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
|
// Copyright 2011 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 regexp
import (
"io"
"regexp/syntax"
)
// A queue is a 'sparse array' holding pending threads of execution.
// See http://research.swtch.com/2008/03/using-uninitialized-memory-for-fun-and.html
type queue struct {
sparse []uint32
dense []entry
}
// A entry is an entry on a queue.
// It holds both the instruction pc and the actual thread.
// Some queue entries are just place holders so that the machine
// knows it has considered that pc. Such entries have t == nil.
type entry struct {
pc uint32
t *thread
}
// A thread is the state of a single path through the machine:
// an instruction and a corresponding capture array.
// See http://swtch.com/~rsc/regexp/regexp2.html
type thread struct {
inst *syntax.Inst
cap []int
}
// A machine holds all the state during an NFA simulation for p.
type machine struct {
re *Regexp // corresponding Regexp
p *syntax.Prog // compiled program
op *onePassProg // compiled onepass program, or notOnePass
q0, q1 queue // two queues for runq, nextq
pool []*thread // pool of available threads
matched bool // whether a match was found
matchcap []int // capture information for the match
// cached inputs, to avoid allocation
inputBytes inputBytes
inputString inputString
inputReader inputReader
}
func (m *machine) newInputBytes(b []byte) input {
m.inputBytes.str = b
return &m.inputBytes
}
func (m *machine) newInputString(s string) input {
m.inputString.str = s
return &m.inputString
}
func (m *machine) newInputReader(r io.RuneReader) input {
m.inputReader.r = r
m.inputReader.atEOT = false
m.inputReader.pos = 0
return &m.inputReader
}
// progMachine returns a new machine running the prog p.
func progMachine(p *syntax.Prog, op *onePassProg) *machine {
m := &machine{p: p, op: op}
n := len(m.p.Inst)
m.q0 = queue{make([]uint32, n), make([]entry, 0, n)}
m.q1 = queue{make([]uint32, n), make([]entry, 0, n)}
ncap := p.NumCap
if ncap < 2 {
ncap = 2
}
m.matchcap = make([]int, ncap)
return m
}
func (m *machine) init(ncap int) {
for _, t := range m.pool {
t.cap = t.cap[:ncap]
}
m.matchcap = m.matchcap[:ncap]
}
// alloc allocates a new thread with the given instruction.
// It uses the free pool if possible.
func (m *machine) alloc(i *syntax.Inst) *thread {
var t *thread
if n := len(m.pool); n > 0 {
t = m.pool[n-1]
m.pool = m.pool[:n-1]
} else {
t = new(thread)
t.cap = make([]int, len(m.matchcap), cap(m.matchcap))
}
t.inst = i
return t
}
// free returns t to the free pool.
func (m *machine) free(t *thread) {
m.inputBytes.str = nil
m.inputString.str = ""
m.inputReader.r = nil
m.pool = append(m.pool, t)
}
// match runs the machine over the input starting at pos.
// It reports whether a match was found.
// If so, m.matchcap holds the submatch information.
func (m *machine) match(i input, pos int) bool {
startCond := m.re.cond
if startCond == ^syntax.EmptyOp(0) { // impossible
return false
}
m.matched = false
for i := range m.matchcap {
m.matchcap[i] = -1
}
runq, nextq := &m.q0, &m.q1
r, r1 := endOfText, endOfText
width, width1 := 0, 0
r, width = i.step(pos)
if r != endOfText {
r1, width1 = i.step(pos + width)
}
var flag syntax.EmptyOp
if pos == 0 {
flag = syntax.EmptyOpContext(-1, r)
} else {
flag = i.context(pos)
}
for {
if len(runq.dense) == 0 {
if startCond&syntax.EmptyBeginText != 0 && pos != 0 {
// Anchored match, past beginning of text.
break
}
if m.matched {
// Have match; finished exploring alternatives.
break
}
if len(m.re.prefix) > 0 && r1 != m.re.prefixRune && i.canCheckPrefix() {
// Match requires literal prefix; fast search for it.
