done

scripts/backfill_prefilter.py

@@ -0,0 +1,72 @@
+"""
+Zero-token backfill:
+  - Mark old posts with prefilter_reason (rt_only / url_only / empty) where applicable
+  - Fill ai_reasoning for posts that have a signal but empty reasoning
+  - Stamp analysis_version='legacy' on anything still missing it
+
+Run once after the schema migration. Safe to re-run.
+"""
+from __future__ import annotations
+
+import asyncio
+import os
+import sys
+from typing import Optional
+
+sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+
+from sqlalchemy import select
+
+from app.database import AsyncSessionLocal
+from app.models import Post
+
+
+def classify(text: str) -> Optional[str]:
+    stripped = (text or "").strip()
+    if not stripped:
+        return "empty"
+    if stripped.startswith("RT: https://") and len(stripped) < 60:
+        return "rt_only"
+    if stripped.startswith("https://") and " " not in stripped:
+        return "url_only"
+    return None
+
+
+PREFILTER_MSG = {
+    "rt_only": "Pre-filtered: retweet with no added commentary.",
+    "url_only": "Pre-filtered: bare URL with no text content.",
+    "empty": "Pre-filtered: empty post body.",
+}
+
+
+async def main():
+    async with AsyncSessionLocal() as db:
+        posts = (await db.execute(select(Post))).scalars().all()
+
+        pre = ai_fix = ver = 0
+        for p in posts:
+            reason = classify(p.text)
+            if reason and not p.prefilter_reason:
+                p.prefilter_reason = reason
+                if not p.ai_reasoning:
+                    p.ai_reasoning = PREFILTER_MSG[reason]
+                    ai_fix += 1
+                pre += 1
+
+            # Any remaining signal'd post with empty reasoning: fill a minimal note
+            if p.signal and not p.ai_reasoning:
+                p.ai_reasoning = "(legacy) reasoning not recorded for this post."
+                ai_fix += 1
+
+            if not p.analysis_version:
+                p.analysis_version = "legacy"
+                ver += 1
+
+        await db.commit()
+        print(f"✅ prefilter_reason set on {pre} posts")
+        print(f"✅ ai_reasoning filled on {ai_fix} posts")
+        print(f"✅ analysis_version stamped 'legacy' on {ver} posts")
+
+
+if __name__ == "__main__":
+    asyncio.run(main())