This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
nitpicks
[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
e612b5a0 9This document describes differences between the v5.16.0 release and the v5.18.0
e08634c5 10release.
6db9054f 11
e612b5a0
RS
12If you are upgrading from an earlier release such as v5.14.0, first read
13L<perl5160delta>, which describes differences between v5.14.0 and v5.16.0.
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
1ea0ff56
RS
40Since some features (like C<~~> or C<my $_>) now emit experimental warnings,
41and you may want to disable them in code that is also run on perls that do not
42recognize these warning categories, consider using the C<if> pragma like this:
43
5c46a73f 44 no if $] >= 5.018, 'warnings', "experimental::feature_name";
1ea0ff56 45
5ed58cbd
RS
46Existing experimental features may begin emitting these warnings, too. Please
47consult L<perlexperiment> for information on which features are considered
48experimental.
49
50=head2 Hash overhaul
51
e612b5a0 52Changes to the implementation of hashes in perl v5.18.0 will be one of the most
484d8ba1
DG
53visible changes to the behavior of existing code.
54
8e74b2ed 55By default, two distinct hash variables with identical keys and values may now
484d8ba1
DG
56provide their contents in a different order where it was previously identical.
57
58When encountering these changes, the key to cleaning up from them is to accept
59that B<hashes are unordered collections> and to act accordingly.
5ed58cbd
RS
60
61=head3 Hash randomization
62
63The seed used by Perl's hash function is now random. This means that the
64order which keys/values will be returned from functions like C<keys()>,
65C<values()>, and C<each()> will differ from run to run.
66
67This change was introduced to make Perl's hashes more robust to algorithmic
68complexity attacks, and also because we discovered that it exposes hash
69ordering dependency bugs and makes them easier to track down.
70
71Toolchain maintainers might want to invest in additional infrastructure to
72test for things like this. Running tests several times in a row and then
73comparing results will make it easier to spot hash order dependencies in
74code. Authors are strongly encouraged not to expose the key order of
75Perl's hashes to insecure audiences.
76
77Further, every hash has its own iteration order, which should make it much
78more difficult to determine what the current hash seed is.
79
0f2be12f
RS
80=head3 New hash functions
81
e612b5a0 82Perl v5.18 includes support for multiple hash functions, and changed
0f2be12f
RS
83the default (to ONE_AT_A_TIME_HARD), you can choose a different
84algorithm by defining a symbol at compile time. For a current list,
e612b5a0 85consult the F<INSTALL> document. Note that as of Perl v5.18 we can
0f2be12f
RS
86only recommend use of the default or SIPHASH. All the others are
87known to have security issues and are for research purposes only.
5ed58cbd 88
f105b7be 89=head3 PERL_HASH_SEED environment variable now takes a hex value
5ed58cbd 90
484d8ba1
DG
91C<PERL_HASH_SEED> no longer accepts an integer as a parameter;
92instead the value is expected to be a binary value encoded in a hex
93string, such as "0xf5867c55039dc724". This is to make the
94infrastructure support hash seeds of arbitrary lengths, which might
5ed58cbd
RS
95exceed that of an integer. (SipHash uses a 16 byte seed).
96
c40c48bb
RS
97=head3 PERL_PERTURB_KEYS environment variable added
98
f105b7be 99The C<PERL_PERTURB_KEYS> environment variable allows one to control the level of
c40c48bb
RS
100randomization applied to C<keys> and friends.
101
f105b7be 102When C<PERL_PERTURB_KEYS> is 0, perl will not randomize the key order at all. The
c40c48bb
RS
103chance that C<keys> changes due to an insert will be the same as in previous
104perls, basically only when the bucket size is changed.
105
f105b7be 106When C<PERL_PERTURB_KEYS> is 1, perl will randomize keys in a non-repeatable
c40c48bb
RS
107way. The chance that C<keys> changes due to an insert will be very high. This
108is the most secure and default mode.
109
f105b7be
KE
110When C<PERL_PERTURB_KEYS> is 2, perl will randomize keys in a repeatable way.
111Repeated runs of the same program should produce the same output every time.
c40c48bb 112
f105b7be
KE
113C<PERL_HASH_SEED> implies a non-default C<PERL_PERTURB_KEYS> setting. Setting
114C<PERL_HASH_SEED=0> (exactly one 0) implies C<PERL_PERTURB_KEYS=0> (hash key
115randomization disabled); settng C<PERL_HASH_SEED> to any other value implies
116C<PERL_PERTURB_KEYS=2> (deterministic and repeatable hash key randomization).
117Specifying C<PERL_PERTURB_KEYS> explicitly to a different level overrides this
c40c48bb
RS
118behavior.
119
5ed58cbd
RS
120=head3 Hash::Util::hash_seed() now returns a string
121
122Hash::Util::hash_seed() now returns a string instead of an integer. This
123is to make the infrastructure support hash seeds of arbitrary lengths
124which might exceed that of an integer. (SipHash uses a 16 byte seed).
125
126=head3 Output of PERL_HASH_SEED_DEBUG has been changed
127
128The environment variable PERL_HASH_SEED_DEBUG now makes perl show both the
f105b7be 129hash function perl was built with, I<and> the seed, in hex, in use for that
5ed58cbd
RS
130process. Code parsing this output, should it exist, must change to accommodate
131the new format. Example of the new format:
132
133 $ PERL_HASH_SEED_DEBUG=1 ./perl -e1
134 HASH_FUNCTION = MURMUR3 HASH_SEED = 0x1476bb9f
135
136=head2 Upgrade to Unicode 6.2
137
2e7bc647 138Perl now supports Unicode 6.2. A list of changes from Unicode
5ed58cbd
RS
1396.1 is at L<http://www.unicode.org/versions/Unicode6.2.0>.
140
141=head2 Character name aliases may now include non-Latin1-range characters
142
143It is possible to define your own names for characters for use in
144C<\N{...}>, C<charnames::vianame()>, etc. These names can now be
145comprised of characters from the whole Unicode range. This allows for
146names to be in your native language, and not just English. Certain
147restrictions apply to the characters that may be used (you can't define
148a name that has punctuation in it, for example). See L<charnames/CUSTOM
149ALIASES>.
150
151=head2 New DTrace probes
152
153The following new DTrace probes have been added:
14731ad1 154
337fb649 155=over 4
14731ad1 156
82d98f72 157=item *
14731ad1 158
5ed58cbd
RS
159C<op-entry>
160
161=item *
162
163C<loading-file>
164
165=item *
166
167C<loaded-file>
168
169=back
170
171=head2 C<${^LAST_FH}>
172
173This new variable provides access to the filehandle that was last read.
174This is the handle used by C<$.> and by C<tell> and C<eof> without
175arguments.
176
177=head2 Regular Expression Set Operations
178
179This is an B<experimental> feature to allow matching against the union,
180intersection, etc., of sets of code points, similar to
181L<Unicode::Regex::Set>. It can also be used to extend C</x> processing
182to [bracketed] character classes, and as a replacement of user-defined
183properties, allowing more complex expressions than they do. See
184L<perlrecharclass/Extended Bracketed Character Classes>.
185
186=head2 Lexical subroutines
187
188This new feature is still considered B<experimental>. To enable it:
189
190 use 5.018;
191 no warnings "experimental::lexical_subs";
192 use feature "lexical_subs";
193
194You can now declare subroutines with C<state sub foo>, C<my sub foo>, and
195C<our sub foo>. (C<state sub> requires that the "state" feature be
196enabled, unless you write it as C<CORE::state sub foo>.)
197
198C<state sub> creates a subroutine visible within the lexical scope in which
199it is declared. The subroutine is shared between calls to the outer sub.
200
201C<my sub> declares a lexical subroutine that is created each time the
202enclosing block is entered. C<state sub> is generally slightly faster than
203C<my sub>.
204
205C<our sub> declares a lexical alias to the package subroutine of the same
206name.
207
208For more information, see L<perlsub/Lexical Subroutines>.
209
210=head2 Computed Labels
211
212The loop controls C<next>, C<last> and C<redo>, and the special C<dump>
213operator, now allow arbitrary expressions to be used to compute labels at run
214time. Previously, any argument that was not a constant was treated as the
215empty string.
216
217=head2 More CORE:: subs
218
219Several more built-in functions have been added as subroutines to the
220CORE:: namespace - namely, those non-overridable keywords that can be
221implemented without custom parsers: C<defined>, C<delete>, C<exists>,
222C<glob>, C<pos>, C<protoytpe>, C<scalar>, C<split>, C<study>, and C<undef>.
223
224As some of these have prototypes, C<prototype('CORE::...')> has been
225changed to not make a distinction between overridable and non-overridable
226keywords. This is to make C<prototype('CORE::pos')> consistent with
227C<prototype(&CORE::pos)>.
228
229=head2 C<kill> with negative signal names
230
231C<kill> has always allowed a negative signal number, which kills the
232process group instead of a single process. It has also allowed signal
233names. But it did not behave consistently, because negative signal names
234were treated as 0. Now negative signals names like C<-INT> are supported
235and treated the same way as -2 [perl #112990].
236
237=head1 Security
238
239=head2 C<Storable> security warning in documentation
240
241The documentation for C<Storable> now includes a section which warns readers
242of the danger of accepting Storable documents from untrusted sources. The
243short version is that deserializing certain types of data can lead to loading
244modules and other code execution. This is documented behavior and wanted
245behavior, but this opens an attack vector for malicious entities.
246
247=head2 C<Locale::Maketext> allowed code injection via a malicious template
248
249If users could provide a translation string to Locale::Maketext, this could be
250used to invoke arbitrary Perl subroutines available in the current process.
251
252This has been fixed, but it is still possible to invoke any method provided by
253C<Locale::Maketext> itself or a subclass that you are using. One of these
254methods in turn will invoke the Perl core's C<sprintf> subroutine.
255
256In summary, allowing users to provide translation strings without auditing
257them is a bad idea.
258
259This vulnerability is documented in CVE-2012-6329.
260
261=head2 Avoid calling memset with a negative count
262
263Poorly written perl code that allows an attacker to specify the count to perl's
264C<x> string repeat operator can already cause a memory exhaustion
e612b5a0 265denial-of-service attack. A flaw in versions of perl before v5.15.5 can escalate
5ed58cbd
RS
266that into a heap buffer overrun; coupled with versions of glibc before 2.16, it
267possibly allows the execution of arbitrary code.
268
269The flaw addressed to this commit has been assigned identifier CVE-2012-5195
270and was researched by Tim Brown.
271
272=head1 Incompatible Changes
273
274=head2 See also: hash overhaul
275
276Some of the changes in the L<hash overhaul|/"Hash overhaul"> are not fully
277compatible with previous versions of perl. Please read that section.
278
279=head2 An unknown character name in C<\N{...}> is now a syntax error
280
281Previously, it warned, and the Unicode REPLACEMENT CHARACTER was
282substituted. Unicode now recommends that this situation be a syntax
283error. Also, the previous behavior led to some confusing warnings and
284behaviors, and since the REPLACEMENT CHARACTER has no use other than as
285a stand-in for some unknown character, any code that has this problem is
286buggy.
287
288=head2 Formerly deprecated characters in C<\N{}> character name aliases are now errors.
289
290Since v5.12.0, it has been deprecated to use certain characters in
291user-defined C<\N{...}> character names. These now cause a syntax
292error. For example, it is now an error to begin a name with a digit,
293such as in
294
295 my $undraftable = "\N{4F}"; # Syntax error!
296
237b9c6d 297or to have commas anywhere in the name. See L<charnames/CUSTOM ALIASES>.
5ed58cbd
RS
298
299=head2 C<\N{BELL}> now refers to U+1F514 instead of U+0007
300
301Unicode 6.0 reused the name "BELL" for a different code point than it
302traditionally had meant. Since Perl v5.14, use of this name still
303referred to U+0007, but would raise a deprecation warning. Now, "BELL"
304refers to U+1F514, and the name for U+0007 is "ALERT". All the
305functions in L<charnames> have been correspondingly updated.
306
307=head2 New Restrictions in Multi-Character Case-Insensitive Matching in Regular Expression Bracketed Character Classes
308
309Unicode has now withdrawn their previous recommendation for regular
310expressions to automatically handle cases where a single character can
237b9c6d 311match multiple characters case-insensitively, for example, the letter
5ed58cbd
RS
312LATIN SMALL LETTER SHARP S and the sequence C<ss>. This is because
313it turns out to be impracticable to do this correctly in all
314circumstances. Because Perl has tried to do this as best it can, it
315will continue to do so. (We are considering an option to turn it off.)
316However, a new restriction is being added on such matches when they
317occur in [bracketed] character classes. People were specifying
318things such as C</[\0-\xff]/i>, and being surprised that it matches the
319two character sequence C<ss> (since LATIN SMALL LETTER SHARP S occurs in
320this range). This behavior is also inconsistent with using a
321property instead of a range: C<\p{Block=Latin1}> also includes LATIN
322SMALL LETTER SHARP S, but C</[\p{Block=Latin1}]/i> does not match C<ss>.
323The new rule is that for there to be a multi-character case-insensitive
324match within a bracketed character class, the character must be
325explicitly listed, and not as an end point of a range. This more
326closely obeys the Principle of Least Astonishment. See
327L<perlrecharclass/Bracketed Character Classes>. Note that a bug [perl
328#89774], now fixed as part of this change, prevented the previous
329behavior from working fully.
330
331=head2 Explicit rules for variable names and identifiers
332
c6c6975e
RS
333Due to an oversight, single character variable names in v5.16 were
334completely unrestricted. This opened the door to several kinds of
335insanity. As of v5.18, these now follow the rules of other identifiers,
336in addition to accepting characters that match the C<\p{POSIX_Punct}>
337property.
5ed58cbd
RS
338
339There are no longer any differences in the parsing of identifiers
340specified as C<$...> or C<${...}>; previously, they were dealt with in
341different parts of the core, and so had slightly different behavior. For
342instance, C<${foo:bar}> was a legal variable name. Since they are now
343both parsed by the same code, that is no longer the case.
