| 1 |
.\" -*- mode: troff; coding: utf-8 -*- |
| 2 |
.\" Automatically generated by Pod::Man v6.0.2 (Pod::Simple 3.45) |
| 3 |
.\" |
| 4 |
.\" Standard preamble: |
| 5 |
.\" ======================================================================== |
| 6 |
.de Sp \" Vertical space (when we can't use .PP) |
| 7 |
.if t .sp .5v |
| 8 |
.if n .sp |
| 9 |
.. |
| 10 |
.de Vb \" Begin verbatim text |
| 11 |
.ft CW |
| 12 |
.nf |
| 13 |
.ne \\$1 |
| 14 |
.. |
| 15 |
.de Ve \" End verbatim text |
| 16 |
.ft R |
| 17 |
.fi |
| 18 |
.. |
| 19 |
.\" \*(C` and \*(C' are quotes in nroff, nothing in troff, for use with C<>. |
| 20 |
.ie n \{\ |
| 21 |
. ds C` "" |
| 22 |
. ds C' "" |
| 23 |
'br\} |
| 24 |
.el\{\ |
| 25 |
. ds C` |
| 26 |
. ds C' |
| 27 |
'br\} |
| 28 |
.\" |
| 29 |
.\" Escape single quotes in literal strings from groff's Unicode transform. |
| 30 |
.ie \n(.g .ds Aq \(aq |
| 31 |
.el .ds Aq ' |
| 32 |
.\" |
| 33 |
.\" If the F register is >0, we'll generate index entries on stderr for |
| 34 |
.\" titles (.TH), headers (.SH), subsections (.SS), items (.Ip), and index |
| 35 |
.\" entries marked with X<> in POD. Of course, you'll have to process the |
| 36 |
.\" output yourself in some meaningful fashion. |
| 37 |
.\" |
| 38 |
.\" Avoid warning from groff about undefined register 'F'. |
| 39 |
.de IX |
| 40 |
.. |
| 41 |
.nr rF 0 |
| 42 |
.if \n(.g .if rF .nr rF 1 |
| 43 |
.if (\n(rF:(\n(.g==0)) \{\ |
| 44 |
. if \nF \{\ |
| 45 |
. de IX |
| 46 |
. tm Index:\\$1\t\\n%\t"\\$2" |
| 47 |
.. |
| 48 |
. if !\nF==2 \{\ |
| 49 |
. nr % 0 |
| 50 |
. nr F 2 |
| 51 |
. \} |
| 52 |
. \} |
| 53 |
.\} |
| 54 |
.rr rF |
| 55 |
.\" |
| 56 |
.\" Required to disable full justification in groff 1.23.0. |
| 57 |
.if n .ds AD l |
| 58 |
.\" ======================================================================== |
| 59 |
.\" |
| 60 |
.IX Title "PERL5420DELTA 1" |
| 61 |
.TH PERL5420DELTA 1 2026-03-09 "perl v5.42.1" "Perl Programmers Reference Guide" |
| 62 |
.\" For nroff, turn off justification. Always turn off hyphenation; it makes |
| 63 |
.\" way too many mistakes in technical documents. |
| 64 |
.if n .ad l |
| 65 |
.nh |
| 66 |
.SH NAME |
| 67 |
perl5420delta \- what is new for perl v5.42.0 |
| 68 |
.SH DESCRIPTION |
| 69 |
.IX Header "DESCRIPTION" |
| 70 |
This document describes differences between the 5.42.0 release and the 5.40.0 |
| 71 |
release. |
| 72 |
.SH "Core Enhancements" |
| 73 |
.IX Header "Core Enhancements" |
| 74 |
.SS "More CORE:: subs" |
| 75 |
.IX Subsection "More CORE:: subs" |
| 76 |
\&\f(CW\*(C`chdir\*(C'\fR has been added as a subroutine to the \f(CW\*(C`CORE::\*(C'\fR namespace. |
| 77 |
.PP |
| 78 |
Previously, code like \f(CW&CORE::chdir($dir)\fR or \f(CWmy $ref = \e&CORE::chdir; |
| 79 |
$ref\->($dir)\fR would throw an error saying \f(CW\*(C`&CORE::chdir cannot be called |
| 80 |
directly\*(C'\fR. These cases are now fully supported. |
| 81 |
.ie n .SS "New pragma ""source::encoding""" |
| 82 |
.el .SS "New pragma \f(CWsource::encoding\fP" |
| 83 |
.IX Subsection "New pragma source::encoding" |
| 84 |
This allows you to declare that the portion of a program for the |
| 85 |
remainder of the lexical scope of this pragma is encoded either entirely |
| 86 |
in ASCII (for \f(CW\*(C`use\ source::encoding\ \*(Aqascii\*(Aq\*(C'\fR) or if UTF\-8 is allowed |
| 87 |
as well (for \f(CW\*(C`use\ source::encoding\ \*(Aqutf8\*(Aq\*(C'\fR). No other encodings are |
| 88 |
accepted. The second form is entirely equivalent to \f(CW\*(C`use\ utf8\*(C'\fR, and |
| 89 |
may be used interchangeably with that. |
| 90 |
.PP |
| 91 |
The purpose of this pragma is to catch cases early where you forgot to |
| 92 |
specify \f(CW\*(C`use\ utf8\*(C'\fR. |
| 93 |
.PP |
| 94 |
\&\f(CW\*(C`use\ source::encoding\ \*(Aqascii\*(Aq\*(C'\fR is automatically enabled within the |
| 95 |
lexical scope of a \f(CW\*(C`use\ v5.41.0\*(C'\fR or higher. |
| 96 |
.PP |
| 97 |
\&\f(CW\*(C`no\ source::encoding\*(C'\fR turns off all this checking for the remainder of |
| 98 |
its lexical scope. The meaning of non\-ASCII characters is then undefined. |
| 99 |
.ie n .SS "New "":writer"" attribute on field variables" |
| 100 |
.el .SS "New \f(CW:writer\fP attribute on field variables" |
| 101 |
.IX Subsection "New :writer attribute on field variables" |
| 102 |
Classes defined using \f(CW\*(C`use feature \*(Aqclass\*(Aq\*(C'\fR are now able to automatically |
| 103 |
create writer accessors for scalar fields, by using the \f(CW\*(C`:writer\*(C'\fR attribute, |
| 104 |
similar to the way that \f(CW\*(C`:reader\*(C'\fR already creates reader accessors. |
| 105 |
.PP |
| 106 |
.Vb 4 |
| 107 |
\& class Point { |
| 108 |
\& field $x :reader :writer :param; |
| 109 |
\& field $y :reader :writer :param; |
| 110 |
\& } |
| 111 |
\& |
| 112 |
\& my $p = Point\->new( x => 20, y => 40 ); |
| 113 |
\& $p\->set_x(60); |
| 114 |
.Ve |
| 115 |
.ie n .SS "New ""any"" and ""all"" operators" |
| 116 |
.el .SS "New \f(CWany\fP and \f(CWall\fP operators" |
| 117 |
.IX Subsection "New any and all operators" |
| 118 |
Two new experimental features have been added, which introduce the |
| 119 |
list\-processing operators \f(CW\*(C`any\*(C'\fR and \f(CW\*(C`all\*(C'\fR. |
| 120 |
.PP |
| 121 |
.Vb 3 |
| 122 |
\& use v5.42; |
| 123 |
\& use feature \*(Aqkeyword_all\*(Aq; |
| 124 |
\& no warnings \*(Aqexperimental::keyword_all\*(Aq; |
| 125 |
\& |
| 126 |
\& my @numbers = ... |
| 127 |
\& |
| 128 |
\& if ( all { $_ % 2 == 0 } @numbers ) { |
| 129 |
\& say "All the numbers are even"; |
| 130 |
\& } |
| 131 |
.Ve |
| 132 |
.PP |
| 133 |
These keywords operate similarly to \f(CW\*(C`grep\*(C'\fR except that they only |
| 134 |
ever return true or false, testing if any (or all) of the elements in |
| 135 |
the list make the testing block yield true. Because of this they can |
| 136 |
short\-circuit, avoiding the need to test any further elements if a given |
| 137 |
element determines the eventual result. |
| 138 |
.PP |
| 139 |
These are inspired by the same\-named functions in the List::Util module, |
| 140 |
except that they are implemented as direct core operators, and thus perform |
| 141 |
faster, and do not produce an additional subroutine call stack frame for |
| 142 |
invoking the code block. |
| 143 |
.PP |
| 144 |
The feature flags enabling those keywords have been named |
| 145 |
\&\f(CW\*(C`keyword_any\*(C'\fR |
| 146 |
and \f(CW\*(C`keyword_all\*(C'\fR |
| 147 |
to avoid confusion with the ability of the \f(CW\*(C`feature\*(C'\fR module |
| 148 |
to refer to all of its features by using the \f(CW\*(C`:all\*(C'\fR |
| 149 |
export tag. [GH #23104 <https://github.com/Perl/perl5/issues/23104>] |
| 150 |
.PP |
| 151 |
The related experimental warning flags are consequently named |
| 152 |
\&\f(CW\*(C`experimental::keyword_any\*(C'\fR and \f(CW\*(C`experimental::keyword_all\*(C'\fR. |
| 153 |
.SS "Apostrophe as a global name separator can be disabled" |
| 154 |
.IX Subsection "Apostrophe as a global name separator can be disabled" |
| 155 |
This was deprecated in Perl 5.38 and removed as scheduled in perl |
| 156 |
5.41.3, but after some discussion has been reinstated by default. |
| 157 |
.PP |
| 158 |
This can be controlled with the \f(CW\*(C`apostrophe_as_package_separator\*(C'\fR |
| 159 |
feature which is enabled by default, but is disabled from the 5.41 |
| 160 |
feature bundle onwards. |
| 161 |
.PP |
| 162 |
If you want to disable use within your own code you can explicitly |
| 163 |
disable the feature: |
| 164 |
.PP |
| 165 |
.Vb 1 |
| 166 |
\& no feature "apostrophe_as_package_separator"; |
| 167 |
.Ve |
| 168 |
.PP |
| 169 |
Note that disabling this feature only prevents use of apostrophe as a |
| 170 |
package separator within code; symbolic references still treat \f(CW\*(C`\*(Aq\*(C'\fR as |
| 171 |
\&\f(CW\*(C`::\*(C'\fR with the feature disabled: |
| 172 |
.PP |
| 173 |
.Vb 7 |
| 174 |
\& my $symref = "My\*(AqModule\*(AqVar"; |
| 175 |
\& # default features |
| 176 |
\& my $x = $My\*(AqModule\*(AqVar; # fine |
| 177 |
\& no feature "apostrophe_as_package_separator"; |
| 178 |
\& no strict "refs"; |
| 179 |
\& my $y = $$symref; # like $My::Module::Var |
| 180 |
\& my $z = $My\*(AqModule\*(AqVar; # syntax error |
| 181 |
.Ve |
| 182 |
.PP |
| 183 |
[GH #22644 <https://github.com/Perl/perl5/issues/22644>] |
| 184 |
.ie n .SS "Lexical method declaration using ""my method""" |
| 185 |
.el .SS "Lexical method declaration using \f(CWmy method\fP" |
| 186 |
.IX Subsection "Lexical method declaration using my method" |
| 187 |
Like \f(CW\*(C`sub\*(C'\fR since Perl version 5.18, \f(CW\*(C`method\*(C'\fR can now be prefixed with the |
| 188 |
\&\f(CW\*(C`my\*(C'\fR keyword. This declares a subroutine that has lexical, rather than |
| 189 |
