chore(android/engine): Remove use of lexical-model catalog
- Dominant language
- Pascal
- Stars
- 534
- Forks
- 143
- Avg merge
- 2d 10h
- Merged PRs (30d)
- 113
Description
This a change from 13.0 to 14.0 where we no longer need to cache the catalog info of available lexical-models
e.g. https://api.keyman.com/model?q
During keyboard installation, a check is already made for the associated lexical model and that gets downloaded/installed as needed.
### TODO
- [x] Remove queries for the catalog info
- [ ] Add a one-time upgrade check going to 14.0 for existing keyboards and installing associated lexical-models (this is a stretch goal, and likely to defer to future release).
- [ ] Cleanup data/Dataset
- [ ] Change ModelPicker UX
- [x] Open Dictionaries option, Keyman should query api.keyman.com for available models for that language.
- [ ] Show spinner in dictionary list until it finishes loading
@mcdurdin can you provide guidance on impacted UX?
1. in "Settings" --> "Languages", there's existing workflow where users can navigate to "Dictionaries" and pick an available model to install (if not installed yet)
2. I believe we will still allow users to navigate to an installed model and display the model info screen
Contributor guide
Assessment
This issue has not been assessed yet.