2 * HTMLtree.c : implementation of access function for an HTML tree.
4 * See Copyright for the status of this software.
12 #ifdef LIBXML_HTML_ENABLED
14 #include <string.h> /* for memset() only ! */
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>
35 /************************************************************************
37 * Getting/Setting encoding meta tags *
39 ************************************************************************/
42 * htmlGetMetaEncoding:
45 * Encoding definition lookup in the Meta tags
47 * Returns the current encoding as flagged in the HTML source
50 htmlGetMetaEncoding(htmlDocPtr doc
) {
52 const xmlChar
*content
;
53 const xmlChar
*encoding
;
63 if ((cur
->type
== XML_ELEMENT_NODE
) && (cur
->name
!= NULL
)) {
64 if (xmlStrEqual(cur
->name
, BAD_CAST
"html"))
66 if (xmlStrEqual(cur
->name
, BAD_CAST
"head"))
68 if (xmlStrEqual(cur
->name
, BAD_CAST
"meta"))
81 if ((cur
->type
== XML_ELEMENT_NODE
) && (cur
->name
!= NULL
)) {
82 if (xmlStrEqual(cur
->name
, BAD_CAST
"head"))
84 if (xmlStrEqual(cur
->name
, BAD_CAST
"meta"))
95 * Search the meta elements
99 if ((cur
->type
== XML_ELEMENT_NODE
) && (cur
->name
!= NULL
)) {
100 if (xmlStrEqual(cur
->name
, BAD_CAST
"meta")) {
101 xmlAttrPtr attr
= cur
->properties
;
103 const xmlChar
*value
;
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")))
115 else if ((value
!= NULL
)
116 && (!xmlStrcasecmp(attr
->name
, BAD_CAST
"content")))
118 if ((http
!= 0) && (content
!= NULL
))
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
) {
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
)
146 if (encoding
!= NULL
) {
147 while ((*encoding
== ' ') || (*encoding
== '\t')) encoding
++;
153 * htmlSetMetaEncoding:
155 * @encoding: the encoding string
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.
161 * Returns 0 in case of success and -1 in case of error
164 htmlSetMetaEncoding(htmlDocPtr doc
, const xmlChar
*encoding
) {
165 htmlNodePtr cur
, meta
= NULL
, head
= NULL
;
166 const xmlChar
*content
= NULL
;
167 char newcontent
[100];
174 /* html isn't a real encoding it's just libxml2 way to get entities */
175 if (!xmlStrcasecmp(encoding
, BAD_CAST
"html"))
178 if (encoding
!= NULL
) {
179 snprintf(newcontent
, sizeof(newcontent
), "text/html; charset=%s",
181 newcontent
[sizeof(newcontent
) - 1] = 0;
189 while (cur
!= NULL
) {
190 if ((cur
->type
== XML_ELEMENT_NODE
) && (cur
->name
!= NULL
)) {
191 if (xmlStrcasecmp(cur
->name
, BAD_CAST
"html") == 0)
193 if (xmlStrcasecmp(cur
->name
, BAD_CAST
"head") == 0)
195 if (xmlStrcasecmp(cur
->name
, BAD_CAST
"meta") == 0)
207 while (cur
!= NULL
) {
208 if ((cur
->type
== XML_ELEMENT_NODE
) && (cur
->name
!= NULL
)) {
209 if (xmlStrcasecmp(cur
->name
, BAD_CAST
"head") == 0)
211 if (xmlStrcasecmp(cur
->name
, BAD_CAST
"meta") == 0) {
222 if (cur
->children
== NULL
)
228 * Search and update all the remaining the meta elements carrying
229 * encoding information
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
;
236 const xmlChar
*value
;
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")))
250 if ((value
!= NULL
) &&
251 (!xmlStrcasecmp(attr
->name
, BAD_CAST
"content")))
254 if ((http
!= 0) && (content
!= NULL
))
259 if ((http
!= 0) && (content
!= NULL
)) {
270 if ((encoding
!= NULL
) && (head
!= NULL
)) {
272 * Create a new Meta element with the right attributes
275 meta
= xmlNewDocNode(doc
, NULL
, BAD_CAST
"meta", NULL
);
276 if (head
->children
== NULL
)
277 xmlAddChild(head
, meta
);
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
);
284 /* remove the meta tag if NULL is passed */
285 if (encoding
== NULL
) {
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
);
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"
307 static const char* htmlBooleanAttrs
[] = {
308 "checked", "compact", "declare", "defer", "disabled", "ismap",
309 "multiple", "nohref", "noresize", "noshade", "nowrap", "readonly",
316 * @name: the name of the attribute to check
318 * Determine if a given attribute is a boolean attribute.