package visibility. See perlclass for more detail. |
| 190 |
.ie n .SS "Lexical method invocation operator ""\->&""" |
| 191 |
.el .SS "Lexical method invocation operator \f(CW\->&\fP" |
| 192 |
.IX Subsection "Lexical method invocation operator ->&" |
| 193 |
Along with the ability to declare methods lexically, this release also permits |
| 194 |
invoking a lexical subroutine as if it were a method, bypassing the usual |
| 195 |
name\-based method resolution. |
| 196 |
.PP |
| 197 |
Combined with lexical method declaration, these two new abilities create the |
| 198 |
effect of having private methods. |
| 199 |
.SS "Switch and Smart Match operator kept, behind a feature" |
| 200 |
.IX Subsection "Switch and Smart Match operator kept, behind a feature" |
| 201 |
The "switch" feature and the smartmatch operator, \f(CW\*(C`~~\*(C'\fR, were introduced in |
| 202 |
v5.10. Their behavior was significantly changed in v5.10.1. When the |
| 203 |
"experiment" system was added in v5.18.0, switch and smartmatch were |
| 204 |
retroactively declared experimental. Over the years, proposals to fix or |
| 205 |
supplement the features have come and gone. |
| 206 |
.PP |
| 207 |
They were deprecated in Perl v5.38.0 and scheduled for removal in |
| 208 |
Perl v5.42.0. After extensive discussion their removal has been indefinitely |
| 209 |
postponed. Using them no longer produces a deprecation warning. |
| 210 |
.PP |
| 211 |
Switch itself still requires the \f(CW\*(C`switch\*(C'\fR feature, which is enabled |
| 212 |
by default for feature bundles from v5.9.5 through to v5.34. Switch |
| 213 |
remains disabled in feature bundles 5.35 and later, but can be |
| 214 |
separately enabled: |
| 215 |
.PP |
| 216 |
.Vb 7 |
| 217 |
\& # no switch here |
| 218 |
\& use v5.10; |
| 219 |
\& # switch here |
| 220 |
\& use v5.36; |
| 221 |
\& # no switch here |
| 222 |
\& use feature "switch"; |
| 223 |
\& # switch here |
| 224 |
.Ve |
| 225 |
.PP |
| 226 |
Smart match now requires the \f(CW\*(C`smartmatch\*(C'\fR feature, which is enabled |
| 227 |
by default and included in all feature bundles up to 5.40. It is |
| 228 |
disabled for the 5.41 feature bundle and later, but can be separately |
| 229 |
enabled: |
| 230 |
.PP |
| 231 |
.Vb 5 |
| 232 |
\& # smartmatch here |
| 233 |
\& use v5.41; |
| 234 |
\& # no smartmatch here |
| 235 |
\& use feature "smartmatch"; |
| 236 |
\& # smartmatch here |
| 237 |
.Ve |
| 238 |
.PP |
| 239 |
[GH #22752 <https://github.com/Perl/perl5/issues/22752>] |
| 240 |
.SS "Unicode 16.0 supported" |
| 241 |
.IX Subsection "Unicode 16.0 supported" |
| 242 |
Perl now supports Unicode 16.0 |
| 243 |
<https://www.unicode.org/versions/Unicode16.0.0/> including the changes |
| 244 |
introduced in 15.1 <https://www.unicode.org/versions/Unicode15.1.0/>. |
| 245 |
.ie n .SS "Assigning logical xor ""^^="" operator" |
| 246 |
.el .SS "Assigning logical xor \f(CW^^=\fP operator" |
| 247 |
.IX Subsection "Assigning logical xor ^^= operator" |
| 248 |
Perl 5.40.0 introduced the logical medium\-precedence exclusive\-or operator |
| 249 |
\&\f(CW\*(C`^^\*(C'\fR. It was not noticed at the time that the assigning variant \f(CW\*(C`^^=\*(C'\fR was |
| 250 |
also missing. This is now added. |
| 251 |
.SH Security |
| 252 |
.IX Header "Security" |
| 253 |
.SS "[CVE\-2024\-56406] Heap buffer overflow vulnerability with tr//" |
| 254 |
.IX Subsection "[CVE-2024-56406] Heap buffer overflow vulnerability with tr//" |
| 255 |
A heap buffer overflow vulnerability was discovered in Perl. |
| 256 |
.PP |
| 257 |
When there are non\-ASCII bytes in the left\-hand\-side of the \f(CW\*(C`tr\*(C'\fR operator, |
| 258 |
\&\f(CWS_do_trans_invmap()\fR can overflow the destination pointer \f(CW\*(C`d\*(C'\fR. |
| 259 |
.PP |
| 260 |
.Vb 2 |
| 261 |
\& $ perl \-e \*(Aq$_ = "\ex{FF}" x 1000000; tr/\exFF/\ex{100}/;\*(Aq |
| 262 |
\& Segmentation fault (core dumped) |
| 263 |
.Ve |
| 264 |
.PP |
| 265 |
It is believed that this vulnerability can enable Denial of Service or |
| 266 |
Arbitrary Code Execution attacks on platforms that lack sufficient defenses. |
| 267 |
.PP |
| 268 |
This problem was discovered by Nathan Mills and assigned |
| 269 |
[CVE\-2024\-56406 <https://lists.security.metacpan.org/cve-announce/msg/28708725/>] |
| 270 |
by the CPAN Security Group <https://security.metacpan.org/>. |
| 271 |
.PP |
| 272 |
The patch to fix this issue (87f42aa0e0096e9a346c9672aa3a0bd3bef8c1dd <https://github.com/Perl/perl5/commit/87f42aa0e0096e9a346c9672aa3a0bd3bef8c1dd>) is applicable to all perls that are vulnerable, including those |
| 273 |
out\-of\-support. |
| 274 |
.SS "[CVE\-2025\-40909] Perl threads have a working directory race condition where file operations may target unintended paths" |
| 275 |
.IX Subsection "[CVE-2025-40909] Perl threads have a working directory race condition where file operations may target unintended paths" |
| 276 |
Perl thread cloning had a working directory race condition where file |
| 277 |
operations may target unintended paths. Perl 5.42 will no longer chdir |
| 278 |
to each handle. |
| 279 |
.PP |
| 280 |
This problem was discovered by Vincent Lefèvre via [GH #23010 <https://github.com/Perl/perl5/issues/23010>] |
| 281 |
and assigned [CVE\-2025\-40909 <https://lists.security.metacpan.org/cve-announce/msg/30017499/>] |
| 282 |
by the CPAN Security Group <https://security.metacpan.org/>. |
| 283 |
.PP |
| 284 |
Fixes were provided via [GH #23019 <https://github.com/Perl/perl5/pull/23019>] |
| 285 |
and [GH #23361 <https://github.com/Perl/perl5/pull/23361>]. |
| 286 |
.SH "Incompatible Changes" |
| 287 |
.IX Header "Incompatible Changes" |
| 288 |
.SS "Removed containing function references for functions without eval" |
| 289 |
.IX Subsection "Removed containing function references for functions without eval" |
| 290 |
Perl 5.40 reintroduced unconditional references from functions to their |
| 291 |
containing functions to fix a bug introduced in Perl 5.18 that broke the |
| 292 |
special behaviour of \f(CW\*(C`eval EXPR\*(C'\fR in package \f(CW\*(C`DB\*(C'\fR which is used by the |
| 293 |
debugger. |
| 294 |
.PP |
| 295 |
In some cases this change led to circular reference chains between closures and |
| 296 |
other existing references, resulting in memory leaks. |
| 297 |
.PP |
| 298 |
This change has been reverted, fixing |
| 299 |
[GH #22547 <https://github.com/Perl/perl5/issues/22547>] but re\-breaking |
| 300 |
[GH #19370 <https://github.com/Perl/perl5/issues/19370>]. |
| 301 |
.PP |
| 302 |
This means the reference loops won\*(Aqt occur, and that lexical variables and |
| 303 |
functions from enclosing functions may not be visible in the debugger. |
| 304 |
.PP |
| 305 |
Note that calling \f(CW\*(C`eval EXPR\*(C'\fR in a function unconditionally causes a function |
| 306 |
to reference its enclosing functions as it always has. |
| 307 |
.SH "Performance Enhancements" |
| 308 |
.IX Header "Performance Enhancements" |
| 309 |
.IP \(bu 4 |
| 310 |
Constant\-folded strings are now shareable via the Copy\-on\-Write mechanism. |
| 311 |
[GH #22163 <https://github.com/Perl/perl5/pull/22163>] |
| 312 |
.Sp |
| 313 |
The following code would previously have allocated eleven string buffers, |
| 314 |
each containing one million "A"s: |
| 315 |
.Sp |
| 316 |
.Vb 1 |
| 317 |
\& my @scalars; push @scalars, ("A" x 1_000_000) for 0..9; |
| 318 |
.Ve |
| 319 |
.Sp |
| 320 |
Now a single buffer is allocated and shared between a CONST OP and |
| 321 |
the ten scalar elements of \f(CW@scalars\fR. |
| 322 |
.Sp |
| 323 |
Note that any code using this sort of constant to simulate memory leaks |
| 324 |
(perhaps in test files) must now permute the string in order to trigger |
| 325 |
a string copy and the allocation of separate buffers. For example, |
| 326 |
\&\f(CW\*(C`("A" x 1_000_000).time\*(C'\fR might be a suitable small change. |
| 327 |
.IP \(bu 4 |
| 328 |
\&\f(CW\*(C`tr///\*(C'\fR now runs at the same speed regardless of the internal |
| 329 |
representation of its operand, as long as the only characters being |
| 330 |
translated are ASCII\-range, for example \f(CW\*(C`tr/A\-Z/a\-z/\*(C'\fR. Previously, if |
| 331 |
the internal encoding was UTF\-8, a slower, more general implementation |
| 332 |
was used. |
| 333 |
.IP \(bu 4 |
| 334 |
Code that uses the \f(CW\*(C`indexed\*(C'\fR function from the builtin module to generate |
| 335 |
a list of index/value pairs out of an array or list which is then passed into |
| 336 |
a two\-variable \f(CW\*(C`foreach\*(C'\fR list to unpack those again is now optimised to be |
| 337 |
more efficient. |
| 338 |
.Sp |
| 339 |
.Vb 1 |
| 340 |
\& my @array = (...); |
| 341 |
\& |
| 342 |
\& foreach my ($idx, $val) (builtin::indexed @array) { |
| 343 |
\& ... |
| 344 |
