From 24ce9e5f51478f1e1cd352d7971aade2ffadded2 Mon Sep 17 00:00:00 2001 From: gfabiano Date: Wed, 21 Jun 2017 01:59:45 +0200 Subject: [PATCH 1/3] [npr] Add new extractor --- youtube_dl/extractor/extractors.py | 5 +- youtube_dl/extractor/npr.py | 135 ++++++++++++++++++++--------- 2 files changed, 97 insertions(+), 43 deletions(-) diff --git a/youtube_dl/extractor/extractors.py b/youtube_dl/extractor/extractors.py index e97691daa..fddbc9ce6 100644 --- a/youtube_dl/extractor/extractors.py +++ b/youtube_dl/extractor/extractors.py @@ -709,7 +709,10 @@ from .npo import ( VPROIE, WNLIE, ) -from .npr import NprIE +from .npr import ( + NprPlaylistIE, + NprVideoIE, +) from .nrk import ( NRKIE, NRKPlaylistIE, diff --git a/youtube_dl/extractor/npr.py b/youtube_dl/extractor/npr.py index 1777aa10b..a5ff6e8c4 100644 --- a/youtube_dl/extractor/npr.py +++ b/youtube_dl/extractor/npr.py @@ -1,33 +1,79 @@ from __future__ import unicode_literals from .common import InfoExtractor -from ..compat import compat_urllib_parse_urlencode -from ..utils import ( - int_or_none, - qualities, -) +from ..utils import int_or_none -class NprIE(InfoExtractor): - _VALID_URL = r'https?://(?:www\.)?npr\.org/player/v2/mediaPlayer\.html\?.*\bid=(?P\d+)' +class NprBaseIE(InfoExtractor): + def extract_info(self, id): + json_data = self._download_json( + 'http://api.npr.org/query', id, query={ + 'id': id, + 'fields': 'titles,audio,show,multimedia,text', + 'format': 'json', + 'apiKey': 'MDAzMzQ2MjAyMDEyMzk4MTU1MDg3ZmM3MQ010', + }) + + return json_data['list']['story'][0] + + def extract_formats(self, media_info): + FORMATS_ = { + 'hls': 'm3u8', 'hlsOnDemand': 'm3u8', 'mediastream': 'mp3', + 'mp3': 'mp3', 'mp4': 'mp4', 'wm': 'wm', 'threegp': '3gp' + } + + formats = [] + for format_id, formats_entry in media_info.get('format', {}).items(): + if not formats_entry: + continue + if isinstance(formats_entry, list): + formats_entry = formats_entry[0] + format_url = formats_entry.get('$text') + if not format_url: + continue + if format_id == 'smil': + formats += self._extract_smil_formats( + format_url, + media_info['id'], + fatal=False + ) + continue + elif format_id == 'm3u8': + # Error 404 for some reason + formats += self._extract_m3u8_formats( + format_url, + media_info['id'], + fatal=False + ) + continue + formats.append({ + 'url': format_url, + 'format_id': format_id, + 'ext': FORMATS_.get(format_id), + }) + return formats + + +class NprPlaylistIE(NprBaseIE): + _VALID_URL = r'https?://(?:www\.)?npr\.org/(?:sections/\w+/\d+/\d+/\d+/|player/v2/mediaPlayer\.html\?.*\bid=)(?P\d+)' _TESTS = [{ - 'url': 'http://www.npr.org/player/v2/mediaPlayer.html?id=449974205', + 'url': 'http://www.npr.org/sections/allsongs/2015/10/21/449974205/new-music-from-beach-house-chairlift-cmj-discoveries-and-more', 'info_dict': { 'id': '449974205', 'title': 'New Music From Beach House, Chairlift, CMJ Discoveries And More' }, 'playlist_count': 7, }, { - 'url': 'http://www.npr.org/player/v2/mediaPlayer.html?action=1&t=1&islist=false&id=446928052&m=446929930&live=1', + 'url': 'http://www.npr.org/sections/deceptivecadence/2015/10/09/446928052/music-from-the-shadows-ancient-armenian-hymns-and-piano-jazz', 'info_dict': { 'id': '446928052', 'title': "Songs We Love: Tigran Hamasyan, 'Your Mercy is Boundless'" }, 'playlist': [{ - 'md5': '12fa60cb2d3ed932f53609d4aeceabf1', + 'md5': 'df2917b738fdd2358a9f0e7e49fcdf2e', 'info_dict': { 'id': '446929930', - 'ext': 'mp3', + # 'ext': 'mp4', 'title': 'Your Mercy is Boundless (Bazum en Qo gtutyunqd)', 'duration': 402, }, @@ -36,47 +82,52 @@ class NprIE(InfoExtractor): def _real_extract(self, url): playlist_id = self._match_id(url) - - config = self._download_json( - 'http://api.npr.org/query?