tpctt / tfhpple Goto Github PK

Hi,
We have code which does parsing of SOAP xml, and we use xmlNodeGetContent api. In valgrind it shows that it leaking memory. What i found in some posts that we need to explicitly free the memory returned by xmlNodeGetContent. Can someone help me on how to do this for below recursive calls?

void getSoapMsgDetails(xmlNode* node)
{
while(node)
{
if(XML_ELEMENT_NODE == node->type)
{
if(0 == strcmp( soapMsgHeader.c_str() ,(const char*)node->name ) )
{
xmlNode childNode=node->children;
while(childNode)
{
if( (XML_ELEMENT_NODE == childNode->type) && ( 0 == ( xmlChildElementCount(childNode) ) ) )
{
std::string name= (const char)childNode->name;
std::string value = (const char*)xmlNodeGetContent(childNode);
soapHeaderAttributes.insert(std::pair< std::string, std::string> (name,value) );
}
childNode=childNode->next;
}
}

        if(0 == strcmp(soapMsgMo.c_str(), (const char*)node->name))
        {
            getSoapMsgMoList(node);
        }

        if(0 == strcmp(soapMsgBody.c_str(), (const char*)node->name))
        {
            node = node->children;
            while(XML_ELEMENT_NODE != node->type)
            {
                node = node->next;
            }
            msgTypeVal = (char*)node->name;
            node = node->parent;
        }
    }
    getSoapMsgDetails(node->children);
    node = node->next;
}

}

in XPathQuery.m line29
xmlChar *nodeContent = xmlNodeGetContent(currentNode);
this nodeContent seems to be leak in next code
if ([resultForNode[@"nodeName"] isEqual:@"text"] && parentResult) {
if this code is excuted the nodeContent is not free

First of all, I really like this library, you doing a great job thank you for maintaining it.
I have a serious problem, and its already one week i'm not able to solve it.
I'm calling on a method which does the following

 private class func getVideoAttribute(forKey key: String, source: String) -> Any? {
        let parser = TFHpple(htmlData: source.data(using: .utf8, allowLossyConversion: false))!
        **if let videos = parser.search(withXPathQuery: "//video")** // Crash happens here.

actually when going inside the error i can see it's crash on this line...

NSArray *PerformHTMLXPathQueryWithEncoding(NSData *document, NSString *query,NSString *encoding)
{
    xmlDocPtr doc;

    /* Load XML document */
    const char *encoded = encoding ? [encoding cStringUsingEncoding:NSUTF8StringEncoding] : NULL;
// Error happens herein the next line
    **doc = htmlReadMemory([document bytes], (int)[document length], "", encoded, HTML_PARSE_NOWARNING | 
    HTML_PARSE_NOERROR);**
    if (doc == NULL) {
        NSLog(@"Unable to parse.");
        return nil;
    }
    NSArray *result = PerformXPathQuery(doc, query);
    xmlFreeDoc(doc);
    return result;
}

any idea way it's happening only on real device ?