[naver] improve extraction

- improve geo-restriction handling
- extract automatic captions
- extract uploader metadata
- extract VLive HLS formats

youtube_dl/extractor/naver.py

@@ -1,68 +1,33 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import re
+
 from .common import InfoExtractor
 from ..utils import (
+    clean_html,
+    dict_get,
     ExtractorError,
+    get_element_by_class,
     int_or_none,
+    try_get,
     update_url_query,
 )
 
 
-class NaverIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:m\.)?tv(?:cast)?\.naver\.com/v/(?P<id>\d+)'
+class NaverBaseIE(InfoExtractor):
+    _CAPTION_EXT_RE = r'\.(?:ttml|vtt)'
 
-    _TESTS = [{
-        'url': 'http://tv.naver.com/v/81652',
-        'info_dict': {
-            'id': '81652',
-            'ext': 'mp4',
-            'title': '[9월 모의고사 해설강의][수학_김상희] 수학 A형 16~20번',
-            'description': '합격불변의 법칙 메가스터디 | 메가스터디 수학 김상희 선생님이 9월 모의고사 수학A형 16번에서 20번까지 해설강의를 공개합니다.',
-            'upload_date': '20130903',
-        },
-    }, {
-        'url': 'http://tv.naver.com/v/395837',
-        'md5': '638ed4c12012c458fefcddfd01f173cd',
-        'info_dict': {
-            'id': '395837',
-            'ext': 'mp4',
-            'title': '9년이 지나도 아픈 기억, 전효성의 아버지',
-            'description': 'md5:5bf200dcbf4b66eb1b350d1eb9c753f7',
-            'upload_date': '20150519',
-        },
-        'skip': 'Georestricted',
-    }, {
-        'url': 'http://tvcast.naver.com/v/81652',
-        'only_matching': True,
-    }]
-
-    def _real_extract(self, url):
-        video_id = self._match_id(url)
-        webpage = self._download_webpage(url, video_id)
-
-        vid = self._search_regex(
-            r'videoId["\']\s*:\s*(["\'])(?P<value>(?:(?!\1).)+)\1', webpage,
-            'video id', fatal=None, group='value')
-        in_key = self._search_regex(
-            r'inKey["\']\s*:\s*(["\'])(?P<value>(?:(?!\1).)+)\1', webpage,
-            'key', default=None, group='value')
-
-        if not vid or not in_key:
-            error = self._html_search_regex(
-                r'(?s)<div class="(?:nation_error|nation_box|error_box)">\s*(?:<!--.*?-->)?\s*<p class="[^"]+">(?P<msg>.+?)</p>\s*</div>',
-                webpage, 'error', default=None)
-            if error:
-                raise ExtractorError(error, expected=True)
-            raise ExtractorError('couldn\'t extract vid and key')
+    def _extract_video_info(self, video_id, vid, key):
         video_data = self._download_json(
             'http://play.rmcnmv.naver.com/vod/play/v2.0/' + vid,
             video_id, query={
-                'key': in_key,
+                'key': key,
             })
         meta = video_data['meta']
         title = meta['subject']
         formats = []
+        get_list = lambda x: try_get(video_data, lambda y: y[x + 's']['list'], list) or []
 
         def extract_formats(streams, stream_type, query={}):
             for stream in streams:
@@ -73,7 +38,7 @@ def extract_formats(streams, stream_type, query={}):
                 encoding_option = stream.get('encodingOption', {})
                 bitrate = stream.get('bitrate', {})
                 formats.append({
-                    'format_id': '%s_%s' % (stream.get('type') or stream_type, encoding_option.get('id') or encoding_option.get('name')),
+                    'format_id': '%s_%s' % (stream.get('type') or stream_type, dict_get(encoding_option, ('name', 'id'))),
                     'url': stream_url,
                     'width': int_or_none(encoding_option.get('width')),
                     'height': int_or_none(encoding_option.get('height')),
@@ -83,7 +48,7 @@ def extract_formats(streams, stream_type, query={}):
                     'protocol': 'm3u8_native' if stream_type == 'HLS' else None,
                 })
 
