3 perl5120delta - what is new for perl v5.12.0
7 UPDATED MODULE LIST NEEDS TO BE GENERATED
11 This document describes differences between the 5.10.0 release and
14 Many of the bug fixes in 5.12.0 are already included in the 5.10.1
17 You can see the list of those changes in the 5.10.1 release notes (L<perl5101delta>).
20 =head1 Core Enhancements
22 =head2 New C<package NAME VERSION> syntax
24 This new syntax allows a module author to set the $VERSION of a namespace
25 when the namespace is declared with 'package'. It eliminates the need
26 for C<our $VERSION = ...> and similar constructs. E.g.
28 package Foo::Bar 1.23;
29 # $Foo::Bar::VERSION == 1.23
31 There are several advantages to this:
37 C<$VERSION> is parsed in exactly the same way as C<use NAME VERSION>
41 C<$VERSION> is set at compile time
45 C<$VERSION> is a version object that provides proper overloading of
46 comparision operators so comparing C<$VERSION> to decimal (1.23) or
47 dotted-decimal (v1.2.3) version numbers works correctly.
51 Eliminates C<$VERSION = ...> and C<eval $VERSION> clutter
55 As it requires VERSION to be a numeric literal or v-string
56 literal, it can be statically parsed by toolchain modules
57 without C<eval> the way MM-E<gt>parse_version does for C<$VERSION = ...>
61 It does not break old code with only C<package NAME>, but code that uses
62 C<package NAME VERSION> will need to be restricted to perl 5.12.0 or newer
63 This is analogous to the change to C<open> from two-args to three-args.
64 Users requiring the latest Perl will benefit, and perhaps after several
65 years, it will become a standard practice.
69 However, C<package NAME VERSION> requires a new, 'strict' version
70 number format. See L<"Version number formats"> for details.
73 =head2 The C<...> operator
75 A new operator, C<...>, nicknamed the Yada Yada operator, has been added.
76 It is intended to mark placeholder code that is not yet implemented.
77 See L<perlop/"Yada Yada Operator">. (chromatic)
79 =head2 Implicit strictures
81 Using the C<use VERSION> syntax with a version number greater or equal
82 to 5.11.0 will lexically enable strictures just like C<use strict>
83 would do (in addition to enabling features.) The following:
92 =head2 Unicode improvements
94 Perl 5.12 comes with Unicode 5.2, the latest version available to
95 us at the time of release. This version of Unicode was released in
96 October 2009. See L<http://www.unicode.org/versions/Unicode5.2.0> for
97 further details about what's changed in this version of the standard.
98 See L<perlunicode> for instructions on installing and using other versions
101 Additionally, Perl's developers have significantly improved Perl's Unicode
102 implementation. For full details, see L</Unicode overhaul> below.
104 =head2 Y2038 compliance
106 Perl's core time-related functions are now Y2038 compliant. (It may not mean much to you, but your kids will love it!)
108 =head2 qr overloading
110 It is now possible to overload the C<qr//> operator, that is,
111 conversion to regexp, like it was already possible to overload
112 conversion to boolean, string or number of objects. It is invoked when
113 an object appears on the right hand side of the C<=~> operator or when
114 it is interpolated into a regexp. See L<overload>.
116 =head2 Pluggable keywords
118 Extension modules can now cleanly hook into the Perl parser to define
119 new kinds of keyword-headed expression and compound statement. The
120 syntax following the keyword is defined entirely by the extension. This
121 allow a completely non-Perl sublanguage to be parsed inline, with the
122 correct ops cleanly generated.
124 See L<perlapi/PL_keyword_plugin> for the mechanism. The Perl core
125 source distribution also includes a new module
126 L<XS::APItest::KeywordRPN>, which implements reverse Polish notation
127 arithmetic via pluggable keywords. This module is mainly used for test
128 purposes, and is not normally installed, but also serves as an example
129 of how to use the new mechanism.
131 Perl's developers consider this feature to be experimental. We may remove
132 it or change it in a backwards-incompatible way in Perl 5.14.
134 =head2 APIs for more internals
136 The lowest layers of the lexer and parts of the pad system now have C
137 APIs available to XS extensions. These are necessary to support proper
138 use of pluggable keywords, but have other uses too. The new APIs are
139 experimental, and only cover a small proportion of what would be
140 necessary to take full advantage of the core's facilities in these
141 areas. It is intended that the Perl 5.13 development cycle will see the
142 addition of a full range of clean, supported interfaces.
144 Perl's developers consider this feature to be experimental. We may remove
145 it or change it in a backwards-incompatible way in Perl 5.14.
147 =head2 Overridable function lookup
149 Where an extension module hooks the creation of rv2cv ops to modify the
150 subroutine lookup process, this now works correctly for bareword
151 subroutine calls. This means that prototypes on subroutines referenced
152 this way will be processed correctly. (Previously bareword subroutine
153 names were initially looked up, for parsing purposes, by an unhookable
154 mechanism, so extensions could only properly influence subroutine names
155 that appeared with an C<&> sigil.)
157 =head2 A proper interface for pluggable Method Resolution Orders
159 As of Perl 5.12.0 there is a new interface for plugging and using method
160 resolution orders other than the default linear depth first search.
161 The C3 method resolution order added in 5.10.0 has been re-implemented as
162 a plugin, without changing its Perl-space interface. See L<perlmroapi> for
167 =head2 C<\N> experimental regex escape
169 Perl now supports C<\N>, a new regex escape which you can think of as
170 the inverse of C<\n>. It will match any character that is not a newline,
171 independently from the presence or absence of the single line match
172 modifier C</s>. It is not usable within a character class. C<\N{3}>
173 means to match 3 non-newlines; C<\N{5,}> means to match at least 5.
174 C<\N{NAME}> still means the character or sequence named C<NAME>, but
175 C<NAME> no longer can be things like C<3>, or C<5,>.
177 This will break a L<custom charnames translator|charnames/CUSTOM
178 TRANSLATORS> which allows numbers for character names, as C<\N{3}> will
179 now mean to match 3 non-newline characters, and not the character whose
180 name is C<3>. (No name defined by the Unicode standard is a number,
181 so only custom translators might be affected.)
183 Perl's developers are somewhat concerned about possible user confusion
184 with the existing C<\N{...}> construct which matches characters by their
185 Unicode name. Consequently, this feature is experimental. We may remove
186 it or change it in a backwards-incompatible way in Perl 5.14.
188 =head2 DTrace support
190 Perl now has some support for DTrace. See "DTrace support" in F<INSTALL>.
192 =head2 Support for C<configure_requires> in CPAN module metadata
194 Both C<CPAN> and C<CPANPLUS> now support the C<configure_requires> keyword
195 in the F<META.yml> metadata file included in most recent CPAN distributions.
196 This allows distribution authors to specify configuration prerequisites that
197 must be installed before running F<Makefile.PL> or F<Build.PL>.
199 See the documentation for C<ExtUtils::MakeMaker> or C<Module::Build> for more
200 on how to specify C<configure_requires> when creating a distribution for CPAN.
202 =head2 C<each> is now more flexible
204 The C<each> function can now operate on arrays.
206 =head2 C<$,> flexibility
208 The variable C<$,> may now be tied.
210 =head2 // in where clauses
212 // now behaves like || in when clauses
214 =head2 Enabling warnings from your shell environment
216 You can now set C<-W> from the C<PERL5OPT> environment variable
218 =head2 C<delete local>
220 C<delete local> now allows you to locally delete a hash entry.
222 =head2 New support for Abstract namespace sockets
224 Abstract namespace sockets are Linux-specific socket type that live in
225 AF_UNIX family, slightly abusing it to be able to use arbitrary
226 character arrays as addresses: They start with nul byte and are not
227 terminated by nul byte, but with the length passed to the socket()
230 =head2 32-bit limit on substr arguments removed
232 The 32-bit limit on C<substr> arguments has now been removed. The full range
233 of the system's signed and unsigned integers is now available for the C<pos>
234 and C<len> arguments.
236 =head1 Potentially Incompatible Changes
238 =head2 Deprecations warn by default
240 Perl now defaults to issuing a warning if a deprecated language feature
243 To disable this feature in a given lexical scope, you should use C<no
244 warnings 'deprecated';> For information about which language features
245 are deprecated and explanations of various deprecation warnings, please
246 see L<perldiag.pod>. See L</Deprecations> below for the list of features
247 and modules Perl's developers have deprecated as part of this release.
249 =head2 Version number formats
251 Acceptable version number formats have been formalized into "strict" and
252 "lax" rules. C<package NAME VERSION> takes a strict version number.
253 C<UNIVERSAL::VERSION> and the L<version> object constructors take lax
254 version numbers. Providing an invalid version will result in a fatal
255 error. The version argument in C<use NAME VERSION> is first parsed as a
256 numeric literal or v-string and then passed to C<UNIVERSAL::VERSION>
257 (and must then pass the "lax" format test).
259 These formats are documented fully in the L<version> module. To a first
260 approximation, a "strict" version number is a positive decimal number
261 (integer or decimal-fraction) without exponentiation or else a
262 dotted-decimal v-string with a leading 'v' character and at least three
263 components. A "lax" version number allows v-strings with fewer than
264 three components or without a leading 'v'. Under "lax" rules, both
265 decimal and dotted-decimal versions may have a trailing "alpha"
266 component separated by an underscore character after a fractional or
267 dotted-decimal component.
269 The L<version> module adds C<version::is_strict> and C<version::is_lax>
270 functions to check a scalar against these rules.
