This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
perldelta: remove an inscrutable note, clean up another
[perl5.git] / pod / perldelta.pod
CommitLineData
44691e6f
AB
1=encoding utf8
2
3=head1 NAME
4
5ed58cbd 5perldelta - what is new for perl v5.18.0
e128ab2c 6
4eabcf70 7=head1 DESCRIPTION
6db9054f 8
5ed58cbd 9This document describes differences between the 5.16.0 release and the 5.18.0
e08634c5 10release.
6db9054f 11
d5f315e8 12If you are upgrading from an earlier release such as 5.14.0, first read
5ed58cbd 13L<perl5160delta>, which describes differences between 5.14.0 and 5.16.0.
3f01b192 14
5ed58cbd 15=head1 Core Enhancements
3f01b192 16
5ed58cbd 17=head2 New mechanism for experimental features
82d98f72 18
5ed58cbd 19Newly-added experimental features will now require this incantation:
82d98f72 20
5ed58cbd
RS
21 no warnings "experimental::feature_name";
22 use feature "feature_name"; # would warn without the prev line
23
24There is a new warnings category, called "experimental", containing
25warnings that the L<feature> pragma emits when enabling experimental
26features.
27
28Newly-added experimental features will also be given special warning IDs,
29which consist of "experimental::" followed by the name of the feature. (The
30plan is to extend this mechanism eventually to all warnings, to allow them
31to be enabled or disabled individually, and not just by category.)
32
33By saying
34
35 no warnings "experimental::feature_name";
36
37you are taking responsibility for any breakage that future changes to, or
38removal of, the feature may cause.
39
40Existing experimental features may begin emitting these warnings, too. Please
41consult L<perlexperiment> for information on which features are considered
42experimental.
43
44=head2 Hash overhaul
45
46Changes to the implementation of hashes in perl 5.18.0 will be one of the most
47visible changes to the behavior of existing code. For the most part, these
48changes will be visible as two distinct hash variables now providing their
49contents in a different order where it was previously identical. When
50encountering these changes, the key to cleaning up from them is to accept that
51B<hashes are unordered collections> and to act accordingly.
52
53=head3 Hash randomization
54
55The seed used by Perl's hash function is now random. This means that the
56order which keys/values will be returned from functions like C<keys()>,
57C<values()>, and C<each()> will differ from run to run.
58
59This change was introduced to make Perl's hashes more robust to algorithmic
60complexity attacks, and also because we discovered that it exposes hash
61ordering dependency bugs and makes them easier to track down.
62
63Toolchain maintainers might want to invest in additional infrastructure to
64test for things like this. Running tests several times in a row and then
65comparing results will make it easier to spot hash order dependencies in
66code. Authors are strongly encouraged not to expose the key order of
67Perl's hashes to insecure audiences.
68
69Further, every hash has its own iteration order, which should make it much
70more difficult to determine what the current hash seed is.
71
72=head3 New hash function: Murmurhash-32
73
74We have switched Perl's hash function to use Murmurhash-32, and added build
75support for several other hash functions. This new function is expected to
76perform equivalently to the old one for shorter strings and is faster for
77hashing longer strings.
78
f105b7be 79=head3 PERL_HASH_SEED environment variable now takes a hex value
5ed58cbd 80
f105b7be 81C<PERL_HASH_SEED> no longer accepts an integer as a parameter; instead the
5ed58cbd
RS
82value is expected to be a binary string encoded in hex. This is to make
83the infrastructure support hash seeds of arbitrary lengths which might
84exceed that of an integer. (SipHash uses a 16 byte seed).
85
c40c48bb
RS
86=head3 PERL_PERTURB_KEYS environment variable added
87
f105b7be 88The C<PERL_PERTURB_KEYS> environment variable allows one to control the level of
c40c48bb
RS
89randomization applied to C<keys> and friends.
90
f105b7be 91When C<PERL_PERTURB_KEYS> is 0, perl will not randomize the key order at all. The
c40c48bb
RS
92chance that C<keys> changes due to an insert will be the same as in previous
93perls, basically only when the bucket size is changed.
94
f105b7be 95When C<PERL_PERTURB_KEYS> is 1, perl will randomize keys in a non-repeatable
c40c48bb
RS
96way. The chance that C<keys> changes due to an insert will be very high. This
97is the most secure and default mode.
98
f105b7be
KE
99When C<PERL_PERTURB_KEYS> is 2, perl will randomize keys in a repeatable way.
100Repeated runs of the same program should produce the same output every time.
c40c48bb
RS
101The chance that keys changes due to an insert will be very high.
102
f105b7be
KE
103C<PERL_HASH_SEED> implies a non-default C<PERL_PERTURB_KEYS> setting. Setting
104C<PERL_HASH_SEED=0> (exactly one 0) implies C<PERL_PERTURB_KEYS=0> (hash key
105randomization disabled); settng C<PERL_HASH_SEED> to any other value implies
106C<PERL_PERTURB_KEYS=2> (deterministic and repeatable hash key randomization).
107Specifying C<PERL_PERTURB_KEYS> explicitly to a different level overrides this
c40c48bb
RS
108behavior.
109
5ed58cbd
RS
110=head3 Hash::Util::hash_seed() now returns a string
111
112Hash::Util::hash_seed() now returns a string instead of an integer. This
113is to make the infrastructure support hash seeds of arbitrary lengths
114which might exceed that of an integer. (SipHash uses a 16 byte seed).
115
116=head3 Output of PERL_HASH_SEED_DEBUG has been changed
117
118The environment variable PERL_HASH_SEED_DEBUG now makes perl show both the
f105b7be 119hash function perl was built with, I<and> the seed, in hex, in use for that
5ed58cbd
RS
120process. Code parsing this output, should it exist, must change to accommodate
121the new format. Example of the new format:
122
123 $ PERL_HASH_SEED_DEBUG=1 ./perl -e1
124 HASH_FUNCTION = MURMUR3 HASH_SEED = 0x1476bb9f
125
126=head2 Upgrade to Unicode 6.2
127
2e7bc647 128Perl now supports Unicode 6.2. A list of changes from Unicode
5ed58cbd
RS
1296.1 is at L<http://www.unicode.org/versions/Unicode6.2.0>.
130
131=head2 Character name aliases may now include non-Latin1-range characters
132
133It is possible to define your own names for characters for use in
134C<\N{...}>, C<charnames::vianame()>, etc. These names can now be
135comprised of characters from the whole Unicode range. This allows for
136names to be in your native language, and not just English. Certain
137restrictions apply to the characters that may be used (you can't define
138a name that has punctuation in it, for example). See L<charnames/CUSTOM
139ALIASES>.
140
141=head2 New DTrace probes
142
143The following new DTrace probes have been added:
14731ad1 144
337fb649 145=over 4
14731ad1 146
82d98f72 147=item *
14731ad1 148
5ed58cbd
RS
149C<op-entry>
150
151=item *
152
153C<loading-file>
154
155=item *
156
157C<loaded-file>
158
159=back
160
161=head2 C<${^LAST_FH}>
162
163This new variable provides access to the filehandle that was last read.
164This is the handle used by C<$.> and by C<tell> and C<eof> without
165arguments.
166
167=head2 Regular Expression Set Operations
168
169This is an B<experimental> feature to allow matching against the union,
170intersection, etc., of sets of code points, similar to
171L<Unicode::Regex::Set>. It can also be used to extend C</x> processing
172to [bracketed] character classes, and as a replacement of user-defined
173properties, allowing more complex expressions than they do. See
174L<perlrecharclass/Extended Bracketed Character Classes>.
175
176=head2 Lexical subroutines
177
178This new feature is still considered B<experimental>. To enable it:
179
180 use 5.018;
181 no warnings "experimental::lexical_subs";
182 use feature "lexical_subs";
183
184You can now declare subroutines with C<state sub foo>, C<my sub foo>, and
185C<our sub foo>. (C<state sub> requires that the "state" feature be
186enabled, unless you write it as C<CORE::state sub foo>.)
187
188C<state sub> creates a subroutine visible within the lexical scope in which
189it is declared. The subroutine is shared between calls to the outer sub.
190
191C<my sub> declares a lexical subroutine that is created each time the
192enclosing block is entered. C<state sub> is generally slightly faster than
193C<my sub>.
194
195C<our sub> declares a lexical alias to the package subroutine of the same
196name.
197
198For more information, see L<perlsub/Lexical Subroutines>.
199
200=head2 Computed Labels
201
202The loop controls C<next>, C<last> and C<redo>, and the special C<dump>
203operator, now allow arbitrary expressions to be used to compute labels at run
204time. Previously, any argument that was not a constant was treated as the
205empty string.
206
207=head2 More CORE:: subs
208
209Several more built-in functions have been added as subroutines to the
210CORE:: namespace - namely, those non-overridable keywords that can be
211implemented without custom parsers: C<defined>, C<delete>, C<exists>,
212C<glob>, C<pos>, C<protoytpe>, C<scalar>, C<split>, C<study>, and C<undef>.
213
214As some of these have prototypes, C<prototype('CORE::...')> has been
215changed to not make a distinction between overridable and non-overridable
216keywords. This is to make C<prototype('CORE::pos')> consistent with
217C<prototype(&CORE::pos)>.
218
219=head2 C<kill> with negative signal names
220
221C<kill> has always allowed a negative signal number, which kills the
222process group instead of a single process. It has also allowed signal
223names. But it did not behave consistently, because negative signal names
224were treated as 0. Now negative signals names like C<-INT> are supported
225and treated the same way as -2 [perl #112990].
226
227=head1 Security
228
229=head2 C<Storable> security warning in documentation
230
231The documentation for C<Storable> now includes a section which warns readers
232of the danger of accepting Storable documents from untrusted sources. The
233short version is that deserializing certain types of data can lead to loading
234modules and other code execution. This is documented behavior and wanted
235behavior, but this opens an attack vector for malicious entities.
236
237=head2 C<Locale::Maketext> allowed code injection via a malicious template
238
239If users could provide a translation string to Locale::Maketext, this could be
240used to invoke arbitrary Perl subroutines available in the current process.
241
242This has been fixed, but it is still possible to invoke any method provided by
243C<Locale::Maketext> itself or a subclass that you are using. One of these
244methods in turn will invoke the Perl core's C<sprintf> subroutine.
245
246In summary, allowing users to provide translation strings without auditing
247them is a bad idea.
248
249This vulnerability is documented in CVE-2012-6329.
250
251=head2 Avoid calling memset with a negative count
252
253Poorly written perl code that allows an attacker to specify the count to perl's
254C<x> string repeat operator can already cause a memory exhaustion
255denial-of-service attack. A flaw in versions of perl before 5.15.5 can escalate
256that into a heap buffer overrun; coupled with versions of glibc before 2.16, it
257possibly allows the execution of arbitrary code.
258
259The flaw addressed to this commit has been assigned identifier CVE-2012-5195
260and was researched by Tim Brown.
261
262=head1 Incompatible Changes
263
264=head2 See also: hash overhaul
265
266Some of the changes in the L<hash overhaul|/"Hash overhaul"> are not fully
267compatible with previous versions of perl. Please read that section.
268
269=head2 An unknown character name in C<\N{...}> is now a syntax error
270
271Previously, it warned, and the Unicode REPLACEMENT CHARACTER was
272substituted. Unicode now recommends that this situation be a syntax
273error. Also, the previous behavior led to some confusing warnings and
274behaviors, and since the REPLACEMENT CHARACTER has no use other than as
275a stand-in for some unknown character, any code that has this problem is
276buggy.
277
278=head2 Formerly deprecated characters in C<\N{}> character name aliases are now errors.
279
280Since v5.12.0, it has been deprecated to use certain characters in
281user-defined C<\N{...}> character names. These now cause a syntax
282error. For example, it is now an error to begin a name with a digit,
283such as in
284
285 my $undraftable = "\N{4F}"; # Syntax error!
286
287or to have commas anywhere in the name. See L<charnames/CUSTOM ALIASES>
288
289=head2 C<\N{BELL}> now refers to U+1F514 instead of U+0007
290
291Unicode 6.0 reused the name "BELL" for a different code point than it
292traditionally had meant. Since Perl v5.14, use of this name still
293referred to U+0007, but would raise a deprecation warning. Now, "BELL"
294refers to U+1F514, and the name for U+0007 is "ALERT". All the
295functions in L<charnames> have been correspondingly updated.
296
297=head2 New Restrictions in Multi-Character Case-Insensitive Matching in Regular Expression Bracketed Character Classes
298
299Unicode has now withdrawn their previous recommendation for regular
300expressions to automatically handle cases where a single character can
301match multiple characters case-insensitively, for example the letter
302LATIN SMALL LETTER SHARP S and the sequence C<ss>. This is because
303it turns out to be impracticable to do this correctly in all
304circumstances. Because Perl has tried to do this as best it can, it
305will continue to do so. (We are considering an option to turn it off.)
306However, a new restriction is being added on such matches when they
307occur in [bracketed] character classes. People were specifying
308things such as C</[\0-\xff]/i>, and being surprised that it matches the
309two character sequence C<ss> (since LATIN SMALL LETTER SHARP S occurs in
310this range). This behavior is also inconsistent with using a
311property instead of a range: C<\p{Block=Latin1}> also includes LATIN
312SMALL LETTER SHARP S, but C</[\p{Block=Latin1}]/i> does not match C<ss>.
