entities.c 29.2 KB
Newer Older
1
/*
2
 * entities.c : implementation for the XML entities handling
3
4
5
 *
 * See Copyright for the status of this software.
 *
6
 * daniel@veillard.com
7
8
 */

9
10
11
12
13
/* To avoid EBCDIC trouble when parsing on zOS */
#if defined(__MVS__)
#pragma convert("ISO8859-1")
#endif

14
#define IN_LIBXML
Bjorn Reese's avatar
Bjorn Reese committed
15
#include "libxml.h"
16
17
18
19
20
21
22
23
24

#include <string.h>
#ifdef HAVE_STDLIB_H
#include <stdlib.h>
#endif
#include <libxml/xmlmemory.h>
#include <libxml/hash.h>
#include <libxml/entities.h>
#include <libxml/parser.h>
25
#include <libxml/parserInternals.h>
26
#include <libxml/xmlerror.h>
27
#include <libxml/globals.h>
28
#include <libxml/dict.h>
29

30
31
#include "save.h"

32
33
34
35
/*
 * The XML predefined entities.
 */

36
37
static xmlEntity xmlEntityLt = {
    NULL, XML_ENTITY_DECL, BAD_CAST "lt",
Daniel Veillard's avatar
Daniel Veillard committed
38
    NULL, NULL, NULL, NULL, NULL, NULL,
39
40
    BAD_CAST "<", BAD_CAST "<", 1,
    XML_INTERNAL_PREDEFINED_ENTITY,
41
    NULL, NULL, NULL, NULL, 0, 1
42
};
43
44
static xmlEntity xmlEntityGt = {
    NULL, XML_ENTITY_DECL, BAD_CAST "gt",
Daniel Veillard's avatar
Daniel Veillard committed
45
    NULL, NULL, NULL, NULL, NULL, NULL,
46
47
    BAD_CAST ">", BAD_CAST ">", 1,
    XML_INTERNAL_PREDEFINED_ENTITY,
48
    NULL, NULL, NULL, NULL, 0, 1
49
50
51
};
static xmlEntity xmlEntityAmp = {
    NULL, XML_ENTITY_DECL, BAD_CAST "amp",
Daniel Veillard's avatar
Daniel Veillard committed
52
    NULL, NULL, NULL, NULL, NULL, NULL,
53
54
    BAD_CAST "&", BAD_CAST "&", 1,
    XML_INTERNAL_PREDEFINED_ENTITY,
55
    NULL, NULL, NULL, NULL, 0, 1
56
57
58
};
static xmlEntity xmlEntityQuot = {
    NULL, XML_ENTITY_DECL, BAD_CAST "quot",
Daniel Veillard's avatar
Daniel Veillard committed
59
    NULL, NULL, NULL, NULL, NULL, NULL,
60
61
    BAD_CAST "\"", BAD_CAST "\"", 1,
    XML_INTERNAL_PREDEFINED_ENTITY,
62
    NULL, NULL, NULL, NULL, 0, 1
63
64
65
};
static xmlEntity xmlEntityApos = {
    NULL, XML_ENTITY_DECL, BAD_CAST "apos",
Daniel Veillard's avatar
Daniel Veillard committed
66
    NULL, NULL, NULL, NULL, NULL, NULL,
67
68
    BAD_CAST "'", BAD_CAST "'", 1,
    XML_INTERNAL_PREDEFINED_ENTITY,
69
    NULL, NULL, NULL, NULL, 0, 1
70
71
};

72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
/**
 * xmlEntitiesErrMemory:
 * @extra:  extra informations
 *
 * Handle an out of memory condition
 */
static void
xmlEntitiesErrMemory(const char *extra)
{
    __xmlSimpleError(XML_FROM_TREE, XML_ERR_NO_MEMORY, NULL, NULL, extra);
}

/**
 * xmlEntitiesErr:
 * @code:  the error code
 * @msg:  the message
 *
 * Handle an out of memory condition
 */
91
static void LIBXML_ATTR_FORMAT(2,0)
92
93
94
95
96
xmlEntitiesErr(xmlParserErrors code, const char *msg)
{
    __xmlSimpleError(XML_FROM_TREE, code, NULL, msg, NULL);
}

97
98
99
/*
 * xmlFreeEntity : clean-up an entity record.
 */
100
101
102
103
104
105
106
107
108
109
110
static void
xmlFreeEntity(xmlEntityPtr entity)
{
    xmlDictPtr dict = NULL;

    if (entity == NULL)
        return;

    if (entity->doc != NULL)
        dict = entity->doc->dict;

111

112
    if ((entity->children) && (entity->owner == 1) &&
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
        (entity == (xmlEntityPtr) entity->children->parent))
        xmlFreeNodeList(entity->children);
    if (dict != NULL) {
        if ((entity->name != NULL) && (!xmlDictOwns(dict, entity->name)))
            xmlFree((char *) entity->name);
        if ((entity->ExternalID != NULL) &&
	    (!xmlDictOwns(dict, entity->ExternalID)))
            xmlFree((char *) entity->ExternalID);
        if ((entity->SystemID != NULL) &&
	    (!xmlDictOwns(dict, entity->SystemID)))
            xmlFree((char *) entity->SystemID);
        if ((entity->URI != NULL) && (!xmlDictOwns(dict, entity->URI)))
            xmlFree((char *) entity->URI);
        if ((entity->content != NULL)
            && (!xmlDictOwns(dict, entity->content)))
            xmlFree((char *) entity->content);
        if ((entity->orig != NULL) && (!xmlDictOwns(dict, entity->orig)))
            xmlFree((char *) entity->orig);
    } else {
        if (entity->name != NULL)
            xmlFree((char *) entity->name);
        if (entity->ExternalID != NULL)
            xmlFree((char *) entity->ExternalID);
        if (entity->SystemID != NULL)
            xmlFree((char *) entity->SystemID);
        if (entity->URI != NULL)
            xmlFree((char *) entity->URI);
        if (entity->content != NULL)
            xmlFree((char *) entity->content);
        if (entity->orig != NULL)
            xmlFree((char *) entity->orig);
    }
145
146
147
148
    xmlFree(entity);
}

