From fac55558ad55344d72cf66033e5f2d2076b2f8cb Mon Sep 17 00:00:00 2001 From: Philipp Hagemeister Date: Mon, 24 Mar 2014 23:21:20 +0100 Subject: [PATCH] [washingtonpost] Add extractor (Fixes #2622) --- test/test_utils.py | 8 ++ youtube_dl/extractor/__init__.py | 1 + youtube_dl/extractor/washingtonpost.py | 103 +++++++++++++++++++++++++ youtube_dl/utils.py | 4 + 4 files changed, 116 insertions(+) create mode 100644 youtube_dl/extractor/washingtonpost.py diff --git a/test/test_utils.py b/test/test_utils.py index e920d661f..2348c0415 100644 --- a/test/test_utils.py +++ b/test/test_utils.py @@ -10,6 +10,7 @@ sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) # Various small unit tests import io +import json import xml.etree.ElementTree #from youtube_dl.utils import htmlentity_transform @@ -36,6 +37,7 @@ from youtube_dl.utils import ( urlencode_postdata, xpath_with_ns, parse_iso8601, + strip_jsonp, ) if sys.version_info < (3, 0): @@ -272,5 +274,11 @@ class TestUtil(unittest.TestCase): self.assertEqual(parse_iso8601('2014-03-23T22:04:26+0000'), 1395612266) self.assertEqual(parse_iso8601('2014-03-23T22:04:26Z'), 1395612266) + def test_strip_jsonp(self): + stripped = strip_jsonp('cb ([ {"id":"532cb",\n\n\n"x":\n3}\n]\n);') + d = json.loads(stripped) + self.assertEqual(d, [{"id": "532cb", "x": 3}]) + + if __name__ == '__main__': unittest.main() diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py index 39b250b10..56b382aed 100644 --- a/youtube_dl/extractor/__init__.py +++ b/youtube_dl/extractor/__init__.py @@ -282,6 +282,7 @@ from .vine import VineIE from .viki import VikiIE from .vk import VKIE from .vube import VubeIE +from .washingtonpost import WashingtonPostIE from .wat import WatIE from .wdr import WDRIE from .weibo import WeiboIE diff --git a/youtube_dl/extractor/washingtonpost.py b/youtube_dl/extractor/washingtonpost.py new file mode 100644 index 000000000..cb8f0887d --- /dev/null +++ b/youtube_dl/extractor/washingtonpost.py @@ -0,0 +1,103 @@ +from __future__ import unicode_literals + +import re + +from .common import InfoExtractor +from ..utils import ( + int_or_none, + strip_jsonp, +) + + +class WashingtonPostIE(InfoExtractor): + _VALID_URL = r'^https?://(?:www\.)?washingtonpost\.com/.*?/(?P[^/]+)/(?:$|[?#])' + _TEST = { + 'url': 'http://www.washingtonpost.com/sf/national/2014/03/22/sinkhole-of-bureaucracy/', + 'playlist': [{ + 'md5': 'c3f4b4922ffa259243f68e928db2db8c', + 'info_dict': { + 'id': 'fc433c38-b146-11e3-b8b3-44b1d1cd4c1f', + 'ext': 'mp4', + 'title': 'Breaking Points: The Paper Mine', + 'duration': 1287, + 'description': 'Overly complicated paper pushing is nothing new to government bureaucracy. But the way federal retirement applications are filed may be the most outdated. David Fahrenthold explains.', + 'uploader': 'The Washington Post', + 'timestamp': 1395527908, + 'upload_date': '20140322', + }, + }, { + 'md5': 'f645a07652c2950cd9134bb852c5f5eb', + 'info_dict': { + 'id': '41255e28-b14a-11e3-b8b3-44b1d1cd4c1f', + 'ext': 'mp4', + 'title': 'The town bureaucracy sustains', + 'description': 'Underneath the friendly town of Boyers is a sea of government paperwork. In a disused limestone mine, hundreds of locals now track, file and process retirement applications for the federal government. We set out to find out what it\'s like to do paperwork 230 feet underground.', + 'duration': 2217, + 'timestamp': 1395528005, + 'upload_date': '20140322', + 'uploader': 'The Washington Post', + }, + }] + } + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + page_id = mobj.group('id') + + webpage = self._download_webpage(url, page_id) + title = self._og_search_title(webpage) + uuids = re.findall(r'data-video-uuid="([^"]+)"', webpage) + entries = [] + for i, uuid in enumerate(uuids, start=1): + vinfo_all = self._download_json( + 'http://www.washingtonpost.com/posttv/c/videojson/%s?resType=jsonp' % uuid, + page_id, + transform_source=strip_jsonp, + note='Downloading information of video %d/%d' % (i, len(uuids)) + ) + vinfo = vinfo_all[0]['contentConfig'] + uploader = vinfo.get('credits', {}).get('source') + timestamp = int_or_none( + vinfo.get('dateConfig', {}).get('dateFirstPublished'), 1000) + + formats = [{ + 'format_id': ( + '%s-%s-%s' % (s.get('type'), s.get('width'), s.get('bitrate')) + if s.get('width') + else s.get('type')), + 'vbr': s.get('bitrate') if s.get('width') != 0 else None, + 'width': s.get('width'), + 'height': s.get('height'), + 'acodec': s.get('audioCodec'), + 'vcodec': s.get('videoCodec') if s.get('width') != 0 else 'none', + 'filesize': s.get('fileSize'), + 'url': s.get('url'), + 'ext': 'mp4', + 'protocol': { + 'MP4': 'http', + 'F4F': 'f4m', + }.get(s.get('type')) + } for s in vinfo.get('streams', [])] + source_media_url = vinfo.get('sourceMediaURL') + if source_media_url: + formats.append({ + 'format_id': 'source_media', + 'url': source_media_url, + }) + self._sort_formats(formats) + entries.append({ + 'id': uuid, + 'title': vinfo['title'], + 'description': vinfo.get('blurb'), + 'uploader': uploader, + 'formats': formats, + 'duration': int_or_none(vinfo.get('videoDuration'), 100), + 'timestamp': timestamp, + }) + + return { + '_type': 'playlist', + 'entries': entries, + 'id': page_id, + 'title': title, + } diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index fd03b0d7c..29c9b1a4c 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -1328,3 +1328,7 @@ US_RATINGS = { 'R': 16, 'NC': 18, } + + +def strip_jsonp(code): + return re.sub(r'(?s)^[a-zA-Z_]+\s*\(\s*(.*)\);\s*?\s*$', r'\1', code)