Page MenuHomePhabricator
Create Task
Maniphest T73101

Cirrus misses many JS pages on some searches (Lucene finds them)
Closed, DuplicatePublic
Actions

Description

Here is another regression I noticed when searching for things in JS pages:

Searching for "Tools/APC" on MediaWiki and User namespaces of Portuguese Wikipedia I get these:

Lucene: Results 1 - 7 of 7
https://pt.wikipedia.org/w/index.php?title=Especial%3APesquisar&profile=advanced&search=%22Tools%2FAPC%22&fulltext=Search&ns2=1&ns8=1&profile=advanced&srbackend=LuceneSearch

Cirrus: Result 1 of 1
https://pt.wikipedia.org/w/index.php?title=Especial%3APesquisar&profile=advanced&search=%22Tools%2FAPC%22&fulltext=Search&ns2=1&ns8=1&profile=advanced&srbackend=CirrusSearch

Version: unspecified
Severity: normal
URL: https://pt.wikipedia.org/w/index.php?title=Especial%3APesquisar&profile=advanced&search=%22Tools%2FAPC%22&fulltext=Search&ns2=1&ns8=1&profile=advanced&srbackend=CirrusSearch
See Also:
https://bugzilla.wikimedia.org/show_bug.cgi?id=68920

Details

Reference
bz71101

Related Objects

Event Timeline

• bzimport raised the priority of this task from to High.Nov 22 2014, 3:43 AM
• bzimport added a project: CirrusSearch.
• bzimport set Reference to bz71101.
• bzimport added a subscriber: Unknown Object (MLST).
He7d3r created this task.Sep 21 2014, 2:35 PM
Nemo_bis added a comment.Sep 22 2014, 10:30 AM

Dupe of bug 68782?

He7d3r added a comment.EditedSep 22 2014, 1:41 PM

Only if we want to expand that one by removing the "intitle". But I'm not really sure they have the same underlying cause.

What I know is that I'm struggling with missing results almost every time I want to check some scripts (T70920, T70782, T70527, T65861). Maybe it should be a single high priority bug, or many separate bugs. I don't know. The fact is that with Lucene timing out and Cirrus results not being trustable, my next option is to use dumps, which is not very practical.

• Manybubbles added a comment.Sep 22 2014, 1:43 PM

Bleh. Sorry Lucene isn't being helpful either. I hadn't realized that. I'll just bump this bug's priority up and try to get a start on it real soon.

He7d3r added a comment.Sep 22 2014, 3:30 PM

Cirrus: There were no results matching the query.
https://pt.wikibooks.org/w/index.php?title=Especial%3APesquisar&profile=advanced&search=loader.load&fulltext=Search&ns2=1&ns8=1&profile=advanced&srbackend=CirrusSearch&uselang=en
(same result with 'insource:"loader.load"' and '"loader.load"' )

Lucene: An error has occurred while searching: HTTP request timed out.
https://pt.wikibooks.org/w/index.php?title=Especial%3APesquisar&profile=advanced&search=loader.load&fulltext=Search&ns2=1&ns8=1&profile=advanced&srbackend=LuceneSearch&uselang=en

Dump ptwikibooks-20140905-pages-meta-current.xml
121 pages from User and MediaWiki namespaces containing "loader.load"
Examples:
https://pt.wikibooks.org/wiki/MediaWiki:Gadget-popups.js
https://pt.wikibooks.org/wiki/MediaWiki:Gadget-WikiMiniAtlas.js
https://pt.wikibooks.org/wiki/MediaWiki:Gadget-rightsfilter.js
https://pt.wikibooks.org/wiki/User:Sarrus/common.js
https://pt.wikibooks.org/wiki/User:Teles/common.js

He7d3r added a comment.Apr 20 2015, 6:34 PM

Come on! How many months will we have to wait to get some useful results when looking for CSS and JS pages?

intitle:".css"
There were no results matching the query.
intitle:".js"
One result. Completely unrelated

See also:

• Deskana unsubscribed.Apr 20 2015, 6:57 PM
Nemo_bis added a subscriber: Discovery-ARCHIVED.Apr 22 2015, 8:11 AM
• demon unsubscribed.Aug 19 2015, 3:44 PM
Restricted Application added a project: Discovery-ARCHIVED. · View Herald TranscriptAug 19 2015, 3:44 PM
Restricted Application added a subscriber: Aklapper. · View Herald Transcript
He7d3r mentioned this in T70920: Cirrus finds no results but Lucene finds 500+ (when searching outside main namespace).Nov 12 2015, 11:44 AM
• Deskana closed this task as a duplicate of T64733: CirrusSearch does not find all JavaScript and CSS pages when using insource and intitle syntax.Dec 5 2015, 7:10 AM
He7d3r mentioned this in T122290: Make "insource" work for JavaScript, CSS and Lua pages.Dec 23 2015, 1:26 PM
He7d3r mentioned this in T64733: CirrusSearch does not find all JavaScript and CSS pages when using insource and intitle syntax.
• Deskana moved this task from Inbox to Resolved/Invalid/Declined/Legacy on the CirrusSearch board.Dec 31 2015, 5:10 AM
Content licensed under Creative Commons Attribution-ShareAlike (CC BY-SA) 4.0 unless otherwise noted; code licensed under GNU General Public License (GPL) 2.0 or later and other open source licenses. By using this site, you agree to the Terms of Use, Privacy Policy, and Code of Conduct. · Wikimedia Foundation · Privacy Policy · Code of Conduct · Terms of Use · Disclaimer · CC-BY-SA · GPL