schema_init.c 103 KB
Newer Older
1
2
3
/* schema_init.c - init builtin schema */
/* $OpenLDAP$ */
/*
Kurt Zeilenga's avatar
Kurt Zeilenga committed
4
 * Copyright 1998-2000 The OpenLDAP Foundation, All Rights Reserved.
5
6
7
8
9
10
11
12
13
14
15
16
17
 * COPYING RESTRICTIONS APPLY, see COPYRIGHT file
 */

#include "portable.h"

#include <stdio.h>

#include <ac/ctype.h>
#include <ac/string.h>
#include <ac/socket.h>

#include "slap.h"
#include "ldap_pvt.h"
18

19
20
#include "ldap_utf8.h"

21
22
23
24
25
26
27
#include "lutil_hash.h"
/* We should replace MD5 with a faster hash */
#define HASH_BYTES				LUTIL_HASH_BYTES
#define HASH_CONTEXT			lutil_HASH_CTX
#define HASH_Init(c)			lutil_HASHInit(c)
#define HASH_Update(c,buf,len)	lutil_HASHUpdate(c,buf,len)
#define HASH_Final(d,c)			lutil_HASHFinal(d,c)
28

29
/* recycled validatation routines */
30
#define berValidate						blobValidate
31
32
33
34

/* unimplemented pretters */
#define dnPretty						NULL
#define integerPretty					NULL
35
36

/* recycled matching routines */
37
#define bitStringMatch					octetStringMatch
38
#define integerMatch					caseIgnoreIA5Match
39
40
41
#define numericStringMatch				caseIgnoreIA5Match
#define objectIdentifierMatch			caseIgnoreIA5Match
#define telephoneNumberMatch			caseIgnoreIA5Match
42
#define telephoneNumberSubstringsMatch	caseIgnoreIA5SubstringsMatch
43
44
#define generalizedTimeMatch			caseIgnoreIA5Match
#define generalizedTimeOrderingMatch	caseIgnoreIA5Match
45
#define uniqueMemberMatch				dnMatch
46

47
48
/* approx matching rules */
#define directoryStringApproxMatchOID	"1.3.6.1.4.1.4203.666.4.4"
Gary Williams's avatar
Gary Williams committed
49
50
51
#define directoryStringApproxMatch	approxMatch
#define directoryStringApproxIndexer	approxIndexer
#define directoryStringApproxFilter	approxFilter
52
#define IA5StringApproxMatchOID			"1.3.6.1.4.1.4203.666.4.5"
Gary Williams's avatar
Gary Williams committed
53
#define IA5StringApproxMatch			approxMatch
54
#define IA5StringApproxIndexer			approxIndexer
Gary Williams's avatar
Gary Williams committed
55
#define IA5StringApproxFilter			approxFilter
56

57
58
59
60
/* orderring matching rules */
#define caseIgnoreOrderingMatch			caseIgnoreMatch
#define caseExactOrderingMatch			caseExactMatch

61
/* unimplemented matching routines */
62
63
64
65
66
67
68
#define caseIgnoreListMatch				NULL
#define caseIgnoreListSubstringsMatch	NULL
#define protocolInformationMatch		NULL
#define integerFirstComponentMatch		NULL

#define OpenLDAPaciMatch				NULL
#define authPasswordMatch				NULL
69
70

/* recycled indexing/filtering routines */
71
72
#define dnIndexer						caseIgnoreIndexer
#define dnFilter						caseIgnoreFilter
73
74
#define integerIndexer					caseIgnoreIA5Indexer
#define integerFilter					caseIgnoreIA5Filter
75

76
77
78
79
80
#define telephoneNumberIndexer			caseIgnoreIA5Indexer
#define telephoneNumberFilter			caseIgnoreIA5Filter
#define telephoneNumberSubstringsIndexer	caseIgnoreIA5SubstringsIndexer
#define telephoneNumberSubstringsFilter		caseIgnoreIA5SubstringsFilter

81
82
83
84
static char *strcasechr( const char *str, int c )
{
	char *lower = strchr( str, TOLOWER(c) );
	char *upper = strchr( str, TOUPPER(c) );
85

86
87
88
89
90
91
92
93
	if( lower && upper ) {
		return lower < upper ? lower : upper;
	} else if ( lower ) {
		return lower;
	} else {
		return upper;
	}
}
94

95
96
97
static int
octetStringMatch(
	int *matchp,
98
	slap_mask_t flags,
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *value,
	void *assertedValue )
{
	int match = value->bv_len - ((struct berval *) assertedValue)->bv_len;

	if( match == 0 ) {
		match = memcmp( value->bv_val,
			((struct berval *) assertedValue)->bv_val,
			value->bv_len );
	}

	*matchp = match;
	return LDAP_SUCCESS;
}

/* Index generation function */
int octetStringIndexer(
118
119
	slap_mask_t use,
	slap_mask_t flags,
120
121
122
123
124
125
126
127
128
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *prefix,
	struct berval **values,
	struct berval ***keysp )
{
	int i;
	size_t slen, mlen;
	struct berval **keys;
129
	HASH_CONTEXT   HASHcontext;
Gary Williams's avatar
Gary Williams committed
130
	unsigned char	HASHdigest[HASH_BYTES];
131
	struct berval digest;
132
133
	digest.bv_val = HASHdigest;
	digest.bv_len = sizeof(HASHdigest);
134
135
136
137
138

	for( i=0; values[i] != NULL; i++ ) {
		/* just count them */
	}

Kurt Zeilenga's avatar
Kurt Zeilenga committed
139
140
141
	/* we should have at least one value at this point */
	assert( i > 0 );

