From 3eb38acb43f30c94a9b93894b1d8171f3e6fa809 Mon Sep 17 00:00:00 2001
From: David Triendl <david@triendl.name>
Date: Fri, 21 Feb 2014 17:28:30 +0100
Subject: [PATCH] [BR] Add "BR" extractor

Extractor for videos from the Bayerischer Rundfunk Mediathek[1]. Currently only
supports videos. Audio and podcasts do not work yet with this extractor.

1: http://br.de/mediathek
---
 youtube_dl/extractor/__init__.py |  1 +
 youtube_dl/extractor/br.py       | 70 ++++++++++++++++++++++++++++++++
 2 files changed, 71 insertions(+)
 create mode 100644 youtube_dl/extractor/br.py

diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py
index e35287f88..989482a9b 100644
--- a/youtube_dl/extractor/__init__.py
+++ b/youtube_dl/extractor/__init__.py
@@ -19,6 +19,7 @@ from .bbccouk import BBCCoUkIE
 from .blinkx import BlinkxIE
 from .bliptv import BlipTVIE, BlipTVUserIE
 from .bloomberg import BloombergIE
+from .br import BRIE
 from .breakcom import BreakIE
 from .brightcove import BrightcoveIE
 from .c56 import C56IE
diff --git a/youtube_dl/extractor/br.py b/youtube_dl/extractor/br.py
new file mode 100644
index 000000000..888c5c239
--- /dev/null
+++ b/youtube_dl/extractor/br.py
@@ -0,0 +1,70 @@
+# coding: utf-8
+
+from .common import InfoExtractor
+
+class BRIE(InfoExtractor):
+
+    IE_DESC = u"Bayerischer Rundfunk Mediathek"
+    _VALID_URL = r"^https?://(?:www\.)?br\.de/mediathek/video/(?:sendungen/)?(?:[a-z0-9\-]+\.html)$"
+    _BASE_URL = u"http://www.br.de"
+
+    _TESTS = []
+
+    def _real_extract(self, url):
+        page = self._download_webpage(url, None)
+        xml_url = self._search_regex(r"return BRavFramework\.register\(BRavFramework\('avPlayer_(?:[a-f0-9-]{36})'\)\.setup\({dataURL:'(/mediathek/video/[a-z0-9/~_.-]+)'}\)\);", page, "XMLURL")
+        xml = self._download_xml(self._BASE_URL + xml_url, None)
+
+        videos = []
+        for xml_video in xml.findall("video"):
+            video = {}
+            video["id"] = xml_video.get("externalId")
+            video["title"] = xml_video.find("title").text
+            video["formats"] = self._extract_formats(xml_video.find("assets"))
+            video["thumbnails"] = self._extract_thumbnails(xml_video.find("teaserImage/variants"))
+            video["thumbnail"] = video["thumbnails"][0]["url"]
+            video["description"] = " ".join(xml_video.find("shareTitle").text.splitlines())
+            video["uploader"] = xml_video.find("author").text
+            video["upload_date"] = "".join(reversed(xml_video.find("broadcastDate").text.split(".")))
+            video["webpage_url"] = xml_video.find("permalink").text
+            videos.append(video)
+
+        if len(videos) > 1:
+            self._downloader.report_warning(u'found multiple videos; please'
+                u'report this with the video URL to http://yt-dl.org/bug')
+        return videos[0]
+
+    def _extract_formats(self, assets):
+        vformats = []
+        for asset in assets.findall("asset"):
+            if asset.find("downloadUrl") is None:
+                continue
+            vformat = {}
+            vformat["url"] = asset.find("downloadUrl").text
+            vformat["ext"] = asset.find("mediaType").text
+            vformat["format_id"] = asset.get("type")
+            vformat["width"] = int(asset.find("frameWidth").text)
+            vformat["height"] = int(asset.find("frameHeight").text)
+            vformat["resolution"] = "%ix%i" % (vformat["width"], vformat["height"])
+            vformat["tbr"] = int(asset.find("bitrateVideo").text)
+            vformat["abr"] = int(asset.find("bitrateAudio").text)
+            vformat["vcodec"] = asset.find("codecVideo").text
+            vformat["container"] = vformat["ext"]
+            vformat["filesize"] = int(asset.find("size").text)
+            vformat["preference"] = vformat["quality"] = -1
+            vformat["format"] = "%s container with %i Kbps %s" % (vformat["container"], vformat["tbr"], vformat["vcodec"])
+            vformats.append(vformat)
+        self._sort_formats(vformats)
+        return vformats
+
+    def _extract_thumbnails(self, variants):
+        thumbnails = []
+        for variant in variants.findall("variant"):
+            thumbnail = {}
+            thumbnail["url"] = self._BASE_URL + variant.find("url").text
+            thumbnail["width"] = int(variant.find("width").text)
+            thumbnail["height"] = int(variant.find("height").text)
+            thumbnail["resolution"] = "%ix%i" % (thumbnail["width"], thumbnail["height"])
+            thumbnails.append(thumbnail)
+        thumbnails.sort(key = lambda x: x["width"] * x["height"], reverse=True)
+        return thumbnails