forked from cockroachdb/cockroach
-
Notifications
You must be signed in to change notification settings - Fork 0
/
scan.go
817 lines (735 loc) · 16.6 KB
/
scan.go
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
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
// Copyright 2015 The Cockroach Authors.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
// implied. See the License for the specific language governing
// permissions and limitations under the License.
package parser
import (
"bytes"
"encoding/hex"
"fmt"
"go/constant"
"go/token"
"strconv"
"strings"
"unicode/utf8"
"github.com/cockroachdb/cockroach/pkg/sql/pgwire/pgerror"
)
const eof = -1
const errUnterminated = "unterminated string"
const errInvalidUTF8 = "invalid UTF-8 byte sequence"
const errInvalidHexNumeric = "invalid hexadecimal numeric literal"
const singleQuote = '\''
// Scanner lexes SQL statements.
type Scanner struct {
in string
pos int
tokBuf sqlSymType
lastTok sqlSymType
nextTok *sqlSymType
lastError struct {
msg string
hint string
detail string
unimplementedFeature string
}
stmts []Statement
identQuote int
stringQuote int
initialized bool
}
// MakeScanner makes a Scanner from str.
func MakeScanner(str string) Scanner {
var s Scanner
s.init(str)
return s
}
func (s *Scanner) init(str string) {
if s.initialized {
panic("scanner already initialized; a scanner cannot be reused.")
}
s.initialized = true
s.in = str
s.identQuote = '"'
}
// Tokens calls f on all tokens of the input until an EOF is encountered.
func (s *Scanner) Tokens(f func(token int)) {
for {
t := s.Lex(&s.tokBuf)
if t == 0 {
return
}
f(t)
}
}
// Lex lexes a token from input.
func (s *Scanner) Lex(lval *sqlSymType) int {
// The core lexing takes place in scan(). Here we do a small bit of post
// processing of the lexical tokens so that the grammar only requires
// one-token lookahead despite SQL requiring multi-token lookahead in some
// cases. These special cases are handled below and the returned tokens are
// adjusted to reflect the lookahead (LA) that occurred.
if s.nextTok != nil {
*lval = *s.nextTok
s.nextTok = nil
} else {
s.scan(lval)
}
switch lval.id {
case NOT, NULLS, WITH, AS:
default:
s.lastTok = *lval
return lval.id
}
s.nextTok = &s.tokBuf
s.scan(s.nextTok)
// If you update these cases, update lookaheadKeywords below.
switch lval.id {
case AS:
switch s.nextTok.id {
case OF:
lval.id = AS_LA
}
case NOT:
switch s.nextTok.id {
case BETWEEN, IN, LIKE, ILIKE, SIMILAR:
lval.id = NOT_LA
}
case WITH:
switch s.nextTok.id {
case TIME, ORDINALITY:
lval.id = WITH_LA
}
}
s.lastTok = *lval
return lval.id
}
// Unimplemented wraps Error, setting lastUnimplementedError.
func (s *Scanner) Unimplemented(feature string) {
s.Error("unimplemented")
s.lastError.unimplementedFeature = feature
}
// UnimplementedWithIssue wraps Error, setting lastUnimplementedError.
func (s *Scanner) UnimplementedWithIssue(issue int) {
s.Error("unimplemented")
s.lastError.unimplementedFeature = fmt.Sprintf("#%d", issue)
s.lastError.hint = fmt.Sprintf("See: https://github.com/cockroachdb/cockroach/issues/%d", issue)
}
// SetHelp marks the "last error" field in the Scanner to become a
// help text. This method is invoked in the error action of the
// parser, so the help text is only produced if the last token
// encountered was HELPTOKEN -- other cases are just syntax errors,
// and in that case we do not want the help text to overwrite the
// lastError field, which was set earlier to contain details about the
// syntax error.
func (s *Scanner) SetHelp(msg HelpMessage) {
if s.lastTok.id == HELPTOKEN {
s.populateHelpMsg(msg.String())
} else {
if msg.Command != "" {
s.lastError.hint = `try \h ` + msg.Command
} else {
s.lastError.hint = `try \hf ` + msg.Function
}
}
}
func (s *Scanner) populateHelpMsg(msg string) {
s.lastError.unimplementedFeature = ""
s.lastError.msg = "help token in input"
s.lastError.hint = msg
}
func (s *Scanner) Error(e string) {
if s.lastTok.id == ERROR {
// This is a tokenizer (lexical) error: just emit the invalid
// input as error.
s.lastError.msg = s.lastTok.str
} else {
// This is a contextual error. Print the provided error message
// and the error context.
s.lastError.msg = fmt.Sprintf("%s at or near \"%s\"", e, s.lastTok.str)
}
// Find the end of the line containing the last token.
