[cxx-abi-dev] mangling of UCNs

[Martin v. Löwis]

scott douglass <sdouglass at arm.com> writes:

> The mangling of UCNs is not specified by the ABI.  I think it should
> be.  Perhaps the idea is that the mangling of UCNs would have to be
> the same as an implementation's C99 implementation?

I think there are two reasons why this is not specified:
- there was no prior experience in the committee,
- there were contradicting approaches.

However, there was consensus that it should be the same as in C.

> Anyway, here's a fairly simple proposal.  

I think any approach involving mangling is more complicated than
necessary. Here is my proposal:

Identifiers are encoded in UTF-8.

That's it. It's nicely integrates C, C++, and, if needed, Java.
It is also the "right thing", IMO.

Regards,
Martin