Completion project
===========================

Uses language models to predict what user is about to write.
It can be used:
 - in T9 or iTap like systems,
 - for auto-completion in text editors,
 - in devices for disabled people (writing with one finger, eye movements, etc.).

Dependecies
-----------
Needs Python3, PyQt4 for Python3, and SIP framework.
It could be installed from these Ubuntu packages:
	- python3-pyqt4
	- python3-sip 
	- python3-sip-dev
It's also recommended to have some NLP toolkit installed to
create ARPA files.

Requires Ubuntu 15.04 (or newer).


Demo
----
First run `make` in the root directory, it should:
	- compile KenLM objects,
	- compile KenLM wrapper and ARPA selector objects,
	- compile executables for building binary files from ARPA
	  for KenLM and ARPA selector,
	- generate SIP bindings sources and compile them to .so
	  Python modules,
	- create binary files for sample language model.

If everything works properly demo could be started by
`./src/python/ui-demo.py b -lm sample-data/povidky.bin -sel sample-data/povidky.sel.bin`.