/*
149
150
151
 * xmlCreateEntity:
 *
 * internal routine doing the entity node strutures allocations
152
153
 */
static xmlEntityPtr
154
155
156
xmlCreateEntity(xmlDictPtr dict, const xmlChar *name, int type,
	        const xmlChar *ExternalID, const xmlChar *SystemID,
	        const xmlChar *content) {
157
158
159
160
    xmlEntityPtr ret;

    ret = (xmlEntityPtr) xmlMalloc(sizeof(xmlEntity));
    if (ret == NULL) {
161
        xmlEntitiesErrMemory("xmlCreateEntity: malloc failed");
162
163
164
165
	return(NULL);
    }
    memset(ret, 0, sizeof(xmlEntity));
    ret->type = XML_ENTITY_DECL;
166
    ret->checked = 0;
167
168
169
170
171

    /*
     * fill the structure.
     */
    ret->etype = (xmlEntityType) type;
172
173
174
175
176
177
178
179
180
181
182
183
184
    if (dict == NULL) {
	ret->name = xmlStrdup(name);
	if (ExternalID != NULL)
	    ret->ExternalID = xmlStrdup(ExternalID);
	if (SystemID != NULL)
	    ret->SystemID = xmlStrdup(SystemID);
    } else {
        ret->name = xmlDictLookup(dict, name, -1);
	if (ExternalID != NULL)
	    ret->ExternalID = xmlDictLookup(dict, ExternalID, -1);
	if (SystemID != NULL)
	    ret->SystemID = xmlDictLookup(dict, SystemID, -1);
    }
185
186
    if (content != NULL) {
        ret->length = xmlStrlen(content);
187
188
189
190
191
	if ((dict != NULL) && (ret->length < 5))
	    ret->content = (xmlChar *)
	                   xmlDictLookup(dict, content, ret->length);
	else
	    ret->content = xmlStrndup(content, ret->length);
192
193
194
195
196
197
198
     } else {
        ret->length = 0;
        ret->content = NULL;
    }
    ret->URI = NULL; /* to be computed by the layer knowing
			the defining entity */
    ret->orig = NULL;
199
    ret->owner = 0;
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

    return(ret);
}

/*
 * xmlAddEntity : register a new entity for an entities table.
 */
static xmlEntityPtr
xmlAddEntity(xmlDtdPtr dtd, const xmlChar *name, int type,
	  const xmlChar *ExternalID, const xmlChar *SystemID,
	  const xmlChar *content) {
    xmlDictPtr dict = NULL;
    xmlEntitiesTablePtr table = NULL;
    xmlEntityPtr ret;

    if (name == NULL)
	return(NULL);
    if (dtd == NULL)
	return(NULL);
    if (dtd->doc != NULL)
        dict = dtd->doc->dict;

    switch (type) {
        case XML_INTERNAL_GENERAL_ENTITY:
        case XML_EXTERNAL_GENERAL_PARSED_ENTITY:
        case XML_EXTERNAL_GENERAL_UNPARSED_ENTITY:
	    if (dtd->entities == NULL)
		dtd->entities = xmlHashCreateDict(0, dict);
	    table = dtd->entities;
	    break;
        case XML_INTERNAL_PARAMETER_ENTITY:
        case XML_EXTERNAL_PARAMETER_ENTITY:
	    if (dtd->pentities == NULL)
		dtd->pentities = xmlHashCreateDict(0, dict);
	    table = dtd->pentities;
	    break;
        case XML_INTERNAL_PREDEFINED_ENTITY:
	    return(NULL);
    }
    if (table == NULL)
	return(NULL);
    ret = xmlCreateEntity(dict, name, type, ExternalID, SystemID, content);
    if (ret == NULL)
        return(NULL);
244
    ret->doc = dtd->doc;
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261

    if (xmlHashAddEntry(table, name, ret)) {
	/*
	 * entity was already defined at another level.
	 */
        xmlFreeEntity(ret);
	return(NULL);
    }
    return(ret);
}

/**
 * xmlGetPredefinedEntity:
 * @name:  the entity name
 *
 * Check whether this name is an predefined entity.
 *
262
 * Returns NULL if not, otherwise the entity
263
264
265
 */
xmlEntityPtr
xmlGetPredefinedEntity(const xmlChar *name) {
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
    if (name == NULL) return(NULL);
    switch (name[0]) {
        case 'l':
	    if (xmlStrEqual(name, BAD_CAST "lt"))
	        return(&xmlEntityLt);
	    break;
        case 'g':
	    if (xmlStrEqual(name, BAD_CAST "gt"))
	        return(&xmlEntityGt);
	    break;
        case 'a':
	    if (xmlStrEqual(name, BAD_CAST "amp"))
	        return(&xmlEntityAmp);
	    if (xmlStrEqual(name, BAD_CAST "apos"))
	        return(&xmlEntityApos);
	    break;
        case 'q':
	    if (xmlStrEqual(name, BAD_CAST "quot"))
	        return(&xmlEntityQuot);
	    break;
	default:
	    break;
    }
    return(NULL);
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
}

