3 perl5120delta - what is new for perl v5.12.0
5 =head1 XXX - THIS DOCUMENT IS ONLY CURRENT THROUGH PERL5114
11 UPDATED MODULE LIST NEEDS TO BE GENERATED
12 ORDERING NEEDS CHECKING
13 HEAVY COPYEDITING IS NEEDED
18 This document describes differences between the 5.10.0 release and
21 Many of the bug fixes in 5.12.0 were already seen in the 5.10.1
22 maintenance release since the two releases were kept closely
23 coordinated (while 5.12.0 was still called 5.11.something).
25 You can see the list of changes in the 5.10.1 release
26 by reading L<perl5101delta>.
28 =head1 Core Enhancements
32 It is now possible to overload the C<qr//> operator, that is,
33 conversion to regexp, like it was already possible to overload
34 conversion to boolean, string or number of objects. It is invoked when
35 an object appears on the right hand side of the C<=~> operator or when
36 it is interpolated into a regexp. See L<overload>.
38 =head2 Pluggable keywords
40 Extension modules can now cleanly hook into the Perl parser to define
41 new kinds of keyword-headed expression and compound statement. The
42 syntax following the keyword is defined entirely by the extension. This
43 allow a completely non-Perl sublanguage to be parsed inline, with the
44 correct ops cleanly generated. This feature is currently considered
47 See L<perlapi/PL_keyword_plugin> for the mechanism. The Perl core
48 source distribution also includes a new module
49 L<XS::APItest::KeywordRPN>, which implements reverse Polish notation
50 arithmetic via pluggable keywords. This module is mainly used for test
51 purposes, and is not normally installed, but also serves as an example
52 of how to use the new mechanism.
54 =head2 APIs for more internals
56 The lowest layers of the lexer and parts of the pad system now have C
57 APIs available to XS extensions. These are necessary to support proper
58 use of pluggable keywords, but have other uses too. The new APIs are
59 experimental, and only cover a small proportion of what would be
60 necessary to take full advantage of the core's facilities in these
61 areas. It is intended that the Perl 5.13 development cycle will see the
62 addition of a full range of clean, supported interfaces.
64 =head2 Overridable function lookup
66 Where an extension module hooks the creation of rv2cv ops to modify the
67 subroutine lookup process, this now works correctly for bareword
68 subroutine calls. This means that prototypes on subroutines referenced
69 this way will be processed correctly. (Previously bareword subroutine
70 names were initially looked up, for parsing purposes, by an unhookable
71 mechanism, so extensions could only properly influence subroutine names
72 that appeared with an C<&> sigil.)
74 =head2 Unicode version
76 Perl is shipped with the latest Unicode version, 5.2, dated October 2009. See
77 L<http://www.unicode.org/versions/Unicode5.2.0> for details about this release
78 of Unicode. See L<perlunicode> for instructions on installing and using
79 older versions of Unicode.
81 =head2 Unicode properties
83 A concerted effort has been made to update Perl to be in sync with the latest
84 Unicode standard. Changes for this include:
86 Perl can now handle every Unicode character property. A new pod,
87 L<perluniprops>, lists all available non-Unihan character properties. By
88 default the Unihan properties and certain others (deprecated and Unicode
89 internal-only ones) are not exposed. See below for more details on
90 these; there is also a section in the pod listing them, and explaining
91 why they are not exposed.
93 Perl now fully supports the Unicode compound-style of using C<=> and C<:>
94 in writing regular expressions: C<\p{property=value}> and
95 C<\p{property:value}> (both of which mean the same thing).
97 Perl now fully supports the Unicode loose matching rules for text
98 between the braces in C<\p{...}> constructs. In addition, Perl allows
99 underscores between digits of numbers.
101 All the Unicode-defined synonyms for properties and property values are
104 C<qr/\X/>, which matches a Unicode logical character, has been expanded to work
105 better with various Asian languages. It now is defined as an I<extended
106 grapheme cluster>. (See L<http://www.unicode.org/reports/tr29/>).
107 Anything matched previously and that made sense will continue to be
108 matched, but in addition:
114 C<\X> will not break apart a C<S<CR LF>> sequence.
118 C<\X> will now match a sequence which includes the C<ZWJ> and C<ZWNJ> characters.
122 C<\X> will now always match at least one character, including an initial mark.
123 Marks generally come after a base character, but it is possible in Unicode to
124 have them in isolation, and C<\X> will now handle that case, for example at the
125 beginning of a line, or after a C<ZWSP>. And this is the part where C<\X>
126 doesn't match the things that it used to that don't make sense. Formerly, for
127 example, you could have the nonsensical case of an accented LF.
131 C<\X> will now match a (Korean) Hangul syllable sequence, and the Thai and Lao
136 Otherwise, this change should be transparent for the non-affected languages.
138 C<\p{...}> matches using the Canonical_Combining_Class property were
139 completely broken in previous Perls. This is now fixed.
141 In previous Perls, the Unicode C<Decomposition_Type=Compat> property and a
142 Perl extension had the same name, which led to neither matching all the
143 correct values (with more than 100 mistakes in one, and several thousand
144 in the other). The Perl extension has now been renamed to be
145 C<Decomposition_Type=Noncanonical> (short: C<dt=noncanon>). It has the same
146 meaning as was previously intended, namely the union of all the
147 non-canonical Decomposition types, with Unicode C<Compat> being just one of
150 C<\p{Uppercase}> and C<\p{Lowercase}> have been brought into line with the
151 Unicode definitions. This means they each match a few more characters
154 C<\p{Cntrl}> now matches the same characters as C<\p{Control}>. This means it
155 no longer will match Private Use (gc=co), Surrogates (gc=cs), nor Format
156 (gc=cf) code points. The Format code points represent the biggest
157 possible problem. All but 36 of them are either officially deprecated
158 or strongly discouraged from being used. Of those 36, likely the most
159 widely used are the soft hyphen (U+00AD), and BOM, ZWSP, ZWNJ, WJ, and
160 similar characters, plus bidirectional controls.
162 C<\p{Alpha}> now matches the same characters as C<\p{Alphabetic}>. The Perl
163 definition included a number of things that aren't really alpha (all
164 marks), while omitting many that were. As a direct consequence, the
165 definitions of C<\p{Alnum}> and C<\p{Word}> which depend on Alpha also change
168 C<\p{Word}> also now doesn't match certain characters it wasn't supposed
169 to, such as fractions.
171 C<\p{Print}> no longer matches the line control characters: Tab, LF, CR,
172 FF, VT, and NEL. This brings it in line with standards and the documentation.
174 C<\p{Decomposition_Type=Canonical}> now includes the Hangul syllables.
176 C<\p{XDigit}> now matches the same characters as C<\p{Hex_Digit}>. This
177 means that in addition to the characters it currently matches,
178 C<[A-Fa-f0-9]>, it will also match the 22 fullwidth equivalents, for
179 example U+FF10: FULLWIDTH DIGIT ZERO.
181 The Numeric type property has been extended to include the Unihan
184 There is a new Perl extension, the 'Present_In', or simply 'In',
185 property. This is an extension of the Unicode Age property, but
186 C<\p{In=5.0}> matches any code point whose usage has been determined
187 I<as of> Unicode version 5.0. The C<\p{Age=5.0}> only matches code points
188 added in I<precisely> version 5.0.
