debian/control: Add Recommends: python3-pyxattr

[youtubedl] / youtube_dl / extractor / bandcamp.py

diff --git a/youtube_dl/extractor/bandcamp.py b/youtube_dl/extractor/bandcamp.py
index 505877b773d45b36be31d8dea8a6a1766d72d4ca..be41bd5a22477fce2aca4a043799574e148fdc57 100644 (file)
--- a/youtube_dl/extractor/bandcamp.py
+++ b/youtube_dl/extractor/bandcamp.py
@@ -1,7 +1,9 @@
 from __future__ import unicode_literals
 
 import json
 from __future__ import unicode_literals
 
 import json

+import random

 import re
 import re

+import time

 
 from .common import InfoExtractor
 from ..compat import (
 
 from .common import InfoExtractor
 from ..compat import (


@@ -10,11 +12,18 @@ from ..compat import (
 )
 from ..utils import (
     ExtractorError,
 )
 from ..utils import (
     ExtractorError,

+    float_or_none,
+    int_or_none,
+    KNOWN_EXTENSIONS,
+    parse_filesize,
+    unescapeHTML,
+    update_url_query,
+    unified_strdate,

 )
 
 
 class BandcampIE(InfoExtractor):
 )
 
 
 class BandcampIE(InfoExtractor):

-    _VALID_URL = r'https?://.*?\.bandcamp\.com/track/(?P<title>.*)'
+    _VALID_URL = r'https?://.*?\.bandcamp\.com/track/(?P<title>[^/?#&]+)'

     _TESTS = [{
         'url': 'http://youtube-dl.bandcamp.com/track/youtube-dl-test-song',
         'md5': 'c557841d5e50261777a6585648adf439',
     _TESTS = [{
         'url': 'http://youtube-dl.bandcamp.com/track/youtube-dl-test-song',
         'md5': 'c557841d5e50261777a6585648adf439',


@@ -27,12 +36,12 @@ class BandcampIE(InfoExtractor):
         '_skip': 'There is a limit of 200 free downloads / month for the test song'
     }, {
         'url': 'http://benprunty.bandcamp.com/track/lanius-battle',
         '_skip': 'There is a limit of 200 free downloads / month for the test song'
     }, {
         'url': 'http://benprunty.bandcamp.com/track/lanius-battle',

-        'md5': '2b68e5851514c20efdff2afc5603b8b4',
+        'md5': '0369ace6b939f0927e62c67a1a8d9fa7',

         'info_dict': {
             'id': '2650410135',
         'info_dict': {
             'id': '2650410135',

-            'ext': 'mp3',
-            'title': 'Lanius (Battle)',
-            'uploader': 'Ben Prunty Music',
+            'ext': 'aiff',
+            'title': 'Ben Prunty - Lanius (Battle)',
+            'uploader': 'Ben Prunty',

         },
     }]
 
         },
     }]
 


@@ -40,32 +49,38 @@ class BandcampIE(InfoExtractor):
         mobj = re.match(self._VALID_URL, url)
         title = mobj.group('title')
         webpage = self._download_webpage(url, title)
         mobj = re.match(self._VALID_URL, url)
         title = mobj.group('title')
         webpage = self._download_webpage(url, title)

+        thumbnail = self._html_search_meta('og:image', webpage, default=None)

         m_download = re.search(r'freeDownloadPage: "(.*?)"', webpage)
         if not m_download:
             m_trackinfo = re.search(r'trackinfo: (.+),\s*?\n', webpage)
             if m_trackinfo:
                 json_code = m_trackinfo.group(1)
                 data = json.loads(json_code)[0]
         m_download = re.search(r'freeDownloadPage: "(.*?)"', webpage)
         if not m_download:
             m_trackinfo = re.search(r'trackinfo: (.+),\s*?\n', webpage)
             if m_trackinfo:
                 json_code = m_trackinfo.group(1)
                 data = json.loads(json_code)[0]

+                track_id = compat_str(data['id'])
+
+                if not data.get('file'):
+                    raise ExtractorError('Not streamable', video_id=track_id, expected=True)

 
                 formats = []
                 for format_id, format_url in data['file'].items():
                     ext, abr_str = format_id.split('-', 1)
                     formats.append({
                         'format_id': format_id,
 
                 formats = []
                 for format_id, format_url in data['file'].items():
                     ext, abr_str = format_id.split('-', 1)
                     formats.append({
                         'format_id': format_id,

