feat(需求): 撞款重复检查改查数据库——①采集明细先落库再更新文件(V111 biz_shop_data_crawl_item行级明细,归档时countryResults展开,同店同日幂等替换) ②撞款扫描换源直接查明细表(不再解析OSS xlsx,修复旧文件表头差异致4家店静默丢弃) ③存量Excel数据自动迁移(日报快照row_payload优先+OSS宽松表头回退:ASIN码/小写兼容) ④移除84家店铺全量空态注入(重复检查基数=已采集店,超管口径8家) ⑤新增明细聚合/宽松解析单测6个

2026-09-07 02:38:39 +08:00
parent 678605e9fb
commit 5f8c469b8e
24 changed files with 638 additions and 160 deletions
@@ -0,0 +1,18 @@
+package com.nanri.aiimage.modules.shopdatacrawl.mapper;
+
+import com.baomidou.mybatisplus.core.mapper.BaseMapper;
+import com.nanri.aiimage.modules.shopdatacrawl.model.entity.ShopDataCrawlItemEntity;
+import org.apache.ibatis.annotations.Delete;
+import org.apache.ibatis.annotations.Mapper;
+import org.apache.ibatis.annotations.Param;
+
+@Mapper
+public interface ShopDataCrawlItemMapper extends BaseMapper<ShopDataCrawlItemEntity> {
+
+    /** 删除某店某批次（每日快照替换语义：先删后插）。 */
+    @Delete("""
+            DELETE FROM biz_shop_data_crawl_item
+            WHERE shop_name = #{shopName} AND business_date = #{businessDate}
+            """)
+    int deleteBatch(@Param("shopName") String shopName, @Param("businessDate") java.time.LocalDate businessDate);
+}
@@ -0,0 +1,32 @@
+package com.nanri.aiimage.modules.shopdatacrawl.model.entity;
+
+import com.baomidou.mybatisplus.annotation.IdType;
+import com.baomidou.mybatisplus.annotation.TableId;
+import com.baomidou.mybatisplus.annotation.TableName;
+import lombok.Data;
+
+import java.time.LocalDate;
+import java.time.LocalDateTime;
+
+@Data
+@TableName("biz_shop_data_crawl_item")
+public class ShopDataCrawlItemEntity {
+
+    @TableId(type = IdType.AUTO)
+    private Long id;
+    private Long shopId;
+    private String shopName;
+    private String groupName;
+    private LocalDate businessDate;
+    private String country;
+    private String asin;
+    private String brand;
+    private String price;
+    private String itemDate;
+    private String unitsSold;
+    private String itemJson;
+    private Long resultId;
+    private Long taskId;
+    private Long dailyFileId;
+    private LocalDateTime createdAt;
+}
@@ -0,0 +1,129 @@
+package com.nanri.aiimage.modules.shopdatacrawl.service;
+
+import com.fasterxml.jackson.databind.ObjectMapper;
+import com.nanri.aiimage.modules.shopdatacrawl.mapper.ShopDataCrawlItemMapper;
+import com.nanri.aiimage.modules.shopdatacrawl.model.dto.ShopDataCrawlCountryResultDto;
+import com.nanri.aiimage.modules.shopdatacrawl.model.dto.ShopDataCrawlRowDto;
+import com.nanri.aiimage.modules.shopdatacrawl.model.entity.ShopDataCrawlItemEntity;
+import com.nanri.aiimage.modules.shopdatacrawl.model.vo.ShopDataCrawlResultItemVo;
+import lombok.RequiredArgsConstructor;
+import lombok.extern.slf4j.Slf4j;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+
+import java.time.LocalDate;
+import java.util.ArrayList;
+import java.util.LinkedHashMap;
+import java.util.List;
+import java.util.Map;
+
+/**
+ * 店铺数据采集行级明细落库：采集数据先落库、再更新 Excel 文件；
+ * 撞款重复检查（shopduplicatecheck）直接查明细表聚合，不再解析 OSS xlsx。
+ * 语义：每店×每日一份累计快照（先删同批再插，幂等可重放）。
+ */
+@Slf4j
+@Service
+@RequiredArgsConstructor
+public class ShopDataCrawlItemStoreService {
+
+    private final ShopDataCrawlItemMapper itemMapper;
+    private final ObjectMapper objectMapper;
+
+    /** 国家站点码（旧文件 sheet 未命中映射则保留原文参与撞款，国家序列为空）。 */
+    public static String normalizeCountry(String raw) {
+        return raw == null ? "" : raw.trim().toUpperCase();
+    }
+
+    /**
+     * 把当日累计快照（countryResults 展开）落为行级明细。
+     * 增量归档调用点在 Excel 组装/上传之前（先落库再更新文件）；
+     * 存量迁移由 shopduplicatecheck 模块构造 Vo 后复用（幂等：先删同批再插）。
+     */
+    @Transactional(rollbackFor = Exception.class)
+    public void saveShopBatchFromSnapshot(String shopName, LocalDate businessDate,
+                                          List<ShopDataCrawlResultItemVo> snapshots,
+                                          Long resultId, Long taskId, Long dailyFileId) {
+        List<ShopDataCrawlItemEntity> rows = new ArrayList<>();
+        for (ShopDataCrawlResultItemVo vo : snapshots) {
+            if (vo == null || Boolean.FALSE.equals(vo.getSuccess()) || vo.getCountryResults() == null) {
+                continue;
+            }
+            for (ShopDataCrawlCountryResultDto countryResult : vo.getCountryResults()) {
+                if (countryResult == null || countryResult.getItems() == null) {
+                    continue;
+                }
+                String country = normalizeCountry(countryResult.getCountry());
+                for (ShopDataCrawlRowDto item : countryResult.getItems()) {
+                    if (item == null || blank(item.getAsin())) {
+                        continue;
+                    }
+                    rows.add(toEntity(shopName, businessDate, country, item, resultId, taskId, dailyFileId));
+                }
+            }
+        }
+        replaceBatch(shopName, businessDate, rows, taskId);
+    }
+
+    private void replaceBatch(String shopName, LocalDate businessDate,
+                              List<ShopDataCrawlItemEntity> rows, Long taskId) {
+        int deleted = itemMapper.deleteBatch(shopName, businessDate);
+        int inserted = 0;
+        for (ShopDataCrawlItemEntity row : rows) {
+            itemMapper.insert(row);
+            inserted++;
+        }
+        log.info("[shop-data-crawl-item] 店铺明细批次落库 shop={} date={} 删除 {} 行，新插 {} 行（taskId={}）",
+                shopName, businessDate, deleted, inserted, taskId);
+    }
+
+    private ShopDataCrawlItemEntity toEntity(String shopName, LocalDate businessDate, String country,
+                                             ShopDataCrawlRowDto item, Long resultId, Long taskId, Long dailyFileId) {
+        ShopDataCrawlItemEntity entity = new ShopDataCrawlItemEntity();
+        entity.setShopName(shopName);
+        entity.setBusinessDate(businessDate);
+        entity.setCountry(country);
+        entity.setAsin(item.getAsin().trim());
+        entity.setBrand(trimToNull(item.getBrand()));
+        entity.setPrice(trimToNull(item.getPrice()));
+        entity.setItemDate(trimToNull(item.getDate()));
+        entity.setUnitsSold(trimToNull(item.getUnitsSold()));
+        entity.setItemJson(extraColumnsJson(item));
+        entity.setResultId(resultId);
+        entity.setTaskId(taskId);
+        entity.setDailyFileId(dailyFileId);
+        return entity;
+    }
+
+    /** 其余展示列（库存销量/销售排名/页面浏览量/商品图）原文兜底为 JSON，明细查询不依赖。 */
+    private String extraColumnsJson(ShopDataCrawlRowDto item) {
+        try {
+            Map<String, Object> extra = new LinkedHashMap<>();
+            if (item.getInventorySales() != null) {
+                extra.put("inventorySales", item.getInventorySales());
+            }
+            if (item.getSalesRank() != null) {
+                extra.put("salesRank", item.getSalesRank());
+            }
+            if (item.getPageViews() != null) {
+                extra.put("pageViews", item.getPageViews());
+            }
+            if (item.getCommodityImage() != null) {
+                extra.put("commodityImage", item.getCommodityImage());
+            }
+            return extra.isEmpty() ? null : objectMapper.writeValueAsString(extra);
+        } catch (Exception ex) {
+            log.warn("[shop-data-crawl-item] 明细附加列序列化失败 shop={} asin={}",
+                    item.getAsin(), ex.getMessage());
+            return null;
+        }
+    }
+
+    private static boolean blank(String value) {
+        return value == null || value.isBlank();
+    }
+
+    private static String trimToNull(String value) {
+        return value == null || value.isBlank() ? null : value.trim();
+    }
+}
@@ -118,6 +118,7 @@ public class ShopDataCrawlTaskService {
    private final TransientPayloadStorageService transientPayloadStorageService;
    private final InstanceMetadata instanceMetadata;
    private final ShopDataCrawlDailyFileService dailyFileService;
+    private final ShopDataCrawlItemStoreService shopDataCrawlItemStoreService;
    private final PlatformTransactionManager transactionManager;
    private final TaskProgressLightAssembler taskProgressLightAssembler;

