feat: label-ocr 默认 GB+US 近7天; refresh_pending 合并无面单订单(近7天)

flows/shared/db.py

@@ -8,7 +8,7 @@
 
 from __future__ import annotations
 
-from datetime import datetime
+from datetime import datetime, timedelta
 from typing import Any
 
 from psycopg2.extras import execute_values
@@ -101,12 +101,16 @@ class SyncDB:
     # ---------- OCR 面单 enrichment ----------
 
     def get_ocr_pending(
-        self, limit: int = 0, destination_country: str | None = None, courier_source: str | None = None,
+        self, limit: int = 0,
+        destination_countries: list[str] | None = None,
+        courier_source: str | None = None,
+        days: int = 0,
     ) -> list[tuple[str, datetime, str, str | None]]:
         """查待 OCR 订单(所有系统):(orderNo, createdAt, labelPdfUrl, destinationCountry)。
 
-        条件:有 labelPdfUrl、ocr_layout 未处理(Null)。
-        limit=0 表示不限。可选按 destinationCountry / courierSource 筛选。
+        条件:有 labelPdfUrl、ocrLayout 未处理(Null)。
+        limit=0 表示不限。可选按 destinationCountries / courierSource 筛选。
+        days>0 时只取近 N 天下单的订单。
         """
         sql = (
             'SELECT "orderNo","createdAt","labelPdfUrl","destinationCountry" '
@@ -114,12 +118,15 @@ class SyncDB:
             'WHERE "labelPdfUrl" IS NOT NULL AND "ocrLayout" IS NULL'
         )
         params: list = []
-        if destination_country:
-            sql += ' AND "destinationCountry"=%s'
-            params.append(destination_country)
+        if destination_countries:
+            sql += ' AND "destinationCountry"=ANY(%s)'
+            params.append(destination_countries)
         if courier_source:
             sql += ' AND "courierSource"=%s'
             params.append(courier_source)
+        if days > 0:
+            sql += ' AND "orderedAt" >= %s'
+            params.append(datetime.now() - timedelta(days=days))
         sql += ' ORDER BY "orderedAt" DESC'
         if limit > 0:
             sql += " LIMIT %s"
@@ -129,6 +136,19 @@ class SyncDB:
                 cur.execute(sql, params)
                 return [(r[0], r[1], r[2], r[3]) for r in cur.fetchall()]
 
+    def get_orders_without_label(self, system_name: str, days: int = 7) -> list[str]:
+        """查近 N 天下单但无 labelPdfUrl 的订单号。"""
+        cutoff = datetime.now() - timedelta(days=days)
+        with self._conn() as conn:
+            with conn.cursor() as cur:
+                cur.execute(
+                    'SELECT DISTINCT "orderNo" FROM "Orders" '
+                    'WHERE "systemName"=%s AND "labelPdfUrl" IS NULL AND "orderedAt" >= %s',
+                    (system_name, cutoff),
+                )
+                rows = cur.fetchall()
+        return [r[0] for r in rows]
+
     def update_ocr_fields(
         self, order_no: str, created_at: datetime, fields: dict[str, str | None]
     ) -> None: