schema_init.c 111 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
 * COPYING RESTRICTIONS APPLY, see COPYRIGHT file
 */

#include "portable.h"

#include <stdio.h>
Kurt Zeilenga's avatar
Kurt Zeilenga committed
11
#include <limits.h>
12
13

#include <ac/ctype.h>
14
#include <ac/errno.h>
15
16
17
18
19
#include <ac/string.h>
#include <ac/socket.h>

#include "slap.h"
#include "ldap_pvt.h"
20

21
22
#include "ldap_utf8.h"

23
24
25
26
27
28
29
#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)
30

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

/* unimplemented pretters */
35
#ifndef USE_LDAP_DN_PARSING
36
#define dnPretty						NULL
37
#endif /* !USE_LDAP_DN_PARSING */
38
#define integerPretty					NULL
39
40

/* recycled matching routines */
41
#define bitStringMatch					octetStringMatch
42
43
44
#define numericStringMatch				caseIgnoreIA5Match
#define objectIdentifierMatch			caseIgnoreIA5Match
#define telephoneNumberMatch			caseIgnoreIA5Match
45
#define telephoneNumberSubstringsMatch	caseIgnoreIA5SubstringsMatch
46
47
#define generalizedTimeMatch			caseIgnoreIA5Match
#define generalizedTimeOrderingMatch	caseIgnoreIA5Match
48
#define uniqueMemberMatch				dnMatch
49

50
51
/* approx matching rules */
#define directoryStringApproxMatchOID	"1.3.6.1.4.1.4203.666.4.4"
Gary Williams's avatar
Gary Williams committed
52
53
54
#define directoryStringApproxMatch	approxMatch
#define directoryStringApproxIndexer	approxIndexer
#define directoryStringApproxFilter	approxFilter
55
#define IA5StringApproxMatchOID			"1.3.6.1.4.1.4203.666.4.5"
Gary Williams's avatar
Gary Williams committed
56
#define IA5StringApproxMatch			approxMatch
57
#define IA5StringApproxIndexer			approxIndexer
Gary Williams's avatar
Gary Williams committed
58
#define IA5StringApproxFilter			approxFilter
59

60
61
62
63
/* orderring matching rules */
#define caseIgnoreOrderingMatch			caseIgnoreMatch
#define caseExactOrderingMatch			caseExactMatch

64
/* unimplemented matching routines */
65
66
67
68
69
70
71
#define caseIgnoreListMatch				NULL
#define caseIgnoreListSubstringsMatch	NULL
#define protocolInformationMatch		NULL
#define integerFirstComponentMatch		NULL

#define OpenLDAPaciMatch				NULL
#define authPasswordMatch				NULL
72
73

/* recycled indexing/filtering routines */
74
75
#define dnIndexer				caseExactIgnoreIndexer
#define dnFilter				caseExactIgnoreFilter
76
77
#define bitStringFilter			octetStringFilter
#define bitStringIndexer		octetStringIndexer
78

79
80
81
82
83
#define telephoneNumberIndexer			caseIgnoreIA5Indexer
#define telephoneNumberFilter			caseIgnoreIA5Filter
#define telephoneNumberSubstringsIndexer	caseIgnoreIA5SubstringsIndexer
#define telephoneNumberSubstringsFilter		caseIgnoreIA5SubstringsFilter

84
85
86
87
/* must match OIDs below */
#define caseExactMatchOID			"2.5.13.5"
#define caseExactSubstringsMatchOID		"2.5.13.7"

88
89
90
91
static char *strcasechr( const char *str, int c )
{
	char *lower = strchr( str, TOLOWER(c) );
	char *upper = strchr( str, TOUPPER(c) );
92

93
94
95
96
97
98
99
100
	if( lower && upper ) {
		return lower < upper ? lower : upper;
	} else if ( lower ) {
		return lower;
	} else {
		return upper;
	}
}
101

