Commit c20a51afe9 for perl

commit c20a51afe9c45935e792ecf20a5e4dea5737e783
Author: Karl Williamson <khw@cpan.org>
Date:   Wed Oct 21 06:18:47 2026 -0600

    regcharclass.pl: Give each macro its own api section

    Commit 7838a4632996eab80c11128f9cb23a31f765e574 added the ability of
    regcharclass.pl to generate perlintern entries for the macros it
    defines.  It placed all of them in the perlintern "character
    classification" section.  I looked at perlintern, and realized some
    would better go in the "casing" section.

    This commit just adds the section for each individual macro, as opposed
    to a global for all of them.  The net result is that perlintern is
    unchanged by this; that is left for the next commit.

diff --git a/regcharclass.h b/regcharclass.h
index d058c3e890..4e8df79109 100644
--- a/regcharclass.h
+++ b/regcharclass.h
@@ -14,10 +14,6 @@
  * changed or removed without notice.
  */

-/*
-=for apidoc_section $classification
-*/
-
 #ifndef PERL_REGCHARCLASS_H_ /* Guard against nested #includes */
 #  define PERL_REGCHARCLASS_H_

@@ -30,6 +26,7 @@
 	\p{VertSpace}
 */
 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_LNBREAK_safe|const U8 * s|const U8 * e|bool is_utf8
 A few Unicode characters or character sequences act as line terminators.
 C<\R> in a regular expression pattern matches them.
@@ -50,6 +47,7 @@ The value returned gives the number of bytes the matched
 character sequence actually occupies.


+
 =cut
 */

@@ -82,6 +80,7 @@ character sequence actually occupies.
 #endif

 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_LNBREAK_utf8_safe|const U8 * s|const U8 * e
 A few Unicode characters or character sequences act as line terminators.
 C<\R> in a regular expression pattern matches them.
@@ -96,6 +95,7 @@ The value returned gives the number of bytes the matched
 character sequence actually occupies.


+
 =cut
 */

@@ -121,6 +121,7 @@ character sequence actually occupies.
 #endif

 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_LNBREAK_latin1_safe|const U8 * s|const U8 * e
 A few Unicode characters or character sequences act as line terminators.
 C<\R> in a regular expression pattern matches them.
@@ -135,6 +136,7 @@ The value returned gives the number of bytes the matched
 character sequence actually occupies.


+
 =cut
 */

@@ -158,6 +160,7 @@ character sequence actually occupies.
 	\p{HorizSpace}
 */
 /*
+=for apidoc_section $classification
 =for apidoc CT|STRLEN|is_HORIZWS_high|const U8 * s
 A few Unicode characters are treated as horizontal space.

@@ -175,6 +178,7 @@ one complete character before calling this; otherwise this
 could read beyond its end.


+
 =cut
 */

@@ -189,6 +193,7 @@ could read beyond its end.
 : ( ( ( 0xE3 == ((const U8*)s)[0] ) && ( 0x80 == ((const U8*)s)[1] ) ) && ( 0x80 == ((const U8*)s)[2] ) ) ? 3 : 0 ) )

 /*
+=for apidoc_section $classification
 =for apidoc CT|STRLEN|is_HORIZWS_cp_high|UV cp
 A few Unicode characters are treated as horizontal space.

@@ -199,6 +204,7 @@ Otherwise, it returns zero.



+
 =cut
 */

@@ -215,6 +221,7 @@ Otherwise, it returns zero.
 	\p{VertSpace}
 */
 /*
+=for apidoc_section $classification
 =for apidoc CT|STRLEN|is_VERTWS_high|const U8 * s
 A few Unicode characters are treated as vertical space.

@@ -232,6 +239,7 @@ one complete character before calling this; otherwise this
 could read beyond its end.