142
143
144
145
146
147
	keys = ch_malloc( sizeof( struct berval * ) * (i+1) );

	slen = strlen( syntax->ssyn_oid );
	mlen = strlen( mr->smr_oid );

	for( i=0; values[i] != NULL; i++ ) {
148
		HASH_Init( &HASHcontext );
149
		if( prefix != NULL && prefix->bv_len > 0 ) {
150
			HASH_Update( &HASHcontext,
151
152
				prefix->bv_val, prefix->bv_len );
		}
153
		HASH_Update( &HASHcontext,
154
			syntax->ssyn_oid, slen );
155
		HASH_Update( &HASHcontext,
156
			mr->smr_oid, mlen );
157
		HASH_Update( &HASHcontext,
158
			values[i]->bv_val, values[i]->bv_len );
159
		HASH_Final( HASHdigest, &HASHcontext );
160
161
162
163
164
165
166
167
168
169
170
171
172

		keys[i] = ber_bvdup( &digest );
	}

	keys[i] = NULL;

	*keysp = keys;

	return LDAP_SUCCESS;
}

/* Index generation function */
int octetStringFilter(
173
174
	slap_mask_t use,
	slap_mask_t flags,
175
176
177
178
179
180
181
182
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *prefix,
	void * assertValue,
	struct berval ***keysp )
{
	size_t slen, mlen;
	struct berval **keys;
183
	HASH_CONTEXT   HASHcontext;
Gary Williams's avatar
Gary Williams committed
184
	unsigned char	HASHdigest[HASH_BYTES];
185
186
	struct berval *value = (struct berval *) assertValue;
	struct berval digest;
187
188
	digest.bv_val = HASHdigest;
	digest.bv_len = sizeof(HASHdigest);
189
190
191
192
193
194

	slen = strlen( syntax->ssyn_oid );
	mlen = strlen( mr->smr_oid );

	keys = ch_malloc( sizeof( struct berval * ) * 2 );

195
	HASH_Init( &HASHcontext );
196
	if( prefix != NULL && prefix->bv_len > 0 ) {
197
		HASH_Update( &HASHcontext,
198
199
			prefix->bv_val, prefix->bv_len );
	}
200
	HASH_Update( &HASHcontext,
201
		syntax->ssyn_oid, slen );
202
	HASH_Update( &HASHcontext,
203
		mr->smr_oid, mlen );
204
	HASH_Update( &HASHcontext,
205
		value->bv_val, value->bv_len );
206
	HASH_Final( HASHdigest, &HASHcontext );
207
208
209
210
211
212
213
214

	keys[0] = ber_bvdup( &digest );
	keys[1] = NULL;

	*keysp = keys;

	return LDAP_SUCCESS;
}
215

216
217
218
219
220
221
222
223
224
225
static int
dnValidate(
	Syntax *syntax,
	struct berval *in )
{
	int rc;
	char *dn;

	if( in->bv_len == 0 ) return LDAP_SUCCESS;

226
	dn = ch_strdup( in->bv_val );
227

228
	if( dn == NULL ) {
Kurt Zeilenga's avatar
Kurt Zeilenga committed
229
		return LDAP_INVALID_SYNTAX;
230

231
	} else if ( strlen( in->bv_val ) != in->bv_len ) {
Kurt Zeilenga's avatar
Kurt Zeilenga committed
232
		rc = LDAP_INVALID_SYNTAX;
233

234
	} else if ( dn_validate( dn ) == NULL ) {
Kurt Zeilenga's avatar
Kurt Zeilenga committed
235
		rc = LDAP_INVALID_SYNTAX;
236
237
238
239
240
241

	} else {
		rc = LDAP_SUCCESS;
	}

	ch_free( dn );
242
243
244
	return rc;
}

245
int
246
247
248
249
250
dnNormalize(
	Syntax *syntax,
	struct berval *val,
	struct berval **normalized )
{
251
	struct berval *out;
252

253
254
255
	if ( val->bv_len != 0 ) {
		char *dn;
		out = ber_bvstr( UTF8normalize( val->bv_val, UTF8_CASEFOLD ) );
Kurt Zeilenga's avatar
Kurt Zeilenga committed
256

257
258
		dn = dn_validate( out->bv_val );

259
		if( dn == NULL ) {
260
261
262
			ber_bvfree( out );
			return LDAP_INVALID_SYNTAX;
		}
263

264
265
266
267
		out->bv_val = dn;
		out->bv_len = strlen( dn );
	} else {
		out = ber_bvdup( val );
268
269
270
271
272
273
274
275
	}

	*normalized = out;
	return LDAP_SUCCESS;
}

static int
dnMatch(
276
	int *matchp,
277
	slap_mask_t flags,
278
279
280
281
282
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *value,
	void *assertedValue )
{
283
	int match;
284
285
	struct berval *asserted = (struct berval *) assertedValue;
	
286
287
288
	match = value->bv_len - asserted->bv_len;

	if( match == 0 ) {
289
#ifdef USE_DN_NORMALIZE
290
		match = strcmp( value->bv_val, asserted->bv_val );
291
#else
292
		match = strcasecmp( value->bv_val, asserted->bv_val );
293
#endif
294
295
	}