/**
 * xmlAddDtdEntity:
 * @doc:  the document
 * @name:  the entity name
 * @type:  the entity type XML_xxx_yyy_ENTITY
 * @ExternalID:  the entity external ID if available
 * @SystemID:  the entity system ID if available
 * @content:  the entity content
 *
 * Register a new entity for this document DTD external subset.
 *
 * Returns a pointer to the entity or NULL in case of error
 */
xmlEntityPtr
xmlAddDtdEntity(xmlDocPtr doc, const xmlChar *name, int type,
	        const xmlChar *ExternalID, const xmlChar *SystemID,
		const xmlChar *content) {
    xmlEntityPtr ret;
    xmlDtdPtr dtd;

    if (doc == NULL) {
313
314
	xmlEntitiesErr(XML_DTD_NO_DOC,
	        "xmlAddDtdEntity: document is NULL");
315
316
317
	return(NULL);
    }
    if (doc->extSubset == NULL) {
318
319
	xmlEntitiesErr(XML_DTD_NO_DTD,
	        "xmlAddDtdEntity: document without external subset");
320
321
322
323
324
325
326
	return(NULL);
    }
    dtd = doc->extSubset;
    ret = xmlAddEntity(dtd, name, type, ExternalID, SystemID, content);
    if (ret == NULL) return(NULL);

    /*
327
     * Link it to the DTD
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
     */
    ret->parent = dtd;
    ret->doc = dtd->doc;
    if (dtd->last == NULL) {
	dtd->children = dtd->last = (xmlNodePtr) ret;
    } else {
        dtd->last->next = (xmlNodePtr) ret;
	ret->prev = dtd->last;
	dtd->last = (xmlNodePtr) ret;
    }
    return(ret);
}

/**
 * xmlAddDocEntity:
 * @doc:  the document
 * @name:  the entity name
 * @type:  the entity type XML_xxx_yyy_ENTITY
 * @ExternalID:  the entity external ID if available
 * @SystemID:  the entity system ID if available
 * @content:  the entity content
 *
 * Register a new entity for this document.
 *
 * Returns a pointer to the entity or NULL in case of error
 */
xmlEntityPtr
xmlAddDocEntity(xmlDocPtr doc, const xmlChar *name, int type,
	        const xmlChar *ExternalID, const xmlChar *SystemID,
	        const xmlChar *content) {
    xmlEntityPtr ret;
    xmlDtdPtr dtd;

    if (doc == NULL) {
362
363
	xmlEntitiesErr(XML_DTD_NO_DOC,
	        "xmlAddDocEntity: document is NULL");
364
365
366
	return(NULL);
    }
    if (doc->intSubset == NULL) {
367
368
	xmlEntitiesErr(XML_DTD_NO_DTD,
	        "xmlAddDocEntity: document without internal subset");
369
370
371
372
373
374
375
	return(NULL);
    }
    dtd = doc->intSubset;
    ret = xmlAddEntity(dtd, name, type, ExternalID, SystemID, content);
    if (ret == NULL) return(NULL);

    /*
376
     * Link it to the DTD
377
378
379
380
381
382
383
384
385
386
387
388
389
     */
    ret->parent = dtd;
    ret->doc = dtd->doc;
    if (dtd->last == NULL) {
	dtd->children = dtd->last = (xmlNodePtr) ret;
    } else {
	dtd->last->next = (xmlNodePtr) ret;
	ret->prev = dtd->last;
	dtd->last = (xmlNodePtr) ret;
    }
    return(ret);
}

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
/**
 * xmlNewEntity:
 * @doc:  the document
 * @name:  the entity name
 * @type:  the entity type XML_xxx_yyy_ENTITY
 * @ExternalID:  the entity external ID if available
 * @SystemID:  the entity system ID if available
 * @content:  the entity content
 *
 * Create a new entity, this differs from xmlAddDocEntity() that if
 * the document is NULL or has no internal subset defined, then an
 * unlinked entity structure will be returned, it is then the responsability
 * of the caller to link it to the document later or free it when not needed
 * anymore.
 *
 * Returns a pointer to the entity or NULL in case of error
 */
xmlEntityPtr
xmlNewEntity(xmlDocPtr doc, const xmlChar *name, int type,
	     const xmlChar *ExternalID, const xmlChar *SystemID,
	     const xmlChar *content) {
    xmlEntityPtr ret;
    xmlDictPtr dict;

    if ((doc != NULL) && (doc->intSubset != NULL)) {
	return(xmlAddDocEntity(doc, name, type, ExternalID, SystemID, content));
    }
    if (doc != NULL)
        dict = doc->dict;
    else
        dict = NULL;
    ret = xmlCreateEntity(dict, name, type, ExternalID, SystemID, content);
    if (ret == NULL)
        return(NULL);
    ret->doc = doc;
    return(ret);
}

