Switch Unicode Escaping to a VSCode-like system (#19990)

This PR rewrites the invisible unicode detection algorithm to more
closely match that of the Monaco editor on the system. It provides a
technique for detecting ambiguous characters and relaxes the detection
of combining marks.

Control characters are in addition detected as invisible in this
implementation whereas they are not on monaco but this is related to
font issues.

Close #19913

Signed-off-by: Andrew Thornton <art27@cantab.net>

1 files changed, 17 insertions, 16 deletions

diff --git a/services/gitdiff/gitdiff.go b/services/gitdiff/gitdiff.go
index 6dd237bbc8..9844992f5b 100644
--- a/services/gitdiff/gitdiff.go
+++ b/services/gitdiff/gitdiff.go
@@ -32,6 +32,7 @@ import (
 	"code.gitea.io/gitea/modules/lfs"
 	"code.gitea.io/gitea/modules/log"
 	"code.gitea.io/gitea/modules/setting"
+	"code.gitea.io/gitea/modules/translation"
 
 	"github.com/sergi/go-diff/diffmatchpatch"
 	stdcharset "golang.org/x/net/html/charset"
@@ -169,11 +170,11 @@ func getDiffLineSectionInfo(treePath, line string, lastLeftIdx, lastRightIdx int
 }
 
 // escape a line's content or return <br> needed for copy/paste purposes
-func getLineContent(content string) DiffInline {
+func getLineContent(content string, locale translation.Locale) DiffInline {
 	if len(content) > 0 {
-		return DiffInlineWithUnicodeEscape(template.HTML(html.EscapeString(content)))
+		return DiffInlineWithUnicodeEscape(template.HTML(html.EscapeString(content)), locale)
 	}
-	return DiffInline{Content: "<br>"}
+	return DiffInline{EscapeStatus: &charset.EscapeStatus{}, Content: "<br>"}
 }
 
 // DiffSection represents a section of a DiffFile.
@@ -267,26 +268,26 @@ func init() {
 
 // DiffInline is a struct that has a content and escape status
 type DiffInline struct {
-	EscapeStatus charset.EscapeStatus
+	EscapeStatus *charset.EscapeStatus
 	Content      template.HTML
 }
 
 // DiffInlineWithUnicodeEscape makes a DiffInline with hidden unicode characters escaped
-func DiffInlineWithUnicodeEscape(s template.HTML) DiffInline {
-	status, content := charset.EscapeControlString(string(s))
+func DiffInlineWithUnicodeEscape(s template.HTML, locale translation.Locale) DiffInline {
+	status, content := charset.EscapeControlHTML(string(s), locale)
 	return DiffInline{EscapeStatus: status, Content: template.HTML(content)}
 }
 
 // DiffInlineWithHighlightCode makes a DiffInline with code highlight and hidden unicode characters escaped
-func DiffInlineWithHighlightCode(fileName, language, code string) DiffInline {
-	status, content := charset.EscapeControlString(highlight.Code(fileName, language, code))
+func DiffInlineWithHighlightCode(fileName, language, code string, locale translation.Locale) DiffInline {
+	status, content := charset.EscapeControlHTML(highlight.Code(fileName, language, code), locale)
 	return DiffInline{EscapeStatus: status, Content: template.HTML(content)}
 }
 
 // GetComputedInlineDiffFor computes inline diff for the given line.
-func (diffSection *DiffSection) GetComputedInlineDiffFor(diffLine *DiffLine) DiffInline {
+func (diffSection *DiffSection) GetComputedInlineDiffFor(diffLine *DiffLine, locale translation.Locale) DiffInline {
 	if setting.Git.DisableDiffHighlight {
-		return getLineContent(diffLine.Content[1:])
+		return getLineContent(diffLine.Content[1:], locale)
 	}
 
 	var (
@@ -303,26 +304,26 @@ func (diffSection *DiffSection) GetComputedInlineDiffFor(diffLine *DiffLine) Dif
 	// try to find equivalent diff line. ignore, otherwise
 	switch diffLine.Type {
 	case DiffLineSection:
-		return getLineContent(diffLine.Content[1:])
+		return getLineContent(diffLine.Content[1:], locale)
 	case DiffLineAdd:
 		compareDiffLine = diffSection.GetLine(DiffLineDel, diffLine.RightIdx)
 		if compareDiffLine == nil {
-			return DiffInlineWithHighlightCode(diffSection.FileName, language, diffLine.Content[1:])
+			return DiffInlineWithHighlightCode(diffSection.FileName, language, diffLine.Content[1:], locale)
 		}
 		diff1 = compareDiffLine.Content
 		diff2 = diffLine.Content
 	case DiffLineDel:
 		compareDiffLine = diffSection.GetLine(DiffLineAdd, diffLine.LeftIdx)
 		if compareDiffLine == nil {
-			return DiffInlineWithHighlightCode(diffSection.FileName, language, diffLine.Content[1:])
+			return DiffInlineWithHighlightCode(diffSection.FileName, language, diffLine.Content[1:], locale)
 		}
 		diff1 = diffLine.Content
 		diff2 = compareDiffLine.Content
 	default:
 		if strings.IndexByte(" +-", diffLine.Content[0]) > -1 {
-			return DiffInlineWithHighlightCode(diffSection.FileName, language, diffLine.Content[1:])
+			return DiffInlineWithHighlightCode(diffSection.FileName, language, diffLine.Content[1:], locale)
 		}
-		return DiffInlineWithHighlightCode(diffSection.FileName, language, diffLine.Content)
+		return DiffInlineWithHighlightCode(diffSection.FileName, language, diffLine.Content, locale)
 	}
 
 	hcd := newHighlightCodeDiff()
@@ -330,7 +331,7 @@ func (diffSection *DiffSection) GetComputedInlineDiffFor(diffLine *DiffLine) Dif
 	// it seems that Gitea doesn't need the line wrapper of Chroma, so do not add them back
 	// if the line wrappers are still needed in the future, it can be added back by "diffToHTML(hcd.lineWrapperTags. ...)"
 	diffHTML := diffToHTML(nil, diffRecord, diffLine.Type)
-	return DiffInlineWithUnicodeEscape(template.HTML(diffHTML))
+	return DiffInlineWithUnicodeEscape(template.HTML(diffHTML), locale)
 }
 
 // DiffFile represents a file diff.