[videa] improve extraction
authorRemita Amine <remitamine@gmail.com>
Thu, 26 Nov 2020 11:56:49 +0000 (12:56 +0100)
committerRemita Amine <remitamine@gmail.com>
Thu, 26 Nov 2020 11:56:49 +0000 (12:56 +0100)
youtube_dl/extractor/videa.py

index a03614cc10918301dbd3196de2123b0f22b2a8b8..ab2c15cdec154bf7d760c588d114eed106221b60 100644 (file)
@@ -1,10 +1,9 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import re
 import random
+import re
 import string
-import struct
 
 from .common import InfoExtractor
 from ..utils import (
@@ -12,13 +11,14 @@ from ..utils import (
     int_or_none,
     mimetype2ext,
     parse_codecs,
+    update_url_query,
     xpath_element,
     xpath_text,
 )
 from ..compat import (
     compat_b64decode,
     compat_ord,
-    compat_parse_qs,
+    compat_struct_pack,
 )
 
 
@@ -28,7 +28,7 @@ class VideaIE(InfoExtractor):
                         videa(?:kid)?\.hu/
                         (?:
                             videok/(?:[^/]+/)*[^?#&]+-|
-                            player\?.*?\bv=|
+                            (?:videojs_)?player\?.*?\bv=|
                             player/v/
                         )
                         (?P<id>[^?#&]+)
@@ -62,6 +62,7 @@ class VideaIE(InfoExtractor):
         'url': 'https://videakid.hu/player/v/8YfIAjxwWGwT8HVQ?autoplay=1',
         'only_matching': True,
     }]
+    _STATIC_SECRET = 'xHb0ZvME5q8CBcoQi6AngerDu3FGO9fkUlwPmLVY_RTzj2hJIS4NasXWKy1td7p'
 
     @staticmethod
     def _extract_urls(webpage):
@@ -69,75 +70,84 @@ class VideaIE(InfoExtractor):
             r'<iframe[^>]+src=(["\'])(?P<url>(?:https?:)?//videa\.hu/player\?.*?\bv=.+?)\1',
             webpage)]
 
-    def rc4(self, ciphertext, key):
+    @staticmethod
+    def rc4(cipher_text, key):
         res = b''
 
-        keyLen = len(key)
+        key_len = len(key)
         S = list(range(256))
 
         j = 0
         for i in range(256):
-            j = (j + S[i] + ord(key[i % keyLen])) % 256
+            j = (j + S[i] + ord(key[i % key_len])) % 256
             S[i], S[j] = S[j], S[i]
 
         i = 0
         j = 0
-        for m in range(len(ciphertext)):
+        for m in range(len(cipher_text)):
             i = (i + 1) % 256
             j = (j + S[i]) % 256
             S[i], S[j] = S[j], S[i]
             k = S[(S[i] + S[j]) % 256]
-            res += struct.pack("B", k ^ compat_ord(ciphertext[m]))
+            res += compat_struct_pack('B', k ^ compat_ord(cipher_text[m]))
 
