1 /*
2 * HTMLtree.c : implementation of access function for an HTML tree.
3 *
4 * See Copyright for the status of this software.
5 *
6 * daniel@veillard.com
7 */
8
9
10 #define IN_LIBXML
11 #include "libxml.h"
12 #ifdef LIBXML_HTML_ENABLED
13
14 #include <string.h> /* for memset() only ! */
15
16 #ifdef HAVE_CTYPE_H
17 #include <ctype.h>
18 #endif
19 #ifdef HAVE_STDLIB_H
20 #include <stdlib.h>
21 #endif
22
23 #include <libxml/xmlmemory.h>
24 #include <libxml/HTMLparser.h>
25 #include <libxml/HTMLtree.h>
26 #include <libxml/entities.h>
27 #include <libxml/valid.h>
28 #include <libxml/xmlerror.h>
29 #include <libxml/parserInternals.h>
30 #include <libxml/globals.h>
31 #include <libxml/uri.h>
32
33 #include "buf.h"
34
35 /************************************************************************
36 * *
37 * Getting/Setting encoding meta tags *
38 * *
39 ************************************************************************/
40
41 /**
42 * htmlGetMetaEncoding:
43 * @doc: the document
44 *
45 * Encoding definition lookup in the Meta tags
46 *
47 * Returns the current encoding as flagged in the HTML source
48 */
49 const xmlChar *
htmlGetMetaEncoding(htmlDocPtr doc)50 htmlGetMetaEncoding(htmlDocPtr doc) {
51 htmlNodePtr cur;
52 const xmlChar *content;
53 const xmlChar *encoding;
54
55 if (doc == NULL)
56 return(NULL);
57 cur = doc->children;
58
59 /*
60 * Search the html
61 */
62 while (cur != NULL) {
63 if ((cur->type == XML_ELEMENT_NODE) && (cur->name != NULL)) {
64 if (xmlStrEqual(cur->name, BAD_CAST"html"))
65 break;
66 if (xmlStrEqual(cur->name, BAD_CAST"head"))
67 goto found_head;
68 if (xmlStrEqual(cur->name, BAD_CAST"meta"))
69 goto found_meta;
70 }
71 cur = cur->next;
72 }
73 if (cur == NULL)
74 return(NULL);
75 cur = cur->children;
76
77 /*
78 * Search the head
79 */
80 while (cur != NULL) {
81 if ((cur->type == XML_ELEMENT_NODE) && (cur->name != NULL)) {
82 if (xmlStrEqual(cur->name, BAD_CAST"head"))
83 break;
84 if (xmlStrEqual(cur->name, BAD_CAST"meta"))
85 goto found_meta;
86 }
87 cur = cur->next;
88 }
89 if (cur == NULL)
90 return(NULL);
91 found_head:
92 cur = cur->children;
93
94 /*
95 * Search the meta elements
96 */
97 found_meta:
98 while (cur != NULL) {
99 if ((cur->type == XML_ELEMENT_NODE) && (cur->name != NULL)) {
100 if (xmlStrEqual(cur->name, BAD_CAST"meta")) {
101 xmlAttrPtr attr = cur->properties;
102 int http;
103 const xmlChar *value;
104
105 content = NULL;
106 http = 0;
107 while (attr != NULL) {
108 if ((attr->children != NULL) &&
109 (attr->children->type == XML_TEXT_NODE) &&
110 (attr->children->next == NULL)) {
111 value = attr->children->content;
112 if ((!xmlStrcasecmp(attr->name, BAD_CAST"http-equiv"))
113 && (!xmlStrcasecmp(value, BAD_CAST"Content-Type")))
114 http = 1;
115 else if ((value != NULL)
116 && (!xmlStrcasecmp(attr->name, BAD_CAST"content")))
117 content = value;
118 if ((http != 0) && (content != NULL))
119 goto found_content;
120 }
121 attr = attr->next;
122 }
123 }
124 }
125 cur = cur->next;
126 }
127 return(NULL);
128
129 found_content:
130 encoding = xmlStrstr(content, BAD_CAST"charset=");
131 if (encoding == NULL)
132 encoding = xmlStrstr(content, BAD_CAST"Charset=");
133 if (encoding == NULL)
134 encoding = xmlStrstr(content, BAD_CAST"CHARSET=");
135 if (encoding != NULL) {
136 encoding += 8;
137 } else {
138 encoding = xmlStrstr(content, BAD_CAST"charset =");
139 if (encoding == NULL)
140 encoding = xmlStrstr(content, BAD_CAST"Charset =");
141 if (encoding == NULL)
142 encoding = xmlStrstr(content, BAD_CAST"CHARSET =");
143 if (encoding != NULL)
144 encoding += 9;
145 }
146 if (encoding != NULL) {
147 while ((*encoding == ' ') || (*encoding == '\t')) encoding++;
148 }
149 return(encoding);
150 }
151
152 /**
153 * htmlSetMetaEncoding:
154 * @doc: the document
155 * @encoding: the encoding string
156 *
157 * Sets the current encoding in the Meta tags
158 * NOTE: this will not change the document content encoding, just
159 * the META flag associated.
