This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Use 'was' not 'is' to describe fixed regressions
[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
79=head3 PERL_HASH_SEED enviornment variable now takes a hex value
80
81PERL_HASH_SEED no longer accepts an integer as a parameter, instead the
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
88PERL_PERTURB_KEYS environment variable allows one to control the level of
89randomization applied to C<keys> and friends.
90
91When PERL_PERTURB_KEYS is 0, perl will not randomize key order at all. The
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
95When PERL_PERTURB_KEYS is 1, perl will randomize keys in a non repeatedable
96way. The chance that C<keys> changes due to an insert will be very high. This
97is the most secure and default mode.
98
99When PERL_PERTURB_KEYS is 2, perl will randomize keys in a repeatedable way.
100Repititive runs of the same program should produce the same output every time.
101The chance that keys changes due to an insert will be very high.
102
103PERL_HASH_SEED implies a non-default PERL_PERTURB_KEYS setting. Setting
104PERL_HASH_SEED=0 (exactly one 0) implies PERL_PERTURB_KEYS=0 (hash key
105randomization disabled), settng PERL_HASH_SEED to any other value, implies
106PERL_PERTURB_KEYS=2 (deterministic/repeatable hash key randomization).
107Specifying PERL_PERTURB_KEYS explicitly to a different level overrides this
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
119hash function perl was built with AND the seed, in hex, in use for that
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
128Perl now supports the final version of Unicode 6.2. Earlier releases in
129the 5.17 series supported Unicode 6.2 beta versions. There were no
130substantive changes in the final Unicode 6.2 version from the most
131recent beta, included in Perl 5.17.4. A list of changes from Unicode
1326.1 is at L<http://www.unicode.org/versions/Unicode6.2.0>.
133
134=head2 Character name aliases may now include non-Latin1-range characters
135
136It is possible to define your own names for characters for use in
137C<\N{...}>, C<charnames::vianame()>, etc. These names can now be
138comprised of characters from the whole Unicode range. This allows for
139names to be in your native language, and not just English. Certain
140restrictions apply to the characters that may be used (you can't define
141a name that has punctuation in it, for example). See L<charnames/CUSTOM
142ALIASES>.
143
144=head2 New DTrace probes
145
146The following new DTrace probes have been added:
14731ad1 147
337fb649 148=over 4
14731ad1 149
82d98f72 150=item *
14731ad1 151
5ed58cbd
RS
152C<op-entry>
153
154=item *
155
156C<loading-file>
157
158=item *
159
160C<loaded-file>
161
162=back
163
164=head2 C<${^LAST_FH}>
165
166This new variable provides access to the filehandle that was last read.
167This is the handle used by C<$.> and by C<tell> and C<eof> without
168arguments.
169
170=head2 Regular Expression Set Operations
171
172This is an B<experimental> feature to allow matching against the union,
173intersection, etc., of sets of code points, similar to
174L<Unicode::Regex::Set>. It can also be used to extend C</x> processing
175to [bracketed] character classes, and as a replacement of user-defined
176properties, allowing more complex expressions than they do. See
177L<perlrecharclass/Extended Bracketed Character Classes>.
178
179=head2 Lexical subroutines
180
181This new feature is still considered B<experimental>. To enable it:
182
183 use 5.018;
184 no warnings "experimental::lexical_subs";
185 use feature "lexical_subs";
186
187You can now declare subroutines with C<state sub foo>, C<my sub foo>, and
188C<our sub foo>. (C<state sub> requires that the "state" feature be
189enabled, unless you write it as C<CORE::state sub foo>.)
190
191C<state sub> creates a subroutine visible within the lexical scope in which
192it is declared. The subroutine is shared between calls to the outer sub.
193
194C<my sub> declares a lexical subroutine that is created each time the
195enclosing block is entered. C<state sub> is generally slightly faster than
196C<my sub>.
197
198C<our sub> declares a lexical alias to the package subroutine of the same
199name.
200
201For more information, see L<perlsub/Lexical Subroutines>.
202
203=head2 Computed Labels
204
205The loop controls C<next>, C<last> and C<redo>, and the special C<dump>
206operator, now allow arbitrary expressions to be used to compute labels at run
207time. Previously, any argument that was not a constant was treated as the
208empty string.
209
210=head2 More CORE:: subs
211
212Several more built-in functions have been added as subroutines to the
213CORE:: namespace - namely, those non-overridable keywords that can be
214implemented without custom parsers: C<defined>, C<delete>, C<exists>,
215C<glob>, C<pos>, C<protoytpe>, C<scalar>, C<split>, C<study>, and C<undef>.
216
217As some of these have prototypes, C<prototype('CORE::...')> has been
218changed to not make a distinction between overridable and non-overridable
219keywords. This is to make C<prototype('CORE::pos')> consistent with
220C<prototype(&CORE::pos)>.
221
222=head2 C<kill> with negative signal names
223
224C<kill> has always allowed a negative signal number, which kills the
225process group instead of a single process. It has also allowed signal
226names. But it did not behave consistently, because negative signal names
227were treated as 0. Now negative signals names like C<-INT> are supported
228and treated the same way as -2 [perl #112990].
229
230=head1 Security
231
232=head2 C<Storable> security warning in documentation
233
234The documentation for C<Storable> now includes a section which warns readers
235of the danger of accepting Storable documents from untrusted sources. The
236short version is that deserializing certain types of data can lead to loading
237modules and other code execution. This is documented behavior and wanted
238behavior, but this opens an attack vector for malicious entities.
239
240=head2 C<Locale::Maketext> allowed code injection via a malicious template
241
242If users could provide a translation string to Locale::Maketext, this could be
243used to invoke arbitrary Perl subroutines available in the current process.
244
245This has been fixed, but it is still possible to invoke any method provided by
246C<Locale::Maketext> itself or a subclass that you are using. One of these
247methods in turn will invoke the Perl core's C<sprintf> subroutine.
248
249In summary, allowing users to provide translation strings without auditing
250them is a bad idea.
251
252This vulnerability is documented in CVE-2012-6329.
253
254=head2 Avoid calling memset with a negative count
255
256Poorly written perl code that allows an attacker to specify the count to perl's
257C<x> string repeat operator can already cause a memory exhaustion
258denial-of-service attack. A flaw in versions of perl before 5.15.5 can escalate
259that into a heap buffer overrun; coupled with versions of glibc before 2.16, it
260possibly allows the execution of arbitrary code.
261
262The flaw addressed to this commit has been assigned identifier CVE-2012-5195
263and was researched by Tim Brown.
264
265=head1 Incompatible Changes
266
267=head2 See also: hash overhaul
268
269Some of the changes in the L<hash overhaul|/"Hash overhaul"> are not fully
270compatible with previous versions of perl. Please read that section.
271
272=head2 An unknown character name in C<\N{...}> is now a syntax error
273
274Previously, it warned, and the Unicode REPLACEMENT CHARACTER was
275substituted. Unicode now recommends that this situation be a syntax
276error. Also, the previous behavior led to some confusing warnings and
277behaviors, and since the REPLACEMENT CHARACTER has no use other than as
278a stand-in for some unknown character, any code that has this problem is
279buggy.
280
281=head2 Formerly deprecated characters in C<\N{}> character name aliases are now errors.
282
283Since v5.12.0, it has been deprecated to use certain characters in
284user-defined C<\N{...}> character names. These now cause a syntax
285error. For example, it is now an error to begin a name with a digit,
286such as in
287
288 my $undraftable = "\N{4F}"; # Syntax error!
289
290or to have commas anywhere in the name. See L<charnames/CUSTOM ALIASES>
291
292=head2 C<\N{BELL}> now refers to U+1F514 instead of U+0007
293
294Unicode 6.0 reused the name "BELL" for a different code point than it
295traditionally had meant. Since Perl v5.14, use of this name still
296referred to U+0007, but would raise a deprecation warning. Now, "BELL"
297refers to U+1F514, and the name for U+0007 is "ALERT". All the
298functions in L<charnames> have been correspondingly updated.
299
300=head2 New Restrictions in Multi-Character Case-Insensitive Matching in Regular Expression Bracketed Character Classes
301
302Unicode has now withdrawn their previous recommendation for regular
303expressions to automatically handle cases where a single character can
304match multiple characters case-insensitively, for example the letter
305LATIN SMALL LETTER SHARP S and the sequence C<ss>. This is because
306it turns out to be impracticable to do this correctly in all
307circumstances. Because Perl has tried to do this as best it can, it
308will continue to do so. (We are considering an option to turn it off.)
309However, a new restriction is being added on such matches when they
310occur in [bracketed] character classes. People were specifying
311things such as C</[\0-\xff]/i>, and being surprised that it matches the
312two character sequence C<ss> (since LATIN SMALL LETTER SHARP S occurs in
313this range). This behavior is also inconsistent with using a
314property instead of a range: C<\p{Block=Latin1}> also includes LATIN
315SMALL LETTER SHARP S, but C</[\p{Block=Latin1}]/i> does not match C<ss>.
316The new rule is that for there to be a multi-character case-insensitive
317match within a bracketed character class, the character must be
318explicitly listed, and not as an end point of a range. This more
319closely obeys the Principle of Least Astonishment. See
320L<perlrecharclass/Bracketed Character Classes>. Note that a bug [perl
321#89774], now fixed as part of this change, prevented the previous
322behavior from working fully.
323
324=head2 Explicit rules for variable names and identifiers
325
326Due to an oversight, length-one variable names in 5.16 were completely
327unrestricted, and opened the door to several kinds of insanity. As of
3285.18, these now follow the rules of other identifiers, in addition
329to accepting characters that match the C<\p{POSIX_Punct}> property.
330
331There are no longer any differences in the parsing of identifiers
332specified as C<$...> or C<${...}>; previously, they were dealt with in
333different parts of the core, and so had slightly different behavior. For
334instance, C<${foo:bar}> was a legal variable name. Since they are now
335both parsed by the same code, that is no longer the case.
336
337=head2 C<\s> in regular expressions now matches a Vertical Tab
338
339No one could recall why C<\s> didn't match C<\cK>, the vertical tab.
340Now it does. Given the extreme rarity of that character, very little
341breakage is expected.
342
343=head2 C</(?{})/> and C</(??{})/> have been heavily reworked
344
345The implementation of this feature has been almost completely rewritten.
346Although its main intent is to fix bugs, some behaviors, especially
347related to the scope of lexical variables, will have changed. This is
348described more fully in the L</Selected Bug Fixes> section.
349
350=head2 Stricter parsing of substitution replacement
351
352It is no longer possible to abuse the way the parser parses C<s///e> like
353this:
354
355 %_=(_,"Just another ");
356 $_="Perl hacker,\n";
357 s//_}->{_/e;print
358
359=head2 C<given> now aliases the global C<$_>
360
361Instead of assigning to an implicit lexical C<$_>, C<given> now makes the
362global C<$_> an alias for its argument, just like C<foreach>. However, it
363still uses lexical C<$_> if there is lexical C<$_> in scope (again, just like
364C<foreach>) [perl #114020].
365
366=head2 Lexical C<$_> is now experimental
367
368Since it was introduced in Perl 5.10, it has caused much confusion with no
369obvious solution:
370
371=over
829397b9
TC
372
373=item *
374
5ed58cbd
RS
375Various modules (e.g., List::Util) expect callback routines to use the
376global C<$_>. C<use List::Util 'first'; my $_; first { $_ == 1 } @list>
377does not work as one would expect.
71e6aba6
RS
378
379=item *
380
5ed58cbd
RS
381A C<my $_> declaration earlier in the same file can cause confusing closure
382warnings.
71e6aba6
RS
383
384=item *
385
5ed58cbd
RS
386The "_" subroutine prototype character allows called subroutines to access
387your lexical C<$_>, so it is not really private after all.
2426c394 388
71e6aba6
RS
389=item *
390
5ed58cbd
RS
391Nevertheless, subroutines with a "(@)" prototype and methods cannot access
392the caller's lexical C<$_>, unless they are written in XS.
71e6aba6
RS
393
394=item *
395
5ed58cbd
RS
396But even XS routines cannot access a lexical C<$_> declared, not in the
397calling subroutine, but in an outer scope, iff that subroutine happened not
398to mention C<$_> or use any operators that default to C<$_>.
399
400=back
401
402It is our hope that lexical C<$_> can be rehabilitated, but this may
403cause changes in its behavior. Please use it with caution until it
404becomes stable.
