feat(scanner): render PDF first page as cover fallback via pdftoppm

extractPDFCover previously only saved embedded raster images from page 1, so vector/text-first-page PDFs (e.g. InDesign exports like Data Structures the Fun Way) ended up with no cover and a dashboard placeholder. When no embedded image is found it now falls back to rendering page 1 with pdftoppm (poppler-utils), saving the same {pdf}.cover.jpg sidecar. Also adds MediaScanner.RescanMediaItem, which re-extracts metadata for a single media item (resolving its on-disk path from library folders) so previously imported books can backfill covers without a full force rescan. Dockerfile installs poppler-utils in the final and test-runner stages.
2026-09-11 23:08:20 -04:00
parent 6aa958c78f
commit 0614795ca2
2 changed files with 126 additions and 42 deletions
@@ -43,7 +43,7 @@ RUN --mount=type=cache,target=/root/go/pkg/mod \
 # This stage is ONLY used for running tests, never deployed to production
 FROM golang:1.26-alpine AS test-runner

-RUN apk --no-cache add ca-certificates curl
+RUN apk --no-cache add ca-certificates curl poppler-utils

 WORKDIR /app

@@ -65,7 +65,7 @@ CMD ["go", "test", "./cmd/server/tests", "-v", "-timeout", "5m", "-parallel=1",
 # Final stage
 FROM alpine:latest

-RUN apk --no-cache add ca-certificates curl
+RUN apk --no-cache add ca-certificates curl poppler-utils

 # Install kepubify for EPUB→KEPUB conversion
 RUN wget -O /usr/bin/kepubify https://github.com/pgaskin/kepubify/releases/latest/download/kepubify-linux-64bit \
@@ -23,6 +23,7 @@ import (
 	"io"
 	"io/fs"
 	"os"
+	"os/exec"
 	"path/filepath"
 	"regexp"
 	"strconv"
@@ -2078,56 +2079,94 @@ func (s *MediaScanner) extractPDFCover(pdfPath string) (string, error) {
 	// Use pdfcpu API to extract images from first page
 	// ExtractImagesFile(inFile, outDir string, selectedPages []string, conf *model.Configuration) error
 	err = pdfcpuapi.ExtractImagesFile(pdfPath, tmpDir, []string{"1"}, nil)
+	if err == nil {
+		// Check for extracted images in the temp directory
+		entries, err := os.ReadDir(tmpDir)
+		if err == nil && len(entries) > 0 {
+			// Find the largest image (likely the cover)
+			var largestImage string
+			var largestSize int64
+
+			for _, entry := range entries {
+				if entry.IsDir() {
+					continue
+				}
+				info, err := entry.Info()
+				if err != nil {
+					continue
+				}
+				// Skip very small files (likely thumbnails or icons)
+				if info.Size() < 1000 {
+					continue
+				}
+				if info.Size() > largestSize {
+					largestImage = filepath.Join(tmpDir, entry.Name())
+					largestSize = info.Size()
+				}
+			}
+
+			if largestImage != "" {
+				// Read the image
+				imageData, err := os.ReadFile(largestImage)
+				if err == nil && len(imageData) > 0 {
+					// Save cover to disk (same pattern as comics: {pdf_path}.cover.jpg)
+					coverPath := pdfPath + ".cover.jpg"
+					if err := os.WriteFile(coverPath, imageData, 0644); err != nil {
+						return "", fmt.Errorf("failed to write cover file: %v", err)
+					}
+
+					return coverPath, nil
+				}
+			}
+		}
+	}
+
+	// No embedded raster cover found (e.g. vector/text first page) - fall back
+	// to rendering the first page with pdftoppm (poppler-utils).
+	return s.renderPDFCoverPage(pdfPath), nil
+}
+
+// renderPDFCoverPage renders the first page of a PDF file to a JPEG image
+// using pdftoppm. It saves the cover next to the PDF ({pdf_path}.cover.jpg).
+// Returns the path to the saved cover, or empty string if rendering failed.
+func (s *MediaScanner) renderPDFCoverPage(pdfPath string) string {
+	if _, err := exec.LookPath("pdftoppm"); err != nil {
+		fmt.Printf("Warning: pdftoppm not available, skipping PDF cover render for %s\n", pdfPath)
+		return ""
+	}
+
+	tmpDir, err := os.MkdirTemp("", "pdf-render-")
 	if err != nil {
-		// No images found or extraction failed - this is OK, just return empty
-		return "", nil
+		fmt.Printf("Warning: failed to create temp dir for PDF cover render %s: %v\n", pdfPath, err)
+		return ""
+	}
+	defer func() {
+		if err := os.RemoveAll(tmpDir); err != nil {
+			fmt.Printf("Warning: failed to remove temp directory %s: %v\n", tmpDir, err)
+		}
+	}()
+
+	outPrefix := filepath.Join(tmpDir, "cover")
+	cmd := exec.Command("pdftoppm", "-jpeg", "-f", "1", "-l", "1", "-singlefile", "-r", "150", pdfPath, outPrefix)
+	if output, err := cmd.CombinedOutput(); err != nil {
+		fmt.Printf("Warning: failed to render PDF cover from %s: %v, output: %s\n", pdfPath, err, string(output))
+		return ""
 	}