160 *
161 * Returns 0 in case of success and -1 in case of error
162 */
163 int
htmlSetMetaEncoding(htmlDocPtr doc,const xmlChar * encoding)164 htmlSetMetaEncoding(htmlDocPtr doc, const xmlChar *encoding) {
165 htmlNodePtr cur, meta = NULL, head = NULL;
166 const xmlChar *content = NULL;
167 char newcontent[100];
168
169 newcontent[0] = 0;
170
171 if (doc == NULL)
172 return(-1);
173
174 /* html isn't a real encoding it's just libxml2 way to get entities */
175 if (!xmlStrcasecmp(encoding, BAD_CAST "html"))
176 return(-1);
177
178 if (encoding != NULL) {
179 snprintf(newcontent, sizeof(newcontent), "text/html; charset=%s",
180 (char *)encoding);
181 newcontent[sizeof(newcontent) - 1] = 0;
182 }
183
184 cur = doc->children;
185
186 /*
187 * Search the html
188 */
189 while (cur != NULL) {
190 if ((cur->type == XML_ELEMENT_NODE) && (cur->name != NULL)) {
191 if (xmlStrcasecmp(cur->name, BAD_CAST"html") == 0)
192 break;
193 if (xmlStrcasecmp(cur->name, BAD_CAST"head") == 0)
194 goto found_head;
195 if (xmlStrcasecmp(cur->name, BAD_CAST"meta") == 0)
196 goto found_meta;
197 }
198 cur = cur->next;
199 }
200 if (cur == NULL)
201 return(-1);
202 cur = cur->children;
203
204 /*
205 * Search the head
206 */
207 while (cur != NULL) {
208 if ((cur->type == XML_ELEMENT_NODE) && (cur->name != NULL)) {
209 if (xmlStrcasecmp(cur->name, BAD_CAST"head") == 0)
210 break;
211 if (xmlStrcasecmp(cur->name, BAD_CAST"meta") == 0) {
212 head = cur->parent;
213 goto found_meta;
214 }
215 }
216 cur = cur->next;
217 }
218 if (cur == NULL)
219 return(-1);
220 found_head:
221 head = cur;
222 if (cur->children == NULL)
223 goto create;
224 cur = cur->children;
225
226 found_meta:
227 /*
228 * Search and update all the remaining the meta elements carrying
229 * encoding informations
230 */
231 while (cur != NULL) {
232 if ((cur->type == XML_ELEMENT_NODE) && (cur->name != NULL)) {
233 if (xmlStrcasecmp(cur->name, BAD_CAST"meta") == 0) {
234 xmlAttrPtr attr = cur->properties;
235 int http;
236 const xmlChar *value;
237
238 content = NULL;
239 http = 0;
240 while (attr != NULL) {
241 if ((attr->children != NULL) &&
242 (attr->children->type == XML_TEXT_NODE) &&
243 (attr->children->next == NULL)) {
244 value = attr->children->content;
245 if ((!xmlStrcasecmp(attr->name, BAD_CAST"http-equiv"))
246 && (!xmlStrcasecmp(value, BAD_CAST"Content-Type")))
247 http = 1;
248 else
249 {
250 if ((value != NULL) &&
251 (!xmlStrcasecmp(attr->name, BAD_CAST"content")))
252 content = value;
253 }
254 if ((http != 0) && (content != NULL))
255 break;
256 }
257 attr = attr->next;
258 }
259 if ((http != 0) && (content != NULL)) {
260 meta = cur;
261 break;
262 }
263
264 }
265 }
266 cur = cur->next;
267 }
268 create:
269 if (meta == NULL) {
270 if ((encoding != NULL) && (head != NULL)) {
271 /*
272 * Create a new Meta element with the right attributes
273 */
274
275 meta = xmlNewDocNode(doc, NULL, BAD_CAST"meta", NULL);
276 if (head->children == NULL)
277 xmlAddChild(head, meta);
278 else
279 xmlAddPrevSibling(head->children, meta);
280 xmlNewProp(meta, BAD_CAST"http-equiv", BAD_CAST"Content-Type");
281 xmlNewProp(meta, BAD_CAST"content", BAD_CAST newcontent);
282 }
283 } else {
284 /* remove the meta tag if NULL is passed */
285 if (encoding == NULL) {
286 xmlUnlinkNode(meta);
287 xmlFreeNode(meta);
288 }
289 /* change the document only if there is a real encoding change */
290 else if (xmlStrcasestr(content, encoding) == NULL) {
291 xmlSetProp(meta, BAD_CAST"content", BAD_CAST newcontent);
292 }
293 }
294
295
296 return(0);
297 }
298
299 /**
300 * booleanHTMLAttrs:
301 *
302 * These are the HTML attributes which will be output
303 * in minimized form, i.e. <option selected="selected"> will be
304 * output as <option selected>, as per XSLT 1.0 16.2 "HTML Output Method"
305 *
306 */
307 static const char* htmlBooleanAttrs[] = {
308 "checked", "compact", "declare", "defer", "disabled", "ismap",
309 "multiple", "nohref", "noresize", "noshade", "nowrap", "readonly",
310 "selected", NULL
311 };
312
313
314 /**
315 * htmlIsBooleanAttr:
316 * @name: the name of the attribute to check
317 *
318 * Determine if a given attribute is a boolean attribute.
319 *
320 * returns: false if the attribute is not boolean, true otherwise.
321 */
322 int
htmlIsBooleanAttr(const xmlChar * name)323 htmlIsBooleanAttr(const xmlChar *name)
324 {
325 int i = 0;
326
327 while (htmlBooleanAttrs[i] != NULL) {
328 if (xmlStrcasecmp((const xmlChar *)htmlBooleanAttrs[i], name) == 0)
329 return 1;
330 i++;
331 }
332 return 0;
333 }
334
335 #ifdef LIBXML_OUTPUT_ENABLED
336 /*
337 * private routine exported from xmlIO.c
338 */
339 xmlOutputBufferPtr
340 xmlAllocOutputBufferInternal(xmlCharEncodingHandlerPtr encoder);
341 /************************************************************************
342 * *
343 * Output error handlers *
344 * *
345 ************************************************************************/
346 /**
347 * htmlSaveErrMemory:
348 * @extra: extra informations
349 *
350 * Handle an out of memory condition
351 */
352 static void
htmlSaveErrMemory(const char * extra)353 htmlSaveErrMemory(const char *extra)
354 {
355 __xmlSimpleError(XML_FROM_OUTPUT, XML_ERR_NO_MEMORY, NULL, NULL, extra);
356 }
357
358 /**
359 * htmlSaveErr:
360 * @code: the error number
361 * @node: the location of the error.
