michael/lean2
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

feat(kernel/normalizer): avoid unnecessary creation of closures for n-ary functions

Browse source 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
This commit is contained in:
Leonardo de Moura 2013-12-17 17:38:04 -08:00
parent 10f28c7bec
commit 23e518001a
1 changed files with 9 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

14
src/kernel/normalizer.cpp
View file

@ -220,18 +220,22 @@ class normalizer::imp {
while (true) {
if (is_closure(f) && is_lambda(to_closure(f).get_expr())) {
// beta reduction
expr const & fv = to_closure(f).get_expr();
expr fv = to_closure(f).get_expr();
value_stack new_s = to_closure(f).get_stack();
while (is_lambda(fv) && i < n) {
new_s = extend(new_s, normalize(arg(a, i), s, k));
i++;
fv = abst_body(fv);
}
{
freset<cache> reset(m_cache);
flet<context> set(m_ctx, to_closure(f).get_context());
value_stack new_s = extend(to_closure(f).get_stack(), normalize(arg(a, i), s, k));
f = normalize(abst_body(fv), new_s, k);
f = normalize(fv, new_s, k);
}
if (i == n - 1) {
if (i == n) {
r = f;
break;
}
i++;
} else {
buffer<expr> new_args;
new_args.push_back(f);