Find correct client rectangles for image inclosed in svg tag

[kimtuck]

I have seen some books where the title page is a single image enclosed in a svg tag, something like

<svg:svg ....> <svg:image ....></svg:image> </svg:svg>

In this case, the getNormalizedRectangles code finds the svg:image tag, and calculates rectangles based on that. But the element that needs to be examined for the rectangle size is the enclosing svg tag.

[danielweck]

@jccr is that something EvidentPoint has already looked into?