From defaf19f5de839777cda1654d42b54850a78d8a7 Mon Sep 17 00:00:00 2001 From: Your Name Date: Fri, 2 Jan 2015 02:20:04 -0600 Subject: [PATCH 1/2] Push api updates to simplify audiomack, add support for albums --- youtube_dl/extractor/__init__.py | 2 +- youtube_dl/extractor/audiomack.py | 110 ++++++++++++++++++++++-------- 2 files changed, 84 insertions(+), 28 deletions(-) diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py index 6c5827f88..e4c51f238 100644 --- a/youtube_dl/extractor/__init__.py +++ b/youtube_dl/extractor/__init__.py @@ -26,7 +26,7 @@ from .arte import ( ArteTVEmbedIE, ) from .atresplayer import AtresPlayerIE -from .audiomack import AudiomackIE +from .audiomack import AudiomackIE, AudiomackAlbumIE from .auengine import AUEngineIE from .azubu import AzubuIE from .bambuser import BambuserIE, BambuserChannelIE diff --git a/youtube_dl/extractor/audiomack.py b/youtube_dl/extractor/audiomack.py index 622b20989..b6f3039c7 100644 --- a/youtube_dl/extractor/audiomack.py +++ b/youtube_dl/extractor/audiomack.py @@ -9,20 +9,21 @@ import time class AudiomackIE(InfoExtractor): - _VALID_URL = r'https?://(?:www\.)?audiomack\.com/song/(?P[\w/-]+)' + _VALID_URL = r'https?://(?:www\.)?audiomack\.com/(song)/(?P[\w/-]+)' IE_NAME = 'audiomack' _TESTS = [ - # hosted on audiomack + # audiomack { 'url': 'http://www.audiomack.com/song/roosh-williams/extraordinary', 'info_dict': { - 'id': 'roosh-williams/extraordinary', - 'ext': 'mp3', - 'title': 'Roosh Williams - Extraordinary' + 'id': '310086', + "ext": "mp3", + "artist": "Roosh Williams", + 'title': 'Extraordinary' } }, - # hosted on soundcloud via audiomack + # audiomack through soundcloud { 'add_ie': ['Soundcloud'], 'url': 'http://www.audiomack.com/song/xclusiveszone/take-kare', @@ -37,33 +38,88 @@ class AudiomackIE(InfoExtractor): }, ] + @staticmethod + def create_song_dictionary(api_response, album_url_tag, track_no=0): + # All keys are the same in audiomack api and InfoExtractor format + entry = {key: api_response[key] for key in ["title", "artist", "id", "url"] if key in api_response} + # Fudge values in the face of missing metadata + if "id" not in entry: + entry["id"] = track_no + if "title" not in entry: + entry["title"] = album_url_tag + return entry + def _real_extract(self, url): - video_id = self._match_id(url) + # URLs end with [uploader name]/[uploader title] + # this title is whatever the user types in, and is rarely + # the proper song title. Real metadata is in the api response + album_url_tag = self._match_id(url) + # Request the extended version of the api for extra fields like artist and title api_response = self._download_json( - "http://www.audiomack.com/api/music/url/song/%s?_=%d" % ( - video_id, time.time()), - video_id) + "http://www.audiomack.com/api/music/url/song/%s?extended=1&_=%d" % ( + album_url_tag, time.time()), + album_url_tag) - if "url" not in api_response: - raise ExtractorError("Unable to deduce api url of song") - realurl = api_response["url"] + # API is inconsistent with errors + if "url" not in api_response or not api_response["url"] or "error" in api_response: + raise ExtractorError("Invalid url %s", url) # Audiomack wraps a lot of soundcloud tracks in their branded wrapper - # - if so, pass the work off to the soundcloud extractor - if SoundcloudIE.suitable(realurl): - return {'_type': 'url', 'url': realurl, 'ie_key': 'Soundcloud'} + # if so, pass the work off to the soundcloud extractor + if SoundcloudIE.suitable(api_response["url"]): + return {'_type': 'url', 'url': api_response["url"], 'ie_key': 'Soundcloud'} - webpage = self._download_webpage(url, video_id) - artist = self._html_search_regex( - r'(.*?)', webpage, "artist") - songtitle = self._html_search_regex( - r'

.*?(.*?)

