linguist-js/dist/analyser/classifiers/byModeline.js

Analyse the programming languages used in a folder or from raw content, using the same rules that GitHub Linguist does.

const modelineRegex = /-\*-|(?:syntax|filetype|ft)\s*=/;
const escapeRegExp = (value) => value.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
export default function byModeline(file, langData, opts) {
    if (!opts.checkModeline || opts.quick)
        return [];
    const firstLine = file.firstLine;
    // Check modeline declaration
    if (!firstLine || !modelineRegex.test(firstLine))
        return [];
    const modelineText = firstLine.toLowerCase().split(modelineRegex)[1] ?? '';
    // Add identified language(s)
    return Object.entries(langData).flatMap(([lang, data]) => {
        const safeLang = escapeRegExp(lang.toLowerCase());
        const matchesLang = new RegExp(`\\b${safeLang}(?![\\w#+*]|-\\*-)`).test(modelineText);
        const aliases = data.aliases ?? [];
        const matchesAlias = aliases.some((alias) => {
            const safeAlias = escapeRegExp(alias.toLowerCase());
            return new RegExp(`\\b${safeAlias}(?![\\w#+*]|-\\*-)`).test(modelineText);
        });
        return matchesLang || matchesAlias ? [lang] : [];
    });
}