diff options
| author | Markus Heiser <markus.heiser@darmarit.de> | 2023-09-19 12:55:21 +0200 |
|---|---|---|
| committer | Markus Heiser <markus.heiser@darmarIT.de> | 2023-09-19 17:06:56 +0200 |
| commit | 043dcbf7c5dbd0778a60c75bb2636cb72ea5793b (patch) | |
| tree | eefdebe1832bb1c35cece6fda1e254ac7d68c9d6 /searx/engines/qwant.py | |
| parent | 3ac7c40b6ad16ebdc448e54e22bd0d735dd7b8c0 (diff) | |
| download | searxng-043dcbf7c5dbd0778a60c75bb2636cb72ea5793b.tar.gz searxng-043dcbf7c5dbd0778a60c75bb2636cb72ea5793b.zip | |
[fix] engine qwant (web-lite) - ignore advertising adds
Closes: https://github.com/searxng/searxng/issues/2812
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
Diffstat (limited to 'searx/engines/qwant.py')
| -rw-r--r-- | searx/engines/qwant.py | 5 |
1 files changed, 4 insertions, 1 deletions
diff --git a/searx/engines/qwant.py b/searx/engines/qwant.py index 654a76337..168eb860e 100644 --- a/searx/engines/qwant.py +++ b/searx/engines/qwant.py @@ -159,9 +159,12 @@ def parse_web_lite(resp): dom = lxml.html.fromstring(resp.text) for item in eval_xpath_list(dom, '//section/article'): + if eval_xpath(item, "./span[contains(@class, 'tooltip')]"): + # ignore randomly interspersed advertising adds + continue results.append( { - 'url': extract_text(eval_xpath(item, './span')), + 'url': extract_text(eval_xpath(item, "./span[contains(@class, 'url partner')]")), 'title': extract_text(eval_xpath(item, './h2/a')), 'content': extract_text(eval_xpath(item, './p')), } |