0x8000 + GetConsoleMode = ctypes.WINFUNCTYPE( + ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, + ctypes.POINTER(ctypes.wintypes.DWORD))( + ("GetConsoleMode", ctypes.windll.kernel32)) + INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value + + 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) + + if not_a_console(h): + return False + + def next_nonbmp_pos(s): + try: + return next(i for i, c in enumerate(s) if ord(c) > 0xffff) + except StopIteration: + return len(s) + + while s: + count = min(next_nonbmp_pos(s), 1024) + + ret = WriteConsoleW( + h, s, count if count else 2, ctypes.byref(written), None) + if ret == 0: + raise OSError('Failed to write string') + if not count: # We just wrote a non-BMP character + assert written.value == 2 + s = s[1:] + else: + assert written.value > 0 + s = s[written.value:] + return True + + +def write_string(s, out=None, encoding=None): if out is None: out = sys.stderr assert type(s) == compat_str + if sys.platform == 'win32' and encoding is None and hasattr(out, 'fileno'): + if _windows_write_string(s, out): + return + 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: + byt = s.encode(encoding or preferredencoding(), 'ignore') + out.write(byt) + elif hasattr(out, 'buffer'): + enc = encoding or getattr(out, 'encoding', None) or preferredencoding() + byt = s.encode(enc, 'ignore') + out.buffer.write(byt) + else: 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() @@ -1092,9 +1226,12 @@ def month_by_name(name): return None -def fix_xml_all_ampersand(xml_str): +def fix_xml_ampersands(xml_str): """Replace all the '&' by '&' in XML""" - return xml_str.replace(u'&', u'&') + return re.sub( + r'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)', + u'&', + xml_str) def setproctitle(title): @@ -1103,11 +1240,11 @@ def setproctitle(title): 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') + title_bytes = title.encode('utf-8') + buf = ctypes.create_string_buffer(len(title_bytes)) + buf.value = title_bytes try: - libc.prctl(15, ctypes.byref(buf), 0, 0, 0) + libc.prctl(15, buf, 0, 0, 0) except AttributeError: return # Strange libc, just skip this @@ -1128,8 +1265,15 @@ class HEADRequest(compat_urllib_request.Request): return "HEAD" -def int_or_none(v): - return v if v is None else int(v) +def int_or_none(v, scale=1, default=None, get_attr=None): + if get_attr: + if v is not None: + v = getattr(v, get_attr, None) + return default if v is None else (int(v) // scale) + + +def float_or_none(v, scale=1, default=None): + return default if v is None else (float(v) / scale) def parse_duration(s): @@ -1137,7 +1281,7 @@ def parse_duration(s): return None m = re.match( - r'(?:(?:(?P[0-9]+):)?(?P[0-9]+):)?(?P[0-9]+)$', s) + r'(?:(?:(?P[0-9]+)[:h])?(?P[0-9]+)[:m])?(?P[0-9]+)s?(?::[0-9]+)?$', s) if not m: return None res = int(m.group('secs')) @@ -1161,3 +1305,152 @@ def check_executable(exe, args=[]): 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): + unicode_escape = codecs.getdecoder('unicode_escape') + return re.sub( + r'\\U[0-9a-fA-F]{8}', + lambda m: unicode_escape(m.group(0))[0], + 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 + + +def read_batch_urls(batch_fd): + def fixup(url): + if not isinstance(url, compat_str): + url = url.decode('utf-8', 'replace') + BOM_UTF8 = u'\xef\xbb\xbf' + if url.startswith(BOM_UTF8): + url = url[len(BOM_UTF8):] + url = url.strip() + if url.startswith(('#', ';', ']')): + return False + return url + + with contextlib.closing(batch_fd) as fd: + return [url for url in map(fixup, fd) if url] + + +def urlencode_postdata(*args, **kargs): + return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii') + + +def parse_xml(s): + class TreeBuilder(xml.etree.ElementTree.TreeBuilder): + def doctype(self, name, pubid, system): + pass # Ignore doctypes + + parser = xml.etree.ElementTree.XMLParser(target=TreeBuilder()) + kwargs = {'parser': parser} if sys.version_info >= (2, 7) else {} + return xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs) + + +if sys.version_info < (3, 0) and sys.platform == 'win32': + def compat_getpass(prompt, *args, **kwargs): + if isinstance(prompt, compat_str): + prompt = prompt.encode(preferredencoding()) + return getpass.getpass(prompt, *args, **kwargs) +else: + compat_getpass = getpass.getpass + + +US_RATINGS = { + 'G': 0, + 'PG': 10, + 'PG-13': 13, + 'R': 16, + 'NC': 18, +} + + +def strip_jsonp(code): + return re.sub(r'(?s)^[a-zA-Z0-9_]+\s*\(\s*(.*)\);?\s*?\s*$', r'\1', code) + + +def qualities(quality_ids): + """ Get a numeric quality value out of a list of possible values """ + def q(qid): + try: + return quality_ids.index(qid) + except ValueError: + return -1 + return q + + +DEFAULT_OUTTMPL = '%(title)s-%(id)s.%(ext)s' + +try: + subprocess_check_output = subprocess.check_output +except AttributeError: + def subprocess_check_output(*args, **kwargs): + assert 'input' not in kwargs + p = subprocess.Popen(*args, stdout=subprocess.PIPE, **kwargs) + output, _ = p.communicate() + ret = p.poll() + if ret: + raise subprocess.CalledProcessError(ret, p.args, output=output) + return output