Laboratoire ICAR » Plateforme TXM

Currently, the XPathResult.groovy script/class API provides only two 'all in memory' model methods:

• getXpathResponse returns the first match of an XPath
• getXpathResponses returns all the string values of the matches of an XPath as an array of strings

To be able to process large XML files with lots of elements matching a given XPath, we need to add to the API a way not to store the intermediate array of strings in memory. The new method can limit the memory consumption used to transmit the result to only the one used for the list of DOM result nodes.

Solution¶

Add a new streamingGetXpathResponses method with the code of getXpathResponses where:

        NodeList nodes = (NodeList) result;
        for (int i = 0; i < nodes.getLength(); i++) {
            //println nodes.item(i)
            xresult.add((nodes.item(i).getNodeValue()));
        }
        return xresult

is replaced by:

        NodeList nodes = (NodeList) result
        return nodes

The Groovy code calling streamingGetXpathResponses can then iterate the results with:

nodes.each { println it.getNodeValue() }