histval_file initval_file refactoring

Finally finished a project started last December.

@MichelJuillard Thanks for this work. I guess it is part of the resolution of dynare#1671 (closed).

Could you possibly also open a merge request on the main Dynare repository with the corresponding changes to the MATLAB code (and ideally also to the documentation)? I'd like to review and accept them in tandem, since they’re strongly coupled. I’m aware that it’s difficult to have the testsuite pass until I accept the preprocessor changes, because of the submodule issue, but let’s ignore that problem for the time being.

Also, should I squash your preprocessor changes into a single commit? (I think so, since they seem to form a single logical unit).

@sebastien and @MichelJuillard Sorry to ask, but where is the Matlab code? Last week I have added, in the enterprise branch, the possibility to set the initval_file with a dseries object saved in a .mat or .m file... I'd like to compare.

The Matlab code is in git@git.dynare.org:MichelJuillard/dynare.git branch initval_histval I still have to resolve some divergency with my local branch. I will make the Merge Request tonight or tomorrow morning, You can squash the commits, I had checked the box on the Merge Request. Stephane, I made a design where I use only dseries for initval and histval and the file loading mechanism included in dseries,

@sebastien I opened a merge request for the Matlab code

merged

mentioned in commit d05ffde6