Given the following document,
<root>a<b>c</b></root>
reparenting the 'c' text node using addPrevSibling(b_node, c_node) cause
the document to end up with two text nodes 'a' and 'c', in other words root
will have 3 nodes 'a', 'c' text nodes and the empty b node. On the other
hand if addNextSibling(a_node, c_node) was used, the text nodes will be
merged correctly and the root node will have 2 nodes 'ac' text node and the
empty b node. Below is a test program that outputs the contents of the
nodes in both cases.
This was tested with version 2.7.8 on ubuntu as well as master
(commit 8fc913fcc9d2c3fd5c76694cb4e631ca1242de37)
#include <stdio.h>
#include <libxml/parser.h>
#include <libxml/tree.h>
int main() {
/*
* this initialize the library and check potential ABI mismatches
* between the version it was compiled for and the actual shared
* library used.
*/
LIBXML_TEST_VERSION;
xmlDocPtr doc1, doc2; /* the resulting document tree */
const char doc_content [] = "<root>a<b>c</b></root>";
doc1 = xmlReadMemory(doc_content, sizeof(doc_content), NULL, NULL, 1);
doc2 = xmlReadMemory(doc_content, sizeof(doc_content), NULL, NULL, 1);
if (!doc1 || !doc2) {
fprintf(stderr, "Failed to parse %s\n", doc_content);
return 1;
}
xmlNodePtr b_node = doc1->children->last;
xmlNodePtr c_node = b_node->last;
xmlAddPrevSibling(b_node, c_node);
b_node = doc2->children->last;
c_node = b_node->last;
xmlAddNextSibling(doc2->children->children, c_node);
for (xmlNodePtr node = doc1->children->children; node; node = node->next)
printf("contents: %s\n", node->content);
for (xmlNodePtr node = doc2->children->children; node; node = node->next)
printf("contents: %s\n", node->content);
return 0;
}
_______________________________________________
xml mailing list, project page http://xmlsoft.org/
[email protected]
https://mail.gnome.org/mailman/listinfo/xml
