[zdf/common] Use API in ZDF extractor.

This also comes with a lot of extra format fields
Fixes #1518

youtube_dl/FileDownloader.py

@@ -1,4 +1,3 @@
-import math
 import os
 import re
 import subprocess
@@ -11,6 +10,7 @@
     ContentTooShortError,
     determine_ext,
     encodeFilename,
+    format_bytes,
     sanitize_open,
     timeconvert,
 )
@@ -53,20 +53,6 @@ def __init__(self, ydl, params):
         self._progress_hooks = []
         self.params = params
 
-    @staticmethod
-    def format_bytes(bytes):
-        if bytes is None:
-            return 'N/A'
-        if type(bytes) is str:
-            bytes = float(bytes)
-        if bytes == 0.0:
-            exponent = 0
-        else:
-            exponent = int(math.log(bytes, 1024.0))
-        suffix = ['B','KiB','MiB','GiB','TiB','PiB','EiB','ZiB','YiB'][exponent]
-        converted = float(bytes) / float(1024 ** exponent)
-        return '%.2f%s' % (converted, suffix)
-
     @staticmethod
     def format_seconds(seconds):
         (mins, secs) = divmod(seconds, 60)
@@ -117,7 +103,7 @@ def calc_speed(start, now, bytes):
     def format_speed(speed):
         if speed is None:
             return '%10s' % '---b/s'
-        return '%10s' % ('%s/s' % FileDownloader.format_bytes(speed))
+        return '%10s' % ('%s/s' % format_bytes(speed))
 
     @staticmethod
     def best_block_size(elapsed_time, bytes):
@@ -525,7 +511,7 @@ def _do_download(self, filename, info_dict):
                 self.to_screen(u'\r[download] File is larger than max-filesize (%s bytes > %s bytes). Aborting.' % (data_len, max_data_len))
                 return False
 
-        data_len_str = self.format_bytes(data_len)
+        data_len_str = format_bytes(data_len)
         byte_counter = 0 + resume_len
         block_size = self.params.get('buffersize', 1024)
         start = time.time()

youtube_dl/YoutubeDL.py

@@ -30,6 +30,7 @@
     DownloadError,
     encodeFilename,
     ExtractorError,
+    format_bytes,
     locked_file,
     MaxDownloadsReached,
     PostProcessingError,
@@ -867,9 +868,11 @@ def format_resolution(format, default='unknown'):
 
     def list_formats(self, info_dict):
         def format_note(fdict):
-            if fdict.get('format_note') is not None:
-                return fdict['format_note']
             res = u''
+            if fdict.get('format_note') is not None:
+                res += fdict['format_note'] + u' '
+            if fdict.get('quality_name') is not None:
+                res += u'%s ' % fdict['quality_name']
             if fdict.get('vcodec') is not None:
                 res += u'%-5s' % fdict['vcodec']
             elif fdict.get('vbr') is not None:
@@ -886,25 +889,30 @@ def format_note(fdict):
                 res += 'audio'
             if fdict.get('abr') is not None:
                 res += u'@%3dk' % fdict['abr']
+            if fdict.get('filesize') is not None:
+                if res:
+                    res += u', '
+                res += format_bytes(fdict['filesize'])
             return res
 
-        def line(format):
-            return (u'%-20s%-10s%-12s%s' % (
+        def line(format, idlen=20):
+            return ((u'%-' + compat_str(idlen + 1) + u's%-10s%-12s%s') % (
                 format['format_id'],
                 format['ext'],
                 self.format_resolution(format),
                 format_note(format),
-                )
-            )
+            ))
 
         formats = info_dict.get('formats', [info_dict])
-        formats_s = list(map(line, formats))
+        idlen = max(len(u'format code'),
+                    max(len(f['format_id']) for f in formats))
+        formats_s = [line(f, idlen) for f in formats]
         if len(formats) > 1:
             formats_s[0] += (' ' if format_note(formats[0]) else '') + '(worst)'
             formats_s[-1] += (' ' if format_note(formats[-1]) else '') + '(best)'
 