@@ -1931,6 +1932,14 @@ public class ShopDataCrawlTaskService {
            // 不再读回旧累计对象（readObjectBytes）并整表重写（replaceCountriesWorkbook）；
            // 历史成员行无 payload 时按结果快照兜底，兼容旧归档数据。
            List<ShopDataCrawlResultItemVo> accumulatedItems = buildDailyFileFromData(baseDailyFile, List.of(snapshot));
+            // 先落库再更新文件（用户约定）：行级明细先于 Excel 组装/上传；
+            // 失败抛异常走现有失败/补偿路径，明细幂等（同店同日先删后插）。
+            LocalDate itemBatchDate = baseDailyFile != null && baseDailyFile.getBusinessDate() != null
+                    ? baseDailyFile.getBusinessDate()
+                    : dailyFileService.currentBusinessDateTime().toLocalDate();
+            shopDataCrawlItemStoreService.saveShopBatchFromSnapshot(
+                    snapshot.getShopName(), itemBatchDate, accumulatedItems,
+                    snapshot.getResultId(), task.getId(), baseDailyFile == null ? null : baseDailyFile.getId());
            int rowCount = excelAssemblyService.writeWorkbook(outputXlsx, accumulatedItems);
            String objectKey = ossStorageService.uploadResultFile(outputXlsx, MODULE_TYPE);
            if (blank(objectKey)) {
@@ -0,0 +1,75 @@
+package com.nanri.aiimage.modules.shopduplicatecheck.mapper;
+
+import com.nanri.aiimage.modules.shopduplicatecheck.model.dto.ShopDailyFileMigrationDto;
+import com.nanri.aiimage.modules.shopduplicatecheck.model.dto.ShopLatestItemRowDto;
+import org.apache.ibatis.annotations.Mapper;
+import org.apache.ibatis.annotations.Param;
+import org.apache.ibatis.annotations.Select;
+
+import java.time.LocalDate;
+import java.util.List;
+
+/**
+ * 撞款扫描的明细表数据源：biz_shop_data_crawl_item（采集先落库）。
+ * 不依赖 OSS xlsx 解析，每店取最新 business_date 批次行聚合。
+ */
+@Mapper
+public interface ShopDuplicateCheckItemMapper {
+
+    /** 每店最新 business_date 批次的全部明细行（撞款统计直接数据源）。 */
+    @Select("""
+            SELECT i.shop_name AS shopName,
+                   i.country AS country,
+                   i.asin AS asin,
+                   i.item_date AS itemDate,
+                   i.price AS price,
+                   i.brand AS brand
+            FROM biz_shop_data_crawl_item i
+            JOIN (
+                SELECT shop_name, MAX(business_date) AS max_date
+                FROM biz_shop_data_crawl_item
+                GROUP BY shop_name
+            ) latest
+              ON latest.shop_name = i.shop_name AND latest.max_date = i.business_date
+            """)
+    List<ShopLatestItemRowDto> selectLatestShopItems();
+
+    /** 幂等判定：该店该日是否已有明细批次。 */
+    @Select("""
+            SELECT COUNT(1) FROM biz_shop_data_crawl_item
+            WHERE shop_name = #{shopName} AND business_date = #{businessDate}
+            """)
+    long countShopBatch(@Param("shopName") String shopName, @Param("businessDate") LocalDate businessDate);
+
+    /** 存量迁移数据源：每店最新已补偿 daily_file（join 最新结果行拿店名/文件地址）。 */
+    @Select("""
+            SELECT ranked.daily_file_id AS dailyFileId,
+                   ranked.business_date AS businessDate,
+                   ranked.latest_result_id AS latestResultId,
+                   fr.source_filename AS shopName,
+                   fr.result_file_url AS resultFileUrl,
+                   fr.task_id AS taskId
+            FROM (
+                SELECT df.id AS daily_file_id,
+                       df.business_date,
+                       df.latest_result_id,
+                       ROW_NUMBER() OVER (
+                           PARTITION BY df.shop_key
+                           ORDER BY df.business_date DESC, df.id DESC) AS row_no
+                FROM biz_shop_data_crawl_daily_file df
+                WHERE df.compensation_done = 1
+            ) ranked
+            JOIN biz_file_result fr ON fr.id = ranked.latest_result_id
+            WHERE ranked.row_no = 1
+              AND fr.module_type = 'SHOP_DATA_CRAWL'
+            """)
+    List<ShopDailyFileMigrationDto> selectLegacyMigrationFiles();
+
+    /** 取每日成员行（row_payload 快照 JSON），无 payload 时返回 null。 */
+    @Select("""
+            SELECT row_payload FROM biz_shop_data_crawl_daily_member
+            WHERE daily_file_id = #{dailyFileId} AND result_id = #{resultId}
+            LIMIT 1
+            """)
+    String selectMemberPayload(@Param("dailyFileId") Long dailyFileId, @Param("resultId") Long resultId);
+}
@@ -0,0 +1,13 @@
+package com.nanri.aiimage.modules.shopduplicatecheck.model.dto;
+
+import java.time.LocalDate;
+
+/** 存量迁移数据源：每店最新已补偿每日文件（含店名/结果文件地址）。 */
+public record ShopDailyFileMigrationDto(
+        Long dailyFileId,
+        LocalDate businessDate,
+        Long latestResultId,
+        String shopName,
+        String resultFileUrl,
+        Long taskId) {
+}
@@ -0,0 +1,11 @@
+package com.nanri.aiimage.modules.shopduplicatecheck.model.dto;
+
+/** 撞款扫描明细行（biz_shop_data_crawl_item 每店最新批次行）。 */
+public record ShopLatestItemRowDto(
+        String shopName,
+        String country,
+        String asin,
+        String itemDate,
+        String price,
+        String brand) {
+}
@@ -5,43 +5,47 @@ import com.fasterxml.jackson.databind.ObjectMapper;
 import com.nanri.aiimage.common.exception.BusinessException;
 import com.nanri.aiimage.common.service.DistributedJobLockService;
 import com.nanri.aiimage.modules.file.service.oss.OssStorageService;
+import com.nanri.aiimage.modules.shopdatacrawl.model.dto.ShopDataCrawlCountryResultDto;
+import com.nanri.aiimage.modules.shopdatacrawl.model.dto.ShopDataCrawlRowDto;
+import com.nanri.aiimage.modules.shopdatacrawl.model.vo.ShopDataCrawlResultItemVo;
+import com.nanri.aiimage.modules.shopdatacrawl.service.ShopDataCrawlItemStoreService;
 import com.nanri.aiimage.modules.shopduplicatecheck.mapper.ShopDataDuplicateScanMapper;
+import com.nanri.aiimage.modules.shopduplicatecheck.mapper.ShopDuplicateCheckItemMapper;
 import com.nanri.aiimage.modules.shopduplicatecheck.mapper.ShopDuplicateCheckSourceMapper;
 import com.nanri.aiimage.modules.shopduplicatecheck.model.dto.ScanFullRowDto;
 import com.nanri.aiimage.modules.shopduplicatecheck.model.dto.ScanLightRowDto;
+import com.nanri.aiimage.modules.shopduplicatecheck.model.dto.ShopDailyFileMigrationDto;
 import com.nanri.aiimage.modules.shopduplicatecheck.model.dto.ShopGroupLabelDto;
-import com.nanri.aiimage.modules.shopduplicatecheck.model.dto.ShopSourceRowDto;
+import com.nanri.aiimage.modules.shopduplicatecheck.model.dto.ShopLatestItemRowDto;
 import com.nanri.aiimage.modules.shopduplicatecheck.model.entity.ShopDataDuplicateScanEntity;
 import com.nanri.aiimage.modules.shopduplicatecheck.model.payload.DuplicateScanPayload;
 import com.nanri.aiimage.modules.shopduplicatecheck.model.payload.DuplicateScanSummary;
 import com.nanri.aiimage.modules.shopduplicatecheck.service.support.DuplicateCheckAggregator;
 import com.nanri.aiimage.modules.shopduplicatecheck.service.support.DuplicateCheckWorkbookParser;
+import com.nanri.aiimage.modules.shopduplicatecheck.service.support.RawRow;
 import com.nanri.aiimage.modules.shopduplicatecheck.service.support.ShopParsed;
 import lombok.extern.slf4j.Slf4j;
 import org.springframework.beans.factory.annotation.Autowired;
 import org.springframework.scheduling.annotation.Scheduled;
 import org.springframework.stereotype.Service;