+
 =cut
 */

@@ -240,6 +248,7 @@ could read beyond its end.
 ( (STRLEN)( ( ( ( 0xE2 == ((const U8*)s)[0] ) && ( 0x80 == ((const U8*)s)[1] ) ) && ( inRANGE_helper_(U8, ((const U8*)s)[2], 0xA8, 0xA9) ) ) ? 3 : 0 ) )

 /*
+=for apidoc_section $classification
 =for apidoc CT|STRLEN|is_VERTWS_cp_high|UV cp
 A few Unicode characters are treated as vertical space.

@@ -250,6 +259,7 @@ Otherwise, it returns zero.



+
 =cut
 */

@@ -263,6 +273,7 @@ Otherwise, it returns zero.
 	\p{XDigit}
 */
 /*
+=for apidoc_section $classification
 =for apidoc CT|STRLEN|is_XDIGIT_high|const U8 * s
 A few Unicode characters are treated as hexadecimal digits.

@@ -280,6 +291,7 @@ one complete character before calling this; otherwise this
 could read beyond its end.


+
 =cut
 */

@@ -292,6 +304,7 @@ could read beyond its end.
 : 0 ) )

 /*
+=for apidoc_section $classification
 =for apidoc CT|STRLEN|is_XDIGIT_cp_high|UV cp
 A few Unicode characters are treated as hexadecimal digits.

@@ -302,6 +315,7 @@ Otherwise, it returns zero.



+
 =cut
 */

@@ -316,6 +330,7 @@ Otherwise, it returns zero.
 	\p{XPerlSpace}
 */
 /*
+=for apidoc_section $classification
 =for apidoc CT|STRLEN|is_XPERLSPACE_high|const U8 * s
 This returns non-zero if the
 sequence of the next few bytes starting at C<s>
@@ -331,6 +346,7 @@ one complete character before calling this; otherwise this
 could read beyond its end.


+
 =cut
 */

@@ -345,6 +361,7 @@ could read beyond its end.
 : ( ( ( 0xE3 == ((const U8*)s)[0] ) && ( 0x80 == ((const U8*)s)[1] ) ) && ( 0x80 == ((const U8*)s)[2] ) ) ? 3 : 0 ) )

 /*
+=for apidoc_section $classification
 =for apidoc CT|STRLEN|is_XPERLSPACE_cp_high|UV cp
 This returns non-zero if code point C<cp> is
 an above-Latin1 range character that matches \p{XPerlSpace}.
@@ -353,6 +370,7 @@ Otherwise, it returns zero.



+
 =cut
 */

@@ -370,6 +388,7 @@ Otherwise, it returns zero.
 	\p{XPerlSpace}
 */
 /*
+=for apidoc_section $classification
 =for apidoc CT|STRLEN|is_SPACE_utf8_safe_backwards|const U8 * s|const U8 * e
 Some Unicode characters are treated as space, both horizontal and vertical.

@@ -383,6 +402,7 @@ The value returned gives the number of bytes the matched
 character actually occupies.


+
 =cut
 */

@@ -416,6 +436,7 @@ character actually occupies.
 	\p{_Perl_Nchar}
 */
 /*
+=for apidoc_section $classification
 =for apidoc CT|STRLEN|is_NONCHAR_utf8_safe|const U8 * s|const U8 * e
 Unicode has 66 noncharacter codepoints.  These are not assigned to characters,
 and are guaranteed never to be, so that an application can treat them as
@@ -432,6 +453,7 @@ The value returned gives the number of bytes the matched
 code point actually occupies.


+
 =cut
 */

@@ -454,6 +476,7 @@ code point actually occupies.
 	0xFFFE - 0xFFFF
 */
 /*
+=for apidoc_section $classification
 =for apidoc eT|STRLEN|is_SHORTER_NON_CHARS_utf8|const U8 * s
 Unicode has 66 noncharacter codepoints.  These are not assigned to characters,
 and are guaranteed never to be, so that an application can treat them as
@@ -483,6 +506,7 @@ one complete code point before calling this; otherwise this
 could read beyond its end.


