data

Data Directory

This directory houses data that is ingested by taxonate at compile time.

Updating the Supported Languages

Prerequisites

You'll need the following software installed on your machine in order to develop and submit changes to the languages supported by taxonate:

• CUE
• Prettier

Overview

The programming languages that taxonate is able to detect are defined by data in the languages.json file, which should not be edited manually!

To ensure consistency and check for potential conflicts, the language definitions should be written using CUE (Configure Unify Execute) within the languages.cue file. Once you've made an edit to the language definitions, you can evaluate the configuration and export the resulting JSON by running the custom CUE command:

$ cue dump

This will emit a pretty version of the JSON data to STDOUT, and will also save a minified copy, overwriting the existing languages.json file.

To verify that the languages.cue and languages.json files are in sync, you can validate that they match and adhere to the defined constraints by running:

$ cue vet languages.json languages.cue

Schema

A language definition has four parts that can be seen in this example:

"python": {
  "name": "Python",
  "globs": [
    "*.py",
    "*.pyw"
  ],
  "interpreters": [
    "python",
    "python2",
    "python3"
  ]
}

1. key is the map key used to identify a language and is what end users will specify via the --language LANGUAGE command line option. The key must be a valid CUE identifier and will end up being lowercased in the final JSON output.

2. name is the language's human friendly display name and should be capitalized accordingly, as a proper noun.

3. globs is an array containing the common pattern(s) that will match filenames belonging to the language.

4. interpreters is an array containing the language's executable program name(s) that would be specified in a script's shebang line as part of the interpreter directive.

NOTE: globs and interpreters are considered to be filetype "markers" used for identification; a minimum of one marker is required for each language definition.