', - webpage, "title") - title = artist + " - " + songtitle + return self.create_song_dictionary(api_response, album_url_tag) - return { - 'id': video_id, - 'title': title, - 'url': realurl, + +class AudiomackAlbumIE(InfoExtractor): + _VALID_URL = r'https?://(?:www\.)?audiomack\.com/album/(?P[\w/-]+)' + IE_NAME = 'audiomack:album' + _TESTS = [ + # Standard album playlist + { + 'url': 'http://www.audiomack.com/album/flytunezcom/tha-tour-part-2-mixtape', + "playlist_count": 15, + 'info_dict': + { + 'id': "812251", + 'title': "Tha Tour: Part 2 (Official Mixtape)" + } + }, + # Album playlist ripped from fakeshoredrive with no metadata + { + "url": "http://www.audiomack.com/album/fakeshoredrive/ppp-pistol-p-project", + "playlist_count": 10 } + ] + + def _real_extract(self, url): + # URLs end with [uploader name]/[uploader title] + # this title is whatever the user types in, and is rarely + # the proper song title. Real metadata is in the api response + album_url_tag = self._match_id(url) + result = {"_type": "playlist", "entries": []} + # There is no one endpoint for album metadata - instead it is included/repeated in each song's metadata + # Therefore we don't know how many songs the album has and must infi-loop until failure + track_no = 0 + while True: + # Get song's metadata + api_response = self._download_json("http://www.audiomack.com/api/music/url/album/%s/%d?extended=1&_=%d" + % (album_url_tag, track_no, time.time()), album_url_tag) + + # Total failure, only occurs when url is totally wrong + # Won't happen in middle of valid playlist (next case) + if "url" not in api_response or "error" in api_response: + raise ExtractorError("Invalid url for track %d of album url %s" % (track_no, url)) + # URL is good but song id doesn't exist - usually means end of playlist + elif not api_response["url"]: + break + else: + # Pull out the album metadata and add to result (if it exists) + for resultkey, apikey in [("id", "album_id"), ("title", "album_title")]: + if apikey in api_response and resultkey not in result: + result[resultkey] = api_response[apikey] + result["entries"].append(AudiomackIE.create_song_dictionary(api_response, album_url_tag, track_no)) + track_no += 1 + return result From ff0813313ac33b6abd03e9322b706dc83c5aeb14 Mon Sep 17 00:00:00 2001 From: Your Name Date: Thu, 8 Jan 2015 18:35:33 -0600 Subject: [PATCH 2/2] Minor style changes --- youtube_dl/extractor/audiomack.py | 54 +++++++++++++++---------------- 1 file changed, 27 insertions(+), 27 deletions(-) diff --git a/youtube_dl/extractor/audiomack.py b/youtube_dl/extractor/audiomack.py index b6f3039c7..73cde78d3 100644 --- a/youtube_dl/extractor/audiomack.py +++ b/youtube_dl/extractor/audiomack.py @@ -9,21 +9,21 @@ import time class AudiomackIE(InfoExtractor): - _VALID_URL = r'https?://(?:www\.)?audiomack\.com/(song)/(?P[\w/-]+)' + _VALID_URL = r'https?://(?:www\.)?audiomack\.com/song/(?P[\w/-]+)' IE_NAME = 'audiomack' _TESTS = [ - # audiomack + # hosted on audiomack { 'url': 'http://www.audiomack.com/song/roosh-williams/extraordinary', 'info_dict': { 'id': '310086', - "ext": "mp3", - "artist": "Roosh Williams", + 'ext': 'mp3', + 'artist': 'Roosh Williams', 'title': 'Extraordinary' } }, - # audiomack through soundcloud + # audiomack wrapper around soundcloud song { 'add_ie': ['Soundcloud'], 'url': 'http://www.audiomack.com/song/xclusiveszone/take-kare', @@ -41,12 +41,12 @@ class AudiomackIE(InfoExtractor): @staticmethod def create_song_dictionary(api_response, album_url_tag, track_no=0): # All keys are the same in audiomack api and InfoExtractor format - entry = {key: api_response[key] for key in ["title", "artist", "id", "url"] if key in api_response} + entry = {key: api_response[key] for key in ['title', 'artist', 'id', 'url'] if key in api_response} # Fudge values in the face of missing metadata - if "id" not in entry: - entry["id"] = track_no - if "title" not in entry: - entry["title"] = album_url_tag + if 'id' not in entry: + entry['id'] = track_no + if 'title' not in entry: + entry['title'] = album_url_tag return entry def _real_extract(self, url): @@ -57,18 +57,18 @@ class AudiomackIE(InfoExtractor): # Request the extended version of the api for extra fields like artist and title api_response = self._download_json( - "http://www.audiomack.com/api/music/url/song/%s?extended=1&_=%d" % ( + 'http://www.audiomack.com/api/music/url/song/%s?extended=1&_=%d' % ( album_url_tag, time.time()), album_url_tag) # API is inconsistent with errors - if "url" not in api_response or not api_response["url"] or "error" in api_response: - raise ExtractorError("Invalid url %s", url) + if 'url' not in api_response or not api_response['url'] or 'error' in api_response: + raise ExtractorError('Invalid url %s', url) # Audiomack wraps a lot of soundcloud tracks in their branded wrapper # if so, pass the work off to the soundcloud extractor - if SoundcloudIE.suitable(api_response["url"]): - return {'_type': 'url', 'url': api_response["url"], 'ie_key': 'Soundcloud'} + if SoundcloudIE.suitable(api_response['url']): + return {'_type': 'url', 'url': api_response['url'], 'ie_key': 'Soundcloud'} return self.create_song_dictionary(api_response, album_url_tag) @@ -80,17 +80,17 @@ class AudiomackAlbumIE(InfoExtractor): # Standard album playlist { 'url': 'http://www.audiomack.com/album/flytunezcom/tha-tour-part-2-mixtape', - "playlist_count": 15, + 'playlist_count': 15, 'info_dict': { - 'id': "812251", - 'title': "Tha Tour: Part 2 (Official Mixtape)" + 'id': '812251', + 'title': 'Tha Tour: Part 2 (Official Mixtape)' } }, # Album playlist ripped from fakeshoredrive with no metadata { - "url": "http://www.audiomack.com/album/fakeshoredrive/ppp-pistol-p-project", - "playlist_count": 10 + 'url': 'http://www.audiomack.com/album/fakeshoredrive/ppp-pistol-p-project', + 'playlist_count': 10 } ] @@ -99,27 +99,27 @@ class AudiomackAlbumIE(InfoExtractor): # this title is whatever the user types in, and is rarely # the proper song title. Real metadata is in the api response album_url_tag = self._match_id(url) - result = {"_type": "playlist", "entries": []} + result = {'_type': 'playlist', 'entries': []} # There is no one endpoint for album metadata - instead it is included/repeated in each song's metadata # Therefore we don't know how many songs the album has and must infi-loop until failure track_no = 0 while True: # Get song's metadata - api_response = self._download_json("http://www.audiomack.com/api/music/url/album/%s/%d?extended=1&_=%d" + api_response = self._download_json('http://www.audiomack.com/api/music/url/album/%s/%d?extended=1&_=%d' % (album_url_tag, track_no, time.time()), album_url_tag) # Total failure, only occurs when url is totally wrong # Won't happen in middle of valid playlist (next case) - if "url" not in api_response or "error" in api_response: - raise ExtractorError("Invalid url for track %d of album url %s" % (track_no, url)) + if 'url' not in api_response or 'error' in api_response: + raise ExtractorError('Invalid url for track %d of album url %s' % (track_no, url)) # URL is good but song id doesn't exist - usually means end of playlist - elif not api_response["url"]: + elif not api_response['url']: break else: # Pull out the album metadata and add to result (if it exists) - for resultkey, apikey in [("id", "album_id"), ("title", "album_title")]: + for resultkey, apikey in [('id', 'album_id'), ('title', 'album_title')]: if apikey in api_response and resultkey not in result: result[resultkey] = api_response[apikey] - result["entries"].append(AudiomackIE.create_song_dictionary(api_response, album_url_tag, track_no)) + result['entries'].append(AudiomackIE.create_song_dictionary(api_response, album_url_tag, track_no)) track_no += 1 return result