Closed IOBYTE closed 6 years ago
For some reason, I don't get notified of new issues when they're created, so I didn't see this until just now just after I published cdecl 4.5 that contains support for C++11's using
declarations. However, I forgot to add support for explaining using
declarations. I'll get around to it.
I also discovered that there are some broken using
cases like the one you gave. :(
OK, I just added support for explain using
name =
gibberish.
I also realized that the breakage isn't limited to using
. It fails to parse array declarations without a name, e.g., explain void f(char[])
. If you insert a name like explain void f(char s[])
then it works.
Unfortunately, it's not an easy fix and will likely require some serious reworking of parts of the grammar so as not to trigger a gazillion more shift/reduce conflicts (as well as new reduce/reduce conflicts).
With 671 unit tests, you'd think I would have caught this earlier; but no. :(
Unfortunately, it's not an easy fix and will likely require some serious reworking of parts of the grammar so as not to trigger a gazillion more shift/reduce conflicts (as well as new reduce/reduce conflicts).
It turns out that the fix was simpler than I thought. It's fixed in 4.6.1.
Are there any plans to add C++ using type aliases?