From 632e5684ce797eb8a7372eb25dd4ce299f2e66de Mon Sep 17 00:00:00 2001
From: Naglis Jonaitis <njonaitis@gmail.com>
Date: Tue, 23 Sep 2014 00:28:19 +0300
Subject: [PATCH] [nfl] Add new extractor. (Closes #3815)

---
 youtube_dl/extractor/__init__.py |   1 +
 youtube_dl/extractor/nfl.py      | 103 +++++++++++++++++++++++++++++++
 2 files changed, 104 insertions(+)
 create mode 100644 youtube_dl/extractor/nfl.py

diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py
index 244d22297..1f1fc0eb2 100644
--- a/youtube_dl/extractor/__init__.py
+++ b/youtube_dl/extractor/__init__.py
@@ -240,6 +240,7 @@
 from .newgrounds import NewgroundsIE
 from .newstube import NewstubeIE
 from .nfb import NFBIE
+from .nfl import NFLIE
 from .nhl import NHLIE, NHLVideocenterIE
 from .niconico import NiconicoIE
 from .ninegag import NineGagIE
diff --git a/youtube_dl/extractor/nfl.py b/youtube_dl/extractor/nfl.py
new file mode 100644
index 000000000..f53596f5e
--- /dev/null
+++ b/youtube_dl/extractor/nfl.py
@@ -0,0 +1,103 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..utils import (
+    ExtractorError,
+    int_or_none,
+    remove_end,
+)
+
+
+class NFLIE(InfoExtractor):
+    IE_NAME = 'nfl.com'
+    _VALID_URL = r'(?x)https?://(?:www\.)?nfl\.com/(?:videos/(?:.+)/|.*?\#video=)(?P<id>\d..[0-9]+)'
+    _PLAYER_CONFIG_URL = 'http://www.nfl.com/static/content/static/config/video/config.json'
+    _TEST = {
+        'url': 'http://www.nfl.com/videos/nfl-game-highlights/0ap3000000398478/Week-3-Redskins-vs-Eagles-highlights',
+        'skip_download': True,  # md5 sum fluctuates
+        'info_dict': {
+            'id': '0ap3000000398478',
+            'ext': 'mp4',
+            'title': 'Week 3: Washington Redskins vs. Philadelphia Eagles highlights',
+            'description': 'md5:56323bfb0ac4ee5ab24bd05fdf3bf478',
+            'upload_date': '20140921',
+            'timestamp': 1411337580,
+            'thumbnail': 're:^https?://.*\.jpg$',
+        }
+    }
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        video_id = mobj.group('id')
+
+        config = self._download_json(self._PLAYER_CONFIG_URL, video_id,
+                                     note='Downloading player config')
+        url_template = 'http://nfl.com{contentURLTemplate:s}'.format(**config)
+        video_data = self._download_json(url_template.format(id=video_id), video_id)
+
+        cdns = config.get('cdns')
+        if not cdns:
+            raise ExtractorError('Failed to get CDN data', expected=True)
+
+        formats = []
+        streams = video_data.get('cdnData', {}).get('bitrateInfo', [])
+        for name, cdn in cdns.items():
+            # LimeLight streams don't seem to work
+            if cdn.get('name') == 'LIMELIGHT':
+                continue
+
+            protocol = cdn.get('protocol')
+            host = remove_end(cdn.get('host', ''), '/')
+            if not (protocol and host):
+                continue
+
+            path_prefix = cdn.get('pathprefix', '')
+            if path_prefix and not path_prefix.endswith('/'):
+                path_prefix = '%s/' % path_prefix
+
+            get_url = lambda p: '{protocol:s}://{host:s}/{prefix:s}{path:}'.format(
+                protocol=protocol,
+                host=host,
+                prefix=path_prefix,
+                path=p,
+            )
+
+            if protocol == 'rtmp':
+                preference = -2
+            elif 'prog' in name.lower():
+                preference = -1
+            else:
+                preference = 0
+
+            for stream in streams:
+                path = stream.get('path')
+                if not path:
+                    continue
+
+                formats.append({
+                    'url': get_url(path),
+                    'vbr': int_or_none(stream.get('rate', 0), 1000),
+                    'preference': preference,
+                    'format_note': name,
+                })
+
+        self._sort_formats(formats)
+
+        thumbnail = None
+        for q in ('xl', 'l', 'm', 's', 'xs'):
+            thumbnail = video_data.get('imagePaths', {}).get(q)
+            if thumbnail:
+                break
+
+        return {
+            'id': video_id,
+            'title': video_data.get('storyHeadline'),
+            'formats': formats,
+            'description': video_data.get('caption'),
+            'duration': video_data.get('duration'),
+            'thumbnail': thumbnail,
+            'timestamp': int_or_none(video_data.get('posted'), 1000),
+        }