[hketv] Improve and simplify (closes #18696)
authorSergey M․ <dstftw@gmail.com>
Sun, 20 Jan 2019 10:42:25 +0000 (17:42 +0700)
committerSergey M․ <dstftw@gmail.com>
Sun, 20 Jan 2019 10:43:11 +0000 (17:43 +0700)
youtube_dl/extractor/hketv.py

index b5790cdee220db906a63d4722622c14317529c76..b57927fc13071dbb56a03557604c5e5c19874833 100644 (file)
@@ -8,8 +8,8 @@ from ..utils import (
     ExtractorError,
     int_or_none,
     merge_dicts,
+    parse_count,
     str_or_none,
-    str_to_int,
     try_get,
     unified_strdate,
     urlencode_postdata,
@@ -30,20 +30,12 @@ class HKETVIE(InfoExtractor):
             'id': '2932360618',
             'ext': 'mp4',
             'title': '喜閱一生(共享閱讀樂) (中、英文字幕可供選擇)',
-            'description': '本節目輯錄了「閱讀滿Fun嘉年華」和「二○一八響應世界閱讀日――悅愛閱讀・愈讀愈愛」的活動花絮,並由學者、作家、演藝界人士等,分享培養子女閱讀興趣和習慣的方法,以及呼籲大家一同分享閱讀的樂趣。',
+            'description': 'md5:d5286d05219ef50e0613311cbe96e560',
             'upload_date': '20181024',
             'duration': 900,
-            'subtitles': {
-                'en': [{
-                    'url': 'https://apps.hkedcity.net/media/mediaplayer/caption.php?f=74395&lang=en',
-                    'ext': 'srt',
-                }],
-                'zh-Hant': [{
-                    'url': 'https://apps.hkedcity.net/media/mediaplayer/caption.php?f=74395&lang=qmt',
-                    'ext': 'srt',
-                }],
-            }
+            'subtitles': 'count:2',
         },
+        'skip': 'Geo restricted to HK',
     }, {
         'url': 'https://www.hkedcity.net/etv/resource/972641418',
         'md5': '1ed494c1c6cf7866a8290edad9b07dc9',
@@ -51,11 +43,15 @@ class HKETVIE(InfoExtractor):
             'id': '972641418',
             'ext': 'mp4',
             'title': '衣冠楚楚 (天使系列之一)',
-            'description': '天國仙境,有兩位可愛的天使小姐妹。她們對幾千年來天使衣著一成不變頗有不滿。她們下望人世間:只見人們穿著七彩繽紛、款式各異的服裝,漂亮極了。天使姐妹決定下凡考察衣著,以設計天使新裝。  下到人間,姐妹試穿各式各樣的衣著,引發連串奇特有趣的情節:她們穿著校服在街上閒逛時,被女警誤認為逃學而送回學校,到校後又被體育老師誤認為是新同學,匆匆忙忙換上運動服後在操場上大顯神通。她們穿著護士服在醫院散步時,又被誤認為當班護士,而投入追尋失蹤病童、治病救人的工作中去。姐妹倆還到過玩具店,與布娃娃們談論衣著。她們也去過服裝設計學校,被當成小模特兒而試穿各式服裝。最令姐妹倆興奮的是一場盛大的民族服裝表演會。身穿盛裝的十二個民族的少女在台上翩翩起舞,各種服飾七彩繽紛、美不勝收。姐妹們情不自禁地穿上民族服裝,小天使變成了少數民族姑娘……最後天使姐妹回到天上,對於天使究竟穿甚麼樣的衣服好,她們還是拿不定主意。  節目通過天使姐妹的奇特經歷,反復示範各式衣服鞋襪的正確讀音及談論衣著時的常用句式,並以盛大的民族服裝表演活動,帶出有關服裝的文化知識。內容豐富而饒有趣味。',
+            'description': 'md5:10bb3d659421e74f58e5db5691627b0f',
             'upload_date': '20070109',
             'duration': 907,
             'subtitles': {},
         },
+        'params': {
+            'geo_verification_proxy': '<HK proxy here>',
+        },
+        'skip': 'Geo restricted to HK',
     }]
 