i := strings.Index(s.in[s.lastTok.pos:], "\n")
if i == -1 {
i = len(s.in)
} else {
i += s.lastTok.pos
}
// Find the beginning of the line containing the last token. Note that
// LastIndex returns -1 if "\n" could not be found.
j := strings.LastIndex(s.in[:s.lastTok.pos], "\n") + 1
// Output everything up to and including the line containing the last token.
var buf bytes.Buffer
fmt.Fprintf(&buf, "source SQL:\n%s\n", s.in[:i])
// Output a caret indicating where the last token starts.
fmt.Fprintf(&buf, "%s^", strings.Repeat(" ", s.lastTok.pos-j))
s.lastError.detail = buf.String()
s.lastError.unimplementedFeature = ""
}
func (s *Scanner) scan(lval *sqlSymType) {
lval.id = 0
lval.pos = s.pos
lval.str = "EOF"
if _, ok := s.skipWhitespace(lval, true); !ok {
return
}
ch := s.next()
if ch == eof {
lval.pos = s.pos
return
}
lval.id = ch
lval.pos = s.pos - 1
lval.str = s.in[lval.pos:s.pos]
switch ch {
case '$':
// placeholder? $[0-9]+
if isDigit(s.peek()) {
s.scanPlaceholder(lval)
return
}
return
case s.identQuote:
// "[^"]"
if s.scanString(lval, s.identQuote, false, true) {
lval.id = IDENT
}
return
case singleQuote:
// '[^']'
if s.scanString(lval, ch, false, true) {
lval.id = SCONST
}
return
case s.stringQuote:
// '[^']'
if s.scanString(lval, s.stringQuote, false, true) {
lval.id = SCONST
}
return
case 'b', 'B':
// Bytes?
if t := s.peek(); t == singleQuote || t == s.stringQuote {
// [bB]'[^']'
s.pos++
if s.scanString(lval, t, true, false) {
lval.id = BCONST
}
return
}
s.scanIdent(lval)
return
case 'r', 'R':
s.scanIdent(lval)
return
case 'e', 'E':
// Escaped string?
if t := s.peek(); t == singleQuote || t == s.stringQuote {
// [eE]'[^']'
s.pos++
if s.scanString(lval, t, true, true) {
lval.id = SCONST
}
return
}
s.scanIdent(lval)
return
case 'x', 'X':
// Hex literal?
if t := s.peek(); t == singleQuote || t == s.stringQuote {
// [xX]'[a-f0-9]'
s.pos++
if s.scanStringOrHex(lval, t, false, false, true) {
lval.id = BCONST
}
return
}
s.scanIdent(lval)
return
case '.':
switch t := s.peek(); {
case t == '.': // ..
s.pos++
lval.id = DOT_DOT
return
case isDigit(t):
s.scanNumber(lval, ch)
return
}
return
case '!':
switch s.peek() {
case '=': // !=
s.pos++
lval.id = NOT_EQUALS
return
case '~': // !~
s.pos++
switch s.peek() {
case '*': // !~*
s.pos++
lval.id = NOT_REGIMATCH
return
}
lval.id = NOT_REGMATCH
return
}
return
case '?':
lval.id = HELPTOKEN
return
case '<':
switch s.peek() {
case '<': // <<
s.pos++
lval.id = LSHIFT
return
case '>': // <>
s.pos++
lval.id = NOT_EQUALS
return
case '=': // <=
s.pos++
lval.id = LESS_EQUALS
return
}
return
case '>':
switch s.peek() {
case '>': // >>
s.pos++
lval.id = RSHIFT
return
case '=': // >=
s.pos++
lval.id = GREATER_EQUALS
return
}
return
case ':':
switch s.peek() {
case ':': // ::
if s.peekN(1) == ':' {
// :::
s.pos += 2
lval.id = TYPEANNOTATE
return
}
s.pos++
lval.id = TYPECAST
return
}
return
case '|':
switch s.peek() {
case '|': // ||
s.pos++
lval.id = CONCAT
return
}
return
case '/':
switch s.peek() {
case '/': // //
s.pos++
lval.id = FLOORDIV
return
}
return
case '~':
switch s.peek() {
case '*': // ~*
s.pos++
lval.id = REGIMATCH
return
}
return
default:
if isDigit(ch) {
s.scanNumber(lval, ch)
return
}
if isIdentStart(ch) {
s.scanIdent(lval)
return
}
}
// Everything else is a single character token which we already initialized
// lval for above.
