【C++】使用libxml解析XML文件

今天药忘吃喽~ 2022-09-20 15:04 347阅读 0赞

OpenCV中使用FileStorage创建和解析XML文件。但FileStorage不能处理节点为空的情况，且对属性也不友好。例子见[这里][Link 1]。

libxml是C接口的处理xml文件的开源代码库。可以在[这里][Link 2]下载。

对于一段如下的xml文件：

<?xml version="1.0"?>
    <Message Version="1.0">
      <Info evaluateType="2" mediaFile="1.avi">1</Info>
      <Items>
        <Item frameNum="0" scale="2" />
        <Item frameNum="1" scale="2" />
        <Item frameNum="2" scale="2" />
    	<Item frameNum="2913" scale="2">
          <Label type="2" l="564" t="178" r="594" b="280" id="15" />
          <Label type="2" l="528" t="184" r="560" b="288" id="11" />
          <Label type="2" l="490" t="182" r="522" b="288" id="16" />
          <Label type="2" l="614" t="358" r="704" b="600" id="4" />
        </Item>
      </Items>
    </Message>

libxml处理的函数如下：

string getLabes(char *szDocName, map<pair<int,int>,Rect > &labels){
    	xmlDocPtr doc;          
    	xmlNodePtr curNode;      
    	xmlChar *szKey;          
    	string filename="";
    
    	doc = xmlReadFile(szDocName,"GB2312",XML_PARSE_RECOVER); //解析文件
    
    	if (doc==NULL){  
    		fprintf(stderr,"Document not parsed successfully.\n");     
    		return filename; 
    
    	} 
    
    	curNode = xmlDocGetRootElement(doc); //确定文档根元素
    	if (NULL == curNode){ 
    		fprintf(stderr,"empty document\n");
    		xmlFreeDoc(doc); 
    		return filename; 
    	} 
    
    	if (xmlStrcmp(curNode->name, BAD_CAST "Message")) {
    		fprintf(stderr,"document of the wrong type, root node != Message"); 
    		xmlFreeDoc(doc);
    		return filename; 
    	} 
    
    	curNode = curNode->xmlChildrenNode;
    	xmlNodePtr propNodePtr = curNode;
    	xmlNodePtr itemNodePtr;  
    
    	while(curNode != NULL){
    		//取出文件名称
    		if (xmlHasProp(curNode,BAD_CAST "mediaFile")) {
    			propNodePtr = curNode;
    
    		} 
    		if (!xmlStrcmp(curNode->name, BAD_CAST "Items")) {
    			itemNodePtr = curNode->xmlChildrenNode;
    		} 
    		curNode = curNode->next;
    	}
    
    	xmlAttrPtr attrPtr = propNodePtr->properties;
    	while (attrPtr != NULL){
    		if (!xmlStrcmp(attrPtr->name, BAD_CAST "mediaFile")){
    			xmlChar* szAttr = xmlGetProp(propNodePtr,BAD_CAST "mediaFile");
    			char* szAttrG = u2g((char*)szAttr);
    			filename = string(szAttrG);
    			//cout<<"get filename: "<<filename<<endl;
    			xmlFree(szAttr);
    		}
    		attrPtr = attrPtr->next;
    	}
    
    	// Item
    	while (itemNodePtr != NULL){
    		int frameNum = 0;
    		int peopleID = 0;
    		Rect rect;
    
    		//<Label type="2" l="620" t="164" r="648" b="234" id="15" />
    		xmlAttrPtr attrPtr = itemNodePtr->properties;
    		while (attrPtr != NULL){
    			if (!xmlStrcmp(attrPtr->name, BAD_CAST "frameNum")){
    				xmlChar* szAttr = xmlGetProp(itemNodePtr,BAD_CAST "frameNum");
    				frameNum = atoi((char*)szAttr);
    				//cout<<"get frameNum: "<<frameNum<<endl;
    				xmlFree(szAttr);
    			}
    			attrPtr = attrPtr->next;
    		}
    
    		// Label
    		xmlNodePtr childNodePtr = itemNodePtr->xmlChildrenNode;
    		while (childNodePtr != NULL){
    			int l=0,t=0,r=0,b=0;
    			attrPtr = childNodePtr->properties; 
    			while(attrPtr != NULL){
    				if (!xmlStrcmp(attrPtr->name, BAD_CAST "l")){
    					xmlChar* szAttr = xmlGetProp(childNodePtr,BAD_CAST "l");
    					l=atoi((char*)szAttr);
    					xmlFree(szAttr);
    				}
    				if (!xmlStrcmp(attrPtr->name, BAD_CAST "t")){
    					xmlChar* szAttr = xmlGetProp(childNodePtr,BAD_CAST "t");
    					t=atoi((char*)szAttr);
    					xmlFree(szAttr);
    				}
    				if (!xmlStrcmp(attrPtr->name, BAD_CAST "r")){
    					xmlChar* szAttr = xmlGetProp(childNodePtr,BAD_CAST "r");
    					r=atoi((char*)szAttr);
    					xmlFree(szAttr);
    				}
    				if (!xmlStrcmp(attrPtr->name, BAD_CAST "b")){
    					xmlChar* szAttr = xmlGetProp(childNodePtr,BAD_CAST "b");
    					b=atoi((char*)szAttr);
    					xmlFree(szAttr);
    				}
    				if (!xmlStrcmp(attrPtr->name, BAD_CAST "id")){
    					xmlChar* szAttr = xmlGetProp(childNodePtr,BAD_CAST "id");
    					peopleID=atoi((char*)szAttr);
    					xmlFree(szAttr);
    				}
    				attrPtr = attrPtr->next;
    			}
    			if(!(l==0&&t==0&&b==0&&r==0)){
    				rect.x=l;rect.y=t;rect.width=r-l;rect.height=b-t;
    				labels[pair<int,int>(frameNum,peopleID)] = rect;
    				//cout<<"get label: "<<l<<" "<<t<<" "<<r<<" "<<b<<" "<<" "<<peopleID<<endl;
    			}
    			childNodePtr = childNodePtr->next;
    		}
    
    		itemNodePtr = itemNodePtr->next;
    	}
    
    	xmlFreeDoc(doc);
    	return filename;
    }

用map的iterator输出：

for(map<pair<int,int>,Rect >::iterator it=labels.begin();
    		it!=labels.end();it++){
    			cout<<(*it).first.first<<" "<<(*it).first.second<<" "
    				<<(*it).second.x<<" "<<(*it).second.y<<" "
    				<<(*it).second.width<<" "<<(*it).second.height<<endl;
    	}

![20140122112220625][]

### （转载请注明作者和出处：[http://blog.csdn.net/xiaowei\_cqu][http_blog.csdn.net_xiaowei_cqu] 未经允许请勿用于商业用途）       ###

[Link 1]: http://www.oschina.net/code/explore/OpenCV-2.2.0/samples/cpp/filestorage.cpp
[Link 2]: http://xmlsoft.org/
[20140122112220625]: /images/20220828/3c3c3f028dea4cf89c2d237ffae24a5e.png
[http_blog.csdn.net_xiaowei_cqu]: http://blog.csdn.net/xiaowei_cqu