428
429
430
431
432
433
434
435
/**
 * xmlGetEntityFromTable:
 * @table:  an entity table
 * @name:  the entity name
 * @parameter:  look for parameter entities
 *
 * Do an entity lookup in the table.
 * returns the corresponding parameter entity, if found.
Daniel Veillard's avatar
Daniel Veillard committed
436
 *
437
438
 * Returns A pointer to the entity structure or NULL if not found.
 */
439
static xmlEntityPtr
440
441
442
443
444
445
446
447
448
449
450
xmlGetEntityFromTable(xmlEntitiesTablePtr table, const xmlChar *name) {
    return((xmlEntityPtr) xmlHashLookup(table, name));
}

/**
 * xmlGetParameterEntity:
 * @doc:  the document referencing the entity
 * @name:  the entity name
 *
 * Do an entity lookup in the internal and external subsets and
 * returns the corresponding parameter entity, if found.
Daniel Veillard's avatar
Daniel Veillard committed
451
 *
452
453
454
455
456
457
458
 * Returns A pointer to the entity structure or NULL if not found.
 */
xmlEntityPtr
xmlGetParameterEntity(xmlDocPtr doc, const xmlChar *name) {
    xmlEntitiesTablePtr table;
    xmlEntityPtr ret;

459
460
    if (doc == NULL)
	return(NULL);
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
    if ((doc->intSubset != NULL) && (doc->intSubset->pentities != NULL)) {
	table = (xmlEntitiesTablePtr) doc->intSubset->pentities;
	ret = xmlGetEntityFromTable(table, name);
	if (ret != NULL)
	    return(ret);
    }
    if ((doc->extSubset != NULL) && (doc->extSubset->pentities != NULL)) {
	table = (xmlEntitiesTablePtr) doc->extSubset->pentities;
	return(xmlGetEntityFromTable(table, name));
    }
    return(NULL);
}

/**
 * xmlGetDtdEntity:
 * @doc:  the document referencing the entity
 * @name:  the entity name
 *
479
 * Do an entity lookup in the DTD entity hash table and
480
 * returns the corresponding entity, if found.
481
 * Note: the first argument is the document node, not the DTD node.
Daniel Veillard's avatar
Daniel Veillard committed
482
 *
483
484
485
486
487
488
 * Returns A pointer to the entity structure or NULL if not found.
 */
xmlEntityPtr
xmlGetDtdEntity(xmlDocPtr doc, const xmlChar *name) {
    xmlEntitiesTablePtr table;

489
490
    if (doc == NULL)
	return(NULL);
491
492
493
494
495
496
497
498
499
500
501
502
503
    if ((doc->extSubset != NULL) && (doc->extSubset->entities != NULL)) {
	table = (xmlEntitiesTablePtr) doc->extSubset->entities;
	return(xmlGetEntityFromTable(table, name));
    }
    return(NULL);
}

/**
 * xmlGetDocEntity:
 * @doc:  the document referencing the entity
 * @name:  the entity name
 *
 * Do an entity lookup in the document entity hash table and
504
 * returns the corresponding entity, otherwise a lookup is done
505
 * in the predefined entities too.
Daniel Veillard's avatar
Daniel Veillard committed
506
 *
507
508
509
 * Returns A pointer to the entity structure or NULL if not found.
 */
xmlEntityPtr
Kurt Roeckx's avatar
Kurt Roeckx committed
510
xmlGetDocEntity(const xmlDoc *doc, const xmlChar *name) {
511
512
513
514
515
516
517
518
519
520
    xmlEntityPtr cur;
    xmlEntitiesTablePtr table;

    if (doc != NULL) {
	if ((doc->intSubset != NULL) && (doc->intSubset->entities != NULL)) {
	    table = (xmlEntitiesTablePtr) doc->intSubset->entities;
	    cur = xmlGetEntityFromTable(table, name);
	    if (cur != NULL)
		return(cur);
	}
521
522
523
524
525
526
527
528
	if (doc->standalone != 1) {
	    if ((doc->extSubset != NULL) &&
		(doc->extSubset->entities != NULL)) {
		table = (xmlEntitiesTablePtr) doc->extSubset->entities;
		cur = xmlGetEntityFromTable(table, name);
		if (cur != NULL)
		    return(cur);
	    }
529
530
	}
    }
531
    return(xmlGetPredefinedEntity(name));
532
533
534
535
536
537
}

/*
 * Macro used to grow the current buffer.
 */
#define growBufferReentrant() {						\
538
    xmlChar *tmp;                                                       \
Aron Xu's avatar
Aron Xu committed
539
    size_t new_size = buffer_size * 2;                                  \
540
541
542
543
544
    if (new_size < buffer_size) goto mem_error;                         \
    tmp = (xmlChar *) xmlRealloc(buffer, new_size);	                \
    if (tmp == NULL) goto mem_error;                                    \
    buffer = tmp;							\
    buffer_size = new_size;						\
545
546
547
}

/**
548
 * xmlEncodeEntitiesInternal:
549
550
 * @doc:  the document containing the string
 * @input:  A string to convert to XML.
551
 * @attr: are we handling an atrbute value
552
553
554
555
556
557
558
559
 *
 * Do a global encoding of a string, replacing the predefined entities
 * and non ASCII values with their entities and CharRef counterparts.
 * Contrary to xmlEncodeEntities, this routine is reentrant, and result
 * must be deallocated.
 *
 * Returns A newly allocated string with the substitution done.
 */
560
561
static xmlChar *
xmlEncodeEntitiesInternal(xmlDocPtr doc, const xmlChar *input, int attr) {
562
563
564
    const xmlChar *cur = input;
    xmlChar *buffer = NULL;
    xmlChar *out = NULL;
565
    size_t buffer_size = 0;
566
567
568
569
570
571
572
573
574
575
576
577
    int html = 0;

    if (input == NULL) return(NULL);
    if (doc != NULL)
        html = (doc->type == XML_HTML_DOCUMENT_NODE);

