I need to render one element from an SVG to an image. It needs to be like it appears in the SVG, without further transforms.
The only way I can think of to achieve this, is to set the SVG's viewbox on the target element. To do this, I need to compute the element's bounding box relative to the SVG viewBox.
I've tried to use getBBox(), but I don't think it can work. Its coordinates are relative to the transformed element: the transforms of the element and its parents aren't applied. It's not possible to apply those transforms to the bbox since the transformed-bbox might not contain the object anymore: for example think what would happen if the bbox need to rotate by 45 degrees.
An alternative could be getBoundingClientRect(). However it represents the bounding box of the element relative to the screen coordinates and again it might not be enough to apply the transforms from the screen to the SVG viewport, since they could transform the bounding box in a way that breaks it.
The last idea I have is to create a new <svg> element (with no transform), place in it one (or more) <g> whose function is to apply all the transform from the original SVG's viewBox to the target element's, then place (a copy of) the target element into it and eventually call getBBox() on the newly created SVG element: if I'm not mistaken, the resulting bbox should be what I need and I can set it as the viewBox of the original SVG. Although it might not work if the target element contains some <use> elements or similar stuff. I think it'll take quite some time to implement this and I'm afraid it will be wasted time since it won't work for whatever reason.
Are there other ways to obtain an element's bounding box relative to the SVG viewBox? Otherwise other ways to correctly render one element from an SVG to an image?