- Cleaned a little bit our parser: removal of old prelude constructs

- arrow.ml* to define basic builder for arrow (node, name, ...)
- machine_code_common similar to corelang but for machine_code (printers, some builders, ...)
- machine_code restricted to the translatation from normalized nodes to machines

Structs defining machines are now in machine_code_types

- Renamed type declarations as lustre_types and machine_code_types

Changed selection of files in odocl

Some issues wrt machine type features.
Work in progress

[general] large modification: added machine types, a second typing phase dealing with machine types (eg uint8)
          typing was transformed as a functor and parametrized by basic types (int/real/bool)
          it can also be applied multiple times on the same program

          Solved the misrenaming of imported nodes (eg int_to_real)

[main] node locals are now sorted according to their dependencies wrt clocks. The produced lustre node with types shall now be compilable

[main] enum typedef in C use the original lustre filename as identifier. This commit cleans the filename to remove dots.

	- mutations that could not be performed (ie. changing an integer constants when no integer constant appear in the program)
	- infinite computation of mutants, because of unproper randomization.

Improved the Cmake Lustre_Compile function

[stateflow] some progress, linking the parsed json to lustrec engine. Some variables are not yet typed. To investigate ...

[general] adding more entry rules for lustre parser to extract expressions, vdecl_list and statement list

[global] reordered local vars, keeping the declared ones before the others. Was mandatory for clocked expressions in compiled automata.