-        extract_formats(video_data.get('videos', {}).get('list', []), 'H264')
+        extract_formats(get_list('video'), 'H264')
         for stream_set in video_data.get('streams', []):
             query = {}
             for param in stream_set.get('keys', []):
@@ -101,28 +66,101 @@ def extract_formats(streams, stream_type, query={}):
                     'mp4', 'm3u8_native', m3u8_id=stream_type, fatal=False))
         self._sort_formats(formats)
 
+        replace_ext = lambda x, y: re.sub(self._CAPTION_EXT_RE, '.' + y, x)
+
+        def get_subs(caption_url):
+            if re.search(self._CAPTION_EXT_RE, caption_url):
+                return [{
+                    'url': replace_ext(caption_url, 'ttml'),
+                }, {
+                    'url': replace_ext(caption_url, 'vtt'),
+                }]
+            else:
+                return [{'url': caption_url}]
+
+        automatic_captions = {}
         subtitles = {}
-        for caption in video_data.get('captions', {}).get('list', []):
+        for caption in get_list('caption'):
             caption_url = caption.get('source')
             if not caption_url:
                 continue
-            subtitles.setdefault(caption.get('language') or caption.get('locale'), []).append({
-                'url': caption_url,
-            })
+            sub_dict = automatic_captions if caption.get('type') == 'auto' else subtitles
+            sub_dict.setdefault(dict_get(caption, ('locale', 'language')), []).extend(get_subs(caption_url))
 
-        upload_date = self._search_regex(
-            r'<span[^>]+class="date".*?(\d{4}\.\d{2}\.\d{2})',
-            webpage, 'upload date', fatal=False)
-        if upload_date:
-            upload_date = upload_date.replace('.', '')
+        user = meta.get('user', {})
 
         return {
             'id': video_id,
             'title': title,
             'formats': formats,
             'subtitles': subtitles,
-            'description': self._og_search_description(webpage),
-            'thumbnail': meta.get('cover', {}).get('source') or self._og_search_thumbnail(webpage),
+            'automatic_captions': automatic_captions,
+            'thumbnail': try_get(meta, lambda x: x['cover']['source']),
             'view_count': int_or_none(meta.get('count')),
-            'upload_date': upload_date,
+            'uploader_id': user.get('id'),
+            'uploader': user.get('name'),
+            'uploader_url': user.get('url'),
         }
+
+
+class NaverIE(NaverBaseIE):
+    _VALID_URL = r'https?://(?:m\.)?tv(?:cast)?\.naver\.com/(?:v|embed)/(?P<id>\d+)'
+    _GEO_BYPASS = False
+    _TESTS = [{
+        'url': 'http://tv.naver.com/v/81652',
+        'info_dict': {
+            'id': '81652',
+            'ext': 'mp4',
+            'title': '[9월 모의고사 해설강의][수학_김상희] 수학 A형 16~20번',
+            'description': '메가스터디 수학 김상희 선생님이 9월 모의고사 수학A형 16번에서 20번까지 해설강의를 공개합니다.',
+            'upload_date': '20130903',
+            'uploader': '메가스터디, 합격불변의 법칙',
+            'uploader_id': 'megastudy',
+        },
+    }, {
+        'url': 'http://tv.naver.com/v/395837',
+        'md5': '8a38e35354d26a17f73f4e90094febd3',
+        'info_dict': {
+            'id': '395837',
+            'ext': 'mp4',
+            'title': '9년이 지나도 아픈 기억, 전효성의 아버지',
+            'description': 'md5:eb6aca9d457b922e43860a2a2b1984d3',
+            'upload_date': '20150519',
+            'uploader': '4가지쇼 시즌2',
+            'uploader_id': 'wrappinguser29',
+        },
+        'skip': 'Georestricted',
+    }, {
+        'url': 'http://tvcast.naver.com/v/81652',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        content = self._download_json(
+            'https://tv.naver.com/api/contents/json/v/' + video_id,
+            video_id, headers=self.geo_verification_headers())
+        player_json = content.get('playerJson') or {}
+
+        vid = player_json.get('videoId')
+        in_key = player_json.get('inKey')
+
+        if not vid or not in_key:
+            player_auth = player_json.get('playerAuth')
+            if player_auth == 'notCountry':
+                self.raise_geo_restricted(countries=['KR'])
+            elif player_auth == 'notLogin':
+                self.raise_login_required()
+            raise ExtractorError('couldn\'t extract vid and key')
+        info = self._extract_video_info(video_id, vid, in_key)
+
+        clip_info_html = content.get('clipInfoHtml')
+        if clip_info_html:
+            info['description'] = clean_html(get_element_by_class('desc', clip_info_html))
+            upload_date = self._search_regex(
+                r'<span[^>]+class="date".*?(\d{4}\.\d{2}\.\d{2})',
+                clip_info_html, 'upload date', fatal=False)
+            if upload_date:
+                info['upload_date'] = upload_date.replace('.', '')
+
+        return info

