feat: replace export url

internal/model/formula/response.go

@@ -22,3 +22,10 @@ type GetFormulaTaskResponse struct {
 type FormulaRecognitionResponse struct {
 	Result string `json:"result"`
 }
+
+// ImageOCRResponse 图片OCR接口返回的响应
+type ImageOCRResponse struct {
+	Latex           string        `json:"latex"`            // LaTeX 格式内容
+	Markdown        string        `json:"markdown"`         // Markdown 格式内容
+	MathML          string        `json:"mathml"`           // MathML 格式（无公式时为空）
+}

internal/service/recognition_service.go

@@ -22,6 +22,7 @@ import (
 	"gitea.com/bitwsd/document_ai/pkg/constant"
 	"gitea.com/bitwsd/document_ai/pkg/httpclient"
 	"gitea.com/bitwsd/document_ai/pkg/oss"
+	"gitea.com/bitwsd/document_ai/pkg/requestid"
 	"gitea.com/bitwsd/document_ai/pkg/utils"
 	"gorm.io/gorm"
 )
@@ -511,8 +512,8 @@ func (s *RecognitionService) processFormulaTask(ctx context.Context, taskID int6
 	// 设置Content-Type头为application/json
 	headers := map[string]string{"Content-Type": "application/json", utils.RequestIDHeaderKey: utils.GetRequestIDFromContext(ctx)}
 
-	// 发送请求时会使用带超时的context
-	resp, err := s.httpClient.RequestWithRetry(ctx, http.MethodPost, "https://cloud.texpixel.com:10443/vlm/formula/predict", bytes.NewReader(jsonData), headers)
+	// 发送请求到新的 OCR 接口
+	resp, err := s.httpClient.RequestWithRetry(ctx, http.MethodPost, "https://cloud.texpixel.com:10443/doc_process/v1/image/ocr", bytes.NewReader(jsonData), headers)
 	if err != nil {
 		if ctx.Err() == context.DeadlineExceeded {
 			log.Error(ctx, "func", "processFormulaTask", "msg", "请求超时")
@@ -532,12 +533,18 @@ func (s *RecognitionService) processFormulaTask(ctx context.Context, taskID int6
 	log.Info(ctx, "func", "processFormulaTask", "msg", "响应内容", "body", body.String())
 
 	// 解析 JSON 响应
-	var formulaResp formula.FormulaRecognitionResponse
-	if err := json.Unmarshal(body.Bytes(), &formulaResp); err != nil {
+	var ocrResp formula.ImageOCRResponse
+	if err := json.Unmarshal(body.Bytes(), &ocrResp); err != nil {
 		log.Error(ctx, "func", "processFormulaTask", "msg", "解析响应JSON失败", "error", err)
 		return err
 	}
-	err = resultDao.Create(tx, dao.RecognitionResult{TaskID: taskID, TaskType: dao.TaskTypeFormula, Latex: formulaResp.Result})
+	err = resultDao.Create(tx, dao.RecognitionResult{
+		TaskID:   taskID,
+		TaskType: dao.TaskTypeFormula,
+		Latex:    ocrResp.Latex,
+		Markdown: ocrResp.Markdown,
+		MathML:   ocrResp.MathML,
+	})
 	if err != nil {
 		log.Error(ctx, "func", "processFormulaTask", "msg", "保存任务结果失败", "error", err)
 		return err
@@ -705,15 +712,19 @@ func (s *RecognitionService) processOneTask(ctx context.Context) {
 	}
 
 	ctx = context.WithValue(ctx, utils.RequestIDKey, task.TaskUUID)
-	log.Info(ctx, "func", "processFormulaQueue", "msg", "获取任务成功", "task_id", taskID)
 
-	err = s.processBaiduOCRTask(ctx, taskID, task.FileURL)
-	if err != nil {
-		log.Error(ctx, "func", "processFormulaQueue", "msg", "处理任务失败", "error", err)
-		return
-	}
+	// 使用 gls 设置 request_id，确保在整个任务处理过程中可用
+	requestid.SetRequestID(task.TaskUUID, func() {
+		log.Info(ctx, "func", "processFormulaQueue", "msg", "获取任务成功", "task_id", taskID)
 
-	log.Info(ctx, "func", "processFormulaQueue", "msg", "处理任务成功", "task_id", taskID)
+		err = s.processFormulaTask(ctx, taskID, task.FileURL)
+		if err != nil {
+			log.Error(ctx, "func", "processFormulaQueue", "msg", "处理任务失败", "error", err)
+			return
+		}
+
+		log.Info(ctx, "func", "processFormulaQueue", "msg", "处理任务成功", "task_id", taskID)
+	})
 }
 
 // processMathpixTask 使用 Mathpix API 处理公式识别任务（用于增强识别）

internal/service/task.go

@@ -3,10 +3,10 @@ package service
 import (
 	"bytes"
 	"context"
+	"encoding/json"
 	"errors"
 	"fmt"
 	"io"
-	"mime/multipart"
 	"net/http"
 	"strings"
 
@@ -151,19 +151,29 @@ func (svc *TaskService) ExportTask(ctx context.Context, req *task.ExportTaskRequ
 		return nil, "", errors.New("markdown not found")
 	}
 
-	// call http://localhost:8055/export
-	body := &bytes.Buffer{}
-	writer := multipart.NewWriter(body)
-	_ = writer.WriteField("markdown_input", markdown)
-	_ = writer.WriteField("type", req.Type)
-	writer.Close()
+	// 获取文件名（去掉扩展名）
+	filename := strings.TrimSuffix(recognitionTask.FileName, "."+strings.ToLower(strings.Split(recognitionTask.FileName, ".")[len(strings.Split(recognitionTask.FileName, "."))-1]))
+	if filename == "" {
+		filename = "texpixel"
+	}
 
-	httpReq, err := http.NewRequestWithContext(ctx, http.MethodPost, "https://cloud.texpixel.com:10443/doc_converter/v1/export", body)
+	// 构建 JSON 请求体
+	requestBody := map[string]string{
+		"markdown": markdown,
+		"filename": filename,
+	}
+	jsonData, err := json.Marshal(requestBody)
+	if err != nil {
+		log.Error(ctx, "func", "ExportTask", "msg", "json marshal failed", "error", err)
+		return nil, "", err
+	}
+
+	httpReq, err := http.NewRequestWithContext(ctx, http.MethodPost, "https://cloud.texpixel.com:10443/doc_process/v1/convert/file", bytes.NewReader(jsonData))
 	if err != nil {
 		log.Error(ctx, "func", "ExportTask", "msg", "create http request failed", "error", err)
 		return nil, "", err
 	}
-	httpReq.Header.Set("Content-Type", writer.FormDataContentType())
+	httpReq.Header.Set("Content-Type", "application/json")
 
 	client := &http.Client{}
 	resp, err := client.Do(httpReq)
@@ -184,16 +194,8 @@ func (svc *TaskService) ExportTask(ctx context.Context, req *task.ExportTaskRequ
 		return nil, "", err
 	}
 
-	// determine content type based on export type
-	var contentType string
-	switch req.Type {
-	case "pdf":
-		contentType = "application/pdf"
-	case "docx":
-		contentType = "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
-	default:
-		contentType = "application/octet-stream"
-	}
+	// 新接口只返回 DOCX 格式
+	contentType := "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
 
 	return fileData, contentType, nil
 }