296
#ifdef NEW_LOGGING
Gary Williams's avatar
Gary Williams committed
297
	LDAP_LOG(( "schema", LDAP_LEVEL_ENTRY,
Kurt Zeilenga's avatar
Kurt Zeilenga committed
298
299
		"dnMatch: %d\n    %s\n    %s\n", match,
		value->bv_val, asserted->bv_val ));
300
#else
301
	Debug( LDAP_DEBUG_ARGS, "dnMatch %d\n\t\"%s\"\n\t\"%s\"\n",
Kurt Zeilenga's avatar
Kurt Zeilenga committed
302
		match, value->bv_val, asserted->bv_val );
303
304
#endif

305
306
307

	*matchp = match;
	return LDAP_SUCCESS;
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
static int
nameUIDValidate(
	Syntax *syntax,
	struct berval *in )
{
	int rc;
	struct berval *dn;

	if( in->bv_len == 0 ) return LDAP_SUCCESS;

	dn = ber_bvdup( in );

	if( dn->bv_val[dn->bv_len-1] == '\'' ) {
		/* assume presence of optional UID */
		ber_len_t i;

		for(i=dn->bv_len-2; i>2; i--) {
			if( dn->bv_val[i] != '0' &&	dn->bv_val[i] != '1' ) {
				break;
			}
		}
		if( dn->bv_val[i] != '\'' ) {
			return LDAP_INVALID_SYNTAX;
		}
		if( dn->bv_val[i-1] != 'B' ) {
			return LDAP_INVALID_SYNTAX;
		}
		if( dn->bv_val[i-2] != '#' ) {
			return LDAP_INVALID_SYNTAX;
		}

		/* trim the UID to allow use of dn_validate */
		dn->bv_val[i-2] = '\0';
	}

	rc = dn_validate( dn->bv_val ) == NULL
		? LDAP_INVALID_SYNTAX : LDAP_SUCCESS;

	ber_bvfree( dn );
	return rc;
}

static int
nameUIDNormalize(
	Syntax *syntax,
	struct berval *val,
	struct berval **normalized )
{
	struct berval *out = ber_bvdup( val );

	if( out->bv_len != 0 ) {
		char *dn;
		ber_len_t dnlen;
		char *uid = NULL;
		ber_len_t uidlen = 0;

		if( out->bv_val[out->bv_len-1] == '\'' ) {
			/* assume presence of optional UID */
			uid = strrchr( out->bv_val, '#' );

			if( uid == NULL ) {
				ber_bvfree( out );
				return LDAP_INVALID_SYNTAX;
			}

			uidlen = out->bv_len - (out->bv_val - uid);
			/* temporarily trim the UID */
			*uid = '\0';
		}

#ifdef USE_DN_NORMALIZE
		dn = dn_normalize( out->bv_val );
#else
		dn = dn_validate( out->bv_val );
#endif

		if( dn == NULL ) {
			ber_bvfree( out );
			return LDAP_INVALID_SYNTAX;
		}

		dnlen = strlen(dn);

		if( uidlen ) {
			/* restore the separator */
			*uid = '#';
			/* shift the UID */
			SAFEMEMCPY( &dn[dnlen], uid, uidlen );
		}

		out->bv_val = dn;
		out->bv_len = dnlen + uidlen;
	}

	*normalized = out;
	return LDAP_SUCCESS;
}

408
409
410
411
412
413
414
415
416
static int
inValidate(
	Syntax *syntax,
	struct berval *in )
{
	/* any value allowed */
	return LDAP_OTHER;
}

417
static int
418
blobValidate(
419
420
421
422
	Syntax *syntax,
	struct berval *in )
{
	/* any value allowed */
423
	return LDAP_SUCCESS;
424
425
}

426
427
428
429
430
431
432
433
434
435
436
437
438
static int
bitStringValidate(
	Syntax *syntax,
	struct berval *in )
{
	ber_len_t i;

	/* very unforgiving validation, requires no normalization
	 * before simplistic matching
	 */
	if( in->bv_len < 3 ) {
		return LDAP_INVALID_SYNTAX;
	}
439

440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
	if( in->bv_val[0] != 'B' ||
		in->bv_val[1] != '\'' ||
		in->bv_val[in->bv_len-1] != '\'' )
	{
		return LDAP_INVALID_SYNTAX;
	}

	for( i=in->bv_len-2; i>1; i-- ) {
		if( in->bv_val[i] != '0' && in->bv_val[i] != '1' ) {
			return LDAP_INVALID_SYNTAX;
		}
	}

	return LDAP_SUCCESS;
}

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
/*
 * Handling boolean syntax and matching is quite rigid.
 * A more flexible approach would be to allow a variety
 * of strings to be normalized and prettied into TRUE
 * and FALSE.
 */
static int
booleanValidate(
	Syntax *syntax,
	struct berval *in )
{
	/* very unforgiving validation, requires no normalization
	 * before simplistic matching
	 */

	if( in->bv_len == 4 ) {
		if( !memcmp( in->bv_val, "TRUE", 4 ) ) {
			return LDAP_SUCCESS;
		}
	} else if( in->bv_len == 5 ) {
		if( !memcmp( in->bv_val, "FALSE", 5 ) ) {
			return LDAP_SUCCESS;
		}
	}

	return LDAP_INVALID_SYNTAX;
}

static int
booleanMatch(
	int *matchp,
487
	slap_mask_t flags,
488
489
490
491
492
493
494
495
496
497
498
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *value,
	void *assertedValue )
{
	/* simplistic matching allowed by rigid validation */
	struct berval *asserted = (struct berval *) assertedValue;
	*matchp = value->bv_len != asserted->bv_len;
	return LDAP_SUCCESS;
}

