]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/lrt.py
Merge tag 'upstream/2017.03.07'
[youtubedl] / youtube_dl / extractor / lrt.py
index d72d470aa8dbb532f80f44796354dabd3de92261..f5c997ef4c79398734b2bd5feb09fd335eaf5ade 100644 (file)
@@ -2,12 +2,11 @@
 from __future__ import unicode_literals
 
 import re
-import json
 
 from .common import InfoExtractor
 from ..utils import (
     determine_ext,
-    js_to_json,
+    int_or_none,
     parse_duration,
     remove_end,
 )
@@ -16,46 +15,72 @@ from ..utils import (
 class LRTIE(InfoExtractor):
     IE_NAME = 'lrt.lt'
     _VALID_URL = r'https?://(?:www\.)?lrt\.lt/mediateka/irasas/(?P<id>[0-9]+)'
-    _TEST = {
+    _TESTS = [{
+        # m3u8 download
         'url': 'http://www.lrt.lt/mediateka/irasas/54391/',
+        'md5': 'fe44cf7e4ab3198055f2c598fc175cb0',
         'info_dict': {
             'id': '54391',
             'ext': 'mp4',
             'title': 'Septynios Kauno dienos',
             'description': 'md5:24d84534c7dc76581e59f5689462411a',
             'duration': 1783,
+            'view_count': int,
+            'like_count': int,
         },
-        'params': {
-            'skip_download': True,  # HLS download
+    }, {
+        # direct mp3 download
+        'url': 'http://www.lrt.lt/mediateka/irasas/1013074524/',
+        'md5': '389da8ca3cad0f51d12bed0c844f6a0a',
+        'info_dict': {
+            'id': '1013074524',
+            'ext': 'mp3',
+            'title': 'Kita tema 2016-09-05 15:05',
+            'description': 'md5:1b295a8fc7219ed0d543fc228c931fb5',
+            'duration': 3008,
+            'view_count': int,
+            'like_count': int,
         },
-
-    }
+    }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
 
         title = remove_end(self._og_search_title(webpage), ' - LRT')
-        thumbnail = self._og_search_thumbnail(webpage)
-        description = self._og_search_description(webpage)
-        duration = parse_duration(self._search_regex(
-            r"'duration':\s*'([^']+)',", webpage,
-            'duration', fatal=False, default=None))
 
         formats = []
-        for js in re.findall(r'(?s)config:\s*(\{.*?\})', webpage):
-            data = json.loads(js_to_json(js))
-            if data['provider'] == 'rtmp':
+        for _, file_url in re.findall(
+                r'file\s*:\s*(["\'])(?P<url>(?:(?!\1).)+)\1', webpage):
+            ext = determine_ext(file_url)
+            if ext not in ('m3u8', 'mp3'):
+                continue
+            # mp3 served as m3u8 produces stuttered media file
+            if ext == 'm3u8' and '.mp3' in file_url:
+                continue
+            if ext == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    file_url, video_id, 'mp4', entry_protocol='m3u8_native',
+                    fatal=False))
+            elif ext == 'mp3':
                 formats.append({
-                    'format_id': 'rtmp',
-                    'ext': determine_ext(data['file']),
-                    'url': data['streamer'],
-                    'play_path': 'mp4:%s' % data['file'],
-                    'preference': -1,
+                    'url': file_url,
+                    'vcodec': 'none',
                 })
-            else:
-                formats.extend(
-                    self._extract_m3u8_formats(data['file'], video_id, 'mp4'))
+        self._sort_formats(formats)
+
+        thumbnail = self._og_search_thumbnail(webpage)
+        description = self._og_search_description(webpage)
+        duration = parse_duration(self._search_regex(
+            r'var\s+record_len\s*=\s*(["\'])(?P<duration>[0-9]+:[0-9]+:[0-9]+)\1',
+            webpage, 'duration', default=None, group='duration'))
+
+        view_count = int_or_none(self._html_search_regex(
+            r'<div[^>]+class=(["\']).*?record-desc-seen.*?\1[^>]*>(?P<count>.+?)</div>',
+            webpage, 'view count', fatal=False, group='count'))
+        like_count = int_or_none(self._search_regex(
+            r'<span[^>]+id=(["\'])flikesCount.*?\1>(?P<count>\d+)<',
+            webpage, 'like count', fatal=False, group='count'))
 
         return {
             'id': video_id,
@@ -64,4 +89,6 @@ class LRTIE(InfoExtractor):
             'thumbnail': thumbnail,
             'description': description,
             'duration': duration,
+            'view_count': view_count,
+            'like_count': like_count,
         }