summaryrefslogtreecommitdiffstats
path: root/yt_dlp/extractor/banbye.py
blob: d10bdf8daafa968875552d042b3636acc623bea8 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
import math
import urllib.parse

from .common import InfoExtractor
from ..utils import (
    InAdvancePagedList,
    format_field,
    traverse_obj,
    unified_timestamp,
)


class BanByeBaseIE(InfoExtractor):
    _API_BASE = 'https://api.banbye.com'
    _CDN_BASE = 'https://cdn.banbye.com'
    _VIDEO_BASE = 'https://banbye.com/watch'

    @staticmethod
    def _extract_playlist_id(url, param='playlist'):
        return urllib.parse.parse_qs(
            urllib.parse.urlparse(url).query).get(param, [None])[0]

    def _extract_playlist(self, playlist_id):
        data = self._download_json(f'{self._API_BASE}/playlists/{playlist_id}', playlist_id)
        return self.playlist_result([
            self.url_result(f'{self._VIDEO_BASE}/{video_id}', BanByeIE)
            for video_id in data['videoIds']], playlist_id, data.get('name'))


class BanByeIE(BanByeBaseIE):
    _VALID_URL = r'https?://(?:www\.)?banbye\.com/(?:en/)?watch/(?P<id>[\w-]+)'
    _TESTS = [{
        'url': 'https://banbye.com/watch/v_ytfmvkVYLE8T',
        'md5': '2f4ea15c5ca259a73d909b2cfd558eb5',
        'info_dict': {
            'id': 'v_ytfmvkVYLE8T',
            'ext': 'mp4',
            'title': 'md5:5ec098f88a0d796f987648de6322ba0f',
            'description': 'md5:4d94836e73396bc18ef1fa0f43e5a63a',
            'uploader': 'wRealu24',
            'channel_id': 'ch_wrealu24',
            'channel_url': 'https://banbye.com/channel/ch_wrealu24',
            'timestamp': 1647604800,
            'upload_date': '20220318',
            'duration': 1931,
            'thumbnail': r're:https?://.*\.webp',
            'tags': 'count:5',
            'like_count': int,
            'dislike_count': int,
            'view_count': int,
            'comment_count': int,
        },
    }, {
        'url': 'https://banbye.com/watch/v_2JjQtqjKUE_F?playlistId=p_Ld82N6gBw_OJ',
        'info_dict': {
            'title': 'Krzysztof Karoń',
            'id': 'p_Ld82N6gBw_OJ',
        },
        'playlist_mincount': 9,
    }, {
        'url': 'https://banbye.com/watch/v_kb6_o1Kyq-CD',
        'info_dict': {
            'id': 'v_kb6_o1Kyq-CD',
            'ext': 'mp4',
            'title': 'Co tak naprawdę dzieje się we Francji?! Czy Warszawa a potem cała Polska będzie drugim Paryżem?!🤔🇵🇱',
            'description': 'md5:82be4c0e13eae8ea1ca8b9f2e07226a8',
            'uploader': 'Marcin Rola - MOIM ZDANIEM!🇵🇱',
            'channel_id': 'ch_QgWnHvDG2fo5',
            'channel_url': 'https://banbye.com/channel/ch_QgWnHvDG2fo5',
            'duration': 597,
            'timestamp': 1688642656,
            'upload_date': '20230706',
            'thumbnail': 'https://cdn.banbye.com/video/v_kb6_o1Kyq-CD/96.webp',
            'tags': ['Paryż', 'Francja', 'Polska', 'Imigranci', 'Morawiecki', 'Tusk'],
            'like_count': int,
            'dislike_count': int,
            'view_count': int,
            'comment_count': int,
        },
    }]

    def _real_extract(self, url):
        video_id = self._match_id(url)
        playlist_id = self._extract_playlist_id(url, 'playlistId')

        if self._yes_playlist(playlist_id, video_id):
            return self._extract_playlist(playlist_id)

        data = self._download_json(f'{self._API_BASE}/videos/{video_id}', video_id)
        thumbnails = [{
            'id': f'{quality}p',
            'url': f'{self._CDN_BASE}/video/{video_id}/{quality}.webp',
        } for quality in [48, 96, 144, 240, 512, 1080]]
        formats = [{
            'format_id': f'http-{quality}p',
            'quality': quality,
            'url': f'{self._CDN_BASE}/video/{video_id}/{quality}.mp4',
        } for quality in data['quality']]

        return {
            'id': video_id,
            'title': data.get('title'),
            'description': data.get('desc'),
            'uploader': traverse_obj(data, ('channel', 'name')),
            'channel_id': data.get('channelId'),
            'channel_url': format_field(data, 'channelId', 'https://banbye.com/channel/%s'),
            'timestamp': unified_timestamp(data.get('publishedAt')),
            'duration': data.get('duration'),
            'tags': data.get('tags'),
            'formats': formats,
            'thumbnails': thumbnails,
            'like_count': data.get('likes'),
            'dislike_count': data.get('dislikes'),
            'view_count': data.get('views'),
            'comment_count': data.get('commentCount'),
        }


class BanByeChannelIE(BanByeBaseIE):
    _VALID_URL = r'https?://(?:www\.)?banbye\.com/(?:en/)?channel/(?P<id>\w+)'
    _TESTS = [{
        'url': 'https://banbye.com/channel/ch_wrealu24',
        'info_dict': {
            'title': 'wRealu24',
            'id': 'ch_wrealu24',
            'description': 'md5:da54e48416b74dfdde20a04867c0c2f6',
        },
        'playlist_mincount': 791,
    }, {
        'url': 'https://banbye.com/channel/ch_wrealu24?playlist=p_Ld82N6gBw_OJ',
        'info_dict': {
            'title': 'Krzysztof Karoń',
            'id': 'p_Ld82N6gBw_OJ',
        },
        'playlist_count': 9,
    }]
    _PAGE_SIZE = 100

    def _real_extract(self, url):
        channel_id = self._match_id(url)
        playlist_id = self._extract_playlist_id(url)

        if playlist_id:
            return self._extract_playlist(playlist_id)

        def page_func(page_num):
            data = self._download_json(f'{self._API_BASE}/videos', channel_id, query={
                'channelId': channel_id,
                'sort': 'new',
                'limit': self._PAGE_SIZE,
                'offset': page_num * self._PAGE_SIZE,
            }, note=f'Downloading page {page_num + 1}')
            return [
                self.url_result(f"{self._VIDEO_BASE}/{video['_id']}", BanByeIE)
                for video in data['items']
            ]

        channel_data = self._download_json(f'{self._API_BASE}/channels/{channel_id}', channel_id)
        entries = InAdvancePagedList(
            page_func,
            math.ceil(channel_data['videoCount'] / self._PAGE_SIZE),
            self._PAGE_SIZE)

        return self.playlist_result(
            entries, channel_id, channel_data.get('name'), channel_data.get('description'))