linguist-js 2.5.1 → 2.5.2

package/dist/index.js

@@ -28,7 +28,6 @@ async function analyse(input, opts = {}) {
     const vendorPaths = [...vendorData, ...docData, ...generatedData];
     // Setup main variables
     const fileAssociations = {};
-    const definiteness = {};
     const extensions = {};
     const overrides = {};
     const results = {
@@ -153,6 +152,8 @@ async function analyse(input, opts = {}) {
     };
     const overridesArray = Object.entries(overrides);
     // List all languages that could be associated with a given file
+    const definiteness = {};
+    const fromShebang = {};
     for (const file of files) {
         let firstLine;
         if (useRawContent) {
@@ -174,22 +175,27 @@ async function analyse(input, opts = {}) {
             for (const [lang, data] of Object.entries(langData)) {
                 const langMatcher = (lang) => `\\b${lang.toLowerCase().replace(/\W/g, '\\$&')}(?![\\w#+*]|-\*-)`;
                 // Check for interpreter match
-                const matchesInterpretor = (_f = data.interpreters) === null || _f === void 0 ? void 0 : _f.some(interpreter => firstLine.match(`\\b${interpreter}\\b`));
+                if (opts.checkShebang && hasShebang) {
+                    const matchesInterpretor = (_f = data.interpreters) === null || _f === void 0 ? void 0 : _f.some(interpreter => firstLine.match(`\\b${interpreter}\\b`));
+                    if (matchesInterpretor)
+                        matches.push(lang);
+                }
                 // Check modeline declaration
-                const matchesLang = firstLine.toLowerCase().match(langMatcher(lang));
-                const matchesAlias = (_g = data.aliases) === null || _g === void 0 ? void 0 : _g.some(lang => firstLine.toLowerCase().match(langMatcher(lang)));
-                // Add language
-                const interpretorCheck = opts.checkShebang && matchesInterpretor;
-                const modelineCheck = opts.checkModeline && (matchesLang || matchesAlias);
-                if (interpretorCheck || modelineCheck) {
-                    matches.push(lang);
+                if (opts.checkModeline && hasModeline) {
+                    const modelineText = firstLine.toLowerCase().replace(/^.*-\*-(.+)-\*-.*$/, '$1');
+                    const matchesLang = modelineText.match(langMatcher(lang));
+                    const matchesAlias = (_g = data.aliases) === null || _g === void 0 ? void 0 : _g.some(lang => modelineText.match(langMatcher(lang)));
+                    if (matchesLang || matchesAlias)
+                        matches.push(lang);
                 }
             }
+            // Add identified language(s)
             if (matches.length) {
-                // Add explicitly-identified language
-                const forcedLang = matches[0];
-                addResult(file, forcedLang);
-                definiteness[file] = true;
+                for (const match of matches)
+                    addResult(file, match);
+                if (matches.length === 1)
+                    definiteness[file] = true;
+                fromShebang[file] = true;
                 continue;
             }
         }
@@ -206,22 +212,34 @@ async function analyse(input, opts = {}) {
         }
         // Search each language
         let skipExts = false;
+        // Check if filename is a match
         for (const lang in langData) {
-            // Check if filename is a match
             const matchesName = (_h = langData[lang].filenames) === null || _h === void 0 ? void 0 : _h.some(name => path_1.default.basename(file.toLowerCase()) === name.toLowerCase());
             if (matchesName) {
                 addResult(file, lang);
                 skipExts = true;
             }
         }
+        // Check if extension is a match
+        const possibleExts = [];
         if (!skipExts)
             for (const lang in langData) {
-                // Check if extension is a match
-                const matchesExt = (_j = langData[lang].extensions) === null || _j === void 0 ? void 0 : _j.some(ext => file.toLowerCase().endsWith(ext.toLowerCase()));
-                if (matchesExt) {
-                    addResult(file, lang);
+                const extMatches = (_j = langData[lang].extensions) === null || _j === void 0 ? void 0 : _j.filter(ext => file.toLowerCase().endsWith(ext.toLowerCase()));
+                if (extMatches === null || extMatches === void 0 ? void 0 : extMatches.length) {
+                    for (const ext of extMatches)
+                        possibleExts.push({ ext, lang });
                 }
             }
+        // Apply more specific extension if available
+        const isComplexExt = (ext) => /\..+\./.test(ext);
+        const hasComplexExt = possibleExts.some(data => isComplexExt(data.ext));
+        for (const { ext, lang } of possibleExts) {
+            if (hasComplexExt && !isComplexExt(ext))
+                continue;
+            if (!hasComplexExt && isComplexExt(ext))
+                continue;
+            addResult(file, lang);
+        }
         // Fallback to null if no language matches
         if (!fileAssociations[file]) {
             addResult(file, null);
@@ -247,9 +265,8 @@ async function analyse(input, opts = {}) {
         if (opts.checkHeuristics)
             for (const heuristics of heuristicsData.disambiguations) {
                 // Make sure the extension matches the current file
-                if (!heuristics.extensions.includes(extensions[file])) {
+                if (!fromShebang[file] && !heuristics.extensions.includes(extensions[file]))
                     continue;
-                }
                 // Load heuristic rules
                 for (const heuristic of heuristics.rules) {
                     // Make sure the language is not an array

package/ext/heuristics.yml

@@ -112,7 +112,7 @@ disambiguations:
   - language: Smalltalk
     pattern: '![\w\s]+methodsFor: '
   - language: 'C#'
-    pattern: '^(\s*namespace\s*[\w\.]+\s*{|\s*\/\/)'
+    pattern: '^(\s*namespace\s*[\w\.]+\s*({|;)|\s*\/\/)'
 - extensions: ['.csc']
   rules:
   - language: GSC

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "linguist-js",
-	"version": "2.5.1",
+	"version": "2.5.2",
 	"description": "Analyse languages used in a folder. Powered by GitHub Linguist, although it doesn't need to be installed.",
 	"main": "dist/index.js",
 	"bin": {
@@ -11,7 +11,7 @@
 		"node": ">=12"
 	},
 	"scripts": {
-		"download-files": "npx tsx build/download-files",
+		"download-files": "npx tsx@2 build/download-files",
 		"prepare": "npm run download-files && npm test && npm run perf",
 		"perf": "tsc && node test/perf",
 		"test": "tsc && node test/folder && echo --- && node test/unit"

package/readme.md

@@ -77,7 +77,7 @@ Running LinguistJS on this folder will return the following JSON:
 ### Notes
 
 - File paths in the output use only forward slashes as delimiters, even on Windows.
-- Do not rely on any language classification output from LinguistJS being unchanged between runs.
+- Unless running in offline mode, do not rely on any language classification output from LinguistJS being unchanged between runs.
   Language data is fetched each run from the latest classifications of [`github-linguist`](https://github.com/github/linguist).
   This data is subject to change at any time and may change the results of a run even when using the same version of Linguist.