Home

Welcome to the Compendium wiki!

As for now this wiki is quite tiny, but it gives some insight on the why and how Compendium has been done, and allows to track the development status.

The analysis process description explains the text processing recipe, from the Lexer to the detectors.

See the changelog for the development status and the eventual breaking changes, and the milestones page for the upcoming releases notes.

Contribute

If you wish to contribute to Compendium, here are the steps in order to setup Compendium environment.

Setup env

Clone the repository:

git@github.com:Ulflander/compendium-js.git

Install the dependencies using npm:

npm install

then, initialise the build process with:

mkdir build && gulp lexicon

Build

Compendium requires gulp for the build process. Install it globally using npm install -g gulp.

Use gulp command to build and watch the source files for changes for live testing and rebuild.

Test

Compendium uses nodeunit to run the unit tests. Install it with npm install nodeunit -g. Use nodeunit test/*.js or node test to run the tests.

Run Penn Treebank benchmark

node test/benchmark/run.js will run Compendium against a selection of the Penn Treebank dataset.