OpenJustice.be is a non-profit legaltech aiming to open up access to legal knowledge (court decisions, law, doctrine, ...) and instill a true digital-native culture in the world of belgian justice digitalisation.
Upload API offers stateless (ie: no storage or recording of data) operations (Text extraction, pseudonymisation, OCR, ...) to the upload interface from the OpenJustice platform.
If offers a Rest API, documented via OpenAPI, and integrations with more specific functions (Apache Tika, Pyghotess, ...)
This API can already:
- pseudonymise a text
- extract metadata from a pdf file
- OCR process a file in multiple ways
Those functionalities already cover all this component should do, any future work looks to improve what's already offered.
Clone the repository to your directory of choice:
> git clone https://github.com/openjusticebe/anon_api.git
> cd anon_apiOnce in the root directory, you can run the API with docker (for testing) or using poetry (for development).
# docker
> docker build -t "api" ./ && docker run --rm -it -p 5000:5000 api
# poetry
> poetry installPoetry installs a local, isolated python environment, to avoid conflicts with your system's python modules. See here.
A recent python version (>=3.7) will also be needed.
# Run locally in debug mode
> poetry run api --debug
These local URI's provide interaction and documentation for the endpoint
- http://127.0.0.1:5005 for root
- http://127.0.0.1:5005/docs for OpenAPI documentation
See config_default.toml file to create a local config.toml file for the API to use.
Configuration can also happen through environmen variables, see main.py for a list.
# Run locally in debug mode with a config file.
> poetry run api --config config.toml --debug- Provide full support for the OpenJustice upload interactions
Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.
Please make sure to update tests as appropriate.
This API has been built up from a previous incarnation, as such some legacy code is still present, such as testing multiple anonymisation algorithms. See below for info
Algorithms are detailed in the config_default.toml file for the moment.
Parameters are not supported yet. A check is done on the required environment variables to know if the
algorithm is available.
A specific module also has to be added to the anon_api/modules folder, which is responsible for querying and applying the results from the selected algorithm.
This component uses BDD (behave). To run it (a local tika server is needed):
poetry run python -m behave tests/behave