320 * returns: false if the attribute is not boolean, true otherwise.
323 htmlIsBooleanAttr(const xmlChar
*name
)
327 while (htmlBooleanAttrs
[i
] != NULL
) {
328 if (xmlStrcasecmp((const xmlChar
*)htmlBooleanAttrs
[i
], name
) == 0)
335 #ifdef LIBXML_OUTPUT_ENABLED
337 * private routine exported from xmlIO.c
340 xmlAllocOutputBufferInternal(xmlCharEncodingHandlerPtr encoder
);
341 /************************************************************************
343 * Output error handlers *
345 ************************************************************************/
348 * @extra: extra information
350 * Handle an out of memory condition
353 htmlSaveErrMemory(const char *extra
)
355 __xmlSimpleError(XML_FROM_OUTPUT
, XML_ERR_NO_MEMORY
, NULL
, NULL
, extra
);
360 * @code: the error number
361 * @node: the location of the error.
362 * @extra: extra information
364 * Handle an out of memory condition
367 htmlSaveErr(int code
, xmlNodePtr node
, const char *extra
)
369 const char *msg
= NULL
;
372 case XML_SAVE_NOT_UTF8
:
373 msg
= "string is not in UTF-8\n";
375 case XML_SAVE_CHAR_INVALID
:
376 msg
= "invalid character value\n";
378 case XML_SAVE_UNKNOWN_ENCODING
:
379 msg
= "unknown encoding %s\n";
381 case XML_SAVE_NO_DOCTYPE
:
382 msg
= "HTML has no DOCTYPE\n";
385 msg
= "unexpected error number\n";
387 __xmlSimpleError(XML_FROM_OUTPUT
, code
, node
, msg
, extra
);
390 /************************************************************************
392 * Dumping HTML tree content to a simple buffer *
394 ************************************************************************/
397 * htmlBufNodeDumpFormat:
398 * @buf: the xmlBufPtr output
400 * @cur: the current node
401 * @format: should formatting spaces been added
403 * Dump an HTML node, recursive behaviour,children are printed too.
405 * Returns the number of byte written or -1 in case of error
408 htmlBufNodeDumpFormat(xmlBufPtr buf
, xmlDocPtr doc
, xmlNodePtr cur
,
412 xmlOutputBufferPtr outbuf
;
420 outbuf
= (xmlOutputBufferPtr
) xmlMalloc(sizeof(xmlOutputBuffer
));
421 if (outbuf
== NULL
) {
422 htmlSaveErrMemory("allocating HTML output buffer");
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
;
433 use
= xmlBufUse(buf
);
434 htmlNodeDumpFormatOutput(outbuf
, doc
, cur
, NULL
, format
);
436 ret
= xmlBufUse(buf
) - use
;
442 * @buf: the HTML buffer output
444 * @cur: the current node
446 * Dump an HTML node, recursive behaviour,children are printed too,
447 * and formatting returns are added.
449 * Returns the number of byte written or -1 in case of error
452 htmlNodeDump(xmlBufferPtr buf
, xmlDocPtr doc
, xmlNodePtr cur
) {
456 if ((buf
== NULL
) || (cur
== NULL
))
460 buffer
= xmlBufFromBuffer(buf
);
464 ret
= htmlBufNodeDumpFormat(buffer
, doc
, cur
, 1);
466 xmlBufBackToBuffer(buffer
);
474 * htmlNodeDumpFileFormat:
475 * @out: the FILE pointer
477 * @cur: the current node
478 * @encoding: the document encoding
479 * @format: should formatting spaces been added
481 * Dump an HTML node, recursive behaviour,children are printed too.
483 * TODO: if encoding == NULL try to save in the doc encoding
485 * returns: the number of byte written or -1 in case of failure.
