« back to all changes in this revision
Viewing changes to dfa.c
- browse files at revision 408.9.52
- compare with another revision
- compare with revision 28
- download diff
- download tarball
- view history from revision 408.9.52
- Committer: Arnold D. Robbins
- Date: 2014-10-28 18:33:36 UTC
- mfrom: (408.5.317)
- mto: (408.27.1) (615.1.1)
- mto: This revision was merged to the branch mainline in revision 531.
- Revision ID: git-v1:0182312b0fc945a20a3d7aeac1488540b5518e3a
Merge branch 'gawk-4.1-stable'
- files modified:
- ChangeLog
added
removed
dfa.c
432
432
slots so far, not counting trans[-1]. */
433
433
int trcount; /* Number of transition tables that have
434
434
actually been built. */
435
int min_trcount; /* Minimum of number of transition tables.
436
Always keep the number, even after freeing
437
the transition tables. It is also the
438
number of initial states. */
435
439
state_num **trans; /* Transition tables for states that can
436
440
never accept. If the transitions for a
437
441
state have not yet been computed, or the
450
454
newline is stored separately and handled
451
455
as a special case. Newline is also used
452
456
as a sentinel at the end of the buffer. */
457
state_num initstate_letter; /* Initial state for letter context. */
458
state_num initstate_others; /* Initial state for other contexts. */
453
459
struct dfamust *musts; /* List of strings, at least one of which
454
460
is known to appear in any r.e. matching
455
461
the dfa. */
2588
2594
2589
2595
/* Build the initial state. */
2590
2596
separate_contexts = state_separate_contexts (&merged);
2591
state_index (d, &merged,
2592
(separate_contexts & CTX_NEWLINE
2593
? CTX_NEWLINE : separate_contexts ^ CTX_ANY));
2597
if (separate_contexts & CTX_NEWLINE)
2598
state_index (d, &merged, CTX_NEWLINE);
2599
d->initstate_others = d->min_trcount
2600
= state_index (d, &merged, separate_contexts ^ CTX_ANY);
2601
if (separate_contexts & CTX_LETTER)
2602
d->initstate_letter = d->min_trcount
2603
= state_index (d, &merged, CTX_LETTER);
2604
else
2605
d->initstate_letter = d->initstate_others;
2606
d->min_trcount++;
2594
2607
2595
2608
free (posalloc);
2596
2609
free (stkalloc);
2927
2940
/* Set an upper limit on the number of transition tables that will ever
2928
2941
exist at once. 1024 is arbitrary. The idea is that the frequently
2929
2942
used transition tables will be quickly rebuilt, whereas the ones that
2930
were only needed once or twice will be cleared away. However, do
2931
not clear the initial state, as it's always used. */
2943
were only needed once or twice will be cleared away. However, do not
2944
clear the initial D->min_trcount states, since they are always used. */
2932
2945
if (d->trcount >= 1024)
2933
2946
{
2934
for (i = 1; i < d->tralloc; ++i)
2947
for (i = d->min_trcount; i < d->tralloc; ++i)
2935
2948
{
2936
2949
free (d->trans[i]);
2937
2950
free (d->fails[i]);
2938
2951
d->trans[i] = d->fails[i] = NULL;
2939
2952
}
2940
d->trcount = 1;
2953
d->trcount = d->min_trcount;
2941
2954
}
2942
2955
2943
2956
++d->trcount;
3308
3321
expression "\\" accepts the codepoint 0x5c, but should not accept the second
3309
3322
byte of the codepoint 0x815c. Then the initial state must skip the bytes
3310
3323
that are not a single byte character nor the first byte of a multibyte
3311
character. */
3324
character.
3325
3326
Given DFA state d, use mbs_to_wchar to advance MBP until it reaches or
3327
exceeds P. If WCP is non-NULL, set *WCP to the final wide character
3328
processed, or if no wide character is processed, set it to WEOF.
3329
Both P and MBP must be no larger than END. */
3312
3330
static unsigned char const *
3313
3331
skip_remains_mb (struct dfa *d, unsigned char const *p,
3314
unsigned char const *mbp, char const *end)
3332
unsigned char const *mbp, char const *end, wint_t *wcp)
3315
3333
{
3316
wint_t wc;
3334
wint_t wc = WEOF;
3317
3335
while (mbp < p)
3318
3336
mbp += mbs_to_wchar (&wc, (char const *) mbp,
3319
3337
end - (char const *) mbp, d);
3338
if (wcp != NULL)
3339
*wcp = wc;
3320
3340
return mbp;
3321
3341
}
3322
3342
3378
3398
{
3379
3399
s1 = s;
3380
3400
3381
if (s == 0)
3401
if (s < d->min_trcount)
3382
3402
{
3383
if (d->states[s].mbps.nelem == 0)
3403
if (d->min_trcount == 1)
3384
3404
{
3385
do
3405
if (d->states[s].mbps.nelem == 0)
3386
3406
{
3387
while (t[*p] == 0)
3388
p++;
3389
p = mbp = skip_remains_mb (d, p, mbp, end);
3407
do
3408
{
3409
while (t[*p] == 0)
3410
p++;
3411
p = mbp = skip_remains_mb (d, p, mbp, end, NULL);
3412
}
3413
while (t[*p] == 0);
3390
3414
}
3391
while (t[*p] == 0);
3415
else
3416
p = mbp = skip_remains_mb (d, p, mbp, end, NULL);
3392
3417
}
3393
3418
else
3394
p = mbp = skip_remains_mb (d, p, mbp, end);
3419
{
3420
wint_t wc;
3421
mbp = skip_remains_mb (d, p, mbp, end, &wc);
3422
3423
/* If d->min_trcount is greater than 1, maybe
3424
transit to another initial state after skip. */
3425
if (p < mbp)
3426
{
3427
int context = wchar_context (wc);
3428
if (context == CTX_LETTER)
3429
s = d->initstate_letter;
3430
else
3431
/* It's CTX_NONE. CTX_NEWLINE cannot happen,
3432
as we assume that a newline is always a
3433
single byte character. */
3434
s = d->initstate_others;
3435
p = mbp;
3436
s1 = s;
3437
}
3438
}
3395
3439
}
3396
3440
3397
3441
if (d->states[s].mbps.nelem == 0)
Loggerhead is a web-based interface for Breezy
Version: 2.0.1