\& } |
| 345 |
.Ve |
| 346 |
.Sp |
| 347 |
|
| 348 |
.Sp |
| 349 |
.Vb 3 |
| 350 |
\& foreach my ($idx, $val) (builtin::indexed LIST...) { |
| 351 |
\& ... |
| 352 |
\& } |
| 353 |
.Ve |
| 354 |
.Sp |
| 355 |
In particular, a temporary list twice the size of the original is no longer |
| 356 |
generated. Instead, the loop iterates down the original array or list |
| 357 |
in\-place directly, in the same way that \f(CW\*(C`foreach (@array)\*(C'\fR or |
| 358 |
\&\f(CW\*(C`foreach (LIST)\*(C'\fR would do. |
| 359 |
.IP \(bu 4 |
| 360 |
The peephole optimizer recognises the following zero\-offset \f(CW\*(C`substr\*(C'\fR patterns |
| 361 |
and swaps in a new dedicated operator (\f(CW\*(C`OP_SUBSTR_LEFT\*(C'\fR). |
| 362 |
[GH #22785 <https://github.com/Perl/perl5/issues/22785>] |
| 363 |
.Sp |
| 364 |
.Vb 2 |
| 365 |
\& substr($x, 0, ...) |
| 366 |
\& substr($x, 0, ..., \*(Aq\*(Aq) |
| 367 |
.Ve |
| 368 |
.IP \(bu 4 |
| 369 |
The stringification of integers by "print" in perlfunc and "say" in perlfunc, |
| 370 |
when coming from an \f(CW\*(C`SVt_IV\*(C'\fR, is now more efficient. |
| 371 |
[GH #22927 <https://github.com/Perl/perl5/issues/22927>] |
| 372 |
.IP \(bu 4 |
| 373 |
String reversal from a single argument, when the string buffer is not |
| 374 |
"swiped", is now done in a single pass and is noticeably faster. |
| 375 |
The extent of the improvement is compiler & hardware dependent. |
| 376 |
[GH #23012 <https://github.com/Perl/perl5/issues/23012>] |
| 377 |
.SH "Modules and Pragmata" |
| 378 |
.IX Header "Modules and Pragmata" |
| 379 |
.SS "Updated Modules and Pragmata" |
| 380 |
.IX Subsection "Updated Modules and Pragmata" |
| 381 |
.IP \(bu 4 |
| 382 |
Archive::Tar has been upgraded from version 3.02_001 to 3.04. |
| 383 |
.IP \(bu 4 |
| 384 |
B::Deparse has been upgraded from version 1.76 to 1.85. |
| 385 |
.IP \(bu 4 |
| 386 |
Benchmark has been upgraded from version 1.25 to 1.27. |
| 387 |
.IP \(bu 4 |
| 388 |
builtin has been upgraded from version 0.014 to 0.019. |
| 389 |
.IP \(bu 4 |
| 390 |
Compress::Raw::Bzip2 has been upgraded from version 2.212 to 2.213. |
| 391 |
.IP \(bu 4 |
| 392 |
Compress::Raw::Zlib has been upgraded from version 2.212 to 2.213. |
| 393 |
.IP \(bu 4 |
| 394 |
Config::Perl::V has been upgraded from version 0.36 to 0.38. |
| 395 |
.IP \(bu 4 |
| 396 |
CPAN has been upgraded from version 2.36 to 2.38. |
| 397 |
.IP \(bu 4 |
| 398 |
CPAN::Meta::YAML has been upgraded from version 0.018 to 0.020. |
| 399 |
.IP \(bu 4 |
| 400 |
Data::Dumper has been upgraded from version 2.189 to 2.192. |
| 401 |
.IP \(bu 4 |
| 402 |
DB has been upgraded from version 1.08 to 1.09. |
| 403 |
.IP \(bu 4 |
| 404 |
DBM_Filter has been upgraded from version 0.06 to 0.07. |
| 405 |
.IP \(bu 4 |
| 406 |
Devel::Peek has been upgraded from version 1.34 to 1.36. |
| 407 |
.IP \(bu 4 |
| 408 |
Devel::PPPort has been upgraded from version 3.72 to 3.73. |
| 409 |
.IP \(bu 4 |
| 410 |
Digest::MD5 has been upgraded from version 2.58_01 to 2.59. |
| 411 |
.IP \(bu 4 |
| 412 |
DynaLoader has been upgraded from version 1.56 to 1.57. |
| 413 |
.IP \(bu 4 |
| 414 |
experimental has been upgraded from version 0.032 to 0.035. |
| 415 |
.IP \(bu 4 |
| 416 |
Exporter has been upgraded from version 5.78 to 5.79. |
| 417 |
.IP \(bu 4 |
| 418 |
ExtUtils::CBuilder has been upgraded from version 0.280240 to 0.280242. |
| 419 |
.IP \(bu 4 |
| 420 |
ExtUtils::MakeMaker has been upgraded from version 7.70 to 7.76. |
| 421 |
.IP \(bu 4 |
| 422 |
ExtUtils::ParseXS has been upgraded from version 3.51 to 3.57. |
| 423 |
.IP \(bu 4 |
| 424 |
ExtUtils::Typemaps has been upgraded from version 3.51 to 3.57. |
| 425 |
.IP \(bu 4 |
| 426 |
Fcntl has been upgraded from version 1.18 to 1.20. |
| 427 |
.IP \(bu 4 |
| 428 |
feature has been upgraded from version 1.89 to 1.97. |
| 429 |
.IP \(bu 4 |
| 430 |
fields has been upgraded from version 2.25 to 2.27. |
| 431 |
.IP \(bu 4 |
| 432 |
File::Spec has been upgraded from version 3.90 to 3.94. |
| 433 |
.IP \(bu 4 |
| 434 |
Getopt::Long has been upgraded from version 2.57 to 2.58. |
| 435 |
.IP \(bu 4 |
| 436 |
HTTP::Tiny has been upgraded from version 0.088 to 0.090. |
| 437 |
.IP \(bu 4 |
| 438 |
IO::Compress has been upgraded from version 2.212 to 2.213. |
| 439 |
.IP \(bu 4 |
| 440 |
IO::Socket::IP has been upgraded from version 0.42 to 0.43. |
| 441 |
.IP \(bu 4 |
| 442 |
IPC::Open3 has been upgraded from version 1.22 to 1.24. |
| 443 |
.IP \(bu 4 |
| 444 |
locale has been upgraded from version 1.12 to 1.13. |
| 445 |
.IP \(bu 4 |
| 446 |
Math::BigInt has been upgraded from version 2.003002 to 2.005002. |
| 447 |
.IP \(bu 4 |
| 448 |
Math::BigInt::FastCalc has been upgraded from version 0.5018 to 0.5020. |
| 449 |
.IP \(bu 4 |
| 450 |
Math::Complex has been upgraded from version 1.62 to 1.63. |
| 451 |
.IP \(bu 4 |
| 452 |
Memoize has been upgraded from version 1.16 to 1.17. |
| 453 |
.IP \(bu 4 |
| 454 |
Module::CoreList has been upgraded from version 5.20240609 to 5.20250702. |
| 455 |
.IP \(bu 4 |
| 456 |
NDBM_File has been upgraded from version 1.17 to 1.18. |
| 457 |
.IP \(bu 4 |
| 458 |
ODBM_File has been upgraded from version 1.18 to 1.20. |
| 459 |
.IP \(bu 4 |
| 460 |
Opcode has been upgraded from version 1.65 to 1.69. |
| 461 |
.IP \(bu 4 |
| 462 |
overload has been upgraded from version 1.37 to 1.40. |
| 463 |
.IP \(bu 4 |
| 464 |
parent has been upgraded from version 0.241 to 0.244. |
| 465 |
.IP \(bu 4 |
| 466 |
perlfaq has been upgraded from version 5.20240218 to 5.20250619. |
| 467 |
.IP \(bu 4 |
| 468 |
Pod::Usage has been upgraded from version 2.03 to 2.05. |
| 469 |
.IP \(bu 4 |
| 470 |
podlators has been upgraded from version 5.01_02 to v6.0.2. |
| 471 |
.IP \(bu 4 |
| 472 |
POSIX has been upgraded from version 2.20 to 2.23. |
| 473 |
.IP \(bu 4 |
| 474 |
re has been upgraded from version 0.47 to 0.48. |
| 475 |
.IP \(bu 4 |
| 476 |
Safe has been upgraded from version 2.46 to 2.47. |
| 477 |
.IP \(bu 4 |
| 478 |
Scalar::Util has been upgraded from version 1.63 to 1.68_01. |
| 479 |
.IP \(bu 4 |
| 480 |
Search::Dict has been upgraded from version 1.07 to 1.08. |
| 481 |
.IP \(bu 4 |
| 482 |
SelfLoader has been upgraded from version 1.27 to 1.28. |
| 483 |
.IP \(bu 4 |
| 484 |
sort has been upgraded from version 2.05 to 2.06. |
| 485 |
.IP \(bu 4 |
| 486 |
Storable has been upgraded from version 3.32 to 3.37. |
| 487 |
.IP \(bu 4 |
| 488 |
strict has been upgraded from version 1.13 to 1.14. |
| 489 |
.IP \(bu 4 |
| 490 |
Term::Table has been upgraded from version 0.018 to 0.024. |
| 491 |
.IP \(bu 4 |
| 492 |
Test::Harness has been upgraded from version 3.48 to 3.50. |
| 493 |
.IP \(bu 4 |
| 494 |
Test::Simple has been upgraded from version 1.302199 to 1.302210. |
| 495 |
.IP \(bu 4 |
| 496 |
Thread has been upgraded from version 3.05 to 3.06. |
| 497 |
.IP \(bu 4 |
| 498 |
threads has been upgraded from version 2.40 to 2.43. |
| 499 |
.IP \(bu 4 |
| 500 |
threads::shared has been upgraded from version 1.69 to 1.70. |
| 501 |
.IP \(bu 4 |
| 502 |
Tie::File has been upgraded from version 1.09 to 1.10. |
| 503 |
.IP \(bu 4 |
| 504 |
Tie::RefHash has been upgraded from version 1.40 to 1.41. |
| 505 |
.IP \(bu 4 |
| 506 |
Time::HiRes has been upgraded from version 1.9777 to 1.9778. |
| 507 |
.IP \(bu 4 |
| 508 |
Time::Piece has been upgraded from version 1.3401_01 to 1.36. |
| 509 |
.IP \(bu 4 |
| 510 |
Unicode::UCD has been upgraded from version 0.78 to 0.81. |
| 511 |
.IP \(bu 4 |
| 512 |
utf8 has been upgraded from version 1.25 to 1.27. |
| 513 |
.IP \(bu 4 |
| 514 |
version has been upgraded from version 0.9930 to 0.9933. |
| 515 |
.IP \(bu 4 |
| 516 |
VMS::Filespec has been upgraded from version 1.13 to 1.15. |
| 517 |
.IP \(bu 4 |
| 518 |
warnings has been upgraded from version 1.69 to 1.74. |
| 519 |
.IP \(bu 4 |
| 520 |
Win32 has been upgraded from version 0.59 to 0.59_01. |
| 521 |
.IP \(bu 4 |
| 522 |
XS::APItest has been upgraded from version 1.36 to 1.43. |
| 523 |
.SH Documentation |
| 524 |
.IX Header "Documentation" |
| 525 |
.SS "Changes to Existing Documentation" |
| 526 |
.IX Subsection "Changes to Existing Documentation" |
| 527 |
We have attempted to update the documentation to reflect the changes |
| 528 |
listed in this document. If you find any we have missed, open an issue |
| 529 |
at <https://github.com/Perl/perl5/issues>. |
| 530 |
.PP |
| 531 |
Additionally, the following selected changes have been made: |
| 532 |
.PP |
| 533 |
\fIperlapi\fR |
| 534 |
.IX Subsection "perlapi" |
| 535 |
.IP \(bu 4 |
| 536 |
Combined the documentation for several groups of related functions |
| 537 |
into single entries. |
| 538 |
.IP \(bu 4 |
| 539 |
All forms of \f(CWgv_fetchmeth()\fR are now documented together. |
| 540 |
.IP \(bu 4 |
| 541 |
