First, thanks for being interested in helping us out! If you find an issue you're interested in, feel free to make a comment about how you're thinking of approaching implementing it in the issue and we can give you feedback. Please also read our code of conduct before getting started.
When you come to implement your new feature, clone the repository and then create a branch off develop
locally and add commits to implement your feature.
If your git history is not so clean, please do rewrite before you submit your PR - if you're not sure if you need to do this, go ahead and submit and we can let you know when you submit.
To submit your changes for review you have to fork the repository, push your new branch there and then create a Pull Request with OpenOversight:develop
as the target.
Use PULL_REQUEST_TEMPLATE.md to create the description for your PR! (The template should populate automatically when you go to open the pull request.)
Whenever you make a commit with git
the name and email saved locally is stored with that commit and will become part of the public history of the project. This can be an unwanted, for example when using a work computer. We recommend changing the email-settings in the github account at https://github.com/settings/emails and selecting "Keep my email addresses private" as well as "Block command line pushes that expose my email". Also find your github-email address of the form <id>+<username>@users.noreply.github.com
in that section. Then you can change the email and username stored with your commits by running the following commands
git config user.email "<your-github-email>"
git config user.name "<your-github-username>"
This will make sure that all commits you make locally are associated with your github account and do not contain any additional identifying information. More detailed information on this topic can be found here.
We use pre-commit for automated linting and style checks. Be sure to install pre-commit and run pre-commit install
in your local version of the repository to install our pre-commit checks. This will make sure your commits are always formatted correctly.
You can run pre-commit run --all-files
or make lint
to run pre-commit over your local codebase, or pre-commit run
to run it only over the currently stages files.
Keep in mind when adding images that alt
tags are required for screen readers. If text outside of the image explains what the image is or is referring to, the tag can be an empty string (alt=""
). The tag can also be empty if the image is decoration and does not add information or context. If the image has text or important information, use the present tense to describe what is happening in the image.
For further reading:
- https://www.a11yproject.com/
- https://www.w3.org/WAI/tutorials/images/decision-tree/
- https://accessibility.huit.harvard.edu/describe-content-images
You can use our Docker-compose environment to stand up a development OpenOversight.
You will need to have Docker installed in order to use the Docker development environment.
To build and run the development environment, simply make dev
. Whenever you want to rebuild the containers, make build
(you should need to do this rarely).
Tests are executed via make test
. If you're switching between the Docker and Vagrant/VirtualBox environments and having trouble getting tests running, make sure to delete any remaining .pyc
files and __pycache__
directories.
To hop into the postgres container, you can do the following:
$ docker exec -it openoversight-postgres-1 bash
# psql -d openoversight-dev -U openoversight
or run make attach
.
Similarly to hop into the web container:
$ docker exec -it openoversight-web-1 bash
Once you're done, make stop
and make clean
to stop and remove the containers respectively.
OpenOversight tries to auto-detect which email implementation to use based on which of the following is configured (in this order):
- Google:
service_account_key.json
exists and is not empty - SMTP:
MAIL_SERVER
andMAIL_PORT
environment variables are set - Simulated: If neither of the previous 2 implementations are configured, emails will only be logged
To send email using a GSuite email account, you will need a Google Cloud Platform service account that is attached to that email address. Here are some general tips for working with service accounts: Link. We would suggest that you do not use a personal email address, but instead one that is used strictly for sending out OpenOversight emails.
You will need to do these two things for the service account to work as a Gmail bot:
- Enable domain-wide delegation for the service account: Link
- Enable the
https://www.googleapis.com/auth/gmail.send
scope in the Gmail API for your service account: Link - Save the service account key file in OpenOversight's base folder as
service_account_key.json
. The file is in the.gitignore
file GitHub will not allow you to save it, provided you've named it correctly.
To send email using SMTP, set the following environment variables in your docker-compose.yml file or .env file:
MAIL_SERVER
MAIL_PORT
MAIL_USE_TLS
MAIL_USERNAME
MAIL_PASSWORD
For more information about these settings, please see the Flask-Mail documentation.
Regardless of implementation, save the email address associated with your service account to a variable named OO_SERVICE_EMAIL
in a .env
file in the base directory of this repository. For development and testing, update the OO_SERVICE_EMAIL
variable in the docker-compose.yml
file.
Example .env
variable:
OO_SERVICE_EMAIL="[email protected]"
In addition to needing a service account email, you also need an admin email address so that users have someone to reach out to if an action is taken on their account that needs to be reversed or addressed.
For production, save the email address associated with your admin account to a variable named OO_HELP_EMAIL
in a .env
file in the base directory of this repository. For development and testing, update the OO_HELP_EMAIL
variable in the docker-compose.yml
file.
Example .env
variable:
OO_HELP_EMAIL="[email protected]"
We use an S3 bucket for image uploads. If you are working on functionality involving image uploads, then you should follow the "S3 Image Hosting" section in DEPLOY.md to make a test S3 bucket on Amazon Web Services.
Once you have done this, you can put your AWS credentials in the following environmental variables:
$ export S3_BUCKET_NAME=openoversight-test
$ export AWS_ACCESS_KEY_ID=testtest
$ export AWS_SECRET_ACCESS_KEY=testtest
$ export AWS_DEFAULT_REGION=us-east-1
Now when you run make dev
as usual in the same session, you will be able to submit images to
your test bucket.
