招标解析算法服务
This commit is contained in:
parent
31521edfd3
commit
a808af2289
|
|
@ -1,7 +1,7 @@
|
|||
ocr:
|
||||
service:
|
||||
url: http://192.168.0.37:9091/extract # ocr 请求地址
|
||||
healthUrl: http://192.168.0.37:9091 # ocr 服务健康检查
|
||||
convertUrl: http://192.168.0.37:10000/extract # word 转pdf 请求地址
|
||||
analysisUrl: http://192.168.0.37:10001/extract # 招标解析算法服务 请求地址
|
||||
timeout: 30000 # ocr 请求超时时间
|
||||
max-connections: 100
|
||||
|
|
@ -0,0 +1,32 @@
|
|||
package com.bonus.common.domain.ocr.vo;
|
||||
|
||||
import com.fasterxml.jackson.annotation.JsonProperty;
|
||||
import lombok.Data;
|
||||
|
||||
import java.util.Map;
|
||||
import java.util.Objects;
|
||||
import java.util.stream.Collectors;
|
||||
|
||||
/**
|
||||
* @className:OcrResponse
|
||||
* @author:cwchen
|
||||
* @date:2025-10-16-9:53
|
||||
* @version:1.0
|
||||
* @description: 招标解析响应结果
|
||||
*/
|
||||
@Data
|
||||
public class AnalysisResponse {
|
||||
|
||||
@JsonProperty("status")
|
||||
private String status; // 状态码
|
||||
|
||||
@JsonProperty("message")
|
||||
private String message; // 消息
|
||||
|
||||
@JsonProperty("data")
|
||||
private Map<String, String> data; // 识别数据
|
||||
|
||||
public boolean isSuccess() {
|
||||
return Objects.equals(status, "success");
|
||||
}
|
||||
}
|
||||
|
|
@ -74,7 +74,7 @@ public class FileUtil {
|
|||
/**
|
||||
* 根据文件名获取MIME类型
|
||||
*/
|
||||
private static String getMimeTypeByFilename(String filename) {
|
||||
public static String getMimeTypeByFilename(String filename) {
|
||||
if (filename == null) {
|
||||
return "application/octet-stream";
|
||||
}
|
||||
|
|
|
|||
|
|
@ -0,0 +1,261 @@
|
|||
package com.bonus.ocr.service;
|
||||
|
||||
import com.bonus.common.domain.ocr.dto.OcrRequest;
|
||||
import com.bonus.common.domain.ocr.vo.AnalysisResponse;
|
||||
import com.fasterxml.jackson.databind.ObjectMapper;
|
||||
import lombok.extern.slf4j.Slf4j;
|
||||
import org.apache.http.HttpEntity;
|
||||
import org.apache.http.client.config.RequestConfig;
|
||||
import org.apache.http.client.methods.CloseableHttpResponse;
|
||||
import org.apache.http.client.methods.HttpPost;
|
||||
import org.apache.http.entity.ContentType;
|
||||
import org.apache.http.entity.mime.HttpMultipartMode;
|
||||
import org.apache.http.entity.mime.MultipartEntityBuilder;
|
||||
import org.apache.http.entity.mime.content.FileBody;
|
||||
import org.apache.http.impl.client.CloseableHttpClient;
|
||||
import org.apache.http.impl.client.HttpClients;
|
||||
import org.apache.http.util.EntityUtils;
|
||||
import org.springframework.beans.factory.annotation.Value;
|
||||
import org.springframework.stereotype.Service;
|
||||
|
||||
import java.io.IOException;
|
||||
import java.nio.charset.StandardCharsets;
|
||||
import java.util.Optional;
|
||||
|
||||
/**
|
||||
* @className: 招标解析算法服务
|
||||
* @author: cwchen
|
||||
* @date: 2025-10-16-10:35
|
||||
* @version: 1.0
|
||||
