[cda] Implement birthday verification (closes #12789)

ChangeLog

@@ -1,6 +1,7 @@
 version <unreleased>
 
 Extractors
++ [cda] Support birthday verification (#12789)
 * [leeco] Fix extraction (#12974)
 
 

test/test_utils.py

@@ -44,6 +44,7 @@
     limit_length,
     mimetype2ext,
     month_by_name,
+    multipart_encode,
     ohdave_rsa_encrypt,
     OnDemandPagedList,
     orderedSet,
@@ -620,6 +621,16 @@ def query_dict(url):
             'http://example.com/path', {'test': '第二行тест'})),
             query_dict('http://example.com/path?test=%E7%AC%AC%E4%BA%8C%E8%A1%8C%D1%82%D0%B5%D1%81%D1%82'))
 
+    def test_multipart_encode(self):
+        self.assertEqual(
+            multipart_encode({b'field': b'value'}, boundary='AAAAAA')[0],
+            b'--AAAAAA\r\nContent-Disposition: form-data; name="field"\r\n\r\nvalue\r\n--AAAAAA--\r\n')
+        self.assertEqual(
+            multipart_encode({'欄位'.encode('utf-8'): '值'.encode('utf-8')}, boundary='AAAAAA')[0],
+            b'--AAAAAA\r\nContent-Disposition: form-data; name="\xe6\xac\x84\xe4\xbd\x8d"\r\n\r\n\xe5\x80\xbc\r\n--AAAAAA--\r\n')
+        self.assertRaises(
+            ValueError, multipart_encode, {b'field': b'value'}, boundary='value')
+
     def test_dict_get(self):
         FALSE_VALUES = {
             'none': None,

youtube_dl/extractor/cda.py

@@ -9,7 +9,10 @@
     ExtractorError,
     float_or_none,
     int_or_none,
+    multipart_encode,
     parse_duration,
+    random_birthday,
+    urljoin,
 )
 
 
@@ -27,7 +30,8 @@ class CDAIE(InfoExtractor):
             'description': 'md5:269ccd135d550da90d1662651fcb9772',
             'thumbnail': r're:^https?://.*\.jpg$',
             'average_rating': float,
-            'duration': 39
+            'duration': 39,
+            'age_limit': 0,
         }
     }, {
         'url': 'http://www.cda.pl/video/57413289',
@@ -41,13 +45,41 @@ class CDAIE(InfoExtractor):
             'uploader': 'crash404',
             'view_count': int,
             'average_rating': float,
-            'duration': 137
+            'duration': 137,
+            'age_limit': 0,
         }
+    }, {
+        # Age-restricted
+        'url': 'http://www.cda.pl/video/1273454c4',
+        'info_dict': {
+            'id': '1273454c4',
+            'ext': 'mp4',
+            'title': 'Bronson (2008) napisy HD 1080p',
+            'description': 'md5:1b6cb18508daf2dc4e0fa4db77fec24c',
+            'height': 1080,
+            'uploader': 'boniek61',
+            'thumbnail': r're:^https?://.*\.jpg$',
+            'duration': 5554,
+            'age_limit': 18,
+            'view_count': int,
+            'average_rating': float,
+        },
     }, {
         'url': 'http://ebd.cda.pl/0x0/5749950c',
         'only_matching': True,
     }]
 
+    def _download_age_confirm_page(self, url, video_id, *args, **kwargs):
+        form_data = random_birthday('rok', 'miesiac', 'dzien')
+        form_data.update({'return': url, 'module': 'video', 'module_id': video_id})
+        data, content_type = multipart_encode(form_data)
+        return self._download_webpage(
+            urljoin(url, '/a/validatebirth'), video_id, *args,
+            data=data, headers={
+                'Referer': url,
+                'Content-Type': content_type,
+            }, **kwargs)
+
     def _real_extract(self, url):
         video_id = self._match_id(url)
         self._set_cookie('cda.pl', 'cda.player', 'html5')
@@ -57,6 +89,13 @@ def _real_extract(self, url):
         if 'Ten film jest dostępny dla użytkowników premium' in webpage:
             raise ExtractorError('This video is only available for premium users.', expected=True)
 
