pystardog 0.17.0

Use Stardog with Python!

pystardog

a Python wrapper for communicating with the Stardog HTTP server.

Docs: http://pystardog.readthedocs.io

Requirements: Python 3.8+

What is it?

This library wraps all the functionality of a client for the Stardog Knowledge Graph, and provides access to a full set of functions such as executing SPARQL queries and many administrative tasks.

The implementation uses the HTTP protocol, since most of Stardog functionality is available using this protocol. For more information, see HTTP Programming in Stardog's documentation.

Installation

pystardog is on PyPI. To install:

pip install pystardog

Quick Example

import stardog

conn_details = {
  'endpoint': 'http://localhost:5820',
  'username': 'admin',
  'password': 'admin'
}

with stardog.Admin(**conn_details) as admin:
  db = admin.new_database('db')

  with stardog.Connection('db', **conn_details) as conn:
    conn.begin()
    conn.add(stardog.content.File('./test/data/example.ttl'))
    conn.commit()
    results = conn.select('select * { ?a ?p ?o }')

  db.drop()

Interactive Tutorial

There is a Jupyter notebook and instructions in the notebooks directory of this repository.

Documentation

Documentation is available at http://pystardog.readthedocs.io

Build the Docs Locally

The docs can be built locally using Sphinx:

cd docs
pip install -r requirements.txt
make html

Autodoc Type Hints

The docs use sphinx-autodoc-typehints which allows you to omit types when documenting argument/returns types of functions. For example:

The following function:

def database(self, name: str) -> "Database":
    """Retrieves an object representing a database.

    :param name: The database name

    :return: the database
    """
    return Database(name, self.client)

will yield the following documentation after Sphinx processes it:

Note Only arguments that have an existing :param: directive in the docstring get their respective :type: directives added. The :rtype: directive is added if and only if no existing :rtype: is found. See the docs for additional information on how the extension works.

Auto Build

Docs can be rebuilt automatically when saving a Python file by utilizing sphinx-autobuild

cd docs
pip install -r requirements.txt requirements-dev.txt
make livehtml

This should make the docs available at http://localhost:8000.

Example output after running make livehtml:

❯ make livehtml
sphinx-autobuild "." "_build"   --watch ../stardog/
[sphinx-autobuild] > sphinx-build /Users/frodo/projects/pystardog/docs /Users/frodo/projects/pystardog/docs/_build
Running Sphinx v6.2.1
loading pickled environment... done
building [mo]: targets for 0 po files that are out of date
writing output...
building [html]: targets for 0 source files that are out of date
updating environment: 0 added, 0 changed, 0 removed
reading sources...
looking for now-outdated files... none found
no targets are out of date.
build succeeded.

The HTML pages are in _build.
[I 230710 15:26:18 server:335] Serving on http://127.0.0.1:8000
[I 230710 15:26:18 handlers:62] Start watching changes
[I 230710 15:26:18 handlers:64] Start detecting changes

Contributing and Development

Contrbutions are always welcome to pystardog.

To make a contribution:

1. Create a new branch off of main. There is no set naming convention for branches but try and keep it descriptive.

   git checkout -b feature/add-support-for-X
   

2. Make your changes. If you are making substantive changes to pystardog, tests should be added to ensure your changes are working as expected. See Running Tests for additional information about running tests.

3. Format your code. All Python code should be formatted using Black. See Formatting Your Code for additional information.

4. Commit and push your code. Similar to branch names, there is no set structure for commit messages but try and keep your commit messages succinct and on topic.

   git commit -am "feat: adds support for feature X"
   git push origin feature/add-support-for-x
   

5. Create a pull request against main. All CircleCI checks should be passing in order to merge your PR. CircleCI will run tests against all supported versions of Python, single node and cluster tests for pystardog, as well as do some static analysis of the code.

Running Tests

Requirements:

• Docker
• Docker Compose
• Valid Stardog License

To run the tests locally, a valid Stardog license is required and placed at dockerfiles/stardog-license-key.bin.

1. Bring a stardog instance using docker-compose. For testing about 90% of the pystardog features, just a single node is sufficient, although we also provide a cluster set up for further testing.

   # Bring a single node instance plus a bunch of Virtual Graphs for testing (Recommended).
   docker-compose -f docker-compose.single-node.yml up -d
   
   # A cluster setup is also provided, if cluster only features are to be implemented and tested.
   docker-compose -f docker-compose.cluster.yml up -d
   

2. Create a virtual environment with the necessary dependencies:

   # Create a virtualenv and activate it
   virtualenv -p $(which python3) venv
   source venv/bin/activate
   
   # Install dependencies
   pip install -r requirements.txt -r test-requirements.txt 
   

3. Run the test suite:

   # Run the basic test suite (covers most of the pystardog functionalities)
   pytest test/test_admin_basic.py test/test_connection.py test/test_utils.py -s 
   

   Note Tests can be targeted against a specific Stardog endpoint by specifying an --endpoint option to pytest. Please note, that the tests will make modifications to the Stardog instance like deleting users, roles, databases, etc. By default, the --endpoint is set to http://localhost:5820, which is where the Dockerized Stardog (defined in the Docker compose files) is configured to be available at.

   pytest test/test_connection.py -k test_queries -s --endpoint https://my-other-stardog:5820
   

Formatting your code

To format all the Python code:

# make sure black is install
virtualenv -p $(which python3) venv
. venv/bin/activate
pip install -r test-requirements.txt

# run black formatter
black .