summaryrefslogtreecommitdiff
path: root/searx/engines/youtube_noapi.py
diff options
context:
space:
mode:
authorMarc Abonce Seguin <marc-abonce@mailbox.org>2019-03-26 20:33:36 -0600
committerMarc Abonce Seguin <marc-abonce@mailbox.org>2019-03-26 21:09:15 -0600
commitf2d49a697124b8f4c6a4df68626b3d29ec959e70 (patch)
treec7af626d656c9245af291f0bbde3c69dcf4ddffa /searx/engines/youtube_noapi.py
parenta80a2d05d14290c57c3cfc59d373e597b1c1c4b5 (diff)
downloadsearxng-f2d49a697124b8f4c6a4df68626b3d29ec959e70.tar.gz
searxng-f2d49a697124b8f4c6a4df68626b3d29ec959e70.zip
[fix] get youtube results from js object
Results are not appearing in the html document anymore, instead they are found inside an object embedded in a script.
Diffstat (limited to 'searx/engines/youtube_noapi.py')
-rw-r--r--searx/engines/youtube_noapi.py70
1 files changed, 34 insertions, 36 deletions
diff --git a/searx/engines/youtube_noapi.py b/searx/engines/youtube_noapi.py
index 9f01841f6..3bf25932b 100644
--- a/searx/engines/youtube_noapi.py
+++ b/searx/engines/youtube_noapi.py
@@ -8,7 +8,8 @@
# @stable no
# @parse url, title, content, publishedDate, thumbnail, embedded
-from lxml import html
+from functools import reduce
+from json import loads
from searx.engines.xpath import extract_text
from searx.utils import list_get
from searx.url_utils import quote_plus
@@ -34,20 +35,6 @@ embedded_url = '<iframe width="540" height="304" ' +\
base_youtube_url = 'https://www.youtube.com/watch?v='
-# specific xpath variables
-results_xpath = "//ol/li/div[contains(@class, 'yt-lockup yt-lockup-tile yt-lockup-video vve-check')]"
-url_xpath = './/h3/a/@href'
-title_xpath = './/div[@class="yt-lockup-content"]/h3/a'
-content_xpath = './/div[@class="yt-lockup-content"]/div[@class="yt-lockup-description yt-ui-ellipsis yt-ui-ellipsis-2"]'
-
-
-# returns extract_text on the first result selected by the xpath or None
-def extract_text_from_dom(result, xpath):
- r = result.xpath(xpath)
- if len(r) > 0:
- return extract_text(r[0])
- return None
-
# do search-request
def request(query, params):
@@ -63,27 +50,38 @@ def request(query, params):
def response(resp):
results = []
- dom = html.fromstring(resp.text)
-
- # parse results
- for result in dom.xpath(results_xpath):
- videoid = list_get(result.xpath('@data-context-item-id'), 0)
- if videoid is not None:
- url = base_youtube_url + videoid
- thumbnail = 'https://i.ytimg.com/vi/' + videoid + '/hqdefault.jpg'
-
- title = extract_text_from_dom(result, title_xpath) or videoid
- content = extract_text_from_dom(result, content_xpath)
-
- embedded = embedded_url.format(videoid=videoid)
-
- # append result
- results.append({'url': url,
- 'title': title,
- 'content': content,
- 'template': 'videos.html',
- 'embedded': embedded,
- 'thumbnail': thumbnail})
+ results_data = resp.text[resp.text.find('ytInitialData'):]
+ results_data = results_data[results_data.find('{'):results_data.find(';\n')]
+
+ results_json = loads(results_data) if results_data else {}
+ sections = results_json.get('contents', {})\
+ .get('twoColumnSearchResultsRenderer', {})\
+ .get('primaryContents', {})\
+ .get('sectionListRenderer', {})\
+ .get('contents', [])
+
+ for section in sections:
+ for video_container in section.get('itemSectionRenderer', {}).get('contents', []):
+ video = video_container.get('videoRenderer', {})
+ videoid = video.get('videoId')
+ if videoid is not None:
+ url = base_youtube_url + videoid
+ thumbnail = 'https://i.ytimg.com/vi/' + videoid + '/hqdefault.jpg'
+ title = video.get('title', {}).get('simpleText', videoid)
+ description_snippet = video.get('descriptionSnippet', {})
+ if 'runs' in description_snippet:
+ content = reduce(lambda a, b: a + b.get('text', ''), description_snippet.get('runs'), '')
+ else:
+ content = description_snippet.get('simpleText', '')
+ embedded = embedded_url.format(videoid=videoid)
+
+ # append result
+ results.append({'url': url,
+ 'title': title,
+ 'content': content,
+ 'template': 'videos.html',
+ 'embedded': embedded,
+ 'thumbnail': thumbnail})
# return results
return results