Hi Tom,

Thanks for writing this up for the mailing list.

Another way of interpreting this change is:

A UTS-18 regex implementation is not locale-dependent.

Since many of the current challenges in the current std::regex are related to locale-dependent behaviour, this represents an opportunity for a huge simplification in specification and implementation.

Best regards,

                               Peter

From: SG16 <sg16-bounces@lists.isocpp.org> On Behalf Of Tom Honermann via SG16
Sent: 22 January 2021 15:45
To: SG16 <sg16@lists.isocpp.org>
Cc: Tom Honermann <tom@honermann.net>; Hana Dusíková <hanicka@hanicka.net>
Subject: [SG16] Changes to Unicode UTS #18, Unicode Regular Expressions, level 3 (tailoring) dropped.

EXTERNAL MAIL

Thanks to Peter Brett for pointing this out to me the other day.  The latest revision of the Unicode UTS#18 covering regular expressions has some significant changes.  In particular, the level 3 specification covering tailoring has been dropped.

The current version of the UTS is available at https://www.unicode.org/reports/tr18/tr18-21.html.

Diff marks relative to the prior version can be viewed at https://www.unicode.org/reports/tr18/tr18-20.html.

These changes appear to have been approved at UTC meeting #162 in May, 2020.  Minutes are available at https://www.unicode.org/L2/L2020/20015.htm; search for "UTS #18".  The relevant papers covering the changes appear to be:

• L2/19-364R2, Feedback on Proposed Update UTS #18, Markus Scherer, 2010-10-08.
• L2/20-017, Comments on Public Review Issues.

• See the link to issue 404.

• L2/20-032, Document for review, Mark Davis and Andy Heninger, 2019-11-08.
• L2/20-051, Response to feedback on UTS18, Mark Davis, 2020-01-14.

The upshot for us is that, when we get back around to looking at std::regex replacements, we won't have to spend time worrying about support for tailoring.

Tom.