Non-present languages in my list

[Tachytaenius]

Describe the bug

When I preview changes (showing more languages and excluding some more repos) to my language counter on my profile, it's convinced that there are tiny trace amounts of PHP and Pawn, which I have never programmed in before, and C++ which I have no non-forked public repositories of (though if I do a search for C++ files under my user that are not in forks, I get a bunch of assembly-related files (but PHP and Pawn do come up totally empty)).

Expected behavior

PHP and Pawn should not be present, and neither should C++, though I think that is on GitHub's end if it comes up with files that aren't C++ in a C++ search?

Screenshots / Live demo link

Additional context

No response

[qwerty541]

Hey, @Tachytaenius! We use GitHub GraphQL API under the hood to fetch stats data. It show that this repository contains PHP: https://github.com/Tachytaenius/game-boy-n-body. You can personally check the data which we receive from GitHub by running the following query in GraphQL explorer:

query userInfo($login: String!) {
  user(login: $login) {
    # fetch only owner repos & not forks
    repositories(ownerAffiliations: OWNER, isFork: false, first: 100) {
      nodes {
        name
        languages(first: 10, orderBy: {field: SIZE, direction: DESC}) {
          edges {
            size
            node {
              color
              name
            }
          }
        }
      }
    }
  }
}

{ "login":"Tachytaenius" }

If it count some other files as PHP by mistake you can contact GitHub support about that. Our team can't do anything, correctness of https://github.com/github-linguist/linguist is not our responsibility. I can only suggest you to try silence these files using configuration of .gitattributes file.

[rickstaa]

In addition to @qwerty541's response, there's an ongoing need to enhance the language algorithm (refer to the GitHub issue here). It retrieves only the initial 100 repositories, leading to inaccuracies for users with an extensive repository count. A pull request has been initiated to address this issue, and upon approval, we'll release it as an environment variable, enabling usage within GitHub actions.

Apart from that, due to GraphQL limitations, we are currently constrained to fetching language data on a repository level rather than a user basis. I've initiated a feature request with GitHub to enhance this behaviour. Your support for this feature request is crucial in drawing attention to the issue at GitHub and increasing the likelihood of a solution.