344
345=head2 C<\s> in regular expressions now matches a Vertical Tab
346
347No one could recall why C<\s> didn't match C<\cK>, the vertical tab.
348Now it does. Given the extreme rarity of that character, very little
349breakage is expected.
350
351=head2 C</(?{})/> and C</(??{})/> have been heavily reworked
352
353The implementation of this feature has been almost completely rewritten.
354Although its main intent is to fix bugs, some behaviors, especially
355related to the scope of lexical variables, will have changed. This is
356described more fully in the L</Selected Bug Fixes> section.
357
358=head2 Stricter parsing of substitution replacement
359
360It is no longer possible to abuse the way the parser parses C<s///e> like
361this:
362
363 %_=(_,"Just another ");
364 $_="Perl hacker,\n";
365 s//_}->{_/e;print
366
367=head2 C<given> now aliases the global C<$_>
368
369Instead of assigning to an implicit lexical C<$_>, C<given> now makes the
370global C<$_> an alias for its argument, just like C<foreach>. However, it
371still uses lexical C<$_> if there is lexical C<$_> in scope (again, just like
372C<foreach>) [perl #114020].
373
64a74324
RS
374=head2 The smartmatch family of features are now experimental
375
376Smart match, added in v5.10.0 and significantly revised in v5.10.1, has been
377a regular point of complaint. Although there are a number of ways in which
378it is useful, it has also proven problematic and confusing for both users and
379implementors of Perl. There have been a number of proposals on how to best
380address the problem. It is clear that smartmatch is almost certainly either
381going to change or go away in the future. Relying on its current behavior
382is not recommended.
383
384Warnings will now be issued when the parser sees C<~~>, C<given>, or C<when>.
385To disable these warnings, you can add this line to the appropriate scope:
386
387 no if $] >= 5.018, "experimental::smartmatch";
388
389Consider, though, replacing the use of these features, as they may change
390behavior again before becoming stable.
391
5ed58cbd
RS
392=head2 Lexical C<$_> is now experimental
393
e612b5a0 394Since it was introduced in Perl v5.10, it has caused much confusion with no
5ed58cbd
RS
395obvious solution:
396
397=over
829397b9
TC
398
399=item *
400
5ed58cbd
RS
401Various modules (e.g., List::Util) expect callback routines to use the
402global C<$_>. C<use List::Util 'first'; my $_; first { $_ == 1 } @list>
403does not work as one would expect.
71e6aba6
RS
404
405=item *
406
5ed58cbd
RS
407A C<my $_> declaration earlier in the same file can cause confusing closure
408warnings.
71e6aba6
RS
409
410=item *
411
5ed58cbd
RS
412The "_" subroutine prototype character allows called subroutines to access
413your lexical C<$_>, so it is not really private after all.
2426c394 414
71e6aba6
RS
415=item *
416
5ed58cbd
RS
417Nevertheless, subroutines with a "(@)" prototype and methods cannot access
418the caller's lexical C<$_>, unless they are written in XS.
71e6aba6
RS
419
420=item *
421
5ed58cbd
RS
422But even XS routines cannot access a lexical C<$_> declared, not in the
423calling subroutine, but in an outer scope, iff that subroutine happened not
424to mention C<$_> or use any operators that default to C<$_>.
425
426=back
427
428It is our hope that lexical C<$_> can be rehabilitated, but this may
429cause changes in its behavior. Please use it with caution until it
430becomes stable.
431
432=head2 readline() with C<$/ = \N> now reads N characters, not N bytes
433
434Previously, when reading from a stream with I/O layers such as
435C<encoding>, the readline() function, otherwise known as the C<< <> >>
436operator, would read I<N> bytes from the top-most layer. [perl #79960]
437
438Now, I<N> characters are read instead.
439
440There is no change in behaviour when reading from streams with no
441extra layers, since bytes map exactly to characters.
442
443=head2 Overridden C<glob> is now passed one argument
444
445C<glob> overrides used to be passed a magical undocumented second argument
446that identified the caller. Nothing on CPAN was using this, and it got in
447the way of a bug fix, so it was removed. If you really need to identify
448the caller, see L<Devel::Callsite> on CPAN.
449
450=head2 Here-doc parsing
451
452The body of a here-document inside a quote-like operator now always begins
453on the line after the "<<foo" marker. Previously, it was documented to
454begin on the line following the containing quote-like operator, but that
455was only sometimes the case [perl #114040].
456
457=head2 Alphanumeric operators must now be separated from the closing
458delimiter of regular expressions
459
460You may no longer write something like:
461
462 m/a/and 1
463
464Instead you must write
465
466 m/a/ and 1
467
468with whitespace separating the operator from the closing delimiter of
469the regular expression. Not having whitespace has resulted in a
470deprecation warning since Perl v5.14.0.
471
472=head2 qw(...) can no longer be used as parentheses
473
474C<qw> lists used to fool the parser into thinking they were always
475surrounded by parentheses. This permitted some surprising constructions
476such as C<foreach $x qw(a b c) {...}>, which should really be written
477C<foreach $x (qw(a b c)) {...}>. These would sometimes get the lexer into
478the wrong state, so they didn't fully work, and the similar C<foreach qw(a
479b c) {...}> that one might expect to be permitted never worked at all.
480
481This side effect of C<qw> has now been abolished. It has been deprecated
e612b5a0 482since Perl v5.13.11. It is now necessary to use real parentheses
5ed58cbd
RS
483everywhere that the grammar calls for them.
484
485=head2 Interaction of lexical and default warnings
486
487Turning on any lexical warnings used first to disable all default warnings
488if lexical warnings were not already enabled:
489
490 $*; # deprecation warning
491 use warnings "void";
492 $#; # void warning; no deprecation warning
493
f105b7be 494Now, the C<debugging>, C<deprecated>, C<glob>, C<inplace> and C<malloc> warnings
5ed58cbd
RS
495categories are left on when turning on lexical warnings (unless they are
496turned off by C<no warnings>, of course).
497
498This may cause deprecation warnings to occur in code that used to be free
499of warnings.
500
501Those are the only categories consisting only of default warnings. Default
f105b7be
KE
502warnings in other categories are still disabled by C<< use warnings "category" >>,
503as we do not yet have the infrastructure for controlling
5ed58cbd
RS
504individual warnings.
505
506=head2 C<state sub> and C<our sub>
507
508Due to an accident of history, C<state sub> and C<our sub> were equivalent
509to a plain C<sub>, so one could even create an anonymous sub with
510C<our sub { ... }>. These are now disallowed outside of the "lexical_subs"
511feature. Under the "lexical_subs" feature they have new meanings described
512in L<perlsub/Lexical Subroutines>.
513
514=head2 Defined values stored in environment are forced to byte strings
515
516A value stored in an environment variable has always been stringified. In this
517release, it is converted to be only a byte string. First, it is forced to be a
518only a string. Then if the string is utf8 and the equivalent of
519C<utf8::downgrade()> works, that result is used; otherwise, the equivalent of
520C<utf8::encode()> is used, and a warning is issued about wide characters
521(L</Diagnostics>).
522
523=head2 C<require> dies for unreadable files
524
525When C<require> encounters an unreadable file, it now dies. It used to
526ignore the file and continue searching the directories in C<@INC>
527[perl #113422].
528
529=head2 C<gv_fetchmeth_*> and SUPER
530
531The various C<gv_fetchmeth_*> XS functions used to treat a package whose
f105b7be
KE
532named ended with C<::SUPER> specially. A method lookup on the C<Foo::SUPER>
533package would be treated as a C<SUPER> method lookup on the C<Foo> package. This
534is no longer the case. To do a C<SUPER> lookup, pass the C<Foo> stash and the
535C<GV_SUPER> flag.
5ed58cbd
RS
536
537=head2 C<split>'s first argument is more consistently interpreted
538
e612b5a0 539After some changes earlier in v5.17, C<split>'s behavior has been
5ed58cbd
RS
540simplified: if the PATTERN argument evaluates to a literal string
541containing one space, it is treated the way that a I<literal> string
542containing one space once was.
543
544=head1 Deprecations
545
546=head2 Deprecated modules
547
548The following modules will be removed from the core distribution in a
549future release, and should be installed from CPAN instead. Distributions
550on CPAN which require these should add them to their prerequisites.
f105b7be 551The core versions of these modules will issue C<"deprecated">-category
5ed58cbd
RS
552warnings.
553
554You can silence these deprecation warnings by installing the modules
555in question from CPAN.
556
557=over
558
559=item L<Archive::Extract>
560
561=item L<B::Lint>
562
563=item L<B::Lint::Debug>
564
565=item L<CPANPLUS> and all included C<CPANPLUS::*> modules
566
567=item L<Devel::InnerPackage>
568
569=item L<encoding>
570
571=item L<Log::Message>
572
573=item L<Log::Message::Config>
574
575=item L<Log::Message::Handlers>
576
577=item L<Log::Message::Item>
578
579=item L<Log::Message::Simple>
580
581=item L<Module::Pluggable>
582
583=item L<Module::Pluggable::Object>
584
585=item L<Object::Accessor>
586
587=item L<Pod::LaTeX>
588
589=item L<Term::UI>
590
591=item L<Term::UI::History>
592
593=back
594
595=head2 Deprecated Utilities
596
597The following utilities will be removed from the core distribution in a
598future release as their associated modules have been deprecated. They
599will remain available with the applicable CPAN distribution.
600
601=over
602
603=item L<cpanp>
604
605=item C<cpanp-run-perl>
606
607=item L<cpan2dist>
608
609These items are part of the C<CPANPLUS> distribution.
610
611=item L<pod2latex>
612
613This item is part of the C<Pod::LaTeX> distribution.
614
615=back
616
617=head2 PL_sv_objcount
618
619This interpreter-global variable used to track the total number of
620Perl objects in the interpreter. It is no longer maintained and will
e612b5a0 621be removed altogether in Perl v5.20.
5ed58cbd
RS
622
623=head2 Five additional characters should be escaped in patterns with C</x>
624
625When a regular expression pattern is compiled with C</x>, Perl treats 6
626characters as white space to ignore, such as SPACE and TAB. However,
627Unicode recommends 11 characters be treated thusly. We will conform
628with this in a future Perl version. In the meantime, use of any of the
629missing characters will raise a deprecation warning, unless turned off.
630The five characters are:
f105b7be
KE
631
632 U+0085 NEXT LINE,
633 U+200E LEFT-TO-RIGHT MARK,
634 U+200F RIGHT-TO-LEFT MARK,
635 U+2028 LINE SEPARATOR,
636
5ed58cbd 637and
f105b7be
KE
638
639 U+2029 PARAGRAPH SEPARATOR.
5ed58cbd
RS
640
641=head2 User-defined charnames with surprising whitespace
642
643A user-defined character name with trailing or multiple spaces in a row is
644likely a typo. This now generates a warning when defined, on the assumption
645that uses of it will be unlikely to include the excess whitespace.
646
647=head2 Various XS-callable functions are now deprecated
648
649All the functions used to classify characters will be removed from a
650future version of Perl, and should not be used. With participating C
651compilers (e.g., gcc), compiling any file that uses any of these will
652generate a warning. These were not intended for public use; there are
653equivalent, faster, macros for most of them.
e612b5a0 654
2e7bc647 655See L<perlapi/Character classes>. The complete list is:
e612b5a0 656
5ed58cbd
RS
657C<is_uni_alnum>, C<is_uni_alnumc>, C<is_uni_alnumc_lc>,
658C<is_uni_alnum_lc>, C<is_uni_alpha>, C<is_uni_alpha_lc>,
659C<is_uni_ascii>, C<is_uni_ascii_lc>, C<is_uni_blank>,
660C<is_uni_blank_lc>, C<is_uni_cntrl>, C<is_uni_cntrl_lc>,
661C<is_uni_digit>, C<is_uni_digit_lc>, C<is_uni_graph>,
662C<is_uni_graph_lc>, C<is_uni_idfirst>, C<is_uni_idfirst_lc>,
663C<is_uni_lower>, C<is_uni_lower_lc>, C<is_uni_print>,
664C<is_uni_print_lc>, C<is_uni_punct>, C<is_uni_punct_lc>,
665C<is_uni_space>, C<is_uni_space_lc>, C<is_uni_upper>,
666C<is_uni_upper_lc>, C<is_uni_xdigit>, C<is_uni_xdigit_lc>,
667C<is_utf8_alnum>, C<is_utf8_alnumc>, C<is_utf8_alpha>,
668C<is_utf8_ascii>, C<is_utf8_blank>, C<is_utf8_char>,
669C<is_utf8_cntrl>, C<is_utf8_digit>, C<is_utf8_graph>,
670C<is_utf8_idcont>, C<is_utf8_idfirst>, C<is_utf8_lower>,
671C<is_utf8_mark>, C<is_utf8_perl_space>, C<is_utf8_perl_word>,
672C<is_utf8_posix_digit>, C<is_utf8_print>, C<is_utf8_punct>,
673C<is_utf8_space>, C<is_utf8_upper>, C<is_utf8_xdigit>,
674C<is_utf8_xidcont>, C<is_utf8_xidfirst>.
675
676In addition these three functions that have never worked properly are
677deprecated:
678C<to_uni_lower_lc>, C<to_uni_title_lc>, and C<to_uni_upper_lc>.
679
f105b7be 680=head2 Certain rare uses of backslashes within regexes are now deprecated
5ed58cbd
RS
681
682There are three pairs of characters that Perl recognizes as
683metacharacters in regular expression patterns: C<{}>, C<[]>, and C<()>.
