If you want to contribute code to xahau-py
, the following sections describe how to set up your developer environment.
To make it easy to manage your Python environment with xahau-py
, including switching between versions, install pyenv
and follow these steps:
-
Install
pyenv
:brew install pyenv
For other installation options, see the
pyenv
README. -
Use
pyenv
to install the optimized version forxahau-py
(currently 3.11.6):pyenv install 3.11.6
-
Set the global version of Python with
pyenv
:pyenv global 3.11.6
To enable autocompletion and other functionality from your shell, add pyenv
to your environment.
These steps assume that you're using a Zsh shell. For other shells, see the pyenv
README.
-
Add
pyenv init
to your Zsh shell:echo -e 'if command -v pyenv 1>/dev/null 2>&1; then\n eval "$(pyenv init -)"\nfi' >> ~/.zshrc
-
Source or restart your terminal:
. ~/.zshrc
To simplify managing library dependencies and the virtual environment, xahau-py
uses poetry
.
-
curl -sSL https://install.python-poetry.org | python3 -
May need to run export PATH="$HOME/.local/bin:$PATH"
to invoke Poetry (see step 3 "Add Poetry to your PATH").
-
Install
poetry
dependencies from pyproject.toml:poetry install
To run linting and other checks, xahau-py
uses pre-commit
.
Note: You only need to install pre-commit
if you want to contribute code to xahau-py
.
-
Install
pre-commit
:pip3 install pre-commit pre-commit install
To run the linter:
poetry run flake8 xahau tests --darglint-ignore-regex="^_(.*)"
# Works for single or multiple unit/integration tests
# Ex: poetry run poe test tests/unit/models/test_response.py tests/integration/transactions/test_account_delete.py
poetry run poe test FILE_PATHS
poetry run poe test_unit
To run integration tests, you'll need a standalone xahaud node running with WS port 6006
and JSON RPC port 5005
. You can run a docker container for this:
docker run -p 5005:5005 -p 6006:6006 --interactive -t --platform linux/amd64 xahauci/xahaud:2024.9.11
Breaking down the command:
docker run -p 5005:5005 -p 6006:6006
starts a Docker container with an open port for admin JsonRPC and WebSocket requests.--interactive
allows you to interact with the container.-t
starts a terminal in the container for you to send commands to.xahauci/xahaud:2024.9.11
is an image that is regularly updated with the latestxahaud
releases Then to actually run the tests, run the command:
poetry run poe test_integration
To run both unit and integration tests and see code coverage:
poetry run poe test_coverage
To see manually code coverage after running unit tests or integration tests:
poetry run coverage report
To switch your python version before running tests:
pyenv local 3.11
poetry env use python3.11
poetry install
Replace python3.11
with whatever version of Python you want to use (you must have it installed with pyenv
for it to work).
You can see the complete reference documentation at xahau-py
docs. You can also generate them locally using poetry
and sphinx
:
# Go to the docs/ folder
cd docs/
# Build the docs
poetry run sphinx-apidoc -o source/ ../xahau
poetry run make html
To see the output:
# Go to docs/_build/html/
cd docs/_build/html/
# Open the index file to view it in a browser:
open index.html
You can view docs builds for xahau-py versions on the ReadTheDocs website here: https://readthedocs.org/projects/xahau-py/builds/
In order to test how a change in docs configuration looks like on ReadTheDocs before merging:
- Publish a branch with your docs configuration changes
- Active and hide the branch by scrolling down on this page: https://readthedocs.org/projects/xahau-py/versions/
- View the page / build results here: https://readthedocs.org/projects/xahau-py/builds/
- Once you're done testing, make the test branch inactive.
- If adding functionality to a new part of the library, create new file with a class that inherits
IntegrationTestCase
fromtests.integration.integration_test_case
to store all individual tests under (ex:class TestWallet(IntegrationTestCase)
). Otherwise, add to an existing file. - Create an async function for each test case (unless the test is only being used for the sync client)
- Include the
@test_async_and_sync
decorator to test against all client types, unless you specifically only want to test with one client. You can also use the decorator to:- Limit tests to sync/async only
- Limit the number of retries
- Use Testnet instead of a standalone network
- Import modules for sync equivalents of any async functions used
- Be sure to reuse pre-made values,
WALLET
,DESTINATION
,TESTNET_WALLET
,TESTNET_DESTINATION
,OFFER
, andPAYMENT_CHANNEL
, fromtests/integrations/reusable_values.py
- Be sure to use condensed functions, like
submit_transaction_async
andsign_and_reliable_submission_async
, fromtests/integrations/it_utils.py
Examples can be found in subfolders of tests/integrations
This should almost always be done using the xrpl-codec-gen
script - if the output needs manual intervention afterwards, consider updating the script instead.
- Clone / pull the latest changes from xahaud - Specifically the
develop
branch is usually the right one. - Clone / pull the latest changes from
xrpl-codec-gen
- From the
xrpl-codec-gen
tool, follow the steps in theREADME.md
to generate a newdefinitions.json
file. - Replace the
definitions.json
file in thexahau-binary-codec
with the newly generated file. - Verify that the changes make sense by inspection before submitting, as there may be updates required for the
xrpl-codec-gen
tool depending on the latest amendments we're updating to match.
- Your changes should have unit and/or integration tests.
- Your changes should pass the linter.
- Your code should pass all the unit and integration tests on Github (which check all versions of Python).
- Open a PR against
main
and ensure that all CI passes. - Get a full code review from one of the maintainers.
- Merge your changes.
- Create a branch off main that properly increments the version in
pyproject.toml
and updates theCHANGELOG
appropriately. We follow Semantic Versioning. - Merge this branch into
main
. - Locally build and download the package.
- Pull main locally.
- Run
poetry build
to build the package locally. - Locally download the package by running
pip install path/to/local/xahau-py/dist/.whl
. - Make sure that this local installation works as intended, and that the changes are reflected properly.
- Run
poetry publish --dry-run
and make sure everything looks good. - Publish the update by running
poetry publish
.- This will require entering PyPI login info.
- Create a new Github release/tag off of this branch.
- Send an email to xahau-announce.
- Post an announcement in the XRPL Discord #python channel with a link to the changes and highlighting key changes.
We have a low-traffic mailing list for announcements of new xahau-py
releases. (About 1 email every couple of weeks)
If you're using the XRP Ledger in production, you should run a xahaud server and subscribe to the xahau-server mailing list as well.