+
 =cut
 */

@@ -518,6 +542,7 @@ could read beyond its end.
 	0x10FFFE - 0x10FFFF
 */
 /*
+=for apidoc_section $classification
 =for apidoc eT|STRLEN|is_LARGER_NON_CHARS_utf8|const U8 * s
 Unicode has 66 noncharacter codepoints.  These are not assigned to characters,
 and are guaranteed never to be, so that an application can treat them as
@@ -547,6 +572,7 @@ one complete code point before calling this; otherwise this
 could read beyond its end.


+
 =cut
 */

@@ -567,6 +593,7 @@ could read beyond its end.
 	\p{_Perl_Surrogate}
 */
 /*
+=for apidoc_section $classification
 =for apidoc CT|STRLEN|is_SURROGATE_utf8|const U8 * s
 Surrogate code points enable UTF-16 to be extended to be able to represent all
 Unicode code points
@@ -585,6 +612,7 @@ one complete code point before calling this; otherwise this
 could read beyond its end.


+
 =cut
 */

@@ -593,6 +621,7 @@ could read beyond its end.
 ( (STRLEN)( ( ( 0xED == ((const U8*)s)[0] ) && ( inRANGE_helper_(U8, ((const U8*)s)[1], 0xA0, 0xBF) ) ) ? 3 : 0 ) )

 /*
+=for apidoc_section $classification
 =for apidoc CT|STRLEN|is_SURROGATE_utf8_safe|const U8 * s|const U8 * e
 Surrogate code points enable UTF-16 to be extended to be able to represent all
 Unicode code points
@@ -607,6 +636,7 @@ The value returned gives the number of bytes the matched
 code point actually occupies.


+
 =cut
 */

@@ -620,6 +650,7 @@ code point actually occupies.
 	\p{_Perl_Quotemeta}
 */
 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_QUOTEMETA_high|const U8 * s
 Most characters in a Perl program represent themselves, but there are a few
 "meta-characters" which mean something else unless escaped in some way.
@@ -638,6 +669,7 @@ one complete character before calling this; otherwise this
 could read beyond its end.


+
 =cut
 */

@@ -700,6 +732,7 @@ could read beyond its end.
 	%regcharclass_multi_char_folds::multi_char_folds('u', 'a')
 */
 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_MULTI_CHAR_FOLD_utf8_safe|const U8 * s|const U8 * e
 The casefold of some Unicode characters expands to two or three characters.

@@ -714,6 +747,7 @@ The value returned gives the number of bytes the matched
 character sequence actually occupies.


+
 =cut
 */

@@ -958,6 +992,7 @@ character sequence actually occupies.
 #endif

 /*
+=for apidoc_section $classification
 =for apidoc QT|UV|what_MULTI_CHAR_FOLD_utf8_safe|const U8 * s|const U8 * e
 The casefold of some Unicode characters expands to two or three characters.

@@ -974,6 +1009,7 @@ The value returned gives the number of bytes the matched
 character sequence actually occupies.


+
 =cut
 */

@@ -1412,6 +1448,7 @@ character sequence actually occupies.
 	%regcharclass_multi_char_folds::multi_char_folds('l', 'a')
 */
 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_MULTI_CHAR_FOLD_latin1_safe|const U8 * s|const U8 * e
 The casefold of some Unicode characters expands to two or three characters.

@@ -1426,6 +1463,7 @@ The value returned gives the number of bytes the matched
 character sequence actually occupies.


+
 =cut
 */

@@ -1447,6 +1485,7 @@ character sequence actually occupies.
 #endif

 /*
+=for apidoc_section $classification
 =for apidoc QT|UV|what_MULTI_CHAR_FOLD_latin1_safe|const U8 * s|const U8 * e
 The casefold of some Unicode characters expands to two or three characters.

@@ -1463,6 +1502,7 @@ The value returned gives the number of bytes the matched
 character sequence actually occupies.


