bonus
/
smart-bid-service
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

企业库ocr 识别修改

This commit is contained in:
cwchen 2025-11-24 10:22:53 +08:00
parent 084f83e3be
commit 4ee54de8ac
3 changed files with 210 additions and 68 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
bonus-admin/src/main/resources/application-ocr.yml
View File

@ -1,6 +1,6 @@
ocr:
service:
url: http://192.168.0.37:8000/extract/ # ocr 请求地址
healthUrl: http://192.168.0.37:8000/ # ocr 服务健康检查
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
timeout: 30000 # ocr 请求超时时间
max-connections: 100

27
bonus-common/src/main/java/com/bonus/common/domain/ocr/vo/OcrResponse.java
View File

@ -3,6 +3,10 @@ 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
@ -14,15 +18,32 @@ import lombok.Data;
public class OcrResponse {
@JsonProperty("status_code")
private Integer statusCode; // 状态码
private Integer statusCode; // 状态码
@JsonProperty("message")
private String message; // 消息
private String message; // 消息
@JsonProperty("data")
private OcrData data; // 识别数据
private Map<String, String> data; // 识别数据
public boolean isSuccess() {
return statusCode != null && statusCode == 200;
}
/**
* 获取识别文本
*/
public String getText() {
if (data == null) return null;
return data.values().stream()
.filter(Objects::nonNull)
.collect(Collectors.joining(" "));
}
/**
* 获取特定字段值
*/
public String getFieldValue(String fieldName) {
return data != null ? data.get(fieldName) : null;
}
}