313The new rule is that for there to be a multi-character case-insensitive
314match within a bracketed character class, the character must be
315explicitly listed, and not as an end point of a range. This more
316closely obeys the Principle of Least Astonishment. See
317L<perlrecharclass/Bracketed Character Classes>. Note that a bug [perl
318#89774], now fixed as part of this change, prevented the previous
319behavior from working fully.
320
321=head2 Explicit rules for variable names and identifiers
322
323Due to an oversight, length-one variable names in 5.16 were completely
324unrestricted, and opened the door to several kinds of insanity. As of
3255.18, these now follow the rules of other identifiers, in addition
326to accepting characters that match the C<\p{POSIX_Punct}> property.
327
328There are no longer any differences in the parsing of identifiers
329specified as C<$...> or C<${...}>; previously, they were dealt with in
330different parts of the core, and so had slightly different behavior. For
331instance, C<${foo:bar}> was a legal variable name. Since they are now
332both parsed by the same code, that is no longer the case.
333
334=head2 C<\s> in regular expressions now matches a Vertical Tab
335
336No one could recall why C<\s> didn't match C<\cK>, the vertical tab.
337Now it does. Given the extreme rarity of that character, very little
338breakage is expected.
339
340=head2 C</(?{})/> and C</(??{})/> have been heavily reworked
341
342The implementation of this feature has been almost completely rewritten.
343Although its main intent is to fix bugs, some behaviors, especially
344related to the scope of lexical variables, will have changed. This is
345described more fully in the L</Selected Bug Fixes> section.
346
347=head2 Stricter parsing of substitution replacement
348
349It is no longer possible to abuse the way the parser parses C<s///e> like
350this:
351
352 %_=(_,"Just another ");
353 $_="Perl hacker,\n";
354 s//_}->{_/e;print
355
356=head2 C<given> now aliases the global C<$_>
357
358Instead of assigning to an implicit lexical C<$_>, C<given> now makes the
359global C<$_> an alias for its argument, just like C<foreach>. However, it
360still uses lexical C<$_> if there is lexical C<$_> in scope (again, just like
361C<foreach>) [perl #114020].
362
363=head2 Lexical C<$_> is now experimental
364
365Since it was introduced in Perl 5.10, it has caused much confusion with no
366obvious solution:
367
368=over
829397b9
TC
369
370=item *
371
5ed58cbd
RS
372Various modules (e.g., List::Util) expect callback routines to use the
373global C<$_>. C<use List::Util 'first'; my $_; first { $_ == 1 } @list>
374does not work as one would expect.
71e6aba6
RS
375
376=item *
377
5ed58cbd
RS
378A C<my $_> declaration earlier in the same file can cause confusing closure
379warnings.
71e6aba6
RS
380
381=item *
382
5ed58cbd
RS
383The "_" subroutine prototype character allows called subroutines to access
384your lexical C<$_>, so it is not really private after all.
2426c394 385
71e6aba6
RS
386=item *
387
5ed58cbd
RS
388Nevertheless, subroutines with a "(@)" prototype and methods cannot access
389the caller's lexical C<$_>, unless they are written in XS.
71e6aba6
RS
390
391=item *
392
5ed58cbd
RS
393But even XS routines cannot access a lexical C<$_> declared, not in the
394calling subroutine, but in an outer scope, iff that subroutine happened not
395to mention C<$_> or use any operators that default to C<$_>.
396
397=back
398
399It is our hope that lexical C<$_> can be rehabilitated, but this may
400cause changes in its behavior. Please use it with caution until it
401becomes stable.
402
403=head2 readline() with C<$/ = \N> now reads N characters, not N bytes
404
405Previously, when reading from a stream with I/O layers such as
406C<encoding>, the readline() function, otherwise known as the C<< <> >>
407operator, would read I<N> bytes from the top-most layer. [perl #79960]
408
409Now, I<N> characters are read instead.
410
411There is no change in behaviour when reading from streams with no
412extra layers, since bytes map exactly to characters.
413
414=head2 Overridden C<glob> is now passed one argument
415
416C<glob> overrides used to be passed a magical undocumented second argument
417that identified the caller. Nothing on CPAN was using this, and it got in
418the way of a bug fix, so it was removed. If you really need to identify
419the caller, see L<Devel::Callsite> on CPAN.
420
421=head2 Here-doc parsing
422
423The body of a here-document inside a quote-like operator now always begins
424on the line after the "<<foo" marker. Previously, it was documented to
425begin on the line following the containing quote-like operator, but that
426was only sometimes the case [perl #114040].
427
428=head2 Alphanumeric operators must now be separated from the closing
429delimiter of regular expressions
430
431You may no longer write something like:
432
433 m/a/and 1
434
435Instead you must write
436
437 m/a/ and 1
438
439with whitespace separating the operator from the closing delimiter of
440the regular expression. Not having whitespace has resulted in a
441deprecation warning since Perl v5.14.0.
442
443=head2 qw(...) can no longer be used as parentheses
444
445C<qw> lists used to fool the parser into thinking they were always
446surrounded by parentheses. This permitted some surprising constructions
447such as C<foreach $x qw(a b c) {...}>, which should really be written
448C<foreach $x (qw(a b c)) {...}>. These would sometimes get the lexer into
449the wrong state, so they didn't fully work, and the similar C<foreach qw(a
450b c) {...}> that one might expect to be permitted never worked at all.
451
452This side effect of C<qw> has now been abolished. It has been deprecated
453since Perl 5.13.11. It is now necessary to use real parentheses
454everywhere that the grammar calls for them.
455
456=head2 Interaction of lexical and default warnings
457
458Turning on any lexical warnings used first to disable all default warnings
459if lexical warnings were not already enabled:
460
461 $*; # deprecation warning
462 use warnings "void";
463 $#; # void warning; no deprecation warning
464
f105b7be 465Now, the C<debugging>, C<deprecated>, C<glob>, C<inplace> and C<malloc> warnings
5ed58cbd
RS
466categories are left on when turning on lexical warnings (unless they are
467turned off by C<no warnings>, of course).
468
469This may cause deprecation warnings to occur in code that used to be free
470of warnings.
471
472Those are the only categories consisting only of default warnings. Default
f105b7be
KE
473warnings in other categories are still disabled by C<< use warnings "category" >>,
474as we do not yet have the infrastructure for controlling
5ed58cbd
RS
475individual warnings.
476
477=head2 C<state sub> and C<our sub>
478
479Due to an accident of history, C<state sub> and C<our sub> were equivalent
480to a plain C<sub>, so one could even create an anonymous sub with
481C<our sub { ... }>. These are now disallowed outside of the "lexical_subs"
482feature. Under the "lexical_subs" feature they have new meanings described
483in L<perlsub/Lexical Subroutines>.
484
485=head2 Defined values stored in environment are forced to byte strings
486
487A value stored in an environment variable has always been stringified. In this
488release, it is converted to be only a byte string. First, it is forced to be a
489only a string. Then if the string is utf8 and the equivalent of
490C<utf8::downgrade()> works, that result is used; otherwise, the equivalent of
491C<utf8::encode()> is used, and a warning is issued about wide characters
492(L</Diagnostics>).
493
494=head2 C<require> dies for unreadable files
495
496When C<require> encounters an unreadable file, it now dies. It used to
497ignore the file and continue searching the directories in C<@INC>
498[perl #113422].
499
500=head2 C<gv_fetchmeth_*> and SUPER
501
502The various C<gv_fetchmeth_*> XS functions used to treat a package whose
f105b7be
KE
503named ended with C<::SUPER> specially. A method lookup on the C<Foo::SUPER>
504package would be treated as a C<SUPER> method lookup on the C<Foo> package. This
505is no longer the case. To do a C<SUPER> lookup, pass the C<Foo> stash and the
506C<GV_SUPER> flag.
5ed58cbd
RS
507
508=head2 C<split>'s first argument is more consistently interpreted
509
510After some changes earlier in 5.17, C<split>'s behavior has been
511simplified: if the PATTERN argument evaluates to a literal string
512containing one space, it is treated the way that a I<literal> string
513containing one space once was.
514
515=head1 Deprecations
516
517=head2 Deprecated modules
518
519The following modules will be removed from the core distribution in a
520future release, and should be installed from CPAN instead. Distributions
521on CPAN which require these should add them to their prerequisites.
f105b7be 522The core versions of these modules will issue C<"deprecated">-category
5ed58cbd
RS
523warnings.
524
525You can silence these deprecation warnings by installing the modules
526in question from CPAN.
527
528=over
529
530=item L<Archive::Extract>
531
532=item L<B::Lint>
533
534=item L<B::Lint::Debug>
535
536=item L<CPANPLUS> and all included C<CPANPLUS::*> modules
537
538=item L<Devel::InnerPackage>
539
540=item L<encoding>
541
542=item L<Log::Message>
543
544=item L<Log::Message::Config>
545
546=item L<Log::Message::Handlers>
547
548=item L<Log::Message::Item>
549
550=item L<Log::Message::Simple>
551
552=item L<Module::Pluggable>
553
554=item L<Module::Pluggable::Object>
555
556=item L<Object::Accessor>
557
558=item L<Pod::LaTeX>
559
560=item L<Term::UI>
561
562=item L<Term::UI::History>
563
564=back
565
566=head2 Deprecated Utilities
567
568The following utilities will be removed from the core distribution in a
569future release as their associated modules have been deprecated. They
570will remain available with the applicable CPAN distribution.
571
572=over
573
574=item L<cpanp>
575
576=item C<cpanp-run-perl>
577
578=item L<cpan2dist>
579
580These items are part of the C<CPANPLUS> distribution.
581
582=item L<pod2latex>
583
584This item is part of the C<Pod::LaTeX> distribution.
585
586=back
587
588=head2 PL_sv_objcount
589
590This interpreter-global variable used to track the total number of
591Perl objects in the interpreter. It is no longer maintained and will
592be removed altogether in Perl 5.20.
593
594=head2 Five additional characters should be escaped in patterns with C</x>
595
596When a regular expression pattern is compiled with C</x>, Perl treats 6
597characters as white space to ignore, such as SPACE and TAB. However,
598Unicode recommends 11 characters be treated thusly. We will conform
599with this in a future Perl version. In the meantime, use of any of the
600missing characters will raise a deprecation warning, unless turned off.
601The five characters are:
f105b7be
KE
602
603 U+0085 NEXT LINE,
604 U+200E LEFT-TO-RIGHT MARK,
605 U+200F RIGHT-TO-LEFT MARK,
606 U+2028 LINE SEPARATOR,
607
5ed58cbd 608and
f105b7be
KE
609
610 U+2029 PARAGRAPH SEPARATOR.
5ed58cbd
RS
611
612=head2 User-defined charnames with surprising whitespace
613
614A user-defined character name with trailing or multiple spaces in a row is
615likely a typo. This now generates a warning when defined, on the assumption
616that uses of it will be unlikely to include the excess whitespace.
617
618=head2 Various XS-callable functions are now deprecated
619
620All the functions used to classify characters will be removed from a
621future version of Perl, and should not be used. With participating C
622compilers (e.g., gcc), compiling any file that uses any of these will
623generate a warning. These were not intended for public use; there are
624equivalent, faster, macros for most of them.
2e7bc647 625See L<perlapi/Character classes>. The complete list is:
5ed58cbd
RS
626C<is_uni_alnum>, C<is_uni_alnumc>, C<is_uni_alnumc_lc>,
627C<is_uni_alnum_lc>, C<is_uni_alpha>, C<is_uni_alpha_lc>,
628C<is_uni_ascii>, C<is_uni_ascii_lc>, C<is_uni_blank>,
629C<is_uni_blank_lc>, C<is_uni_cntrl>, C<is_uni_cntrl_lc>,
630C<is_uni_digit>, C<is_uni_digit_lc>, C<is_uni_graph>,
631C<is_uni_graph_lc>, C<is_uni_idfirst>, C<is_uni_idfirst_lc>,
632C<is_uni_lower>, C<is_uni_lower_lc>, C<is_uni_print>,
633C<is_uni_print_lc>, C<is_uni_punct>, C<is_uni_punct_lc>,
634C<is_uni_space>, C<is_uni_space_lc>, C<is_uni_upper>,
635C<is_uni_upper_lc>, C<is_uni_xdigit>, C<is_uni_xdigit_lc>,
636C<is_utf8_alnum>, C<is_utf8_alnumc>, C<is_utf8_alpha>,
637C<is_utf8_ascii>, C<is_utf8_blank>, C<is_utf8_char>,
638C<is_utf8_cntrl>, C<is_utf8_digit>, C<is_utf8_graph>,
639C<is_utf8_idcont>, C<is_utf8_idfirst>, C<is_utf8_lower>,
640C<is_utf8_mark>, C<is_utf8_perl_space>, C<is_utf8_perl_word>,
641C<is_utf8_posix_digit>, C<is_utf8_print>, C<is_utf8_punct>,
642C<is_utf8_space>, C<is_utf8_upper>, C<is_utf8_xdigit>,
643C<is_utf8_xidcont>, C<is_utf8_xidfirst>.