684These can be used as well to delimit patterns, as in:
685
686 m{foo}
687 s(foo)(bar)
688
689Since they are metacharacters, they have special meaning to regular
690expression patterns, and it turns out that you can't turn off that
691special meaning by the normal means of preceding them with a backslash,
f105b7be 692if you use them, paired, within a pattern delimited by them. For
5ed58cbd
RS
693example, in
694
695 m{foo\{1,3\}}
696
697the backslashes do not change the behavior, and this matches
698S<C<"f o">> followed by one to three more occurrences of C<"o">.
699
700Usages like this, where they are interpreted as metacharacters, are
701exceedingly rare; we think there are none, for example, in all of CPAN.
702Hence, this deprecation should affect very little code. It does give
703notice, however, that any such code needs to change, which will in turn
704allow us to change the behavior in future Perl versions so that the
705backslashes do have an effect, and without fear that we are silently
706breaking any existing code.
707
d5f315e8
KW
708=head2 Splitting the tokens C<(?> and C<(*> in regular expressions
709
710A deprecation warning is now raised if the C<(> and C<?> are separated
711by white space or comments in C<(?...)> regular expression constructs.
712Similarly, if the C<(> and C<*> are separated in C<(*VERB...)>
713constructs.
714
e0a1dec5
LT
715=head2 Pre-PerlIO IO implementations
716
717Perl supports being built without PerlIO proper, using a stdio or sfio
718wrapper instead. A perl build like this will not support IO layers and
719thus Unicode IO, making it rather handicapped.
720
721PerlIO supports a C<stdio> layer if stdio use is desired, and similarly a
722sfio layer could be produced.
723
5ed58cbd
RS
724=head1 Future Deprecations
725
726=over
71e6aba6
RS
727
728=item *
729
4263dd11 730Platforms without support infrastructure
5ed58cbd
RS
731
732Both Windows CE and z/OS have been historically under-maintained, and are
733currently neither successfully building nor regularly being smoke tested.
734Efforts are underway to change this situation, but it should not be taken for
735granted that the platforms are safe and supported. If they do not become
736buildable and regularly smoked, support for them may be actively removed in
737future releases. If you have an interest in these platforms and you can lend
738your time, expertise, or hardware to help support these platforms, please let
739the perl development effort know by emailing C<perl5-porters@perl.org>.
740
741Some platforms that appear otherwise entirely dead are also on the short list
e612b5a0 742for removal between now and v5.20.0:
5ed58cbd
RS
743
744=over
745
746=item DG/UX
747
748=item NeXT
749
750=back
1993add8 751
ec985017
RS
752We also think it likely that current versions of Perl will no longer
753build AmigaOS, DJGPP, NetWare (natively), OS/2 and Plan 9. If you
754are using Perl on such a platform and have an interest in ensuring
755Perl's future on them, please contact us.
756
757We believe that Perl has long been unable to build on mixed endian
758architectures (such as PDP-11s), and intend to remove any remain
759support code. Similarly, code supporting the long umaintained GNU
760dld will be removed soon if no-one makes themselves known as an
761active user.
762
1993add8
RS
763=item *
764
5ed58cbd
RS
765Swapping of $< and $>
766
767For more information about this future deprecation, see L<the relevant RT
768ticket|https://rt.perl.org/rt3/Ticket/Display.html?id=96212>.
71e6aba6
RS
769
770=item *
771
5ed58cbd 772C<microperl>, long broken and of unclear present purpose, will be removed.
71e6aba6
RS
773
774=item *
775
5ed58cbd
RS
776Revamping C<< "\Q" >> semantics in double-quotish strings when combined with
777other escapes.
778
779There are several bugs and inconsistencies involving combinations
780of C<\Q> and escapes like C<\x>, C<\L>, etc., within a C<\Q...\E> pair.
781These need to be fixed, and doing so will necessarily change current
782behavior. The changes have not yet been settled.
71e6aba6 783
d5f315e8
KW
784=item *
785
786Use of C<$^>, where C<^> stands for any actual (non-printing) C0 control
787character will be disallowed in a future Perl version. Use C<${^}>
788instead (where again C<^> stands for a control character),
789or better, C<$^A> , where C<^> this time is a caret (CIRCUMFLEX ACCENT),
790and C<A> stands for any of the characters listed at the end of
791L<perlebcdic/OPERATOR DIFFERENCES>.
792
337fb649 793=back
2426c394 794
5ed58cbd
RS
795=head1 Performance Enhancements
796
797=over 4
2426c394 798
5ed58cbd 799=item *
2426c394 800
5ed58cbd
RS
801Lists of lexical variable declarations (C<my($x, $y)>) are now optimised
802down to a single op and are hence faster than before.
2426c394 803
5ed58cbd 804=item *
2426c394 805
5ed58cbd
RS
806A new C preprocessor define C<NO_TAINT_SUPPORT> was added that, if set,
807disables Perl's taint support altogether. Using the -T or -t command
808line flags will cause a fatal error. Beware that both core tests as
809well as many a CPAN distribution's tests will fail with this change. On
810the upside, it provides a small performance benefit due to reduced
811branching.
2426c394 812
5ed58cbd
RS
813B<Do not enable this unless you know exactly what you are getting yourself
814into.>
815
816=item *
817
818C<pack> with constant arguments is now constant folded in most cases
819[perl #113470].
820
821=item *
822
823Speed up in regular expression matching against Unicode properties. The
824largest gain is for C<\X>, the Unicode "extended grapheme cluster." The
825gain for it is about 35% - 40%. Bracketed character classes, e.g.,
826C<[0-9\x{100}]> containing code points above 255 are also now faster.
827
828=item *
829
830On platforms supporting it, several former macros are now implemented as static
831inline functions. This should speed things up slightly on non-GCC platforms.
832
833=item *
834
66f62cf6
RS
835The optimisation of hashes in boolean context has been extended to
836affect C<scalar(%hash)>, C<%hash ? ... : ...>, and C<sub { %hash || ... }>.
5ed58cbd
RS
837
838=item *
839
f105b7be 840Filetest operators manage the stack in a fractionally more efficient manner.
5ed58cbd
RS
841
842=item *
843
844Globs used in a numeric context are now numified directly in most cases,
f105b7be 845rather than being numified via stringification.
5ed58cbd
RS
846
847=item *
848
849The C<x> repetition operator is now folded to a single constant at compile
850time if called in scalar context with constant operands and no parentheses
851around the left operand.
852
853=back
854
855=head1 Modules and Pragmata
856
857=head2 New Modules and Pragmata
2426c394 858
337fb649 859=over 4
982110e0 860
82d98f72 861=item *
2426c394 862
5ed58cbd
RS
863L<Config::Perl::V> version 0.16 has been added as a dual-lifed module.
864It provides structured data retrieval of C<perl -V> output including
865information only known to the C<perl> binary and not available via L<Config>.
866
867=back
868
869=head2 Updated Modules and Pragmata
870
9863976f 871For a complete list of updates, run:
5ed58cbd 872
f105b7be 873 $ corelist --diff 5.16.0 5.18.0
5ed58cbd 874
e612b5a0 875You can substitute your favorite version in place of C<5.16.0>, too.
5ed58cbd 876
5ed58cbd
RS
877=head2 Removed Modules and Pragmata
878
879=over
33392251
BF
880
881=item *
882
5ed58cbd
RS
883L<Version::Requirements> has been removed from the core distribution. It is
884available under a different name: L<CPAN::Meta::Requirements>.
2426c394 885
337fb649 886=back
2426c394 887
5ed58cbd 888=head1 Documentation
19718730 889
5ed58cbd
RS
890=head2 Changes to Existing Documentation
891
892=head3 L<perlcheat>
82d98f72 893
5a6a30f4 894=over 4
b7c7d786 895
5ed58cbd
RS
896=item *
897
898L<perlcheat> has been reorganized, and a few new sections were added.
899
900=back
901
902=head3 L<perldata>
903
904=over 4
82d98f72 905
5ed58cbd 906=item *
d2d1e842 907
5ed58cbd
RS
908Now explicitly documents the behaviour of hash initializer lists that
909contain duplicate keys.
f355e93d 910
5a6a30f4 911=back
f355e93d 912
5ed58cbd 913=head3 L<perldiag>
19718730 914
19718730 915=over 4
e14ac59b 916
5ed58cbd
RS
917=item *
918
919The explanation of symbolic references being prevented by "strict refs"
920now doesn't assume that the reader knows what symbolic references are.
921
922=back
9f351b45 923
5ed58cbd 924=head3 L<perlfaq>
9f351b45 925
5ed58cbd 926=over 4
9f351b45 927
5ed58cbd 928=item *
7cf3104f 929
5ed58cbd 930L<perlfaq> has been synchronized with version 5.0150040 from CPAN.
12719193 931
6253ee75 932=back
216cf7fc 933
5ed58cbd 934=head3 L<perlfunc>
f5b73711 935
5ed58cbd
RS
936=over 4
937
938=item *
a75569c0 939
5ed58cbd 940The return value of C<pipe> is now documented.
a75569c0 941
5ed58cbd 942=item *
a75569c0 943
5ed58cbd
RS
944Clarified documentation of C<our>.
945
946=back
947
948=head3 L<perlop>
949
950=over 4
951
952=item *
953
954Loop control verbs (C<dump>, C<goto>, C<next>, C<last> and C<redo>) have always
955had the same precedence as assignment operators, but this was not documented
956until now.
957
958=back
959
960=head3 Diagnostics
961
962The following additions or changes have been made to diagnostic output,
963including warnings and fatal error messages. For the complete list of
964diagnostic messages, see L<perldiag>.
965
5ed58cbd
RS
966=head2 New Diagnostics
967
5ed58cbd
RS
968=head3 New Errors
969
970=over 4
971
972=item *
973
974L<Unterminated delimiter for here document|perldiag/"Unterminated delimiter for here document">
975
976This message now occurs when a here document label has an initial quotation
977mark but the final quotation mark is missing.
978
979This replaces a bogus and misleading error message about not finding the label
980itself [perl #114104].
981
982=item *
983
984L<panic: child pseudo-process was never scheduled|perldiag/"panic: child pseudo-process was never scheduled">
985
986This error is thrown when a child pseudo-process in the ithreads implementation
987on Windows was not scheduled within the time period allowed and therefore was
988not able to initialize properly [perl #88840].
989
990=item *
991
992L<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/">
993
994This error has been added for C<(?&0)>, which is invalid. It used to
995produce an incomprehensible error message [perl #101666].
996
997=item *
998
999L<Can't use an undefined value as a subroutine reference|perldiag/"Can't use an undefined value as %s reference">
1000
1001Calling an undefined value as a subroutine now produces this error message.
1002It used to, but was accidentally disabled, first in Perl 5.004 for
e612b5a0 1003non-magical variables, and then in Perl v5.14 for magical (e.g., tied)
5ed58cbd
RS
1004variables. It has now been restored. In the mean time, undef was treated
1005as an empty string [perl #113576].
1006
1007=item *
1008
1009L<Experimental "%s" subs not enabled|perldiag/"Experimental "%s" subs not enabled">
1010
1011To use lexical subs, you must first enable them:
1012
1013 no warnings 'experimental::lexical_subs';
1014 use feature 'lexical_subs';
1015 my sub foo { ... }
1016
1017=back
1018
1019=head3 New Warnings
1020
1021=over 4
1022
1023=item *
1024
5ed58cbd
RS
1025Strings with code points over 0xFF may not be mapped into in-memory file
1026handles
1027
1028=item *
1029
1030L<'%s' resolved to '\o{%s}%d'|perldiag/"'%s' resolved to '\o{%s}%d'">
1031
1032=item *
1033
1034L<'Trailing white-space in a charnames alias definition is deprecated'|perldiag/"Trailing white-space in a charnames alias definition is deprecated">
1035
1036=item *
1037
1038L<'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">
1039
1040=item *
1041
1042L<'Passing malformed UTF-8 to "%s" is deprecated'|perldiag/"Passing malformed UTF-8 to "%s" is deprecated">
1043
1044=item *
1045
1046L<Subroutine "&%s" is not available|perldiag/"Subroutine "&%s" is not available">
1047
1048(W closure) During compilation, an inner named subroutine or eval is
1049attempting to capture an outer lexical subroutine that is not currently
1050available. This can happen for one of two reasons. First, the lexical
1051subroutine may be declared in an outer anonymous subroutine that has not
1052yet been created. (Remember that named subs are created at compile time,
1053while anonymous subs are created at run-time.) For example,
1054
1055 sub { my sub a {...} sub f { \&a } }
1056
1057At the time that f is created, it can't capture the current the "a" sub,
1058since the anonymous subroutine hasn't been created yet. Conversely, the
1059following won't give a warning since the anonymous subroutine has by now
1060been created and is live:
1061
1062 sub { my sub a {...} eval 'sub f { \&a }' }->();
1063
1064The second situation is caused by an eval accessing a variable that has
1065gone out of scope, for example,
1066
1067 sub f {
1068 my sub a {...}
1069 sub { eval '\&a' }
1070 }
1071 f()->();
1072
1073Here, when the '\&a' in the eval is being compiled, f() is not currently
1074being executed, so its &a is not available for capture.
1075
1076=item *
1077
1078L<"%s" subroutine &%s masks earlier declaration in same %s|perldiag/"%s" subroutine &%s masks earlier declaration in same %s>
1079
1080(W misc) A "my" or "state" subroutine has been redeclared in the
1081current scope or statement, effectively eliminating all access to
1082the previous instance. This is almost always a typographical error.
1083Note that the earlier subroutine will still exist until the end of
1084the scope or until all closure references to it are destroyed.
1085
1086=item *
1087
1088L<The %s feature is experimental|perldiag/"The %s feature is experimental">
1089
1090(S experimental) This warning is emitted if you enable an experimental
1091feature via C<use feature>. Simply suppress the warning if you want
1092to use the feature, but know that in doing so you are taking the risk
1093of using an experimental feature which may change or be removed in a
1094future Perl version:
1095
1096 no warnings "experimental::lexical_subs";
1097 use feature "lexical_subs";
1098
1099=item *
1100
1101L<sleep(%u) too large|perldiag/"sleep(%u) too large">
1102
1103(W overflow) You called C<sleep> with a number that was larger than it can
1104reliably handle and C<sleep> probably slept for less time than requested.
