[soompi] Add new extractor for tv.soompi.com

diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py
index 6f8c261d5b416bcc7fd4f7a85185ece235030b57..2a5cf95470c99886eadbc5306fb31f516d7d8b80 100644 (file)
--- a/youtube_dl/extractor/__init__.py
+++ b/youtube_dl/extractor/__init__.py
@@ -482,6 +482,10 @@ from .smotri import (
 from .snotr import SnotrIE
 from .sockshare import SockshareIE
 from .sohu import SohuIE
+from .soompi import (
+    SoompiIE,
+    SoompiShowIE,
+)
 from .soundcloud import (
     SoundcloudIE,
     SoundcloudSetIE,

diff --git a/youtube_dl/extractor/soompi.py b/youtube_dl/extractor/soompi.py
new file mode 100644 (file)
index 0000000..5ecf40b
--- /dev/null
+++ b/youtube_dl/extractor/soompi.py
@@ -0,0 +1,130 @@
+# encoding: utf-8
+from __future__ import unicode_literals
+
+import re
+import json
+import base64
+import xml.etree.ElementTree
+
+# Soompi uses the same subtitle encryption as crunchyroll
+from .crunchyroll import CrunchyrollIE
+
+
+class SoompiIE(CrunchyrollIE):
+    IE_NAME = 'soompi'
+    _VALID_URL = r'^https?://tv\.soompi\.com/en/watch/(?P<id>[0-9]+)'
+    _TESTS = [{
+        'url': 'http://tv.soompi.com/en/watch/23363',
+        'info_dict': {
+            'id': '23363',
+            'ext': 'mp4',
+            'title': 'Liar Game CM1',
+            'description': '15sec'
+        },
+        'params': {
+            'skip_download': True,
+        },
+    }]
+
+    def _get_episodes(self, webpage, episode_filter=None):
+        episodes = json.loads(
+            self._search_regex(r'\s+VIDEOS\s+= (\[.+?\]);', webpage, "episodes meta"))
+        return [ep for ep in episodes if episode_filter is None or episode_filter(ep)]
+
+    def _get_subtitles(self, video_id, show_format_xml):
+        subtitles = {}
+        subtitle_info_nodes = show_format_xml.findall('./{default}preload/subtitles/subtitle')
+        subtitle_nodes = show_format_xml.findall('./{default}preload/subtitle')
+
+        sub_langs = {}
+        for i in subtitle_info_nodes:
+            sub_langs[i.attrib["id"]] = i.attrib["title"]
+
+        for s in subtitle_nodes:
+            lang_code = sub_langs.get(s.attrib["id"], None)
+            if lang_code is None:
+                continue
+
+            sub_id = int(s.attrib["id"])
+            iv = base64.b64decode(s.find("iv").text)
+            data = base64.b64decode(s.find("data").text)
+            subtitle = self._decrypt_subtitles(data, iv, sub_id).decode('utf-8')
+            sub_root = xml.etree.ElementTree.fromstring(subtitle)
+
+            subtitles[lang_code] = [{
+                'ext': 'srt', 'data': self._convert_subtitles_to_srt(sub_root)
+            }, {
+                'ext': 'ass', 'data': self._convert_subtitles_to_ass(sub_root)
+            }]
+        return subtitles
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+
+        webpage = self._download_webpage(
+            url, video_id, note="Downloading episode page",
+            errnote="Video may not be available for your location")
+        vid_formats = re.findall(r"\?quality=q([0-9]+)", webpage)
+
+        show_meta = json.loads(
+            self._search_regex(r'\s+var show = (\{.+?\});', webpage, "show meta"))
+        episodes = self._get_episodes(
+            webpage, episode_filter=lambda x: x['id'] == video_id)
+
+        title = episodes[0]["name"]
+        description = episodes[0]["description"]
+        duration = int(episodes[0]["duration"])
+        slug = show_meta["slug"]
+
+        formats = []
+        show_format_xml = None
+        for vf in vid_formats:
+            show_format_url = "http://tv.soompi.com/en/show/%s/%s-config.xml?mode=hls&quality=q%s" \
+                              % (slug, video_id, vf)
+            show_format_xml = self._download_xml(
+                show_format_url, video_id, note="Downloading q%s show xml" % vf)
+            avail_formats = self._extract_m3u8_formats(
+                show_format_xml.find('./{default}preload/stream_info/file').text,
+                video_id, ext="mp4", m3u8_id=vf, preference=int(vf))
+            formats.extend(avail_formats)
+        self._sort_formats(formats)
+
+        subtitles = self.extract_subtitles(video_id, show_format_xml)
+
+        return {
+            'id': video_id,
+            'title': title,
+            'description': description,
+            'duration': duration,
+            'formats': formats,
+            'subtitles': subtitles
+        }
+
+
+class SoompiShowIE(SoompiIE):
+    IE_NAME = 'soompi:show'
+    _VALID_URL = r'^https?://tv\.soompi\.com/en/shows/(?P<id>[0-9a-zA-Z\-_]+)'
+    _TESTS = [{
+        'url': 'http://tv.soompi.com/en/shows/liar-game',
+        'info_dict': {
+            'id': 'liar-game',
+            'title': 'Liar Game',
+            'description': 'md5:52c02bce0c1a622a95823591d0589b66',
+        },
+        'playlist_count': 14,
+    }]
+
+    def _real_extract(self, url):
+        show_id = self._match_id(url)
+
+        webpage = self._download_webpage(url, show_id, note="Downloading show page")
+        title = self._og_search_title(webpage).replace("SoompiTV | ", "")
+        description = self._og_search_description(webpage)
+
+        episodes = self._get_episodes(webpage)
+        entries = []
+        for ep in episodes:
+            entries.append(self.url_result(
+                'http://tv.soompi.com/en/watch/%s' % ep['id'], 'Soompi', ep['id']))
+
+        return self.playlist_result(entries, show_id, title, description)