aboutsummaryrefslogtreecommitdiff
path: root/libgo/go/text/scanner/scanner.go
blob: f46f63d0ee6b438aa374c7b0727c9f4298b035a5 (plain)
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
// Copyright 2009 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.

// Package scanner provides a scanner and tokenizer for UTF-8-encoded text.
// It takes an io.Reader providing the source, which then can be tokenized
// through repeated calls to the Scan function.  For compatibility with
// existing tools, the NUL character is not allowed (implementation
// restriction).
//
// By default, a Scanner skips white space and Go comments and recognizes all
// literals as defined by the Go language specification.  It may be
// customized to recognize only a subset of those literals and to recognize
// different white space characters.
//
// Basic usage pattern:
//
//	var s scanner.Scanner
//	s.Init(src)
//	tok := s.Scan()
//	for tok != scanner.EOF {
//		// do something with tok
//		tok = s.Scan()
//	}
//
package scanner

import (
	"bytes"
	"fmt"
	"io"
	"os"
	"unicode"
	"unicode/utf8"
)

// TODO(gri): Consider changing this to use the new (token) Position package.

// A source position is represented by a Position value.
// A position is valid if Line > 0.
type Position struct {
	Filename string // filename, if any
	Offset   int    // byte offset, starting at 0
	Line     int    // line number, starting at 1
	Column   int    // column number, starting at 1 (character count per line)
}

// IsValid returns true if the position is valid.
func (pos *Position) IsValid() bool { return pos.Line > 0 }

func (pos Position) String() string {
	s := pos.Filename
	if pos.IsValid() {
		if s != "" {
			s += ":"
		}
		s += fmt.Sprintf("%d:%d", pos.Line, pos.Column)
	}
	if s == "" {
		s = "???"
	}
	return s
}

// Predefined mode bits to control recognition of tokens. For instance,
// to configure a Scanner such that it only recognizes (Go) identifiers,
// integers, and skips comments, set the Scanner's Mode field to:
//
//	ScanIdents | ScanInts | SkipComments
//
const (
	ScanIdents     = 1 << -Ident
	ScanInts       = 1 << -Int
	ScanFloats     = 1 << -Float // includes Ints
	ScanChars      = 1 << -Char
	ScanStrings    = 1 << -String
	ScanRawStrings = 1 << -RawString
	ScanComments   = 1 << -Comment
	SkipComments   = 1 << -skipComment // if set with ScanComments, comments become white space
	GoTokens       = ScanIdents | ScanFloats | ScanChars | ScanStrings | ScanRawStrings | ScanComments | SkipComments
)

// The result of Scan is one of the following tokens or a Unicode character.
const (
	EOF = -(iota + 1)
	Ident
	Int
	Float
	Char
	String
	RawString
	Comment
	skipComment
)

var tokenString = map[rune]string{
	EOF:       "EOF",
	Ident:     "Ident",
	Int:       "Int",
	Float:     "Float",
	Char:      "Char",
	String:    "String",
	RawString: "RawString",
	Comment:   "Comment",
}

// TokenString returns a (visible) string for a token or Unicode character.
func TokenString(tok rune) string {
	if s, found := tokenString[tok]; found {
		return s
	}
	return fmt.Sprintf("%q", string(tok))
}

// GoWhitespace is the default value for the Scanner's Whitespace field.
// Its value selects Go's white space characters.
const GoWhitespace = 1<<'\t' | 1<<'\n' | 1<<'\r' | 1<<' '

const bufLen = 1024 // at least utf8.UTFMax

// A Scanner implements reading of Unicode characters and tokens from an io.Reader.
type Scanner struct {
	// Input
	src io.Reader

	// Source buffer
	srcBuf [bufLen + 1]byte // +1 for sentinel for common case of s.next()
	srcPos int              // reading position (srcBuf index)
	srcEnd int              // source end (srcBuf index)