+import java.time.LocalDate;
 import java.time.LocalDateTime;
 import java.time.format.DateTimeFormatter;
 import java.util.ArrayList;
 import java.util.HashMap;
-import java.util.HashSet;
+import java.util.LinkedHashMap;
 import java.util.LinkedHashSet;
 import java.util.List;
 import java.util.Locale;
 import java.util.Map;
 import java.util.Set;
-import java.util.concurrent.ExecutionException;
-import java.util.concurrent.ExecutorService;
-import java.util.concurrent.Executors;
-import java.util.concurrent.Future;
-import java.util.concurrent.TimeUnit;
 import java.util.concurrent.atomic.AtomicLong;

 /**
- * 撞款扫描：每日 00:00 定时 + force 同步全量重扫。逐店取最新结果文件、拉取并解析 xlsx、
- * 聚合成 shops/items/summary 落库 shop_data_duplicate_scan（格式与 Python 现行版本一致）。
+ * 撞款扫描：每日 00:00 定时 + force 同步全量重扫。
+ * 数据源 = 采集明细表（biz_shop_data_crawl_item，采集先落库再更新文件），
+ * 直接查库聚合 shops/items/summary 落库 shop_data_duplicate_scan（输出契约不变）。
 * 双实例通过 Redis 分布式锁防重；扫描失败落 FAILED 行并上抛，force 场景由端点转 409/500。
 */
@Service
@@ -53,7 +57,6 @@ public class ShopDataDuplicateCheckScanService {

    private static final long MAX_RESULT_BYTES = 256L * 1024 * 1024;
    private static final long MEM_CACHE_MAX_BYTES = 64L * 1024 * 1024;
-    private static final int PARSE_THREADS = 4;
    private static final java.time.Duration SCAN_LOCK_TTL = java.time.Duration.ofHours(6);

    private final ShopDataDuplicateScanMapper scanMapper;
@@ -61,6 +64,8 @@ public class ShopDataDuplicateCheckScanService {
    private final OssStorageService ossStorageService;
    private final DistributedJobLockService distributedJobLockService;
    private final ObjectMapper objectMapper;
+    private final ShopDuplicateCheckItemMapper itemMapper;
+    private final ShopDataCrawlItemStoreService itemStoreService;

