search

github-linguist / linguist

Language Savant. If your repository's language is being reported incorrectly, send us a pull request!
MIT License
12.33k stars 4.27k forks source link

consider defining TeX / LaTeX a "programming language" #516

Closed davidwhogg closed 11 years ago

davidwhogg commented 11 years ago

While I recognize that TeX is a kind of "markup" language, many of us have repos that are nearly all LaTeX (or TeX), and they get classified (on their top page on github) randomly to any trace "programming" language that linguist can find. (For instance, https://github.com/davidwhogg/DataAnalysisRecipes gets classified "Python".)

I, for one, welcome one of two possible github-overlord moves:

This issue came up in evaluation of @dfm's "open-source report card", but really it is a persistent issue if github wants to be a repository of all things open, not just code but also documents.

In a (lovely, frankly) web conversation with @xpaulbettsx he brought up the problem that LaTeX can't be rendered by github because it can shell out. Maybe that is, itself, a reason we ought to classify it as a "programming" language? Just grasping at straws! :nyantocat:

dfm commented 11 years ago

I think that this is a great idea!

+1

(P.S. sweet nyantocat)

aaronpuchert commented 11 years ago

The same goes for PostScript. Maybe one should consider adding a class between programming and markup for these. Most TeX files aren't real programs, but neither pure data like HTML: TeX is Turing-complete, and you can write real programs in it. However, the second solution seems to be the way to go, since there are many "data-only" repositories. Making TeX a programming language would only solve this issue for TeX repos. And I don't think this is a good idea at all, because most TeX files are indeed no programs. The question is how this would distort the statistics for existing ("programming") repositories. (consider projects with extensive TeX documentation)

hai-nguyen-van commented 11 years ago

It has been shown that LaTeX is Turing-complete (http://tug.org/TUGboat/Articles/tb11-3/tb29greene.pdf). Thus, it satisfies all properties of current programming languages : equivalent to any partial recursive function, universal machine existence, recursion theorem and s-m-n theorem. Even if it's main goal is not programming it is able to compute as well as Church's conjecture.

tnm commented 11 years ago

Although we're going to leave TeX as markup, I'm comfortable adding detection for it: https://github.com/github/linguist/commit/8b8123a3c1251fdae7e6c2d704bcd77bdfdbeadd.

Thanks.

davidwhogg commented 10 years ago

thanks for doing this. Love @github