644
645In addition these three functions that have never worked properly are
646deprecated:
647C<to_uni_lower_lc>, C<to_uni_title_lc>, and C<to_uni_upper_lc>.
648
f105b7be 649=head2 Certain rare uses of backslashes within regexes are now deprecated
5ed58cbd
RS
650
651There are three pairs of characters that Perl recognizes as
652metacharacters in regular expression patterns: C<{}>, C<[]>, and C<()>.
653These can be used as well to delimit patterns, as in:
654
655 m{foo}
656 s(foo)(bar)
657
658Since they are metacharacters, they have special meaning to regular
659expression patterns, and it turns out that you can't turn off that
660special meaning by the normal means of preceding them with a backslash,
f105b7be 661if you use them, paired, within a pattern delimited by them. For
5ed58cbd
RS
662example, in
663
664 m{foo\{1,3\}}
665
666the backslashes do not change the behavior, and this matches
667S<C<"f o">> followed by one to three more occurrences of C<"o">.
668
669Usages like this, where they are interpreted as metacharacters, are
670exceedingly rare; we think there are none, for example, in all of CPAN.
671Hence, this deprecation should affect very little code. It does give
672notice, however, that any such code needs to change, which will in turn
673allow us to change the behavior in future Perl versions so that the
674backslashes do have an effect, and without fear that we are silently
675breaking any existing code.
676
d5f315e8
KW
677=head2 Splitting the tokens C<(?> and C<(*> in regular expressions
678
679A deprecation warning is now raised if the C<(> and C<?> are separated
680by white space or comments in C<(?...)> regular expression constructs.
681Similarly, if the C<(> and C<*> are separated in C<(*VERB...)>
682constructs.
683
e0a1dec5
LT
684=head2 Pre-PerlIO IO implementations
685
686Perl supports being built without PerlIO proper, using a stdio or sfio
687wrapper instead. A perl build like this will not support IO layers and
688thus Unicode IO, making it rather handicapped.
689
690PerlIO supports a C<stdio> layer if stdio use is desired, and similarly a
691sfio layer could be produced.
692
5ed58cbd
RS
693=head1 Future Deprecations
694
695=over
71e6aba6
RS
696
697=item *
698
4263dd11 699Platforms without support infrastructure
5ed58cbd
RS
700
701Both Windows CE and z/OS have been historically under-maintained, and are
702currently neither successfully building nor regularly being smoke tested.
703Efforts are underway to change this situation, but it should not be taken for
704granted that the platforms are safe and supported. If they do not become
705buildable and regularly smoked, support for them may be actively removed in
706future releases. If you have an interest in these platforms and you can lend
707your time, expertise, or hardware to help support these platforms, please let
708the perl development effort know by emailing C<perl5-porters@perl.org>.
709
710Some platforms that appear otherwise entirely dead are also on the short list
711for removal between now and 5.20.0:
712
713=over
714
715=item DG/UX
716
717=item NeXT
718
719=back
1993add8
RS
720
721=item *
722
5ed58cbd
RS
723Swapping of $< and $>
724
725For more information about this future deprecation, see L<the relevant RT
726ticket|https://rt.perl.org/rt3/Ticket/Display.html?id=96212>.
71e6aba6
RS
727
728=item *
729
5ed58cbd 730C<microperl>, long broken and of unclear present purpose, will be removed.
71e6aba6
RS
731
732=item *
733
5ed58cbd
RS
734Revamping C<< "\Q" >> semantics in double-quotish strings when combined with
735other escapes.
736
737There are several bugs and inconsistencies involving combinations
738of C<\Q> and escapes like C<\x>, C<\L>, etc., within a C<\Q...\E> pair.
739These need to be fixed, and doing so will necessarily change current
740behavior. The changes have not yet been settled.
71e6aba6 741
d5f315e8
KW
742=item *
743
744Use of C<$^>, where C<^> stands for any actual (non-printing) C0 control
745character will be disallowed in a future Perl version. Use C<${^}>
746instead (where again C<^> stands for a control character),
747or better, C<$^A> , where C<^> this time is a caret (CIRCUMFLEX ACCENT),
748and C<A> stands for any of the characters listed at the end of
749L<perlebcdic/OPERATOR DIFFERENCES>.
750
337fb649 751=back
2426c394 752
5ed58cbd
RS
753=head1 Performance Enhancements
754
755=over 4
2426c394 756
5ed58cbd 757=item *
2426c394 758
5ed58cbd
RS
759Lists of lexical variable declarations (C<my($x, $y)>) are now optimised
760down to a single op and are hence faster than before.
2426c394 761
5ed58cbd 762=item *
2426c394 763
5ed58cbd
RS
764A new C preprocessor define C<NO_TAINT_SUPPORT> was added that, if set,
765disables Perl's taint support altogether. Using the -T or -t command
766line flags will cause a fatal error. Beware that both core tests as
767well as many a CPAN distribution's tests will fail with this change. On
768the upside, it provides a small performance benefit due to reduced
769branching.
2426c394 770
5ed58cbd
RS
771B<Do not enable this unless you know exactly what you are getting yourself
772into.>
773
774=item *
775
776C<pack> with constant arguments is now constant folded in most cases
777[perl #113470].
778
779=item *
780
781Speed up in regular expression matching against Unicode properties. The
782largest gain is for C<\X>, the Unicode "extended grapheme cluster." The
783gain for it is about 35% - 40%. Bracketed character classes, e.g.,
784C<[0-9\x{100}]> containing code points above 255 are also now faster.
785
786=item *
787
788On platforms supporting it, several former macros are now implemented as static
789inline functions. This should speed things up slightly on non-GCC platforms.
790
791=item *
792
66f62cf6
RS
793The optimisation of hashes in boolean context has been extended to
794affect C<scalar(%hash)>, C<%hash ? ... : ...>, and C<sub { %hash || ... }>.
5ed58cbd
RS
795
796=item *
797
f105b7be 798Filetest operators manage the stack in a fractionally more efficient manner.
5ed58cbd
RS
799
800=item *
801
802Globs used in a numeric context are now numified directly in most cases,
f105b7be 803rather than being numified via stringification.
5ed58cbd
RS
804
805=item *
806
807The C<x> repetition operator is now folded to a single constant at compile
808time if called in scalar context with constant operands and no parentheses
809around the left operand.
810
811=back
812
813=head1 Modules and Pragmata
814
815=head2 New Modules and Pragmata
2426c394 816
337fb649 817=over 4
982110e0 818
82d98f72 819=item *
2426c394 820
5ed58cbd
RS
821L<Config::Perl::V> version 0.16 has been added as a dual-lifed module.
822It provides structured data retrieval of C<perl -V> output including
823information only known to the C<perl> binary and not available via L<Config>.
824
825=back
826
827=head2 Updated Modules and Pragmata
828
829This is only an overview of selected module updates. For a complete
830list of updates, run:
831
f105b7be 832 $ corelist --diff 5.16.0 5.18.0
5ed58cbd 833
f105b7be 834You can substitute your favorite version in place of 5.16.0, too.
5ed58cbd
RS
835
836=over 4
33392251
BF
837
838=item *
839
5ed58cbd
RS
840L<XXX> has been upgraded from version A.xx to B.yy.
841
842=back
843
844=head2 Removed Modules and Pragmata
845
846=over
33392251
BF
847
848=item *
849
5ed58cbd
RS
850L<Version::Requirements> has been removed from the core distribution. It is
851available under a different name: L<CPAN::Meta::Requirements>.
2426c394 852
337fb649 853=back
2426c394 854
5ed58cbd 855=head1 Documentation
19718730 856
5ed58cbd
RS
857=head2 Changes to Existing Documentation
858
859=head3 L<perlcheat>
82d98f72 860
5a6a30f4 861=over 4
b7c7d786 862
5ed58cbd
RS
863=item *
864
865L<perlcheat> has been reorganized, and a few new sections were added.
866
867=back
868
869=head3 L<perldata>
870
871=over 4
82d98f72 872
5ed58cbd 873=item *
d2d1e842 874
5ed58cbd
RS
875Now explicitly documents the behaviour of hash initializer lists that
876contain duplicate keys.
f355e93d 877
5a6a30f4 878=back
f355e93d 879
5ed58cbd 880=head3 L<perldiag>
19718730 881
19718730 882=over 4
e14ac59b 883
5ed58cbd
RS
884=item *
885
886The explanation of symbolic references being prevented by "strict refs"
887now doesn't assume that the reader knows what symbolic references are.
888
889=back
9f351b45 890
5ed58cbd 891=head3 L<perlfaq>
9f351b45 892
5ed58cbd 893=over 4
9f351b45 894
5ed58cbd 895=item *
7cf3104f 896
5ed58cbd 897L<perlfaq> has been synchronized with version 5.0150040 from CPAN.
12719193 898
6253ee75 899=back
216cf7fc 900
5ed58cbd 901=head3 L<perlfunc>
f5b73711 902
5ed58cbd
RS
903=over 4
904
905=item *
a75569c0 906
5ed58cbd 907The return value of C<pipe> is now documented.
a75569c0 908
5ed58cbd 909=item *
a75569c0 910
5ed58cbd
RS
911Clarified documentation of C<our>.
912
913=back
914
915=head3 L<perlop>
916
917=over 4
918
919=item *
920
921Loop control verbs (C<dump>, C<goto>, C<next>, C<last> and C<redo>) have always
922had the same precedence as assignment operators, but this was not documented
923until now.
924
925=back
926
927=head3 Diagnostics
928
929The following additions or changes have been made to diagnostic output,
930including warnings and fatal error messages. For the complete list of
931diagnostic messages, see L<perldiag>.
932
933XXX New or changed warnings emitted by the core's C<C> code go here. Also
934include any changes in L<perldiag> that reconcile it to the C<C> code.
935
936=head2 New Diagnostics
937
938XXX Newly added diagnostic messages go under here, separated into New Errors
939and New Warnings
940
941=head3 New Errors
942
943=over 4
944
945=item *
946
947L<Unterminated delimiter for here document|perldiag/"Unterminated delimiter for here document">
948
949This message now occurs when a here document label has an initial quotation
950mark but the final quotation mark is missing.
951
952This replaces a bogus and misleading error message about not finding the label
953itself [perl #114104].
954
955=item *
956
957L<panic: child pseudo-process was never scheduled|perldiag/"panic: child pseudo-process was never scheduled">
958
959This error is thrown when a child pseudo-process in the ithreads implementation
960on Windows was not scheduled within the time period allowed and therefore was
961not able to initialize properly [perl #88840].
962
963=item *
964
965L<Group name must start with a non-digit word character in regex; marked by <-- HERE in mE<sol>%sE<sol>|perldiag/"Group name must start with a non-digit word character in regex; marked by <-- HERE in m/%s/">
966
967This error has been added for C<(?&0)>, which is invalid. It used to
968produce an incomprehensible error message [perl #101666].
969
970=item *
971
972L<Can't use an undefined value as a subroutine reference|perldiag/"Can't use an undefined value as %s reference">
973
974Calling an undefined value as a subroutine now produces this error message.
975It used to, but was accidentally disabled, first in Perl 5.004 for
976non-magical variables, and then in Perl 5.14 for magical (e.g., tied)
977variables. It has now been restored. In the mean time, undef was treated
978as an empty string [perl #113576].
979
980=item *
981
982L<Experimental "%s" subs not enabled|perldiag/"Experimental "%s" subs not enabled">
983
984To use lexical subs, you must first enable them:
985
986 no warnings 'experimental::lexical_subs';
987 use feature 'lexical_subs';
988 my sub foo { ... }
989
990=back
991
992=head3 New Warnings
993
994=over 4
995
996=item *
997
998XXX: This needs more detail.
999
1000Strings with code points over 0xFF may not be mapped into in-memory file
1001handles
1002
1003=item *
1004
1005L<'%s' resolved to '\o{%s}%d'|perldiag/"'%s' resolved to '\o{%s}%d'">
1006
1007=item *
1008
1009L<'Trailing white-space in a charnames alias definition is deprecated'|perldiag/"Trailing white-space in a charnames alias definition is deprecated">
1010
1011=item *
1012
1013L<'A sequence of multiple spaces in a charnames alias definition is deprecated'|perldiag/"A sequence of multiple spaces in a charnames alias definition is deprecated">
1014
1015=item *
1016
1017L<'Passing malformed UTF-8 to "%s" is deprecated'|perldiag/"Passing malformed UTF-8 to "%s" is deprecated">
1018
1019=item *
1020
1021L<Subroutine "&%s" is not available|perldiag/"Subroutine "&%s" is not available">
1022
1023(W closure) During compilation, an inner named subroutine or eval is
1024attempting to capture an outer lexical subroutine that is not currently
1025available. This can happen for one of two reasons. First, the lexical
1026subroutine may be declared in an outer anonymous subroutine that has not
1027yet been created. (Remember that named subs are created at compile time,
1028while anonymous subs are created at run-time.) For example,
1029
1030 sub { my sub a {...} sub f { \&a } }
1031
1032At the time that f is created, it can't capture the current the "a" sub,
1033since the anonymous subroutine hasn't been created yet. Conversely, the
1034following won't give a warning since the anonymous subroutine has by now
1035been created and is live:
1036
1037 sub { my sub a {...} eval 'sub f { \&a }' }->();
1038
1039The second situation is caused by an eval accessing a variable that has
1040gone out of scope, for example,
1041
1042 sub f {
1043 my sub a {...}
1044 sub { eval '\&a' }
1045 }
1046 f()->();
1047
1048Here, when the '\&a' in the eval is being compiled, f() is not currently
1049being executed, so its &a is not available for capture.
