Skip to content

Latest commit

 

History

History
98 lines (63 loc) · 3.12 KB

development.md

File metadata and controls

98 lines (63 loc) · 3.12 KB
Raw

Project Development Tips

This document describes a few things that may be useful for development.

When Pulling Changes From Master

Code changes to the project's dependencies or database (tables or functions) are not reflected in the Docker stack until you manually rebuild the corresponding Docker images and replace the database Docker Volume, which is a persistent storage container attached to the ephemeral Docker stack container.

To perform both of these steps:

make dev_down
make dev_drop_database
make dev_build
make dev_up

If you don't want to lose the contents of your database ... well we don't currently have a tool to automatically export/import the data. But we need it! See Issue #299, and feel free to add this feature.

In-project guide to using Docker:

Available here: doc/docker.md

Create a Local User

You can create user accounts manually in your local database if you'd like to do so for any manual testing. Here's how:

  1. Choose a password and compute its bcrypt hash:

In the project directory, run

docker exec -ti pipenv run python

to open the python interactive terminal with the project's dependencies loaded. Then run

$ from werkzeug.security import generate_password_hash
$ generate_password_hash('your_password')
  1. Choose an email address and insert it and your password hash into the database:

In the project directory, run

make dev_psql

This launches the PSQL interactive environment in the project's postgres docker container. In this terminal, run:

$ SELECT * FROM USERS_CREATE('your_email', 'your_hashed_password', 'your name', 'your group name', true);

providing your actual email and hashed password each in the single-quotes.

This runs a custom SQL function which inserts your user credentials.

Manually Testing Backend API calls

See design.md for the specifications of each API endpoint.

To test the endpoint calls for /auth_token and /users respectively, run:

$ curl -X GET -i -H "Content-Type: application/json" "localhost:5000/api/auth_token" --data '{"email":"your_email", "password":"your_password"}'

which should return something like this:

{
  "auth_token": "[long hash string]"
}

the auth_token string provides a claim that a particular user is logged in. Subsequent endpoints that require authorization will read the auth_token string to verify the logged-in user's credentials.

To run the /users/ POST endpoint, run

$ curl -X POST -i -H "Content-Type: application/json" -H "Authorization: Bearer [auth-string]" localhost:5000/api/users --data '{"email":"new_email", "password":"new_password", "admin":false}'

replacing [auth-string] with the auth_token value you just received. This should return:

{ "admin": false, "email": "new_email", "timestamp": "[timestamp]" }

Mock Endpoints that make 3rd-party requests:

We have an alternate Docker configuration that mocks the /api/search and /api/oeci endpoint behavior. See src/frontend/developerUtils/developerUtils.md