1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
|
package sanitize
import (
"bytes"
"encoding/hex"
"fmt"
"strconv"
"strings"
"time"
"unicode/utf8"
)
// Part is either a string or an int. A string is raw SQL. An int is a
// argument placeholder.
type Part any
type Query struct {
Parts []Part
}
// utf.DecodeRune returns the utf8.RuneError for errors. But that is actually rune U+FFFD -- the unicode replacement
// character. utf8.RuneError is not an error if it is also width 3.
//
// https://github.com/jackc/pgx/issues/1380
const replacementcharacterwidth = 3
func (q *Query) Sanitize(args ...any) (string, error) {
argUse := make([]bool, len(args))
buf := &bytes.Buffer{}
for _, part := range q.Parts {
var str string
switch part := part.(type) {
case string:
str = part
case int:
argIdx := part - 1
if argIdx >= len(args) {
return "", fmt.Errorf("insufficient arguments")
}
arg := args[argIdx]
switch arg := arg.(type) {
case nil:
str = "null"
case int64:
str = strconv.FormatInt(arg, 10)
case float64:
str = strconv.FormatFloat(arg, 'f', -1, 64)
case bool:
str = strconv.FormatBool(arg)
case []byte:
str = QuoteBytes(arg)
case string:
str = QuoteString(arg)
case time.Time:
str = arg.Truncate(time.Microsecond).Format("'2006-01-02 15:04:05.999999999Z07:00:00'")
default:
return "", fmt.Errorf("invalid arg type: %T", arg)
}
argUse[argIdx] = true
default:
return "", fmt.Errorf("invalid Part type: %T", part)
}
buf.WriteString(str)
}
for i, used := range argUse {
if !used {
return "", fmt.Errorf("unused argument: %d", i)
}
}
return buf.String(), nil
}
func NewQuery(sql string) (*Query, error) {
l := &sqlLexer{
src: sql,
stateFn: rawState,
}
for l.stateFn != nil {
l.stateFn = l.stateFn(l)
}
query := &Query{Parts: l.parts}
return query, nil
}
func QuoteString(str string) string {
return "'" + strings.ReplaceAll(str, "'", "''") + "'"
}
func QuoteBytes(buf []byte) string {
return `'\x` + hex.EncodeToString(buf) + "'"
}
type sqlLexer struct {
src string
start int
pos int
nested int // multiline comment nesting level.
stateFn stateFn
parts []Part
}
type stateFn func(*sqlLexer) stateFn
func rawState(l *sqlLexer) stateFn {
for {
r, width := utf8.DecodeRuneInString(l.src[l.pos:])
l.pos += width
switch r {
case 'e', 'E':
nextRune, width := utf8.DecodeRuneInString(l.src[l.pos:])
if nextRune == '\'' {
l.pos += width
return escapeStringState
}
case '\'':
return singleQuoteState
case '"':
return doubleQuoteState
case '$':
nextRune, _ := utf8.DecodeRuneInString(l.src[l.pos:])
if '0' <= nextRune && nextRune <= '9' {
if l.pos-l.start > 0 {
l.parts = append(l.parts, l.src[l.start:l.pos-width])
}
l.start = l.pos
return placeholderState
}
case '-':
nextRune, width := utf8.DecodeRuneInString(l.src[l.pos:])
if nextRune == '-' {
l.pos += width
return oneLineCommentState
}
case '/':
nextRune, width := utf8.DecodeRuneInString(l.src[l.pos:])
if nextRune == '*' {
l.pos += width
return multilineCommentState
}
case utf8.RuneError:
if width != replacementcharacterwidth {
if l.pos-l.start > 0 {
l.parts = append(l.parts, l.src[l.start:l.pos])
l.start = l.pos
}
return nil
}
}
}
}
func singleQuoteState(l *sqlLexer) stateFn {
for {
r, width := utf8.DecodeRuneInString(l.src[l.pos:])
l.pos += width
switch r {
case '\'':
nextRune, width := utf8.DecodeRuneInString(l.src[l.pos:])
if nextRune != '\'' {
return rawState
}
l.pos += width
case utf8.RuneError:
if width != replacementcharacterwidth {
if l.pos-l.start > 0 {
l.parts = append(l.parts, l.src[l.start:l.pos])
l.start = l.pos
}
return nil
}
}
}
}
func doubleQuoteState(l *sqlLexer) stateFn {
for {
r, width := utf8.DecodeRuneInString(l.src[l.pos:])
l.pos += width
switch r {
case '"':
nextRune, width := utf8.DecodeRuneInString(l.src[l.pos:])
if nextRune != '"' {
return rawState
}
l.pos += width
case utf8.RuneError:
if width != replacementcharacterwidth {
if l.pos-l.start > 0 {
l.parts = append(l.parts, l.src[l.start:l.pos])
l.start = l.pos
}
return nil
}
}
}
}
// placeholderState consumes a placeholder value. The $ must have already has
// already been consumed. The first rune must be a digit.
func placeholderState(l *sqlLexer) stateFn {
num := 0
for {
r, width := utf8.DecodeRuneInString(l.src[l.pos:])
l.pos += width
if '0' <= r && r <= '9' {
num *= 10
num += int(r - '0')
} else {
l.parts = append(l.parts, num)
l.pos -= width
l.start = l.pos
return rawState
}
}
}
func escapeStringState(l *sqlLexer) stateFn {
for {
r, width := utf8.DecodeRuneInString(l.src[l.pos:])
l.pos += width
switch r {
case '\\':
_, width = utf8.DecodeRuneInString(l.src[l.pos:])
l.pos += width
case '\'':
nextRune, width := utf8.DecodeRuneInString(l.src[l.pos:])
if nextRune != '\'' {
return rawState
}
l.pos += width
case utf8.RuneError:
if width != replacementcharacterwidth {
if l.pos-l.start > 0 {
l.parts = append(l.parts, l.src[l.start:l.pos])
l.start = l.pos
}
return nil
}
}
}
}
func oneLineCommentState(l *sqlLexer) stateFn {
for {
r, width := utf8.DecodeRuneInString(l.src[l.pos:])
l.pos += width
switch r {
case '\\':
_, width = utf8.DecodeRuneInString(l.src[l.pos:])
l.pos += width
case '\n', '\r':
return rawState
case utf8.RuneError:
if width != replacementcharacterwidth {
if l.pos-l.start > 0 {
l.parts = append(l.parts, l.src[l.start:l.pos])
l.start = l.pos
}
return nil
}
}
}
}
func multilineCommentState(l *sqlLexer) stateFn {
for {
r, width := utf8.DecodeRuneInString(l.src[l.pos:])
l.pos += width
switch r {
case '/':
nextRune, width := utf8.DecodeRuneInString(l.src[l.pos:])
if nextRune == '*' {
l.pos += width
l.nested++
}
case '*':
nextRune, width := utf8.DecodeRuneInString(l.src[l.pos:])
if nextRune != '/' {
continue
}
l.pos += width
if l.nested == 0 {
return rawState
}
l.nested--
case utf8.RuneError:
if width != replacementcharacterwidth {
if l.pos-l.start > 0 {
l.parts = append(l.parts, l.src[l.start:l.pos])
l.start = l.pos
}
return nil
}
}
}
}
// SanitizeSQL replaces placeholder values with args. It quotes and escapes args
// as necessary. This function is only safe when standard_conforming_strings is
// on.
func SanitizeSQL(sql string, args ...any) (string, error) {
query, err := NewQuery(sql)
if err != nil {
return "", err
}
return query.Sanitize(args...)
}
|
