]>
Commit | Line | Data |
---|---|---|
27940ca0 PN |
1 | from .common import InfoExtractor |
2 | from ..compat import ( | |
27940ca0 PN |
3 | compat_str, |
4 | compat_urllib_parse_urlparse, | |
5 | ) | |
6 | from ..utils import ( | |
7 | urljoin, | |
8 | int_or_none, | |
3c3a07ee | 9 | parse_codecs, |
4dfbf869 | 10 | parse_qs, |
27940ca0 | 11 | try_get, |
27940ca0 PN |
12 | ) |
13 | ||
14 | ||
15 | def _raw_id(src_url): | |
16 | return compat_urllib_parse_urlparse(src_url).path.split('/')[-1] | |
17 | ||
18 | ||
19 | class SeznamZpravyIE(InfoExtractor): | |
3c3a07ee | 20 | _VALID_URL = r'https?://(?:www\.)?seznamzpravy\.cz/iframe/player\?.*\bsrc=' |
bfd973ec | 21 | _EMBED_REGEX = [r'<iframe\b[^>]+\bsrc=(["\'])(?P<url>(?:https?:)?//(?:www\.)?seznamzpravy\.cz/iframe/player\?.*?)\1'] |
27940ca0 | 22 | _TESTS = [{ |
3c3a07ee | 23 | 'url': 'https://www.seznamzpravy.cz/iframe/player?duration=241&serviceSlug=zpravy&src=https%3A%2F%2Fv39-a.sdn.szn.cz%2Fv_39%2Fvmd%2F5999c902ea707c67d8e267a9%3Ffl%3Dmdk%2C432f65a0%7C&itemType=video&autoPlay=false&title=Sv%C4%9Bt%20bez%20obalu%3A%20%C4%8Ce%C5%A1t%C3%AD%20voj%C3%A1ci%20na%20mis%C3%ADch%20(kr%C3%A1tk%C3%A1%20verze)&series=Sv%C4%9Bt%20bez%20obalu&serviceName=Seznam%20Zpr%C3%A1vy&poster=%2F%2Fd39-a.sdn.szn.cz%2Fd_39%2Fc_img_F_I%2FR5puJ.jpeg%3Ffl%3Dcro%2C0%2C0%2C1920%2C1080%7Cres%2C1200%2C%2C1%7Cjpg%2C80%2C%2C1&width=1920&height=1080&cutFrom=0&cutTo=0&splVersion=VOD&contentId=170889&contextId=35990&showAdvert=true&collocation=&autoplayPossible=true&embed=&isVideoTooShortForPreroll=false&isVideoTooLongForPostroll=true&videoCommentOpKey=&videoCommentId=&version=4.0.76&dotService=zpravy&gemiusPrismIdentifier=bVc1ZIb_Qax4W2v5xOPGpMeCP31kFfrTzj0SqPTLh_b.Z7&zoneIdPreroll=seznam.pack.videospot&skipOffsetPreroll=5§ionPrefixPreroll=%2Fzpravy', |
27940ca0 PN |
24 | 'info_dict': { |
25 | 'id': '170889', | |
26 | 'ext': 'mp4', | |
27 | 'title': 'Svět bez obalu: Čeští vojáci na misích (krátká verze)', | |
3c3a07ee S |
28 | 'thumbnail': r're:^https?://.*\.jpe?g', |
29 | 'duration': 241, | |
30 | 'series': 'Svět bez obalu', | |
31 | }, | |
32 | 'params': { | |
33 | 'skip_download': True, | |
34 | }, | |
35 | }, { | |
36 | # with Location key | |
37 | 'url': 'https://www.seznamzpravy.cz/iframe/player?duration=null&serviceSlug=zpravy&src=https%3A%2F%2Flive-a.sdn.szn.cz%2Fv_39%2F59e468fe454f8472a96af9fa%3Ffl%3Dmdk%2C5c1e2840%7C&itemType=livevod&autoPlay=false&title=P%C5%99edseda%20KDU-%C4%8CSL%20Pavel%20B%C4%9Blobr%C3%A1dek%20ve%20volebn%C3%AD%20V%C3%BDzv%C4%9B%20Seznamu&series=V%C3%BDzva&serviceName=Seznam%20Zpr%C3%A1vy&poster=%2F%2Fd39-a.sdn.szn.cz%2Fd_39%2Fc_img_G_J%2FjTBCs.jpeg%3Ffl%3Dcro%2C0%2C0%2C1280%2C720%7Cres%2C1200%2C%2C1%7Cjpg%2C80%2C%2C1&width=16&height=9&cutFrom=0&cutTo=0&splVersion=VOD&contentId=185688&contextId=38489&showAdvert=true&collocation=&hideFullScreen=false&hideSubtitles=false&embed=&isVideoTooShortForPreroll=false&isVideoTooShortForPreroll2=false&isVideoTooLongForPostroll=false&fakePostrollZoneID=seznam.clanky.zpravy.preroll&fakePrerollZoneID=seznam.clanky.zpravy.preroll&videoCommentId=&trim=default_16x9&noPrerollVideoLength=30&noPreroll2VideoLength=undefined&noMidrollVideoLength=0&noPostrollVideoLength=999999&autoplayPossible=true&version=5.0.41&dotService=zpravy&gemiusPrismIdentifier=zD3g7byfW5ekpXmxTVLaq5Srjw5i4hsYo0HY1aBwIe..27&zoneIdPreroll=seznam.pack.videospot&skipOffsetPreroll=5§ionPrefixPreroll=%2Fzpravy%2Fvyzva&zoneIdPostroll=seznam.pack.videospot&skipOffsetPostroll=5§ionPrefixPostroll=%2Fzpravy%2Fvyzva®ression=false', | |
38 | 'info_dict': { | |
39 | 'id': '185688', | |
40 | 'ext': 'mp4', | |
41 | 'title': 'Předseda KDU-ČSL Pavel Bělobrádek ve volební Výzvě Seznamu', | |
42 | 'thumbnail': r're:^https?://.*\.jpe?g', | |
43 | 'series': 'Výzva', | |
44 | }, | |
45 | 'params': { | |
46 | 'skip_download': True, | |
47 | }, | |
27940ca0 PN |
48 | }] |
49 | ||
50 | def _extract_sdn_formats(self, sdn_url, video_id): | |
51 | sdn_data = self._download_json(sdn_url, video_id) | |
3c3a07ee S |
52 | |
53 | if sdn_data.get('Location'): | |
54 | sdn_url = sdn_data['Location'] | |
55 | sdn_data = self._download_json(sdn_url, video_id) | |
56 | ||
27940ca0 PN |
57 | formats = [] |
58 | mp4_formats = try_get(sdn_data, lambda x: x['data']['mp4'], dict) or {} | |
3c3a07ee S |
59 | for format_id, format_data in mp4_formats.items(): |
60 | relative_url = format_data.get('url') | |
27940ca0 PN |
61 | if not relative_url: |
62 | continue | |
63 | ||
64 | try: | |
3c3a07ee | 65 | width, height = format_data.get('resolution') |
27940ca0 PN |
66 | except (TypeError, ValueError): |
67 | width, height = None, None | |
68 | ||
3c3a07ee S |
69 | f = { |
70 | 'url': urljoin(sdn_url, relative_url), | |
71 | 'format_id': 'http-%s' % format_id, | |
72 | 'tbr': int_or_none(format_data.get('bandwidth'), scale=1000), | |
27940ca0 PN |
73 | 'width': int_or_none(width), |
74 | 'height': int_or_none(height), | |
3c3a07ee S |
75 | } |
76 | f.update(parse_codecs(format_data.get('codec'))) | |
77 | formats.append(f) | |
78 | ||
79 | pls = sdn_data.get('pls', {}) | |
27940ca0 | 80 | |
3c3a07ee S |
81 | def get_url(format_id): |
82 | return try_get(pls, lambda x: x[format_id]['url'], compat_str) | |
83 | ||
84 | dash_rel_url = get_url('dash') | |
27940ca0 | 85 | if dash_rel_url: |
3c3a07ee S |
86 | formats.extend(self._extract_mpd_formats( |
87 | urljoin(sdn_url, dash_rel_url), video_id, mpd_id='dash', | |
88 | fatal=False)) | |
27940ca0 | 89 | |
3c3a07ee | 90 | hls_rel_url = get_url('hls') |
27940ca0 | 91 | if hls_rel_url: |
3c3a07ee S |
92 | formats.extend(self._extract_m3u8_formats( |
93 | urljoin(sdn_url, hls_rel_url), video_id, ext='mp4', | |
