Commit graph

12 commits

Author SHA1 Message Date
Marc Durdin
b7ce99480f feat(web): prediction casing follows current layer
Fixes #6455.

This involved adding an extra hint to the data passed into the lexical
model layer, being the CasingForm that the current layer wants to use.

Given there are several entry points to the lexical model layer, this
meant a few places where we needed to check the current layer.

I refactored the very long function `InputProcessor.processKeyEvent` to
extract out the `buildAlternates` function. There should be no
functional changes to the `buildAlternates` code. But it makes
`processKeyEvent` significantly easier to read.
2022-04-01 16:20:50 +11:00
Eddie Antonio Santos
fea6026f71
test(common/lmlayer): add regression test for searchTermToKey 2020-04-06 09:25:50 -06:00
Eddie Antonio Santos
13c5a4cdd5
Trie to appease IE... 2019-07-30 09:47:15 -06:00
Eddie Antonio Santos
ada6af46f6
Fix old dummy model test case. 2019-07-25 16:05:04 -06:00
Eddie Antonio Santos
c938767da7
Fix in-browser test. 2019-06-12 15:26:52 -06:00
Eddie Antonio Santos
7d09c0097f
Change integration tests to use tries instead of the simple wordlist. 2019-05-28 11:24:07 -06:00
Eddie Antonio Santos
75ef303422
Update english-1000 in integration test. 2019-04-29 20:46:53 -06:00
Joshua A. Horton
3b1e715f4b Adjusts predictive model spec for 'configuring' based on 'capabilities' 2019-03-06 10:15:14 +07:00
Joshua A. Horton
263af9c721 Addresses many of Eddie's PR review comments. 2019-03-06 08:24:57 +07:00
Joshua A. Horton
2bf999edee Patches up the integrated tests to work with the model loading change. 2019-02-28 15:52:24 +07:00
Joshua A. Horton
2f5b48f625 Reinstates the basic headless prediction test. 2019-02-28 15:04:36 +07:00
Joshua A. Horton
6218b808dd Retools how the LMLayer adds models, fixes most headless cases. 2019-02-28 14:45:42 +07:00