lukeye/doc-parser-server
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

新增可搜索状态字段

Browse Source
This commit is contained in:
luke 2025-06-20 17:13:18 +08:00
parent 4f439361b7
commit f2bd42005a
13 changed files with 76 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
config/mysql/init.txt
View File

@ -88,6 +88,7 @@ CREATE TABLE IF NOT EXISTS oss_upload_record (
id BIGINT AUTO_INCREMENT PRIMARY KEY COMMENT '主键',
file_name VARCHAR(255) NOT NULL COMMENT '文件名',
local_rela_file_path VARCHAR(512) NOT NULL DEFAULT 'UNKNOWN' COMMENT '本地文件相对路径',
searchable_status INT NOT NULL DEFAULT 0 COMMENT '可搜索状态：0-初始状态，1-可搜索, 2-导入失败, 99-禁止搜索'
bucket_name VARCHAR(128) NOT NULL COMMENT '目标 bucket 名',
uploader VARCHAR(128) NOT NULL COMMENT '上传人用户名',
upload_time DATETIME NOT NULL COMMENT '上传时间',

22
src/main/java/com/knowledge/base/domain/common/enums/SearchableStatusEnum.java Normal file
View File

@ -0,0 +1,22 @@
package com.knowledge.base.domain.common.enums;
/**
* @author Luke.ye
* @date 2025/6/20 16:41
*/
public enum SearchableStatusEnum {
INIT(0, "未开始进行导入"),
SUCCESS(1, "成功"),
FAILED(2, "导入失败"),
PROHIBITED(99, "用户禁止文件被检索"),
;
public int code;
private String desc;
SearchableStatusEnum(int code, String desc) {
this.code = code;
this.desc = desc;
}
}

1
src/main/java/com/knowledge/base/domain/doc/model/OSRecordDO.java
View File

@ -9,6 +9,7 @@ public class OSRecordDO {
private Long id;
private String fileName;
private String localRelaFilePath;
private int searchableStatus;
private String bucketName;
private String uploader;
private String objectPath;

8
src/main/java/com/knowledge/base/domain/doc/repository/iface/OSRecordRepository.java
View File

@ -25,6 +25,14 @@ public interface OSRecordRepository {
*/
Optional<OSRecord> getLatestRecordByRelaPath(String localRelaFilePath);
/**
* 更新上传记录中文档的可搜索状态
* @param localRelaFilePath
* @param searchableStatus
* @return
*/
boolean updateSearchableStatusByRelaPath(String localRelaFilePath, int searchableStatus);
/**
* 批量获取
* @param recordIds

1
src/main/java/com/knowledge/base/domain/doc/repository/po/OSRecord.java
View File

@ -14,6 +14,7 @@ public class OSRecord {
private String fileName;
private String localRelaFilePath;
private int searchableStatus;
private String bucketName;
private String uploader;
private String objectPath;

10
src/main/java/com/knowledge/base/domain/doc/service/iface/FileDomainService.java
View File

@ -32,6 +32,14 @@ public interface FileDomainService {
*/
Optional<OSRecordDO> getLatestRecordByRelaPath(String localRelaFilePath);
/**
* 更新上传记录中文档的可搜索状态
* @param localRelaFilePath
* @param searchableStatus
* @return
*/
boolean updateSearchableStatusByRelaPath(String localRelaFilePath, int searchableStatus);
/**
* 批量获取上传记录
* @param recordIds
@ -45,4 +53,6 @@ public interface FileDomainService {
* @return
*/
boolean batchRemoveOSFiles(List<Long> recordIds);
}

8
src/main/java/com/knowledge/base/domain/doc/service/impl/AbstractBaseFileImporter.java
View File

@ -8,6 +8,7 @@ import cn.hutool.json.JSONUtil;
import com.google.common.collect.Lists;
import com.google.common.collect.Maps;
import com.knowledge.base.domain.common.enums.DocMetaPropEnum;
import com.knowledge.base.domain.common.enums.SearchableStatusEnum;
import com.knowledge.base.domain.doc.model.FileEsField;
import com.knowledge.base.domain.doc.model.FileEsModel;
import com.knowledge.base.domain.doc.model.OSRecordDO;
@ -143,9 +144,10 @@ public abstract class AbstractBaseFileImporter implements DocumentImporter {
// === Step 5: 构建待写入文档 ===
FileEsModel fileEsModel = buildDocument(fileNameWithSuffix, localRelaFilePath, content, localMTime, extInfo);
// === Step 6: 写入 Elasticsearch 并更新缓存 ===
esGateway.saveDoc(fileEsModel);
logger.info("导入成功: {}", localRelaFilePath);
// === Step 6: 写入 Elasticsearch 并更新 导入状态&缓存 ===
boolean saved = esGateway.saveDoc(fileEsModel);
logger.info("导入结束: relativePah: {}, status: {}", localRelaFilePath, saved);
fileDomainService.updateSearchableStatusByRelaPath(localRelaFilePath, saved ? SearchableStatusEnum.SUCCESS.code : SearchableStatusEnum.FAILED.code);
String metaJson = transToMetaJson(BeanUtil.beanToMap(fileEsModel, false, true));
fileCacheService.cacheMeta(localRelaFilePath, metaJson, ConstantConfig.FILE_META_CACHE_EXPIRED_MINUTES);

5
src/main/java/com/knowledge/base/domain/doc/service/impl/FileDomainServiceImpl.java
View File