1050
1051=item *
1052
1053L<"%s" subroutine &%s masks earlier declaration in same %s|perldiag/"%s" subroutine &%s masks earlier declaration in same %s>
1054
1055(W misc) A "my" or "state" subroutine has been redeclared in the
1056current scope or statement, effectively eliminating all access to
1057the previous instance. This is almost always a typographical error.
1058Note that the earlier subroutine will still exist until the end of
1059the scope or until all closure references to it are destroyed.
1060
1061=item *
1062
1063L<The %s feature is experimental|perldiag/"The %s feature is experimental">
1064
1065(S experimental) This warning is emitted if you enable an experimental
1066feature via C<use feature>. Simply suppress the warning if you want
1067to use the feature, but know that in doing so you are taking the risk
1068of using an experimental feature which may change or be removed in a
1069future Perl version:
1070
1071 no warnings "experimental::lexical_subs";
1072 use feature "lexical_subs";
1073
1074=item *
1075
1076L<sleep(%u) too large|perldiag/"sleep(%u) too large">
1077
1078(W overflow) You called C<sleep> with a number that was larger than it can
1079reliably handle and C<sleep> probably slept for less time than requested.
1080
1081=item *
1082
1083L<Wide character in setenv|perldiag/"Wide character in %s">
1084
1085Attempts to put wide characters into environment variables via C<%ENV> now
1086provoke this warning.
1087
1088=item *
1089
1090"L<Invalid negative number (%s) in chr|perldiag/"Invalid negative number (%s) in chr">"
1091
1092C<chr()> now warns when passed a negative value [perl #83048].
1093
1094=item *
1095
1096"L<Integer overflow in srand|perldiag/"Integer overflow in srand">"
1097
1098C<srand()> now warns when passed a value that doesn't fit in a C<UV> (since the
1099value will be truncated rather than overflowing) [perl #40605].
1100
1101=item *
1102
1103"L<-i used with no filenames on the command line, reading from STDIN|perldiag/"-i used with no filenames on the command line, reading from STDIN">"
1104
1105Running perl with the C<-i> flag now warns if no input files are provided on
1106the command line [perl #113410].
1107
1108=back
1109
1110=head2 Changes to Existing Diagnostics
1111
1112=over 4
1113
1114=item *
1115
1116L<$* is no longer supported|perldiag/"$* is no longer supported">
1117
1118The warning that use of C<$*> and C<$#> is no longer supported is now
1119generated for every location that references them. Previously it would fail
1120to be generated if another variable using the same typeglob was seen first
1121(e.g. C<@*> before C<$*>), and would not be generated for the second and
1122subsequent uses. (It's hard to fix the failure to generate warnings at all
1123without also generating them every time, and warning every time is
1124consistent with the warnings that C<$[> used to generate.)
1125
1126=item *
1127
1128The warnings for C<\b{> and C<\B{> were added. They are a deprecation
1129warning which should be turned off by that category. One should not
1130have to turn off regular regexp warnings as well to get rid of these.
1131
1132=item *
1133
1134L<Constant(%s): Call to &{$^H{%s}} did not return a defined value|perldiag/Constant(%s): Call to &{$^H{%s}} did not return a defined value>
1135
1136Constant overloading that returns C<undef> results in this error message.
1137For numeric constants, it used to say "Constant(undef)". "undef" has been
1138replaced with the number itself.
1139
1140=item *
1141
1142The error produced when a module cannot be loaded now includes a hint that
1143the module may need to be installed: "Can't locate hopping.pm in @INC (you
1144may need to install the hopping module) (@INC contains: ...)"
1145
1146=item *
1147
1148L<vector argument not supported with alpha versions|perldiag/vector argument not supported with alpha versions>
1149
1150This warning was not suppressable, even with C<no warnings>. Now it is
1151suppressible, and has been moved from the "internal" category to the
1152"printf" category.
1153
1154=item *
1155
1156C<< Can't do {n,m} with n > m in regex; marked by <-- HERE in m/%s/ >>
1157
1158This fatal error has been turned into a warning that reads:
1159
1160L<< Quantifier {n,m} with n > m can't match in regex | perldiag/Quantifier {n,m} with n > m can't match in regex >>
1161
1162(W regexp) Minima should be less than or equal to maxima. If you really want
1163your regexp to match something 0 times, just put {0}.
1164
1165=item *
1166
1167The "Runaway prototype" warning that occurs in bizarre cases has been
1168removed as being unhelpful and inconsistent.
1169
1170=item *
1171
1172The "Not a format reference" error has been removed, as the only case in
1173which it could be triggered was a bug.
1174
1175=item *
1176
1177The "Unable to create sub named %s" error has been removed for the same
1178reason.
1179
1180=item *
1181
1182The 'Can't use "my %s" in sort comparison' error has been downgraded to a
1183warning, '"my %s" used in sort comparison' (with 'state' instead of 'my'
1184for state variables). In addition, the heuristics for guessing whether
1185lexical $a or $b has been misused have been improved to generate fewer
1186false positives. Lexical $a and $b are no longer disallowed if they are
1187outside the sort block. Also, a named unary or list operator inside the
1188sort block no longer causes the $a or $b to be ignored [perl #86136].
1189
1190=back
1191
1192=head1 Utility Changes
1193
1194=head3 L<h2xs>
1195
1196=over 4
1197
1198=item *
1199
1200F<h2xs> no longer produces invalid code for empty defines. [perl #20636]
1201
1202=back
1203
1204=head1 Configuration and Compilation
1205
1206=over 4
1207
1208=item *
1209
1210Added C<useversionedarchname> option to Configure
1211
1212When set, it includes 'api_versionstring' in 'archname'. E.g.
1213x86_64-linux-5.13.6-thread-multi. It is unset by default.
1214
1215This feature was requested by Tim Bunce, who observed that
f105b7be 1216C<INSTALL_BASE> creates a library structure that does not
5ed58cbd
RS
1217differentiate by perl version. Instead, it places architecture
1218specific files in "$install_base/lib/perl5/$archname". This makes
f105b7be 1219it difficult to use a common C<INSTALL_BASE> library path with
5ed58cbd
RS
1220multiple versions of perl.
1221
f105b7be 1222By setting C<-Duseversionedarchname>, the $archname will be
c2959982 1223distinct for architecture I<and> API version, allowing mixed use of
f105b7be 1224C<INSTALL_BASE>.
5ed58cbd
RS
1225
1226=item *
1227
ff772877
RS
1228Add a C<PERL_NO_INLINE_FUNCTIONS> option
1229
f105b7be 1230If C<PERL_NO_INLINE_FUNCTIONS> is defined, don't include "inline.h"
ff772877
RS
1231
1232This permits test code to include the perl headers for definitions without
1233creating a link dependency on the perl library (which may not exist yet).
1234
1235=item *
1236
5ed58cbd
RS
1237Configure will honour the external C<MAILDOMAIN> environment variable, if set.
1238
1239=item *
1240
1241C<installman> no longer ignores the silent option
1242
1243=item *
1244
1245Both C<META.yml> and C<META.json> files are now included in the distribution.
1246
1247=item *
1248
1249F<Configure> will now correctly detect C<isblank()> when compiling with a C++
1250compiler.
1251
1252=item *
1253
1254The pager detection in F<Configure> has been improved to allow responses which
1255specify options after the program name, e.g. B</usr/bin/less -R>, if the user
1256accepts the default value. This helps B<perldoc> when handling ANSI escapes
1257[perl #72156].
1258
1259=back
1260
1261=head1 Testing
1262
1263=over 4
1264
1265=item *
1266
1267The test suite now has a section for tests that require very large amounts
1268of memory. These tests won't run by default; they can be enabled by
1269setting the C<PERL_TEST_MEMORY> environment variable to the number of
1270gibibytes of memory that may be safely used.
1271
1272=back
1273
1274=head1 Platform Support
1275
1276=head2 Discontinued Platforms
1277
1278=over 4
1279
1280=item BeOS
1281
1282BeOS was an operating system for personal computers developed by Be Inc,
1283initially for their BeBox hardware. The OS Haiku was written as an open
1284source replacement for/continuation of BeOS, and its perl port is current and
1285actively maintained.
1286
1287=item UTS Global
1288
1289Support code relating to UTS global has been removed. UTS was a mainframe
1290version of System V created by Amdahl, subsequently sold to UTS Global. The
1291port has not been touched since before Perl 5.8.0, and UTS Global is now
1292defunct.
1293
1294=item VM/ESA
1295
1296Support for VM/ESA has been removed. The port was tested on 2.3.0, which
1297IBM ended service on in March 2002. 2.4.0 ended service in June 2003, and
1298was superseded by Z/VM. The current version of Z/VM is V6.2.0, and scheduled
1299for end of service on 2015/04/30.
1300
1301=item MPE/IX
1302
1303Support for MPE/IX has been removed.
1304
1305=item EPOC
1306
1307Support code relating to EPOC has been removed. EPOC was a family of
1308operating systems developed by Psion for mobile devices. It was the
1309predecessor of Symbian. The port was last updated in April 2002.
1310
1311=item Rhapsody
1312
1313Support for Rhapsody has been removed.
1314
1315=back
1316
1317=head2 Platform-Specific Notes
1318
1319=head3 AIX
1320
1321Configure now always adds C<-qlanglvl=extc99> to the CC flags on AIX when
1322using xlC. This will make it easier to compile a number of XS-based modules
1323that assume C99 [perl #113778].
1324
1325=head3 clang++
1326
1327There is now a workaround for a compiler bug that prevented compiling
1328with clang++ since Perl 5.15.7 [perl #112786].
1329
1330=head3 C++
1331
1332When compiling the Perl core as C++ (which is only semi-supported), the
1333mathom functions are now compiled as C<extern "C">, to ensure proper
1334binary compatibility. (However, binary compatibility isn't generally
1335guaranteed anyway in the situations where this would matter.)
1336
1337=head3 Darwin
1338
1339Stop hardcoding an alignment on 8 byte boundaries to fix builds using
1340-Dusemorebits.
1341
1342=head3 Haiku
1343
1344Perl should now work out of the box on Haiku R1 Alpha 4.
1345
1346=head3 MidnightBSD
1347
1348C<libc_r> was removed from recent versions of MidnightBSD and older versions
1349work better with C<pthread>. Threading is now enabled using C<pthread> which
1350corrects build errors with threading enabled on 0.4-CURRENT.
1351
1352=head3 Solaris
1353
1354In Configure, avoid running sed commands with flags not supported on Solaris.
1355
1356=head3 VMS
1357
1358=over
1359
1360=item *
1361
1362Where possible, the case of filenames and command-line arguments is now
1363preserved by enabling the CRTL features C<DECC$EFS_CASE_PRESERVE> and
1364C<DECC$ARGV_PARSE_STYLE> at start-up time. The latter only takes effect
1365when extended parse is enabled in the process from which Perl is run.
1366
1367=item *
1368
1369The character set for Extended Filename Syntax (EFS) is now enabled by default
1370on VMS. Among other things, this provides better handling of dots in directory
05f5908f 1371names, multiple dots in filenames, and spaces in filenames. To obtain the old
5ed58cbd
RS
1372behavior, set the logical name C<DECC$EFS_CHARSET> to C<DISABLE>.
1373
1374=item *
1375
05f5908f 1376Fixed linking on builds configured with C<-Dusemymalloc=y>.
5ed58cbd
RS
1377
1378=item *
1379
05f5908f
CB
1380Experimental support for building Perl with the HP C++ compiler is available
1381by configuring with C<-Dusecxx>.
5ed58cbd
RS
1382
1383=item *
1384
1385All C header files from the top-level directory of the distribution are now
1386installed on VMS, providing consistency with a long-standing practice on other
1387platforms. Previously only a subset were installed, which broke non-core
1388extension builds for extensions that depended on the missing include files.
1389
1390=item *
1391
1392Quotes are now removed from the command verb (but not the parameters) for
1393commands spawned via C<system>, backticks, or a piped C<open>. Previously,
1394quotes on the verb were passed through to DCL, which would fail to recognize
1395the command. Also, if the verb is actually a path to an image or command
1396procedure on an ODS-5 volume, quoting it now allows the path to contain spaces.
1397
1398=item *
1399
1400The B<a2p> build has been fixed for the HP C++ compiler on OpenVMS.
1401
1402=back
1403
1404=head3 Win32
1405
1406=over
1407
1408=item *
1409
1410Perl can now be built using Microsoft's Visual C++ 2012 compiler by specifying
1411CCTYPE=MSVC110 (or MSVC110FREE if you are using the free Express edition for
1412Windows Desktop) in F<win32/Makefile>.
1413
1414=item *
1415
f105b7be 1416The option to build without C<USE_SOCKETS_AS_HANDLES> has been removed.
5ed58cbd
RS
1417
1418=item *
1419
1420Fixed a problem where perl could crash while cleaning up threads (including the
1421main thread) in threaded debugging builds on Win32 and possibly other platforms
1422[perl #114496].
