A lightweight test helper package
Fix a different utf-8 bug!
+24
+8
internal/diff/chardiff_test.go
+8
internal/diff/chardiff_test.go
···
90
90
added: []byte("0"),
91
91
wantHasChanged: true,
92
92
},
93
+
{
94
+
// Regression: identical invalid UTF-8 bytes must produce all-unchanged segments.
95
+
// \x80 is a bare continuation byte — invalid UTF-8 on its own.
96
+
name: "identical invalid UTF-8",
97
+
removed: []byte("\x80"),
98
+
added: []byte("\x80"),
99
+
wantAllUnchanged: true,
100
+
},
93
101
}
94
102
95
103
for _, tt := range tests {
+13
internal/diff/chars.go
+13
internal/diff/chars.go
···
1
1
package diff
2
2
3
3
import (
4
+
"bytes"
4
5
"unicode/utf8"
5
6
)
6
7
···
21
22
// Strips and restores trailing \n before diffing.
22
23
// If either side exceeds 500 runes, returns a single-segment fallback (whole-line).
23
24
func CharDiff(removed, added []byte) InlineChange {
25
+
// Identical inputs: short-circuit before any UTF-8 handling. This also
26
+
// covers identical invalid UTF-8, where fallback would wrongly return
27
+
// Changed:true segments.
28
+
if bytes.Equal(removed, added) {
29
+
seg := Segment{Text: append([]byte(nil), removed...), Changed: false}
30
+
31
+
return InlineChange{
32
+
Removed: []Segment{seg},
33
+
Added: []Segment{{Text: append([]byte(nil), added...), Changed: false}},
34
+
}
35
+
}
36
+
24
37
// Strip trailing newline, remember whether each side had one.
25
38
removedNL := len(removed) > 0 && removed[len(removed)-1] == '\n'
26
39
addedNL := len(added) > 0 && added[len(added)-1] == '\n'