    /** 进程内读缓存：轻量行 id 判活，行变化才全量读（对应 Python _duplicate_scan_memory）。 */
    private final AtomicLong cachedRowId = new AtomicLong(-1L);
@@ -71,12 +76,16 @@ public class ShopDataDuplicateCheckScanService {
                                             ShopDuplicateCheckSourceMapper sourceMapper,
                                             OssStorageService ossStorageService,
                                             DistributedJobLockService distributedJobLockService,
-                                             ObjectMapper objectMapper) {
+                                             ObjectMapper objectMapper,
+                                             ShopDuplicateCheckItemMapper itemMapper,
+                                             ShopDataCrawlItemStoreService itemStoreService) {
        this.scanMapper = scanMapper;
        this.sourceMapper = sourceMapper;
        this.ossStorageService = ossStorageService;
        this.distributedJobLockService = distributedJobLockService;
        this.objectMapper = objectMapper;
+        this.itemMapper = itemMapper;
+        this.itemStoreService = itemStoreService;
    }

    /** 读侧视图：scanned_at 为最新 SUCCESS 行 created_at（yyyy-MM-dd HH:mm:ss）。 */
@@ -171,6 +180,8 @@ public class ShopDataDuplicateCheckScanService {

    private void runScanAndSave(DistributedJobLockService.LockHandle lock) {
        try {
+            // 存量 Excel 数据先同步进 MySQL 明细（幂等），之后扫描直接查库聚合。
+            migrateLegacyItemsOnce();
            List<ShopParsed> parsedShops = collectShopParsed(lock);
            DuplicateCheckAggregator.Aggregate aggregate =
                    DuplicateCheckAggregator.aggregate(parsedShops, "job");
@@ -215,183 +226,190 @@ public class ShopDataDuplicateCheckScanService {
        return message == null || message.isBlank() ? ex.getClass().getSimpleName() : message;
    }

-    /** 逐店并发拉取/解析结果文件；单店失败仅告警跳过，不中断整体。
-     *  扫描基线与店铺表全量对齐：无结果文件的店以空态(0 记录)注入，保证撞款控制台可见全部店铺。 */
+    /**
+     * 撞款扫描数据源：直接查采集明细表（biz_shop_data_crawl_item，每店最新 business_date 批次）。
+     * 用户约定：不再解析 OSS xlsx（旧 4 家店因 xlsx 格式差异被静默丢弃的根因），
+     * 重复检查基数 = 已采集落库的店（超管应见 8 家）。
+     */
    private List<ShopParsed> collectShopParsed(DistributedJobLockService.LockHandle lock) {
-        List<ShopSourceRowDto> rows = sourceMapper.selectLatestResultRows();
-        List<ShopParsed> parsedShops = new ArrayList<>();
+        List<ShopLatestItemRowDto> rows = itemMapper.selectLatestShopItems();
        if (rows.isEmpty()) {
+            log.info("[shop-duplicate-check] 明细表暂无采集数据，本次扫描不出库");
+            return List.of();
+        }
+        // 店铺名去重（分组标签来自店铺管理 biz_shop_manage，与页面口径一致）。
+        Set<String> shopNames = new LinkedHashSet<>();
+        for (ShopLatestItemRowDto row : rows) {
+            if (row != null) {
+                shopNames.add(normalizeShopName(row.shopName()));
+            }
+        }
+        Map<String, String> groupLabels = new HashMap<>();
+        for (ShopGroupLabelDto dto : sourceMapper.selectGroupLabels(new ArrayList<>(shopNames))) {
+            groupLabels.put(shopKey(dto.getShopName()), dto.getGroupName() == null ? "" : dto.getGroupName());
+        }
+        List<ShopParsed> parsedShops = buildParsedShops(rows, groupLabels);
+        log.info("[shop-duplicate-check] 明细数据源扫描：{} 家店铺，{} 行明细", parsedShops.size(), rawRowsCount(parsedShops));
        return parsedShops;
    }
-        Map<String, String> groupLabels = loadGroupLabels(rows);
-        Set<String> parsedNames = new HashSet<>();
-        ExecutorService executor = Executors.newFixedThreadPool(Math.min(PARSE_THREADS, rows.size()));
-        try {
-            List<Future<ShopParsed>> futures = new ArrayList<>(rows.size());
-            for (ShopSourceRowDto row : rows) {
-                futures.add(executor.submit(() -> parseShop(row, groupLabels)));
-            }
-            for (Future<ShopParsed> future : futures) {
-                if (lock != null) {
-                    lock.renew(SCAN_LOCK_TTL);
-                }
-                try {
-                    ShopParsed item = future.get();
-                    if (item != null) {
-                        parsedShops.add(item);
-                        parsedNames.add(normalizeShopName(item.shopName()));
-                    }
-                } catch (ExecutionException ex) {
-                    log.warn("[shop-duplicate-check] 并发解析店铺结果文件异常: {}", ex.getCause() == null
-                            ? ex.getMessage() : ex.getCause().getMessage());
-                } catch (InterruptedException ex) {
-                    Thread.currentThread().interrupt();
-                    log.warn("[shop-duplicate-check] 解析线程被中断");
-                    break;
-                }
-            }
-        } finally {
-            executor.shutdown();
-            try {
-                if (!executor.awaitTermination(30, TimeUnit.SECONDS)) {
-                    executor.shutdownNow();
-                }
-            } catch (InterruptedException ex) {
-                Thread.currentThread().interrupt();
-                executor.shutdownNow();
-            }
-        }
-        // 基线扩展：店铺表全量中尚未解析到结果文件行的店 → 空态注入（0 记录、分组标签来自店铺表）。
-        List<ShopGroupLabelDto> allShops = sourceMapper.selectAllShopLabels();
-        int injectedEmptyShops = 0;
-        for (ShopGroupLabelDto all : allShops) {
-            String name = normalizeShopName(all.getShopName());
-            if (name.isEmpty() || parsedNames.contains(name)) {
+
+    /** 明细行 → 按店 ShopParsed（纯函数，便于单测）：空 ASIN 行过滤、国家去重大写、分组标签注入。 */
+    static List<ShopParsed> buildParsedShops(List<ShopLatestItemRowDto> rows, Map<String, String> groupLabels) {
+        Map<String, List<ShopLatestItemRowDto>> byShop = new LinkedHashMap<>();
+        for (ShopLatestItemRowDto row : rows) {
+            if (row == null) {
                continue;
            }
-            String groupName = all.getGroupName() == null ? "" : all.getGroupName();
-            parsedShops.add(new ShopParsed(name, groupName, List.of(), List.of()));
-            injectedEmptyShops++;
+            String name = row.shopName() == null ? "未命名" : (row.shopName().trim().isEmpty() ? "未命名" : row.shopName().trim());
+            byShop.computeIfAbsent(name, k -> new ArrayList<>()).add(row);
        }
-        if (injectedEmptyShops > 0) {
-            log.info("[shop-duplicate-check] 扫描基线含空态店铺 {} 家（无结果文件，以 0 记录参与撞款视图）", injectedEmptyShops);
+        Map<String, String> safeGroups = groupLabels == null ? Map.of() : groupLabels;
+        List<ShopParsed> parsedShops = new ArrayList<>();
+        for (Map.Entry<String, List<ShopLatestItemRowDto>> entry : byShop.entrySet()) {
+            String shopName = entry.getKey();
+            List<RawRow> rawRows = new ArrayList<>();
+            Set<String> countries = new LinkedHashSet<>();
+            for (ShopLatestItemRowDto row : entry.getValue()) {
+                if (blank(row.asin())) {
+                    continue;
+                }
+                String country = row.country() == null ? "" : row.country().trim().toUpperCase(Locale.ROOT);
+                rawRows.add(new RawRow(
+                        row.asin().trim(),
+                        nvl(row.itemDate()),
+                        nvl(row.price()),
+                        nvl(row.brand()),
+                        country));
+                if (!country.isEmpty()) {
+                    countries.add(country);
+                }
+            }
+            parsedShops.add(new ShopParsed(shopName,
+                    safeGroups.getOrDefault(shopKey(shopName), ""),
+                    new ArrayList<>(countries),
+                    rawRows));
        }
        return parsedShops;
    }

