XMLTools.selectNodes() problem

Technical Q&A
1

I am hitting a website that supports xml downloads.

here is a smple link:

var url = http://clinicaltrials.gov/search?intr=Cetuximab&displayxml=true
XMLTools.loadXML(url, "ctTool.parseXML(xmlDoc, xmlText)");

I process this in the following method:

parseXML: function(xmlDoc, xmlText) {
    	if (xmlDoc == null) {
    		return;
    	}
    	var studies = XMLTools.selectNodes(xmlDoc, "//clinical_study");
    	Log.logInfo("Number of studies found: " + studies.length);
    	var trialList = [];
    	for (var i = 0; i < studies.length; i++) {
    		var study = studies[i];
    		var trial = new Object();
    		Log.logInfo("found study: " + XMLTools.selectString(study, "//nct_id"));
    		trial.NCTID = XMLTools.selectString(study, "//nct_id");
    		trial.title = XMLTools.selectString(study, "//title");
    		trial.url = XMLTools.selectString(study, "//url");
    		trialList[trialList.length] = trial;
    	}
}

The loop seems to get stuck on the first element. I get a count of 20 “clinical_trial” elements back in the xmlDoc (which is correct) but in the Log.logInfo() I get the first study back 20 times.

Thoughts?
-pf

2

In XPath, “//” is always document-relative, even if executed in the scope of a specific node. You want “.//” to search under the target node.