From b91d322b646edf753ca9f3659563e4d6c525d3f6 Mon Sep 17 00:00:00 2001 From: Matthieu Sozeau Date: Fri, 20 May 2016 03:40:46 +0200 Subject: Typeclasses: refine the eauto tactic - Treat shelved dependent subgoals that might not be resolved after some proof search correctly by restarting their resolution as soon as possible (if they are typeclasses in only_classes mode). - Treat dependencies between goals better, avoiding backtracking more often when dependencies allow. --- tactics/class_tactics.mli | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'tactics/class_tactics.mli') diff --git a/tactics/class_tactics.mli b/tactics/class_tactics.mli index 6b4ff8b5e..e9d915de1 100644 --- a/tactics/class_tactics.mli +++ b/tactics/class_tactics.mli @@ -36,6 +36,7 @@ val autoapply : constr -> Hints.hint_db_name -> tactic type newautoinfo = { search_depth : int list; last_tac : Pp.std_ppcmds Lazy.t; + search_dep : bool; search_cut : Hints.hints_path; search_hints : Hints.Hint_db.t } @@ -43,8 +44,8 @@ val new_hints_tac : bool -> Hints.hint_db list -> newautoinfo -> (newautoinfo -> unit Proofview.tactic) -> unit Proofview.tactic -val make_autogoal' : ?st:Names.transparent_state -> - bool -> +val make_autogoal' : ?st:Names.transparent_state -> + bool -> bool -> Hints.hints_path -> int -> ([ `NF ], 'c) Proofview.Goal.t -> newautoinfo val new_eauto_tac : ?st:Names.transparent_state -> -- cgit v1.2.3