102
103
104
static int
octetStringMatch(
	int *matchp,
105
	slap_mask_t flags,
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
	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(
125
126
	slap_mask_t use,
	slap_mask_t flags,
127
128
129
130
131
132
133
134
135
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *prefix,
	struct berval **values,
	struct berval ***keysp )
{
	int i;
	size_t slen, mlen;
	struct berval **keys;
136
	HASH_CONTEXT   HASHcontext;
Gary Williams's avatar
Gary Williams committed
137
	unsigned char	HASHdigest[HASH_BYTES];
138
	struct berval digest;
139
140
	digest.bv_val = HASHdigest;
	digest.bv_len = sizeof(HASHdigest);
141
142
143
144
145

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

Kurt Zeilenga's avatar
Kurt Zeilenga committed
146
147
148
	/* we should have at least one value at this point */
	assert( i > 0 );

149
150
151
152
153
154
	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++ ) {
155
		HASH_Init( &HASHcontext );
156
		if( prefix != NULL && prefix->bv_len > 0 ) {
157
			HASH_Update( &HASHcontext,
158
159
				prefix->bv_val, prefix->bv_len );
		}
160
		HASH_Update( &HASHcontext,
161
			syntax->ssyn_oid, slen );
162
		HASH_Update( &HASHcontext,
163
			mr->smr_oid, mlen );
164
		HASH_Update( &HASHcontext,
165
			values[i]->bv_val, values[i]->bv_len );
166
		HASH_Final( HASHdigest, &HASHcontext );
167
168
169
170
171
172
173
174
175
176
177
178
179

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

	keys[i] = NULL;

	*keysp = keys;

	return LDAP_SUCCESS;
}

/* Index generation function */
int octetStringFilter(
180
181
	slap_mask_t use,
	slap_mask_t flags,
182
183
184
185
186
187
188
189
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *prefix,
	void * assertValue,
	struct berval ***keysp )
{
	size_t slen, mlen;
	struct berval **keys;
190
	HASH_CONTEXT   HASHcontext;
Gary Williams's avatar
Gary Williams committed
191
	unsigned char	HASHdigest[HASH_BYTES];
192
193
	struct berval *value = (struct berval *) assertValue;
	struct berval digest;
194
195
	digest.bv_val = HASHdigest;
	digest.bv_len = sizeof(HASHdigest);
196
197
198
199
200
201

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

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

202
	HASH_Init( &HASHcontext );
203
	if( prefix != NULL && prefix->bv_len > 0 ) {
204
		HASH_Update( &HASHcontext,
205
206
			prefix->bv_val, prefix->bv_len );
	}
207
	HASH_Update( &HASHcontext,
208
		syntax->ssyn_oid, slen );
209
	HASH_Update( &HASHcontext,
210
		mr->smr_oid, mlen );
211
	HASH_Update( &HASHcontext,
212
		value->bv_val, value->bv_len );
213
	HASH_Final( HASHdigest, &HASHcontext );
214
215
216
217
218
219
220
221

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

	*keysp = keys;

	return LDAP_SUCCESS;
}
222

223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
#ifdef USE_LDAP_DN_PARSING
static int
dnValidate(
	Syntax *syntax,
	struct berval *in )
{
	int		rc;
	LDAPDN		*dn = NULL;

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

	rc = ldap_str2dn( in->bv_val, &dn, LDAP_DN_FORMAT_LDAPV3 );
	ldapava_free_dn( dn );
	
	if ( rc != LDAP_SUCCESS ) {
		return( LDAP_INVALID_SYNTAX );
	}

	return( LDAP_SUCCESS );
}

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
static void
AVA_Sort( LDAPRDN *rdn, int iAVA )
{
	int		i;
	LDAPAVA		*ava_in = rdn[ iAVA ][ 0 ];
	
	for ( i = 0; i < iAVA; i++ ) {
		LDAPAVA		*ava = rdn[ i ][ 0 ];
		int		a, j;

		a = strcmp( ava_in->la_attr->bv_val, ava->la_attr->bv_val );

		if ( a > 0 ) {
			break;
		}

		while ( a == 0 ) {
			int		v, d;

			d = ava_in->la_value->bv_len - ava->la_value->bv_len;

			v = memcmp( ava_in->la_value->bv_val, 
					ava->la_value->bv_val,
					d <= 0 ? ava_in->la_value->bv_len 
						: ava->la_value->bv_len );

			if ( v == 0 && d != 0 ) {
				v = d;
			}

			if ( v <= 0 ) {
				/* 
				 * got it!
				 */
				break;
			}

			if ( ++i == iAVA ) {
				/*
				 * already sorted
				 */
				return;
			}

			ava = rdn[ i ][ 0 ];
			a = strcmp( ava_in->la_value->bv_val, 
					ava->la_value->bv_val );
		}

