A prototype implementation of the EODH QA Checker concept.
The aim of this prototype is to illustrate the concept of the QA Checker and its interfaces. It is not intended to define a requirement for how the system should be implemented.
Clone the repository, and install from the repository top-level directory with:
$ pip install -e .
The EODH QA Checker should enable the annotation of STAC collections/items in the EODH Catalogue with the results of processes (referred to as QA Check Workflows) intended to verify/validate the quality of datasets.
Each QA Check Workflow provides the means to check a particular aspect of a collection/item’s quality. Each collection then has a defined set of QA Check Workflows that should run against it at defined time intervals or triggered by defined events – to be handled by the EODH Event and Notification Service (ENS) - this is not implemented in this prototype.
QA Check Workflows in this prototype are Python classes, which are subclasses of eodh_qa_prototype.qa_check_workflows.base.BaseQACheckWorkflow. The example repository of QA Check Workflows in the module eodh_qa_prototype.qa_check_workflows.
Collections are assigned a set of QA Check Workflows to run against the collection as a whole, or individual items in the collection. In the prototype this is defined by eodh_qa_prototype/etc/collectionqa.yaml.
Assigned QA check workflows are run by the QA Checker Runner. In the prototype this is the eodh_qa_prototype.runner.QACheckRunner class.
Run the illustrative example with:
$ python example_scripts/example_workflow.py
This runs all the QA Check Workflows assigned to a Sentinel-2 STAC item.