1105
1106=item *
1107
1108L<Wide character in setenv|perldiag/"Wide character in %s">
1109
1110Attempts to put wide characters into environment variables via C<%ENV> now
1111provoke this warning.
1112
1113=item *
1114
1115"L<Invalid negative number (%s) in chr|perldiag/"Invalid negative number (%s) in chr">"
1116
1117C<chr()> now warns when passed a negative value [perl #83048].
1118
1119=item *
1120
1121"L<Integer overflow in srand|perldiag/"Integer overflow in srand">"
1122
1123C<srand()> now warns when passed a value that doesn't fit in a C<UV> (since the
1124value will be truncated rather than overflowing) [perl #40605].
1125
1126=item *
1127
1128"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">"
1129
1130Running perl with the C<-i> flag now warns if no input files are provided on
1131the command line [perl #113410].
1132
1133=back
1134
1135=head2 Changes to Existing Diagnostics
1136
1137=over 4
1138
1139=item *
1140
1141L<$* is no longer supported|perldiag/"$* is no longer supported">
1142
1143The warning that use of C<$*> and C<$#> is no longer supported is now
1144generated for every location that references them. Previously it would fail
1145to be generated if another variable using the same typeglob was seen first
1146(e.g. C<@*> before C<$*>), and would not be generated for the second and
1147subsequent uses. (It's hard to fix the failure to generate warnings at all
1148without also generating them every time, and warning every time is
1149consistent with the warnings that C<$[> used to generate.)
1150
1151=item *
1152
1153The warnings for C<\b{> and C<\B{> were added. They are a deprecation
1154warning which should be turned off by that category. One should not
1155have to turn off regular regexp warnings as well to get rid of these.
1156
1157=item *
1158
1159L<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>
1160
1161Constant overloading that returns C<undef> results in this error message.
1162For numeric constants, it used to say "Constant(undef)". "undef" has been
1163replaced with the number itself.
1164
1165=item *
1166
1167The error produced when a module cannot be loaded now includes a hint that
1168the module may need to be installed: "Can't locate hopping.pm in @INC (you
1169may need to install the hopping module) (@INC contains: ...)"
1170
1171=item *
1172
1173L<vector argument not supported with alpha versions|perldiag/vector argument not supported with alpha versions>
1174
1175This warning was not suppressable, even with C<no warnings>. Now it is
1176suppressible, and has been moved from the "internal" category to the
1177"printf" category.
1178
1179=item *
1180
1181C<< Can't do {n,m} with n > m in regex; marked by <-- HERE in m/%s/ >>
1182
1183This fatal error has been turned into a warning that reads:
1184
1185L<< Quantifier {n,m} with n > m can't match in regex | perldiag/Quantifier {n,m} with n > m can't match in regex >>
1186
1187(W regexp) Minima should be less than or equal to maxima. If you really want
1188your regexp to match something 0 times, just put {0}.
1189
1190=item *
1191
1192The "Runaway prototype" warning that occurs in bizarre cases has been
1193removed as being unhelpful and inconsistent.
1194
1195=item *
1196
1197The "Not a format reference" error has been removed, as the only case in
1198which it could be triggered was a bug.
1199
1200=item *
1201
1202The "Unable to create sub named %s" error has been removed for the same
1203reason.
1204
1205=item *
1206
1207The 'Can't use "my %s" in sort comparison' error has been downgraded to a
1208warning, '"my %s" used in sort comparison' (with 'state' instead of 'my'
1209for state variables). In addition, the heuristics for guessing whether
1210lexical $a or $b has been misused have been improved to generate fewer
1211false positives. Lexical $a and $b are no longer disallowed if they are
1212outside the sort block. Also, a named unary or list operator inside the
1213sort block no longer causes the $a or $b to be ignored [perl #86136].
1214
1215=back
1216
1217=head1 Utility Changes
1218
1219=head3 L<h2xs>
1220
1221=over 4
1222
1223=item *
1224
1225F<h2xs> no longer produces invalid code for empty defines. [perl #20636]
1226
1227=back
1228
1229=head1 Configuration and Compilation
1230
1231=over 4
1232
1233=item *
1234
1235Added C<useversionedarchname> option to Configure
1236
1237When set, it includes 'api_versionstring' in 'archname'. E.g.
1238x86_64-linux-5.13.6-thread-multi. It is unset by default.
1239
1240This feature was requested by Tim Bunce, who observed that
f105b7be 1241C<INSTALL_BASE> creates a library structure that does not
5ed58cbd
RS
1242differentiate by perl version. Instead, it places architecture
1243specific files in "$install_base/lib/perl5/$archname". This makes
f105b7be 1244it difficult to use a common C<INSTALL_BASE> library path with
5ed58cbd
RS
1245multiple versions of perl.
1246
f105b7be 1247By setting C<-Duseversionedarchname>, the $archname will be
c2959982 1248distinct for architecture I<and> API version, allowing mixed use of
f105b7be 1249C<INSTALL_BASE>.
5ed58cbd
RS
1250
1251=item *
1252
ff772877
RS
1253Add a C<PERL_NO_INLINE_FUNCTIONS> option
1254
f105b7be 1255If C<PERL_NO_INLINE_FUNCTIONS> is defined, don't include "inline.h"
ff772877
RS
1256
1257This permits test code to include the perl headers for definitions without
1258creating a link dependency on the perl library (which may not exist yet).
1259
1260=item *
1261
5ed58cbd
RS
1262Configure will honour the external C<MAILDOMAIN> environment variable, if set.
1263
1264=item *
1265
1266C<installman> no longer ignores the silent option
1267
1268=item *
1269
1270Both C<META.yml> and C<META.json> files are now included in the distribution.
1271
1272=item *
1273
1274F<Configure> will now correctly detect C<isblank()> when compiling with a C++
1275compiler.
1276
1277=item *
1278
1279The pager detection in F<Configure> has been improved to allow responses which
1280specify options after the program name, e.g. B</usr/bin/less -R>, if the user
1281accepts the default value. This helps B<perldoc> when handling ANSI escapes
1282[perl #72156].
1283
1284=back
1285
1286=head1 Testing
1287
1288=over 4
1289
1290=item *
1291
1292The test suite now has a section for tests that require very large amounts
1293of memory. These tests won't run by default; they can be enabled by
1294setting the C<PERL_TEST_MEMORY> environment variable to the number of
1295gibibytes of memory that may be safely used.
1296
1297=back
1298
1299=head1 Platform Support
1300
1301=head2 Discontinued Platforms
1302
1303=over 4
1304
1305=item BeOS
1306
1307BeOS was an operating system for personal computers developed by Be Inc,
1308initially for their BeBox hardware. The OS Haiku was written as an open
1309source replacement for/continuation of BeOS, and its perl port is current and
1310actively maintained.
1311
1312=item UTS Global
1313
1314Support code relating to UTS global has been removed. UTS was a mainframe
1315version of System V created by Amdahl, subsequently sold to UTS Global. The
e612b5a0 1316port has not been touched since before Perl v5.8.0, and UTS Global is now
5ed58cbd
RS
1317defunct.
1318
1319=item VM/ESA
1320
1321Support for VM/ESA has been removed. The port was tested on 2.3.0, which
1322IBM ended service on in March 2002. 2.4.0 ended service in June 2003, and
1323was superseded by Z/VM. The current version of Z/VM is V6.2.0, and scheduled
1324for end of service on 2015/04/30.
1325
1326=item MPE/IX
1327
1328Support for MPE/IX has been removed.
1329
1330=item EPOC
1331
1332Support code relating to EPOC has been removed. EPOC was a family of
1333operating systems developed by Psion for mobile devices. It was the
1334predecessor of Symbian. The port was last updated in April 2002.
1335
1336=item Rhapsody
1337
1338Support for Rhapsody has been removed.
1339
1340=back
1341
1342=head2 Platform-Specific Notes
1343
1344=head3 AIX
1345
1346Configure now always adds C<-qlanglvl=extc99> to the CC flags on AIX when
1347using xlC. This will make it easier to compile a number of XS-based modules
1348that assume C99 [perl #113778].
1349
1350=head3 clang++
1351
1352There is now a workaround for a compiler bug that prevented compiling
e612b5a0 1353with clang++ since Perl v5.15.7 [perl #112786].
5ed58cbd
RS
1354
1355=head3 C++
1356
1357When compiling the Perl core as C++ (which is only semi-supported), the
1358mathom functions are now compiled as C<extern "C">, to ensure proper
1359binary compatibility. (However, binary compatibility isn't generally
1360guaranteed anyway in the situations where this would matter.)
1361
1362=head3 Darwin
1363
1364Stop hardcoding an alignment on 8 byte boundaries to fix builds using
1365-Dusemorebits.
1366
1367=head3 Haiku
1368
1369Perl should now work out of the box on Haiku R1 Alpha 4.
1370
1371=head3 MidnightBSD
1372
1373C<libc_r> was removed from recent versions of MidnightBSD and older versions
1374work better with C<pthread>. Threading is now enabled using C<pthread> which
1375corrects build errors with threading enabled on 0.4-CURRENT.
1376
1377=head3 Solaris
1378
1379In Configure, avoid running sed commands with flags not supported on Solaris.
1380
1381=head3 VMS
1382
1383=over
1384
1385=item *
1386
1387Where possible, the case of filenames and command-line arguments is now
1388preserved by enabling the CRTL features C<DECC$EFS_CASE_PRESERVE> and
1389C<DECC$ARGV_PARSE_STYLE> at start-up time. The latter only takes effect
1390when extended parse is enabled in the process from which Perl is run.
1391
1392=item *
1393
1394The character set for Extended Filename Syntax (EFS) is now enabled by default
1395on VMS. Among other things, this provides better handling of dots in directory
05f5908f 1396names, multiple dots in filenames, and spaces in filenames. To obtain the old
5ed58cbd
RS
1397behavior, set the logical name C<DECC$EFS_CHARSET> to C<DISABLE>.
1398
1399=item *
1400
05f5908f 1401Fixed linking on builds configured with C<-Dusemymalloc=y>.
5ed58cbd
RS
1402
1403=item *
1404
05f5908f
CB
1405Experimental support for building Perl with the HP C++ compiler is available
1406by configuring with C<-Dusecxx>.
5ed58cbd
RS
1407
1408=item *
1409
1410All C header files from the top-level directory of the distribution are now
1411installed on VMS, providing consistency with a long-standing practice on other
1412platforms. Previously only a subset were installed, which broke non-core
1413extension builds for extensions that depended on the missing include files.
1414
1415=item *
1416
1417Quotes are now removed from the command verb (but not the parameters) for
1418commands spawned via C<system>, backticks, or a piped C<open>. Previously,
1419quotes on the verb were passed through to DCL, which would fail to recognize
1420the command. Also, if the verb is actually a path to an image or command
1421procedure on an ODS-5 volume, quoting it now allows the path to contain spaces.
1422
1423=item *
1424
1425The B<a2p> build has been fixed for the HP C++ compiler on OpenVMS.
1426
1427=back
1428
1429=head3 Win32
1430
1431=over
1432
1433=item *
1434
1435Perl can now be built using Microsoft's Visual C++ 2012 compiler by specifying
1436CCTYPE=MSVC110 (or MSVC110FREE if you are using the free Express edition for
1437Windows Desktop) in F<win32/Makefile>.
1438
1439=item *
1440
f105b7be 1441The option to build without C<USE_SOCKETS_AS_HANDLES> has been removed.
5ed58cbd
RS
1442
1443=item *
1444
1445Fixed a problem where perl could crash while cleaning up threads (including the
1446main thread) in threaded debugging builds on Win32 and possibly other platforms
1447[perl #114496].
1448
1449=item *
1450
1451A rare race condition that would lead to L<sleep|perlfunc/sleep> taking more
1452time than requested, and possibly even hanging, has been fixed [perl #33096].
1453
1454=item *
1455
1456C<link> on Win32 now attempts to set C<$!> to more appropriate values
1457based on the Win32 API error code. [perl #112272]
1458
1459Perl no longer mangles the environment block, e.g. when launching a new
1460sub-process, when the environment contains non-ASCII characters. Known
1461problems still remain, however, when the environment contains characters
1462outside of the current ANSI codepage (e.g. see the item about Unicode in
1463C<%ENV> in L<http://perl5.git.perl.org/perl.git/blob/HEAD:/Porting/todo.pod>).
1464[perl #113536]
1465
1466=item *
1467
1468Building perl with some Windows compilers used to fail due to a problem
1469with miniperl's C<glob> operator (which uses the C<perlglob> program)
1470deleting the PATH environment variable [perl #113798].
1471
1472=item *
1473
f105b7be 1474A new makefile option, C<USE_64_BIT_INT>, has been added to the Windows
5ed58cbd
RS
1475makefiles. Set this to "define" when building a 32-bit perl if you want
1476it to use 64-bit integers.
1477
1478Machine code size reductions, already made to the DLLs of XS modules in
e612b5a0 1479Perl v5.17.2, have now been extended to the perl DLL itself.
5ed58cbd 1480
e612b5a0 1481Building with VC++ 6.0 was inadvertently broken in Perl v5.17.2 but has
5ed58cbd
RS
1482now been fixed again.
1483
1484=back
1485
1486=head3 WinCE
1487
1488Building on WinCE is now possible once again, although more work is required
1489to fully restore a clean build.
1490
1491=head1 Internal Changes
1492
1493=over
1494
1495=item *
1496
4263dd11 1497Synonyms for the misleadingly named C<av_len()> have been created:
5ed58cbd
RS
1498C<av_top_index()> and C<av_tindex>. All three of these return the
1499number of the highest index in the array, not the number of elements it
1500contains.