\&\f(CW\*(C`gv_autoload4\*(C'\fR is now documented with \f(CW\*(C`gv_autoload_pv\*(C'\fR and additional notes added. |
| 542 |
The long \f(CW\*(C`Perl_\*(C'\fR forms are now listed when available. |
| 543 |
.PP |
| 544 |
\fIperldata\fR |
| 545 |
.IX Subsection "perldata" |
| 546 |
.IP \(bu 4 |
| 547 |
Binary and octal floating\-point constants (such as \f(CW\*(C`012.345p\-2\*(C'\fR and |
| 548 |
\&\f(CW\*(C`0b101.11p\-1\*(C'\fR) are now documented. This feature was first introduced in perl |
| 549 |
5.22.0 together with hexadecimal floating\-point constants and had a few bug |
| 550 |
fixes in perl 5.28.0, but it was never formally documented. |
| 551 |
[GH #18664 <https://github.com/Perl/perl5/issues/18664>] |
| 552 |
.PP |
| 553 |
\fIperlfunc\fR |
| 554 |
.IX Subsection "perlfunc" |
| 555 |
.IP \(bu 4 |
| 556 |
Clarified the description of \f(CW\*(C`ref\*(C'\fR and \f(CW\*(C`reftype\*(C'\fR in relation |
| 557 |
to built\-in types and class names. |
| 558 |
.IP \(bu 4 |
| 559 |
Clarified that perl \f(CW\*(C`sort\*(C'\fR is stable (and has been since v5.8.0). |
| 560 |
.IP \(bu 4 |
| 561 |
The recommended alternatives to the \f(CWrand()\fR function were updated to modern modules |
| 562 |
recommended by the CPAN Security Group <https://security.metacpan.org/>. |
| 563 |
[GH #22873 <https://github.com/Perl/perl5/pull/22873>] |
| 564 |
.PP |
| 565 |
\fIperlgov\fR |
| 566 |
.IX Subsection "perlgov" |
| 567 |
.IP \(bu 4 |
| 568 |
The list of Steering Council and Core Team members have been updated, following the conclusion of |
| 569 |
the latest election on 2024\-07\-17. |
| 570 |
.PP |
| 571 |
\fIperlguts\fR |
| 572 |
.IX Subsection "perlguts" |
| 573 |
.IP \(bu 4 |
| 574 |
Added some description of "real" \f(CW\*(C`AV\*(C'\fRs compared to "fake" \f(CW\*(C`AV\*(C'\fRs. |
| 575 |
.IP \(bu 4 |
| 576 |
Documentation was updated to reflect that mixing \f(CW\*(C`Newx\*(C'\fR, \f(CW\*(C`Renew\*(C'\fR, and |
| 577 |
\&\f(CW\*(C`Safefree\*(C'\fR vs \f(CW\*(C`malloc\*(C'\fR, \f(CW\*(C`realloc\*(C'\fR, and \f(CW\*(C`free\*(C'\fR are not allowed, and mixing |
| 578 |
pointers between the 2 classes of APIs is not allowed. Updates made in |
| 579 |
perlguts and perlclib. |
| 580 |
.IP \(bu 4 |
| 581 |
Additional caveats have been added to the description of \f(CW\*(C`TARG\*(C'\fR. |
| 582 |
.PP |
| 583 |
\fIperlop\fR |
| 584 |
.IX Subsection "perlop" |
| 585 |
.IP \(bu 4 |
| 586 |
Portions of perlop are supposed to be ordered so that all the operators |
| 587 |
wth the same precedence are in a single section, and the sections are |
| 588 |
ordered so that the highest precedence operators appear first. |
| 589 |
This ordering has now been restored. Other reorganization was done to |
| 590 |
improve clarity, with more basic operations described before ones that |
| 591 |
depend on them. |
| 592 |
.IP \(bu 4 |
| 593 |
The documentation for here\-docs has been cleaned up and reorganized. |
| 594 |
Indented here\-docs were formerly documented separately, now the two |
| 595 |
types have interwoven documentation which is more compact, and easier to |
| 596 |
understand. |
| 597 |
.IP \(bu 4 |
| 598 |
The documentation of the \f(CW\*(C`xor\*(C'\fR operator has been expanded. |
| 599 |
.IP \(bu 4 |
| 600 |
Outdated advice about using relational string operators in UTF\-8 locales |
| 601 |
has been removed. Use Unicode::Collate for the best results, but |
| 602 |
these operators will give adequate results on many platforms. |
| 603 |
.IP \(bu 4 |
| 604 |
Normalized alignment of verbatim sections, fixing how they are displayed by |
| 605 |
some Pod viewers that strip indentation. |
| 606 |
.PP |
| 607 |
\fIperlvar\fR |
| 608 |
.IX Subsection "perlvar" |
| 609 |
.IP \(bu 4 |
| 610 |
Entries for \f(CW$#\fR and \f(CW$*\fR have been amended to note that use of them result in a compilation |
| 611 |
error, not a warning. |
| 612 |
.SH Diagnostics |
| 613 |
.IX Header "Diagnostics" |
| 614 |
The following additions or changes have been made to diagnostic output, |
| 615 |
including warnings and fatal error messages. For the complete list of |
| 616 |
diagnostic messages, see perldiag. |
| 617 |
.SS "New Diagnostics" |
| 618 |
.IX Subsection "New Diagnostics" |
| 619 |
\fINew Errors\fR |
| 620 |
.IX Subsection "New Errors" |
| 621 |
.IP \(bu 4 |
| 622 |
Use of non\-ASCII character 0x%X illegal when \*(Aquse source::encoding "ascii"\*(Aq is in effect |
| 623 |
.Sp |
| 624 |
(F) This pragma forbids non\-ASCII characters within its scope. |
| 625 |
.IP \(bu 4 |
| 626 |
Undefined subroutine &%s called, close to label \*(Aq%s\*(Aq |
| 627 |
.Sp |
| 628 |
(F) The subroutine indicated hasn\*(Aqt been defined, or if it was, it has |
| 629 |
since been undefined. |
| 630 |
.Sp |
| 631 |
This error could also indicate a mistyped package separator, when a |
| 632 |
single colon was typed instead of two colons. For example, \f(CWFoo:bar()\fR |
| 633 |
would be parsed as the label \f(CW\*(C`Foo\*(C'\fR followed by an unqualified function |
| 634 |
name: \f(CW\*(C`foo: bar()\*(C'\fR. [GH #22860 <https://github.com/Perl/perl5/issues/22860>] |
| 635 |
.PP |
| 636 |
\fINew Warnings\fR |
| 637 |
.IX Subsection "New Warnings" |
| 638 |
.IP \(bu 4 |
| 639 |
_\|_CLASS_\|_ is experimental |
| 640 |
.Sp |
| 641 |
(S experimental::class) This warning is emitted if you use the \f(CW\*(C`_\|_CLASS_\|_\*(C'\fR |
| 642 |
keyword of \f(CW\*(C`use feature \*(Aqclass\*(Aq\*(C'\fR. This keyword is currently |
| 643 |
experimental and its behaviour may change in future releases of Perl. |
| 644 |
.IP \(bu 4 |
| 645 |
%s() attempted on handle \f(CW%s\fR opened with \fBopen()\fR |
| 646 |
.Sp |
| 647 |
(W io) You called \fBreaddir()\fR, \fBtelldir()\fR, \fBseekdir()\fR, \fBrewinddir()\fR or |
| 648 |
\&\fBclosedir()\fR on a handle that was opened with \fBopen()\fR. If you want to |
| 649 |
use these functions to traverse the contents of a directory, you need |
| 650 |
to open the handle with \fBopendir()\fR. |
| 651 |
.Sp |
| 652 |
[GH #22394 <https://github.com/Perl/perl5/issues/22394>] |
| 653 |
.IP \(bu 4 |
| 654 |
Possible precedence problem between ! and \f(CW%s\fR |
| 655 |
.Sp |
| 656 |
(W precedence) You wrote something like |
| 657 |
.Sp |
| 658 |
.Vb 4 |
| 659 |
\& !$x < $y # parsed as: (!$x) < $y |
| 660 |
\& !$x eq $y # parsed as: (!$x) eq $y |
| 661 |
\& !$x =~ /regex/ # parsed as: (!$x) =~ /regex/ |
| 662 |
\& !$obj isa Some::Class # parsed as: (!$obj) isa Some::Class |
| 663 |
.Ve |
| 664 |
.Sp |
| 665 |
but because \f(CW\*(C`!\*(C'\fR has higher precedence than comparison operators, \f(CW\*(C`=~\*(C'\fR, and |
| 666 |
\&\f(CW\*(C`isa\*(C'\fR, this is interpreted as comparing/matching the logical negation of the |
| 667 |
first operand, instead of negating the result of the comparison/match. |
| 668 |
.Sp |
| 669 |
To disambiguate, either use a negated comparison/binding operator: |
| 670 |
.Sp |
| 671 |
.Vb 3 |
| 672 |
\& $x >= $y |
| 673 |
\& $x ne $y |
| 674 |
\& $x !~ /regex/ |
| 675 |
.Ve |
| 676 |
.Sp |
| 677 |
\&... or parentheses: |
| 678 |
.Sp |
| 679 |
.Vb 4 |
| 680 |
\& !($x < $y) |
| 681 |
\& !($x eq $y) |
| 682 |
\& !($x =~ /regex/) |
| 683 |
\& !($obj isa Some::Class) |
| 684 |
.Ve |
| 685 |
.Sp |
| 686 |
\&... or the low\-precedence \f(CW\*(C`not\*(C'\fR operator: |
| 687 |
.Sp |
| 688 |
.Vb 4 |
| 689 |
\& not $x < $y |
| 690 |
\& not $x eq $y |
| 691 |
\& not $x =~ /regex/ |
| 692 |
\& not $obj isa Some::Class |
| 693 |
.Ve |
| 694 |
.Sp |
| 695 |
(If you did mean to compare the boolean result of negating the first operand, |
| 696 |
parenthesize as \f(CW\*(C`(!$x) < $y\*(C'\fR, \f(CW\*(C`(!$x) eq $y\*(C'\fR, etc.) |
| 697 |
.Sp |
| 698 |
Note: this warning does not trigger for code like \f(CW\*(C`!!$x == $y\*(C'\fR, |
| 699 |
i.e. where double negation (\f(CW\*(C`!!\*(C'\fR) is used as a convert\-to\-boolean operator. |
| 700 |
.SS "Changes to Existing Diagnostics" |
| 701 |
.IX Subsection "Changes to Existing Diagnostics" |
| 702 |
.IP \(bu 4 |
| 703 |
%s() attempted on invalid dirhandle \f(CW%s\fR |
| 704 |
.Sp |
| 705 |
This was consolidated from separate messages for \fBreaddir()\fR, \fBtelldir()\fR, |
| 706 |
\&\fBseekdir()\fR, \fBrewinddir()\fR and \fBclosedir()\fR as part of refactoring for |
| 707 |
[GH #22394 <https://github.com/Perl/perl5/issues/22394>]. |
| 708 |
.IP \(bu 4 |
| 709 |
Useless use of \f(CW%s\fR in void context |
| 710 |
.Sp |
| 711 |
This warning now triggers for use of a chained comparison like \f(CW\*(C`0 < $x < 1\*(C'\fR. |
| 712 |
[GH #22969 <https://github.com/Perl/perl5/issues/22969>] |
| 713 |
.IP \(bu 4 |
| 714 |
Use of uninitialized value%s |
| 715 |
.Sp |
| 716 |
Prevent this warning when accessing a function parameter in \f(CW@_\fR that |
| 717 |
is an lvalue reference to an untied hash element where the key was |