405
406=head2 readline() with C<$/ = \N> now reads N characters, not N bytes
407
408Previously, when reading from a stream with I/O layers such as
409C<encoding>, the readline() function, otherwise known as the C<< <> >>
410operator, would read I<N> bytes from the top-most layer. [perl #79960]
411
412Now, I<N> characters are read instead.
413
414There is no change in behaviour when reading from streams with no
415extra layers, since bytes map exactly to characters.
416
417=head2 Overridden C<glob> is now passed one argument
418
419C<glob> overrides used to be passed a magical undocumented second argument
420that identified the caller. Nothing on CPAN was using this, and it got in
421the way of a bug fix, so it was removed. If you really need to identify
422the caller, see L<Devel::Callsite> on CPAN.
423
424=head2 Here-doc parsing
425
426The body of a here-document inside a quote-like operator now always begins
427on the line after the "<<foo" marker. Previously, it was documented to
428begin on the line following the containing quote-like operator, but that
429was only sometimes the case [perl #114040].
430
431=head2 Alphanumeric operators must now be separated from the closing
432delimiter of regular expressions
433
434You may no longer write something like:
435
436 m/a/and 1
437
438Instead you must write
439
440 m/a/ and 1
441
442with whitespace separating the operator from the closing delimiter of
443the regular expression. Not having whitespace has resulted in a
444deprecation warning since Perl v5.14.0.
445
446=head2 qw(...) can no longer be used as parentheses
447
448C<qw> lists used to fool the parser into thinking they were always
449surrounded by parentheses. This permitted some surprising constructions
450such as C<foreach $x qw(a b c) {...}>, which should really be written
451C<foreach $x (qw(a b c)) {...}>. These would sometimes get the lexer into
452the wrong state, so they didn't fully work, and the similar C<foreach qw(a
453b c) {...}> that one might expect to be permitted never worked at all.
454
455This side effect of C<qw> has now been abolished. It has been deprecated
456since Perl 5.13.11. It is now necessary to use real parentheses
457everywhere that the grammar calls for them.
458
459=head2 Interaction of lexical and default warnings
460
461Turning on any lexical warnings used first to disable all default warnings
462if lexical warnings were not already enabled:
463
464 $*; # deprecation warning
465 use warnings "void";
466 $#; # void warning; no deprecation warning
467
468Now, the debugging, deprecated, glob, inplace and malloc warnings
469categories are left on when turning on lexical warnings (unless they are
470turned off by C<no warnings>, of course).
471
472This may cause deprecation warnings to occur in code that used to be free
473of warnings.
474
475Those are the only categories consisting only of default warnings. Default
476warnings in other categories are still disabled by C<use warnings
477"category">, as we do not yet have the infrastructure for controlling
478individual warnings.
479
480=head2 C<state sub> and C<our sub>
481
482Due to an accident of history, C<state sub> and C<our sub> were equivalent
483to a plain C<sub>, so one could even create an anonymous sub with
484C<our sub { ... }>. These are now disallowed outside of the "lexical_subs"
485feature. Under the "lexical_subs" feature they have new meanings described
486in L<perlsub/Lexical Subroutines>.
487
488=head2 Defined values stored in environment are forced to byte strings
489
490A value stored in an environment variable has always been stringified. In this
491release, it is converted to be only a byte string. First, it is forced to be a
492only a string. Then if the string is utf8 and the equivalent of
493C<utf8::downgrade()> works, that result is used; otherwise, the equivalent of
494C<utf8::encode()> is used, and a warning is issued about wide characters
495(L</Diagnostics>).
496
497=head2 C<require> dies for unreadable files
498
499When C<require> encounters an unreadable file, it now dies. It used to
500ignore the file and continue searching the directories in C<@INC>
501[perl #113422].
502
503=head2 C<gv_fetchmeth_*> and SUPER
504
505The various C<gv_fetchmeth_*> XS functions used to treat a package whose
506named ended with ::SUPER specially. A method lookup on the Foo::SUPER
507package would be treated as a SUPER method lookup on the Foo package. This
508is no longer the case. To do a SUPER lookup, pass the Foo stash and the
509GV_SUPER flag.
510
511=head2 C<split>'s first argument is more consistently interpreted
512
513After some changes earlier in 5.17, C<split>'s behavior has been
514simplified: if the PATTERN argument evaluates to a literal string
515containing one space, it is treated the way that a I<literal> string
516containing one space once was.
517
518=head1 Deprecations
519
520=head2 Deprecated modules
521
522The following modules will be removed from the core distribution in a
523future release, and should be installed from CPAN instead. Distributions
524on CPAN which require these should add them to their prerequisites.
525The core versions of these modules will issue "deprecated"-category
526warnings.
527
528You can silence these deprecation warnings by installing the modules
529in question from CPAN.
530
531=over
532
533=item L<Archive::Extract>
534
535=item L<B::Lint>
536
537=item L<B::Lint::Debug>
538
539=item L<CPANPLUS> and all included C<CPANPLUS::*> modules
540
541=item L<Devel::InnerPackage>
542
543=item L<encoding>
544
545=item L<Log::Message>
546
547=item L<Log::Message::Config>
548
549=item L<Log::Message::Handlers>
550
551=item L<Log::Message::Item>
552
553=item L<Log::Message::Simple>
554
555=item L<Module::Pluggable>
556
557=item L<Module::Pluggable::Object>
558
559=item L<Object::Accessor>
560
561=item L<Pod::LaTeX>
562
563=item L<Term::UI>
564
565=item L<Term::UI::History>
566
567=back
568
569=head2 Deprecated Utilities
570
571The following utilities will be removed from the core distribution in a
572future release as their associated modules have been deprecated. They
573will remain available with the applicable CPAN distribution.
574
575=over
576
577=item L<cpanp>
578
579=item C<cpanp-run-perl>
580
581=item L<cpan2dist>
582
583These items are part of the C<CPANPLUS> distribution.
584
585=item L<pod2latex>
586
587This item is part of the C<Pod::LaTeX> distribution.
588
589=back
590
591=head2 PL_sv_objcount
592
593This interpreter-global variable used to track the total number of
594Perl objects in the interpreter. It is no longer maintained and will
595be removed altogether in Perl 5.20.
596
597=head2 Five additional characters should be escaped in patterns with C</x>
598
599When a regular expression pattern is compiled with C</x>, Perl treats 6
600characters as white space to ignore, such as SPACE and TAB. However,
601Unicode recommends 11 characters be treated thusly. We will conform
602with this in a future Perl version. In the meantime, use of any of the
603missing characters will raise a deprecation warning, unless turned off.
604The five characters are:
605U+0085 NEXT LINE,
606U+200E LEFT-TO-RIGHT MARK,
607U+200F RIGHT-TO-LEFT MARK,
608U+2028 LINE SEPARATOR,
609and
610U+2029 PARAGRAPH SEPARATOR.
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.
625See L<perlapi/Character classes>. The complete list (including some
626that were deprecated in 5.17.7) is:
627C<is_uni_alnum>, C<is_uni_alnumc>, C<is_uni_alnumc_lc>,
628C<is_uni_alnum_lc>, C<is_uni_alpha>, C<is_uni_alpha_lc>,
629C<is_uni_ascii>, C<is_uni_ascii_lc>, C<is_uni_blank>,
630C<is_uni_blank_lc>, C<is_uni_cntrl>, C<is_uni_cntrl_lc>,
631C<is_uni_digit>, C<is_uni_digit_lc>, C<is_uni_graph>,
632C<is_uni_graph_lc>, C<is_uni_idfirst>, C<is_uni_idfirst_lc>,
633C<is_uni_lower>, C<is_uni_lower_lc>, C<is_uni_print>,
634C<is_uni_print_lc>, C<is_uni_punct>, C<is_uni_punct_lc>,
635C<is_uni_space>, C<is_uni_space_lc>, C<is_uni_upper>,
636C<is_uni_upper_lc>, C<is_uni_xdigit>, C<is_uni_xdigit_lc>,
637C<is_utf8_alnum>, C<is_utf8_alnumc>, C<is_utf8_alpha>,
638C<is_utf8_ascii>, C<is_utf8_blank>, C<is_utf8_char>,
639C<is_utf8_cntrl>, C<is_utf8_digit>, C<is_utf8_graph>,
640C<is_utf8_idcont>, C<is_utf8_idfirst>, C<is_utf8_lower>,
641C<is_utf8_mark>, C<is_utf8_perl_space>, C<is_utf8_perl_word>,
642C<is_utf8_posix_digit>, C<is_utf8_print>, C<is_utf8_punct>,
643C<is_utf8_space>, C<is_utf8_upper>, C<is_utf8_xdigit>,
644C<is_utf8_xidcont>, C<is_utf8_xidfirst>.
645
646In addition these three functions that have never worked properly are
647deprecated:
648C<to_uni_lower_lc>, C<to_uni_title_lc>, and C<to_uni_upper_lc>.
649
650=head2 Certain rare uses of backslashes within regexes are now deprectated
651
652There are three pairs of characters that Perl recognizes as
653metacharacters in regular expression patterns: C<{}>, C<[]>, and C<()>.
654These can be used as well to delimit patterns, as in:
655
656 m{foo}
657 s(foo)(bar)
658
659Since they are metacharacters, they have special meaning to regular
660expression patterns, and it turns out that you can't turn off that
661special meaning by the normal means of preceding them with a backslash,
662if you use them, paired, within a pattern delimitted by them. For
663example, in
664
665 m{foo\{1,3\}}
666
667the backslashes do not change the behavior, and this matches
668S<C<"f o">> followed by one to three more occurrences of C<"o">.
669
670Usages like this, where they are interpreted as metacharacters, are
671exceedingly rare; we think there are none, for example, in all of CPAN.
672Hence, this deprecation should affect very little code. It does give
673notice, however, that any such code needs to change, which will in turn
674allow us to change the behavior in future Perl versions so that the
675backslashes do have an effect, and without fear that we are silently
676breaking any existing code.
677
d5f315e8
KW
678=head2 Splitting the tokens C<(?> and C<(*> in regular expressions
679
680A deprecation warning is now raised if the C<(> and C<?> are separated
681by white space or comments in C<(?...)> regular expression constructs.
682Similarly, if the C<(> and C<*> are separated in C<(*VERB...)>
683constructs.
684
5ed58cbd
RS
685=head1 Future Deprecations
686
687=over
71e6aba6
RS
688
689=item *
690
5ed58cbd
RS
691Platforms with out support infrastructure
692
693Both Windows CE and z/OS have been historically under-maintained, and are
694currently neither successfully building nor regularly being smoke tested.
695Efforts are underway to change this situation, but it should not be taken for
696granted that the platforms are safe and supported. If they do not become
697buildable and regularly smoked, support for them may be actively removed in
698future releases. If you have an interest in these platforms and you can lend
699your time, expertise, or hardware to help support these platforms, please let
700the perl development effort know by emailing C<perl5-porters@perl.org>.
701
702Some platforms that appear otherwise entirely dead are also on the short list
703for removal between now and 5.20.0:
704
705=over
706
707=item DG/UX
708
709=item NeXT
710
711=back
1993add8
RS
712
713=item *
714
5ed58cbd
RS
715Swapping of $< and $>
716
717For more information about this future deprecation, see L<the relevant RT
718ticket|https://rt.perl.org/rt3/Ticket/Display.html?id=96212>.
71e6aba6
RS
719
720=item *
721
5ed58cbd
RS
722sfio, stdio
723
724Perl supports being built without PerlIO proper, using a stdio or sfio
725wrapper instead. A perl build like this will not support IO layers and
726thus Unicode IO, making it rather handicapped.
727
728PerlIO supports a C<stdio> layer if stdio use is desired, and similarly a
729sfio layer could be produced.
71e6aba6
RS
730
731=item *
732
5ed58cbd 733C<microperl>, long broken and of unclear present purpose, will be removed.
71e6aba6
RS
734
735=item *
736
5ed58cbd
RS
737Revamping C<< "\Q" >> semantics in double-quotish strings when combined with
738other escapes.
739
740There are several bugs and inconsistencies involving combinations
741of C<\Q> and escapes like C<\x>, C<\L>, etc., within a C<\Q...\E> pair.
742These need to be fixed, and doing so will necessarily change current
743behavior. The changes have not yet been settled.