+
 =cut
 */

@@ -1499,6 +1539,7 @@ character sequence actually occupies.
 	%regcharclass_multi_char_folds::multi_char_folds('u', '3')
 */
 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_THREE_CHAR_FOLD_utf8_safe|const U8 * s|const U8 * e
 The casefold of a few Unicode characters expands to three characters.

@@ -1513,6 +1554,7 @@ The value returned gives the number of bytes the matched
 character sequence actually occupies.


+
 =cut
 */

@@ -1545,6 +1587,7 @@ character sequence actually occupies.
 	%regcharclass_multi_char_folds::multi_char_folds('l', '3')
 */
 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_THREE_CHAR_FOLD_latin1_safe|const U8 * s|const U8 * e
 The casefold of a few Unicode characters expands to three characters.

@@ -1559,6 +1602,7 @@ The value returned gives the number of bytes the matched
 character sequence actually occupies.


+
 =cut
 */

@@ -1575,6 +1619,7 @@ character sequence actually occupies.
 	%regcharclass_multi_char_folds::multi_char_folds('u', 'h')
 */
 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_THREE_CHAR_FOLD_HEAD_utf8_safe|const U8 * s|const U8 * e
 The casefold of a few Unicode characters expands to three characters.

@@ -1590,6 +1635,7 @@ The value returned gives the number of bytes the matched
 character sequence actually occupies.


+
 =cut
 */

@@ -1669,6 +1715,7 @@ character sequence actually occupies.
 	%regcharclass_multi_char_folds::multi_char_folds('l', 'h')
 */
 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_THREE_CHAR_FOLD_HEAD_latin1_safe|const U8 * s|const U8 * e
 The casefold of a few Unicode characters expands to three characters.

@@ -1684,6 +1731,7 @@ The value returned gives the number of bytes the matched
 character sequence actually occupies.


+
 =cut
 */

@@ -1707,6 +1755,7 @@ character sequence actually occupies.
 	\p{_Perl_Folds_To_Multi_Char}
 */
 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_FOLDS_TO_MULTI_utf8|const U8 * s
 The casefold of some Unicode characters expands to two or three characters.

@@ -1725,6 +1774,7 @@ one complete character before calling this; otherwise this
 could read beyond its end.


+
 =cut
 */

@@ -1761,6 +1811,7 @@ could read beyond its end.
 	\p{_Perl_Problematic_Locale_Folds}
 */
 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_PROBLEMATIC_LOCALE_FOLD_utf8|const U8 * s
 The casefold of most Unicode characters is straight forward, but for some, it
 is problematic in some way.
@@ -1780,6 +1831,7 @@ one complete character before calling this; otherwise this
 could read beyond its end.


+
 =cut
 */

@@ -1810,6 +1862,7 @@ could read beyond its end.
 #endif

 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_PROBLEMATIC_LOCALE_FOLD_cp|UV cp
 The casefold of most Unicode characters is straight forward, but for some, it
 is problematic in some way.
@@ -1822,6 +1875,7 @@ Otherwise, it returns zero.



+
 =cut
 */

@@ -1850,6 +1904,7 @@ Otherwise, it returns zero.
 	\p{_Perl_Problematic_Locale_Foldeds_Start}
 */
 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_PROBLEMATIC_LOCALE_FOLDEDS_START_utf8|const U8 * s
 The casefold of most Unicode characters is straight forward, but for some, it
 is problematic in some way.
@@ -1871,6 +1926,7 @@ one complete character before calling this; otherwise this
 could read beyond its end.


+
 =cut
 */

@@ -1901,6 +1957,7 @@ could read beyond its end.
 #endif

 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_PROBLEMATIC_LOCALE_FOLDEDS_START_cp|UV cp
 The casefold of most Unicode characters is straight forward, but for some, it
 is problematic in some way.
@@ -1915,6 +1972,7 @@ Otherwise, it returns zero.



+
 =cut
 */

@@ -1943,6 +2001,7 @@ Otherwise, it returns zero.
 	\p{_Perl_PatWS}
 */
 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_PATWS_safe|const U8 * s|const U8 * e|bool is_utf8
 Not all characters that are nominally space characters in Unicode are
 considered as such in regular expression patterns.  Sometimes we need to
@@ -1964,6 +2023,7 @@ The value returned gives the number of bytes the matched
 character actually occupies.