		/*
		 * move ahead
		 */
		for ( j = iAVA; j > i; j-- ) {
			rdn[ j ][ 0 ] = rdn[ j - 1 ][ 0 ];
		}
		rdn[ i ][ 0 ] = ava_in;

		return;
	}
}

/*
 * In-place normalization of the structural representation 
 * of a distinguished name
 */
static int
DN_Normalize( LDAPDN *dn )
{
	int 		iRDN;
	int 		rc;

	assert( dn );

	for ( iRDN = 0; dn[ iRDN ]; iRDN++ ) {
		LDAPRDN		*rdn = dn[ iRDN ][ 0 ];
		int		iAVA;

		for ( iAVA = 0; rdn[ iAVA ]; iAVA++ ) {
			LDAPAVA			*ava = rdn[ iAVA ][ 0 ];
			AttributeDescription	*ad = NULL;
			const char		*text = NULL;
			slap_syntax_transform_func *nf = NULL;
			struct berval		*bv = NULL;

			rc = slap_bv2ad( ava->la_attr, &ad, &text );
			if ( rc != LDAP_SUCCESS ) {
				return( LDAP_INVALID_SYNTAX );
			}

			/* 
			 * FIXME: is this required? 
			 */
			ber_bvfree( ava->la_attr );
			ava->la_attr = ber_bvdup( &ad->ad_cname );
				
			/*
			 * FIXME: What is this intended for?
			 */
			nf = ad->ad_type->sat_syntax->ssyn_normalize;
			if ( !nf ) {
				break;
			}
				
			rc = ( *nf )( ad->ad_type->sat_syntax,
				 ava->la_value, &bv );
			
			if ( rc != LDAP_SUCCESS ) {
				return( LDAP_INVALID_SYNTAX );
			}

			/*
			 * FIXME: shouldn't this happen inside 
			 * ssyn_normalize if the syntax is case 
			 * insensitive?
			 */
			if ( !( ava->la_flags & LDAP_AVA_BINARY ) ) {
				struct berval *s = bv;
				
				bv = ber_bvstr( UTF8normalize( bv, 
						UTF8_CASEFOLD ) );
				ber_bvfree( s );
			}

			ber_bvfree( ava->la_value );
			ava->la_value = bv;

			AVA_Sort( rdn, iAVA );
		}
	}

	return( LDAP_SUCCESS );
}

379
380
381
382
383
384
385
386
int
dnNormalize(
	Syntax *syntax,
	struct berval *val,
	struct berval **normalized )
{
	struct berval *out = NULL;

387
388
	Debug( LDAP_DEBUG_TRACE, ">>> dnNormalize: %s\n", val->bv_val, 0, 0 );

389
390
391
	if ( val->bv_len != 0 ) {
		LDAPDN		*dn = NULL;
		char		*dn_out = NULL;
392
		int		rc;
393

394
395
396
		/*
		 * Go to structural representation
		 */
397
398
399
400
401
		rc = ldap_str2dn( val->bv_val, &dn, LDAP_DN_FORMAT_LDAPV3 );
		if ( rc != LDAP_SUCCESS ) {
			return( LDAP_INVALID_SYNTAX );
		}

402
403
404
		/*
		 * Add schema-aware normalization stuff
		 */
405
406
		if ( DN_Normalize( dn ) != LDAP_SUCCESS ) {
			goto error_return;
407
408
		}

409
410
411
		/*
		 * Back to string representation
		 */
412
413
414
		rc = ldap_dn2str( dn, &dn_out, LDAP_DN_FORMAT_LDAPV3 );

		if ( rc != LDAP_SUCCESS ) {
415
416
error_return:;
			ldapava_free_dn( dn );
417
418
419
			return( LDAP_INVALID_SYNTAX );
		}

420
421
		ldapava_free_dn( dn );

422
423
424
425
426
427
		out = ber_bvstr( dn_out );

