Added support for XPath and DOMParser

[christophe-pietquin]

This merge request adds support for XPath in order to allow testing REST/XML APIs in Newman using XPath

[shamasis]

@christophe-pietquin - i see the value in XPath (and probably JPath), but XMLDom is too heavy to be loaded everytime (and we are already struggling to remove backbone and other less relevant packages now.) We have cheerio to do the same thing as XMLDom and is already very popular. Let me know your thoughts.

[christophe-pietquin]

@shamasis I understand your concern about heaviness. The question is than how would you write a test script with xpath assertions that would run in both Postman and Newman if using cheerio?

This merge request allow us to create test script like the following one which detects if it is ran from Postman or Newman and uses the proper objects to execute the xpath queries:

xmlDoc = new DOMParser().parseFromString(responseBody, 'text/xml');
var xpathImpl;
var XPathResultImpl;
if (typeof xpath != "undefined") {
   xpathImpl = xpath;
XPathResultImpl = xpath.XPathResult;

} else { 
xpathImpl = document; 
 XPathResultImpl = XPathResult;
}
var docs = xpathImpl.evaluate("/response/results/total", xmlDoc, null, XPathResultImpl.STRING_TYPE, null); 
tests["total number of matching documents is returned"] = docs.stringValue.toLowerCase() === '1';