1501
1502=item *
1503
1504SvUPGRADE() is no longer an expression. Originally this macro (and its
1505underlying function, sv_upgrade()) were documented as boolean, although
1506in reality they always croaked on error and never returned false. In 2005
1507the documentation was updated to specify a void return value, but
1508SvUPGRADE() was left always returning 1 for backwards compatibility. This
1509has now been removed, and SvUPGRADE() is now a statement with no return
1510value.
1511
1512So this is now a syntax error:
1513
1514 if (!SvUPGRADE(sv)) { croak(...); }
1515
1516If you have code like that, simply replace it with
1517
1518 SvUPGRADE(sv);
1519
237b9c6d 1520or to avoid compiler warnings with older perls, possibly
5ed58cbd
RS
1521
1522 (void)SvUPGRADE(sv);
1523
1524=item *
1525
1526Perl has a new copy-on-write mechanism that allows any SvPOK scalar to be
1527upgraded to a copy-on-write scalar. A reference count on the string buffer
d16360cf
RS
1528is stored in the string buffer itself. This feature is B<not enabled by
1529default>.
5ed58cbd 1530
d16360cf
RS
1531It can be enabled in a perl build by running F<Configure> with
1532B<-Accflags=-DPERL_NEW_COPY_ON_WRITE>, and we would encourage XS authors
1533to try their code with such an enabled perl, and provide feedback.
1534Unfortunately, there is not yet a good guide to updating XS code to cope
1535with COW. Until such a document is available, consult the perl5-porters
1536mailing list.
5ed58cbd 1537
d16360cf
RS
1538It breaks a few XS modules by allowing copy-on-write scalars to go
1539through code paths that never encountered them before.
5ed58cbd
RS
1540
1541=item *
1542
1543Copy-on-write no longer uses the SvFAKE and SvREADONLY flags. Hence,
1544SvREADONLY indicates a true read-only SV.
1545
1546Use the SvIsCOW macro (as before) to identify a copy-on-write scalar.
1547
1548=item *
1549
f105b7be 1550C<PL_glob_index> is gone.
5ed58cbd
RS
1551
1552=item *
1553
1554The private Perl_croak_no_modify has had its context parameter removed. It is
1555now has a void prototype. Users of the public API croak_no_modify remain
1556unaffected.
1557
1558=item *
1559
1560Copy-on-write (shared hash key) scalars are no longer marked read-only.
1561C<SvREADONLY> returns false on such an SV, but C<SvIsCOW> still returns
1562true.
1563
1564=item *
1565
1566A new op type, C<OP_PADRANGE> has been introduced. The perl peephole
1567optimiser will, where possible, substitute a single padrange op for a
1568pushmark followed by one or more pad ops, and possibly also skipping list
1569and nextstate ops. In addition, the op can carry out the tasks associated
f105b7be 1570with the RHS of a C<< my(...) = @_ >> assignment, so those ops may be optimised
5ed58cbd
RS
1571away too.
1572
1573=item *
1574
1575Case-insensitive matching inside a [bracketed] character class with a
1576multi-character fold no longer excludes one of the possibilities in the
1577circumstances that it used to. [perl #89774].
1578
1579=item *
1580
1581C<PL_formfeed> has been removed.
1582
1583=item *
1584
1585The regular expression engine no longer reads one byte past the end of the
1586target string. While for all internally well-formed scalars this should
1587never have been a problem, this change facilitates clever tricks with
1588string buffers in CPAN modules. [perl #73542]
1589
1590=item *
1591
1592Inside a BEGIN block, C<PL_compcv> now points to the currently-compiling
1593subroutine, rather than the BEGIN block itself.
1594
1595=item *
1596
1597C<mg_length> has been deprecated.
1598
1599=item *
1600
1601C<sv_len> now always returns a byte count and C<sv_len_utf8> a character
1602count. Previously, C<sv_len> and C<sv_len_utf8> were both buggy and would
1603sometimes returns bytes and sometimes characters. C<sv_len_utf8> no longer
be12dd22 1604assumes that its argument is in UTF-8. Neither of these creates UTF-8 caches
5ed58cbd
RS
1605for tied or overloaded values or for non-PVs any more.
1606
1607=item *
1608
1609C<sv_mortalcopy> now copies string buffers of shared hash key scalars when
1610called from XS modules [perl #79824].
1611
1612=item *
1613
1614C<RXf_SPLIT> and C<RXf_SKIPWHITE> are no longer used. They are now
1615#defined as 0.
1616
1617=item *
1618
1619The new C<RXf_MODIFIES_VARS> flag can be set by custom regular expression
1620engines to indicate that the execution of the regular expression may cause
1621variables to be modified. This lets C<s///> know to skip certain
1622optimisations. Perl's own regular expression engine sets this flag for the
1623special backtracking verbs that set $REGMARK and $REGERROR.
1624
1625=item *
1626
1627The APIs for accessing lexical pads have changed considerably.
1628
1629C<PADLIST>s are now longer C<AV>s, but their own type instead.
1630C<PADLIST>s now contain a C<PAD> and a C<PADNAMELIST> of C<PADNAME>s,
1631rather than C<AV>s for the pad and the list of pad names. C<PAD>s,
1632C<PADNAMELIST>s, and C<PADNAME>s are to be accessed as such through the
1633newly added pad API instead of the plain C<AV> and C<SV> APIs. See
1634L<perlapi> for details.
1635
1636=item *
1637
1638In the regex API, the numbered capture callbacks are passed an index
1639indicating what match variable is being accessed. There are special
1640index values for the C<$`, $&, $&> variables. Previously the same three
1641values were used to retrieve C<${^PREMATCH}, ${^MATCH}, ${^POSTMATCH}>
1642too, but these have now been assigned three separate values. See
1643L<perlreapi/Numbered capture callbacks>.
1644
1645=item *
1646
1647C<PL_sawampersand> was previously a boolean indicating that any of
1648C<$`, $&, $&> had been seen; it now contains three one-bit flags
1649indicating the presence of each of the variables individually.
1650
1651=item *
1652
1653The C<CV *> typemap entry now supports C<&{}> overloading and typeglobs,
1654just like C<&{...}> [perl #96872].
1655
1656=item *
1657
1658The C<SVf_AMAGIC> flag to indicate overloading is now on the stash, not the
1659object. It is now set automatically whenever a method or @ISA changes, so
1660its meaning has changed, too. It now means "potentially overloaded". When
1661the overload table is calculated, the flag is automatically turned off if
1662there is no overloading, so there should be no noticeable slowdown.
1663
1664The staleness of the overload tables is now checked when overload methods
1665are invoked, rather than during C<bless>.
1666
1667"A" magic is gone. The changes to the handling of the C<SVf_AMAGIC> flag
1668eliminate the need for it.
1669
1670C<PL_amagic_generation> has been removed as no longer necessary. For XS
1671modules, it is now a macro alias to C<PL_na>.
1672
1673The fallback overload setting is now stored in a stash entry separate from
1674overloadedness itself.
1675
1676=item *
1677
1678The character-processing code has been cleaned up in places. The changes
1679should be operationally invisible.
1680
1681=item *
1682
e612b5a0 1683The C<study> function was made a no-op in v5.16. It was simply disabled via
5ed58cbd
RS
1684a C<return> statement; the code was left in place. Now the code supporting
1685what C<study> used to do has been removed.
1686
1687=item *
1688
1689Under threaded perls, there is no longer a separate PV allocated for every
1690COP to store its package name (C<< cop->stashpv >>). Instead, there is an
1691offset (C<< cop->stashoff >>) into the new C<PL_stashpad> array, which
1692holds stash pointers.
1693
1694=item *
1695
1696In the pluggable regex API, the C<regexp_engine> struct has acquired a new
1697field C<op_comp>, which is currently just for perl's internal use, and
f105b7be 1698should be initialized to NULL by other regex plugin modules.
5ed58cbd
RS
1699
1700=item *
1701
7779650e 1702A new function C<alloccopstash> has been added to the API, but is considered
5ed58cbd
RS
1703experimental. See L<perlapi>.
1704
1705=item *
1706
1707Perl used to implement get magic in a way that would sometimes hide bugs in
4263dd11 1708code that could call mg_get() too many times on magical values. This hiding of
5ed58cbd
RS
1709errors no longer occurs, so long-standing bugs may become visible now. If
1710you see magic-related errors in XS code, check to make sure it, together
1711with the Perl API functions it uses, calls mg_get() only once on SvGMAGICAL()
1712values.
1713
1714=item *
1715
1716OP allocation for CVs now uses a slab allocator. This simplifies
1717memory management for OPs allocated to a CV, so cleaning up after a
1718compilation error is simpler and safer [perl #111462][perl #112312].
1719
1720=item *
1721
f105b7be 1722C<PERL_DEBUG_READONLY_OPS> has been rewritten to work with the new slab
5ed58cbd
RS
1723allocator, allowing it to catch more violations than before.
1724
1725=item *
1726
f105b7be
KE
1727The old slab allocator for ops, which was only enabled for C<PERL_IMPLICIT_SYS>
1728and C<PERL_DEBUG_READONLY_OPS>, has been retired.
5ed58cbd
RS
1729
1730=back
1731
1732=head1 Selected Bug Fixes
1733
1734=over 4
1735
1736=item *
1737
1738Here-doc terminators no longer require a terminating newline character when
1739they occur at the end of a file. This was already the case at the end of a
1740string eval [perl #65838].
1741
1742=item *
1743
f105b7be 1744C<-DPERL_GLOBAL_STRUCT> builds now free the global struct B<after>
5ed58cbd
RS
1745they've finished using it.
1746
1747=item *
1748
1749A trailing '/' on a path in @INC will no longer have an additional '/'
1750appended.
1751
1752=item *
1753
1754The C<:crlf> layer now works when unread data doesn't fit into its own
1755buffer. [perl #112244].
1756
1757=item *
1758
1759C<ungetc()> now handles UTF-8 encoded data. [perl #116322].
1760
1761=item *
1762
1763A bug in the core typemap caused any C types that map to the T_BOOL core
1764typemap entry to not be set, updated, or modified when the T_BOOL variable was
1765used in an OUTPUT: section with an exception for RETVAL. T_BOOL in an INPUT:
1766section was not affected. Using a T_BOOL return type for an XSUB (RETVAL)
1767was not affected. A side effect of fixing this bug is, if a T_BOOL is specified
1768in the OUTPUT: section (which previous did nothing to the SV), and a read only
1769SV (literal) is passed to the XSUB, croaks like "Modification of a read-only
1770value attempted" will happen. [perl #115796]
1771
1772=item *
1773
1774On many platforms, providing a directory name as the script name caused perl
1775to do nothing and report success. It should now universally report an error
1776and exit nonzero. [perl #61362]
1777
1778=item *
1779
1780C<sort {undef} ...> under fatal warnings no longer crashes. It had
e612b5a0 1781begun crashing in Perl v5.16.
5ed58cbd
RS
1782
1783=item *
1784
1785Stashes blessed into each other
1786(C<bless \%Foo::, 'Bar'; bless \%Bar::, 'Foo'>) no longer result in double
e612b5a0 1787frees. This bug started happening in Perl v5.16.
5ed58cbd
RS
1788
1789=item *
1790
1791Numerous memory leaks have been fixed, mostly involving fatal warnings and
1792syntax errors.
1793
1794=item *
1795
1796Some failed regular expression matches such as C<'f' =~ /../g> were not
1797resetting C<pos>. Also, "match-once" patterns (C<m?...?g>) failed to reset
1798it, too, when invoked a second time [perl #23180].
1799
1800=item *
1801
1802Accessing C<$&> after a pattern match now works if it had not been seen
1803before the match. I.e., this applies to C<${'&'}> (under C<no strict>) and
1804C<eval '$&'>. The same applies to C<$'> and C<$`> [perl #4289].
1805
1806=item *
1807
1808Several bugs involving C<local *ISA> and C<local *Foo::> causing stale
1809MRO caches have been fixed.
1810
1811=item *
1812
1813Defining a subroutine when its typeglob has been aliased no longer results
e612b5a0 1814in stale method caches. This bug was introduced in Perl v5.10.
5ed58cbd
RS
1815
1816=item *
1817
1818Localising a typeglob containing a subroutine when the typeglob's package
1819has been deleted from its parent stash no longer produces an error. This
e612b5a0 1820bug was introduced in Perl v5.14.
5ed58cbd
RS
1821
1822=item *
1823
1824Under some circumstances, C<local *method=...> would fail to reset method
1825caches upon scope exit.
1826
1827=item *
1828
1829C</[.foo.]/> is no longer an error, but produces a warning (as before) and
1830is treated as C</[.fo]/> [perl #115818].
1831
1832=item *
1833
1834C<goto $tied_var> now calls FETCH before deciding what type of goto
1835(subroutine or label) this is.
1836
1837=item *
1838
1839Renaming packages through glob assignment
1840(C<*Foo:: = *Bar::; *Bar:: = *Baz::>) in combination with C<m?...?> and
1841C<reset> no longer makes threaded builds crash.
1842
1843=item *
1844
1845A number of bugs related to assigning a list to hash have been fixed. Many of
1846these involve lists with repeated keys like C<(1, 1, 1, 1)>.
1847
1848=over 4
1849
1850=item *
1851
1852The expression C<scalar(%h = (1, 1, 1, 1))> now returns C<4>, not C<2>.
1853
1854=item *
1855
1856The return value of C<%h = (1, 1, 1)> in list context was wrong. Previously
1857this would return C<(1, undef, 1)>, now it returns C<(1, undef)>.
1858
1859=item *
1860
1861Perl now issues the same warning on C<($s, %h) = (1, {})> as it does for
1862C<(%h) = ({})>, "Reference found where even-sized list expected".
1863
1864=item *
1865
1866A number of additional edge cases in list assignment to hashes were
1867corrected. For more details see commit 23b7025ebc.
1868
1869=back
1870
1871=item *
1872
1873Attributes applied to lexical variables no longer leak memory.