* @description: 招标解析算法服务实现类
|
||||
*/
|
||||
@Service(value = "AnalysisOcrService")
|
||||
@Slf4j
|
||||
public class AnalysisOcrService {
|
||||
|
||||
private static final String UTF_8 = "UTF-8";
|
||||
private static final String FILE_PART_NAME = "file";
|
||||
private static final String TYPE_PART_NAME = "type";
|
||||
private static final String FIELDS_JSON_PART_NAME = "fields_json";
|
||||
|
||||
@Value("${ocr.service.analysisUrl}")
|
||||
private String ocrServiceUrl;
|
||||
|
||||
@Value("${ocr.service.timeout}")
|
||||
private int timeout;
|
||||
|
||||
private final CloseableHttpClient httpClient;
|
||||
private final ObjectMapper objectMapper;
|
||||
|
||||
public AnalysisOcrService() {
|
||||
RequestConfig requestConfig = RequestConfig.custom()
|
||||
.setConnectTimeout(timeout)
|
||||
.setSocketTimeout(timeout)
|
||||
.setConnectionRequestTimeout(timeout)
|
||||
.build();
|
||||
|
||||
this.httpClient = HttpClients.custom()
|
||||
.setDefaultRequestConfig(requestConfig)
|
||||
.build();
|
||||
this.objectMapper = new ObjectMapper();
|
||||
}
|
||||
|
||||
/**
|
||||
* 调用招标解析服务
|
||||
*
|
||||
* @param ocrRequest 招标解析请求参数
|
||||
* @return OCR响应结果
|
||||
* @throws IOException 当招标解析算法服务调用失败时抛出
|
||||
*/
|
||||
public AnalysisResponse callOcrService(OcrRequest ocrRequest) throws IOException {
|
||||
validateOcrRequest(ocrRequest);
|
||||
|
||||
HttpPost httpPost = null;
|
||||
try {
|
||||
httpPost = createHttpPost(ocrRequest);
|
||||
return executeOcrRequest(httpPost);
|
||||
} catch (IOException e) {
|
||||
log.error("调用招标解析算法服务失败", e);
|
||||
return null;
|
||||
} finally {
|
||||
cleanupResources(ocrRequest, httpPost);
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* 验证招标解析算法请求参数
|
||||
*/
|
||||
private void validateOcrRequest(OcrRequest ocrRequest) {
|
||||
if (ocrRequest == null) {
|
||||
throw new IllegalArgumentException("招标解析算法请求参数不能为空");
|
||||
}
|
||||
if (ocrRequest.getFile() == null || !ocrRequest.getFile().exists()) {
|
||||
throw new IllegalArgumentException("招标解析文件不能为空或文件不存在");
|
||||
}
|
||||
if (ocrRequest.getType() == null || ocrRequest.getType().trim().isEmpty()) {
|
||||
throw new IllegalArgumentException("招标解析类型不能为空");
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* 创建HTTP POST请求
|
||||
*/
|
||||
private HttpPost createHttpPost(OcrRequest ocrRequest) {
|
||||
HttpPost httpPost = new HttpPost(ocrServiceUrl);
|
||||
|
||||
MultipartEntityBuilder builder = MultipartEntityBuilder.create();
|
||||
builder.setCharset(StandardCharsets.UTF_8);
|
||||
builder.setMode(HttpMultipartMode.BROWSER_COMPATIBLE);
|
||||
|
||||
// 添加文件字段
|
||||
builder.addPart(FILE_PART_NAME,
|
||||
new FileBody(ocrRequest.getFile(),
|
||||
ContentType.MULTIPART_FORM_DATA,
|
||||
ocrRequest.getFile().getName()));
|
||||
|
||||
// 添加类型字段
|
||||
builder.addTextBody(TYPE_PART_NAME,
|
||||
ocrRequest.getType(),
|
||||
ContentType.TEXT_PLAIN.withCharset(UTF_8));
