【发布时间】：2012-02-11 12:01:09
【问题描述】：
我正在尝试从xmlTextReaderPtr 构建DOM tree。在我的最终程序中，它将用于处理来自大型 XML 文件（如 I did in java）的带有 xslt 的小型 DOM 树。我找不到使用命名空间/前缀的正确方法（我应该如何创建元素和属性？）。这是一个非常基本的 C 代码：我正在从流中构建 DOM，然后转储 DOM。我应该如何修改我的代码来处理命名空间/前缀？非常感谢！
#include <stdio.h>
#include <libxml/parser.h>
#include <libxml/tree.h>
#include <libxml/xmlreader.h>
/* dump XML node */
static void print_element_names(xmlNode * a_node)
{
xmlNode *cur_node = NULL;
for (cur_node = a_node; cur_node; cur_node = xmlNextElementSibling(cur_node))
{
if (cur_node->type == XML_ELEMENT_NODE)
{
xmlAttrPtr attr;
printf("node type: Element, name: %s", cur_node->name);
if(cur_node->ns!=0 && cur_node->ns->href!=0) printf(" with namespace: %s", cur_node->ns->href);
printf("\n");
for(attr = cur_node->properties; NULL != attr; attr = attr->next)
{
xmlChar* v=xmlGetProp( cur_node,attr->name);
printf(" @%s=%s ", attr->name,v);
xmlFree(v);
}
}
print_element_names(cur_node->children);
}
}
int main(int argc,char** argv)
{
LIBXML_TEST_VERSION;
xmlTextReaderPtr reader;
xmlDocPtr doc = NULL;
xmlNodePtr current=NULL;
xmlNsPtr ns=NULL;
/* read from stdin */
reader=xmlReaderForFd(fileno(stdin),0,"UTF-8",0);
for(;;)
{
int nodeType;
int ret = xmlTextReaderRead(reader);
if(ret<=0) break;
nodeType=xmlTextReaderNodeType(reader);
switch(nodeType)
{
case XML_READER_TYPE_ELEMENT:
{
xmlNsPtr ns=0;
xmlNodePtr node;
if(doc==NULL)
{
doc=xmlNewDoc( BAD_CAST "1.0");
}
if(xmlTextReaderConstNamespaceUri(reader)!=0)
{
/** how should I handle the attributes' namespaces & prefix here ? */
xmlNsPtr ns=xmlSearchNs(doc,current,xmlTextReaderConstNamespaceUri(reader));
node=xmlNewNode(ns, xmlTextReaderConstName(reader));
if(ns==0)
{
ns=xmlNewNs(node,
xmlTextReaderConstPrefix(reader),
xmlTextReaderConstNamespaceUri(reader)
);
}
}
else
{
node=xmlNewNode(0, xmlTextReaderConstName(reader));
}
if(current==NULL)
{
xmlDocSetRootElement(doc,node);
}
else
{
xmlAddChild(current,node);
}
current=node;
if(xmlTextReaderIsEmptyElement(reader))
{
current= current->parent;
}
if(xmlTextReaderHasAttributes(reader))
{
int i;
int n_att=xmlTextReaderAttributeCount(reader);
for(i=0;i< n_att;++i)
{
const xmlChar* k;
xmlChar* v;
xmlTextReaderMoveToAttributeNo(reader,i);
k = xmlTextReaderConstName(reader);
v = xmlTextReaderValue(reader);
/** how should I handle the attributes' namespaces & prefix here ? */
xmlNewProp(node,k, v);
xmlFree(v);
}
xmlTextReaderMoveToElement(reader);
}
break;
}
case XML_READER_TYPE_END_ELEMENT:
{
current= current->parent;
break;
}
case XML_READER_TYPE_SIGNIFICANT_WHITESPACE:
case XML_READER_TYPE_TEXT:
{
const xmlChar* v = xmlTextReaderConstValue(reader);
xmlNodePtr node= xmlNewDocText(doc,v);
xmlAddChild(current,node);
break;
}
default:
{
fprintf(stderr,"Ignoring node Type %d\n",nodeType);
break;
}
}
}
if(doc!=NULL)
{
print_element_names(xmlDocGetRootElement(doc));
xmlDocDump(stderr,doc);
xmlFreeDoc(doc);
}
xmlFreeTextReader(reader);
xmlCleanupParser();
xmlMemoryDump();
return 0;
}
这是我的测试文件：
<?xml version="1.0"?>
<a xmlns="http://urn1.org" xmlns:ns1="http://urn2.org" ns1:test="ok">azdazd
<b xmlns:ns2="http://urn3.org" xmlns:ns3="http://urn4.org" ns3:test="OK"/>
azd
<ns1:b test="ok"/>
xaz
</a>
非常感谢，
【问题讨论】：
标签： c xml namespaces stream libxml2