-	// Check for extracted images in the temp directory
-	entries, err := os.ReadDir(tmpDir)
-	if err != nil || len(entries) == 0 {
-		return "", nil
-	}
-
-	// Find the largest image (likely the cover)
-	var largestImage string
-	var largestSize int64
-
-	for _, entry := range entries {
-		if entry.IsDir() {
-			continue
-		}
-		info, err := entry.Info()
-		if err != nil {
-			continue
-		}
-		// Skip very small files (likely thumbnails or icons)
-		if info.Size() < 1000 {
-			continue
-		}
-		if info.Size() > largestSize {
-			largestImage = filepath.Join(tmpDir, entry.Name())
-			largestSize = info.Size()
-		}
-	}
-
-	if largestImage == "" {
-		return "", nil
-	}
-
-	// Read the image
-	imageData, err := os.ReadFile(largestImage)
-	if err != nil || len(imageData) == 0 {
-		return "", nil
+	imageData, err := os.ReadFile(outPrefix + ".jpg")
+	if err != nil || len(imageData) < 1000 {
+		fmt.Printf("Warning: PDF cover render produced no usable image for %s\n", pdfPath)
+		return ""
 	}

 	// Save cover to disk (same pattern as comics: {pdf_path}.cover.jpg)
 	coverPath := pdfPath + ".cover.jpg"
 	if err := os.WriteFile(coverPath, imageData, 0644); err != nil {
-		return "", fmt.Errorf("failed to write cover file: %v", err)
+		fmt.Printf("Warning: failed to write rendered PDF cover for %s: %v\n", pdfPath, err)
+		return ""
 	}

-	return coverPath, nil
+	return coverPath
 }

 // ComicInfo represents metadata from ComicInfo.xml
@@ -2618,6 +2657,51 @@ func (s *MediaScanner) updateMediaItem(ctx context.Context, mediaItemID pgtype.U
 	return err
 }

+// RescanMediaItem re-extracts metadata for a single media item and updates it.
+// It is the per-book rescan used by the Edit Metadata dialog and backfills
+// covers for items imported before the PDF render fallback existed.
+func (s *MediaScanner) RescanMediaItem(ctx context.Context, mediaItemID pgtype.UUID) error {
+	item, err := s.db.GetMediaItem(ctx, mediaItemID)
+	if err != nil {
+		return fmt.Errorf("media item not found: %w", err)
+	}
+
+	folders, err := s.db.GetLibraryFolders(ctx, item.LibraryID)
+	if err != nil || len(folders) == 0 {
+		return fmt.Errorf("no library folders found for library")
+	}
+
+	folderPaths := make([]string, 0, len(folders))
+	for _, folder := range folders {
+		folderPaths = append(folderPaths, folder.FolderPath)
+	}
+	s.folders = folderPaths
+
+	var fullPath string
+	for _, folder := range folders {
+		candidate := filepath.Join(folder.FolderPath, item.FilePath)
+		if _, err := os.Stat(candidate); err == nil {
+			fullPath = candidate
+			break
+		}
+	}
+	if fullPath == "" {
+		return fmt.Errorf("media file not found on disk: %s", item.FilePath)
+	}
+
+	info, err := os.Stat(fullPath)
+	if err != nil {
+		return fmt.Errorf("failed to stat media file: %w", err)
+	}
+
+	if err := s.updateMediaItem(ctx, mediaItemID, fullPath, info); err != nil {
+		return fmt.Errorf("failed to update media item: %w", err)
+	}
+	s.recomputeHashInfo(ctx, mediaItemID, item.LibraryID, fullPath)
+
+	return nil
+}
+
 func (s *MediaScanner) getMediaItemByFilePath(ctx context.Context, filePath string, libraryID pgtype.UUID) (database.MediaItems, error) {
 	return s.db.GetMediaItemByFilePath(ctx, database.GetMediaItemByFilePathParams{
 		FilePath:  s.getRelativePath(filePath),