fix(review-feedback-961): address latest review comments

Author: ZhuLinsen

data_provider/base.py

@@ -902,12 +902,54 @@ def get_daily_data(
 
         # Normalize code (strip SH/SZ prefix etc.)
         stock_code = normalize_stock_code(stock_code)
+        should_retry_with_akshare_ambiguous_fallback = (
+            stock_code.isdigit() and len(stock_code) == 5
+        )
 
         fetchers = self._get_fetchers_snapshot()
         errors = []
         total_fetchers = len(fetchers)
         request_start = time.time()
 
+        def _run_fetcher_chain(target_code: str, log_prefix: str = "数据源") -> Tuple[Optional[pd.DataFrame], Optional[str], List[str], bool]:
+            local_errors: List[str] = []
+            for attempt, fetcher in enumerate(fetchers, start=1):
+                try:
+                    logger.info(f"[{log_prefix}尝试 {attempt}/{total_fetchers}] [{fetcher.name}] 获取 {target_code}...")
+                    df = self._call_fetcher_method(
+                        fetcher,
+                        "get_daily_data",
+                        stock_code=target_code,
+                        start_date=start_date,
+                        end_date=end_date,
+                        days=days,
+                    )
+
+                    if df is not None and not df.empty:
+                        elapsed = time.time() - request_start
+                        logger.info(
+                            f"[{log_prefix}完成] {target_code} 使用 [{fetcher.name}] 获取成功: "
+                            f"rows={len(df)}, elapsed={elapsed:.2f}s"
+                        )
+                        return df, fetcher.name, local_errors, False
+
+                except Exception as e:
+                    error_type, error_reason = summarize_exception(e)
+                    error_msg = f"[{fetcher.name}] ({error_type}) {error_reason}"
+                    logger.warning(
+                        f"[{log_prefix}失败 {attempt}/{total_fetchers}] [{fetcher.name}] {target_code}: "
+                        f"error_type={error_type}, reason={error_reason}"
+                    )
+                    local_errors.append(error_msg)
+                    if attempt < total_fetchers:
+                        next_fetcher = fetchers[attempt]
+                        if log_prefix == "数据源":
+                            logger.info(f"[数据源切换] {target_code}: [{fetcher.name}] -> [{next_fetcher.name}]")
+                        else:
+                            logger.info(f"[{log_prefix}切换] {target_code}: [{fetcher.name}] -> [{next_fetcher.name}]")
+
+            return None, None, local_errors, True
+
         # 快速路径：美股指数与美股股票直接路由到 YfinanceFetcher
         if is_us_index_code(stock_code) or is_us_stock_code(stock_code):
             for attempt, fetcher in enumerate(fetchers, start=1):
@@ -947,39 +989,24 @@ def get_daily_data(
             logger.error(f"[数据源终止] {stock_code} 获取失败: elapsed={elapsed:.2f}s\n{error_summary}")
             raise DataFetchError(error_summary)
 
-        for attempt, fetcher in enumerate(fetchers, start=1):
-            try:
-                logger.info(f"[数据源尝试 {attempt}/{total_fetchers}] [{fetcher.name}] 获取 {stock_code}...")
-                df = self._call_fetcher_method(
-                    fetcher,
-                    "get_daily_data",
-                    stock_code=stock_code,
-                    start_date=start_date,
-                    end_date=end_date,
-                    days=days
+        df, source_name, chain_errors, _ = _run_fetcher_chain(stock_code)
+        if df is not None:
+            return df, source_name
+        errors.extend(chain_errors)
+
+        if should_retry_with_akshare_ambiguous_fallback:
+            fallback_stock_code = stock_code.zfill(6)
+            if fallback_stock_code != stock_code:
+                logger.info(
+                    f"[数据源回退] {stock_code} 首轮无可用数据，尝试补零后按 A 股路径重试 {fallback_stock_code}"
                 )
-                
-                if df is not None and not df.empty:
-                    elapsed = time.time() - request_start
-                    logger.info(
-                        f"[数据源完成] {stock_code} 使用 [{fetcher.name}] 获取成功: "
-                        f"rows={len(df)}, elapsed={elapsed:.2f}s"
-                    )
-                    return df, fetcher.name
-                    
-            except Exception as e:
-                error_type, error_reason = summarize_exception(e)
-                error_msg = f"[{fetcher.name}] ({error_type}) {error_reason}"
-                logger.warning(
-                    f"[数据源失败 {attempt}/{total_fetchers}] [{fetcher.name}] {stock_code}: "
-                    f"error_type={error_type}, reason={error_reason}"
+                fallback_df, fallback_source_name, fallback_errors, _ = _run_fetcher_chain(
+                    fallback_stock_code,
+                    log_prefix="数据源回退",
                 )
-                errors.append(error_msg)
-                if attempt < total_fetchers:
-                    next_fetcher = fetchers[attempt]
-                    logger.info(f"[数据源切换] {stock_code}: [{fetcher.name}] -> [{next_fetcher.name}]")
-                # 继续尝试下一个数据源
-                continue
+                if fallback_df is not None:
+                    return fallback_df, fallback_source_name
+                errors.extend([f"fallback:{msg}" for msg in fallback_errors])
 
