init f0ckm

debug/backfill_phash.mjs

@@ -0,0 +1,78 @@
+import db from "../src/inc/sql.mjs";
+import queue from "../src/inc/queue.mjs";
+import fs from "fs";
+import path from "path";
+
+const __dirname = path.dirname(new URL(import.meta.url).pathname);
+const PROJECT_ROOT = path.resolve(__dirname, '..');
+const BATCH_SIZE = 50;
+
+async function backfill() {
+    console.log("Starting PHash backfill...");
+
+    try {
+        // Count total items needing backfill
+        const countResult = await db`SELECT count(*) as count FROM items WHERE (phash IS NULL OR phash = '') AND active = true`;
+        const total = countResult[0].count;
+        console.log(`Found ${total} items to process.`);
+
+        let processed = 0;
+        let errors = 0;
+
+        while (true) {
+            const items = await db`
+                SELECT id, dest 
+                FROM items 
+                WHERE (phash IS NULL OR phash = '') 
+                AND active = true 
+                ORDER BY id DESC 
+                LIMIT ${BATCH_SIZE}
+            `;
+
+            if (items.length === 0) break;
+
+            for (const item of items) {
+                // Correctly resolve path relative to project root
+                const filePath = path.join(PROJECT_ROOT, 'public', 'b', item.dest);
+
+                if (!fs.existsSync(filePath)) {
+                    console.log(`[SKIP] File not found: ${item.dest} (ID: ${item.id})`);
+                    // Mark as MISSING so we don't pick it up again
+                    await db`UPDATE items SET phash = 'MISSING' WHERE id = ${item.id}`;
+                    processed++;
+                    continue;
+                }
+
+                try {
+                    console.log(`[PROCESSING] ID ${item.id}: ${item.dest}`);
+                    const phash = await queue.generatePHash(filePath);
+
+                    if (phash) {
+                        await db`UPDATE items SET phash = ${phash} WHERE id = ${item.id}`;
+                        console.log(`[SUCCESS] ID ${item.id}: Updated PHash.`);
+                    } else {
+                        console.log(`[FAILED] ID ${item.id}: Could not generate PHash.`);
+                        // Mark as ERROR so we don't pick it up again
+                        await db`UPDATE items SET phash = 'ERROR' WHERE id = ${item.id}`;
+                        errors++;
+                    }
+                } catch (err) {
+                    console.error(`[ERROR] ID ${item.id}:`, err);
+                    await db`UPDATE items SET phash = 'ERROR' WHERE id = ${item.id}`;
+                    errors++;
+                }
+                processed++;
+            }
+            console.log(`Progress: ${processed} completed (Remaining in batch loop...)`);
+        }
+
+        console.log("Backfill complete!");
+        process.exit(0);
+
+    } catch (err) {
+        console.error("Fatal error:", err);
+        process.exit(1);
+    }
+}
+
+backfill();