         header_line = line({
             'format_id': u'format code', 'ext': u'extension',
-            '_resolution': u'resolution', 'format_note': u'note'})
+            '_resolution': u'resolution', 'format_note': u'note'}, idlen=idlen)
         self.to_screen(u'[info] Available formats for %s:\n%s\n%s' %
                        (info_dict['id'], header_line, u"\n".join(formats_s)))

youtube_dl/extractor/common.py

@@ -76,6 +76,8 @@ class InfoExtractor(object):
                     * acodec    Name of the audio codec in use
                     * vbr       Average video bitrate in KBit/s
                     * vcodec    Name of the video codec in use
+                    * quality_name Human-readable name of the video quality.
+                    * filesize  The number of bytes, if known in advance
     webpage_url:    The url to the video webpage, if given to youtube-dl it
                     should allow to get the same result again. (It will be set
                     by YoutubeDL if it's missing)

youtube_dl/extractor/zdf.py

@@ -1,75 +1,100 @@
+import operator
 import re
 
 from .common import InfoExtractor
 from ..utils import (
-    determine_ext,
-    ExtractorError,
+    parse_xml_doc,
+    unified_strdate,
 )
 
 
 class ZDFIE(InfoExtractor):
     _VALID_URL = r'^http://www\.zdf\.de\/ZDFmediathek(?P<hash>#)?\/(.*beitrag\/video\/)(?P<video_id>[^/\?]+)(?:\?.*)?'
-    _MEDIA_STREAM = r'<a href="(?P<video_url>.+(?P<media_type>.streaming).+/zdf/(?P<quality>[^\/]+)/[^"]*)".+class="play".+>'
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
-        if mobj is None:
-            raise ExtractorError(u'Invalid URL: %s' % url)
         video_id = mobj.group('video_id')
 
-        if mobj.group('hash'):
-            url = url.replace(u'#', u'', 1)
+        xml_url = u'http://www.zdf.de/ZDFmediathek/xmlservice/web/beitragsDetails?ak=web&id=%s' % video_id
+        info_xml = self._download_webpage(
+            xml_url, video_id, note=u'Downloading video info')
+        doc = parse_xml_doc(info_xml)
 
-        html = self._download_webpage(url, video_id)
-        streams = [m.groupdict() for m in re.finditer(self._MEDIA_STREAM, html)]
-        if streams is None:
-            raise ExtractorError(u'No media url found.')
+        title = doc.find('.//information/title').text
+        description = doc.find('.//information/detail').text
+        uploader_node = doc.find('.//details/originChannelTitle')
+        uploader = None if uploader_node is None else uploader_node.text
+        duration_str = doc.find('.//details/length').text
+        duration_m = re.match(r'''(?x)^
+            (?P<hours>[0-9]{2})
+            :(?P<minutes>[0-9]{2})
+            :(?P<seconds>[0-9]{2})
+            (?:\.(?P<ms>[0-9]+)?)
+            ''', duration_str)
+        duration = (
+            (
+                (int(duration_m.group('hours')) * 60 * 60) +
+                (int(duration_m.group('minutes')) * 60) +
+                int(duration_m.group('seconds'))
+            )
+            if duration_m
+            else None
+        )
+        upload_date = unified_strdate(doc.find('.//details/airtime').text)
 
-        # s['media_type'] == 'wstreaming' -> use 'Windows Media Player' and mms url
-        # s['media_type'] == 'hstreaming' -> use 'Quicktime' and rtsp url
-        # choose first/default media type and highest quality for now
-        def stream_pref(s):
-            TYPE_ORDER = ['ostreaming', 'hstreaming', 'wstreaming']
+        def xml_to_format(fnode):
+            video_url = fnode.find('url').text
+            is_available = u'http://www.metafilegenerator' not in video_url
+
+            format_id = fnode.attrib['basetype']
+            format_m = re.match(r'''(?x)
+                (?P<vcodec>[^_]+)_(?P<acodec>[^_]+)_(?P<container>[^_]+)_
+                (?P<proto>[^_]+)_(?P<index>[^_]+)_(?P<indexproto>[^_]+)
+            ''', format_id)
+
+            PROTO_ORDER = ['http', 'rtmp', 'rtsp']
             try:
-                type_pref = TYPE_ORDER.index(s['media_type'])
+                proto_pref = -PROTO_ORDER.index(format_m.group('proto'))
             except ValueError:
-                type_pref = 999
+                proto_pref = 999
 
