Add Github action for autogenerating table of contents json files

[Blargian]

Summary

Adds a Github action for generating table of contents .json files. These will be used to automatically generate table of contents tables on the landing pages.

Updates the current script to be able to:

• accept arg --dir
• generate one table of contents file for all files in a folder and it's subfolders (eg - as we would need for interfaces/formats with arg --single-toc)
• generate one table of contents per folder (the usual use case, default)

Future improvements:

• Extend the script to be able to accept a single file which will list directories to generate for and whether to do a single toc or one per subdirectory. eg:

/en/cloud/manage
/en/interfaces/formats --single-toc

Our folder structure's aren't great so for the time being we will add folders one by one.

Checklist

• Delete items not relevant to your PR
• URL changes should add a redirect to the old URL via https://github.com/ClickHouse/clickhouse-docs/blob/main/docusaurus.config.js
• If adding a new integration page, also add an entry to the integrations list here: https://github.com/ClickHouse/clickhouse-docs/blob/main/docs/en/integrations/index.mdx

[Blargian]

Notes for self:

So, it works - toc.json is generated. This will work for docs on the ClickHouse/clickhouse-docs repo, we can just commit the added files when the workflow runs. However, the problem is that we have any files from the ClickHouse/ClickHouse repo in .gitignore. So the question now is do we make toc.json which is generated in these folders an exception and commit them to this repo? This is problematic I guess as we copy content across during build - they would get overwritten I assume. On the other hand, we wanted to improve that aspect of the build process and use something like rsync (see #3087)

Alternative solutions:

• copy all table_of_contents files to their own folder maybe named as interfaces_formats_toc.json for example rather than just one toc.json in each folder. I don't like that solution much.
• have blank toc.json files in the ClickHouse/ClickHouse docs folders and just populate them from the action.

update 1: Have ended up going for a solution where we will store the table of contents files in a specific directory on this repo. Still need to figure out how we will generate the links to the files in the table of contents relative to the file which will use toc.json

For example:

• We have docs/en/interfaces/formats and we add to the github action line python -u ./scripts/table-of-contents-generator/toc_gen.py --dir="docs/en/interfaces/formats" --single-toc --out="table-of-contents-files" --ignore "_snippets"
• This outputs docs_en_interfaces_formats_toc.json at the table-of-contents-files directory.
• However to be useful it should contain, as an example, for Arrow some property {"link" : "Arrow/Arrow.md"}. I already added a dir property which would have docs/interfaces/formats/Arrow/Arrow.md so I guess it's a case of just knowing where to cut that string. I could add another arg to the script such that you could pass --base "docs/interfaces/formats" and it could use that to know where to cut the string.