499
500
501
502
503
504
505
506
507
static int
UTF8StringValidate(
	Syntax *syntax,
	struct berval *in )
{
	ber_len_t count;
	int len;
	unsigned char *u = in->bv_val;

508
509
	if( !in->bv_len ) return LDAP_INVALID_SYNTAX;

510
	for( count = in->bv_len; count > 0; count-=len, u+=len ) {
511
512
513
514
		/* get the length indicated by the first byte */
		len = LDAP_UTF8_CHARLEN( u );

		/* should not be zero */
515
		if( len == 0 ) return LDAP_INVALID_SYNTAX;
516
517
518

		/* make sure len corresponds with the offset
			to the next character */
519
		if( LDAP_UTF8_OFFSET( u ) != len ) return LDAP_INVALID_SYNTAX;
520
521
	}

522
	if( count != 0 ) return LDAP_INVALID_SYNTAX;
523

524
	return LDAP_SUCCESS;
525
526
527
528
529
530
531
532
533
}

static int
UTF8StringNormalize(
	Syntax *syntax,
	struct berval *val,
	struct berval **normalized )
{
	struct berval *newval;
534
	char *p, *q, *s;
535

536
	newval = ch_malloc( sizeof( struct berval ) );
537

538
	p = val->bv_val;
539

540
541
542
	/* Ignore initial whitespace */
	while ( ldap_utf8_isspace( p ) ) {
		LDAP_UTF8_INCR( p );
Kurt Zeilenga's avatar
Kurt Zeilenga committed
543
	}
544

545
546
	if( *p == '\0' ) {
		ch_free( newval );
Kurt Zeilenga's avatar
Kurt Zeilenga committed
547
548
		return LDAP_INVALID_SYNTAX;
	}
549

550
551
552
	newval->bv_val = ch_strdup( p );
	p = q = newval->bv_val;
	s = NULL;
553

554
555
	while ( *p ) {
		int len;
556

557
558
559
560
561
		if ( ldap_utf8_isspace( p ) ) {
			len = LDAP_UTF8_COPY(q,p);
			s=q;
			p+=len;
			q+=len;
562

563
564
565
566
			/* Ignore the extra whitespace */
			while ( ldap_utf8_isspace( p ) ) {
				LDAP_UTF8_INCR( p );
			}
Kurt Zeilenga's avatar
Kurt Zeilenga committed
567
		} else {
568
569
570
571
			len = LDAP_UTF8_COPY(q,p);
			s=NULL;
			p+=len;
			q+=len;
Kurt Zeilenga's avatar
Kurt Zeilenga committed
572
		}
573
574
575
576
	}

	assert( *newval->bv_val );
	assert( newval->bv_val < p );
Kurt Zeilenga's avatar
Kurt Zeilenga committed
577
	assert( q <= p );
578

579
580
581
582
583
584
585
586
587
588
589
	/* cannot start with a space */
	assert( !ldap_utf8_isspace(newval->bv_val) );

	/*
	 * If the string ended in space, backup the pointer one
	 * position.  One is enough because the above loop collapsed
	 * all whitespace to a single space.
	 */

	if ( s != NULL ) {
		q = s;
Kurt Zeilenga's avatar
Kurt Zeilenga committed
590
	}
591

592
593
594
595
596
597
598
599
	/* cannot end with a space */
	assert( !ldap_utf8_isspace( LDAP_UTF8_PREV(q) ) );

	/* null terminate */
	*q = '\0';

	newval->bv_len = q - newval->bv_val;
	*normalized = newval;
600

601
	return LDAP_SUCCESS;
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
/* Returns Unicode cannonically normalized copy of a substring assertion
 * Skipping attribute description */
SubstringsAssertion *
UTF8SubstringsassertionNormalize(
	SubstringsAssertion *sa,
	char casefold )
{
	SubstringsAssertion *nsa;
	int i;

	nsa = (SubstringsAssertion *)ch_calloc( 1, sizeof(SubstringsAssertion) );
	if( nsa == NULL ) {
		return NULL;
	}

	if( sa->sa_initial != NULL ) {
		nsa->sa_initial = ber_bvstr( UTF8normalize( sa->sa_initial->bv_val, casefold ) );
		if( nsa->sa_initial == NULL ) {
			goto err;
		}
	}

	if( sa->sa_any != NULL ) {
		for( i=0; sa->sa_any[i] != NULL; i++ ) {
			/* empty */
		}
		nsa->sa_any = (struct berval **)ch_malloc( (i + 1) * sizeof(struct berval *) );
		for( i=0; sa->sa_any[i] != NULL; i++ ) {
			nsa->sa_any[i] = ber_bvstr( UTF8normalize( sa->sa_any[i]->bv_val, casefold ) );
			if( nsa->sa_any[i] == NULL ) {
				goto err;
			}
		}
		nsa->sa_any[i] = NULL;
	}

	if( sa->sa_final != NULL ) {
		nsa->sa_final = ber_bvstr( UTF8normalize( sa->sa_final->bv_val, casefold ) );
		if( nsa->sa_final == NULL ) {
			goto err;
		}
	}

	return nsa;

err:
	ch_free( nsa->sa_final );
	ber_bvecfree( nsa->sa_any );
	ch_free( nsa->sa_initial );
	ch_free( nsa );
	return NULL;
}

657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
/* Strip characters with the 8th bit set */
char *
strip8bitChars(
	char *in )      
{
	char *p = in, *q;
  
	if( in == NULL ) {
		return NULL;
	}
	while( *p ) {
		if( *p & 0x80 ) {
			q = p;
			while( *++q & 0x80 ) {
				/* empty */
			}
			p = memmove(p, q, strlen(q) + 1);
		} else {
			p++;
		}
	}
	return in;
}