	// Source position
	srcBufOffset int // byte offset of srcBuf[0] in source
	line         int // line count
	column       int // character count
	lastLineLen  int // length of last line in characters (for correct column reporting)
	lastCharLen  int // length of last character in bytes

	// Token text buffer
	// Typically, token text is stored completely in srcBuf, but in general
	// the token text's head may be buffered in tokBuf while the token text's
	// tail is stored in srcBuf.
	tokBuf bytes.Buffer // token text head that is not in srcBuf anymore
	tokPos int          // token text tail position (srcBuf index); valid if >= 0
	tokEnd int          // token text tail end (srcBuf index)

	// One character look-ahead
	ch rune // character before current srcPos

	// Error is called for each error encountered. If no Error
	// function is set, the error is reported to os.Stderr.
	Error func(s *Scanner, msg string)

	// ErrorCount is incremented by one for each error encountered.
	ErrorCount int

	// The Mode field controls which tokens are recognized. For instance,
	// to recognize Ints, set the ScanInts bit in Mode. The field may be
	// changed at any time.
	Mode uint

	// The Whitespace field controls which characters are recognized
	// as white space. To recognize a character ch <= ' ' as white space,
	// set the ch'th bit in Whitespace (the Scanner's behavior is undefined
	// for values ch > ' '). The field may be changed at any time.
	Whitespace uint64

	// Start position of most recently scanned token; set by Scan.
	// Calling Init or Next invalidates the position (Line == 0).
	// The Filename field is always left untouched by the Scanner.
	// If an error is reported (via Error) and Position is invalid,
	// the scanner is not inside a token. Call Pos to obtain an error
	// position in that case.
	Position
}

// Init initializes a Scanner with a new source and returns s.
// Error is set to nil, ErrorCount is set to 0, Mode is set to GoTokens,
// and Whitespace is set to GoWhitespace.
func (s *Scanner) Init(src io.Reader) *Scanner {
	s.src = src

	// initialize source buffer
	// (the first call to next() will fill it by calling src.Read)
	s.srcBuf[0] = utf8.RuneSelf // sentinel
	s.srcPos = 0
	s.srcEnd = 0

	// initialize source position
	s.srcBufOffset = 0
	s.line = 1
	s.column = 0
	s.lastLineLen = 0
	s.lastCharLen = 0

	// initialize token text buffer
	// (required for first call to next()).
	s.tokPos = -1

	// initialize one character look-ahead
	s.ch = -1 // no char read yet

	// initialize public fields
	s.Error = nil
	s.ErrorCount = 0
	s.Mode = GoTokens
	s.Whitespace = GoWhitespace
	s.Line = 0 // invalidate token position

	return s
}

// TODO(gri): The code for next() and the internal scanner state could benefit
//            from a rethink. While next() is optimized for the common ASCII
//            case, the "corrections" needed for proper position tracking undo
//            some of the attempts for fast-path optimization.

// next reads and returns the next Unicode character. It is designed such
// that only a minimal amount of work needs to be done in the common ASCII
// case (one test to check for both ASCII and end-of-buffer, and one test
// to check for newlines).
func (s *Scanner) next() rune {
	ch, width := rune(s.srcBuf[s.srcPos]), 1

