daniel added a comment.

Implementation Note: Ranking is currently done by an in-memory sort on 5000 
prefix matches. This is needed to avoid a file-sort on the MySQL server:  A 
composite index covering the text (or search_key) field plus the weight would 
not be able to utility the index on the weight for prefix matches. Composite 
indexes only work with exact matches. A multidimensional index would be needed 
to do ranked prefix matches in MySQL.

The proposed solution is to drop support for ranking based on SQL; Ranking 
would be fully supported in the implementation based on Elastic search.


TASK DETAIL
  https://phabricator.wikimedia.org/T86949

REPLY HANDLER ACTIONS
  Reply to comment or attach files, or !close, !claim, !unsubscribe or !assign 
<username>.

EMAIL PREFERENCES
  https://phabricator.wikimedia.org/settings/panel/emailpreferences/

To: daniel
Cc: Aklapper, daniel, Wikidata-bugs, aude



_______________________________________________
Wikidata-bugs mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/wikidata-bugs

Reply via email to