681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
#if defined(SLAPD_APPROX_MULTISTRING)

#if defined(SLAPD_APPROX_INITIALS)
#define SLAPD_APPROX_DELIMITER "._ "
#define SLAPD_APPROX_WORDLEN 2
#else
#define SLAPD_APPROX_DELIMITER " "
#define SLAPD_APPROX_WORDLEN 1
#endif

static int
approxMatch(
	int *matchp,
	slap_mask_t flags,
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *value,
	void *assertedValue )
{
	char *val, *assertv, **values, **words, *c;
	int i, count, len, nextchunk=0, nextavail=0;
702
	size_t avlen;
703

704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
	/* Yes, this is necessary */
	val = UTF8normalize( value->bv_val, UTF8_NOCASEFOLD );
	if( val == NULL ) {
		*matchp = 1;
		return LDAP_SUCCESS;
	}
	strip8bitChars( val );

	/* Yes, this is necessary */
	assertv = UTF8normalize( ((struct berval *)assertedValue)->bv_val,
				 UTF8_NOCASEFOLD );
	if( assertv == NULL ) {
		free( val );
		*matchp = 1;
		return LDAP_SUCCESS;
	}
	strip8bitChars( assertv );
	avlen = strlen( assertv );
722
723
724
725
726
727
728
729
730
731
732
733

	/* Isolate how many words there are */
	for( c=val,count=1; *c; c++ ) {
		c = strpbrk( c, SLAPD_APPROX_DELIMITER );
		if ( c == NULL ) break;
		*c = '\0';
		count++;
	}

	/* Get a phonetic copy of each word */
	words = (char **)ch_malloc( count * sizeof(char *) );
	values = (char **)ch_malloc( count * sizeof(char *) );
Gary Williams's avatar
Gary Williams committed
734
	for( c=val,i=0;	 i<count;  i++,c+=strlen(c)+1 ) {
735
736
737
738
		words[i] = c;
		values[i] = phonetic(c);
	}

739
	/* Work through the asserted value's words, to see if at least some
740
741
	   of the words are there, in the same order. */
	len = 0;
742
	while ( nextchunk < avlen ) {
743
		len = strcspn( assertv + nextchunk, SLAPD_APPROX_DELIMITER);
744
745
746
747
		if( len == 0 ) {
			nextchunk++;
			continue;
		}
748
#if defined(SLAPD_APPROX_INITIALS)
749
		else if( len == 1 ) {
750
751
			/* Single letter words need to at least match one word's initial */
			for( i=nextavail; i<count; i++ )
752
753
				if( !strncasecmp( assertv+nextchunk, words[i], 1 )) {
					nextavail=i+1;
754
					break;
755
				}
756
757
		}
#endif
758
		else {
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
			/* Isolate the next word in the asserted value and phonetic it */
			assertv[nextchunk+len] = '\0';
			val = phonetic( assertv + nextchunk );

			/* See if this phonetic chunk is in the remaining words of *value */
			for( i=nextavail; i<count; i++ ){
				if( !strcmp( val, values[i] ) ){
					nextavail = i+1;
					break;
				}
			}
		}

		/* This chunk in the asserted value was NOT within the *value. */
		if( i >= count ) {
			nextavail=-1;
			break;
		}

		/* Go on to the next word in the asserted value */
		nextchunk += len+1;
	}

	/* If some of the words were seen, call it a match */
	if( nextavail > 0 ) {
		*matchp = 0;
	}
	else {
		*matchp = 1;
	}

	/* Cleanup allocs */
791
	free( assertv );
792
793
794
795
796
	for( i=0; i<count; i++ ) {
		ch_free( values[i] );
	}
	ch_free( values );
	ch_free( words );
797
	free( val );
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816

	return LDAP_SUCCESS;
}

int 
approxIndexer(
	slap_mask_t use,
	slap_mask_t flags,
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *prefix,
	struct berval **values,
	struct berval ***keysp )
{
	char *val, *c;
	int i,j, len, wordcount, keycount=0;
	struct berval **newkeys, **keys=NULL;

	for( j=0; values[j] != NULL; j++ ) {
817
818
819
820
		/* Yes, this is necessary */
		val = UTF8normalize( values[j]->bv_val, UTF8_NOCASEFOLD );
		strip8bitChars( val );

821
		/* Isolate how many words there are. There will be a key for each */
Gary Williams's avatar
Gary Williams committed
822
		for( wordcount=0,c=val;	 *c;  c++) {
823
824
825
826
827
828
829
830
831
			len = strcspn(c, SLAPD_APPROX_DELIMITER);
			if( len >= SLAPD_APPROX_WORDLEN ) wordcount++;
			c+= len;
			if (*c == '\0') break;
			*c = '\0';
		}

		/* Allocate/increase storage to account for new keys */
		newkeys = (struct berval **)ch_malloc( (keycount + wordcount + 1) 
Kurt Zeilenga's avatar
Kurt Zeilenga committed
832
			* sizeof(struct berval *) );
833
834
835
836
837
		memcpy( newkeys, keys, keycount * sizeof(struct berval *) );
		if( keys ) ch_free( keys );
		keys = newkeys;

