This project contains the Dockerfile and images assets required to create the accessibility-assessment image that is used in CI for the assessment of captured pages using axe, pa11y and vnu. The violations found during the assessment are pushed to ELK.
You will need to install docker to proceed with this guide.
The following instructions give reviewers of the point in time accessibility audit early sight of the accessibility assessment findings. Work is ongoing to implement this functionality in the build.
- Build the accessibility-assessment image from the docker directory:
docker build -t accessibility-assessment:1.0.0 .- Start your local ELK stack from the apps directory:
docker-compose up -dGo to http://localhost:5601 in your browser and wait for kibana to initialise
-
In Kibana, navigate to Management -> Saved Ojbects and click Import. Load the Index Saved Object and the visualisations and dashboards. This file contains +700 saved objects so may take a few minutes to load.
-
Ensure that you have the
JENKINS_USERNAMEandJENKINS_API_KEYenvironment variables set in your local development environment. These values are used to retrieve a given test suite's pages from the jenkins api. -
Pick the test suite job name that you'd like to assess from jenkins(?) and execute:
./assess-test-suite.sh your-ui-test-job-nameNOTE: Running time will depend on the number of unique pages that were captured during the UI test run. As a rough guide, expect the assessment to take approximately 4 seconds per page.
Execute the following command:
docker-compose killAll of the data loaded into ElasticSearch during the assessment will be persisted in apps/esdata/ for when you next start your local ELK stack.
As mentioned in the previous section, all alerts and saved objects will be written to disk in apps/esdata/.
If you wish to clear down Elastic Search, simply delete everything in this directory. I.e. rm -r apps/esdata/*
This code is open source software licensed under the Apache 2.0 License.