488 htmlNodeDumpFileFormat(FILE *out
, xmlDocPtr doc
,
489 xmlNodePtr cur
, const char *encoding
, int format
) {
490 xmlOutputBufferPtr buf
;
491 xmlCharEncodingHandlerPtr handler
= NULL
;
496 if (encoding
!= NULL
) {
499 enc
= xmlParseCharEncoding(encoding
);
500 if (enc
!= XML_CHAR_ENCODING_UTF8
) {
501 handler
= xmlFindCharEncodingHandler(encoding
);
503 htmlSaveErr(XML_SAVE_UNKNOWN_ENCODING
, NULL
, encoding
);
507 * Fallback to HTML or ASCII when the encoding is unspecified
510 handler
= xmlFindCharEncodingHandler("HTML");
512 handler
= xmlFindCharEncodingHandler("ascii");
516 * save the content to a temp buffer.
518 buf
= xmlOutputBufferCreateFile(out
, handler
);
519 if (buf
== NULL
) return(0);
521 htmlNodeDumpFormatOutput(buf
, doc
, cur
, NULL
, format
);
523 ret
= xmlOutputBufferClose(buf
);
529 * @out: the FILE pointer
531 * @cur: the current node
533 * Dump an HTML node, recursive behaviour,children are printed too,
534 * and formatting returns are added.
537 htmlNodeDumpFile(FILE *out
, xmlDocPtr doc
, xmlNodePtr cur
) {
538 htmlNodeDumpFileFormat(out
, doc
, cur
, NULL
, 1);
542 * htmlDocDumpMemoryFormat:
544 * @mem: OUT: the memory pointer
545 * @size: OUT: the memory length
546 * @format: should formatting spaces been added
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.
552 htmlDocDumpMemoryFormat(xmlDocPtr cur
, xmlChar
**mem
, int *size
, int format
) {
553 xmlOutputBufferPtr buf
;
554 xmlCharEncodingHandlerPtr handler
= NULL
;
555 const char *encoding
;
559 if ((mem
== NULL
) || (size
== NULL
))
567 encoding
= (const char *) htmlGetMetaEncoding(cur
);
569 if (encoding
!= NULL
) {
572 enc
= xmlParseCharEncoding(encoding
);
573 if (enc
!= XML_CHAR_ENCODING_UTF8
) {
574 handler
= xmlFindCharEncodingHandler(encoding
);
576 htmlSaveErr(XML_SAVE_UNKNOWN_ENCODING
, NULL
, encoding
);
581 * Fallback to HTML or ASCII when the encoding is unspecified
584 handler
= xmlFindCharEncodingHandler("HTML");
586 handler
= xmlFindCharEncodingHandler("ascii");
589 buf
= xmlAllocOutputBufferInternal(handler
);
596 htmlDocContentDumpFormatOutput(buf
, cur
, NULL
, format
);
598 xmlOutputBufferFlush(buf
);
599 if (buf
->conv
!= NULL
) {
600 *size
= xmlBufUse(buf
->conv
);
601 *mem
= xmlStrndup(xmlBufContent(buf
->conv
), *size
);
603 *size
= xmlBufUse(buf
->buffer
);
604 *mem
= xmlStrndup(xmlBufContent(buf
->buffer
), *size
);
606 (void)xmlOutputBufferClose(buf
);
612 * @mem: OUT: the memory pointer
613 * @size: OUT: the memory length
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.
619 htmlDocDumpMemory(xmlDocPtr cur
, xmlChar
**mem
, int *size
) {
620 htmlDocDumpMemoryFormat(cur
, mem
, size
, 1);
624 /************************************************************************
626 * Dumping HTML tree content to an I/O output buffer *
628 ************************************************************************/
630 void xmlNsListDumpOutput(xmlOutputBufferPtr buf
, xmlNsPtr cur
);
634 * @buf: the HTML buffer output
636 * @encoding: the encoding string
638 * TODO: check whether encoding is needed
640 * Dump the HTML document DTD, if any.
643 htmlDtdDumpOutput(xmlOutputBufferPtr buf
, xmlDocPtr doc
,
644 const char *encoding ATTRIBUTE_UNUSED
) {
645 xmlDtdPtr cur
= doc
->intSubset
;
648 htmlSaveErr(XML_SAVE_NO_DOCTYPE
, (xmlNodePtr
) doc
, NULL
);
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
);
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
);
665 xmlOutputBufferWriteString(buf
, ">\n");
669 * htmlAttrDumpOutput:
670 * @buf: the HTML buffer output
672 * @cur: the attribute pointer
674 * Dump an HTML attribute
677 htmlAttrDumpOutput(xmlOutputBufferPtr buf
, xmlDocPtr doc
, xmlAttrPtr cur
) {
681 * The html output method should not escape a & character
682 * occurring in an attribute value immediately followed by
683 * a { character (see Section B.7.1 of the HTML 4.0 Recommendation).
