I agree that Hubert's last option is the one we want here.

> encoding schemes that support escape sequences or some other form of
markup should be excluded

Could you give an example of such encoding scheme?

- Victor

On Wed, Feb 26, 2020 at 7:22 AM Tom Honermann <tom_at_[hidden]> wrote:

> On 2/26/20 2:39 AM, Corentin Jabot via Lib wrote:
>
>
>
> On Wed, 26 Feb 2020 at 02:38, Hubert Tong via SG16 <sg16_at_[hidden]>
> wrote:
>
>> In [format.string.std], the meaning of "Unicode encoding" added by
>> P1868R2 is unclear. One interpretation of what was meant by "Unicode
>> encoding" is "UCS encoding scheme" (as defined by ISO/IEC 10646). Another
>> interpretation is an encoding scheme capable of encoding all UCS code
>> points that have been assigned to characters. Yet another interpretation is
>> an encoding scheme capable of encoding all UCS scalar values.
>>
>
> Any encoding scheme capable of encoding any UCS codepoint - where a UCS
> codepoint is tautologically any value in the UCS codespace (U+0 -
> U+10FFFF)
>
> I'd like Victor to weigh in, but I think Hubert's last interpretation
> meets minimum requirements. It might be simplified to "any encoding scheme
> that encodes (only) UCS scalar values" (encoding schemes that support
> escape sequences or some other form of markup should be excluded).
>
> I think the definition/description of "Unicode Character Encoding Schemes"
> in UTR #17 section 5 (
> https://www.unicode.org/reports/tr17/#CharacterEncodingScheme) fits what
> we're after.
>
> Tom.
>
>
>
>> --
>> SG16 mailing list
>> SG16_at_[hidden]
>> https://lists.isocpp.org/mailman/listinfo.cgi/sg16
>>
>
> _______________________________________________
> Lib mailing listLib_at_[hidden]
> Subscription: https://lists.isocpp.org/mailman/listinfo.cgi/lib
> Link to this post: http://lists.isocpp.org/lib/2020/02/15467.php
>
>
>