fix(sub): keep every Clash scalar on its own line

The YAML encoder writes multi-line strings as block scalars and gets
CR and the Unicode line breaks (NEL, LS, PS) wrong inside them, so a
remark, path or header name with such characters could continue as YAML
outside its scalar. Strings with a non-printable rune or invalid UTF-8
are now written double-quoted with escapes, map keys included.
This commit is contained in:
MHSanaei
2026-10-31 08:56:24 +00:00
parent 534afe13ab
commit 8359a0f72e
2 changed files with 73 additions and 5 deletions
+44 -5
View File
@@ -3,7 +3,10 @@ package sub
import (
"reflect"
"regexp"
"strconv"
"strings"
"unicode"
"unicode/utf8"
"github.com/goccy/go-yaml"
)
@@ -41,30 +44,66 @@ func yamlScalarIsAmbiguous(s string) bool {
return yamlNonStringWords[s] || yamlNonStringNumber.MatchString(s)
}
// yamlEscapedString is written double-quoted on one line: the encoder's block scalars
// mishandle CR and the Unicode line breaks, so the text could continue as YAML.
type yamlEscapedString string
func (s yamlEscapedString) MarshalYAML() ([]byte, error) {
return []byte(strconv.Quote(string(s))), nil
}
func yamlNeedsEscaping(s string) bool {
if !utf8.ValidString(s) {
return true
}
return strings.ContainsFunc(s, func(r rune) bool { return !unicode.IsPrint(r) })
}
func quoteYAMLString(s string) any {
switch {
case yamlNeedsEscaping(s):
return yamlEscapedString(s)
case yamlScalarIsAmbiguous(s):
return yamlQuotedString(s)
}
return s
}
func quoteAmbiguousYAMLScalars(v any) any {
if v == nil {
return nil
}
if s, ok := v.(string); ok {
if yamlScalarIsAmbiguous(s) {
return yamlQuotedString(s)
}
return s
return quoteYAMLString(s)
}
rv := reflect.ValueOf(v)
switch rv.Kind() {
case reflect.Map:
out := make(map[string]any, rv.Len())
escapedKeys := false
iter := rv.MapRange()
for iter.Next() {
key, ok := iter.Key().Interface().(string)
if !ok {
return v
}
escapedKeys = escapedKeys || yamlNeedsEscaping(key)
out[key] = quoteAmbiguousYAMLScalars(iter.Value().Interface())
}
return out
if !escapedKeys {
return out
}
// Only marshaler-typed keys escape, and only a map[any]any key keeps its type.
keyed := make(map[any]any, len(out))
for key, value := range out {
if yamlNeedsEscaping(key) {
keyed[yamlEscapedString(key)] = value
} else {
keyed[key] = value
}
}
return keyed
case reflect.Slice, reflect.Array:
if rv.Kind() == reflect.Slice && rv.Type().Elem().Kind() == reflect.Uint8 {
return v
+29
View File
@@ -127,3 +127,32 @@ func TestQuotedScalarEscapesQuotes(t *testing.T) {
t.Errorf("password round-tripped to %#v\n%s", got, out)
}
}
// Node-reported paths, remarks and header names reach the profile; a line break the
// encoder writes bare would start a top-level key of the subscriber's Clash config.
func TestLineBreaksStayInsideTheirScalar(t *testing.T) {
for _, value := range []string{
"/ws\r\n\r\nmode:\r direct", "/ws\n\n\n\n\n\n\rmode:\r direct", "a\nb\nc\u0085mode: direct",
"a\nb\nc
mode: direct", "a\nb\nc
mode: direct", "2026-01-01\n\n\rmode: direct",
} {
profile := map[string]any{"proxies": []any{map[string]any{
"name": "n",
"ws-opts": map[string]any{"path": value, "headers": map[string]any{value: "v"}},
}}}
out, err := marshalClashYAML(profile)
if err != nil {
t.Fatalf("marshal %q: %v", value, err)
}
if strings.ContainsAny(string(out), "\r\u0085

") || strings.Contains(string(out), "\nmode:") {
t.Fatalf("value %q escaped its scalar:\n%q", value, out)
}
var back map[string]any
if err := yaml.Unmarshal(out, &back); err != nil {
t.Fatalf("read back %q: %v\n%q", value, err, out)
}
opts := back["proxies"].([]any)[0].(map[string]any)["ws-opts"].(map[string]any)
if opts["path"] != value || opts["headers"].(map[string]any)[value] != "v" {
t.Fatalf("value %q did not round-trip:\n%q", value, out)
}
}
}