684 * This is implemented in xmlEncodeEntitiesReentrant
690 xmlOutputBufferWriteString(buf
, " ");
691 if ((cur
->ns
!= NULL
) && (cur
->ns
->prefix
!= NULL
)) {
692 xmlOutputBufferWriteString(buf
, (const char *)cur
->ns
->prefix
);
693 xmlOutputBufferWriteString(buf
, ":");
695 xmlOutputBufferWriteString(buf
, (const char *)cur
->name
);
696 if ((cur
->children
!= NULL
) && (!htmlIsBooleanAttr(cur
->name
))) {
697 value
= xmlNodeListGetString(doc
, cur
->children
, 0);
699 xmlOutputBufferWriteString(buf
, "=");
700 if ((cur
->ns
== NULL
) && (cur
->parent
!= NULL
) &&
701 (cur
->parent
->ns
== NULL
) &&
702 ((!xmlStrcasecmp(cur
->name
, BAD_CAST
"href")) ||
703 (!xmlStrcasecmp(cur
->name
, BAD_CAST
"action")) ||
704 (!xmlStrcasecmp(cur
->name
, BAD_CAST
"src")) ||
705 ((!xmlStrcasecmp(cur
->name
, BAD_CAST
"name")) &&
706 (!xmlStrcasecmp(cur
->parent
->name
, BAD_CAST
"a"))))) {
708 xmlChar
*tmp
= value
;
710 while (IS_BLANK_CH(*tmp
)) tmp
++;
713 * the < and > have already been escaped at the entity level
714 * And doing so here breaks server side includes
716 escaped
= xmlURIEscapeStr(tmp
, BAD_CAST
"@/:=?;#%&,+<>");
717 if (escaped
!= NULL
) {
718 xmlBufWriteQuotedString(buf
->buffer
, escaped
);
721 xmlBufWriteQuotedString(buf
->buffer
, value
);
724 xmlBufWriteQuotedString(buf
->buffer
, value
);
728 xmlOutputBufferWriteString(buf
, "=\"\"");
734 * htmlNodeDumpFormatOutput:
735 * @buf: the HTML buffer output
737 * @cur: the current node
738 * @encoding: the encoding string (unused)
739 * @format: should formatting spaces been added
741 * Dump an HTML node, recursive behaviour,children are printed too.
744 htmlNodeDumpFormatOutput(xmlOutputBufferPtr buf
, xmlDocPtr doc
,
745 xmlNodePtr cur
, const char *encoding ATTRIBUTE_UNUSED
,
749 const htmlElemDesc
* info
;
753 if ((cur
== NULL
) || (buf
== NULL
)) {
760 case XML_HTML_DOCUMENT_NODE
:
761 case XML_DOCUMENT_NODE
:
762 if (((xmlDocPtr
) cur
)->intSubset
!= NULL
) {
763 htmlDtdDumpOutput(buf
, (xmlDocPtr
) cur
, NULL
);
765 if (cur
->children
!= NULL
) {
771 case XML_ELEMENT_NODE
:
773 * Get specific HTML info for that node.
