Debian Codegraph @

upstream/debian-unstable/man1/perl5280delta.1

raw
1 .\" -*- mode: troff; coding: utf-8 -*-
2 .\" Automatically generated by Pod::Man 5.0102 (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 .\"
57 .IX Title "PERL5280DELTA 1"
58 .TH PERL5280DELTA 1 2025-11-16 "perl v5.40.1" "Perl Programmers Reference Guide"
59 .\" For nroff, turn off justification. Always turn off hyphenation; it makes
60 .\" way too many mistakes in technical documents.
61 .if n .ad l
62 .nh
63 .SH NAME
64 perl5280delta \- what is new for perl v5.28.0
65 .SH DESCRIPTION
66 .IX Header "DESCRIPTION"
67 This document describes differences between the 5.26.0 release and the 5.28.0
68 release.
69 .PP
70 If you are upgrading from an earlier release such as 5.24.0, first read
71 perl5260delta, which describes differences between 5.24.0 and 5.26.0.
72 .SH "Core Enhancements"
73 .IX Header "Core Enhancements"
74 .SS "Unicode 10.0 is supported"
75 .IX Subsection "Unicode 10.0 is supported"
76 A list of changes is at
77 <http://www.unicode.org/versions/Unicode10.0.0>.
78 .ie n .SS """delete"" on key/value hash slices"
79 .el .SS "\f(CWdelete\fP on key/value hash slices"
80 .IX Subsection "delete on key/value hash slices"
81 \&\f(CW\*(C`delete\*(C'\fR can now be used on
82 key/value hash slices,
83 returning the keys along with the deleted values.
84 [GH #15982] <https://github.com/Perl/perl5/issues/15982>
85 .SS "Experimentally, there are now alphabetic synonyms for some regular expression assertions"
86 .IX Subsection "Experimentally, there are now alphabetic synonyms for some regular expression assertions"
87 If you find it difficult to remember how to write certain of the pattern
88 assertions, there are now alphabetic synonyms.
89 .PP
90 .Vb 7
91 \& CURRENT NEW SYNONYMS
92 \& \-\-\-\-\-\- \-\-\-\-\-\-\-\-\-\-\-\-
93 \& (?=...) (*pla:...) or (*positive_lookahead:...)
94 \& (?!...) (*nla:...) or (*negative_lookahead:...)
95 \& (?<=...) (*plb:...) or (*positive_lookbehind:...)
96 \& (?<!...) (*nlb:...) or (*negative_lookbehind:...)
97 \& (?>...) (*atomic:...)
98 .Ve
99 .PP
100 These are considered experimental, so using any of these will raise
101 (unless turned off) a warning in the \f(CW\*(C`experimental::alpha_assertions\*(C'\fR
102 category.
103 .SS "Mixed Unicode scripts are now detectable"
104 .IX Subsection "Mixed Unicode scripts are now detectable"
105 A mixture of scripts, such as Cyrillic and Latin, in a string is often
106 the sign of a spoofing attack. A new regular expression construct
107 now allows for easy detection of these. For example, you can say
108 .PP
109 .Vb 1
110 \& qr/(*script_run: \ed+ \eb )/x
111 .Ve
112 .PP
113 And the digits matched will all be from the same set of 10. You won't
114 get a look-alike digit from a different script that has a different
115 value than what it appears to be.
116 .PP
117 Or:
118 .PP
119 .Vb 1
120 \& qr/(*sr: \eb \ew+ \eb )/x
121 .Ve
122 .PP
123 makes sure that all the characters come from the same script.
124 .PP
125 You can also combine script runs with \f(CW\*(C`(?>...)\*(C'\fR (or
126 \&\f(CW\*(C`*atomic:...)\*(C'\fR).
127 .PP
128 Instead of writing:
129 .PP
130 .Vb 1
131 \& (*sr:(?<...))
132 .Ve
133 .PP
134 you can now run:
135 .PP
136 .Vb 3
137 \& (*asr:...)
138 \& # or
139 \& (*atomic_script_run:...)
140 .Ve
141 .PP
142 This is considered experimental, so using it will raise (unless turned
143 off) a warning in the \f(CW\*(C`experimental::script_run\*(C'\fR category.
144 .PP
145 See "Script Runs" in perlre.
146 .ie n .SS "In-place editing with ""perl \-i"" is now safer"
147 .el .SS "In-place editing with \f(CWperl \-i\fP is now safer"
148 .IX Subsection "In-place editing with perl -i is now safer"
149 Previously in-place editing (\f(CW\*(C`perl \-i\*(C'\fR) would delete or rename the
150 input file as soon as you started working on a new file.
151 .PP
152 Without backups this would result in loss of data if there was an
153 error, such as a full disk, when writing to the output file.
154 .PP
155 This has changed so that the input file isn't replaced until the
156 output file has been completely written and successfully closed.
157 .PP
158 This works by creating a work file in the same directory, which is
159 renamed over the input file once the output file is complete.
160 .PP
161 Incompatibilities:
162 .IP \(bu 4
163 Since this renaming needs to only happen once, if you create a thread
164 or child process, that renaming will only happen in the original
165 thread or process.
166 .IP \(bu 4
167 If you change directories while processing a file, and your operating
168 system doesn't provide the \f(CWunlinkat()\fR, \f(CWrenameat()\fR and \f(CWfchmodat()\fR
169 functions, the final rename step may fail.
170 .PP
171 [GH #15216] <https://github.com/Perl/perl5/issues/15216>
172 .SS "Initialisation of aggregate state variables"
173 .IX Subsection "Initialisation of aggregate state variables"
174 A persistent lexical array or hash variable can now be initialized,
175 by an expression such as \f(CW\*(C`state @a = qw(x y z)\*(C'\fR. Initialization of a
176 list of persistent lexical variables is still not possible.
177 .SS "Full-size inode numbers"
178 .IX Subsection "Full-size inode numbers"
179 On platforms where inode numbers are of a type larger than perl's native
180 integer numerical types, stat will preserve the full
181 content of large inode numbers by returning them in the form of strings of
182 decimal digits. Exact comparison of inode numbers can thus be achieved by
183 comparing with \f(CW\*(C`eq\*(C'\fR rather than \f(CW\*(C`==\*(C'\fR. Comparison with \f(CW\*(C`==\*(C'\fR, and other
184 numerical operations (which are usually meaningless on inode numbers),
185 work as well as they did before, which is to say they fall back to
186 floating point, and ultimately operate on a fairly useless rounded inode
187 number if the real inode number is too big for the floating point format.
188 .ie n .SS "The ""sprintf"" %j format size modifier is now available with pre\-C99 compilers"
189 .el .SS "The \f(CWsprintf\fP \f(CW%j\fP format size modifier is now available with pre\-C99 compilers"
190 .IX Subsection "The sprintf %j format size modifier is now available with pre-C99 compilers"
191 The actual size used depends on the platform, so remains unportable.
192 .SS "Close-on-exec flag set atomically"
193 .IX Subsection "Close-on-exec flag set atomically"
194 When opening a file descriptor, perl now generally opens it with its
195 close-on-exec flag already set, on platforms that support doing so.
196 This improves thread safety, because it means that an \f(CW\*(C`exec\*(C'\fR initiated
197 by one thread can no longer cause a file descriptor in the process
198 of being opened by another thread to be accidentally passed to the
199 executed program.
200 .PP
201 Additionally, perl now sets the close-on-exec flag more reliably, whether
202 it does so atomically or not. Most file descriptors were getting the
203 flag set, but some were being missed.
204 .SS "String\- and number-specific bitwise ops are no longer experimental"
205 .IX Subsection "String- and number-specific bitwise ops are no longer experimental"
206 The new string-specific (\f(CW\*(C`&. |. ^. ~.\*(C'\fR) and number-specific (\f(CW\*(C`& | ^ ~\*(C'\fR)
207 bitwise operators introduced in Perl 5.22 that are available within the
208 scope of \f(CW\*(C`use feature \*(Aqbitwise\*(Aq\*(C'\fR are no longer experimental.
209 Because the number-specific ops are spelled the same way as the existing
210 operators that choose their behaviour based on their operands, these
211 operators must still be enabled via the "bitwise" feature, in either of
212 these two ways:
213 .PP
214 .Vb 1
215 \& use feature "bitwise";
216 \&
217 \& use v5.28; # "bitwise" now included
218 .Ve
219 .PP
220 They are also now enabled by the \fB\-E\fR command-line switch.
221 .PP
222 The "bitwise" feature no longer emits a warning. Existing code that
223 disables the "experimental::bitwise" warning category that the feature
224 previously used will continue to work.
225 .PP
226 One caveat that module authors ought to be aware of is that the numeric
227 operators now pass a fifth TRUE argument to overload methods. Any methods
228 that check the number of operands may croak if they do not expect so many.
229 XS authors in particular should be aware that this:
230 .PP
231 .Vb 2
232 \& SV *
233 \& bitop_handler (lobj, robj, swap)
234 .Ve
235 .PP
236 may need to be changed to this:
237 .PP
238 .Vb 2
239 \& SV *
240 \& bitop_handler (lobj, robj, swap, ...)
241 .Ve
242 .SS "Locales are now thread-safe on systems that support them"
243 .IX Subsection "Locales are now thread-safe on systems that support them"
244 These systems include Windows starting with Visual Studio 2005, and in
245 POSIX 2008 systems.
246 .PP
247 The implication is that you are now free to use locales and change them
248 in a threaded environment. Your changes affect only your thread.
249 See "Multi-threaded operation" in perllocale
250 .ie n .SS "New read-only predefined variable ""${^SAFE_LOCALES}"""
251 .el .SS "New read-only predefined variable \f(CW${^SAFE_LOCALES}\fP"
252 .IX Subsection "New read-only predefined variable ${^SAFE_LOCALES}"
253 This variable is 1 if the Perl interpreter is operating in an
254 environment where it is safe to use and change locales (see
255 perllocale.) This variable is true when the perl is
256 unthreaded, or compiled in a platform that supports thread-safe locale
257 operation (see previous item).
258 .SH Security
259 .IX Header "Security"
260 .SS "[CVE\-2017\-12837] Heap buffer overflow in regular expression compiler"
261 .IX Subsection "[CVE-2017-12837] Heap buffer overflow in regular expression compiler"
262 Compiling certain regular expression patterns with the case-insensitive
263 modifier could cause a heap buffer overflow and crash perl. This has now been
264 fixed.
265 [GH #16021] <https://github.com/Perl/perl5/issues/16021>
266 .SS "[CVE\-2017\-12883] Buffer over-read in regular expression parser"
267 .IX Subsection "[CVE-2017-12883] Buffer over-read in regular expression parser"
268 For certain types of syntax error in a regular expression pattern, the error
269 message could either contain the contents of a random, possibly large, chunk of
270 memory, or could crash perl. This has now been fixed.
271 [GH #16025] <https://github.com/Perl/perl5/issues/16025>
272 .ie n .SS "[CVE\-2017\-12814] $ENV{$key} stack buffer overflow on Windows"
273 .el .SS "[CVE\-2017\-12814] \f(CW$ENV{$key}\fP stack buffer overflow on Windows"
274 .IX Subsection "[CVE-2017-12814] $ENV{$key} stack buffer overflow on Windows"
275 A possible stack buffer overflow in the \f(CW%ENV\fR code on Windows has been fixed
276 by removing the buffer completely since it was superfluous anyway.
277 [GH #16051] <https://github.com/Perl/perl5/issues/16051>
278 .SS "Default Hash Function Change"
279 .IX Subsection "Default Hash Function Change"
280 Perl 5.28.0 retires various older hash functions which are not viewed as
281 sufficiently secure for use in Perl. We now support four general purpose
282 hash functions, Siphash (2\-4 and 1\-3 variants), and Zaphod32, and StadtX
283 hash. In addition we support SBOX32 (a form of tabular hashing) for hashing
284 short strings, in conjunction with any of the other hash functions provided.
285 .PP
286 By default Perl is configured to support SBOX hashing of strings up to 24
287 characters, in conjunction with StadtX hashing on 64 bit builds, and
288 Zaphod32 hashing for 32 bit builds.
289 .PP
290 You may control these settings with the following options to Configure:
291 .PP
292 .Vb 4
293 \& \-DPERL_HASH_FUNC_SIPHASH
294 \& \-DPERL_HASH_FUNC_SIPHASH13
295 \& \-DPERL_HASH_FUNC_STADTX
296 \& \-DPERL_HASH_FUNC_ZAPHOD32
297 .Ve
298 .PP
299 To disable SBOX hashing you can use
300 .PP
301 .Vb 1
302 \& \-DPERL_HASH_USE_SBOX32_ALSO=0
303 .Ve
304 .PP
305 And to set the maximum length to use SBOX32 hashing on with:
306 .PP
307 .Vb 1
308 \& \-DSBOX32_MAX_LEN=16
309 .Ve
310 .PP
311 The maximum length allowed is 256. There probably isn't much point
312 in setting it higher than the default.