1423
1424=item *
1425
1426A rare race condition that would lead to L<sleep|perlfunc/sleep> taking more
1427time than requested, and possibly even hanging, has been fixed [perl #33096].
1428
1429=item *
1430
1431C<link> on Win32 now attempts to set C<$!> to more appropriate values
1432based on the Win32 API error code. [perl #112272]
1433
1434Perl no longer mangles the environment block, e.g. when launching a new
1435sub-process, when the environment contains non-ASCII characters. Known
1436problems still remain, however, when the environment contains characters
1437outside of the current ANSI codepage (e.g. see the item about Unicode in
1438C<%ENV> in L<http://perl5.git.perl.org/perl.git/blob/HEAD:/Porting/todo.pod>).
1439[perl #113536]
1440
1441=item *
1442
1443Building perl with some Windows compilers used to fail due to a problem
1444with miniperl's C<glob> operator (which uses the C<perlglob> program)
1445deleting the PATH environment variable [perl #113798].
1446
1447=item *
1448
f105b7be 1449A new makefile option, C<USE_64_BIT_INT>, has been added to the Windows
5ed58cbd
RS
1450makefiles. Set this to "define" when building a 32-bit perl if you want
1451it to use 64-bit integers.
1452
1453Machine code size reductions, already made to the DLLs of XS modules in
1454Perl 5.17.2, have now been extended to the perl DLL itself.
1455
1456Building with VC++ 6.0 was inadvertently broken in Perl 5.17.2 but has
1457now been fixed again.
1458
1459=back
1460
1461=head3 WinCE
1462
1463Building on WinCE is now possible once again, although more work is required
1464to fully restore a clean build.
1465
1466=head1 Internal Changes
1467
1468=over
1469
1470=item *
1471
4263dd11 1472Synonyms for the misleadingly named C<av_len()> have been created:
5ed58cbd
RS
1473C<av_top_index()> and C<av_tindex>. All three of these return the
1474number of the highest index in the array, not the number of elements it
1475contains.
1476
1477=item *
1478
1479SvUPGRADE() is no longer an expression. Originally this macro (and its
1480underlying function, sv_upgrade()) were documented as boolean, although
1481in reality they always croaked on error and never returned false. In 2005
1482the documentation was updated to specify a void return value, but
1483SvUPGRADE() was left always returning 1 for backwards compatibility. This
1484has now been removed, and SvUPGRADE() is now a statement with no return
1485value.
1486
1487So this is now a syntax error:
1488
1489 if (!SvUPGRADE(sv)) { croak(...); }
1490
1491If you have code like that, simply replace it with
1492
1493 SvUPGRADE(sv);
1494
1495or to to avoid compiler warnings with older perls, possibly
1496
1497 (void)SvUPGRADE(sv);
1498
1499=item *
1500
1501Perl has a new copy-on-write mechanism that allows any SvPOK scalar to be
1502upgraded to a copy-on-write scalar. A reference count on the string buffer
1503is stored in the string buffer itself.
1504
1505This breaks a few XS modules by allowing copy-on-write scalars to go
1506through code paths that never encountered them before.
1507
1508This behaviour can still be disabled by running F<Configure> with
1509B<-Accflags=-DPERL_NO_COW>. This option will probably be removed in Perl
15105.20.
1511
1512=item *
1513
1514Copy-on-write no longer uses the SvFAKE and SvREADONLY flags. Hence,
1515SvREADONLY indicates a true read-only SV.
1516
1517Use the SvIsCOW macro (as before) to identify a copy-on-write scalar.
1518
1519=item *
1520
f105b7be 1521C<PL_glob_index> is gone.
5ed58cbd
RS
1522
1523=item *
1524
1525The private Perl_croak_no_modify has had its context parameter removed. It is
1526now has a void prototype. Users of the public API croak_no_modify remain
1527unaffected.
1528
1529=item *
1530
1531Copy-on-write (shared hash key) scalars are no longer marked read-only.
1532C<SvREADONLY> returns false on such an SV, but C<SvIsCOW> still returns
1533true.
1534
1535=item *
1536
1537A new op type, C<OP_PADRANGE> has been introduced. The perl peephole
1538optimiser will, where possible, substitute a single padrange op for a
1539pushmark followed by one or more pad ops, and possibly also skipping list
1540and nextstate ops. In addition, the op can carry out the tasks associated
f105b7be 1541with the RHS of a C<< my(...) = @_ >> assignment, so those ops may be optimised
5ed58cbd
RS
1542away too.
1543
1544=item *
1545
1546Case-insensitive matching inside a [bracketed] character class with a
1547multi-character fold no longer excludes one of the possibilities in the
1548circumstances that it used to. [perl #89774].
1549
1550=item *
1551
1552C<PL_formfeed> has been removed.
1553
1554=item *
1555
1556The regular expression engine no longer reads one byte past the end of the
1557target string. While for all internally well-formed scalars this should
1558never have been a problem, this change facilitates clever tricks with
1559string buffers in CPAN modules. [perl #73542]
1560
1561=item *
1562
1563Inside a BEGIN block, C<PL_compcv> now points to the currently-compiling
1564subroutine, rather than the BEGIN block itself.
1565
1566=item *
1567
1568C<mg_length> has been deprecated.
1569
1570=item *
1571
1572C<sv_len> now always returns a byte count and C<sv_len_utf8> a character
1573count. Previously, C<sv_len> and C<sv_len_utf8> were both buggy and would
1574sometimes returns bytes and sometimes characters. C<sv_len_utf8> no longer
1575assumes that its argument is in UTF8. Neither of these creates UTF8 caches
1576for tied or overloaded values or for non-PVs any more.
1577
1578=item *
1579
1580C<sv_mortalcopy> now copies string buffers of shared hash key scalars when
1581called from XS modules [perl #79824].
1582
1583=item *
1584
1585C<RXf_SPLIT> and C<RXf_SKIPWHITE> are no longer used. They are now
1586#defined as 0.
1587
1588=item *
1589
1590The new C<RXf_MODIFIES_VARS> flag can be set by custom regular expression
1591engines to indicate that the execution of the regular expression may cause
1592variables to be modified. This lets C<s///> know to skip certain
1593optimisations. Perl's own regular expression engine sets this flag for the
1594special backtracking verbs that set $REGMARK and $REGERROR.
1595
1596=item *
1597
1598The APIs for accessing lexical pads have changed considerably.
1599
1600C<PADLIST>s are now longer C<AV>s, but their own type instead.
1601C<PADLIST>s now contain a C<PAD> and a C<PADNAMELIST> of C<PADNAME>s,
1602rather than C<AV>s for the pad and the list of pad names. C<PAD>s,
1603C<PADNAMELIST>s, and C<PADNAME>s are to be accessed as such through the
1604newly added pad API instead of the plain C<AV> and C<SV> APIs. See
1605L<perlapi> for details.
1606
1607=item *
1608
1609In the regex API, the numbered capture callbacks are passed an index
1610indicating what match variable is being accessed. There are special
1611index values for the C<$`, $&, $&> variables. Previously the same three
1612values were used to retrieve C<${^PREMATCH}, ${^MATCH}, ${^POSTMATCH}>
1613too, but these have now been assigned three separate values. See
1614L<perlreapi/Numbered capture callbacks>.
1615
1616=item *
1617
1618C<PL_sawampersand> was previously a boolean indicating that any of
1619C<$`, $&, $&> had been seen; it now contains three one-bit flags
1620indicating the presence of each of the variables individually.
1621
1622=item *
1623
1624The C<CV *> typemap entry now supports C<&{}> overloading and typeglobs,
1625just like C<&{...}> [perl #96872].
1626
1627=item *
1628
1629The C<SVf_AMAGIC> flag to indicate overloading is now on the stash, not the
1630object. It is now set automatically whenever a method or @ISA changes, so
1631its meaning has changed, too. It now means "potentially overloaded". When
1632the overload table is calculated, the flag is automatically turned off if
1633there is no overloading, so there should be no noticeable slowdown.
1634
1635The staleness of the overload tables is now checked when overload methods
1636are invoked, rather than during C<bless>.
1637
1638"A" magic is gone. The changes to the handling of the C<SVf_AMAGIC> flag
1639eliminate the need for it.
1640
1641C<PL_amagic_generation> has been removed as no longer necessary. For XS
1642modules, it is now a macro alias to C<PL_na>.
1643
1644The fallback overload setting is now stored in a stash entry separate from
1645overloadedness itself.
1646
1647=item *
1648
1649The character-processing code has been cleaned up in places. The changes
1650should be operationally invisible.
1651
1652=item *
1653
1654The C<study> function was made a no-op in 5.16. It was simply disabled via
1655a C<return> statement; the code was left in place. Now the code supporting
1656what C<study> used to do has been removed.
1657
1658=item *
1659
1660Under threaded perls, there is no longer a separate PV allocated for every
1661COP to store its package name (C<< cop->stashpv >>). Instead, there is an
1662offset (C<< cop->stashoff >>) into the new C<PL_stashpad> array, which
1663holds stash pointers.
1664
1665=item *
1666
1667In the pluggable regex API, the C<regexp_engine> struct has acquired a new
1668field C<op_comp>, which is currently just for perl's internal use, and
f105b7be 1669should be initialized to NULL by other regex plugin modules.
5ed58cbd
RS
1670
1671=item *
1672
1673A new function C<alloccoptash> has been added to the API, but is considered
1674experimental. See L<perlapi>.
1675
1676=item *
1677
1678Perl used to implement get magic in a way that would sometimes hide bugs in
4263dd11 1679code that could call mg_get() too many times on magical values. This hiding of
5ed58cbd
RS
1680errors no longer occurs, so long-standing bugs may become visible now. If
1681you see magic-related errors in XS code, check to make sure it, together
1682with the Perl API functions it uses, calls mg_get() only once on SvGMAGICAL()
1683values.
1684
1685=item *
1686
1687OP allocation for CVs now uses a slab allocator. This simplifies
1688memory management for OPs allocated to a CV, so cleaning up after a
1689compilation error is simpler and safer [perl #111462][perl #112312].
1690
1691=item *
1692
f105b7be 1693C<PERL_DEBUG_READONLY_OPS> has been rewritten to work with the new slab
5ed58cbd
RS
1694allocator, allowing it to catch more violations than before.
1695
1696=item *
1697
f105b7be
KE
1698The old slab allocator for ops, which was only enabled for C<PERL_IMPLICIT_SYS>
1699and C<PERL_DEBUG_READONLY_OPS>, has been retired.
5ed58cbd
RS
1700
1701=back
1702
1703=head1 Selected Bug Fixes
1704
1705=over 4
1706
1707=item *
1708
1709Here-doc terminators no longer require a terminating newline character when
1710they occur at the end of a file. This was already the case at the end of a
1711string eval [perl #65838].
1712
1713=item *
1714
f105b7be 1715C<-DPERL_GLOBAL_STRUCT> builds now free the global struct B<after>
5ed58cbd
RS
1716they've finished using it.
1717
1718=item *
1719
1720A trailing '/' on a path in @INC will no longer have an additional '/'
1721appended.
1722
1723=item *
1724
1725The C<:crlf> layer now works when unread data doesn't fit into its own
1726buffer. [perl #112244].
1727
1728=item *
1729
1730C<ungetc()> now handles UTF-8 encoded data. [perl #116322].
1731
1732=item *
1733
1734A bug in the core typemap caused any C types that map to the T_BOOL core
1735typemap entry to not be set, updated, or modified when the T_BOOL variable was
1736used in an OUTPUT: section with an exception for RETVAL. T_BOOL in an INPUT:
1737section was not affected. Using a T_BOOL return type for an XSUB (RETVAL)
1738was not affected. A side effect of fixing this bug is, if a T_BOOL is specified
1739in the OUTPUT: section (which previous did nothing to the SV), and a read only
1740SV (literal) is passed to the XSUB, croaks like "Modification of a read-only
1741value attempted" will happen. [perl #115796]
1742
1743=item *
1744
1745On many platforms, providing a directory name as the script name caused perl
1746to do nothing and report success. It should now universally report an error
1747and exit nonzero. [perl #61362]
1748
1749=item *
1750
1751C<sort {undef} ...> under fatal warnings no longer crashes. It had
1752begun crashing in Perl 5.16.
1753
1754=item *
1755
1756Stashes blessed into each other
1757(C<bless \%Foo::, 'Bar'; bless \%Bar::, 'Foo'>) no longer result in double
1758frees. This bug started happening in Perl 5.16.
1759
1760=item *
1761
1762Numerous memory leaks have been fixed, mostly involving fatal warnings and
1763syntax errors.
1764
1765=item *
1766
1767Some failed regular expression matches such as C<'f' =~ /../g> were not
1768resetting C<pos>. Also, "match-once" patterns (C<m?...?g>) failed to reset
1769it, too, when invoked a second time [perl #23180].
1770
1771=item *
1772
1773Accessing C<$&> after a pattern match now works if it had not been seen
1774before the match. I.e., this applies to C<${'&'}> (under C<no strict>) and
1775C<eval '$&'>. The same applies to C<$'> and C<$`> [perl #4289].
