[npo] Improve extraction and update tests

youtube_dl/extractor/npo.py

@@ -3,15 +3,19 @@
 import re
 
 from .common import InfoExtractor
-from ..compat import compat_HTTPError
+from ..compat import (
+    compat_HTTPError,
+    compat_str,
+)
 from ..utils import (
+    determine_ext,
+    ExtractorError,
     fix_xml_ampersands,
     orderedSet,
     parse_duration,
     qualities,
     strip_jsonp,
     unified_strdate,
-    ExtractorError,
 )
 
 
@@ -40,8 +44,7 @@ class NPOIE(NPOBaseIE):
                         (?P<id>[^/?#]+)
                 '''
 
-    _TESTS = [
+    _TESTS = [{
-        {
         'url': 'http://www.npo.nl/nieuwsuur/22-06-2014/VPWON_1220719',
         'md5': '4b3f9c429157ec4775f2c9cb7b911016',
         'info_dict': {
@@ -51,8 +54,7 @@ class NPOIE(NPOBaseIE):
             'description': 'Dagelijks tussen tien en elf: nieuws, sport en achtergronden.',
             'upload_date': '20140622',
         },
-        },
+    }, {
-        {
         'url': 'http://www.npo.nl/de-mega-mike-mega-thomas-show/27-02-2009/VARA_101191800',
         'md5': 'da50a5787dbfc1603c4ad80f31c5120b',
         'info_dict': {
@@ -63,46 +65,42 @@ class NPOIE(NPOBaseIE):
             'upload_date': '20090227',
             'duration': 2400,
         },
-        },
+    }, {
-        {
         'url': 'http://www.npo.nl/tegenlicht/25-02-2013/VPWON_1169289',
         'md5': 'f8065e4e5a7824068ed3c7e783178f2c',
         'info_dict': {
             'id': 'VPWON_1169289',
             'ext': 'm4v',
-                'title': 'Tegenlicht: De toekomst komt uit Afrika',
+            'title': 'Tegenlicht: Zwart geld. De toekomst komt uit Afrika',
             'description': 'md5:52cf4eefbc96fffcbdc06d024147abea',
             'upload_date': '20130225',
             'duration': 3000,
         },
-        },
+    }, {
-        {
         'url': 'http://www.npo.nl/de-nieuwe-mens-deel-1/21-07-2010/WO_VPRO_043706',
         'info_dict': {
             'id': 'WO_VPRO_043706',
-                'ext': 'wmv',
+            'ext': 'm4v',
             'title': 'De nieuwe mens - Deel 1',
             'description': 'md5:518ae51ba1293ffb80d8d8ce90b74e4b',
             'duration': 4680,
         },
         'params': {
-                # mplayer mms download
             'skip_download': True,
         }
-        },
+    }, {
         # non asf in streams
-        {
         'url': 'http://www.npo.nl/hoe-gaat-europa-verder-na-parijs/10-01-2015/WO_NOS_762771',
-            'md5': 'b3da13de374cbe2d5332a7e910bef97f',
         'info_dict': {
             'id': 'WO_NOS_762771',
             'ext': 'mp4',
             'title': 'Hoe gaat Europa verder na Parijs?',
         },
-        },
+        'params': {
-        {
+            'skip_download': True,
+        }
+    }, {
         'url': 'http://www.ntr.nl/Aap-Poot-Pies/27/detail/Aap-poot-pies/VPWON_1233944#content',
-            'md5': '01c6a2841675995da1f0cf776f03a9c3',
         'info_dict': {
             'id': 'VPWON_1233944',
             'ext': 'm4v',
@@ -111,10 +109,11 @@ class NPOIE(NPOBaseIE):
             'upload_date': '20150508',
             'duration': 599,
         },
-        },
+        'params': {
-        {
+            'skip_download': True,
+        }
+    }, {
         'url': 'http://www.omroepwnl.nl/video/fragment/vandaag-de-dag-verkiezingen__POMS_WNL_853698',
-            'md5': 'd30cd8417b8b9bca1fdff27428860d08',
         'info_dict': {
             'id': 'POW_00996502',
             'ext': 'm4v',
@@ -123,20 +122,23 @@ class NPOIE(NPOBaseIE):
             'upload_date': '20150508',
             'duration': 462,
         },
-        },
+        'params': {
-        {
+            'skip_download': True,
+        }
+    }, {
         'url': 'http://www.zapp.nl/de-bzt-show/gemist/KN_1687547',
         'only_matching': True,
-        },
+    }, {
-        {
         'url': 'http://www.zapp.nl/de-bzt-show/filmpjes/POMS_KN_7315118',
         'only_matching': True,
-        },
+    }, {
-        {
         'url': 'http://www.zapp.nl/beste-vrienden-quiz/extra-video-s/WO_NTR_1067990',
         'only_matching': True,
-        }
+    }, {
-    ]
+        # live stream
+        'url': 'npo:LI_NL1_4188102',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
@@ -165,27 +167,41 @@ def _get_info(self, video_id):
         token = self._get_token(video_id)
 
         formats = []
+        urls = set()
 
