mirror of
https://github.com/QwenLM/qwen-code.git
synced 2026-11-07 06:19:43 -10:00
* fix(test): stop measuring model latency anywhere under CI (#11271) The prompt-latency probe in qwen-serve-baseline sends 20 real prompts through the shared OpenAI-compatible gateway and asserts p99 < 60s. One gateway-queued prompt in twenty fails it, and each vitest retry re-issues all 20 prompts into the same degraded window, so a single slow window fails every attempt and turns the shard red. Since the pool skip (#11004) the probe runs on exactly one E2E leg — macOS shard 2/2, the only shard holding the file — and that leg failed twice within six hours on unrelated commits (runs 34070970091 and 34088422718, both full-duration with retries consumed) while every Linux leg, macOS shard 1/2, and the OpenTUI leg stayed green. The measured quantity is gateway weather, not the daemon; the same argument the pool skip made for host contention applies to the shared gateway. Skip the probe whenever CI is set (any populated CI marker, matching the repo's other CI checks), keeping the self-hosted disjunct so a pool-shaped shell outside CI keeps its specific skip reason. The probe still runs off CI on a credential, and QWEN_BASELINE_ENABLE_PROMPT_LATENCY=1 force-runs it anywhere. The skip reason distinguishes CI gateway contention from pool host contention. * fix(test): tolerate transient model-serving errors in the acp plan-mode case (#11271) The CI-wide prompt-latency skip rested on a misattribution: in both cited runs the latency probe ran and passed, and the macOS leg went red on cli/acp-integration.test.ts > blocks write tools in plan mode — killed by a transient model-serving -32603 returned over ACP, with the retry attempt then tripping ENOTEMPTY on the previous attempt's leftover .qwen-home. Revert the CI skip, re-issue the plan-mode session/prompt on the observed gateway -32603 shape (bounded at 3 attempts, 2s apart, inside the existing per-request timeout), and let the TestRig.setup directory reset retry through the mid-delete refill race the way globalSetup's teardown already does. * test(e2e): isolate protocol checks from model service * fix(test): surface fake-server setup failures in json-output teardown (#11271) Co-authored-by: Qwen-Coder <[email protected]> * fix(test): close ACP teardown race and tighten fake-server E2E (#11271) - Move per-agent QWEN_HOME out of rig.testDir: the agent keeps writing there ~300ms after exit, racing global teardown's recursive rm (ENOTEMPTY). cleanup() now removes it with retries. - Drop the plan-mode test's permissionHandler: the plan-mode guard short-circuits before any permission request, so it could never run and its comment claimed coverage the test does not have. - Name the collected tool-call events in the plan-mode assertion so a wire-shape drift is distinguishable from a plan-mode let-through. - Stub loopback NO_PROXY in json-output tests: an inherited HTTP(S)_PROXY otherwise tunnels the fake-server POST and the cases time out. Two-chunk fixture restores multi-delta accumulator coverage. - Make the auth-mismatch case hermetic via fakeModelArgs instead of the ambient OPENAI_* secrets. Co-authored-by: Qwen-Coder <[email protected]> * fix(test): reclaim leaked ACP qwen homes and unify scratch teardown (#11271) --------- Co-authored-by: yiliang114 <[email protected]> Co-authored-by: 易良 <[email protected]> Co-authored-by: Qwen-Coder <[email protected]>
348 lines
10 KiB
TypeScript
348 lines
10 KiB
TypeScript
/**
|
|
* @license
|
|
* Copyright 2025 Google LLC
|
|
* SPDX-License-Identifier: Apache-2.0
|
|
*/
|
|
|
|
import { expect, describe, it, beforeEach, afterEach, vi } from 'vitest';
|
|
import {
|
|
CONTAINER_SANDBOX_NO_PROXY,
|
|
fakeServerHostOptions,
|
|
IS_CONTAINER_SANDBOX,
|
|
TestRig,
|
|
} from '../test-helper.js';
|
|
import {
|
|
startFakeOpenAIServer,
|
|
type FakeOpenAIServer,
|
|
} from '../fake-openai-server.js';
|
|
|
|
describe('JSON output', () => {
|
|
let rig: TestRig;
|
|
let fakeServer: FakeOpenAIServer;
|
|
|
|
beforeEach(async () => {
|
|
rig = new TestRig();
|
|
// The spawned CLI routes through any inherited HTTP(S)_PROXY; without a
|
|
// loopback NO_PROXY entry the fake-server POST is tunnelled and these
|
|
// tests hang. Mirrors test-helper.ts:993-995.