-                        'url': format_url,
+                        'url': self._proto_relative_url(format_url, 'http:'),

                         'ext': ext,
                         'vcodec': 'none',
                         'acodec': ext,
                         'ext': ext,
                         'vcodec': 'none',
                         'acodec': ext,

-                        'abr': int(abr_str),
+                        'abr': int_or_none(abr_str),

                     })
 
                 self._sort_formats(formats)
 
                 return {
                     })
 
                 self._sort_formats(formats)
 
                 return {

-                    'id': compat_str(data['id']),
+                    'id': track_id,

                     'title': data['title'],
                     'title': data['title'],

+                    'thumbnail': thumbnail,

                     'formats': formats,
                     'formats': formats,

-                    'duration': float(data['duration']),
+                    'duration': float_or_none(data.get('duration')),

                 }
             else:
                 raise ExtractorError('No free songs found')
                 }
             else:
                 raise ExtractorError('No free songs found')


@@ -75,41 +90,74 @@ class BandcampIE(InfoExtractor):
             r'(?ms)var TralbumData = .*?[{,]\s*id: (?P<id>\d+),?$',
             webpage, 'video id')
 
             r'(?ms)var TralbumData = .*?[{,]\s*id: (?P<id>\d+),?$',
             webpage, 'video id')
 

-        download_webpage = self._download_webpage(download_link, video_id, 'Downloading free downloads page')
-        # We get the dictionary of the track from some javascript code
-        all_info = self._parse_json(self._search_regex(
-            r'(?sm)items: (.*?),$', download_webpage, 'items'), video_id)
-        info = all_info[0]
-        # We pick mp3-320 for now, until format selection can be easily implemented.
-        mp3_info = info['downloads']['mp3-320']
-        # If we try to use this url it says the link has expired
-        initial_url = mp3_info['url']
-        m_url = re.match(
-            r'(?P<server>http://(.*?)\.bandcamp\.com)/download/track\?enc=mp3-320&fsig=(?P<fsig>.*?)&id=(?P<id>.*?)&ts=(?P<ts>.*)$',
-            initial_url)
-        # We build the url we will use to get the final track url
-        # This url is build in Bandcamp in the script download_bunde_*.js
-        request_url = '%s/statdownload/track?enc=mp3-320&fsig=%s&id=%s&ts=%s&.rand=665028774616&.vrs=1' % (m_url.group('server'), m_url.group('fsig'), video_id, m_url.group('ts'))
-        final_url_webpage = self._download_webpage(request_url, video_id, 'Requesting download url')
-        # If we could correctly generate the .rand field the url would be
-        # in the "download_url" key
-        final_url = self._search_regex(
-            r'"retry_url":"(.*?)"', final_url_webpage, 'final video URL')
+        download_webpage = self._download_webpage(
+            download_link, video_id, 'Downloading free downloads page')
+
+        blob = self._parse_json(
+            self._search_regex(
+                r'data-blob=(["\'])(?P<blob>{.+?})\1', download_webpage,
+                'blob', group='blob'),
+            video_id, transform_source=unescapeHTML)
+
+        info = blob['digital_items'][0]
+
+        downloads = info['downloads']
+        track = info['title']
+
+        artist = info.get('artist')
+        title = '%s - %s' % (artist, track) if artist else track
+
+        download_formats = {}
+        for f in blob['download_formats']:
+            name, ext = f.get('name'), f.get('file_extension')
+            if all(isinstance(x, compat_str) for x in (name, ext)):
+                download_formats[name] = ext.strip('.')
+
+        formats = []
+        for format_id, f in downloads.items():
+            format_url = f.get('url')
+            if not format_url:
+                continue
+            # Stat URL generation algorithm is reverse engineered from
+            # download_*_bundle_*.js
+            stat_url = update_url_query(
+                format_url.replace('/download/', '/statdownload/'), {
+                    '.rand': int(time.time() * 1000 * random.random()),
+                })
+            format_id = f.get('encoding_name') or format_id
+            stat = self._download_json(
+                stat_url, video_id, 'Downloading %s JSON' % format_id,
+                transform_source=lambda s: s[s.index('{'):s.rindex('}') + 1],
+                fatal=False)
+            if not stat:
+                continue
+            retry_url = stat.get('retry_url')
+            if not isinstance(retry_url, compat_str):
+                continue
+            formats.append({
+                'url': self._proto_relative_url(retry_url, 'http:'),
+                'ext': download_formats.get(format_id),
+                'format_id': format_id,
+                'format_note': f.get('description'),
+                'filesize': parse_filesize(f.get('size_mb')),
+                'vcodec': 'none',
+            })
+        self._sort_formats(formats)