190 A number of properties did not have the correct values for unassigned
191 code points. This is now fixed. The affected properties are
192 Bidi_Class, East_Asian_Width, Joining_Type, Decomposition_Type,
193 Hangul_Syllable_Type, Numeric_Type, and Line_Break.
195 The Default_Ignorable_Code_Point, ID_Continue, and ID_Start properties
196 have been updated to their current Unicode definitions.
198 Certain properties that are supposed to be Unicode internal-only were
199 erroneously exposed by previous Perls. Use of these in regular
200 expressions will now generate, if enabled, a deprecated warning message.
201 The properties are: Other_Alphabetic, Other_Default_Ignorable_Code_Point,
202 Other_Grapheme_Extend, Other_ID_Continue, Other_ID_Start, Other_Lowercase,
203 Other_Math, and Other_Uppercase.
205 An installation can now fairly easily change which Unicode properties
206 Perl understands. As mentioned above, certain properties are by default
207 turned off. These include all the Unihan properties (which should be
208 accessible via the CPAN module Unicode::Unihan) and any deprecated or
209 Unicode internal-only property that Perl has never exposed.
211 The generated files in the C<lib/unicore/To> directory are now more
212 clearly marked as being stable, directly usable by applications.
213 New hash entries in them give the format of the normal entries,
214 which allows for easier machine parsing. Perl can generate files
215 in this directory for any property, though most are suppressed. An
216 installation can choose to change which get written. Instructions
217 are in L<perluniprops>.
219 =head2 Regular Expressions
221 U+0FFFF is now a legal character in regular expressions.
223 =head2 A proper interface for pluggable Method Resolution Orders
225 As of Perl 5.11.0 there is a new interface for plugging and using method
226 resolution orders other than the default (linear depth first search).
227 The C3 method resolution order added in 5.10.0 has been re-implemented as
228 a plugin, without changing its Perl-space interface. See L<perlmroapi> for
231 =head2 The C<overloading> pragma
233 This pragma allows you to lexically disable or enable overloading
234 for some or all operations. (Yuval Kogman)
236 =head2 C<\N> experimental regex escape
238 A new regex escape has been added, C<\N>. It will match any character that
239 is not a newline, independently from the presence or absence of the single
240 line match modifier C</s>. It is not usable within a character class.
241 C<\N{3}> means to match 3 non-newlines; C<\N{5,}> means to match at least 5.
242 C<\N{NAME}> still means the character or sequence named C<NAME>, but C<NAME> no
243 longer can be things like C<3>, or C<5,>.
244 Compatibility with Unicode names is preserved, as none look like these, but it
245 has been possible to create custom names that do look like them, and those will
246 no longer work. (Rafael Garcia-Suarez)
248 This will break a L<custom charnames translator|charnames/CUSTOM TRANSLATORS>
249 which allows numbers for character names, as C<\N{3}> will now mean to match 3
250 non-newline characters, and not the character whose name is C<3>. (No standard
251 name is a number, so only a custom translator would be affected.)
253 This escape is experimental, subject to change, because there is some concern
254 about possible confusion with the previous meaning of C<\N{...}>
256 =head2 \N{...} now compiles better, always forces UTF-8 internal representation.
258 There were several problems that have been fixed with recognizing C<\N{...}>
259 constructs. As part of this, any scalar or regex that has either a
260 C<\N{I<name>}> or C<\N{U+I<wide hex char>}> in its definition will be stored in
261 UTF-8 format. (This was true previously for all occurences of C<\N{I<name>}>
262 that did not use a custom translator, but now it's always true.)
264 =head2 Implicit strictures
266 Using the C<use VERSION> syntax with a version number greater or equal
267 to 5.11.0 will also lexically enable strictures just like C<use strict>
268 would do (in addition to enabling features.) So, the following:
277 =head2 Parallel tests
279 The core distribution can now run its regression tests in parallel on
280 Unix-like platforms. Instead of running C<make test>, set C<TEST_JOBS> in
281 your environment to the number of tests to run in parallel, and run
282 C<make test_harness>. On a Bourne-like shell, this can be done as
284 TEST_JOBS=3 make test_harness # Run 3 tests in parallel
286 An environment variable is used, rather than parallel make itself, because
287 L<TAP::Harness> needs to be able to schedule individual non-conflicting test
288 scripts itself, and there is no standard interface to C<make> utilities to
289 interact with their job schedulers.
291 Note that currently some test scripts may fail when run in parallel (most
292 notably C<ext/IO/t/io_dir.t>). If necessary run just the failing scripts
293 again sequentially and see if the failures go away.
295 =head2 The C<...> operator
297 A new operator, C<...>, nicknamed the Yada Yada operator, has been added.
298 It is intended to mark placeholder code that is not yet implemented.
299 See L<perlop/"Yada Yada Operator">. (chromatic)
301 =head2 DTrace support
303 Some support for DTrace has been added. See "DTrace support" in F<INSTALL>.
305 =head2 Support for C<configure_requires> in CPAN module metadata
307 Both C<CPAN> and C<CPANPLUS> now support the C<configure_requires> keyword
308 in the F<META.yml> metadata file included in most recent CPAN distributions.
309 This allows distribution authors to specify configuration prerequisites that
310 must be installed before running F<Makefile.PL> or F<Build.PL>.
312 See the documentation for C<ExtUtils::MakeMaker> or C<Module::Build> for more
313 on how to specify C<configure_requires> when creating a distribution for CPAN.
315 =head2 C<each> is now more flexible
317 The C<each> function can now operate on arrays.
319 =head2 Y2038 compliance
321 Perl's core time-related functions are now Y2038 compliant. (With 29
324 =head2 C<$,> flexibility
326 The variable C<$,> may now be tied.
328 =head2 // in where clauses
330 // now behaves like || in when clauses
332 =head2 Enabling warnings from your shell environment
334 You can now set C<-W> from the C<PERL5OPT> environment variable
336 =head2 C<delete local>
338 C<delete local> now allows you to locally delete a hash entry.
340 =head2 New support for Abstract namespace sockets
342 Abstract namespace sockets are Linux-specific socket type that live in
343 AF_UNIX family, slightly abusing it to be able to use arbitrary
344 character arrays as addresses: They start with nul byte and are not
345 terminated by nul byte, but with the length passed to the socket()
348 =head2 New C<package NAME VERSION> syntax
350 This new syntax allows a module author to set the $VERSION of a namespace
351 when the namespace is declared with 'package'. It eliminates the need
352 for C<our $VERSION = ...> and similar constructs. E.g.
354 package Foo::Bar 1.23;
355 # $Foo::Bar::VERSION == 1.23
357 There are several advantages to this:
363 C<$VERSION> is parsed in exactly the same way as C<use NAME VERSION>
367 C<$VERSION> is set at compile time
371 C<$VERSION> is a version object that provides proper overloading of
372 comparision operators so comparing C<$VERSION> to decimal (1.23) or
373 dotted-decimal (v1.2.3) version numbers works correctly.
377 Eliminates C<$VERSION = ...> and C<eval $VERSION> clutter
381 As it requires VERSION to be a numeric literal or v-string
382 literal, it can be statically parsed by toolchain modules
383 without C<eval> the way MM-E<gt>parse_version does for C<$VERSION = ...>
387 It does not break old code with only C<package NAME>, but code that uses
388 C<package NAME VERSION> will need to be restricted to perl 5.12.0 or newer
389 This is analogous to the change to C<open> from two-args to three-args.
390 Users requiring the latest Perl will benefit, and perhaps after several
391 years, it will become a standard practice.