|
|
const noProxy = IS_CONTAINER_SANDBOX
|
|
? CONTAINER_SANDBOX_NO_PROXY
|
|
: '127.0.0.1,localhost';
|
|
vi.stubEnv('NO_PROXY', noProxy);
|
|
vi.stubEnv('no_proxy', noProxy);
|
|
fakeServer = await startFakeOpenAIServer(
|
|
({ body }) =>
|
|
body['stream'] === true
|
|
? { contentChunks: ['Par', 'is'] }
|
|
: { content: '{"selected_memories":[]}' },
|
|
fakeServerHostOptions(),
|
|
);
|
|
await rig.setup('json-output-test');
|
|
});
|
|
|
|
afterEach(async () => {
|
|
await fakeServer?.close();
|
|
vi.unstubAllEnvs();
|
|
await rig.cleanup();
|
|
});
|
|
|
|
const fakeModelArgs = () => [
|
|
'--auth-type',
|
|
'openai',
|
|
'--model',
|
|
'fake-model',
|
|
'--openai-base-url',
|
|
fakeServer.baseUrl,
|
|
'--openai-api-key',
|
|
'fake-key',
|
|
];
|
|
|
|
it('should return a valid JSON array with result message containing response and stats', async () => {
|
|
const result = await rig.run(
|
|
'What is the capital of France?',
|
|
'--output-format',
|
|
'json',
|
|
...fakeModelArgs(),
|
|
);
|
|
const parsed = JSON.parse(result);
|
|
|
|
// The output should be an array of messages
|
|
expect(Array.isArray(parsed)).toBe(true);
|
|
expect(parsed.length).toBeGreaterThan(0);
|
|
|
|
// Find the result message (should be the last message)
|
|
const resultMessage = parsed.find(
|
|
(msg: unknown) =>
|
|
typeof msg === 'object' &&
|
|
msg !== null &&
|
|
'type' in msg &&
|
|
msg.type === 'result',
|
|
);
|
|
|
|
expect(resultMessage).toBeDefined();
|
|
expect(resultMessage).toHaveProperty('is_error');
|
|
expect(resultMessage.is_error).toBe(false);
|
|
expect(resultMessage).toHaveProperty('result');
|
|
expect(typeof resultMessage.result).toBe('string');
|
|
expect(resultMessage.result.toLowerCase()).toContain('paris');
|
|
|
|
// Stats may be present if available
|
|
if ('stats' in resultMessage) {
|
|
expect(typeof resultMessage.stats).toBe('object');
|
|
}
|
|
});
|
|
|
|
it('should return line-delimited JSON messages for stream-json output format', async () => {
|
|
const result = await rig.run(
|
|
'What is the capital of France?',
|
|
'--output-format',
|
|
'stream-json',
|
|
...fakeModelArgs(),
|
|
);
|
|
|
|
// Stream-json output is line-delimited JSON (one JSON object per line)
|
|
const lines = result
|
|
.trim()
|
|
.split('\n')
|
|
.filter((line) => line.trim());
|
|
expect(lines.length).toBeGreaterThan(0);
|
|
|
|
// Parse each line as a JSON object
|
|
const messages: unknown[] = [];
|
|
for (const line of lines) {
|
|
try {
|
|
const parsed = JSON.parse(line);
|
|
messages.push(parsed);
|
|
} catch (parseError) {
|
|
throw new Error(
|
|
`Failed to parse JSON line: ${line}. Error: ${parseError}`,
|
|
);
|
|
}
|
|
}
|
|
|
|
// Should have at least system, assistant, and result messages
|
|
expect(messages.length).toBeGreaterThanOrEqual(3);
|
|
|
|
// Find system message
|
|
const systemMessage = messages.find(
|
|
(msg: unknown) =>
|
|
typeof msg === 'object' &&
|
|
msg !== null &&
|
|
'type' in msg &&
|
|
msg.type === 'system',
|
|
);
|
|
expect(systemMessage).toBeDefined();
|
|
expect(systemMessage).toHaveProperty('subtype');
|
|
expect(systemMessage).toHaveProperty('session_id');
|
|
|
|
// Find assistant message
|
|
const assistantMessage = messages.find(
|
|
(msg: unknown) =>
|