-    private Map<String, String> loadGroupLabels(List<ShopSourceRowDto> rows) {
-        Set<String> names = new LinkedHashSet<>();
-        for (ShopSourceRowDto row : rows) {
-            String name = normalizeShopName(row.getShopName());
-            if (!name.isEmpty()) {
-                names.add(name);
+    private static int rawRowsCount(List<ShopParsed> shops) {
+        int total = 0;
+        for (ShopParsed shop : shops) {
+            total += shop.rows().size();
        }
-        }
-        Map<String, String> labels = new HashMap<>();
-        if (names.isEmpty()) {
-            return labels;
-        }
-        for (ShopGroupLabelDto dto : sourceMapper.selectGroupLabels(new ArrayList<>(names))) {
-            labels.put(shopKey(dto.getShopName()), dto.getGroupName() == null ? "" : dto.getGroupName());
-        }
-        return labels;
+        return total;
    }

-    private ShopParsed parseShop(ShopSourceRowDto row, Map<String, String> groupLabels) {
-        long resultId = row.getResultId() == null ? 0L : row.getResultId();
-        if (resultId <= 0) {
-            return null;
+    /**
+     * 存量数据迁移（幂等）：每店最新每日文件的采集数据 → 行级明细表。
+     * 优先 daily_member.row_payload（快照 JSON，无需拉 OSS）；老归档无 payload 时
+     * 回退解析 OSS xlsx（宽松表头），保证 8 家历史店全部入库。
+     */
+    private void migrateLegacyItemsOnce() {
+        List<ShopDailyFileMigrationDto> files = itemMapper.selectLegacyMigrationFiles();
+        if (files.isEmpty()) {
+            log.info("[shop-data-crawl-item] 存量明细迁移：无待迁移每日文件");
+            return;
        }
-        String shopName = normalizeShopName(row.getShopName());
-        if (row.getResultFileUrl() == null || row.getResultFileUrl().isBlank()) {
-            return null;
+        int migrated = 0;
+        int skipped = 0;
+        int failed = 0;
+        for (ShopDailyFileMigrationDto file : files) {
+            LocalDate businessDate = file.businessDate();
+            String shopName = normalizeShopName(file.shopName());
+            if (businessDate == null || file.latestResultId() == null) {
+                log.warn("[shop-data-crawl-item] 存量迁移跳过（缺日期/结果行） shop={}", shopName);
+                failed++;
+                continue;
            }
            try {
-            if (row.getResultFileSize() != null && row.getResultFileSize() > MAX_RESULT_BYTES) {
-                log.warn("[shop-data-crawl] 结果文件过大跳过 result_id={} size={}", resultId, row.getResultFileSize());
-                return null;
+                if (itemMapper.countShopBatch(shopName, businessDate) > 0) {
+                    skipped++;
+                    continue;
                }
-            byte[] bytes;
-            try {
-                bytes = ossStorageService.readObjectBytesBounded(row.getResultFileUrl(), MAX_RESULT_BYTES);
-            } catch (IllegalArgumentException ex) {
-                log.warn("[shop-data-crawl] 结果文件过大跳过 result_id={}: {}", resultId, ex.getMessage());
-                return null;
+                List<ShopDataCrawlResultItemVo> snapshots = buildLegacySnapshot(file);
+                if (snapshots == null) {
+                    log.warn("[shop-data-crawl-item] 存量迁移无数据源 shop={} date={}（无快照且文件不可解析）",
+                            shopName, businessDate);
+                    failed++;
+                    continue;
                }
-            List<com.nanri.aiimage.modules.shopduplicatecheck.service.support.RawRow> parsed;
-            try {
-                parsed = DuplicateCheckWorkbookParser.parse(bytes);
+                itemStoreService.saveShopBatchFromSnapshot(shopName, businessDate, snapshots,
+                        file.latestResultId(), file.taskId(), file.dailyFileId());
+                migrated++;
            } catch (Exception ex) {
-                log.warn("[shop-data-crawl] 解析结果文件失败 result_id={}: {}", resultId,
-                        ex.getMessage() == null ? ex.toString() : ex.getMessage());
-                return null;
+                log.error("[shop-data-crawl-item] 存量迁移失败 shop={} date={} msg={}",
+                        shopName, businessDate, ex.getMessage());
+                failed++;
            }
-            List<String> countryCodes = parseCountryCodesJson(row.getCountryCodesJson());
-            if (countryCodes.isEmpty()) {
-                countryCodes = parseCountryCodesRequest(row.getRequestJson());
        }
-            String groupName = groupLabels.getOrDefault(shopKey(shopName), "");
-            return new ShopParsed(shopName, groupName, countryCodes, parsed);
+        log.info("[shop-data-crawl-item] 存量明细迁移完成：迁移 {} 家，跳过 {} 家（已有明细），失败 {} 家",
+                migrated, skipped, failed);
+    }
+
+    /** 构造迁移快照：row_payload 优先，回退 OSS xlsx 宽松解析；均不可用返回 null。 */
+    private List<ShopDataCrawlResultItemVo> buildLegacySnapshot(ShopDailyFileMigrationDto file) throws Exception {
+        String payload = itemMapper.selectMemberPayload(file.dailyFileId(), file.latestResultId());
+        if (payload != null && !payload.isBlank()) {
+            try {
+                return List.of(objectMapper.readValue(payload, ShopDataCrawlResultItemVo.class));
            } catch (Exception ex) {
