P1949 C++ Identifier Syntax using Unicode Standard Annex 31

[wg21bot]

P1949R0 C++ Identifier Syntax using Unicode Standard Annex 31 (Steve Downey)

[tahonermann]

SG16 in Belfast: Needs revision.

SG16 reviewed a draft of this paper in Belfast during our discussions of NL029. The author was present. We're looking forward to a revision for C++23 that addresses C compatibility, implementation experience, and that has wording.

SG16 minutes are available with the minutes for the discussion of NL029 at:

• http://wiki.edg.com/bin/view/Wg21belfast/SG16NBNL029

[wg21bot]

P1949R1 C++ Identifier Syntax using Unicode Standard Annex 31 (Steve Downey, Zach Laine, Tom Honermann, Peter Bindels)

[tahonermann]

Removed the SG7 label per email discussion; SG7 related concerns will be discussed via P1953 (#690).

[tahonermann]

SG16 approved P1949R1 in Prague. Minutes are at http://wiki.edg.com/bin/view/Wg21prague/SG16P1949R1.

P1949R1: Implementations should implicitly normalize identifiers to NFC

SF	F	N	A	SA
1	2	2	3	0

Consensus? No

P1949R1: non-NFC normalized identifiers should be ill-formed.

SF	F	N	A	SA
4	4	0	0	0

Consensus? Yes

P1949R1: We agree with the stated UAX31 conformance requirements.

SF	F	N	A	SA
2	4	1	0	0

Consensus? Yes

P1949R1: Encourage exploring incorporation of UAX39; exclusion of confusables.

SF	F	N	A	SA
0	3	2	0	1

Consensus? Weak - authors prerogative

P1949R1: Forward to EWG

SF	F	N	A	SA
4	2	1	0	0

Adding the needs-revision tag to address SG16 guidance and to provide wording. SG16 guidance was to specify that non-NFC normalized identifiers are ill-formed.

[erichkeane]

EWGI in Prague

Given the time constraints of the committee, should we spend additional committee effort on
P1949?

SF	F	N	A	SA
12	3	1	0	0

Should implementers be allowed to extend the list of allowable characters as a conforming
extension?

SF	F	N	A	SA
1	1	7	4	0

We believe that P1949R1 has been sufficiently developed and (pending the feedback from the
room and SG16 are merged) this paper is ready for EWG.

SF	F	N	A	SA
8	6	1	0	0

[jfbastien]

D1949R2 C++ Identifier Syntax using Unicode Standard Annex 31

Paper is the same as P1949R1 from the mailing, plus wording and NFC.

D1949R2 is tentatively ready to go to CWG for C++23.

SF	F	N	A	SA
12	8	1	0	0

[wg21bot]

P1949R2 C++ Identifier Syntax using Unicode Standard Annex 31 (Steve Downey, Zach Laine, Tom Honermann, Peter Bindels)

[wg21bot]

P1949R3 C++ Identifier Syntax using Unicode Standard Annex 31 (Steve Downey, Zach Laine, Tom Honermann, Peter Bindels, Jens Maurer)

[jfbastien]

This has been updated since marked at "tentatively ready", so we'll need to revisit.

[tahonermann]

Removing the SG16 label as this proposal is now in EWG's hands. I trust the EWG chair will involve SG16 should further review warrant it.

[wg21bot]

P1949R5 C++ Identifier Syntax using Unicode Standard Annex 31 (Steve Downey, Zach Laine, Tom Honermann, Peter Bindels, Jens Maurer)

[wg21bot]

P1949R6 C++ Identifier Syntax using Unicode Standard Annex 31 (Steve Downey, Zach Laine, Tom Honermann, Peter Bindels, Jens Maurer)

[jfbastien]

R6 was discussed in today's EWG telecon.

POLL: P1949R6 is tentatively ready to move to CWG

SF	F	N	A	SA
13	6	2	0	0

[jfbastien]

EWG February 2021 polls:

Poll: Forward P1949R6 "C++ Identifier Syntax using Unicode Standard Annex 31" to Core.

SF	F	N	A	SA
19	6	3	0	0

Poll outcome: ✅ consensus.

Salient comments:

• The allow-list approach isn’t sustainable. There are too many characters allowed that can be actively harmful, such as RTL modifiers.
• Follows Unicode recommendations. WG21 shouldn’t be in the business of assigning meaning to Unicode characters, Unicode is authoritative in this case.
• Avoids surprising behavior that is difficult to defend.
• Good international language support encourages wider adoption and this wider adoption implies a larger global pool of resources for the development of the C++ ecosystem.

[jensmaurer]

CWG 2021-04-12: Approved D1949R7 for plenary vote.

[wg21bot]

P1949R7 C++ Identifier Syntax using Unicode Standard Annex 31 (Steve Downey, Zach Laine, Tom Honermann, Peter Bindels, Jens Maurer)

[AaronBallman]

This paper was discussed at the Jun 4, 2021 SG22 meeting. SG22 agreed with the direction (but no official poll was taken). A WG14 member has agreed to help bring a companion paper to WG14 on this topic, so SG22 does not need to see this paper again.

[wg21bot]

Adopted 2021-06.

[jensmaurer]

WG14 accepted http://www.open-std.org/jtc1/sc22/wg14/www/docs/n2836.pdf
(N2836 Identifier Syntax using Unicode Standard Annex 31) for C23.