Display SVG files as images instead of text (#14101)

* Change to display SVG files as images

* Remove unsafe styles from SVG CSP

* Add integration test to test SVG headers

* Add config setting to disable SVG rendering

* Add test for img tag when loading SVG image

* Remove the Raw view button for svg files since we don't fully support this

* Fix copyright year

* Rename and move config setting

* Add setting to cheat sheet in docs

* Fix so that comment matches cheat sheet

* Add allowing styles in CSP based on pull request feedback

* Re-enable raw button since we show SVG styles now

* Change so that SVG files are editable

* Add UI to toggle between source and rendered image for SVGs

* Change to show blame button for SVG images

* Fix to update ctx data

* Add test for DetectContentType when file is longer than sniffLen

Co-authored-by: Jonathan Tran <jon@allspice.io>
Co-authored-by: Kyle D <kdumontnu@gmail.com>

2 files changed, 96 insertions, 5 deletions

diff --git a/modules/base/tool.go b/modules/base/tool.go
index 7ac572b85b..c497bee44a 100644
--- a/modules/base/tool.go
+++ b/modules/base/tool.go
@@ -15,6 +15,7 @@ import (
 	"net/http"
 	"os"
 	"path/filepath"
+	"regexp"
 	"runtime"
 	"strconv"
 	"strings"
@@ -28,6 +29,15 @@ 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(`(?s)\A\s*(?:<!--.*?-->\s*)*<svg\b`)
+var svgTagInXMLRegex = regexp.MustCompile(`(?s)\A<\?xml\b.*?\?>\s*(?:<!--.*?-->\s*)*<svg\b`)
+
 // EncodeMD5 encodes string to md5 hex value.
 func EncodeMD5(str string) string {
 	m := md5.New()
@@ -265,32 +275,61 @@ 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(http.DetectContentType(data), "text/")
+	return strings.Contains(DetectContentType(data), "text/")
 }
 
 // IsImageFile detects if data is an image format
 func IsImageFile(data []byte) bool {
-	return strings.Contains(http.DetectContentType(data), "image/")
+	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(http.DetectContentType(data), "application/pdf")
+	return strings.Contains(DetectContentType(data), "application/pdf")
 }
 
 // IsVideoFile detects if data is an video format
 func IsVideoFile(data []byte) bool {
-	return strings.Contains(http.DetectContentType(data), "video/")
+	return strings.Contains(DetectContentType(data), "video/")
 }
 
 // IsAudioFile detects if data is an video format
 func IsAudioFile(data []byte) bool {
-	return strings.Contains(http.DetectContentType(data), "audio/")
+	return strings.Contains(DetectContentType(data), "audio/")
 }
 
 // EntryIcon returns the octicon class for displaying files/directories
diff --git a/modules/base/tool_test.go b/modules/base/tool_test.go
index 0b708dafdb..cda1685da7 100644
--- a/modules/base/tool_test.go
+++ b/modules/base/tool_test.go
@@ -183,11 +183,63 @@ func TestIsLetter(t *testing.T) {
 	assert.False(t, IsLetter('$'))
 }
 
+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>`)))
+}
+
 func TestIsTextFile(t *testing.T) {
 	assert.True(t, IsTextFile([]byte{}))
 	assert.True(t, IsTextFile([]byte("lorem ipsum")))
 }
 
+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(`<?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.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 TestFormatNumberSI(t *testing.T) {
 	assert.Equal(t, "125", FormatNumberSI(int(125)))
 	assert.Equal(t, "1.3k", FormatNumberSI(int64(1317)))