1776
1777=item *
1778
1779Several bugs involving C<local *ISA> and C<local *Foo::> causing stale
1780MRO caches have been fixed.
1781
1782=item *
1783
1784Defining a subroutine when its typeglob has been aliased no longer results
1785in stale method caches. This bug was introduced in Perl 5.10.
1786
1787=item *
1788
1789Localising a typeglob containing a subroutine when the typeglob's package
1790has been deleted from its parent stash no longer produces an error. This
1791bug was introduced in Perl 5.14.
1792
1793=item *
1794
1795Under some circumstances, C<local *method=...> would fail to reset method
1796caches upon scope exit.
1797
1798=item *
1799
1800C</[.foo.]/> is no longer an error, but produces a warning (as before) and
1801is treated as C</[.fo]/> [perl #115818].
1802
1803=item *
1804
1805C<goto $tied_var> now calls FETCH before deciding what type of goto
1806(subroutine or label) this is.
1807
1808=item *
1809
1810Renaming packages through glob assignment
1811(C<*Foo:: = *Bar::; *Bar:: = *Baz::>) in combination with C<m?...?> and
1812C<reset> no longer makes threaded builds crash.
1813
1814=item *
1815
1816A number of bugs related to assigning a list to hash have been fixed. Many of
1817these involve lists with repeated keys like C<(1, 1, 1, 1)>.
1818
1819=over 4
1820
1821=item *
1822
1823The expression C<scalar(%h = (1, 1, 1, 1))> now returns C<4>, not C<2>.
1824
1825=item *
1826
1827The return value of C<%h = (1, 1, 1)> in list context was wrong. Previously
1828this would return C<(1, undef, 1)>, now it returns C<(1, undef)>.
1829
1830=item *
1831
1832Perl now issues the same warning on C<($s, %h) = (1, {})> as it does for
1833C<(%h) = ({})>, "Reference found where even-sized list expected".
1834
1835=item *
1836
1837A number of additional edge cases in list assignment to hashes were
1838corrected. For more details see commit 23b7025ebc.
1839
1840=back
1841
1842=item *
1843
1844Attributes applied to lexical variables no longer leak memory.
1845[perl #114764]
1846
1847=item *
1848
1849C<dump>, C<goto>, C<last>, C<next>, C<redo> or C<require> followed by a
1850bareword (or version) and then an infix operator is no longer a syntax
1851error. It used to be for those infix operators (like C<+>) that have a
1852different meaning where a term is expected. [perl #105924]
1853
1854=item *
1855
1856C<require a::b . 1> and C<require a::b + 1> no longer produce erroneous
1857ambiguity warnings. [perl #107002]
1858
1859=item *
1860
1861Class method calls are now allowed on any string, and not just strings
1862beginning with an alphanumeric character. [perl #105922]
1863
1864=item *
1865
1866An empty pattern created with C<qr//> used in C<m///> no longer triggers
1867the "empty pattern reuses last pattern" behaviour. [perl #96230]
1868
1869=item *
1870
1871Tying a hash during iteration no longer results in a memory leak.
1872
1873=item *
1874
1875Freeing a tied hash during iteration no longer results in a memory leak.
1876
1877=item *
1878
1879List assignment to a tied array or hash that dies on STORE no longer
1880results in a memory leak.
1881
1882=item *
1883
1884If the hint hash (C<%^H>) is tied, compile-time scope entry (which copies
1885the hint hash) no longer leaks memory if FETCH dies. [perl #107000]
1886
1887=item *
1888
1889Constant folding no longer inappropriately triggers the special
1890C<split " "> behaviour. [perl #94490]
1891
1892=item *
1893
1894C<defined scalar(@array)>, C<defined do { &foo }>, and similar constructs
1895now treat the argument to C<defined> as a simple scalar. [perl #97466]
1896
1897=item *
1898
1899Running a custom debugging that defines no C<*DB::DB> glob or provides a
1900subroutine stub for C<&DB::DB> no longer results in a crash, but an error
1901instead. [perl #114990]
1902
1903=item *
1904
1905C<reset ""> now matches its documentation. C<reset> only resets C<m?...?>
1906patterns when called with no argument. An empty string for an argument now
1907does nothing. (It used to be treated as no argument.) [perl #97958]
1908
1909=item *
1910
1911C<printf> with an argument returning an empty list no longer reads past the
1912end of the stack, resulting in erratic behaviour. [perl #77094]
1913
1914=item *
1915
1916C<--subname> no longer produces erroneous ambiguity warnings.
1917[perl #77240]
1918
1919=item *
1920
1921C<v10> is now allowed as a label or package name. This was inadvertently
1922broken when v-strings were added in Perl 5.6. [perl #56880]
1923
1924=item *
1925
1926C<length>, C<pos>, C<substr> and C<sprintf> could be confused by ties,
1927overloading, references and typeglobs if the stringification of such
1928changed the internal representation to or from UTF8. [perl #114410]
1929
1930=item *
1931
1932utf8::encode now calls FETCH and STORE on tied variables. utf8::decode now
1933calls STORE (it was already calling FETCH).
1934
1935=item *
1936
1937C<$tied =~ s/$non_utf8/$utf8/> no longer loops infinitely if the tied
1938variable returns a Latin-1 string, shared hash key scalar, or reference or
2ae351f8 1939typeglob that stringifies as ASCII or Latin-1. This was a regression from
5ed58cbd
RS
19405.12.x.
1941
1942=item *
1943
1944C<s///> without /e is now better at detecting when it needs to forego
1945certain optimisations, fixing some buggy cases:
1946
1947=over
1948
1949=item *
1950
1951Match variables in certain constructs (C<&&>, C<||>, C<..> and others) in
1952the replacement part; e.g., C<s/(.)/$l{$a||$1}/g>. [perl #26986]
1953
1954=item *
1955
1956Aliases to match variables in the replacement.
1957
1958=item *
1959
1960C<$REGERROR> or C<$REGMARK> in the replacement. [perl #49190]
1961
1962=item *
1963
1964An empty pattern (C<s//$foo/>) that causes the last-successful pattern to
1965be used, when that pattern contains code blocks that modify the variables
1966in the replacement.
1967
1968=back
1969
1970=item *
1971
1972The taintedness of the replacement string no longer affects the taintedness
1973of the return value of C<s///e>.
1974
1975=item *
1976
1977The C<$|> autoflush variable is created on-the-fly when needed. If this
1978happened (e.g., if it was mentioned in a module or eval) when the
1979currently-selected filehandle was a typeglob with an empty IO slot, it used
1980to crash. [perl #115206]
1981
1982=item *
1983
1984Line numbers at the end of a string eval are no longer off by one.
1985[perl #114658]
1986
1987=item *
1988
1989@INC filters (subroutines returned by subroutines in @INC) that set $_ to a
1990copy-on-write scalar no longer cause the parser to modify that string
1991buffer in place.
1992
1993=item *
1994
1995C<length($object)> no longer returns the undefined value if the object has
1996string overloading that returns undef. [perl #115260]
1997
1998=item *
1999
2000The use of C<PL_stashcache>, the stash name lookup cache for method calls, has
2001been restored,
2002
2003Commit da6b625f78f5f133 in August 2011 inadvertently broke the code that looks
2004up values in C<PL_stashcache>. As it's a only cache, quite correctly everything
2005carried on working without it.
2006
2007=item *
2008
2009The error "Can't localize through a reference" had disappeared in 5.16.0
2010when C<local %$ref> appeared on the last line of an lvalue subroutine.
2011This error disappeared for C<\local %$ref> in perl 5.8.1. It has now
2012been restored.
2013
2014=item *
2015
2016The parsing of here-docs has been improved significantly, fixing several
2017parsing bugs and crashes and one memory leak, and correcting wrong
2018subsequent line numbers under certain conditions.
2019
2020=item *
2021
2022Inside an eval, the error message for an unterminated here-doc no longer
2023has a newline in the middle of it [perl #70836].
2024
2025=item *
2026
2027A substitution inside a substitution pattern (C<s/${s|||}//>) no longer
2028confuses the parser.
2029
2030=item *
2031
2032It may be an odd place to allow comments, but C<s//"" # hello/e> has
2033always worked, I<unless> there happens to be a null character before the
2034first #. Now it works even in the presence of nulls.
2035
2036=item *
2037
2038An invalid range in C<tr///> or C<y///> no longer results in a memory leak.
2039
2040=item *
2041
2042String eval no longer treats a semicolon-delimited quote-like operator at
2043the very end (C<eval 'q;;'>) as a syntax error.
2044
2045=item *
2046
2047C<< warn {$_ => 1} + 1 >> is no longer a syntax error. The parser used to
2048get confused with certain list operators followed by an anonymous hash and
2049then an infix operator that shares its form with a unary operator.
2050
2051=item *
2052
2053C<(caller $n)[6]> (which gives the text of the eval) used to return the
2054actual parser buffer. Modifying it could result in crashes. Now it always
2055returns a copy. The string returned no longer has "\n;" tacked on to the
2056end. The returned text also includes here-doc bodies, which used to be
2057omitted.
2058
2059=item *
2060
2061Reset the utf8 position cache when accessing magical variables to avoid the
2062string buffer and the utf8 position cache getting out of sync
2063[perl #114410].
2064
2065=item *
2066
2067Various cases of get magic being called twice for magical utf8 strings have been
2068fixed.
2069
2070=item *
2071
2072This code (when not in the presence of C<$&> etc)
2073
2074 $_ = 'x' x 1_000_000;
2075 1 while /(.)/;
2076
2077used to skip the buffer copy for performance reasons, but suffered from C<$1>
2078etc changing if the original string changed. That's now been fixed.
2079
2080=item *
2081
2082Perl doesn't use PerlIO anymore to report out of memory messages, as PerlIO
2083might attempt to allocate more memory.
2084
2085=item *
2086
2087In a regular expression, if something is quantified with C<{n,m}> where
2088C<S<n E<gt> m>>, it can't possibly match. Previously this was a fatal
2089error, but now is merely a warning (and that something won't match).
2090[perl #82954].
2091
2092=item *
2093
2094It used to be possible for formats defined in subroutines that have
2095subsequently been undefined and redefined to close over variables in the
2096wrong pad (the newly-defined enclosing sub), resulting in crashes or
2097"Bizarre copy" errors.
2098
2099=item *
2100
2101Redefinition of XSUBs at run time could produce warnings with the wrong
2102line number.
2103
2104=item *
2105
2106The %vd sprintf format does not support version objects for alpha versions.
2107It used to output the format itself (%vd) when passed an alpha version, and
2108also emit an "Invalid conversion in printf" warning. It no longer does,
2109but produces the empty string in the output. It also no longer leaks
2110memory in this case.
2111
2112=item *
2113
2114C<< $obj->SUPER::method >> calls in the main package could fail if the
2115SUPER package had already been accessed by other means.
2116
2117=item *
2118
f105b7be 2119Stash aliasing (C<< *foo:: = *bar:: >>) no longer causes SUPER calls to ignore
5ed58cbd
RS
2120changes to methods or @ISA or use the wrong package.
2121
2122=item *
2123
2124Method calls on packages whose names end in ::SUPER are no longer treated
2125as SUPER method calls, resulting in failure to find the method.
2126Furthermore, defining subroutines in such packages no longer causes them to
2127be found by SUPER method calls on the containing package [perl #114924].
2128
2129=item *
2130
2131C<\w> now matches the code points U+200C (ZERO WIDTH NON-JOINER) and U+200D
2132(ZERO WIDTH JOINER). C<\W> no longer matches these. This change is because
2133Unicode corrected their definition of what C<\w> should match.
2134
2135=item *
2136
2137C<dump LABEL> no longer leaks its label.
2138
2139=item *
2140
2141Constant folding no longer changes the behaviour of functions like C<stat()>
2142and C<truncate()> that can take either filenames or handles.
2143C<stat 1 ? foo : bar> nows treats its argument as a file name (since it is an
2144arbitrary expression), rather than the handle "foo".
2145
2146=item *
2147
2148C<truncate FOO, $len> no longer falls back to treating "FOO" as a file name if
2149the filehandle has been deleted. This was broken in Perl 5.16.0.
2150
2151=item *
2152
2153Subroutine redefinitions after sub-to-glob and glob-to-glob assignments no
2154longer cause double frees or panic messages.
2155
2156=item *
2157
2158C<s///> now turns vstrings into plain strings when performing a substitution,
2159even if the resulting string is the same (C<s/a/a/>).
2160
2161=item *
2162
2163Prototype mismatch warnings no longer erroneously treat constant subs as having
2164no prototype when they actually have "".
2165
2166=item *
2167
2168Constant subroutines and forward declarations no longer prevent prototype
2169mismatch warnings from omitting the sub name.
2170
2171=item *
2172
2173C<undef> on a subroutine now clears call checkers.
2174
2175=item *
2176
2177The C<ref> operator started leaking memory on blessed objects in Perl 5.16.0.
2178This has been fixed [perl #114340].
2179
2180=item *
2181
2182C<use> no longer tries to parse its arguments as a statement, making
2183C<use constant { () };> a syntax error [perl #114222].
2184
2185=item *
2186
2187On debugging builds, "uninitialized" warnings inside formats no longer cause
2188assertion failures.
