X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/99b46abf79266c2bf5eabc291dc4f366e61b5dde..81657e6063f6dd0b3ee6243d446b8df1b13adfbc:/youtube_dl/extractor/openload.py diff --git a/youtube_dl/extractor/openload.py b/youtube_dl/extractor/openload.py index 4e80ca9..eaaaf8a 100644 --- a/youtube_dl/extractor/openload.py +++ b/youtube_dl/extractor/openload.py @@ -1,19 +1,249 @@ # coding: utf-8 -from __future__ import unicode_literals, division +from __future__ import unicode_literals -import math +import json +import os +import re +import subprocess +import tempfile from .common import InfoExtractor -from ..compat import compat_chr +from ..compat import ( + compat_urlparse, + compat_kwargs, +) from ..utils import ( - decode_png, + check_executable, determine_ext, + encodeArgument, ExtractorError, + get_element_by_id, + get_exe_version, + is_outdated_version, + std_headers, ) +def cookie_to_dict(cookie): + cookie_dict = { + 'name': cookie.name, + 'value': cookie.value, + } + if cookie.port_specified: + cookie_dict['port'] = cookie.port + if cookie.domain_specified: + cookie_dict['domain'] = cookie.domain + if cookie.path_specified: + cookie_dict['path'] = cookie.path + if cookie.expires is not None: + cookie_dict['expires'] = cookie.expires + if cookie.secure is not None: + cookie_dict['secure'] = cookie.secure + if cookie.discard is not None: + cookie_dict['discard'] = cookie.discard + try: + if (cookie.has_nonstandard_attr('httpOnly') or + cookie.has_nonstandard_attr('httponly') or + cookie.has_nonstandard_attr('HttpOnly')): + cookie_dict['httponly'] = True + except TypeError: + pass + return cookie_dict + + +def cookie_jar_to_list(cookie_jar): + return [cookie_to_dict(cookie) for cookie in cookie_jar] + + +class PhantomJSwrapper(object): + """PhantomJS wrapper class + + This class is experimental. + """ + + _TEMPLATE = r''' + phantom.onError = function(msg, trace) {{ + var msgStack = ['PHANTOM ERROR: ' + msg]; + if(trace && trace.length) {{ + msgStack.push('TRACE:'); + trace.forEach(function(t) {{ + msgStack.push(' -> ' + (t.file || t.sourceURL) + ': ' + t.line + + (t.function ? ' (in function ' + t.function +')' : '')); + }}); + }} + console.error(msgStack.join('\n')); + phantom.exit(1); + }}; + var page = require('webpage').create(); + var fs = require('fs'); + var read = {{ mode: 'r', charset: 'utf-8' }}; + var write = {{ mode: 'w', charset: 'utf-8' }}; + JSON.parse(fs.read("{cookies}", read)).forEach(function(x) {{ + phantom.addCookie(x); + }}); + page.settings.resourceTimeout = {timeout}; + page.settings.userAgent = "{ua}"; + page.onLoadStarted = function() {{ + page.evaluate(function() {{ + delete window._phantom; + delete window.callPhantom; + }}); + }}; + var saveAndExit = function() {{ + fs.write("{html}", page.content, write); + fs.write("{cookies}", JSON.stringify(phantom.cookies), write); + phantom.exit(); + }}; + page.onLoadFinished = function(status) {{ + if(page.url === "") {{ + page.setContent(fs.read("{html}", read), "{url}"); + }} + else {{ + {jscode} + }} + }}; + page.open(""); + ''' + + _TMP_FILE_NAMES = ['script', 'html', 'cookies'] + + @staticmethod + def _version(): + return get_exe_version('phantomjs', version_re=r'([0-9.]+)') + + def __init__(self, extractor, required_version=None, timeout=10000): + self._TMP_FILES = {} + + self.exe = check_executable('phantomjs', ['-v']) + if not self.exe: + raise ExtractorError('PhantomJS executable not found in PATH, ' + 'download it from http://phantomjs.org', + expected=True) + + self.extractor = extractor + + if required_version: + version = self._version() + if is_outdated_version(version, required_version): + self.extractor._downloader.report_warning( + 'Your copy of PhantomJS is outdated, update it to version ' + '%s or newer if you encounter any errors.' % required_version) + + self.options = { + 'timeout': timeout, + } + for name in self._TMP_FILE_NAMES: + tmp = tempfile.NamedTemporaryFile(delete=False) + tmp.close() + self._TMP_FILES[name] = tmp + + def __del__(self): + for name in self._TMP_FILE_NAMES: + try: + os.remove(self._TMP_FILES[name].name) + except (IOError, OSError, KeyError): + pass + + def _save_cookies(self, url): + cookies = cookie_jar_to_list(self.extractor._downloader.cookiejar) + for cookie in cookies: + if 'path' not in cookie: + cookie['path'] = '/' + if 