362 * @extra: extra informations
363 *
364 * Handle an out of memory condition
365 */
366 static void
htmlSaveErr(int code,xmlNodePtr node,const char * extra)367 htmlSaveErr(int code, xmlNodePtr node, const char *extra)
368 {
369 const char *msg = NULL;
370
371 switch(code) {
372 case XML_SAVE_NOT_UTF8:
373 msg = "string is not in UTF-8\n";
374 break;
375 case XML_SAVE_CHAR_INVALID:
376 msg = "invalid character value\n";
377 break;
378 case XML_SAVE_UNKNOWN_ENCODING:
379 msg = "unknown encoding %s\n";
380 break;
381 case XML_SAVE_NO_DOCTYPE:
382 msg = "HTML has no DOCTYPE\n";
383 break;
384 default:
385 msg = "unexpected error number\n";
386 }
387 __xmlSimpleError(XML_FROM_OUTPUT, code, node, msg, extra);
388 }
389
390 /************************************************************************
391 * *
392 * Dumping HTML tree content to a simple buffer *
393 * *
394 ************************************************************************/
395
396 /**
397 * htmlBufNodeDumpFormat:
398 * @buf: the xmlBufPtr output
399 * @doc: the document
400 * @cur: the current node
401 * @format: should formatting spaces been added
402 *
403 * Dump an HTML node, recursive behaviour,children are printed too.
404 *
405 * Returns the number of byte written or -1 in case of error
406 */
407 static size_t
htmlBufNodeDumpFormat(xmlBufPtr buf,xmlDocPtr doc,xmlNodePtr cur,int format)408 htmlBufNodeDumpFormat(xmlBufPtr buf, xmlDocPtr doc, xmlNodePtr cur,
409 int format) {
410 size_t use;
411 int ret;
412 xmlOutputBufferPtr outbuf;
413
414 if (cur == NULL) {
415 return (-1);
416 }
417 if (buf == NULL) {
418 return (-1);
419 }
420 outbuf = (xmlOutputBufferPtr) xmlMalloc(sizeof(xmlOutputBuffer));
421 if (outbuf == NULL) {
422 htmlSaveErrMemory("allocating HTML output buffer");
423 return (-1);
424 }
425 memset(outbuf, 0, (size_t) sizeof(xmlOutputBuffer));
426 outbuf->buffer = buf;
427 outbuf->encoder = NULL;
428 outbuf->writecallback = NULL;
429 outbuf->closecallback = NULL;
430 outbuf->context = NULL;
431 outbuf->written = 0;
432
433 use = xmlBufUse(buf);
434 htmlNodeDumpFormatOutput(outbuf, doc, cur, NULL, format);
435 xmlFree(outbuf);
436 ret = xmlBufUse(buf) - use;
437 return (ret);
438 }
439
440 /**
441 * htmlNodeDump:
442 * @buf: the HTML buffer output
443 * @doc: the document
444 * @cur: the current node
445 *
446 * Dump an HTML node, recursive behaviour,children are printed too,
447 * and formatting returns are added.
448 *
449 * Returns the number of byte written or -1 in case of error
450 */
451 int
htmlNodeDump(xmlBufferPtr buf,xmlDocPtr doc,xmlNodePtr cur)452 htmlNodeDump(xmlBufferPtr buf, xmlDocPtr doc, xmlNodePtr cur) {
453 xmlBufPtr buffer;
454 size_t ret;
455
456 if ((buf == NULL) || (cur == NULL))
457 return(-1);
458
459 xmlInitParser();
460 buffer = xmlBufFromBuffer(buf);
461 if (buffer == NULL)
462 return(-1);
463
464 ret = htmlBufNodeDumpFormat(buffer, doc, cur, 1);
465
466 xmlBufBackToBuffer(buffer);
467
468 if (ret > INT_MAX)
469 return(-1);
470 return((int) ret);
471 }
472
473 /**
474 * htmlNodeDumpFileFormat:
475 * @out: the FILE pointer
476 * @doc: the document
477 * @cur: the current node
478 * @encoding: the document encoding
479 * @format: should formatting spaces been added
480 *
481 * Dump an HTML node, recursive behaviour,children are printed too.
482 *
483 * TODO: if encoding == NULL try to save in the doc encoding
484 *
485 * returns: the number of byte written or -1 in case of failure.
486 */
487 int
htmlNodeDumpFileFormat(FILE * out,xmlDocPtr doc,xmlNodePtr cur,const char * encoding,int format)488 htmlNodeDumpFileFormat(FILE *out, xmlDocPtr doc,
489 xmlNodePtr cur, const char *encoding, int format) {
490 xmlOutputBufferPtr buf;
491 xmlCharEncodingHandlerPtr handler = NULL;
492 int ret;
493
494 xmlInitParser();
495
496 if (encoding != NULL) {
497 xmlCharEncoding enc;
498
499 enc = xmlParseCharEncoding(encoding);
500 if (enc != XML_CHAR_ENCODING_UTF8) {
501 handler = xmlFindCharEncodingHandler(encoding);
502 if (handler == NULL)
503 htmlSaveErr(XML_SAVE_UNKNOWN_ENCODING, NULL, encoding);
504 }
505 } else {
506 /*
507 * Fallback to HTML or ASCII when the encoding is unspecified
508 */
509 if (handler == NULL)
510 handler = xmlFindCharEncodingHandler("HTML");
511 if (handler == NULL)
512 handler = xmlFindCharEncodingHandler("ascii");
513 }
514
515 /*
516 * save the content to a temp buffer.
517 */
518 buf = xmlOutputBufferCreateFile(out, handler);
519 if (buf == NULL) return(0);
520
521 htmlNodeDumpFormatOutput(buf, doc, cur, encoding, format);
522
523 ret = xmlOutputBufferClose(buf);
524 return(ret);
525 }
526
527 /**
528 * htmlNodeDumpFile:
529 * @out: the FILE pointer
530 * @doc: the document
531 * @cur: the current node
532 *
533 * Dump an HTML node, recursive behaviour,children are printed too,
534 * and formatting returns are added.
