3d82f0e5d5
After creating a ref, the Node-side ingest.blob worker enqueues a follow-up job for the Python void-workers (Plan 4) to OCR / extract text. Other kinds (file) get no follow-up. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
74 lines
3.4 KiB
JavaScript
74 lines
3.4 KiB
JavaScript
import { describe, it, expect, beforeEach, afterEach } from 'vitest';
|
|
import fs from 'node:fs/promises';
|
|
import path from 'node:path';
|
|
import os from 'node:os';
|
|
import { resetDb } from '../../helpers/db.js';
|
|
import { migrateUp } from '../../../lib/db/migrate.js';
|
|
import { stopBoss, waitForJob } from '../../helpers/boss.js';
|
|
import * as queue from '../../../lib/jobs/queue.js';
|
|
import { registerWorkers } from '../../../lib/jobs/index.js';
|
|
import * as spaces from '../../../lib/db/repos/spaces.js';
|
|
import * as refs from '../../../lib/db/repos/refs.js';
|
|
import * as jobsRepo from '../../../lib/db/repos/jobs.js';
|
|
|
|
let tmpRoot;
|
|
beforeEach(async () => {
|
|
tmpRoot = await fs.mkdtemp(path.join(os.tmpdir(), 'void-blobs-'));
|
|
process.env.BLOB_ROOT = tmpRoot;
|
|
await resetDb(); await migrateUp(); await queue.start(); await registerWorkers();
|
|
});
|
|
afterEach(async () => { await stopBoss(); });
|
|
|
|
describe('ingest.blob worker', () => {
|
|
it('creates a ref pointing at the blob (kind=file)', async () => {
|
|
const sp = await spaces.create({ slug: 'b', name: 'B' }, { kind: 'user', id: null });
|
|
const upTmp = path.join(tmpRoot, 'up.tmp');
|
|
await fs.writeFile(upTmp, Buffer.from('hello blob'));
|
|
const id = await queue.enqueue('ingest.blob', {
|
|
space_id: sp.id, tmp_path: upTmp, filename: 'hello.txt', content_type: 'text/plain'
|
|
});
|
|
const j = await waitForJob('ingest.blob', id, { timeoutMs: 10_000 });
|
|
expect(j.state).toBe('completed');
|
|
const rows = await refs.list({ space_id: sp.id });
|
|
expect(rows[0].kind).toBe('file');
|
|
expect(rows[0].blob_path).toBeTruthy();
|
|
expect(rows[0].title).toBe('hello.txt');
|
|
});
|
|
|
|
it('classifies image content_type as image kind', async () => {
|
|
const sp = await spaces.create({ slug: 'b2', name: 'B2' }, { kind: 'user', id: null });
|
|
const upTmp = path.join(tmpRoot, 'pic.tmp');
|
|
await fs.writeFile(upTmp, Buffer.from([0x89, 0x50, 0x4e, 0x47])); // PNG magic
|
|
const id = await queue.enqueue('ingest.blob', {
|
|
space_id: sp.id, tmp_path: upTmp, filename: 'x.png', content_type: 'image/png'
|
|
});
|
|
await waitForJob('ingest.blob', id, { timeoutMs: 10_000 });
|
|
const rows = await refs.list({ space_id: sp.id });
|
|
expect(rows[0].kind).toBe('image');
|
|
});
|
|
|
|
it('enqueues extract.pdf after creating a pdf ref', async () => {
|
|
const sp = await spaces.create({ slug: 'bpdf', name: 'BPdf' }, { kind: 'user', id: null });
|
|
const upTmp = path.join(tmpRoot, 'doc.tmp');
|
|
await fs.writeFile(upTmp, Buffer.from('%PDF-1.4 ...'));
|
|
const id = await queue.enqueue('ingest.blob', {
|
|
space_id: sp.id, tmp_path: upTmp, filename: 'a.pdf', content_type: 'application/pdf'
|
|
});
|
|
await waitForJob('ingest.blob', id, { timeoutMs: 10_000 });
|
|
const pending = await jobsRepo.list({ name: 'extract.pdf' });
|
|
expect(pending.length).toBeGreaterThan(0);
|
|
});
|
|
|
|
it('enqueues extract.image after creating an image ref', async () => {
|
|
const sp = await spaces.create({ slug: 'bimg', name: 'BImg' }, { kind: 'user', id: null });
|
|
const upTmp = path.join(tmpRoot, 'pic.tmp');
|
|
await fs.writeFile(upTmp, Buffer.from([0x89, 0x50, 0x4e, 0x47]));
|
|
const id = await queue.enqueue('ingest.blob', {
|
|
space_id: sp.id, tmp_path: upTmp, filename: 'a.png', content_type: 'image/png'
|
|
});
|
|
await waitForJob('ingest.blob', id, { timeoutMs: 10_000 });
|
|
const pending = await jobsRepo.list({ name: 'extract.image' });
|
|
expect(pending.length).toBeGreaterThan(0);
|
|
});
|
|
});
|