-            QUALITY_ORDER = ['veryhigh', '300']
+            quality = fnode.find('./quality').text
+            QUALITY_ORDER = ['veryhigh', '300', 'high', 'med', 'low']
             try:
-                quality_pref = QUALITY_ORDER.index(s['quality'])
+                quality_pref = -QUALITY_ORDER.index(quality)
             except ValueError:
                 quality_pref = 999
 
-            return (type_pref, quality_pref)
+            abr = int(fnode.find('./audioBitrate').text) // 1000
+            vbr = int(fnode.find('./videoBitrate').text) // 1000
+            pref = (is_available, proto_pref, quality_pref, vbr, abr)
 
-        sorted_streams = sorted(streams, key=stream_pref)
-        if not sorted_streams:
-            raise ExtractorError(u'No stream found.')
-        stream = sorted_streams[0]
+            return {
+                'format_id': format_id,
+                'url': video_url,
+                'ext': format_m.group('container'),
+                'acodec': format_m.group('acodec'),
+                'vcodec': format_m.group('vcodec'),
+                'abr': abr,
+                'vbr': vbr,
+                'width': int(fnode.find('./width').text),
+                'height': int(fnode.find('./height').text),
+                'quality_name': quality,
+                'filesize': int(fnode.find('./filesize').text),
+                'format_note': None if is_available else u'(unavailable)',
+                '_pref': pref,
+            }
 
-        media_link = self._download_webpage(
-            stream['video_url'],
-            video_id,
-            u'Get stream URL')
-
-        #MMS_STREAM = r'href="(?P<video_url>mms://[^"]*)"'
-        RTSP_STREAM = r'(?P<video_url>rtsp://[^"]*.mp4)'
-
-        mobj = re.search(self._MEDIA_STREAM, media_link)
-        if mobj is None:
-            mobj = re.search(RTSP_STREAM, media_link)
-            if mobj is None:
-                raise ExtractorError(u'Cannot extract mms:// or rtsp:// URL')
-        video_url = mobj.group('video_url')
-
-        title = self._html_search_regex(
-            r'<h1(?: class="beitragHeadline")?>(.*?)</h1>',
-            html, u'title')
+        format_nodes = doc.findall('.//formitaeten/formitaet')
+        formats = sorted(map(xml_to_format, format_nodes),
+                         key=operator.itemgetter('_pref'))
 
         return {
             'id': video_id,
-            'url': video_url,
             'title': title,
-            'ext': determine_ext(video_url)
+            'formats': formats,
+            'description': description,
+            'uploader': uploader,
+            'duration': duration,
+            'upload_date': upload_date,
         }

youtube_dl/utils.py

@@ -8,6 +8,7 @@
 import io
 import json
 import locale
+import math
 import os
 import pipes
 import platform
@@ -16,6 +17,7 @@
 import socket
 import sys
 import traceback
+import xml.etree.ElementTree
 import zlib
 
 try:
@@ -1006,3 +1008,22 @@ def unsmuggle_url(smug_url):
     jsond = compat_parse_qs(sdata)[u'__youtubedl_smuggle'][0]
     data = json.loads(jsond)
     return url, data
+
+
+def parse_xml_doc(s):
+    assert isinstance(s, type(u''))
+    return xml.etree.ElementTree.fromstring(s.encode('utf-8'))
+
+
+def format_bytes(bytes):
+    if bytes is None:
+        return u'N/A'
+    if type(bytes) is str:
+        bytes = float(bytes)
+    if bytes == 0.0:
+        exponent = 0
+    else:
+        exponent = int(math.log(bytes, 1024.0))
+    suffix = [u'B', u'KiB', u'MiB', u'GiB', u'TiB', u'PiB', u'EiB', u'ZiB', u'YiB'][exponent]
+    converted = float(bytes) / float(1024 ** exponent)
+    return u'%.2f%s' % (converted, suffix)