    /*
     * allocate an translation buffer.
     */
    buffer_size = 1000;
    buffer = (xmlChar *) xmlMalloc(buffer_size * sizeof(xmlChar));
    if (buffer == NULL) {
578
        xmlEntitiesErrMemory("xmlEncodeEntities: malloc failed");
579
580
581
582
583
	return(NULL);
    }
    out = buffer;

    while (*cur != '\0') {
584
585
        size_t indx = out - buffer;
        if (indx + 100 > buffer_size) {
586
587

	    growBufferReentrant();
588
	    out = &buffer[indx];
589
590
591
592
593
594
	}

	/*
	 * By default one have to encode at least '<', '>', '"' and '&' !
	 */
	if (*cur == '<') {
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
	    const xmlChar *end;

	    /*
	     * Special handling of server side include in HTML attributes
	     */
	    if (html && attr &&
	        (cur[1] == '!') && (cur[2] == '-') && (cur[3] == '-') &&
	        ((end = xmlStrstr(cur, BAD_CAST "-->")) != NULL)) {
	        while (cur != end) {
		    *out++ = *cur++;
		    indx = out - buffer;
		    if (indx + 100 > buffer_size) {
			growBufferReentrant();
			out = &buffer[indx];
		    }
		}
		*out++ = *cur++;
		*out++ = *cur++;
		*out++ = *cur++;
		continue;
	    }
616
617
618
619
620
621
622
623
624
625
	    *out++ = '&';
	    *out++ = 'l';
	    *out++ = 't';
	    *out++ = ';';
	} else if (*cur == '>') {
	    *out++ = '&';
	    *out++ = 'g';
	    *out++ = 't';
	    *out++ = ';';
	} else if (*cur == '&') {
626
627
628
629
	    /*
	     * Special handling of &{...} construct from HTML 4, see
	     * http://www.w3.org/TR/html401/appendix/notes.html#h-B.7.1
	     */
630
631
	    if (html && attr && (cur[1] == '{') &&
	        (strchr((const char *) cur, '}'))) {
632
633
634
635
636
637
638
639
640
641
642
	        while (*cur != '}') {
		    *out++ = *cur++;
		    indx = out - buffer;
		    if (indx + 100 > buffer_size) {
			growBufferReentrant();
			out = &buffer[indx];
		    }
		}
		*out++ = *cur++;
		continue;
	    }
643
644
645
646
647
648
	    *out++ = '&';
	    *out++ = 'a';
	    *out++ = 'm';
	    *out++ = 'p';
	    *out++ = ';';
	} else if (((*cur >= 0x20) && (*cur < 0x80)) ||
649
	    (*cur == '\n') || (*cur == '\t') || ((html) && (*cur == '\r'))) {
650
651
652
653
654
	    /*
	     * default case, just copy !
	     */
	    *out++ = *cur;
	} else if (*cur >= 0x80) {
655
	    if (((doc != NULL) && (doc->encoding != NULL)) || (html)) {
656
		/*
657
		 * Bjørn Reese <br@sseusa.com> provided the patch
658
659
660
661
662
663
664
	        xmlChar xc;
	        xc = (*cur & 0x3F) << 6;
	        if (cur[1] != 0) {
		    xc += *(++cur) & 0x3F;
		    *out++ = xc;
	        } else
		 */
665
		*out++ = *cur;
666
667
668
669
	    } else {
		/*
		 * We assume we have UTF-8 input.
		 */
670
		char buf[11], *ptr;
671
672
673
		int val = 0, l = 1;

		if (*cur < 0xC0) {
674
		    xmlEntitiesErr(XML_CHECK_NOT_UTF8,
675
			    "xmlEncodeEntities: input not UTF-8");
676
677
		    if (doc != NULL)
			doc->encoding = xmlStrdup(BAD_CAST "ISO-8859-1");
678
679
680
681
		    snprintf(buf, sizeof(buf), "&#%d;", *cur);
		    buf[sizeof(buf) - 1] = 0;
		    ptr = buf;
		    while (*ptr != 0) *out++ = *ptr++;
682
		    cur++;
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
		    continue;
		} else if (*cur < 0xE0) {
                    val = (cur[0]) & 0x1F;
		    val <<= 6;
		    val |= (cur[1]) & 0x3F;
		    l = 2;
		} else if (*cur < 0xF0) {
                    val = (cur[0]) & 0x0F;
		    val <<= 6;
		    val |= (cur[1]) & 0x3F;
		    val <<= 6;
		    val |= (cur[2]) & 0x3F;
		    l = 3;
		} else if (*cur < 0xF8) {
                    val = (cur[0]) & 0x07;
		    val <<= 6;
		    val |= (cur[1]) & 0x3F;
		    val <<= 6;
		    val |= (cur[2]) & 0x3F;
		    val <<= 6;
		    val |= (cur[3]) & 0x3F;
		    l = 4;
		}
		if ((l == 1) || (!IS_CHAR(val))) {
707
		    xmlEntitiesErr(XML_ERR_INVALID_CHAR,
708
			"xmlEncodeEntities: char out of range\n");
709
710
		    if (doc != NULL)
			doc->encoding = xmlStrdup(BAD_CAST "ISO-8859-1");
711
712
713
714
715
716
717
718
719
720
		    snprintf(buf, sizeof(buf), "&#%d;", *cur);
		    buf[sizeof(buf) - 1] = 0;
		    ptr = buf;
		    while (*ptr != 0) *out++ = *ptr++;
		    cur++;
		    continue;
		}
		/*
		 * We could do multiple things here. Just save as a char ref
		 */
721
		snprintf(buf, sizeof(buf), "&#x%X;", val);
722
723
724
725
726
727
		buf[sizeof(buf) - 1] = 0;
		ptr = buf;
		while (*ptr != 0) *out++ = *ptr++;
		cur += l;
		continue;
	    }
728
	} else if (IS_BYTE_CHAR(*cur)) {
729
	    char buf[11], *ptr;
730
731
732
733
734
735
736
737