395 However, C<package NAME VERSION> requires a new, 'strict' version
396 number format. See L<"Version number formats"> for details.
398 =head1 Incompatible Changes
400 =head2 Version number formats
402 Acceptable version number formats have been formalized into "strict" and
403 "lax" rules. C<package NAME VERSION> takes a strict version number.
404 C<UNIVERSAL::VERSION> and the L<version> object constructors take lax
405 version numbers. Providing an invalid version will result in a fatal
406 error. The version argument in C<use NAME VERSION> is first parsed as a
407 numeric literal or v-string and then passed to C<UNIVERSAL::VERSION>
408 (and must then pass the "lax" format test).
410 These formats are documented fully in the L<version> module. To a first
411 approximation, a "strict" version number is a positive decimal number
412 (integer or decimal-fraction) without exponentiation or else a
413 dotted-decimal v-string with a leading 'v' character and at least three
414 components. A "lax" version number allows v-strings with fewer than
415 three components or without a leading 'v'. Under "lax" rules, both
416 decimal and dotted-decimal versions may have a trailing "alpha"
417 component separated by an underscore character after a fractional or
418 dotted-decimal component.
420 The L<version> module adds C<version::is_strict> and C<version::is_lax>
421 functions to check a scalar against these rules.
423 =head2 @INC reorganization
425 In @INC, ARCHLIB and PRIVLIB now occur after after the current version's
426 site_perl and vendor_perl.
428 =head2 Switch statement changes
430 The handling of complex expressions by the C<given>/C<when> switch
431 statement has been enhanced. These enhancements are also available in
432 5.10.1 and subsequent 5.10 releases. There are two new cases where
433 C<when> now interprets its argument as a boolean, instead of an
434 expression to be used in a smart match:
436 =head2 flip-flop operators
438 The C<..> and C<...> flip-flop operators are now evaluated in boolean
439 context, following their usual semantics; see L<perlop/"Range Operators">.
441 Note that, as in perl 5.10.0, C<when (1..10)> will not work to test
442 whether a given value is an integer between 1 and 10; you should use
443 C<when ([1..10])> instead (note the array reference).
445 However, contrary to 5.10.0, evaluating the flip-flop operators in boolean
446 context ensures it can now be useful in a C<when()>, notably for
447 implementing bistable conditions, like in:
449 when (/^=begin/ .. /^=end/) {
453 =head2 defined-or operator
455 A compound expression involving the defined-or operator, as in
456 C<when (expr1 // expr2)>, will be treated as boolean if the first
457 expression is boolean. (This just extends the existing rule that applies
458 to the regular or operator, as in C<when (expr1 || expr2)>.)
460 =head2 Smart match changes
462 This section details more changes brought to the semantics to
463 the smart match operator, that naturally also modify the behaviour
464 of the switch statements where smart matching is implicitly used.
465 These changes were also made for the 5.10.1 release, and will remain in
466 subsequent 5.10 releases.
468 =head3 Changes to type-based dispatch
470 The smart match operator C<~~> is no longer commutative. The behaviour of
471 a smart match now depends primarily on the type of its right hand
472 argument. Moreover, its semantics have been adjusted for greater
473 consistency or usefulness in several cases. While the general backwards
474 compatibility is maintained, several changes must be noted:
480 Code references with an empty prototype are no longer treated specially.
481 They are passed an argument like the other code references (even if they
482 choose to ignore it).
486 C<%hash ~~ sub {}> and C<@array ~~ sub {}> now test that the subroutine
487 returns a true value for each key of the hash (or element of the
488 array), instead of passing the whole hash or array as a reference to
493 Due to the commutativity breakage, code references are no longer
494 treated specially when appearing on the left of the C<~~> operator,
495 but like any vulgar scalar.
499 C<undef ~~ %hash> is always false (since C<undef> can't be a key in a
500 hash). No implicit conversion to C<""> is done (as was the case in perl
505 C<$scalar ~~ @array> now always distributes the smart match across the
506 elements of the array. It's true if one element in @array verifies
507 C<$scalar ~~ $element>. This is a generalization of the old behaviour
508 that tested whether the array contained the scalar.
512 The full dispatch table for the smart match operator is given in
513 L<perlsyn/"Smart matching in detail">.
515 =head3 Smart match and overloading
517 According to the rule of dispatch based on the rightmost argument type,
518 when an object overloading C<~~> appears on the right side of the
519 operator, the overload routine will always be called (with a 3rd argument
520 set to a true value, see L<overload>.) However, when the object will
521 appear on the left, the overload routine will be called only when the
522 rightmost argument is a simple scalar. This way, distributivity of smart
523 match across arrays is not broken, as well as the other behaviours with
524 complex types (coderefs, hashes, regexes). Thus, writers of overloading
525 routines for smart match mostly need to worry only with comparing
526 against a scalar, and possibly with stringification overloading; the
527 other common cases will be automatically handled consistently.
529 C<~~> will now refuse to work on objects that do not overload it (in order
530 to avoid relying on the object's underlying structure). (However, if the
531 object overloads the stringification or the numification operators, and
532 if overload fallback is active, it will be used instead, as usual.)
534 =head2 Labels can't be keywords
536 Labels used as targets for the C<goto>, C<last>, C<next> or C<redo>
537 statements cannot be keywords anymore. This restriction will prevent
538 potential confusion between the C<goto LABEL> and C<goto EXPR> syntaxes:
539 for example, a statement like C<goto print> would jump to a label whose
540 name would be the return value of C<print()>, (usually 1), instead of a
541 label named C<print>. Moreover, the other control flow statements
542 would just ignore any keyword passed to them as a label name. Since
543 such labels cannot be defined anymore, this kind of error will be
546 =head2 Other incompatible changes
552 The definitions of a number of Unicode properties have changed to match those
553 of the current Unicode standard. These are listed above under L</Unicode
554 properties>. This could break code that is expecting the old definitions.
558 The boolkeys op moved to the group of hash ops. This breaks binary
563 Filehandles are blessed directly into C<IO::Handle>, as C<FileHandle> is
564 merely a wrapper around C<IO::Handle>.
566 The previous behaviour was to bless Filehandles into L<FileHandle>
567 (an empty proxy class) if it was loaded into memory and otherwise
568 to bless them into C<IO::Handle>.
572 The semantics of C<use feature :5.10*> have changed slightly.
573 See L<"Modules and Pragmata"> for more information.
577 The version control system used for the development of the perl
578 interpreter has been switched from Perforce to git. This is mainly an
579 internal issue that only affects people actively working on the perl core;
580 but it may have minor external visibility, for example in some of details
581 of the output of C<perl -V>. See L<perlrepository> for more information.
585 The internal structure of the C<ext/> directory in the perl source has
586 been reorganised. In general, a module C<Foo::Bar> whose source was
587 stored under F<ext/Foo/Bar/> is now located under F<ext/Foo-Bar/>. Also,
588 nearly all dual-life modules have been moved from F<lib/> to F<ext/>. This
589 is purely a source tarball change, and should make no difference to the
590 compilation or installation of perl, unless you have a very customised build
591 process that explicitly relies on this structure, or which hard-codes the
592 C<nonxs_ext> F<Configure> parameter. Specifically, this change does not by
593 default alter the location of any files in the final installation.
