From 490d8bb242d33213b948485f5b94c55e22cf86a6 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jan=20Kundr=C3=A1t?= Date: Thu, 21 Nov 2024 11:32:44 +0100 Subject: [PATCH 03/18] remove a misleading comment MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Organization: Wires The whole intention within action's input/output handling here was to put some emphasis on the fact that we aren't tracking the input/output nodes directly. However, looking at all the other classes this is a bit redundant, we're using a pattern like this all the time. Just drop the comment. Change-Id: Ibd9bf9f1e83c650dda3bc43ef48e61dd6d95da5a Signed-off-by: Mattias Walström --- src/SchemaNode.cpp | 3 --- 1 file changed, 3 deletions(-) diff --git a/src/SchemaNode.cpp b/src/SchemaNode.cpp index 81e938f..9934cea 100644 --- a/src/SchemaNode.cpp +++ b/src/SchemaNode.cpp @@ -640,9 +640,6 @@ bool List::isUserOrdered() const */ ActionRpcInput ActionRpc::input() const { - // I need a lysc_node* for ActionRpcInput, but m_node->input is a lysp_node_action_inout. lysp_node_action_inout is - // still just a lysc_node, so I'll just convert to lysc_node. - // This is not very pretty, but I don't want to introduce another member for ActionRpcInput and ActionRpcOutput. return ActionRpcInput{reinterpret_cast(&reinterpret_cast(m_node)->input), m_ctx}; } -- 2.43.0