advance := i.index(m.re, pos)
if advance < 0 {
break
}
pos += advance
r, width = i.step(pos)
r1, width1 = i.step(pos + width)
}
}
if !m.matched {
if len(m.matchcap) > 0 {
m.matchcap[0] = pos
}
m.add(runq, uint32(m.p.Start), pos, m.matchcap, flag, nil)
}
flag = syntax.EmptyOpContext(r, r1)
m.step(runq, nextq, pos, pos+width, r, flag)
if width == 0 {
break
}
if len(m.matchcap) == 0 && m.matched {
// Found a match and not paying attention
// to where it is, so any match will do.
break
}
pos += width
r, width = r1, width1
if r != endOfText {
r1, width1 = i.step(pos + width)
}
runq, nextq = nextq, runq
}
m.clear(nextq)
return m.matched
}
// clear frees all threads on the thread queue.
func (m *machine) clear(q *queue) {
for _, d := range q.dense {
if d.t != nil {
// m.free(d.t)
m.pool = append(m.pool, d.t)
}
}
q.dense = q.dense[:0]
}
// step executes one step of the machine, running each of the threads
// on runq and appending new threads to nextq.
// The step processes the rune c (which may be endOfText),
// which starts at position pos and ends at nextPos.
// nextCond gives the setting for the empty-width flags after c.
func (m *machine) step(runq, nextq *queue, pos, nextPos int, c rune, nextCond syntax.EmptyOp) {
longest := m.re.longest
for j := 0; j < len(runq.dense); j++ {
d := &runq.dense[j]
t := d.t
if t == nil {
continue
}
if longest && m.matched && len(t.cap) > 0 && m.matchcap[0] < t.cap[0] {
// m.free(t)
m.pool = append(m.pool, t)
continue
}
i := t.inst
add := false
switch i.Op {
default:
panic("bad inst")
case syntax.InstMatch:
if len(t.cap) > 0 && (!longest || !m.matched || m.matchcap[1] < pos) {
t.cap[1] = pos
copy(m.matchcap, t.cap)
}
if !longest {
// First-match mode: cut off all lower-priority threads.
for _, d := range runq.dense[j+1:] {
if d.t != nil {
// m.free(d.t)
m.pool = append(m.pool, d.t)
}
}
runq.dense = runq.dense[:0]
}
m.matched = true
case syntax.InstRune:
add = i.MatchRune(c)
case syntax.InstRune1:
add = c == i.Rune[0]
case syntax.InstRuneAny:
add = true
case syntax.InstRuneAnyNotNL:
add = c != '\n'
}
if add {
t = m.add(nextq, i.Out, nextPos, t.cap, nextCond, t)
}
if t != nil {
// m.free(t)
m.pool = append(m.pool, t)
}
}
runq.dense = runq.dense[:0]
}
// add adds an entry to q for pc, unless the q already has such an entry.
// It also recursively adds an entry for all instructions reachable from pc by following
// empty-width conditions satisfied by cond. pos gives the current position
// in the input.
func (m *machine) add(q *queue, pc uint32, pos int, cap []int, cond syntax.EmptyOp, t *thread) *thread {
if pc == 0 {
return t
}
if j := q.sparse[pc]; j < uint32(len(q.dense)) && q.dense[j].pc == pc {
return t
}
j := len(q.dense)
q.dense = q.dense[:j+1]
d := &q.dense[j]
d.t = nil
d.pc = pc
q.sparse[pc] = uint32(j)
i := &m.p.Inst[pc]
switch i.Op {
default:
panic("unhandled")
case syntax.InstFail:
// nothing
case syntax.InstAlt, syntax.InstAltMatch:
t = m.add(q, i.Out, pos, cap, cond, t)
t = m.add(q, i.Arg, pos, cap, cond, t)
case syntax.InstEmptyWidth:
if syntax.EmptyOp(i.Arg)&^cond == 0 {
t = m.add(q, i.Out, pos, cap, cond, t)
}
case syntax.InstNop:
t = m.add(q, i.Out, pos, cap, cond, t)
case syntax.InstCapture:
if int(i.Arg) < len(cap) {
opos := cap[i.Arg]
cap[i.Arg] = pos
m.add(q, i.Out, pos, cap, cond, nil)
cap[i.Arg] = opos
} else {
t = m.add(q, i.Out, pos, cap, cond, t)
}
case syntax.InstMatch, syntax.InstRune, syntax.InstRune1, syntax.InstRuneAny, syntax.InstRuneAnyNotNL:
if t == nil {
t = m.alloc(i)
} else {
t.inst = i
}
if len(cap) > 0 && &t.cap[0] != &cap[0] {
copy(t.cap, cap)
}
d.t = t
t = nil
}
return t
}
// onepass runs the machine over the input starting at pos.