272 =head2 @INC reorganization
274 In C<@INC>, C<ARCHLIB> and C<PRIVLIB> now occur after after the current
275 version's C<site_perl> and C<vendor_perl>. Modules installed into
276 C<site_perl> and C<vendor_perl> will now be loaded in preference to
277 those installed in C<ARCHLIB> and C<PRIVLIB>.
279 =head2 Switch statement changes
281 The C<given>/C<when> switch statement handles complex statements better
282 than Perl 5.10.0 did (These enhancements are also available in
283 5.10.1 and subsequent 5.10 releases.) There are two new cases where
284 C<when> now interprets its argument as a boolean, instead of an
285 expression to be used in a smart match:
289 =item flip-flop operators
291 The C<..> and C<...> flip-flop operators are now evaluated in boolean
292 context, following their usual semantics; see L<perlop/"Range Operators">.
294 Note that, as in perl 5.10.0, C<when (1..10)> will not work to test
295 whether a given value is an integer between 1 and 10; you should use
296 C<when ([1..10])> instead (note the array reference).
298 However, contrary to 5.10.0, evaluating the flip-flop operators in boolean
299 context ensures it can now be useful in a C<when()>, notably for
300 implementing bistable conditions, like in:
302 when (/^=begin/ .. /^=end/) {
306 =item defined-or operator
308 A compound expression involving the defined-or operator, as in
309 C<when (expr1 // expr2)>, will be treated as boolean if the first
310 expression is boolean. (This just extends the existing rule that applies
311 to the regular or operator, as in C<when (expr1 || expr2)>.)
315 =head2 Smart match changes
317 Since Perl 5.10.0, Perl's developers have made a number of changes to
318 the smart match operator. These, of course, also alter the behaviour
319 of the switch statements where smart matching is implicitly used.
320 These changes were also made for the 5.10.1 release, and will remain in
321 subsequent 5.10 releases.
323 =head3 Changes to type-based dispatch
325 The smart match operator C<~~> is no longer commutative. The behaviour of
326 a smart match now depends primarily on the type of its right hand
327 argument. Moreover, its semantics have been adjusted for greater
328 consistency or usefulness in several cases. While the general backwards
329 compatibility is maintained, several changes must be noted:
335 Code references with an empty prototype are no longer treated specially.
336 They are passed an argument like the other code references (even if they
337 choose to ignore it).
341 C<%hash ~~ sub {}> and C<@array ~~ sub {}> now test that the subroutine
342 returns a true value for each key of the hash (or element of the
343 array), instead of passing the whole hash or array as a reference to
348 Due to the commutativity breakage, code references are no longer
349 treated specially when appearing on the left of the C<~~> operator,
350 but like any vulgar scalar.
354 C<undef ~~ %hash> is always false (since C<undef> can't be a key in a
355 hash). No implicit conversion to C<""> is done (as was the case in perl
360 C<$scalar ~~ @array> now always distributes the smart match across the
361 elements of the array. It's true if one element in @array verifies
362 C<$scalar ~~ $element>. This is a generalization of the old behaviour
363 that tested whether the array contained the scalar.
367 The full dispatch table for the smart match operator is given in
368 L<perlsyn/"Smart matching in detail">.
370 =head3 Smart match and overloading
372 According to the rule of dispatch based on the rightmost argument type,
373 when an object overloading C<~~> appears on the right side of the
374 operator, the overload routine will always be called (with a 3rd argument
375 set to a true value, see L<overload>.) However, when the object will
376 appear on the left, the overload routine will be called only when the
377 rightmost argument is a simple scalar. This way, distributivity of smart
378 match across arrays is not broken, as well as the other behaviours with
379 complex types (coderefs, hashes, regexes). Thus, writers of overloading
380 routines for smart match mostly need to worry only with comparing
381 against a scalar, and possibly with stringification overloading; the
382 other common cases will be automatically handled consistently.
384 C<~~> will now refuse to work on objects that do not overload it (in order
385 to avoid relying on the object's underlying structure). (However, if the
386 object overloads the stringification or the numification operators, and
387 if overload fallback is active, it will be used instead, as usual.)
389 =head2 Other potentially incompatible changes
395 The definitions of a number of Unicode properties have changed to match
396 those of the current Unicode standard. These are listed above under
397 L</Unicode overhaul>. This change may break code that expects the old definitions.
401 The boolkeys op has moved to the group of hash ops. This breaks binary
406 Filehandles are now always blessed into C<IO::File>.
408 The previous behaviour was to bless Filehandles into L<FileHandle>
409 (an empty proxy class) if it was loaded into memory and otherwise
410 to bless them into C<IO::Handle>.
414 The semantics of C<use feature :5.10*> have changed slightly.
415 See L<"Modules and Pragmata"> for more information.
419 Perl's developers now use git, rather than Perforce. This should be
420 a purely internal change only relevant to people actively working on
421 the core. However, you may see minor difference in perl as a consequence
422 of the change. For example in some of details of the output of C<perl
423 -V>. See L<perlrepository> for more information.
427 As part of the C<Test::Harness> 2.x to 3.x upgrade, the experimental
428 C<Test::Harness::Straps> module has been removed.
429 See L</"Modules and Pragmata"> for more details.
433 As part of the C<ExtUtils::MakeMaker> upgrade, the
434 C<ExtUtils::MakeMaker::bytes> and C<ExtUtils::MakeMaker::vmsish> modules
435 have been removed from this distribution.
439 C<Module::CoreList> no longer contains the C<%:patchlevel> hash.
444 C<length undef> now returns undef.
448 Unsupported private C API functions are now declared "static" to prevent
449 leakage to Perl's public API.
453 To support the bootstrapping process, F<miniperl> no longer builds with
454 UTF-8 support in the regexp engine.
456 This allows a build to complete with PERL_UNICODE set and a UTF-8 locale.
457 Without this there's a bootstrapping problem, as miniperl can't load the UTF-8
458 components of the regexp engine, because they're not yet built.
462 F<miniperl>'s @INC is now restricted to just C<-I...>, the split of
463 C<$ENV{PERL5LIB}>, and "C<.>"
467 A space or a newline is now required after a C<"#line XXX"> directive.
471 Tied filehandles now have an additional method EOF which provides the EOF type
475 To better match all other flow control statements, C<foreach> may no
476 longer be used as an attribute.
483 From time to time, Perl's developers find it necessary to deprecate
484 features or modules we've previously shipped as part of the core
485 distribution. We are well aware of the pain and frustration that a
486 backwards-incompatible change to Perl can cause for developers building
487 or maintaining software in Perl. You can be sure that when we deprecate
488 a functionality or syntax, it isn't a choice we make lightly. Sometimes,
489 we choose to deprecate functionality or syntax because it was found to
490 be poorly designed or implemented. Sometimes, this is because they're
491 holding back other features or causing performance problems. Sometimes,
492 the reasons are more complex. Wherever possible, we try to keep deprecated
493 functionality available to developers in its previous form for at least
494 one major release. So long as a deprecated feature isn't actively
495 disrupting our ability to maintain and extend Perl, we'll try to leave
496 it in place as long as possible.
498 The following items are now deprecated:
504 C<suidperl> is no longer part of Perl. It used to provide a mechanism to
505 emulate setuid permission bits on systems that don't support it properly.
508 =item Use of C<:=> to mean an empty attribute list
510 An accident of Perl's parser meant that these constructions were all
517 with the C<:> being treated as the start of an attribute list, which
518 ends before the C<=>. As whitespace is not significant here, all are
519 parsed as an empty attribute list, hence all the above are equivalent
520 to, and better written as
524 because no attribute processing is done for an empty list.
526 As is, this meant that C<:=> cannot be used as a new token, without
527 silently changing the meaning of existing code. Hence that particular
528 form is now deprecated, and will become a syntax error. If it is
529 absolutely necessary to have empty attribute lists (for example,
530 because of a code generator) then avoid the warning by adding a space
533 =item C<< UNIVERSAL->import() >>
535 The method C<< UNIVERSAL->import() >> is now deprecated. Attempting to
536 pass import arguments to a C<use UNIVERSAL> statement will result in a
540 =item Use of "goto" to jump into a construct
542 Using C<goto> to jump from an outer scope into an inner scope is now
543 deprecated. This rare use case was causing problems in the
544 implementation of scopes.
546 =item Custom character names in \N{name} that don't look like names
548 In C<\N{I<name>}>, I<name> can be just about anything. The standard Unicode
549 names have a very limited domain, but a custom name translator could create
550 names that are, for example, made up entirely of punctuation symbols. It is
551 now deprecated to make names that don't begin with an alphabetic character, and
552 aren't alphanumeric or contain other than a very few other characters,
553 namely spaces, dashes, parentheses and colons. Because of the added meaning of
554 C<\N> (See L</C<\N> experimental regex escape>), names that look like curly
555 brace -enclosed quantifiers won't work. For example, C<\N{3,4}> now means to
556 match 3 to 4 non-newlines; before a custom name C<3,4> could have been created.
558 =item Deprecated Modules
560 The following modules will be removed from the core distribution in a future
561 release, and should be installed from CPAN instead. Distributions on CPAN
562 which require these should add them to their prerequisites. The core versions
563 of these modules warnings will issue a deprecation warning.
565 If you ship a packaged version of Perl, either alone or as part of a larger
566 system, then you should carefully consider the reprecussions of core module
567 deprecations. You may want to consider shipping your default build of
568 Perl with packages for some or all deprecated modules which install into
569 C<vendor> or C<site> perl library directories. This will inhibit the
570 deprecation warnings.
