Skip to content

Latest commit

 

History

History
220 lines (154 loc) · 12.8 KB

CONTRIBUTING.md

File metadata and controls

220 lines (154 loc) · 12.8 KB
Raw

Contributing

Hi there! We're thrilled that you'd like to contribute to this project. Your help is essential for keeping it great.

Contributions to this project are released to the public under the project's open source license.

This project adheres to the Contributor Covenant Code of Conduct. By participating, you are expected to uphold this code.

The majority of contributions won't need to touch any Ruby code at all.

Dependencies

Linguist is a Ruby library so you will need a recent version of Ruby installed. There are known problems with the macOS/XCode supplied version of Ruby that causes problems installing some of the dependencies. Accordingly, we highly recommend you install a version of Ruby using Homebrew, rbenv, rvm, ruby-build, asdf or other packaging system, before attempting to install Linguist and the dependencies.

Linguist uses the charlock_holmes character encoding detection library which in turn uses ICU, and the libgit2 bindings for Ruby provided by rugged. Bundler v1.10.0 or newer is required for installing the Ruby gem dependencies. Docker is also required when adding or updating grammars. These components have their own dependencies - icu4c, and cmake and pkg-config respectively - which you may need to install before you can install Linguist.

On macOS with Homebrew the instructions below under Getting started will install these dependencies for you.

On Ubuntu:

apt-get install cmake pkg-config libicu-dev docker.io ruby ruby-dev zlib1g-dev build-essential libssl-dev

The latest version of Bundler can be installed with gem install bundler.

Getting started

Before you can start contributing to Linguist, you'll need to set up your environment first. Clone the repo and run script/bootstrap to install its dependencies.

git clone https://github.com/github/linguist.git
cd linguist/
script/bootstrap

To run Linguist from the cloned repository:

bundle exec bin/github-linguist --breakdown

Adding an extension to a language

We try only to add new extensions once they have some usage on GitHub. In most cases we prefer that each new file extension be in use in at least 200 unique :user/:repo repositories before supporting them in Linguist (but see #5756 for a temporary change in the criteria).

To add support for a new extension:

  1. Add your extension to the language entry in languages.yml. Keep the extensions in alphabetical order, sorted case-sensitively (uppercase before lowercase). The exception is the primary extension: it should always be first.
  2. Add at least one sample for your extension to the samples directory in the correct subdirectory. We prefer examples of real-world code showing common usage. The more representative of the structure of the language, the better.
  3. Open a pull request, linking to a GitHub search result showing in-the-wild usage. If you are adding a sample, please state clearly the license covering the code. If possible, link to the original source of the sample. If you wrote the sample specifically for the PR and are happy for it to be included under the MIT license that covers Linguist, you can state this instead.

Additionally, if this extension is already listed in languages.yml and associated with another language, then sometimes a few more steps will need to be taken:

  1. Make sure that example .yourextension files are present in the samples directory for each language that uses .yourextension.
  2. Test the performance of the Bayesian classifier with a relatively large number (1000s) of sample .yourextension files (ping @lildude to help with this). This ensures we're not misclassifying files.
  3. If the Bayesian classifier does a bad job with the sample .yourextension files then a heuristic may need to be written to help.

See My Linguist PR has been merged but GitHub doesn't reflect my changes for details on when your changes will appear on GitHub after your PR has been merged.

Adding a language

We try only to add languages once they have some usage on GitHub. In most cases we prefer that each new file extension be in use in at least 200 unique :user/:repo repositories before supporting them in Linguist (but see #5756 for a temporary change in the criteria).

To add support for a new language:

  1. Add an entry for your language to languages.yml. Omit the language_id field for now.
  2. Add a syntax-highlighting grammar for your language using: 
    script/add-grammar https://github.com/JaneSmith/MyGrammar
    This command will analyze the grammar and, if no problems are found, add it to the repository. If problems are found, please report them to the grammar maintainer as you will otherwise be unable to add it. Please only add grammars that have one of these licenses.
  3. Add samples for your language to the samples directory in the correct subdirectory.
  4. Generate a unique ID for your language by running script/update-ids.
  5. Open a pull request, linking to GitHub search results showing in-the-wild usage. Please state clearly the license covering the code in the samples. Link directly to the original source if possible. If you wrote the sample specifically for the PR and are happy for it to be included under the MIT license that covers Linguist, you can state this instead.

