Dict Proto

Prototype of a digital Latin-Bulgarian dictionary with morphological analysis.

The app presents a simple search interface. Upon entering a form of a Latin word it's analyzed by Whitaker's Words. The output of this analysis is used to determine the possible lemmas which are then retrieved from the dictionary database.

The database of this prototype includes a limited number of Latin lemmas and their Bulgarian meanings.

A list of all mentions of the form in the database is also generated.

Installation

We use Whitaker's Words for the morphological analysis. To use this project you need to have it installed on your system. Follow the instructions in the Whitaker's Words repository.

Install pipenv. Then clone the project and in the project directory run

pipenv install

to install dependencies and enter the virtualenv shell.

Usage

Configure paths for the words binary from Whitaker's Words in config.yaml. Copy the example configuration file to your own personal config (do not put this personal config in git). The copy is where you write your personal configuration.

cp config.example.yaml config.yaml

Open config.yaml file and update the paths. words.cwd is the path to the directory where the data files are located (the main directory of words). words.bin is the binary, for example words.exe. On windows you might need to specify full path to the executable file.

Then run

pipenv run python hello.py

to start the flask server.

Building and deploying

To build the docker image, you must first build Whitaker's Words for linux.

Clone the words project into ./words directory in this project then follow their build instructions.

When this is done, run

docker-compose build

to build the image.