+        need_confirm_age = False
+        if self._html_search_regex(r'(<form[^>]+action="/a/validatebirth")',
+                                   webpage, 'birthday validate form', default=None):
+            webpage = self._download_age_confirm_page(
+                url, video_id, note='Confirming age')
+            need_confirm_age = True
+
         formats = []
 
         uploader = self._search_regex(r'''(?x)
@@ -81,6 +120,7 @@ def _real_extract(self, url):
             'thumbnail': self._og_search_thumbnail(webpage),
             'formats': formats,
             'duration': None,
+            'age_limit': 18 if need_confirm_age else 0,
         }
 
         def extract_format(page, version):
@@ -121,7 +161,12 @@ def extract_format(page, version):
         for href, resolution in re.findall(
                 r'<a[^>]+data-quality="[^"]+"[^>]+href="([^"]+)"[^>]+class="quality-btn"[^>]*>([0-9]+p)',
                 webpage):
-            webpage = self._download_webpage(
+            if need_confirm_age:
+                handler = self._download_age_confirm_page
+            else:
+                handler = self._download_webpage
+
+            webpage = handler(
                 self._BASE_URL + href, video_id,
                 'Downloading %s version information' % resolution, fatal=False)
             if not webpage:
@@ -129,6 +174,7 @@ def extract_format(page, version):
                 # invalid version is requested.
                 self.report_warning('Unable to download %s version information' % resolution)
                 continue
+
             extract_format(webpage, resolution)
 
         self._sort_formats(formats)

youtube_dl/extractor/videopress.py

@@ -1,7 +1,6 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import random
 import re
 
 from .common import InfoExtractor
@@ -11,6 +10,7 @@
     float_or_none,
     parse_age_limit,
     qualities,
+    random_birthday,
     try_get,
     unified_timestamp,
     urljoin,
@@ -47,13 +47,10 @@ def _extract_urls(webpage):
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
+        query = random_birthday('birth_year', 'birth_month', 'birth_day')
         video = self._download_json(
             'https://public-api.wordpress.com/rest/v1.1/videos/%s' % video_id,
-            video_id, query={
+            video_id, query=query)
-                'birth_month': random.randint(1, 12),
-                'birth_day': random.randint(1, 31),
-                'birth_year': random.randint(1950, 1995),
-            })
 
         title = video['title']
 

youtube_dl/utils.py

@@ -11,6 +11,7 @@
 import ctypes
 import datetime
 import email.utils
+import email.header
 import errno
 import functools
 import gzip
@@ -2097,6 +2098,58 @@ def update_Request(req, url=None, data=None, headers={}, query={}):
     return new_req
 
 
+def try_multipart_encode(data, boundary):
+    content_type = 'multipart/form-data; boundary=%s' % boundary
+
+    out = b''
+    for k, v in data.items():
+        out += b'--' + boundary.encode('ascii') + b'\r\n'
+        if isinstance(k, compat_str):
+            k = k.encode('utf-8')
+        if isinstance(v, compat_str):
+            v = v.encode('utf-8')
+        # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
+        # suggests sending UTF-8 directly. Firefox sends UTF-8, too
+        content = b'Content-Disposition: form-data; name="%s"\r\n\r\n' % k + v + b'\r\n'
+        if boundary.encode('ascii') in content:
+            raise ValueError('Boundary overlaps with data')
+        out += content
+
+    out += b'--' + boundary.encode('ascii') + b'--\r\n'
+
+    return out, content_type
+
+
+def multipart_encode(data, boundary=None):
+    '''
+    Encode a dict to RFC 7578-compliant form-data
+
+    data:
+        A dict where keys and values can be either Unicode or bytes-like
+        objects.
+    boundary:
+        If specified a Unicode object, it's used as the boundary. Otherwise
+        a random boundary is generated.
+
+    Reference: https://tools.ietf.org/html/rfc7578
+    '''
+    has_specified_boundary = boundary is not None
+
+    while True:
+        if boundary is None:
+            boundary = '---------------' + str(random.randrange(0x0fffffff, 0xffffffff))
+
+        try:
+            out, content_type = try_multipart_encode(data, boundary)
+            break
+        except ValueError:
+            if has_specified_boundary:
+                raise
+            boundary = None
+
+    return out, content_type
+
+
 def dict_get(d, key_or_keys, default=None, skip_false_values=True):
     if isinstance(key_or_keys, (list, tuple)):
         for key in key_or_keys:
@@ -3760,3 +3813,11 @@ def write_xattr(path, key, value):
                         "Couldn't find a tool to set the xattrs. "
                         "Install either the python 'xattr' module, "
                         "or the 'xattr' binary.")
+
+
+def random_birthday(year_field, month_field, day_field):
+    return {
+        year_field: str(random.randint(1950, 1995)),
+        month_field: str(random.randint(1, 12)),
+        day_field: str(random.randint(1, 31)),
+    }