-        pubopties = metadata.get('pubopties')
-        if pubopties:
         quality = qualities(['adaptive', 'wmv_sb', 'h264_sb', 'wmv_bb', 'h264_bb', 'wvc1_std', 'h264_std'])
         items = self._download_json(
-                'http://ida.omroep.nl/app.php/%s' % video_id,
+            'http://ida.omroep.nl/app.php/%s' % video_id, video_id,
             'Downloading formats JSON', query={
                 'adaptive': 'yes',
                 'token': token,
             })['items'][0]
         for num, item in enumerate(items):
             item_url = item.get('url')
-                if not item_url:
+            if not item_url or item_url in urls:
                 continue
+            urls.add(item_url)
             format_id = self._search_regex(
                 r'video/ida/([^/]+)', item_url, 'format id',
                 default=None)
+
+            def add_format_url(format_url):
+                formats.append({
+                    'url': format_url,
+                    'format_id': format_id,
+                    'quality': quality(format_id),
+                })
+
+            # Example: http://www.npo.nl/de-nieuwe-mens-deel-1/21-07-2010/WO_VPRO_043706
+            if item.get('contentType') == 'url':
+                add_format_url(item_url)
+                continue
+
             try:
                 stream_info = self._download_json(
                     item_url + '&type=json', video_id,
-                        'Downloading %s stream JSON' % item.get('label') or format_id or num)
+                    'Downloading %s stream JSON'
+                    % item.get('label') or format_id or num)
             except ExtractorError as ee:
                 if isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404:
                     error = (self._parse_json(
@@ -194,49 +210,71 @@ def _get_info(self, video_id):
                     if error:
                         raise ExtractorError(error, expected=True)
                 raise
-                if stream_info.get('error_code', 0) or stream_info.get('errorcode', 0):
+            # Stream URL instead of JSON, example: npo:LI_NL1_4188102
+            if isinstance(stream_info, compat_str):
+                if not stream_info.startswith('http'):
                     continue
+                video_url = stream_info
+            # JSON
+            else:
                 video_url = stream_info.get('url')
-                if not video_url:
+            if not video_url or video_url in urls:
                 continue
-                if stream_info.get('family') == 'adaptive':
+            urls.add(item_url)
+            if determine_ext(video_url) == 'm3u8':
                 formats.extend(self._extract_m3u8_formats(
                     video_url, video_id, ext='mp4',
-                        entry_protocol='m3u8_native', m3u8_id='hls',
+                    entry_protocol='m3u8_native', m3u8_id='hls', fatal=False))
-                        fatal=False))
             else:
-                    formats.append({
+                add_format_url(video_url)
-                        'url': video_url,
-                        'format_id': format_id,
-                        'quality': quality(format_id),
-                    })
 
-        streams = metadata.get('streams')
+        is_live = metadata.get('medium') == 'live'
-        if streams:
+
-            for i, stream in enumerate(streams):
+        if not is_live:
+            for num, stream in enumerate(metadata.get('streams', [])):
                 stream_url = stream.get('url')
-                if not stream_url:
+                if not stream_url or stream_url in urls:
                     continue
-                if '.asf' not in stream_url:
+                urls.add(stream_url)
-                    formats.append({
+                # smooth streaming is not supported
-                        'url': stream_url,
+                stream_type = stream.get('type', '').lower()
-                        'quality': stream.get('kwaliteit'),
+                if stream_type in ['ss', 'ms']:
-                    })
                     continue
+                if stream_type == 'hds':
+                    f4m_formats = self._extract_f4m_formats(
+                        stream_url, video_id, fatal=False)
+                    # f4m downloader downloads only piece of live stream
+                    for f4m_format in f4m_formats:
+                        f4m_format['preference'] = -1
+                    formats.extend(f4m_formats)
+                elif stream_type == 'hls':
+                    formats.extend(self._extract_m3u8_formats(
+                        stream_url, video_id, ext='mp4', fatal=False))
+                # Example: http://www.npo.nl/de-nieuwe-mens-deel-1/21-07-2010/WO_VPRO_043706
+                elif '.asf' in stream_url:
                     asx = self._download_xml(
                         stream_url, video_id,
-                    'Downloading stream %d ASX playlist' % i,
+                        'Downloading stream %d ASX playlist' % num,
-                    transform_source=fix_xml_ampersands)
+                        transform_source=fix_xml_ampersands, fatal=False)
+                    if not asx:
+                        continue
                     ref = asx.find('./ENTRY/Ref')
                     if ref is None:
                         continue
                     video_url = ref.get('href')
-                if not video_url:
+                    if not video_url or video_url in urls:
                         continue
+                    urls.add(video_url)
                     formats.append({
                         'url': video_url,
                         'ext': stream.get('formaat', 'asf'),
                         'quality': stream.get('kwaliteit'),
+                        'preference': -10,
+                    })
+                else:
+                    formats.append({
+                        'url': stream_url,
+                        'quality': stream.get('kwaliteit'),
                     })
 
