Development

Get the sources

Anonymous:

git clone https://pagure.io/pagure.git

Contributors:

git clone ssh://git@pagure.io/pagure.git

Dependencies

Install the build dependencies of pagure:

sudo dnf install git python-virtualenv libgit2-devel \
                 libjpeg-devel gcc libffi-devel redhat-rpm-config

The python dependencies of pagure are listed in the file requirements.txt at the top level of the sources.

virtualenv pagure_env
source ./pagure_env/bin/activate
pip install pygit2==<version of libgit2 found>.* # e.g. 0.23.*
pip install -r requirements.txt

Note

working in a virtualenv is tricky due to the dependency on pygit2 and thus on libgit2 but the pygit2 documentation has a solution for this.

How to run pagure

There are several options when it comes to a development environment. Vagrant will provide you with a virtual machine which you can develop on, you can use a container to run pagure or you can install it directly on your host machine. The README has detailed instructions for the different options.

Run pagure for development

Adjust the configuration file (secret key, database URL, admin group…) See Configuration for more detailed information about the configuration.

Create the database scheme:

./createdb.py

Create the folder that will receive the different git repositories:

mkdir {repos,docs,forks,tickets,requests,remotes}

Run the server:

./runserver.py

If you want to change some configuration key you can create a file, place the configuration change in it and use it with

./runserver.py -c <config_file>

For example, create the file config with in it:

from datetime import timedelta
# Makes the admin session longer
ADMIN_SESSION_LIFETIME = timedelta(minutes=20000000)

# Use a postgresql database instead of sqlite
DB_URL = 'postgresql://user:pass@localhost/pagure'
# Change the OpenID endpoint
FAS_OPENID_ENDPOINT = 'https://id.stg.fedoraproject.org'

APP_URL = '*'
EVENTSOURCE_SOURCE = 'http://localhost:8080'
EVENTSOURCE_PORT = '8080'
DOC_APP_URL = '*'

# Avoid sending email when developing
EMAIL_SEND = False

and run the server with:

./runserver.py -c config

To get some profiling information you can also run it as:

./runserver.py --profile

You should be able to access the server at http://localhost:5000

Every time you save a file, the project will be automatically restarted so you can see your change immediately.

Create a pull-request for testing

When working on pagure, it is pretty often that one wanted to work on a feature or a bug related to pull-requests needs to create one.

Making a pull-request for development purposes isn’t hard, if you remember that since you’re running a local instance, the git repos created in your pagure instance are also local.

So here are in a few steps that one could perform to create a pull-request in a local pagure instance.

  • Create a project on your pagure instance, let’s say it will be called test

  • Create a folder clones somewhere in your system (you probably do not want it in the repos folder created above, next to it is fine though):

    mkdir clones

  • Clone the repo of the test project into this clones folder and move into it:

    cd clones
git clone ~/path/to/pagure/repos/test.git
cd test

  • Add and commit some files:

    echo "*~" > .gitignore
git add .gitignore
git commit -m "Add a .gitignore file"
echo "BSD" > LICENSE
git add LICENSE
git commit -m "Add a LICENSE file"

  • Push these changes:

    git push -u origin master

  • Create a new branch and add a commit in it:

    git branch new_branch
git checkout new_branch
touch test
git add test
git commit -m "Add file: test"

  • Push this new branch:

    git push -u origin new_branch

Then go back to your pagure instance running in your web-browser, check the test project. You should see two branches: master and new_branch. From there you should be able to open a new pull-request, either from the front page or via the File Pull Request button in the Pull Requests page.

Coding standards

We are trying to make the code PEP8-compliant. There is a flake8 tool that can automatically check your source.

We run the source code through black as part of the tests, so you may have to do some adjustments or run it yourself (which is simple: black /path/to/pagure).

Note

flake8 and black are available in Fedora:

dnf install python3-flake8 python3-black

or

yum install python3-flake8 python3-black

Send patch

The easiest way to work on pagure is to make your own branch in git, make your changes to this branch, commit whenever you want, rebase on master, whenever you need and when you are done, send the patch either by email, via the trac or a pull-request (using git or github).

The workflow would therefore be something like:

git branch <my_shiny_feature>
git checkout <my_shiny_feature>
<work>
git commit file1 file2
<more work>
git commit file3 file4
git checkout master
git pull
git checkout <my_shiny_feature>
git rebase master
git format-patch -2

This will create two patch files that you can send by email to submit in a ticket on pagure, by email or after forking the project on pagure by submitting a pull-request (in which case the last step above git format-patch -2 is not needed.

Note

Though not required, it’s a good idea to begin the commit message with a single short (less than 50 character) line summarizing the change, followed by a blank line and then a more thorough description. The text up to the first blank line in a commit message is treated as the commit title, and that title is used throughout Git. For example, git-format-patch turns a commit into email, and it uses the title on the Subject line and the rest of the commit in the body. Pagure uses lines that contain only ‘Fixes #number’ as references to issues. If for example a commit message of a pagure patch has a line ‘Fixes #3547’ and a pullrequest (PR) gets created in pagure, this PR will be linked to from https://pagure.io/pagure/issue/3547

Unit-tests

Pagure has a number of unit-tests.

We aim at having a full (100%) coverage of the whole code (including the Flask application) and of course a smart coverage as in we want to check that the functions work the way we want but also that they fail when we expect it and the way we expect it.

Tests checking that function are failing when/how we want are as important as tests checking they work the way they are intended to.

So here are a few steps that one could perform to run unit-tests in a local pagure instance.

  • Install the dependencies:

    pip install -r requirements-testing.txt

  • Run it:

    tox ./test/

If you want to run a single interpreter, cou can use:

tox -e py38 ./test/

Each unit-tests files (located under tests/) can be called by alone, allowing easier debugging of the tests. For example:

pytest-3 tests/test_pagure_lib.py

Note

In order to have coverage information you might have to install python-coverage

dnf install python3-pytest-cov

To run the unit-tests, there is also a container available with all the dependencies needed. Use the following command to run the tests

$ ./dev/run-tests-container.py

This command will build a fedora based container and execute the test suite. You can also limit the tests to unit-test files or single tests similar to the options described above. You need set the environment variables REPO and BRANCH if the tests are not yet available in the upstream pagure master branch.