Add Image Diff for SVG files (#14867)

* Added type sniffer.

* Switched content detection from base to typesniffer.

* Added GuessContentType to Blob.

* Moved image info logic to client.
Added support for SVG images in diff.

* Restore old blocked svg behaviour.

* Added missing image formats.

* Execute image diff only when container is visible.

* add margin to spinner

* improve BIN tag on image diffs

* Default to render view.

* Show image diff on incomplete diff.

Co-authored-by: silverwind <[email protected]>
Co-authored-by: Lunny Xiao <[email protected]>
Co-authored-by: Lauris BH <[email protected]>

Author: 4 people

modules/avatar/avatar.go

@@ -10,8 +10,9 @@ import (
 	"image"
 	"image/color/palette"
 
-	// Enable PNG support:
-	_ "image/png"
+	_ "image/gif"  // for processing gif images
+	_ "image/jpeg" // for processing jpeg images
+	_ "image/png"  // for processing png images
 
 	"code.gitea.io/gitea/modules/setting"
 	"code.gitea.io/gitea/modules/util"

modules/base/tool.go

@@ -12,10 +12,8 @@ import (
 	"encoding/hex"
 	"errors"
 	"fmt"
-	"net/http"
 	"os"
 	"path/filepath"
-	"regexp"
 	"runtime"
 	"strconv"
 	"strings"
@@ -30,15 +28,6 @@ import (
 	"github.com/dustin/go-humanize"
 )
 
-// Use at most this many bytes to determine Content Type.
-const sniffLen = 512
-
-// SVGMimeType MIME type of SVG images.
-const SVGMimeType = "image/svg+xml"
-
-var svgTagRegex = regexp.MustCompile(`(?si)\A\s*(?:(<!--.*?-->|<!DOCTYPE\s+svg([\s:]+.*?>|>))\s*)*<svg[\s>\/]`)
-var svgTagInXMLRegex = regexp.MustCompile(`(?si)\A<\?xml\b.*?\?>\s*(?:(<!--.*?-->|<!DOCTYPE\s+svg([\s:]+.*?>|>))\s*)*<svg[\s>\/]`)
-
 // EncodeMD5 encodes string to md5 hex value.
 func EncodeMD5(str string) string {
 	m := md5.New()
@@ -276,63 +265,6 @@ func IsLetter(ch rune) bool {
 	return 'a' <= ch && ch <= 'z' || 'A' <= ch && ch <= 'Z' || ch == '_' || ch >= 0x80 && unicode.IsLetter(ch)
 }
 
-// DetectContentType extends http.DetectContentType with more content types.
-func DetectContentType(data []byte) string {
-	ct := http.DetectContentType(data)
-
-	if len(data) > sniffLen {
-		data = data[:sniffLen]
-	}
-
-	if setting.UI.SVG.Enabled &&
-		((strings.Contains(ct, "text/plain") || strings.Contains(ct, "text/html")) && svgTagRegex.Match(data) ||
-			strings.Contains(ct, "text/xml") && svgTagInXMLRegex.Match(data)) {
-
-		// SVG is unsupported.  https://github.com/golang/go/issues/15888
-		return SVGMimeType
-	}
-	return ct
-}
-
-// IsRepresentableAsText returns true if file content can be represented as
-// plain text or is empty.
-func IsRepresentableAsText(data []byte) bool {
-	return IsTextFile(data) || IsSVGImageFile(data)
-}
-
-// IsTextFile returns true if file content format is plain text or empty.
-func IsTextFile(data []byte) bool {
-	if len(data) == 0 {
-		return true
-	}
-	return strings.Contains(DetectContentType(data), "text/")
-}
-
-// IsImageFile detects if data is an image format
-func IsImageFile(data []byte) bool {
-	return strings.Contains(DetectContentType(data), "image/")
-}
-
-// IsSVGImageFile detects if data is an SVG image format
-func IsSVGImageFile(data []byte) bool {
-	return strings.Contains(DetectContentType(data), SVGMimeType)
-}
-
-// IsPDFFile detects if data is a pdf format
-func IsPDFFile(data []byte) bool {
-	return strings.Contains(DetectContentType(data), "application/pdf")
-}
-
-// IsVideoFile detects if data is an video format
-func IsVideoFile(data []byte) bool {
-	return strings.Contains(DetectContentType(data), "video/")
-}
-
-// IsAudioFile detects if data is an video format
-func IsAudioFile(data []byte) bool {
-	return strings.Contains(DetectContentType(data), "audio/")
-}
-
 // EntryIcon returns the octicon class for displaying files/directories
 func EntryIcon(entry *git.TreeEntry) string {
 	switch {

