Search code examples
c++xmlrecursionpugixml

PugiXML extracting data from recursive xml_tree_walker

I'm trying to extract data from an XML that looks as shown in the image below using pugiXML traverse function. i'm able to print out all the data from the XML file using this function, however im not able to extract and store these into my variables. The main issue is that im not able to get node.name() and node.value() at the same time. node.name() comes in 1 iteration node.value() comes in the iteration after.

    const char* node_types[] =
{
    "null","Element", "Name", "InstancePath", "InstancePathDepth", "FieldAmount", "DataType", "TypeName", "DataSize", "ArrayTypeName", "LowerBound", "UpperBound", "BitPosition", "IsArray", "IsElementary", "IsProgram", "IsResource", "IsString", "IsStruct", "IsTask"
};

    // tag::impl[]
    struct simple_walker : pugi::xml_tree_walker
    {
        virtual bool for_each(pugi::xml_node& node)
        {
            
            for (int i = 0; i < depth(); ++i) std::cout << "  "; // indentation             
                
            
                std::cout << node.name()<<" "<< node.value() << "\n";
    
    
            return true; // continue traversal
        }
    };

enter image description hereenter image description here

The output is great, same as in the XML file.

Solution

  • You can skip elements that aren't of type node_element and use node.child_value() roughly like this:

        virtual bool for_each(pugi::xml_node& node)
    {
        if (node.type() != pugi::node_element)
            return true;

        for (int i = 0; i < depth(); ++i) std::cout << "  "; // indentation             

        std::cout << node.name()<<" "<< node.child_value() << "\n";

        return true; // continue traversal
    }

    This is necessary because text (PCDATA) nodes are separate from element nodes: https://pugixml.org/docs/manual.html#node_pcdata