776 info
= htmlTagLookup(cur
->name
);
780 xmlOutputBufferWriteString(buf
, "<");
781 if ((cur
->ns
!= NULL
) && (cur
->ns
->prefix
!= NULL
)) {
782 xmlOutputBufferWriteString(buf
, (const char *)cur
->ns
->prefix
);
783 xmlOutputBufferWriteString(buf
, ":");
785 xmlOutputBufferWriteString(buf
, (const char *)cur
->name
);
787 xmlNsListDumpOutput(buf
, cur
->nsDef
);
788 attr
= cur
->properties
;
789 while (attr
!= NULL
) {
790 htmlAttrDumpOutput(buf
, doc
, attr
);
794 if ((info
!= NULL
) && (info
->empty
)) {
795 xmlOutputBufferWriteString(buf
, ">");
796 } else if (cur
->children
== NULL
) {
797 if ((info
!= NULL
) && (info
->saveEndTag
!= 0) &&
798 (xmlStrcmp(BAD_CAST info
->name
, BAD_CAST
"html")) &&
799 (xmlStrcmp(BAD_CAST info
->name
, BAD_CAST
"body"))) {
800 xmlOutputBufferWriteString(buf
, ">");
802 xmlOutputBufferWriteString(buf
, "></");
803 if ((cur
->ns
!= NULL
) && (cur
->ns
->prefix
!= NULL
)) {
804 xmlOutputBufferWriteString(buf
,
805 (const char *)cur
->ns
->prefix
);
806 xmlOutputBufferWriteString(buf
, ":");
808 xmlOutputBufferWriteString(buf
, (const char *)cur
->name
);
809 xmlOutputBufferWriteString(buf
, ">");
812 xmlOutputBufferWriteString(buf
, ">");
813 if ((format
) && (info
!= NULL
) && (!info
->isinline
) &&
814 (cur
->children
->type
!= HTML_TEXT_NODE
) &&
815 (cur
->children
->type
!= HTML_ENTITY_REF_NODE
) &&
816 (cur
->children
!= cur
->last
) &&
817 (cur
->name
!= NULL
) &&
818 (cur
->name
[0] != 'p')) /* p, pre, param */
819 xmlOutputBufferWriteString(buf
, "\n");
824 if ((format
) && (cur
->next
!= NULL
) &&
825 (info
!= NULL
) && (!info
->isinline
)) {
826 if ((cur
->next
->type
!= HTML_TEXT_NODE
) &&
827 (cur
->next
->type
!= HTML_ENTITY_REF_NODE
) &&
828 (cur
->parent
!= NULL
) &&
829 (cur
->parent
->name
!= NULL
) &&
830 (cur
->parent
->name
[0] != 'p')) /* p, pre, param */
831 xmlOutputBufferWriteString(buf
, "\n");
836 case XML_ATTRIBUTE_NODE
:
837 htmlAttrDumpOutput(buf
, doc
, (xmlAttrPtr
) cur
);
841 if (cur
->content
== NULL
)
843 if (((cur
->name
== (const xmlChar
*)xmlStringText
) ||
844 (cur
->name
!= (const xmlChar
*)xmlStringTextNoenc
)) &&
845 ((cur
->parent
== NULL
) ||
846 ((xmlStrcasecmp(cur
->parent
->name
, BAD_CAST
"script")) &&
847 (xmlStrcasecmp(cur
->parent
->name
, BAD_CAST
"style"))))) {
850 buffer
= xmlEncodeEntitiesReentrant(doc
, cur
->content
);
851 if (buffer
!= NULL
) {
852 xmlOutputBufferWriteString(buf
, (const char *)buffer
);
856 xmlOutputBufferWriteString(buf
, (const char *)cur
->content
);
860 case HTML_COMMENT_NODE
:
861 if (cur
->content
!= NULL
) {
862 xmlOutputBufferWriteString(buf
, "<!--");
863 xmlOutputBufferWriteString(buf
, (const char *)cur
->content
);
864 xmlOutputBufferWriteString(buf
, "-->");
869 if (cur
->name
!= NULL
) {
870 xmlOutputBufferWriteString(buf
, "<?");
871 xmlOutputBufferWriteString(buf
, (const char *)cur
->name
);
872 if (cur
->content
!= NULL
) {
873 xmlOutputBufferWriteString(buf
, " ");
874 xmlOutputBufferWriteString(buf
,
875 (const char *)cur
->content
);
877 xmlOutputBufferWriteString(buf
, ">");
881 case HTML_ENTITY_REF_NODE
:
882 xmlOutputBufferWriteString(buf
, "&");
883 xmlOutputBufferWriteString(buf
, (const char *)cur
->name
);
884 xmlOutputBufferWriteString(buf
, ";");
887 case HTML_PRESERVE_NODE
:
888 if (cur
->content
!= NULL
) {
889 xmlOutputBufferWriteString(buf
, (const char *)cur
->content
);
900 if (cur
->next
!= NULL
) {
906 * The parent should never be NULL here but we want to handle
907 * corrupted documents gracefully.