modules/base/tool_test.go

@@ -5,7 +5,6 @@
 package base
 
 import (
-	"encoding/base64"
 	"os"
 	"testing"
 	"time"
@@ -246,97 +245,6 @@ func TestIsLetter(t *testing.T) {
 	assert.False(t, IsLetter(0x93))
 }
 
-func TestDetectContentTypeLongerThanSniffLen(t *testing.T) {
-	// Pre-condition: Shorter than sniffLen detects SVG.
-	assert.Equal(t, "image/svg+xml", DetectContentType([]byte(`<!-- Comment --><svg></svg>`)))
-	// Longer than sniffLen detects something else.
-	assert.Equal(t, "text/plain; charset=utf-8", DetectContentType([]byte(`<!--
-Comment Comment Comment Comment Comment Comment Comment Comment Comment Comment
-Comment Comment Comment Comment Comment Comment Comment Comment Comment Comment
-Comment Comment Comment Comment Comment Comment Comment Comment Comment Comment
-Comment Comment Comment Comment Comment Comment Comment Comment Comment Comment
-Comment Comment Comment Comment Comment Comment Comment Comment Comment Comment
-Comment Comment Comment Comment Comment Comment Comment Comment Comment Comment
-Comment Comment Comment --><svg></svg>`)))
-}
-
-// IsRepresentableAsText
-
-func TestIsTextFile(t *testing.T) {
-	assert.True(t, IsTextFile([]byte{}))
-	assert.True(t, IsTextFile([]byte("lorem ipsum")))
-}
-
-func TestIsImageFile(t *testing.T) {
-	png, _ := base64.StdEncoding.DecodeString("iVBORw0KGgoAAAANSUhEUgAAAAoAAAAKCAYAAACNMs+9AAAAG0lEQVQYlWN4+vTpf3SMDTAMBYXYBLFpHgoKAeiOf0SGE9kbAAAAAElFTkSuQmCC")
-	assert.True(t, IsImageFile(png))
-	assert.False(t, IsImageFile([]byte("plain text")))
-}
-
-func TestIsSVGImageFile(t *testing.T) {
-	assert.True(t, IsSVGImageFile([]byte("<svg></svg>")))
-	assert.True(t, IsSVGImageFile([]byte("    <svg></svg>")))
-	assert.True(t, IsSVGImageFile([]byte(`<svg width="100"></svg>`)))
-	assert.True(t, IsSVGImageFile([]byte("<svg/>")))
-	assert.True(t, IsSVGImageFile([]byte(`<?xml version="1.0" encoding="UTF-8"?><svg></svg>`)))
-	assert.True(t, IsSVGImageFile([]byte(`<!-- Comment -->
-	<svg></svg>`)))
-	assert.True(t, IsSVGImageFile([]byte(`<!-- Multiple -->
-	<!-- Comments -->
-	<svg></svg>`)))
-	assert.True(t, IsSVGImageFile([]byte(`<!-- Multiline
-	Comment -->
-	<svg></svg>`)))
-	assert.True(t, IsSVGImageFile([]byte(`<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1 Basic//EN"
-	"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11-basic.dtd">
-	<svg></svg>`)))
-	assert.True(t, IsSVGImageFile([]byte(`<?xml version="1.0" encoding="UTF-8"?>
-	<!-- Comment -->
-	<svg></svg>`)))
-	assert.True(t, IsSVGImageFile([]byte(`<?xml version="1.0" encoding="UTF-8"?>
-	<!-- Multiple -->
-	<!-- Comments -->
-	<svg></svg>`)))
-	assert.True(t, IsSVGImageFile([]byte(`<?xml version="1.0" encoding="UTF-8"?>
-	<!-- Multline
-	Comment -->
-	<svg></svg>`)))
-	assert.True(t, IsSVGImageFile([]byte(`<?xml version="1.0" encoding="UTF-8"?>
-	<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN" "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">
-	<!-- Multline
-	Comment -->
-	<svg></svg>`)))
-	assert.False(t, IsSVGImageFile([]byte{}))
-	assert.False(t, IsSVGImageFile([]byte("svg")))
-	assert.False(t, IsSVGImageFile([]byte("<svgfoo></svgfoo>")))
-	assert.False(t, IsSVGImageFile([]byte("text<svg></svg>")))
-	assert.False(t, IsSVGImageFile([]byte("<html><body><svg></svg></body></html>")))
-	assert.False(t, IsSVGImageFile([]byte(`<script>"<svg></svg>"</script>`)))
-	assert.False(t, IsSVGImageFile([]byte(`<!-- <svg></svg> inside comment -->
-	<foo></foo>`)))
-	assert.False(t, IsSVGImageFile([]byte(`<?xml version="1.0" encoding="UTF-8"?>
-	<!-- <svg></svg> inside comment -->
-	<foo></foo>`)))
-}
-
-func TestIsPDFFile(t *testing.T) {
-	pdf, _ := base64.StdEncoding.DecodeString("JVBERi0xLjYKJcOkw7zDtsOfCjIgMCBvYmoKPDwvTGVuZ3RoIDMgMCBSL0ZpbHRlci9GbGF0ZURlY29kZT4+CnN0cmVhbQp4nF3NPwsCMQwF8D2f4s2CNYk1baF0EHRwOwg4iJt/NsFb/PpevUE4Mjwe")
-	assert.True(t, IsPDFFile(pdf))
-	assert.False(t, IsPDFFile([]byte("plain text")))
-}
-
-func TestIsVideoFile(t *testing.T) {
-	mp4, _ := base64.StdEncoding.DecodeString("AAAAGGZ0eXBtcDQyAAAAAGlzb21tcDQyAAEI721vb3YAAABsbXZoZAAAAADaBlwX2gZcFwAAA+gA")
-	assert.True(t, IsVideoFile(mp4))
-	assert.False(t, IsVideoFile([]byte("plain text")))
-}
-
-func TestIsAudioFile(t *testing.T) {
-	mp3, _ := base64.StdEncoding.DecodeString("SUQzBAAAAAABAFRYWFgAAAASAAADbWFqb3JfYnJhbmQAbXA0MgBUWFhYAAAAEQAAA21pbm9yX3Zl")
-	assert.True(t, IsAudioFile(mp3))
-	assert.False(t, IsAudioFile([]byte("plain text")))
-}
-
 // TODO: Test EntryIcon
 
