From: Vladimír Vondruš Date: Thu, 7 May 2020 15:24:26 +0000 (+0200) Subject: documentation/doxygen: it calculates a hash of this func differently now. X-Git-Url: https://www.chiark.greenend.org.uk/ucgi/~cjwatson/git?a=commitdiff_plain;h=cdbaa02303f212ab4be0ea4afb398a6bfe18e066;p=blog.git documentation/doxygen: it calculates a hash of this func differently now. Um. I smell a hidden bug or A REGRESSION INCOMING. --- diff --git a/documentation/test_doxygen/cpp_function_attributes/structFoo.html b/documentation/test_doxygen/cpp_function_attributes/structFoo.html index 6a0a60ac..85ba0c58 100644 --- a/documentation/test_doxygen/cpp_function_attributes/structFoo.html +++ b/documentation/test_doxygen/cpp_function_attributes/structFoo.html @@ -64,7 +64,7 @@
Random type and constexpr together.
- auto baz() -> decltype(auto) constexpr noexcept + auto baz() -> decltype(auto) constexpr noexcept
decltype(auto) and constexpr together
@@ -106,9 +106,9 @@

Random type and constexpr together.

This is okay.

-
+

- decltype(auto) Foo::baz() constexpr noexcept + decltype(auto) Foo::baz() constexpr noexcept

decltype(auto) and constexpr together

For some reason, due to decltype(auto), Doxygen swaps the order, causing the constexpr to be hard to detect. Don't even ask how it handles trailing return types. It's just HORRIBLE.