	} else {
		out = ber_bvdup( val );
	}

428
429
	Debug( LDAP_DEBUG_TRACE, "<<< dnNormalize: %s\n", out->bv_val, 0, 0 );

430
431
432
433
434
435
436
437
438
	*normalized = out;

	return( LDAP_SUCCESS );
}

int
dnPretty(
	Syntax *syntax,
	struct berval *val,
439
	struct berval **pretty)
440
441
442
443
444
445
446
447
{
	struct berval *out = NULL;

	if ( val->bv_len != 0 ) {
		LDAPDN		*dn = NULL;
		char		*dn_out = NULL;
		int		rc;

Kurt Zeilenga's avatar
Kurt Zeilenga committed
448
449
		/* FIXME: should be liberal in what we accept */
		rc = ldap_str2dn( val->bv_val, &dn, LDAP_DN_FORMAT_LDAPV3 );
450
451
452
453
		if ( rc != LDAP_SUCCESS ) {
			return( LDAP_INVALID_SYNTAX );
		}

Kurt Zeilenga's avatar
Kurt Zeilenga committed
454
455
456
457
458
459
460
461
462
463
464
465
		/* Add LDAPDN pretty code here
		 *   pretty'ing MUST preserve values
		 *   but MAY use alternative representation of value.
		 *
		 * That is, in addition to changes made by the
		 * str->LDAPDN->str processing, this code can
		 *	 convert values from BER to LDAP string representation
		 *	 	(or other direction if appropriate)
		 *	 "pretty" values
		 *	 normalize attribute descriptions
		 *	 reorder AVAs in RDNs
		 */
466

Kurt Zeilenga's avatar
Kurt Zeilenga committed
467
468
469
		/* FIXME: not sure why the default isn't pretty */
		rc = ldap_dn2str( dn, &dn_out,
			LDAP_DN_FORMAT_LDAPV3 | LDAP_DN_PRETTY );
470
471
472
473
474
475
476
477
478
479
480
481
		ldapava_free_dn( dn );

		if ( rc != LDAP_SUCCESS ) {
			return( LDAP_INVALID_SYNTAX );
		}

		out = ber_bvstr( dn_out );

	} else {
		out = ber_bvdup( val );
	}

482
	*pretty = out;
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

	return( LDAP_SUCCESS );
}

int
dnMatch(
	int *matchp,
	slap_mask_t flags,
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *value,
	void *assertedValue )
{
	int match;
	struct berval *asserted = (struct berval *) assertedValue;
	
	match = value->bv_len - asserted->bv_len;

	if ( match == 0 ) {
#ifdef USE_DN_NORMALIZE
		match = strcmp( value->bv_val, asserted->bv_val );
		fprintf(stderr, "USE_DN_NORMALIZE :(\n");
#else
		match = strcasecmp( value->bv_val, asserted->bv_val );
		fprintf(stderr, "!USE_DN_NORMALIZE :)\n");
#endif
	}

#ifdef NEW_LOGGING
	LDAP_LOG(( "schema", LDAP_LEVEL_ENTRY,
		"dnMatch: %d\n    %s\n    %s\n", match,
		value->bv_val, asserted->bv_val ));
#else
	Debug( LDAP_DEBUG_ARGS, "dnMatch %d\n\t\"%s\"\n\t\"%s\"\n",
		match, value->bv_val, asserted->bv_val );
#endif

	*matchp = match;
	return( LDAP_SUCCESS );
}

#else /* !USE_LDAP_DN_PARSING */

526
527
528
529
530
531
532
533
534
535
static int
dnValidate(
	Syntax *syntax,
	struct berval *in )
{
	int rc;
	char *dn;

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

536
	dn = ch_strdup( in->bv_val );
537

538
	if( dn == NULL ) {
Kurt Zeilenga's avatar
Kurt Zeilenga committed
539
		return LDAP_INVALID_SYNTAX;
540

541
	} else if ( strlen( in->bv_val ) != in->bv_len ) {
Kurt Zeilenga's avatar
Kurt Zeilenga committed
542
		rc = LDAP_INVALID_SYNTAX;
543