313 .SH "Incompatible Changes"
314 .IX Header "Incompatible Changes"
315 .SS "Subroutine attribute and signature order"
316 .IX Subsection "Subroutine attribute and signature order"
317 The experimental subroutine signatures feature has been changed so that
318 subroutine attributes must now come before the signature rather than
319 after. This is because attributes like \f(CW\*(C`:lvalue\*(C'\fR can affect the
320 compilation of code within the signature, for example:
321 .PP
322 .Vb 1
323 \& sub f :lvalue ($a = do { $x = "abc"; return substr($x,0,1)}) { ...}
324 .Ve
325 .PP
326 Note that this the second time they have been flipped:
327 .PP
328 .Vb 2
329 \& sub f :lvalue ($a, $b) { ... }; # 5.20; 5.28 onwards
330 \& sub f ($a, $b) :lvalue { ... }; # 5.22 \- 5.26
331 .Ve
332 .SS "Comma-less variable lists in formats are no longer allowed"
333 .IX Subsection "Comma-less variable lists in formats are no longer allowed"
334 Omitting the commas between variables passed to formats is no longer
335 allowed. This has been deprecated since Perl 5.000.
336 .ie n .SS "The "":locked"" and "":unique"" attributes have been removed"
337 .el .SS "The \f(CW:locked\fP and \f(CW:unique\fP attributes have been removed"
338 .IX Subsection "The :locked and :unique attributes have been removed"
339 These have been no-ops and deprecated since Perl 5.12 and 5.10,
340 respectively.
341 .ie n .SS """\eN{}"" with nothing between the braces is now illegal"
342 .el .SS "\f(CW\eN{}\fP with nothing between the braces is now illegal"
343 .IX Subsection "N{} with nothing between the braces is now illegal"
344 This has been deprecated since Perl 5.24.
345 .SS "Opening the same symbol as both a file and directory handle is no longer allowed"
346 .IX Subsection "Opening the same symbol as both a file and directory handle is no longer allowed"
347 Using \f(CWopen()\fR and \f(CWopendir()\fR to associate both a filehandle and a dirhandle
348 to the same symbol (glob or scalar) has been deprecated since Perl 5.10.
349 .ie n .SS "Use of bare ""<<"" to mean ""<<"""""" is no longer allowed"
350 .el .SS "Use of bare \f(CW<<\fP to mean \f(CW<<""""\fP is no longer allowed"
351 .IX Subsection "Use of bare << to mean <<"""" is no longer allowed"
352 Use of a bare terminator has been deprecated since Perl 5.000.
353 .SS "Setting $/ to a reference to a non-positive integer no longer allowed"
354 .IX Subsection "Setting $/ to a reference to a non-positive integer no longer allowed"
355 This used to work like setting it to \f(CW\*(C`undef\*(C'\fR, but has been deprecated
356 since Perl 5.20.
357 .ie n .SS "Unicode code points with values exceeding ""IV_MAX"" are now fatal"
358 .el .SS "Unicode code points with values exceeding \f(CWIV_MAX\fP are now fatal"
359 .IX Subsection "Unicode code points with values exceeding IV_MAX are now fatal"
360 This was deprecated since Perl 5.24.
361 .ie n .SS "The ""B::OP::terse"" method has been removed"
362 .el .SS "The \f(CWB::OP::terse\fP method has been removed"
363 .IX Subsection "The B::OP::terse method has been removed"
364 Use \f(CW\*(C`B::Concise::b_terse\*(C'\fR instead.
365 .SS "Use of inherited AUTOLOAD for non-methods is no longer allowed"
366 .IX Subsection "Use of inherited AUTOLOAD for non-methods is no longer allowed"
367 This was deprecated in Perl 5.004.
368 .SS "Use of strings with code points over 0xFF is not allowed for bitwise string operators"
369 .IX Subsection "Use of strings with code points over 0xFF is not allowed for bitwise string operators"
370 Code points over \f(CW0xFF\fR do not make sense for bitwise operators and such
371 an operation will now croak, except for a few remaining cases. See
372 perldeprecation.
373 .PP
374 This was deprecated in Perl 5.24.
375 .ie n .SS "Setting ""${^ENCODING}"" to a defined value is now illegal"
376 .el .SS "Setting \f(CW${^ENCODING}\fP to a defined value is now illegal"
377 .IX Subsection "Setting ${^ENCODING} to a defined value is now illegal"
378 This has been deprecated since Perl 5.22 and a no-op since Perl 5.26.
379 .ie n .SS "Backslash no longer escapes colon in PATH for the ""\-S"" switch"
380 .el .SS "Backslash no longer escapes colon in PATH for the \f(CW\-S\fP switch"
381 .IX Subsection "Backslash no longer escapes colon in PATH for the -S switch"
382 Previously the \f(CW\*(C`\-S\*(C'\fR switch incorrectly treated backslash ("\e") as an
383 escape for colon when traversing the \f(CW\*(C`PATH\*(C'\fR environment variable.
384 [GH #15584] <https://github.com/Perl/perl5/issues/15584>
385 .SS "the \-DH (DEBUG_H) misfeature has been removed"
386 .IX Subsection "the -DH (DEBUG_H) misfeature has been removed"
387 On a perl built with debugging support, the \f(CW\*(C`H\*(C'\fR flag to the \f(CW\*(C`\-D\*(C'\fR
388 debugging option has been removed. This was supposed to dump hash values,
389 but has been broken for many years.
390 .SS "Yada-yada is now strictly a statement"
391 .IX Subsection "Yada-yada is now strictly a statement"
392 By the time of its initial stable release in Perl 5.12, the \f(CW\*(C`...\*(C'\fR
393 (yada-yada) operator was explicitly intended to serve as a statement,
394 not an expression. However, the original implementation was confused
395 on this point, leading to inconsistent parsing. The operator was
396 accidentally accepted in a few situations where it did not serve as a
397 complete statement, such as
398 .PP
399 .Vb 2
400 \& ... . "foo";
401 \& ... if $a < $b;
402 .Ve
403 .PP
404 The parsing has now been made consistent, permitting yada-yada only as
405 a statement. Affected code can use \f(CW\*(C`do{...}\*(C'\fR to put a yada-yada into
406 an arbitrary expression context.
407 .SS "Sort algorithm can no longer be specified"
408 .IX Subsection "Sort algorithm can no longer be specified"
409 Since Perl 5.8, the sort pragma has had subpragmata \f(CW\*(C`_mergesort\*(C'\fR,
410 \&\f(CW\*(C`_quicksort\*(C'\fR, and \f(CW\*(C`_qsort\*(C'\fR that can be used to specify which algorithm
411 perl should use to implement the sort builtin.
412 This was always considered a dubious feature that might not last,
413 hence the underscore spellings, and they were documented as not being
414 portable beyond Perl 5.8. These subpragmata have now been deleted,
415 and any attempt to use them is an error. The sort pragma otherwise
416 remains, and the algorithm-neutral \f(CW\*(C`stable\*(C'\fR subpragma can be used to
417 control sorting behaviour.
418 [GH #13234] <https://github.com/Perl/perl5/issues/13234>
419 .SS "Over-radix digits in floating point literals"
420 .IX Subsection "Over-radix digits in floating point literals"
421 Octal and binary floating point literals used to permit any hexadecimal
422 digit to appear after the radix point. The digits are now restricted
423 to those appropriate for the radix, as digits before the radix point
424 always were.
425 .ie n .SS "Return type of unpackstring()"
426 .el .SS "Return type of \f(CWunpackstring()\fP"
427 .IX Subsection "Return type of unpackstring()"
428 The return types of the C API functions \f(CWunpackstring()\fR and
429 \&\f(CWunpack_str()\fR have changed from \f(CW\*(C`I32\*(C'\fR to \f(CW\*(C`SSize_t\*(C'\fR, in order to
430 accommodate datasets of more than two billion items.
431 .SH Deprecations
432 .IX Header "Deprecations"
433 .ie n .SS "Use of ""vec"" on strings with code points above 0xFF is deprecated"
434 .el .SS "Use of \f(CWvec\fP on strings with code points above 0xFF is deprecated"
435 .IX Subsection "Use of vec on strings with code points above 0xFF is deprecated"
436 Such strings are represented internally in UTF\-8, and \f(CW\*(C`vec\*(C'\fR is a
437 bit-oriented operation that will likely give unexpected results on those
438 strings.
439 .ie n .SS "Some uses of unescaped ""{"" in regexes are no longer fatal"
440 .el .SS "Some uses of unescaped \f(CW""{""\fP in regexes are no longer fatal"
441 .IX Subsection "Some uses of unescaped ""{"" in regexes are no longer fatal"
442 Perl 5.26.0 fatalized some uses of an unescaped left brace, but an
443 exception was made at the last minute, specifically crafted to be a
444 minimal change to allow GNU Autoconf to work. That tool is heavily
445 depended upon, and continues to use the deprecated usage. Its use of an
446 unescaped left brace is one where we have no intention of repurposing
447 \&\f(CW"{"\fR to be something other than itself.
448 .PP
449 That exception is now generalized to include various other such cases
450 where the \f(CW"{"\fR will not be repurposed.
451 .PP
452 Note that these uses continue to raise a deprecation message.
453 .ie n .SS "Use of unescaped ""{"" immediately after a ""("" in regular expression patterns is deprecated"
454 .el .SS "Use of unescaped \f(CW""{""\fP immediately after a \f(CW""(""\fP in regular expression patterns is deprecated"
455 .IX Subsection "Use of unescaped ""{"" immediately after a ""("" in regular expression patterns is deprecated"
456 Using unescaped left braces is officially deprecated everywhere, but it
457 is not enforced in contexts where their use does not interfere with
458 expected extensions to the language. A deprecation is added in this
459 release when the brace appears immediately after an opening parenthesis.
460 Before this, even if the brace was part of a legal quantifier, it was
461 not interpreted as such, but as the literal characters, unlike other
462 quantifiers that follow a \f(CW"("\fR which are considered errors. Now,
463 their use will raise a deprecation message, unless turned off.
464 .ie n .SS "Assignment to $[ will be fatal in Perl 5.30"
465 .el .SS "Assignment to \f(CW$[\fP will be fatal in Perl 5.30"
466 .IX Subsection "Assignment to $[ will be fatal in Perl 5.30"
467 Assigning a non-zero value to \f(CW$[\fR has been deprecated
468 since Perl 5.12, but was never given a deadline for removal. This has
469 now been scheduled for Perl 5.30.
470 .SS "\fBhostname()\fP won't accept arguments in Perl 5.32"
471 .IX Subsection "hostname() won't accept arguments in Perl 5.32"
472 Passing arguments to \f(CWSys::Hostname::hostname()\fR was already deprecated,
473 but didn't have a removal date. This has now been scheduled for Perl
474 5.32. [GH #14662] <https://github.com/Perl/perl5/issues/14662>
475 .SS "Module removals"
476 .IX Subsection "Module removals"
477 The following modules will be removed from the core distribution in a
478 future release, and will at that time need to be installed from CPAN.
479 Distributions on CPAN which require these modules will need to list them as
480 prerequisites.
481 .PP
482 The core versions of these modules will now issue \f(CW"deprecated"\fR\-category
483 warnings to alert you to this fact. To silence these deprecation warnings,
484 install the modules in question from CPAN.
485 .PP
486 Note that these are (with rare exceptions) fine modules that you are encouraged
487 to continue to use. Their disinclusion from core primarily hinges on their
488 necessity to bootstrapping a fully functional, CPAN-capable Perl installation,
489 not usually on concerns over their design.
490 .IP B::Debug 4
491 .IX Item "B::Debug"
492 .PD 0
493 .IP "Locale::Codes and its associated Country, Currency and Language modules" 4
494 .IX Item "Locale::Codes and its associated Country, Currency and Language modules"
495 .PD
496 .SH "Performance Enhancements"
497 .IX Header "Performance Enhancements"
498 .IP \(bu 4
499 The start up overhead for creating regular expression patterns with
500 Unicode properties (\f(CW\*(C`\ep{...}\*(C'\fR) has been greatly reduced in most cases.
501 .IP \(bu 4
502 Many string concatenation expressions are now considerably faster, due
503 to the introduction internally of a \f(CW\*(C`multiconcat\*(C'\fR opcode which combines
504 multiple concatenations, and optionally a \f(CW\*(C`=\*(C'\fR or \f(CW\*(C`.=\*(C'\fR, into a single
505 action. For example, apart from retrieving \f(CW$s\fR, \f(CW$a\fR and \f(CW$b\fR, this
506 whole expression is now handled as a single op:
507 .Sp
508 .Vb 1
509 \& $s .= "a=$a b=$b\en"
510 .Ve
511 .Sp
512 As a special case, if the LHS of an assignment is a lexical variable or
513 \&\f(CW\*(C`my $s\*(C'\fR, the op itself handles retrieving the lexical variable, which
514 is faster.