71e6aba6 744
d5f315e8
KW
745=item *
746
747Use of C<$^>, where C<^> stands for any actual (non-printing) C0 control
748character will be disallowed in a future Perl version. Use C<${^}>
749instead (where again C<^> stands for a control character),
750or better, C<$^A> , where C<^> this time is a caret (CIRCUMFLEX ACCENT),
751and C<A> stands for any of the characters listed at the end of
752L<perlebcdic/OPERATOR DIFFERENCES>.
753
337fb649 754=back
2426c394 755
5ed58cbd
RS
756=head1 Performance Enhancements
757
758=over 4
2426c394 759
5ed58cbd 760=item *
2426c394 761
5ed58cbd
RS
762Lists of lexical variable declarations (C<my($x, $y)>) are now optimised
763down to a single op and are hence faster than before.
2426c394 764
5ed58cbd 765=item *
2426c394 766
5ed58cbd
RS
767A new C preprocessor define C<NO_TAINT_SUPPORT> was added that, if set,
768disables Perl's taint support altogether. Using the -T or -t command
769line flags will cause a fatal error. Beware that both core tests as
770well as many a CPAN distribution's tests will fail with this change. On
771the upside, it provides a small performance benefit due to reduced
772branching.
2426c394 773
5ed58cbd
RS
774B<Do not enable this unless you know exactly what you are getting yourself
775into.>
776
777=item *
778
779C<pack> with constant arguments is now constant folded in most cases
780[perl #113470].
781
782=item *
783
784Speed up in regular expression matching against Unicode properties. The
785largest gain is for C<\X>, the Unicode "extended grapheme cluster." The
786gain for it is about 35% - 40%. Bracketed character classes, e.g.,
787C<[0-9\x{100}]> containing code points above 255 are also now faster.
788
789=item *
790
791On platforms supporting it, several former macros are now implemented as static
792inline functions. This should speed things up slightly on non-GCC platforms.
793
794=item *
795
796Apply the optimisation of hashes in boolean context, such as in C<if> or
797C<and>, to constructs in non-void context.
798
799=item *
800
801Extend the optimisation of hashes in boolean context to C<scalar(%hash)>,
802C<%hash ? ... : ...>, and C<sub { %hash || ... }>.
803
804=item *
805
806Filetest ops manage the stack in a fractionally more efficient manner.
807
808=item *
809
810Globs used in a numeric context are now numified directly in most cases,
811rather than being numerified via stringification.
812
813=item *
814
815The C<x> repetition operator is now folded to a single constant at compile
816time if called in scalar context with constant operands and no parentheses
817around the left operand.
818
819=back
820
821=head1 Modules and Pragmata
822
823=head2 New Modules and Pragmata
2426c394 824
337fb649 825=over 4
982110e0 826
82d98f72 827=item *
2426c394 828
5ed58cbd
RS
829L<Config::Perl::V> version 0.16 has been added as a dual-lifed module.
830It provides structured data retrieval of C<perl -V> output including
831information only known to the C<perl> binary and not available via L<Config>.
832
833=back
834
835=head2 Updated Modules and Pragmata
836
837This is only an overview of selected module updates. For a complete
838list of updates, run:
839
840 $ corelist --diff 5.14.0 5.16.0
841
842You can substitute your favorite version in place of 5.14.0, too.
843
844=over 4
33392251
BF
845
846=item *
847
5ed58cbd
RS
848L<XXX> has been upgraded from version A.xx to B.yy.
849
850=back
851
852=head2 Removed Modules and Pragmata
853
854=over
33392251
BF
855
856=item *
857
5ed58cbd
RS
858L<Version::Requirements> has been removed from the core distribution. It is
859available under a different name: L<CPAN::Meta::Requirements>.
2426c394 860
337fb649 861=back
2426c394 862
5ed58cbd 863=head1 Documentation
19718730 864
5ed58cbd
RS
865=head2 Changes to Existing Documentation
866
867=head3 L<perlcheat>
82d98f72 868
5a6a30f4 869=over 4
b7c7d786 870
5ed58cbd
RS
871=item *
872
873L<perlcheat> has been reorganized, and a few new sections were added.
874
875=back
876
877=head3 L<perldata>
878
879=over 4
82d98f72 880
5ed58cbd 881=item *
d2d1e842 882
5ed58cbd
RS
883Now explicitly documents the behaviour of hash initializer lists that
884contain duplicate keys.
f355e93d 885
5a6a30f4 886=back
f355e93d 887
5ed58cbd 888=head3 L<perldiag>
19718730 889
19718730 890=over 4
e14ac59b 891
5ed58cbd
RS
892=item *
893
894The explanation of symbolic references being prevented by "strict refs"
895now doesn't assume that the reader knows what symbolic references are.
896
897=back
9f351b45 898
5ed58cbd 899=head3 L<perlfaq>
9f351b45 900
5ed58cbd 901=over 4
9f351b45 902
5ed58cbd 903=item *
7cf3104f 904
5ed58cbd 905L<perlfaq> has been synchronized with version 5.0150040 from CPAN.
12719193 906
6253ee75 907=back
216cf7fc 908
5ed58cbd 909=head3 L<perlfunc>
f5b73711 910
5ed58cbd
RS
911=over 4
912
913=item *
a75569c0 914
5ed58cbd 915The return value of C<pipe> is now documented.
a75569c0 916
5ed58cbd 917=item *
a75569c0 918
5ed58cbd
RS
919Clarified documentation of C<our>.
920
921=back
922
923=head3 L<perlop>
924
925=over 4
926
927=item *
928
929Loop control verbs (C<dump>, C<goto>, C<next>, C<last> and C<redo>) have always
930had the same precedence as assignment operators, but this was not documented
931until now.
932
933=back
934
935=head3 Diagnostics
936
937The following additions or changes have been made to diagnostic output,
938including warnings and fatal error messages. For the complete list of
939diagnostic messages, see L<perldiag>.
940
941XXX New or changed warnings emitted by the core's C<C> code go here. Also
942include any changes in L<perldiag> that reconcile it to the C<C> code.
943
944=head2 New Diagnostics
945
946XXX Newly added diagnostic messages go under here, separated into New Errors
947and New Warnings
948
949=head3 New Errors
950
951=over 4
952
953=item *
954
955L<Unterminated delimiter for here document|perldiag/"Unterminated delimiter for here document">
956
957This message now occurs when a here document label has an initial quotation
958mark but the final quotation mark is missing.
959
960This replaces a bogus and misleading error message about not finding the label
961itself [perl #114104].
962
963=item *
964
965L<panic: child pseudo-process was never scheduled|perldiag/"panic: child pseudo-process was never scheduled">
966
967This error is thrown when a child pseudo-process in the ithreads implementation
968on Windows was not scheduled within the time period allowed and therefore was
969not able to initialize properly [perl #88840].
970
971=item *
972
973L<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/">
974
975This error has been added for C<(?&0)>, which is invalid. It used to
976produce an incomprehensible error message [perl #101666].
977
978=item *
979
980L<Can't use an undefined value as a subroutine reference|perldiag/"Can't use an undefined value as %s reference">
981
982Calling an undefined value as a subroutine now produces this error message.
983It used to, but was accidentally disabled, first in Perl 5.004 for
984non-magical variables, and then in Perl 5.14 for magical (e.g., tied)
985variables. It has now been restored. In the mean time, undef was treated
986as an empty string [perl #113576].
987
988=item *
989
990L<Experimental "%s" subs not enabled|perldiag/"Experimental "%s" subs not enabled">
991
992To use lexical subs, you must first enable them:
993
994 no warnings 'experimental::lexical_subs';
995 use feature 'lexical_subs';
996 my sub foo { ... }
997
998=back
999
1000=head3 New Warnings
1001
1002=over 4
1003
1004=item *
1005
1006XXX: This needs more detail.
1007
1008Strings with code points over 0xFF may not be mapped into in-memory file
1009handles
1010
1011=item *
1012
1013L<'%s' resolved to '\o{%s}%d'|perldiag/"'%s' resolved to '\o{%s}%d'">
1014
1015=item *
1016
1017L<'Trailing white-space in a charnames alias definition is deprecated'|perldiag/"Trailing white-space in a charnames alias definition is deprecated">
1018
1019=item *
1020
1021L<'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">
1022
1023=item *
1024
1025L<'Passing malformed UTF-8 to "%s" is deprecated'|perldiag/"Passing malformed UTF-8 to "%s" is deprecated">
1026
1027=item *
1028
1029L<Subroutine "&%s" is not available|perldiag/"Subroutine "&%s" is not available">
1030
1031(W closure) During compilation, an inner named subroutine or eval is
1032attempting to capture an outer lexical subroutine that is not currently
1033available. This can happen for one of two reasons. First, the lexical
1034subroutine may be declared in an outer anonymous subroutine that has not
1035yet been created. (Remember that named subs are created at compile time,
1036while anonymous subs are created at run-time.) For example,
1037
1038 sub { my sub a {...} sub f { \&a } }
1039
1040At the time that f is created, it can't capture the current the "a" sub,
1041since the anonymous subroutine hasn't been created yet. Conversely, the
1042following won't give a warning since the anonymous subroutine has by now
1043been created and is live:
1044
1045 sub { my sub a {...} eval 'sub f { \&a }' }->();
1046
1047The second situation is caused by an eval accessing a variable that has
1048gone out of scope, for example,
1049
1050 sub f {
1051 my sub a {...}
1052 sub { eval '\&a' }
1053 }
1054 f()->();
1055
1056Here, when the '\&a' in the eval is being compiled, f() is not currently
1057being executed, so its &a is not available for capture.
1058
1059=item *
1060
1061L<"%s" subroutine &%s masks earlier declaration in same %s|perldiag/"%s" subroutine &%s masks earlier declaration in same %s>
1062
1063(W misc) A "my" or "state" subroutine has been redeclared in the
1064current scope or statement, effectively eliminating all access to
1065the previous instance. This is almost always a typographical error.
1066Note that the earlier subroutine will still exist until the end of
1067the scope or until all closure references to it are destroyed.
1068
1069=item *
1070
1071L<The %s feature is experimental|perldiag/"The %s feature is experimental">
1072
1073(S experimental) This warning is emitted if you enable an experimental
1074feature via C<use feature>. Simply suppress the warning if you want
1075to use the feature, but know that in doing so you are taking the risk
1076of using an experimental feature which may change or be removed in a
1077future Perl version:
1078
1079 no warnings "experimental::lexical_subs";
1080 use feature "lexical_subs";
1081
1082=item *
1083
1084L<sleep(%u) too large|perldiag/"sleep(%u) too large">
1085
1086(W overflow) You called C<sleep> with a number that was larger than it can
1087reliably handle and C<sleep> probably slept for less time than requested.
1088
1089=item *
1090
1091L<Wide character in setenv|perldiag/"Wide character in %s">
1092
1093Attempts to put wide characters into environment variables via C<%ENV> now
1094provoke this warning.
1095
1096=item *
1097
1098"L<Invalid negative number (%s) in chr|perldiag/"Invalid negative number (%s) in chr">"
1099
1100C<chr()> now warns when passed a negative value [perl #83048].
1101
1102=item *
1103
1104"L<Integer overflow in srand|perldiag/"Integer overflow in srand">"
1105
1106C<srand()> now warns when passed a value that doesn't fit in a C<UV> (since the
1107value will be truncated rather than overflowing) [perl #40605].
1108
1109=item *
1110
1111"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">"
1112
1113Running perl with the C<-i> flag now warns if no input files are provided on
1114the command line [perl #113410].
1115
1116=back
1117
1118=head2 Changes to Existing Diagnostics
1119
1120=over 4
1121
1122=item *
1123
1124L<$* is no longer supported|perldiag/"$* is no longer supported">
1125
1126The warning that use of C<$*> and C<$#> is no longer supported is now
1127generated for every location that references them. Previously it would fail
1128to be generated if another variable using the same typeglob was seen first
1129(e.g. C<@*> before C<$*>), and would not be generated for the second and
1130subsequent uses. (It's hard to fix the failure to generate warnings at all
1131without also generating them every time, and warning every time is
1132consistent with the warnings that C<$[> used to generate.)
1133
1134=item *
1135
1136The warnings for C<\b{> and C<\B{> were added. They are a deprecation
1137warning which should be turned off by that category. One should not
1138have to turn off regular regexp warnings as well to get rid of these.
