Ensures that let-in's in arities of inductive types work well. Maybe not

very useful in practice but as soon as let-in's were not forbidden in
the internal data structure, better to do it. Moreover, this gets
closer to the view were inductive definitions are uniformly built from
"contexts". (checker not changed!)

git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12273 85f007b7-540e-0410-9357-904b9bb8a0f7

1 files changed, 2 insertions, 0 deletions

diff --git a/kernel/indtypes.ml b/kernel/indtypes.ml
index 4aac096fc..ccf9b3f6c 100644
--- a/kernel/indtypes.ml
+++ b/kernel/indtypes.ml
@@ -588,6 +588,7 @@ let build_inductive env env_ar params isrecord isfinite inds nmr recargs cst =
   (* Compute the set of used section variables *)
   let hyps =  used_section_variables env inds in
   let nparamargs = rel_context_nhyps params in
+  let nparamdecls = rel_context_length params in
   (* Check one inductive *)
   let build_one_packet (id,cnames,lc,(ar_sign,ar_kind)) recarg =
     (* Type of constructors in normal form *)
@@ -629,6 +630,7 @@ let build_inductive env env_ar params isrecord isfinite inds nmr recargs cst =
 	mind_arity = arkind;
 	mind_arity_ctxt = ar_sign;
 	mind_nrealargs = rel_context_nhyps ar_sign - nparamargs;
+	mind_nrealargs_ctxt = rel_context_length ar_sign - nparamdecls;
 	mind_kelim = kelim;
 	mind_consnames = Array.of_list cnames;
 	mind_consnrealdecls = consnrealargs;