1874[perl #114764]
1875
1876=item *
1877
1878C<dump>, C<goto>, C<last>, C<next>, C<redo> or C<require> followed by a
1879bareword (or version) and then an infix operator is no longer a syntax
1880error. It used to be for those infix operators (like C<+>) that have a
1881different meaning where a term is expected. [perl #105924]
1882
1883=item *
1884
1885C<require a::b . 1> and C<require a::b + 1> no longer produce erroneous
1886ambiguity warnings. [perl #107002]
1887
1888=item *
1889
1890Class method calls are now allowed on any string, and not just strings
1891beginning with an alphanumeric character. [perl #105922]
1892
1893=item *
1894
1895An empty pattern created with C<qr//> used in C<m///> no longer triggers
1896the "empty pattern reuses last pattern" behaviour. [perl #96230]
1897
1898=item *
1899
1900Tying a hash during iteration no longer results in a memory leak.
1901
1902=item *
1903
1904Freeing a tied hash during iteration no longer results in a memory leak.
1905
1906=item *
1907
1908List assignment to a tied array or hash that dies on STORE no longer
1909results in a memory leak.
1910
1911=item *
1912
1913If the hint hash (C<%^H>) is tied, compile-time scope entry (which copies
1914the hint hash) no longer leaks memory if FETCH dies. [perl #107000]
1915
1916=item *
1917
1918Constant folding no longer inappropriately triggers the special
1919C<split " "> behaviour. [perl #94490]
1920
1921=item *
1922
1923C<defined scalar(@array)>, C<defined do { &foo }>, and similar constructs
1924now treat the argument to C<defined> as a simple scalar. [perl #97466]
1925
1926=item *
1927
1928Running a custom debugging that defines no C<*DB::DB> glob or provides a
1929subroutine stub for C<&DB::DB> no longer results in a crash, but an error
1930instead. [perl #114990]
1931
1932=item *
1933
1934C<reset ""> now matches its documentation. C<reset> only resets C<m?...?>
1935patterns when called with no argument. An empty string for an argument now
1936does nothing. (It used to be treated as no argument.) [perl #97958]
1937
1938=item *
1939
1940C<printf> with an argument returning an empty list no longer reads past the
1941end of the stack, resulting in erratic behaviour. [perl #77094]
1942
1943=item *
1944
1945C<--subname> no longer produces erroneous ambiguity warnings.
1946[perl #77240]
1947
1948=item *
1949
1950C<v10> is now allowed as a label or package name. This was inadvertently
e612b5a0 1951broken when v-strings were added in Perl v5.6. [perl #56880]
5ed58cbd
RS
1952
1953=item *
1954
1955C<length>, C<pos>, C<substr> and C<sprintf> could be confused by ties,
1956overloading, references and typeglobs if the stringification of such
be12dd22 1957changed the internal representation to or from UTF-8. [perl #114410]
5ed58cbd
RS
1958
1959=item *
1960
1961utf8::encode now calls FETCH and STORE on tied variables. utf8::decode now
1962calls STORE (it was already calling FETCH).
1963
1964=item *
1965
1966C<$tied =~ s/$non_utf8/$utf8/> no longer loops infinitely if the tied
1967variable returns a Latin-1 string, shared hash key scalar, or reference or
2ae351f8 1968typeglob that stringifies as ASCII or Latin-1. This was a regression from
e612b5a0 1969v5.12.
5ed58cbd
RS
1970
1971=item *
1972
1973C<s///> without /e is now better at detecting when it needs to forego
1974certain optimisations, fixing some buggy cases:
1975
1976=over
1977
1978=item *
1979
1980Match variables in certain constructs (C<&&>, C<||>, C<..> and others) in
1981the replacement part; e.g., C<s/(.)/$l{$a||$1}/g>. [perl #26986]
1982
1983=item *
1984
1985Aliases to match variables in the replacement.
1986
1987=item *
1988
1989C<$REGERROR> or C<$REGMARK> in the replacement. [perl #49190]
1990
1991=item *
1992
1993An empty pattern (C<s//$foo/>) that causes the last-successful pattern to
1994be used, when that pattern contains code blocks that modify the variables
1995in the replacement.
1996
1997=back
1998
1999=item *
2000
2001The taintedness of the replacement string no longer affects the taintedness
2002of the return value of C<s///e>.
2003
2004=item *
2005
2006The C<$|> autoflush variable is created on-the-fly when needed. If this
2007happened (e.g., if it was mentioned in a module or eval) when the
2008currently-selected filehandle was a typeglob with an empty IO slot, it used
2009to crash. [perl #115206]
2010
2011=item *
2012
2013Line numbers at the end of a string eval are no longer off by one.
2014[perl #114658]
2015
2016=item *
2017
2018@INC filters (subroutines returned by subroutines in @INC) that set $_ to a
2019copy-on-write scalar no longer cause the parser to modify that string
2020buffer in place.
2021
2022=item *
2023
2024C<length($object)> no longer returns the undefined value if the object has
2025string overloading that returns undef. [perl #115260]
2026
2027=item *
2028
2029The use of C<PL_stashcache>, the stash name lookup cache for method calls, has
2030been restored,
2031
2032Commit da6b625f78f5f133 in August 2011 inadvertently broke the code that looks
2033up values in C<PL_stashcache>. As it's a only cache, quite correctly everything
2034carried on working without it.
2035
2036=item *
2037
e612b5a0 2038The error "Can't localize through a reference" had disappeared in v5.16.0
5ed58cbd 2039when C<local %$ref> appeared on the last line of an lvalue subroutine.
e612b5a0 2040This error disappeared for C<\local %$ref> in perl v5.8.1. It has now
5ed58cbd
RS
2041been restored.
2042
2043=item *
2044
2045The parsing of here-docs has been improved significantly, fixing several
2046parsing bugs and crashes and one memory leak, and correcting wrong
2047subsequent line numbers under certain conditions.
2048
2049=item *
2050
2051Inside an eval, the error message for an unterminated here-doc no longer
2052has a newline in the middle of it [perl #70836].
2053
2054=item *
2055
2056A substitution inside a substitution pattern (C<s/${s|||}//>) no longer
2057confuses the parser.
2058
2059=item *
2060
2061It may be an odd place to allow comments, but C<s//"" # hello/e> has
2062always worked, I<unless> there happens to be a null character before the
2063first #. Now it works even in the presence of nulls.
2064
2065=item *
2066
2067An invalid range in C<tr///> or C<y///> no longer results in a memory leak.
2068
2069=item *
2070
2071String eval no longer treats a semicolon-delimited quote-like operator at
2072the very end (C<eval 'q;;'>) as a syntax error.
2073
2074=item *
2075
2076C<< warn {$_ => 1} + 1 >> is no longer a syntax error. The parser used to
2077get confused with certain list operators followed by an anonymous hash and
2078then an infix operator that shares its form with a unary operator.
2079
2080=item *
2081
2082C<(caller $n)[6]> (which gives the text of the eval) used to return the
2083actual parser buffer. Modifying it could result in crashes. Now it always
2084returns a copy. The string returned no longer has "\n;" tacked on to the
2085end. The returned text also includes here-doc bodies, which used to be
2086omitted.
2087
2088=item *
2089
be12dd22
RS
2090Reset the UTF-8 position cache when accessing magical variables to avoid the
2091string buffer and the UTF-8 position cache getting out of sync
5ed58cbd
RS
2092[perl #114410].
2093
2094=item *
2095
be12dd22
RS
2096Various cases of get magic being called twice for magical UTF-8
2097strings have been fixed.
5ed58cbd
RS
2098
2099=item *
2100
2101This code (when not in the presence of C<$&> etc)
2102
2103 $_ = 'x' x 1_000_000;
2104 1 while /(.)/;
2105
2106used to skip the buffer copy for performance reasons, but suffered from C<$1>
2107etc changing if the original string changed. That's now been fixed.
2108
2109=item *
2110
2111Perl doesn't use PerlIO anymore to report out of memory messages, as PerlIO
2112might attempt to allocate more memory.
2113
2114=item *
2115
2116In a regular expression, if something is quantified with C<{n,m}> where
2117C<S<n E<gt> m>>, it can't possibly match. Previously this was a fatal
2118error, but now is merely a warning (and that something won't match).
2119[perl #82954].
2120
2121=item *
2122
2123It used to be possible for formats defined in subroutines that have
2124subsequently been undefined and redefined to close over variables in the
2125wrong pad (the newly-defined enclosing sub), resulting in crashes or
2126"Bizarre copy" errors.
2127
2128=item *
2129
2130Redefinition of XSUBs at run time could produce warnings with the wrong
2131line number.
2132
2133=item *
2134
2135The %vd sprintf format does not support version objects for alpha versions.
2136It used to output the format itself (%vd) when passed an alpha version, and
2137also emit an "Invalid conversion in printf" warning. It no longer does,
2138but produces the empty string in the output. It also no longer leaks
2139memory in this case.
2140
2141=item *
2142
2143C<< $obj->SUPER::method >> calls in the main package could fail if the
2144SUPER package had already been accessed by other means.
2145
2146=item *
2147
f105b7be 2148Stash aliasing (C<< *foo:: = *bar:: >>) no longer causes SUPER calls to ignore
5ed58cbd
RS
2149changes to methods or @ISA or use the wrong package.
2150
2151=item *
2152
2153Method calls on packages whose names end in ::SUPER are no longer treated
2154as SUPER method calls, resulting in failure to find the method.
2155Furthermore, defining subroutines in such packages no longer causes them to
2156be found by SUPER method calls on the containing package [perl #114924].
2157
2158=item *
2159
2160C<\w> now matches the code points U+200C (ZERO WIDTH NON-JOINER) and U+200D
2161(ZERO WIDTH JOINER). C<\W> no longer matches these. This change is because
2162Unicode corrected their definition of what C<\w> should match.
2163
2164=item *
2165
2166C<dump LABEL> no longer leaks its label.
2167
2168=item *
2169
2170Constant folding no longer changes the behaviour of functions like C<stat()>
2171and C<truncate()> that can take either filenames or handles.
2172C<stat 1 ? foo : bar> nows treats its argument as a file name (since it is an
2173arbitrary expression), rather than the handle "foo".
2174
2175=item *
2176
2177C<truncate FOO, $len> no longer falls back to treating "FOO" as a file name if
e612b5a0 2178the filehandle has been deleted. This was broken in Perl v5.16.0.
5ed58cbd
RS
2179
2180=item *
2181
2182Subroutine redefinitions after sub-to-glob and glob-to-glob assignments no
2183longer cause double frees or panic messages.
2184
2185=item *
2186
2187C<s///> now turns vstrings into plain strings when performing a substitution,
2188even if the resulting string is the same (C<s/a/a/>).
2189
2190=item *
2191
2192Prototype mismatch warnings no longer erroneously treat constant subs as having
2193no prototype when they actually have "".
2194
2195=item *
2196
2197Constant subroutines and forward declarations no longer prevent prototype
2198mismatch warnings from omitting the sub name.
2199
2200=item *
2201
2202C<undef> on a subroutine now clears call checkers.
2203
2204=item *
2205
e612b5a0 2206The C<ref> operator started leaking memory on blessed objects in Perl v5.16.0.
5ed58cbd
RS
2207This has been fixed [perl #114340].
2208
2209=item *
2210
2211C<use> no longer tries to parse its arguments as a statement, making
2212C<use constant { () };> a syntax error [perl #114222].
2213
2214=item *
2215
2216On debugging builds, "uninitialized" warnings inside formats no longer cause
2217assertion failures.
2218
2219=item *
2220
2221On debugging builds, subroutines nested inside formats no longer cause
2222assertion failures [perl #78550].
2223
2224=item *
2225
2226Formats and C<use> statements are now permitted inside formats.
2227
2228=item *
2229
2230C<print $x> and C<sub { print $x }-E<gt>()> now always produce the same output.
2231It was possible for the latter to refuse to close over $x if the variable was
2232not active; e.g., if it was defined outside a currently-running named
2233subroutine.
2234
2235=item *
2236
2237Similarly, C<print $x> and C<print eval '$x'> now produce the same output.
2238This also allows "my $x if 0" variables to be seen in the debugger [perl
2239#114018].
2240
2241=item *
2242
2243Formats called recursively no longer stomp on their own lexical variables, but
2244each recursive call has its own set of lexicals.
2245
2246=item *
2247
2248Attempting to free an active format or the handle associated with it no longer
2249results in a crash.
2250
2251=item *
2252
2253Format parsing no longer gets confused by braces, semicolons and low-precedence
2254operators. It used to be possible to use braces as format delimiters (instead
2255of C<=> and C<.>), but only sometimes. Semicolons and low-precedence operators
2256in format argument lines no longer confuse the parser into ignoring the line's
2257return value. In format argument lines, braces can now be used for anonymous
2258hashes, instead of being treated always as C<do> blocks.
2259
2260=item *
2261
2262Formats can now be nested inside code blocks in regular expressions and other
2263quoted constructs (C</(?{...})/> and C<qq/${...}/>) [perl #114040].
2264
2265=item *
2266
2267Formats are no longer created after compilation errors.
2268
2269=item *
2270
2271Under debugging builds, the B<-DA> command line option started crashing in Perl
e612b5a0 2272v5.16.0. It has been fixed [perl #114368].
5ed58cbd
RS
2273
2274=item *
2275
2276A potential deadlock scenario involving the premature termination of a pseudo-
2277forked child in a Windows build with ithreads enabled has been fixed. This
2278resolves the common problem of the F<t/op/fork.t> test hanging on Windows [perl
2279#88840].
2280
2281=item *
2282
5ed58cbd
RS
2283The code which generates errors from C<require()> could potentially read one or
2284two bytes before the start of the filename for filenames less than three bytes
2285long and ending C</\.p?\z/>. This has now been fixed. Note that it could
2286never have happened with module names given to C<use()> or C<require()> anyway.
