cacahuate 2.0.0

The process virtual machine

The process virtual machine

This project defines storage for an abstract _process_ in a company, and implements a virtual machine that keeps track of the execution of instances of the process.

Develop

You will need the redis and mongo databases, and rabbitmq for this to work.

git clone https://github.com/tracsa/cacahuate.git && cd cacahuate
virtualenv -p /usr/bin/python3 .env
echo "export CACAHUATE_SETTINGS=$(pwd)/settings_develop.py" >> .env/bin/activate
echo "export FLASK_APP=cacahuate.http.wsgi" >> .env/bin/activate
echo "export FLASK_DEBUG=1" >> .env/bin/activate
touch settings_develop.py
source .env/bin/activate
pip install -r requirements.txt
pytest

Installation

pip install cacahuate

Cacahuated

This is the daemon in charge of moving pointers in the process, run with:

cacahuated

The Cacahuate REST API

In this same repository you will find a flask application that exposes a REST api for controling Cacahuate.

How to run

FLASK_APP=cacahuate.http.wsgi flask run

You can use any wsgi-compliant server, like gunicorn, to run this:

gunicorn cacahuate.http.wsgi:app

TODO

• cambiar api de LoginProvider y HierarchyProvider, no deben guardar usuarios

• auth backends no deben crear usuarios, solo devolver objetos calificados

• date input solo debe recibir yyyy-mm-dd

• hacer validación intensiva de xml: - dependencias existen arriba en el xml - id de nodos son únicos - condicionales validan - params existen

• guardar en un store privado por hash los xml al iniciar proceso, obsoleta el sistema de versiones (borrar toda referencia a la versión) - correr validaciones al mover un xml a este storage

• atrapar excepciones de template de jinja y añadir mensaje al titulo

• todo uso de jinja debe tener try