The dependency on the clock was lost when the enum is translated into
an integer (because some expression were still refering to the enum type).

The fix was to index VarMap.t by string instead of Soc.var_expr in ActionsDeps.

Remove a lot of warnings (considered as errors by dune).

Indeed, in order to sort Soc.gao, I currently process in 3 stages ;
1. compute a total ordering of Actions.t according to the deps
2. transform Actions.t into Soc.gao (List.map)
3. factorize Soc.gao by looking at consecutive gao that have the same guard

With this change, between steps 1 and 2, I reorder the list of actions
so that group of independent actions are sorted according to their
clock. The idea is of course to provide to step 3 more opportunities
of factorization.

The result is not very convincing yet.

I've also tried to sort the actions wrt their clock before performing
the toposort. Seems not too bad.

I've kept the old behavior by default though.

Indeed it is possible when each branch of the ite updates no memory.
This is done in the new L2lOptimIte module.

For the time being, it does detect when the node has no memory. It
only looks at the declaration: nodes have memory, and not functions.

I should infer that information and raise warnings or errors if what
I infer is not compatible with waht is declared (will come later).

Also split ActionsDep into ActionsDep and Action.

Also fix a bug in L2lsplit where deeply nested (>2) merge were not
splitted.

Indeed, computing the dependencies and choosing an order for actions
(that may optimize test openning in the near futurea) are 2 different
things.

The main difference is that I use a map instead of list to store
visited nodes.  The compilation time of normal.ec is divided by 2!!
(35s->15s)

We do not really see the difference in the time of the non-reg tests
since the execution of normal.ec still exceeds the 10s timeouts.

I took the data type definition from SocExecValue to define a Data
module that is a strict extension of the Lutin Data module.

The duplicated Genlex module is now also identical to the one of
Lurette.

Quite boring, but simple thanks to the ocaml type system.

nb : programs do run, but I did not check that they run correctly...

Also fix a regression introduced in the previous change where
incorrect ec code was generated for diese.

Indeed, the data structure in output of the lis2loc parser was very
similar to Lic.t/LicPrg.t. Hence, this shunt.

At this stage it compiles, but it's untested and does not work.