test: fix deno test helpers

src/apps/cli/testdeno/helpers/cli.ts

@@ -39,27 +39,73 @@ function concatChunks(chunks: Uint8Array[]): Uint8Array {
     return out;
 }
 
+function formatTeeCommand(args: string[]): string {
+    return ["node", CLI_DIST, ...args].map((part) => JSON.stringify(part)).join(" ");
+}
+
+function createLineTeeWriter(
+    pid: number,
+    streamName: "stdout" | "stderr",
+    writer: (chunk: Uint8Array) => void
+): { write: (chunk: Uint8Array) => void; close: () => void } {
+    const enc = new TextEncoder();
+    const dec = new TextDecoder();
+    let pending = "";
+    let headerWritten = false;
+    const emitLine = (line: string) => {
+        if (!headerWritten) {
+            writer(enc.encode(`[CLI tee pid=${pid}:${streamName}]\n`));
+            headerWritten = true;
+        }
+        writer(enc.encode(`[CLI tee pid=${pid}:${streamName}] ${line}\n`));
+    };
+
+    const flush = (final = false) => {
+        let index = pending.indexOf("\n");
+        while (index >= 0) {
+            const line = pending.slice(0, index).replace(/\r$/, "");
+            pending = pending.slice(index + 1);
+            emitLine(line);
+            index = pending.indexOf("\n");
+        }
+        if (final && pending.length > 0) {
+            emitLine(pending.replace(/\r$/, ""));
+            pending = "";
+        }
+    };
+
+    return {
+        write(chunk: Uint8Array) {
+            pending += dec.decode(chunk, { stream: true });
+            flush(false);
+        },
+        close() {
+            pending += dec.decode();
+            flush(true);
+        },
+    };
+}
+
 async function collectStream(
     stream: ReadableStream<Uint8Array>,
-    teeTarget: WritableStream<Uint8Array> | null
+    teeTarget: { write: (chunk: Uint8Array) => void; close: () => void } | null
 ): Promise<Uint8Array> {
     const reader = stream.getReader();
     const chunks: Uint8Array[] = [];
-    const writer = teeTarget?.getWriter();
     try {
         while (true) {
             const { done, value } = await reader.read();
             if (done) break;
             if (value) {
                 chunks.push(value);
-                if (writer) {
+                if (teeTarget) {
-                    await writer.write(value);
+                    teeTarget.write(value);
                 }
             }
         }
     } finally {
-        if (writer) {
+        if (teeTarget) {
-            writer.releaseLock();
+            teeTarget.close();
         }
         reader.releaseLock();
     }
@@ -76,8 +122,18 @@ async function runNodeCommand(args: string[], stdinData?: Uint8Array): Promise<C
         stderr: "piped",
     }).spawn();
 
-    const stdoutPromise = collectStream(child.stdout, TEE_ENABLED ? Deno.stdout.writable : null);
+    if (TEE_ENABLED) {
-    const stderrPromise = collectStream(child.stderr, TEE_ENABLED ? Deno.stderr.writable : null);
+        Deno.stdout.writeSync(new TextEncoder().encode(`[CLI tee pid=${child.pid}] process: ${formatTeeCommand(cliArgs)}\n`));
+    }
+
+    const stdoutPromise = collectStream(
+        child.stdout,
+        TEE_ENABLED ? createLineTeeWriter(child.pid, "stdout", (chunk) => Deno.stdout.writeSync(chunk)) : null
+    );
+    const stderrPromise = collectStream(
+        child.stderr,
+        TEE_ENABLED ? createLineTeeWriter(child.pid, "stderr", (chunk) => Deno.stderr.writeSync(chunk)) : null
+    );
 
     if (stdinData) {
         const w = child.stdin.getWriter();

