to 3.0: enable escape and escape the JSON control characters. (#22607)

Enable escape character and escape the JSON control characters.

```
create table t1 (a int primary key, b json);

insert into t1 values(1, '{"key1":"你好\\t不\\r好\\f呀\\n\\\\"}');
insert into t1 values(2, '{"key2":"谢谢\\t你，\\r我非常\\f好\\n\\\\"}');

1. select * from t1 into outfile "/tmp/gg/t1.csv" fields terminated by "," enclosed by '\'';
2. load data infile "/tmp/gg/t1.csv" into table t1 fields terminated by "," enclosed by '\'' ignore 1 lines;
```

All JSON control characters in the JSON column of t1 should be escaped when we select data out to the outfile.

Approved by: @XuPeng-SH, @heni02, @daviszhen

Author: gouhongshen

pkg/frontend/export.go

@@ -311,13 +311,53 @@ func formatJsonString(str string, flag bool, terminatedBy string) string {
 	return tmp
 }
 
+func escapeJSONControlChars(s string) string {
+	var builder strings.Builder
+	builder.Grow(len(s))
+
+	// \\t
+	// \\\\t
+	// abc\t
+	// {"a": "abc", "b": "abc\t"}
+	for i := 0; i < len(s); {
+		if i+2 > len(s) {
+			builder.WriteString(s[i:])
+			break
+		}
+		switch s[i : i+2] {
+		case `\t`:
+			i += 2
+			builder.WriteString(`\\t`)
+		case `\n`:
+			i += 2
+			builder.WriteString(`\\n`)
+		case `\r`:
+			i += 2
+			builder.WriteString(`\\r`)
+		case `\f`:
+			i += 2
+			builder.WriteString(`\\f`)
+		case `\\`:
+			i += 2
+			builder.WriteString(`\\\\`)
+		default:
+			builder.WriteString(s[i : i+1])
+			i++
+		}
+	}
+
+	return builder.String()
+}
+
 func constructByte(ctx context.Context, obj FeSession, bat *batch.Batch, index int32, ByteChan chan *BatchByte, ep *ExportConfig) {
 	ses := obj.(*Session)
 	symbol := ep.Symbol
 	closeby := ep.userConfig.Fields.EnclosedBy.Value
 	terminated := ep.userConfig.Fields.Terminated.Value
 	flag := ep.ColumnFlag
 
+	escape = closeby
+
 	buffer := &bytes.Buffer{}
 
 	for i := 0; i < bat.RowCount(); i++ {
@@ -329,7 +369,12 @@ func constructByte(ctx context.Context, obj FeSession, bat *batch.Batch, index i
 			switch vec.GetType().Oid { //get col
 			case types.T_json:
 				val := types.DecodeJson(vec.GetBytesAt(i))
-				formatOutputString(ep, []byte(formatJsonString(val.String(), flag[j], terminated)), symbol[j], closeby, flag[j], buffer)
+				formatOutputString(ep, []byte(formatJsonString(
+					escapeJSONControlChars(
+						val.String(),
+					),
+					flag[j], terminated),
+				), symbol[j], closeby, flag[j], buffer)
 			case types.T_bool:
 				val := vector.GetFixedAtNoTypeCheck[bool](vec, i)
 				if val {

pkg/frontend/export_test.go

@@ -17,6 +17,7 @@ package frontend
 import (
 	"bytes"
 	"context"
+	"strings"
 	"testing"
 