| 718 |
undefined. This warning is still produced at the point of call. |
| 719 |
[GH #22423 <https://github.com/Perl/perl5/issues/22423>] |
| 720 |
.SH "Utility Changes" |
| 721 |
.IX Header "Utility Changes" |
| 722 |
.SS \fIPorting/test\-dist\-modules.pl\fP |
| 723 |
.IX Subsection "Porting/test-dist-modules.pl" |
| 724 |
.IP \(bu 4 |
| 725 |
Separate installation (without overwriting installed modules) is now the default. |
| 726 |
.IP \(bu 4 |
| 727 |
Documentation is significantly enhanced. |
| 728 |
.SH "Configuration and Compilation" |
| 729 |
.IX Header "Configuration and Compilation" |
| 730 |
.IP \(bu 4 |
| 731 |
Fix compilation on platforms (e.g. "Gentoo Prefix") with only a C locale [GH #22569 <https://github.com/Perl/perl5/issues/22569>] |
| 732 |
Bug first reported downstream bugs.gentoo.org/939014 <https://bugs.gentoo.org/939014> |
| 733 |
.IP \(bu 4 |
| 734 |
The (mostly undocumented) configuration macro \f(CW\*(C`PERL_STRICT_CR\*(C'\fR has been |
| 735 |
removed. When enabled (e.g. with \f(CW\*(C`./Configure \-A ccflags=\-DPERL_STRICT_CR\*(C'\fR), |
| 736 |
it would make the perl parser throw a fatal error when it encountered a CR |
| 737 |
(carriage return) character in source files. The default (and now only) |
| 738 |
behavior of the perl parser is to strip CRs paired with newline characters and |
| 739 |
otherwise treat them as whitespace. |
| 740 |
.Sp |
| 741 |
(\f(CW\*(C`PERL_STRICT_CR\*(C'\fR was originally introduced in perl 5.005 to optionally |
| 742 |
restore backward compatibility with perl 5.004, which had made CR in source |
| 743 |
files an error. Before that, CR was accepted, but retained literally in quoted |
| 744 |
multi\-line constructs such as here\-documents, even at the end of a line.) |
| 745 |
.IP \(bu 4 |
| 746 |
Similarly, the (even less documented) configuration macro \f(CW\*(C`PERL_CR_FILTER\*(C'\fR has |
| 747 |
been removed. When enabled, it would install a default source filter to strip |
| 748 |
carriage returns from source code before the parser proper got to see it. |
| 749 |
.SH Testing |
| 750 |
.IX Header "Testing" |
| 751 |
Tests were added and changed to reflect the other additions and |
| 752 |
changes in this release. Furthermore, these significant changes were |
| 753 |
made: |
| 754 |
.IP \(bu 4 |
| 755 |
A new \fIt/run/todo.t\fR test script was added as a place for TODO tests |
| 756 |
for known unfixed bugs. Patches are welcome to add to this file. |
| 757 |
.IP \(bu 4 |
| 758 |
Added testing of the perl headers against the C++ compiler |
| 759 |
corresponding to the C compiler perl is being built with. |
| 760 |
[GH #22232 <https://github.com/Perl/perl5/issues/22232>] |
| 761 |
.SH "Platform Support" |
| 762 |
.IX Header "Platform Support" |
| 763 |
.SS "Platform\-Specific Notes" |
| 764 |
.IX Subsection "Platform-Specific Notes" |
| 765 |
.IP "arm64 Darwin" 4 |
| 766 |
.IX Item "arm64 Darwin" |
| 767 |
Fix arm64 darwin hints when using use64bitall with Configure [GH #22672 <https://github.com/Perl/perl5/issues/22672>] |
| 768 |
.IP Android 4 |
| 769 |
.IX Item "Android" |
| 770 |
Changes to \fIperl_langinfo.h\fR for Android [GH #22650 <https://github.com/Perl/perl5/issues/22650>] related to [GH #22627 <https://github.com/Perl/perl5/issues/22627>]. |
| 771 |
.IP Cygwin 4 |
| 772 |
.IX Item "Cygwin" |
| 773 |
\&\fIcygwin.c\fR: fix several silly/terrible C errors. [GH #22724 <https://github.com/Perl/perl5/issues/22724>] |
| 774 |
.Sp |
| 775 |
Supply an explicit base address for \f(CW\*(C`cygperl*.dll\*(C'\fR that cannot |
| 776 |
conflict with those generated by \f(CW\*(C`\-\-enable\-auto\-image\-base\*(C'\fR. [GH #22695 <https://github.com/Perl/perl5/issues/22695>][GH #22104 <https://github.com/Perl/perl5/issues/22104>] |
| 777 |
.IP "MacOS (Darwin)" 4 |
| 778 |
.IX Item "MacOS (Darwin)" |
| 779 |
Collation of strings using locales on MacOS 15 (Darwin 24) and up has |
| 780 |
been turned off due to a failed assertion in its libc. |
| 781 |
.Sp |
| 782 |
If earlier versions are also experiencing issues (such as failures in |
| 783 |
\&\fIlocale.t\fR), you can explicitly disable locale collation by adding the |
| 784 |
\&\f(CW\*(C`\-Accflags=\-DNO_LOCALE_COLLATE\*(C'\fR option to your invocation of \f(CW\*(C`./Configure\*(C'\fR, |
| 785 |
or just \f(CW\*(C`\-DNO_LOCALE_COLLATE\*(C'\fR to the \f(CW\*(C`ccflags\*(C'\fR and \f(CW\*(C`cppflags\*(C'\fR |
| 786 |
variables in \fIconfig.sh\fR. |
| 787 |
.SH "Internal Changes" |
| 788 |
.IX Header "Internal Changes" |
| 789 |
.IP \(bu 4 |
| 790 |
The "\f(CW\*(C`sv_strftime_ints\*(C'\fR" in perlapi function is introduced. This is an |
| 791 |
enhanced version of "\f(CW\*(C`my_strftime\*(C'\fR" in perlapi, which is retained for |
| 792 |
backwards compatibility. Both are to call \fBstrftime\fR\|(3) when you have |
| 793 |
the year, month, hour, etc. The new function handles UTF8ness for you, |
| 794 |
and allows you to specify if you want the possibility of daylight |
| 795 |
savings time to be considered. \f(CW\*(C`my_strftime\*(C'\fR never considers DST. |
| 796 |
.IP \(bu 4 |
| 797 |
The \f(CW\*(C`bytes_to_utf8\*(C'\fR, \f(CW\*(C`bytes_from_utf8\*(C'\fR, and \f(CW\*(C`bytes_from_utf8_loc\*(C'\fR |
| 798 |
functions are no longer experimental. |
| 799 |
.IP \(bu 4 |
| 800 |
Calls to \fBcall_argv()\fR with the \f(CW\*(C`G_DISCARD\*(C'\fR flag |
| 801 |
set also ensure the SV parameters constructed from the \f(CW\*(C`argv\*(C'\fR |
| 802 |
parameter are released before \f(CWcall_argv()\fR returns. Previously they |
| 803 |
were released on the next FREETMPS. [GH #22255 <https://github.com/Perl/perl5/issues/22255>] |
| 804 |
.IP \(bu 4 |
| 805 |
When built with the \f(CW\*(C`\-DDEBUGGING\*(C'\fR compile option, perl API functions that |
| 806 |
take pointers to distinct types of SVs (AVs, HVs or CVs) will check the |
| 807 |
\&\f(CWSvTYPE()\fR of the passed values to ensure they are valid. Additionally, |
| 808 |
internal code within core functions that attempts to extract AVs, HVs or CVs |
| 809 |
from reference values passed in will also perform such checks. |
| 810 |
.Sp |
| 811 |
While this has been entirely tested by normal Perl CI testing, there may |
| 812 |
still be some corner\-cases where these constraints are violated in |
| 813 |
otherwise\-valid calls. These may require further investigation if they are |
| 814 |
found, and specific code to be adjusted to account for it. |
| 815 |
.IP \(bu 4 |
| 816 |
The \f(CWop_dump()\fR function has been expanded to include additional information |
| 817 |
about the recent \f(CW\*(C`OP_METHSTART\*(C'\fR and \f(CW\*(C`OP_INITFIELD\*(C'\fR ops, as well as for |
| 818 |
\&\f(CW\*(C`OP_ARGCHECK\*(C'\fR and \f(CW\*(C`OP_ARGELEM\*(C'\fR which had not been done previously. |
| 819 |
.IP \(bu 4 |
| 820 |
\&\f(CWop_dump()\fR now also has the facility to print extra debugging information |
| 821 |
about custom operators, if those operators register a helper function via the |
| 822 |
new \f(CW\*(C`xop_dump\*(C'\fR element of the \f(CW\*(C`XOP\*(C'\fR structure. For more information, see the |
| 823 |
relevant additions to perlguts. |
| 824 |
.IP \(bu 4 |
| 825 |
New API functions are introduced to convert strings encoded in UTF\-8 to |
| 826 |
their ordinal code point equivalent. These are safe to use by default, |
| 827 |
and generally more convenient to use than the existing ones. |
| 828 |
.Sp |
| 829 |
"\f(CW\*(C`utf8_to_uv\*(C'\fR" in perlapi and "\f(CW\*(C`utf8_to_uv_or_die\*(C'\fR" in perlapi replace |
| 830 |
"\f(CW\*(C`utf8_to_uvchr\*(C'\fR" in perlapi (which is retained for backwards |
| 831 |
compatibility), but you should convert to use the new forms, as likely |
| 832 |
you aren\*(Aqt using the old one safely. |
| 833 |
.Sp |
| 834 |
To convert in the opposite direction, you can now use |
| 835 |
"\f(CW\*(C`uv_to_utf8\*(C'\fR" in perlapi. This is not a new function, but a new synonym |
| 836 |
for "\f(CW\*(C`uvchr_to_utf8\*(C'\fR" in perlapi. It is added so you don\*(Aqt have to learn |
| 837 |
two sets of names. |
| 838 |
.Sp |
| 839 |
There are also two new functions, "\f(CW\*(C`strict_utf8_to_uv\*(C'\fR" in perlapi and |
| 840 |
"\f(CW\*(C`c9strict_utf8_to_uv\*(C'\fR" in perlapi which do the same thing except when |
| 841 |
the input string represents a code point that Unicode doesn\*(Aqt accept as |
| 842 |
legal for interchange, using either the strict original definition |
| 843 |
(\f(CW\*(C`strict_utf8_to_uv\*(C'\fR), or the looser one given by |
| 844 |
Unicode Corrigendum #9 <https://www.unicode.org/versions/corrigendum9.html> |
| 845 |
(\f(CW\*(C`c9strict_utf8_to_uv\*(C'\fR). When the input string represents one of the |
| 846 |
restricted code points, these functions return the Unicode |
| 847 |