544
	} else if ( dn_validate( dn ) == NULL ) {
Kurt Zeilenga's avatar
Kurt Zeilenga committed
545
		rc = LDAP_INVALID_SYNTAX;
546
547
548
549
550
551

	} else {
		rc = LDAP_SUCCESS;
	}

	ch_free( dn );
552
553
554
	return rc;
}

555
int
556
557
558
559
560
dnNormalize(
	Syntax *syntax,
	struct berval *val,
	struct berval **normalized )
{
561
	struct berval *out;
562

563
564
	if ( val->bv_len != 0 ) {
		char *dn;
565
		out = ber_bvstr( UTF8normalize( val, UTF8_CASEFOLD ) );
Kurt Zeilenga's avatar
Kurt Zeilenga committed
566

567
568
		dn = dn_validate( out->bv_val );

569
		if( dn == NULL ) {
570
571
572
			ber_bvfree( out );
			return LDAP_INVALID_SYNTAX;
		}
573

574
575
576
577
		out->bv_val = dn;
		out->bv_len = strlen( dn );
	} else {
		out = ber_bvdup( val );
578
579
580
581
582
583
	}

	*normalized = out;
	return LDAP_SUCCESS;
}

584
int
585
dnMatch(
586
	int *matchp,
587
	slap_mask_t flags,
588
589
590
591
592
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *value,
	void *assertedValue )
{
593
	int match;
594
595
	struct berval *asserted = (struct berval *) assertedValue;
	
596
597
598
	match = value->bv_len - asserted->bv_len;

	if( match == 0 ) {
599
#ifdef USE_DN_NORMALIZE
600
		match = strcmp( value->bv_val, asserted->bv_val );
601
#else
602
		match = strcasecmp( value->bv_val, asserted->bv_val );
603
#endif
604
605
	}

606
#ifdef NEW_LOGGING
Gary Williams's avatar
Gary Williams committed
607
	LDAP_LOG(( "schema", LDAP_LEVEL_ENTRY,
Kurt Zeilenga's avatar
Kurt Zeilenga committed
608
609
		"dnMatch: %d\n    %s\n    %s\n", match,
		value->bv_val, asserted->bv_val ));
610
#else
611
	Debug( LDAP_DEBUG_ARGS, "dnMatch %d\n\t\"%s\"\n\t\"%s\"\n",
Kurt Zeilenga's avatar
Kurt Zeilenga committed
612
		match, value->bv_val, asserted->bv_val );
613
614
#endif

615
616
617

	*matchp = match;
	return LDAP_SUCCESS;
618
}
619

620
621
#endif /* !USE_LDAP_DN_PARSING */

622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
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;
			}
		}
Stig Venaas's avatar
Stig Venaas committed
643
644
645
646
		if( dn->bv_val[i] != '\'' ||
		    dn->bv_val[i-1] != 'B' ||
		    dn->bv_val[i-2] != '#' ) {
			ber_bvfree( dn );
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
			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;
}

717
718
719
720
721
722
723
724
725
static int
inValidate(
	Syntax *syntax,
	struct berval *in )
{
	/* any value allowed */
	return LDAP_OTHER;
}

726
static int
727
blobValidate(
728
729
730
731
	Syntax *syntax,
	struct berval *in )
{
	/* any value allowed */
732
	return LDAP_SUCCESS;
733
734
}

735
736
737
738
739
740
741
742
743
744
745
746
747
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;
	}
748

749
750
751
752
753
754
755
756
757
758
	/*
	 * rfc 2252 section 6.3 Bit String
	 * bitstring = "'" *binary-digit "'"
	 * binary-digit = "0" / "1"
	 * example: '0101111101'B
	 */
	
	if( in->bv_val[0] != '\'' ||
		in->bv_val[in->bv_len-2] != '\'' ||
		in->bv_val[in->bv_len-1] != 'B' )
759
760
761
762
	{
		return LDAP_INVALID_SYNTAX;
	}

763
	for( i=in->bv_len-3; i>0; i-- ) {
764
765
766
767
768
769
770
771
		if( in->bv_val[i] != '0' && in->bv_val[i] != '1' ) {
			return LDAP_INVALID_SYNTAX;
		}
	}

	return LDAP_SUCCESS;
}

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
static int
bitStringNormalize(
	Syntax *syntax,
	struct berval *val,
	struct berval **normalized )
{
	/*
     * A normalized bitString is has no extaneous (leading) zero bits.
	 * That is, '00010'B is normalized to '10'B
	 * However, as a special case, '0'B requires no normalization.
     */
	struct berval *newval;
	char *p;

