done

scripts/backfill_signals.py

@@ -0,0 +1,86 @@
+"""
+把 AI 信号写回数据库。
+跳过纯RT/URL帖子和已有signal的帖子。
+并发执行提速，自动限速避免API超限。
+用法: python scripts/backfill_signals.py [--limit 500] [--overwrite]
+"""
+import asyncio
+import argparse
+import sys
+import os
+sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+
+from app.services.analysis import analyze_post
+from app.database import AsyncSessionLocal
+from app.models import Post
+from sqlalchemy import select
+
+CONCURRENCY = 5   # 并发数，避免触发API限速
+
+
+async def process_one(post: Post, semaphore: asyncio.Semaphore, overwrite: bool) -> str:
+    """分析单条帖子并写回DB，返回状态字符串"""
+    if not overwrite and post.signal is not None:
+        return "skip"
+
+    async with semaphore:
+        analysis = await analyze_post(post.text)
+        await asyncio.sleep(0.1)  # 轻度限速
+
+    async with AsyncSessionLocal() as db:
+        result = await db.execute(select(Post).where(Post.id == post.id))
+        p = result.scalar_one_or_none()
+        if p:
+            p.sentiment         = analysis["sentiment"]
+            p.signal            = analysis["signal"]
+            p.ai_confidence     = analysis["confidence"]
+            p.ai_reasoning      = analysis["reasoning"]
+            p.relevant          = analysis["relevant"]
+            p.prefilter_reason  = analysis.get("prefilter_reason")
+            p.analysis_version  = analysis.get("analysis_version")
+            if analysis["relevant"] and analysis["asset"] and not p.price_impact_asset:
+                p.price_impact_asset = analysis["asset"]
+            await db.commit()
+
+    sig = analysis["signal"]
+    return f"{sig}:{analysis['confidence']}%"
+
+
+async def main(limit: int, overwrite: bool):
+    async with AsyncSessionLocal() as db:
+        result = await db.execute(
+            select(Post).order_by(Post.published_at.desc()).limit(limit)
+        )
+        posts = result.scalars().all()
+
+    total = len(posts)
+    print(f"共 {total} 条帖子，并发={CONCURRENCY}，overwrite={overwrite}\n")
+
+    semaphore = asyncio.Semaphore(CONCURRENCY)
+    tasks = [process_one(p, semaphore, overwrite) for p in posts]
+
+    done = 0
+    buy = short = sell = hold = skipped = 0
+
+    for coro in asyncio.as_completed(tasks):
+        status = await coro
+        done += 1
+        if status == "skip":
+            skipped += 1
+        elif status.startswith("buy"):   buy += 1
+        elif status.startswith("short"): short += 1
+        elif status.startswith("sell"):  sell += 1
+        else:                            hold += 1
+
+        if done % 20 == 0 or done == total:
+            print(f"  进度 {done}/{total} | BUY={buy} SHORT={short} SELL={sell} HOLD={hold} SKIP={skipped}")
+
+    print(f"\n✅ 完成: BUY={buy} SHORT={short} SELL={sell} HOLD={hold} SKIP={skipped}")
+
+
+if __name__ == "__main__":
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--limit", type=int, default=400)
+    ap.add_argument("--overwrite", action="store_true", help="重新分析已有signal的帖子")
+    args = ap.parse_args()
+    asyncio.run(main(args.limit, args.overwrite))