}
func (s *Scanner) peek() int {
if s.pos >= len(s.in) {
return eof
}
return int(s.in[s.pos])
}
func (s *Scanner) peekN(n int) int {
pos := s.pos + n
if pos >= len(s.in) {
return eof
}
return int(s.in[pos])
}
func (s *Scanner) next() int {
ch := s.peek()
if ch != eof {
s.pos++
}
return ch
}
func (s *Scanner) skipWhitespace(lval *sqlSymType, allowComments bool) (newline, ok bool) {
newline = false
for {
ch := s.peek()
if ch == '\n' {
s.pos++
newline = true
continue
}
if ch == ' ' || ch == '\t' || ch == '\r' || ch == '\f' {
s.pos++
continue
}
if allowComments {
if present, cok := s.scanComment(lval); !cok {
return false, false
} else if present {
continue
}
}
break
}
return newline, true
}
func (s *Scanner) scanComment(lval *sqlSymType) (present, ok bool) {
start := s.pos
ch := s.peek()
if ch == '/' {
s.pos++
if s.peek() != '*' {
s.pos--
return false, true
}
s.pos++
depth := 1
for {
switch s.next() {
case '*':
if s.peek() == '/' {
s.pos++
depth--
if depth == 0 {
return true, true
}
continue
}
case '/':
if s.peek() == '*' {
s.pos++
depth++
continue
}
case eof:
lval.id = ERROR
lval.pos = start
lval.str = "unterminated comment"
return false, false
}
}
}
if ch == '-' {
s.pos++
if s.peek() != '-' {
s.pos--
return false, true
}
for {
switch s.next() {
case eof, '\n':
return true, true
}
}
}
return false, true
}
func (s *Scanner) scanIdent(lval *sqlSymType) {
start := s.pos - 1
for ; isIdentMiddle(s.peek()); s.pos++ {
}
lval.str = Name(s.in[start:s.pos]).Normalize()
if id, ok := keywords[strings.ToUpper(lval.str)]; ok {
lval.id = id
} else {
lval.id = IDENT
}
}
func (s *Scanner) scanNumber(lval *sqlSymType, ch int) {
start := s.pos - 1
isHex := false
hasDecimal := ch == '.'
hasExponent := false
for {
ch := s.peek()
if isHex && isHexDigit(ch) || isDigit(ch) {
s.pos++
continue
}
if ch == 'x' || ch == 'X' {
if isHex || s.in[start] != '0' || s.pos != start+1 {
lval.id = ERROR
lval.str = errInvalidHexNumeric
return
}
s.pos++
isHex = true
continue
}
if isHex {
break
}
if ch == '.' {
if hasDecimal || hasExponent {
break
}
s.pos++
if s.peek() == '.' {
// Found ".." while scanning a number: back up to the end of the
// integer.
s.pos--
break
}
hasDecimal = true
continue
}
if ch == 'e' || ch == 'E' {
if hasExponent {
break
}
hasExponent = true
s.pos++
ch = s.peek()
if ch == '-' || ch == '+' {
s.pos++
}
ch = s.peek()
if !isDigit(ch) {
lval.id = ERROR
lval.str = "invalid floating point literal"
return
}
continue
}
break
}
lval.str = s.in[start:s.pos]
if hasDecimal || hasExponent {
lval.id = FCONST
floatConst := constant.MakeFromLiteral(lval.str, token.FLOAT, 0)
if floatConst.Kind() == constant.Unknown {
lval.id = ERROR
lval.str = fmt.Sprintf("could not make constant float from literal %q", lval.str)
return
}
lval.union.val = &NumVal{Value: floatConst, OrigString: lval.str}
} else {
if isHex && s.pos == start+2 {
lval.id = ERROR
lval.str = errInvalidHexNumeric
return
}
// Strip off leading zeros from non-hex (decimal) literals so that
// constant.MakeFromLiteral doesn't inappropriately interpret the
// string as an octal literal. Note: we can't use strings.TrimLeft
// here, because it will truncate '0' to ''.
if !isHex {
for len(lval.str) > 1 && lval.str[0] == '0' {
lval.str = lval.str[1:]
}
}
lval.id = ICONST
intConst := constant.MakeFromLiteral(lval.str, token.INT, 0)
if intConst.Kind() == constant.Unknown {
lval.id = ERROR
lval.str = fmt.Sprintf("could not make constant int from literal %q", lval.str)
return
}
lval.union.val = &NumVal{Value: intConst, OrigString: lval.str}
}
}
func (s *Scanner) scanPlaceholder(lval *sqlSymType) {
start := s.pos
for isDigit(s.peek()) {
s.pos++
}
lval.str = s.in[start:s.pos]
uval, err := strconv.ParseUint(lval.str, 10, 64)
if err == nil && uval > 1<<63 {
err = pgerror.NewErrorf(pgerror.CodeNumericValueOutOfRangeError, "integer value out of range: %d", uval)
}
if err != nil {
lval.id = ERROR
lval.str = err.Error()
return
}
// uval is now in the range [0, 1<<63]. Casting to an int64 leaves the range
// [0, 1<<63 - 1] intact and moves 1<<63 to -1<<63 (a.k.a. math.MinInt64).