515 .Sp
516 In general, the more the expression includes a mix of constant strings and
517 variable expressions, the longer the expression, and the more it mixes
518 together non\-utf8 and utf8 strings, the more marked the performance
519 improvement. For example on a \f(CW\*(C`x86_64\*(C'\fR system, this code has been
520 benchmarked running four times faster:
521 .Sp
522 .Vb 4
523 \& my $s;
524 \& my $a = "ab\ex{100}cde";
525 \& my $b = "fghij";
526 \& my $c = "\ex{101}klmn";
527 \&
528 \& for my $i (1..10_000_000) {
529 \& $s = "\ex{100}wxyz";
530 \& $s .= "foo=$a bar=$b baz=$c";
531 \& }
532 .Ve
533 .Sp
534 In addition, \f(CW\*(C`sprintf\*(C'\fR expressions which have a constant format
535 containing only \f(CW%s\fR and \f(CW\*(C`%%\*(C'\fR format elements, and which have a fixed
536 number of arguments, are now also optimised into a \f(CW\*(C`multiconcat\*(C'\fR op.
537 .IP \(bu 4
538 The \f(CWref()\fR builtin is now much faster in boolean context, since it no
539 longer bothers to construct a temporary string like \f(CW\*(C`Foo=ARRAY(0x134af48)\*(C'\fR.
540 .IP \(bu 4
541 \&\f(CWkeys()\fR in void and scalar contexts is now more efficient.
542 .IP \(bu 4
543 The common idiom of comparing the result of \fBindex()\fR with \-1 is now
544 specifically optimised, e.g.
545 .Sp
546 .Vb 1
547 \& if (index(...) != \-1) { ... }
548 .Ve
549 .IP \(bu 4
550 \&\f(CWfor()\fR loops and similar constructs are now more efficient in most cases.
551 .IP \(bu 4
552 File::Glob has been modified to remove unnecessary backtracking and
553 recursion, thanks to Russ Cox. See <https://research.swtch.com/glob>
554 for more details.
555 .IP \(bu 4
556 The XS-level \f(CWSvTRUE()\fR API function is now more efficient.
557 .IP \(bu 4
558 Various integer-returning ops are now more efficient in scalar/boolean context.
559 .IP \(bu 4
560 Slightly improved performance when parsing stash names.
561 [GH #15689] <https://github.com/Perl/perl5/issues/15689>
562 .IP \(bu 4
563 Calls to \f(CW\*(C`require\*(C'\fR for an already loaded module are now slightly faster.
564 [GH #16175] <https://github.com/Perl/perl5/issues/16175>
565 .IP \(bu 4
566 The performance of pattern matching \f(CW\*(C`[[:ascii:]]\*(C'\fR and \f(CW\*(C`[[:^ascii:]]\*(C'\fR
567 has been improved significantly except on EBCDIC platforms.
568 .IP \(bu 4
569 Various optimizations have been applied to matching regular expression
570 patterns, so under the right circumstances, significant performance
571 gains may be noticed. But in an application with many varied patterns,
572 little overall improvement likely will be seen.
573 .IP \(bu 4
574 Other optimizations have been applied to UTF\-8 handling, but these are
575 not typically a major factor in most applications.
576 .SH "Modules and Pragmata"
577 .IX Header "Modules and Pragmata"
578 Key highlights in this release across several modules:
579 .SS "Removal of use vars"
580 .IX Subsection "Removal of use vars"
581 The usage of \f(CW\*(C`use vars\*(C'\fR has been discouraged since the introduction of
582 \&\f(CW\*(C`our\*(C'\fR in Perl 5.6.0. Where possible the usage of this pragma has now been
583 removed from the Perl source code.
584 .PP
585 This had a slight effect (for the better) on the output of WARNING_BITS in
586 B::Deparse.
587 .SS "Use of DynaLoader changed to XSLoader in many modules"
588 .IX Subsection "Use of DynaLoader changed to XSLoader in many modules"
589 XSLoader is more modern, and most modules already require perl 5.6 or
590 greater, so no functionality is lost by switching. In some cases, we have
591 also made changes to the local implementation that may not be reflected in
592 the version on CPAN due to a desire to maintain more backwards
593 compatibility.
594 .SS "Updated Modules and Pragmata"
595 .IX Subsection "Updated Modules and Pragmata"
596 .IP \(bu 4
597 Archive::Tar has been upgraded from version 2.24 to 2.30.
598 .Sp
599 This update also handled CVE\-2018\-12015: directory traversal
600 vulnerability.
601 [cpan #125523] <https://rt.cpan.org/Ticket/Display.html?id=125523>
602 .IP \(bu 4
603 arybase has been upgraded from version 0.12 to 0.15.
604 .IP \(bu 4
605 Attribute::Handlers has been upgraded from version 0.99 to 1.01.
606 .IP \(bu 4
607 attributes has been upgraded from version 0.29 to 0.33.
608 .IP \(bu 4
609 B has been upgraded from version 1.68 to 1.74.
610 .IP \(bu 4
611 B::Concise has been upgraded from version 0.999 to 1.003.
612 .IP \(bu 4
613 B::Debug has been upgraded from version 1.24 to 1.26.
614 .Sp
615 NOTE: B::Debug is deprecated and may be removed from a future version
616 of Perl.
617 .IP \(bu 4
618 B::Deparse has been upgraded from version 1.40 to 1.48.
619 .Sp
620 It includes many bug fixes, and in particular, it now deparses variable
621 attributes correctly:
622 .Sp
623 .Vb 2
624 \& my $x :foo; # used to deparse as
625 \& # \*(Aqattributes\*(Aq\->import(\*(Aqmain\*(Aq, \e$x, \*(Aqfoo\*(Aq), my $x;
626 .Ve
627 .IP \(bu 4
628 base has been upgraded from version 2.25 to 2.27.
629 .IP \(bu 4
630 bignum has been upgraded from version 0.47 to 0.49.
631 .IP \(bu 4
632 blib has been upgraded from version 1.06 to 1.07.
633 .IP \(bu 4
634 bytes has been upgraded from version 1.05 to 1.06.
635 .IP \(bu 4
636 Carp has been upgraded from version 1.42 to 1.50.
637 .Sp
638 If a package on the call stack contains a constant named \f(CW\*(C`ISA\*(C'\fR, Carp no
639 longer throws a "Not a GLOB reference" error.
640 .Sp
641 Carp, when generating stack traces, now attempts to work around
642 longstanding bugs resulting from Perl's non-reference-counted stack.
643 [GH #9282] <https://github.com/Perl/perl5/issues/9282>
644 .Sp
645 Carp has been modified to avoid assuming that objects cannot be
646 overloaded without the overload module loaded (this can happen with
647 objects created by XS modules). Previously, infinite recursion would
648 result if an XS-defined overload method itself called Carp.
649 [GH #16407] <https://github.com/Perl/perl5/issues/16407>
650 .Sp
651 Carp now avoids using \f(CW\*(C`overload::StrVal\*(C'\fR, partly because older versions
652 of overload (included with perl 5.14 and earlier) load Scalar::Util
653 at run time, which will fail if Carp has been invoked after a syntax error.
654 .IP \(bu 4
655 charnames has been upgraded from version 1.44 to 1.45.
656 .IP \(bu 4
657 Compress::Raw::Zlib has been upgraded from version 2.074 to 2.076.
658 .Sp
659 This addresses a security vulnerability in older versions of the 'zlib' library
660 (which is bundled with Compress-Raw-Zlib).
661 .IP \(bu 4
662 Config::Extensions has been upgraded from version 0.01 to 0.02.
663 .IP \(bu 4
664 Config::Perl::V has been upgraded from version 0.28 to 0.29.
665 .IP \(bu 4
666 CPAN has been upgraded from version 2.18 to 2.20.
667 .IP \(bu 4
668 Data::Dumper has been upgraded from version 2.167 to 2.170.
669 .Sp
670 Quoting of glob names now obeys the Useqq option
671 [GH #13274] <https://github.com/Perl/perl5/issues/13274>.
672 .Sp
673 Attempts to set an option to \f(CW\*(C`undef\*(C'\fR through a combined getter/setter
674 method are no longer mistaken for getter calls
675 [GH #12135] <https://github.com/Perl/perl5/issues/12135>.
676 .IP \(bu 4
677 Devel::Peek has been upgraded from version 1.26 to 1.27.
678 .IP \(bu 4
679 Devel::PPPort has been upgraded from version 3.35 to 3.40.
680 .Sp
681 Devel::PPPort has moved from cpan-first to perl-first maintenance
682 .Sp
683 Primary responsibility for the code in Devel::PPPort has moved into core perl.
684 In a practical sense there should be no change except that hopefully it will
685 stay more up to date with changes made to symbols in perl, rather than needing
686 to be updated after the fact.
687 .IP \(bu 4
688 Digest::SHA has been upgraded from version 5.96 to 6.01.
689 .IP \(bu 4
690 DirHandle has been upgraded from version 1.04 to 1.05.
691 .IP \(bu 4
692 DynaLoader has been upgraded from version 1.42 to 1.45.
693 .Sp
694 Its documentation now shows the use of \f(CW\*(C`_\|_PACKAGE_\|_\*(C'\fR and direct object
695 syntax
696 [GH #16190] <https://github.com/Perl/perl5/issues/16190>.
697 .IP \(bu 4
698 Encode has been upgraded from version 2.88 to 2.97.
699 .IP \(bu 4
700 encoding has been upgraded from version 2.19 to 2.22.
701 .IP \(bu 4
702 Errno has been upgraded from version 1.28 to 1.29.
703 .IP \(bu 4
704 experimental has been upgraded from version 0.016 to 0.019.
705 .IP \(bu 4
706 Exporter has been upgraded from version 5.72 to 5.73.
707 .IP \(bu 4
708 ExtUtils::CBuilder has been upgraded from version 0.280225 to 0.280230.
709 .IP \(bu 4
710 ExtUtils::Constant has been upgraded from version 0.23 to 0.25.
711 .IP \(bu 4
712 ExtUtils::Embed has been upgraded from version 1.34 to 1.35.
713 .IP \(bu 4
714 ExtUtils::Install has been upgraded from version 2.04 to 2.14.
715 .IP \(bu 4
716 ExtUtils::MakeMaker has been upgraded from version 7.24 to 7.34.
717 .IP \(bu 4
718 ExtUtils::Miniperl has been upgraded from version 1.06 to 1.08.
719 .IP \(bu 4
720 ExtUtils::ParseXS has been upgraded from version 3.34 to 3.39.
721 .IP \(bu 4
722 ExtUtils::Typemaps has been upgraded from version 3.34 to 3.38.
723 .IP \(bu 4
724 ExtUtils::XSSymSet has been upgraded from version 1.3 to 1.4.
725 .IP \(bu 4
726 feature has been upgraded from version 1.47 to 1.52.
727 .IP \(bu 4
728 fields has been upgraded from version 2.23 to 2.24.
729 .IP \(bu 4
730 File::Copy has been upgraded from version 2.32 to 2.33.
731 .Sp
732 It will now use the sub-second precision variant of \fButime()\fR supplied by
733 Time::HiRes where available.
734 [GH #16225] <https://github.com/Perl/perl5/issues/16225>.
735 .IP \(bu 4
736 File::Fetch has been upgraded from version 0.52 to 0.56.
737 .IP \(bu 4
738 File::Glob has been upgraded from version 1.28 to 1.31.
739 .IP \(bu 4
740 File::Path has been upgraded from version 2.12_01 to 2.15.
741 .IP \(bu 4
742 File::Spec and Cwd have been upgraded from version 3.67 to 3.74.
743 .IP \(bu 4
744 File::stat has been upgraded from version 1.07 to 1.08.
745 .IP \(bu 4
746 FileCache has been upgraded from version 1.09 to 1.10.
747 .IP \(bu 4
748 Filter::Simple has been upgraded from version 0.93 to 0.95.
749 .IP \(bu 4
750 Filter::Util::Call has been upgraded from version 1.55 to 1.58.
751 .IP \(bu 4
752 GDBM_File has been upgraded from version 1.15 to 1.17.
753 .Sp
754 Its documentation now explains that \f(CW\*(C`each\*(C'\fR and \f(CW\*(C`delete\*(C'\fR don't mix in
755 hashes tied to this module
756 [GH #12894] <https://github.com/Perl/perl5/issues/12894>.