572 Alternatively, you may want to consider patching F<lib/deprecate.pm>
573 to provide deprecation warnings specific to your packaging system or
574 distribution of Perl, consistent with how your packaging system or
575 distribution manages a staged transition from a release where the
576 installation of a single package provides the given functionality, to a later
577 release where the system administrator needs to know to install multiple
578 packages to get that same functionality.
584 =item L<Pod::Plainer>
590 Switch is buggy and should be avoided. You may find Perl's new
591 C<given>/C<when> feature a suitable replacement. See L<perlsyn/"Switch
592 statements"> for more information.
596 =item Assignment to $[
598 =item Use of the attribute :locked on subroutines
600 =item Use of "locked" with the attributes pragma
602 =item Use of "unique" with the attributes pragma
606 C<Perl_pmflag> is no longer part of Perl's public API. Calling it now
607 generates a deprecation warning, and it will be removed in a future
608 release. Although listed as part of the API, it was never documented,
609 and only ever used in F<toke.c>, and prior to 5.10, F<regcomp.c>. In
610 core, it has been replaced by a static function.
612 =item Numerous Perl 4-era libraries
614 F<termcap.pl>, F<tainted.pl>, F<stat.pl>, F<shellwords.pl>, F<pwd.pl>,
615 F<open3.pl>, F<open2.pl>, F<newgetopt.pl>, F<look.pl>, F<find.pl>,
616 F<finddepth.pl>, F<importenv.pl>, F<hostname.pl>, F<getopts.pl>,
617 F<getopt.pl>, F<getcwd.pl>, F<flush.pl>, F<fastcwd.pl>, F<exceptions.pl>,
618 F<ctime.pl>, F<complete.pl>, F<cacheout.pl>, F<bigrat.pl>, F<bigint.pl>,
619 F<bigfloat.pl>, F<assert.pl>, F<abbrev.pl>, F<dotsh.pl>, and
620 F<timelocal.pl> are all now deprecated. Using them will incur a warning.
625 =head1 Unicode overhaul
627 Perl's developers have made a concerted effort to update Perl to be in
628 sync with the latest Unicode standard. Changes for this include:
630 Perl can now handle every Unicode character property. New documentation,
631 L<perluniprops>, lists all available non-Unihan character properties. By
632 default, perl does not expose Unihan, deprecated or Unicode-internal
633 properties. See below for more details on these; there is also a section
634 in the pod listing them, and explaining why they are not exposed.
636 Perl now fully supports the Unicode compound-style of using C<=> and C<:>
637 in writing regular expressions: C<\p{property=value}> and
638 C<\p{property:value}> (both of which mean the same thing).
640 Perl now fully supports the Unicode loose matching rules for text
641 between the braces in C<\p{...}> constructs. In addition, Perl allows
642 underscores between digits of numbers.
644 Perl now accepts all the Unicode-defined synonyms for properties and property values.
646 C<qr/\X/>, which matches a Unicode logical character, has been expanded to work
647 better with various Asian languages. It now is defined as an I<extended
648 grapheme cluster>. (See L<http://www.unicode.org/reports/tr29/>).
649 Anything matched previously and that made sense will continue to be
650 accepted. Additionally:
656 C<\X> will not break apart a C<S<CR LF>> sequence.
660 C<\X> will now match a sequence which includes the C<ZWJ> and C<ZWNJ> characters.
664 C<\X> will now always match at least one character, including an initial mark.
665 Marks generally come after a base character, but it is possible in Unicode to
666 have them in isolation, and C<\X> will now handle that case, for example at the
667 beginning of a line, or after a C<ZWSP>. And this is the part where C<\X>
668 doesn't match the things that it used to that don't make sense. Formerly, for
669 example, you could have the nonsensical case of an accented LF.
673 C<\X> will now match a (Korean) Hangul syllable sequence, and the Thai and Lao
678 Otherwise, this change should be transparent for the non-affected languages.
680 C<\p{...}> matches using the Canonical_Combining_Class property were
681 completely broken in previous releases of Perl. They should now work correctly.
683 Before Perl 5.12, the Unicode C<Decomposition_Type=Compat> property and a
684 Perl extension had the same name, which led to neither matching all the
685 correct values (with more than 100 mistakes in one, and several thousand
686 in the other). The Perl extension has now been renamed to be
687 C<Decomposition_Type=Noncanonical> (short: C<dt=noncanon>). It has the same
688 meaning as was previously intended, namely the union of all the
689 non-canonical Decomposition types, with Unicode C<Compat> being just one of
692 C<\p{Decomposition_Type=Canonical}> now includes the Hangul syllables.
694 C<\p{Uppercase}> and C<\p{Lowercase}> now work as the Unicode standard says they should.
695 This means they each match a few more characters than they used to.
697 C<\p{Cntrl}> now matches the same characters as C<\p{Control}>. This means it
698 no longer will match Private Use (gc=co), Surrogates (gc=cs), nor Format
699 (gc=cf) code points. The Format code points represent the biggest
700 possible problem. All but 36 of them are either officially deprecated
701 or strongly discouraged from being used. Of those 36, likely the most
702 widely used are the soft hyphen (U+00AD), and BOM, ZWSP, ZWNJ, WJ, and
703 similar characters, plus bidirectional controls.
705 C<\p{Alpha}> now matches the same characters as C<\p{Alphabetic}>. Before 5.12, Perl's definition
706 definition included a number of things that aren't really alpha (all
707 marks) while omitting many that were. The
708 definitions of C<\p{Alnum}> and C<\p{Word}> depend on Alpha's definition and have changed accordingly.
710 C<\p{Word}> no longer incorrectly matches non-word characters such as fractions.
712 C<\p{Print}> no longer matches the line control characters: Tab, LF, CR,
713 FF, VT, and NEL. This brings it in line with standards and the documentation.
715 C<\p{XDigit}> now matches the same characters as C<\p{Hex_Digit}>. This
716 means that in addition to the characters it currently matches,
717 C<[A-Fa-f0-9]>, it will also match the 22 fullwidth equivalents, for
718 example U+FF10: FULLWIDTH DIGIT ZERO.
720 The Numeric type property has been extended to include the Unihan
723 There is a new Perl extension, the 'Present_In', or simply 'In',
724 property. This is an extension of the Unicode Age property, but
725 C<\p{In=5.0}> matches any code point whose usage has been determined
726 I<as of> Unicode version 5.0. The C<\p{Age=5.0}> only matches code points
727 added in I<precisely> version 5.0.
729 A number of properties now have the correct values for unassigned
730 code points. The affected properties are
731 Bidi_Class, East_Asian_Width, Joining_Type, Decomposition_Type,
732 Hangul_Syllable_Type, Numeric_Type, and Line_Break.
734 The Default_Ignorable_Code_Point, ID_Continue, and ID_Start properties
735 are now up to date with current Unicode definitions.
737 Earlier versions of Perl erroneously exposed certain properties that are supposed to be Unicode internal-only.
738 Use of these in regular expressions will now generate, if enabled, a deprecation warning message.
739 The properties are: Other_Alphabetic, Other_Default_Ignorable_Code_Point,
740 Other_Grapheme_Extend, Other_ID_Continue, Other_ID_Start, Other_Lowercase,
741 Other_Math, and Other_Uppercase.
743 It is now possible to change which Unicode properties Perl understands
744 on a per-installation basis. As mentioned above, certain properties
745 are turned off by default. These include all the Unihan properties
746 (which should be accessible via the CPAN module Unicode::Unihan) and any
747 deprecated or Unicode internal-only property that Perl has never exposed.
749 The generated files in the C<lib/unicore/To> directory are now more
750 clearly marked as being stable, directly usable by applications.
751 New hash entries in them give the format of the normal entries,
752 which allows for easier machine parsing. Perl can generate files
753 in this directory for any property, though most are suppressed.
754 You can find instructions for changing which are written in L<perluniprops>.
758 =head1 Modules and Pragmata
760 =head2 Notable new Modules and Pragmata
766 C<autodie> is a new lexically-scoped alternative for the C<Fatal> module.
767 The bundled version is 2.06_01. Note that in this release, using a string
768 eval when C<autodie> is in effect can cause the autodie behaviour to leak
769 into the surrounding scope. See L<autodie/"BUGS"> for more details.
773 C<Compress::Raw::Bzip2>
777 C<overloading> allows you to lexically disable or enable overloading
778 for some or all operations. (Yuval Kogman)
782 C<parent> establishes an ISA relationship with base classes at compile
783 time. It provides the key feature of C<base> without further unwanted
792 =head2 Notable changes to Modules and Pragmata
798 C<charnames> now contains the Unicode F<NameAliases.txt> database file.
799 This has the effect of adding some extra C<\N> character names that
800 formerly wouldn't have been recognised; for example, C<"\N{LATIN CAPITAL
805 In C<feature>, the meaning of the C<:5.10> and C<:5.10.X> feature bundles has
806 changed slightly. The last component, if any (i.e. C<X>) is simply ignored.
807 This is predicated on the assumption that new features will not, in
808 general, be added to maintenance releases. So C<:5.10> and C<:5.10.X>
809 have identical effect. This is a change to the behaviour documented for
812 C<feature> now includes the C<unicode_strings> feature:
814 use feature "unicode_strings";
816 This pragma turns on Unicode semantics for the case-changing operations
817 (C<uc>, C<lc>, C<ucfirst>, C<lcfirst>) on strings that don't have the
818 internal UTF-8 flag set, but that contain single-byte characters between
823 C<mro> is now implemented as an XS extension. The documented interface has not
824 changed. Code relying on the implementation detail that some C<mro::>
825 methods happened to be available at all times gets to "keep both pieces".
