properly check image format for storing and distance hashing

config.go

@@ -34,7 +34,7 @@ import (
 )
 
 const (
-	VERSION    string = "0.3.10"
+	VERSION    string = "0.3.11"
 	Baseuri    string = "https://www.kleinanzeigen.de"
 	Listuri    string = "/s-bestandsliste.html"
 	Defaultdir string = "."

go.mod

@@ -32,6 +32,7 @@ require (
 	github.com/mitchellh/reflectwalk v1.0.2 // indirect
 	github.com/nfnt/resize v0.0.0-20180221191011-83c6a9932646 // indirect
 	github.com/pelletier/go-toml v1.9.5 // indirect
+	golang.org/x/image v0.23.0 // indirect
 	golang.org/x/net v0.23.0 // indirect
 	golang.org/x/sys v0.21.0 // indirect
 	gopkg.in/yaml.v3 v3.0.1 // indirect

go.sum

@@ -67,6 +67,8 @@ github.com/stretchr/testify v1.8.4/go.mod h1:sz/lmYIOXD/1dqDmKjjqLyZ2RngseejIcXl
 github.com/tlinden/yadu v0.1.3 h1:5cRCUmj+l5yvlM2irtpFBIJwVV2DPEgYSaWvF19FtcY=
 github.com/tlinden/yadu v0.1.3/go.mod h1:l3bRmHKL9zGAR6pnBHY2HRPxBecf7L74BoBgOOpTcUA=
 golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w=
+golang.org/x/image v0.23.0 h1:HseQ7c2OpPKTPVzNjG5fwJsOTCiiwS4QdsYi5XU6H68=
+golang.org/x/image v0.23.0/go.mod h1:wJJBTdLfCCf3tiHa1fNxpZmUI4mmoZvwMCPP0ddoNKY=
 golang.org/x/net v0.0.0-20180218175443-cbe0f9307d01/go.mod h1:mL1N/T3taQHkDXs73rZJwtUhF3w3ftmwwsq0BUmARs4=
 golang.org/x/net v0.0.0-20181114220301-adae6a3d119a/go.mod h1:mL1N/T3taQHkDXs73rZJwtUhF3w3ftmwwsq0BUmARs4=
 golang.org/x/net v0.0.0-20190606173856-1492cefac77f/go.mod h1:HSz+uSET+XFnRR8LxR5pz3Of3rY3CfYBVs4xY44aLks=

image.go

@@ -20,11 +20,16 @@ package main
 import (
 	"bytes"
 	"fmt"
-	"image/jpeg"
+	"image"
+	_ "image/gif"
+	_ "image/jpeg"
+	_ "image/png"
 	"log/slog"
 	"os"
 	"path/filepath"
 
+	_ "golang.org/x/image/webp"
+
 	"github.com/corona10/goimagehash"
 )
 
@@ -35,6 +40,7 @@ type Image struct {
 	Hash     *goimagehash.ImageHash
 	Data     *bytes.Reader
 	URI      string
+	Mime     string
 }
 
 // used for logging to avoid printing Data
@@ -49,21 +55,49 @@ func (img *Image) LogValue() slog.Value {
 // holds all images of an ad
 type Cache []*goimagehash.ImageHash
 
-func NewImage(buf *bytes.Reader, filename, uri string) *Image {
+// filename comes from the scraper, it contains directory/base w/o suffix
+func NewImage(buf *bytes.Reader, filename, uri string) (*Image, error) {
+	_, imgconfig, err := image.DecodeConfig(buf)
+	if err != nil {
+		return nil, fmt.Errorf("failed to decode image: %w", err)
+	}
+
+	buf.Seek(0, 0)
+
+	if imgconfig == "jpeg" {
+		// we're  using the  format as file  extension, but  have used
+		// "jpg" in the past, so to be backwards compatible, stay with
+		// it.
+		imgconfig = "jpg"
+	}
+
+	if imgconfig == "" {
+		return nil, fmt.Errorf("failed to process image: unknown or unsupported image format (supported: jpg,png,gif,webp)")
+	}
+
+	filename += "." + imgconfig
+
 	img := &Image{
 		Filename: filename,
 		URI:      uri,
 		Data:     buf,
+		Mime:     imgconfig,
 	}
 
-	return img
+	slog.Debug("image MIME", "mime", img.Mime)
+
+	return img, nil
 }
 
 // Calculate diff hash of the image
 func (img *Image) CalcHash() error {
-	jpgdata, err := jpeg.Decode(img.Data)
+	jpgdata, format, err := image.Decode(img.Data)
 	if err != nil {
-		return fmt.Errorf("failed to decode JPEG image: %w", err)
+		return fmt.Errorf("failed to decode image: %w", err)
+	}
+
+	if format == "" {
+		return fmt.Errorf("failed to decode image: unknown or unsupported image format (supported: jpg,png,gif,webp)")
 	}
 
 	hash1, err := goimagehash.DifferenceHash(jpgdata)
@@ -78,6 +112,10 @@ func (img *Image) CalcHash() error {
 
 // checks if 2 images are similar enough to be considered the same
 func (img *Image) Similar(hash *goimagehash.ImageHash) bool {
+	if img.Mime != "jpeg" {
+		return false
+	}
+
 	distance, err := img.Hash.Distance(hash)
 	if err != nil {
 		slog.Debug("failed to compute diff hash distance", "error", err)
@@ -133,12 +171,19 @@ func ReadImages(addir string, dont bool) (Cache, error) {
 
 			reader := bytes.NewReader(data.Bytes())
 
-			img := NewImage(reader, filename, "")
+			img, err := NewImage(reader, filename, "")
+			if err != nil {
+				return nil, err
+			}
+
 			if err := img.CalcHash(); err != nil {
 				return nil, err
 			}
 
-			slog.Debug("Caching image from file system", "image", img, "hash", img.Hash.ToString())
+			if img.Hash != nil {
+				slog.Debug("Caching image from file system", "image", img, "hash", img.Hash.ToString())
+			}
+
 			cache = append(cache, img.Hash)
 		}
 	}

scrape.go

@@ -170,7 +170,9 @@ func ScrapeImages(fetch *Fetcher, advertisement *Ad, addir string) error {
 
 	for _, imguri := range advertisement.Images {
 		imguri := imguri
-		file := filepath.Join(adpath, fmt.Sprintf("%d.jpg", img))
+
+		// we append the suffix later in NewImage() based on image format
+		file := filepath.Join(adpath, fmt.Sprintf("%d", img))
 
 		egroup.Go(func() error {
 			// wait a little
@@ -192,7 +194,11 @@ func ScrapeImages(fetch *Fetcher, advertisement *Ad, addir string) error {
 
 			reader := bytes.NewReader(buf.Bytes())
 
-			image := NewImage(reader, file, imguri)
+			image, err := NewImage(reader, file, imguri)
+			if err != nil {
+				return err
+			}
+
 			err = image.CalcHash()
 			if err != nil {
 				return err
@@ -211,7 +217,7 @@ func ScrapeImages(fetch *Fetcher, advertisement *Ad, addir string) error {
 				return fmt.Errorf("failed to seek(0) on image reader: %w", err)
 			}
 
-			err = WriteImage(file, reader)
+			err = WriteImage(image.Filename, reader)
 			if err != nil {
 				return err
 			}