When producing pdf directly, PostScript code cannot be used, which means that pstricks macros do not work. Now, it is still possible to include pstricks code using external tools (running LaTeX on files built from the document's preamble and the figure's contents), but not as seamlessly as we could want.
My aim is to fix this, by implementing a PostScript interpreter in TeX. Before starting such a large project (the part of the PostScript standard that deals only with the programming structures is ~150 pages long), I would like to make sure that this will not be in vain.
If I understand correctly, pstricks macros eventually insert the PostScript code as \special with a certain structure, so in principle I should be able to redefine that primitive to send PostScript special to my interpreter.
The step that is not clear to me is how to go from there to actually producing graphics: presumably I would be using pgf. This leads me to a few related questions:
Is it enough to simply hook into
\special, or should I also change thepstricksmacros that inserts spacing around figures, for instance?Is
pgfpowerful enough to express all elementary graphical operations of PostScript?Will I need a general purpose (e)ps to pdf converter?
pst-funcalone. There are other ODE solving possibilities etc. too. Do you plan to replicate the PSTricks overhead in the interpreter? Though I don't get how that would work, sounds just amazing. I'll wait for Herbert's response but still great initiative. – percusse Apr 23 '13 at 00:43