From 19c9367aed504b1728ac6d9e41ee171996e117b3 Mon Sep 17 00:00:00 2001
From: oiweiwei <oiweiweiei@gmail.com>
Date: Wed, 29 Jan 2025 16:03:04 +0100
Subject: [PATCH 1/2] encoding/json: improve decoder alloc count
---
src/encoding/json/scanner.go | 17 ++++++++++++-----
src/encoding/json/stream.go | 6 +++---
2 files changed, 15 insertions(+), 8 deletions(-)
diff --git a/src/encoding/json/scanner.go b/src/encoding/json/scanner.go
index da6ea2ac8f843f..d8799cb2ebe974 100644
--- a/src/encoding/json/scanner.go
+++ b/src/encoding/json/scanner.go
@@ -45,11 +45,18 @@ func checkValid(data []byte, scan *scanner) error {
// A SyntaxError is a description of a JSON syntax error.
// [Unmarshal] will return a SyntaxError if the JSON can't be parsed.
type SyntaxError struct {
- msg string // description of error
- Offset int64 // error occurred after reading Offset bytes
+ msg string // description of error
+ Offset int64 // error occurred after reading Offset bytes
+ invalidCharContext string // invalid character error context
+ invalidChar byte // the invalid character
}
-func (e *SyntaxError) Error() string { return e.msg }
+func (e *SyntaxError) Error() string {
+ if e.invalidCharContext != "" {
+ return "invalid character " + quoteChar(e.invalidChar) + " " + e.invalidCharContext
+ }
+ return e.msg
+}
// A scanner is a JSON scanning state machine.
// Callers call scan.reset and then pass bytes in one at a time
@@ -168,7 +175,7 @@ func (s *scanner) eof() int {
return scanEnd
}
if s.err == nil {
- s.err = &SyntaxError{"unexpected end of JSON input", s.bytes}
+ s.err = &SyntaxError{"unexpected end of JSON input", s.bytes, "", 0}
}
return scanError
}
@@ -590,7 +597,7 @@ func stateError(s *scanner, c byte) int {
// error records an error and switches to the error state.
func (s *scanner) error(c byte, context string) int {
s.step = stateError
- s.err = &SyntaxError{"invalid character " + quoteChar(c) + " " + context, s.bytes}
+ s.err = &SyntaxError{invalidCharContext: context, invalidChar: c, Offset: s.bytes}
return scanError
}
diff --git a/src/encoding/json/stream.go b/src/encoding/json/stream.go
index e2d9470bcc7fca..896b824d086807 100644
--- a/src/encoding/json/stream.go
+++ b/src/encoding/json/stream.go
@@ -312,7 +312,7 @@ func (dec *Decoder) tokenPrepareForDecode() error {
return err
}
if c != ',' {
- return &SyntaxError{"expected comma after array element", dec.InputOffset()}
+ return &SyntaxError{"expected comma after array element", dec.InputOffset(), "", 0}
}
dec.scanp++
dec.tokenState = tokenArrayValue
@@ -322,7 +322,7 @@ func (dec *Decoder) tokenPrepareForDecode() error {
return err
}
if c != ':' {
- return &SyntaxError{"expected colon after object key", dec.InputOffset()}
+ return &SyntaxError{"expected colon after object key", dec.InputOffset(), "", 0}
}
dec.scanp++
dec.tokenState = tokenObjectValue
@@ -475,7 +475,7 @@ func (dec *Decoder) tokenError(c byte) (Token, error) {
case tokenObjectComma:
context = " after object key:value pair"
}
- return nil, &SyntaxError{"invalid character " + quoteChar(c) + context, dec.InputOffset()}
+ return nil, &SyntaxError{invalidChar: c, invalidCharContext: context, Offset: dec.InputOffset()}
}
// More reports whether there is another element in the
From b0c5899d8fbeb968807fd3375151e48f30885330 Mon Sep 17 00:00:00 2001
From: oiweiwei <oiweiweiei@gmail.com>
Date: Thu, 30 Jan 2025 22:34:15 +0100
Subject: [PATCH 2/2] fix: ut
---
src/encoding/json/decode_test.go | 41 ++++++++++++++++---------------
src/encoding/json/scanner_test.go | 4 +--
src/encoding/json/stream.go | 12 ++++-----
src/encoding/json/stream_test.go | 8 +++---
4 files changed, 33 insertions(+), 32 deletions(-)
diff --git a/src/encoding/json/decode_test.go b/src/encoding/json/decode_test.go
index 8aad11b8bfbce2..c5b73802c2f4be 100644
--- a/src/encoding/json/decode_test.go
+++ b/src/encoding/json/decode_test.go
@@ -467,21 +467,21 @@ var unmarshalTests = []struct {
{CaseName: Name(""), in: `{"alphabet": "xyz"}`, ptr: new(U), err: fmt.Errorf("json: unknown field \"alphabet\""), disallowUnknownFields: true},
// syntax errors
- {CaseName: Name(""), in: `{"X": "foo", "Y"}`, err: &SyntaxError{"invalid character '}' after object key", 17}},