	if ch >= utf8.RuneSelf {
		// uncommon case: not ASCII or not enough bytes
		for s.srcPos+utf8.UTFMax > s.srcEnd && !utf8.FullRune(s.srcBuf[s.srcPos:s.srcEnd]) {
			// not enough bytes: read some more, but first
			// save away token text if any
			if s.tokPos >= 0 {
				s.tokBuf.Write(s.srcBuf[s.tokPos:s.srcPos])
				s.tokPos = 0
				// s.tokEnd is set by Scan()
			}
			// move unread bytes to beginning of buffer
			copy(s.srcBuf[0:], s.srcBuf[s.srcPos:s.srcEnd])
			s.srcBufOffset += s.srcPos
			// read more bytes
			// (an io.Reader must return io.EOF when it reaches
			// the end of what it is reading - simply returning
			// n == 0 will make this loop retry forever; but the
			// error is in the reader implementation in that case)
			i := s.srcEnd - s.srcPos
			n, err := s.src.Read(s.srcBuf[i:bufLen])
			s.srcPos = 0
			s.srcEnd = i + n
			s.srcBuf[s.srcEnd] = utf8.RuneSelf // sentinel
			if err != nil {
				if s.srcEnd == 0 {
					if s.lastCharLen > 0 {
						// previous character was not EOF
						s.column++
					}
					s.lastCharLen = 0
					return EOF
				}
				if err != io.EOF {
					s.error(err.Error())
				}
				// If err == EOF, we won't be getting more
				// bytes; break to avoid infinite loop. If
				// err is something else, we don't know if
				// we can get more bytes; thus also break.
				break
			}
		}
		// at least one byte
		ch = rune(s.srcBuf[s.srcPos])
		if ch >= utf8.RuneSelf {
			// uncommon case: not ASCII
			ch, width = utf8.DecodeRune(s.srcBuf[s.srcPos:s.srcEnd])
			if ch == utf8.RuneError && width == 1 {
				// advance for correct error position
				s.srcPos += width
				s.lastCharLen = width
				s.column++
				s.error("illegal UTF-8 encoding")
				return ch
			}
		}
	}

	// advance
	s.srcPos += width
	s.lastCharLen = width
	s.column++

	// special situations
	switch ch {
	case 0:
		// implementation restriction for compatibility with other tools
		s.error("illegal character NUL")
	case '\n':
		s.line++
		s.lastLineLen = s.column
		s.column = 0
	}

	return ch
}

// Next reads and returns the next Unicode character.
// It returns EOF at the end of the source. It reports
// a read error by calling s.Error, if not nil; otherwise
// it prints an error message to os.Stderr. Next does not
// update the Scanner's Position field; use Pos() to
// get the current position.
func (s *Scanner) Next() rune {
	s.tokPos = -1 // don't collect token text
	s.Line = 0    // invalidate token position
	ch := s.Peek()
	s.ch = s.next()
	return ch
}

// Peek returns the next Unicode character in the source without advancing
// the scanner. It returns EOF if the scanner's position is at the last
// character of the source.
func (s *Scanner) Peek() rune {
	if s.ch < 0 {
		s.ch = s.next()
	}
	return s.ch
}

func (s *Scanner) error(msg string) {
	s.ErrorCount++
	if s.Error != nil {
		s.Error(s, msg)
		return
	}
	pos := s.Position
	if !pos.IsValid() {
		pos = s.Pos()
	}
	fmt.Fprintf(os.Stderr, "%s: %s\n", pos, msg)
}

func (s *Scanner) scanIdentifier() rune {
	ch := s.next() // read character after first '_' or letter
	for ch == '_' || unicode.IsLetter(ch) || unicode.IsDigit(ch) {
		ch = s.next()
	}
	return ch
}

func digitVal(ch rune) int {
	switch {
	case '0' <= ch && ch <= '9':
		return int(ch - '0')
	case 'a' <= ch && ch <= 'f':
		return int(ch - 'a' + 10)
	case 'A' <= ch && ch <= 'F':
		return int(ch - 'A' + 10)
	}
	return 16 // larger than any legal digit val
}

func isDecimal(ch rune) bool { return '0' <= ch && ch <= '9' }

func (s *Scanner) scanMantissa(ch rune) rune {
	for isDecimal(ch) {
		ch = s.next()
	}
	return ch
}

func (s *Scanner) scanFraction(ch rune) rune {
	if ch == '.' {
		ch = s.scanMantissa(s.next())
	}
	return ch
}