757 .Sp
758 It will now retry opening with an acceptable block size if asking gdbm
759 to default the block size failed
760 [GH #13232] <https://github.com/Perl/perl5/issues/13232>.
761 .IP \(bu 4
762 Getopt::Long has been upgraded from version 2.49 to 2.5.
763 .IP \(bu 4
764 Hash::Util::FieldHash has been upgraded from version 1.19 to 1.20.
765 .IP \(bu 4
766 I18N::Langinfo has been upgraded from version 0.13 to 0.17.
767 .Sp
768 This module is now available on all platforms, emulating the system
769 \&\fBnl_langinfo\fR\|(3) on systems that lack it. Some caveats apply, as
770 detailed in its documentation, the most severe being
771 that, except for MS Windows, the \f(CW\*(C`CODESET\*(C'\fR item is not implemented on
772 those systems, always returning \f(CW""\fR.
773 .Sp
774 It now sets the UTF\-8 flag in its returned scalar if the string contains
775 legal non-ASCII UTF\-8, and the locale is UTF\-8
776 [GH #15131] <https://github.com/Perl/perl5/issues/15131>.
777 .Sp
778 This update also fixes a bug in which the underlying locale was ignored
779 for the \f(CW\*(C`RADIXCHAR\*(C'\fR (always was returned as a dot) and the \f(CW\*(C`THOUSEP\*(C'\fR
780 (always empty). Now the locale-appropriate values are returned.
781 .IP \(bu 4
782 I18N::LangTags has been upgraded from version 0.42 to 0.43.
783 .IP \(bu 4
784 if has been upgraded from version 0.0606 to 0.0608.
785 .IP \(bu 4
786 IO has been upgraded from version 1.38 to 1.39.
787 .IP \(bu 4
788 IO::Socket::IP has been upgraded from version 0.38 to 0.39.
789 .IP \(bu 4
790 IPC::Cmd has been upgraded from version 0.96 to 1.00.
791 .IP \(bu 4
792 JSON::PP has been upgraded from version 2.27400_02 to 2.97001.
793 .IP \(bu 4
794 The \f(CW\*(C`libnet\*(C'\fR distribution has been upgraded from version 3.10 to 3.11.
795 .IP \(bu 4
796 List::Util has been upgraded from version 1.46_02 to 1.49.
797 .IP \(bu 4
798 Locale::Codes has been upgraded from version 3.42 to 3.56.
799 .Sp
800 \&\fBNOTE\fR: Locale::Codes scheduled to be removed from core in Perl 5.30.
801 .IP \(bu 4
802 Locale::Maketext has been upgraded from version 1.28 to 1.29.
803 .IP \(bu 4
804 Math::BigInt has been upgraded from version 1.999806 to 1.999811.
805 .IP \(bu 4
806 Math::BigInt::FastCalc has been upgraded from version 0.5005 to 0.5006.
807 .IP \(bu 4
808 Math::BigRat has been upgraded from version 0.2611 to 0.2613.
809 .IP \(bu 4
810 Module::CoreList has been upgraded from version 5.20170530 to 5.20180622.
811 .IP \(bu 4
812 mro has been upgraded from version 1.20 to 1.22.
813 .IP \(bu 4
814 Net::Ping has been upgraded from version 2.55 to 2.62.
815 .IP \(bu 4
816 NEXT has been upgraded from version 0.67 to 0.67_01.
817 .IP \(bu 4
818 ODBM_File has been upgraded from version 1.14 to 1.15.
819 .IP \(bu 4
820 Opcode has been upgraded from version 1.39 to 1.43.
821 .IP \(bu 4
822 overload has been upgraded from version 1.28 to 1.30.
823 .IP \(bu 4
824 PerlIO::encoding has been upgraded from version 0.25 to 0.26.
825 .IP \(bu 4
826 PerlIO::scalar has been upgraded from version 0.26 to 0.29.
827 .IP \(bu 4
828 PerlIO::via has been upgraded from version 0.16 to 0.17.
829 .IP \(bu 4
830 Pod::Functions has been upgraded from version 1.11 to 1.13.
831 .IP \(bu 4
832 Pod::Html has been upgraded from version 1.2202 to 1.24.
833 .Sp
834 A title for the HTML document will now be automatically generated by
835 default from a "NAME" section in the POD document, as it used to be
836 before the module was rewritten to use Pod::Simple::XHTML to do the
837 core of its job
838 [GH #11954] <https://github.com/Perl/perl5/issues/11954>.
839 .IP \(bu 4
840 Pod::Perldoc has been upgraded from version 3.28 to 3.2801.
841 .IP \(bu 4
842 The \f(CW\*(C`podlators\*(C'\fR distribution has been upgraded from version 4.09 to 4.10.
843 .Sp
844 Man page references and function names now follow the Linux man page
845 formatting standards, instead of the Solaris standard.
846 .IP \(bu 4
847 POSIX has been upgraded from version 1.76 to 1.84.
848 .Sp
849 Some more cautions were added about using locale-specific functions in
850 threaded applications.
851 .IP \(bu 4
852 re has been upgraded from version 0.34 to 0.36.
853 .IP \(bu 4
854 Scalar::Util has been upgraded from version 1.46_02 to 1.50.
855 .IP \(bu 4
856 SelfLoader has been upgraded from version 1.23 to 1.25.
857 .IP \(bu 4
858 Socket has been upgraded from version 2.020_03 to 2.027.
859 .IP \(bu 4
860 sort has been upgraded from version 2.02 to 2.04.
861 .IP \(bu 4
862 Storable has been upgraded from version 2.62 to 3.08.
863 .IP \(bu 4
864 Sub::Util has been upgraded from version 1.48 to 1.49.
865 .IP \(bu 4
866 subs has been upgraded from version 1.02 to 1.03.
867 .IP \(bu 4
868 Sys::Hostname has been upgraded from version 1.20 to 1.22.
869 .IP \(bu 4
870 Term::ReadLine has been upgraded from version 1.16 to 1.17.
871 .IP \(bu 4
872 Test has been upgraded from version 1.30 to 1.31.
873 .IP \(bu 4
874 Test::Harness has been upgraded from version 3.38 to 3.42.
875 .IP \(bu 4
876 Test::Simple has been upgraded from version 1.302073 to 1.302133.
877 .IP \(bu 4
878 threads has been upgraded from version 2.15 to 2.22.
879 .Sp
880 The documentation now better describes the problems that arise when
881 returning values from threads, and no longer warns about creating threads
882 in \f(CW\*(C`BEGIN\*(C'\fR blocks.
883 [GH #11563] <https://github.com/Perl/perl5/issues/11563>
884 .IP \(bu 4
885 threads::shared has been upgraded from version 1.56 to 1.58.
886 .IP \(bu 4
887 Tie::Array has been upgraded from version 1.06 to 1.07.
888 .IP \(bu 4
889 Tie::StdHandle has been upgraded from version 4.4 to 4.5.
890 .IP \(bu 4
891 Time::gmtime has been upgraded from version 1.03 to 1.04.
892 .IP \(bu 4
893 Time::HiRes has been upgraded from version 1.9741 to 1.9759.
894 .IP \(bu 4
895 Time::localtime has been upgraded from version 1.02 to 1.03.
896 .IP \(bu 4
897 Time::Piece has been upgraded from version 1.31 to 1.3204.
898 .IP \(bu 4
899 Unicode::Collate has been upgraded from version 1.19 to 1.25.
900 .IP \(bu 4
901 Unicode::Normalize has been upgraded from version 1.25 to 1.26.
902 .IP \(bu 4
903 Unicode::UCD has been upgraded from version 0.68 to 0.70.
904 .Sp
905 The function \f(CW\*(C`num\*(C'\fR now accepts an optional parameter to help in
906 diagnosing error returns.
907 .IP \(bu 4
908 User::grent has been upgraded from version 1.01 to 1.02.
909 .IP \(bu 4
910 User::pwent has been upgraded from version 1.00 to 1.01.
911 .IP \(bu 4
912 utf8 has been upgraded from version 1.19 to 1.21.
913 .IP \(bu 4
914 vars has been upgraded from version 1.03 to 1.04.
915 .IP \(bu 4
916 version has been upgraded from version 0.9917 to 0.9923.
917 .IP \(bu 4
918 VMS::DCLsym has been upgraded from version 1.08 to 1.09.
919 .IP \(bu 4
920 VMS::Stdio has been upgraded from version 2.41 to 2.44.
921 .IP \(bu 4
922 warnings has been upgraded from version 1.37 to 1.42.
923 .Sp
924 It now includes new functions with names ending in \f(CW\*(C`_at_level\*(C'\fR, allowing
925 callers to specify the exact call frame.
926 [GH #16257] <https://github.com/Perl/perl5/issues/16257>
927 .IP \(bu 4
928 XS::Typemap has been upgraded from version 0.15 to 0.16.
929 .IP \(bu 4
930 XSLoader has been upgraded from version 0.27 to 0.30.
931 .Sp
932 Its documentation now shows the use of \f(CW\*(C`_\|_PACKAGE_\|_\*(C'\fR, and direct object
933 syntax for example \f(CW\*(C`DynaLoader\*(C'\fR usage
934 [GH #16190] <https://github.com/Perl/perl5/issues/16190>.
935 .Sp
936 Platforms that use \f(CW\*(C`mod2fname\*(C'\fR to edit the names of loadable
937 libraries now look for bootstrap (.bs) files under the correct,
938 non-edited name.
939 .SS "Removed Modules and Pragmata"
940 .IX Subsection "Removed Modules and Pragmata"
941 .IP \(bu 4
942 The \f(CW\*(C`VMS::stdio\*(C'\fR compatibility shim has been removed.
943 .SH Documentation
944 .IX Header "Documentation"
945 .SS "Changes to Existing Documentation"
946 .IX Subsection "Changes to Existing Documentation"
947 We have attempted to update the documentation to reflect the changes
948 listed in this document. If you find any we have missed, send email
949 to [email protected] <mailto:[email protected]>.
950 .PP
951 Additionally, the following selected changes have been made:
952 .PP
953 \fIperlapi\fR
954 .IX Subsection "perlapi"
955 .IP \(bu 4
956 The API functions \f(CWperl_parse()\fR, \f(CWperl_run()\fR, and \f(CWperl_destruct()\fR
957 are now documented comprehensively, where previously the only
958 documentation was a reference to the perlembed tutorial.
959 .IP \(bu 4
960 The documentation of \f(CWnewGIVENOP()\fR has been belatedly updated to
961 account for the removal of lexical \f(CW$_\fR.
962 .IP \(bu 4
963 The API functions \f(CWnewCONSTSUB()\fR and \f(CWnewCONSTSUB_flags()\fR are
964 documented much more comprehensively than before.
965 .PP
966 \fIperldata\fR
967 .IX Subsection "perldata"
968 .IP \(bu 4
969 The section "Truth and Falsehood" in perlsyn has been moved into
970 perldata.
971 .PP
972 \fIperldebguts\fR
973 .IX Subsection "perldebguts"
974 .IP \(bu 4
975 The description of the conditions under which \f(CWDB::sub()\fR will be called
976 has been clarified.
977 [GH #16055] <https://github.com/Perl/perl5/issues/16055>
978 .PP
979 \fIperldiag\fR
980 .IX Subsection "perldiag"
981 .IP \(bu 4
982 "Variable length lookbehind not implemented in regex m/%s/" in perldiag
983 .Sp
984 This now gives more ideas as to workarounds to the issue that was
985 introduced in Perl 5.18 (but not documented explicitly in its perldelta)
986 for the fact that some Unicode \f(CW\*(C`/i\*(C'\fR rules cause a few sequences such as
987 .Sp
988 .Vb 1
989 \& (?<!st)
990 .Ve
991 .Sp
992 to be considered variable length, and hence disallowed.
993 .IP \(bu 4
994 "Use of state \f(CW$_\fR is experimental" in perldiag
995 .Sp
996 This entry has been removed, as the experimental support of this construct was
997 removed in perl 5.24.0.
998 .IP \(bu 4
999 The diagnostic \f(CW\*(C`Initialization of state variables in list context
1000 currently forbidden\*(C'\fR has changed to \f(CW\*(C`Initialization of state variables
1001 in list currently forbidden\*(C'\fR, because list-context initialization of
1002 single aggregate state variables is now permitted.
