layout | permalink | description | title | width |
---|---|---|---|---|
default |
dev |
how to contribute |
developping matchID |
is-10 |
You should first make stop
matchID and have a clean docker environnement, if you first ran matchID with make start
.
Then :
make dev
The interest of using docker
for developping or debugging is that you are near to production architecture. You should consult the automation corner to understand the components of matchID.
the frontend code should be in ../frontend/src
, it's in Vue.js.
When you start developping you still can go to http://localhost/matchID
.
The only difference is that your changes in the code will appear live, with the node.js
server.
Any error will appear live in your browser navigation
Any change on the code in code/*.py
will have a live effect, thanx to the werkzeug
which checks any changes.
When you have a bug in your python code (and python can't compile), you'll have to correct it and docker
will restart quickly (but you may wait 30 seconds).
When debugging the backend you should try :
docker logs -f matchid-backend
Then any error or dirty print
will be prompted here.
For logs and print, the werkzeug
has a buffer of 100 lines, to you have to flush it or wait for faster debug.
Every change of the yaml
configuration is read whenever an api call is done. So, if you change conf.yml
or connectors.yml
, or matchID_validation.yml
,
or any recipe or dataset if you reordered your projects configuration from server side, you should just make an api call :
curl -XGET http://localhost/api/v0/conf/
This will moreover a useful way to check if your changes have been read, as it return the configuration of all the projects.
We use the GitHub code pipeline, so you have any contribution you should fork the project and then use the pull request method.