- {CaseName: Name(""), in: `[1, 2, 3+]`, err: &SyntaxError{"invalid character '+' after array element", 9}},
- {CaseName: Name(""), in: `{"X":12x}`, err: &SyntaxError{"invalid character 'x' after object key:value pair", 8}, useNumber: true},
+ {CaseName: Name(""), in: `{"X": "foo", "Y"}`, err: &SyntaxError{invalidChar: '}', invalidCharContext: "after object key", Offset: 17}},
+ {CaseName: Name(""), in: `[1, 2, 3+]`, err: &SyntaxError{invalidChar: '+', invalidCharContext: "after array element", Offset: 9}},
+ {CaseName: Name(""), in: `{"X":12x}`, err: &SyntaxError{invalidChar: 'x', invalidCharContext: "after object key:value pair", Offset: 8}, useNumber: true},
{CaseName: Name(""), in: `[2, 3`, err: &SyntaxError{msg: "unexpected end of JSON input", Offset: 5}},
- {CaseName: Name(""), in: `{"F3": -}`, ptr: new(V), err: &SyntaxError{msg: "invalid character '}' in numeric literal", Offset: 9}},
+ {CaseName: Name(""), in: `{"F3": -}`, ptr: new(V), err: &SyntaxError{invalidChar: '}', invalidCharContext: "in numeric literal", Offset: 9}},
// raw value errors
- {CaseName: Name(""), in: "\x01 42", err: &SyntaxError{"invalid character '\\x01' looking for beginning of value", 1}},
- {CaseName: Name(""), in: " 42 \x01", err: &SyntaxError{"invalid character '\\x01' after top-level value", 5}},
- {CaseName: Name(""), in: "\x01 true", err: &SyntaxError{"invalid character '\\x01' looking for beginning of value", 1}},
- {CaseName: Name(""), in: " false \x01", err: &SyntaxError{"invalid character '\\x01' after top-level value", 8}},
- {CaseName: Name(""), in: "\x01 1.2", err: &SyntaxError{"invalid character '\\x01' looking for beginning of value", 1}},
- {CaseName: Name(""), in: " 3.4 \x01", err: &SyntaxError{"invalid character '\\x01' after top-level value", 6}},
- {CaseName: Name(""), in: "\x01 \"string\"", err: &SyntaxError{"invalid character '\\x01' looking for beginning of value", 1}},
- {CaseName: Name(""), in: " \"string\" \x01", err: &SyntaxError{"invalid character '\\x01' after top-level value", 11}},
+ {CaseName: Name(""), in: "\x01 42", err: &SyntaxError{invalidChar: '\x01', invalidCharContext: "looking for beginning of value", Offset: 1}},
+ {CaseName: Name(""), in: " 42 \x01", err: &SyntaxError{invalidChar: '\x01', invalidCharContext: "after top-level value", Offset: 5}},
+ {CaseName: Name(""), in: "\x01 true", err: &SyntaxError{invalidChar: '\x01', invalidCharContext: "looking for beginning of value", Offset: 1}},
+ {CaseName: Name(""), in: " false \x01", err: &SyntaxError{invalidChar: '\x01', invalidCharContext: "after top-level value", Offset: 8}},
+ {CaseName: Name(""), in: "\x01 1.2", err: &SyntaxError{invalidChar: '\x01', invalidCharContext: "looking for beginning of value", Offset: 1}},
+ {CaseName: Name(""), in: " 3.4 \x01", err: &SyntaxError{invalidChar: '\x01', invalidCharContext: "after top-level value", Offset: 6}},
+ {CaseName: Name(""), in: "\x01 \"string\"", err: &SyntaxError{invalidChar: '\x01', invalidCharContext: "looking for beginning of value", Offset: 1}},
+ {CaseName: Name(""), in: " \"string\" \x01", err: &SyntaxError{invalidChar: '\x01', invalidCharContext: "after top-level value", Offset: 11}},
// array tests
{CaseName: Name(""), in: `[1, 2, 3]`, ptr: new([3]int), out: [3]int{1, 2, 3}},
@@ -1096,8 +1096,9 @@ var unmarshalTests = []struct {
in: `invalid`,
ptr: new(Number),
err: &SyntaxError{
- msg: "invalid character 'i' looking for beginning of value",
- Offset: 1,
+ invalidChar: 'i',
+ invalidCharContext: "looking for beginning of value",
+ Offset: 1,
},
},
{
@@ -1178,7 +1179,7 @@ var unmarshalTests = []struct {
CaseName: Name(""),
in: `[1,2,true,4,5}`,
ptr: new([]int),
- err: &SyntaxError{msg: "invalid character '}' after array element", Offset: 14},
+ err: &SyntaxError{invalidChar: '}', invalidCharContext: "after array element", Offset: 14},
},
{
CaseName: Name(""),
@@ -2589,23 +2590,23 @@ func TestUnmarshalErrorAfterMultipleJSON(t *testing.T) {
}{{
CaseName: Name(""),
in: `1 false null :`,
- err: &SyntaxError{"invalid character ':' looking for beginning of value", 14},
+ err: &SyntaxError{invalidChar: ':', invalidCharContext: "looking for beginning of value", Offset: 14},