// It reports whether a match was found.
// If so, m.matchcap holds the submatch information.
func (m *machine) onepass(i input, pos int) bool {
startCond := m.re.cond
if startCond == ^syntax.EmptyOp(0) { // impossible
return false
}
m.matched = false
for i := range m.matchcap {
m.matchcap[i] = -1
}
r, r1 := endOfText, endOfText
width, width1 := 0, 0
r, width = i.step(pos)
if r != endOfText {
r1, width1 = i.step(pos + width)
}
var flag syntax.EmptyOp
if pos == 0 {
flag = syntax.EmptyOpContext(-1, r)
} else {
flag = i.context(pos)
}
pc := m.op.Start
inst := m.op.Inst[pc]
// If there is a simple literal prefix, skip over it.
if pos == 0 && syntax.EmptyOp(inst.Arg)&^flag == 0 &&
len(m.re.prefix) > 0 && i.canCheckPrefix() {
// Match requires literal prefix; fast search for it.
if i.hasPrefix(m.re) {
pos += len(m.re.prefix)
r, width = i.step(pos)
r1, width1 = i.step(pos + width)
flag = i.context(pos)
pc = int(m.re.prefixEnd)
} else {
return m.matched
}
}
for {
inst = m.op.Inst[pc]
pc = int(inst.Out)
switch inst.Op {
default:
panic("bad inst")
case syntax.InstMatch:
m.matched = true
if len(m.matchcap) > 0 {
m.matchcap[0] = 0
m.matchcap[1] = pos
}
return m.matched
case syntax.InstRune:
if !inst.MatchRune(r) {
return m.matched
}
case syntax.InstRune1:
if r != inst.Rune[0] {
return m.matched
}
case syntax.InstRuneAny:
// Nothing
case syntax.InstRuneAnyNotNL:
if r == '\n' {
return m.matched
}
// peek at the input rune to see which branch of the Alt to take
case syntax.InstAlt, syntax.InstAltMatch:
pc = int(onePassNext(&inst, r))
continue
case syntax.InstFail:
return m.matched
case syntax.InstNop:
continue
case syntax.InstEmptyWidth:
if syntax.EmptyOp(inst.Arg)&^flag != 0 {
return m.matched
}
continue
case syntax.InstCapture:
if int(inst.Arg) < len(m.matchcap) {
m.matchcap[inst.Arg] = pos
}
continue
}
if width == 0 {
break
}
flag = syntax.EmptyOpContext(r, r1)
pos += width
r, width = r1, width1
if r != endOfText {
r1, width1 = i.step(pos + width)
}
}
return m.matched
}
// empty is a non-nil 0-element slice,
// so doExecute can avoid an allocation
// when 0 captures are requested from a successful match.
var empty = make([]int, 0)
// doExecute finds the leftmost match in the input and returns
// the position of its subexpressions.
func (re *Regexp) doExecute(r io.RuneReader, b []byte, s string, pos int, ncap int) []int {
m := re.get()
var i input
if r != nil {
i = m.newInputReader(r)
} else if b != nil {
i = m.newInputBytes(b)
} else {
i = m.newInputString(s)
}
if m.op != notOnePass {
if !m.onepass(i, pos) {
re.put(m)
return nil
}
} else {
m.init(ncap)
if !m.match(i, pos) {
re.put(m)
return nil
}
}
if ncap == 0 {
re.put(m)
return empty // empty but not nil
}
cap := make([]int, len(m.matchcap))
copy(cap, m.matchcap)
re.put(m)
return cap
}
|