[3sat] Add support (Fixes #1001)

diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py
index 1a5f68e1555faa820061173ea7157b965c0a5350..e46698736e840b2a0874cf0336a6f855aebea31f 100644 (file)
--- a/youtube_dl/extractor/__init__.py
+++ b/youtube_dl/extractor/__init__.py
@@ -11,6 +11,7 @@ from .comedycentral import ComedyCentralIE
 from .cspan import CSpanIE
 from .dailymotion import DailymotionIE
 from .depositfiles import DepositFilesIE
+from .dreisat import DreiSatIE
 from .eighttracks import EightTracksIE
 from .escapist import EscapistIE
 from .facebook import FacebookIE

diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py
index 655836ff6efca2ef95e80631d528901d6ae28d46..8db72ba7afb480cbd9a894be78c0e1774556af52 100644 (file)
--- a/youtube_dl/extractor/common.py
+++ b/youtube_dl/extractor/common.py
@@ -36,6 +36,8 @@ class InfoExtractor(object):
     The following fields are optional:
 
     format:         The video format, defaults to ext (used for --get-format)
+    thumbnails:     A list of dictionaries (with the entries "resolution" and
+                    "url") for the varying thumbnails
     thumbnail:      Full URL to a video thumbnail image.
     description:    One-line video description.
     uploader:       Full name of the video uploader.

diff --git a/youtube_dl/extractor/dreisat.py b/youtube_dl/extractor/dreisat.py
new file mode 100644 (file)
index 0000000..56227c7
--- /dev/null
+++ b/youtube_dl/extractor/dreisat.py
@@ -0,0 +1,85 @@
+# coding: utf-8
+
+import re
+import xml.etree.ElementTree
+
+from .common import InfoExtractor
+from ..utils import (
+    determine_ext,
+    ExtractorError,
+    unified_strdate,
+)
+
+
+class DreiSatIE(InfoExtractor):
+    IE_NAME = '3sat'
+    _VALID_URL = r'(?:http://)?(?:www\.)?3sat.de/mediathek/index.php\?(?:(?:mode|display)=[^&]+&)*obj=(?P<id>[0-9]+)$'
+    _TEST = {
+        u"url": u"http://www.3sat.de/mediathek/index.php?obj=36983",
+        u'file': u'36983.webm',
+        u'md5': u'57c97d0469d71cf874f6815aa2b7c944',
+        u'info_dict': {
+            u"title": u"Kaffeeland Schweiz",
+            u"description": u"Über 80 Kaffeeröstereien liefern in der Schweiz das Getränk, in das das Land so vernarrt ist: Mehr als 1000 Tassen trinkt ein Schweizer pro Jahr. SCHWEIZWEIT nimmt die Kaffeekultur unter die...", 
+            u"uploader": u"3sat",
+            u"upload_date": u"20130622"
+        }
+    }
+
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        video_id = mobj.group('id')
+        details_url = 'http://www.3sat.de/mediathek/xmlservice/web/beitragsDetails?ak=web&id=%s' % video_id
+        details_xml = self._download_webpage(details_url, video_id, note=u'Downloading video details')
+        details_doc = xml.etree.ElementTree.fromstring(details_xml.encode('utf-8'))
+
+        thumbnail_els = details_doc.findall('.//teaserimage')
+        thumbnails = [{
+            'width': te.attrib['key'].partition('x')[0],
+            'height': te.attrib['key'].partition('x')[2],
+            'url': te.text,
+        } for te in thumbnail_els]
+
+        information_el = details_doc.find('.//information')
+        video_title = information_el.find('./title').text
+        video_description = information_el.find('./detail').text
+
+        details_el = details_doc.find('.//details')
+        video_uploader = details_el.find('./channel').text
+        upload_date = unified_strdate(details_el.find('./airtime').text)
+
+        format_els = details_doc.findall('.//formitaet')
+        formats = [{
+            'format_id': fe.attrib['basetype'],
+            'width': int(fe.find('./width').text),
+            'height': int(fe.find('./height').text),
+            'url': fe.find('./url').text,
+            'filesize': int(fe.find('./filesize').text),
+            'video_bitrate': int(fe.find('./videoBitrate').text),
+            '3sat_qualityname': fe.find('./quality').text,
+        } for fe in format_els
+            if not fe.find('./url').text.startswith('http://www.metafilegenerator.de/')]
+
+        def _sortkey(format):
+            qidx = ['low', 'med', 'high', 'veryhigh'].index(format['3sat_qualityname'])
+            prefer_http = 1 if 'rtmp' in format['url'] else 0
+            return (qidx, prefer_http, format['video_bitrate'])
+        formats.sort(key=_sortkey)
+
+        info = {
+            'id': video_id,
+            'title': video_title,
+            'formats': formats,
+            'description': video_description,
+            'thumbnails': thumbnails,
+            'thumbnail': thumbnails[-1]['url'],
+            'uploader': video_uploader,
+            'upload_date': upload_date,
+        }
+
+        # TODO: Remove when #980 has been merged
+        info['url'] = formats[-1]['url']
+        info['ext'] = determine_ext(formats[-1]['url'])
+
+        return self.video_result(info)
\ No newline at end of file

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 9137a4f70be1423c20b9c2107f15db2c91abdada..b9bff5fde87d91a5956e978c98880f05034ac6ab 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -623,7 +623,7 @@ def unified_strdate(date_str):
     date_str = date_str.replace(',',' ')
     # %z (UTC offset) is only supported in python>=3.2
     date_str = re.sub(r' (\+|-)[\d]*$', '', date_str)
-    format_expressions = ['%d %B %Y', '%B %d %Y', '%b %d %Y', '%Y-%m-%d', '%d/%m/%Y', '%Y/%m/%d %H:%M:%S']
+    format_expressions = ['%d %B %Y', '%B %d %Y', '%b %d %Y', '%Y-%m-%d', '%d/%m/%Y', '%Y/%m/%d %H:%M:%S', '%d.%m.%Y %H:%M']
     for expression in format_expressions:
         try:
             upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
@@ -631,6 +631,13 @@ def unified_strdate(date_str):
             pass
     return upload_date
 
+def determine_ext(url):
+    guess = url.partition(u'?')[0].rpartition(u'.')[2]
+    if re.match(r'^[A-Za-z0-9]+$', guess):
+        return guess
+    else:
+        return u'unknown_video'
+
 def date_from_str(date_str):
     """
     Return a datetime object from a string in the format YYYYMMDD or