-            log.warn("[shop-data-crawl] 拉取结果文件失败 result_id={}: {}", resultId,
-                    ex.getMessage() == null ? ex.toString() : ex.getMessage());
+                log.warn("[shop-data-crawl-item] 迁移快照 JSON 解析失败，回退文件解析 shop={} date={} msg={}",
+                        file.shopName(), file.businessDate(), ex.getMessage());
+            }
+        }
+        if (file.resultFileUrl() == null || file.resultFileUrl().isBlank()) {
            return null;
        }
+        byte[] bytes = ossStorageService.readObjectBytesBounded(file.resultFileUrl(), MAX_RESULT_BYTES);
+        List<RawRow> rawRows = DuplicateCheckWorkbookParser.parse(bytes);
+        return List.of(rawRowsToSnapshot(file, rawRows));
+    }
+
+    /** RawRow（宽松解析结果）→ 快照 VO（countryResults 按国家聚合）。 */
+    private ShopDataCrawlResultItemVo rawRowsToSnapshot(ShopDailyFileMigrationDto file, List<RawRow> rawRows) {
+        ShopDataCrawlResultItemVo vo = new ShopDataCrawlResultItemVo();
+        vo.setShopName(normalizeShopName(file.shopName()));
+        vo.setSuccess(true);
+        vo.setResultId(file.latestResultId());
+        vo.setTaskId(file.taskId());
+        Map<String, List<ShopDataCrawlRowDto>> byCountry = new LinkedHashMap<>();
+        for (RawRow row : rawRows) {
+            if (row == null || blank(row.asin())) {
+                continue;
+            }
+            ShopDataCrawlRowDto dto = new ShopDataCrawlRowDto();
+            dto.setAsin(row.asin().trim());
+            dto.setDate(nvl(row.date()));
+            dto.setPrice(nvl(row.price()));
+            dto.setBrand(nvl(row.brand()));
+            String country = ShopDataCrawlItemStoreService.normalizeCountry(row.country());
+            byCountry.computeIfAbsent(country, k -> new ArrayList<>()).add(dto);
+        }
+        List<ShopDataCrawlCountryResultDto> countryResults = new ArrayList<>();
+        for (Map.Entry<String, List<ShopDataCrawlRowDto>> entry : byCountry.entrySet()) {
+            ShopDataCrawlCountryResultDto cr = new ShopDataCrawlCountryResultDto();
+            cr.setCountry(entry.getKey());
+            cr.setItems(entry.getValue());
+            countryResults.add(cr);
+        }
+        vo.setCountryResults(countryResults);
+        return vo;
+    }
+
+    private static boolean blank(String value) {
+        return value == null || value.isBlank();
+    }
+
+    private static String nvl(String value) {
+        return value == null ? "" : value.trim();
    }

    private String normalizeShopName(String raw) {
        return raw == null ? "未命名" : (raw.trim().isEmpty() ? "未命名" : raw.trim());
    }

-    private String shopKey(String name) {
+    private static String shopKey(String name) {
        return (name == null ? "" : name.trim()).toLowerCase(Locale.ROOT);
    }
-
-    private List<String> parseCountryCodesJson(String jsonValue) {
-        if (jsonValue == null || jsonValue.isBlank()) {
-            return List.of();
-        }
-        try {
-            JsonNode node = objectMapper.readTree(jsonValue);
-            if (node != null && node.isArray()) {
-                return normalizeCodes(node);
-            }
-        } catch (Exception ex) {
-            log.warn("[shop-data-crawl] 解析 country_codes_json 失败: {}", ex.getMessage());
-        }
-        return List.of();
-    }
-
-    private List<String> parseCountryCodesRequest(String requestJson) {
-        if (requestJson == null || requestJson.isBlank()) {
-            return List.of();
-        }
-        try {
-            JsonNode node = objectMapper.readTree(requestJson);
-            if (node == null || !node.isObject()) {
-                return List.of();
-            }
-            JsonNode raw = node.get("countryCodes");
-            if (raw == null) {
-                raw = node.get("country_codes");
-            }
-            if (raw != null && raw.isArray()) {
-                return normalizeCodes(raw);
-            }
-        } catch (Exception ex) {
-            log.warn("[shop-data-crawl] 解析 request_json 国家代码失败: {}", ex.getMessage());
-        }
-        return List.of();
-    }
-
-    private List<String> normalizeCodes(JsonNode array) {
-        List<String> codes = new ArrayList<>();
-        for (JsonNode item : array) {
-            String value = item.asText();
-            if (value != null && !value.trim().isEmpty()) {
-                codes.add(value.trim().toUpperCase(Locale.ROOT));
-            }
-        }
-        return codes;
-    }
 }
@@ -68,8 +68,9 @@ public final class DuplicateCheckWorkbookParser {
            }
            for (Map.Entry<Integer, String> entry : headerMap.entrySet()) {
                String header = cell(entry.getValue());
+                String normalized = header.replaceAll("\\s+", "").toUpperCase();
                int index = entry.getKey();
-                if ("ASIN".equals(header)) {
+                if (isAsinHeader(normalized)) {
                    asinCol = index;
                } else if ("日期".equals(header)) {
                    dateCol = index;
@@ -81,6 +82,11 @@ public final class DuplicateCheckWorkbookParser {
            }
        }