In addition, if your new language defines an extension that's already listed in languages.yml (such as .foo) then sometimes a few more steps will need to be taken:

  1. Make sure that example .foo files are present in the samples directory for each language that uses .foo.
  2. Test the performance of the Bayesian classifier with a relatively large number (1000s) of sample .foo files (ping @lildude to help with this). This ensures we're not misclassifying files.
  3. If the Bayesian classifier does a bad job with the sample .foo files, then a heuristic may need to be written to help.

Remember, the goal here is to try and avoid false positives!

See My Linguist PR has been merged but GitHub doesn't reflect my changes for details on when your changes will appear on GitHub after your PR has been merged.

Fixing a misclassified language

Most languages are detected by their file extension defined in languages.yml. For disambiguating between files with common extensions, Linguist applies some heuristics and a statistical classifier. This process can help differentiate between, for example, .h files which could be either C, C++, or Obj-C.

Misclassifications can often be solved by either adding a new filename or extension for the language or adding more samples to make the classifier smarter.

Fixing syntax highlighting

Syntax highlighting in GitHub is performed using TextMate-compatible grammars. These are the same grammars used by TextMate, Sublime Text, and Atom. Every language in languages.yml is mapped to its corresponding TextMate scopeName. This scope name will be used when picking up a grammar for highlighting.

Assuming your code is being detected as the right language, in most cases syntax highlighting problems are due to a bug in the language grammar rather than a bug in Linguist. vendor/README.md lists all the grammars we use for syntax highlighting on GitHub.com. Find the one corresponding to your code's programming language and submit a bug report upstream. If you can, try to reproduce the highlighting problem in the text editor that the grammar is designed for (TextMate, Sublime Text, or Atom) and include that information in your bug report.

You can also try to fix the bug yourself and submit a pull-request. TextMate's documentation offers a good introduction on how to work with TextMate-compatible grammars. Note that Linguist uses PCRE regular expressions, while TextMate uses Oniguruma. Although they are mostly compatible there might be some differences in syntax and semantics between the two. Linguist's grammar compiler will highlight any problems when the grammar is updated.

Once the bug has been fixed upstream, we'll pick it up for GitHub in the next release of Linguist.

See My Linguist PR has been merged but GitHub doesn't reflect my changes for details on when the upstream changes will appear on GitHub.

Changing the source of a syntax highlighting grammar

We'd like to ensure Linguist and GitHub.com are using the latest and greatest grammars that are consistent with the current usage but understand that sometimes a grammar can lag behind the evolution of a language or even stop being developed. This often results in someone grasping the opportunity to create a newer and better and more actively maintained grammar, and we'd love to use it and pass on its functionality to our users.

Switching the source of a grammar is really easy:

script/add-grammar --replace MyGrammar https://github.com/PeterPan/MyGrammar

This command will analyze the grammar and, if no problems are found, add it to the repository. If problems are found, please report these problems to the grammar maintainer as you will not be able to add the grammar if problems are found.

Please only add grammars that have one of these licenses.

Please then open a pull request for the updated grammar.

See My Linguist PR has been merged but GitHub doesn't reflect my changes for details on when your changes will appear on GitHub after your PR has been merged.

Changing the color associated with a language

Many of the colors associated with the languages within Linguist have been in place for a very long time. The colors were often chosen based on the colors used by the language at the time and since then users will have become familiar with those colors as they appear on GitHub.com. If you would like to change the color of a language, we ask that you propose your suggested color change to the wider community for your language to gain consensus before submitting a pull request. Please do this in a community forum or repository used and known by the wider community of that language, not the Linguist repository.

Once you've received consensus that the community is happy with your proposed color change, please feel free to open a PR making the change and link to the public discussion where this was agreed by the community. If there are official branding guidelines to support the colour choice, please link to those too.

Testing

You can run the tests locally with:

bundle exec rake test

You can test the classifier locally with:

bundle exec script/cross-validation --test

Sometimes getting the tests running can be too much work, especially if you don't have much Ruby experience. It's okay: be lazy and let GitHub Actions run the tests for you. Just open a pull request and the bot will start cranking away.

Here's our current build status: Actions Status

Maintainers

Linguist is maintained with ❤️ by:

  • @Alhadis
  • @larsbrinkhoff
  • @lildude (GitHub staff)
  • @pchaigno

As Linguist is a production dependency for GitHub we have a couple of workflow restrictions:

  • Anyone with commit rights can merge Pull Requests provided that there is a 👍 from a GitHub staff member.
  • Releases are performed by GitHub staff so we can ensure GitHub.com always stays up to date with the latest release of Linguist and there are no regressions in production.