535 */
536 void
htmlNodeDumpFile(FILE * out,xmlDocPtr doc,xmlNodePtr cur)537 htmlNodeDumpFile(FILE *out, xmlDocPtr doc, xmlNodePtr cur) {
538 htmlNodeDumpFileFormat(out, doc, cur, NULL, 1);
539 }
540
541 /**
542 * htmlDocDumpMemoryFormat:
543 * @cur: the document
544 * @mem: OUT: the memory pointer
545 * @size: OUT: the memory length
546 * @format: should formatting spaces been added
547 *
548 * Dump an HTML document in memory and return the xmlChar * and it's size.
549 * It's up to the caller to free the memory.
550 */
551 void
htmlDocDumpMemoryFormat(xmlDocPtr cur,xmlChar ** mem,int * size,int format)552 htmlDocDumpMemoryFormat(xmlDocPtr cur, xmlChar**mem, int *size, int format) {
553 xmlOutputBufferPtr buf;
554 xmlCharEncodingHandlerPtr handler = NULL;
555 const char *encoding;
556
557 xmlInitParser();
558
559 if ((mem == NULL) || (size == NULL))
560 return;
561 if (cur == NULL) {
562 *mem = NULL;
563 *size = 0;
564 return;
565 }
566
567 encoding = (const char *) htmlGetMetaEncoding(cur);
568
569 if (encoding != NULL) {
570 xmlCharEncoding enc;
571
572 enc = xmlParseCharEncoding(encoding);
573 if (enc != XML_CHAR_ENCODING_UTF8) {
574 handler = xmlFindCharEncodingHandler(encoding);
575 if (handler == NULL)
576 htmlSaveErr(XML_SAVE_UNKNOWN_ENCODING, NULL, encoding);
577
578 }
579 } else {
580 /*
581 * Fallback to HTML or ASCII when the encoding is unspecified
582 */
583 if (handler == NULL)
584 handler = xmlFindCharEncodingHandler("HTML");
585 if (handler == NULL)
586 handler = xmlFindCharEncodingHandler("ascii");
587 }
588
589 buf = xmlAllocOutputBufferInternal(handler);
590 if (buf == NULL) {
591 *mem = NULL;
592 *size = 0;
593 return;
594 }
595
596 htmlDocContentDumpFormatOutput(buf, cur, NULL, format);
597
598 xmlOutputBufferFlush(buf);
599 if (buf->conv != NULL) {
600 *size = xmlBufUse(buf->conv);
601 *mem = xmlStrndup(xmlBufContent(buf->conv), *size);
602 } else {
603 *size = xmlBufUse(buf->buffer);
604 *mem = xmlStrndup(xmlBufContent(buf->buffer), *size);
605 }
606 (void)xmlOutputBufferClose(buf);
607 }
608
609 /**
610 * htmlDocDumpMemory:
611 * @cur: the document
612 * @mem: OUT: the memory pointer
613 * @size: OUT: the memory length
614 *
615 * Dump an HTML document in memory and return the xmlChar * and it's size.
616 * It's up to the caller to free the memory.
617 */
618 void
htmlDocDumpMemory(xmlDocPtr cur,xmlChar ** mem,int * size)619 htmlDocDumpMemory(xmlDocPtr cur, xmlChar**mem, int *size) {
620 htmlDocDumpMemoryFormat(cur, mem, size, 1);
621 }
622
623
624 /************************************************************************
625 * *
626 * Dumping HTML tree content to an I/O output buffer *
627 * *
628 ************************************************************************/
629
630 void xmlNsListDumpOutput(xmlOutputBufferPtr buf, xmlNsPtr cur);
631
632 /**
633 * htmlDtdDumpOutput:
634 * @buf: the HTML buffer output
635 * @doc: the document
636 * @encoding: the encoding string
637 *
638 * TODO: check whether encoding is needed
639 *
640 * Dump the HTML document DTD, if any.
641 */
642 static void
htmlDtdDumpOutput(xmlOutputBufferPtr buf,xmlDocPtr doc,const char * encoding ATTRIBUTE_UNUSED)643 htmlDtdDumpOutput(xmlOutputBufferPtr buf, xmlDocPtr doc,
644 const char *encoding ATTRIBUTE_UNUSED) {
645 xmlDtdPtr cur = doc->intSubset;
646
647 if (cur == NULL) {
648 htmlSaveErr(XML_SAVE_NO_DOCTYPE, (xmlNodePtr) doc, NULL);
649 return;
650 }
651 xmlOutputBufferWriteString(buf, "<!DOCTYPE ");
652 xmlOutputBufferWriteString(buf, (const char *)cur->name);
653 if (cur->ExternalID != NULL) {
654 xmlOutputBufferWriteString(buf, " PUBLIC ");
655 xmlBufWriteQuotedString(buf->buffer, cur->ExternalID);
656 if (cur->SystemID != NULL) {
657 xmlOutputBufferWriteString(buf, " ");
658 xmlBufWriteQuotedString(buf->buffer, cur->SystemID);
659 }
660 } else if (cur->SystemID != NULL &&
661 xmlStrcmp(cur->SystemID, BAD_CAST "about:legacy-compat")) {
662 xmlOutputBufferWriteString(buf, " SYSTEM ");
663 xmlBufWriteQuotedString(buf->buffer, cur->SystemID);
664 }
665 xmlOutputBufferWriteString(buf, ">\n");
666 }
667
668 /**
669 * htmlAttrDumpOutput:
670 * @buf: the HTML buffer output
671 * @doc: the document
672 * @cur: the attribute pointer
673 * @encoding: the encoding string
674 *
675 * Dump an HTML attribute
676 */
677 static void
htmlAttrDumpOutput(xmlOutputBufferPtr buf,xmlDocPtr doc,xmlAttrPtr cur,const char * encoding ATTRIBUTE_UNUSED)678 htmlAttrDumpOutput(xmlOutputBufferPtr buf, xmlDocPtr doc, xmlAttrPtr cur,
679 const char *encoding ATTRIBUTE_UNUSED) {
680 xmlChar *value;
681
682 /*
683 * The html output method should not escape a & character
684 * occurring in an attribute value immediately followed by
685 * a { character (see Section B.7.1 of the HTML 4.0 Recommendation).