1139
1140=item *
1141
1142L<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>
1143
1144Constant overloading that returns C<undef> results in this error message.
1145For numeric constants, it used to say "Constant(undef)". "undef" has been
1146replaced with the number itself.
1147
1148=item *
1149
1150The error produced when a module cannot be loaded now includes a hint that
1151the module may need to be installed: "Can't locate hopping.pm in @INC (you
1152may need to install the hopping module) (@INC contains: ...)"
1153
1154=item *
1155
1156L<vector argument not supported with alpha versions|perldiag/vector argument not supported with alpha versions>
1157
1158This warning was not suppressable, even with C<no warnings>. Now it is
1159suppressible, and has been moved from the "internal" category to the
1160"printf" category.
1161
1162=item *
1163
1164C<< Can't do {n,m} with n > m in regex; marked by <-- HERE in m/%s/ >>
1165
1166This fatal error has been turned into a warning that reads:
1167
1168L<< Quantifier {n,m} with n > m can't match in regex | perldiag/Quantifier {n,m} with n > m can't match in regex >>
1169
1170(W regexp) Minima should be less than or equal to maxima. If you really want
1171your regexp to match something 0 times, just put {0}.
1172
1173=item *
1174
1175The "Runaway prototype" warning that occurs in bizarre cases has been
1176removed as being unhelpful and inconsistent.
1177
1178=item *
1179
1180The "Not a format reference" error has been removed, as the only case in
1181which it could be triggered was a bug.
1182
1183=item *
1184
1185The "Unable to create sub named %s" error has been removed for the same
1186reason.
1187
1188=item *
1189
1190The 'Can't use "my %s" in sort comparison' error has been downgraded to a
1191warning, '"my %s" used in sort comparison' (with 'state' instead of 'my'
1192for state variables). In addition, the heuristics for guessing whether
1193lexical $a or $b has been misused have been improved to generate fewer
1194false positives. Lexical $a and $b are no longer disallowed if they are
1195outside the sort block. Also, a named unary or list operator inside the
1196sort block no longer causes the $a or $b to be ignored [perl #86136].
1197
1198=back
1199
1200=head1 Utility Changes
1201
1202=head3 L<h2xs>
1203
1204=over 4
1205
1206=item *
1207
1208F<h2xs> no longer produces invalid code for empty defines. [perl #20636]
1209
1210=back
1211
1212=head1 Configuration and Compilation
1213
1214=over 4
1215
1216=item *
1217
1218Added C<useversionedarchname> option to Configure
1219
1220When set, it includes 'api_versionstring' in 'archname'. E.g.
1221x86_64-linux-5.13.6-thread-multi. It is unset by default.
1222
1223This feature was requested by Tim Bunce, who observed that
1224INSTALL_BASE creates a library structure that does not
1225differentiate by perl version. Instead, it places architecture
1226specific files in "$install_base/lib/perl5/$archname". This makes
1227it difficult to use a common INSTALL_BASE library path with
1228multiple versions of perl.
1229
1230By setting -Duseversionedarchname, the $archname will be
c2959982 1231distinct for architecture I<and> API version, allowing mixed use of
5ed58cbd
RS
1232INSTALL_BASE.
1233
1234=item *
1235
ff772877
RS
1236Add a C<PERL_NO_INLINE_FUNCTIONS> option
1237
1238If PERL_NO_INLINE_FUNCTIONS is defined, don't include "inline.h"
1239
1240This permits test code to include the perl headers for definitions without
1241creating a link dependency on the perl library (which may not exist yet).
1242
1243=item *
1244
5ed58cbd
RS
1245Configure will honour the external C<MAILDOMAIN> environment variable, if set.
1246
1247=item *
1248
1249C<installman> no longer ignores the silent option
1250
1251=item *
1252
1253Both C<META.yml> and C<META.json> files are now included in the distribution.
1254
1255=item *
1256
1257F<Configure> will now correctly detect C<isblank()> when compiling with a C++
1258compiler.
1259
1260=item *
1261
1262The pager detection in F<Configure> has been improved to allow responses which
1263specify options after the program name, e.g. B</usr/bin/less -R>, if the user
1264accepts the default value. This helps B<perldoc> when handling ANSI escapes
1265[perl #72156].
1266
1267=back
1268
1269=head1 Testing
1270
1271=over 4
1272
1273=item *
1274
1275The test suite now has a section for tests that require very large amounts
1276of memory. These tests won't run by default; they can be enabled by
1277setting the C<PERL_TEST_MEMORY> environment variable to the number of
1278gibibytes of memory that may be safely used.
1279
1280=back
1281
1282=head1 Platform Support
1283
1284=head2 Discontinued Platforms
1285
1286=over 4
1287
1288=item BeOS
1289
1290BeOS was an operating system for personal computers developed by Be Inc,
1291initially for their BeBox hardware. The OS Haiku was written as an open
1292source replacement for/continuation of BeOS, and its perl port is current and
1293actively maintained.
1294
1295=item UTS Global
1296
1297Support code relating to UTS global has been removed. UTS was a mainframe
1298version of System V created by Amdahl, subsequently sold to UTS Global. The
1299port has not been touched since before Perl 5.8.0, and UTS Global is now
1300defunct.
1301
1302=item VM/ESA
1303
1304Support for VM/ESA has been removed. The port was tested on 2.3.0, which
1305IBM ended service on in March 2002. 2.4.0 ended service in June 2003, and
1306was superseded by Z/VM. The current version of Z/VM is V6.2.0, and scheduled
1307for end of service on 2015/04/30.
1308
1309=item MPE/IX
1310
1311Support for MPE/IX has been removed.
1312
1313=item EPOC
1314
1315Support code relating to EPOC has been removed. EPOC was a family of
1316operating systems developed by Psion for mobile devices. It was the
1317predecessor of Symbian. The port was last updated in April 2002.
1318
1319=item Rhapsody
1320
1321Support for Rhapsody has been removed.
1322
1323=back
1324
1325=head2 Platform-Specific Notes
1326
1327=head3 AIX
1328
1329Configure now always adds C<-qlanglvl=extc99> to the CC flags on AIX when
1330using xlC. This will make it easier to compile a number of XS-based modules
1331that assume C99 [perl #113778].
1332
1333=head3 clang++
1334
1335There is now a workaround for a compiler bug that prevented compiling
1336with clang++ since Perl 5.15.7 [perl #112786].
1337
1338=head3 C++
1339
1340When compiling the Perl core as C++ (which is only semi-supported), the
1341mathom functions are now compiled as C<extern "C">, to ensure proper
1342binary compatibility. (However, binary compatibility isn't generally
1343guaranteed anyway in the situations where this would matter.)
1344
1345=head3 Darwin
1346
1347Stop hardcoding an alignment on 8 byte boundaries to fix builds using
1348-Dusemorebits.
1349
1350=head3 Haiku
1351
1352Perl should now work out of the box on Haiku R1 Alpha 4.
1353
1354=head3 MidnightBSD
1355
1356C<libc_r> was removed from recent versions of MidnightBSD and older versions
1357work better with C<pthread>. Threading is now enabled using C<pthread> which
1358corrects build errors with threading enabled on 0.4-CURRENT.
1359
1360=head3 Solaris
1361
1362In Configure, avoid running sed commands with flags not supported on Solaris.
1363
1364=head3 VMS
1365
1366=over
1367
1368=item *
1369
1370Where possible, the case of filenames and command-line arguments is now
1371preserved by enabling the CRTL features C<DECC$EFS_CASE_PRESERVE> and
1372C<DECC$ARGV_PARSE_STYLE> at start-up time. The latter only takes effect
1373when extended parse is enabled in the process from which Perl is run.
1374
1375=item *
1376
1377The character set for Extended Filename Syntax (EFS) is now enabled by default
1378on VMS. Among other things, this provides better handling of dots in directory
1379names, multiple dots in filenames,and spaces in filenames. To obtain the old
1380behavior, set the logical name C<DECC$EFS_CHARSET> to C<DISABLE>.
1381
1382=item *
1383
1384Fix linking on builds configured with -Dusemymalloc=y.
1385
1386=item *
1387
1388It should now be possible to compile Perl as C++ on VMS.
1389
1390=item *
1391
1392All C header files from the top-level directory of the distribution are now
1393installed on VMS, providing consistency with a long-standing practice on other
1394platforms. Previously only a subset were installed, which broke non-core
1395extension builds for extensions that depended on the missing include files.
1396
1397=item *
1398
1399Quotes are now removed from the command verb (but not the parameters) for
1400commands spawned via C<system>, backticks, or a piped C<open>. Previously,
1401quotes on the verb were passed through to DCL, which would fail to recognize
1402the command. Also, if the verb is actually a path to an image or command
1403procedure on an ODS-5 volume, quoting it now allows the path to contain spaces.
1404
1405=item *
1406
1407The B<a2p> build has been fixed for the HP C++ compiler on OpenVMS.
1408
1409=back
1410
1411=head3 Win32
1412
1413=over
1414
1415=item *
1416
1417Perl can now be built using Microsoft's Visual C++ 2012 compiler by specifying
1418CCTYPE=MSVC110 (or MSVC110FREE if you are using the free Express edition for
1419Windows Desktop) in F<win32/Makefile>.
1420
1421=item *
1422
1423The option to build without USE_SOCKETS_AS_HANDLES has been removed.
1424
1425=item *
1426
1427Fixed a problem where perl could crash while cleaning up threads (including the
1428main thread) in threaded debugging builds on Win32 and possibly other platforms
1429[perl #114496].
1430
1431=item *
1432
1433A rare race condition that would lead to L<sleep|perlfunc/sleep> taking more
1434time than requested, and possibly even hanging, has been fixed [perl #33096].
1435
1436=item *
1437
1438C<link> on Win32 now attempts to set C<$!> to more appropriate values
1439based on the Win32 API error code. [perl #112272]
1440
1441Perl no longer mangles the environment block, e.g. when launching a new
1442sub-process, when the environment contains non-ASCII characters. Known
1443problems still remain, however, when the environment contains characters
1444outside of the current ANSI codepage (e.g. see the item about Unicode in
1445C<%ENV> in L<http://perl5.git.perl.org/perl.git/blob/HEAD:/Porting/todo.pod>).
1446[perl #113536]
1447
1448=item *
1449
1450Building perl with some Windows compilers used to fail due to a problem
1451with miniperl's C<glob> operator (which uses the C<perlglob> program)
1452deleting the PATH environment variable [perl #113798].
1453
1454=item *
1455
1456A new makefile option, USE_64_BIT_INT, has been added to the Windows
1457makefiles. Set this to "define" when building a 32-bit perl if you want
1458it to use 64-bit integers.
1459
1460Machine code size reductions, already made to the DLLs of XS modules in
1461Perl 5.17.2, have now been extended to the perl DLL itself.
1462
1463Building with VC++ 6.0 was inadvertently broken in Perl 5.17.2 but has
1464now been fixed again.
1465
1466=back
1467
1468=head3 WinCE
1469
1470Building on WinCE is now possible once again, although more work is required
1471to fully restore a clean build.
1472
1473=head1 Internal Changes
1474
1475=over
1476
1477=item *
1478
1479Synonyms for the misleadingly named C<av_len()> has been created:
1480C<av_top_index()> and C<av_tindex>. All three of these return the
1481number of the highest index in the array, not the number of elements it
1482contains.
1483
1484=item *
1485
1486SvUPGRADE() is no longer an expression. Originally this macro (and its
1487underlying function, sv_upgrade()) were documented as boolean, although
1488in reality they always croaked on error and never returned false. In 2005
1489the documentation was updated to specify a void return value, but
1490SvUPGRADE() was left always returning 1 for backwards compatibility. This
1491has now been removed, and SvUPGRADE() is now a statement with no return
1492value.
1493
1494So this is now a syntax error:
1495
1496 if (!SvUPGRADE(sv)) { croak(...); }
1497
1498If you have code like that, simply replace it with
1499
1500 SvUPGRADE(sv);
1501
1502or to to avoid compiler warnings with older perls, possibly
1503
1504 (void)SvUPGRADE(sv);
1505
1506=item *
1507
1508Perl has a new copy-on-write mechanism that allows any SvPOK scalar to be
1509upgraded to a copy-on-write scalar. A reference count on the string buffer
1510is stored in the string buffer itself.
