Replace notebooks with pipelines that use SQL Runner

[inglesp]

This is more general issue around notebooks that run on the server, that encompasses some others, including:
opensafely-core/cohort-extractor#830
opensafely-core/cohort-extractor#831

There are several notebooks that are intended to be run on the server, including the database notebooks, and the ones in this repo. Applications and status of notebooks is described in this doc.

The current process for running the database notebooks is described in the team manual, and in the raw-data-plausibility-checks repo README for other notebooks.

We need to

1. Define a common process for running notebooks on the server and how/where to publish outputs
2. Find a way to automate the running of certain notebooks - typically ones that should rerun after a database build (database notebooks, vaccine names notebook, maybe others?). This could possibly be triggered by a database build, if we can identify that - perhaps by job-runner going into maintenance mode? Or if the notebooks are not too heavy-duty, a weekly (or whatever) cron job.

---

Will has started a discussion in #issue-db-notebooks-release and also in this thread.

[inglesp]

Was a draft issue by @rebkwok

[inglesp]

After discussion with @wjchulme, we should start with the TPP schema notebook.

[iaindillingham]

I've broken this issue into several smaller issues. First, some conversion issues:

• Convert database notebooks to use SQL Runner #20, which mirrors Take ownership of database notebooks cohort-extractor#831
• Convert vaccine names notebook to use SQL Runner #22, which mirrors Document vaccine names notebook cohort-extractor#830
• Convert raw data plausibility checks to use SQL Runner #23

Then, some process issues:

• Run database notebooks regularly #21
• I may create similar issues, for vaccine names and raw data plausibility checks 🙂

Then, a governance issue:

• Document use of SQL Runner #16