597 As part of the C<Test::Harness> 2.x to 3.x upgrade, the experimental
598 C<Test::Harness::Straps> module has been removed.
599 See L</"Updated Modules"> for more details.
603 As part of the C<ExtUtils::MakeMaker> upgrade, the
604 C<ExtUtils::MakeMaker::bytes> and C<ExtUtils::MakeMaker::vmsish> modules
605 have been removed from this distribution.
609 C<Module::CoreList> no longer contains the C<%:patchlevel> hash.
613 This one is actually a change introduced in 5.10.0, but it was missed
614 from that release's perldelta, so it is mentioned here instead.
616 A bugfix related to the handling of the C</m> modifier and C<qr> resulted
617 in a change of behaviour between 5.8.x and 5.10.0:
619 # matches in 5.8.x, doesn't match in 5.10.0
620 $re = qr/^bar/; "foo\nbar" =~ /$re/m;
624 C<length undef> now returns undef.
628 Unsupported private C API functions are now declared "static" to prevent
629 leakage to Perl's public API.
633 To support the bootstrapping process, F<miniperl> no longer builds with
634 UTF-8 support in the regexp engine.
636 This allows a build to complete with PERL_UNICODE set and a UTF-8 locale.
637 Without this there's a bootstrapping problem, as miniperl can't load the UTF-8
638 components of the regexp engine, because they're not yet built.
642 F<miniperl>'s @INC is now restricted to just C<-I...>, the split of
643 C<$ENV{PERL5LIB}>, and "C<.>"
647 A space or a newline is now required after a C<"#line XXX"> directive.
651 Tied filehandles now have an additional method EOF which provides the EOF type
655 To better match all other flow control statements, C<foreach> may no
656 longer be used as an attribute.
662 From time to time, Perl's developers find it necessary to deprecate
663 features or modules we've previously shipped as part of the core
664 distribution. We are well aware of the pain and frustration that a
665 backwards-incompatible change to Perl can cause for developers building
666 or maintaining software in Perl. You can be sure that when we deprecate
667 a functionality or syntax, it isn't a choice we make lightly. Sometimes,
668 we choose to deprecate functionality or syntax because it was found to
669 be poorly designed or implemented. Sometimes, this is because they're
670 holding back other features or causing performance problems. Sometimes,
671 the reasons are more complex. Wherever possible, we try to keep deprecated
672 functionality available to developers in its previous form for at least
673 one major release. So long as a deprecated feature isn't actively
674 disrupting our ability to maintain and extend Perl, we'll try to leave
675 it in place as long as possible.
677 The following items are now deprecated.
681 =item Use of C<:=> to mean an empty attribute list is now deprecated.
683 An accident of Perl's parser meant that these constructions were all
690 with the C<:> being treated as the start of an attribute list, which
691 ends before the C<=>. As whitespace is not significant here, all are
692 parsed as an empty attribute list, hence all the above are equivalent
693 to, and better written as
697 because no attribute processing is done for an empty list.
699 As is, this meant that C<:=> cannot be used as a new token, without
700 silently changing the meaning of existing code. Hence that particular
701 form is now deprecated, and will become a syntax error. If it is
702 absolutely necessary to have empty attribute lists (for example,
703 because of a code generator) then avoid the warning by adding a space
706 =item C<< UNIVERSAL->import() >>
708 The method C<< UNIVERSAL->import() >> is now deprecated. Attempting to
709 pass import arguments to a C<use UNIVERSAL> statement will result in a
712 =item Use of "goto" to jump into a construct is deprecated
714 Using C<goto> to jump from an outer scope into an inner scope is now
715 deprecated. This rare use case was causing problems in the
716 implementation of scopes.
718 =item Custom character names in \N{name} should look like names
720 In C<\N{I<name>}>, I<name> can be just about anything. The standard Unicode
721 names have a very limited domain, but a custom name translator could create
722 names that are, for example, made up entirely of punctuation symbols. It is
723 now deprecated to make names that don't begin with an alphabetic character, and
724 aren't alphanumeric or contain other than a very few other characters,
725 namely spaces, dashes, parentheses and colons. Because of the added meaning of
726 C<\N> (See L</C<\N> experimental regex escape>), names that look like curly
727 brace -enclosed quantifiers won't work. For example, C<\N{3,4}> now means to
728 match 3 to 4 non-newlines; before a custom name C<3,4> could have been created.
730 =item Deprecated Modules
732 The following modules will be removed from the core distribution in a future
733 release, and should be installed from CPAN instead. Distributions on CPAN
734 which require these should add them to their prerequisites. The core versions
735 of these modules warnings will issue a deprecation warning.
737 If you ship a packaged version of Perl, either alone or as part of a larger
738 system, then you should carefully consider the reprecussions of core module
739 deprecations. You may want to consider shipping your default build of
740 Perl with packages for some or all deprecated modules which install into
741 C<vendor> or C<site> perl library directories. This will inhibit the
742 deprecation warnings.
744 Alternatively, you may want to consider patching F<lib/deprecate.pm>
745 to provide deprecation warnings specific to your packaging system or
746 distribution of Perl.
752 =item L<Pod::Plainer>
758 Switch is buggy and should be avoided. See L<perlsyn/"Switch
759 statements"> for its replacement.
765 C<suidperl> has been removed. It used to provide a mechanism to
766 emulate setuid permission bits on systems that don't support it properly.
768 =item Assignment to $[
772 Remove attrs, which has been deprecated since 1999-10-02.
774 =item Use of the attribute :locked on subroutines.
776 =item Use of "locked" with the attributes pragma.
778 =item Use of "unique" with the attributes pragma.
780 =item Numerous Perl 4-era libraries:
782 F<termcap.pl>, F<tainted.pl>, F<stat.pl>, F<shellwords.pl>, F<pwd.pl>,
783 F<open3.pl>, F<open2.pl>, F<newgetopt.pl>, F<look.pl>, F<find.pl>,
784 F<finddepth.pl>, F<importenv.pl>, F<hostname.pl>, F<getopts.pl>,
785 F<getopt.pl>, F<getcwd.pl>, F<flush.pl>, F<fastcwd.pl>, F<exceptions.pl>,
786 F<ctime.pl>, F<complete.pl>, F<cacheout.pl>, F<bigrat.pl>, F<bigint.pl>,
787 F<bigfloat.pl>, F<assert.pl>, F<abbrev.pl>, F<dotsh.pl>, and
788 F<timelocal.pl> are all now deprecated. Using them will incur a warning.
792 =head1 Modules and Pragmata
794 =head2 Dual-lifed modules moved
796 Dual-lifed modules maintained primarily in the Perl core now live in dist/.
797 Dual-lifed modules maintained primarily on CPAN now live in cpan/
799 In previous releases of Perl, it was customary to enumerate all module
800 changes in this section of the C<perldelta> file. From 5.11.0 forward
801 only notable updates (such as new or deprecated modules ) will be listed
802 in this section. For a complete reference to the versions of modules
803 shipped in a given release of perl, please see L<Module::CoreList>.