\&\f(CW\*(C`REPLACEMENT CHARACTER\*(C'\fR instead. |
| 848 |
.Sp |
| 849 |
Also "\f(CW\*(C`extended_utf8_to_uv\*(C'\fR" in perlapi is a synonym for \f(CW\*(C`utf8_to_uv\*(C'\fR, for use |
| 850 |
when you want to emphasize that the entire range of Perl extended UTF\-8 |
| 851 |
is acceptable. |
| 852 |
.Sp |
| 853 |
There are also replacement functions for the three more specialized |
| 854 |
conversion functions that you are unlikely to need to use. Again, the |
| 855 |
old forms are kept for backwards compatibility, but you should convert |
| 856 |
to use the new forms. |
| 857 |
.Sp |
| 858 |
"\f(CW\*(C`utf8_to_uv_flags\*(C'\fR" in perlapi replaces "\f(CW\*(C`utf8n_to_uvchr\*(C'\fR" in perlapi. |
| 859 |
.Sp |
| 860 |
"\f(CW\*(C`utf8_to_uv_errors\*(C'\fR" in perlapi replaces "\f(CW\*(C`utf8n_to_uvchr_error\*(C'\fR" in perlapi. |
| 861 |
.Sp |
| 862 |
"\f(CW\*(C`utf8_to_uv_msgs\*(C'\fR" in perlapi replaces |
| 863 |
"\f(CW\*(C`utf8n_to_uvchr_msgs\*(C'\fR" in perlapi. |
| 864 |
.Sp |
| 865 |
Also added are the inverse functions "\f(CW\*(C`uv_to_utf8_flags\*(C'\fR" in perlapi |
| 866 |
and "\f(CW\*(C`uv_to_utf8_msgs\*(C'\fR" in perlapi, which are synonyms for the existing |
| 867 |
functions, "\f(CW\*(C`uvchr_to_utf8_flags\*(C'\fR" in perlapi and |
| 868 |
"\f(CW\*(C`uvchr_to_utf8_flags_msgs\*(C'\fR" in perlapi respectively. These are provided only |
| 869 |
so you don\*(Aqt have to learn two sets of names. |
| 870 |
.IP \(bu 4 |
| 871 |
Three new API functions are introduced to convert strings encoded in |
| 872 |
UTF\-8 to native bytes format (if possible). These are easier to use |
| 873 |
than the existing ones, and they avoid unnecessary memory allocations. |
| 874 |
The functions are "\f(CW\*(C`utf8_to_bytes_overwrite\*(C'\fR" in perlapi which is used |
| 875 |
when it is ok for the input string to be overwritten with the converted |
| 876 |
result; and "\f(CW\*(C`utf8_to_bytes_new_pv\*(C'\fR" in perlapi and |
| 877 |
"\f(CW\*(C`utf8_to_bytes_temp_pv\*(C'\fR" in perlapi when the original string must be |
| 878 |
preserved intact. \f(CW\*(C`utf8_to_bytes_temp_pv\*(C'\fR returns the result in a |
| 879 |
temporary using perlapi/\f(CW\*(C`SAVEFREEPV\*(C'\fR that will automatically be |
| 880 |
destroyed. With \f(CW\*(C`utf8_to_bytes_new_pv\*(C'\fR, you are responsible for |
| 881 |
freeing the newly allocated memory that is returned if the conversion is |
| 882 |
successful. |
| 883 |
.Sp |
| 884 |
The latter two functions are designed to replace |
| 885 |
"\f(CW\*(C`bytes_from_utf8\*(C'\fR" in perlapi which creates memory unnecessarily, or |
| 886 |
unnecessarily large. |
| 887 |
.IP \(bu 4 |
| 888 |
New API functions \f(CWvalid_identifier_pve()\fR, |
| 889 |
\&\f(CWvalid_identifier_pvn()\fR and |
| 890 |
\&\f(CWvalid_identifier_sv()\fR have been added, which |
| 891 |
test if a string would be considered by Perl to be a valid identifier name. |
| 892 |
.IP \(bu 4 |
| 893 |
When assigning from an SVt_IV into a SVt_NV (or vice versa), providing that |
| 894 |
both are "bodyless" types, Perl_sv_setsv_flags will now just change the |
| 895 |
destination type to match the source type. Previously, an SVt_IV would have |
| 896 |
been upgraded to a SVt_PVNV to store an NV, and an SVt_NV would have been |
| 897 |
upgraded to a SVt_PVIV to store an IV. This change prevents the need to |
| 898 |
allocate \- and later free \- the relevant body struct. |
| 899 |
.IP \(bu 4 |
| 900 |
Two new API functions are introduced to convert strings encoded in |
| 901 |
native bytes format to UTF\-8. These return the string unchanged if its |
| 902 |
UTF\-8 representation is the same as the original. Otherwise, new memory |
| 903 |
is allocated to contain the converted string. This is in contrast to |
| 904 |
the existing "\f(CW\*(C`bytes_to_utf8\*(C'\fR" in perlapi which always allocates new |
| 905 |
memory. The new functions are "\f(CW\*(C`bytes_to_utf8_free_me\*(C'\fR" in perlapi and |
| 906 |
"\f(CW\*(C`bytes_to_utf8_temp_pv\*(C'\fR" in perlapi. |
| 907 |
"\f(CW\*(C`bytes_to_utf8_temp_pv\*(C'\fR" in perlapi arranges for the new memory to |
| 908 |
automatically be freed. With \f(CW\*(C`bytes_to_utf8_free_me\*(C'\fR, you are |
| 909 |
responsible for freeing any newly allocated memory. |
| 910 |
.IP \(bu 4 |
| 911 |
The way that subroutine signatures are parsed by the parser grammar has been |
| 912 |
changed. |
| 913 |
.Sp |
| 914 |
Previously, when parsing individual signature parameters, the parser would |
| 915 |
accumulate an \f(CW\*(C`OP_ARGELEM\*(C'\fR optree fragment for each parameter on the parser |
| 916 |
stack, collecting them in an \f(CW\*(C`OP_LIST\*(C'\fR sequence, before finally building the |
| 917 |
complete argument handling optree itself, in a large action block defined |
| 918 |
directly in \fIperly.y\fR. |
| 919 |
.Sp |
| 920 |
In the new approach, all the optree generation is handled by newly\-defined |
| 921 |
functions in \fIop.c\fR which are called by the action blocks in the parser. |
| 922 |
These do not keep state on the parser stack, but instead in a dedicated memory |
| 923 |
structure referenced by the main \f(CW\*(C`PL_parser\*(C'\fR structure. This is intended to |
| 924 |
be largely opaque to other code, and accessed only via the new functions. |
| 925 |
.Sp |
| 926 |
This new arrangement is intended to allow more flexible code generation and |
| 927 |
additional features to be developed in the future. |
| 928 |
.IP \(bu 4 |
| 929 |
Three new API functions have been added to interact with the regexp global |
| 930 |
match position stored in an SV. These are \f(CWsv_regex_global_pos_get()\fR, |
| 931 |
\&\f(CWsv_regex_global_pos_set()\fR and \f(CWsv_regex_global_pos_clear()\fR. Using these |
| 932 |
API functions avoids XS modules needing to know about or interact directly |
| 933 |
with the way this position is currently stored, which involves the |
| 934 |
\&\f(CW\*(C`PERL_MAGIC_regex_global\*(C'\fR magic type. |
| 935 |
.IP \(bu 4 |
| 936 |
New \f(CW\*(C`SvVSTRING\*(C'\fR API macro |
| 937 |
.Sp |
| 938 |
A new API macro has been added, which is used to obtain the second string |
| 939 |
buffer out of a "vstring" SV, in a manner similar to the \f(CW\*(C`SvPV\*(C'\fR macro which |
| 940 |
obtains the regular string buffer out of a regular SV. |
| 941 |
.Sp |
| 942 |
.Vb 2 |
| 943 |
\& STRLEN len; |
| 944 |
\& const char *vstr_pv = SvVSTRING(sv, vstr_len); |
| 945 |
.Ve |
| 946 |
.Sp |
| 947 |
See "\f(CW\*(C`SvVSTRING\*(C'\fR" in perlapi. |
| 948 |
.SH "Selected Bug Fixes" |
| 949 |
.IX Header "Selected Bug Fixes" |
| 950 |
.IP \(bu 4 |
| 951 |
Fix null pointer dereference in S_SvREFCNT_dec [GH #16627 <https://github.com/Perl/perl5/issues/16627>]. |
| 952 |
.IP \(bu 4 |
| 953 |
Fix feature \*(Aqclass\*(Aq Segmentation fault in DESTROY [GH #22278 <https://github.com/Perl/perl5/issues/22278>]. |
| 954 |
.IP \(bu 4 |
| 955 |
\&\f(CW\*(C`chdir\*(C'\fR now returns real booleans (as its documentation describes), not |
| 956 |
integers. This means the result of a failed \f(CW\*(C`chdir\*(C'\fR now stringifies to \f(CW\*(Aq\*(Aq\fR, |
| 957 |
not \f(CW\*(Aq0\*(Aq\fR. |
| 958 |
.Sp |
| 959 |
[GH #22365 <https://github.com/Perl/perl5/issues/22365>] |
| 960 |
.IP \(bu 4 |
| 961 |
Compound assignment operators return lvalues that can be further modified: |
| 962 |
.Sp |
| 963 |
.Vb 4 |
| 964 |
\& ($x &= $y) += $z; |
| 965 |
\& # Equivalent to: |
| 966 |
\& # $x &= $y; |
| 967 |
\& # $x += $z; |
| 968 |
.Ve |
| 969 |
.Sp |
| 970 |
However, the separate numeric/string bitwise operators provided by the |
| 971 |
\&\f(CW\*(C`bitwise\*(C'\fR feature, \f(CW\*(C`&= ^= |= &.= ^.= |.=\*(C'\fR, |
| 972 |
did not return such lvalues: |
| 973 |
.Sp |
| 974 |
.Vb 4 |
| 975 |
\& use feature qw(bitwise); |
| 976 |
\& ($x &= $y) += $z; |
| 977 |
\& # Used to die: |
| 978 |
\& # Can\*(Aqt modify numeric bitwise and (&) in addition (+) at ... |
| 979 |
.Ve |
| 980 |
.Sp |
| 981 |
This has been corrected. [GH #22412 <https://github.com/Perl/perl5/issues/22412>] |
| 982 |
.IP \(bu 4 |
| 983 |
Starting in v5.39.8, "\f(CW\*(C`strftime\*(C'\fR" in POSIX would crash or produce odd errors |
| 984 |
(such as \f(CW\*(C`Out of memory in perl:util:safesysmalloc\*(C'\fR) when given a format |
| 985 |
string that wasn\*(Aqt actually a string, but a number, \f(CW\*(C`undef\*(C'\fR, or an object |
| 986 |
(even one with overloaded string conversion). |
| 987 |
.Sp |
| 988 |
Now \f(CW\*(C`strftime\*(C'\fR stringifies its first argument, as before. |
| 989 |
[GH #22498 <https://github.com/Perl/perl5/issues/22498>] |
| 990 |
.Sp |
| 991 |
Also, fix \f(CWPOSIX::strftime()\fR [GH #22369 <https://github.com/Perl/perl5/issues/22369>]. |
| 992 |
.IP \(bu 4 |
| 993 |
\&\f(CW\*(C`pack("p", ...)\*(C'\fR and \f(CW\*(C`pack("P", ...)\*(C'\fR now \fBSvPV_force()\fR the supplied |