686 * This is implemented in xmlEncodeEntitiesReentrant
687 */
688
689 if (cur == NULL) {
690 return;
691 }
692 xmlOutputBufferWriteString(buf, " ");
693 if ((cur->ns != NULL) && (cur->ns->prefix != NULL)) {
694 xmlOutputBufferWriteString(buf, (const char *)cur->ns->prefix);
695 xmlOutputBufferWriteString(buf, ":");
696 }
697 xmlOutputBufferWriteString(buf, (const char *)cur->name);
698 if ((cur->children != NULL) && (!htmlIsBooleanAttr(cur->name))) {
699 value = xmlNodeListGetString(doc, cur->children, 0);
700 if (value) {
701 xmlOutputBufferWriteString(buf, "=");
702 if ((cur->ns == NULL) && (cur->parent != NULL) &&
703 (cur->parent->ns == NULL) &&
704 ((!xmlStrcasecmp(cur->name, BAD_CAST "href")) ||
705 (!xmlStrcasecmp(cur->name, BAD_CAST "action")) ||
706 (!xmlStrcasecmp(cur->name, BAD_CAST "src")) ||
707 ((!xmlStrcasecmp(cur->name, BAD_CAST "name")) &&
708 (!xmlStrcasecmp(cur->parent->name, BAD_CAST "a"))))) {
709 xmlChar *tmp = value;
710 /* xmlURIEscapeStr() escapes '"' so it can be safely used. */
711 xmlBufCCat(buf->buffer, "\"");
712
713 while (IS_BLANK_CH(*tmp)) tmp++;
714
715 /* URI Escape everything, except server side includes. */
716 for ( ; ; ) {
717 xmlChar *escaped;
718 xmlChar endChar;
719 xmlChar *end = NULL;
720 xmlChar *start = (xmlChar *)xmlStrstr(tmp, BAD_CAST "<!--");
721 if (start != NULL) {
722 end = (xmlChar *)xmlStrstr(tmp, BAD_CAST "-->");
723 if (end != NULL) {
724 *start = '\0';
725 }
726 }
727
728 /* Escape the whole string, or until start (set to '\0'). */
729 escaped = xmlURIEscapeStr(tmp, BAD_CAST"@/:=?;#%&,+");
730 if (escaped != NULL) {
731 xmlBufCat(buf->buffer, escaped);
732 xmlFree(escaped);
733 } else {
734 xmlBufCat(buf->buffer, tmp);
735 }
736
737 if (end == NULL) { /* Everything has been written. */
738 break;
739 }
740
741 /* Do not escape anything within server side includes. */
742 *start = '<'; /* Restore the first character of "<!--". */
743 end += 3; /* strlen("-->") */
744 endChar = *end;
745 *end = '\0';
746 xmlBufCat(buf->buffer, start);
747 *end = endChar;
748 tmp = end;
749 }
750
751 xmlBufCCat(buf->buffer, "\"");
752 } else {
753 xmlBufWriteQuotedString(buf->buffer, value);
754 }
755 xmlFree(value);
756 } else {
757 xmlOutputBufferWriteString(buf, "=\"\"");
758 }
759 }
760 }
761
762 /**
763 * htmlAttrListDumpOutput:
764 * @buf: the HTML buffer output
765 * @doc: the document
766 * @cur: the first attribute pointer
767 * @encoding: the encoding string
768 *
769 * Dump a list of HTML attributes
770 */
771 static void
htmlAttrListDumpOutput(xmlOutputBufferPtr buf,xmlDocPtr doc,xmlAttrPtr cur,const char * encoding)772 htmlAttrListDumpOutput(xmlOutputBufferPtr buf, xmlDocPtr doc, xmlAttrPtr cur, const char *encoding) {
773 if (cur == NULL) {
774 return;
775 }
776 while (cur != NULL) {
777 htmlAttrDumpOutput(buf, doc, cur, encoding);
778 cur = cur->next;
779 }
780 }
781
782
783
784 /**
785 * htmlNodeListDumpOutput:
786 * @buf: the HTML buffer output
787 * @doc: the document
788 * @cur: the first node
789 * @encoding: the encoding string
790 * @format: should formatting spaces been added
791 *
792 * Dump an HTML node list, recursive behaviour,children are printed too.
793 */
794 static void
htmlNodeListDumpOutput(xmlOutputBufferPtr buf,xmlDocPtr doc,xmlNodePtr cur,const char * encoding,int format)795 htmlNodeListDumpOutput(xmlOutputBufferPtr buf, xmlDocPtr doc,
796 xmlNodePtr cur, const char *encoding, int format) {
797 if (cur == NULL) {
798 return;
799 }
800 while (cur != NULL) {
801 htmlNodeDumpFormatOutput(buf, doc, cur, encoding, format);
802 cur = cur->next;
803 }
804 }
805
806 /**
807 * htmlNodeDumpFormatOutput:
808 * @buf: the HTML buffer output
809 * @doc: the document
810 * @cur: the current node
811 * @encoding: the encoding string
812 * @format: should formatting spaces been added
813 *
814 * Dump an HTML node, recursive behaviour,children are printed too.
815 */
816 void
htmlNodeDumpFormatOutput(xmlOutputBufferPtr buf,xmlDocPtr doc,xmlNodePtr cur,const char * encoding,int format)817 htmlNodeDumpFormatOutput(xmlOutputBufferPtr buf, xmlDocPtr doc,
818 xmlNodePtr cur, const char *encoding, int format) {
819 const htmlElemDesc * info;
820
821 xmlInitParser();
822
823 if ((cur == NULL) || (buf == NULL)) {
824 return;
825 }
826 /*
827 * Special cases.