805 =head2 New Modules and Pragmata
813 This is a new lexically-scoped alternative for the C<Fatal> module.
814 The bundled version is 2.06_01. Note that in this release, using a string
815 eval when C<autodie> is in effect can cause the autodie behaviour to leak
816 into the surrounding scope. See L<autodie/"BUGS"> for more details.
820 C<Compress::Raw::Bzip2>
822 This has been added to the core (version 2.020).
828 This pragma establishes an ISA relationship with base classes at compile
829 time. It provides the key feature of C<base> without further unwanted
836 This has been added to the core (version 1.39).
840 =head2 Pragmata Changes
848 See L</"The C<overloading> pragma"> above.
854 The C<attrs> pragma has been removed. It had been marked as deprecated since
861 The Unicode F<NameAliases.txt> database file has been added. This has the
862 effect of adding some extra C<\N> character names that formerly wouldn't
863 have been recognised; for example, C<"\N{LATIN CAPITAL LETTER GHA}">.
869 The meaning of the C<:5.10> and C<:5.10.X> feature bundles has
870 changed slightly. The last component, if any (i.e. C<X>) is simply ignored.
871 This is predicated on the assumption that new features will not, in
872 general, be added to maintenance releases. So C<:5.10> and C<:5.10.X>
873 have identical effect. This is a change to the behaviour documented for
880 Upgraded from version 1.00 to 1.01. Performance for single inheritance is 40%
881 faster - see L</"Performance Enhancements"> below.
883 C<mro> is now implemented as an XS extension. The documented interface has not
884 changed. Code relying on the implementation detail that some C<mro::>
885 methods happened to be available at all times gets to "keep both pieces".
891 Supports %.0f formatting internally.
897 Allow overloading of 'qr'.
903 Upgraded from version 1.19 to 1.20.
909 This pragma no longer suppresses C<Use of uninitialized value in range
910 (or flip)> warnings. [perl #71204]
916 Upgraded from 1.13 to 1.14. Added the C<unicode_strings> feature:
918 use feature "unicode_strings";
920 This pragma turns on Unicode semantics for the case-changing operations
921 (C<uc>, C<lc>, C<ucfirst>, C<lcfirst>) on strings that don't have the
922 internal UTF-8 flag set, but that contain single-byte characters between
929 Upgraded from version 1.74 to 1.75.
935 Upgraded from version 0.02 to 0.03.
937 This version introduces the C<stash_name> method to allow subclasses of
938 C<less> to pick where in %^H to store their stash.
944 Upgraded from version 0.77 to 0.81.
946 This version adds support for L</Version number formats> as described earlier
947 in this document and in its own documentation.
953 Upgraded from 1.07 to 1.09.
955 Added new C<warnings::fatal_enabled()> function.
956 This version adds the C<illegalproto> warning category. See also L</New or
957 Changed Diagnostics> for this change.
962 =head2 Updated Modules
966 =item XXX TODO RECALCULATE THIS VS 5.10.0
970 =head2 Removed Modules and Pragmata
978 Removed from the Perl core. Prior version was 'undef'.
984 =head2 New Documentation
992 This contains instructions on how to build perl for the Haiku platform.
998 This describes the new interface for pluggable Method Resolution Orders.
1004 This document, by Richard Foley, provides an introduction to the use of
1005 performance and optimization techniques which can be used with particular
1006 reference to perl programs.
1012 This describes how to access the perl source using the I<git> version
1017 L<perlpolicy> extends the "Social contract about contributed modules" into
1018 the beginnings of a document on Perl porting policies.
1022 =head2 Changes to Existing Documentation
1024 The various large F<Changes*> files (which listed every change made to perl
1025 over the last 18 years) have been removed, and replaced by a small file,
1026 also called F<Changes>, which just explains how that same information may
1027 be extracted from the git version control system.
1029 The file F<Porting/patching.pod> has been deleted, as it mainly described
1030 interacting with the old Perforce-based repository, which is now obsolete.
1031 Information still relevant has been moved to L<perlrepository>.
1033 L<perlapi>, L<perlintern>, L<perlmodlib> and L<perltoc> are now all
1034 generated at build time, rather than being shipped as part of the release.
1040 Documented -X overloading.
1044 Documented that C<when()> treats specially most of the filetest operators
1048 Documented C<when> as a syntax modifier
1052 Eliminated "Old Perl threads tutorial", which described 5005 threads.
1054 F<pod/perlthrtut.pod> is the same material reworked for ithreads.
1058 Correct previous documentation: v-strings are not deprecated
1060 With version objects, we need them to use MODULE VERSION syntax. This
1061 patch removes the deprecation notice.
1065 Added security contact information to L<perlsec>
1067 A significant fraction of the core documentation has been updated to clarify
1068 the behavior of Perl's Unicode handling.
1070 Much of the remaining core documentation has been reviewed and edited
1071 for clarity, consistent use of language, and to fix the spelling of Tom
1072 Christiansen's name.
1074 The Pod specification (L<perlpodspec>) has been updated to bring the
1075 specification in line with modern usage already supported by most Pod
1076 systems. A parameter string may now follow the format name in a
1077 "begin/end" region. Links to URIs with a text description are now
1078 allowed. The usage of C<LE<lt>"section"E<gt>> has been marked as
1081 L<if.pm|if> has been documented in L<perlfunc/use> as a means to get
1082 conditional loading of modules despite the implicit BEGIN block around
1087 The documentation for C<$1> in perlvar.pod has been clarified.
1091 C<\N{U+I<wide hex char>}> is now documented.
1095 =head1 Performance Enhancements
1101 A new internal cache means that C<isa()> will often be faster.
1105 The implementation of C<C3> Method Resolution Order has been optimised -
1106 linearisation for classes with single inheritance is 40% faster. Performance
1107 for multiple inheritance is unchanged.
1111 Under C<use locale>, the locale-relevant information is now cached on
1112 read-only values, such as the list returned by C<keys %hash>. This makes
1113 operations such as C<sort keys %hash> in the scope of C<use locale> much
1118 Empty C<DESTROY> methods are no longer called.
1122 Faster C<Perl_sv_utf8_upgrade()>
1126 Speed up C<keys> on empty hash
1130 C<if (%foo)> has been optimized to be faster than C<if (keys %foo)>
1134 Reversing an array to itself (as in C<@a = reverse @a>) in void context
1135 now happens in-place and is several orders of magnitude faster than it
1136 used to be. It will also preserve non-existent elements whenever
1137 possible, i.e. for non magical arrays or tied arrays with C<EXISTS> and
1142 =head1 Installation and Configuration Improvements
1144 =head2 F<ext/> reorganisation
1146 The layout of directories in F<ext> has been revised. Specifically, all
1147 extensions are now flat, and at the top level, with C</> in pathnames
1148 replaced by C<->, so that F<ext/Data/Dumper/> is now F<ext/Data-Dumper/>,
1149 etc. The names of the extensions as specified to F<Configure>, and as
1150 reported by C<%Config::Config> under the keys C<dynamic_ext>,
1151 C<known_extensions>, C<nonxs_ext> and C<static_ext> have not changed, and
1152 still use C</>. Hence this change will not have any affect once perl is
1153 installed. C<Safe> has been split out from being part of C<Opcode>, and
1154 C<mro> is now an extension in its own right.
1156 Nearly all dual-life modules have been moved from F<lib> to F<ext>, and will
1157 now appear as known C<nonxs_ext>. This will made no difference to the
1158 structure of an installed perl, nor will the modules installed differ,
1159 unless you run F<Configure> with options to specify an exact list of
1160 extensions to build. In this case, you will rapidly become aware that you
1161 need to add to your list, because various modules needed to complete the
1162 build, such as C<ExtUtils::ParseXS>, have now become extensions, and
1163 without them the build will fail well before it attempts to run the
1166 =head2 Configuration improvements
1168 If C<vendorlib> and C<vendorarch> are the same, then they are only added to
1171 C<$Config{usedevel}> and the C-level C<PERL_USE_DEVEL> are now defined if
1172 perl is built with C<-Dusedevel>.