%s' % compat_urllib_parse_urlencode({ - 'id': playlist_id, - 'fields': 'titles,audio,show', - 'format': 'json', - 'apiKey': 'MDAzMzQ2MjAyMDEyMzk4MTU1MDg3ZmM3MQ010', - }), playlist_id) - - story = config['list']['story'][0] - - KNOWN_FORMATS = ('threegp', 'mp4', 'mp3') - quality = qualities(KNOWN_FORMATS) + story = self.extract_info(playlist_id) entries = [] for audio in story.get('audio', []): title = audio.get('title', {}).get('$text') duration = int_or_none(audio.get('duration', {}).get('$text')) - formats = [] - for format_id, formats_entry in audio.get('format', {}).items(): - if not formats_entry: - continue - if isinstance(formats_entry, list): - formats_entry = formats_entry[0] - format_url = formats_entry.get('$text') - if not format_url: - continue - if format_id in KNOWN_FORMATS: - formats.append({ - 'url': format_url, - 'format_id': format_id, - 'ext': formats_entry.get('type'), - 'quality': quality(format_id), - }) - self._sort_formats(formats) entries.append({ 'id': audio['id'], 'title': title, 'duration': duration, - 'formats': formats, + 'formats': self.extract_formats(audio), }) playlist_title = story.get('title', {}).get('$text') return self.playlist_result(entries, playlist_id, playlist_title) + + +class NprVideoIE(NprBaseIE): + _VALID_URL = r'https?://(?:www\.)?npr\.org/event/music/(?P\d+)' + + _TEST = { + 'url': 'http://www.npr.org/event/music/533198237/tigers-jaw-tiny-desk-concert', + 'md5': '5b385e0e96c2731261df9a4ed1ff2cba', + 'info_dict': { + 'id': '533201718', + 'display_id': '533198237', + 'title': 'Tigers Jaw: Tiny Desk Concert', + 'ext': 'mp4', + 'width': 624, + 'height': 351, + } + } + + def _real_extract(self, url): + display_id = self._match_id(url) + story = self.extract_info(display_id) + + video = story.get('multimedia')[0] + + return { + 'url': url, + 'display_id': display_id, + 'title': story.get('title', {}).get('$text'), + 'id': video.get('id'), + 'duration': int_or_none(video.get('duration', {}).get('$text')), + 'formats': self.extract_formats(video), + 'width': int_or_none(video.get('width', {}).get('$text')), + 'height': int_or_none(video.get('height', {}).get('$text')), + } From b97e947558b02814da2b8379b1efacf76fba4942 Mon Sep 17 00:00:00 2001 From: gfabiano Date: Wed, 21 Jun 2017 23:19:02 +0200 Subject: [PATCH 2/3] [npr] Requested fixes --- youtube_dl/extractor/extractors.py | 2 +- youtube_dl/extractor/npr.py | 65 +++++++++++++++++++----------- 2 files changed, 42 insertions(+), 25 deletions(-) diff --git a/youtube_dl/extractor/extractors.py b/youtube_dl/extractor/extractors.py index fddbc9ce6..a6817c1c1 100644 --- a/youtube_dl/extractor/extractors.py +++ b/youtube_dl/extractor/extractors.py @@ -710,7 +710,7 @@ from .npo import ( WNLIE, ) from .npr import ( - NprPlaylistIE, + NprIE, NprVideoIE, ) from .nrk import ( diff --git a/youtube_dl/extractor/npr.py b/youtube_dl/extractor/npr.py index a5ff6e8c4..795b7623f 100644 --- a/youtube_dl/extractor/npr.py +++ b/youtube_dl/extractor/npr.py @@ -1,14 +1,19 @@ from __future__ import unicode_literals +import re + from .common import InfoExtractor -from ..utils import int_or_none +from ..utils import ( + int_or_none, + ExtractorError +) class NprBaseIE(InfoExtractor): - def extract_info(self, id): + def _extract_info(self, id_): json_data = self._download_json( - 'http://api.npr.org/query', id, query={ - 'id': id, + 'http://api.npr.org/query', id_, query={ + 'id': id_, 'fields': 'titles,audio,show,multimedia,text', 'format': 'json', 'apiKey': 'MDAzMzQ2MjAyMDEyMzk4MTU1MDg3ZmM3MQ010', @@ -16,7 +21,7 @@ class NprBaseIE(InfoExtractor): return json_data['list']['story'][0] - def extract_formats(self, media_info): + def _extract_formats(self, media_info): FORMATS_ = { 'hls': 'm3u8', 'hlsOnDemand': 'm3u8', 'mediastream': 'mp3', 'mp3': 'mp3', 'mp4': 'mp4', 'wm': 'wm', 'threegp': '3gp' @@ -39,22 +44,32 @@ class NprBaseIE(InfoExtractor): ) continue elif format_id == 'm3u8': - # Error 404 for some reason - formats += self._extract_m3u8_formats( - format_url, - media_info['id'], - fatal=False + mobj = re.match( + r'(?Phttps?:.+?akamaihd.+?-n-)(?P(?:,*\d+)+)', + format_url ) + if mobj is None: + continue + m3u8_base = mobj.group('baseurl') + qualities = re.findall(r'(\d+)', mobj.group('qualities')) + + for quality in qualities: + formats += self._extract_m3u8_formats( + m3u8_base + '%s.mp4/master.m3u8' % quality, + media_info['id'], + fatal=False + ) continue formats.append({ 'url': format_url, 'format_id': format_id, 'ext': FORMATS_.get(format_id), }) + self._sort_formats(formats) return formats -class NprPlaylistIE(NprBaseIE): +class NprIE(NprBaseIE): _VALID_URL = r'https?://(?:www\.)