first commit

2026-04-20 23:05:59 +08:00
commit 9a72566753
33 changed files with 2138 additions and 0 deletions
@@ -0,0 +1,144 @@
+"""
+历史帖子价格回溯
+对 DB 中没有价格数据的帖子，从 Binance 拉历史 K 线计算涨跌幅
+"""
+
+import asyncio
+import logging
+from datetime import datetime, timezone, timedelta
+from typing import Optional
+
+import httpx
+from sqlalchemy import select
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.config import settings
+from app.models import Post
+
+logger = logging.getLogger(__name__)
+
+# 每次从 Binance 拉多少分钟的 1m K 线（覆盖 1h 涨跌幅计算需要至少 60 根）
+FETCH_WINDOW_MINUTES = 90
+
+
+async def _fetch_klines(symbol: str, start_ms: int, limit: int = 90) -> list:
+    url = (
+        f"{settings.binance_rest_url}/api/v3/klines"
+        f"?symbol={symbol}&interval=1m&startTime={start_ms}&limit={limit}"
+    )
+    async with httpx.AsyncClient(timeout=15) as client:
+        resp = await client.get(url)
+        resp.raise_for_status()
+    return resp.json()
+
+
+def _price_at(klines: list, target_ms: int) -> Optional[float]:
+    """找最接近 target_ms 的收盘价"""
+    best = None
+    best_diff = float("inf")
+    for row in klines:
+        diff = abs(row[0] - target_ms)
+        if diff < best_diff:
+            best_diff = diff
+            best = float(row[4])  # close
+    return best
+
+
+def _pct_change(klines: list, from_ms: int, delta_minutes: int) -> Optional[float]:
+    from_price = _price_at(klines, from_ms)
+    to_price = _price_at(klines, from_ms + delta_minutes * 60 * 1000)
+    if from_price and to_price and from_price != 0:
+        return round((to_price - from_price) / from_price * 100, 4)
+    return None
+
+
+async def backfill_price_impact(db_session_factory, asset: str = "BTC") -> None:
+    """
+    对 DB 中 relevant=True 但没有价格数据的帖子补充价格回溯。
+    对 relevant=False 的帖子，做简单判断（标题含 crypto/bitcoin/btc 关键词则标记为相关）。
+    """
+    symbol = "BTCUSDT" if asset == "BTC" else "ETHUSDT"
+
+    async with db_session_factory() as db:
+        # 拿所有没有价格数据的帖子
+        result = await db.execute(
+            select(Post)
+            .where(Post.price_at_post == None)
+            .order_by(Post.published_at.asc())
+        )
+        posts = result.scalars().all()
+
+    logger.info("找到 %d 条帖子需要价格回溯 (asset=%s)", len(posts), asset)
+    if not posts:
+        return
+
+    # 关键词判断是否与加密相关（快速，不用 Claude API）
+    crypto_keywords = [
+        "bitcoin", "btc", "crypto", "cryptocurrency", "blockchain",
+        "ethereum", "eth", "digital currency", "defi", "coinbase",
+        "sec", "regulation", "tariff", "dollar", "inflation", "fed",
+        "economy", "market", "trade", "sanctions", "iran", "china",
+    ]
+
+    def is_relevant(text: str) -> bool:
+        t = text.lower()
+        return any(kw in t for kw in crypto_keywords)
+
+    saved = 0
+    errors = 0
+
+    for i, post in enumerate(posts):
+        try:
+            published_at = post.published_at
+            if published_at.tzinfo is None:
+                published_at = published_at.replace(tzinfo=timezone.utc)
+
+            start_ms = int(published_at.timestamp() * 1000)
+
+            # 判断相关性（用关键词，不消耗 Claude API）
+            relevant = is_relevant(post.text)
+            sentiment = "neutral"
+            if relevant:
+                t = post.text.lower()
+                bullish_kw = ["great", "win", "winning", "strong", "best", "love", "beautiful", "tremendous", "amazing", "pro-crypto", "bitcoin reserve"]
+                bearish_kw = ["bad", "terrible", "war", "crisis", "sanction", "ban", "regulate", "crack", "fraud", "scam"]
+                if any(k in t for k in bullish_kw):
+                    sentiment = "bullish"
+                elif any(k in t for k in bearish_kw):
+                    sentiment = "bearish"
+
+            # 拉 Binance 历史价格
+            klines = await _fetch_klines(symbol, start_ms, limit=FETCH_WINDOW_MINUTES)
+
+            price_at_post = _price_at(klines, start_ms)
+            m5 = _pct_change(klines, start_ms, 5)
+            m15 = _pct_change(klines, start_ms, 15)
+            m1h = _pct_change(klines, start_ms, 60)
+
+            # 更新帖子
+            async with db_session_factory() as db:
+                result = await db.execute(select(Post).where(Post.id == post.id))
+                p = result.scalar_one_or_none()
+                if p:
+                    p.relevant = relevant
+                    p.sentiment = sentiment
+                    p.price_impact_asset = asset if relevant else None
+                    p.price_at_post = price_at_post
+                    p.price_impact_m5 = m5 if relevant else None
+                    p.price_impact_m15 = m15 if relevant else None
+                    p.price_impact_m1h = m1h if relevant else None
+                    await db.commit()
+                    saved += 1
+
+            if (i + 1) % 20 == 0:
+                logger.info("进度: %d/%d 已处理，已保存 %d 条", i + 1, len(posts), saved)
+
+            # 避免触发 Binance 限速（1200 requests/min）
+            await asyncio.sleep(0.1)
+
+        except Exception as exc:
+            errors += 1
+            logger.error("帖子 id=%d 回溯失败: %s", post.id, exc)
+            await asyncio.sleep(1)
+
+    logger.info("✅ 价格回溯完成: 共处理 %d 条，成功 %d 条，失败 %d 条", len(posts), saved, errors)