247
bonus-ocr/src/main/java/com/bonus/ocr/service/OcrService.java
View File

@ -17,37 +17,38 @@ import org.apache.http.entity.mime.MultipartEntityBuilder;
import org.apache.http.entity.mime.HttpMultipartMode;
import org.apache.http.entity.mime.content.FileBody;
import org.apache.http.entity.ContentType;
import java.nio.charset.StandardCharsets;
import java.io.IOException;
import java.nio.charset.StandardCharsets;
import java.util.Map;
import java.util.Optional;
/**
* @className:OcrService
* @author:cwchen
* @date:2025-10-16-10:35
* @version:1.0
* @className: OcrService
* @author: cwchen
* @date: 2025-10-16-10:35
* @version: 1.0
* @description: OCR服务实现类
*/
@Service(value = "OcrService")
@Slf4j
public class OcrService {
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("${ocr.service.url}")
private String ocrServiceUrl;
@Value("${ocr.service.healthUrl}")
private String healthUrl;
@Value("${ocr.service.timeout}")
private int timeout;
private final CloseableHttpClient httpClient;
private final ObjectMapper objectMapper;
public OcrService() {
// 配置HTTP客户端
RequestConfig requestConfig = RequestConfig.custom()
.setConnectTimeout(timeout)
.setSocketTimeout(timeout)
@ -62,71 +63,181 @@ public class OcrService {
/**
* 调用OCR服务
*
* @param ocrRequest OCR请求参数
* @return OCR响应结果
* @throws IOException 当OCR服务调用失败时抛出
*/
public OcrResponse callOcrService(OcrRequest ocrRequest) throws IOException {
validateOcrRequest(ocrRequest);
HttpPost httpPost = null;
try {
httpPost = createHttpPost(ocrRequest);
return executeOcrRequest(httpPost);
} catch (IOException e) {
log.error("调用OCR服务失败", e);
throw new IOException("OCR服务调用失败: " + e.getMessage(), e);
} finally {
cleanupResources(ocrRequest, httpPost);
}
}
/**
* 验证OCR请求参数
*/
private void validateOcrRequest(OcrRequest ocrRequest) {
if (ocrRequest == null) {
throw new IllegalArgumentException("OCR请求参数不能为空");
}
if (ocrRequest.getFile() == null || !ocrRequest.getFile().exists()) {
throw new IllegalArgumentException("OCR文件不能为空或文件不存在");
}
if (ocrRequest.getType() == null || ocrRequest.getType().trim().isEmpty()) {
throw new IllegalArgumentException("OCR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OCR请求
*/
private OcrResponse executeOcrRequest(HttpPost httpPost) throws IOException {
log.info("开始调用OCR服务识别");
try (CloseableHttpResponse response = httpClient.execute(httpPost)) {
return processHttpResponse(response);
}
}
/**
* 处理HTTP响应
*/
private OcrResponse processHttpResponse(CloseableHttpResponse response) throws IOException {
int statusCode = response.getStatusLine().getStatusCode();
String responseBody = getResponseBody(response);
log.info("OCR服务响应状态: {}", statusCode);
log.debug("OCR服务响应内容: {}", responseBody); // 改为debug级别，避免日志过大
// 检查HTTP状态码
if (statusCode != 200) {
log.error("OCR服务HTTP请求失败，状态码: {}, 响应: {}", statusCode, responseBody);
throw new IOException("OCR服务HTTP请求失败，状态码: " + statusCode);
}
OcrResponse ocrResponse = parseResponseBody(responseBody);
handleOcrResult(ocrResponse);
return ocrResponse;
}
/**
* 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 OcrResponse parseResponseBody(String responseBody) throws IOException {
try {
// 创建 multipart entity builder
MultipartEntityBuilder builder = MultipartEntityBuilder.create();
builder.setCharset(StandardCharsets.UTF_8);
builder.setMode(HttpMultipartMode.BROWSER_COMPATIBLE);
return objectMapper.readValue(responseBody, OcrResponse.class);
} catch (IOException e) {
log.error("解析OCR响应失败，响应内容: {}", responseBody, e);
throw new IOException("解析OCR响应失败: " + e.getMessage(), e);
}
}
// 添加文件字段
builder.addPart("file",
new FileBody(ocrRequest.getFile(), ContentType.MULTIPART_FORM_DATA, ocrRequest.getFile().getName()));
// 添加类型字段
builder.addTextBody("type", ocrRequest.getType(),
ContentType.TEXT_PLAIN.withCharset("UTF-8"));
// 添加解析内容字段
builder.addTextBody("fields_json", ocrRequest.getFields_json(),
ContentType.TEXT_PLAIN.withCharset("UTF-8"));
/**
* 处理OCR识别结果
*/
private void handleOcrResult(OcrResponse ocrResponse) {
if (ocrResponse.isSuccess()) {
log.info("OCR识别成功");
logOcrResults(ocrResponse);
} else {
log.warn("OCR识别失败: {}", ocrResponse.getMessage());
}
}
// 设置请求实体 - 注意：这里会自动设置正确的 Content-Type 和 boundary
httpPost.setEntity(builder.build());
// 设置 Accept 头
httpPost.setHeader("Accept", "application/json");
log.info("OCR服务开始识别");
// 执行请求
try (CloseableHttpResponse response = httpClient.execute(httpPost)) {
HttpEntity entity = response.getEntity();
String responseBody = EntityUtils.toString(entity, "UTF-8");
log.info("OCR服务响应状态: {}", response.getStatusLine().getStatusCode());
log.info("OCR服务响应内容: {}", responseBody);
// 解析响应
OcrResponse ocrResponse = objectMapper.readValue(responseBody, OcrResponse.class);
if (ocrResponse.isSuccess()) {
log.info("OCR识别成功");
// 获取识别结果
if (ocrResponse.getData() != null && ocrResponse.getData().getChatRes() != null) {
Map<String, String> chatRes = ocrResponse.getData().getChatRes();
chatRes.forEach((k, v) -> {
log.info("识别结果 - key: {}, value: {}", k, v);
});
/**
* 记录OCR识别结果
*/
private void logOcrResults(OcrResponse ocrResponse) {
Optional.ofNullable(ocrResponse.getData())
.ifPresent(data -> {
if (log.isInfoEnabled()) {
data.forEach((key, value) ->
log.info("识别结果 - key: {}, value: {}", key, value));
}
} else {
log.warn("OCR识别失败: {}", ocrResponse.getMessage());
}
});
}
return ocrResponse;
}
// return null;
} catch (Exception e) {
log.error("调用OCR服务失败", e);
return null;
// throw new IOException("OCR服务调用失败: " + e.getMessage(), e);
} finally {
// 清理临时文件
if (ocrRequest.getFile() != null && ocrRequest.getFile().exists()) {
/**
* 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);
}
httpPost.releaseConnection();
}
}
@ -135,9 +246,19 @@ public class OcrService {
*/
public void close() {
try {
httpClient.close();
if (httpClient != null) {
httpClient.close();
log.info("OCR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();
}
}