 func TestSetupGiteaRoot(t *testing.T) {

modules/git/blob.go

@@ -10,6 +10,8 @@ import (
 	"encoding/base64"
 	"io"
 	"io/ioutil"
+
+	"code.gitea.io/gitea/modules/typesniffer"
 )
 
 // This file contains common functions between the gogit and !gogit variants for git Blobs
@@ -82,3 +84,14 @@ func (b *Blob) GetBlobContentBase64() (string, error) {
 	}
 	return string(out), nil
 }
+
+// GuessContentType guesses the content type of the blob.
+func (b *Blob) GuessContentType() (typesniffer.SniffedType, error) {
+	r, err := b.DataAsync()
+	if err != nil {
+		return typesniffer.SniffedType{}, err
+	}
+	defer r.Close()
+
+	return typesniffer.DetectContentTypeFromReader(r)
+}

modules/git/commit.go

@@ -11,13 +11,7 @@ import (
 	"container/list"
 	"errors"
 	"fmt"
-	"image"
-	"image/color"
-	_ "image/gif"  // for processing gif images
-	_ "image/jpeg" // for processing jpeg images
-	_ "image/png"  // for processing png images
 	"io"
-	"net/http"
 	"os/exec"
 	"strconv"
 	"strings"
@@ -81,70 +75,6 @@ func (c *Commit) ParentCount() int {
 	return len(c.Parents)
 }
 
