[openload] separate PhantomJS code from extractor

[youtube-dl] / youtube_dl / extractor / openload.py

diff --git a/youtube_dl/extractor/openload.py b/youtube_dl/extractor/openload.py
index 4468f31fcae074090346d134180fee98752b7822..ac5e0bb08afd84602720d9525a27e63aac84e4c4 100644 (file)
--- a/youtube_dl/extractor/openload.py
+++ b/youtube_dl/extractor/openload.py
@@ -4,15 +4,16 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..compat import compat_chr
 from ..utils import (
-    encode_base_n,
+    determine_ext,
     ExtractorError,
+    get_element_by_id,
+    PhantomJSwrapper,
 )
 
 
 class OpenloadIE(InfoExtractor):
-    _VALID_URL = r'https://openload.(?:co|io)/(?:f|embed)/(?P<id>[a-zA-Z0-9-]+)'
+    _VALID_URL = r'https?://(?:openload\.(?:co|io)|oload\.tv)/(?:f|embed)/(?P<id>[a-zA-Z0-9-_]+)'
 
     _TESTS = [{
         'url': 'https://openload.co/f/kUEfGclsU9o',
@@ -21,7 +22,23 @@ 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',
@@ -29,79 +46,61 @@ class OpenloadIE(InfoExtractor):
     }, {
         'url': 'https://openload.io/f/ZAn6oz-VZGE/',
         'only_matching': True,
+    }, {
+        'url': 'https://openload.co/f/_-ztPaZtMhM/',
+        'only_matching': True,
+    }, {
+        # unavailable via https://openload.co/f/Sxz5sADo82g/, different layout
+        # for title and ext
+        'url': 'https://openload.co/embed/Sxz5sADo82g/',
+        'only_matching': True,
+    }, {
+        'url': 'https://oload.tv/embed/KnG-kKZdcfY/',
+        'only_matching': True,
     }]
 
-    @staticmethod
-    def openload_level2_debase(m):
-        radix, num = int(m.group(1)) + 27, int(m.group(2))
-        return '"' + encode_base_n(num, radix) + '"'
+    _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'
 
-    @classmethod
-    def openload_level2(cls, txt):
-        # The function name is ǃ \u01c3
-        # Using escaped unicode literals does not work in Python 3.2
-        return re.sub(r'ǃ\((\d+),(\d+)\)', cls.openload_level2_debase, txt, re.UNICODE).replace('"+"', '')
+    @staticmethod
+    def _extract_urls(webpage):
+        return re.findall(
+            r'<iframe[^>]+src=["\']((?:https?://)?(?:openload\.(?:co|io)|oload\.tv)/embed/[a-zA-Z0-9-_]+)',
+            webpage)
 
-    # Openload uses a variant of aadecode
-    # openload_decode and related functions are originally written by
-    # vitas@matfyz.cz and released with public domain
-    # See https://github.com/rg3/youtube-dl/issues/8489
-    @classmethod
-    def openload_decode(cls, txt):
-        symbol_table = [
-            ('_', '(ﾟДﾟ) [ﾟΘﾟ]'),
-            ('a', '(ﾟДﾟ) [ﾟωﾟﾉ]'),
-            ('b', '(ﾟДﾟ) [ﾟΘﾟﾉ]'),
-            ('c', '(ﾟДﾟ) [\'c\']'),
-            ('d', '(ﾟДﾟ) [ﾟｰﾟﾉ]'),
-            ('e', '(ﾟДﾟ) [ﾟДﾟﾉ]'),
-            ('f', '(ﾟДﾟ) [1]'),
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        url = 'https://openload.co/embed/%s/' % video_id
+        headers = {
+            'User-Agent': self._USER_AGENT,
+        }
 
-            ('o', '(ﾟДﾟ) [\'o\']'),
-            ('u', '(oﾟｰﾟo)'),
-            ('c', '(ﾟДﾟ) [\'c\']'),
+        phantom = PhantomJSwrapper(self)
+        webpage, _ = phantom.get(url, video_id=video_id, headers=headers)
 
-            ('7', '((ﾟｰﾟ) + (o^_^o))'),
-            ('6', '((o^_^o) +(o^_^o) +(c^_^o))'),
-            ('5', '((ﾟｰﾟ) + (ﾟΘﾟ))'),
-            ('4', '(-~3)'),
-            ('3', '(-~-~1)'),
-            ('2', '(-~1)'),
-            ('1', '(-~0)'),
-            ('0', '((c^_^o)-(c^_^o))'),
-        ]
-        delim = '(ﾟДﾟ)[ﾟεﾟ]+'
-        ret = ''
-        for aachar in txt.split(delim):
-            for val, pat in symbol_table:
-                aachar = aachar.replace(pat, val)
-            aachar = aachar.replace('+ ', '')
-            m = re.match(r'^\d+', aachar)
-            if m:
-                ret += compat_chr(int(m.group(0), 8))
-            else:
-                m = re.match(r'^u([\da-f]+)', aachar)
-                if m:
-                    ret += compat_chr(int(m.group(1), 16))
-        return cls.openload_level2(ret)
+        if 'File not found' in webpage or 'deleted by the owner' in webpage:
+            raise ExtractorError('File not found', expected=True, video_id=video_id)
 
-    def _real_extract(self, url):
-        video_id = self._match_id(url)
-        webpage = self._download_webpage(url, video_id)
+        decoded_id = get_element_by_id('streamurl', webpage)
 
-        if 'File not found' in webpage:
-            raise ExtractorError('File not found', expected=True)
+        video_url = 'https://openload.co/stream/%s?mime=true' % decoded_id
 
-        code = self._search_regex(
-            r'<video[^>]+>\s*<script[^>]+>([^<]+)</script>',
-            webpage, 'JS code')
+        title = self._og_search_title(webpage, default=None) or self._search_regex(
+            r'<span[^>]+class=["\']title["\'][^>]*>([^<]+)', webpage,
+            'title', default=None) or self._html_search_meta(
+            'description', webpage, 'title', fatal=True)
 
-        video_url = self._search_regex(
-            r'return\s+"(https?://[^"]+)"', self.openload_decode(code), 'video URL')
+        entries = self._parse_html5_media_entries(url, webpage, video_id)
+        entry = entries[0] if entries else {}
+        subtitles = entry.get('subtitles')
 
-        return {
+        info_dict = {
             'id': video_id,
-            'title': self._og_search_title(webpage),
-            'thumbnail': self._og_search_thumbnail(webpage),
+            'title': title,
+            '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, 'mp4'),
+            'subtitles': subtitles,
+            'http_headers': headers,
         }
+        return info_dict