+
 =cut
 */

@@ -1989,6 +2049,7 @@ character actually occupies.
 	0xD000 - 0xD7FF
 */
 /*
+=for apidoc_section $classification
 =for apidoc CT|STRLEN|is_HANGUL_ED_utf8_safe|const U8 * s|const U8 * e
 For convenience of the implementation, certain Korean Hangul syllable characters
 have to be treated specially.
@@ -2003,6 +2064,7 @@ The value returned gives the number of bytes the matched
 character actually occupies.


+
 =cut
 */

@@ -2016,6 +2078,7 @@ character actually occupies.
 	\p{_Perl_Word_But_NonCont}
 */
 /*
+=for apidoc_section $classification
 =for apidoc QT|STRLEN|is_WORD_BUT_NONCONT_safe|const U8 * s|const U8 * e|bool is_utf8
 Some Unicode characters that match C<\w> can't be used in Perl names (such as
 identifiers and labels).
@@ -2036,6 +2099,7 @@ The value returned gives the number of bytes the matched
 character actually occupies.


+
 =cut
 */

@@ -4878,6 +4942,6 @@ character actually occupies.
  * 94644b4e53ef4f2590ef86a261dcd22ffadf79247fcc7f31c4ebf5b19111fd25 lib/unicore/mktables
  * a0079d7556b20c2de1fdc3d492797a91b8ec8a06006f94460006185cb6380962 lib/unicore/version
  * 0a6b5ab33bb1026531f816efe81aea1a8ffcd34a27cbea37dd6a70a63d73c844 regen/charset_translations.pl
- * 6f2138d6540593d5132aa1f3ff92bb6a69e674acf39bbe1f6d72532e1ebdb2d2 regen/regcharclass.pl
+ * 1a3589a581e9b3c9ad4c13efdcf71d8953df2f63cf06a081a6b209640ed1799e regen/regcharclass.pl
  * 1ce43daf54162d0a267412453ce2a613f60f3a7631365775cc52d34213251a83 regen/regcharclass_multi_char_folds.pl
  * ex: set ro ft=c: */
diff --git a/regen/regcharclass.pl b/regen/regcharclass.pl
index 560a2a5e8e..eabe1fce28 100755
--- a/regen/regcharclass.pl
+++ b/regen/regcharclass.pl
@@ -1700,7 +1700,12 @@ sub make_macro {
     my $def_fmt="$macro%s($argstr)";
     my $optree= $self->$method( %opts, type => $type, ret_type => $ret_type );
     my $api_ret = ($returns_cp) ? 'UV' : 'STRLEN';
-    my $pod = "=for apidoc $opts{flags}T|$api_ret|$macro$doc_arglist\n$doc_text";
+    my $apidoc_section = '$classification';
+    my $pod = <<~EOT;
+        =for apidoc_section $apidoc_section
+        =for apidoc $opts{flags}T|$api_ret|$macro$doc_arglist
+        $doc_text
+        EOT

     return $self->render( $optree, ($type =~ /^cp/) ? 1 : 0, \%opts, $def_fmt ),
            $pod;
@@ -1730,10 +1735,6 @@ EOF
     );

     print $out_fh <<~EOT;
-        /*
-        =for apidoc_section \$classification
-        */
-
         #ifndef PERL_REGCHARCLASS_H_ /* Guard against nested #includes */
         #  define PERL_REGCHARCLASS_H_

@@ -1802,9 +1803,8 @@ EOF
                     flags => $flags,
                 );

-                my $first_line = $api =~ s/\n.*//rs;
-
-                if (! $seen{$first_line}++) {
+                $api =~ m/(=for apidoc .*)/;
+                if (! $seen{$1}++) {
                     print $out_fh <<~EOT;
                         /*
                         $api