Hynes/Void-Homelab
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(jobs): blob worker fans out to extract.pdf / extract.image

Browse Source 
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>
This commit is contained in:
root
2026-06-01 09:34:06 +10:00
parent f2035c1de6
commit 3d82f0e5d5
2 changed files with 33 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

25
tests/jobs/workers/blob.test.js
View File

@@ -9,6 +9,7 @@ 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 () => {
@@ -45,4 +46,28 @@ describe('ingest.blob worker', () => {
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);
});
});