829 C<diagnostics> now supports %.0f formatting internally.
833 C<overload> now allow overloading of 'qr'.
837 C<diagnostics> no longer suppresses C<Use of uninitialized value in range
838 (or flip)> warnings. [perl #71204]
840 C<less> now includes the C<stash_name> method to allow subclasses of
841 C<less> to pick where in %^H to store their stash.
845 C<version> now has support for L</Version number formats> as described earlier
846 in this document and in its own documentation.
850 C<warnings> has a new C<warnings::fatal_enabled()> function. It also includes a new C<illegalproto> warning category. See also L</New or Changed Diagnostics> for this change.
855 =head2 Removed Modules and Pragmata
863 The C<attrs> pragma has been removed. It had been marked as deprecated
868 C<Devel::DProf::V> is no longer part of the Perl core.
872 =head2 Deprecated Modules and Pragmata
874 See L</Deprecated Modules> above.
878 =head2 New Documentation
884 L<perlhaiku> contains instructions on how to build perl for the Haiku platform.
888 L<perlmroapi> describes the new interface for pluggable Method Resolution Orders.
892 L<perlperf>, by Richard Foley, provides an introduction to the use of
893 performance and optimization techniques which can be used with particular
894 reference to perl programs.
898 L<perlrepository> describes how to access the perl source using the I<git> version
903 L<perlpolicy> extends the "Social contract about contributed modules" into
904 the beginnings of a document on Perl porting policies.
908 =head2 Changes to Existing Documentation
916 The various large F<Changes*> files (which listed every change made to perl
917 over the last 18 years) have been removed, and replaced by a small file,
918 also called F<Changes>, which just explains how that same information may
919 be extracted from the git version control system.
923 F<Porting/patching.pod> has been deleted, as it mainly described
924 interacting with the old Perforce-based repository, which is now obsolete.
925 Information still relevant has been moved to L<perlrepository>.
930 The syntax C<unless (EXPR) BLOCK else BLOCK> is now documented as valid, as
931 is the syntax C<unless (EXPR) BLOCK elsif (EXPR) BLOCK ... else BLOCK>,
932 although actually using the latter may not be the best idea for the
933 readability of your source code.
938 Documented -X overloading.
942 Documented that C<when()> treats specially most of the filetest operators
946 Documented C<when> as a syntax modifier.
950 Eliminated "Old Perl threads tutorial", which described 5005 threads.
952 F<pod/perlthrtut.pod> is the same material reworked for ithreads.
956 Correct previous documentation: v-strings are not deprecated
958 With version objects, we need them to use MODULE VERSION syntax. This
959 patch removes the deprecation notice.
963 Security contact information is now part of L<perlsec>.
967 A significant fraction of the core documentation has been updated to clarify
968 the behavior of Perl's Unicode handling.
970 Much of the remaining core documentation has been reviewed and edited
971 for clarity, consistent use of language, and to fix the spelling of Tom
976 The Pod specification (L<perlpodspec>) has been updated to bring the
977 specification in line with modern usage already supported by most Pod
978 systems. A parameter string may now follow the format name in a
979 "begin/end" region. Links to URIs with a text description are now
980 allowed. The usage of C<LE<lt>"section"E<gt>> has been marked as
985 L<if.pm|if> has been documented in L<perlfunc/use> as a means to get
986 conditional loading of modules despite the implicit BEGIN block around
991 The documentation for C<$1> in perlvar.pod has been clarified.
995 C<\N{U+I<wide hex char>}> is now documented.
999 =head1 Selected Performance Enhancements
1005 A new internal cache means that C<isa()> will often be faster.
1009 The implementation of C<C3> Method Resolution Order has been optimised -
1010 linearisation for classes with single inheritance is 40% faster. Performance
1011 for multiple inheritance is unchanged.
1015 Under C<use locale>, the locale-relevant information is now cached on
1016 read-only values, such as the list returned by C<keys %hash>. This makes
1017 operations such as C<sort keys %hash> in the scope of C<use locale> much
1022 Empty C<DESTROY> methods are no longer called.
1026 C<Perl_sv_utf8_upgrade()> is now faster.
1030 C<keys> on empty hash is now faster.
1034 C<if (%foo)> has been optimized to be faster than C<if (keys %foo)>.
1038 Reversing an array to itself (as in C<@a = reverse @a>) in void context
1039 now happens in-place and is several orders of magnitude faster than it
1040 used to be. It will also preserve non-existent elements whenever
1041 possible, i.e. for non magical arrays or tied arrays with C<EXISTS> and
1046 =head1 Installation and Configuration Improvements
1052 L<perlapi>, L<perlintern>, L<perlmodlib> and L<perltoc> are now all
1053 generated at build time, rather than being shipped as part of the release.
1057 If C<vendorlib> and C<vendorarch> are the same, then they are only added to
1062 C<$Config{usedevel}> and the C-level C<PERL_USE_DEVEL> are now defined if
1063 perl is built with C<-Dusedevel>.
1067 F<Configure> will enable use of C<-fstack-protector>, to provide protection
1068 against stack-smashing attacks, if the compiler supports it.
1072 F<Configure> will now determine the correct prototypes for re-entrant
1073 functions and for C<gconvert> if you are using a C++ compiler rather
1078 On Unix, if you build from a tree containing a git repository, the
1079 configuration process will note the commit hash you have checked out, for
1080 display in the output of C<perl -v> and C<perl -V>. Unpushed local commits
1081 are automatically added to the list of local patches displayed by
1086 Perl now supports SystemTap's C<dtrace> compatibility layer and an
1087 issue with linking C<miniperl> has been fixed in the process.
1091 perldoc now uses C<less -R> instead of C<less> for improved behaviour
1092 in the face of C<groff>'s new usage of ANSI escape codes.
1097 C<perl -V> now reports use of the compile-time options C<USE_PERL_ATOF> and
1098 C<USE_ATTRIBUTES_FOR_PERLIO>.
1102 As part of the flattening of F<ext>, all extensions on all platforms are
1103 built by F<make_ext.pl>. This replaces the Unix-specific
1104 F<ext/util/make_ext>, VMS-specific F<make_ext.com> and Win32-specific
1105 F<win32/buildext.pl>.
1109 =head1 Internal Changes
1111 Each release of Perl sees numerous internal changes which shouldn't
1112 affect day to day usage but may still be notable for developers working
1113 with Perl's source code.
1119 The J.R.R. Tolkien quotes at the head of C source file have been checked and
1120 proper citations added, thanks to a patch from Tom Christiansen.
1124 The internal structure of the dual-life modules traditionally found in
1125 the F<lib/> and F<ext/> directories y in the perl source has changed
1126 significantly. Where possible, dual-lifed modules have been extracted
1127 from F<lib/> and F<ext/>.
1129 Dual-lifed modules maintained by Perl's developers as part of the Perl
1130 core now live in F<dist/>. Dual-lifed modules maintained primarily on
1131 CPAN now live in F<cpan/>. When reporting a bug in a module located
1132 under F<cpan/>, please send your bug report directly to the module's
1133 bug tracker or author, rather than Perl's bug tracker.
1137 C<\N{...}> now compiles better, always forces UTF-8 internal representation
1139 Perl's developers have fixed several problems with the recognition of C<\N{...}>
1140 constructs. As part of this, perl will store any scalar or regex containing
1141 C<\N{I<name>}> or C<\N{U+I<wide hex char>}> in its definition in
1142 UTF-8 format. (This was true previously for all occurences of C<\N{I<name>}>
1143 that did not use a custom translator, but now it's always true.)
1147 Perl_magic_setmglob now knows about globs, fixing RT #71254.
1151 C<SVt_RV> no longer exists. RVs are now stored in IVs.
1155 REGEXPs are now first class.
1159 C<Perl_vcroak()> now accepts a null first argument. In addition, a full audit
1160 was made of the "not NULL" compiler annotations, and those for several
1161 other internal functions were corrected.
1165 New macros C<dSAVEDERRNO>, C<dSAVE_ERRNO>, C<SAVE_ERRNO>, C<RESTORE_ERRNO>
1166 have been added to formalise the temporary saving of the C<errno>
1171 The function C<Perl_sv_insert_flags> has been added to augment
1176 The function C<Perl_newSV_type(type)> has been added, equivalent to
1177 C<Perl_newSV()> followed by C<Perl_sv_upgrade(type)>.
1181 The function C<Perl_newSVpvn_flags()> has been added, equivalent to
1182 C<Perl_newSVpvn()> and then performing the action relevant to the flag.
1184 Two flag bits are currently supported.
1190 C<SVf_UTF8> will call C<SvUTF8_on()> for you. (Note that this does not convert an
1191 sequence of ISO 8859-1 characters to UTF-8). A wrapper, C<newSVpvn_utf8()>
1192 is available for this.
1196 C<SVs_TEMP> now calls C<Perl_sv_2mortal()> on the new SV.
1200 There is also a wrapper that takes constant strings, C<newSVpvs_flags()>.
1204 The function C<Perl_croak_xs_usage> has been added as a wrapper to
1209 Perl now exports the functions C<PerlIO_find_layer> and C<PerlIO_list_alloc>.
1213 C<PL_na> has been exterminated from the core code, replaced by local STRLEN
1214 temporaries, or C<*_nolen()> calls. Either approach is faster than C<PL_na>,
1215 which is a pointer dereference into the interpreter structure under ithreads,
1216 and a global variable otherwise.
