Gitea: Language statistics hidden languages

[lafriks]

Currently we use languages hard coded in Gitea code. We should probably switch to using go-enry method to get language type and include only programming and markup languages in stats:
https://github.com/go-enry/go-enry/blob/dc6fc02209b37a02b64eed383a5f281059e86eac/common.go#L453

[CirnoT]

Markup should be included too (HTML, CSS)

[silverwind]

FWIW, GitHub does not hide either HTML or CSS from their stats.

[CirnoT]

Yes, hence why should most likely allow Markup along with Programming type.

[silverwind]

Ah, misunderstood you. I wonder if we would have to include more then we currently exclude.

[CirnoT]

No, just two types
https://github.com/go-enry/go-enry/blob/dc6fc02209b37a02b64eed383a5f281059e86eac/common.go#L444

[6543]

@lafriks if we only include programming languages, then https://gitea.com/gitea/blog would be a go project ... - dont know if this is the right way ...

Idear: can we add a reposetting "excluded lang in status: xy, abc, ..."

[silverwind]

Idear: can we add a reposetting "excluded lang in status: xy, abc, ..."

I'd just do it via .gitattributes, that way it's portable and would work both on GitHub and Gitea. The relevant attribute to mark a glob undetectable is -linguist-detectable.

[lafriks]

@6543 no it would be HTML project :) We do already exclude Markdown: https://github.com/go-gitea/gitea/blob/dc812f8ba5bf1c123fa948afed15c4309da8fb45/modules/git/repo_language_stats.go#L33

[lafriks]

first we need .gitattributes parser ;) tho it should probably be part of go-git project

[silverwind]

go-enry indicates in the README that it wants to support .gitattributes, maybe it shoud be there:

https://github.com/go-enry/go-enry#divergences-from-linguist