1003 .PP
1004 \fIperlembed\fR
1005 .IX Subsection "perlembed"
1006 .IP \(bu 4
1007 The examples in perlembed have been made more portable in the way
1008 they exit, and the example that gets an exit code from the embedded Perl
1009 interpreter now gets it from the right place. The examples that pass
1010 a constructed argv to Perl now show the mandatory null \f(CW\*(C`argv[argc]\*(C'\fR.
1011 .IP \(bu 4
1012 An example in perlembed used the string value of \f(CW\*(C`ERRSV\*(C'\fR as a
1013 format string when calling \fBcroak()\fR. If that string contains format
1014 codes such as \f(CW%s\fR this could crash the program.
1015 .Sp
1016 This has been changed to a call to \fBcroak_sv()\fR.
1017 .Sp
1018 An alternative could have been to supply a trivial format string:
1019 .Sp
1020 .Vb 1
1021 \& croak("%s", SvPV_nolen(ERRSV));
1022 .Ve
1023 .Sp
1024 or as a special case for \f(CW\*(C`ERRSV\*(C'\fR simply:
1025 .Sp
1026 .Vb 1
1027 \& croak(NULL);
1028 .Ve
1029 .PP
1030 \fIperlfunc\fR
1031 .IX Subsection "perlfunc"
1032 .IP \(bu 4
1033 There is now a note that warnings generated by built-in functions are
1034 documented in perldiag and warnings.
1035 [GH #12642] <https://github.com/Perl/perl5/issues/12642>
1036 .IP \(bu 4
1037 The documentation for the \f(CW\*(C`exists\*(C'\fR operator no longer says that
1038 autovivification behaviour "may be fixed in a future release".
1039 We've determined that we're not going to change the default behaviour.
1040 [GH #15231] <https://github.com/Perl/perl5/issues/15231>
1041 .IP \(bu 4
1042 A couple of small details in the documentation for the \f(CW\*(C`bless\*(C'\fR operator
1043 have been clarified.
1044 [GH #14684] <https://github.com/Perl/perl5/issues/14684>
1045 .IP \(bu 4
1046 The description of \f(CW@INC\fR hooks in the documentation for \f(CW\*(C`require\*(C'\fR
1047 has been corrected to say that filter subroutines receive a useless
1048 first argument.
1049 [GH #12569] <https://github.com/Perl/perl5/issues/12569>
1050 .IP \(bu 4
1051 The documentation of \f(CW\*(C`ref\*(C'\fR has been rewritten for clarity.
1052 .IP \(bu 4
1053 The documentation of \f(CW\*(C`use\*(C'\fR now explains what syntactically qualifies
1054 as a version number for its module version checking feature.
1055 .IP \(bu 4
1056 The documentation of \f(CW\*(C`warn\*(C'\fR has been updated to reflect that since Perl
1057 5.14 it has treated complex exception objects in a manner equivalent
1058 to \f(CW\*(C`die\*(C'\fR.
1059 [GH #13641] <https://github.com/Perl/perl5/issues/13641>
1060 .IP \(bu 4
1061 The documentation of \f(CW\*(C`die\*(C'\fR and \f(CW\*(C`warn\*(C'\fR has been revised for clarity.
1062 .IP \(bu 4
1063 The documentation of \f(CW\*(C`each\*(C'\fR has been improved, with a slightly more
1064 explicit description of the sharing of iterator state, and with
1065 caveats regarding the fragility of while-each loops.
1066 [GH #16334] <https://github.com/Perl/perl5/issues/16334>
1067 .IP \(bu 4
1068 Clarification to \f(CW\*(C`require\*(C'\fR was added to explain the differences between
1069 .Sp
1070 .Vb 2
1071 \& require Foo::Bar;
1072 \& require "Foo/Bar.pm";
1073 .Ve
1074 .PP
1075 \fIperlgit\fR
1076 .IX Subsection "perlgit"
1077 .IP \(bu 4
1078 The precise rules for identifying \f(CW\*(C`smoke\-me\*(C'\fR branches are now stated.
1079 .PP
1080 \fIperlguts\fR
1081 .IX Subsection "perlguts"
1082 .IP \(bu 4
1083 The section on reference counting in perlguts has been heavily revised,
1084 to describe references in the way a programmer needs to think about them
1085 rather than in terms of the physical data structures.
1086 .IP \(bu 4
1087 Improve documentation related to UTF\-8 multibytes.
1088 .PP
1089 \fIperlintern\fR
1090 .IX Subsection "perlintern"
1091 .IP \(bu 4
1092 The internal functions \f(CWnewXS_len_flags()\fR and \f(CWnewATTRSUB_x()\fR are
1093 now documented.
1094 .PP
1095 \fIperlobj\fR
1096 .IX Subsection "perlobj"
1097 .IP \(bu 4
1098 The documentation about \f(CW\*(C`DESTROY\*(C'\fR methods has been corrected, updated,
1099 and revised, especially in regard to how they interact with exceptions.
1100 [GH #14083] <https://github.com/Perl/perl5/issues/14083>
1101 .PP
1102 \fIperlop\fR
1103 .IX Subsection "perlop"
1104 .IP \(bu 4
1105 The description of the \f(CW\*(C`x\*(C'\fR operator in perlop has been clarified.
1106 [GH #16253] <https://github.com/Perl/perl5/issues/16253>
1107 .IP \(bu 4
1108 perlop has been updated to note that \f(CW\*(C`qw\*(C'\fR's whitespace rules differ
1109 from that of \f(CW\*(C`split\*(C'\fR's in that only ASCII whitespace is used.
1110 .IP \(bu 4
1111 The general explanation of operator precedence and associativity has
1112 been corrected and clarified.
1113 [GH #15153] <https://github.com/Perl/perl5/issues/15153>
1114 .IP \(bu 4
1115 The documentation for the \f(CW\*(C`\e\*(C'\fR referencing operator now explains the
1116 unusual context that it supplies to its operand.
1117 [GH #15932] <https://github.com/Perl/perl5/issues/15932>
1118 .PP
1119 \fIperlrequick\fR
1120 .IX Subsection "perlrequick"
1121 .IP \(bu 4
1122 Clarifications on metacharacters and character classes
1123 .PP
1124 \fIperlretut\fR
1125 .IX Subsection "perlretut"
1126 .IP \(bu 4
1127 Clarify metacharacters.
1128 .PP
1129 \fIperlrun\fR
1130 .IX Subsection "perlrun"
1131 .IP \(bu 4
1132 Clarify the differences between \fB\-M\fR and \fB\-m\fR.
1133 [GH #15998] <https://github.com/Perl/perl5/issues/15998>
1134 .PP
1135 \fIperlsec\fR
1136 .IX Subsection "perlsec"
1137 .IP \(bu 4
1138 The documentation about set-id scripts has been updated and revised.
1139 [GH #10289] <https://github.com/Perl/perl5/issues/10289>
1140 .IP \(bu 4
1141 A section about using \f(CW\*(C`sudo\*(C'\fR to run Perl scripts has been added.
1142 .PP
1143 \fIperlsyn\fR
1144 .IX Subsection "perlsyn"
1145 .IP \(bu 4
1146 The section "Truth and Falsehood" in perlsyn has been removed from
1147 that document, where it didn't belong, and merged into the existing
1148 paragraph on the same topic in perldata.
1149 .IP \(bu 4
1150 The means to disambiguate between code blocks and hash constructors,
1151 already documented in perlref, are now documented in perlsyn too.
1152 [GH #15918] <https://github.com/Perl/perl5/issues/15918>
1153 .PP
1154 \fIperluniprops\fR
1155 .IX Subsection "perluniprops"
1156 .IP \(bu 4
1157 perluniprops has been updated to note that \f(CW\*(C`\ep{Word}\*(C'\fR now includes
1158 code points matching the \f(CW\*(C`\ep{Join_Control}\*(C'\fR property. The change to
1159 the property was made in Perl 5.18, but not documented until now. There
1160 are currently only two code points that match this property U+200C (ZERO
1161 WIDTH NON-JOINER) and U+200D (ZERO WIDTH JOINER).
1162 .IP \(bu 4
1163 For each binary table or property, the documentation now includes which
1164 characters in the range \f(CW\*(C`\ex00\-\exFF\*(C'\fR it matches, as well as a list of
1165 the first few ranges of code points matched above that.
1166 .PP
1167 \fIperlvar\fR
1168 .IX Subsection "perlvar"
1169 .IP \(bu 4
1170 The entry for \f(CW$+\fR in perlvar has been expanded upon to describe handling of
1171 multiply-named capturing groups.
1172 .PP
1173 \fIperlfunc, perlop, perlsyn\fR
1174 .IX Subsection "perlfunc, perlop, perlsyn"
1175 .IP \(bu 4
1176 In various places, improve the documentation of the special cases
1177 in the condition expression of a while loop, such as implicit \f(CW\*(C`defined\*(C'\fR
1178 and assignment to \f(CW$_\fR.
1179 [GH #16334] <https://github.com/Perl/perl5/issues/16334>
1180 .SH Diagnostics
1181 .IX Header "Diagnostics"
1182 The following additions or changes have been made to diagnostic output,
1183 including warnings and fatal error messages. For the complete list of
1184 diagnostic messages, see perldiag.
1185 .SS "New Diagnostics"
1186 .IX Subsection "New Diagnostics"
1187 \fINew Errors\fR
1188 .IX Subsection "New Errors"
1189 .IP \(bu 4
1190 Can't "goto" into a "given" block
1191 .Sp
1192 (F) A "goto" statement was executed to jump into the middle of a \f(CW\*(C`given\*(C'\fR
1193 block. You can't get there from here. See "goto" in perlfunc.
1194 .IP \(bu 4
1195 Can't "goto" into a binary or list expression
1196 .Sp
1197 Use of \f(CW\*(C`goto\*(C'\fR to jump into the parameter of a binary or list operator has
1198 been prohibited, to prevent crashes and stack corruption.
1199 [GH #15914] <https://github.com/Perl/perl5/issues/15914>
1200 .Sp
1201 You may only enter the \fIfirst\fR argument of an operator that takes a fixed
1202 number of arguments, since this is a case that will not cause stack
1203 corruption.
1204 [GH #16415] <https://github.com/Perl/perl5/issues/16415>
1205 .PP
1206 \fINew Warnings\fR
1207 .IX Subsection "New Warnings"
1208 .IP \(bu 4
1209 Old package separator used in string
1210 .Sp
1211 (W syntax) You used the old package separator, "'", in a variable
1212 named inside a double-quoted string; e.g., \f(CW"In $name\*(Aqs house"\fR. This
1213 is equivalent to \f(CW"In $name::s house"\fR. If you meant the former, put
1214 a backslash before the apostrophe (\f(CW"In $name\e\*(Aqs house"\fR).
1215 .IP \(bu 4
1216 "Locale '%s' contains (at least) the following characters which
1217 have unexpected meanings: \f(CW%s\fR The Perl program will use the expected
1218 meanings" in perldiag
1219 .SS "Changes to Existing Diagnostics"
1220 .IX Subsection "Changes to Existing Diagnostics"
1221 .IP \(bu 4
1222 A false-positive warning that was issued when using a
1223 numerically-quantified sub-pattern in a recursive regex has been
1224 silenced. [GH #16106] <https://github.com/Perl/perl5/issues/16106>
1225 .IP \(bu 4
1226 The warning about useless use of a concatenation operator in void context
1227 is now generated for expressions with multiple concatenations, such as
1228 \&\f(CW\*(C`$a.$b.$c\*(C'\fR, which used to mistakenly not warn.
1229 [GH #3990] <https://github.com/Perl/perl5/issues/3990>
1230 .IP \(bu 4
1231 Warnings that a variable or subroutine "masks earlier declaration in same
1232 \&...", or that an \f(CW\*(C`our\*(C'\fR variable has been redeclared, have been moved to a
1233 new warnings category "shadow". Previously they were in category "misc".
1234 .IP \(bu 4
1235 The deprecation warning from \f(CWSys::Hostname::hostname()\fR saying that
1236 it doesn't accept arguments now states the Perl version in which the
1237 warning will be upgraded to an error.
1238 [GH #14662] <https://github.com/Perl/perl5/issues/14662>
1239 .IP \(bu 4
1240 The perldiag entry for the error regarding a set-id script has been
1241 expanded to make clear that the error is reporting a specific security
1242 vulnerability, and to advise how to fix it.
1243 .IP \(bu 4
1244 The \f(CW\*(C`Unable to flush stdout\*(C'\fR error message was missing a trailing
1245 newline. [debian #875361]
1246 .SH "Utility Changes"
1247 .IX Header "Utility Changes"
1248 .SS perlbug
1249 .IX Subsection "perlbug"
1250 .IP \(bu 4
1251 \&\f(CW\*(C`\-\-help\*(C'\fR and \f(CW\*(C`\-\-version\*(C'\fR options have been added.