1220 C<Perl_mg_free()> used to leave freed memory accessible via C<SvMAGIC()> on
1221 the scalar. It now updates the linked list to remove each piece of magic
1226 Under ithreads, the regex in C<PL_reg_curpm> is now reference counted. This
1227 eliminates a lot of hackish workarounds to cope with it not being reference
1232 C<Perl_mg_magical()> would sometimes incorrectly turn on C<SvRMAGICAL()>.
1233 This has been fixed.
1237 The I<public> IV and NV flags are now not set if the string value has
1238 trailing "garbage". This behaviour is consistent with not setting the
1239 public IV or NV flags if the value is out of range for the type.
1243 Uses of C<Nullav>, C<Nullcv>, C<Nullhv>, C<Nullop>, C<Nullsv> etc have been
1244 replaced by C<NULL> in the core code, and non-dual-life modules, as C<NULL>
1245 is clearer to those unfamiliar with the core code.
1249 A macro C<MUTABLE_PTR(p)> has been added, which on (non-pedantic) gcc will
1250 not cast away C<const>, returning a C<void *>. Macros C<MUTABLE_SV(av)>,
1251 C<MUTABLE_SV(cv)> etc build on this, casting to C<AV *> etc without
1252 casting away C<const>. This allows proper compile-time auditing of
1253 C<const> correctness in the core, and helped picked up some errors (now
1258 Macros C<mPUSHs()> and C<mXPUSHs()> have been added, for pushing SVs on the
1259 stack and mortalizing them.
1263 Use of the private structure C<mro_meta> has changed slightly. Nothing
1264 outside the core should be accessing this directly anyway.
1268 A new tool, F<Porting/expand-macro.pl> has been added, that allows you
1269 to view how a C preprocessor macro would be expanded when compiled.
1270 This is handy when trying to decode the macro hell that is the perl
1277 =head2 Testing improvements
1281 =item Parallel tests
1283 The core distribution can now run its regression tests in parallel on
1284 Unix-like platforms. Instead of running C<make test>, set C<TEST_JOBS> in
1285 your environment to the number of tests to run in parallel, and run
1286 C<make test_harness>. On a Bourne-like shell, this can be done as
1288 TEST_JOBS=3 make test_harness # Run 3 tests in parallel
1290 An environment variable is used, rather than parallel make itself, because
1291 L<TAP::Harness> needs to be able to schedule individual non-conflicting test
1292 scripts itself, and there is no standard interface to C<make> utilities to
1293 interact with their job schedulers.
1295 Note that currently some test scripts may fail when run in parallel (most
1296 notably C<ext/IO/t/io_dir.t>). If necessary run just the failing scripts
1297 again sequentially and see if the failures go away.
1299 =item Test harness flexibility
1301 It's now possible to override C<PERL5OPT> and friends in F<t/TEST>
1305 Several tests that have the potential to hang forever if they fail now
1306 incorporate a "watchdog" functionality that will kill them after a timeout,
1307 which helps ensure that C<make test> and C<make test_harness> run to
1308 completion automatically. (Jerry Hedden).
1315 Perl's developers have added a number of new tests to the core.
1316 In addition to the items listed below, many modules updated from CPAN
1317 incorporate new tests.
1323 Significant cleanups to core tests to ensure that language and
1324 interpreter features are not used before they're tested.
1328 C<make test_porting> now runs a number of important pre-commit checks
1329 which might be of use to anyone working on the Perl core.
1333 F<t/porting/podcheck.t> automatically checks the well-formedness of
1334 POD found in all .pl, .pm and .pod files in the F<MANIFEST>, other than in
1335 dual-lifed modules which are primarily maintained outside the Perl core.
1339 F<t/porting/manifest.t> now tests that all files listed in MANIFEST are present.
1343 F<t/op/while_readdir.t> tests that a bare readdir in while loop sets $_.
1347 F<t/comp/retainedlines.t> checks that the debugger can retain source lines from C<eval>.
1351 F<t/io/perlio_fail.t> checks that bad layers fail.
1355 F<t/io/perlio_leaks.t> checks that PerlIO layers are not leaking.
1359 F<t/io/perlio_open.t> checks that certain special forms of open work.
1363 F<t/io/perlio.t> includes general PerlIO tests.
1367 F<t/io/pvbm.t> checks that there is no unexpected interaction between the internal types
1368 C<PVBM> and C<PVGV>.
1372 F<t/mro/package_aliases.t> checks that mro works properly in the presence of aliased packages.
1376 F<t/op/dbm.t> tests C<dbmopen> and C<dbmclose>.
1380 F<t/op/index_thr.t> tests the interaction of C<index> and threads.
1384 F<t/op/pat_thr.t> tests the interaction of esoteric patterns and threads.
1388 F<t/op/qr_gc.t> tests that C<qr> doesn't leak.
1392 F<t/op/reg_email_thr.t> tests the interaction of regex recursion and threads.
1396 F<t/op/regexp_qr_embed_thr.t> tests the interaction of patterns with embedded C<qr//> and threads.
1400 F<t/op/regexp_unicode_prop.t> tests Unicode properties in regular expressions.
1404 F<t/op/regexp_unicode_prop_thr.t> tests the interaction of Unicode properties and threads.
1408 F<t/op/reg_nc_tie.t> tests the tied methods of C<Tie::Hash::NamedCapture>.
1412 F<t/op/reg_posixcc.t> checks that POSIX character classes behave consistently.
1418 checks that exportable C<re> functions in F<universal.c> work.
1422 F<t/op/setpgrpstack.t> checks that C<setpgrp> works.
1426 F<t/op/substr_thr.t> tests the interaction of C<substr> and threads.
1430 F<t/op/upgrade.t> checks that upgrading and assigning scalars works.
1434 F<t/uni/lex_utf8.t> checks that Unicode in the lexer works.
1438 F<t/uni/tie.t> checks that Unicode and C<tie> work.
1442 F<t/comp/final_line_num.t> tests whether line numbers are correct at EOF
1446 F<t/comp/form_scope.t> tests format scoping.
1450 F<t/comp/line_debug.t> tests whether C<< @{"_<$file"} >> works.
1454 F<t/op/filetest_t.t> tests if -t file test works.
1458 F<t/op/qr.t> tests C<qr>.
1462 F<t/op/utf8cache.t> tests malfunctions of the utf8 cache.
1466 F<t/re/uniprops.t> test unicodes C<\p{}> regex constructs.
1470 F<t/op/filehandle.t> tests some suitably portable filetest operators
1471 to check that they work as expected, particularly in the light of some
1472 internal changes made in how filehandles are blessed.
1476 F<t/op/time_loop.t> tests that unix times greater than C<2**63>, which
1477 can now be handed to C<gmtime> and C<localtime>, do not cause an internal
1478 overflow or an excessively long loop.
1483 =head1 New or Changed Diagnostics
1485 =head2 New Diagnostics
1491 SV allocation tracing has been added to the diagnostics enabled by C<-Dm>.
1492 The tracing can alternatively output via the C<PERL_MEM_LOG> mechanism, if
1493 that was enabled when the F<perl> binary was compiled.
1497 Smartmatch resolution tracing has been added as a new diagnostic. Use C<-DM> to
1502 A new debugging flag C<-DB> now dumps subroutine definitions, leaving
1503 C<-Dx> for its original purpose of dumping syntax trees.
1507 Perl 5.12 provides a number of new diagnostic messages to help you write
1508 better code. See L<perldiag> for details of these new messages.
1514 C<Bad plugin affecting keyword '%s'>
1518 C<gmtime(%.0f) too large>
1522 C<Lexing code attempted to stuff non-Latin-1 character into Latin-1 input>
1526 C<Lexing code internal error (%s)>
1530 C<localtime(%.0f) too large>
1534 C<Overloaded dereference did not return a reference>
1538 C<Overloaded qr did not return a REGEXP>
1542 C<Perl_pmflag() is deprecated, and will be removed from the XS API>
1546 C<lvalue attribute ignored after the subroutine has been defined>
1548 This new warning is issued when one attempts to mark a subroutine as
1549 lvalue after it has been defined.
1553 Perl now warns you if C<++> or C<--> are unable to change the value because it's
1554 beyond the limit of representation.
1556 This uses a new warnings category: "imprecision".
1560 C<lc>, C<uc>, C<lcfirst>, and C<ucfirst> warn when passed undef.
1564 C<Show constant in "Useless use of a constant in void context">
1568 C<Prototype after '%s'>
1572 C<panic: sv_chop %s>
1574 This new fatal error occurs when the C routine C<Perl_sv_chop()> was
1575 passed a position that is not within the scalar's string buffer. This
1576 could be caused by buggy XS code, and at this point recovery is not
1582 The fatal error C<Malformed UTF-8 returned by \N> is now produced if the
1583 C<charnames> handler returns malformed UTF-8.
1587 If an unresolved named character or sequence was encountered when compiling a
1588 regex pattern then the fatal error C<\\N{NAME} must be resolved by the lexer>
1589 is now produced. This can happen, for example, when using a single-quotish
1590 context like C<$re = '\N{SPACE}'; $re;>. See L<perldiag> for more examples of
1591 how the lexer can get bypassed.
