no ref
- added dir prop, calculated by i18next from language (using the dir
function)
- tweaked a few styles to use me/ms/pe/ps instead of mr/ml/pr/pl
- added updated test that checks that stemming works in English, and added tests for partial and full-word searching with RTL content.
no ref
According to the flexsearch documentation, https://github.com/nextapps-de/flexsearch?tab=readme-ov-file#cjk-word-break-chinese-japanese-korean for searching CJK text, need to pass in a custom encode function for better search results.
This enhancement for CJK will only take effect when the ghost site locale is set to one of `zh`, `zh-Hans`, `zh-Hant`, `ja`, `ko`.
Co-authored-by: Cathy Sarisky <42299862+cathysarisky@users.noreply.github.com>