         self._sort_formats(formats)
@@ -250,28 +288,28 @@ def _get_info(self, video_id):
 
         return {
             'id': video_id,
-            'title': title,
+            'title': self._live_title(title) if is_live else title,
             'description': metadata.get('info'),
             'thumbnail': metadata.get('images', [{'url': None}])[-1]['url'],
             'upload_date': unified_strdate(metadata.get('gidsdatum')),
             'duration': parse_duration(metadata.get('tijdsduur')),
             'formats': formats,
             'subtitles': subtitles,
+            'is_live': is_live,
         }
 
 
 class NPOLiveIE(NPOBaseIE):
     IE_NAME = 'npo.nl:live'
-    _VALID_URL = r'https?://(?:www\.)?npo\.nl/live/(?P<id>.+)'
+    _VALID_URL = r'https?://(?:www\.)?npo\.nl/live/(?P<id>[^/?#&]+)'
 
     _TEST = {
         'url': 'http://www.npo.nl/live/npo-1',
         'info_dict': {
-            'id': 'LI_NEDERLAND1_136692',
+            'id': 'LI_NL1_4188102',
             'display_id': 'npo-1',
             'ext': 'mp4',
-            'title': 're:^Nederland 1 [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
+            'title': 're:^NPO 1 [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
-            'description': 'Livestream',
             'is_live': True,
         },
         'params': {
@@ -287,58 +325,12 @@ def _real_extract(self, url):
         live_id = self._search_regex(
             r'data-prid="([^"]+)"', webpage, 'live id')
 
-        metadata = self._download_json(
-            'http://e.omroep.nl/metadata/%s' % live_id,
-            display_id, transform_source=strip_jsonp)
-
-        token = self._get_token(display_id)
-
-        formats = []
-
-        streams = metadata.get('streams')
-        if streams:
-            for stream in streams:
-                stream_type = stream.get('type').lower()
-                # smooth streaming is not supported
-                if stream_type in ['ss', 'ms']:
-                    continue
-                stream_info = self._download_json(
-                    'http://ida.omroep.nl/aapi/?stream=%s&token=%s&type=jsonp'
-                    % (stream.get('url'), token),
-                    display_id, 'Downloading %s JSON' % stream_type)
-                if stream_info.get('error_code', 0) or stream_info.get('errorcode', 0):
-                    continue
-                stream_url = self._download_json(
-                    stream_info['stream'], display_id,
-                    'Downloading %s URL' % stream_type,
-                    'Unable to download %s URL' % stream_type,
-                    transform_source=strip_jsonp, fatal=False)
-                if not stream_url:
-                    continue
-                if stream_type == 'hds':
-                    f4m_formats = self._extract_f4m_formats(stream_url, display_id)
-                    # f4m downloader downloads only piece of live stream
-                    for f4m_format in f4m_formats:
-                        f4m_format['preference'] = -1
-                    formats.extend(f4m_formats)
-                elif stream_type == 'hls':
-                    formats.extend(self._extract_m3u8_formats(stream_url, display_id, 'mp4'))
-                else:
-                    formats.append({
-                        'url': stream_url,
-                        'preference': -10,
-                    })
-
-        self._sort_formats(formats)
-
         return {
+            '_type': 'url_transparent',
+            'url': 'npo:%s' % live_id,
+            'ie_key': NPOIE.ie_key(),
             'id': live_id,
             'display_id': display_id,
-            'title': self._live_title(metadata['titel']),
-            'description': metadata['info'],
-            'thumbnail': metadata.get('images', [{'url': None}])[-1]['url'],
-            'formats': formats,
-            'is_live': True,
         }