1252 .SH "Configuration and Compilation"
1253 .IX Header "Configuration and Compilation"
1254 .IP \(bu 4
1255 C89 requirement
1256 .Sp
1257 Perl has been documented as requiring a C89 compiler to build since October
1258 1998. A variety of simplifications have now been made to Perl's internals to
1259 rely on the features specified by the C89 standard. We believe that this
1260 internal change hasn't altered the set of platforms that Perl builds on, but
1261 please report a bug if Perl now has new problems building on your platform.
1262 .IP \(bu 4
1263 On GCC, \f(CW\*(C`\-Werror=pointer\-arith\*(C'\fR is now enabled by default,
1264 disallowing arithmetic on void and function pointers.
1265 .IP \(bu 4
1266 Where an HTML version of the documentation is installed, the HTML
1267 documents now use relative links to refer to each other. Links from
1268 the index page of perlipc to the individual section documents are
1269 now correct.
1270 [GH #11941] <https://github.com/Perl/perl5/issues/11941>
1271 .IP \(bu 4
1272 \&\fIlib/unicore/mktables\fR now correctly canonicalizes the names of the
1273 dependencies stored in the files it generates.
1274 .Sp
1275 \&\fIregen/mk_invlists.pl\fR, unlike the other \fIregen/*.pl\fR scripts, used
1276 \&\f(CW$0\fR to name itself in the dependencies stored in the files it
1277 generates. It now uses a literal so that the path stored in the
1278 generated files doesn't depend on how \fIregen/mk_invlists.pl\fR is
1279 invoked.
1280 .Sp
1281 This lack of canonical names could cause test failures in \fIt/porting/regen.t\fR.
1282 [GH #16446] <https://github.com/Perl/perl5/issues/16446>
1283 .IP \(bu 4
1284 New probes
1285 .RS 4
1286 .IP HAS_BUILTIN_ADD_OVERFLOW 2
1287 .IX Item "HAS_BUILTIN_ADD_OVERFLOW"
1288 .PD 0
1289 .IP HAS_BUILTIN_MUL_OVERFLOW 2
1290 .IX Item "HAS_BUILTIN_MUL_OVERFLOW"
1291 .IP HAS_BUILTIN_SUB_OVERFLOW 2
1292 .IX Item "HAS_BUILTIN_SUB_OVERFLOW"
1293 .IP HAS_THREAD_SAFE_NL_LANGINFO_L 2
1294 .IX Item "HAS_THREAD_SAFE_NL_LANGINFO_L"
1295 .IP HAS_LOCALECONV_L 2
1296 .IX Item "HAS_LOCALECONV_L"
1297 .IP HAS_MBRLEN 2
1298 .IX Item "HAS_MBRLEN"
1299 .IP HAS_MBRTOWC 2
1300 .IX Item "HAS_MBRTOWC"
1301 .IP HAS_MEMRCHR 2
1302 .IX Item "HAS_MEMRCHR"
1303 .IP HAS_NANOSLEEP 2
1304 .IX Item "HAS_NANOSLEEP"
1305 .IP HAS_STRNLEN 2
1306 .IX Item "HAS_STRNLEN"
1307 .IP HAS_STRTOLD_L 2
1308 .IX Item "HAS_STRTOLD_L"
1309 .IP I_WCHAR 2
1310 .IX Item "I_WCHAR"
1311 .RE
1312 .RS 4
1313 .RE
1314 .PD
1315 .SH Testing
1316 .IX Header "Testing"
1317 .IP \(bu 4
1318 Testing of the XS-APItest directory is now done in parallel, where
1319 applicable.
1320 .IP \(bu 4
1321 Perl now includes a default \fI.travis.yml\fR file for Travis CI testing
1322 on github mirrors.
1323 [GH #14558] <https://github.com/Perl/perl5/issues/14558>
1324 .IP \(bu 4
1325 The watchdog timer count in \fIre/pat_psycho.t\fR can now be overridden.
1326 .Sp
1327 This test can take a long time to run, so there is a timer to keep
1328 this in check (currently, 5 minutes). This commit adds checking
1329 the environment variable \f(CW\*(C`PERL_TEST_TIME_OUT_FACTOR\*(C'\fR; if set,
1330 the time out setting is multiplied by its value.
1331 .IP \(bu 4
1332 \&\fIharness\fR no longer waits for 30 seconds when running \fIt/io/openpid.t\fR.
1333 [GH #13535] <https://github.com/Perl/perl5/issues/13535>
1334 [GH #16420] <https://github.com/Perl/perl5/issues/16420>
1335 .SH Packaging
1336 .IX Header "Packaging"
1337 For the past few years we have released perl using three different archive
1338 formats: bzip (\f(CW\*(C`.bz2\*(C'\fR), LZMA2 (\f(CW\*(C`.xz\*(C'\fR) and gzip (\f(CW\*(C`.gz\*(C'\fR). Since xz compresses
1339 better and decompresses faster, and gzip is more compatible and uses less
1340 memory, we have dropped the \f(CW\*(C`.bz2\*(C'\fR archive format with this release.
1341 (If this poses a problem, do let us know; see "Reporting Bugs", below.)
1342 .SH "Platform Support"
1343 .IX Header "Platform Support"
1344 .SS "Discontinued Platforms"
1345 .IX Subsection "Discontinued Platforms"
1346 .IP "PowerUX / Power MAX OS" 4
1347 .IX Item "PowerUX / Power MAX OS"
1348 Compiler hints and other support for these apparently long-defunct
1349 platforms has been removed.
1350 .SS "Platform-Specific Notes"
1351 .IX Subsection "Platform-Specific Notes"
1352 .IP CentOS 4
1353 .IX Item "CentOS"
1354 Compilation on CentOS 5 is now fixed.
1355 .IP Cygwin 4
1356 .IX Item "Cygwin"
1357 A build with the quadmath library can now be done on Cygwin.
1358 .IP Darwin 4
1359 .IX Item "Darwin"
1360 Perl now correctly uses reentrant functions, like \f(CW\*(C`asctime_r\*(C'\fR, on
1361 versions of Darwin that have support for them.
1362 .IP FreeBSD 4
1363 .IX Item "FreeBSD"
1364 FreeBSD's \fI/usr/share/mk/sys.mk\fR specifies \f(CW\*(C`\-O2\*(C'\fR for
1365 architectures other than ARM and MIPS. By default, perl is now compiled
1366 with the same optimization levels.
1367 .IP VMS 4
1368 .IX Item "VMS"
1369 Several fix-ups for \fIconfigure.com\fR, marking function VMS has
1370 (or doesn't have).
1371 .Sp
1372 CRTL features can now be set by embedders before invoking Perl by using
1373 the \f(CW\*(C`decc$feature_set\*(C'\fR and \f(CW\*(C`decc$feature_set_value\*(C'\fR functions.
1374 Previously any attempt to set features after image initialization were
1375 ignored.
1376 .IP Windows 4
1377 .IX Item "Windows"
1378 .RS 4
1379 .PD 0
1380 .IP \(bu 4
1381 .PD
1382 Support for compiling perl on Windows using Microsoft Visual Studio 2017
1383 (containing Visual C++ 14.1) has been added.
1384 .IP \(bu 4
1385 Visual C++ compiler version detection has been improved to work on non-English
1386 language systems.
1387 .IP \(bu 4
1388 We now set \f(CW$Config{libpth}\fR correctly for 64\-bit builds using Visual C++
1389 versions earlier than 14.1.
1390 .RE
1391 .RS 4
1392 .RE
1393 .SH "Internal Changes"
1394 .IX Header "Internal Changes"
1395 .IP \(bu 4
1396 A new optimisation phase has been added to the compiler,
1397 \&\f(CWoptimize_optree()\fR, which does a top-down scan of a complete optree
1398 just before the peephole optimiser is run. This phase is not currently
1399 hookable.
1400 .IP \(bu 4
1401 An \f(CW\*(C`OP_MULTICONCAT\*(C'\fR op has been added. At \f(CWoptimize_optree()\fR time, a
1402 chain of \f(CW\*(C`OP_CONCAT\*(C'\fR and \f(CW\*(C`OP_CONST\*(C'\fR ops, together optionally with an
1403 \&\f(CW\*(C`OP_STRINGIFY\*(C'\fR and/or \f(CW\*(C`OP_SASSIGN\*(C'\fR, are combined into a single
1404 \&\f(CW\*(C`OP_MULTICONCAT\*(C'\fR op. The op is of type \f(CW\*(C`UNOP_AUX\*(C'\fR, and the aux array
1405 contains the argument count, plus a pointer to a constant string and a set
1406 of segment lengths. For example with
1407 .Sp
1408 .Vb 1
1409 \& my $x = "foo=$foo, bar=$bar\en";
1410 .Ve
1411 .Sp
1412 the constant string would be \f(CW"foo=, bar=\en"\fR and the segment lengths
1413 would be (4,6,1). If the string contains characters such as \f(CW\*(C`\ex80\*(C'\fR, whose
1414 representation changes under utf8, two sets of strings plus lengths are
1415 precomputed and stored.
1416 .IP \(bu 4
1417 Direct access to \f(CW\*(C`PL_keyword_plugin\*(C'\fR is not
1418 safe in the presence of multithreading. A new
1419 \&\f(CW\*(C`wrap_keyword_plugin\*(C'\fR function has been
1420 added to allow XS modules to safely define custom keywords even when
1421 loaded from a thread, analogous to \f(CW\*(C`PL_check\*(C'\fR /
1422 \&\f(CW\*(C`wrap_op_checker\*(C'\fR.
1423 .IP \(bu 4
1424 The \f(CW\*(C`PL_statbuf\*(C'\fR interpreter variable has been removed.
1425 .IP \(bu 4
1426 The deprecated function \f(CWto_utf8_case()\fR, accessible from XS code, has
1427 been removed.
1428 .IP \(bu 4
1429 A new function
1430 \&\f(CWis_utf8_invariant_string_loc()\fR
1431 has been added that is like
1432 \&\f(CWis_utf8_invariant_string()\fR
1433 but takes an extra pointer parameter into which is stored the location
1434 of the first variant character, if any are found.
1435 .IP \(bu 4
1436 A new function, \f(CWPerl_langinfo()\fR has been
1437 added. It is an (almost) drop-in replacement for the system
1438 \&\f(CWnl_langinfo(3)\fR, but works on platforms that lack that; as well as
1439 being more thread-safe, and hiding some gotchas with locale handling
1440 from the caller. Code that uses this, needn't use \f(CWlocaleconv(3)\fR
1441 (and be affected by the gotchas) to find the decimal point, thousands
1442 separator, or currency symbol. See "Perl_langinfo" in perlapi.
1443 .IP \(bu 4
1444 A new API function \f(CWsv_rvunweaken()\fR has
1445 been added to complement \f(CWsv_rvweaken()\fR.
1446 The implementation was taken from "unweaken" in Scalar::Util.
1447 .IP \(bu 4
1448 A new flag, \f(CW\*(C`SORTf_UNSTABLE\*(C'\fR, has been added. This will allow a
1449 future commit to make mergesort unstable when the user specifies ‘no
1450 sort stable’, since it has been decided that mergesort should remain
1451 stable by default.
1452 .IP \(bu 4
1453 XS modules can now automatically get reentrant versions of system
1454 functions on threaded perls.
1455 .Sp
1456 By adding
1457 .Sp
1458 .Vb 1
1459 \& #define PERL_REENTRANT
1460 .Ve
1461 .Sp
1462 near the beginning of an \f(CW\*(C`XS\*(C'\fR file, it will be compiled so that
1463 whatever reentrant functions perl knows about on that system will
1464 automatically and invisibly be used instead of the plain, non-reentrant
1465 versions. For example, if you write \f(CWgetpwnam()\fR in your code, on a
1466 system that has \f(CWgetpwnam_r()\fR all calls to the former will be translated
1467 invisibly into the latter. This does not happen except on threaded
1468 perls, as they aren't needed otherwise. Be aware that which functions
1469 have reentrant versions varies from system to system.
1470 .IP \(bu 4
1471 The \f(CW\*(C`PERL_NO_OP_PARENT\*(C'\fR build define is no longer supported, which means
1472 that perl is now always built with \f(CW\*(C`PERL_OP_PARENT\*(C'\fR enabled.
1473 .IP \(bu 4
1474 The format and content of the non\-utf8 transliteration table attached to
1475 the \f(CW\*(C`op_pv\*(C'\fR field of \f(CW\*(C`OP_TRANS\*(C'\fR/\f(CW\*(C`OP_TRANSR\*(C'\fR ops has changed. It's now a
1476 \&\f(CW\*(C`struct OPtrans_map\*(C'\fR.