     _CC_LANGS = {
@@ -69,117 +65,127 @@ class HKETVIE(InfoExtractor):
         '\u0e44\u0e17\u0e22': 'th',
         '\u0627\u0631\u062f\u0648': 'ur',
     }
+    _FORMAT_HEIGHTS = {
+        'SD': 360,
+        'HD': 720,
+    }
+    _APPS_BASE_URL = 'https://apps.hkedcity.net'
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
-        title = self._html_search_meta('ed_title', webpage, fatal=True)
 
-        file_id = self._html_search_regex(r'post_var\["file_id"\]\s*=\s*(.+?);', webpage, 'file ID')
-        curr_url = self._html_search_regex(r'post_var\["curr_url"\]\s*=\s*"(.+?)";', webpage, 'curr URL')
+        title = (
+            self._html_search_meta(
+                ('ed_title', 'search.ed_title'), webpage, default=None) or
+            self._search_regex(
+                r'data-favorite_title_(?:eng|chi)=(["\'])(?P<id>(?:(?!\1).)+)\1',
+                webpage, 'title', default=None, group='url') or
+            self._html_search_regex(
+                r'<h1>([^<]+)</h1>', webpage, 'title', default=None) or
+            self._og_search_title(webpage)
+        )
+
+        file_id = self._search_regex(
+            r'post_var\[["\']file_id["\']\s*\]\s*=\s*(.+?);',
+            webpage, 'file ID')
+        curr_url = self._search_regex(
+            r'post_var\[["\']curr_url["\']\s*\]\s*=\s*"(.+?)";',
+            webpage, 'curr URL')
         data = {
             'action': 'get_info',
             'curr_url': curr_url,
             'file_id': file_id,
             'video_url': file_id,
         }
-        _APPS_BASE_URL = 'https://apps.hkedcity.net'
-        handler_url = _APPS_BASE_URL + '/media/play/handler.php'
 
         response = self._download_json(
-            handler_url, video_id,
+            self._APPS_BASE_URL + '/media/play/handler.php', video_id,
             data=urlencode_postdata(data),
-            headers=merge_dicts({'Content-Type': 'application/x-www-form-urlencoded'},
-                                self.geo_verification_headers()))
+            headers=merge_dicts({
+                'Content-Type': 'application/x-www-form-urlencoded'},
+                self.geo_verification_headers()))
 
         result = response['result']
 
+        if not response.get('success') or not response.get('access'):
+            error = clean_html(response.get('access_err_msg'))
+            if 'Video streaming is not available in your country' in error:
+                self.raise_geo_restricted(
+                    msg=error, countries=self._GEO_COUNTRIES)
+            else:
+                raise ExtractorError(error, expected=True)
+
         formats = []
-        subtitles = {}
 