|
||||
|
||||
// 添加解析内容字段
|
||||
if (ocrRequest.getFields_json() != null) {
|
||||
builder.addTextBody(FIELDS_JSON_PART_NAME,
|
||||
ocrRequest.getFields_json(),
|
||||
ContentType.TEXT_PLAIN.withCharset(UTF_8));
|
||||
}
|
||||
|
||||
httpPost.setEntity(builder.build());
|
||||
httpPost.setHeader("Accept", "application/json");
|
||||
|
||||
return httpPost;
|
||||
}
|
||||
|
||||
/**
|
||||
* 执行招标解析算法请求
|
||||
*/
|
||||
private AnalysisResponse executeOcrRequest(HttpPost httpPost) throws IOException {
|
||||
log.info("开始调用招标解析算法服务");
|
||||
|
||||
try (CloseableHttpResponse response = httpClient.execute(httpPost)) {
|
||||
return processHttpResponse(response);
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* 处理HTTP响应
|
||||
*/
|
||||
private AnalysisResponse processHttpResponse(CloseableHttpResponse response) throws IOException {
|
||||
int statusCode = response.getStatusLine().getStatusCode();
|
||||
String responseBody = getResponseBody(response);
|
||||
|
||||
log.info("招标解析算法服务响应状态: {}", statusCode);
|
||||
log.debug("招标解析算法响应内容: {}", responseBody); // 改为debug级别，避免日志过大
|
||||
|
||||
// 检查HTTP状态码
|
||||
if (statusCode != 200) {
|
||||
log.error("招标解析算法HTTP请求失败，状态码: {}, 响应: {}", statusCode, responseBody);
|
||||
return null;
|
||||
}
|
||||
|
||||
AnalysisResponse AnalysisResponse = parseResponseBody(responseBody);
|
||||
return AnalysisResponse;
|
||||
}
|
||||
|
||||
/**
|
||||
* 获取响应体
|
||||
*/
|
||||
private String getResponseBody(CloseableHttpResponse response) throws IOException {
|
||||
HttpEntity entity = response.getEntity();
|
||||
return EntityUtils.toString(entity, UTF_8);
|
||||
}
|
||||
|
||||
/**
|
||||
* 解析响应体
|
||||
*/
|
||||
private AnalysisResponse parseResponseBody(String responseBody) throws IOException {
|
||||
try {
|
||||
return objectMapper.readValue(responseBody, AnalysisResponse.class);
|
||||
} catch (IOException e) {
|
||||
log.error("解析招标解析算法响应失败，响应内容: {}", responseBody, e);
|
||||
return null;
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* 处理OCR识别结果
|
||||
*/
|
||||
private void handleOcrResult(AnalysisResponse AnalysisResponse) {
|
||||
if (AnalysisResponse.isSuccess()) {
|
||||
log.info("OCR识别成功");
|
||||
logOcrResults(AnalysisResponse);
|
||||
} else {
|
||||
log.warn("OCR识别失败", AnalysisResponse.getMessage());
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* 记录OCR识别结果
|
||||
*/
|
||||
private void logOcrResults(AnalysisResponse AnalysisResponse) {
|
||||
Optional.ofNullable(AnalysisResponse.getData())
|
||||
.ifPresent(data -> {
|
||||
if (log.isInfoEnabled()) {
|
||||
data.forEach((key, value) ->
|
||||
log.info("识别结果 - key: {}, value: {}", key, value));
|
||||
}
|
||||
});
|
||||
}
|
||||
|
||||
/**
|
||||
* 清理资源
|
||||
*/
|
||||
private void cleanupResources(OcrRequest ocrRequest, HttpPost httpPost) {
|
||||
// 清理HTTP连接
|
||||
if (httpPost != null) {
|
||||
httpPost.releaseConnection();