}, {
CaseName: Name(""),
in: `1 [] [,]`,
- err: &SyntaxError{"invalid character ',' looking for beginning of value", 7},
+ err: &SyntaxError{invalidChar: ',', invalidCharContext: "looking for beginning of value", Offset: 7},
}, {
CaseName: Name(""),
in: `1 [] [true:]`,
- err: &SyntaxError{"invalid character ':' after array element", 11},
+ err: &SyntaxError{invalidChar: ':', invalidCharContext: "after array element", Offset: 11},
}, {
CaseName: Name(""),
in: `1 {} {"x"=}`,
- err: &SyntaxError{"invalid character '=' after object key", 14},
+ err: &SyntaxError{invalidChar: '=', invalidCharContext: "after object key", Offset: 14},
}, {
CaseName: Name(""),
in: `falsetruenul#`,
- err: &SyntaxError{"invalid character '#' in literal null (expecting 'l')", 13},
+ err: &SyntaxError{invalidChar: '#', invalidCharContext: "in literal null (expecting 'l')", Offset: 13},
}}
for _, tt := range tests {
t.Run(tt.Name, func(t *testing.T) {
diff --git a/src/encoding/json/scanner_test.go b/src/encoding/json/scanner_test.go
index 068439dcaca858..896dbbd67d4daa 100644
--- a/src/encoding/json/scanner_test.go
+++ b/src/encoding/json/scanner_test.go
@@ -190,8 +190,8 @@ func TestIndentErrors(t *testing.T) {
in string
err error
}{
- {Name(""), `{"X": "foo", "Y"}`, &SyntaxError{"invalid character '}' after object key", 17}},
- {Name(""), `{"X": "foo" "Y": "bar"}`, &SyntaxError{"invalid character '\"' after object key:value pair", 13}},
+ {Name(""), `{"X": "foo", "Y"}`, &SyntaxError{invalidChar: '}', invalidCharContext: "after object key", Offset: 17}},
+ {Name(""), `{"X": "foo" "Y": "bar"}`, &SyntaxError{invalidChar: '"', invalidCharContext: "after object key:value pair", Offset: 13}},
}
for _, tt := range tests {
t.Run(tt.Name, func(t *testing.T) {
diff --git a/src/encoding/json/stream.go b/src/encoding/json/stream.go
index 896b824d086807..3ccb86678acaef 100644
--- a/src/encoding/json/stream.go
+++ b/src/encoding/json/stream.go
@@ -463,17 +463,17 @@ func (dec *Decoder) tokenError(c byte) (Token, error) {
var context string
switch dec.tokenState {
case tokenTopValue:
- context = " looking for beginning of value"
+ context = "looking for beginning of value"
case tokenArrayStart, tokenArrayValue, tokenObjectValue:
- context = " looking for beginning of value"
+ context = "looking for beginning of value"
case tokenArrayComma:
- context = " after array element"
+ context = "after array element"
case tokenObjectKey:
- context = " looking for beginning of object key string"
+ context = "looking for beginning of object key string"
case tokenObjectColon:
- context = " after object key"
+ context = "after object key"
case tokenObjectComma:
- context = " after object key:value pair"
+ context = "after object key:value pair"
}
return nil, &SyntaxError{invalidChar: c, invalidCharContext: context, Offset: dec.InputOffset()}
}
diff --git a/src/encoding/json/stream_test.go b/src/encoding/json/stream_test.go
index 46f9407c881c36..bb4f5a595fc63e 100644
--- a/src/encoding/json/stream_test.go
+++ b/src/encoding/json/stream_test.go
@@ -444,18 +444,18 @@ func TestDecodeInStream(t *testing.T) {
{CaseName: Name(""), json: ` [{"a": 1} {"a": 2}] `, expTokens: []any{
Delim('['),
decodeThis{map[string]any{"a": float64(1)}},
- decodeThis{&SyntaxError{"expected comma after array element", 11}},
+ decodeThis{&SyntaxError{"expected comma after array element", 11, "", 0}},
}},
{CaseName: Name(""), json: `{ "` + strings.Repeat("a", 513) + `" 1 }`, expTokens: []any{
Delim('{'), strings.Repeat("a", 513),
- decodeThis{&SyntaxError{"expected colon after object key", 518}},
+ decodeThis{&SyntaxError{"expected colon after object key", 518, "", 0}},
}},
{CaseName: Name(""), json: `{ "\a" }`, expTokens: []any{
Delim('{'),
- &SyntaxError{"invalid character 'a' in string escape code", 3},
+ &SyntaxError{invalidChar: 'a', invalidCharContext: "in string escape code", Offset: 3},
}},
{CaseName: Name(""), json: ` \a`, expTokens: []any{
- &SyntaxError{"invalid character '\\\\' looking for beginning of value", 1},
+ &SyntaxError{invalidChar: '\\', invalidCharContext: "looking for beginning of value", Offset: 1},
}},
}
for _, tt := range tests {