Merge 2021-10 CWG Motion 4

P2316R2 Consistent character literal encoding

Author: tkoeppe

source/preprocessor.tex

@@ -478,32 +478,12 @@
 expression even though it is unsigned in translation phase
 7\iref{lex.phases}.
 \end{note}
-This includes interpreting \grammarterm{character-literal}s, which may involve
-converting escape sequences into execution character set members.
-Whether the numeric value for these \grammarterm{character-literal}s
-matches the value obtained when an identical \grammarterm{character-literal}
-occurs in an expression
-(other than within a
-\tcode{\#if}
-or
-\tcode{\#elif}
-directive)
-is \impldef{numeric values of \grammarterm{character-literal}s in \tcode{\#if}
-directives}.
+This includes interpreting \grammarterm{character-literal}s
+according to the rules in \ref{lex.ccon}.
 \begin{note}
-Thus, the constant expression in the following
-\tcode{\#if}
-directive and
-\keyword{if} statement\iref{stmt.if}
-is not guaranteed to evaluate to the same value in these two
-contexts:
-\begin{codeblock}
-#if 'z' - 'a' == 25
-if ('z' - 'a' == 25)
-\end{codeblock}
+The associated character encodings of literals are the same
+in \tcode{\#if} and \tcode{\#elif} directives and in any expression.
 \end{note}
-Also, whether a single-character \grammarterm{character-literal} may have a negative
-value is \impldef{negative value of \grammarterm{character-literal} in preprocessor}.
 Each subexpression with type
 \tcode{bool}
 is subjected to integral promotion before processing continues.