func (s *Scanner) scanExponent(ch rune) rune {
	if ch == 'e' || ch == 'E' {
		ch = s.next()
		if ch == '-' || ch == '+' {
			ch = s.next()
		}
		ch = s.scanMantissa(ch)
	}
	return ch
}

func (s *Scanner) scanNumber(ch rune) (rune, rune) {
	// isDecimal(ch)
	if ch == '0' {
		// int or float
		ch = s.next()
		if ch == 'x' || ch == 'X' {
			// hexadecimal int
			ch = s.next()
			for digitVal(ch) < 16 {
				ch = s.next()
			}
		} else {
			// octal int or float
			seenDecimalDigit := false
			for isDecimal(ch) {
				if ch > '7' {
					seenDecimalDigit = true
				}
				ch = s.next()
			}
			if s.Mode&ScanFloats != 0 && (ch == '.' || ch == 'e' || ch == 'E') {
				// float
				ch = s.scanFraction(ch)
				ch = s.scanExponent(ch)
				return Float, ch
			}
			// octal int
			if seenDecimalDigit {
				s.error("illegal octal number")
			}
		}
		return Int, ch
	}
	// decimal int or float
	ch = s.scanMantissa(ch)
	if s.Mode&ScanFloats != 0 && (ch == '.' || ch == 'e' || ch == 'E') {
		// float
		ch = s.scanFraction(ch)
		ch = s.scanExponent(ch)
		return Float, ch
	}
	return Int, ch
}

func (s *Scanner) scanDigits(ch rune, base, n int) rune {
	for n > 0 && digitVal(ch) < base {
		ch = s.next()
		n--
	}
	if n > 0 {
		s.error("illegal char escape")
	}
	return ch
}

func (s *Scanner) scanEscape(quote rune) rune {
	ch := s.next() // read character after '/'
	switch ch {
	case 'a', 'b', 'f', 'n', 'r', 't', 'v', '\\', quote:
		// nothing to do
		ch = s.next()
	case '0', '1', '2', '3', '4', '5', '6', '7':
		ch = s.scanDigits(ch, 8, 3)
	case 'x':
		ch = s.scanDigits(s.next(), 16, 2)
	case 'u':
		ch = s.scanDigits(s.next(), 16, 4)
	case 'U':
		ch = s.scanDigits(s.next(), 16, 8)
	default:
		s.error("illegal char escape")
	}
	return ch
}

func (s *Scanner) scanString(quote rune) (n int) {
	ch := s.next() // read character after quote
	for ch != quote {
		if ch == '\n' || ch < 0 {
			s.error("literal not terminated")
			return
		}
		if ch == '\\' {
			ch = s.scanEscape(quote)
		} else {
			ch = s.next()
		}
		n++
	}
	return
}

func (s *Scanner) scanRawString() {
	ch := s.next() // read character after '`'
	for ch != '`' {
		if ch < 0 {
			s.error("literal not terminated")
			return
		}
		ch = s.next()
	}
}

func (s *Scanner) scanChar() {
	if s.scanString('\'') != 1 {
		s.error("illegal char literal")
	}
}

func (s *Scanner) scanComment(ch rune) rune {
	// ch == '/' || ch == '*'
	if ch == '/' {
		// line comment
		ch = s.next() // read character after "//"
		for ch != '\n' && ch >= 0 {
			ch = s.next()
		}
		return ch
	}

	// general comment
	ch = s.next() // read character after "/*"
	for {
		if ch < 0 {
			s.error("comment not terminated")
			break
		}
		ch0 := ch
		ch = s.next()
		if ch0 == '*' && ch == '/' {
			ch = s.next()
			break
		}
	}
	return ch
}

// Scan reads the next token or Unicode character from source and returns it.
// It only recognizes tokens t for which the respective Mode bit (1<<-t) is set.
// It returns EOF at the end of the source. It reports scanner errors (read and
// token errors) by calling s.Error, if not nil; otherwise it prints an error
// message to os.Stderr.
func (s *Scanner) Scan() rune {
	ch := s.Peek()