	/* start at the first bit */
	p = &val->bv_val[1];

	/* Find the first non-zero bit */
	while ( *p == '0' ) p++;

	newval = (struct berval *) ch_malloc( sizeof(struct berval) );

	if( *p == '\'' ) {
		/* no non-zero bits */
		newval->bv_val = ch_strdup("\'0\'B");
		newval->bv_len = sizeof("\'0\'B") - 1;
		goto done;
	}

	newval->bv_val = ch_malloc( val->bv_len + 1 );

	newval->bv_val[0] = '\'';
	newval->bv_len = 1;

	for( ; *p != '\0'; p++ ) {
		newval->bv_val[newval->bv_len++] = *p;
	}

	newval->bv_val[newval->bv_len] = '\0';

done:
	*normalized = newval;
	return LDAP_SUCCESS;
}

817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
/*
 * 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,
848
	slap_mask_t flags,
849
850
851
852
853
854
855
856
857
858
859
	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;
}

860
861
862
863
864
865
866
867
868
static int
UTF8StringValidate(
	Syntax *syntax,
	struct berval *in )
{
	ber_len_t count;
	int len;
	unsigned char *u = in->bv_val;

869
870
	if( !in->bv_len ) return LDAP_INVALID_SYNTAX;

871
	for( count = in->bv_len; count > 0; count-=len, u+=len ) {
872
873
874
875
		/* get the length indicated by the first byte */
		len = LDAP_UTF8_CHARLEN( u );

		/* should not be zero */
876
		if( len == 0 ) return LDAP_INVALID_SYNTAX;
877
878
879

		/* make sure len corresponds with the offset
			to the next character */
880
		if( LDAP_UTF8_OFFSET( u ) != len ) return LDAP_INVALID_SYNTAX;
881
882
	}

883
	if( count != 0 ) return LDAP_INVALID_SYNTAX;
884

885
	return LDAP_SUCCESS;
886
887
888
889
890
891
892
893
894
}

static int
UTF8StringNormalize(
	Syntax *syntax,
	struct berval *val,
	struct berval **normalized )
{
	struct berval *newval;
895
	char *p, *q, *s;
896

897
	newval = ch_malloc( sizeof( struct berval ) );
898

899
	p = val->bv_val;
900

901
902
903
	/* Ignore initial whitespace */
	while ( ldap_utf8_isspace( p ) ) {
		LDAP_UTF8_INCR( p );
Kurt Zeilenga's avatar
Kurt Zeilenga committed
904
	}
905

906
907
	if( *p == '\0' ) {
		ch_free( newval );
Kurt Zeilenga's avatar
Kurt Zeilenga committed
908
909
		return LDAP_INVALID_SYNTAX;
	}
910

911
912
913
	newval->bv_val = ch_strdup( p );
	p = q = newval->bv_val;
	s = NULL;
914

915
916
	while ( *p ) {
		int len;
917

918
919
920
921
922
		if ( ldap_utf8_isspace( p ) ) {
			len = LDAP_UTF8_COPY(q,p);
			s=q;
			p+=len;
			q+=len;
923

924
925
926
927
			/* Ignore the extra whitespace */
			while ( ldap_utf8_isspace( p ) ) {
				LDAP_UTF8_INCR( p );
			}
Kurt Zeilenga's avatar
Kurt Zeilenga committed
928
		} else {
929
930
931
932
			len = LDAP_UTF8_COPY(q,p);
			s=NULL;
			p+=len;
			q+=len;
Kurt Zeilenga's avatar
Kurt Zeilenga committed
933
		}
934
935
936
937
	}

	assert( *newval->bv_val );
	assert( newval->bv_val < p );
Kurt Zeilenga's avatar
Kurt Zeilenga committed
938
	assert( q <= p );
939

940
941
942
943
944
945
946
947
948
949
950
	/* 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
951
	}
952

953
954
955
956
957
958
959
960
	/* 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;
961