1595 C<Invalid hexadecimal number in \\N{U+...}> is a new fatal error triggered when
1596 the character constant represented by C<...> is not a valid hexadecimal
1601 The new meaning of C<\N> as C<[^\n]> is not valid in a bracketed character
1602 class, just like C<.> in a character class loses its special meaning, and will
1603 cause the fatal error C<\\N in a character class must be a named character: \\N{...}>.
1607 The rules on what is legal for the C<...> in C<\N{...}> have been tightened
1608 up so that unless the C<...> begins with an alphabetic character and continues
1609 with a combination of alphanumerics, dashes, spaces, parentheses or colons
1610 then the warning C<Deprecated character(s) in \\N{...} starting at '%s'> is
1615 The warning C<Using just the first characters returned by \N{}> will be
1616 issued if the C<charnames> handler returns a sequence of characters which
1617 exceeds the limit of the number of characters that can be used. The message
1618 will indicate which characters were used and which were discarded.
1624 =head2 Changed Diagnostics
1626 A number of existing diagnostic messages have been improved or corrected:
1632 A new warning category C<illegalproto> allows finer-grained control of
1633 warnings around function prototypes.
1639 =item C<Illegal character in prototype for %s : %s>
1641 =item C<Prototype after '%c' for %s : %s>
1645 have been moved from the C<syntax> top-level warnings category into a new
1646 first-level category, C<illegalproto>. These two warnings are currently the
1647 only ones emitted during parsing of an invalid/illegal prototype, so one
1650 no warnings 'illegalproto';
1652 to suppress only those, but not other syntax-related warnings. Warnings where
1653 prototypes are changed, ignored, or not met are still in the C<prototype>
1654 category as before. (Matt S. Trout)
1658 C<Deep recursion on subroutine "%s">
1660 It is now possible to change the depth threshold for this warning from the
1661 default of 100, by recompiling the F<perl> binary, setting the C
1662 pre-processor macro C<PERL_SUB_DEPTH_WARN> to the desired value.
1666 C<Illegal character in prototype> warning is now more precise
1667 when reporting illegal characters after _
1671 mro merging error messages are now very similar to those produced by L<Algorithm::C3>.
1675 Amelioration of the error message "Unrecognized character %s in column %d"
1677 Changes the error message to "Unrecognized character %s; marked by E<lt>--
1678 HERE after %sE<lt>-- HERE near column %d". This should make it a little
1679 simpler to spot and correct the suspicious character.
1683 Perl now explicitly points to C<$.> when it causes an uninitialized warning for
1684 ranges in scalar context.
1688 C<split> now warns when called in void context.
1692 C<printf>-style functions called with too few arguments will now issue the
1693 warning C<"Missing argument in %s"> [perl #71000]
1697 Perl now properly returns a syntax error instead of segfaulting
1698 if C<each>, C<keys>, or C<values> is used without an argument.
1702 C<tell()> now fails properly if called without an argument and when no
1703 previous file was read.
1705 C<tell()> now returns C<-1>, and sets errno to C<EBADF>, thus restoring
1706 the 5.8.x behaviour.
1710 C<overload> no longer implicitly unsets fallback on repeated 'use
1715 POSIX::strftime() can now handle Unicode characters in the format string.
1719 The Windows select() implementation now supports all empty C<fd_set>s
1724 The C<syntax> category was removed from 5 warnings that should only be in
1729 Three fatal C<pack>/C<unpack> error messages have been normalized to
1734 C<Unicode character is illegal> has been rephrased to be more accurate
1736 It now reads C<Unicode non-character is illegal in interchange> and the
1737 perldiag documentation has been expanded a bit.
1741 Currently, all but the first of the several characters that the C<charnames>
1742 handler may return are discarded when used in a regular expression pattern
1743 bracketed character class. If this happens then the warning C<Using just the
1744 first character returned by \N{} in character class> will be issued.
1748 The warning C<Missing right brace on \\N{} or unescaped left brace after \\N.
1749 Assuming the latter> will be issued if Perl encounters a C<\N{> but doesn't
1750 find a matching C<}>. In this case Perl doesn't know if it was mistakenly
1751 omitted, or if "match non-newline" followed by "match a C<{>" was desired.
1752 It assumes the latter because that is actually a valid interpretation as
1753 written, unlike the other case. If you meant the former, you need to add the
1754 matching right brace. If you did mean the latter, you can silence this
1755 warning by writing instead C<\N\{>.
1759 C<gmtime> and C<localtime> called with numbers smaller than they can reliably
1760 handle will now issue the warnings C<gmtime(%.0f) too small> and
1761 C<localtime(%.0f) too small>.
1765 The following diagnostic messages have been removed:
1775 C<Can't locate package %s for the parents of %s>
1777 In general this warning it only got produced in
1778 conjunction with other warnings, and removing it allowed an ISA lookup
1779 optimisation to be added.
1783 C<v-string in use/require is non-portable>
1787 =head1 Utility Changes
1793 F<h2ph> now looks in C<include-fixed> too, which is a recent addition to gcc's
1798 F<h2xs> no longer incorrectly treats enum values like macros (Daniel Burr).
1799 It also now handles C++ style constants (C<//>) properly in enums. (A patch from
1800 Rainer Weikusat was used; Daniel Burr also proposed a similar fix).
1804 F<perl5db.pl> now supports C<LVALUE> subroutines. Additionally, the debugger now correctly handles proxy constant subroutines, and subroutine stubs.
1808 F<perlbug> now uses C<%Module::CoreList::bug_tracker> to print out
1809 upstream bug tracker URLs. If a user identifies a particular module
1810 as the topic of their bug report and we're able to divine ithe URL for
1811 its upstream bug tracker, perlbug now provide a message to the user
1812 explaining that the core copies the CPAN version directly, and provide
1813 the URL for reporting the bug directly to the upstream author.
1815 F<perlbug> no longer reports "Message sent" when it hasn't actually sent the message
1819 F<perlthanks> is a new utility for sending non-bug-reports to the
1820 authors and maintainers of Perl. Getting nothing but bug reports can
1821 become a bit demoralising. If Perl 5.12 works well for you, please try
1822 out F<perlthanks>. It will make the developers smile.
1826 Perl's developers have fixed bugs in F<a2p> having to do with the
1827 C<match()> operator in list context.
1831 =head1 Selected Bug Fixes
1837 U+0FFFF is now a legal character in regular expressions.
1841 pp_qr now always returns a new regexp SV. Resolves RT #69852.
1843 Instead of returning a(nother) reference to the (pre-compiled) regexp in the
1844 optree, use reg_temp_copy() to create a copy of it, and return a reference to
1845 that. This resolves issues about Regexp::DESTROY not being called in a timely
1846 fashion (the original bug tracked by RT #69852), as well as bugs related to
1847 blessing regexps, and of assigning to regexps, as described in correspondence
1848 added to the ticket.
1850 It transpires that we also need to undo the SvPVX() sharing when ithreads
1851 cloning a Regexp SV, because mother_re is set to NULL, instead of a cloned
1852 copy of the mother_re. This change might fix bugs with regexps and threads in
1853 certain other situations, but as yet neither tests nor bug reports have
1854 indicated any problems, so it might not actually be an edge case that it's
1859 Several compilation errors and segfaults when perl was built with C<-Dmad> were fixed.
1863 Fixes for lexer API changes in 5.11.2 which broke NYTProf's savesrc option.
1867 C<-t> should only return TRUE for file handles connected to a TTY
1869 The Microsoft C version of C<isatty()> returns TRUE for all
1870 character mode devices, including the F</dev/null>-style "nul"
1871 device and printers like "lpt1".
1875 Fixed a regression caused by commit fafafbaf which caused a panic during
1876 parameter passing [perl #70171]
1880 On systems which in-place edits without backup files, -i'*' now works as
1881 the documentation says it does [perl #70802]
1885 Saving and restoring magic flags no longer loses readonly flag.
1889 The malformed syntax C<grep EXPR LIST> (note the missing comma) no longer
1890 causes abrupt and total failure.
1894 Regular expressions compiled with C<qr{}> literals properly set C<$'> when
1899 Using named subroutines with C<sort> should no longer lead to bus errors [perl
1904 Numerous bugfixes catch small issues caused by the recently-added Lexer API.
1908 Smart match against C<@_> sometimes gave false negatives. [perl #71078]
1912 C<$@> may now be assigned a read-only value (without error or busting
1917 C<sort> called recursively from within an active comparison subroutine no
1918 longer causes a bus error if run multiple times. [perl #71076]
1922 Tie::Hash::NamedCapture::* will not abort if passed bad input (RT #71828)
1926 @_ and $_ no longer leak under threads (RT #34342 and #41138, also
1931 C<-I> on shebang line now adds directories in front of @INC
1932 as documented, and as does C<-I> when specified on the command-line.
1936 C<kill> is now fatal when called on non-numeric process identifiers.
1937 Previously, an C<undef> process identifier would be interpreted as a
1938 request to kill process 0, which would terminate the current process
1939 group on POSIX systems. Since process identifiers are always integers,
1940 killing a non-numeric process is now fatal.
1944 5.10.0 inadvertently disabled an optimisation, which caused a measurable
1945 performance drop in list assignment, such as is often used to assign
1946 function parameters from C<@_>. The optimisation has been re-instated, and
1947 the performance regression fixed. (This fix is also present in 5.10.1)
1951 Fixed memory leak on C<while (1) { map 1, 1 }> [RT #53038].
1955 Some potential coredumps in PerlIO fixed [RT #57322,54828].
1959 The debugger now works with lvalue subroutines.
1963 The debugger's C<m> command was broken on modules that defined constants
1968 C<crypt> and string complement could return tainted values for untainted
1969 arguments [RT #59998].