         # 所有数据源都失败
         error_summary = f"所有数据源获取 {stock_code} 失败:\n" + "\n".join(errors)

docs/CHANGELOG.md

@@ -15,6 +15,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/).
 - [修复] 移除 `HistoryItem` 与 `ReportSummary` 响应 Schema 中 `sentiment_score` 的 `ge=0/le=100` 约束（fixes #942）——历史库中存储的超范围负值或大于 100 的情绪评分不再触发 Pydantic ValidationError，历史列表与详情接口恢复正常返回。
 - [修复] 修复输入 5 位裸数字代码（如 '02714'）被误路由到港股路径导致分析错误股票的问题：`normalize_stock_code` 现在对首位为 '0'、补全后落入深市 A 股 000–003 前缀范围且不在已知港股白名单内的 5 位代码自动补前导 '0' 并打印歧义警告；如本意为港股，应使用 'HK02714' 或 '02714.HK' 格式消除歧义。
 - [修复] 修复 5 位裸数字港股代码（如 '02319' 蒙牛乳业）因不在旧 STOCK_NAME_MAP 白名单中被误补全为 A 股代码的回归问题：`normalize_stock_code` 现在对所有 5 位纯数字代码保持"纯语法"规范化，不再做市场推断或自动补零，由上层 DataFetcherManager / 管线在首轮查询无数据时决定是否以补零 A 股代码重试；如需明确指定港股，请使用 'HK02319' 或 '02319.HK' 格式。
+- [修复] 修复 5 位裸数字代码误路由：`normalize_stock_code` 保持纯语法处理，不再在规范化层自动补零；`DataFetcherManager.get_daily_data` 先按港股通道取数，若 5 位裸码在首轮仍无数据，再尝试补零后按 6 位 A 股码重试（如 02714→002714）。
 
 ## [3.12.0] - 2026-04-01
 

tests/test_stock_code_five_digit_ambiguity.py

@@ -7,14 +7,15 @@
 
 Covered acceptance criteria:
 1. '002714' → '002714'  (6-digit A-share, unchanged)
-2. '02714'  → '02714'   (5-digit, kept as-is; upstream decides routing)
+2. '02714'  → '02714'   (5-digit, kept as-is; manager 在港股路径无数据时可尝试补零到 '002714')
 3. '00700'  → '00700'   (known HK, kept as-is)
 4. 'HK02714' → 'HK02714' (explicit HK prefix, stays HK)
 5. '02714.HK' → 'HK02714' (explicit HK suffix, stays HK)
 6. '02319'  → '02319'   (real HK code not in old STOCK_NAME_MAP, must not be padded)
 """
 
 import sys
+import pandas as pd
 import unittest
 from unittest.mock import MagicMock
 
@@ -25,14 +26,41 @@
     sys.modules["json_repair"] = MagicMock()
 
 try:
-    from data_provider.base import normalize_stock_code, _KNOWN_HK_BARE_CODES
+    from data_provider.base import DataFetchError, DataFetcherManager, normalize_stock_code, _KNOWN_HK_BARE_CODES
     _IMPORTS_OK = True
     _IMPORT_ERROR = ""
 except ImportError as e:
     _IMPORTS_OK = False
     _IMPORT_ERROR = str(e)
 
 
+def _sample_df() -> pd.DataFrame:
+    return pd.DataFrame(
+        {
+            "date": ["2026-03-01", "2026-03-02"],
+            "open": [1.0, 1.1],
+            "high": [1.2, 1.2],
+            "low": [0.9, 1.0],
+            "close": [1.1, 1.05],
+            "volume": [1000, 1000],
+            "amount": [10000, 10000],
+            "pct_chg": [0.5, 0.5],
+        }
+    )
+
+
+class _MockDailyFetcher:
+    def __init__(self, name: str, priority: int, result_provider):
+        self.name = name
+        self.priority = priority
+        self._result_provider = result_provider
+        self.calls = []
+
+    def get_daily_data(self, stock_code: str, start_date: str | None = None, end_date: str | None = None, days: int = 30):
+        self.calls.append(stock_code)
+        return self._result_provider(stock_code=stock_code, start_date=start_date, end_date=end_date, days=days)
+
+
 @unittest.skipIf(not _IMPORTS_OK, f"imports failed: {_IMPORT_ERROR}")
 class TestFiveDigitAmbiguity(unittest.TestCase):
     """
@@ -114,6 +142,30 @@ def test_known_hk_bare_codes_contains_02319(self):
         """'02319' (蒙牛乳业, not in old STOCK_NAME_MAP) must be in the HK whitelist."""
         self.assertIn("02319", _KNOWN_HK_BARE_CODES)
 
+    def test_manager_fallback_to_ashare_after_hk_empty(self):
+        """港股路径无数据时会补零重试 A 股候选码。"""
+
+        def result_provider(stock_code: str, **_):
+            if stock_code == "002714":
+                return _sample_df()
+            return None
+
+        fetcher = _MockDailyFetcher("AkshareFetcher", 1, result_provider)
+        manager = DataFetcherManager(fetchers=[fetcher])
+        df, source = manager.get_daily_data("02714", start_date="2026-03-01", end_date="2026-03-02")
+        self.assertEqual(source, "AkshareFetcher")
+        self.assertEqual(fetcher.calls, ["02714", "002714"])
+        self.assertEqual(len(df), 2)
+
+    def test_manager_does_not_retry_padded_for_explicit_hk_code(self):
+        """显式 HK 代码不会触发 6 位补零兜底。"""
+
+        fetcher = _MockDailyFetcher("AkshareFetcher", 1, lambda **_: None)
+        manager = DataFetcherManager(fetchers=[fetcher])
+        with self.assertRaises(DataFetchError):
+            manager.get_daily_data("HK02714", start_date="2026-03-01", end_date="2026-03-02")
+        self.assertEqual(fetcher.calls, ["HK02714"])
+
     # --- Other normalize_stock_code behaviours must be preserved ---
 
     def test_sh_prefix_stripped(self):