s-barannikov wrote:

> > I made an assumption but never bothered to check with others: are we going 
> > to support multibyte encodings right away? I assumed not.
> 
> We don't have a reliable way in the code currently to identify whether an 
> encoding is multibyte or not. Such a mechanism would be needed to generate 
> appropriate diagnostics if we were to limit support.

Unless I'm missing something, this PR assumes the execution encoding is 
single-byte. In particular, convertBasicCharacter takes single byte. (It 
returns single character, but that is fine, since all conversion specifiers are 
in the basic character set, and thus should be one-byte encodable.)
To make it support mutlibyte characters, we would need to know how to advance 
the format string iterator (read: determine multibyte character "length") when 
advancing through the format string. That's a big change...


https://github.com/llvm/llvm-project/pull/196568
_______________________________________________
llvm-branch-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits

Reply via email to