test(speech): reproduce model installation without bzip2

packages/server/src/server/speech/providers/local/sherpa/model-downloader.test.ts

@@ -8,6 +8,7 @@ import {
   existsSync,
   truncateSync,
   readdirSync,
+  symlinkSync,
 } from "node:fs";
 import { execFileSync } from "node:child_process";
 import { tmpdir } from "node:os";
@@ -23,7 +24,71 @@ function makeTmpDir(): string {
 
 const logger = pino({ level: "silent" });
 
+// A tar.bz2 containing the three Kokoro files ("complete model data") and espeak-ng-data/.
+// Generated with Python tarfile + bz2, so testing extraction needs no system compressor.
+const compressedModelArchive = Buffer.from(
+  "QlpoOTFBWSZTWYGlJLIAAOjfgcqBQAP/ICEAAAD+r99AABAICDAA1mQkopo0P1JkeoMTBDE9RgyiADQZAAAACSVPTU9TT1B6g0aGJ6hkyafQaedKK3SkotKCAUvkgkMWhnICQFABoykPAQBIYAZHteVO8PnEPulpWBWogKIMAi5XOBFWQPCYTsU1U7snVwYKXg/axQ0QgDBKDgsClwZB7z0TLGwJV+3hIFscrlfwje/SLh4QDkqFxkXTaE+x7Zh/DDrAILu8MdygDAzOzD88IBQpa8wkDqd3cEH8XckU4UJCBpSSyA==",
+  "base64",
+);
+
 describe("sherpa model downloader", () => {
+  test.skipIf(process.platform !== "linux")(
+    "installs a cached bzip2 model without a system bzip2 executable",
+    async () => {
+      const root = makeTmpDir();
+      const modelsDir = path.join(root, "models");
+      const spec = getSherpaOnnxModelSpec("kokoro-en-v0_19");
+      const archive = path.join(
+        modelsDir,
+        ".downloads",
+        path.basename(new URL(spec.archiveUrl).pathname),
+      );
+      const originalPath = process.env.PATH;
+      try {
+        const binDir = path.join(root, "bin");
+        mkdirSync(binDir);
+        symlinkSync(
+          execFileSync("which", ["tar"], { encoding: "utf8" }).trim(),
+          path.join(binDir, "tar"),
+        );
+        mkdirSync(path.dirname(archive), { recursive: true });
+        writeFileSync(archive, compressedModelArchive);
+        // GNU tar is available, but the bzip2 executable it invokes is not.
+        process.env.PATH = binDir;
+        const modelDir = await ensureSherpaOnnxModel({ modelsDir, modelId: spec.id, logger });
+        expect(readFileSync(path.join(modelDir, "model.onnx"), "utf8")).toBe("complete model data");
+        expect(existsSync(path.join(modelDir, "espeak-ng-data"))).toBe(true);
+        expect(existsSync(archive)).toBe(false);
+      } finally {
+        if (originalPath === undefined) delete process.env.PATH;
+        else process.env.PATH = originalPath;
+        rmSync(root, { recursive: true, force: true });
+      }
+    },
+  );
+  test.skipIf(process.platform !== "linux")(
+    "installs the same cached bzip2 archive when system tools are available",
+    async () => {
+      const root = makeTmpDir();
+      const modelsDir = path.join(root, "models");
+      const spec = getSherpaOnnxModelSpec("kokoro-en-v0_19");
+      const archive = path.join(
+        modelsDir,
+        ".downloads",
+        path.basename(new URL(spec.archiveUrl).pathname),
+      );
+      try {
+        mkdirSync(path.dirname(archive), { recursive: true });
+        writeFileSync(archive, compressedModelArchive);
+        const modelDir = await ensureSherpaOnnxModel({ modelsDir, modelId: spec.id, logger });
+        expect(readFileSync(path.join(modelDir, "model.onnx"), "utf8")).toBe("complete model data");
+        expect(existsSync(archive)).toBe(false);
+      } finally {
+        rmSync(root, { recursive: true, force: true });
+      }
+    },
+  );
+
   test("does not expose a partial model when tar extraction fails", async () => {
     const root = makeTmpDir();
     const modelsDir = path.join(root, "models");