Automated Section renumber and grammar extraction

[github-actions]

renumber sections. Add grammar

[jskeet]

We should not merge this. The table is deliberately indented as it's part of a list bullet. We may need Smarten changes to handle this...

[BillWagner]

ping @Nigel-Ecma

[BillWagner]

@jskeet @Nigel-Ecma

One option (for the short term) is to turn off the smarten task. We'd continue to update the TOC and the grammar by merging these automated PRs, but a few non-smart quotes would sneak in. As long as we fix the root issue before we submit the final, we should be fine. Thoughts?

[jskeet]

Maybe convert it from auto-run to manually-invoked - then we can run it once in a while (e.g. before submission) and pull/fix/push to undo anything we don't want. I think I'd still like the task itself to be run in a CI way rather than by individuals.

[Nigel-Ecma]

@jskeet @BillWagner – I've just checked in a Markdig fix for HTML blocks in lists. Smarten is again working without roundtrip errors on draft-v6.

[jskeet]

Thanks @Nigel-Ecma!

[BillWagner]

Thanks @Nigel-Ecma

I'll make a PR that updates the smarten tar file.

[BillWagner]

@jskeet

The job runs and updates this branch on each update. Now that smarten isn't part of the "update on merge" workflow, this can be merged.

More importantly, when we merge the better-betterness PR, we'll want to update the TOC.