Hosted in this repository is the open-sourced backend system under the hood of the iOS custom keyboard application Nixacernis Keyboard.
For maximum compatibility with mainstream academic circle, the following translations are mostly taken via reference from Wikipedia pages. Consult Reference section to know more.
| Chinese | English | Note |
|---|---|---|
| 拼音 | pinyin | |
| 声母 | initial | |
| 韵母 | final | |
| 零声母 | zero onset | |
| 逐序格式 | LUTP | Short for Line Up to Pick, a desginated representation of keybaord layouts |
Nixacernis keyboard has totally 18 keys. For convenience, each of them is binded with a number from 0 to 17. (0-indexed for convenience in under hood data processing)
| First Column | Second Column | Third Column | Fourth Column | Fifth Column | Sixth Column | |
|---|---|---|---|---|---|---|
| First Row | 0 | 1 | 2 | 3 | 4 | 5 |
| Second Row | 6 | 7 | 8 | 9 | 10 | 11 |
| Third Row | 12 | 13 | 14 | 15 | 16 | 17 |
Note that since the project is still under heavy development, it's very likely that current APIs may undergo dramatic changes in the future. No backward compatibility is promised, and use them to your own risk.
Go to /transliterate/<query_string> for transliteration.
The format of <query_string>, represented as regular expression, is [1-18](-[1-18])*.
For example, if the user consecutively typed keys corresponding to number 1, 3, 8, 3, 11, then a
GETrequest is sent to url<hosted_url>/transliterate/1-3-8-3-11.
Baed on data past froward, transliteration is conducted in the server side and result is packed in JSON format, routed back to client.
After user typed some words, you may prefer to update the word's priority so that it shows up more likely next time. We have a wrapped up API for doing this. Just send a POST request to url <hosted_url>/increment/<word>.
Currently this feature is done by increase the word count by 1 under the hood. Future API update may introduce increment of word priority with different granularity.
Python 3.6 and Django 2.2
$ pip install django bidict$ git clone https://github.com/MapleCCC/Nixacernis-Keyboard-Backend.git
$ cd nixacernis-keyboard-backend/
# Install dependencies
$ pip install -r requirements.txt
# Set up initial database
$ python manage.py migrate
# Set up server at localhost:8000
$ python manage.py runserverNote that the above commands set up only a naive development server as a built-in feature provided by Django. It's not intended for production deployment.
Instead, use Gunicorn as recommended by Django:
$ gunicorn nixacernis.wsgi# To autoformat `*.py` files under the whole repository
$ make autoformat
# To check linting errors among all `*.py` files under repository
$ make lint
# To preview README file via GitHub style rendering
$ make preview
# To eradicate redundant files for a pristine build state
# NOTE that this command may erase local database that's not version controlled.
$ make clean
# To launch a shell for working around
$ make shellThere is a heroku app instance setup for automatic CI and testing. Go to https://nixacernis-keyboard-backend.herokuapp.com/ to have a check.
Wikipedia pages for Pinyin, Pinyin table, Standard Chinese Phonology, Standard Chinese, Historical Chinese Phonology, etc,.