 
         return {
             'id': video_id,
 
         return {
             'id': video_id,

-            'title': info['title'],
-            'ext': 'mp3',
-            'vcodec': 'none',
-            'url': final_url,
-            'thumbnail': info.get('thumb_url'),
+            'title': title,
+            'thumbnail': info.get('thumb_url') or thumbnail,

             'uploader': info.get('artist'),
             'uploader': info.get('artist'),

+            'artist': artist,
+            'track': track,
+            'formats': formats,

         }
 
 
 class BandcampAlbumIE(InfoExtractor):
     IE_NAME = 'Bandcamp:album'
         }
 
 
 class BandcampAlbumIE(InfoExtractor):
     IE_NAME = 'Bandcamp:album'

-    _VALID_URL = r'https?://(?:(?P<subdomain>[^.]+)\.)?bandcamp\.com(?:/album/(?P<album_id>[^?#]+)|/?(?:$|[?#]))'
+    _VALID_URL = r'https?://(?:(?P<subdomain>[^.]+)\.)?bandcamp\.com(?:/album/(?P<album_id>[^/?#&]+))?'

 
     _TESTS = [{
         'url': 'http://blazo.bandcamp.com/album/jazz-format-mixtape-vol-1',
 
     _TESTS = [{
         'url': 'http://blazo.bandcamp.com/album/jazz-format-mixtape-vol-1',


@@ -156,22 +204,58 @@ class BandcampAlbumIE(InfoExtractor):
             'uploader_id': 'dotscale',
         },
         'playlist_mincount': 7,
             'uploader_id': 'dotscale',
         },
         'playlist_mincount': 7,

+    }, {
+        # with escaped quote in title
+        'url': 'https://jstrecords.bandcamp.com/album/entropy-ep',
+        'info_dict': {
+            'title': '"Entropy" EP',
+            'uploader_id': 'jstrecords',
+            'id': 'entropy-ep',
+        },
+        'playlist_mincount': 3,
+    }, {
+        # not all tracks have songs
+        'url': 'https://insulters.bandcamp.com/album/we-are-the-plague',
+        'info_dict': {
+            'id': 'we-are-the-plague',
+            'title': 'WE ARE THE PLAGUE',
+            'uploader_id': 'insulters',
+        },
+        'playlist_count': 2,

     }]
 
     }]
 

+    @classmethod
+    def suitable(cls, url):
+        return (False
+                if BandcampWeeklyIE.suitable(url) or BandcampIE.suitable(url)
+                else super(BandcampAlbumIE, cls).suitable(url))
+

     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         uploader_id = mobj.group('subdomain')
         album_id = mobj.group('album_id')
         playlist_id = album_id or uploader_id
         webpage = self._download_webpage(url, playlist_id)
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         uploader_id = mobj.group('subdomain')
         album_id = mobj.group('album_id')
         playlist_id = album_id or uploader_id
         webpage = self._download_webpage(url, playlist_id)

-        tracks_paths = re.findall(r'<a href="(.*?)" itemprop="url">', webpage)
-        if not tracks_paths:
+        track_elements = re.findall(
+            r'(?s)<div[^>]*>(.*?<a[^>]+href="([^"]+?)"[^>]+itemprop="url"[^>]*>.*?)</div>', webpage)
+        if not track_elements:

             raise ExtractorError('The page doesn\'t contain any tracks')
             raise ExtractorError('The page doesn\'t contain any tracks')