src/apps/cli/testdeno/helpers/dataset.ts

@@ -0,0 +1,123 @@
+export type DeterministicDatasetConfig = {
+    rootDir: string;
+    datasetDirName: string;
+    seed: string;
+    mdCount: number;
+    mdMinSizeBytes: number;
+    mdMaxSizeBytes: number;
+    binCount: number;
+    binSizeBytes: number;
+};
+
+export type DatasetEntry = {
+    kind: "md" | "bin";
+    relativePath: string;
+    absolutePath: string;
+    size: number;
+};
+
+export type DeterministicDataset = {
+    rootDir: string;
+    datasetDirName: string;
+    seed: string;
+    entries: DatasetEntry[];
+    totalFiles: number;
+    totalBytes: number;
+    mdCount: number;
+    binCount: number;
+};
+
+function fnv1a32(input: string): number {
+    let hash = 0x811c9dc5;
+    for (let i = 0; i < input.length; i++) {
+        hash ^= input.charCodeAt(i) & 0xff;
+        hash = Math.imul(hash, 0x01000193);
+    }
+    return hash >>> 0;
+}
+
+function createXorshift32(seed: number): () => number {
+    let state = seed >>> 0;
+    if (state === 0) {
+        state = 0x9e3779b9;
+    }
+    return () => {
+        state ^= state << 13;
+        state ^= state >>> 17;
+        state ^= state << 5;
+        return state >>> 0;
+    };
+}
+
+function createTextBytes(size: number, fileIndex: number, seed: string): Uint8Array {
+    const template =
+        `# Bench file ${fileIndex}\n` +
+        `seed: ${seed}\n` +
+        "lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.\n";
+
+    const templateBytes = new TextEncoder().encode(template);
+    const out = new Uint8Array(size);
+    for (let i = 0; i < size; i++) {
+        out[i] = templateBytes[i % templateBytes.length];
+    }
+    return out;
+}
+
+function toPath(rootDir: string, relativePath: string): string {
+    return `${rootDir}/${relativePath}`;
+}
+
+export async function createDeterministicDataset(config: DeterministicDatasetConfig): Promise<DeterministicDataset> {
+    if (config.mdCount < 0 || config.binCount < 0) {
+        throw new Error("mdCount and binCount must be non-negative");
+    }
+    if (config.mdMinSizeBytes <= 0 || config.mdMaxSizeBytes <= 0 || config.binSizeBytes <= 0) {
+        throw new Error("all size values must be positive");
+    }
+    if (config.mdMinSizeBytes > config.mdMaxSizeBytes) {
+        throw new Error("mdMinSizeBytes must be <= mdMaxSizeBytes");
+    }
+
+    const datasetRoot = toPath(config.rootDir, config.datasetDirName);
+    const mdDir = `${datasetRoot}/md`;
+    const binDir = `${datasetRoot}/bin`;
+    await Deno.mkdir(mdDir, { recursive: true });
+    await Deno.mkdir(binDir, { recursive: true });
+
+    const nextRandom = createXorshift32(fnv1a32(config.seed));
+    const mdRange = config.mdMaxSizeBytes - config.mdMinSizeBytes + 1;
+    const entries: DatasetEntry[] = [];
+
+    for (let index = 0; index < config.mdCount; index++) {
+        const size = config.mdMinSizeBytes + (nextRandom() % mdRange);
+        const relativePath = `${config.datasetDirName}/md/file-${String(index).padStart(4, "0")}.md`;
+        const absolutePath = toPath(config.rootDir, relativePath);
+        const body = createTextBytes(size, index, config.seed);
+        await Deno.writeFile(absolutePath, body);
+        entries.push({ kind: "md", relativePath, absolutePath, size });
+    }
+
+    for (let index = 0; index < config.binCount; index++) {
+        const size = config.binSizeBytes;
+        const relativePath = `${config.datasetDirName}/bin/file-${String(index).padStart(4, "0")}.bin`;
+        const absolutePath = toPath(config.rootDir, relativePath);
+        const body = new Uint8Array(size);
+        for (let i = 0; i < size; i++) {
+            body[i] = nextRandom() & 0xff;
+        }
+        await Deno.writeFile(absolutePath, body);
+        entries.push({ kind: "bin", relativePath, absolutePath, size });
+    }
+
+    const totalBytes = entries.reduce((sum, e) => sum + e.size, 0);
+    return {
+        rootDir: config.rootDir,
+        datasetDirName: config.datasetDirName,
+        seed: config.seed,
+        entries,
+        totalFiles: entries.length,
+        totalBytes,
+        mdCount: config.mdCount,
+        binCount: config.binCount,
+    };
+}

src/apps/cli/testdeno/helpers/docker.ts

@@ -27,29 +27,53 @@ function parseCommand(command: string): { bin: string; prefix: string[] } {
     return { bin: parts[0], prefix: parts.slice(1) };
 }
 
-async function runCommand(bin: string, args: string[]): Promise<{ code: number; stdout: string; stderr: string }> {
+async function collectStream(
-    const cmd = new Deno.Command(bin, {
+    stream: ReadableStream<Uint8Array>,
-        args,
+    teeTarget: ((chunk: Uint8Array) => void) | null
-        stdin: "null",
+): Promise<Uint8Array> {
-        stdout: "piped",
+    const reader = stream.getReader();
-        stderr: "piped",
+    const chunks: Uint8Array[] = [];
-    });
     try {
-        const { code, stdout, stderr } = await cmd.output();
+        while (true) {
+            const { done, value } = await reader.read();
+            if (done) break;
+            if (!value) continue;
+            chunks.push(value);
+            if (teeTarget) {
+                teeTarget(value);
+            }
+        }
+    } finally {
+        reader.releaseLock();
+    }
+
+    const total = chunks.reduce((sum, chunk) => sum + chunk.length, 0);
+    const out = new Uint8Array(total);
+    let offset = 0;
+    for (const chunk of chunks) {
+        out.set(chunk, offset);
+        offset += chunk.length;
+    }
+    return out;
+}
+
+async function runCommand(bin: string, args: string[]): Promise<{ code: number; stdout: string; stderr: string }> {
+    try {
+        const child = new Deno.Command(bin, {
+            args,
+            stdin: "null",
+            stdout: "piped",
+            stderr: "piped",
+        }).spawn();
+        const stdoutPromise = collectStream(child.stdout, DOCKER_TEE ? (chunk) => Deno.stdout.writeSync(chunk) : null);
+        const stderrPromise = collectStream(child.stderr, DOCKER_TEE ? (chunk) => Deno.stderr.writeSync(chunk) : null);
+        const [status, stdout, stderr] = await Promise.all([child.status, stdoutPromise, stderrPromise]);
         const dec = new TextDecoder();
         const result = {
-            code,
+            code: status.code,
             stdout: dec.decode(stdout),
             stderr: dec.decode(stderr),
         };
-        if (DOCKER_TEE) {
-            if (result.stdout.trim().length > 0) {
-                console.log(`[docker:${bin}] ${result.stdout.trimEnd()}`);
-            }
-            if (result.stderr.trim().length > 0) {
-                console.error(`[docker:${bin}] ${result.stderr.trimEnd()}`);
-            }
-        }
         return result;
     } catch (err) {
         if (err instanceof Deno.errors.NotFound) {