diff options
| author |  Emilio Jesus Gallego Arias <e+git@x80.org> | 2017-11-19 02:55:51 +0100 |
|---|---|---|
| committer | Emilio Jesus Gallego Arias <e+git@x80.org> | 2017-11-19 03:03:36 +0100 |
| commit | 8cada511701d8893bab5553470ab721b33713043 (patch) | |
| tree | cb9cc69cdb9c89b973c923f753e85b68a1ac27f3 /plugins/funind | |
| parent | edf1a8f36f75861b822081b3825357e122b6937d (diff) | |
[proof] Attempt to deprecate some V82 parts of the proof API.
I followed what seems to be the intention of the code, with the
original intention of remove the global imperative proof state.
However, I fully fail to see why the new API is better than the old
one. In fact the opposite seems the contrary.
Still big parts of the "new proof engine" seem unfinished, and I'm
afraid I am not the right person to know what direction things should
take.
Diffstat (limited to 'plugins/funind')
| -rw-r--r-- | plugins/funind/recdef.ml | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/plugins/funind/recdef.ml b/plugins/funind/recdef.ml index 2fdc3bc37..b8d41d539 100644 --- a/plugins/funind/recdef.ml +++ b/plugins/funind/recdef.ml @@ -1225,8 +1225,8 @@ let whole_start (concl_tac:tactic) nb_args is_mes func input_type relation rec_a let get_current_subgoals_types () = let p = Proof_global.give_me_the_proof () in - let { Evd.it=sgs ; sigma=sigma } = Proof.V82.subgoals p in - sigma, List.map (Goal.V82.abstract_type sigma) sgs + let sgs,_,_,_,sigma = Proof.proof p in + sigma, List.map (Goal.V82.abstract_type sigma) sgs exception EmptySubgoals let build_and_l sigma l = |