[RFC nft] evaluate: kill anon sets with one element

[Florian Westphal]

pretends that
ip saddr { 1.1.1.1 }

is
ip saddr 1.1.1.1

Needs more work (breaks dumps in test cases afaics).
Is that a good idea in first place?

I see various new users adopting

{ single-value }

probably due to copying/adapting a "{ foo, bar }" rule to "{ foo }"...
---
 src/evaluate.c | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/src/evaluate.c b/src/evaluate.c
index 0bda431d5a16..e922b7f9fb26 100644
--- a/src/evaluate.c
+++ b/src/evaluate.c
@@ -1251,11 +1251,13 @@ static int expr_evaluate_set_elem(struct eval_ctx *ctx, 
struct expr **expr)
 static int expr_evaluate_set(struct eval_ctx *ctx, struct expr **expr)
 {
        struct expr *set = *expr, *i, *next;
+       unsigned int count = 0;
 
        list_for_each_entry_safe(i, next, &set->expressions, list) {
                if (list_member_evaluate(ctx, &i) < 0)
                        return -1;
 
+               count++;
                if (i->ops->type == EXPR_SET_ELEM &&
                    i->key->ops->type == EXPR_SET_REF)
                        return expr_error(ctx->msgs, i,
@@ -1286,6 +1288,22 @@ static int expr_evaluate_set(struct eval_ctx *ctx, 
struct expr **expr)
                        set->set_flags |= NFT_SET_INTERVAL;
        }
 
+       if (!ctx->set && count == 1) {
+               i = list_first_entry(&set->expressions, struct expr, list);
+               if (i->ops->type == EXPR_SET_ELEM) {
+                       switch (i->key->ops->type) {
+                       case EXPR_RANGE:
+                       case EXPR_VALUE:
+                               *expr = i->key;
+                               i->key = NULL;
+                               expr_free(set);
+                               return 0;
+                       default:
+                               break;
+                       }
+               }
+       }
+
        set->set_flags |= NFT_SET_CONSTANT;
 
        set->dtype = ctx->ectx.dtype;
-- 
2.20.1