|
typeof msg === 'object' &&
|
|
msg !== null &&
|
|
'type' in msg &&
|
|
msg.type === 'assistant',
|
|
);
|
|
expect(assistantMessage).toBeDefined();
|
|
expect(assistantMessage).toHaveProperty('message');
|
|
expect(assistantMessage).toHaveProperty('session_id');
|
|
|
|
// Find result message (should be the last message)
|
|
const resultMessage = messages[messages.length - 1] as {
|
|
type: string;
|
|
is_error: boolean;
|
|
result: string;
|
|
};
|
|
expect(resultMessage).toBeDefined();
|
|
expect(
|
|
typeof resultMessage === 'object' &&
|
|
resultMessage !== null &&
|
|
'type' in resultMessage &&
|
|
resultMessage.type === 'result',
|
|
).toBe(true);
|
|
expect(resultMessage).toHaveProperty('is_error');
|
|
expect(resultMessage.is_error).toBe(false);
|
|
expect(resultMessage).toHaveProperty('result');
|
|
expect(typeof resultMessage.result).toBe('string');
|
|
expect(resultMessage.result.toLowerCase()).toContain('paris');
|
|
});
|
|
|
|
it('should include stream events when using stream-json with include-partial-messages', async () => {
|
|
const result = await rig.run(
|
|
'What is the capital of France?',
|
|
'--output-format',
|
|
'stream-json',
|
|
'--include-partial-messages',
|
|
...fakeModelArgs(),
|
|
);
|
|
|
|
// Stream-json output is line-delimited JSON (one JSON object per line)
|
|
const lines = result
|
|
.trim()
|
|
.split('\n')
|
|
.filter((line) => line.trim());
|
|
expect(lines.length).toBeGreaterThan(0);
|
|
|
|
// Parse each line as a JSON object
|
|
const messages: unknown[] = [];
|
|
for (const line of lines) {
|
|
try {
|
|
const parsed = JSON.parse(line);
|
|
messages.push(parsed);
|
|
} catch (parseError) {
|
|
throw new Error(
|
|
`Failed to parse JSON line: ${line}. Error: ${parseError}`,
|
|
);
|
|
}
|
|
}
|
|
|
|
// Should have more messages than without include-partial-messages
|
|
// because we're including stream events
|
|
expect(messages.length).toBeGreaterThan(3);
|
|
|
|
// Find stream_event messages
|
|
const streamEvents = messages.filter(
|
|
(msg: unknown) =>
|
|
typeof msg === 'object' &&
|
|
msg !== null &&
|
|
'type' in msg &&
|
|
msg.type === 'stream_event',
|
|
);
|
|
expect(streamEvents.length).toBeGreaterThan(0);
|
|
|
|
// Verify stream event structure
|
|
const firstStreamEvent = streamEvents[0];
|
|
expect(firstStreamEvent).toHaveProperty('event');
|
|
expect(firstStreamEvent).toHaveProperty('session_id');
|
|
expect(firstStreamEvent).toHaveProperty('uuid');
|
|
|
|
// Check for expected stream event types
|
|
const eventTypes = streamEvents.map((event: unknown) =>
|
|
typeof event === 'object' &&
|
|
event !== null &&
|
|
'event' in event &&
|
|
typeof event.event === 'object' &&
|
|
event.event !== null &&
|
|
'type' in event.event
|
|
? event.event.type
|
|
: null,
|
|
);
|
|
|
|
// Should have message_start event
|
|
expect(eventTypes).toContain('message_start');
|
|
|
|
// Should have content_block_start event
|
|
expect(eventTypes).toContain('content_block_start');
|
|
|
|
// Should have content_block_delta events
|
|
expect(eventTypes).toContain('content_block_delta');
|
|
|
|
// Should have content_block_stop event
|
|
expect(eventTypes).toContain('content_block_stop');
|
|
|
|
// Should have message_stop event
|
|