| 994 |
SV unless it is read only. This will remove CoW from the SV and |
| 995 |
prevents code that writes through the generated pointer from modifying |
| 996 |
the value of other SVs that happen the share the same CoWed string |
| 997 |
buffer. |
| 998 |
.Sp |
| 999 |
Note: this does not make \f(CW\*(C`pack("p",... )\*(C'\fR safe, if the SV is magical |
| 1000 |
then any writes to the buffer will likely be discarded on the next |
| 1001 |
read. [GH #22380 <https://github.com/Perl/perl5/issues/22380>] |
| 1002 |
.IP \(bu 4 |
| 1003 |
Enforce \f(CW\*(C`no feature "bareword_filehandles"\*(C'\fR for bareword file handles |
| 1004 |
that have strictness removed because they are used in \fBopen()\fR with a |
| 1005 |
"dup" mode, such as in \f(CW\*(C`open my $fh, ">&", THISHANDLE\*(C'\fR. [GH #22568 <https://github.com/Perl/perl5/issues/22568>] |
| 1006 |
.IP \(bu 4 |
| 1007 |
Using \f(CW\*(C`goto\*(C'\fR to tail call, or using the \fBcall_sv()\fR and related APIs to |
| 1008 |
call, any of \fBtrim()\fR, \fBrefaddr()\fR, \fBreftype()\fR, \fBceil()\fR, \fBfloor()\fR or |
| 1009 |
\&\fBstringify()\fR in the \f(CW\*(C`builtin::\*(C'\fR package would crash or assert due to a |
| 1010 |
\&\f(CW\*(C`TARG\*(C'\fR handling bug. [GH #22542 <https://github.com/Perl/perl5/issues/22542>] |
| 1011 |
.IP \(bu 4 |
| 1012 |
Fix \fBsv_gets()\fR to accept a \f(CW\*(C`SSize_t\*(C'\fR append offset instead of \f(CW\*(C`I32\*(C'\fR. |
| 1013 |
This prevents integer overflows when appending to a large \f(CW\*(C`SV\*(C'\fR for |
| 1014 |
\&\f(CW\*(C`readpipe\*(C'\fR aka \f(CW\*(C`qx//\*(C'\fR and \f(CW\*(C`readline\*(C'\fR. |
| 1015 |
<https://www.perlmonks.org/?node_id=11161665> |
| 1016 |
.IP \(bu 4 |
| 1017 |
Fixed an issue where \f(CW\*(C`utf8n_to_uvchr\*(C'\fR failed to correctly identify |
| 1018 |
certain invalid UTF\-8 sequences as invalid. Specifically, sequences |
| 1019 |
that start with continuation bytes or unassigned bytes could cause |
| 1020 |
unexpected behavior or a panic. This fix ensures that such invalid |
| 1021 |
sequences are now properly detected and handled. This correction |
| 1022 |
also resolves related issues in modules that handle UTF\-8 processing, |
| 1023 |
such as \f(CW\*(C`Encode.pm\*(C'\fR. |
| 1024 |
.IP \(bu 4 |
| 1025 |
The perl parser would erroneously parse some POD directives |
| 1026 |
as if they were \f(CW\*(C`=cut\*(C'\fR. Some other POD directives |
| 1027 |
whose names start with \fIcut\fR, prematurely terminating an embedded POD section. |
| 1028 |
The following cases were affected: \fIcut\fR followed by a digit (e.g. |
| 1029 |
\&\f(CW\*(C`=cut2studio\*(C'\fR), \fIcut\fR followed by an underscore (e.g. \f(CW\*(C`=cut_grass\*(C'\fR), and in |
| 1030 |
string \f(CW\*(C`eval\*(C'\fR, any identifier starting with \fIcut\fR (e.g. \f(CW\*(C`=cute\*(C'\fR). |
| 1031 |
[GH #22759 <https://github.com/Perl/perl5/issues/22759>] |
| 1032 |
.IP \(bu 4 |
| 1033 |
Builds with \f(CW\*(C`\-msse\*(C'\fR and quadmath on 32\-bit x86 systems would crash |
| 1034 |
with a misaligned access early in the build. [GH #22577 <https://github.com/Perl/perl5/issues/22577>] |
| 1035 |
.IP \(bu 4 |
| 1036 |
On threaded builds on POSIX\-like systems, if the perl signal handler |
| 1037 |
receives a signal, we now resend the signal to the main perl thread. Previously |
| 1038 |
this would crash. [GH #22487 <https://github.com/Perl/perl5/issues/22487>] |
| 1039 |
.IP \(bu 4 |
| 1040 |
Declaring a lexically scoped array or hash using \f(CW\*(C`state\*(C'\fR within a subroutine |
| 1041 |
and then immediately returning no longer triggers a "Bizarre copy of HASH/ARRAY |
| 1042 |
in subroutine exit" error. [GH #18630 <https://github.com/Perl/perl5/issues/18630>] |
| 1043 |
.IP \(bu 4 |
| 1044 |
\&\f(CWbuiltin::trim()\fR didn\*(Aqt properly clear \f(CW\*(C`TARG\*(C'\fR which could result in |
| 1045 |
out of date cached numeric versions of the value being used on a |
| 1046 |
second evaluation. Properly clear any cached values. [GH #22784 <https://github.com/Perl/perl5/issues/22784>] |
| 1047 |
.IP \(bu 4 |
| 1048 |
"shmread" in perlfunc and "shmwrite" in perlfunc are no longer limited to 31\-bit |
| 1049 |
values and can use all the available bits on a platform for their POS and |
| 1050 |
SIZE arguments. |
| 1051 |
[GH #22895 <https://github.com/Perl/perl5/issues/22895>] |
| 1052 |
.IP \(bu 4 |
| 1053 |
"shmread" in perlfunc is now better behaved if VAR is not a plain string. If VAR |
| 1054 |
is a tied variable, it calls \f(CW\*(C`STORE\*(C'\fR once; previously, it would also call |
| 1055 |
\&\f(CW\*(C`FETCH\*(C'\fR, but without using the result. If VAR is a reference, the referenced |
| 1056 |
entity has its refcount properly decremented when VAR is turned into a string; |
| 1057 |
previously, it would leak memory. |
| 1058 |
[GH #22898 <https://github.com/Perl/perl5/issues/22898>] |
| 1059 |
.IP \(bu 4 |
| 1060 |
The \f(CW$SIG{_\|_DIE_\|_}\fR and \f(CW$SIG{_\|_WARN_\|_}\fR handlers can no longer be invoked |
| 1061 |
recursively, either deliberately or by accident, as described in |
| 1062 |
"%SIG" in perlvar. That is, when an exception (or warning) triggers a call to a |
| 1063 |
\&\f(CW$SIG{_\|_DIE_\|_}\fR (or \f(CW$SIG{_\|_WARN_\|_}\fR) handler, further exceptions (or |
| 1064 |
warnings) are processed directly, ignoring \f(CW%SIG\fR until the original |
| 1065 |
\&\f(CW$SIG{_\|_DIE_\|_}\fR (or \f(CW$SIG{_\|_WARN_\|_}\fR) handler call returns. |
| 1066 |
[GH #14527 <https://github.com/Perl/perl5/issues/14527>], [GH #22984 <https://github.com/Perl/perl5/issues/22984>], [GH #22987 <https://github.com/Perl/perl5/issues/22987>] |
| 1067 |
.IP \(bu 4 |
| 1068 |
The \f(CWObjectFIELDS()\fR for an object and \f(CW\*(C`xhv_class_fields\*(C'\fR for the |
| 1069 |
object\*(Aqs stash weren\*(Aqt always NULL or not\-NULL, confusing \f(CWsv_dump()\fR |
| 1070 |
(and hence Devel::Peek\*(Aqs \f(CWDump()\fR) into crashing on an object with no |
| 1071 |
defined fields in some cases. [GH #22959 <https://github.com/Perl/perl5/issues/22959>] |
| 1072 |
.IP \(bu 4 |
| 1073 |
When comparing strings when using a UTF\-8 locale, the behavior was |
| 1074 |
previously undefined if either or both contained an above\-Unicode code |
| 1075 |
point, such as 0x110000. Now all such code points will collate the same |
| 1076 |
as the highest Unicode code point, U+10FFFF. [GH #22989 <https://github.com/Perl/perl5/issues/22989>] |
| 1077 |
.IP \(bu 4 |
| 1078 |
In regexes, the contents of \f(CW\*(C`\eg{...}\*(C'\fR backreferences are now properly |
| 1079 |
validated. Previously, \f(CW\*(C`\eg{1 FOO}\*(C'\fR was silently parsed as \f(CW\*(C`\eg{1}\*(C'\fR, ignoring |
| 1080 |
everything after the first number. |
| 1081 |
[GH #23050 <https://github.com/Perl/perl5/issues/23050>] |
| 1082 |
.IP \(bu 4 |
| 1083 |
A run\-time pattern which contained a code block which recursed back to the |
| 1084 |
same bit of code which ran that match, could cause a crash. |
| 1085 |
[GH #22869 <https://github.com/Perl/perl5/issues/22869>] |
| 1086 |
.Sp |
| 1087 |
For example: |
| 1088 |
.Sp |
| 1089 |
.Vb 3 |
| 1090 |
\& my $r = qr/... (?{ foo() if ... }) .../; |
| 1091 |
\& sub foo { $string =~ $r } |
| 1092 |
\& foo() |
| 1093 |
.Ve |
| 1094 |
.IP \(bu 4 |
| 1095 |
In some cases an \f(CW\*(C`eval\*(C'\fR would not add integer parts to the source |
| 1096 |
lines saved by the debugger. [GH #23151 <https://github.com/Perl/perl5/issues/23151>] |
| 1097 |
.IP \(bu 4 |
| 1098 |
In debugging mode, perl saves source lines from all files |
| 1099 |
(plus an indication of whether each line is breakable) |
| 1100 |
for use by the debugger. The internal storage format has been |
| 1101 |
optimized to use less memory. |
| 1102 |
This should save 24 bytes per stored line for 64\-bit |
| 1103 |
systems, more for \f(CW\*(C`\-Duselongdouble\*(C'\fR or \f(CW\*(C`\-Dusequadmath\*(C'\fR builds. |
| 1104 |
Discussed in [GH #23171 <https://github.com/Perl/perl5/issues/23171>]. |
| 1105 |
.IP \(bu 4 |
| 1106 |
Ensure cloning the save stack for fork emulation doesn\*(Aqt duplicate |
| 1107 |
freeing the RExC state. [GH #23022 <https://github.com/Perl/perl5/issues/23022>] |
| 1108 |
.IP \(bu 4 |
| 1109 |
Smartmatch against a code reference that uses a loop exit such as |
| 1110 |
\&\f(CW\*(C`last\*(C'\fR would crash perl. [GH #16608 <https://github.com/Perl/perl5/issues/16608>] |
| 1111 |
.IP \(bu 4 |
| 1112 |
Class initializers and \f(CW\*(C`ADJUST\*(C'\fR blocks, per perlclass, that |
| 1113 |
called \f(CW\*(C`last\*(C'\fR or other loop exits would crash perl. Same cause as |
| 1114 |
for [GH #16608 <https://github.com/Perl/perl5/issues/16608>]. |
| 1115 |
.IP \(bu 4 |
| 1116 |
Exceptions thrown and caught entirely within a \f(CW\*(C`defer {}\*(C'\fR or \f(CW\*(C`finally {}\*(C'\fR |