1511
1512This breaks a few XS modules by allowing copy-on-write scalars to go
1513through code paths that never encountered them before.
1514
1515This behaviour can still be disabled by running F<Configure> with
1516B<-Accflags=-DPERL_NO_COW>. This option will probably be removed in Perl
15175.20.
1518
1519=item *
1520
1521Copy-on-write no longer uses the SvFAKE and SvREADONLY flags. Hence,
1522SvREADONLY indicates a true read-only SV.
1523
1524Use the SvIsCOW macro (as before) to identify a copy-on-write scalar.
1525
1526=item *
1527
1528PL_glob_index is gone.
1529
1530=item *
1531
1532The private Perl_croak_no_modify has had its context parameter removed. It is
1533now has a void prototype. Users of the public API croak_no_modify remain
1534unaffected.
1535
1536=item *
1537
1538Copy-on-write (shared hash key) scalars are no longer marked read-only.
1539C<SvREADONLY> returns false on such an SV, but C<SvIsCOW> still returns
1540true.
1541
1542=item *
1543
1544A new op type, C<OP_PADRANGE> has been introduced. The perl peephole
1545optimiser will, where possible, substitute a single padrange op for a
1546pushmark followed by one or more pad ops, and possibly also skipping list
1547and nextstate ops. In addition, the op can carry out the tasks associated
1548with the RHS of a my(...) = @_ assignment, so those ops may be optimised
1549away too.
1550
1551=item *
1552
1553Case-insensitive matching inside a [bracketed] character class with a
1554multi-character fold no longer excludes one of the possibilities in the
1555circumstances that it used to. [perl #89774].
1556
1557=item *
1558
1559C<PL_formfeed> has been removed.
1560
1561=item *
1562
1563The regular expression engine no longer reads one byte past the end of the
1564target string. While for all internally well-formed scalars this should
1565never have been a problem, this change facilitates clever tricks with
1566string buffers in CPAN modules. [perl #73542]
1567
1568=item *
1569
1570Inside a BEGIN block, C<PL_compcv> now points to the currently-compiling
1571subroutine, rather than the BEGIN block itself.
1572
1573=item *
1574
1575C<mg_length> has been deprecated.
1576
1577=item *
1578
1579C<sv_len> now always returns a byte count and C<sv_len_utf8> a character
1580count. Previously, C<sv_len> and C<sv_len_utf8> were both buggy and would
1581sometimes returns bytes and sometimes characters. C<sv_len_utf8> no longer
1582assumes that its argument is in UTF8. Neither of these creates UTF8 caches
1583for tied or overloaded values or for non-PVs any more.
1584
1585=item *
1586
1587C<sv_mortalcopy> now copies string buffers of shared hash key scalars when
1588called from XS modules [perl #79824].
1589
1590=item *
1591
1592C<RXf_SPLIT> and C<RXf_SKIPWHITE> are no longer used. They are now
1593#defined as 0.
1594
1595=item *
1596
1597The new C<RXf_MODIFIES_VARS> flag can be set by custom regular expression
1598engines to indicate that the execution of the regular expression may cause
1599variables to be modified. This lets C<s///> know to skip certain
1600optimisations. Perl's own regular expression engine sets this flag for the
1601special backtracking verbs that set $REGMARK and $REGERROR.
1602
1603=item *
1604
1605The APIs for accessing lexical pads have changed considerably.
1606
1607C<PADLIST>s are now longer C<AV>s, but their own type instead.
1608C<PADLIST>s now contain a C<PAD> and a C<PADNAMELIST> of C<PADNAME>s,
1609rather than C<AV>s for the pad and the list of pad names. C<PAD>s,
1610C<PADNAMELIST>s, and C<PADNAME>s are to be accessed as such through the
1611newly added pad API instead of the plain C<AV> and C<SV> APIs. See
1612L<perlapi> for details.
1613
1614=item *
1615
1616In the regex API, the numbered capture callbacks are passed an index
1617indicating what match variable is being accessed. There are special
1618index values for the C<$`, $&, $&> variables. Previously the same three
1619values were used to retrieve C<${^PREMATCH}, ${^MATCH}, ${^POSTMATCH}>
1620too, but these have now been assigned three separate values. See
1621L<perlreapi/Numbered capture callbacks>.
1622
1623=item *
1624
1625C<PL_sawampersand> was previously a boolean indicating that any of
1626C<$`, $&, $&> had been seen; it now contains three one-bit flags
1627indicating the presence of each of the variables individually.
1628
1629=item *
1630
1631The C<CV *> typemap entry now supports C<&{}> overloading and typeglobs,
1632just like C<&{...}> [perl #96872].
1633
1634=item *
1635
1636The C<SVf_AMAGIC> flag to indicate overloading is now on the stash, not the
1637object. It is now set automatically whenever a method or @ISA changes, so
1638its meaning has changed, too. It now means "potentially overloaded". When
1639the overload table is calculated, the flag is automatically turned off if
1640there is no overloading, so there should be no noticeable slowdown.
1641
1642The staleness of the overload tables is now checked when overload methods
1643are invoked, rather than during C<bless>.
1644
1645"A" magic is gone. The changes to the handling of the C<SVf_AMAGIC> flag
1646eliminate the need for it.
1647
1648C<PL_amagic_generation> has been removed as no longer necessary. For XS
1649modules, it is now a macro alias to C<PL_na>.
1650
1651The fallback overload setting is now stored in a stash entry separate from
1652overloadedness itself.
1653
1654=item *
1655
1656The character-processing code has been cleaned up in places. The changes
1657should be operationally invisible.
1658
1659=item *
1660
1661The C<study> function was made a no-op in 5.16. It was simply disabled via
1662a C<return> statement; the code was left in place. Now the code supporting
1663what C<study> used to do has been removed.
1664
1665=item *
1666
1667Under threaded perls, there is no longer a separate PV allocated for every
1668COP to store its package name (C<< cop->stashpv >>). Instead, there is an
1669offset (C<< cop->stashoff >>) into the new C<PL_stashpad> array, which
1670holds stash pointers.
1671
1672=item *
1673
1674In the pluggable regex API, the C<regexp_engine> struct has acquired a new
1675field C<op_comp>, which is currently just for perl's internal use, and
1676should be initialised to NULL by other regex plugin modules.
1677
1678=item *
1679
1680A new function C<alloccoptash> has been added to the API, but is considered
1681experimental. See L<perlapi>.
1682
1683=item *
1684
1685Perl used to implement get magic in a way that would sometimes hide bugs in
1686code could call mg_get() too many times on magical values. This hiding of
1687errors no longer occurs, so long-standing bugs may become visible now. If
1688you see magic-related errors in XS code, check to make sure it, together
1689with the Perl API functions it uses, calls mg_get() only once on SvGMAGICAL()
1690values.
1691
1692=item *
1693
1694OP allocation for CVs now uses a slab allocator. This simplifies
1695memory management for OPs allocated to a CV, so cleaning up after a
1696compilation error is simpler and safer [perl #111462][perl #112312].
1697
1698=item *
1699
1700PERL_DEBUG_READONLY_OPS has been rewritten to work with the new slab
1701allocator, allowing it to catch more violations than before.
1702
1703=item *
1704
1705The old slab allocator for ops, which was only enabled for PERL_IMPLICIT_SYS
1706and PERL_DEBUG_READONLY_OPS, has been retired.
1707
1708=back
1709
1710=head1 Selected Bug Fixes
1711
1712=over 4
1713
1714=item *
1715
1716Here-doc terminators no longer require a terminating newline character when
1717they occur at the end of a file. This was already the case at the end of a
1718string eval [perl #65838].
1719
1720=item *
1721
1722-DPERL_GLOBAL_STRUCT builds now free the global struct B<after>
1723they've finished using it.
1724
1725=item *
1726
1727A trailing '/' on a path in @INC will no longer have an additional '/'
1728appended.
1729
1730=item *
1731
1732The C<:crlf> layer now works when unread data doesn't fit into its own
1733buffer. [perl #112244].
1734
1735=item *
1736
1737C<ungetc()> now handles UTF-8 encoded data. [perl #116322].
1738
1739=item *
1740
1741A bug in the core typemap caused any C types that map to the T_BOOL core
1742typemap entry to not be set, updated, or modified when the T_BOOL variable was
1743used in an OUTPUT: section with an exception for RETVAL. T_BOOL in an INPUT:
1744section was not affected. Using a T_BOOL return type for an XSUB (RETVAL)
1745was not affected. A side effect of fixing this bug is, if a T_BOOL is specified
1746in the OUTPUT: section (which previous did nothing to the SV), and a read only
1747SV (literal) is passed to the XSUB, croaks like "Modification of a read-only
1748value attempted" will happen. [perl #115796]
1749
1750=item *
1751
1752On many platforms, providing a directory name as the script name caused perl
1753to do nothing and report success. It should now universally report an error
1754and exit nonzero. [perl #61362]
1755
1756=item *
1757
1758C<sort {undef} ...> under fatal warnings no longer crashes. It had
1759begun crashing in Perl 5.16.
1760
1761=item *
1762
1763Stashes blessed into each other
1764(C<bless \%Foo::, 'Bar'; bless \%Bar::, 'Foo'>) no longer result in double
1765frees. This bug started happening in Perl 5.16.
1766
1767=item *
1768
1769Numerous memory leaks have been fixed, mostly involving fatal warnings and
1770syntax errors.
1771
1772=item *
1773
1774Some failed regular expression matches such as C<'f' =~ /../g> were not
1775resetting C<pos>. Also, "match-once" patterns (C<m?...?g>) failed to reset
1776it, too, when invoked a second time [perl #23180].
1777
1778=item *
1779
1780Accessing C<$&> after a pattern match now works if it had not been seen
1781before the match. I.e., this applies to C<${'&'}> (under C<no strict>) and
1782C<eval '$&'>. The same applies to C<$'> and C<$`> [perl #4289].
1783
1784=item *
1785
1786Several bugs involving C<local *ISA> and C<local *Foo::> causing stale
1787MRO caches have been fixed.
1788
1789=item *
1790
1791Defining a subroutine when its typeglob has been aliased no longer results
1792in stale method caches. This bug was introduced in Perl 5.10.
1793
1794=item *
1795
1796Localising a typeglob containing a subroutine when the typeglob's package
1797has been deleted from its parent stash no longer produces an error. This
1798bug was introduced in Perl 5.14.
1799
1800=item *
1801
1802Under some circumstances, C<local *method=...> would fail to reset method
1803caches upon scope exit.
1804
1805=item *
1806
1807C</[.foo.]/> is no longer an error, but produces a warning (as before) and
1808is treated as C</[.fo]/> [perl #115818].
1809
1810=item *
1811
1812C<goto $tied_var> now calls FETCH before deciding what type of goto
1813(subroutine or label) this is.
1814
1815=item *
1816
1817Renaming packages through glob assignment
1818(C<*Foo:: = *Bar::; *Bar:: = *Baz::>) in combination with C<m?...?> and
1819C<reset> no longer makes threaded builds crash.
1820
1821=item *
1822
1823A number of bugs related to assigning a list to hash have been fixed. Many of
1824these involve lists with repeated keys like C<(1, 1, 1, 1)>.
1825
1826=over 4
1827
1828=item *
1829
1830The expression C<scalar(%h = (1, 1, 1, 1))> now returns C<4>, not C<2>.
1831
1832=item *
1833
1834The return value of C<%h = (1, 1, 1)> in list context was wrong. Previously
1835this would return C<(1, undef, 1)>, now it returns C<(1, undef)>.
1836
1837=item *
1838
1839Perl now issues the same warning on C<($s, %h) = (1, {})> as it does for
1840C<(%h) = ({})>, "Reference found where even-sized list expected".
1841
1842=item *
1843
1844A number of additional edge cases in list assignment to hashes were
1845corrected. For more details see commit 23b7025ebc.
1846
1847=back
1848
1849=item *
1850
1851Attributes applied to lexical variables no longer leak memory.