@ -78,6 +78,11 @@ public class FileDomainServiceImpl implements FileDomainService {
return Optional.of(BeanConvertUtil.convert(osRecordOpt.get(), OSRecordDO.class));
}
@Override
public boolean updateSearchableStatusByRelaPath(String localRelaFilePath, int searchableStatus) {
return osRecordRepository.updateSearchableStatusByRelaPath(localRelaFilePath, searchableStatus);
}
@Override
public List<OSRecordDO> batchQueryOSRecord(List<Long> recordIds) {
return BeanConvertUtil.convertList(osRecordRepository.batchQueryOSRecord(recordIds), OSRecordDO.class);

1
src/main/java/com/knowledge/base/infrastructure/north/controller/FileQueryController.java
View File

@ -116,6 +116,7 @@ public class FileQueryController {
item.put("fileName", dto.getFileName());
item.put("url", dto.getUrl());
item.put("recordId", dto.getId());
item.put("searchableStatus", dto.getSearchableStatus());
// 格式化上传时间
if (StrUtil.isNotBlank(dto.getUploadTime())) {

2
src/main/java/com/knowledge/base/infrastructure/north/controller/FileWriteController.java
View File

@ -6,6 +6,7 @@ import cn.hutool.core.util.StrUtil;
import com.knowledge.base.application.service.DocAppService;
import com.knowledge.base.application.service.UserAppService;
import com.knowledge.base.domain.common.enums.DocMetaPropEnum;
import com.knowledge.base.domain.common.enums.SearchableStatusEnum;
import com.knowledge.base.domain.doc.service.FileImporterDispatcher;
import com.knowledge.base.domain.doc.service.iface.DocumentImporter;
import com.knowledge.base.infrastructure.cache.iface.FileCacheService;
@ -206,6 +207,7 @@ public class FileWriteController {
OSRecordDTO dto = new OSRecordDTO();
dto.setFileName(originFileNameWithSuffix);
dto.setLocalRelaFilePath(allowSaveToLocal ? localRelaFilePath : ConstantConfig.UNKNOWN_LOCAL_RELA_PATH);
dto.setSearchableStatus(allowSaveToLocal ? SearchableStatusEnum.INIT.code : SearchableStatusEnum.PROHIBITED.code);
dto.setUrl(url);
dto.setUploader(uploader);
dto.setBucketName(bucket);

4
src/main/java/com/knowledge/base/infrastructure/north/dto/doc/OSRecordDTO.java
View File

@ -8,10 +8,12 @@ public class OSRecordDTO {
private String fileName;
/**
* 本地相对路径文件名（不为空，则文件可搜索）
* 本地相对路径文件名
*/
private String localRelaFilePath;
private int searchableStatus;
private String bucketName;
private String uploader;
private String objectPath;

11
src/main/java/com/knowledge/base/infrastructure/repository/persistence/doc/OSRecordRepositoryImpl.java
View File

@ -4,6 +4,7 @@ import cn.hutool.core.collection.CollectionUtil;
import cn.hutool.core.util.StrUtil;
import com.baomidou.mybatisplus.core.conditions.query.LambdaQueryWrapper;
import com.baomidou.mybatisplus.core.conditions.query.QueryWrapper;
import com.baomidou.mybatisplus.core.conditions.update.UpdateWrapper;
import com.baomidou.mybatisplus.extension.plugins.pagination.Page;
import com.google.common.collect.Lists;
import com.knowledge.base.domain.common.model.PageResult;
@ -50,6 +51,16 @@ public class OSRecordRepositoryImpl implements OSRecordRepository {
return Optional.ofNullable(fileLatestRecord);
}
@Override
public boolean updateSearchableStatusByRelaPath(String localRelaFilePath, int searchableStatus) {
UpdateWrapper<OSRecord> updateWrapper = new UpdateWrapper<>();
updateWrapper.eq("local_rela_file_path", localRelaFilePath)
.set("searchable_status", searchableStatus);
int rows = osRecordMapper.update(null, updateWrapper);
return rows > 0;
}
@Override
public List<OSRecord> batchQueryOSRecord(List<Long> recordIds) {
if(CollectionUtil.isEmpty(recordIds)) {

8
src/main/java/com/knowledge/base/infrastructure/south/es/FileElasticsearchGateway.java
View File

@ -11,6 +11,7 @@ import org.elasticsearch.action.delete.DeleteRequest;
import org.elasticsearch.action.get.GetRequest;
import org.elasticsearch.action.get.GetResponse;
import org.elasticsearch.action.index.IndexRequest;
import org.elasticsearch.action.index.IndexResponse;
import org.elasticsearch.client.RequestOptions;
import org.elasticsearch.client.RestHighLevelClient;
import org.elasticsearch.index.query.BoolQueryBuilder;
@ -19,6 +20,7 @@ import org.elasticsearch.index.query.QueryBuilders;
import org.elasticsearch.index.query.TermQueryBuilder;
import org.elasticsearch.index.reindex.BulkByScrollResponse;
import org.elasticsearch.index.reindex.DeleteByQueryRequest;
import org.elasticsearch.rest.RestStatus;
import org.elasticsearch.search.SearchHit;
import org.elasticsearch.search.builder.SearchSourceBuilder;
import org.elasticsearch.action.search.SearchRequest;
@ -180,14 +182,16 @@ public class FileElasticsearchGateway {
/**
* 写入/更新文档
*/
public void saveDoc(FileEsModel docModel) {
public boolean saveDoc(FileEsModel docModel) {
try {
String docId = docModel.buildDocId();
Map<String, Object> source = BeanUtil.beanToMap(docModel, false, true);
IndexRequest request = new IndexRequest(INDEX_NAME).id(docId).source(source);
esClient.index(request, RequestOptions.DEFAULT);
IndexResponse response = esClient.index(request, RequestOptions.DEFAULT);
return response.status().getStatus() == RestStatus.OK.getStatus();
} catch (Exception e) {
log.error("ES 写入文档失败: doc={}", docModel, e);
return false;
}
}