refactor: replace pdftoppm with go-fitz for in-process PDF rendering

Switch PDF page rendering from external pdftoppm/pdftocairo subprocess calls
to github.com/gen2brain/go-fitz (MuPDF wrapper), eliminating the poppler-utils
runtime dependency. Enable CGO in Dockerfile builder stage and install gcc/musl-dev
for the static MuPDF link; runtime image remains unchanged.

internal/service/pdf_recognition_service.go

@@ -6,14 +6,13 @@ import (
 	"encoding/base64"
 	"encoding/json"
 	"fmt"
+	"image/png"
 	"io"
 	"net/http"
-	"os"
-	"os/exec"
-	"path/filepath"
-	"sort"
 	"time"
 
+	"github.com/gen2brain/go-fitz"
+
 	pdfmodel "gitea.com/texpixel/document_ai/internal/model/pdf"
 	"gitea.com/texpixel/document_ai/internal/storage/cache"
 	"gitea.com/texpixel/document_ai/internal/storage/dao"
@@ -221,8 +220,7 @@ func (s *PDFRecognitionService) processPDFTask(ctx context.Context, taskID int64
 	}
 
 	processPages := len(pageImages)
-	log.Info(ctx, "func", "processPDFTask", "msg", "开始处理PDF",
-		"task_id", taskID, "process_pages", processPages)
+	log.Info(ctx, "func", "processPDFTask", "msg", "开始处理PDF", "task_id", taskID, "process_pages", processPages)
 
 	// 逐页 OCR，结果收集
 	var pages []dao.PDFPageContent
@@ -236,8 +234,7 @@ func (s *PDFRecognitionService) processPDFTask(ctx context.Context, taskID int64
 			PageNumber: i + 1,
 			Markdown:   ocrResult.Markdown,
 		})
-		log.Info(ctx, "func", "processPDFTask", "msg", "页面OCR完成",
-			"page", i+1, "total", processPages)
+		log.Info(ctx, "func", "processPDFTask", "msg", "页面OCR完成", "page", i+1, "total", processPages)
 	}
 
 	// 序列化并写入 DB（单行）
@@ -262,47 +259,40 @@ func (s *PDFRecognitionService) processPDFTask(ctx context.Context, taskID int64
 	return nil
 }
 
-// renderPDFPages 使用 pdftoppm 将 PDF 渲染为 PNG 字节切片，最多渲染 maxPages 页
+// renderPDFPages 使用 go-fitz 将 PDF 渲染为 PNG 字节切片，最多渲染 maxPages 页
 func renderPDFPages(ctx context.Context, pdfBytes []byte, maxPages int) ([][]byte, error) {
-	tmpDir, err := os.MkdirTemp("", "pdf-ocr-*")
+	doc, err := fitz.NewFromMemory(pdfBytes)
 	if err != nil {
-		return nil, fmt.Errorf("创建临时目录失败: %w", err)
+		return nil, fmt.Errorf("打开PDF失败: %w", err)
 	}
-	defer os.RemoveAll(tmpDir)
+	defer doc.Close()
 
-	pdfPath := filepath.Join(tmpDir, "input.pdf")
-	if err := os.WriteFile(pdfPath, pdfBytes, 0600); err != nil {
-		return nil, fmt.Errorf("写入临时PDF失败: %w", err)
+	total := doc.NumPage()
+	if total == 0 {
+		return nil, fmt.Errorf("PDF不包含任何页面")
+	}
+	if maxPages > 0 && total > maxPages {
+		total = maxPages
 	}
 
-	outPrefix := filepath.Join(tmpDir, "page")
-	cmd := exec.CommandContext(ctx, "pdftoppm",
-		"-r", "150",
-		"-png",
-		"-l", fmt.Sprintf("%d", maxPages),
-		pdfPath,
-		outPrefix,
-	)
-	if out, err := cmd.CombinedOutput(); err != nil {
-		return nil, fmt.Errorf("pdftoppm失败: %w, output: %s", err, string(out))
-	}
-
-	files, err := filepath.Glob(filepath.Join(tmpDir, "page-*.png"))
-	if err != nil {
-		return nil, fmt.Errorf("查找渲染输出文件失败: %w", err)
-	}
-	if len(files) == 0 {
-		return nil, fmt.Errorf("pdftoppm未输出任何页面")
-	}
-	sort.Strings(files)
-
-	pages := make([][]byte, 0, len(files))
-	for _, f := range files {
-		data, err := os.ReadFile(f)
-		if err != nil {
-			return nil, fmt.Errorf("读取页面图片失败: %w", err)
+	pages := make([][]byte, 0, total)
+	for i := 0; i < total; i++ {
+		select {
+		case <-ctx.Done():
+			return nil, ctx.Err()
+		default:
 		}
-		pages = append(pages, data)
+
+		img, err := doc.Image(i)
+		if err != nil {
+			return nil, fmt.Errorf("渲染第%d页失败: %w", i+1, err)
+		}
+
+		var buf bytes.Buffer
+		if err := png.Encode(&buf, img); err != nil {
+			return nil, fmt.Errorf("编码第%d页PNG失败: %w", i+1, err)
+		}
+		pages = append(pages, buf.Bytes())
 	}
 
 	return pages, nil