1852[perl #114764]
1853
1854=item *
1855
1856C<dump>, C<goto>, C<last>, C<next>, C<redo> or C<require> followed by a
1857bareword (or version) and then an infix operator is no longer a syntax
1858error. It used to be for those infix operators (like C<+>) that have a
1859different meaning where a term is expected. [perl #105924]
1860
1861=item *
1862
1863C<require a::b . 1> and C<require a::b + 1> no longer produce erroneous
1864ambiguity warnings. [perl #107002]
1865
1866=item *
1867
1868Class method calls are now allowed on any string, and not just strings
1869beginning with an alphanumeric character. [perl #105922]
1870
1871=item *
1872
1873An empty pattern created with C<qr//> used in C<m///> no longer triggers
1874the "empty pattern reuses last pattern" behaviour. [perl #96230]
1875
1876=item *
1877
1878Tying a hash during iteration no longer results in a memory leak.
1879
1880=item *
1881
1882Freeing a tied hash during iteration no longer results in a memory leak.
1883
1884=item *
1885
1886List assignment to a tied array or hash that dies on STORE no longer
1887results in a memory leak.
1888
1889=item *
1890
1891If the hint hash (C<%^H>) is tied, compile-time scope entry (which copies
1892the hint hash) no longer leaks memory if FETCH dies. [perl #107000]
1893
1894=item *
1895
1896Constant folding no longer inappropriately triggers the special
1897C<split " "> behaviour. [perl #94490]
1898
1899=item *
1900
1901C<defined scalar(@array)>, C<defined do { &foo }>, and similar constructs
1902now treat the argument to C<defined> as a simple scalar. [perl #97466]
1903
1904=item *
1905
1906Running a custom debugging that defines no C<*DB::DB> glob or provides a
1907subroutine stub for C<&DB::DB> no longer results in a crash, but an error
1908instead. [perl #114990]
1909
1910=item *
1911
1912C<reset ""> now matches its documentation. C<reset> only resets C<m?...?>
1913patterns when called with no argument. An empty string for an argument now
1914does nothing. (It used to be treated as no argument.) [perl #97958]
1915
1916=item *
1917
1918C<printf> with an argument returning an empty list no longer reads past the
1919end of the stack, resulting in erratic behaviour. [perl #77094]
1920
1921=item *
1922
1923C<--subname> no longer produces erroneous ambiguity warnings.
1924[perl #77240]
1925
1926=item *
1927
1928C<v10> is now allowed as a label or package name. This was inadvertently
1929broken when v-strings were added in Perl 5.6. [perl #56880]
1930
1931=item *
1932
1933C<length>, C<pos>, C<substr> and C<sprintf> could be confused by ties,
1934overloading, references and typeglobs if the stringification of such
1935changed the internal representation to or from UTF8. [perl #114410]
1936
1937=item *
1938
1939utf8::encode now calls FETCH and STORE on tied variables. utf8::decode now
1940calls STORE (it was already calling FETCH).
1941
1942=item *
1943
1944C<$tied =~ s/$non_utf8/$utf8/> no longer loops infinitely if the tied
1945variable returns a Latin-1 string, shared hash key scalar, or reference or
2ae351f8 1946typeglob that stringifies as ASCII or Latin-1. This was a regression from
5ed58cbd
RS
19475.12.x.
1948
1949=item *
1950
1951C<s///> without /e is now better at detecting when it needs to forego
1952certain optimisations, fixing some buggy cases:
1953
1954=over
1955
1956=item *
1957
1958Match variables in certain constructs (C<&&>, C<||>, C<..> and others) in
1959the replacement part; e.g., C<s/(.)/$l{$a||$1}/g>. [perl #26986]
1960
1961=item *
1962
1963Aliases to match variables in the replacement.
1964
1965=item *
1966
1967C<$REGERROR> or C<$REGMARK> in the replacement. [perl #49190]
1968
1969=item *
1970
1971An empty pattern (C<s//$foo/>) that causes the last-successful pattern to
1972be used, when that pattern contains code blocks that modify the variables
1973in the replacement.
1974
1975=back
1976
1977=item *
1978
1979The taintedness of the replacement string no longer affects the taintedness
1980of the return value of C<s///e>.
1981
1982=item *
1983
1984The C<$|> autoflush variable is created on-the-fly when needed. If this
1985happened (e.g., if it was mentioned in a module or eval) when the
1986currently-selected filehandle was a typeglob with an empty IO slot, it used
1987to crash. [perl #115206]
1988
1989=item *
1990
1991Line numbers at the end of a string eval are no longer off by one.
1992[perl #114658]
1993
1994=item *
1995
1996@INC filters (subroutines returned by subroutines in @INC) that set $_ to a
1997copy-on-write scalar no longer cause the parser to modify that string
1998buffer in place.
1999
2000=item *
2001
2002C<length($object)> no longer returns the undefined value if the object has
2003string overloading that returns undef. [perl #115260]
2004
2005=item *
2006
2007The use of C<PL_stashcache>, the stash name lookup cache for method calls, has
2008been restored,
2009
2010Commit da6b625f78f5f133 in August 2011 inadvertently broke the code that looks
2011up values in C<PL_stashcache>. As it's a only cache, quite correctly everything
2012carried on working without it.
2013
2014=item *
2015
2016The error "Can't localize through a reference" had disappeared in 5.16.0
2017when C<local %$ref> appeared on the last line of an lvalue subroutine.
2018This error disappeared for C<\local %$ref> in perl 5.8.1. It has now
2019been restored.
2020
2021=item *
2022
2023The parsing of here-docs has been improved significantly, fixing several
2024parsing bugs and crashes and one memory leak, and correcting wrong
2025subsequent line numbers under certain conditions.
2026
2027=item *
2028
2029Inside an eval, the error message for an unterminated here-doc no longer
2030has a newline in the middle of it [perl #70836].
2031
2032=item *
2033
2034A substitution inside a substitution pattern (C<s/${s|||}//>) no longer
2035confuses the parser.
2036
2037=item *
2038
2039It may be an odd place to allow comments, but C<s//"" # hello/e> has
2040always worked, I<unless> there happens to be a null character before the
2041first #. Now it works even in the presence of nulls.
2042
2043=item *
2044
2045An invalid range in C<tr///> or C<y///> no longer results in a memory leak.
2046
2047=item *
2048
2049String eval no longer treats a semicolon-delimited quote-like operator at
2050the very end (C<eval 'q;;'>) as a syntax error.
2051
2052=item *
2053
2054C<< warn {$_ => 1} + 1 >> is no longer a syntax error. The parser used to
2055get confused with certain list operators followed by an anonymous hash and
2056then an infix operator that shares its form with a unary operator.
2057
2058=item *
2059
2060C<(caller $n)[6]> (which gives the text of the eval) used to return the
2061actual parser buffer. Modifying it could result in crashes. Now it always
2062returns a copy. The string returned no longer has "\n;" tacked on to the
2063end. The returned text also includes here-doc bodies, which used to be
2064omitted.
2065
2066=item *
2067
2068Reset the utf8 position cache when accessing magical variables to avoid the
2069string buffer and the utf8 position cache getting out of sync
2070[perl #114410].
2071
2072=item *
2073
2074Various cases of get magic being called twice for magical utf8 strings have been
2075fixed.
2076
2077=item *
2078
2079This code (when not in the presence of C<$&> etc)
2080
2081 $_ = 'x' x 1_000_000;
2082 1 while /(.)/;
2083
2084used to skip the buffer copy for performance reasons, but suffered from C<$1>
2085etc changing if the original string changed. That's now been fixed.
2086
2087=item *
2088
2089Perl doesn't use PerlIO anymore to report out of memory messages, as PerlIO
2090might attempt to allocate more memory.
2091
2092=item *
2093
2094In a regular expression, if something is quantified with C<{n,m}> where
2095C<S<n E<gt> m>>, it can't possibly match. Previously this was a fatal
2096error, but now is merely a warning (and that something won't match).
2097[perl #82954].
2098
2099=item *
2100
2101It used to be possible for formats defined in subroutines that have
2102subsequently been undefined and redefined to close over variables in the
2103wrong pad (the newly-defined enclosing sub), resulting in crashes or
2104"Bizarre copy" errors.
2105
2106=item *
2107
2108Redefinition of XSUBs at run time could produce warnings with the wrong
2109line number.
2110
2111=item *
2112
2113The %vd sprintf format does not support version objects for alpha versions.
2114It used to output the format itself (%vd) when passed an alpha version, and
2115also emit an "Invalid conversion in printf" warning. It no longer does,
2116but produces the empty string in the output. It also no longer leaks
2117memory in this case.
2118
2119=item *
2120
2121C<< $obj->SUPER::method >> calls in the main package could fail if the
2122SUPER package had already been accessed by other means.
2123
2124=item *
2125
2126Stash aliasing (C<*foo:: = *bar::>) no longer causes SUPER calls to ignore
2127changes to methods or @ISA or use the wrong package.
2128
2129=item *
2130
2131Method calls on packages whose names end in ::SUPER are no longer treated
2132as SUPER method calls, resulting in failure to find the method.
2133Furthermore, defining subroutines in such packages no longer causes them to
2134be found by SUPER method calls on the containing package [perl #114924].
2135
2136=item *
2137
2138C<\w> now matches the code points U+200C (ZERO WIDTH NON-JOINER) and U+200D
2139(ZERO WIDTH JOINER). C<\W> no longer matches these. This change is because
2140Unicode corrected their definition of what C<\w> should match.
2141
2142=item *
2143
2144C<dump LABEL> no longer leaks its label.
2145
2146=item *
2147
2148Constant folding no longer changes the behaviour of functions like C<stat()>
2149and C<truncate()> that can take either filenames or handles.
2150C<stat 1 ? foo : bar> nows treats its argument as a file name (since it is an
2151arbitrary expression), rather than the handle "foo".
2152
2153=item *
2154
2155C<truncate FOO, $len> no longer falls back to treating "FOO" as a file name if
2156the filehandle has been deleted. This was broken in Perl 5.16.0.
2157
2158=item *
2159
2160Subroutine redefinitions after sub-to-glob and glob-to-glob assignments no
2161longer cause double frees or panic messages.
2162
2163=item *
2164
2165C<s///> now turns vstrings into plain strings when performing a substitution,
2166even if the resulting string is the same (C<s/a/a/>).
2167
2168=item *
2169
2170Prototype mismatch warnings no longer erroneously treat constant subs as having
2171no prototype when they actually have "".
2172
2173=item *
2174
2175Constant subroutines and forward declarations no longer prevent prototype
2176mismatch warnings from omitting the sub name.
2177
2178=item *
2179
2180C<undef> on a subroutine now clears call checkers.
2181
2182=item *
2183
2184The C<ref> operator started leaking memory on blessed objects in Perl 5.16.0.
2185This has been fixed [perl #114340].
2186
2187=item *
2188
2189C<use> no longer tries to parse its arguments as a statement, making
2190C<use constant { () };> a syntax error [perl #114222].
2191
2192=item *
2193
2194On debugging builds, "uninitialized" warnings inside formats no longer cause
2195assertion failures.
2196
2197=item *
2198
2199On debugging builds, subroutines nested inside formats no longer cause
2200assertion failures [perl #78550].
2201
2202=item *
2203
2204Formats and C<use> statements are now permitted inside formats.
2205
2206=item *
2207
2208C<print $x> and C<sub { print $x }-E<gt>()> now always produce the same output.
2209It was possible for the latter to refuse to close over $x if the variable was
2210not active; e.g., if it was defined outside a currently-running named
2211subroutine.
2212
2213=item *
2214
2215Similarly, C<print $x> and C<print eval '$x'> now produce the same output.
2216This also allows "my $x if 0" variables to be seen in the debugger [perl
2217#114018].
2218
2219=item *
2220
2221Formats called recursively no longer stomp on their own lexical variables, but
2222each recursive call has its own set of lexicals.
2223
2224=item *
2225
2226Attempting to free an active format or the handle associated with it no longer
2227results in a crash.
2228
2229=item *
2230
2231Format parsing no longer gets confused by braces, semicolons and low-precedence
2232operators. It used to be possible to use braces as format delimiters (instead
2233of C<=> and C<.>), but only sometimes. Semicolons and low-precedence operators
2234in format argument lines no longer confuse the parser into ignoring the line's
2235return value. In format argument lines, braces can now be used for anonymous
2236hashes, instead of being treated always as C<do> blocks.
2237
2238=item *
2239
2240Formats can now be nested inside code blocks in regular expressions and other
2241quoted constructs (C</(?{...})/> and C<qq/${...}/>) [perl #114040].
2242
2243=item *
2244
2245Formats are no longer created after compilation errors.