1477 .IP \(bu 4
1478 A new compiler \f(CW\*(C`#define\*(C'\fR, \f(CW\*(C`dTHX_DEBUGGING\*(C'\fR. has been added. This is
1479 useful for XS or C code that only need the thread context because their
1480 debugging statements that get compiled only under \f(CW\*(C`\-DDEBUGGING\*(C'\fR need
1481 one.
1482 .IP \(bu 4
1483 A new API function "Perl_setlocale" in perlapi has been added.
1484 .IP \(bu 4
1485 "sync_locale" in perlapi has been revised to return a boolean as to
1486 whether the system was using the global locale or not.
1487 .IP \(bu 4
1488 A new kind of magic scalar, called a "nonelem" scalar, has been introduced.
1489 It is stored in an array to denote a non-existent element, whenever such an
1490 element is accessed in a potential lvalue context. It replaces the
1491 existing "defelem" (deferred element) magic wherever this is possible,
1492 being significantly more efficient. This means that
1493 \&\f(CWsome_sub($sparse_array[$nonelem])\fR no longer has to create a new magic
1494 defelem scalar each time, as long as the element is within the array.
1495 .Sp
1496 It partially fixes the rare bug of deferred elements getting out of synch
1497 with their arrays when the array is shifted or unshifted.
1498 [GH #16364] <https://github.com/Perl/perl5/issues/16364>
1499 .SH "Selected Bug Fixes"
1500 .IX Header "Selected Bug Fixes"
1501 .IP \(bu 4
1502 List assignment (\f(CW\*(C`aassign\*(C'\fR) could in some rare cases allocate an
1503 entry on the mortals stack and leave the entry uninitialized, leading to
1504 possible crashes.
1505 [GH #16017] <https://github.com/Perl/perl5/issues/16017>
1506 .IP \(bu 4
1507 Attempting to apply an attribute to an \f(CW\*(C`our\*(C'\fR variable where a
1508 function of that name already exists could result in a NULL pointer
1509 being supplied where an SV was expected, crashing perl.
1510 [perl #131597] <https://rt.perl.org/Ticket/Display.html?id=131597>
1511 .IP \(bu 4
1512 \&\f(CW\*(C`split \*(Aq \*(Aq\*(C'\fR now correctly handles the argument being split when in the
1513 scope of the \f(CW\*(C`unicode_strings\*(C'\fR feature. Previously, when a string using the single-byte internal
1514 representation contained characters that are whitespace by Unicode rules but
1515 not by ASCII rules, it treated those characters as part of fields rather
1516 than as field separators.
1517 [GH #15904] <https://github.com/Perl/perl5/issues/15904>
1518 .IP \(bu 4
1519 Several built-in functions previously had bugs that could cause them to
1520 write to the internal stack without allocating room for the item being
1521 written. In rare situations, this could have led to a crash. These bugs have
1522 now been fixed, and if any similar bugs are introduced in future, they will
1523 be detected automatically in debugging builds.
1524 .Sp
1525 These internal stack usage checks introduced are also done
1526 by the \f(CW\*(C`entersub\*(C'\fR operator when calling XSUBs. This means we can
1527 report which XSUB failed to allocate enough stack space.
1528 [GH #16126] <https://github.com/Perl/perl5/issues/16126>
1529 .IP \(bu 4
1530 Using a symbolic ref with postderef syntax as the key in a hash lookup was
1531 yielding an assertion failure on debugging builds.
1532 [GH #16029] <https://github.com/Perl/perl5/issues/16029>
1533 .IP \(bu 4
1534 Array and hash variables whose names begin with a caret now admit indexing
1535 inside their curlies when interpolated into strings, as in \f(CW"${^CAPTURE[0]}"\fR to index \f(CW\*(C`@{^CAPTURE}\*(C'\fR.
1536 [GH #16050] <https://github.com/Perl/perl5/issues/16050>
1537 .IP \(bu 4
1538 Fetching the name of a glob that was previously UTF\-8 but wasn't any
1539 longer would return that name flagged as UTF\-8.
1540 [GH #15971] <https://github.com/Perl/perl5/issues/15971>
1541 .IP \(bu 4
1542 The perl \f(CWsprintf()\fR function (via the underlying C function
1543 \&\f(CWPerl_sv_vcatpvfn_flags()\fR) has been heavily reworked to fix many minor
1544 bugs, including the integer wrapping of large width and precision
1545 specifiers and potential buffer overruns. It has also been made faster in
1546 many cases.
1547 .IP \(bu 4
1548 Exiting from an \f(CW\*(C`eval\*(C'\fR, whether normally or via an exception, now always
1549 frees temporary values (possibly calling destructors) \fIbefore\fR setting
1550 \&\f(CW$@\fR. For example:
1551 .Sp
1552 .Vb 3
1553 \& sub DESTROY { eval { die "died in DESTROY"; } }
1554 \& eval { bless []; };
1555 \& # $@ used to be equal to "died in DESTROY" here; it\*(Aqs now "".
1556 .Ve
1557 .IP \(bu 4
1558 Fixed a duplicate symbol failure with \f(CW\*(C`\-flto \-mieee\-fp\*(C'\fR builds.
1559 \&\fIpp.c\fR defined \f(CW\*(C`_LIB_VERSION\*(C'\fR which \f(CW\*(C`\-lieee\*(C'\fR already defines.
1560 [GH #16086] <https://github.com/Perl/perl5/issues/16086>
1561 .IP \(bu 4
1562 The tokenizer no longer consumes the exponent part of a floating
1563 point number if it's incomplete.
1564 [GH #16073] <https://github.com/Perl/perl5/issues/16073>
1565 .IP \(bu 4
1566 On non-threaded builds, for \f(CW\*(C`m/$null/\*(C'\fR where \f(CW$null\fR is an empty
1567 string is no longer treated as if the \f(CW\*(C`/o\*(C'\fR flag was present when the
1568 previous matching match operator included the \f(CW\*(C`/o\*(C'\fR flag. The
1569 rewriting used to implement this behavior could confuse the
1570 interpreter. This matches the behaviour of threaded builds.
1571 [GH #14668] <https://github.com/Perl/perl5/issues/14668>
1572 .IP \(bu 4
1573 Parsing a \f(CW\*(C`sub\*(C'\fR definition could cause a use after free if the \f(CW\*(C`sub\*(C'\fR
1574 keyword was followed by whitespace including newlines (and comments.)
1575 [GH #16097] <https://github.com/Perl/perl5/issues/16097>
1576 .IP \(bu 4
1577 The tokenizer now correctly adjusts a parse pointer when skipping
1578 whitespace in a \f(CW\*(C`${identifier}\*(C'\fR construct.
1579 [perl #131949] <https://rt.perl.org/Public/Bug/Display.html?id=131949>
1580 .IP \(bu 4
1581 Accesses to \f(CW\*(C`${^LAST_FH}\*(C'\fR no longer assert after using any of a
1582 variety of I/O operations on a non-glob.
1583 [GH #15372] <https://github.com/Perl/perl5/issues/15372>
1584 .IP \(bu 4
1585 The XS-level \f(CWCopy()\fR, \f(CWMove()\fR, \f(CWZero()\fR macros and their variants now
1586 assert if the pointers supplied are \f(CW\*(C`NULL\*(C'\fR. ISO C considers
1587 supplying NULL pointers to the functions these macros are built upon
1588 as undefined behaviour even when their count parameters are zero.
1589 Based on these assertions and the original bug report three macro
1590 calls were made conditional.
1591 [GH #16079] <https://github.com/Perl/perl5/issues/16079>
1592 [GH #16112] <https://github.com/Perl/perl5/issues/16112>
1593 .IP \(bu 4
1594 Only the \f(CW\*(C`=\*(C'\fR operator is permitted for defining defaults for
1595 parameters in subroutine signatures. Previously other assignment
1596 operators, e.g. \f(CW\*(C`+=\*(C'\fR, were also accidentally permitted.
1597 [GH #16084] <https://github.com/Perl/perl5/issues/16084>
1598 .IP \(bu 4
1599 Package names are now always included in \f(CW\*(C`:prototype\*(C'\fR warnings
1600 [perl #131833] <https://rt.perl.org/Public/Bug/Display.html?id=131833>
1601 .IP \(bu 4
1602 The \f(CW\*(C`je_old_stack_hwm\*(C'\fR field, previously only found in the \f(CW\*(C`jmpenv\*(C'\fR
1603 structure on debugging builds, has been added to non-debug builds as
1604 well. This fixes an issue with some CPAN modules caused by the size of
1605 this structure varying between debugging and non-debugging builds.
1606 [GH #16122] <https://github.com/Perl/perl5/issues/16122>
1607 .IP \(bu 4
1608 The arguments to the \f(CWninstr()\fR macro are now correctly parenthesized.
1609 .IP \(bu 4
1610 A NULL pointer dereference in the \f(CWS_regmatch()\fR function has been
1611 fixed.
1612 [perl #132017] <https://rt.perl.org/Public/Bug/Display.html?id=132017>
1613 .IP \(bu 4
1614 Calling exec PROGRAM LIST with an empty \f(CW\*(C`LIST\*(C'\fR
1615 has been fixed. This should call \f(CWexecvp()\fR with an empty \f(CW\*(C`argv\*(C'\fR array
1616 (containing only the terminating \f(CW\*(C`NULL\*(C'\fR pointer), but was instead just
1617 returning false (and not setting \f(CW$!\fR).
1618 [GH #16075] <https://github.com/Perl/perl5/issues/16075>
1619 .IP \(bu 4
1620 The \f(CW\*(C`gv_fetchmeth_sv\*(C'\fR C function stopped working properly in Perl 5.22 when
1621 fetching a constant with a UTF\-8 name if that constant subroutine was stored in
1622 the stash as a simple scalar reference, rather than a full typeglob. This has
1623 been corrected.
1624 .IP \(bu 4
1625 Single-letter debugger commands followed by an argument which starts with
1626 punctuation (e.g. \f(CW\*(C`p$^V\*(C'\fR and \f(CW\*(C`x@ARGV\*(C'\fR) now work again. They had been
1627 wrongly requiring a space between the command and the argument.
1628 [GH #13342] <https://github.com/Perl/perl5/issues/13342>
1629 .IP \(bu 4
1630 splice now throws an exception
1631 ("Modification of a read-only value attempted") when modifying a read-only
1632 array. Until now it had been silently modifying the array. The new behaviour
1633 is consistent with the behaviour of push and
1634 unshift.
1635 [GH #15923] <https://github.com/Perl/perl5/issues/15923>
1636 .IP \(bu 4
1637 \&\f(CWstat()\fR, \f(CWlstat()\fR, and file test operators now fail if given a
1638 filename containing a nul character, in the same way that \f(CWopen()\fR
1639 already fails.
1640 .IP \(bu 4
1641 \&\f(CWstat()\fR, \f(CWlstat()\fR, and file test operators now reliably set \f(CW$!\fR when
1642 failing due to being applied to a closed or otherwise invalid file handle.
1643 .IP \(bu 4
1644 File test operators for Unix permission bits that don't exist on a
1645 particular platform, such as \f(CW\*(C`\-k\*(C'\fR (sticky bit) on Windows, now check that
1646 the file being tested exists before returning the blanket false result,
1647 and yield the appropriate errors if the argument doesn't refer to a file.
1648 .IP \(bu 4
1649 Fixed a 'read before buffer' overrun when parsing a range starting with
1650 \&\f(CW\*(C`\eN{}\*(C'\fR at the beginning of the character set for the transliteration
1651 operator.
1652 [GH #16189] <https://github.com/Perl/perl5/issues/16189>
1653 .IP \(bu 4
1654 Fixed a leaked scalar when parsing an empty \f(CW\*(C`\eN{}\*(C'\fR at compile-time.
1655 [GH #16189] <https://github.com/Perl/perl5/issues/16189>
1656 .IP \(bu 4
1657 Calling \f(CW\*(C`do $path\*(C'\fR on a directory or block device now yields a meaningful
1658 error code in \f(CW$!\fR.
1659 [GH #14841] <https://github.com/Perl/perl5/issues/14841>
1660 .IP \(bu 4
1661 Regexp substitution using an overloaded replacement value that provides
1662 a tainted stringification now correctly taints the resulting string.
1663 [GH #12495] <https://github.com/Perl/perl5/issues/12495>
1664 .IP \(bu 4
1665 Lexical sub declarations in \f(CW\*(C`do\*(C'\fR blocks such as \f(CW\*(C`do { my sub lex; 123 }\*(C'\fR
1666 could corrupt the stack, erasing items already on the stack in the
1667 enclosing statement. This has been fixed.