962
	return LDAP_SUCCESS;
963
964
}

965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
/* 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 ) {
981
		nsa->sa_initial = ber_bvstr( UTF8normalize( sa->sa_initial, casefold ) );
982
983
984
985
986
987
988
989
990
991
992
		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++ ) {
993
			nsa->sa_any[i] = ber_bvstr( UTF8normalize( sa->sa_any[i], casefold ) );
994
995
996
997
998
999
1000
1001
			if( nsa->sa_any[i] == NULL ) {
				goto err;
			}
		}
		nsa->sa_any[i] = NULL;
	}

	if( sa->sa_final != NULL ) {
1002
		nsa->sa_final = ber_bvstr( UTF8normalize( sa->sa_final, casefold ) );
1003
1004
1005
1006
1007
1008
1009
1010
		if( nsa->sa_final == NULL ) {
			goto err;
		}
	}

	return nsa;

err:
1011
	ber_bvfree( nsa->sa_final );
1012
	ber_bvecfree( nsa->sa_any );
1013
	ber_bvfree( nsa->sa_initial );
1014
1015
1016
1017
	ch_free( nsa );
	return NULL;
}

1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
/* 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;
}

1042
#ifndef SLAPD_APPROX_OLDSINGLESTRING
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060

#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 )
{
1061
	char *val, *nval, *assertv, **values, **words, *c;
1062
	int i, count, len, nextchunk=0, nextavail=0;
1063
	size_t avlen;
1064

1065
	/* Yes, this is necessary */
1066
	nval = UTF8normalize( value, UTF8_NOCASEFOLD );
1067
	if( nval == NULL ) {
1068
1069
1070
		*matchp = 1;
		return LDAP_SUCCESS;
	}
1071
	strip8bitChars( nval );
1072
1073

	/* Yes, this is necessary */
1074
	assertv = UTF8normalize( ((struct berval *)assertedValue),
1075
1076
				 UTF8_NOCASEFOLD );
	if( assertv == NULL ) {
1077
		ch_free( nval );
1078
1079
1080
1081
1082
		*matchp = 1;
		return LDAP_SUCCESS;
	}
	strip8bitChars( assertv );
	avlen = strlen( assertv );
1083
1084

	/* Isolate how many words there are */
1085
	for( c=nval,count=1; *c; c++ ) {
1086
1087
1088
1089
1090
1091
1092
1093
1094
		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 *) );
1095
	for( c=nval,i=0;  i<count;  i++,c+=strlen(c)+1 ) {
1096
1097
1098
1099
		words[i] = c;
		values[i] = phonetic(c);
	}

1100
	/* Work through the asserted value's words, to see if at least some
1101
1102
	   of the words are there, in the same order. */
	len = 0;
1103
	while ( nextchunk < avlen ) {
1104
		len = strcspn( assertv + nextchunk, SLAPD_APPROX_DELIMITER);
1105
1106
1107
1108
		if( len == 0 ) {
			nextchunk++;
			continue;
		}
1109
#if defined(SLAPD_APPROX_INITIALS)
1110
		else if( len == 1 ) {
1111
1112
			/* Single letter words need to at least match one word's initial */
			for( i=nextavail; i<count; i++ )
1113
1114
				if( !strncasecmp( assertv+nextchunk, words[i], 1 )) {
					nextavail=i+1;
1115
					break;
1116
				}
1117
1118
		}
#endif
1119
		else {
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
			/* 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;
				}
			}
1131
			ch_free( val );
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
		}

		/* 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 */
1153
	free( assertv );
1154
1155
1156
1157
1158
	for( i=0; i<count; i++ ) {
		ch_free( values[i] );
	}
	ch_free( values );
	ch_free( words );
1159
	ch_free( nval );
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178

	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++ ) {
1179
		/* Yes, this is necessary */
1180
		val = UTF8normalize( values[j], UTF8_NOCASEFOLD );
1181
1182
		strip8bitChars( val );

