Added proof_state and functions to control it in controller_itp.
Tried to add a test case for it inside the command p to put a big
P next to a goal that is proved.
It is currently buggy. Needs investigating and cleaning.

Struggling with names. Removed the print_task which looked at the driver
before printing for uniformity. If I dont do that the printing function
I use to have a safe naming environnment inside tasks are not the same
as the printing function we use for display. So hypothesis displayed dont
actually exists and other bugs.

I will try to see if my changes (in particular clean_environnment and
gen_ident) are necessary.

Modifications in Pretty to get real unique names for hypothesis.

I think we need LIFO and not FIFO.

or not before parsing with wrap. So we give a Tformula instead of Tterm
constructor for those transformations.

+ modify the command that print the sessino tree to specify where we are in the tree with "**"
+ add some more comprehensible error messages to wraper function for parsing

"parsing" of these strings is delayed in the transformations themselves,
because in case such an argument must be interpreted as a term, the
task itself is needed do perform name resolution.

Incidentally, saving transformations with arguments in sessions is not an
issue anymore

Only works with hypothesis of the form forall x. A -> B.
Also cleaning the code a little.

like to add more from Ident and Pretty but keeped minimal changes in those
files.
TODOs to be discussed.

Added cut.
Added exists.
Should work by calling ttr "case" term.
Added a function typing_terms that should not be used and which exists
to avoid duplicating parse_transformation_arg.

Extracted a function that only parses arguments and return a term from
test_transformation_with_args.

For support for open term, one should somehow rebuild a namespace from the given task

Adapted work on nearest_goal_right to be able to use it why3shell.
Now should be able to apply transformation and proof on current goal. Also
able to print the current tree and goal. Requires testing.