In addition, as XSLT becomes more pervasive--particularly in the Web services arena--the speed at which an engine can perform transformations is critical. Hardware-based XSLT engines, such as DataPower's XA35, are crucial to the success of your XML-based projects. But to take advantage of such products, an editor must be able to use an alternative external XSLT engine.
Full Debugger
An XSL transformation may include data manipulation, such as mathematical calculations, iterations over sequences of elements and even the restructuring of data elements within the original XML document. These processes are no different than those used with C or Java code, so it's imperative that your XSLT editor include a full debugger. When things go wrong, you'll want to be able to walk through the transformation. Though "print" statements littered throughout your code is one method of debugging, it's not efficient. A full debugger with step capabilities, break points and watch points makes the debugging process immeasurably easier.
A large portion of XSLT consists of finding specific elements within an XML document, then performing some sort of manipulation on the element or subelements. To search efficiently through an XML document, which can be large, many editors use XPath.
XPath queries can be complicated and may involve the use of specific functions, such as local-name() and position(). Although XSL provides some mechanisms for manipulating elements, it is rare to find XSLT without XPath, which provides the basic syntactical constructs necessary to select and address nested elements.