'domain' not in cookie: + cookie['domain'] = compat_urlparse.urlparse(url).netloc + with open(self._TMP_FILES['cookies'].name, 'wb') as f: + f.write(json.dumps(cookies).encode('utf-8')) + + def _load_cookies(self): + with open(self._TMP_FILES['cookies'].name, 'rb') as f: + cookies = json.loads(f.read().decode('utf-8')) + for cookie in cookies: + if cookie['httponly'] is True: + cookie['rest'] = {'httpOnly': None} + if 'expiry' in cookie: + cookie['expire_time'] = cookie['expiry'] + self.extractor._set_cookie(**compat_kwargs(cookie)) + + def get(self, url, html=None, video_id=None, note=None, note2='Executing JS on webpage', headers={}, jscode='saveAndExit();'): + """ + Downloads webpage (if needed) and executes JS + + Params: + url: website url + html: optional, html code of website + video_id: video id + note: optional, displayed when downloading webpage + note2: optional, displayed when executing JS + headers: custom http headers + jscode: code to be executed when page is loaded + + Returns tuple with: + * downloaded website (after JS execution) + * anything you print with `console.log` (but not inside `page.execute`!) + + In most cases you don't need to add any `jscode`. + It is executed in `page.onLoadFinished`. + `saveAndExit();` is mandatory, use it instead of `phantom.exit()` + It is possible to wait for some element on the webpage, for example: + var check = function() { + var elementFound = page.evaluate(function() { + return document.querySelector('#b.done') !== null; + }); + if(elementFound) + saveAndExit(); + else + window.setTimeout(check, 500); + } + + page.evaluate(function(){ + document.querySelector('#a').click(); + }); + check(); + """ + if 'saveAndExit();' not in jscode: + raise ExtractorError('`saveAndExit();` not found in `jscode`') + if not html: + html = self.extractor._download_webpage(url, video_id, note=note, headers=headers) + with open(self._TMP_FILES['html'].name, 'wb') as f: + f.write(html.encode('utf-8')) + + self._save_cookies(url) + + replaces = self.options + replaces['url'] = url + user_agent = headers.get('User-Agent') or std_headers['User-Agent'] + replaces['ua'] = user_agent.replace('"', '\\"') + replaces['jscode'] = jscode + + for x in self._TMP_FILE_NAMES: + replaces[x] = self._TMP_FILES[x].name.replace('\\', '\\\\').replace('"', '\\"') + + with open(self._TMP_FILES['script'].name, 'wb') as f: + f.write(self._TEMPLATE.format(**replaces).encode('utf-8')) + + if video_id is None: + self.extractor.to_screen('%s' % (note2,)) + else: + self.extractor.to_screen('%s: %s' % (video_id, note2)) + + p = subprocess.Popen([ + self.exe, '--ssl-protocol=any', + self._TMP_FILES['script'].name + ], stdout=subprocess.PIPE, stderr=subprocess.PIPE) + out, err = p.communicate() + if p.returncode != 0: + raise ExtractorError( + 'Executing JS failed\n:' + encodeArgument(err)) + with open(self._TMP_FILES['html'].name, 'rb') as f: + html = f.read().decode('utf-8') + + self._load_cookies() + + return (html, encodeArgument(out)) + + class OpenloadIE(InfoExtractor): - _VALID_URL = r'https://openload.(?:co|io)/(?:f|embed)/(?P[a-zA-Z0-9-_]+)' + _VALID_URL = r'https?://(?:www\.)?(?:openload\.(?:co|io|link)|oload\.(?:tv|stream))/(?:f|embed)/(?P[a-zA-Z0-9-_]+)' _TESTS = [{ 'url': 'https://openload.co/f/kUEfGclsU9o', @@ -22,7 +252,23 @@ class OpenloadIE(InfoExtractor): 'id': 'kUEfGclsU9o', 'ext': 'mp4', 'title': 'skyrim_no-audio_1080.mp4', - 'thumbnail': 're:^https?://.*\.jpg$', + 'thumbnail': r're:^https?://.*\.jpg$', + }, + }, { + 'url': 'https://openload.co/embed/rjC09fkPLYs', + 'info_dict': { + 'id': 'rjC09fkPLYs', + 'ext': 'mp4', + 'title': 'movie.mp4', + 'thumbnail': r're:^https?://.*\.jpg$', + 'subtitles': { + 'en': [{ + 'ext': 'vtt', + }], + }, + }, + 'params': { + 'skip_download': True, # test subtitles only }, }, { 'url': 'https://openload.co/embed/kUEfGclsU9o/skyrim_no-audio_1080.mp4', @@ -38,86 +284,80 @@ class OpenloadIE(InfoExtractor): # for title and ext 'url': 'https://openload.co/embed/Sxz5sADo82g/', 'only_matching': True, + }, { + # unavailable via https://openload.co/embed/e-Ixz9ZR5L0/ but available + # via https://openload.co/f/e-Ixz9ZR5L0/ + 'url': 'https://openload.co/f/e-Ixz9ZR5L0/', + 'only_matching': True, + }, { + 'url': 'https://oload.tv/embed/KnG-kKZdcfY/', + 'only_matching': True, + }, { + 'url': 'http://www.openload.link/f/KnG-kKZdcfY', + 'only_matching': True, + }, { + 'url': 'https://oload.stream/f/KnG-kKZdcfY', + 'only_matching': True, }] + _USER_AGENT = 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/56.0.2924.87 Safari/537.36' + + @staticmethod + def _extract_urls(webpage): + return re.findall( + r']+src=["\']((?:https?://)?