2246
2247=item *
2248
2249Under debugging builds, the B<-DA> command line option started crashing in Perl
22505.16.0. It has been fixed [perl #114368].
2251
2252=item *
2253
2254A potential deadlock scenario involving the premature termination of a pseudo-
2255forked child in a Windows build with ithreads enabled has been fixed. This
2256resolves the common problem of the F<t/op/fork.t> test hanging on Windows [perl
2257#88840].
2258
2259=item *
2260
2261The microperl build, broken since Perl 5.15.7, has now been restored.
2262
2263=item *
2264
2265The code which generates errors from C<require()> could potentially read one or
2266two bytes before the start of the filename for filenames less than three bytes
2267long and ending C</\.p?\z/>. This has now been fixed. Note that it could
2268never have happened with module names given to C<use()> or C<require()> anyway.
2269
2270=item *
2271
2272The handling of pathnames of modules given to C<require()> has been made
2273thread-safe on VMS.
2274
2275=item *
2276
d85cd26b
RS
2277Non-blocking sockets have been fixed on VMS.
2278
2279=item *
2280
5ed58cbd
RS
2281A bug in the compilation of a C</(?{})/> expression which affected the TryCatch
2282test suite has been fixed [perl #114242].
2283
2284=item *
2285
2286Pod can now be nested in code inside a quoted construct outside of a string
2287eval. This used to work only within string evals [perl #114040].
2288
2289=item *
2290
2291C<goto ''> now looks for an empty label, producing the "goto must have
2292label" error message, instead of exiting the program [perl #111794].
2293
2294=item *
2295
2296C<goto "\0"> now dies with "Can't find label" instead of "goto must have
2297label".
2298
2299=item *
2300
2301The C function C<hv_store> used to result in crashes when used on C<%^H>
2302[perl #111000].
2303
2304=item *
2305
2306A call checker attached to a closure prototype via C<cv_set_call_checker>
2307is now copied to closures cloned from it. So C<cv_set_call_checker> now
2308works inside an attribute handler for a closure.
2309
2310=item *
2311
2312Writing to C<$^N> used to have no effect. Now it croaks with "Modification
2313of a read-only value" by default, but that can be overridden by a custom
2314regular expression engine, as with C<$1> [perl #112184].
2315
2316=item *
2317
2318C<undef> on a control character glob (C<undef *^H>) no longer emits an
2319erroneous warning about ambiguity [perl #112456].
2320
2321=item *
2322
2323For efficiency's sake, many operators and built-in functions return the
2324same scalar each time. Lvalue subroutines and subroutines in the CORE::
2325namespace were allowing this implementation detail to leak through.
2326C<print &CORE::uc("a"), &CORE::uc("b")> used to print "BB". The same thing
2327would happen with an lvalue subroutine returning the return value of C<uc>.
2328Now the value is copied in such cases.
2329
2330=item *
2331
2332C<method {}> syntax with an empty block or a block returning an empty list
2333used to crash or use some random value left on the stack as its invocant.
2334Now it produces an error.
2335
2336=item *
2337
2338C<vec> now works with extremely large offsets (E<gt>2 GB) [perl #111730].
2339
2340=item *
2341
2342Changes to overload settings now take effect immediately, as do changes to
2343inheritance that affect overloading. They used to take effect only after
2344C<bless>.
2345
2346Objects that were created before a class had any overloading used to remain
2347non-overloaded even if the class gained overloading through C<use overload>
2348or @ISA changes, and even after C<bless>. This has been fixed
2349[perl #112708].
2350
2351=item *
2352
2353Classes with overloading can now inherit fallback values.
2354
2355=item *
2356
2357Overloading was not respecting a fallback value of 0 if there were
2358overloaded objects on both sides of an assignment operator like C<+=>
2359[perl #111856].
2360
2361=item *
2362
2363C<pos> now croaks with hash and array arguments, instead of producing
2364erroneous warnings.
2365
2366=item *
2367
2368C<while(each %h)> now implies C<while(defined($_ = each %h))>, like
2369C<readline> and C<readdir>.
2370
2371=item *
2372
2373Subs in the CORE:: namespace no longer crash after C<undef *_> when called
2374with no argument list (C<&CORE::time> with no parentheses).
2375
2376=item *
2377
2378C<unpack> no longer produces the "'/' must follow a numeric type in unpack"
2379error when it is the data that are at fault [perl #60204].
2380
2381=item *
2382
2383C<join> and C<"@array"> now call FETCH only once on a tied C<$">
2384[perl #8931].
2385
2386=item *
2387
2388Some subroutine calls generated by compiling core ops affected by a
2389C<CORE::GLOBAL> override had op checking performed twice. The checking
2390is always idempotent for pure Perl code, but the double checking can
2391matter when custom call checkers are involved.
2392
2393=item *
2394
2395A race condition used to exist around fork that could cause a signal sent to
2396the parent to be handled by both parent and child. Signals are now blocked
2397briefly around fork to prevent this from happening [perl #82580].
2398
2399=item *
2400
2401The implementation of code blocks in regular expressions, such as C<(?{})>
2402and C<(??{})>, has been heavily reworked to eliminate a whole slew of bugs.
2403The main user-visible changes are:
2404
2405=over 4
2406
2407=item *
2408
2409Code blocks within patterns are now parsed in the same pass as the
2410surrounding code; in particular it is no longer necessary to have balanced
2411braces: this now works:
2412
2413 /(?{ $x='{' })/
2414
2415This means that this error message is no longer generated:
2416
2417 Sequence (?{...}) not terminated or not {}-balanced in regex
2418
2419but a new error may be seen:
2420
2421 Sequence (?{...}) not terminated with ')'
2422
2423In addition, literal code blocks within run-time patterns are only
2424compiled once, at perl compile-time:
2425
2426 for my $p (...) {
2427 # this 'FOO' block of code is compiled once,
2428 # at the same time as the surrounding 'for' loop
2429 /$p{(?{FOO;})/;
2430 }
2431
2432=item *
2433
2434Lexical variables are now sane as regards scope, recursion and closure
2435behavior. In particular, C</A(?{B})C/> behaves (from a closure viewpoint)
2436exactly like C</A/ && do { B } && /C/>, while C<qr/A(?{B})C/> is like
2437C<sub {/A/ && do { B } && /C/}>. So this code now works how you might
2438expect, creating three regexes that match 0, 1, and 2:
2439
2440 for my $i (0..2) {
2441 push @r, qr/^(??{$i})$/;
2442 }
2443 "1" =~ $r[1]; # matches
2444
2445=item *
2446
2447The C<use re 'eval'> pragma is now only required for code blocks defined
2448at runtime; in particular in the following, the text of the C<$r> pattern is
2449still interpolated into the new pattern and recompiled, but the individual
2450compiled code-blocks within C<$r> are reused rather than being recompiled,
2451and C<use re 'eval'> isn't needed any more:
2452
2453 my $r = qr/abc(?{....})def/;
2454 /xyz$r/;
2455
2456=item *
2457
2458Flow control operators no longer crash. Each code block runs in a new
2459dynamic scope, so C<next> etc. will not see
2460any enclosing loops. C<return> returns a value
2461from the code block, not from any enclosing subroutine.
2462
2463=item *
2464
2465Perl normally caches the compilation of run-time patterns, and doesn't
2466recompile if the pattern hasn't changed, but this is now disabled if
2467required for the correct behavior of closures. For example:
2468
2469 my $code = '(??{$x})';
2470 for my $x (1..3) {
2471 # recompile to see fresh value of $x each time
2472 $x =~ /$code/;
2473 }
2474
2475=item *
2476
2477The C</msix> and C<(?msix)> etc. flags are now propagated into the return
2478value from C<(??{})>; this now works:
2479
2480 "AB" =~ /a(??{'b'})/i;
2481
2482=item *
2483
2484Warnings and errors will appear to come from the surrounding code (or for
2485run-time code blocks, from an eval) rather than from an C<re_eval>:
2486
2487 use re 'eval'; $c = '(?{ warn "foo" })'; /$c/;
2488 /(?{ warn "foo" })/;
2489
2490formerly gave:
2491
2492 foo at (re_eval 1) line 1.
2493 foo at (re_eval 2) line 1.
2494
2495and now gives:
2496
2497 foo at (eval 1) line 1.
2498 foo at /some/prog line 2.
2499
2500=back
2501
2502=item *
2503
2504Perl now works as well as can be expected on all releases of Unicode so
2505far. In v5.16, it worked on Unicodes 6.0 and 6.1, but there were
2506various bugs for earlier releases; the older the release the more
2507problems.
2508
2509=item *
2510
2511C<vec> no longer produces "uninitialized" warnings in lvalue context
2512[perl #9423].
2513
2514=item *
2515
2516An optimization involving fixed strings in regular expressions could cause
2517a severe performance penalty in edge cases. This has been fixed
2518[perl #76546].
2519
2520=item *
2521
2522In certain cases, including empty subpatterns within a regular expression (such
2523as C<(?:)> or C<(?:|)>) could disable some optimizations. This has been fixed.
2524
2525=item *
2526
2527The "Can't find an opnumber" message that C<prototype> produces when passed
2528a string like "CORE::nonexistent_keyword" now passes UTF-8 and embedded
2529NULs through unchanged [perl #97478].
2530
2531=item *
2532
2533C<prototype> now treats magical variables like C<$1> the same way as
2534non-magical variables when checking for the CORE:: prefix, instead of
2535treating them as subroutine names.
2536
2537=item *
2538
2539Under threaded perls, a runtime code block in a regular expression could
2540corrupt the package name stored in the op tree, resulting in bad reads
2541in C<caller>, and possibly crashes [perl #113060].
2542
2543=item *
2544
2545Referencing a closure prototype (C<\&{$_[1]}> in an attribute handler for a
2546closure) no longer results in a copy of the subroutine (or assertion
2547failures on debugging builds).
2548
2549=item *
2550
2551C<eval '__PACKAGE__'> now returns the right answer on threaded builds if
2552the current package has been assigned over (as in
2553C<*ThisPackage:: = *ThatPackage::>) [perl #78742].
2554
2555=item *
2556
2557If a package is deleted by code that it calls, it is possible for C<caller>
2558to see a stack frame belonging to that deleted package. C<caller> could
2559crash if the stash's memory address was reused for a scalar and a
2560substitution was performed on the same scalar [perl #113486].
2561
2562=item *
2563
2564C<UNIVERSAL::can> no longer treats its first argument differently
2565depending on whether it is a string or number internally.
2566
2567=item *
2568
2569C<open> with C<< <& >> for the mode checks to see whether the third argument is
2570a number, in determining whether to treat it as a file descriptor or a handle
2571name. Magical variables like C<$1> were always failing the numeric check and
2572being treated as handle names.
2573
2574=item *
2575
2576C<warn>'s handling of magical variables (C<$1>, ties) has undergone several
2577fixes. C<FETCH> is only called once now on a tied argument or a tied C<$@>
2578[perl #97480]. Tied variables returning objects that stringify as "" are
2579no longer ignored. A tied C<$@> that happened to return a reference the
2580I<previous> time it was used is no longer ignored.
2581
2582=item *
2583
2584C<warn ""> now treats C<$@> with a number in it the same way, regardless of
2585whether it happened via C<$@=3> or C<$@="3">. It used to ignore the
2586former. Now it appends "\t...caught", as it has always done with
2587C<$@="3">.
2588
2589=item *
2590
2591Numeric operators on magical variables (e.g., S<C<$1 + 1>>) used to use
2592floating point operations even where integer operations were more appropriate,
2593resulting in loss of accuracy on 64-bit platforms [perl #109542].
2594
2595=item *
2596
2597Unary negation no longer treats a string as a number if the string happened
2598to be used as a number at some point. So, if C<$x> contains the string "dogs",
2599C<-$x> returns "-dogs" even if C<$y=0+$x> has happened at some point.
2600
2601=item *
2602
2603In Perl 5.14, C<-'-10'> was fixed to return "10", not "+10". But magical
2604variables (C<$1>, ties) were not fixed till now [perl #57706].
2605
2606=item *
2607
2608Unary negation now treats strings consistently, regardless of the internal
2609C<UTF8> flag.