Running make dev
will create the database and persist it into your local filesystem.
You can access your PostgreSQL development database via psql using:
psql -h localhost -d openoversight-dev -U openoversight --password
with the password terriblepassword
.
In the event that you need to create or delete the test data, you can do that with
$ python test_data.py --populate
to create the data
or
$ python test_data.py --cleanup
to delete the data
Within the database we use timestamptz
fields for timestamps. To make sure that you are setting timestamps in the correct timezone, set the environment variable TIMEZONE
to your respective Olson-style timezone so that you can make sure any DST conversions are handled by PostgreSQL.
You'll first have to start the Docker instance for the OpenOversight app using the command make start
. To do this, you'll need to be in the base folder of the repository (the one that houses the Makefile
).
$ make start
docker-compose build
...
docker-compose up -d
[+] Running 2/0
✔ Container openoversight-postgres-1 Running 0.0s
✔ Container openoversight-web-1 Running
From here on out, we'll be using the Flask CLI. First we need to 'stamp' the current version of the database:
$ docker exec -it openoversight-web-1 bash # 'openoversight-web-1' is the name of the app container seen in the step above
$ flask db stamp head
$ flask db migrate -m "[THE NAME OF YOUR MIGRATION IN ALL LOWER CASE]" # NOTE: Slugs are limited to 40 characters and will be truncated after the limit
(Hint: If you get errors when running flask
commands, e.g. because of differing Python versions, you may need to run the commands in the docker container by prefacing them as so: docker exec -it openoversight_web_1 flask db stamp head
)
Next make your changes to the database models in OpenOversight/app/models/database.py
. You'll then generate the migrations:
$ flask db migrate -m "[THE NAME OF YOUR MIGRATION IN ALL LOWER CASE]"
And then you should inspect/edit the migrations. You can then apply the migrations:
$ flask db upgrade
You can also downgrade the database using:
flask db downgrade
One way to avoid hitting version incompatibility errors when running flask
commands is to use a virtualenv. See Python Packaging user guidelines for instructions on installing virtualenv. After installing virtualenv, you can create a virtual environment by navigating to the OpenOversight directory and running the below
python3 -m virtualenv env
Confirm you're in the virtualenv by running
which python
The response should point to your env
directory.
If you want to exit the virtualenv, run
deactivate
To reactivate the virtualenv, run
source env/bin/activate
While in the virtualenv, you can install project dependencies by running
pip install -r requirements.txt
and
pip install -r requirements-dev.txt
In addition to generating database migrations, the Flask CLI can be used to run additional commands:
$ flask --help
Usage: flask [OPTIONS] COMMAND [ARGS]...
A general utility script for Flask applications.
Provides commands from Flask, extensions, and the application. Loads the
application defined in the FLASK_APP environment variable, or from a
wsgi.py file. Setting the ENV environment variable to 'development'
will enable debug mode.
$ export FLASK_APP=hello.py
$ export ENV=development
$ flask run
Options:
--version Show the flask version
--help Show this message and exit.
Commands:
bulk-add-officers Bulk adds officers.
db Perform database migrations.
link-images-to-department Link existing images to first department
link-officers-to-department Links officers and units to first department
make-admin-user Add confirmed administrator account
routes Show the routes for the app.
run Runs a development server.
shell Runs a shell in the app context.
In development, you can make an administrator account without having to confirm your email:
$ flask make-admin-user
Username: redshiftzero
Email: [email protected]
Password:
Type your password again:
Administrator redshiftzero successfully added
In docker-compose.yml
, below the line specifying the port number, add the following lines to the web
service:
stdin_open: true
tty: true
Also in docker-compose.yml
, below the line specifying the ENV
, add the following to the environment
portion of the web
service:
FLASK_DEBUG: 0
The above line disables the werkzeug reloader, which can otherwise cause a bug when you place a breakpoint in code that loads at import time, such as classes. The werkzeug reloader will start one pdb process at import time and one when you navigate to the class. This makes it impossible to interact with the pdb prompt, but we can fix it by disabling the reloader.
To set a breakpoint in OpenOversight, first import the pdb module by adding import pdb
to the file you want to debug. Call pdb.set_trace()
on its own line wherever you want to break for debugging.
Next, in your terminal run docker ps
to find the container id of the openoversight_web
image, then run docker attach ${container_id}
to connect to the debugger in your terminal. You can now use pdb prompts to step through the app.
If you want to run an individual test in debug mode, use the below command.
docker-compose run --rm web pytest --pdb -v tests/ -k <test_name_here>
where <test_name_here>
is the name of a single test function, such as test_ac_cannot_add_new_officer_not_in_their_dept
Similarly, you can run all the tests in a file by specifying the file path:
docker-compose run --rm web pytest --pdb -v path/to/test/file
where path/to/test/file
is the relative file path, minus the initial OpenOversight
, such as
tests/routes/test_officer_and_department.py
.
Again, add import pdb
to the file you want to debug, then write pdb.set_trace()
wherever you want to drop a breakpoint. Once the test is up and running in your terminal, you can debug it using pdb prompts.