909 if (cur
->parent
== NULL
)
913 if ((cur
->type
== XML_HTML_DOCUMENT_NODE
) ||
914 (cur
->type
== XML_DOCUMENT_NODE
)) {
915 xmlOutputBufferWriteString(buf
, "\n");
917 if ((format
) && (cur
->ns
== NULL
))
918 info
= htmlTagLookup(cur
->name
);
922 if ((format
) && (info
!= NULL
) && (!info
->isinline
) &&
923 (cur
->last
->type
!= HTML_TEXT_NODE
) &&
924 (cur
->last
->type
!= HTML_ENTITY_REF_NODE
) &&
925 (cur
->children
!= cur
->last
) &&
926 (cur
->name
!= NULL
) &&
927 (cur
->name
[0] != 'p')) /* p, pre, param */
928 xmlOutputBufferWriteString(buf
, "\n");
930 xmlOutputBufferWriteString(buf
, "</");
931 if ((cur
->ns
!= NULL
) && (cur
->ns
->prefix
!= NULL
)) {
932 xmlOutputBufferWriteString(buf
, (const char *)cur
->ns
->prefix
);
933 xmlOutputBufferWriteString(buf
, ":");
935 xmlOutputBufferWriteString(buf
, (const char *)cur
->name
);
936 xmlOutputBufferWriteString(buf
, ">");
938 if ((format
) && (info
!= NULL
) && (!info
->isinline
) &&
939 (cur
->next
!= NULL
)) {
940 if ((cur
->next
->type
!= HTML_TEXT_NODE
) &&
941 (cur
->next
->type
!= HTML_ENTITY_REF_NODE
) &&
942 (cur
->parent
!= NULL
) &&
943 (cur
->parent
->name
!= NULL
) &&
944 (cur
->parent
->name
[0] != 'p')) /* p, pre, param */
945 xmlOutputBufferWriteString(buf
, "\n");
953 * htmlNodeDumpOutput:
954 * @buf: the HTML buffer output
956 * @cur: the current node
957 * @encoding: the encoding string (unused)
959 * Dump an HTML node, recursive behaviour,children are printed too,
960 * and formatting returns/spaces are added.
963 htmlNodeDumpOutput(xmlOutputBufferPtr buf
, xmlDocPtr doc
,
964 xmlNodePtr cur
, const char *encoding ATTRIBUTE_UNUSED
) {
965 htmlNodeDumpFormatOutput(buf
, doc
, cur
, NULL
, 1);
969 * htmlDocContentDumpFormatOutput:
970 * @buf: the HTML buffer output
972 * @encoding: the encoding string (unused)
973 * @format: should formatting spaces been added
975 * Dump an HTML document.
978 htmlDocContentDumpFormatOutput(xmlOutputBufferPtr buf
, xmlDocPtr cur
,
979 const char *encoding ATTRIBUTE_UNUSED
,
981 htmlNodeDumpFormatOutput(buf
, cur
, (xmlNodePtr
) cur
, NULL
, format
);
985 * htmlDocContentDumpOutput:
986 * @buf: the HTML buffer output
988 * @encoding: the encoding string (unused)
990 * Dump an HTML document. Formatting return/spaces are added.
993 htmlDocContentDumpOutput(xmlOutputBufferPtr buf
, xmlDocPtr cur
,
994 const char *encoding ATTRIBUTE_UNUSED
) {
995 htmlNodeDumpFormatOutput(buf
, cur
, (xmlNodePtr
) cur
, NULL
, 1);
998 /************************************************************************
1000 * Saving functions front-ends *
1002 ************************************************************************/
1007 * @cur: the document
1009 * Dump an HTML document to an open FILE.
1011 * returns: the number of byte written or -1 in case of failure.
1014 htmlDocDump(FILE *f
, xmlDocPtr cur
) {
1015 xmlOutputBufferPtr buf
;
1016 xmlCharEncodingHandlerPtr handler
= NULL
;
1017 const char *encoding
;
1022 if ((cur
== NULL
) || (f
== NULL
)) {
1026 encoding
= (const char *) htmlGetMetaEncoding(cur
);
1028 if (encoding
!= NULL
) {
1029 xmlCharEncoding enc
;
1031 enc
= xmlParseCharEncoding(encoding
);
1032 if (enc
!= XML_CHAR_ENCODING_UTF8
) {
1033 handler
= xmlFindCharEncodingHandler(encoding
);
1034 if (handler
== NULL
)
1035 htmlSaveErr(XML_SAVE_UNKNOWN_ENCODING
, NULL
, encoding
);
1039 * Fallback to HTML or ASCII when the encoding is unspecified
1041 if (handler
== NULL
)
1042 handler
= xmlFindCharEncodingHandler("HTML");
1043 if (handler
== NULL
)
1044 handler
= xmlFindCharEncodingHandler("ascii");
1047 buf
= xmlOutputBufferCreateFile(f
, handler
);
1048 if (buf
== NULL
) return(-1);
1049 htmlDocContentDumpOutput(buf
, cur
, NULL
);
1051 ret
= xmlOutputBufferClose(buf
);
1057 * @filename: the filename (or URL)
1058 * @cur: the document
1060 * Dump an HTML document to a file. If @filename is "-" the stdout file is
1062 * returns: the number of byte written or -1 in case of failure.