1668 [GH #16243] <https://github.com/Perl/perl5/issues/16243>
1669 .IP \(bu 4
1670 \&\f(CW\*(C`pack\*(C'\fR and \f(CW\*(C`unpack\*(C'\fR can now handle repeat counts and lengths that
1671 exceed two billion.
1672 [GH #13179] <https://github.com/Perl/perl5/issues/13179>
1673 .IP \(bu 4
1674 Digits past the radix point in octal and binary floating point literals
1675 now have the correct weight on platforms where a floating point
1676 significand doesn't fit into an integer type.
1677 .IP \(bu 4
1678 The canonical truth value no longer has a spurious special meaning as a
1679 callable subroutine. It used to be a magic placeholder for a missing
1680 \&\f(CW\*(C`import\*(C'\fR or \f(CW\*(C`unimport\*(C'\fR method, but is now treated like any other string
1681 \&\f(CW1\fR.
1682 [GH #14902] <https://github.com/Perl/perl5/issues/14902>
1683 .IP \(bu 4
1684 \&\f(CW\*(C`system\*(C'\fR now reduces its arguments to strings in the parent process, so
1685 any effects of stringifying them (such as overload methods being called
1686 or warnings being emitted) are visible in the way the program expects.
1687 [GH #13561] <https://github.com/Perl/perl5/issues/13561>
1688 .IP \(bu 4
1689 The \f(CWreadpipe()\fR built-in function now checks at compile time that
1690 it has only one parameter expression, and puts it in scalar context,
1691 thus ensuring that it doesn't corrupt the stack at runtime.
1692 [GH #2793] <https://github.com/Perl/perl5/issues/2793>
1693 .IP \(bu 4
1694 \&\f(CW\*(C`sort\*(C'\fR now performs correct reference counting when aliasing \f(CW$a\fR and
1695 \&\f(CW$b\fR, thus avoiding premature destruction and leakage of scalars if they
1696 are re-aliased during execution of the sort comparator.
1697 [GH #11422] <https://github.com/Perl/perl5/issues/11422>
1698 .IP \(bu 4
1699 \&\f(CW\*(C`reverse\*(C'\fR with no operand, reversing \f(CW$_\fR by default, is no longer in
1700 danger of corrupting the stack.
1701 [GH #16291] <https://github.com/Perl/perl5/issues/16291>
1702 .IP \(bu 4
1703 \&\f(CW\*(C`exec\*(C'\fR, \f(CW\*(C`system\*(C'\fR, et al are no longer liable to have their argument
1704 lists corrupted by reentrant calls and by magic such as tied scalars.
1705 [GH #15660] <https://github.com/Perl/perl5/issues/15660>
1706 .IP \(bu 4
1707 Perl's own \f(CW\*(C`malloc\*(C'\fR no longer gets confused by attempts to allocate
1708 more than a gigabyte on a 64\-bit platform.
1709 [GH #13273] <https://github.com/Perl/perl5/issues/13273>
1710 .IP \(bu 4
1711 Stacked file test operators in a sort comparator expression no longer
1712 cause a crash.
1713 [GH #15626] <https://github.com/Perl/perl5/issues/15626>
1714 .IP \(bu 4
1715 An identity \f(CW\*(C`tr///\*(C'\fR transformation on a reference is no longer mistaken
1716 for that reference for the purposes of deciding whether it can be
1717 assigned to.
1718 [GH #15812] <https://github.com/Perl/perl5/issues/15812>
1719 .IP \(bu 4
1720 Lengthy hexadecimal, octal, or binary floating point literals no
1721 longer cause undefined behaviour when parsing digits that are of such
1722 low significance that they can't affect the floating point value.
1723 [GH #16114] <https://github.com/Perl/perl5/issues/16114>
1724 .IP \(bu 4
1725 \&\f(CW\*(C`open $$scalarref...\*(C'\fR and similar invocations no longer leak the file
1726 handle.
1727 [GH #12593] <https://github.com/Perl/perl5/issues/12593>
1728 .IP \(bu 4
1729 Some convoluted kinds of regexp no longer cause an arithmetic overflow
1730 when compiled.
1731 [GH #16113] <https://github.com/Perl/perl5/issues/16113>
1732 .IP \(bu 4
1733 The default typemap, by avoiding \f(CW\*(C`newGVgen\*(C'\fR, now no longer leaks when
1734 XSUBs return file handles (\f(CW\*(C`PerlIO *\*(C'\fR or \f(CW\*(C`FILE *\*(C'\fR).
1735 [GH #12593] <https://github.com/Perl/perl5/issues/12593>
1736 .IP \(bu 4
1737 Creating a \f(CW\*(C`BEGIN\*(C'\fR block as an XS subroutine with a prototype no longer
1738 crashes because of the early freeing of the subroutine.
1739 .IP \(bu 4
1740 The \f(CW\*(C`printf\*(C'\fR format specifier \f(CW\*(C`%.0f\*(C'\fR no longer rounds incorrectly
1741 [GH #9125] <https://github.com/Perl/perl5/issues/9125>,
1742 and now shows the correct sign for a negative zero.
1743 .IP \(bu 4
1744 Fixed an issue where the error \f(CW\*(C`Scalar value @arrayname[0] better
1745 written as $arrayname\*(C'\fR would give an error \f(CW\*(C`Cannot printf Inf with \*(Aqc\*(Aq\*(C'\fR
1746 when arrayname starts with \f(CW\*(C`Inf\*(C'\fR.
1747 [GH #16335] <https://github.com/Perl/perl5/issues/16335>
1748 .IP \(bu 4
1749 The Perl implementation of \f(CWgetcwd()\fR in \f(CW\*(C`Cwd\*(C'\fR in the PathTools
1750 distribution now behaves the same as XS implementation on errors: it
1751 returns an error, and sets \f(CW$!\fR.
1752 [GH #16338] <https://github.com/Perl/perl5/issues/16338>
1753 .IP \(bu 4
1754 Vivify array elements when putting them on the stack.
1755 Fixes [GH #5310] <https://github.com/Perl/perl5/issues/5310>
1756 (reported in April 2002).
1757 .IP \(bu 4
1758 Fixed parsing of braced subscript after parens. Fixes
1759 [GH #4688] <https://github.com/Perl/perl5/issues/4688>
1760 (reported in December 2001).
1761 .IP \(bu 4
1762 \&\f(CW\*(C`tr/non_utf8/long_non_utf8/c\*(C'\fR could give the wrong results when the
1763 length of the replacement character list was greater than 0x7fff.
1764 .IP \(bu 4
1765 \&\f(CW\*(C`tr/non_utf8/non_utf8/cd\*(C'\fR failed to add the implied
1766 \&\f(CW\*(C`\ex{100}\-\ex{7fffffff}\*(C'\fR to the search character list.
1767 .IP \(bu 4
1768 Compilation failures within "perl-within-perl" constructs, such as with
1769 string interpolation and the right part of \f(CW\*(C`s///e\*(C'\fR, now cause
1770 compilation to abort earlier.
1771 .Sp
1772 Previously compilation could continue in order to report other errors,
1773 but the failed sub-parse could leave partly parsed constructs on the
1774 parser shift-reduce stack, confusing the parser, leading to perl
1775 crashes.
1776 [GH #14739] <https://github.com/Perl/perl5/issues/14739>
1777 .IP \(bu 4
1778 On threaded perls where the decimal point (radix) character is not a
1779 dot, it has been possible for a race to occur between threads when one
1780 needs to use the real radix character (such as with \f(CW\*(C`sprintf\*(C'\fR). This has
1781 now been fixed by use of a mutex on systems without thread-safe locales,
1782 and the problem just doesn't come up on those with thread-safe locales.
1783 .IP \(bu 4
1784 Errors while compiling a regex character class could sometime trigger an
1785 assertion failure.
1786 [GH #16172] <https://github.com/Perl/perl5/issues/16172>
1787 .SH Acknowledgements
1788 .IX Header "Acknowledgements"
1789 Perl 5.28.0 represents approximately 13 months of development since Perl
1790 5.26.0 and contains approximately 730,000 lines of changes across 2,200
1791 files from 77 authors.
1792 .PP
1793 Excluding auto-generated files, documentation and release tools, there were
1794 approximately 580,000 lines of changes to 1,300 .pm, .t, .c and .h files.
1795 .PP
1796 Perl continues to flourish into its fourth decade thanks to a vibrant
1797 community of users and developers. The following people are known to have
1798 contributed the improvements that became Perl 5.28.0:
1799 .PP
1800 Aaron Crane, Abigail, Ævar Arnfjörð Bjarmason, Alberto Simões, Alexandr
1801 Savca, Andrew Fresh, Andy Dougherty, Andy Lester, Aristotle Pagaltzis, Ask
1802 Bjørn Hansen, Chris 'BinGOs' Williams, Craig A. Berry, Dagfinn Ilmari
1803 Mannsåker, Dan Collins, Daniel Dragan, David Cantrell, David Mitchell,
1804 Dmitry Ulanov, Dominic Hargreaves, E. Choroba, Eric Herman, Eugen Konkov,
1805 Father Chrysostomos, Gene Sullivan, George Hartzell, Graham Knop, Harald
1806 Jörg, H.Merijn Brand, Hugo van der Sanden, Jacques Germishuys, James E
1807 Keenan, Jarkko Hietaniemi, Jerry D. Hedden, J. Nick Koston, John Lightsey,
1808 John Peacock, John P. Linderman, John SJ Anderson, Karen Etheridge, Karl
1809 Williamson, Ken Brown, Ken Cotterill, Leon Timmermans, Lukas Mai, Marco
1810 Fontani, Marc-Philip Werner, Matthew Horsfall, Neil Bowers, Nicholas Clark,
1811 Nicolas R., Niko Tyni, Pali, Paul Marquess, Peter John Acklam, Reini Urban,
1812 Renee Baecker, Ricardo Signes, Robin Barker, Sawyer X, Scott Lanning, Sergey
1813 Aleynikov, Shirakata Kentaro, Shoichi Kaji, Slaven Rezic, Smylers, Steffen
1814 Müller, Steve Hay, Sullivan Beck, Thomas Sibley, Todd Rinaldo, Tomasz
1815 Konojacki, Tom Hukins, Tom Wyant, Tony Cook, Vitali Peil, Yves Orton,
1816 Zefram.
1817 .PP
1818 The list above is almost certainly incomplete as it is automatically
1819 generated from version control history. In particular, it does not include
1820 the names of the (very much appreciated) contributors who reported issues to
1821 the Perl bug tracker.
1822 .PP
1823 Many of the changes included in this version originated in the CPAN modules
1824 included in Perl's core. We're grateful to the entire CPAN community for
1825 helping Perl to flourish.
1826 .PP
1827 For a more complete list of all of Perl's historical contributors, please
1828 see the \fIAUTHORS\fR file in the Perl source distribution.
1829 .SH "Reporting Bugs"
1830 .IX Header "Reporting Bugs"
1831 If you find what you think is a bug, you might check the perl bug database
1832 at <https://rt.perl.org/> . There may also be information at
1833 <http://www.perl.org/> , the Perl Home Page.
1834 .PP
1835 If you believe you have an unreported bug, please run the perlbug program
1836 included with your release. Be sure to trim your bug down to a tiny but
1837 sufficient test case. Your bug report, along with the output of \f(CW\*(C`perl \-V\*(C'\fR,
1838 will be sent off to [email protected] to be analysed by the Perl porting team.
1839 .PP
1840 If the bug you are reporting has security implications which make it
1841 inappropriate to send to a publicly archived mailing list, then see
1842 "SECURITY VULNERABILITY CONTACT INFORMATION" in perlsec
1843 for details of how to report the issue.
1844 .SH "Give Thanks"
1845 .IX Header "Give Thanks"
1846 If you wish to thank the Perl 5 Porters for the work we had done in Perl 5,
1847 you can do so by running the \f(CW\*(C`perlthanks\*(C'\fR program:
1848 .PP
1849 .Vb 1
1850 \& perlthanks
1851 .Ve
1852 .PP
1853 This will send an email to the Perl 5 Porters list with your show of thanks.
1854 .SH "SEE ALSO"
1855 .IX Header "SEE ALSO"
1856 The \fIChanges\fR file for an explanation of how to view exhaustive details on
1857 what changed.
1858 .PP
1859 The \fIINSTALL\fR file for how to build Perl.
1860 .PP
1861 The \fIREADME\fR file for general stuff.
1862 .PP
1863 The \fIArtistic\fR and \fICopying\fR files for copyright information.