Page MenuHomePhabricator

Improve Ferret engine indexing performance for large blocks of text
ClosedPublic

Authored by epriestley on Sep 26 2017, 2:15 AM.
Tags
None
Referenced Files
F14050088: D18647.diff
Thu, Nov 14, 4:24 PM
F14043254: D18647.diff
Tue, Nov 12, 8:45 AM
F14037748: D18647.diff
Sun, Nov 10, 6:45 PM
F14024068: D18647.diff
Thu, Nov 7, 5:35 AM
F13959449: D18647.id44766.diff
Oct 14 2024, 7:28 PM
Unknown Object (File)
Oct 1 2024, 8:15 AM
Unknown Object (File)
Sep 26 2024, 8:21 PM
Unknown Object (File)
Sep 19 2024, 12:14 PM
Subscribers
None

Details

Summary

See PHI87. Ref T12974. Currently, we do a lot more work here than we need to: we call phutil_utf8_strtolower() on each token, but can do it once at the beginning on the whole block.

Additionally, since ngrams don't care about order, we only need to convert unique tokens into ngrams. This saves us some phutil_utf8v(). These calls can be slow for large inputs.

Test Plan

Diff Detail

Repository
rP Phabricator
Lint
Lint Not Applicable
Unit
Tests Not Applicable