828 */
829 if (cur->type == XML_DTD_NODE)
830 return;
831 if ((cur->type == XML_HTML_DOCUMENT_NODE) ||
832 (cur->type == XML_DOCUMENT_NODE)){
833 htmlDocContentDumpOutput(buf, (xmlDocPtr) cur, encoding);
834 return;
835 }
836 if (cur->type == XML_ATTRIBUTE_NODE) {
837 htmlAttrDumpOutput(buf, doc, (xmlAttrPtr) cur, encoding);
838 return;
839 }
840 if (cur->type == HTML_TEXT_NODE) {
841 if (cur->content != NULL) {
842 if (((cur->name == (const xmlChar *)xmlStringText) ||
843 (cur->name != (const xmlChar *)xmlStringTextNoenc)) &&
844 ((cur->parent == NULL) ||
845 ((xmlStrcasecmp(cur->parent->name, BAD_CAST "script")) &&
846 (xmlStrcasecmp(cur->parent->name, BAD_CAST "style"))))) {
847 xmlChar *buffer;
848
849 buffer = xmlEncodeEntitiesReentrant(doc, cur->content);
850 if (buffer != NULL) {
851 xmlOutputBufferWriteString(buf, (const char *)buffer);
852 xmlFree(buffer);
853 }
854 } else {
855 xmlOutputBufferWriteString(buf, (const char *)cur->content);
856 }
857 }
858 return;
859 }
860 if (cur->type == HTML_COMMENT_NODE) {
861 if (cur->content != NULL) {
862 xmlOutputBufferWriteString(buf, "<!--");
863 xmlOutputBufferWriteString(buf, (const char *)cur->content);
864 xmlOutputBufferWriteString(buf, "-->");
865 }
866 return;
867 }
868 if (cur->type == HTML_PI_NODE) {
869 if (cur->name == NULL)
870 return;
871 xmlOutputBufferWriteString(buf, "<?");
872 xmlOutputBufferWriteString(buf, (const char *)cur->name);
873 if (cur->content != NULL) {
874 xmlOutputBufferWriteString(buf, " ");
875 xmlOutputBufferWriteString(buf, (const char *)cur->content);
876 }
877 xmlOutputBufferWriteString(buf, ">");
878 return;
879 }
880 if (cur->type == HTML_ENTITY_REF_NODE) {
881 xmlOutputBufferWriteString(buf, "&");
882 xmlOutputBufferWriteString(buf, (const char *)cur->name);
883 xmlOutputBufferWriteString(buf, ";");
884 return;
885 }
886 if (cur->type == HTML_PRESERVE_NODE) {
887 if (cur->content != NULL) {
888 xmlOutputBufferWriteString(buf, (const char *)cur->content);
889 }
890 return;
891 }
892
893 /*
894 * Get specific HTML info for that node.
895 */
896 if (cur->ns == NULL)
897 info = htmlTagLookup(cur->name);
898 else
899 info = NULL;
900
901 xmlOutputBufferWriteString(buf, "<");
902 if ((cur->ns != NULL) && (cur->ns->prefix != NULL)) {
903 xmlOutputBufferWriteString(buf, (const char *)cur->ns->prefix);
904 xmlOutputBufferWriteString(buf, ":");
905 }
906 xmlOutputBufferWriteString(buf, (const char *)cur->name);
907 if (cur->nsDef)
908 xmlNsListDumpOutput(buf, cur->nsDef);
909 if (cur->properties != NULL)
910 htmlAttrListDumpOutput(buf, doc, cur->properties, encoding);
911
912 if ((info != NULL) && (info->empty)) {
913 xmlOutputBufferWriteString(buf, ">");
914 if ((format) && (!info->isinline) && (cur->next != NULL)) {
915 if ((cur->next->type != HTML_TEXT_NODE) &&
916 (cur->next->type != HTML_ENTITY_REF_NODE) &&
917 (cur->parent != NULL) &&
918 (cur->parent->name != NULL) &&
919 (cur->parent->name[0] != 'p')) /* p, pre, param */
920 xmlOutputBufferWriteString(buf, "\n");
921 }
922 return;
923 }
924 if (((cur->type == XML_ELEMENT_NODE) || (cur->content == NULL)) &&
925 (cur->children == NULL)) {
926 if ((info != NULL) && (info->saveEndTag != 0) &&
927 (xmlStrcmp(BAD_CAST info->name, BAD_CAST "html")) &&
928 (xmlStrcmp(BAD_CAST info->name, BAD_CAST "body"))) {
929 xmlOutputBufferWriteString(buf, ">");
930 } else {
931 xmlOutputBufferWriteString(buf, "></");
932 if ((cur->ns != NULL) && (cur->ns->prefix != NULL)) {
933 xmlOutputBufferWriteString(buf, (const char *)cur->ns->prefix);
934 xmlOutputBufferWriteString(buf, ":");
935 }
936 xmlOutputBufferWriteString(buf, (const char *)cur->name);
937 xmlOutputBufferWriteString(buf, ">");
938 }
939 if ((format) && (cur->next != NULL) &&
940 (info != NULL) && (!info->isinline)) {
941 if ((cur->next->type != HTML_TEXT_NODE) &&
942 (cur->next->type != HTML_ENTITY_REF_NODE) &&
943 (cur->parent != NULL) &&
944 (cur->parent->name != NULL) &&
945 (cur->parent->name[0] != 'p')) /* p, pre, param */
946 xmlOutputBufferWriteString(buf, "\n");
947 }
948 return;
949 }
950 xmlOutputBufferWriteString(buf, ">");
951 if ((cur->type != XML_ELEMENT_NODE) &&