+        /** 表头宽松匹配（兼容旧版/大小写/「ASIN码」类变体）：含 ASIN 即视为 ASIN 列。 */
+        private static boolean isAsinHeader(String normalized) {
+            return normalized.contains("ASIN");
+        }
+
        @Override
        public void onRow(String sheetName, Integer sheetNo, int rowIndex,
                          Map<Integer, String> headerMap, Map<Integer, String> rowMap) {
@@ -0,0 +1,25 @@
+-- 店铺数据采集行级明细表：采集数据先落库再更新文件，撞款重复检查直接查库（不再解析 Excel）。
+-- 语义：每店×每日一份快照（countryResults 展开为行）；撞款扫描取每家店最新 business_date 的明细。
+CREATE TABLE IF NOT EXISTS biz_shop_data_crawl_item (
+    id BIGINT NOT NULL AUTO_INCREMENT,
+    shop_id BIGINT NULL COMMENT '店铺 ID（biz_shop_manage.id）',
+    shop_name VARCHAR(255) NOT NULL COMMENT '规范化店铺名（与结果文件 source_filename 一致）',
+    group_name VARCHAR(255) NULL COMMENT '归档时店铺分组名快照（多分组「、」连接）',
+    business_date DATE NOT NULL COMMENT '采集批次日期（每日文件日期）',
+    country VARCHAR(8) NOT NULL COMMENT '国家站点码（UK/DE/FR/ES/IT，展示层统一中文）',
+    asin VARCHAR(32) NOT NULL COMMENT 'ASIN',
+    brand VARCHAR(512) NULL COMMENT '品牌',
+    price VARCHAR(64) NULL COMMENT '价格（原始文本）',
+    item_date VARCHAR(64) NULL COMMENT '原「日期」列原文（如 2026年8月15日 上午10:59）',
+    units_sold VARCHAR(64) NULL COMMENT '售出件数（原始文本）',
+    item_json JSON NULL COMMENT '其余列原文兜底（库存销量/销售排名/浏览量/商品图等）',
+    result_id BIGINT NULL COMMENT '来源结果文件行 id（biz_file_result.id）',
+    task_id BIGINT NULL COMMENT '来源任务 id（biz_file_task.id）',
+    daily_file_id BIGINT NULL COMMENT '来源每日文件行 id（biz_shop_data_crawl_daily_file.id）',
+    created_at DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP,
+    PRIMARY KEY (id),
+    UNIQUE KEY uk_shop_data_crawl_item_batch (shop_name, business_date, country, asin),
+    KEY idx_shop_data_crawl_item_asin (asin),
+    KEY idx_shop_data_crawl_item_shop_date (shop_name, business_date),
+    KEY idx_shop_data_crawl_item_biz_date (business_date)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci COMMENT='店铺数据采集行级明细（快照批次）';
@@ -136,6 +136,7 @@ class ShopDataCrawlChunkUpsertTest {
                transientPayloadStorageService,
                instanceMetadata,
                dailyFileService,
+                mock(ShopDataCrawlItemStoreService.class),
                null,
                mock(TaskProgressLightAssembler.class));

@@ -172,6 +172,7 @@ class ShopDataCrawlCleanupTest {
                transientPayloadStorageService,
                instanceMetadata,
                dailyFileService,
+                mock(ShopDataCrawlItemStoreService.class),
                null,
                mock(TaskProgressLightAssembler.class));
        ReflectionTestUtils.setField(service, "staleTimeoutMinutes", 30L);
@@ -145,6 +145,7 @@ class ShopDataCrawlDailyFileIncrementalTest {
                transientPayloadStorageService,
                instanceMetadata,
                dailyFileService,
+                mock(ShopDataCrawlItemStoreService.class),
                null,
                mock(TaskProgressLightAssembler.class));

@@ -151,6 +151,7 @@ class ShopDataCrawlDailyFileJobSplitTest {
                transientPayloadStorageService,
                instanceMetadata,
                dailyFileService,
+                mock(ShopDataCrawlItemStoreService.class),
                null,
                mock(TaskProgressLightAssembler.class));

@@ -151,6 +151,7 @@ class ShopDataCrawlDailyFileLockTest {
                transientPayloadStorageService,
                instanceMetadata,
                dailyFileService,
+                mock(ShopDataCrawlItemStoreService.class),
                null,
                mock(TaskProgressLightAssembler.class));

@@ -133,6 +133,7 @@ class ShopDataCrawlLightweightProgressTest {
                transientPayloadStorageService,
                instanceMetadata,
                dailyFileService,
+                mock(ShopDataCrawlItemStoreService.class),
                null,
                mock(TaskProgressLightAssembler.class));

@@ -132,6 +132,7 @@ class ShopDataCrawlOwnerColumnTest {
                transientPayloadStorageService,
                instanceMetadata,
                dailyFileService,
+                mock(ShopDataCrawlItemStoreService.class),
                null,
                mock(TaskProgressLightAssembler.class));

@@ -145,6 +145,7 @@ class ShopDataCrawlProgressQueryTest {
                transientPayloadStorageService,
                instanceMetadata,
                dailyFileService,
+                mock(ShopDataCrawlItemStoreService.class),
                null,
                mock(TaskProgressLightAssembler.class));

@@ -132,6 +132,7 @@ class ShopDataCrawlRowDedupKeyTest {
                transientPayloadStorageService,
                instanceMetadata,
                dailyFileService,
+                mock(ShopDataCrawlItemStoreService.class),
                null,
                mock(TaskProgressLightAssembler.class));

@@ -133,6 +133,7 @@ class ShopDataCrawlScopeCounterTest {
                transientPayloadStorageService,
                instanceMetadata,
                dailyFileService,
+                mock(ShopDataCrawlItemStoreService.class),
                null,
                mock(TaskProgressLightAssembler.class));

@@ -134,6 +134,7 @@ class ShopDataCrawlScopeMergeTest {
                transientPayloadStorageService,
                instanceMetadata,
                dailyFileService,
+                mock(ShopDataCrawlItemStoreService.class),
                null,
                mock(TaskProgressLightAssembler.class));

@@ -128,6 +128,7 @@ class ShopDataCrawlTaskServiceChunkTest {
                transientPayloadStorageService,
                instanceMetadata,
                dailyFileService,
+                mock(ShopDataCrawlItemStoreService.class),
                null,
                mock(TaskProgressLightAssembler.class));

