Files
promptfoo--promptfoo/test/server/services/evalJobService.test.ts
T
wehub-resource-sync 0d3cb498a3
Deploy local.promptfoo.app / Deploy to Cloudflare Pages (push) Waiting to run
Test and Publish Multi-arch Docker Image / test (push) Waiting to run
Test and Publish Multi-arch Docker Image / build-docker-and-push-digests (map[digest-suffix:linux-amd64 platform:linux/amd64 runner:ubuntu-latest]) (push) Blocked by required conditions
Test and Publish Multi-arch Docker Image / build-docker-and-push-digests (map[digest-suffix:linux-arm64 platform:linux/arm64 runner:ubuntu-24.04-arm]) (push) Blocked by required conditions
Test and Publish Multi-arch Docker Image / merge-docker-digests (push) Blocked by required conditions
Test and Publish Multi-arch Docker Image / Attest Multi-arch Image (push) Blocked by required conditions
Validate Renovate Config / Validate Renovate Configuration (push) Waiting to run
CI / Shell Format Check (push) Has been cancelled
CI / Check Ruby (3.4) (push) Has been cancelled
CI / CI Config (push) Has been cancelled
CI / Test on Node ${{ matrix.node }} and ${{ matrix.os }}${{ matrix.shard && format(' (shard {0}/3)', matrix.shard) || '' }} (push) Has been cancelled
CI / Build on Node ${{ matrix.node }} (push) Has been cancelled
CI / Style Check (push) Has been cancelled
CI / Generate Assets (push) Has been cancelled
CI / Check Python (3.14) (push) Has been cancelled
CI / Check Python (3.9) (push) Has been cancelled
CI / Build Docs (push) Has been cancelled
CI / Code Scan Action (push) Has been cancelled
CI / Site tests (push) Has been cancelled
CI / webui tests (push) Has been cancelled
CI / Run Integration Tests (push) Has been cancelled
CI / Run Smoke Tests (push) Has been cancelled
CI / Go Tests (push) Has been cancelled
CI / Share Test (push) Has been cancelled
CI / Redteam (Production API) (push) Has been cancelled
CI / Redteam (Staging API) (push) Has been cancelled
CI / GitHub Actions Lint (push) Has been cancelled
CI / Check Ruby (3.0) (push) Has been cancelled
release-please / release-please (push) Has been cancelled
release-please / build (push) Has been cancelled
release-please / publish-npm (push) Has been cancelled
release-please / publish-npm-backfill (push) Has been cancelled
release-please / docker (push) Has been cancelled
release-please / publish-code-scan-action (push) Has been cancelled
release-please / attest-code-scan-action (push) Has been cancelled
chore: import upstream snapshot with attribution
2026-07-13 13:24:08 +08:00

137 lines
4.1 KiB
TypeScript

import { describe, expect, it } from 'vitest';
import { EvalJobService } from '../../../src/server/services/evalJobService';
describe('EvalJobService', () => {
it('creates an in-progress job', () => {
const service = new EvalJobService();
expect(service.create('job-1')).toEqual({
evalId: null,
status: 'in-progress',
progress: 0,
total: 0,
result: null,
logs: [],
});
});
it('tracks progress, completion, and logs', () => {
const service = new EvalJobService();
service.create('job-1');
expect(service.setProgress('job-1', 2, 5)).toBe(true);
expect(service.appendLog('job-1', 'working')).toBe(true);
expect(service.complete('job-1', { results: [] } as never, 'eval-1')).toBe(true);
expect(service.get('job-1')).toMatchObject({
evalId: 'eval-1',
status: 'complete',
progress: 2,
total: 5,
result: { results: [] },
logs: ['working'],
});
});
it('returns defensive snapshots', () => {
const service = new EvalJobService();
service.create('job-1');
const result = { results: [{ output: { nested: 'original' } }] } as never;
service.complete('job-1', result, 'eval-1');
const snapshot = service.get('job-1');
expect(snapshot).toBeDefined();
snapshot?.logs.push('mutated outside service');
(snapshot?.result as any).results[0].output.nested = 'mutated outside service';
(result as any).results[0].output.nested = 'mutated after completion';
expect(service.get('job-1')?.logs).toEqual([]);
expect((service.get('job-1')?.result as any).results[0].output.nested).toBe('original');
});
it('stores JSON-safe snapshots for function-backed prompts', () => {
const service = new EvalJobService();
service.create('job-1');
expect(() =>
service.complete(
'job-1',
{
prompts: [{ function: () => 'generated prompt', label: 'dynamic prompt' }],
results: [],
} as never,
'eval-1',
),
).not.toThrow();
expect(service.get('job-1')?.result).toEqual({
prompts: [{ label: 'dynamic prompt' }],
results: [],
});
});
it('stores JSON-safe snapshots for circular references and bigint values', () => {
const service = new EvalJobService();
service.create('job-1');
const output: { count: bigint; self?: unknown } = { count: 42n };
output.self = output;
expect(() =>
service.complete(
'job-1',
{
results: [{ response: { output } }],
} as never,
'eval-1',
),
).not.toThrow();
expect(service.get('job-1')?.result).toEqual({
results: [{ response: { output: { count: '42' } } }],
});
});
it('supports replacing and appending failure logs', () => {
const service = new EvalJobService();
service.create('job-1');
service.appendLog('job-1', 'before failure');
expect(service.fail('job-1', ['first failure'])).toBe(true);
expect(service.get('job-1')?.logs).toEqual(['first failure']);
expect(service.fail('job-1', ['second failure'], { append: true })).toBe(true);
expect(service.get('job-1')).toMatchObject({
evalId: null,
status: 'error',
result: null,
logs: ['first failure', 'second failure'],
});
});
it('can preserve completed results when marking a job as failed', () => {
const service = new EvalJobService();
service.create('job-1');
service.complete('job-1', { results: [] } as never, 'eval-1');
expect(
service.fail('job-1', ['cancelled after completion'], {
append: true,
resetResult: false,
}),
).toBe(true);
expect(service.get('job-1')).toMatchObject({
evalId: 'eval-1',
status: 'error',
result: { results: [] },
logs: ['cancelled after completion'],
});
});
it('returns false when updating a missing job', () => {
const service = new EvalJobService();
expect(service.setProgress('missing', 1, 1)).toBe(false);
expect(service.complete('missing', null, null)).toBe(false);
expect(service.fail('missing', ['missing'])).toBe(false);
expect(service.appendLog('missing', 'missing')).toBe(false);
});
});