|
||||
}
|
||||
|
||||
// 清理临时文件
|
||||
cleanupTempFile(ocrRequest);
|
||||
}
|
||||
|
||||
/**
|
||||
* 清理临时文件
|
||||
*/
|
||||
private void cleanupTempFile(OcrRequest ocrRequest) {
|
||||
if (ocrRequest.getFile() != null && ocrRequest.getFile().exists()) {
|
||||
try {
|
||||
boolean deleted = ocrRequest.getFile().delete();
|
||||
if (!deleted) {
|
||||
log.warn("临时文件删除失败: {}", ocrRequest.getFile().getAbsolutePath());
|
||||
} else {
|
||||
log.debug("临时文件已删除: {}", ocrRequest.getFile().getAbsolutePath());
|
||||
}
|
||||
} catch (SecurityException e) {
|
||||
log.error("删除临时文件时发生安全异常: {}", ocrRequest.getFile().getAbsolutePath(), e);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* 关闭HTTP客户端
|
||||
*/
|
||||
public void close() {
|
||||
try {
|
||||
if (httpClient != null) {
|
||||
httpClient.close();
|
||||
log.info("招标解析算法服务HTTP客户端已关闭");
|
||||
}
|
||||
} catch (IOException e) {
|
||||
log.error("关闭HTTP客户端失败", e);
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* 销毁方法，用于Spring容器关闭时调用
|
||||
*/
|
||||
public void destroy() {
|
||||
close();
|
||||
}
|
||||
}
|
||||
|
|
@ -1,6 +1,12 @@
|
|||
package com.bonus.rabbitmq.consumer;
|
||||
|
||||
import com.bonus.common.domain.ocr.dto.OcrRequest;
|
||||
import com.bonus.common.domain.ocr.vo.AnalysisResponse;
|
||||
import com.bonus.common.domain.rabbitmq.dto.RabbitMqMessage;
|
||||
import com.bonus.common.utils.FileUtil;
|
||||
import com.bonus.file.config.MinioConfig;
|
||||
import com.bonus.file.util.MinioUtil;
|
||||
import com.bonus.ocr.service.AnalysisOcrService;
|
||||
import com.fasterxml.jackson.databind.ObjectMapper;
|
||||
import com.rabbitmq.client.Channel;
|
||||
import lombok.extern.slf4j.Slf4j;
|
||||
|
|
@ -8,8 +14,11 @@ import org.springframework.amqp.core.Message;
|
|||
import org.springframework.amqp.rabbit.annotation.RabbitListener;
|
||||
import org.springframework.stereotype.Component;
|
||||
|
||||
import javax.annotation.Resource;
|
||||
import java.io.File;
|
||||
import java.io.IOException;
|
||||
import java.nio.charset.StandardCharsets;
|
||||
import java.util.Objects;
|
||||
|
||||
/**
|
||||
* @className:RabbitMQConsumerService
|
||||
|
|
@ -24,6 +33,15 @@ public class RabbitMQConsumerService {
|
|||
|
||||
private final ObjectMapper objectMapper = new ObjectMapper();
|
||||
|
||||
@Resource
|
||||
private MinioConfig minioConfig;
|
||||
|
||||
@Resource
|
||||
private MinioUtil minioUtil;
|
||||
|
||||
@Resource(name = "AnalysisOcrService")
|
||||
private AnalysisOcrService analysisOcrService;
|
||||
|
||||
@RabbitListener(
|
||||
queues = "myQueue",
|
||||
containerFactory = "multiConsumerFactory" // 使用上面配置的工厂，保证按顺序消费
|
||||
|
|
@ -34,7 +52,6 @@ public class RabbitMQConsumerService {
|
|||
|
||||
log.info("🎯 RabbitMQConsumerService 收到原始消息 - deliveryTag: {}, body: {}",
|
||||
deliveryTag, rawBody);
|
||||
|
||||
RabbitMqMessage message;
|
||||
try {
|
||||