@@ -0,0 +1,52 @@
+package com.nanri.aiimage.modules.shopduplicatecheck.service;
+
+import com.nanri.aiimage.modules.shopduplicatecheck.model.dto.ShopLatestItemRowDto;
+import com.nanri.aiimage.modules.shopduplicatecheck.service.support.RawRow;
+import com.nanri.aiimage.modules.shopduplicatecheck.service.support.ShopParsed;
+import org.junit.jupiter.api.Test;
+
+import java.util.List;
+import java.util.Map;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+import static org.junit.jupiter.api.Assertions.assertTrue;
+
+/**
+ * 撞款明细数据源（biz_shop_data_crawl_item）→ ShopParsed 转换：
+ * 按店分组、空 ASIN 行过滤、国家去重大写、分组标签注入。
+ */
+class ShopDataDuplicateCheckScanServiceTest {
+
+    @Test
+    void buildParsedShops_groupsByShopAndNormalizesCountry() {
+        List<ShopLatestItemRowDto> rows = List.of(
+                new ShopLatestItemRowDto("店A", "uk", "B0001", "2026-09-01 10:00", "9.90", "品牌甲"),
+                new ShopLatestItemRowDto("店A", "DE", "B0001", "2026-09-01 11:00", "9.90", "品牌甲"),
+                new ShopLatestItemRowDto("店B", "FR", "B0002", "2026-09-02 09:00", "19.9", "品牌乙"),
+                new ShopLatestItemRowDto("店B", "FR", "", "2026-09-02 09:01", "1.00", "空ASIN行"),
+                new ShopLatestItemRowDto(null, "UK", "B0003", "", "", ""));
+        List<ShopParsed> parsed = ShopDataDuplicateCheckScanService.buildParsedShops(
+                rows, Map.of("店b", "华东组"));
+
+        assertEquals(3, parsed.size(), "按店分组（含未命名兜底）");
+        ShopParsed shopA = parsed.stream().filter(s -> s.shopName().equals("店A")).findFirst().orElseThrow();
+        assertEquals(List.of("UK", "DE"), shopA.countryCodes(), "国家码去重大写");
+        assertEquals(2, shopA.rows().size());
+        RawRow ukRow = shopA.rows().get(0);
+        assertEquals("B0001", ukRow.asin());
+        assertEquals("UK", ukRow.country());
+
+        ShopParsed shopB = parsed.stream().filter(s -> s.shopName().equals("店B")).findFirst().orElseThrow();
+        assertEquals(1, shopB.rows().size(), "空 ASIN 行过滤");
+        assertEquals("华东组", shopB.groupName(), "分组标签注入（shopKey 大小写不敏感）");
+
+        ShopParsed unnamed = parsed.stream().filter(s -> s.shopName().equals("未命名")).findFirst().orElseThrow();
+        assertEquals(1, unnamed.rows().size(), "无店铺名兜底未命名");
+        assertEquals("", unnamed.rows().get(0).brand(), "缺省字段空串兜底");
+    }
+
+    @Test
+    void buildParsedShops_emptyInputReturnsEmpty() {
+        assertTrue(ShopDataDuplicateCheckScanService.buildParsedShops(List.of(), Map.of()).isEmpty());
+    }
+}
@@ -0,0 +1,78 @@
+package com.nanri.aiimage.modules.shopduplicatecheck.service.support;
+
+import org.apache.poi.xssf.streaming.SXSSFWorkbook;
+import org.apache.poi.xssf.usermodel.XSSFWorkbook;
+import org.junit.jupiter.api.Test;
+
+import java.io.ByteArrayOutputStream;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+
+/**
+ * 宽松表头解析（兼容旧版文件）：ASIN 列大小写/空白/「ASIN码」类变体均可识别，
+ * 是存量 Excel 数据迁移入库的前提（旧版 4 家店因表头差异被整店丢弃的根因）。
+ */
+class DuplicateCheckWorkbookParserTest {
+
+    private static byte[] workbookWith(boolean stream, String sheetName, String[] headers, String[] values) throws Exception {
+        try (ByteArrayOutputStream out = new ByteArrayOutputStream()) {
+            if (stream) {
+                SXSSFWorkbook workbook = new SXSSFWorkbook();
+                write(workbook, sheetName, headers, values);
+                workbook.write(out);
+                workbook.dispose();
+            } else {
+                XSSFWorkbook workbook = new XSSFWorkbook();
+                write(workbook, sheetName, headers, values);
+                workbook.write(out);
+                workbook.close();
+            }
+            return out.toByteArray();
+        }
+    }
+
+    private static void write(org.apache.poi.ss.usermodel.Workbook workbook, String sheetName,
+                              String[] headers, String[] values) {
+        org.apache.poi.ss.usermodel.Sheet sheet = workbook.createSheet(sheetName);
+        org.apache.poi.ss.usermodel.Row header = sheet.createRow(0);
+        for (int i = 0; i < headers.length; i++) {
+            header.createCell(i).setCellValue(headers[i]);
+        }
+        org.apache.poi.ss.usermodel.Row row = sheet.createRow(1);
+        for (int i = 0; i < values.length; i++) {
+            row.createCell(i).setCellValue(values[i]);
+        }
+    }
+
+    @Test
+    void parsesLegacyAsinCodeHeader() throws Exception {
+        byte[] bytes = workbookWith(false, "英国",
+                new String[]{"ASIN码", "日期", "价格", "品牌"},
+                new String[]{"B0OLD001", "2026年8月15日 上午10:59", "£34.46", "DOLGABA"});
+        var rows = DuplicateCheckWorkbookParser.parse(bytes);
+        assertEquals(1, rows.size(), "「ASIN码」表头变体可识别");
+        assertEquals("B0OLD001", rows.get(0).asin());
+        assertEquals("UK", rows.get(0).country(), "中文 sheet 名映射 UK");
+        assertEquals("2026年8月15日 上午10:59", rows.get(0).date());
+        assertEquals("DOLGABA", rows.get(0).brand());
+    }
+
+    @Test
+    void parsesLowercaseAsinHeader() throws Exception {
+        byte[] bytes = workbookWith(true, "asin",
+                new String[]{"asin", "date", "price", "brand"},
+                new String[]{"B0LOW001", "2026-08-15 10:59", "9.99", "X"});
+        var rows = DuplicateCheckWorkbookParser.parse(bytes);
+        assertEquals(1, rows.size(), "小写 asin 表头可识别");
+        assertEquals("B0LOW001", rows.get(0).asin());
+        assertEquals(rows.get(0).country(), "", "未知 sheet 名不丢行（国家为空）");
+    }
+
+    @Test
+    void skipsSheetWithoutAsinColumn() throws Exception {
+        byte[] bytes = workbookWith(false, "德国",
+                new String[]{"日期", "价格"},
+                new String[]{"2026-08-15", "9.99"});
+        assertEquals(0, DuplicateCheckWorkbookParser.parse(bytes).size(), "无 ASIN 列的 sheet 跳过");
+    }
+}