-        if response.get('success') and response.get('access'):
-            width = int_or_none(result.get('width'))
-            height = int_or_none(result.get('height'))
-
-            playlist0 = try_get(result, lambda x: x['playlist'][0], dict)
-            fmts = playlist0.get('sources')
-            for fmt in fmts:
-                file_path = fmt.get('file')
-                if file_path:
-                    file_url = urljoin(_APPS_BASE_URL, file_path)
-                    # If we ever wanted to provide the final resolved URL that
-                    # does not require cookies, albeit with a shorter lifespan:
-                    #     urlh = self._downloader.urlopen(file_url)
-                    #     resolved_url = urlh.geturl()
-
-                    label = fmt.get('label')
-                    w = None
-                    h = None
-                    if label == 'HD':
-                        h = 720
-                    elif label == 'SD':
-                        h = 360
-                    if h:
-                        if width and height:
-                            w = h * width // height
-                        else:
-                            w = h * 4 // 3
-
-                    formats.append({
-                        'format_id': label,
-                        'ext': fmt.get('type'),
-                        'url': file_url,
-                        'width': w,
-                        'height': h,
-                    })
-
-            tracks = playlist0.get('tracks', [])
-            for track in tracks:
-                if not isinstance(track, dict):
-                    continue
-                track_kind = str_or_none(track.get('kind'))
-                if not track_kind or not isinstance(track_kind, compat_str):
-                    continue
-                if track_kind.lower() not in ('captions', 'subtitles'):
-                    continue
-                track_url = urljoin(_APPS_BASE_URL, track.get('file'))
-                if not track_url:
-                    continue
-                track_label = track.get('label')
-                subtitles.setdefault(self._CC_LANGS.get(track_label, track_label), []).append({
+        width = int_or_none(result.get('width'))
+        height = int_or_none(result.get('height'))
+
+        playlist0 = result['playlist'][0]
+        for fmt in playlist0['sources']:
+            file_url = urljoin(self._APPS_BASE_URL, fmt.get('file'))
+            if not file_url:
+                continue
+            # If we ever wanted to provide the final resolved URL that
+            # does not require cookies, albeit with a shorter lifespan:
+            #     urlh = self._downloader.urlopen(file_url)
+            #     resolved_url = urlh.geturl()
+            label = fmt.get('label')
+            h = self._FORMAT_HEIGHTS.get(label)
+            w = h * width // height if h and width and height else None
+            formats.append({
+                'format_id': label,
+                'ext': fmt.get('type'),
+                'url': file_url,
+                'width': w,
+                'height': h,
+            })
+        self._sort_formats(formats)
+
+        subtitles = {}
+        tracks = try_get(playlist0, lambda x: x['tracks'], list) or []
+        for track in tracks:
+            if not isinstance(track, dict):
+                continue
+            track_kind = str_or_none(track.get('kind'))
+            if not track_kind or not isinstance(track_kind, compat_str):
+                continue
+            if track_kind.lower() not in ('captions', 'subtitles'):
+                continue
+            track_url = urljoin(self._APPS_BASE_URL, track.get('file'))
+            if not track_url:
+                continue
+            track_label = track.get('label')
+            subtitles.setdefault(self._CC_LANGS.get(
+                track_label, track_label), []).append({
                     'url': self._proto_relative_url(track_url),
                     'ext': 'srt',
                 })
 
-        else:
-            error = clean_html(response.get('access_err_msg'))
-            if 'Video streaming is not available in your country' in error:
-                self.raise_geo_restricted(msg=error, countries=self._GEO_COUNTRIES)
-            else:
-                raise ExtractorError(error)
-
         # Likes
         emotion = self._download_json(
-            'https://emocounter.hkedcity.net/handler.php',
-            video_id,
+            'https://emocounter.hkedcity.net/handler.php', video_id,
             data=urlencode_postdata({
                 'action': 'get_emotion',
                 'data[bucket_id]': 'etv',
                 'data[identifier]': video_id,
             }),
             headers={'Content-Type': 'application/x-www-form-urlencoded'},
-            fatal=False)
-        like_count = int_or_none(try_get(emotion, lambda x: x['data']['emotion_data'][0]['count']))
+            fatal=False) or {}
+        like_count = int_or_none(try_get(
+            emotion, lambda x: x['data']['emotion_data'][0]['count']))
 
         return {
             'id': video_id,
             'title': title,
-            'description': self._html_search_meta('description', webpage, fatal=False),
-            'upload_date': unified_strdate(self._html_search_meta('ed_date', webpage, fatal=False), day_first=False),
+            'description': self._html_search_meta(
+                'description', webpage, fatal=False),
+            'upload_date': unified_strdate(self._html_search_meta(
+                'ed_date', webpage, fatal=False), day_first=False),
             'duration': int_or_none(result.get('length')),
             'formats': formats,
             'subtitles': subtitles,
-            'thumbnail': urljoin(_APPS_BASE_URL, result.get('image')),
-            'view_count': str_to_int(result.get('view_count')),
+            'thumbnail': urljoin(self._APPS_BASE_URL, result.get('image')),
+            'view_count': parse_count(result.get('view_count')),
             'like_count': like_count,
         }