Reduce friction to Behaviour Driven Development (BDD) through scripting

As discussed in the last TCoP https://confluence.skatelescope.org/display/SE/2021-02-09+-+Testing+CoP+Meeting and is being described in the "BDD testing guide" in the developer portal, we will need to create tools that reduce the friction in creating and managing BDD tests.

In particular we'd like:

• for a developer to be able to write the Gherkin test scenario in VSCode, store them into git, develop their automation, and finally upload them into Jira XTP project so that everybody can search/see them in a single location (Jira XTP project);
• for a tester (such as AIV engineers) to be able to write gherkin tests in Jira/xray, (this can already be done)
• for a tester to be able to identify a test or a test set, export them into a git repo, somehow manage development of their automation or reuse of existing automation, run them directly or have the CI pipeline run them, manually or automatically upload the test results into Jira/Xray.

To do so we need scripts to manage interaction with Jira/Xray, and in particular:

• a script to download tests of a test set, exported as .feature files; each test is labeled with a Jira key. (at the moment this is not much work to do manually: GB: true, but we need to automate whatever is worthwhile doing it, and I think this is easy to automate)
• a script to upload a test or a .feature file into Jira and update the Jira representation of the test; if the tests do not exist then new Jira issues need to be created, otherwise existing ones need to be updated. (which is master: Jira or git .feature file? GB: I would choose the git representation)
• a script to upload test executions in a format that is compatible with what can be produced by pytest-bdd and what can be accepted by Xray APIs (see https://docs.getxray.app/display/XRAYCLOUD/REST+API). Such an upload has to link the results of execution of each test with the test issue in Jira. If it is missing it needs to be created, possibly empty.

Ideally, as a bonus, the scripts would do a synchronization of information, not just an upload or download. Synchronization in the sense that the most recent version wins (for example, instead of uploading a test scenario and blindly overwrite the Jira representation, the script might check the modification time of both representations, identify the most recent one, and make sure that that is the one being replicated in both locations).

We also need a mechanism that automatically removes test executions that are older than a given number of days. This could be a scritp run by cron or something else. There should be somebody who is responsible for managing such a retention policy.