Reconcile stale running scrapes and allow per-scrape delete

media-scrape/src/engine.ts

@@ -208,6 +208,25 @@ export class ScrapeEngine {
         return { ...meta };
     }
 
+    /**
+     * A scrape left `running` or `queued` on disk has no live job (the daemon that owned it is
+     * gone). Pause it so the UI shows the truth and Resume can pick it up.
+     */
+    reconcileStaleJobs(): string[] {
+        const stale: string[] = [];
+
+        for (const meta of listScrapes(this.options.rootDir)) {
+            if (meta.state !== "running" && meta.state !== "queued") continue;
+
+            meta.state = "paused";
+            meta.updatedAt = new Date(this.now()).toISOString();
+            writeScrapeMeta(join(this.options.rootDir, meta.id), meta);
+            stale.push(meta.id);
+        }
+
+        return stale;
+    }
+
     /** Cache previews for rows a finished or paused scrape never reached. Returns the count. */
     async backfillPreviews(scrapeId: string): Promise<number> {
         const meta = this.active.get(scrapeId) ?? readScrapeMeta(join(this.options.rootDir, scrapeId));

media-scrape/test/run.ts

@@ -25,7 +25,13 @@ import { betterCandidate } from "../src/dedupe.js";
 import { aspectCondition, resolutionCondition } from "../src/filter-presets.js";
 import { isNearLimit, planRetention } from "../src/retention.js";
 import { isoFromSnowflake, snowflakeFromIso } from "../src/snowflake.js";
-import { dirSizeBytes, listScrapes, ScrapeStore } from "../src/store.js";
+import {
+    dirSizeBytes,
+    listScrapes,
+    readScrapeMeta,
+    ScrapeStore,
+    writeScrapeMeta
+} from "../src/store.js";
 import type { CoreAttachment, CoreMessage } from "../src/types.js";
 
 let failures = 0;
@@ -877,6 +883,25 @@ void (async () => {
         assert.equal(existsSync(join(dir, `${poster?.hash}.webp`)), true);
     });
 
+    await test("stale running scrapes are paused on reconcile", async () => {
+        const backend = new FakeBackend([{ messages: [message("1400")] }]);
+        const engine = engineFor(root, backend);
+        const started = await engine.start({ channelId: "170", guildId: null, mode: "links", backend });
+        await engine.processQueue();
+
+        // Simulate a daemon that died mid-job: the meta still says running.
+        const dir = join(root, started.id);
+        const meta = readScrapeMeta(dir);
+        assert.ok(meta);
+        meta.state = "running";
+        writeScrapeMeta(dir, meta);
+
+        const fresh = new ScrapeEngine({ rootDir: root, previews: false });
+        const stale = fresh.reconcileStaleJobs();
+        assert.deepEqual(stale, [started.id]);
+        assert.equal(readScrapeMeta(dir)?.state, "paused");
+    });
+
     await test("markPreview stores the animated gif reference", () => {
         const dir = join(root, "anim-store");
         mkdirSync(dir, { recursive: true });