expect(eventTypes).toContain('message_stop');
|
|
|
|
// Verify that we still have the complete assistant message
|
|
const assistantMessage = messages.find(
|
|
(msg: unknown) =>
|
|
typeof msg === 'object' &&
|
|
msg !== null &&
|
|
'type' in msg &&
|
|
msg.type === 'assistant',
|
|
);
|
|
expect(assistantMessage).toBeDefined();
|
|
expect(assistantMessage).toHaveProperty('message');
|
|
|
|
// Verify that we still have the result message
|
|
const resultMessage = messages[messages.length - 1] as {
|
|
type: string;
|
|
is_error: boolean;
|
|
result: string;
|
|
};
|
|
expect(resultMessage).toBeDefined();
|
|
expect(
|
|
typeof resultMessage === 'object' &&
|
|
resultMessage !== null &&
|
|
'type' in resultMessage &&
|
|
resultMessage.type === 'result',
|
|
).toBe(true);
|
|
expect(resultMessage).toHaveProperty('is_error');
|
|
expect(resultMessage.is_error).toBe(false);
|
|
expect(resultMessage).toHaveProperty('result');
|
|
expect(resultMessage.result.toLowerCase()).toContain('paris');
|
|
});
|
|
|
|
it('should return a JSON error for enforced auth mismatch before running', async () => {
|
|
await rig.setup('json-output-auth-mismatch', {
|
|
settings: {
|
|
security: { auth: { enforcedType: 'qwen-oauth' } },
|
|
},
|
|
});
|
|
|
|
let thrown: Error | undefined;
|
|
try {
|
|
await rig.run('Hello', '--output-format', 'json', ...fakeModelArgs());
|
|
expect.fail('Expected process to exit with error');
|
|
} catch (e) {
|
|
thrown = e as Error;
|
|
}
|
|
|
|
expect(thrown).toBeDefined();
|
|
const message = (thrown as Error).message;
|
|
|
|
// The error JSON is written to stdout as a CLIResultMessageError
|
|
// Extract stdout from the error message
|
|
const stdoutMatch = message.match(/Stdout:\n([\s\S]*?)(?:\n\nStderr:|$)/);
|
|
expect(
|
|
stdoutMatch,
|
|
'Expected to find stdout in the error message',
|
|
).toBeTruthy();
|
|
|
|
const stdout = stdoutMatch![1];
|
|
let parsed: unknown[];
|
|
try {
|
|
// Parse the JSON array from stdout
|
|
parsed = JSON.parse(stdout);
|
|
} catch (parseError) {
|
|
console.error('Failed to parse the following JSON:', stdout);
|
|
throw new Error(
|
|
`Test failed: Could not parse JSON from stdout. Details: ${parseError}`,
|
|
);
|
|
}
|
|
|
|
// The output should be an array of messages
|
|
expect(Array.isArray(parsed)).toBe(true);
|
|
expect(parsed.length).toBeGreaterThan(0);
|
|
|
|
// Find the result message with error
|
|
const resultMessage = parsed.find(
|
|
(msg: unknown) =>
|
|
typeof msg === 'object' &&
|
|
msg !== null &&
|
|
'type' in msg &&
|
|
msg.type === 'result' &&
|
|
'is_error' in msg &&
|
|
msg.is_error === true,
|
|
) as {
|
|
type: string;
|
|
is_error: boolean;
|
|
subtype: string;
|
|
error?: { message: string; type?: string };
|
|
};
|
|
|
|
expect(resultMessage).toBeDefined();
|
|
expect(resultMessage.is_error).toBe(true);
|
|
expect(resultMessage).toHaveProperty('subtype');
|
|
expect(resultMessage.subtype).toBe('error_during_execution');
|
|
expect(resultMessage).toHaveProperty('error');
|
|
expect(resultMessage.error).toBeDefined();
|
|
expect(resultMessage.error?.message).toContain(
|
|
'configured auth type is qwen-oauth',
|
|
);
|
|
expect(resultMessage.error?.message).toContain(
|
|
'current auth type is openai',
|
|
);
|
|
});
|
|
});
|