Character sets & encodings (C++23)

Adopts Unicode UAX #31 (XID_Start/XID_Continue, NFC) for identifiers, disallowing e.g. emoji (DR).

Makes concatenation of string literals with different encoding prefixes ill-formed.

Requires diagnostic messages to preserve source characters rather than mangling them into an implementation-defined form.

Rewrites the description of character sets and encodings (translation/execution encodings, code units) in Unicode terms matching implementations.

Requires a character literal to have the same value in #if as in the same C++ expression.

Adds named universal character escapes: \N{LATIN SMALL LETTER A WITH ACUTE}.

Adds delimited escape sequences \x{…}, \o{…}, \u{…}, avoiding the greedy-hex problem.

Requires implementations to accept UTF-8 as a portable source-file encoding.

Makes wide character literals with multiple or non-encodable characters ill-formed.

Relaxes the requirement that wchar_t can hold every character of every locale, matching 16-bit wchar_t platforms (DR).

Restores initialization of char/unsigned char arrays from u8 string literals, for C compatibility (DR).

Replaces references to ISO/IEC 10646 with references to The Unicode Standard throughout.