libxml2 predicates in xpath expression are not always recognized

Question

I appeal to you because I have problems in using the libxml2 library that does not take into account certain parameters in my xpath expressions.

Here is an example of xml file that I am trying to parse:

<?xml version="1.0" encoding="ISO-8859-1"?>

<bookstore>

    <book title="Harry Potter" lang="eng" version="1">
       <price>29.99</price>
    </book>

    <book title="Learning XML" lang="eng" version="2">
        <price>38.95</price>
    </book>

    <book title="Learning C" lang="eng" version="2">  
       <price>39.95</price>
    </book>

</bookstore>

Suppose I want to extract all the books whose native language is English and whose version is the first edition.

I'll use if I'm not mistaken the following XPath expression :

//book[@lang='eng' and @version='1']

and the following instructions in my code :

xmlChar * xpath_expression = "//book[@lang='eng' and @version='1']";
xmlXPathObjectPtr xpathRes = xmlXPathEvalExpression(xpath_expression, ctxt);

The problem is that I get as a result, the list of books as if I'd just do the following request:

//book

I wonder if my version is buggy knowing that I have the latest for my debian squeeze (2.7.8.dfsg-2 + squeeze7)...

score 0 · Answer 1 · answered Jun 20 '13 at 19:30

This is most certainly not a bug in libxml2. You probably made an error elsewhere. The following code only prints "Harry Potter":

#include <stdio.h>
#include <libxml/xpath.h>

int main()
{
    static const char xml[] =
        "<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?>\n"
        "<bookstore>\n"
        "    <book title=\"Harry Potter\" lang=\"eng\" version=\"1\">\n"
        "       <price>29.99</price>\n"
        "    </book>\n"
        "    <book title=\"Learning XML\" lang=\"eng\" version=\"2\">\n"
        "        <price>38.95</price>\n"
        "    </book>\n"
        "    <book title=\"Learning C\" lang=\"eng\" version=\"2\">  \n"
        "       <price>39.95</price>\n"
        "    </book>\n"
        "</bookstore>\n";

    xmlDocPtr doc = xmlParseMemory(xml, sizeof(xml));
    xmlXPathContextPtr ctxt = xmlXPathNewContext(doc);
    xmlChar *expression = BAD_CAST "//book[@lang='eng' and @version='1']";
    xmlXPathObjectPtr res = xmlXPathEvalExpression(expression, ctxt);
    xmlNodeSetPtr nodeset = res->nodesetval;

    for (int i = 0; i < nodeset->nodeNr; i++) {
        xmlNodePtr node = nodeset->nodeTab[i];
        xmlChar *title = xmlGetProp(node, BAD_CAST "title");
        printf("%s\n", title);
    }   

    xmlXPathFreeObject(res);
    xmlXPathFreeContext(ctxt);
    xmlFreeDoc(doc);

    return 0;
}

I've realized that the returned response is ambiguous because although I get a single element, I can still access the other elements by parsing the returned node. — guitarooman, Jun 21 '13 at 07:17
That's how libxml2 works. You get the result of an XPath query in an `xmlXPathObject` which in this case contains a node-set. The nodes in this node-set point to the nodes in the original XML document. — nwellnhof, Jun 21 '13 at 18:30

libxml2 predicates in xpath expression are not always recognized

1 Answers1