Make emojier use its own data instead of relying on ibus.
- IBus data originated from unicode official data, use unicode data reduce the dependency on emoji feature.
- It's not build data from run time gives us more control on when to update the data. Especially if font does not cover the emoji, it will just display a blank space. (At this point unicode-emoji 15.0 is out, but noto doesn't fully support it yet).
- It is never a part of kimpanel infra, so also move it out of kimpanel directory.
There are also other some small improvements:
- The category name string is now under our own control, this also fixes the "&" accelerator issue in the side bar.
- dict loading order now always load en. Otherwise certain language may only shows a small number of emoji. Can be tested via
LANGUAGE=ti_ER LANG=ti_ER.UTF-8 plasma-emojier
Edited by Xuetian Weng