lval.union.val = &NumVal{Value: constant.MakeUint64(uval)}
lval.id = PLACEHOLDER
}
func (s *Scanner) scanString(lval *sqlSymType, ch int, allowEscapes, requireUTF8 bool) bool {
return s.scanStringOrHex(lval, ch, allowEscapes, requireUTF8, false)
}
func (s *Scanner) scanStringOrHex(
lval *sqlSymType, ch int, allowEscapes, requireUTF8 bool, decodeHex bool,
) bool {
var buf []byte
var runeTmp [utf8.UTFMax]byte
start := s.pos
outer:
for {
switch s.next() {
case ch:
buf = append(buf, s.in[start:s.pos-1]...)
if s.peek() == ch {
// Double quote is translated into a single quote that is part of the
// string.
start = s.pos
s.pos++
continue
}
if newline, ok := s.skipWhitespace(lval, false); !ok {
return false
} else if !newline {
break outer
}
// SQL allows joining adjacent strings separated by whitespace as long as
// that whitespace contains at least one newline. Kind of strange to
// require the newline, but that is the standard.
if s.peek() != ch {
break outer
}
s.pos++
start = s.pos
continue
case '\\':
t := s.peek()
if allowEscapes {
buf = append(buf, s.in[start:s.pos-1]...)
if t == ch {
start = s.pos
s.pos++
continue
}
switch t {
case 'a', 'b', 'f', 'n', 'r', 't', 'v', 'x', 'X', 'u', 'U', '\\',
'0', '1', '2', '3', '4', '5', '6', '7':
var tmp string
if t == 'X' && len(s.in[s.pos:]) >= 3 {
// UnquoteChar doesn't handle 'X' so we create a temporary string
// for it to parse.
tmp = "\\x" + s.in[s.pos+1:s.pos+3]
} else {
tmp = s.in[s.pos-1:]
}
v, multibyte, tail, err := strconv.UnquoteChar(tmp, byte(ch))
if err != nil {
lval.id = ERROR
lval.str = err.Error()
return false
}
if v < utf8.RuneSelf || !multibyte {
buf = append(buf, byte(v))
} else {
n := utf8.EncodeRune(runeTmp[:], v)
buf = append(buf, runeTmp[:n]...)
}
s.pos += len(tmp) - len(tail) - 1
start = s.pos
continue
}
// If we end up here, it's a redundant escape - simply drop the
// backslash. For example, e'\"' is equivalent to e'"', and
// e'\d\b' to e'd\b'. This is what Postgres does:
// http://www.postgresql.org/docs/9.4/static/sql-syntax-lexical.html#SQL-SYNTAX-STRINGS-ESCAPE
start = s.pos
}
case eof:
lval.id = ERROR
lval.str = errUnterminated
return false
}
}
var decB []byte
if decodeHex {
decB = make([]byte, len(buf)/2)
_, err := hex.Decode(decB, buf)
if err != nil {
// Either the string has an odd number of characters or contains one or
// more invalid bytes.
lval.id = ERROR
lval.str = "invalid hexadecimal bytes literal"
return false
}
buf = decB
}
if requireUTF8 && !utf8.Valid(buf) {
lval.id = ERROR
lval.str = errInvalidUTF8
return false
}
lval.str = string(buf)
return true
}
func isDigit(ch int) bool {
return ch >= '0' && ch <= '9'
}
func isHexDigit(ch int) bool {
return (ch >= '0' && ch <= '9') ||
(ch >= 'a' && ch <= 'f') ||
(ch >= 'A' && ch <= 'F')
}
var lookaheadKeywords = map[string]struct{}{
"BETWEEN": {},
"ILIKE": {},
"IN": {},
"LIKE": {},
"OF": {},
"ORDINALITY": {},
"SIMILAR": {},
"TIME": {},
}
func isNonKeywordBareIdentifier(s string) bool {
if len(s) == 0 || !isIdentStart(int(s[0])) {
return false
}
for i := 1; i < len(s); i++ {
if !isIdentMiddle(int(s[i])) {
return false
}
}
upper := strings.ToUpper(s)
if _, ok := reservedKeywords[upper]; ok {
return false
}
if _, ok := lookaheadKeywords[upper]; ok {
return false
}
return Name(s).Normalize() == s
}
func isIdentStart(ch int) bool {
return (ch >= 'A' && ch <= 'Z') ||
(ch >= 'a' && ch <= 'z') ||
(ch >= 128 && ch <= 255) ||
(ch == '_')
}
func isIdentMiddle(ch int) bool {
return isIdentStart(ch) || isDigit(ch) || ch == '$'
}