fix: 恒指回填改https+分页(腾讯接口单次上限2000根,超限向前翻页拉更早历史)

deploy/profile-scripts/backfill_hk_index.py

@@ -27,20 +27,47 @@ DB_PATH = Path(_MOFIN_ROOT) / "data" / "mofin.db"
 
 UA = "Mozilla/5.0"
 INDEX_CODE = "hkHSI"
+PAGE_SIZE = 2000  # 腾讯接口单次上限（实测 count=2000 ok, 2500 报 param error）
 
 
-def fetch_hsi(count=3000):
-    """从腾讯接口拉 hkHSI 日K（count 根）"""
-    url = (f"http://ifzq.gtimg.cn/appstock/app/fqkline/get?"
-           f"param={INDEX_CODE},day,,,{count},qfq")
+def _fetch_page(count, end_date=""):
+    """拉一页：param=hkHSI,day,start,end,count,qfq。end_date 空=截至最新"""
+    url = (f"https://web.ifzq.gtimg.cn/appstock/app/fqkline/get?"
+           f"param={INDEX_CODE},day,,{end_date},{count},qfq")
     req = urllib.request.Request(url, headers={"User-Agent": UA})
     opener = urllib.request.build_opener(urllib.request.ProxyHandler({}))
     with opener.open(req, timeout=30) as r:
         text = r.read().decode("utf-8", errors="replace")
     data = json.loads(text)
-    node = data.get("data", {}).get(INDEX_CODE, {})
-    bars = node.get("qfqday") or node.get("day") or []
-    return bars
+    node = (data.get("data") or {}).get(INDEX_CODE, {})
+    return node.get("qfqday") or node.get("day") or []
+
+
+def fetch_hsi(count=3000):
+    """分页拉取 hkHSI 日K（单次上限 2000，超过则向前翻页）"""
+    from datetime import datetime as _dt, timedelta as _td
+    import time as _t
+    uniq = {}
+    end_date = ""
+    while len(uniq) < count:
+        want = min(PAGE_SIZE, count - len(uniq) + 5)
+        page = _fetch_page(want, end_date)
+        if not page:
+            break
+        new_cnt = 0
+        for b in page:
+            if b[0] not in uniq:
+                uniq[b[0]] = b
+                new_cnt += 1
+        if new_cnt == 0:
+            break  # 没有新数据，历史到头
+        # 向前翻页：以本页最早日期的前一天为 end
+        earliest = page[0][0]
+        end_date = (_dt.strptime(earliest, "%Y-%m-%d") - _td(days=1)).strftime("%Y-%m-%d")
+        if len(page) < 100:  # 不足一页说明历史拉完
+            break
+        _t.sleep(1)  # 翻页间隔，遵守访问频率
+    return [uniq[d] for d in sorted(uniq)]
 
 
 def main():