(?:openload\.(?:co|io)|oload\.tv)/embed/[a-zA-Z0-9-_]+)', + webpage) + def _real_extract(self, url): video_id = self._match_id(url) - webpage = self._download_webpage(url, video_id) - - if 'File not found' in webpage: - raise ExtractorError('File not found', expected=True) - - # The following extraction logic is proposed by @Belderak and @gdkchan - # and declared to be used freely in youtube-dl - # See https://github.com/rg3/youtube-dl/issues/9706 - - numbers_js = self._download_webpage( - 'https://openload.co/assets/js/obfuscator/n.js', video_id, - note='Downloading signature numbers') - signums = self._search_regex( - r'window\.signatureNumbers\s*=\s*[\'"](?P[a-z]+)[\'"]', - numbers_js, 'signature numbers', group='data') - - linkimg_uri = self._search_regex( - r']+id="linkimg"[^>]+src="([^"]+)"', webpage, 'link image') - linkimg = self._request_webpage( - linkimg_uri, video_id, note=False).read() - - width, height, pixels = decode_png(linkimg) - - output = '' - for y in range(height): - for x in range(width): - r, g, b = pixels[y][3 * x:3 * x + 3] - if r == 0 and g == 0 and b == 0: - break - else: - output += compat_chr(r) - output += compat_chr(g) - output += compat_chr(b) - - img_str_length = len(output) // 200 - img_str = [[0 for x in range(img_str_length)] for y in range(10)] - - sig_str_length = len(signums) // 260 - sig_str = [[0 for x in range(sig_str_length)] for y in range(10)] - - for i in range(10): - for j in range(img_str_length): - begin = i * img_str_length * 20 + j * 20 - img_str[i][j] = output[begin:begin + 20] - for j in range(sig_str_length): - begin = i * sig_str_length * 26 + j * 26 - sig_str[i][j] = signums[begin:begin + 26] - - parts = [] - # TODO: find better names for str_, chr_ and sum_ - str_ = '' - for i in [2, 3, 5, 7]: - str_ = '' - sum_ = float(99) - for j in range(len(sig_str[i])): - for chr_idx in range(len(img_str[i][j])): - if sum_ > float(122): - sum_ = float(98) - chr_ = compat_chr(int(math.floor(sum_))) - if sig_str[i][j][chr_idx] == chr_ and j >= len(str_): - sum_ += float(2.5) - str_ += img_str[i][j][chr_idx] - parts.append(str_.replace(',', '')) - - video_url = 'https://openload.co/stream/%s~%s~%s~%s' % (parts[3], parts[1], parts[2], parts[0]) + url_pattern = 'https://openload.co/%%s/%s/' % video_id + headers = { + 'User-Agent': self._USER_AGENT, + } + + for path in ('embed', 'f'): + page_url = url_pattern % path + last = path == 'f' + webpage = self._download_webpage( + page_url, video_id, 'Downloading %s webpage' % path, + headers=headers, fatal=last) + if not webpage: + continue + if 'File not found' in webpage or 'deleted by the owner' in webpage: + if not last: + continue + raise ExtractorError('File not found', expected=True, video_id=video_id) + break + + phantom = PhantomJSwrapper(self, required_version='2.0') + webpage, _ = phantom.get(page_url, html=webpage, video_id=video_id, headers=headers) + + decoded_id = (get_element_by_id('streamurl', webpage) or + get_element_by_id('streamuri', webpage) or + get_element_by_id('streamurj', webpage)) + + if not decoded_id: + raise ExtractorError('Can\'t find stream URL', video_id=video_id) + + video_url = 'https://openload.co/stream/%s?mime=true' % decoded_id title = self._og_search_title(webpage, default=None) or self._search_regex( r']+class=["\']title["\'][^>]*>([^<]+)', webpage, 'title', default=None) or self._html_search_meta( 'description', webpage, 'title', fatal=True) - return { + entries = self._parse_html5_media_entries(page_url, webpage, video_id) + entry = entries[0] if entries else {} + subtitles = entry.get('subtitles') + + info_dict = { 'id': video_id, 'title': title, - 'thumbnail': self._og_search_thumbnail(webpage, default=None), + 'thumbnail': entry.get('thumbnail') or self._og_search_thumbnail(webpage, default=None), 'url': video_url, # Seems all videos have extensions in their titles - 'ext': determine_ext(title), + 'ext': determine_ext(title, 'mp4'), + 'subtitles': subtitles, + 'http_headers': headers, } + return info_dict