The below equation looks terrible in JEuclid, while the MathML is correct (er, except for the annotation being in the wrong place - I wonder how that happened - probably doesn't matter though).
If JEuclid doesn't do the longdiv notation, I'm not sure if it would be possible to do some pre-transformation to make it into something that might work.
(Note: This may also benefit from LaTeX rendering support if we add that and if it includes this structure.)
It turns out the rendering is OK, but it doesn't make the bounding box big enough to actually fit in the line above, so it gets cropped from the result image.
To fix this I have done as follows with each prior to processing:
Replace it with an mrow containing an mspace (1) and an mover, which contains the original menclose and another mspace (2).
mspace 2 is always empty (no attributes). mspace 1 is empty if the longdiv is the first thing in its parent container (so as not to add blank space at the start if you begin with this structure for some reason), otherwise has width 4px.
The below equation looks terrible in JEuclid, while the MathML is correct (er, except for the annotation being in the wrong place - I wonder how that happened - probably doesn't matter though).
If JEuclid doesn't do the longdiv notation, I'm not sure if it would be possible to do some pre-transformation to make it into something that might work.
(Note: This may also benefit from LaTeX rendering support if we add that and if it includes this structure.)