[format.string] Further clarify description of cases where formatting is

described in terms of a call to to_chars.

Author: zygoloid

source/utilities.tex

@@ -19979,24 +19979,23 @@
 \end{floattable}
 
 \pnum
-Formatting of objects of non-string types
-is done as if by calling \tcode{to_chars} (unless otherwise specified)
+The meaning of some non-string presentation types
+is defined in terms of a call to \tcode{to_chars}.
+In such cases,
+let \range{first}{last} be a range
+large enough to hold the \tcode{to_chars} output
+and \tcode{value} be the formatting argument value.
+Formatting is done as if by calling \tcode{to_chars} as specified
 and copying the output through the output iterator of the format context.
 \begin{note}
 Additional padding and adjustments are performed
 prior to copying the output through the output iterator
 as specified by the format specifiers.
 \end{note}
 
-\pnum
-Let \range{first}{last} be a range
-large enough to hold the \tcode{to_chars} output
-and \tcode{value} be the formatting argument value.
-
 \pnum
 The available integer presentation types
-for integral types other than \tcode{bool} and \tcode{charT},
-and their mapping to \tcode{to_chars},
+for integral types other than \tcode{bool} and \tcode{charT}
 are specified in \tref{format.type.int}.
 \begin{example}
 \begin{codeblock}
@@ -20101,9 +20100,17 @@
 \end{floattable}
 
 \pnum
-The available floating-point presentation types and their mapping to
-\tcode{to_chars} (for values other than infinity and NaN) are
+The available floating-point presentation types and their meanings
+for values other than infinity and NaN are
 specified in \tref{format.type.float}.
+For lower-case presentation types, infinity and NaN are formatted as
+\tcode{inf} and \tcode{nan}, respectively.
+For upper-case presentation types, infinity and NaN are formatted as
+\tcode{INF} and \tcode{NAN}, respectively.
+\begin{note}
+In either case, a sign is included
+if indicated by the \fmtgrammarterm{sign} option.
+\end{note}
 %
 \begin{floattable}{Meaning of \fmtgrammarterm{type} options for floating-point types}{format.type.float}{lp{.8\hsize}}
 \topline
@@ -20182,16 +20189,6 @@
 \\
 \end{floattable}
 
-\pnum
-For lower-case presentation types, infinity and NaN are formatted as
-\tcode{inf} and \tcode{nan}, respectively.
-For upper-case presentation types, infinity and NaN are formatted as
-\tcode{INF} and \tcode{NAN}, respectively.
-\begin{note}
-In either case, a sign is included
-if indicated by the \fmtgrammarterm{sign} option.
-\end{note}
-
 \pnum
 The available pointer presentation types and their mapping to
 \tcode{to_chars} are specified in \tref{format.type.ptr}.