	    snprintf(buf, sizeof(buf), "&#%d;", *cur);
	    buf[sizeof(buf) - 1] = 0;
            ptr = buf;
	    while (*ptr != 0) *out++ = *ptr++;
	}
	cur++;
    }
738
    *out = 0;
739
    return(buffer);
740
741

mem_error:
742
    xmlEntitiesErrMemory("xmlEncodeEntities: realloc failed");
743
744
    xmlFree(buffer);
    return(NULL);
745
746
}

747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
/**
 * xmlEncodeAttributeEntities:
 * @doc:  the document containing the string
 * @input:  A string to convert to XML.
 *
 * Do a global encoding of a string, replacing the predefined entities
 * and non ASCII values with their entities and CharRef counterparts for
 * attribute values.
 *
 * Returns A newly allocated string with the substitution done.
 */
xmlChar *
xmlEncodeAttributeEntities(xmlDocPtr doc, const xmlChar *input) {
    return xmlEncodeEntitiesInternal(doc, input, 1);
}

/**
 * xmlEncodeEntitiesReentrant:
 * @doc:  the document containing the string
 * @input:  A string to convert to XML.
 *
 * Do a global encoding of a string, replacing the predefined entities
 * and non ASCII values with their entities and CharRef counterparts.
 * Contrary to xmlEncodeEntities, this routine is reentrant, and result
 * must be deallocated.
 *
 * Returns A newly allocated string with the substitution done.
 */
xmlChar *
xmlEncodeEntitiesReentrant(xmlDocPtr doc, const xmlChar *input) {
    return xmlEncodeEntitiesInternal(doc, input, 0);
}

780
781
782
783
784
785
786
787
788
789
790
/**
 * xmlEncodeSpecialChars:
 * @doc:  the document containing the string
 * @input:  A string to convert to XML.
 *
 * Do a global encoding of a string, replacing the predefined entities
 * this routine is reentrant, and result must be deallocated.
 *
 * Returns A newly allocated string with the substitution done.
 */
xmlChar *
Kurt Roeckx's avatar
Kurt Roeckx committed
791
xmlEncodeSpecialChars(const xmlDoc *doc ATTRIBUTE_UNUSED, const xmlChar *input) {
792
793
794
    const xmlChar *cur = input;
    xmlChar *buffer = NULL;
    xmlChar *out = NULL;
795
    size_t buffer_size = 0;
796
    if (input == NULL) return(NULL);
797
798
799
800
801
802
803

    /*
     * allocate an translation buffer.
     */
    buffer_size = 1000;
    buffer = (xmlChar *) xmlMalloc(buffer_size * sizeof(xmlChar));
    if (buffer == NULL) {
804
        xmlEntitiesErrMemory("xmlEncodeSpecialChars: malloc failed");
805
806
807
808
809
	return(NULL);
    }
    out = buffer;

    while (*cur != '\0') {
810
811
        size_t indx = out - buffer;
        if (indx + 10 > buffer_size) {
812
813

	    growBufferReentrant();
814
	    out = &buffer[indx];
815
816
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
	}

	/*
	 * By default one have to encode at least '<', '>', '"' and '&' !
	 */
	if (*cur == '<') {
	    *out++ = '&';
	    *out++ = 'l';
	    *out++ = 't';
	    *out++ = ';';
	} else if (*cur == '>') {
	    *out++ = '&';
	    *out++ = 'g';
	    *out++ = 't';
	    *out++ = ';';
	} else if (*cur == '&') {
	    *out++ = '&';
	    *out++ = 'a';
	    *out++ = 'm';
	    *out++ = 'p';
	    *out++ = ';';
	} else if (*cur == '"') {
	    *out++ = '&';
	    *out++ = 'q';
	    *out++ = 'u';
	    *out++ = 'o';
	    *out++ = 't';
	    *out++ = ';';
843
844
845
846
847
848
	} else if (*cur == '\r') {
	    *out++ = '&';
	    *out++ = '#';
	    *out++ = '1';
	    *out++ = '3';
	    *out++ = ';';
849
850
851
852
853
854
855
856
857
	} else {
	    /*
	     * Works because on UTF-8, all extended sequences cannot
	     * result in bytes in the ASCII range.
	     */
	    *out++ = *cur;
	}
	cur++;
    }
858
    *out = 0;
859
    return(buffer);
860
861
862
863
864

mem_error:
    xmlEntitiesErrMemory("xmlEncodeSpecialChars: realloc failed");
    xmlFree(buffer);
    return(NULL);
865
866
867
868
869
870
}