2189
2190=item *
2191
2192On debugging builds, subroutines nested inside formats no longer cause
2193assertion failures [perl #78550].
2194
2195=item *
2196
2197Formats and C<use> statements are now permitted inside formats.
2198
2199=item *
2200
2201C<print $x> and C<sub { print $x }-E<gt>()> now always produce the same output.
2202It was possible for the latter to refuse to close over $x if the variable was
2203not active; e.g., if it was defined outside a currently-running named
2204subroutine.
2205
2206=item *
2207
2208Similarly, C<print $x> and C<print eval '$x'> now produce the same output.
2209This also allows "my $x if 0" variables to be seen in the debugger [perl
2210#114018].
2211
2212=item *
2213
2214Formats called recursively no longer stomp on their own lexical variables, but
2215each recursive call has its own set of lexicals.
2216
2217=item *
2218
2219Attempting to free an active format or the handle associated with it no longer
2220results in a crash.
2221
2222=item *
2223
2224Format parsing no longer gets confused by braces, semicolons and low-precedence
2225operators. It used to be possible to use braces as format delimiters (instead
2226of C<=> and C<.>), but only sometimes. Semicolons and low-precedence operators
2227in format argument lines no longer confuse the parser into ignoring the line's
2228return value. In format argument lines, braces can now be used for anonymous
2229hashes, instead of being treated always as C<do> blocks.
2230
2231=item *
2232
2233Formats can now be nested inside code blocks in regular expressions and other
2234quoted constructs (C</(?{...})/> and C<qq/${...}/>) [perl #114040].
2235
2236=item *
2237
2238Formats are no longer created after compilation errors.
2239
2240=item *
2241
2242Under debugging builds, the B<-DA> command line option started crashing in Perl
22435.16.0. It has been fixed [perl #114368].
2244
2245=item *
2246
2247A potential deadlock scenario involving the premature termination of a pseudo-
2248forked child in a Windows build with ithreads enabled has been fixed. This
2249resolves the common problem of the F<t/op/fork.t> test hanging on Windows [perl
2250#88840].
2251
2252=item *
2253
2254The microperl build, broken since Perl 5.15.7, has now been restored.
2255
2256=item *
2257
2258The code which generates errors from C<require()> could potentially read one or
2259two bytes before the start of the filename for filenames less than three bytes
2260long and ending C</\.p?\z/>. This has now been fixed. Note that it could
2261never have happened with module names given to C<use()> or C<require()> anyway.
2262
2263=item *
2264
2265The handling of pathnames of modules given to C<require()> has been made
2266thread-safe on VMS.
2267
2268=item *
2269
d85cd26b
RS
2270Non-blocking sockets have been fixed on VMS.
2271
2272=item *
2273
5ed58cbd
RS
2274A bug in the compilation of a C</(?{})/> expression which affected the TryCatch
2275test suite has been fixed [perl #114242].
2276
2277=item *
2278
2279Pod can now be nested in code inside a quoted construct outside of a string
2280eval. This used to work only within string evals [perl #114040].
2281
2282=item *
2283
2284C<goto ''> now looks for an empty label, producing the "goto must have
2285label" error message, instead of exiting the program [perl #111794].
2286
2287=item *
2288
2289C<goto "\0"> now dies with "Can't find label" instead of "goto must have
2290label".
2291
2292=item *
2293
2294The C function C<hv_store> used to result in crashes when used on C<%^H>
2295[perl #111000].
2296
2297=item *
2298
2299A call checker attached to a closure prototype via C<cv_set_call_checker>
2300is now copied to closures cloned from it. So C<cv_set_call_checker> now
2301works inside an attribute handler for a closure.
2302
2303=item *
2304
2305Writing to C<$^N> used to have no effect. Now it croaks with "Modification
2306of a read-only value" by default, but that can be overridden by a custom
2307regular expression engine, as with C<$1> [perl #112184].
2308
2309=item *
2310
2311C<undef> on a control character glob (C<undef *^H>) no longer emits an
2312erroneous warning about ambiguity [perl #112456].
2313
2314=item *
2315
2316For efficiency's sake, many operators and built-in functions return the
2317same scalar each time. Lvalue subroutines and subroutines in the CORE::
2318namespace were allowing this implementation detail to leak through.
2319C<print &CORE::uc("a"), &CORE::uc("b")> used to print "BB". The same thing
2320would happen with an lvalue subroutine returning the return value of C<uc>.
2321Now the value is copied in such cases.
2322
2323=item *
2324
2325C<method {}> syntax with an empty block or a block returning an empty list
2326used to crash or use some random value left on the stack as its invocant.
2327Now it produces an error.
2328
2329=item *
2330
2331C<vec> now works with extremely large offsets (E<gt>2 GB) [perl #111730].
2332
2333=item *
2334
2335Changes to overload settings now take effect immediately, as do changes to
2336inheritance that affect overloading. They used to take effect only after
2337C<bless>.
2338
2339Objects that were created before a class had any overloading used to remain
2340non-overloaded even if the class gained overloading through C<use overload>
2341or @ISA changes, and even after C<bless>. This has been fixed
2342[perl #112708].
2343
2344=item *
2345
2346Classes with overloading can now inherit fallback values.
2347
2348=item *
2349
2350Overloading was not respecting a fallback value of 0 if there were
2351overloaded objects on both sides of an assignment operator like C<+=>
2352[perl #111856].
2353
2354=item *
2355
2356C<pos> now croaks with hash and array arguments, instead of producing
2357erroneous warnings.
2358
2359=item *
2360
2361C<while(each %h)> now implies C<while(defined($_ = each %h))>, like
2362C<readline> and C<readdir>.
2363
2364=item *
2365
2366Subs in the CORE:: namespace no longer crash after C<undef *_> when called
2367with no argument list (C<&CORE::time> with no parentheses).
2368
2369=item *
2370
2371C<unpack> no longer produces the "'/' must follow a numeric type in unpack"
2372error when it is the data that are at fault [perl #60204].
2373
2374=item *
2375
2376C<join> and C<"@array"> now call FETCH only once on a tied C<$">
2377[perl #8931].
2378
2379=item *
2380
2381Some subroutine calls generated by compiling core ops affected by a
2382C<CORE::GLOBAL> override had op checking performed twice. The checking
2383is always idempotent for pure Perl code, but the double checking can
2384matter when custom call checkers are involved.
2385
2386=item *
2387
2388A race condition used to exist around fork that could cause a signal sent to
2389the parent to be handled by both parent and child. Signals are now blocked
2390briefly around fork to prevent this from happening [perl #82580].
2391
2392=item *
2393
2394The implementation of code blocks in regular expressions, such as C<(?{})>
2395and C<(??{})>, has been heavily reworked to eliminate a whole slew of bugs.
2396The main user-visible changes are:
2397
2398=over 4
2399
2400=item *
2401
2402Code blocks within patterns are now parsed in the same pass as the
2403surrounding code; in particular it is no longer necessary to have balanced
2404braces: this now works:
2405
2406 /(?{ $x='{' })/
2407
2408This means that this error message is no longer generated:
2409
2410 Sequence (?{...}) not terminated or not {}-balanced in regex
2411
2412but a new error may be seen:
2413
2414 Sequence (?{...}) not terminated with ')'
2415
2416In addition, literal code blocks within run-time patterns are only
2417compiled once, at perl compile-time:
2418
2419 for my $p (...) {
2420 # this 'FOO' block of code is compiled once,
2421 # at the same time as the surrounding 'for' loop
2422 /$p{(?{FOO;})/;
2423 }
2424
2425=item *
2426
2427Lexical variables are now sane as regards scope, recursion and closure
2428behavior. In particular, C</A(?{B})C/> behaves (from a closure viewpoint)
2429exactly like C</A/ && do { B } && /C/>, while C<qr/A(?{B})C/> is like
2430C<sub {/A/ && do { B } && /C/}>. So this code now works how you might
2431expect, creating three regexes that match 0, 1, and 2:
2432
2433 for my $i (0..2) {
2434 push @r, qr/^(??{$i})$/;
2435 }
2436 "1" =~ $r[1]; # matches
2437
2438=item *
2439
2440The C<use re 'eval'> pragma is now only required for code blocks defined
2441at runtime; in particular in the following, the text of the C<$r> pattern is
2442still interpolated into the new pattern and recompiled, but the individual
2443compiled code-blocks within C<$r> are reused rather than being recompiled,
2444and C<use re 'eval'> isn't needed any more:
2445
2446 my $r = qr/abc(?{....})def/;
2447 /xyz$r/;
2448
2449=item *
2450
2451Flow control operators no longer crash. Each code block runs in a new
2452dynamic scope, so C<next> etc. will not see
2453any enclosing loops. C<return> returns a value
2454from the code block, not from any enclosing subroutine.
2455
2456=item *
2457
2458Perl normally caches the compilation of run-time patterns, and doesn't
2459recompile if the pattern hasn't changed, but this is now disabled if
2460required for the correct behavior of closures. For example:
2461
2462 my $code = '(??{$x})';
2463 for my $x (1..3) {
2464 # recompile to see fresh value of $x each time
2465 $x =~ /$code/;
2466 }
2467
2468=item *
2469
2470The C</msix> and C<(?msix)> etc. flags are now propagated into the return
2471value from C<(??{})>; this now works:
2472
2473 "AB" =~ /a(??{'b'})/i;
2474
2475=item *
2476
2477Warnings and errors will appear to come from the surrounding code (or for
2478run-time code blocks, from an eval) rather than from an C<re_eval>:
2479
2480 use re 'eval'; $c = '(?{ warn "foo" })'; /$c/;
2481 /(?{ warn "foo" })/;
2482
2483formerly gave:
2484
2485 foo at (re_eval 1) line 1.
2486 foo at (re_eval 2) line 1.
2487
2488and now gives:
2489
2490 foo at (eval 1) line 1.
2491 foo at /some/prog line 2.
2492
2493=back
2494
2495=item *
2496
2e7bc647
KW
2497Perl now can be recompiled to use any Unicode version. In v5.16, it
2498worked on Unicodes 6.0 and 6.1, but there were various bugs if earlier
2499releases were used; the older the release the more problems.
5ed58cbd
RS
2500
2501=item *
2502
2503C<vec> no longer produces "uninitialized" warnings in lvalue context
2504[perl #9423].
2505
2506=item *
2507
2508An optimization involving fixed strings in regular expressions could cause
2509a severe performance penalty in edge cases. This has been fixed
2510[perl #76546].
2511
2512=item *
2513
2514In certain cases, including empty subpatterns within a regular expression (such
2515as C<(?:)> or C<(?:|)>) could disable some optimizations. This has been fixed.
2516
2517=item *
2518
2519The "Can't find an opnumber" message that C<prototype> produces when passed
2520a string like "CORE::nonexistent_keyword" now passes UTF-8 and embedded
2521NULs through unchanged [perl #97478].
2522
2523=item *
2524
2525C<prototype> now treats magical variables like C<$1> the same way as
2526non-magical variables when checking for the CORE:: prefix, instead of
2527treating them as subroutine names.
2528
2529=item *
2530
2531Under threaded perls, a runtime code block in a regular expression could
2532corrupt the package name stored in the op tree, resulting in bad reads
2533in C<caller>, and possibly crashes [perl #113060].
2534
2535=item *
2536
2537Referencing a closure prototype (C<\&{$_[1]}> in an attribute handler for a
2538closure) no longer results in a copy of the subroutine (or assertion
2539failures on debugging builds).
2540
2541=item *
2542
2543C<eval '__PACKAGE__'> now returns the right answer on threaded builds if
2544the current package has been assigned over (as in
2545C<*ThisPackage:: = *ThatPackage::>) [perl #78742].
2546
2547=item *
2548
2549If a package is deleted by code that it calls, it is possible for C<caller>
2550to see a stack frame belonging to that deleted package. C<caller> could
2551crash if the stash's memory address was reused for a scalar and a
2552substitution was performed on the same scalar [perl #113486].
2553
2554=item *
2555
2556C<UNIVERSAL::can> no longer treats its first argument differently
2557depending on whether it is a string or number internally.
2558
2559=item *
2560
2561C<open> with C<< <& >> for the mode checks to see whether the third argument is
2562a number, in determining whether to treat it as a file descriptor or a handle
2563name. Magical variables like C<$1> were always failing the numeric check and
2564being treated as handle names.
2565
2566=item *
2567
2568C<warn>'s handling of magical variables (C<$1>, ties) has undergone several
2569fixes. C<FETCH> is only called once now on a tied argument or a tied C<$@>
2570[perl #97480]. Tied variables returning objects that stringify as "" are
2571no longer ignored. A tied C<$@> that happened to return a reference the
2572I<previous> time it was used is no longer ignored.
2573
2574=item *
2575
2576C<warn ""> now treats C<$@> with a number in it the same way, regardless of
2577whether it happened via C<$@=3> or C<$@="3">. It used to ignore the
2578former. Now it appends "\t...caught", as it has always done with
2579C<$@="3">.
2580
2581=item *
2582
2583Numeric operators on magical variables (e.g., S<C<$1 + 1>>) used to use
2584floating point operations even where integer operations were more appropriate,
2585resulting in loss of accuracy on 64-bit platforms [perl #109542].