	// reset token text position
	s.tokPos = -1
	s.Line = 0

redo:
	// skip white space
	for s.Whitespace&(1<<uint(ch)) != 0 {
		ch = s.next()
	}

	// start collecting token text
	s.tokBuf.Reset()
	s.tokPos = s.srcPos - s.lastCharLen

	// set token position
	// (this is a slightly optimized version of the code in Pos())
	s.Offset = s.srcBufOffset + s.tokPos
	if s.column > 0 {
		// common case: last character was not a '\n'
		s.Line = s.line
		s.Column = s.column
	} else {
		// last character was a '\n'
		// (we cannot be at the beginning of the source
		// since we have called next() at least once)
		s.Line = s.line - 1
		s.Column = s.lastLineLen
	}

	// determine token value
	tok := ch
	switch {
	case unicode.IsLetter(ch) || ch == '_':
		if s.Mode&ScanIdents != 0 {
			tok = Ident
			ch = s.scanIdentifier()
		} else {
			ch = s.next()
		}
	case isDecimal(ch):
		if s.Mode&(ScanInts|ScanFloats) != 0 {
			tok, ch = s.scanNumber(ch)
		} else {
			ch = s.next()
		}
	default:
		switch ch {
		case '"':
			if s.Mode&ScanStrings != 0 {
				s.scanString('"')
				tok = String
			}
			ch = s.next()
		case '\'':
			if s.Mode&ScanChars != 0 {
				s.scanChar()
				tok = Char
			}
			ch = s.next()
		case '.':
			ch = s.next()
			if isDecimal(ch) && s.Mode&ScanFloats != 0 {
				tok = Float
				ch = s.scanMantissa(ch)
				ch = s.scanExponent(ch)
			}
		case '/':
			ch = s.next()
			if (ch == '/' || ch == '*') && s.Mode&ScanComments != 0 {
				if s.Mode&SkipComments != 0 {
					s.tokPos = -1 // don't collect token text
					ch = s.scanComment(ch)
					goto redo
				}
				ch = s.scanComment(ch)
				tok = Comment
			}
		case '`':
			if s.Mode&ScanRawStrings != 0 {
				s.scanRawString()
				tok = String
			}
			ch = s.next()
		default:
			ch = s.next()
		}
	}

	// end of token text
	s.tokEnd = s.srcPos - s.lastCharLen

	s.ch = ch
	return tok
}

// Pos returns the position of the character immediately after
// the character or token returned by the last call to Next or Scan.
func (s *Scanner) Pos() (pos Position) {
	pos.Filename = s.Filename
	pos.Offset = s.srcBufOffset + s.srcPos - s.lastCharLen
	switch {
	case s.column > 0:
		// common case: last character was not a '\n'
		pos.Line = s.line
		pos.Column = s.column
	case s.lastLineLen > 0:
		// last character was a '\n'
		pos.Line = s.line - 1
		pos.Column = s.lastLineLen
	default:
		// at the beginning of the source
		pos.Line = 1
		pos.Column = 1
	}
	return
}

// TokenText returns the string corresponding to the most recently scanned token.
// Valid after calling Scan().
func (s *Scanner) TokenText() string {
	if s.tokPos < 0 {
		// no token text
		return ""
	}

	if s.tokEnd < 0 {
		// if EOF was reached, s.tokEnd is set to -1 (s.srcPos == 0)
		s.tokEnd = s.tokPos
	}

	if s.tokBuf.Len() == 0 {
		// common case: the entire token text is still in srcBuf
		return string(s.srcBuf[s.tokPos:s.tokEnd])
	}

	// part of the token text was saved in tokBuf: save the rest in
	// tokBuf as well and return its content
	s.tokBuf.Write(s.srcBuf[s.tokPos:s.tokEnd])
	s.tokPos = s.tokEnd // ensure idempotency of TokenText() call
	return s.tokBuf.String()
}