GitLab

As a side effect, provers-detection-data.conf no longer needs to be preprocessed.

why3 tests/test-eval.why -P alt-ergo -T T --eval c1 --eval c2 --eval c3

in a series of tests on available examples, lightweight polymorphism
encoding techniques combined with a more agressive generation of
monomorphic instances (via Discriminate) show better results than
our earlier defaults.

In the same commit:
- remove Encoding_explicit (unsound), Encoding_decoexp (too naive),
  and Encoding_instantiate (subsumed by Discriminate)
- rename Encoding_decorate to Encoding_tags_full and Encoding_guard
  to Encoding_guards_full
- move Encoding_guards_full specific functions from Libencoding to
  Encoding_guards_full
- do not apply type protection in "encoding_tptp" and remove
  Protect_finite which is not needed anymore.

[GTK sourceview] why.lang renamed to why3.lang

This fixes the bug with .dep cleaning, when we don't build the IDE.

Also, can we merge the replayer into why3session already, please?

- all programs with sessions are in examples/
- all programs without sessions are in examples/in_progress/
  (if you have private sessions for those, just move them there)
- all pure logical problems are in logic/
  (to simplify bench scripts and gallery building; they are few anyway)
- all OCaml programs are in examples/use_api/
- all strange stuff is in examples/misc/
  (most of it should probably go)
- Claude's solutions for Foveoos 2011 are in examples/foveoos11-cm/
  (why do we need two sets of solutions for quite simple problems?)
- hoare_logic, bitvectors, vacid_0_binary_heaps are in examples/

Bench scripts and documentation are updated.
Also, bench/bench is simplified a little bit.

the Set signature is moved to a separate Extset module,
instead of being included into Extmap.S. The module hierarchy
becomes simpler, we don't shadow  OCaml standard modules anymore,
and the bug #15270 is fixed, too.