2586
2587=item *
2588
2589Unary negation no longer treats a string as a number if the string happened
2590to be used as a number at some point. So, if C<$x> contains the string "dogs",
2591C<-$x> returns "-dogs" even if C<$y=0+$x> has happened at some point.
2592
2593=item *
2594
2595In Perl 5.14, C<-'-10'> was fixed to return "10", not "+10". But magical
2596variables (C<$1>, ties) were not fixed till now [perl #57706].
2597
2598=item *
2599
2600Unary negation now treats strings consistently, regardless of the internal
2601C<UTF8> flag.
2602
2603=item *
2604
2605A regression introduced in Perl v5.16.0 involving
2606C<tr/I<SEARCHLIST>/I<REPLACEMENTLIST>/> has been fixed. Only the first
2607instance is supposed to be meaningful if a character appears more than
2608once in C<I<SEARCHLIST>>. Under some circumstances, the final instance
2609was overriding all earlier ones. [perl #113584]
2610
2611=item *
2612
2613Regular expressions like C<qr/\87/> previously silently inserted a NUL
2614character, thus matching as if it had been written C<qr/\00087/>. Now it
2615matches as if it had been written as C<qr/87/>, with a message that the
2616sequence C<"\8"> is unrecognized.
2617
2618=item *
2619
2620C<__SUB__> now works in special blocks (C<BEGIN>, C<END>, etc.).
2621
2622=item *
2623
2624Thread creation on Windows could theoretically result in a crash if done
2625inside a C<BEGIN> block. It still does not work properly, but it no longer
2626crashes [perl #111610].
2627
2628=item *
2629
2630C<\&{''}> (with the empty string) now autovivifies a stub like any other
2631sub name, and no longer produces the "Unable to create sub" error
2632[perl #94476].
2633
2634=item *
2635
2636A regression introduced in v5.14.0 has been fixed, in which some calls
2637to the C<re> module would clobber C<$_> [perl #113750].
2638
2639=item *
2640
2641C<do FILE> now always either sets or clears C<$@>, even when the file can't be
2642read. This ensures that testing C<$@> first (as recommended by the
2643documentation) always returns the correct result.
2644
2645=item *
2646
2647The array iterator used for the C<each @array> construct is now correctly
2648reset when C<@array> is cleared (RT #75596). This happens for example when the
2649array is globally assigned to, as in C<@array = (...)>, but not when its
2650B<values> are assigned to. In terms of the XS API, it means that C<av_clear()>
2651will now reset the iterator.
2652
2653This mirrors the behaviour of the hash iterator when the hash is cleared.
2654
2655=item *
2656
2657C<< $class->can >>, C<< $class->isa >>, and C<< $class->DOES >> now return
2658correct results, regardless of whether that package referred to by C<$class>
2659exists [perl #47113].
2660
2661=item *
2662
2663Arriving signals no longer clear C<$@> [perl #45173].
2664
2665=item *
2666
2667Allow C<my ()> declarations with an empty variable list [perl #113554].
2668
2669=item *
2670
2671During parsing, subs declared after errors no longer leave stubs
2672[perl #113712].
2673
2674=item *
2675
2676Closures containing no string evals no longer hang on to their containing
2677subroutines, allowing variables closed over by outer subroutines to be
2678freed when the outer sub is freed, even if the inner sub still exists
2679[perl #89544].
2680
2681=item *
2682
2683Duplication of in-memory filehandles by opening with a "<&=" or ">&=" mode
2684stopped working properly in 5.16.0. It was causing the new handle to
2685reference a different scalar variable. This has been fixed [perl #113764].
2686
2687=item *
2688
2689C<qr//> expressions no longer crash with custom regular expression engines
2690that do not set C<offs> at regular expression compilation time
2691[perl #112962].
2692
2693=item *
2694
2695C<delete local> no longer crashes with certain magical arrays and hashes
2696[perl #112966].
2697
2698=item *
2699
2700C<local> on elements of certain magical arrays and hashes used not to
2701arrange to have the element deleted on scope exit, even if the element did
2702not exist before C<local>.
2703
2704=item *
2705
2706C<scalar(write)> no longer returns multiple items [perl #73690].
2707
2708=item *
2709
2710String to floating point conversions no longer misparse certain strings under
2711C<use locale> [perl #109318].
2712
2713=item *
2714
2715C<@INC> filters that die no longer leak memory [perl #92252].
2716
2717=item *
2718
2719The implementations of overloaded operations are now called in the correct
2720context. This allows, among other things, being able to properly override
2721C<< <> >> [perl #47119].
2722
2723=item *
2724
2725Specifying only the C<fallback> key when calling C<use overload> now behaves
2726properly [perl #113010].
2727
2728=item *
2729
2730C<< sub foo { my $a = 0; while ($a) { ... } } >> and
2731C<< sub foo { while (0) { ... } } >> now return the same thing [perl #73618].
2732
2733=item *
2734
2735String negation now behaves the same under C<use integer;> as it does
2736without [perl #113012].
2737
2738=item *
2739
2740C<chr> now returns the Unicode replacement character (U+FFFD) for -1,
2741regardless of the internal representation. -1 used to wrap if the argument
2742was tied or a string internally.
2743
2744=item *
2745
2746Using a C<format> after its enclosing sub was freed could crash as of
2747perl 5.12.0, if the format referenced lexical variables from the outer sub.
2748
2749=item *
2750
2751Using a C<format> after its enclosing sub was undefined could crash as of
2752perl 5.10.0, if the format referenced lexical variables from the outer sub.
2753
2754=item *
2755
2756Using a C<format> defined inside a closure, which format references
2757lexical variables from outside, never really worked unless the C<write>
2758call was directly inside the closure. In 5.10.0 it even started crashing.
2759Now the copy of that closure nearest the top of the call stack is used to
2760find those variables.
2761
2762=item *
2763
2764Formats that close over variables in special blocks no longer crash if a
2765stub exists with the same name as the special block before the special
2766block is compiled.
2767
2768=item *
2769
2770The parser no longer gets confused, treating C<eval foo ()> as a syntax
2771error if preceded by C<print;> [perl #16249].
2772
2773=item *
2774
2775The return value of C<syscall> is no longer truncated on 64-bit platforms
2776[perl #113980].
2777
2778=item *
2779
2780Constant folding no longer causes C<print 1 ? FOO : BAR> to print to the
2781FOO handle [perl #78064].
2782
2783=item *
2784
2785C<do subname> now calls the named subroutine and uses the file name it
2786returns, instead of opening a file named "subname".
2787
2788=item *
2789
2790Subroutines looked up by rv2cv check hooks (registered by XS modules) are
2791now taken into consideration when determining whether C<foo bar> should be
2792the sub call C<foo(bar)> or the method call C<< "bar"->foo >>.
2793
2794=item *
2795
2796C<CORE::foo::bar> is no longer treated specially, allowing global overrides
2797to be called directly via C<CORE::GLOBAL::uc(...)> [perl #113016].
2798
2799=item *
2800
2801Calling an undefined sub whose typeglob has been undefined now produces the
2802customary "Undefined subroutine called" error, instead of "Not a CODE
2803reference".
2804
2805=item *
2806
2807Two bugs involving @ISA have been fixed. C<*ISA = *glob_without_array> and
2808C<undef *ISA; @{*ISA}> would prevent future modifications to @ISA from
2809updating the internal caches used to look up methods. The
2810*glob_without_array case was a regression from Perl 5.12.
2811
2812=item *
2813
2814Regular expression optimisations sometimes caused C<$> with C</m> to
2815produce failed or incorrect matches [perl #114068].
2816
2817=item *
2818
2819C<__SUB__> now works in a C<sort> block when the enclosing subroutine is
2820predeclared with C<sub foo;> syntax [perl #113710].
2821
2822=item *
2823
2824Unicode properties only apply to Unicode code points, which leads to
2825some subtleties when regular expressions are matched against
2826above-Unicode code points. There is a warning generated to draw your
2827attention to this. However, this warning was being generated
2828inappropriately in some cases, such as when a program was being parsed.
2829Non-Unicode matches such as C<\w> and C<[:word;]> should not generate the
2830warning, as their definitions don't limit them to apply to only Unicode
2831code points. Now the message is only generated when matching against
2832C<\p{}> and C<\P{}>. There remains a bug, [perl #114148], for the very
2833few properties in Unicode that match just a single code point. The
2834warning is not generated if they are matched against an above-Unicode
2835code point.
2836
2837=item *
2838
2839Uninitialized warnings mentioning hash elements would only mention the
2840element name if it was not in the first bucket of the hash, due to an
2841off-by-one error.
2842
2843=item *
2844
2845A regular expression optimizer bug could cause multiline "^" to behave
2846incorrectly in the presence of line breaks, such that
2847C<"/\n\n" =~ m#\A(?:^/$)#im> would not match [perl #115242].
2848
2849=item *
2850
2851Failed C<fork> in list context no longer corrupts the stack.
2852C<@a = (1, 2, fork, 3)> used to gobble up the 2 and assign C<(1, undef, 3)>
2853if the C<fork> call failed.
2854
2855=item *
2856
2857Numerous memory leaks have been fixed, mostly involving tied variables that
2858die, regular expression character classes and code blocks, and syntax
2859errors.
2860
2861=item *
2862
2863Assigning a regular expression (C<${qr//}>) to a variable that happens to
2864hold a floating point number no longer causes assertion failures on
2865debugging builds.
2866
2867=item *
2868
2869Assigning a regular expression to a scalar containing a number no longer
f105b7be 2870causes subsequent numification to produce random numbers.
5ed58cbd
RS
2871
2872=item *
2873
2874Assigning a regular expression to a magic variable no longer wipes away the
2ae351f8 2875magic. This was a regression from 5.10.
5ed58cbd
RS
2876
2877=item *
2878
2879Assigning a regular expression to a blessed scalar no longer results in
2ae351f8 2880crashes. This was also a regression from 5.10.
5ed58cbd
RS
2881
2882=item *
2883
2884Regular expression can now be assigned to tied hash and array elements with
2885flattening into strings.
2886
2887=item *
2888
f105b7be 2889Numifying a regular expression no longer results in an uninitialized
5ed58cbd
RS
2890warning.
2891
2892=item *
2893
2894Negative array indices no longer cause EXISTS methods of tied variables to
2ae351f8 2895be ignored. This was a regression from 5.12.
5ed58cbd
RS
2896
2897=item *
2898
2899Negative array indices no longer result in crashes on arrays tied to
2900non-objects.
2901
2902=item *
2903
2904C<$byte_overload .= $utf8> no longer results in doubly-encoded UTF8 if the
2905left-hand scalar happened to have produced a UTF8 string the last time
2906overloading was invoked.
2907
2908=item *
2909
2910C<goto &sub> now uses the current value of @_, instead of using the array
2911the subroutine was originally called with. This means
2912C<local @_ = (...); goto &sub> now works [perl #43077].
2913
2914=item *
2915
2916If a debugger is invoked recursively, it no longer stomps on its own
2917lexical variables. Formerly under recursion all calls would share the same
2918set of lexical variables [perl #115742].
2919
2920=item *
2921
2922C<*_{ARRAY}> returned from a subroutine no longer spontaneously
2923becomes empty.
2924
2925=back
2926
2927=head1 Known Problems
2928
2929=over 4
2930
2931=item *
2932
2933XXX: the imperfect behavior of the ** deprecation
2934
2935=back
2936
2937=head1 Acknowledgements
a75569c0 2938
5ed58cbd 2939XXX Generate this with:
a75569c0 2940
5ed58cbd 2941 perl Porting/acknowledgements.pl v5.18.0..HEAD
f5b73711 2942
44691e6f
AB
2943=head1 Reporting Bugs
2944
e08634c5
SH
2945If you find what you think is a bug, you might check the articles recently
2946posted to the comp.lang.perl.misc newsgroup and the perl bug database at
2947http://rt.perl.org/perlbug/ . There may also be information at
2948http://www.perl.org/ , the Perl Home Page.
44691e6f 2949
e08634c5
SH
2950If you believe you have an unreported bug, please run the L<perlbug> program
2951included with your release. Be sure to trim your bug down to a tiny but
2952sufficient test case. Your bug report, along with the output of C<perl -V>,
2953will be sent off to perlbug@perl.org to be analysed by the Perl porting team.
44691e6f
AB
2954
2955If the bug you are reporting has security implications, which make it
e08634c5
SH
2956inappropriate to send to a publicly archived mailing list, then please send it
2957to perl5-security-report@perl.org. This points to a closed subscription
2958unarchived mailing list, which includes all the core committers, who will be
2959able to help assess the impact of issues, figure out a resolution, and help
f9001595 2960co-ordinate the release of patches to mitigate or fix the problem across all
e08634c5
SH
2961platforms on which Perl is supported. Please only use this address for
2962security issues in the Perl core, not for modules independently distributed on
2963CPAN.
44691e6f
AB
2964
2965=head1 SEE ALSO
2966
e08634c5
SH
2967The F<Changes> file for an explanation of how to view exhaustive details on
2968what changed.
44691e6f
AB
2969
2970The F<INSTALL> file for how to build Perl.
2971
2972The F<README> file for general stuff.
2973
2974The F<Artistic> and F<Copying> files for copyright information.
2975
2976=cut