2287
2288=item *
2289
2290The handling of pathnames of modules given to C<require()> has been made
2291thread-safe on VMS.
2292
2293=item *
2294
d85cd26b
RS
2295Non-blocking sockets have been fixed on VMS.
2296
2297=item *
2298
5ed58cbd
RS
2299A bug in the compilation of a C</(?{})/> expression which affected the TryCatch
2300test suite has been fixed [perl #114242].
2301
2302=item *
2303
2304Pod can now be nested in code inside a quoted construct outside of a string
2305eval. This used to work only within string evals [perl #114040].
2306
2307=item *
2308
2309C<goto ''> now looks for an empty label, producing the "goto must have
2310label" error message, instead of exiting the program [perl #111794].
2311
2312=item *
2313
2314C<goto "\0"> now dies with "Can't find label" instead of "goto must have
2315label".
2316
2317=item *
2318
2319The C function C<hv_store> used to result in crashes when used on C<%^H>
2320[perl #111000].
2321
2322=item *
2323
2324A call checker attached to a closure prototype via C<cv_set_call_checker>
2325is now copied to closures cloned from it. So C<cv_set_call_checker> now
2326works inside an attribute handler for a closure.
2327
2328=item *
2329
2330Writing to C<$^N> used to have no effect. Now it croaks with "Modification
2331of a read-only value" by default, but that can be overridden by a custom
2332regular expression engine, as with C<$1> [perl #112184].
2333
2334=item *
2335
2336C<undef> on a control character glob (C<undef *^H>) no longer emits an
2337erroneous warning about ambiguity [perl #112456].
2338
2339=item *
2340
2341For efficiency's sake, many operators and built-in functions return the
2342same scalar each time. Lvalue subroutines and subroutines in the CORE::
2343namespace were allowing this implementation detail to leak through.
2344C<print &CORE::uc("a"), &CORE::uc("b")> used to print "BB". The same thing
2345would happen with an lvalue subroutine returning the return value of C<uc>.
2346Now the value is copied in such cases.
2347
2348=item *
2349
2350C<method {}> syntax with an empty block or a block returning an empty list
2351used to crash or use some random value left on the stack as its invocant.
2352Now it produces an error.
2353
2354=item *
2355
2356C<vec> now works with extremely large offsets (E<gt>2 GB) [perl #111730].
2357
2358=item *
2359
2360Changes to overload settings now take effect immediately, as do changes to
2361inheritance that affect overloading. They used to take effect only after
2362C<bless>.
2363
2364Objects that were created before a class had any overloading used to remain
2365non-overloaded even if the class gained overloading through C<use overload>
2366or @ISA changes, and even after C<bless>. This has been fixed
2367[perl #112708].
2368
2369=item *
2370
2371Classes with overloading can now inherit fallback values.
2372
2373=item *
2374
2375Overloading was not respecting a fallback value of 0 if there were
2376overloaded objects on both sides of an assignment operator like C<+=>
2377[perl #111856].
2378
2379=item *
2380
2381C<pos> now croaks with hash and array arguments, instead of producing
2382erroneous warnings.
2383
2384=item *
2385
2386C<while(each %h)> now implies C<while(defined($_ = each %h))>, like
2387C<readline> and C<readdir>.
2388
2389=item *
2390
2391Subs in the CORE:: namespace no longer crash after C<undef *_> when called
2392with no argument list (C<&CORE::time> with no parentheses).
2393
2394=item *
2395
2396C<unpack> no longer produces the "'/' must follow a numeric type in unpack"
2397error when it is the data that are at fault [perl #60204].
2398
2399=item *
2400
2401C<join> and C<"@array"> now call FETCH only once on a tied C<$">
2402[perl #8931].
2403
2404=item *
2405
2406Some subroutine calls generated by compiling core ops affected by a
2407C<CORE::GLOBAL> override had op checking performed twice. The checking
2408is always idempotent for pure Perl code, but the double checking can
2409matter when custom call checkers are involved.
2410
2411=item *
2412
2413A race condition used to exist around fork that could cause a signal sent to
2414the parent to be handled by both parent and child. Signals are now blocked
2415briefly around fork to prevent this from happening [perl #82580].
2416
2417=item *
2418
2419The implementation of code blocks in regular expressions, such as C<(?{})>
2420and C<(??{})>, has been heavily reworked to eliminate a whole slew of bugs.
2421The main user-visible changes are:
2422
2423=over 4
2424
2425=item *
2426
2427Code blocks within patterns are now parsed in the same pass as the
2428surrounding code; in particular it is no longer necessary to have balanced
2429braces: this now works:
2430
2431 /(?{ $x='{' })/
2432
2433This means that this error message is no longer generated:
2434
2435 Sequence (?{...}) not terminated or not {}-balanced in regex
2436
2437but a new error may be seen:
2438
2439 Sequence (?{...}) not terminated with ')'
2440
2441In addition, literal code blocks within run-time patterns are only
2442compiled once, at perl compile-time:
2443
2444 for my $p (...) {
2445 # this 'FOO' block of code is compiled once,
2446 # at the same time as the surrounding 'for' loop
2447 /$p{(?{FOO;})/;
2448 }
2449
2450=item *
2451
2452Lexical variables are now sane as regards scope, recursion and closure
2453behavior. In particular, C</A(?{B})C/> behaves (from a closure viewpoint)
2454exactly like C</A/ && do { B } && /C/>, while C<qr/A(?{B})C/> is like
2455C<sub {/A/ && do { B } && /C/}>. So this code now works how you might
2456expect, creating three regexes that match 0, 1, and 2:
2457
2458 for my $i (0..2) {
2459 push @r, qr/^(??{$i})$/;
2460 }
2461 "1" =~ $r[1]; # matches
2462
2463=item *
2464
2465The C<use re 'eval'> pragma is now only required for code blocks defined
2466at runtime; in particular in the following, the text of the C<$r> pattern is
2467still interpolated into the new pattern and recompiled, but the individual
2468compiled code-blocks within C<$r> are reused rather than being recompiled,
2469and C<use re 'eval'> isn't needed any more:
2470
2471 my $r = qr/abc(?{....})def/;
2472 /xyz$r/;
2473
2474=item *
2475
2476Flow control operators no longer crash. Each code block runs in a new
2477dynamic scope, so C<next> etc. will not see
2478any enclosing loops. C<return> returns a value
2479from the code block, not from any enclosing subroutine.
2480
2481=item *
2482
2483Perl normally caches the compilation of run-time patterns, and doesn't
2484recompile if the pattern hasn't changed, but this is now disabled if
2485required for the correct behavior of closures. For example:
2486
2487 my $code = '(??{$x})';
2488 for my $x (1..3) {
2489 # recompile to see fresh value of $x each time
2490 $x =~ /$code/;
2491 }
2492
2493=item *
2494
2495The C</msix> and C<(?msix)> etc. flags are now propagated into the return
2496value from C<(??{})>; this now works:
2497
2498 "AB" =~ /a(??{'b'})/i;
2499
2500=item *
2501
2502Warnings and errors will appear to come from the surrounding code (or for
2503run-time code blocks, from an eval) rather than from an C<re_eval>:
2504
2505 use re 'eval'; $c = '(?{ warn "foo" })'; /$c/;
2506 /(?{ warn "foo" })/;
2507
2508formerly gave:
2509
2510 foo at (re_eval 1) line 1.
2511 foo at (re_eval 2) line 1.
2512
2513and now gives:
2514
2515 foo at (eval 1) line 1.
2516 foo at /some/prog line 2.
2517
2518=back
2519
2520=item *
2521
2e7bc647
KW
2522Perl now can be recompiled to use any Unicode version. In v5.16, it
2523worked on Unicodes 6.0 and 6.1, but there were various bugs if earlier
2524releases were used; the older the release the more problems.
5ed58cbd
RS
2525
2526=item *
2527
2528C<vec> no longer produces "uninitialized" warnings in lvalue context
2529[perl #9423].
2530
2531=item *
2532
2533An optimization involving fixed strings in regular expressions could cause
2534a severe performance penalty in edge cases. This has been fixed
2535[perl #76546].
2536
2537=item *
2538
2539In certain cases, including empty subpatterns within a regular expression (such
2540as C<(?:)> or C<(?:|)>) could disable some optimizations. This has been fixed.
2541
2542=item *
2543
2544The "Can't find an opnumber" message that C<prototype> produces when passed
2545a string like "CORE::nonexistent_keyword" now passes UTF-8 and embedded
2546NULs through unchanged [perl #97478].
2547
2548=item *
2549
2550C<prototype> now treats magical variables like C<$1> the same way as
2551non-magical variables when checking for the CORE:: prefix, instead of
2552treating them as subroutine names.
2553
2554=item *
2555
2556Under threaded perls, a runtime code block in a regular expression could
2557corrupt the package name stored in the op tree, resulting in bad reads
2558in C<caller>, and possibly crashes [perl #113060].
2559
2560=item *
2561
2562Referencing a closure prototype (C<\&{$_[1]}> in an attribute handler for a
2563closure) no longer results in a copy of the subroutine (or assertion
2564failures on debugging builds).
2565
2566=item *
2567
2568C<eval '__PACKAGE__'> now returns the right answer on threaded builds if
2569the current package has been assigned over (as in
2570C<*ThisPackage:: = *ThatPackage::>) [perl #78742].
2571
2572=item *
2573
2574If a package is deleted by code that it calls, it is possible for C<caller>
2575to see a stack frame belonging to that deleted package. C<caller> could
2576crash if the stash's memory address was reused for a scalar and a
2577substitution was performed on the same scalar [perl #113486].
2578
2579=item *
2580
2581C<UNIVERSAL::can> no longer treats its first argument differently
2582depending on whether it is a string or number internally.
2583
2584=item *
2585
2586C<open> with C<< <& >> for the mode checks to see whether the third argument is
2587a number, in determining whether to treat it as a file descriptor or a handle
2588name. Magical variables like C<$1> were always failing the numeric check and
2589being treated as handle names.
2590
2591=item *
2592
2593C<warn>'s handling of magical variables (C<$1>, ties) has undergone several
2594fixes. C<FETCH> is only called once now on a tied argument or a tied C<$@>
2595[perl #97480]. Tied variables returning objects that stringify as "" are
2596no longer ignored. A tied C<$@> that happened to return a reference the
2597I<previous> time it was used is no longer ignored.
2598
2599=item *
2600
2601C<warn ""> now treats C<$@> with a number in it the same way, regardless of
2602whether it happened via C<$@=3> or C<$@="3">. It used to ignore the
2603former. Now it appends "\t...caught", as it has always done with
2604C<$@="3">.
2605
2606=item *
2607
2608Numeric operators on magical variables (e.g., S<C<$1 + 1>>) used to use
2609floating point operations even where integer operations were more appropriate,
2610resulting in loss of accuracy on 64-bit platforms [perl #109542].
2611
2612=item *
2613
2614Unary negation no longer treats a string as a number if the string happened
2615to be used as a number at some point. So, if C<$x> contains the string "dogs",
2616C<-$x> returns "-dogs" even if C<$y=0+$x> has happened at some point.
2617
2618=item *
2619
e612b5a0 2620In Perl v5.14, C<-'-10'> was fixed to return "10", not "+10". But magical
5ed58cbd
RS
2621variables (C<$1>, ties) were not fixed till now [perl #57706].
2622
2623=item *
2624
2625Unary negation now treats strings consistently, regardless of the internal
2626C<UTF8> flag.
2627
2628=item *
2629
2630A regression introduced in Perl v5.16.0 involving
2631C<tr/I<SEARCHLIST>/I<REPLACEMENTLIST>/> has been fixed. Only the first
2632instance is supposed to be meaningful if a character appears more than
2633once in C<I<SEARCHLIST>>. Under some circumstances, the final instance
2634was overriding all earlier ones. [perl #113584]
2635
2636=item *
2637
2638Regular expressions like C<qr/\87/> previously silently inserted a NUL
2639character, thus matching as if it had been written C<qr/\00087/>. Now it
2640matches as if it had been written as C<qr/87/>, with a message that the
2641sequence C<"\8"> is unrecognized.
2642
2643=item *
2644
2645C<__SUB__> now works in special blocks (C<BEGIN>, C<END>, etc.).
2646
2647=item *
2648
2649Thread creation on Windows could theoretically result in a crash if done
2650inside a C<BEGIN> block. It still does not work properly, but it no longer
2651crashes [perl #111610].
2652
2653=item *
2654
2655C<\&{''}> (with the empty string) now autovivifies a stub like any other
2656sub name, and no longer produces the "Unable to create sub" error
2657[perl #94476].
2658
2659=item *
2660
2661A regression introduced in v5.14.0 has been fixed, in which some calls
2662to the C<re> module would clobber C<$_> [perl #113750].
2663
2664=item *
2665
2666C<do FILE> now always either sets or clears C<$@>, even when the file can't be
2667read. This ensures that testing C<$@> first (as recommended by the
2668documentation) always returns the correct result.
2669
2670=item *
2671
2672The array iterator used for the C<each @array> construct is now correctly
237b9c6d
SK
2673reset when C<@array> is cleared [perl #75596]. This happens, for example, when
2674the array is globally assigned to, as in C<@array = (...)>, but not when its
5ed58cbd
RS
2675B<values> are assigned to. In terms of the XS API, it means that C<av_clear()>
2676will now reset the iterator.
2677
2678This mirrors the behaviour of the hash iterator when the hash is cleared.
2679
2680=item *
2681
2682C<< $class->can >>, C<< $class->isa >>, and C<< $class->DOES >> now return
2683correct results, regardless of whether that package referred to by C<$class>
2684exists [perl #47113].
2685
2686=item *
2687
2688Arriving signals no longer clear C<$@> [perl #45173].
2689
2690=item *
2691
2692Allow C<my ()> declarations with an empty variable list [perl #113554].
