Store all built-in languages

While we want the CodeQL Action to work with third-party language support, having a list of all built-in languages can help us create better type-level checks to ensure that we don't miss things that we want to customize for each of our built-in languages.

.github/workflows/script/tsconfig.json

@@ -0,0 +1,11 @@
+{
+  "extends": "../../../tsconfig.json",
+  "compilerOptions": {
+    "lib": ["esnext"],
+    "rootDir": "../../..",
+    "sourceMap": false,
+    "noEmit": true,
+  },
+  "include": ["./*.ts"],
+  "exclude": ["node_modules"]
+}

.github/workflows/script/update-builtin-languages.ts

@@ -0,0 +1,88 @@
+#!/usr/bin/env npx tsx
+
+/**
+ * Updates src/languages/builtin.json by querying the CodeQL CLI for:
+ * - Languages that have default queries (via codeql-extractor.yml)
+ * - Language aliases (via `codeql resolve languages --format=betterjson --extractor-include-aliases`)
+ *
+ * Usage:
+ *   npx tsx .github/workflows/script/update-builtin-languages.ts [path-to-codeql]
+ *
+ * If no path is given, falls back to "codeql".
+ */
+
+import { execFileSync } from "node:child_process";
+import * as fs from "node:fs";
+import * as path from "node:path";
+
+import * as yaml from "yaml";
+
+const codeqlPath = process.argv[2] || "codeql";
+
+// Step 1: Resolve all language extractor directories.
+const resolveJson: Record<string, string[]> = JSON.parse(
+  execFileSync(
+    codeqlPath,
+    ["resolve", "languages", "--format=json"],
+    { encoding: "utf8" },
+  ),
+);
+
+// Step 2: For each language, read codeql-extractor.yml and check default_queries.
+const languages: string[] = [];
+
+for (const [language, dirs] of Object.entries(resolveJson)) {
+  const extractorDir = dirs[0];
+  const extractorYmlPath = path.join(extractorDir, "codeql-extractor.yml");
+
+  if (!fs.existsSync(extractorYmlPath)) {
+    throw new Error(`Extractor YAML not found for language '${language}' at expected path: ${extractorYmlPath}`);
+  }
+
+  const extractorYml = yaml.parse(fs.readFileSync(extractorYmlPath, "utf8"));
+  const defaultQueries: unknown[] | undefined = extractorYml.default_queries;
+
+  if (Array.isArray(defaultQueries) && defaultQueries.length > 0) {
+    console.log(`  ✅ ${language}: included (default_queries: ${JSON.stringify(defaultQueries)})`);
+    languages.push(language);
+  } else {
+    console.log(`  ❌ ${language}: excluded (no default queries)`);
+  }
+}
+
+languages.sort();
+
+// Step 3: Resolve aliases, filtered to only those targeting included languages.
+const betterjsonOutput = JSON.parse(
+  execFileSync(
+    codeqlPath,
+    ["resolve", "languages", "--format=betterjson", "--extractor-include-aliases"],
+    { encoding: "utf8" },
+  ),
+);
+
+const languageSet = new Set(languages);
+const aliases: Record<string, string> = Object.fromEntries(
+  Object.entries((betterjsonOutput.aliases ?? {}) as Record<string, string>)
+    .filter(([, target]) => languageSet.has(target))
+    .sort(([a], [b]) => a.localeCompare(b)),
+);
+
+// Step 4: Write builtin.json.
+const outputPath = path.join(
+  __dirname,
+  "..",
+  "..",
+  "..",
+  "src",
+  "languages",
+  "builtin.json",
+);
+
+const content = JSON.stringify({ languages, aliases }, null, 2) + "\n";
+fs.mkdirSync(path.dirname(outputPath), { recursive: true });
+fs.writeFileSync(outputPath, content);
+
+console.log(`\nWrote ${outputPath}`);
+console.log(`  Languages: ${languages.join(", ")}`);
+console.log(`  Aliases: ${Object.keys(aliases).join(", ")}`);

.github/workflows/update-bundle.yml

@@ -63,13 +63,10 @@ jobs:
         with:
           tools: https://github.com/github/codeql-action/releases/download/${{ github.event.release.tag_name }}/codeql-bundle-linux64.tar.gz
 
-      - name: Update language aliases
+      - name: Update built-in languages
+        run: npx tsx .github/workflows/script/update-builtin-languages.ts "$CODEQL_PATH"
         env:
           CODEQL_PATH: ${{ steps.setup-codeql.outputs.codeql-path }}
-        run: |
-          "$CODEQL_PATH" resolve languages --format=betterjson --extractor-include-aliases \
-            | jq -S '.aliases // {}' \
-            > src/known-language-aliases.json
 
       - name: Bump Action minor version if new CodeQL minor version series
         id: bump-action-version