952 (cur->content != NULL)) {
953 /*
954 * Uses the OutputBuffer property to automatically convert
955 * invalids to charrefs
956 */
957
958 xmlOutputBufferWriteString(buf, (const char *) cur->content);
959 }
960 if (cur->children != NULL) {
961 if ((format) && (info != NULL) && (!info->isinline) &&
962 (cur->children->type != HTML_TEXT_NODE) &&
963 (cur->children->type != HTML_ENTITY_REF_NODE) &&
964 (cur->children != cur->last) &&
965 (cur->name != NULL) &&
966 (cur->name[0] != 'p')) /* p, pre, param */
967 xmlOutputBufferWriteString(buf, "\n");
968 htmlNodeListDumpOutput(buf, doc, cur->children, encoding, format);
969 if ((format) && (info != NULL) && (!info->isinline) &&
970 (cur->last->type != HTML_TEXT_NODE) &&
971 (cur->last->type != HTML_ENTITY_REF_NODE) &&
972 (cur->children != cur->last) &&
973 (cur->name != NULL) &&
974 (cur->name[0] != 'p')) /* p, pre, param */
975 xmlOutputBufferWriteString(buf, "\n");
976 }
977 xmlOutputBufferWriteString(buf, "</");
978 if ((cur->ns != NULL) && (cur->ns->prefix != NULL)) {
979 xmlOutputBufferWriteString(buf, (const char *)cur->ns->prefix);
980 xmlOutputBufferWriteString(buf, ":");
981 }
982 xmlOutputBufferWriteString(buf, (const char *)cur->name);
983 xmlOutputBufferWriteString(buf, ">");
984 if ((format) && (info != NULL) && (!info->isinline) &&
985 (cur->next != NULL)) {
986 if ((cur->next->type != HTML_TEXT_NODE) &&
987 (cur->next->type != HTML_ENTITY_REF_NODE) &&
988 (cur->parent != NULL) &&
989 (cur->parent->name != NULL) &&
990 (cur->parent->name[0] != 'p')) /* p, pre, param */
991 xmlOutputBufferWriteString(buf, "\n");
992 }
993 }
994
995 /**
996 * htmlNodeDumpOutput:
997 * @buf: the HTML buffer output
998 * @doc: the document
999 * @cur: the current node
1000 * @encoding: the encoding string
1001 *
1002 * Dump an HTML node, recursive behaviour,children are printed too,
1003 * and formatting returns/spaces are added.
1004 */
1005 void
htmlNodeDumpOutput(xmlOutputBufferPtr buf,xmlDocPtr doc,xmlNodePtr cur,const char * encoding)1006 htmlNodeDumpOutput(xmlOutputBufferPtr buf, xmlDocPtr doc,
1007 xmlNodePtr cur, const char *encoding) {
1008 htmlNodeDumpFormatOutput(buf, doc, cur, encoding, 1);
1009 }
1010
1011 /**
1012 * htmlDocContentDumpFormatOutput:
1013 * @buf: the HTML buffer output
1014 * @cur: the document
1015 * @encoding: the encoding string
1016 * @format: should formatting spaces been added
1017 *
1018 * Dump an HTML document.
1019 */
1020 void
htmlDocContentDumpFormatOutput(xmlOutputBufferPtr buf,xmlDocPtr cur,const char * encoding,int format)1021 htmlDocContentDumpFormatOutput(xmlOutputBufferPtr buf, xmlDocPtr cur,
1022 const char *encoding, int format) {
1023 int type;
1024
1025 xmlInitParser();
1026
1027 if ((buf == NULL) || (cur == NULL))
1028 return;
1029
1030 /*
1031 * force to output the stuff as HTML, especially for entities
1032 */
1033 type = cur->type;
1034 cur->type = XML_HTML_DOCUMENT_NODE;
1035 if (cur->intSubset != NULL) {
1036 htmlDtdDumpOutput(buf, cur, NULL);
1037 }
1038 if (cur->children != NULL) {
1039 htmlNodeListDumpOutput(buf, cur, cur->children, encoding, format);
1040 }
1041 xmlOutputBufferWriteString(buf, "\n");
1042 cur->type = (xmlElementType) type;
1043 }
1044
1045 /**
1046 * htmlDocContentDumpOutput:
1047 * @buf: the HTML buffer output
1048 * @cur: the document
1049 * @encoding: the encoding string
1050 *
1051 * Dump an HTML document. Formating return/spaces are added.
1052 */
1053 void
htmlDocContentDumpOutput(xmlOutputBufferPtr buf,xmlDocPtr cur,const char * encoding)1054 htmlDocContentDumpOutput(xmlOutputBufferPtr buf, xmlDocPtr cur,
1055 const char *encoding) {
1056 htmlDocContentDumpFormatOutput(buf, cur, encoding, 1);
1057 }
1058
1059 /************************************************************************
1060 * *
1061 * Saving functions front-ends *
1062 * *
1063 ************************************************************************/
1064
1065 /**
1066 * htmlDocDump:
1067 * @f: the FILE*
1068 * @cur: the document
1069 *
1070 * Dump an HTML document to an open FILE.
1071 *
1072 * returns: the number of byte written or -1 in case of failure.
