feat(quant): M7.3 研究行情口径显式化（默认不复权 none，可切 qfq）

- DailyBarRepository.get_range_many / stream_range_many_columns 增加 adjust 参数
  （默认 'none'）→ SQL 层过滤口径，消除 stock_daily 混 source/adjust 污染因子的风险
- ResearchSpec / SelectionQuery 增加 price_adjustment（none|qfq），随 config_snapshot
  落库可溯源；ResearchService._load_daily 与 SelectionService 装配按口径取数
- tests/test_price_adjustment.py：repo 读取按 adjust 过滤（none/qfq 各自命中）、
  spec 默认与字段记录；全量 pytest 通过

backend/app/infrastructure/persistence/sqlalchemy/repositories/market_impl.py

@@ -158,16 +158,24 @@ class SqlAlchemyDailyBarRepository:
         ).all()
         return [DailyBar.model_validate(r, from_attributes=True) for r in rows]
 
-    def get_range_many(self, symbols: Sequence[str], start: date, end: date) -> list[DailyBar]:
-        rows = self._session.scalars(
+    def get_range_many(
+        self,
+        symbols: Sequence[str],
+        start: date,
+        end: date,
+        adjust: str = "none",
+    ) -> list[DailyBar]:
+        stmt = (
             select(StockDailyModel)
             .where(
                 StockDailyModel.symbol.in_(list(symbols)),
                 StockDailyModel.trade_date >= start,
                 StockDailyModel.trade_date <= end,
+                StockDailyModel.adjust == adjust,  # 研究主口径：不复权（v2 §8）
             )
             .order_by(StockDailyModel.trade_date)
-        ).all()
+        )
+        rows = self._session.scalars(stmt).all()
         return [DailyBar.model_validate(r, from_attributes=True) for r in rows]
 
     def stream_range_many_columns(
@@ -176,6 +184,7 @@ class SqlAlchemyDailyBarRepository:
         start: date,
         end: date,
         columns: Sequence[str],
+        adjust: str = "none",
     ) -> Iterator[tuple]:
         """流式返回 (symbol, trade_date_iso, *float_cols) 元组，分批拉取。
 
@@ -193,6 +202,7 @@ class SqlAlchemyDailyBarRepository:
                 StockDailyModel.symbol.in_(list(symbols)),
                 StockDailyModel.trade_date >= start,
                 StockDailyModel.trade_date <= end,
+                StockDailyModel.adjust == adjust,
             )
             .order_by(StockDailyModel.symbol, StockDailyModel.trade_date)
             .execution_options(yield_per=20000)