1183
		/* Isolate how many words there are. There will be a key for each */
Gary Williams's avatar
Gary Williams committed
1184
		for( wordcount=0,c=val;	 *c;  c++) {
1185
1186
1187
1188
1189
1190
1191
1192
1193
			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
1194
			* sizeof(struct berval *) );
1195
1196
1197
1198
1199
		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
1200
		for( c=val,i=0;	 i<wordcount;  c+=len+1	 ) {
1201
1202
1203
1204
1205
1206
1207
1208
1209
			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++;
		}

1210
		free( val );
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
	}
	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;

1232
	/* Yes, this is necessary */
1233
	val = UTF8normalize( ((struct berval *)assertValue),
1234
1235
1236
1237
1238
1239
1240
1241
1242
			     UTF8_NOCASEFOLD );
	if( val == NULL ) {
		keys = (struct berval **)ch_malloc( sizeof(struct berval *) );
		keys[0] = NULL;
		*keysp = keys;
		return LDAP_SUCCESS;
	}
	strip8bitChars( val );

1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
	/* 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
1256
	for( c=val,i=0;	 i<count; c+=len+1 ) {
1257
1258
		len = strlen(c);
		if( len < SLAPD_APPROX_WORDLEN ) continue;
1259
		keys[i] = ber_bvstr( phonetic( c ) );
1260
1261
1262
		i++;
	}

1263
	free( val );
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284

	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;
1285
	char *s, *t;
1286

1287
	/* Yes, this is necessary */
1288
	s = UTF8normalize( value, UTF8_NOCASEFOLD );
1289
1290
1291
1292
1293
1294
	if( s == NULL ) {
		*matchp = 1;
		return LDAP_SUCCESS;
	}

	/* Yes, this is necessary */
1295
	t = UTF8normalize( ((struct berval *)assertedValue),
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
			   UTF8_NOCASEFOLD );
	if( t == NULL ) {
		free( s );
		*matchp = -1;
		return LDAP_SUCCESS;
	}

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

	free( s );
	free( t );
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328

	*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;
1329
	char *s;
1330
1331

	for( i=0; values[i] != NULL; i++ ) {
Kurt Zeilenga's avatar
Kurt Zeilenga committed
1332
		/* empty - just count them */
1333
	}
Kurt Zeilenga's avatar
Kurt Zeilenga committed
1334
1335

	/* we should have at least one value at this point */
1336
1337
1338
1339
1340
1341
	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++ ) {
1342
		/* Yes, this is necessary */
1343
		s = UTF8normalize( values[i], UTF8_NOCASEFOLD );
1344
1345
1346
1347

		/* strip 8-bit chars and run through phonetic() */
		keys[i] = ber_bvstr( phonetic( strip8bitChars( s ) ) );
		free( s );
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
	}
	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;
1367
	char *s;
1368
1369
1370

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

1371
	/* Yes, this is necessary */
1372
	s = UTF8normalize( ((struct berval *)assertValue),
1373
1374
1375
1376
1377
1378
1379
1380
1381
			     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;
	}
1382
1383
1384
1385
1386
1387
1388

	*keysp = keys;
	return LDAP_SUCCESS;
}
#endif


1389
static int
1390
caseExactMatch(
1391
	int *matchp,
1392
	slap_mask_t flags,
1393
1394
1395
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *value,
1396
	void *assertedValue )
1397
{
1398
1399
1400
	*matchp = UTF8normcmp( value->bv_val,
		((struct berval *) assertedValue)->bv_val,
		UTF8_NOCASEFOLD );
1401
	return LDAP_SUCCESS;
1402
1403
}

1404
static int
1405
caseExactIgnoreSubstringsMatch(
1406
	int *matchp,
1407
	slap_mask_t flags,
1408
1409
1410
1411
1412
1413
	Syntax *syntax,
	MatchingRule *mr,
	struct berval *value,
	void *assertedValue )
{
	int match = 0;
Pierangelo Masarati's avatar
Pierangelo Masarati committed
1414
	SubstringsAssertion *sub = NULL;
1415
	struct berval left;
1416
1417
	int i;
	ber_len_t inlen=0;
1418
1419
1420
1421
	char *nav, casefold;

	casefold = strcmp( mr->smr_oid, caseExactSubstringsMatchOID )
		? UTF8_CASEFOLD : UTF8_NOCASEFOLD;
1422

1423
	nav = UTF8normalize( value, casefold );
1424
1425
1426