| 1117 |
block no longer stop the outer run\-loop. |
| 1118 |
.Sp |
| 1119 |
Code such as the following would stop running the contents of the \f(CW\*(C`defer\*(C'\fR |
| 1120 |
block once the inner exception in the inner \f(CW\*(C`try\*(C'\fR/\f(CW\*(C`catch\*(C'\fR block was caught. |
| 1121 |
This has now been fixed, and runs as expected. ([GH #23064 <https://github.com/Perl/perl5/issues/23064>]). |
| 1122 |
.Sp |
| 1123 |
.Vb 3 |
| 1124 |
\& defer { |
| 1125 |
\& try { die "It breaks\en"; } |
| 1126 |
\& catch ($e) { warn $e } |
| 1127 |
\& |
| 1128 |
\& say "This line would never run"; |
| 1129 |
\& } |
| 1130 |
.Ve |
| 1131 |
.IP \(bu 4 |
| 1132 |
"readline" in perlfunc now clears the error flag if an error occurs when |
| 1133 |
reading and that error is \f(CW\*(C`EAGAIN\*(C'\fR or \f(CW\*(C`EWOULDBLOCK\*(C'\fR. This allows |
| 1134 |
code that depended on \f(CW\*(C`readline\*(C'\fR to clear all errors to ignore |
| 1135 |
these relatively harmless errors. [GH #22883 <https://github.com/Perl/perl5/issues/22883>] |
| 1136 |
.IP \(bu 4 |
| 1137 |
\&\f(CW\*(C`open\*(C'\fR automatically creates an anonymous temporary file |
| 1138 |
when passed \f(CW\*(C`undef\*(C'\fR as a filename: |
| 1139 |
.Sp |
| 1140 |
.Vb 1 |
| 1141 |
\& open(my $fh, "+>", undef) or die ... |
| 1142 |
.Ve |
| 1143 |
.Sp |
| 1144 |
This is supposed to work only when the undefined value is the one returned by |
| 1145 |
the \f(CW\*(C`undef\*(C'\fR function. |
| 1146 |
.Sp |
| 1147 |
In perls before 5.41.3, this caused a problem due to the fact that the same |
| 1148 |
undefined value can be generated by lookups of non\-existent hash keys or array |
| 1149 |
elements, which can lead to bugs in user\-level code (reported as [GH #22385 <https://github.com/Perl/perl5/issues/22385>]). |
| 1150 |
.Sp |
| 1151 |
In 5.41.3, additional checks based on the syntax tree of the call site were |
| 1152 |
added, which fixed this issue for some number of common cases, though not all |
| 1153 |
of them, at the cost of breaking the ability of APIs that wrap \f(CW\*(C`open\*(C'\fR to |
| 1154 |
expose its anonymous file mode. A notable example of such an API is autodie. |
| 1155 |
.Sp |
| 1156 |
This release reverts to the old problem in preference to the new one for the |
| 1157 |
time being. |
| 1158 |
.SH Obituaries |
| 1159 |
.IX Header "Obituaries" |
| 1160 |
.SS "Abe Timmerman" |
| 1161 |
.IX Subsection "Abe Timmerman" |
| 1162 |
Abe Timmerman (ABELTJE) passed away on August 15, 2024. |
| 1163 |
.PP |
| 1164 |
Since 2002, Abe built and maintained the Test::Smoke project: "a |
| 1165 |
set of scripts and modules that try to run the Perl core tests on as |
| 1166 |
many configurations as possible and combine the results into an easy to |
| 1167 |
read report". Smoking Perl on as many platforms and configurations as |
| 1168 |
possible has been instrumental in finding bugs and developing patches |
| 1169 |
for those bugs. |
| 1170 |
.PP |
| 1171 |
Abe was a regular attendee of the Perl Toolchain Summit (née Perl QA |
| 1172 |
Hackathon), the Dutch Perl Workshop and the Amsterdam.pm user group |
| 1173 |
meetings. With his kindness, his smile and his laugh, he helped make |
| 1174 |
Perl and its community better. |
| 1175 |
.PP |
| 1176 |
Abeltje\*(Aqs memorial card said "Grab every opportunity to have a drink of |
| 1177 |
bubbly. This is an opportunity". We\*(Aqll miss you Abe, and we\*(Aqll have a |
| 1178 |
drink of bubbly in your honor. |
| 1179 |
.SS "Andrew Main" |
| 1180 |
.IX Subsection "Andrew Main" |
| 1181 |
Andrew Main (ZEFRAM) passed away on March 10, 2025. |
| 1182 |
.PP |
| 1183 |
Zefram was a brilliant person, seemingly knowledgeable in everything |
| 1184 |
and happy to impart his knowledge and share his striking insights with a |
| 1185 |
gentle, technical demeanor that often failed to convey the genuine care |
| 1186 |
with which he communicated. |
| 1187 |
.PP |
| 1188 |
It would be impossible to overstate the impact that Zefram has had on |
| 1189 |
both the language and culture of Perl over the years. From his countless |
| 1190 |
contributions to the code\-base, to his often quirky but always distinctive |
| 1191 |
appearances at conferences and gatherings, his influence and memory are |
| 1192 |
sure to endure long into the future. |
| 1193 |
.PP |
| 1194 |
Zefram wished to have no designated memorial location in |
| 1195 |
meatspace. His designated memorial location in cyberspace is |
| 1196 |
<http://www.fysh.org/~zefram/personal/>. |
| 1197 |
.SH Acknowledgements |
| 1198 |
.IX Header "Acknowledgements" |
| 1199 |
Perl 5.42.0 represents approximately 13 months of development since Perl |
| 1200 |
5.40.0 and contains approximately 280,000 lines of changes across 1,600 |
| 1201 |
files from 65 authors. |
| 1202 |
.PP |
| 1203 |
Excluding auto\-generated files, documentation and release tools, there were |
| 1204 |
approximately 94,000 lines of changes to 860 .pm, .t, .c and .h files. |
| 1205 |
.PP |
| 1206 |
Perl continues to flourish into its fourth decade thanks to a vibrant |
| 1207 |
community of users and developers. The following people are known to have |
| 1208 |
contributed the improvements that became Perl 5.42.0: |
| 1209 |
.PP |
| 1210 |
Aaron Dill, Andrei Horodniceanu, Andrew Ruthven, Antanas Vaitkus, Aristotle |
| 1211 |
Pagaltzis, Branislav Zahradník, brian d foy, Chad Granum, Chris \*(AqBinGOs\*(Aq |
| 1212 |
Williams, Craig A. Berry, Dabrien \*(AqDabe\*(Aq Murphy, Dagfinn Ilmari Mannsåker, |
| 1213 |
Dan Book, Daniel Dragan, Dan Jacobson, David Cantrell, David Mitchell, E. |
| 1214 |
Choroba, Ed J, Ed Sabol, Elvin Aslanov, Eric Herman, Erik Huelsmann, Gianni |
| 1215 |
Ceccarelli, Graham Knop, hbmaclean, H.Merijn Brand, iabyn, James E Keenan, |
| 1216 |
James Raspass, Johan Vromans, Karen Etheridge, Karl Williamson, Leon |
| 1217 |
Timmermans, Lukas Mai, Marek Rouchal, Marin Tsanov, Mark Fowler, Masahiro |
| 1218 |
Honma, Max Maischein, Paul Evans, Paul Johnson, Paul Marquess, Peter |
| 1219 |
Eisentraut, Peter John Acklam, Philippe Bruhat (BooK), pyrrhlin, Reini |
| 1220 |
Urban, Richard Leach, Robert Rothenberg, Robin Ragged, Russ Allbery, Scott |
| 1221 |
Baker, Sergei Zhmylev, Sevan Janiyan, Sisyphus, Štěpán Němec, Steve Hay, |
| 1222 |
TAKAI Kousuke, Thibault Duponchelle, Todd Rinaldo, Tony Cook, Unicode |
| 1223 |
Consortium, Vladimír Marek, Yves Orton. |
| 1224 |
.PP |
| 1225 |
The list above is almost certainly incomplete as it is automatically |
| 1226 |
generated from version control history. In particular, it does not include |
| 1227 |
the names of the (very much appreciated) contributors who reported issues to |
| 1228 |
the Perl bug tracker. |
| 1229 |
.PP |
| 1230 |
Many of the changes included in this version originated in the CPAN modules |
| 1231 |
included in Perl\*(Aqs core. We\*(Aqre grateful to the entire CPAN community for |
| 1232 |
helping Perl to flourish. |
| 1233 |
.PP |
| 1234 |
For a more complete list of all of Perl\*(Aqs historical contributors, please |
| 1235 |
see the \fIAUTHORS\fR file in the Perl source distribution. |
| 1236 |
.SH "Reporting Bugs" |
| 1237 |
.IX Header "Reporting Bugs" |
| 1238 |
If you find what you think is a bug, you might check the perl bug database |
| 1239 |
at <https://github.com/Perl/perl5/issues>. There may also be information at |
| 1240 |
<https://www.perl.org/>, the Perl Home Page. |
| 1241 |
.PP |
| 1242 |
If you believe you have an unreported bug, please open an issue at |
| 1243 |
<https://github.com/Perl/perl5/issues>. Be sure to trim your bug down to a |
| 1244 |
tiny but sufficient test case. |
| 1245 |
.PP |
| 1246 |
If the bug you are reporting has security implications which make it |
| 1247 |
inappropriate to send to a public issue tracker, then see |
| 1248 |
"SECURITY VULNERABILITY CONTACT INFORMATION" in perlsec |
| 1249 |
for details of how to report the issue. |
| 1250 |
.SH "Give Thanks" |
| 1251 |
.IX Header "Give Thanks" |
| 1252 |
If you wish to thank the Perl 5 Porters for the work we had done in Perl 5, |
| 1253 |
you can do so by running the \f(CW\*(C`perlthanks\*(C'\fR program: |
| 1254 |
.PP |
| 1255 |
.Vb 1 |
| 1256 |
\& perlthanks |
| 1257 |
.Ve |
| 1258 |
.PP |
| 1259 |
This will send an email to the Perl 5 Porters list with your show of thanks. |
| 1260 |
.SH "SEE ALSO" |
| 1261 |
.IX Header "SEE ALSO" |
| 1262 |
The \fIChanges\fR file for an explanation of how to view exhaustive details on |
| 1263 |
what changed. |
| 1264 |
.PP |
| 1265 |
The \fIINSTALL\fR file for how to build Perl. |
| 1266 |
.PP |
| 1267 |
The \fIREADME\fR file for general stuff. |
| 1268 |
.PP |
| 1269 |
The \fIArtistic\fR and \fICopying\fR files for copyright information. |