// 手动反序列化 JSON 为 RabbitMqMessage，避免类型映射问题导致监听方法不执行
|
||||
|
|
@ -56,14 +73,7 @@ public class RabbitMQConsumerService {
|
|||
try {
|
||||
log.info("🛠 开始处理消息内容 - ID: {}, 任务: {}, 业务数据: {}",
|
||||
messageId, taskName, message.getBusinessData());
|
||||
|
||||
// 模拟少量业务处理耗时，避免长时间阻塞
|
||||
/*try {
|
||||
Thread.sleep(500);
|
||||
} catch (InterruptedException e) {
|
||||
Thread.currentThread().interrupt();
|
||||
}*/
|
||||
|
||||
processBusiness(message);
|
||||
// 处理成功，手动确认消息
|
||||
channel.basicAck(deliveryTag, false);
|
||||
log.info("✅ 消息处理完成并确认 - ID: {}, 投递标签: {}", messageId, deliveryTag);
|
||||
|
|
@ -78,4 +88,75 @@ public class RabbitMQConsumerService {
|
|||
}
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* 处理招标解析算法服务业务
|
||||
* @param message
|
||||
* @return void
|
||||
* @author cwchen
|
||||
* @date 2025/11/29 13:25
|
||||
*/
|
||||
private void processBusiness(RabbitMqMessage message) {
|
||||
String uploadPath = message.getUploadPath();
|
||||
File fileFromMinio = getFileFromMinio(uploadPath);
|
||||
AnalysisResponse ocrResponse = performAnalysisRecognition(fileFromMinio);
|
||||
|
||||
}
|
||||
|
||||
/**
|
||||
* 从minio中获取文件
|
||||
* @param uploadPath
|
||||
* @return File
|
||||
* @author cwchen
|
||||
* @date 2025/11/29 13:23
|
||||
*/
|
||||
private File getFileFromMinio(String uploadPath) {
|
||||
try {
|
||||
File file = minioUtil.getFileFromMinio(minioConfig.getBucketName(), uploadPath);
|
||||
if (file == null || !file.exists()) {
|
||||
throw new RuntimeException("Minio文件不存在: " + uploadPath);
|
||||
}
|
||||
return file;
|
||||
} catch (Exception e) {
|
||||
throw new RuntimeException("获取Minio文件失败: " + uploadPath, e);
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* 调用算法服务
|
||||
* @param file
|
||||
* @return AnalysisResponse
|
||||
* @author cwchen
|
||||
* @date 2025/11/29 13:30
|
||||
*/
|
||||
private AnalysisResponse performAnalysisRecognition(File file) {
|
||||
try {
|
||||
OcrRequest ocrRequest = buildOcrRequest(file);
|
||||
AnalysisResponse ocrResponse = analysisOcrService.callOcrService(ocrRequest);
|
||||
// 修复：检查 招标解析算法服务 响应是否为 null
|
||||
if (Objects.isNull(ocrResponse)) {
|
||||
throw new RuntimeException("招标解析算法服务返回结果为空");
|
||||
}
|
||||
log.info("OCR识别成功 - 数据: {}", ocrResponse.getData());
|
||||
return ocrResponse;
|
||||
|
||||
} catch (Exception e) {
|
||||
log.error("OCR识别失败", e);
|
||||
throw new RuntimeException("OCR识别失败: " + e.getMessage(), e);
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* 构建招标解析算法服务请求
|
||||
* @param file
|
||||
* @return OcrRequest
|
||||
* @author cwchen
|
||||
* @date 2025/11/29 13:29
|
||||
*/
|
||||
private OcrRequest buildOcrRequest(File file) {
|
||||
OcrRequest ocrRequest = new OcrRequest();
|
||||
ocrRequest.setFile(file);
|
||||
ocrRequest.setType(FileUtil.getMimeTypeByFilename(file.getName()));
|
||||
return ocrRequest;
|
||||
}
|
||||
}
|
||||
Loading…
Reference in New Issue