1065 htmlSaveFile(const char *filename
, xmlDocPtr cur
) {
1066 xmlOutputBufferPtr buf
;
1067 xmlCharEncodingHandlerPtr handler
= NULL
;
1068 const char *encoding
;
1071 if ((cur
== NULL
) || (filename
== NULL
))
1076 encoding
= (const char *) htmlGetMetaEncoding(cur
);
1078 if (encoding
!= NULL
) {
1079 xmlCharEncoding enc
;
1081 enc
= xmlParseCharEncoding(encoding
);
1082 if (enc
!= XML_CHAR_ENCODING_UTF8
) {
1083 handler
= xmlFindCharEncodingHandler(encoding
);
1084 if (handler
== NULL
)
1085 htmlSaveErr(XML_SAVE_UNKNOWN_ENCODING
, NULL
, encoding
);
1089 * Fallback to HTML or ASCII when the encoding is unspecified
1091 if (handler
== NULL
)
1092 handler
= xmlFindCharEncodingHandler("HTML");
1093 if (handler
== NULL
)
1094 handler
= xmlFindCharEncodingHandler("ascii");
1098 * save the content to a temp buffer.
1100 buf
= xmlOutputBufferCreateFilename(filename
, handler
, cur
->compression
);
1101 if (buf
== NULL
) return(0);
1103 htmlDocContentDumpOutput(buf
, cur
, NULL
);
1105 ret
= xmlOutputBufferClose(buf
);
1110 * htmlSaveFileFormat:
1111 * @filename: the filename
1112 * @cur: the document
1113 * @format: should formatting spaces been added
1114 * @encoding: the document encoding
1116 * Dump an HTML document to a file using a given encoding.
1118 * returns: the number of byte written or -1 in case of failure.
1121 htmlSaveFileFormat(const char *filename
, xmlDocPtr cur
,
1122 const char *encoding
, int format
) {
1123 xmlOutputBufferPtr buf
;
1124 xmlCharEncodingHandlerPtr handler
= NULL
;
1127 if ((cur
== NULL
) || (filename
== NULL
))
1132 if (encoding
!= NULL
) {
1133 xmlCharEncoding enc
;
1135 enc
= xmlParseCharEncoding(encoding
);
1136 if (enc
!= XML_CHAR_ENCODING_UTF8
) {
1137 handler
= xmlFindCharEncodingHandler(encoding
);
1138 if (handler
== NULL
)
1139 htmlSaveErr(XML_SAVE_UNKNOWN_ENCODING
, NULL
, encoding
);
1141 htmlSetMetaEncoding(cur
, (const xmlChar
*) encoding
);
1143 htmlSetMetaEncoding(cur
, (const xmlChar
*) "UTF-8");
1146 * Fallback to HTML or ASCII when the encoding is unspecified
1148 if (handler
== NULL
)
1149 handler
= xmlFindCharEncodingHandler("HTML");
1150 if (handler
== NULL
)
1151 handler
= xmlFindCharEncodingHandler("ascii");
1155 * save the content to a temp buffer.
1157 buf
= xmlOutputBufferCreateFilename(filename
, handler
, 0);
1158 if (buf
== NULL
) return(0);
1160 htmlDocContentDumpFormatOutput(buf
, cur
, encoding
, format
);
1162 ret
= xmlOutputBufferClose(buf
);
1168 * @filename: the filename
1169 * @cur: the document
1170 * @encoding: the document encoding
1172 * Dump an HTML document to a file using a given encoding
1173 * and formatting returns/spaces are added.
1175 * returns: the number of byte written or -1 in case of failure.
1178 htmlSaveFileEnc(const char *filename
, xmlDocPtr cur
, const char *encoding
) {
1179 return(htmlSaveFileFormat(filename
, cur
, encoding
, 1));
1182 #endif /* LIBXML_OUTPUT_ENABLED */
1184 #define bottom_HTMLtree
1185 #include "elfgcchack.h"
1186 #endif /* LIBXML_HTML_ENABLED */