1073 */
1074 int
htmlDocDump(FILE * f,xmlDocPtr cur)1075 htmlDocDump(FILE *f, xmlDocPtr cur) {
1076 xmlOutputBufferPtr buf;
1077 xmlCharEncodingHandlerPtr handler = NULL;
1078 const char *encoding;
1079 int ret;
1080
1081 xmlInitParser();
1082
1083 if ((cur == NULL) || (f == NULL)) {
1084 return(-1);
1085 }
1086
1087 encoding = (const char *) htmlGetMetaEncoding(cur);
1088
1089 if (encoding != NULL) {
1090 xmlCharEncoding enc;
1091
1092 enc = xmlParseCharEncoding(encoding);
1093 if (enc != XML_CHAR_ENCODING_UTF8) {
1094 handler = xmlFindCharEncodingHandler(encoding);
1095 if (handler == NULL)
1096 htmlSaveErr(XML_SAVE_UNKNOWN_ENCODING, NULL, encoding);
1097 }
1098 } else {
1099 /*
1100 * Fallback to HTML or ASCII when the encoding is unspecified
1101 */
1102 if (handler == NULL)
1103 handler = xmlFindCharEncodingHandler("HTML");
1104 if (handler == NULL)
1105 handler = xmlFindCharEncodingHandler("ascii");
1106 }
1107
1108 buf = xmlOutputBufferCreateFile(f, handler);
1109 if (buf == NULL) return(-1);
1110 htmlDocContentDumpOutput(buf, cur, NULL);
1111
1112 ret = xmlOutputBufferClose(buf);
1113 return(ret);
1114 }
1115
1116 /**
1117 * htmlSaveFile:
1118 * @filename: the filename (or URL)
1119 * @cur: the document
1120 *
1121 * Dump an HTML document to a file. If @filename is "-" the stdout file is
1122 * used.
1123 * returns: the number of byte written or -1 in case of failure.
1124 */
1125 int
htmlSaveFile(const char * filename,xmlDocPtr cur)1126 htmlSaveFile(const char *filename, xmlDocPtr cur) {
1127 xmlOutputBufferPtr buf;
1128 xmlCharEncodingHandlerPtr handler = NULL;
1129 const char *encoding;
1130 int ret;
1131
1132 if ((cur == NULL) || (filename == NULL))
1133 return(-1);
1134
1135 xmlInitParser();
1136
1137 encoding = (const char *) htmlGetMetaEncoding(cur);
1138
1139 if (encoding != NULL) {
1140 xmlCharEncoding enc;
1141
1142 enc = xmlParseCharEncoding(encoding);
1143 if (enc != XML_CHAR_ENCODING_UTF8) {
1144 handler = xmlFindCharEncodingHandler(encoding);
1145 if (handler == NULL)
1146 htmlSaveErr(XML_SAVE_UNKNOWN_ENCODING, NULL, encoding);
1147 }
1148 } else {
1149 /*
1150 * Fallback to HTML or ASCII when the encoding is unspecified
1151 */
1152 if (handler == NULL)
1153 handler = xmlFindCharEncodingHandler("HTML");
1154 if (handler == NULL)
1155 handler = xmlFindCharEncodingHandler("ascii");
1156 }
1157
1158 /*
1159 * save the content to a temp buffer.
1160 */
1161 buf = xmlOutputBufferCreateFilename(filename, handler, cur->compression);
1162 if (buf == NULL) return(0);
1163
1164 htmlDocContentDumpOutput(buf, cur, NULL);
1165
1166 ret = xmlOutputBufferClose(buf);
1167 return(ret);
1168 }
1169
1170 /**
1171 * htmlSaveFileFormat:
1172 * @filename: the filename
1173 * @cur: the document
1174 * @format: should formatting spaces been added
1175 * @encoding: the document encoding
1176 *
1177 * Dump an HTML document to a file using a given encoding.
1178 *
1179 * returns: the number of byte written or -1 in case of failure.
1180 */
1181 int
htmlSaveFileFormat(const char * filename,xmlDocPtr cur,const char * encoding,int format)1182 htmlSaveFileFormat(const char *filename, xmlDocPtr cur,
1183 const char *encoding, int format) {
1184 xmlOutputBufferPtr buf;
1185 xmlCharEncodingHandlerPtr handler = NULL;
1186 int ret;
1187
1188 if ((cur == NULL) || (filename == NULL))
1189 return(-1);
1190
1191 xmlInitParser();
1192
1193 if (encoding != NULL) {
1194 xmlCharEncoding enc;
1195
1196 enc = xmlParseCharEncoding(encoding);
1197 if (enc != XML_CHAR_ENCODING_UTF8) {
1198 handler = xmlFindCharEncodingHandler(encoding);
1199 if (handler == NULL)
1200 htmlSaveErr(XML_SAVE_UNKNOWN_ENCODING, NULL, encoding);
1201 }
1202 htmlSetMetaEncoding(cur, (const xmlChar *) encoding);
1203 } else {
1204 htmlSetMetaEncoding(cur, (const xmlChar *) "UTF-8");
1205
1206 /*
1207 * Fallback to HTML or ASCII when the encoding is unspecified
1208 */
1209 if (handler == NULL)
1210 handler = xmlFindCharEncodingHandler("HTML");
1211 if (handler == NULL)
1212 handler = xmlFindCharEncodingHandler("ascii");
1213 }
1214
1215 /*
1216 * save the content to a temp buffer.
1217 */
1218 buf = xmlOutputBufferCreateFilename(filename, handler, 0);
1219 if (buf == NULL) return(0);
1220
1221 htmlDocContentDumpFormatOutput(buf, cur, encoding, format);
1222
1223 ret = xmlOutputBufferClose(buf);
1224 return(ret);
1225 }
1226
1227 /**
1228 * htmlSaveFileEnc:
1229 * @filename: the filename
1230 * @cur: the document
1231 * @encoding: the document encoding
1232 *
1233 * Dump an HTML document to a file using a given encoding
1234 * and formatting returns/spaces are added.
1235 *
1236 * returns: the number of byte written or -1 in case of failure.
1237 */
1238 int
htmlSaveFileEnc(const char * filename,xmlDocPtr cur,const char * encoding)1239 htmlSaveFileEnc(const char *filename, xmlDocPtr cur, const char *encoding) {
1240 return(htmlSaveFileFormat(filename, cur, encoding, 1));
1241 }
1242
1243 #endif /* LIBXML_OUTPUT_ENABLED */
1244
1245 #define bottom_HTMLtree
1246 #include "elfgcchack.h"
1247 #endif /* LIBXML_HTML_ENABLED */
1248