1973 The C<-i>I<.suffix> command-line switch now recreates the file using
1974 restricted permissions, before changing its mode to match the original
1975 file. This eliminates a potential race condition [RT #60904].
1979 On some Unix systems, the value in C<$?> would not have the top bit set
1980 (C<$? & 128>) even if the child core dumped.
1984 Under some circumstances, C<$^R> could incorrectly become undefined
1989 In the XS API, various hash functions, when passed a pre-computed hash where
1990 the key is UTF-8, might result in an incorrect lookup.
1994 XS code including F<XSUB.h> before F<perl.h> gave a compile-time error
1999 C<< $object-E<gt>isa('Foo') >> would report false if the package C<Foo> didn't
2000 exist, even if the object's C<@ISA> contained C<Foo>.
2004 Various bugs in the new-to 5.10.0 mro code, triggered by manipulating
2005 C<@ISA>, have been found and fixed.
2009 Bitwise operations on references could crash the interpreter, e.g.
2010 C<$x=\$y; $x |= "foo"> [RT #54956].
2014 Patterns including alternation might be sensitive to the internal UTF-8
2015 representation, e.g.
2017 my $byte = chr(192);
2018 my $utf8 = chr(192); utf8::upgrade($utf8);
2019 $utf8 =~ /$byte|X}/i; # failed in 5.10.0
2023 Within UTF8-encoded Perl source files (i.e. where C<use utf8> is in
2024 effect), double-quoted literal strings could be corrupted where a C<\xNN>,
2025 C<\0NNN> or C<\N{}> is followed by a literal character with ordinal value
2026 greater than 255 [RT #59908].
2030 C<B::Deparse> failed to correctly deparse various constructs:
2031 C<readpipe STRING> [RT #62428], C<CORE::require(STRING)> [RT #62488],
2032 C<sub foo(_)> [RT #62484].
2036 Using C<setpgrp> with no arguments could corrupt the perl stack.
2040 The block form of C<eval> is now specifically trappable by C<Safe> and
2041 C<ops>. Previously it was erroneously treated like string C<eval>.
2045 In 5.10.0, the two characters C<[~> were sometimes parsed as the smart
2046 match operator (C<~~>) [RT #63854].
2050 In 5.10.0, the C<*> quantifier in patterns was sometimes treated as
2051 C<{0,32767}> [RT #60034, #60464]. For example, this match would fail:
2053 ("ab" x 32768) =~ /^(ab)*$/
2057 C<shmget> was limited to a 32 bit segment size on a 64 bit OS [RT #63924].
2061 Using C<next> or C<last> to exit a C<given> block no longer produces a
2062 spurious warning like the following:
2064 Exiting given via last at foo.pl line 123
2068 On Windows, C<'.\foo'> and C<'..\foo'> were treated differently than
2069 C<'./foo'> and C<'../foo'> by C<do> and C<require> [RT #63492].
2073 Assigning a format to a glob could corrupt the format; e.g.:
2075 *bar=*foo{FORMAT}; # foo format now bad
2079 Attempting to coerce a typeglob to a string or number could cause an
2080 assertion failure. The correct error message is now generated,
2081 C<Can't coerce GLOB to I<$type>>.
2085 Under C<use filetest 'access'>, C<-x> was using the wrong access mode. This
2086 has been fixed [RT #49003].
2090 C<length> on a tied scalar that returned a Unicode value would not be
2091 correct the first time. This has been fixed.
2095 Using an array C<tie> inside in array C<tie> could SEGV. This has been
2100 A race condition inside C<PerlIOStdio_close()> has been identified and
2101 fixed. This used to cause various threading issues, including SEGVs.
2105 In C<unpack>, the use of C<()> groups in scalar context was internally
2106 placing a list on the interpreter's stack, which manifested in various
2107 ways, including SEGVs. This is now fixed [RT #50256].
2111 Magic was called twice in C<substr>, C<\&$x>, C<tie $x, $m> and C<chop>.
2112 These have all been fixed.
2116 A 5.10.0 optimisation to clear the temporary stack within the implicit
2117 loop of C<s///ge> has been reverted, as it turned out to be the cause of
2118 obscure bugs in seemingly unrelated parts of the interpreter [commit
2123 The line numbers for warnings inside C<elsif> are now correct.
2127 The C<..> operator now works correctly with ranges whose ends are at or
2128 close to the values of the smallest and largest integers.
2132 C<binmode STDIN, ':raw'> could lead to segmentation faults on some platforms.
2133 This has been fixed [RT #54828].
2137 An off-by-one error meant that C<index $str, ...> was effectively being
2138 executed as C<index "$str\0", ...>. This has been fixed [RT #53746].
2142 Various leaks associated with named captures in regexes have been fixed
2147 A weak reference to a hash would leak. This was affecting C<DBI>
2152 Using (?|) in a regex could cause a segfault [RT #59734].
2156 Use of a UTF-8 C<tr//> within a closure could cause a segfault [RT #61520].
2160 Calling C<Perl_sv_chop()> or otherwise upgrading an SV could result in an
2161 unaligned 64-bit access on the SPARC architecture [RT #60574].
2165 In the 5.10.0 release, C<inc_version_list> would incorrectly list
2166 C<5.10.*> after C<5.8.*>; this affected the C<@INC> search order
2171 In 5.10.0, C<pack "a*", $tainted_value> returned a non-tainted value
2176 In 5.10.0, C<printf> and C<sprintf> could produce the fatal error
2177 C<panic: utf8_mg_pos_cache_update> when printing UTF-8 strings
2182 In the 5.10.0 release, a dynamically created C<AUTOLOAD> method might be
2183 missed (method cache issue) [RT #60220,60232].
2187 In the 5.10.0 release, a combination of C<use feature> and C<//ee> could
2188 cause a memory leak [RT #63110].
2192 C<-C> on the shebang (C<#!>) line is once more permitted if it is also
2193 specified on the command line. C<-C> on the shebang line used to be a
2194 silent no-op I<if> it was not also on the command line, so perl 5.10.0
2195 disallowed it, which broke some scripts. Now perl checks whether it is
2196 also on the command line and only dies if it is not [RT #67880].
2200 In 5.10.0, certain types of re-entrant regular expression could crash,
2201 or cause the following assertion failure [RT #60508]:
2203 Assertion rx->sublen >= (s - rx->subbeg) + i failed
2207 Perl now includes previously missing files from the Unicode 5.1 Character Database.
2211 Perl now honors C<TMPDIR> when opening an anonymous temporary file.
2216 =head1 Platform Specific Changes
2218 Perl is incredibly portable. In general, if a platform has a C compiler,
2219 someone has ported Perl to it (or will soon). We're happy to announce
2220 that Perl 5.12 includes support for several new platforms. At the same
2221 time, it's time to bid farewell to some (very) old friends.
2223 =head2 New Platforms
2229 Perl's developers have merged patches from Haiku's maintainers. Perl should now
2234 Perl should now build on MirOS BSD.
2238 =head2 Discontinued Platforms
2251 =head2 Updated Platforms
2255 =item Darwin (Mac OS X)
2261 Skip testing the be_BY.CP1131 locale on Darwin 10 (Mac OS X 10.6),
2262 as it's still buggy.
2266 Correct infelicities in the regexp used to identify buggy locales
2267 on Darwin 8 and 9 (Mac OS X 10.4 and 10.5, respectively).
2277 Fix thread library selection [perl #69686]
2287 Initial support for mingw64 is now available.
2291 Various bits of Perl's build infrastructure are no longer converted to
2292 win32 line endings at release time. If this hurts you, please report the
2293 problem with the L<perlbug> program included with perl.
2297 Always add a manifest resource to C<perl.exe> to specify the C<trustInfo>
2298 settings for Windows Vista and later. Without this setting Windows
2299 will treat C<perl.exe> as a legacy application and apply various
2300 heuristics like redirecting access to protected file system areas
2301 (like the "Program Files" folder) to the users "VirtualStore"
2302 instead of generating a proper "permission denied" error.
2304 For VC8 and VC9 this manifest setting is automatically generated by
2305 the compiler/linker (together with the binding information for their
2306 respective runtime libraries); for all other compilers we need to
2307 embed the manifest resource explicitly in the external resource file.
2309 This change also requests the Microsoft Common-Controls version 6.0
2310 (themed controls introduced in Windows XP) via the dependency list
2311 in the assembly manifest. For VC8 and VC9 this is specified using the
2312 C</manifestdependency> linker commandline option instead.
2316 Improved message window handling means that C<alarm> and C<kill> messages
2317 will no longer be dropped under race conditions.
2327 Perl now suppoorts IPv6 on cygwin 1.7 and newer.
2337 C<-UDEBUGGING> is now the default on VMS.
2339 Like it has been everywhere else for ages and ages. Also make
2340 command-line selection of -UDEBUGGING and -DDEBUGGING work in
2341 configure.com; before the only way to turn it off was by saying
2342 no in answer to the interactive question.
2346 The default pipe buffer size on VMS has been updated to 8192 on 64-bit
2351 Reads from the in-memory temporary files of C<PerlIO::scalar> used to fail
2352 if C<$/> was set to a numeric reference (to indicate record-style reads).
