Run Jupyter Kernels only if necessary
In the process of creating a LaTeX document, many edit-compile cycles happen.
To speed up the compilation, it is customary to run only necessary tools.
Is it possible to cache the Jupyter kernel results, store a hash value for each Jupyter cell block, which allows to detect changes to Jupyter cell blocks and a "simple" pdflatex run is sufficient, if no changes are detected, to compile the document using cached results?
In the case of using several Jupyter kernels, it should be possible to run only the necessary subset of Jupyter kernels to update the document.
Edited by kiryph