gtv.org extractor, I guess

haruhi_dl/extractor/extractors.py

@@ -438,6 +438,7 @@ from .googlesearch import GoogleSearchIE
 from .goshgay import GoshgayIE
 from .gputechconf import GPUTechConfIE
 from .groupon import GrouponIE
+from .gtv import GtvIE
 from .guardian import (
     GuardianAudioIE,
     GuardianVideoIE,

haruhi_dl/extractor/gtv.py

@@ -0,0 +1,75 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import base64
+import re
+
+from .common import InfoExtractor
+from ..utils import (
+    NO_DEFAULT,
+)
+
+
+class GtvIE(InfoExtractor):
+    IE_NAME = 'gtv'
+    IE_DESC = 'GTV.org'
+    _VALID_URL = r'https?://(?:www\.)?gtv\.org/video/id=(?P<id>[a-f\d]+)'
+    _TESTS = [{
+        'url': 'https://gtv.org/video/id=5edc48b087564418749581c0',
+        'info_dict': {
+            'id': '5edc48b087564418749581c0',
+            'ext': 'm3u8',
+            'title': 'Hongkongers Commemorate Tiananmen Square Massacre',
+        },
+    }, {
+        'url': 'https://www.gtv.org/video/id=5ed91fb759b1dc11aa3d99db',
+        'info_dict': {
+            'id': '5ed91fb759b1dc11aa3d99db',
+            'ext': 'm3u8',
+            'title': 'Tiananmen : à Taïwan et en Corée du Sud, des cérémonies',
+        },
+    }]
+    # this code is based on a terrible idea to "parse" gRPC data with regexes
+    # that randomly may or may not work, depending on a specific video
+    _WORKING = False
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        headers = {
+            'Content-Type': 'application/grpc-web-text',
+            'Accept': 'application/grpc-web-text',
+        }
+        data = self._download_webpage(
+            'https://app.gtv.org/grpc.video.VideoService/video',
+            video_id, 'Downloading video metadata',
+            data=base64.b64encode(b'\x00\x00\x00\x00\x1a\x0a\x18' + bytes(video_id.encode('utf-8'))),
+            headers=headers)
+        # decoding base64-encoded gRPC data
+        data = base64.b64decode(bytes(data.encode('utf-8')))
+        self.to_screen(data)
+
+        def search(regex, name, default=NO_DEFAULT):
+            val = self._search_regex(re.compile(regex), data, name, default=default)
+            if val:
+                return val.decode('utf-8')
+
+        formats = []
+
+        m3u_url = search(br'\xaa\x01>(/.+?\.m3u8)\xfa', 'm3u8 url')
+        if m3u_url:
+            formats.extend(self._extract_m3u8_formats('https://filegroup.gtv.org' + m3u_url, video_id))
+
+        mp4_url = search(br'\".(/.+?\.mp4)\*A', 'mp4 url')
+        if mp4_url:
+            formats.append({
+                'url': 'https://filegroup.gtv.org' + mp4_url,
+                'ext': 'mp4',
+            })
+
+        self._sort_formats(formats)
+
+        return {
+            'id': video_id,
+            'formats': formats,
+            'title': search(br':.([^\n]+)\n?P.X', 'video title'),
+        }