-func isImageFile(data []byte) (string, bool) {
-	contentType := http.DetectContentType(data)
-	if strings.Contains(contentType, "image/") {
-		return contentType, true
-	}
-	return contentType, false
-}
-
-// IsImageFile is a file image type
-func (c *Commit) IsImageFile(name string) bool {
-	blob, err := c.GetBlobByPath(name)
-	if err != nil {
-		return false
-	}
-
-	dataRc, err := blob.DataAsync()
-	if err != nil {
-		return false
-	}
-	defer dataRc.Close()
-	buf := make([]byte, 1024)
-	n, _ := dataRc.Read(buf)
-	buf = buf[:n]
-	_, isImage := isImageFile(buf)
-	return isImage
-}
-
-// ImageMetaData represents metadata of an image file
-type ImageMetaData struct {
-	ColorModel color.Model
-	Width      int
-	Height     int
-	ByteSize   int64
-}
-
-// ImageInfo returns information about the dimensions of an image
-func (c *Commit) ImageInfo(name string) (*ImageMetaData, error) {
-	if !c.IsImageFile(name) {
-		return nil, nil
-	}
-
-	blob, err := c.GetBlobByPath(name)
-	if err != nil {
-		return nil, err
-	}
-	reader, err := blob.DataAsync()
-	if err != nil {
-		return nil, err
-	}
-	defer reader.Close()
-	config, _, err := image.DecodeConfig(reader)
-	if err != nil {
-		return nil, err
-	}
-
-	metadata := ImageMetaData{
-		ColorModel: config.ColorModel,
-		Width:      config.Width,
-		Height:     config.Height,
-		ByteSize:   blob.Size(),
-	}
-	return &metadata, nil
-}
-
 // GetCommitByPath return the commit of relative path object.
 func (c *Commit) GetCommitByPath(relpath string) (*Commit, error) {
 	return c.repo.getCommitByPathWithID(c.ID, relpath)

modules/indexer/code/bleve.go

@@ -16,12 +16,12 @@ import (
 
 	"code.gitea.io/gitea/models"
 	"code.gitea.io/gitea/modules/analyze"
-	"code.gitea.io/gitea/modules/base"
 	"code.gitea.io/gitea/modules/charset"
 	"code.gitea.io/gitea/modules/git"
 	"code.gitea.io/gitea/modules/log"
 	"code.gitea.io/gitea/modules/setting"
 	"code.gitea.io/gitea/modules/timeutil"
+	"code.gitea.io/gitea/modules/typesniffer"
 	"code.gitea.io/gitea/modules/util"
 
 	"github.com/blevesearch/bleve/v2"
@@ -211,7 +211,7 @@ func (b *BleveIndexer) addUpdate(batchWriter git.WriteCloserError, batchReader *
 	fileContents, err := ioutil.ReadAll(io.LimitReader(batchReader, size))
 	if err != nil {
 		return err
-	} else if !base.IsTextFile(fileContents) {
+	} else if !typesniffer.DetectContentType(fileContents).IsText() {
 		// FIXME: UTF-16 files will probably fail here
 		return nil
 	}

modules/indexer/code/elastic_search.go

@@ -16,12 +16,12 @@ import (
 
 	"code.gitea.io/gitea/models"
 	"code.gitea.io/gitea/modules/analyze"
-	"code.gitea.io/gitea/modules/base"
 	"code.gitea.io/gitea/modules/charset"
 	"code.gitea.io/gitea/modules/git"
 	"code.gitea.io/gitea/modules/log"
 	"code.gitea.io/gitea/modules/setting"
 	"code.gitea.io/gitea/modules/timeutil"
+	"code.gitea.io/gitea/modules/typesniffer"
 
 	"github.com/go-enry/go-enry/v2"
 	jsoniter "github.com/json-iterator/go"
@@ -210,7 +210,7 @@ func (b *ElasticSearchIndexer) addUpdate(batchWriter git.WriteCloserError, batch
 	fileContents, err := ioutil.ReadAll(io.LimitReader(batchReader, size))
 	if err != nil {
 		return nil, err
-	} else if !base.IsTextFile(fileContents) {
+	} else if !typesniffer.DetectContentType(fileContents).IsText() {
 		// FIXME: UTF-16 files will probably fail here
 		return nil, nil
 	}