mirror of
https://github.com/thelounge/thelounge.git
synced 2024-05-27 19:02:23 +02:00
b2797c3a11
Our search was using a linear search over the message text data. This is very inefficient for large databases and barely usable. We can add an FTS index, trading storage for speed. As it's setup, this only supports English, but then we get fancy stemming so that say "work" matches itself as well as "working". This could be reduced to just stripping funny chars, with less good search in the English case. |
||
|---|---|---|
| .. | ||
| client | ||
| commands | ||
| fixtures | ||
| models | ||
| plugins | ||
| shared | ||
| src | ||
| tests | ||
| .mocharc.yml | ||
| .nycrc-mocha.json | ||
| .nycrc-report.json | ||
| client.ts | ||
| server.ts | ||
| tsconfig.json | ||
| util.ts | ||