		/* Get a phonetic copy of each word */
Gary Williams's avatar
Gary Williams committed
838
		for( c=val,i=0;	 i<wordcount;  c+=len+1	 ) {
839
840
841
842
843
844
845
846
847
			len = strlen( c );
			if( len < SLAPD_APPROX_WORDLEN ) continue;
			keys[keycount] = (struct berval *)ch_malloc( sizeof(struct berval) );
			keys[keycount]->bv_val = phonetic( c );
			keys[keycount]->bv_len = strlen( keys[keycount]->bv_val );
			keycount++;
			i++;
		}

848
		free( val );
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
	}
	keys[keycount] = NULL;
	*keysp = keys;

	return LDAP_SUCCESS;
}

int 
approxFilter(
	slap_mask_t use,
	slap_mask_t flags,
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *prefix,
	void * assertValue,
	struct berval ***keysp )
{
	char *val, *c;
	int i, count, len;
	struct berval **keys;

870
871
872
873
874
875
876
877
878
879
880
	/* Yes, this is necessary */
	val = UTF8normalize( ((struct berval *)assertValue)->bv_val,
			     UTF8_NOCASEFOLD );
	if( val == NULL ) {
		keys = (struct berval **)ch_malloc( sizeof(struct berval *) );
		keys[0] = NULL;
		*keysp = keys;
		return LDAP_SUCCESS;
	}
	strip8bitChars( val );

881
882
883
884
885
886
887
888
889
890
891
892
893
	/* Isolate how many words there are. There will be a key for each */
	for( count=0,c=val;  *c;  c++) {
		len = strcspn(c, SLAPD_APPROX_DELIMITER);
		if( len >= SLAPD_APPROX_WORDLEN ) count++;
		c+= len;
		if (*c == '\0') break;
		*c = '\0';
	}

	/* Allocate storage for new keys */
	keys = (struct berval **)ch_malloc( (count + 1) * sizeof(struct berval *) );

	/* Get a phonetic copy of each word */
Gary Williams's avatar
Gary Williams committed
894
	for( c=val,i=0;	 i<count; c+=len+1 ) {
895
896
		len = strlen(c);
		if( len < SLAPD_APPROX_WORDLEN ) continue;
897
		keys[i] = ber_bvstr( phonetic( c ) );
898
899
900
		i++;
	}

901
	free( val );
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922

	keys[count] = NULL;
	*keysp = keys;

	return LDAP_SUCCESS;
}


#else
/* No other form of Approximate Matching is defined */

static int
approxMatch(
	int *matchp,
	slap_mask_t flags,
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *value,
	void *assertedValue )
{
	char *vapprox, *avapprox;
923
	char *s, *t;
924

925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
	/* Yes, this is necessary */
	s = UTF8normalize( value->bv_val, UTF8_NOCASEFOLD );
	if( s == NULL ) {
		*matchp = 1;
		return LDAP_SUCCESS;
	}

	/* Yes, this is necessary */
	t = UTF8normalize( ((struct berval *)assertedValue)->bv_val,
			   UTF8_NOCASEFOLD );
	if( t == NULL ) {
		free( s );
		*matchp = -1;
		return LDAP_SUCCESS;
	}

	vapprox = phonetic( strip8bitChars( s ) );
	avapprox = phonetic( strip8bitChars( t ) );

	free( s );
	free( t );
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966

	*matchp = strcmp( vapprox, avapprox );

	ch_free( vapprox );
	ch_free( avapprox );

	return LDAP_SUCCESS;
}

int 
approxIndexer(
	slap_mask_t use,
	slap_mask_t flags,
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *prefix,
	struct berval **values,
	struct berval ***keysp )
{
	int i;
	struct berval **keys;
967
	char *s;
968
969

	for( i=0; values[i] != NULL; i++ ) {
Kurt Zeilenga's avatar
Kurt Zeilenga committed
970
		/* empty - just count them */
971
	}
Kurt Zeilenga's avatar
Kurt Zeilenga committed
972
973

	/* we should have at least one value at this point */
974
975
976
977
978
979
	assert( i > 0 );

	keys = (struct berval **)ch_malloc( sizeof( struct berval * ) * (i+1) );

	/* Copy each value and run it through phonetic() */
	for( i=0; values[i] != NULL; i++ ) {
980
981
982
983
984
985
		/* Yes, this is necessary */
		s = UTF8normalize( values[i]->bv_val, UTF8_NOCASEFOLD );

		/* strip 8-bit chars and run through phonetic() */
		keys[i] = ber_bvstr( phonetic( strip8bitChars( s ) ) );
		free( s );
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
	}
	keys[i] = NULL;

	*keysp = keys;
	return LDAP_SUCCESS;
}


int 
approxFilter(
	slap_mask_t use,
	slap_mask_t flags,
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *prefix,
	void * assertValue,
	struct berval ***keysp )
{
	struct berval **keys;
1005
	char *s;
1006
1007
1008

	keys = (struct berval **)ch_malloc( sizeof( struct berval * ) * 2 );

1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
	/* Yes, this is necessary */
	s = UTF8normalize( ((struct berval *)assertValue)->bv_val,
			     UTF8_NOCASEFOLD );
	if( s == NULL ) {
		keys[0] = NULL;
	} else {
		/* strip 8-bit chars and run through phonetic() */
		keys[0] = ber_bvstr( phonetic( strip8bitChars( s ) ) );
		free( s );
		keys[1] = NULL;
	}
1020
1021
1022
1023
1024
1025
1026