2357 VMS now supports C<getgrgid>.
2361 Many improvements and cleanups have been made to the VMS file name handling
2362 and conversion code.
2366 Enabling the C<PERL_VMS_POSIX_EXIT> logical name now encodes a POSIX exit
2367 status in a VMS condition value for better interaction with GNV's bash
2368 shell and other utilities that depend on POSIX exit values. See
2369 L<perlvms/"$?"> for details.
2373 C<File::Copy> now detects Unix compatibility mode on VMS.
2379 Removed F<libbsd> for AIX 5L and 6.1. Only C<flock()> was used from F<libbsd>.
2381 Removed F<libgdbm> for AIX 5L and 6.1. The F<libgdbm> is delivered as an
2382 optional package with the AIX Toolbox. Unfortunately the 64 bit version
2385 Hints changes mean that AIX 4.2 should work again.
2389 On Cygwin we now strip the last number from the DLL. This has been the
2390 behaviour in the cygwin.com build for years. The hints files have been
2396 The hints files now identify the correct threading libraries on FreeBSD 7
2401 We now work around a bizarre preprocessor bug in the Irix 6.5 compiler:
2402 C<cc -E -> unfortunately goes into K&R mode, but C<cc -E file.c> doesn't.
2406 Hints now supports versions 5.*.
2410 Various changes from Stratus have been merged in.
2414 There is now support for Symbian S60 3.2 SDK and S60 5.0 SDK.
2419 =head1 Known Problems
2421 This is a list of some significant unfixed bugs, which are regressions
2422 from either 5.10.x or 5.8.x.
2428 C<List::Util::first> misbehaves in the presence of a lexical C<$_>
2429 (typically introduced by C<my $_> or implicitly by C<given>). The variable
2430 which gets set for each iteration is the package variable C<$_>, not the
2431 lexical C<$_> [RT #67694].
2433 A similar issue may occur in other modules that provide functions which
2434 take a block as their first argument, like
2436 foo { ... $_ ...} list
2440 Some regexes may run much more slowly when run in a child thread compared
2441 with the thread the pattern was compiled into [RT #55600].
2445 Things like C<"\N{LATIN SMALL LIGATURE FF}" =~ /\N{LATIN SMALL LETTER F}+/>
2446 will appear to hang as they get into a very long running loop [RT #72998].
2450 Several porters have reported mysterious crashes when Perl's entire test suite is run after a build on certain Windows 2000 systems. When run by hand, the individual tests reportedly work fine.
2460 This one is actually a change introduced in 5.10.0, but it was missed
2461 from that release's perldelta, so it is mentioned here instead.
2463 A bugfix related to the handling of the C</m> modifier and C<qr> resulted
2464 in a change of behaviour between 5.8.x and 5.10.0:
2466 # matches in 5.8.x, doesn't match in 5.10.0
2467 $re = qr/^bar/; "foo\nbar" =~ /$re/m;
2471 =head1 Acknowledgements
2473 Perl 5.12.0 represents approximately two years of development since
2474 Perl 5.10.0 and contains over 750,000 lines of changes across over
2475 3000 files from over 200 authors and committers.
2477 Perl continues to flourish into its third decade thanks to a vibrant community of users and developers. The following people are known to have contributed the improvements that became Perl 5.12.0:
2479 Aaron Crane, Abe Timmerman, Abhijit Menon-Sen, Abigail, Adam Russell,
2480 Adriano Ferreira, Ævar Arnfjörð Bjarmason, Alan Grover, Alexandr
2481 Ciornii, Alex Davies, Alex Vandiver, Andreas Koenig, Andrew Rodland,
2482 andrew@sundale.net, Andy Armstrong, Andy Dougherty, Jose AUGUSTE-ETIENNE,
2483 Benjamin Smith, Ben Morrow, bharanee rathna, Bo Borgerson, Bo Lindbergh,
2484 Brad Gilbert, Bram, Brendan O'Dea, brian d foy, Charles Bailey,
2485 Chip Salzenberg, Chris 'BinGOs' Williams, Christoph Lamprecht, Chris
2486 Williams, chromatic, Claes Jakobsson, Craig A. Berry, Dan Dascalescu,
2487 Daniel Frederick Crisman, Daniel M. Quinlan, Dan Jacobson, Dan Kogai,
2488 Dave Mitchell, Dave Rolsky, David Cantrell, David Dick, David Golden,
2489 David Mitchell, David M. Syzdek, David Nicol, David Wheeler, Dennis
2490 Kaarsemaker, Dintelmann, Peter, Dominic Dunlop, Dr.Ruud, Duke Leto,
2491 Enrico Sorcinelli, Eric Brine, Father Chrysostomos, Florian Ragwitz,
2492 Frank Wiegand, Gabor Szabo, Gene Sullivan, Geoffrey T. Dairiki, George
2493 Greer, Gerard Goossen, Gisle Aas, Goro Fuji, Graham Barr, Green, Paul,
2494 Hans Dieter Pearcey, Harmen, H. Merijn Brand, Hugo van der Sanden,
2495 Ian Goodacre, Igor Sutton, Ingo Weinhold, James Bence, James Mastros,
2496 Jan Dubois, Jari Aalto, Jarkko Hietaniemi, Jay Hannah, Jerry Hedden,
2497 Jesse Vincent, Jim Cromie, Jody Belka, John E. Malmberg, John Malmberg,
2498 John Peacock, John Peacock via RT, John P. Linderman, John Wright,
2499 Josh ben Jore, Jos I. Boumans, Karl Williamson, Kenichi Ishigaki, Ken
2500 Williams, Kevin Brintnall, Kevin Ryde, Kurt Starsinic, Leon Brocard,
2501 Lubomir Rintel, Luke Ross, Marcel Grünauer, Marcus Holland-Moritz, Mark
2502 Jason Dominus, Marko Asplund, Martin Hasch, Mashrab Kuvatov, Matt Kraai,
2503 Matt S Trout, Max Maischein, Michael Breen, Michael Cartmell, Michael
2504 G Schwern, Michael Witten, Mike Giroux, Milosz Tanski, Moritz Lenz,
2505 Nicholas Clark, Nick Cleaton, Niko Tyni, Offer Kaye, Osvaldo Villalon,
2506 Paul Fenwick, Paul Gaborit, Paul Green, Paul Johnson, Paul Marquess,
2507 Philip Hazel, Philippe Bruhat, Rafael Garcia-Suarez, Rainer Tammer,
2508 Rajesh Mandalemula, Reini Urban, Renée Bäcker, Ricardo Signes,
2509 Ricardo SIGNES, Richard Foley, Rich Rauenzahn, Rick Delaney, Risto
2510 Kankkunen, Robert May, Roberto C. Sanchez, Robin Barker, SADAHIRO
2511 Tomoyuki, Salvador Ortiz Garcia, Sam Vilain, Scott Lanning, Sébastien
2512 Aperghis-Tramoni, Sérgio Durigan Júnior, Shlomi Fish, Simon 'corecode'
2513 Schubert, Sisyphus, Slaven Rezic, Smylers, Steffen Müller, Steffen
2514 Ullrich, Stepan Kasal, Steve Hay, Steven Schubiger, Steve Peters, Tels,
2515 The Doctor, Tim Bunce, Tim Jenness, Todd Rinaldo, Tom Christiansen,
2516 Tom Hukins, Tom Wyant, Tony Cook, Torsten Schoenfeld, Tye McQueen,
2517 Vadim Konovalov, Vincent Pit, Hio YAMASHINA, Yasuhiro Matsumoto,
2518 Yitzchak Scott-Thoennes, Yuval Kogman, Yves Orton, Zefram, Zsban Ambrus
2520 This is woefully incomplete as it's automatically generated from version
2521 control history. In particular, it doesn't include the names of the
2522 (very much appreciated) contributors who reported issues in previous
2523 versions of Perl that helped make Perl 5.12.0 better. For a more complete
2524 list of all of Perl's historical contributors, please see the C<AUTHORS>
2525 file in the Perl 5.12.0 distribution.
2527 Many of the changes included in this version originated in the CPAN
2528 modules included in Perl's core. We're grateful to the entire CPAN
2529 community for helping Perl to flourish.
2531 =head1 Reporting Bugs
2533 If you find what you think is a bug, you might check the articles
2534 recently posted to the comp.lang.perl.misc newsgroup and the perl
2535 bug database at L<http://rt.perl.org/perlbug/>. There may also be
2536 information at L<http://www.perl.org/>, the Perl Home Page.
2538 If you believe you have an unreported bug, please run the B<perlbug>
2539 program included with your release. Be sure to trim your bug down
2540 to a tiny but sufficient test case. Your bug report, along with the
2541 output of C<perl -V>, will be sent off to perlbug@perl.org to be
2542 analyzed by the Perl porting team.
2544 If the bug you are reporting has security implications, which make it
2545 inappropriate to send to a publicly archived mailing list, then please send
2546 it to perl5-security-report@perl.org. This points to a closed subscription
2547 unarchived mailing list, which includes all the core committers, who be able
2548 to help assess the impact of issues, figure out a resolution, and help
2549 co-ordinate the release of patches to mitigate or fix the problem across all
2550 platforms on which Perl is supported. Please only use this address for
2551 security issues in the Perl core, not for modules independently
2552 distributed on CPAN.
2556 The F<Changes> file for an explanation of how to view exhaustive details
2559 The F<INSTALL> file for how to build Perl.
2561 The F<README> file for general stuff.
2563 The F<Artistic> and F<Copying> files for copyright information.