2610
2611=item *
2612
2613A regression introduced in Perl v5.16.0 involving
2614C<tr/I<SEARCHLIST>/I<REPLACEMENTLIST>/> has been fixed. Only the first
2615instance is supposed to be meaningful if a character appears more than
2616once in C<I<SEARCHLIST>>. Under some circumstances, the final instance
2617was overriding all earlier ones. [perl #113584]
2618
2619=item *
2620
2621Regular expressions like C<qr/\87/> previously silently inserted a NUL
2622character, thus matching as if it had been written C<qr/\00087/>. Now it
2623matches as if it had been written as C<qr/87/>, with a message that the
2624sequence C<"\8"> is unrecognized.
2625
2626=item *
2627
2628C<__SUB__> now works in special blocks (C<BEGIN>, C<END>, etc.).
2629
2630=item *
2631
2632Thread creation on Windows could theoretically result in a crash if done
2633inside a C<BEGIN> block. It still does not work properly, but it no longer
2634crashes [perl #111610].
2635
2636=item *
2637
2638C<\&{''}> (with the empty string) now autovivifies a stub like any other
2639sub name, and no longer produces the "Unable to create sub" error
2640[perl #94476].
2641
2642=item *
2643
2644A regression introduced in v5.14.0 has been fixed, in which some calls
2645to the C<re> module would clobber C<$_> [perl #113750].
2646
2647=item *
2648
2649C<do FILE> now always either sets or clears C<$@>, even when the file can't be
2650read. This ensures that testing C<$@> first (as recommended by the
2651documentation) always returns the correct result.
2652
2653=item *
2654
2655The array iterator used for the C<each @array> construct is now correctly
2656reset when C<@array> is cleared (RT #75596). This happens for example when the
2657array is globally assigned to, as in C<@array = (...)>, but not when its
2658B<values> are assigned to. In terms of the XS API, it means that C<av_clear()>
2659will now reset the iterator.
2660
2661This mirrors the behaviour of the hash iterator when the hash is cleared.
2662
2663=item *
2664
2665C<< $class->can >>, C<< $class->isa >>, and C<< $class->DOES >> now return
2666correct results, regardless of whether that package referred to by C<$class>
2667exists [perl #47113].
2668
2669=item *
2670
2671Arriving signals no longer clear C<$@> [perl #45173].
2672
2673=item *
2674
2675Allow C<my ()> declarations with an empty variable list [perl #113554].
2676
2677=item *
2678
2679During parsing, subs declared after errors no longer leave stubs
2680[perl #113712].
2681
2682=item *
2683
2684Closures containing no string evals no longer hang on to their containing
2685subroutines, allowing variables closed over by outer subroutines to be
2686freed when the outer sub is freed, even if the inner sub still exists
2687[perl #89544].
2688
2689=item *
2690
2691Duplication of in-memory filehandles by opening with a "<&=" or ">&=" mode
2692stopped working properly in 5.16.0. It was causing the new handle to
2693reference a different scalar variable. This has been fixed [perl #113764].
2694
2695=item *
2696
2697C<qr//> expressions no longer crash with custom regular expression engines
2698that do not set C<offs> at regular expression compilation time
2699[perl #112962].
2700
2701=item *
2702
2703C<delete local> no longer crashes with certain magical arrays and hashes
2704[perl #112966].
2705
2706=item *
2707
2708C<local> on elements of certain magical arrays and hashes used not to
2709arrange to have the element deleted on scope exit, even if the element did
2710not exist before C<local>.
2711
2712=item *
2713
2714C<scalar(write)> no longer returns multiple items [perl #73690].
2715
2716=item *
2717
2718String to floating point conversions no longer misparse certain strings under
2719C<use locale> [perl #109318].
2720
2721=item *
2722
2723C<@INC> filters that die no longer leak memory [perl #92252].
2724
2725=item *
2726
2727The implementations of overloaded operations are now called in the correct
2728context. This allows, among other things, being able to properly override
2729C<< <> >> [perl #47119].
2730
2731=item *
2732
2733Specifying only the C<fallback> key when calling C<use overload> now behaves
2734properly [perl #113010].
2735
2736=item *
2737
2738C<< sub foo { my $a = 0; while ($a) { ... } } >> and
2739C<< sub foo { while (0) { ... } } >> now return the same thing [perl #73618].
2740
2741=item *
2742
2743String negation now behaves the same under C<use integer;> as it does
2744without [perl #113012].
2745
2746=item *
2747
2748C<chr> now returns the Unicode replacement character (U+FFFD) for -1,
2749regardless of the internal representation. -1 used to wrap if the argument
2750was tied or a string internally.
2751
2752=item *
2753
2754Using a C<format> after its enclosing sub was freed could crash as of
2755perl 5.12.0, if the format referenced lexical variables from the outer sub.
2756
2757=item *
2758
2759Using a C<format> after its enclosing sub was undefined could crash as of
2760perl 5.10.0, if the format referenced lexical variables from the outer sub.
2761
2762=item *
2763
2764Using a C<format> defined inside a closure, which format references
2765lexical variables from outside, never really worked unless the C<write>
2766call was directly inside the closure. In 5.10.0 it even started crashing.
2767Now the copy of that closure nearest the top of the call stack is used to
2768find those variables.
2769
2770=item *
2771
2772Formats that close over variables in special blocks no longer crash if a
2773stub exists with the same name as the special block before the special
2774block is compiled.
2775
2776=item *
2777
2778The parser no longer gets confused, treating C<eval foo ()> as a syntax
2779error if preceded by C<print;> [perl #16249].
2780
2781=item *
2782
2783The return value of C<syscall> is no longer truncated on 64-bit platforms
2784[perl #113980].
2785
2786=item *
2787
2788Constant folding no longer causes C<print 1 ? FOO : BAR> to print to the
2789FOO handle [perl #78064].
2790
2791=item *
2792
2793C<do subname> now calls the named subroutine and uses the file name it
2794returns, instead of opening a file named "subname".
2795
2796=item *
2797
2798Subroutines looked up by rv2cv check hooks (registered by XS modules) are
2799now taken into consideration when determining whether C<foo bar> should be
2800the sub call C<foo(bar)> or the method call C<< "bar"->foo >>.
2801
2802=item *
2803
2804C<CORE::foo::bar> is no longer treated specially, allowing global overrides
2805to be called directly via C<CORE::GLOBAL::uc(...)> [perl #113016].
2806
2807=item *
2808
2809Calling an undefined sub whose typeglob has been undefined now produces the
2810customary "Undefined subroutine called" error, instead of "Not a CODE
2811reference".
2812
2813=item *
2814
2815Two bugs involving @ISA have been fixed. C<*ISA = *glob_without_array> and
2816C<undef *ISA; @{*ISA}> would prevent future modifications to @ISA from
2817updating the internal caches used to look up methods. The
2818*glob_without_array case was a regression from Perl 5.12.
2819
2820=item *
2821
2822Regular expression optimisations sometimes caused C<$> with C</m> to
2823produce failed or incorrect matches [perl #114068].
2824
2825=item *
2826
2827C<__SUB__> now works in a C<sort> block when the enclosing subroutine is
2828predeclared with C<sub foo;> syntax [perl #113710].
2829
2830=item *
2831
2832Unicode properties only apply to Unicode code points, which leads to
2833some subtleties when regular expressions are matched against
2834above-Unicode code points. There is a warning generated to draw your
2835attention to this. However, this warning was being generated
2836inappropriately in some cases, such as when a program was being parsed.
2837Non-Unicode matches such as C<\w> and C<[:word;]> should not generate the
2838warning, as their definitions don't limit them to apply to only Unicode
2839code points. Now the message is only generated when matching against
2840C<\p{}> and C<\P{}>. There remains a bug, [perl #114148], for the very
2841few properties in Unicode that match just a single code point. The
2842warning is not generated if they are matched against an above-Unicode
2843code point.
2844
2845=item *
2846
2847Uninitialized warnings mentioning hash elements would only mention the
2848element name if it was not in the first bucket of the hash, due to an
2849off-by-one error.
2850
2851=item *
2852
2853A regular expression optimizer bug could cause multiline "^" to behave
2854incorrectly in the presence of line breaks, such that
2855C<"/\n\n" =~ m#\A(?:^/$)#im> would not match [perl #115242].
2856
2857=item *
2858
2859Failed C<fork> in list context no longer corrupts the stack.
2860C<@a = (1, 2, fork, 3)> used to gobble up the 2 and assign C<(1, undef, 3)>
2861if the C<fork> call failed.
2862
2863=item *
2864
2865Numerous memory leaks have been fixed, mostly involving tied variables that
2866die, regular expression character classes and code blocks, and syntax
2867errors.
2868
2869=item *
2870
2871Assigning a regular expression (C<${qr//}>) to a variable that happens to
2872hold a floating point number no longer causes assertion failures on
2873debugging builds.
2874
2875=item *
2876
2877Assigning a regular expression to a scalar containing a number no longer
2878causes subsequent nummification to produce random numbers.
2879
2880=item *
2881
2882Assigning a regular expression to a magic variable no longer wipes away the
2ae351f8 2883magic. This was a regression from 5.10.
5ed58cbd
RS
2884
2885=item *
2886
2887Assigning a regular expression to a blessed scalar no longer results in
2ae351f8 2888crashes. This was also a regression from 5.10.
5ed58cbd
RS
2889
2890=item *
2891
2892Regular expression can now be assigned to tied hash and array elements with
2893flattening into strings.
2894
2895=item *
2896
2897Nummifying a regular expression no longer results in an uninitialized
2898warning.
2899
2900=item *
2901
2902Negative array indices no longer cause EXISTS methods of tied variables to
2ae351f8 2903be ignored. This was a regression from 5.12.
5ed58cbd
RS
2904
2905=item *
2906
2907Negative array indices no longer result in crashes on arrays tied to
2908non-objects.
2909
2910=item *
2911
2912C<$byte_overload .= $utf8> no longer results in doubly-encoded UTF8 if the
2913left-hand scalar happened to have produced a UTF8 string the last time
2914overloading was invoked.
2915
2916=item *
2917
2918C<goto &sub> now uses the current value of @_, instead of using the array
2919the subroutine was originally called with. This means
2920C<local @_ = (...); goto &sub> now works [perl #43077].
2921
2922=item *
2923
2924If a debugger is invoked recursively, it no longer stomps on its own
2925lexical variables. Formerly under recursion all calls would share the same
2926set of lexical variables [perl #115742].
2927
2928=item *
2929
2930C<*_{ARRAY}> returned from a subroutine no longer spontaneously
2931becomes empty.
2932
2933=back
2934
2935=head1 Known Problems
2936
2937=over 4
2938
2939=item *
2940
2941XXX: the imperfect behavior of the ** deprecation
2942
2943=back
2944
2945=head1 Acknowledgements
a75569c0 2946
5ed58cbd 2947XXX Generate this with:
a75569c0 2948
5ed58cbd 2949 perl Porting/acknowledgements.pl v5.18.0..HEAD
f5b73711 2950
44691e6f
AB
2951=head1 Reporting Bugs
2952
e08634c5
SH
2953If you find what you think is a bug, you might check the articles recently
2954posted to the comp.lang.perl.misc newsgroup and the perl bug database at
2955http://rt.perl.org/perlbug/ . There may also be information at
2956http://www.perl.org/ , the Perl Home Page.
44691e6f 2957
e08634c5
SH
2958If you believe you have an unreported bug, please run the L<perlbug> program
2959included with your release. Be sure to trim your bug down to a tiny but
2960sufficient test case. Your bug report, along with the output of C<perl -V>,
2961will be sent off to perlbug@perl.org to be analysed by the Perl porting team.
44691e6f
AB
2962
2963If the bug you are reporting has security implications, which make it
e08634c5
SH
2964inappropriate to send to a publicly archived mailing list, then please send it
2965to perl5-security-report@perl.org. This points to a closed subscription
2966unarchived mailing list, which includes all the core committers, who will be
2967able to help assess the impact of issues, figure out a resolution, and help
f9001595 2968co-ordinate the release of patches to mitigate or fix the problem across all
e08634c5
SH
2969platforms on which Perl is supported. Please only use this address for
2970security issues in the Perl core, not for modules independently distributed on
2971CPAN.
44691e6f
AB
2972
2973=head1 SEE ALSO
2974
e08634c5
SH
2975The F<Changes> file for an explanation of how to view exhaustive details on
2976what changed.
44691e6f
AB
2977
2978The F<INSTALL> file for how to build Perl.
2979
2980The F<README> file for general stuff.
2981
2982The F<Artistic> and F<Copying> files for copyright information.
2983
2984=cut