	*keysp = keys;
	return LDAP_SUCCESS;
}
#endif


1027
static int
1028
caseExactMatch(
1029
	int *matchp,
1030
	slap_mask_t flags,
1031
1032
1033
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *value,
1034
	void *assertedValue )
1035
{
1036
1037
1038
	*matchp = UTF8normcmp( value->bv_val,
		((struct berval *) assertedValue)->bv_val,
		UTF8_NOCASEFOLD );
1039
	return LDAP_SUCCESS;
1040
1041
}

1042
static int
1043
caseExactSubstringsMatch(
1044
	int *matchp,
1045
	slap_mask_t flags,
1046
1047
1048
1049
1050
1051
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *value,
	void *assertedValue )
{
	int match = 0;
1052
1053
	SubstringsAssertion *sub;
	struct berval left;
1054
1055
	int i;
	ber_len_t inlen=0;
1056
1057
1058
1059
1060
1061
1062
1063
1064
	char *nav;

	nav = UTF8normalize( value->bv_val, UTF8_NOCASEFOLD );
	if( nav == NULL ) {
		match = 1;
		goto done;
	}
	left.bv_val = nav;
	left.bv_len = strlen( nav );
Kurt Zeilenga's avatar
Kurt Zeilenga committed
1065

1066
1067
1068
1069
1070
	sub = UTF8SubstringsassertionNormalize( assertedValue, UTF8_NOCASEFOLD );
	if( sub == NULL ) {
		match = -1;
		goto done;
	}
1071

1072
	/* Add up asserted input length */
1073
1074
1075
1076
	if( sub->sa_initial ) {
		inlen += sub->sa_initial->bv_len;
	}
	if( sub->sa_any ) {
1077
1078
		for(i=0; sub->sa_any[i] != NULL; i++) {
			inlen += sub->sa_any[i]->bv_len;
1079
1080
1081
1082
1083
1084
1085
		}
	}
	if( sub->sa_final ) {
		inlen += sub->sa_final->bv_len;
	}

	if( sub->sa_initial ) {
Kurt Zeilenga's avatar
Kurt Zeilenga committed
1086
1087
1088
1089
1090
		if( inlen > left.bv_len ) {
			match = 1;
			goto done;
		}

1091
1092
		match = strncmp( sub->sa_initial->bv_val, left.bv_val,
			sub->sa_initial->bv_len );
1093
1094
1095
1096
1097
1098
1099

		if( match != 0 ) {
			goto done;
		}

		left.bv_val += sub->sa_initial->bv_len;
		left.bv_len -= sub->sa_initial->bv_len;
1100
		inlen -= sub->sa_initial->bv_len;
Kurt Zeilenga's avatar
Kurt Zeilenga committed
1101
	}
1102

Kurt Zeilenga's avatar
Kurt Zeilenga committed
1103
	if( sub->sa_final ) {
1104
1105
1106
1107
		if( inlen > left.bv_len ) {
			match = 1;
			goto done;
		}
1108

1109
1110
1111
		match = strncmp( sub->sa_final->bv_val,
			&left.bv_val[left.bv_len - sub->sa_final->bv_len],
			sub->sa_final->bv_len );
1112
1113
1114
1115
1116
1117

		if( match != 0 ) {
			goto done;
		}

		left.bv_len -= sub->sa_final->bv_len;
1118
		inlen -= sub->sa_final->bv_len;
1119
1120
1121
	}

	if( sub->sa_any ) {
1122
1123
1124
1125
1126
		for(i=0; sub->sa_any[i]; i++) {
			ber_len_t idx;
			char *p;

retry:
Kurt Zeilenga's avatar
Kurt Zeilenga committed
1127
1128
1129
1130
1131
1132
			if( inlen > left.bv_len ) {
				/* not enough length */
				match = 1;
				goto done;
			}

1133
1134
1135
1136
			if( sub->sa_any[i]->bv_len == 0 ) {
				continue;
			}

1137
			p = strchr( left.bv_val, *sub->sa_any[i]->bv_val );
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160

			if( p == NULL ) {
				match = 1;
				goto done;
			}

			idx = p - left.bv_val;
			assert( idx < left.bv_len );

			if( idx >= left.bv_len ) {
				/* this shouldn't happen */
				return LDAP_OTHER;
			}

			left.bv_val = p;
			left.bv_len -= idx;

			if( sub->sa_any[i]->bv_len > left.bv_len ) {
				/* not enough left */
				match = 1;
				goto done;
			}

1161
1162
1163
			match = strncmp( left.bv_val,
				sub->sa_any[i]->bv_val,
				sub->sa_any[i]->bv_len );
1164
1165

			if( match != 0 ) {
1166
1167
				left.bv_val++;
				left.bv_len--;
1168
1169
1170
1171
1172
				goto retry;
			}

			left.bv_val += sub->sa_any[i]->bv_len;
			left.bv_len -= sub->sa_any[i]->bv_len;
1173
			inlen -= sub->sa_any[i]->bv_len;
1174
		}
1175
1176
1177
	}

done:
1178
1179
1180
1181
1182
1183
1184
	free( nav );
	if( sub != NULL ) {
		ch_free( sub->sa_final );
		ber_bvecfree( sub->sa_any );
		ch_free( sub->sa_initial );
		ch_free( sub );
	}
1185
1186
1187
1188
	*matchp = match;
	return LDAP_SUCCESS;
}