 	"github.com/prashantv/gostub"
@@ -301,3 +302,91 @@ func Test_exportDataToCSVFile(t *testing.T) {
 		convey.So(exportDataFromResultSetToCSVFile(ep), convey.ShouldBeNil)
 	})
 }
+
+func TestEscapeJSONControl(t *testing.T) {
+	tests := []struct {
+		name     string
+		input    string
+		expected string
+	}{
+		{
+			name:     "empty string",
+			input:    "",
+			expected: "",
+		},
+		{
+			name:     "normal characters",
+			input:    `{"a":"hello world 123"}`,
+			expected: `{"a":"hello world 123"}`,
+		},
+		{
+			name:     "UTF-8 characters",
+			input:    `{"a":"hello 世界 😊"}`,
+			expected: `{"a":"hello 世界 😊"}`,
+		},
+		{
+			name:     "single newline",
+			input:    `{"a":"abc` + `\n` + `def"}`,
+			expected: `{"a":"abc\\ndef"}`,
+		},
+		{
+			name:     "single carriage return",
+			input:    `{"a":"abc` + `\r` + `def"}`,
+			expected: `{"a":"abc\\rdef"}`,
+		},
+		{
+			name:     "single tab",
+			input:    `{"a":"abc` + `\t` + `def"}`,
+			expected: `{"a":"abc\\tdef"}`,
+		},
+		{
+			name:     "single form feed",
+			input:    `{"a":"abc` + `\f` + `def"}`,
+			expected: `{"a":"abc\\fdef"}`,
+		},
+		{
+			name:     "multiple newlines",
+			input:    `{"a":"abc` + `\n\n` + `def"}`,
+			expected: `{"a":"abc\\n\\ndef"}`,
+		},
+		{
+			name:     "escaped sequences unchanged",
+			input:    `{"a":"abc\\n\\r\\t\\b\\fdef"}`,
+			expected: `{"a":"abc\\\\n\\\\r\\\\t\\\\b\\\\fdef"}`,
+		},
+		{
+			name:     "mixed escaped and actual",
+			input:    `{"a":"abc\\n` + `\t` + `def\\r"}`,
+			expected: `{"a":"abc\\\\n\\tdef\\\\r"}`,
+		},
+		{
+			name:     "control at start",
+			input:    `\n` + `{"a":"def"}`,
+			expected: `\\n{"a":"def"}`,
+		},
+		{
+			name:     "control at end",
+			input:    `{"a":"def"}` + `\n`,
+			expected: `{"a":"def"}\\n`,
+		},
+		{
+			name:     "long string with controls",
+			input:    strings.Repeat("a", 10000) + `\n` + strings.Repeat("b", 10000) + `\t` + strings.Repeat("c", 10000),
+			expected: strings.Repeat("a", 10000) + `\\n` + strings.Repeat("b", 10000) + `\\t` + strings.Repeat("c", 10000),
+		},
+		{
+			name:     "backslash literal",
+			input:    `{"a":"abc\\def"}`,
+			expected: `{"a":"abc\\\\def"}`,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			res := escapeJSONControlChars(tt.input)
+			if res != tt.expected {
+				t.Errorf("EscapeJSONControl(%v) =\n%v,\nwant\n%v", tt.input, res, tt.expected)
+			}
+		})
+	}
+}

test/distributed/cases/load_data/load_data.result

@@ -1063,4 +1063,18 @@ load data infile '$resources/into_outfile/load_data/t1.csv' into table t1 ignore
 select count(*) from t1;
 count(*)
 3
+create table t2 (a int primary key, b json);
+insert into t2 values(1, '{"key1":"你好\\t不\\r好\\f呀\\n\\\\"}');
+insert into t2 values(2, '{"key2":"谢谢\\t你，\\r我非常\\f好\\n\\\\"}');
+select * from t2 order by a asc;
+a    b
+1    {"key1": "你好\t不\r好\f呀\n\\"}
+2    {"key2": "谢谢\t你，\r我非常\f好\n\\"}
+select * from t2 into outfile '$resources/into_outfile/load_data/t2.csv';
+truncate table t2;
+load data infile '$resources/into_outfile/load_data/t2.csv' into table t2 ignore 1 lines;
+select * from t2 order by a asc;
+a    b
+1    {"key1": "你好\t不\r好\f呀\n\\"}
+2    {"key2": "谢谢\t你，\r我非常\f好\n\\"}
 drop database test;

test/distributed/cases/load_data/load_data.sql

@@ -625,4 +625,17 @@ truncate table t1;
 load data infile '$resources/into_outfile/load_data/t1.csv' into table t1 ignore 1 lines;
 select count(*) from t1;
 
+
+create table t2 (a int primary key, b json);
+
+insert into t2 values(1, '{"key1":"你好\\t不\\r好\\f呀\\n\\\\"}');
+insert into t2 values(2, '{"key2":"谢谢\\t你，\\r我非常\\f好\\n\\\\"}');
+select * from t2 order by a asc;
+
+select * from t2 into outfile '$resources/into_outfile/load_data/t2.csv';
+truncate table t2;
+
+load data infile '$resources/into_outfile/load_data/t2.csv' into table t2 ignore 1 lines;
+select * from t2 order by a asc;
+
 drop database test;