-        return res
+        return res.decode()
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
-        webpage = self._download_webpage(url, video_id, fatal=True)
-        error = self._search_regex(r'<p class="error-text">([^<]+)</p>', webpage, 'error', default=None)
-        if error:
-            raise ExtractorError(error, expected=True)
-
-        video_src_params_raw = self._search_regex(r'<iframe[^>]+id="videa_player_iframe"[^>]+src="/player\?([^"]+)"', webpage, 'video_src_params')
-        video_src_params = compat_parse_qs(video_src_params_raw)
-        player_page = self._download_webpage("https://videa.hu/videojs_player?%s" % video_src_params_raw, video_id, fatal=True)
-        nonce = self._search_regex(r'_xt\s*=\s*"([^"]+)"', player_page, 'nonce')
-        random_seed = ''.join(random.choice(string.ascii_uppercase + string.ascii_lowercase + string.digits) for _ in range(8))
-        static_secret = 'xHb0ZvME5q8CBcoQi6AngerDu3FGO9fkUlwPmLVY_RTzj2hJIS4NasXWKy1td7p'
+        query = {'v': video_id}
+        player_page = self._download_webpage(
+            'https://videa.hu/player', video_id, query=query)
+
+        nonce = self._search_regex(
+            r'_xt\s*=\s*"([^"]+)"', player_page, 'nonce')
         l = nonce[:32]
         s = nonce[32:]
         result = ''
         for i in range(0, 32):
-            result += s[i - (static_secret.index(l[i]) - 31)]
-
-        video_src_params['_s'] = random_seed
-        video_src_params['_t'] = result[:16]
-        encryption_key_stem = result[16:] + random_seed
-
-        [b64_info, handle] = self._download_webpage_handle(
-            'http://videa.hu/videaplayer_get_xml.php', video_id,
-            query=video_src_params, fatal=True)
-
-        encrypted_info = compat_b64decode(b64_info)
-        key = encryption_key_stem + handle.info()['x-videa-xs']
-        info_str = self.rc4(encrypted_info, key).decode('utf8')
-        info = self._parse_xml(info_str, video_id)
-
-        video = xpath_element(info, './/video', 'video', fatal=True)
-        sources = xpath_element(info, './/video_sources', 'sources', fatal=True)
-        hash_values = xpath_element(info, './/hash_values', 'hash_values', fatal=True)
+            result += s[i - (self._STATIC_SECRET.index(l[i]) - 31)]
+
+        random_seed = ''.join(random.choice(string.ascii_letters + string.digits) for _ in range(8))
+        query['_s'] = random_seed
+        query['_t'] = result[:16]
+
+        b64_info, handle = self._download_webpage_handle(
+            'http://videa.hu/videaplayer_get_xml.php', video_id, query=query)
+        if b64_info.startswith('<?xml'):
+            info = self._parse_xml(b64_info, video_id)
+        else:
+            key = result[16:] + random_seed + handle.headers['x-videa-xs']
+            info = self._parse_xml(self.rc4(
+                compat_b64decode(b64_info), key), video_id)
+
+        video = xpath_element(info, './video', 'video')
+        if not video:
+            raise ExtractorError(xpath_element(
+                info, './error', fatal=True), expected=True)
+        sources = xpath_element(
+            info, './video_sources', 'sources', fatal=True)
+        hash_values = xpath_element(
+            info, './hash_values', 'hash values', fatal=True)
 
         title = xpath_text(video, './title', fatal=True)
 
         formats = []
         for source in sources.findall('./video_source'):
             source_url = source.text
-            if not source_url:
+            source_name = source.get('name')
+            source_exp = source.get('exp')
+            if not (source_url and source_name and source_exp):
                 continue
-            source_url += '?md5=%s&expires=%s' % (hash_values.find('hash_value_%s' % source.get('name')).text, source.get('exp'))
+            hash_value = xpath_text(hash_values, 'hash_value_' + source_name)
+            if not hash_value:
+                continue
+            source_url = update_url_query(source_url, {
+                'md5': hash_value,
+                'expires': source_exp,
+            })
             f = parse_codecs(source.get('codecs'))
             f.update({
-                'url': source_url,
+                'url': self._proto_relative_url(source_url),
                 'ext': mimetype2ext(source.get('mimetype')) or 'mp4',
                 'format_id': source.get('name'),
                 'width': int_or_none(source.get('width')),
@@ -146,8 +156,7 @@ class VideaIE(InfoExtractor):
             formats.append(f)
         self._sort_formats(formats)
 
-        thumbnail = xpath_text(video, './poster_src')
-        duration = int_or_none(xpath_text(video, './duration'))
+        thumbnail = self._proto_relative_url(xpath_text(video, './poster_src'))
 
         age_limit = None
         is_adult = xpath_text(video, './is_adult_content', default=None)
@@ -158,7 +167,7 @@ class VideaIE(InfoExtractor):
             'id': video_id,
             'title': title,
             'thumbnail': thumbnail,
-            'duration': duration,
+            'duration': int_or_none(xpath_text(video, './duration')),
             'age_limit': age_limit,
             'formats': formats,
         }