1189
/* Index generation function */
1190
int caseExactIndexer(
1191
1192
	slap_mask_t use,
	slap_mask_t flags,
1193
1194
1195
1196
1197
1198
1199
1200
1201
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *prefix,
	struct berval **values,
	struct berval ***keysp )
{
	int i;
	size_t slen, mlen;
	struct berval **keys;
1202
	HASH_CONTEXT   HASHcontext;
Gary Williams's avatar
Gary Williams committed
1203
	unsigned char	HASHdigest[HASH_BYTES];
1204
	struct berval digest;
1205
1206
	digest.bv_val = HASHdigest;
	digest.bv_len = sizeof(HASHdigest);
1207
1208

	for( i=0; values[i] != NULL; i++ ) {
Kurt Zeilenga's avatar
Kurt Zeilenga committed
1209
		/* empty - just count them */
1210
1211
	}

Kurt Zeilenga's avatar
Kurt Zeilenga committed
1212
1213
1214
	/* we should have at least one value at this point */
	assert( i > 0 );

1215
1216
1217
1218
1219
1220
	keys = ch_malloc( sizeof( struct berval * ) * (i+1) );

	slen = strlen( syntax->ssyn_oid );
	mlen = strlen( mr->smr_oid );

	for( i=0; values[i] != NULL; i++ ) {
1221
		struct berval *value;
Kurt Zeilenga's avatar
Kurt Zeilenga committed
1222
1223
		value = ber_bvstr( UTF8normalize( values[i]->bv_val,
			UTF8_NOCASEFOLD ) );
1224

1225
		HASH_Init( &HASHcontext );
1226
		if( prefix != NULL && prefix->bv_len > 0 ) {
1227
			HASH_Update( &HASHcontext,
1228
1229
				prefix->bv_val, prefix->bv_len );
		}
1230
		HASH_Update( &HASHcontext,
1231
			syntax->ssyn_oid, slen );
1232
		HASH_Update( &HASHcontext,
1233
			mr->smr_oid, mlen );
1234
		HASH_Update( &HASHcontext,
1235
			value->bv_val, value->bv_len );
1236
		HASH_Final( HASHdigest, &HASHcontext );
1237

1238
1239
		ber_bvfree( value );

1240
1241
1242
1243
1244
1245
1246
1247
1248
		keys[i] = ber_bvdup( &digest );
	}

	keys[i] = NULL;
	*keysp = keys;
	return LDAP_SUCCESS;
}

/* Index generation function */
1249
int caseExactFilter(
1250
1251
	slap_mask_t use,
	slap_mask_t flags,
1252
1253
1254
1255
1256
1257
1258
1259
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *prefix,
	void * assertValue,
	struct berval ***keysp )
{
	size_t slen, mlen;
	struct berval **keys;
1260
	HASH_CONTEXT   HASHcontext;
Gary Williams's avatar
Gary Williams committed
1261
	unsigned char	HASHdigest[HASH_BYTES];
1262
1263
	struct berval *value;
	struct berval digest;
1264
1265
	digest.bv_val = HASHdigest;
	digest.bv_len = sizeof(HASHdigest);
1266
1267
1268
1269

	slen = strlen( syntax->ssyn_oid );
	mlen = strlen( mr->smr_oid );

1270
1271
1272
1273
1274
1275
1276
1277
	value = ber_bvstr( UTF8normalize( ((struct berval *) assertValue)->bv_val,
		UTF8_NOCASEFOLD ) );
	/* This usually happens if filter contains bad UTF8 */
	if( value == NULL ) {
		keys = ch_malloc( sizeof( struct berval * ) );
		keys[0] = NULL;
		return LDAP_SUCCESS;
	}
1278
1279
1280

	keys = ch_malloc( sizeof( struct berval * ) * 2 );

1281
	HASH_Init( &HASHcontext );
1282
	if( prefix != NULL && prefix->bv_len > 0 ) {
1283
		HASH_Update( &HASHcontext,
1284
1285
			prefix->bv_val, prefix->bv_len );
	}
1286
	HASH_Update( &HASHcontext,
1287
		syntax->ssyn_oid, slen );
1288
	HASH_Update( &HASHcontext,
1289
		mr->smr_oid, mlen );
1290
	HASH_Update( &HASHcontext,
1291
		value->bv_val, value->bv_len );
1292
	HASH_Final( HASHdigest, &HASHcontext );
1293
1294
1295
1296

	keys[0] = ber_bvdup( &digest );
	keys[1] = NULL;

1297
1298
	ber_bvfree( value );

1299
1300
1301
	*keysp = keys;
	return LDAP_SUCCESS;
}
1302

1303
/* Substrings Index generation function */
1304
int caseExactSubstringsIndexer(
1305
1306
	slap_mask_t use,
	slap_mask_t flags,
1307
1308
1309
1310
1311
1312
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *prefix,
	struct berval **values,
	struct berval ***keysp )
{
1313
	ber_len_t i, nkeys;
1314
1315
	size_t slen, mlen;
	struct berval **keys;
Kurt Zeilenga's avatar
Kurt Zeilenga committed
1316
	struct berval **nvalues;
1317

1318
	HASH_CONTEXT   HASHcontext;
Gary Williams's avatar
Gary Williams committed
1319
	unsigned char	HASHdigest[HASH_BYTES];
1320
	struct berval digest;
1321
1322
	digest.bv_val = HASHdigest;
	digest.bv_len = sizeof(HASHdigest);
1323
1324

	nkeys=0;
1325

Kurt Zeilenga's avatar
Kurt Zeilenga committed
1326
	for( i=0; values[i] != NULL; i++ ) {
Kurt Zeilenga's avatar
Kurt Zeilenga committed
1327
		/* empty - just count them */
Kurt Zeilenga's avatar
Kurt Zeilenga committed
1328
	}
Kurt Zeilenga's avatar
Kurt Zeilenga committed
1329
1330