Proof engine: support for nesting tactic-in-term within other tactics.

Tactic-in-term can be called from within a tactic itself. We have to
preserve the preexisting future_goals (if called from pretyping) and
we have to inform of the existence of pending goals, using
future_goals which is the only way to tell it in the absence of being
part of an encapsulating proofview.

This fixes #6313.

Conversely, future goals, created by pretyping, can call ltac:(giveup) or
ltac:(shelve), and this has to be remembered. So, we do it.

1 files changed, 20 insertions, 0 deletions

diff --git a/test-suite/bugs/closed/6313.v b/test-suite/bugs/closed/6313.v
new file mode 100644
index 000000000..1715e7f90
--- /dev/null
+++ b/test-suite/bugs/closed/6313.v
@@ -0,0 +1,20 @@
+(* Former open goals in nested proofs were lost *)
+
+Inductive foo := a | b | c.
+Goal foo -> foo.
+  intro x.
+  simple refine (match x with
+                 | a => _
+                 | b => ltac:(exact b)
+                 | c => _
+                 end); [exact a|exact c].
+Abort.
+
+(* Another example *)
+
+Goal (True/\0=0 -> True) -> True.
+  intro f.
+  refine
+   (f ltac:(split; only 1:exact I)).
+  reflexivity.
+Qed.