Performance Profiling of XSLT Stylesheet
Whether you are trying to identify a performance issue that is causing your
production XSLT transformation to not meet customer expectations, or you are trying to
proactively identify issues prior to deploying your XSLT transformation, using the
XSLT profiler feature is essential to helping you save time and ultimately ensure a
better performing, more scalable XSLT transformation.
The profiling is available for the XSLT processors that are
bundled with oXygen (Saxon 6.5.5, Saxon EE, and Xalan 2.7.1). This allows you to
profile XSLT 1.0, 2.0, or 3.0 stylesheets.
Invocation Tree View
Using the Invocation Tree view, also known as a call tree, you can examine how
style instructions are processed in a top down manner. The profiling result shows
the duration time for each of the style instructions (including the time needed for
its child instructions and, when an instruction is expanded, how the instruction
time is composed from the child instruction times).
Using the profiler Hotspots view, you can immediately detect the time the
processor spent on each instruction. The hotspot displays the inherent time of the
instruction (the total time processing that instruction, minus the time processing
its child instructions). When a hotspot is expanded, you can see all the different
paths that instructions were called from (a reverse invocation tree), the number of
invocations, the instruction time on that path, and the percentage that path
contributed to the instruction's total execution time.