[polskieradio] Improve extraction (Closes #9813)

This commit is contained in:
Sergey M․ 2016-06-25 23:18:40 +07:00
parent 2d185706ea
commit 0463b77a1f
No known key found for this signature in database
GPG key ID: 2C393E0F18A9236D

View file

@ -1,74 +1,91 @@
# coding: utf-8
from __future__ import unicode_literals
from .common import InfoExtractor
from ..utils import int_or_none
import re
import calendar
from datetime import datetime
from .common import InfoExtractor
from ..compat import (
compat_str,
compat_urllib_parse_unquote,
)
from ..utils import (
int_or_none,
strip_or_none,
unified_timestamp,
)
class PolskieRadioIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?polskieradio\.pl/[0-9]+/[0-9]+/Artykul/(?P<id>[0-9]+),.+'
_VALID_URL = r'https?://(?:www\.)?polskieradio\.pl/\d+/\d+/Artykul/(?P<id>[0-9]+)'
_TESTS = [{
'url': 'http://www.polskieradio.pl/7/5102/Artykul/1587943,Prof-Andrzej-Nowak-o-historii-nie-da-sie-myslec-beznamietnie',
'md5': '2984ee6ce9046d91fc233bc1a864a09a',
'info_dict': {
'id': '1587943',
'ext': 'mp3',
'title': 'Prof. Andrzej Nowak: o historii nie da się myśleć beznamiętnie',
'description': 'md5:12f954edbf3120c5e7075e17bf9fc5c5',
'release_date': '20160227',
'upload_date': '20160227',
},
'playlist': [{
'md5': '2984ee6ce9046d91fc233bc1a864a09a',
'info_dict': {
'id': '1540576',
'ext': 'mp3',
'title': 'md5:d4623290d4ac983bf924061c75c23a0d',
'timestamp': 1456594200,
'duration': 2364
}
'upload_date': '20160227',
'duration': 2364,
},
}],
}, {
'url': 'http://www.polskieradio.pl/265/5217/Artykul/1635803,Euro-2016-nie-ma-miejsca-na-blad-Polacy-graja-ze-Szwajcaria-o-cwiercfinal',
'info_dict': {
'id': '1635803',
'title': 'Euro 2016: nie ma miejsca na błąd. Polacy grają ze Szwajcarią o ćwierćfinał',
'description': 'md5:01cb7d0cad58664095d72b51a1ebada2',
},
'playlist_mincount': 12,
}, {
'url': 'http://polskieradio.pl/9/305/Artykul/1632955,Bardzo-popularne-slowo-remis',
'md5': '68a393e25b942c1a76872f56d303a31a',
'info_dict': {
'id': '1632955',
'ext': 'mp3',
'title': 'Bardzo popularne słowo: remis',
'description': 'md5:3b58dfae614100abc0f175a0b26d5680',
'release_date': '20160617',
'upload_date': '20160617',
'timestamp': 1466184900,
'duration': 393
}
'only_matching': True,
}, {
'url': 'http://www.polskieradio.pl/7/5102/Artykul/1587943',
'only_matching': True,
}]
def _real_extract(self, url):
video_id = self._match_id(url)
webpage = self._download_webpage(url, video_id)
metadata_string = self._html_search_regex(r'<span class="play pr-media-play" data-media=(\{.+\})>', webpage, 'metadata')
metadata = self._parse_json(metadata_string, video_id)
playlist_id = self._match_id(url)
title = self._og_search_title(webpage)
if title is not None:
title = title.strip()
webpage = self._download_webpage(url, playlist_id)
description = self._og_search_description(webpage)
if description is not None:
description = description.strip()
content = self._search_regex(
r'(?s)<div[^>]+class="audio atarticle"[^>]*>(.+?)<script>',
webpage, 'content')
release_date = self._html_search_regex(r'Data emisji:[^0-9]+([0-9]{1,2}\.[0-9]{2}\.[0-9]{4})', webpage, 'release date', fatal=False)
if release_date is not None:
release_date = datetime.strptime(release_date, '%d.%m.%Y').strftime('%Y%m%d')
timestamp = unified_timestamp(self._html_search_regex(
r'(?s)<span[^>]+id="datetime2"[^>]*>(.+?)</span>',
webpage, 'timestamp', fatal=False))
upload_datetime = self._html_search_regex(r'<span id="datetime2" class="time">\s+(.+)\s+</span>', webpage, 'release time', fatal=False)
if upload_datetime is not None:
timestamp = calendar.timegm(datetime.strptime(upload_datetime, '%d.%m.%Y %H:%M').timetuple())
else:
timestamp = None
entries = []
return {
'id': video_id,
'title': title,
'description': description,
'display_id': metadata.get('id'),
'duration': int_or_none(metadata.get('length')),
'url': self._proto_relative_url(metadata.get('file'), 'http:'),
'release_date': release_date,
'timestamp': timestamp
}
media_urls = set()
for data_media in re.findall(r'<[^>]+data-media=({[^>]+})', content):
media = self._parse_json(data_media, playlist_id, fatal=False)
if not media.get('file') or not media.get('desc'):
continue
media_url = self._proto_relative_url(media['file'], 'http:')
if media_url in media_urls:
continue
media_urls.add(media_url)
entries.append({
'id': compat_str(media['id']),
'url': media_url,
'title': compat_urllib_parse_unquote(media['desc']),
'duration': int_or_none(media.get('length')),
'vcodec': 'none' if media.get('provider') == 'audio' else None,
'timestamp': timestamp,
})
title = self._og_search_title(webpage).strip()
description = strip_or_none(self._og_search_description(webpage))
return self.playlist_result(entries, playlist_id, title, description)