task-48: invalid ASIN 记录改为批量 INSERT IGNORE/upsert,唯一键扩为 (data_value, brand)
This commit is contained in:
+10
-27
@@ -35,9 +35,9 @@ import com.nanri.aiimage.modules.collectdata.model.vo.CollectDataTaskSummaryVo;
|
||||
import com.nanri.aiimage.modules.collectdata.util.CollectDataBatchQuery;
|
||||
import com.nanri.aiimage.modules.collectdata.util.CollectDataBrandBatchFilter;
|
||||
import com.nanri.aiimage.modules.collectdata.util.CollectDataExtraJsonCodec;
|
||||
import com.nanri.aiimage.modules.collectdata.util.CollectDataInvalidAsinBatchWriter;
|
||||
import com.nanri.aiimage.modules.collectdata.util.CollectDataParseLimits;
|
||||
import com.nanri.aiimage.modules.invalidasin.mapper.InvalidAsinDataMapper;
|
||||
import com.nanri.aiimage.modules.invalidasin.model.entity.InvalidAsinDataEntity;
|
||||
import com.nanri.aiimage.modules.file.service.LocalFileStorageService;
|
||||
import com.nanri.aiimage.modules.file.service.oss.OssStorageService;
|
||||
import com.nanri.aiimage.modules.task.mapper.FileResultMapper;
|
||||
@@ -140,6 +140,9 @@ public class CollectDataService {
|
||||
/** 品牌检查批次过滤器:空品牌批次跳过远程请求,分类语义与旧实现等价。 */
|
||||
private final CollectDataBrandBatchFilter brandBatchFilter;
|
||||
|
||||
/** invalid ASIN 批量写入器:按批次 INSERT IGNORE,替代逐行插入。 */
|
||||
private final CollectDataInvalidAsinBatchWriter invalidAsinBatchWriter;
|
||||
|
||||
@Value("${aiimage.collect-data.stale-timeout-minutes:30}")
|
||||
private long staleTimeoutMinutes;
|
||||
|
||||
@@ -727,35 +730,15 @@ public class CollectDataService {
|
||||
|
||||
private List<CollectDataResultRowVo> filterByBrandCheck(List<CollectDataResultRowVo> rows, CollectDataStats stats) {
|
||||
CollectDataBrandBatchFilter.BrandBatchOutcome outcome = brandBatchFilter.filter(rows);
|
||||
for (CollectDataResultRowVo row : outcome.rejected()) {
|
||||
stats.brandRejectedCount++;
|
||||
insertInvalidAsin(row);
|
||||
}
|
||||
for (CollectDataResultRowVo row : outcome.queryFailed()) {
|
||||
stats.brandQueryFailedCount++;
|
||||
insertInvalidAsin(row);
|
||||
}
|
||||
stats.brandRejectedCount += outcome.rejected().size();
|
||||
stats.brandQueryFailedCount += outcome.queryFailed().size();
|
||||
List<CollectDataResultRowVo> invalidRows = new ArrayList<>(outcome.rejected().size() + outcome.queryFailed().size());
|
||||
invalidRows.addAll(outcome.rejected());
|
||||
invalidRows.addAll(outcome.queryFailed());
|
||||
invalidAsinBatchWriter.writeBatch(invalidRows);
|
||||
return outcome.accepted();
|
||||
}
|
||||
|
||||
private void insertInvalidAsin(CollectDataResultRowVo row) {
|
||||
if (row == null || row.getAsin() == null || row.getAsin().isBlank()) {
|
||||
return;
|
||||
}
|
||||
String brand = normalizeBrand(row.getBrand());
|
||||
if (brand.isBlank()) {
|
||||
return;
|
||||
}
|
||||
InvalidAsinDataEntity entity = new InvalidAsinDataEntity();
|
||||
entity.setDataValue(row.getAsin());
|
||||
entity.setBrand(brand);
|
||||
entity.setRecordSource("AUTO");
|
||||
try {
|
||||
invalidAsinDataMapper.insert(entity);
|
||||
} catch (DuplicateKeyException ignored) {
|
||||
}
|
||||
}
|
||||
|
||||
private void upsertResultItem(Long taskId, Long resultId, String scopeKey, CollectDataResultRowVo row) {
|
||||
String itemKey = "asin:" + row.getAsin();
|
||||
String scopeHash = hash(scopeKey);
|
||||
|
||||
+94
@@ -0,0 +1,94 @@
|
||||
package com.nanri.aiimage.modules.collectdata.util;
|
||||
|
||||
import com.nanri.aiimage.modules.collectdata.model.vo.CollectDataResultRowVo;
|
||||
import com.nanri.aiimage.modules.invalidasin.mapper.InvalidAsinDataMapper;
|
||||
import com.nanri.aiimage.modules.invalidasin.model.entity.InvalidAsinDataEntity;
|
||||
import lombok.extern.slf4j.Slf4j;
|
||||
import org.springframework.beans.factory.annotation.Value;
|
||||
import org.springframework.stereotype.Component;
|
||||
|
||||
import java.util.ArrayList;
|
||||
import java.util.List;
|
||||
import java.util.Locale;
|
||||
import java.util.regex.Pattern;
|
||||
|
||||
/**
|
||||