1174 F<Configure> will enable use of C<-fstack-protector>, to provide protection
1175 against stack-smashing attacks, if the compiler supports it.
1177 F<Configure> will now determine the correct prototypes for re-entrant
1178 functions and for C<gconvert> if you are using a C++ compiler rather
1181 On Unix, if you build from a tree containing a git repository, the
1182 configuration process will note the commit hash you have checked out, for
1183 display in the output of C<perl -v> and C<perl -V>. Unpushed local commits
1184 are automatically added to the list of local patches displayed by
1187 USE_ATTRIBUTES_FOR_PERLIO is now reported in the compile-time options
1188 listed by the C<-V> switch.
1190 =head2 Compilation improvements
1192 As part of the flattening of F<ext>, all extensions on all platforms are
1193 built by F<make_ext.pl>. This replaces the Unix-specific
1194 F<ext/util/make_ext>, VMS-specific F<make_ext.com> and Win32-specific
1195 F<win32/buildext.pl>.
1197 =head1 Changed Internals
1203 C<Perl_pmflag> has been removed from the public API. Calling it now
1204 generates a deprecation warning, and it will be removed in a future
1205 release. Although listed as part of the API, it was never documented,
1206 and only ever used in F<toke.c>, and prior to 5.10, F<regcomp.c>. In
1207 core, it has been replaced by a static function.
1211 Perl_magic_setmglob now knows about globs, fixing RT #71254.
1215 TODO: C<SVt_RV> is gone. RVs are now stored in IVs
1219 TODO: REGEXPs are first class
1223 TODO: OOK is reworked, such that an OOKed scalar is PV not PVIV
1227 The J.R.R. Tolkien quotes at the head of C source file have been checked and
1228 proper citations added, thanks to a patch from Tom Christiansen.
1232 C<Perl_vcroak()> now accepts a null first argument. In addition, a full audit
1233 was made of the "not NULL" compiler annotations, and those for several
1234 other internal functions were corrected.
1238 New macros C<dSAVEDERRNO>, C<dSAVE_ERRNO>, C<SAVE_ERRNO>, C<RESTORE_ERRNO>
1239 have been added to formalise the temporary saving of the C<errno>
1244 The function C<Perl_sv_insert_flags> has been added to augment
1249 The function C<Perl_newSV_type(type)> has been added, equivalent to
1250 C<Perl_newSV()> followed by C<Perl_sv_upgrade(type)>.
1254 The function C<Perl_newSVpvn_flags()> has been added, equivalent to
1255 C<Perl_newSVpvn()> and then performing the action relevant to the flag.
1257 Two flag bits are currently supported.
1265 This will call C<SvUTF8_on()> for you. (Note that this does not convert an
1266 sequence of ISO 8859-1 characters to UTF-8). A wrapper, C<newSVpvn_utf8()>
1267 is available for this.
1273 Call C<Perl_sv_2mortal()> on the new SV.
1277 There is also a wrapper that takes constant strings, C<newSVpvs_flags()>.
1281 The function C<Perl_croak_xs_usage> has been added as a wrapper to
1286 The functions C<PerlIO_find_layer> and C<PerlIO_list_alloc> are now
1291 C<PL_na> has been exterminated from the core code, replaced by local STRLEN
1292 temporaries, or C<*_nolen()> calls. Either approach is faster than C<PL_na>,
1293 which is a pointer dereference into the interpreter structure under ithreads,
1294 and a global variable otherwise.
1298 C<Perl_mg_free()> used to leave freed memory accessible via C<SvMAGIC()> on
1299 the scalar. It now updates the linked list to remove each piece of magic
1304 Under ithreads, the regex in C<PL_reg_curpm> is now reference counted. This
1305 eliminates a lot of hackish workarounds to cope with it not being reference
1310 C<Perl_mg_magical()> would sometimes incorrectly turn on C<SvRMAGICAL()>.
1311 This has been fixed.
1315 The I<public> IV and NV flags are now not set if the string value has
1316 trailing "garbage". This behaviour is consistent with not setting the
1317 public IV or NV flags if the value is out of range for the type.
1321 SV allocation tracing has been added to the diagnostics enabled by C<-Dm>.
1322 The tracing can alternatively output via the C<PERL_MEM_LOG> mechanism, if
1323 that was enabled when the F<perl> binary was compiled.
1327 Smartmatch resolution tracing has been added as a new diagnostic. Use C<-DM> to
1332 A new debugging flag C<-DB> now dumps subroutine definitions, leaving
1333 C<-Dx> for its original purpose of dumping syntax trees.
1337 Uses of C<Nullav>, C<Nullcv>, C<Nullhv>, C<Nullop>, C<Nullsv> etc have been
1338 replaced by C<NULL> in the core code, and non-dual-life modules, as C<NULL>
1339 is clearer to those unfamiliar with the core code.
1343 A macro C<MUTABLE_PTR(p)> has been added, which on (non-pedantic) gcc will
1344 not cast away C<const>, returning a C<void *>. Macros C<MUTABLE_SV(av)>,
1345 C<MUTABLE_SV(cv)> etc build on this, casting to C<AV *> etc without
1346 casting away C<const>. This allows proper compile-time auditing of
1347 C<const> correctness in the core, and helped picked up some errors (now
1352 Macros C<mPUSHs()> and C<mXPUSHs()> have been added, for pushing SVs on the
1353 stack and mortalizing them.
1357 Use of the private structure C<mro_meta> has changed slightly. Nothing
1358 outside the core should be accessing this directly anyway.
1362 A new tool, F<Porting/expand-macro.pl> has been added, that allows you
1363 to view how a C preprocessor macro would be expanded when compiled.
1364 This is handy when trying to decode the macro hell that is the perl
1373 Many modules updated from CPAN incorporate new tests.
1374 Several tests that have the potential to hang forever if they fail now
1375 incorporate a "watchdog" functionality that will kill them after a timeout,
1376 which helps ensure that C<make test> and C<make test_harness> run to
1377 completion automatically. (Jerry Hedden).
1379 Some core-specific tests have been added:
1385 Significant cleanups to core tests to ensure that language and
1386 interpreter features are not used before they're tested.
1390 C<make test_porting> now runs a number of important pre-commit checks
1391 which might be of use to anyone working on the Perl core.
1395 F<t/porting/podcheck.t> automatically checks the well-formedness of
1396 POD found in all .pl, .pm and .pod files in the F<MANIFEST>, other than in
1397 dual-lifed modules which are primarily maintained outside the Perl core.