2693
2694=item *
2695
2696During parsing, subs declared after errors no longer leave stubs
2697[perl #113712].
2698
2699=item *
2700
2701Closures containing no string evals no longer hang on to their containing
2702subroutines, allowing variables closed over by outer subroutines to be
2703freed when the outer sub is freed, even if the inner sub still exists
2704[perl #89544].
2705
2706=item *
2707
2708Duplication of in-memory filehandles by opening with a "<&=" or ">&=" mode
e612b5a0 2709stopped working properly in v5.16.0. It was causing the new handle to
5ed58cbd
RS
2710reference a different scalar variable. This has been fixed [perl #113764].
2711
2712=item *
2713
2714C<qr//> expressions no longer crash with custom regular expression engines
2715that do not set C<offs> at regular expression compilation time
2716[perl #112962].
2717
2718=item *
2719
2720C<delete local> no longer crashes with certain magical arrays and hashes
2721[perl #112966].
2722
2723=item *
2724
2725C<local> on elements of certain magical arrays and hashes used not to
2726arrange to have the element deleted on scope exit, even if the element did
2727not exist before C<local>.
2728
2729=item *
2730
2731C<scalar(write)> no longer returns multiple items [perl #73690].
2732
2733=item *
2734
2735String to floating point conversions no longer misparse certain strings under
2736C<use locale> [perl #109318].
2737
2738=item *
2739
2740C<@INC> filters that die no longer leak memory [perl #92252].
2741
2742=item *
2743
2744The implementations of overloaded operations are now called in the correct
2745context. This allows, among other things, being able to properly override
2746C<< <> >> [perl #47119].
2747
2748=item *
2749
2750Specifying only the C<fallback> key when calling C<use overload> now behaves
2751properly [perl #113010].
2752
2753=item *
2754
2755C<< sub foo { my $a = 0; while ($a) { ... } } >> and
2756C<< sub foo { while (0) { ... } } >> now return the same thing [perl #73618].
2757
2758=item *
2759
2760String negation now behaves the same under C<use integer;> as it does
2761without [perl #113012].
2762
2763=item *
2764
2765C<chr> now returns the Unicode replacement character (U+FFFD) for -1,
2766regardless of the internal representation. -1 used to wrap if the argument
2767was tied or a string internally.
2768
2769=item *
2770
2771Using a C<format> after its enclosing sub was freed could crash as of
e612b5a0 2772perl v5.12.0, if the format referenced lexical variables from the outer sub.
5ed58cbd
RS
2773
2774=item *
2775
2776Using a C<format> after its enclosing sub was undefined could crash as of
e612b5a0 2777perl v5.10.0, if the format referenced lexical variables from the outer sub.
5ed58cbd
RS
2778
2779=item *
2780
2781Using a C<format> defined inside a closure, which format references
2782lexical variables from outside, never really worked unless the C<write>
e612b5a0 2783call was directly inside the closure. In v5.10.0 it even started crashing.
5ed58cbd
RS
2784Now the copy of that closure nearest the top of the call stack is used to
2785find those variables.
2786
2787=item *
2788
2789Formats that close over variables in special blocks no longer crash if a
2790stub exists with the same name as the special block before the special
2791block is compiled.
2792
2793=item *
2794
2795The parser no longer gets confused, treating C<eval foo ()> as a syntax
2796error if preceded by C<print;> [perl #16249].
2797
2798=item *
2799
2800The return value of C<syscall> is no longer truncated on 64-bit platforms
2801[perl #113980].
2802
2803=item *
2804
2805Constant folding no longer causes C<print 1 ? FOO : BAR> to print to the
2806FOO handle [perl #78064].
2807
2808=item *
2809
2810C<do subname> now calls the named subroutine and uses the file name it
2811returns, instead of opening a file named "subname".
2812
2813=item *
2814
2815Subroutines looked up by rv2cv check hooks (registered by XS modules) are
2816now taken into consideration when determining whether C<foo bar> should be
2817the sub call C<foo(bar)> or the method call C<< "bar"->foo >>.
2818
2819=item *
2820
2821C<CORE::foo::bar> is no longer treated specially, allowing global overrides
2822to be called directly via C<CORE::GLOBAL::uc(...)> [perl #113016].
2823
2824=item *
2825
2826Calling an undefined sub whose typeglob has been undefined now produces the
2827customary "Undefined subroutine called" error, instead of "Not a CODE
2828reference".
2829
2830=item *
2831
2832Two bugs involving @ISA have been fixed. C<*ISA = *glob_without_array> and
2833C<undef *ISA; @{*ISA}> would prevent future modifications to @ISA from
2834updating the internal caches used to look up methods. The
e612b5a0 2835*glob_without_array case was a regression from Perl v5.12.
5ed58cbd
RS
2836
2837=item *
2838
2839Regular expression optimisations sometimes caused C<$> with C</m> to
2840produce failed or incorrect matches [perl #114068].
2841
2842=item *
2843
2844C<__SUB__> now works in a C<sort> block when the enclosing subroutine is
2845predeclared with C<sub foo;> syntax [perl #113710].
2846
2847=item *
2848
2849Unicode properties only apply to Unicode code points, which leads to
2850some subtleties when regular expressions are matched against
2851above-Unicode code points. There is a warning generated to draw your
2852attention to this. However, this warning was being generated
2853inappropriately in some cases, such as when a program was being parsed.
2854Non-Unicode matches such as C<\w> and C<[:word;]> should not generate the
2855warning, as their definitions don't limit them to apply to only Unicode
2856code points. Now the message is only generated when matching against
2857C<\p{}> and C<\P{}>. There remains a bug, [perl #114148], for the very
2858few properties in Unicode that match just a single code point. The
2859warning is not generated if they are matched against an above-Unicode
2860code point.
2861
2862=item *
2863
2864Uninitialized warnings mentioning hash elements would only mention the
2865element name if it was not in the first bucket of the hash, due to an
2866off-by-one error.
2867
2868=item *
2869
2870A regular expression optimizer bug could cause multiline "^" to behave
2871incorrectly in the presence of line breaks, such that
2872C<"/\n\n" =~ m#\A(?:^/$)#im> would not match [perl #115242].
2873
2874=item *
2875
2876Failed C<fork> in list context no longer corrupts the stack.
2877C<@a = (1, 2, fork, 3)> used to gobble up the 2 and assign C<(1, undef, 3)>
2878if the C<fork> call failed.
2879
2880=item *
2881
2882Numerous memory leaks have been fixed, mostly involving tied variables that
2883die, regular expression character classes and code blocks, and syntax
2884errors.
2885
2886=item *
2887
2888Assigning a regular expression (C<${qr//}>) to a variable that happens to
2889hold a floating point number no longer causes assertion failures on
2890debugging builds.
2891
2892=item *
2893
2894Assigning a regular expression to a scalar containing a number no longer
f105b7be 2895causes subsequent numification to produce random numbers.
5ed58cbd
RS
2896
2897=item *
2898
2899Assigning a regular expression to a magic variable no longer wipes away the
e612b5a0 2900magic. This was a regression from v5.10.
5ed58cbd
RS
2901
2902=item *
2903
2904Assigning a regular expression to a blessed scalar no longer results in
e612b5a0 2905crashes. This was also a regression from v5.10.
5ed58cbd
RS
2906
2907=item *
2908
2909Regular expression can now be assigned to tied hash and array elements with
2910flattening into strings.
2911
2912=item *
2913
f105b7be 2914Numifying a regular expression no longer results in an uninitialized
5ed58cbd
RS
2915warning.
2916
2917=item *
2918
2919Negative array indices no longer cause EXISTS methods of tied variables to
e612b5a0 2920be ignored. This was a regression from v5.12.
5ed58cbd
RS
2921
2922=item *
2923
2924Negative array indices no longer result in crashes on arrays tied to
2925non-objects.
2926
2927=item *
2928
be12dd22
RS
2929C<$byte_overload .= $utf8> no longer results in doubly-encoded UTF-8 if the
2930left-hand scalar happened to have produced a UTF-8 string the last time
5ed58cbd
RS
2931overloading was invoked.
2932
2933=item *
2934
2935C<goto &sub> now uses the current value of @_, instead of using the array
2936the subroutine was originally called with. This means
2937C<local @_ = (...); goto &sub> now works [perl #43077].
2938
2939=item *
2940
2941If a debugger is invoked recursively, it no longer stomps on its own
2942lexical variables. Formerly under recursion all calls would share the same
2943set of lexical variables [perl #115742].
2944
2945=item *
2946
2947C<*_{ARRAY}> returned from a subroutine no longer spontaneously
2948becomes empty.
2949
2950=back
2951
2952=head1 Known Problems
2953
2954=over 4
2955
2956=item *
2957
7f50b25b 2958There are no known regressions. Please report any bugs you find!
5ed58cbd
RS
2959
2960=back
2961
2962=head1 Acknowledgements
a75569c0 2963
dd38235c
RS
2964Perl v5.18.0 represents approximately 12 months of development since
2965Perl v5.16.0 and contains approximately 400,000 lines of changes across
29662,100 files from 113 authors.
2967
2968Perl continues to flourish into its third decade thanks to a vibrant
2969community of users and developers. The following people are known to
2970have contributed the improvements that became Perl v5.18.0:
2971
2972Aaron Crane, Aaron Trevena, Abhijit Menon-Sen, Adrian M. Enache, Alan
2973Haggai Alavi, Alexandr Ciornii, Andrew Tam, Andy Dougherty, Anton Nikishaev,
2974Aristotle Pagaltzis, Augustina Blair, Bob Ernst, Brad Gilbert, Breno G. de
2975Oliveira, Brian Carlson, Brian Fraser, Charlie Gonzalez, Chip Salzenberg, Chris
2976'BinGOs' Williams, Christian Hansen, Colin Kuskie, Craig A. Berry, Dagfinn
2977Ilmari Mannsåker, Daniel Dragan, Daniel Perrett, Darin McBride, Dave Rolsky,
2978David Golden, David Leadbeater, David Mitchell, David Nicol, Dominic
2979Hargreaves, E. Choroba, Eric Brine, Evan Miller, Father Chrysostomos, Florian
2980Ragwitz, François Perrad, George Greer, Goro Fuji, H.Merijn Brand, Herbert
2981Breunung, Hugo van der Sanden, Igor Zaytsev, James E Keenan, Jan Dubois,
2982Jasmine Ahuja, Jerry D. Hedden, Jess Robinson, Jesse Luehrs, Joaquin Ferrero,
2983Joel Berger, John Goodyear, John Peacock, Karen Etheridge, Karl Williamson,
2984Karthik Rajagopalan, Kent Fredric, Leon Timmermans, Lucas Holt, Lukas Mai,
2985Marcus Holland-Moritz, Markus Jansen, Martin Hasch, Matthew Horsfall, Max
2986Maischein, Michael G Schwern, Michael Schroeder, Moritz Lenz, Nicholas Clark,
2987Niko Tyni, Oleg Nesterov, Patrik Hägglund, Paul Green, Paul Johnson, Paul
2988Marquess, Peter Martini, Rafael Garcia-Suarez, Reini Urban, Renee Baecker,
2989Rhesa Rozendaal, Ricardo Signes, Robin Barker, Ronald J. Kimball, Ruslan
2990Zakirov, Salvador Fandiño, Sawyer X, Scott Lanning, Sergey Alekseev, Shawn M
2991Moore, Shirakata Kentaro, Shlomi Fish, Sisyphus, Smylers, Steffen Müller,
2992Steve Hay, Steve Peters, Steven Schubiger, Sullivan Beck, Sven Strickroth,
2993Sébastien Aperghis-Tramoni, Thomas Sibley, Tobias Leich, Tom Wyant, Tony Cook,
2994Vadim Konovalov, Vincent Pit, Volker Schatz, Walt Mankowski, Yves Orton,
2995Zefram.
2996
2997The list above is almost certainly incomplete as it is automatically generated
2998from version control history. In particular, it does not include the names of
2999the (very much appreciated) contributors who reported issues to the Perl bug
3000tracker.
3001
3002Many of the changes included in this version originated in the CPAN modules
3003included in Perl's core. We're grateful to the entire CPAN community for
3004helping Perl to flourish.
3005
3006For a more complete list of all of Perl's historical contributors, please see
3007the F<AUTHORS> file in the Perl source distribution.
f5b73711 3008
44691e6f
AB
3009=head1 Reporting Bugs
3010
e08634c5
SH
3011If you find what you think is a bug, you might check the articles recently
3012posted to the comp.lang.perl.misc newsgroup and the perl bug database at
3013http://rt.perl.org/perlbug/ . There may also be information at
3014http://www.perl.org/ , the Perl Home Page.
44691e6f 3015
e08634c5
SH
3016If you believe you have an unreported bug, please run the L<perlbug> program
3017included with your release. Be sure to trim your bug down to a tiny but
3018sufficient test case. Your bug report, along with the output of C<perl -V>,
3019will be sent off to perlbug@perl.org to be analysed by the Perl porting team.
44691e6f
AB
3020
3021If the bug you are reporting has security implications, which make it
e08634c5
SH
3022inappropriate to send to a publicly archived mailing list, then please send it
3023to perl5-security-report@perl.org. This points to a closed subscription
3024unarchived mailing list, which includes all the core committers, who will be
3025able to help assess the impact of issues, figure out a resolution, and help
f9001595 3026co-ordinate the release of patches to mitigate or fix the problem across all
e08634c5
SH
3027platforms on which Perl is supported. Please only use this address for
3028security issues in the Perl core, not for modules independently distributed on
3029CPAN.
44691e6f
AB
3030
3031=head1 SEE ALSO
3032
e08634c5
SH
3033The F<Changes> file for an explanation of how to view exhaustive details on
3034what changed.
44691e6f
AB
3035
3036The F<INSTALL> file for how to build Perl.
3037
3038The F<README> file for general stuff.
3039
3040The F<Artistic> and F<Copying> files for copyright information.
3041
3042=cut