* invalid ASIN 批量写入器:把不合规 ASIN 行按批次批量 INSERT IGNORE 写入
|
||||
* biz_invalid_asin_data,替代逐行 insert + 捕获唯一键异常的旧路径。
|
||||
* 幂等由唯一键 (data_value, brand) + INSERT IGNORE 语义保证;批量失败时
|
||||
* 跳过该批(记录计数,不中断提交流程),恢复后继续后续批次。
|
||||
* 空输入、无合法 data_value/brand 的行、null 行均安全跳过,不发起调用。
|
||||
*/
|
||||
@Slf4j
|
||||
@Component
|
||||
public class CollectDataInvalidAsinBatchWriter {
|
||||
|
||||
private static final Pattern WHITESPACE_PATTERN = Pattern.compile("\\s+");
|
||||
|
||||
private final InvalidAsinDataMapper invalidAsinDataMapper;
|
||||
private final int batchSize;
|
||||
|
||||
public CollectDataInvalidAsinBatchWriter(InvalidAsinDataMapper invalidAsinDataMapper,
|
||||
@Value("${aiimage.collect-data.invalid-asin-batch-size:100}") int batchSize) {
|
||||
this.invalidAsinDataMapper = invalidAsinDataMapper;
|
||||
this.batchSize = Math.max(1, batchSize);
|
||||
}
|
||||
|
||||
/** 批量写入并返回实际尝试写入的行数(INSERT IGNORE 忽略重复,幂等)。 */
|
||||
public int writeBatch(List<CollectDataResultRowVo> rows) {
|
||||
if (rows == null || rows.isEmpty()) {
|
||||
return 0;
|
||||
}
|
||||
int written = 0;
|
||||
for (int start = 0; start < rows.size(); start += batchSize) {
|
||||
int end = Math.min(start + batchSize, rows.size());
|
||||
List<CollectDataResultRowVo> batch = rows.subList(start, end);
|
||||
List<InvalidAsinDataEntity> entities = new ArrayList<>(batch.size());
|
||||
for (CollectDataResultRowVo row : batch) {
|
||||
InvalidAsinDataEntity entity = toEntity(row);
|
||||
if (entity != null) {
|
||||
entities.add(entity);
|
||||
}
|
||||
}
|
||||
if (entities.isEmpty()) {
|
||||
continue;
|
||||
}
|
||||
try {
|
||||
written += invalidAsinDataMapper.insertBatchIgnore(entities);
|
||||
} catch (RuntimeException ex) {
|
||||
log.warn("[collect-data] invalid asin batch insert failed, skip batch err={}", ex.getMessage());
|
||||
}
|
||||
}
|
||||
return written;
|
||||
}
|
||||
|
||||
private InvalidAsinDataEntity toEntity(CollectDataResultRowVo row) {
|
||||
if (row == null) {
|
||||
return null;
|
||||
}
|
||||
String asin = normalize(row.getAsin());
|
||||
String brand = normalize(row.getBrand()).toLowerCase(Locale.ROOT);
|
||||
if (asin.isBlank() || brand.isBlank()) {
|
||||
return null;
|
||||
}
|
||||
InvalidAsinDataEntity entity = new InvalidAsinDataEntity();
|
||||
entity.setDataValue(asin);
|
||||
entity.setBrand(brand);
|
||||
entity.setRecordSource("AUTO");
|
||||
return entity;
|
||||
}
|
||||
|
||||
private static String normalize(String value) {
|
||||
if (value == null) {
|
||||
return "";
|
||||
}
|
||||
String normalized = value.replace(String.valueOf((char) 0xFEFF), "")
|
||||
.replace((char) 0x3000, ' ')
|
||||
.replace("\r\n", " ")
|
||||
.replace("\r", " ")
|
||||
.replace("\n", " ")
|
||||
.replace("\t", " ")
|
||||
.trim();
|
||||
return WHITESPACE_PATTERN.matcher(normalized).replaceAll(" ");
|
||||
}
|
||||
}
|
||||
+17
@@ -2,8 +2,25 @@ package com.nanri.aiimage.modules.invalidasin.mapper;
|
||||
|
||||
import com.baomidou.mybatisplus.core.mapper.BaseMapper;
|
||||
import com.nanri.aiimage.modules.invalidasin.model.entity.InvalidAsinDataEntity;
|
||||
import org.apache.ibatis.annotations.Insert;
|
||||
import org.apache.ibatis.annotations.Mapper;
|
||||
import org.apache.ibatis.annotations.Param;
|
||||
|
||||
import java.util.List;
|
||||
|
||||
@Mapper
|
||||
public interface InvalidAsinDataMapper extends BaseMapper<InvalidAsinDataEntity> {
|
||||
|
||||
/** 批量 INSERT IGNORE:命中唯一键 (data_value, brand) 的重复行静默跳过,幂等。 */
|
||||
@Insert("""
|
||||
<script>
|
||||
INSERT IGNORE INTO biz_invalid_asin_data
|
||||
(data_value, brand, record_source, created_at, updated_at)
|
||||
VALUES
|
||||
<foreach collection="rows" item="row" separator=",">
|
||||
(#{row.dataValue}, #{row.brand}, #{row.recordSource}, #{row.createdAt}, #{row.updatedAt})
|
||||
</foreach>
|
||||
</script>
|
||||
""")
|
||||
int insertBatchIgnore(@Param("rows") List<InvalidAsinDataEntity> rows);
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user