94 | m3u8_id='hls', fatal=False)) | |
27940ca0 | 95 | |
27940ca0 PN |
96 | return formats |
97 | ||
98 | def _real_extract(self, url): | |
4dfbf869 | 99 | params = parse_qs(url) |
3c3a07ee | 100 | |
27940ca0 | 101 | src = params['src'][0] |
3c3a07ee | 102 | title = params['title'][0] |
27940ca0 | 103 | video_id = params.get('contentId', [_raw_id(src)])[0] |
3c3a07ee S |
104 | formats = self._extract_sdn_formats(src + 'spl2,2,VOD', video_id) |
105 | ||
106 | duration = int_or_none(params.get('duration', [None])[0]) | |
107 | series = params.get('series', [None])[0] | |
108 | thumbnail = params.get('poster', [None])[0] | |
27940ca0 PN |
109 | |
110 | return { | |
111 | 'id': video_id, | |
3c3a07ee S |
112 | 'title': title, |
113 | 'thumbnail': thumbnail, | |
114 | 'duration': duration, | |
115 | 'series': series, | |
116 | 'formats': formats, | |
27940ca0 PN |
117 | } |
118 | ||
119 | ||
120 | class SeznamZpravyArticleIE(InfoExtractor): | |
3c3a07ee | 121 | _VALID_URL = r'https?://(?:www\.)?(?:seznam\.cz/zpravy|seznamzpravy\.cz)/clanek/(?:[^/?#&]+)-(?P<id>\d+)' |
27940ca0 PN |
122 | _API_URL = 'https://apizpravy.seznam.cz/' |
123 | ||
124 | _TESTS = [{ | |
125 | # two videos on one page, with SDN URL | |
126 | 'url': 'https://www.seznamzpravy.cz/clanek/jejich-svet-na-nas-utoci-je-lepsi-branit-se-na-jejich-pisecku-rika-reziser-a-major-v-zaloze-marhoul-35990', | |
27940ca0 | 127 | 'info_dict': { |
3c3a07ee S |
128 | 'id': '35990', |
129 | 'title': 'md5:6011c877a36905f28f271fcd8dcdb0f2', | |
130 | 'description': 'md5:933f7b06fa337a814ba199d3596d27ba', | |
131 | }, | |
132 | 'playlist_count': 2, | |
27940ca0 PN |
133 | }, { |
134 | # video with live stream URL | |
135 | 'url': 'https://www.seznam.cz/zpravy/clanek/znovu-do-vlady-s-ano-pavel-belobradek-ve-volebnim-specialu-seznamu-38489', | |
136 | 'info_dict': { | |
3c3a07ee S |
137 | 'id': '38489', |
138 | 'title': 'md5:8fa1afdc36fd378cf0eba2b74c5aca60', | |
139 | 'description': 'md5:428e7926a1a81986ec7eb23078004fb4', | |
140 | }, | |
141 | 'playlist_count': 1, | |
27940ca0 PN |
142 | }] |
143 | ||
3c3a07ee S |
144 | def _real_extract(self, url): |
145 | article_id = self._match_id(url) | |
27940ca0 | 146 | |
3c3a07ee | 147 | webpage = self._download_webpage(url, article_id) |
27940ca0 | 148 | |
3c3a07ee | 149 | info = self._search_json_ld(webpage, article_id, default={}) |
27940ca0 | 150 | |
3c3a07ee S |
151 | title = info.get('title') or self._og_search_title(webpage, fatal=False) |
152 | description = info.get('description') or self._og_search_description(webpage) | |
27940ca0 | 153 | |
3c3a07ee | 154 | return self.playlist_result([ |
2c9d3b99 | 155 | self.url_result(entry_url, ie=SeznamZpravyIE.ie_key()) |
bfd973ec | 156 | for entry_url in SeznamZpravyIE._extract_embed_urls(url, webpage)], |
3c3a07ee | 157 | article_id, title, description) |