1401 F<t/porting/manifest.t> now tests that all files listed in MANIFEST are present.
1405 F<t/op/while_readdir.t>
1407 Test that a bare readdir in while loop sets $_.
1411 F<t/comp/retainedlines.t>
1413 Check that the debugger can retain source lines from C<eval>.
1417 F<t/io/perlio_fail.t>
1419 Check that bad layers fail.
1423 F<t/io/perlio_leaks.t>
1425 Check that PerlIO layers are not leaking.
1429 F<t/io/perlio_open.t>
1431 Check that certain special forms of open work.
1437 General PerlIO tests.
1443 Check that there is no unexpected interaction between the internal types
1444 C<PVBM> and C<PVGV>.
1448 F<t/mro/package_aliases.t>
1450 Check that mro works properly in the presence of aliased packages.
1456 Tests for C<dbmopen> and C<dbmclose>.
1462 Tests for the interaction of C<index> and threads.
1468 Tests for the interaction of esoteric patterns and threads.
1474 Test that C<qr> doesn't leak.
1478 F<t/op/reg_email_thr.t>
1480 Tests for the interaction of regex recursion and threads.
1484 F<t/op/regexp_qr_embed_thr.t>
1486 Tests for the interaction of patterns with embedded C<qr//> and threads.
1490 F<t/op/regexp_unicode_prop.t>
1492 Tests for Unicode properties in regular expressions.
1496 F<t/op/regexp_unicode_prop_thr.t>
1498 Tests for the interaction of Unicode properties and threads.
1502 F<t/op/reg_nc_tie.t>
1504 Test the tied methods of C<Tie::Hash::NamedCapture>.
1508 F<t/op/reg_posixcc.t>
1510 Check that POSIX character classes behave consistently.
1516 Check that exportable C<re> functions in F<universal.c> work.
1520 F<t/op/setpgrpstack.t>
1522 Check that C<setpgrp> works.
1526 F<t/op/substr_thr.t>
1528 Tests for the interaction of C<substr> and threads.
1534 Check that upgrading and assigning scalars works.
1540 Check that Unicode in the lexer works.
1546 Check that Unicode and C<tie> work.
1550 F<t/comp/final_line_num.t>
1552 See if line numbers are correct at EOF
1556 F<t/comp/form_scope.t>
1558 See if format scoping works
1562 F<t/comp/line_debug.t>
1564 See if C<< @{"_<$file"} >> works
1568 F<t/op/filetest_t.t>
1570 See if -t file test works
1582 Tests malfunctions of utf8 cache
1588 Test unicode \p{} regex constructs
1592 =head2 Testing improvements
1598 It's now possible to override C<PERL5OPT> and friends in F<t/TEST>
1603 =head1 New or Changed Diagnostics
1605 Several new diagnostics, see L<perldiag> for details.
1611 C<Bad plugin affecting keyword '%s'>
1615 C<gmtime(%.0f) too large>
1619 C<Lexing code attempted to stuff non-Latin-1 character into Latin-1 input>
1623 C<Lexing code internal error (%s)>
1627 C<localtime(%.0f) too large>
1631 C<Overloaded dereference did not return a reference>
1635 C<Overloaded qr did not return a REGEXP>
1639 C<Perl_pmflag() is deprecated, and will be removed from the XS API>
1643 New warning category C<illegalproto>
1647 Illegal character in prototype for %s : %s
1648 Prototype after '%c' for %s : %s
1650 have been moved from the C<syntax> top-level warnings category into a new
1651 first-level category, C<illegalproto>. These two warnings are currently the
1652 only ones emitted during parsing of an invalid/illegal prototype, so one
1655 no warnings 'illegalproto';
1657 to suppress only those, but not other syntax-related warnings. Warnings where
1658 prototypes are changed, ignored, or not met are still in the C<prototype>
1659 category as before. (Matt S. Trout)
1663 lvalue attribute ignored after the subroutine has been defined
1665 This new warning is issued when one attempts to mark a subroutine as
1666 lvalue after it has been defined.
1670 warn if C<++> or C<--> are unable to change the value because it's
1671 beyond the limit of representation
1673 This uses a new warnings category: "imprecision".
1677 C<lc>, C<uc>, C<lcfirst>, and C<ucfirst> warn when passed undef.
1681 Show constant in "Useless use of a constant in void context"
1685 Make the new warning report undef constants as undef
1689 Add a new warning, "Prototype after '%s'"
1693 Tweak the "Illegal character in prototype" warning so it's more precise
1694 when reporting illegal characters after _
1698 Correct the unintended interpolation of C<$\> in regex
1702 Make overflow warnings in C<gmtime> and C<localtime> only occur when
1703 warnings are enabled
1707 Improve mro merging error messages.
1709 They are now very similar to those produced by Algorithm::C3.
1713 Amelioration of the error message "Unrecognized character %s in column %d"
1715 Changes the error message to "Unrecognized character %s; marked by E<lt>--
1716 HERE after %sE<lt>-- HERE near column %d". This should make it a little
1717 simpler to spot and correct the suspicious character.
1721 Explicitely point to C<$.> when it causes an uninitialized warning for
1722 ranges in scalar context
1726 C<split> now warns when called in void context
1730 C<printf>-style functions called with too few arguments will now issue the
1731 warning C<"Missing argument in %s"> [perl #71000]
1735 C<panic: sv_chop %s>
1737 This new fatal error occurs when the C routine C<Perl_sv_chop()> was
1738 passed a position that is not within the scalar's string buffer. This
1739 could be caused by buggy XS code, and at this point recovery is not
1744 C<Deep recursion on subroutine "%s">
1746 It is now possible to change the depth threshold for this warning from the
1747 default of 100, by recompiling the F<perl> binary, setting the C
1748 pre-processor macro C<PERL_SUB_DEPTH_WARN> to the desired value.
1752 Perl now properly returns a syntax error instead of segfaulting
1753 if C<each>, C<keys>, or C<values> is used without an argument.
1757 C<tell()> now fails properly if called without an argument and when no
1758 previous file was read.
1760 C<tell()> now returns C<-1>, and sets errno to C<EBADF>, thus restoring
1761 the 5.8.x behaviour.
1765 C<overload> no longer implicitly unsets fallback on repeated 'use
1770 POSIX::strftime() can now handle Unicode characters in the format string.
1774 The Windows select() implementation now supports all empty C<fd_set>s
1779 The "syntax" category was removed from 5 warnings that should only be in
1784 Three fatal C<pack>/C<unpack> error messages have been normalized to
1789 "Unicode character is illegal" has been rephrased to be more accurate
1791 It now reads C<Unicode non-character is illegal in interchange> and the
1792 perldiag documentation has been expanded a bit.
1796 Perl now defaults to issuing a warning if a deprecated language feature
1799 To disable this feature in a given lexical scope, you should use C<no
1800 warnings 'deprecated';> For information about which language features
1801 are deprecated and explanations of various deprecation warnings, please
1806 The following diagnostics have been removed:
1816 C<Can't locate package %s for the parents of %s>
1818 This warning has been removed. In general, it only got produced in
1819 conjunction with other warnings, and removing it allowed an ISA lookup
1820 optimisation to be added.
