X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Futils.py;h=9e4492d402c225d53071d1424005ff5be0577681;hb=a88d461dff67205fcec684426afbcbeb4b0e7cf5;hp=4293a77f569879964ce9019c44f8f85bfcaad982;hpb=60f5c9fb191b1ffc11e69dece379a8ad2de22207;p=youtube-dl diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index 4293a77f5..9e4492d40 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -22,7 +22,6 @@ import locale import math import operator import os -import pipes import platform import random import re @@ -36,6 +35,7 @@ import xml.etree.ElementTree import zlib from .compat import ( + compat_HTMLParseError, compat_HTMLParser, compat_basestring, compat_chr, @@ -365,9 +365,9 @@ def get_elements_by_attribute(attribute, value, html, escape_value=True): retlist = [] for m in re.finditer(r'''(?xs) <([a-zA-Z0-9:._-]+) - (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'))*? + (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*? \s+%s=['"]?%s['"]? - (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'))*? + (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*? \s*> (?P.*?) @@ -409,8 +409,12 @@ def extract_attributes(html_element): 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() + try: + parser.feed(html_element) + parser.close() + # Older Python may throw HTMLParseError in case of malformed HTML + except compat_HTMLParseError: + pass return parser.attrs @@ -592,7 +596,7 @@ def unescapeHTML(s): assert type(s) == compat_str return re.sub( - r'&([^;]+;)', lambda m: _htmlentity_transform(m.group(1)), s) + r'&([^&;]+;)', lambda m: _htmlentity_transform(m.group(1)), s) def get_subprocess_encoding(): @@ -932,14 +936,6 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler): except zlib.error: return zlib.decompress(data) - @staticmethod - def addinfourl_wrapper(stream, headers, url, code): - if hasattr(compat_urllib_request.addinfourl, 'getcode'): - return compat_urllib_request.addinfourl(stream, headers, url, code) - ret = compat_urllib_request.addinfourl(stream, headers, url) - ret.code = code - return ret - def http_request(self, req): # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not # always respected by websites, some tend to give out URLs with non percent-encoded @@ -991,13 +987,13 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler): break else: raise original_ioerror - resp = self.addinfourl_wrapper(uncompressed, old_resp.headers, old_resp.url, old_resp.code) + resp = compat_urllib_request.addinfourl(uncompressed, old_resp.headers, old_resp.url, old_resp.code) resp.msg = old_resp.msg del resp.headers['Content-encoding'] # deflate if resp.headers.get('Content-encoding', '') == 'deflate': gz = io.BytesIO(self.deflate(resp.read())) - resp = self.addinfourl_wrapper(gz, old_resp.headers, old_resp.url, old_resp.code) + resp = compat_urllib_request.addinfourl(gz, old_resp.headers, old_resp.url, old_resp.code) resp.msg = old_resp.msg del resp.headers['Content-encoding'] # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see @@ -1187,7 +1183,7 @@ def unified_timestamp(date_str, day_first=True): if date_str is None: return None - date_str = date_str.replace(',', ' ') + date_str = re.sub(r'[,|]', '', date_str) pm_delta = 12 if re.search(r'(?i)PM', date_str) else 0 timezone, date_str = extract_timezone(date_str) @@ -1538,7 +1534,7 @@ def shell_quote(args): if isinstance(a, bytes): # We may get a filename encoded with 'encodeFilename' a = a.decode(encoding) - quoted_args.append(pipes.quote(a)) + quoted_args.append(compat_shlex_quote(a)) return ' '.join(quoted_args) @@ -1819,6 +1815,10 @@ def float_or_none(v, scale=1, invscale=1, default=None): return default +def bool_or_none(v, default=None): + return v if isinstance(v, bool) else default + + def strip_or_none(v): return None if v is None else v.strip() @@ -2211,7 +2211,12 @@ def parse_age_limit(s): def strip_jsonp(code): return re.sub( - r'(?s)^[a-zA-Z0-9_.$]+\s*\(\s*(.*)\);?\s*?(?://[^\n]*)*$', r'\1', code) + r'''(?sx)^ + (?:window\.)?(?P[a-zA-Z0-9_.$]+) + (?:\s*&&\s*(?P=func_name))? + \s*\(\s*(?P.*)\);? + \s*?(?://[^\n]*)*$''', + r'\g', code) def js_to_json(code): @@ -2732,6 +2737,8 @@ def cli_option(params, command_option, param): def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None): param = params.get(param) + if param is None: + return [] assert isinstance(param, bool) if separator: return [command_option + separator + (true_value if param else false_value)] @@ -3815,6 +3822,219 @@ def write_xattr(path, key, value): "or the 'xattr' binary.") +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 not cookie.expires is None: + cookie_dict['expires'] = cookie.expires + if not cookie.secure is None: + cookie_dict['secure'] = cookie.secure + if not cookie.discard is 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""" + + _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.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, + } + self._TMP_FILES = {} + 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: + 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(**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)) + + def random_birthday(year_field, month_field, day_field): return { year_field: str(random.randint(1950, 1995)),