+        # Only tracks with duration info have songs

         entries = [
         entries = [

-            self.url_result(compat_urlparse.urljoin(url, t_path), ie=BandcampIE.ie_key())
-            for t_path in tracks_paths]
-        title = self._search_regex(
-            r'album_title\s*:\s*"(.*?)"', webpage, 'title', fatal=False)
+            self.url_result(
+                compat_urlparse.urljoin(url, t_path),
+                ie=BandcampIE.ie_key(),
+                video_title=self._search_regex(
+                    r'<span\b[^>]+\bitemprop=["\']name["\'][^>]*>([^<]+)',
+                    elem_content, 'track title', fatal=False))
+            for elem_content, t_path in track_elements
+            if self._html_search_meta('duration', elem_content, default=None)]
+
+        title = self._html_search_regex(
+            r'album_title\s*:\s*"((?:\\.|[^"\\])+?)"',
+            webpage, 'title', fatal=False)
+        if title:
+            title = title.replace(r'\"', '"')

         return {
             '_type': 'playlist',
             'uploader_id': uploader_id,
         return {
             '_type': 'playlist',
             'uploader_id': uploader_id,


@@ -179,3 +263,92 @@ class BandcampAlbumIE(InfoExtractor):
             'title': title,
             'entries': entries,
         }
             'title': title,
             'entries': entries,
         }

+
+
+class BandcampWeeklyIE(InfoExtractor):
+    IE_NAME = 'Bandcamp:weekly'
+    _VALID_URL = r'https?://(?:www\.)?bandcamp\.com/?\?(?:.*?&)?show=(?P<id>\d+)'
+    _TESTS = [{
+        'url': 'https://bandcamp.com/?show=224',
+        'md5': 'b00df799c733cf7e0c567ed187dea0fd',
+        'info_dict': {
+            'id': '224',
+            'ext': 'opus',
+            'title': 'BC Weekly April 4th 2017 - Magic Moments',
+            'description': 'md5:5d48150916e8e02d030623a48512c874',
+            'duration': 5829.77,
+            'release_date': '20170404',
+            'series': 'Bandcamp Weekly',
+            'episode': 'Magic Moments',
+            'episode_number': 208,
+            'episode_id': '224',
+        }
+    }, {
+        'url': 'https://bandcamp.com/?blah/blah@&show=228',
+        'only_matching': True
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+
+        blob = self._parse_json(
+            self._search_regex(
+                r'data-blob=(["\'])(?P<blob>{.+?})\1', webpage,
+                'blob', group='blob'),
+            video_id, transform_source=unescapeHTML)
+
+        show = blob['bcw_show']
+
+        # This is desired because any invalid show id redirects to `bandcamp.com`
+        # which happens to expose the latest Bandcamp Weekly episode.
+        show_id = int_or_none(show.get('show_id')) or int_or_none(video_id)
+
+        formats = []
+        for format_id, format_url in show['audio_stream'].items():
+            if not isinstance(format_url, compat_str):
+                continue
+            for known_ext in KNOWN_EXTENSIONS:
+                if known_ext in format_id:
+                    ext = known_ext
+                    break
+            else:
+                ext = None
+            formats.append({
+                'format_id': format_id,
+                'url': format_url,
+                'ext': ext,
+                'vcodec': 'none',
+            })
+        self._sort_formats(formats)
+
+        title = show.get('audio_title') or 'Bandcamp Weekly'
+        subtitle = show.get('subtitle')
+        if subtitle:
+            title += ' - %s' % subtitle
+
+        episode_number = None
+        seq = blob.get('bcw_seq')
+
+        if seq and isinstance(seq, list):
+            try:
+                episode_number = next(
+                    int_or_none(e.get('episode_number'))
+                    for e in seq
+                    if isinstance(e, dict) and int_or_none(e.get('id')) == show_id)
+            except StopIteration:
+                pass
+
+        return {
+            'id': video_id,
+            'title': title,
+            'description': show.get('desc') or show.get('short_desc'),
+            'duration': float_or_none(show.get('audio_duration')),
+            'is_live': False,
+            'release_date': unified_strdate(show.get('published_date')),
+            'series': 'Bandcamp Weekly',
+            'episode': show.get('subtitle'),
+            'episode_number': episode_number,
+            'episode_id': compat_str(video_id),
+            'formats': formats
+        }