/**
 * xmlCreateEntitiesTable:
 *
 * create and initialize an empty entities hash table.
871
 * This really doesn't make sense and should be deprecated
872
873
874
875
876
877
878
879
 *
 * Returns the xmlEntitiesTablePtr just created or NULL in case of error.
 */
xmlEntitiesTablePtr
xmlCreateEntitiesTable(void) {
    return((xmlEntitiesTablePtr) xmlHashCreate(0));
}

880
881
882
883
884
885
886
887
/**
 * xmlFreeEntityWrapper:
 * @entity:  An entity
 * @name:  its name
 *
 * Deallocate the memory used by an entities in the hash table.
 */
static void
Nick Wellnhofer's avatar
Nick Wellnhofer committed
888
xmlFreeEntityWrapper(void *entity, const xmlChar *name ATTRIBUTE_UNUSED) {
889
    if (entity != NULL)
Nick Wellnhofer's avatar
Nick Wellnhofer committed
890
	xmlFreeEntity((xmlEntityPtr) entity);
891
892
}

893
894
895
896
897
898
899
900
/**
 * xmlFreeEntitiesTable:
 * @table:  An entity table
 *
 * Deallocate the memory used by an entities hash table.
 */
void
xmlFreeEntitiesTable(xmlEntitiesTablePtr table) {
Nick Wellnhofer's avatar
Nick Wellnhofer committed
901
    xmlHashFree(table, xmlFreeEntityWrapper);
902
903
}

904
#ifdef LIBXML_TREE_ENABLED
905
906
907
908
909
/**
 * xmlCopyEntity:
 * @ent:  An entity
 *
 * Build a copy of an entity
Daniel Veillard's avatar
Daniel Veillard committed
910
 *
911
912
 * Returns the new xmlEntitiesPtr or NULL in case of error.
 */
Nick Wellnhofer's avatar
Nick Wellnhofer committed
913
914
915
static void *
xmlCopyEntity(void *payload, const xmlChar *name ATTRIBUTE_UNUSED) {
    xmlEntityPtr ent = (xmlEntityPtr) payload;
916
917
918
919
    xmlEntityPtr cur;

    cur = (xmlEntityPtr) xmlMalloc(sizeof(xmlEntity));
    if (cur == NULL) {
920
        xmlEntitiesErrMemory("xmlCopyEntity:: malloc failed");
921
922
923
	return(NULL);
    }
    memset(cur, 0, sizeof(xmlEntity));
924
    cur->type = XML_ENTITY_DECL;
925
926
927
928
929
930
931
932
933
934
935
936

    cur->etype = ent->etype;
    if (ent->name != NULL)
	cur->name = xmlStrdup(ent->name);
    if (ent->ExternalID != NULL)
	cur->ExternalID = xmlStrdup(ent->ExternalID);
    if (ent->SystemID != NULL)
	cur->SystemID = xmlStrdup(ent->SystemID);
    if (ent->content != NULL)
	cur->content = xmlStrdup(ent->content);
    if (ent->orig != NULL)
	cur->orig = xmlStrdup(ent->orig);
937
938
    if (ent->URI != NULL)
	cur->URI = xmlStrdup(ent->URI);
939
940
941
942
943
944
945
946
    return(cur);
}

/**
 * xmlCopyEntitiesTable:
 * @table:  An entity table
 *
 * Build a copy of an entity table.
Daniel Veillard's avatar
Daniel Veillard committed
947
 *
948
949
950
951
 * Returns the new xmlEntitiesTablePtr or NULL in case of error.
 */
xmlEntitiesTablePtr
xmlCopyEntitiesTable(xmlEntitiesTablePtr table) {
Nick Wellnhofer's avatar
Nick Wellnhofer committed
952
    return(xmlHashCopy(table, xmlCopyEntity));
953
}
954
#endif /* LIBXML_TREE_ENABLED */
955

956
#ifdef LIBXML_OUTPUT_ENABLED
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998

/**
 * xmlDumpEntityContent:
 * @buf:  An XML buffer.
 * @content:  The entity content.
 *
 * This will dump the quoted string value, taking care of the special
 * treatment required by %
 */
static void
xmlDumpEntityContent(xmlBufferPtr buf, const xmlChar *content) {
    if (buf->alloc == XML_BUFFER_ALLOC_IMMUTABLE) return;
    if (xmlStrchr(content, '%')) {
        const xmlChar * base, *cur;

	xmlBufferCCat(buf, "\"");
	base = cur = content;
	while (*cur != 0) {
	    if (*cur == '"') {
		if (base != cur)
		    xmlBufferAdd(buf, base, cur - base);
		xmlBufferAdd(buf, BAD_CAST "&quot;", 6);
		cur++;
		base = cur;
	    } else if (*cur == '%') {
		if (base != cur)
		    xmlBufferAdd(buf, base, cur - base);
		xmlBufferAdd(buf, BAD_CAST "&#x25;", 6);
		cur++;
		base = cur;
	    } else {
		cur++;
	    }
	}
	if (base != cur)
	    xmlBufferAdd(buf, base, cur - base);
	xmlBufferCCat(buf, "\"");
    } else {
        xmlBufferWriteQuotedString(buf, content);
    }
}