youtube_dl/extractor/vlive.py

@@ -6,22 +6,18 @@
 import itertools
 
 from .common import InfoExtractor
-from ..compat import (
-    compat_urllib_parse_urlencode,
-    compat_str,
-)
+from .naver import NaverBaseIE
+from ..compat import compat_str
 from ..utils import (
-    dict_get,
     ExtractorError,
-    float_or_none,
-    int_or_none,
+    merge_dicts,
     remove_start,
     try_get,
     urlencode_postdata,
 )
 
 
-class VLiveIE(InfoExtractor):
+class VLiveIE(NaverBaseIE):
     IE_NAME = 'vlive'
     _VALID_URL = r'https?://(?:(?:www|m)\.)?vlive\.tv/video/(?P<id>[0-9]+)'
     _NETRC_MACHINE = 'vlive'
@@ -34,6 +30,7 @@ class VLiveIE(InfoExtractor):
             'title': "[V LIVE] Girl's Day's Broadcast",
             'creator': "Girl's Day",
             'view_count': int,
+            'uploader_id': 'muploader_a',
         },
     }, {
         'url': 'http://www.vlive.tv/video/16937',
@@ -44,6 +41,7 @@ class VLiveIE(InfoExtractor):
             'creator': 'EXO',
             'view_count': int,
             'subtitles': 'mincount:12',
+            'uploader_id': 'muploader_j',
         },
         'params': {
             'skip_download': True,
@@ -187,45 +185,9 @@ def _replay(self, video_id, webpage, long_video_id, key):
                     'This video is only available for CH+ subscribers')
             long_video_id, key = video_info['vid'], video_info['inkey']
 
-        playinfo = self._download_json(
-            'http://global.apis.naver.com/rmcnmv/rmcnmv/vod_play_videoInfo.json?%s'
-            % compat_urllib_parse_urlencode({
-                'videoId': long_video_id,
-                'key': key,
-                'ptc': 'http',
-                'doct': 'json',  # document type (xml or json)
-                'cpt': 'vtt',  # captions type (vtt or ttml)
-            }), video_id)
-
-        formats = [{
-            'url': vid['source'],
-            'format_id': vid.get('encodingOption', {}).get('name'),
-            'abr': float_or_none(vid.get('bitrate', {}).get('audio')),
-            'vbr': float_or_none(vid.get('bitrate', {}).get('video')),
-            'width': int_or_none(vid.get('encodingOption', {}).get('width')),
-            'height': int_or_none(vid.get('encodingOption', {}).get('height')),
-            'filesize': int_or_none(vid.get('size')),
-        } for vid in playinfo.get('videos', {}).get('list', []) if vid.get('source')]
-        self._sort_formats(formats)
-
-        view_count = int_or_none(playinfo.get('meta', {}).get('count'))
-
-        subtitles = {}
-        for caption in playinfo.get('captions', {}).get('list', []):
-            lang = dict_get(caption, ('locale', 'language', 'country', 'label'))
-            if lang and caption.get('source'):
-                subtitles[lang] = [{
-                    'ext': 'vtt',
-                    'url': caption['source']}]
-
-        info = self._get_common_fields(webpage)
-        info.update({
-            'id': video_id,
-            'formats': formats,
-            'view_count': view_count,
-            'subtitles': subtitles,
-        })
-        return info
+        return merge_dicts(
+            self._get_common_fields(webpage),
+            self._extract_video_info(video_id, long_video_id, key))
 
     def _download_init_page(self, video_id):
         return self._download_webpage(