?npr\.org/(?:sections/\w+/\d+/\d+/\d+/|player/v2/mediaPlayer\.html\?.*\bid=)(?P\d+)' _TESTS = [{ 'url': 'http://www.npr.org/sections/allsongs/2015/10/21/449974205/new-music-from-beach-house-chairlift-cmj-discoveries-and-more', @@ -73,7 +88,7 @@ class NprPlaylistIE(NprBaseIE): 'md5': 'df2917b738fdd2358a9f0e7e49fcdf2e', 'info_dict': { 'id': '446929930', - # 'ext': 'mp4', + 'ext': 'mp4', 'title': 'Your Mercy is Boundless (Bazum en Qo gtutyunqd)', 'duration': 402, }, @@ -82,7 +97,7 @@ class NprPlaylistIE(NprBaseIE): def _real_extract(self, url): playlist_id = self._match_id(url) - story = self.extract_info(playlist_id) + story = self._extract_info(playlist_id) entries = [] for audio in story.get('audio', []): @@ -92,7 +107,7 @@ class NprPlaylistIE(NprBaseIE): 'id': audio['id'], 'title': title, 'duration': duration, - 'formats': self.extract_formats(audio), + 'formats': self._extract_formats(audio), }) playlist_title = story.get('title', {}).get('$text') @@ -106,28 +121,30 @@ class NprVideoIE(NprBaseIE): 'url': 'http://www.npr.org/event/music/533198237/tigers-jaw-tiny-desk-concert', 'md5': '5b385e0e96c2731261df9a4ed1ff2cba', 'info_dict': { - 'id': '533201718', - 'display_id': '533198237', + 'id': '533198237', 'title': 'Tigers Jaw: Tiny Desk Concert', 'ext': 'mp4', 'width': 624, 'height': 351, - } + }, + 'expected_warnings': ['HTTP Error 404'], } def _real_extract(self, url): - display_id = self._match_id(url) - story = self.extract_info(display_id) + video_id = self._match_id(url) + story = self._extract_info(video_id) + + title = story.get('title', {}).get('$text') + if title is None: + ExtractorError('Fail extracting title') video = story.get('multimedia')[0] return { - 'url': url, - 'display_id': display_id, - 'title': story.get('title', {}).get('$text'), - 'id': video.get('id'), + 'id': video_id, + 'title': title, 'duration': int_or_none(video.get('duration', {}).get('$text')), - 'formats': self.extract_formats(video), + 'formats': self._extract_formats(video), 'width': int_or_none(video.get('width', {}).get('$text')), 'height': int_or_none(video.get('height', {}).get('$text')), } From f824539760486e9d29abd89974ea644189a5df99 Mon Sep 17 00:00:00 2001 From: gfabiano Date: Fri, 23 Jun 2017 17:35:01 +0200 Subject: [PATCH 3/3] [npr] Fix m3u8 extraction --- youtube_dl/extractor/npr.py | 28 +++++++++++----------------- 1 file changed, 11 insertions(+), 17 deletions(-) diff --git a/youtube_dl/extractor/npr.py b/youtube_dl/extractor/npr.py index 795b7623f..4769d6143 100644 --- a/youtube_dl/extractor/npr.py +++ b/youtube_dl/extractor/npr.py @@ -1,7 +1,5 @@ from __future__ import unicode_literals -import re - from .common import InfoExtractor from ..utils import ( int_or_none, @@ -44,18 +42,14 @@ class NprBaseIE(InfoExtractor): ) continue elif format_id == 'm3u8': - mobj = re.match( - r'(?Phttps?:.+?akamaihd.+?-n-)(?P(?:,*\d+)+)', - format_url - ) - if mobj is None: - continue - m3u8_base = mobj.group('baseurl') - qualities = re.findall(r'(\d+)', mobj.group('qualities')) - - for quality in qualities: + try: formats += self._extract_m3u8_formats( - m3u8_base + '%s.mp4/master.m3u8' % quality, + format_url.replace('200000', '2000000'), + media_info['id'] + ) + except ExtractorError: + formats += self._extract_m3u8_formats( + format_url, media_info['id'], fatal=False ) @@ -119,13 +113,13 @@ class NprVideoIE(NprBaseIE): _TEST = { 'url': 'http://www.npr.org/event/music/533198237/tigers-jaw-tiny-desk-concert', - 'md5': '5b385e0e96c2731261df9a4ed1ff2cba', + 'md5': '2ca640c9725579ea7d020dd23b9cffc2', 'info_dict': { 'id': '533198237', 'title': 'Tigers Jaw: Tiny Desk Concert', - 'ext': 'mp4', - 'width': 624, - 'height': 351, + 'ext': 'm3u8', + 'width': 1280, + 'height': 720, }, 'expected_warnings': ['HTTP Error 404'], }