999
1000
1001
1002
1003
1004
1005
1006
1007
/**
 * xmlDumpEntityDecl:
 * @buf:  An XML buffer.
 * @ent:  An entity table
 *
 * This will dump the content of the entity table as an XML DTD definition
 */
void
xmlDumpEntityDecl(xmlBufferPtr buf, xmlEntityPtr ent) {
1008
    if ((buf == NULL) || (ent == NULL)) return;
1009
1010
1011
1012
1013
1014
1015
1016
    switch (ent->etype) {
	case XML_INTERNAL_GENERAL_ENTITY:
	    xmlBufferWriteChar(buf, "<!ENTITY ");
	    xmlBufferWriteCHAR(buf, ent->name);
	    xmlBufferWriteChar(buf, " ");
	    if (ent->orig != NULL)
		xmlBufferWriteQuotedString(buf, ent->orig);
	    else
1017
		xmlDumpEntityContent(buf, ent->content);
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
	    xmlBufferWriteChar(buf, ">\n");
	    break;
	case XML_EXTERNAL_GENERAL_PARSED_ENTITY:
	    xmlBufferWriteChar(buf, "<!ENTITY ");
	    xmlBufferWriteCHAR(buf, ent->name);
	    if (ent->ExternalID != NULL) {
		 xmlBufferWriteChar(buf, " PUBLIC ");
		 xmlBufferWriteQuotedString(buf, ent->ExternalID);
		 xmlBufferWriteChar(buf, " ");
		 xmlBufferWriteQuotedString(buf, ent->SystemID);
	    } else {
		 xmlBufferWriteChar(buf, " SYSTEM ");
		 xmlBufferWriteQuotedString(buf, ent->SystemID);
	    }
	    xmlBufferWriteChar(buf, ">\n");
	    break;
	case XML_EXTERNAL_GENERAL_UNPARSED_ENTITY:
	    xmlBufferWriteChar(buf, "<!ENTITY ");
	    xmlBufferWriteCHAR(buf, ent->name);
	    if (ent->ExternalID != NULL) {
		 xmlBufferWriteChar(buf, " PUBLIC ");
		 xmlBufferWriteQuotedString(buf, ent->ExternalID);
		 xmlBufferWriteChar(buf, " ");
		 xmlBufferWriteQuotedString(buf, ent->SystemID);
	    } else {
		 xmlBufferWriteChar(buf, " SYSTEM ");
		 xmlBufferWriteQuotedString(buf, ent->SystemID);
	    }
	    if (ent->content != NULL) { /* Should be true ! */
		xmlBufferWriteChar(buf, " NDATA ");
		if (ent->orig != NULL)
		    xmlBufferWriteCHAR(buf, ent->orig);
		else
		    xmlBufferWriteCHAR(buf, ent->content);
	    }
	    xmlBufferWriteChar(buf, ">\n");
	    break;
	case XML_INTERNAL_PARAMETER_ENTITY:
	    xmlBufferWriteChar(buf, "<!ENTITY % ");
	    xmlBufferWriteCHAR(buf, ent->name);
	    xmlBufferWriteChar(buf, " ");
	    if (ent->orig == NULL)
1060
		xmlDumpEntityContent(buf, ent->content);
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
	    else
		xmlBufferWriteQuotedString(buf, ent->orig);
	    xmlBufferWriteChar(buf, ">\n");
	    break;
	case XML_EXTERNAL_PARAMETER_ENTITY:
	    xmlBufferWriteChar(buf, "<!ENTITY % ");
	    xmlBufferWriteCHAR(buf, ent->name);
	    if (ent->ExternalID != NULL) {
		 xmlBufferWriteChar(buf, " PUBLIC ");
		 xmlBufferWriteQuotedString(buf, ent->ExternalID);
		 xmlBufferWriteChar(buf, " ");
		 xmlBufferWriteQuotedString(buf, ent->SystemID);
	    } else {
		 xmlBufferWriteChar(buf, " SYSTEM ");
		 xmlBufferWriteQuotedString(buf, ent->SystemID);
	    }
	    xmlBufferWriteChar(buf, ">\n");
	    break;
	default:
1080
1081
	    xmlEntitiesErr(XML_DTD_UNKNOWN_ENTITY,
		"xmlDumpEntitiesDecl: internal: unknown type entity type");
1082
1083
1084
    }
}

1085
1086
1087
1088
1089
1090
1091
1092
/**
 * xmlDumpEntityDeclScan:
 * @ent:  An entity table
 * @buf:  An XML buffer.
 *
 * When using the hash table scan function, arguments need to be reversed
 */
static void
Nick Wellnhofer's avatar
Nick Wellnhofer committed
1093
1094
1095
xmlDumpEntityDeclScan(void *ent, void *buf,
                      const xmlChar *name ATTRIBUTE_UNUSED) {
    xmlDumpEntityDecl((xmlBufferPtr) buf, (xmlEntityPtr) ent);
1096
}
Daniel Veillard's avatar
Daniel Veillard committed
1097

1098
1099
1100
1101
1102
1103
1104
1105
1106
/**
 * xmlDumpEntitiesTable:
 * @buf:  An XML buffer.
 * @table:  An entity table
 *
 * This will dump the content of the entity table as an XML DTD definition
 */
void
xmlDumpEntitiesTable(xmlBufferPtr buf, xmlEntitiesTablePtr table) {
Nick Wellnhofer's avatar
Nick Wellnhofer committed
1107
    xmlHashScan(table, xmlDumpEntityDeclScan, buf);
1108
}
1109
#endif /* LIBXML_OUTPUT_ENABLED */
1110
1111
#define bottom_entities
#include "elfgcchack.h"