Faster check for duplicated test cases

By avoiding `std::set`, we save 1 allocation per checked test case,
and the sort + pairwise check has much nicer cache behaviour, leading
to much faster check.

Overall, the check is only a tiny part of the startup overhead, but
the improvement is large enough that it ends up measurable in binaries
with many tests (e.g. `SelfTest`).

1 files changed, 6 insertions, 0 deletions

diff --git a/src/catch2/catch_test_case_info.cpp b/src/catch2/catch_test_case_info.cpp
index 9d64e532..d775b89a 100644
--- a/src/catch2/catch_test_case_info.cpp
+++ b/src/catch2/catch_test_case_info.cpp
@@ -259,4 +259,10 @@ namespace Catch {
         return lhs.tags < rhs.tags;
     }
 
+    bool operator==( TestCaseInfo const& lhs, TestCaseInfo const& rhs ) {
+        return lhs.name == rhs.name
+            && lhs.className == rhs.className
+            && lhs.tags == rhs.tags;
+    }
+
 } // end namespace Catch