1824 C<v-string in use/require is non-portable>
1828 =head1 Utility Changes
1836 Now looks in C<include-fixed> too, which is a recent addition to gcc's
1843 No longer incorrectly treats enum values like macros (Daniel Burr).
1845 Now handles C++ style constants (C<//>) properly in enums. (A patch from
1846 Rainer Weikusat was used; Daniel Burr also proposed a similar fix).
1852 C<LVALUE> subroutines now work under the debugger.
1854 The debugger now correctly handles proxy constant subroutines, and
1861 F<perlbug> now uses C<%Module::CoreList::bug_tracker> to print out upstream bug
1864 Where the user names a module that their bug report is about, and we know the
1865 URL for its upstream bug tracker, provide a message to the user explaining
1866 that the core copies the CPAN version directly, and provide the URL for
1867 reporting the bug directly to upstream.
1873 Perl 5.11.0 added a new utility F<perlthanks>, which is a variant of
1874 F<perlbug>, but for sending non-bug-reports to the authors and maintainers
1875 of Perl. Getting nothing but bug reports can become a bit demoralising:
1876 we'll see if this changes things.
1882 No longer reports "Message sent" when it hasn't actually sent the message
1888 Fixed bugs with the match() operator in list context, remove mention of
1893 =head1 Selected Bug Fixes
1899 Ensure that pp_qr returns a new regexp SV each time. Resolves RT #69852.
1901 Instead of returning a(nother) reference to the (pre-compiled) regexp in the
1902 optree, use reg_temp_copy() to create a copy of it, and return a reference to
1903 that. This resolves issues about Regexp::DESTROY not being called in a timely
1904 fashion (the original bug tracked by RT #69852), as well as bugs related to
1905 blessing regexps, and of assigning to regexps, as described in correspondence
1906 added to the ticket.
1908 It transpires that we also need to undo the SvPVX() sharing when ithreads
1909 cloning a Regexp SV, because mother_re is set to NULL, instead of a cloned
1910 copy of the mother_re. This change might fix bugs with regexps and threads in
1911 certain other situations, but as yet neither tests nor bug reports have
1912 indicated any problems, so it might not actually be an edge case that it's
1917 Several compilation errors and segfaults when perl was built with C<-Dmad> were fixed.
1921 Fixes for lexer API changes in 5.11.2 which broke NYTProf's savesrc option.
1925 C<-t> should only return TRUE for file handles connected to a TTY
1927 The Microsoft C version of C<isatty()> returns TRUE for all
1928 character mode devices, including the F</dev/null>-style "nul"
1929 device and printers like "lpt1".
1933 Fixed a regression caused by commit fafafbaf which caused a panic during
1934 parameter passing [perl #70171]
1938 On systems which in-place edits without backup files, -i'*' now works as
1939 the documentation says it does [perl #70802]
1943 Saving and restoring magic flags no longer loses readonly flag.
1947 The malformed syntax C<grep EXPR LIST> (note the missing comma) no longer
1948 causes abrupt and total failure.
1952 Regular expressions compiled with C<qr{}> literals properly set C<$'> when
1957 Using named subroutines with C<sort> should no longer lead to bus errors [perl
1962 Numerous bugfixes catch small issues caused by the recently-added Lexer API.
1966 Smart match against C<@_> sometimes gave false negatives. [perl #71078]
1970 C<$@> may now be assigned a read-only value (without error or busting
1975 C<sort> called recursively from within an active comparison subroutine no
1976 longer causes a bus error if run multiple times. [perl #71076]
1980 Tie::Hash::NamedCapture::* will not abort if passed bad input (RT #71828)
1984 @_ and $_ no longer leak under threads (RT #34342 and #41138, also
1989 C<-I> on shebang line now adds directories in front of @INC
1990 as documented, and as does C<-I> when specified on the command-line.
1994 C<kill> is now fatal when called on non-numeric process identifiers.
1995 Previously, an C<undef> process identifier would be interpreted as a
1996 request to kill process 0, which would terminate the current process
1997 group on POSIX systems. Since process identifiers are always integers,
1998 killing a non-numeric process is now fatal.
2002 5.10.0 inadvertently disabled an optimisation, which caused a measurable
2003 performance drop in list assignment, such as is often used to assign
2004 function parameters from C<@_>. The optimisation has been re-instated, and
2005 the performance regression fixed. (This fix is also present in 5.10.1)
2009 Fixed memory leak on C<while (1) { map 1, 1 }> [RT #53038].
2013 Some potential coredumps in PerlIO fixed [RT #57322,54828].
2017 The debugger now works with lvalue subroutines.
2021 The debugger's C<m> command was broken on modules that defined constants
2026 C<crypt> and string complement could return tainted values for untainted
2027 arguments [RT #59998].
2031 The C<-i>I<.suffix> command-line switch now recreates the file using
2032 restricted permissions, before changing its mode to match the original
2033 file. This eliminates a potential race condition [RT #60904].
2037 On some Unix systems, the value in C<$?> would not have the top bit set
2038 (C<$? & 128>) even if the child core dumped.
2042 Under some circumstances, C<$^R> could incorrectly become undefined
2047 In the XS API, various hash functions, when passed a pre-computed hash where
2048 the key is UTF-8, might result in an incorrect lookup.
2052 XS code including F<XSUB.h> before F<perl.h> gave a compile-time error
2057 C<< $object-E<gt>isa('Foo') >> would report false if the package C<Foo> didn't
2058 exist, even if the object's C<@ISA> contained C<Foo>.
2062 Various bugs in the new-to 5.10.0 mro code, triggered by manipulating
2063 C<@ISA>, have been found and fixed.
2067 Bitwise operations on references could crash the interpreter, e.g.
2068 C<$x=\$y; $x |= "foo"> [RT #54956].
2072 Patterns including alternation might be sensitive to the internal UTF-8
2073 representation, e.g.
2075 my $byte = chr(192);
2076 my $utf8 = chr(192); utf8::upgrade($utf8);
2077 $utf8 =~ /$byte|X}/i; # failed in 5.10.0
2081 Within UTF8-encoded Perl source files (i.e. where C<use utf8> is in
2082 effect), double-quoted literal strings could be corrupted where a C<\xNN>,
2083 C<\0NNN> or C<\N{}> is followed by a literal character with ordinal value
2084 greater than 255 [RT #59908].
2088 C<B::Deparse> failed to correctly deparse various constructs:
2089 C<readpipe STRING> [RT #62428], C<CORE::require(STRING)> [RT #62488],
2090 C<sub foo(_)> [RT #62484].
2094 Using C<setpgrp> with no arguments could corrupt the perl stack.
2098 The block form of C<eval> is now specifically trappable by C<Safe> and
2099 C<ops>. Previously it was erroneously treated like string C<eval>.
2103 In 5.10.0, the two characters C<[~> were sometimes parsed as the smart
2104 match operator (C<~~>) [RT #63854].
2108 In 5.10.0, the C<*> quantifier in patterns was sometimes treated as
2109 C<{0,32767}> [RT #60034, #60464]. For example, this match would fail:
2111 ("ab" x 32768) =~ /^(ab)*$/
2115 C<shmget> was limited to a 32 bit segment size on a 64 bit OS [RT #63924].
2119 Using C<next> or C<last> to exit a C<given> block no longer produces a
2120 spurious warning like the following:
2122 Exiting given via last at foo.pl line 123
2126 On Windows, C<'.\foo'> and C<'..\foo'> were treated differently than
2127 C<'./foo'> and C<'../foo'> by C<do> and C<require> [RT #63492].
2131 Assigning a format to a glob could corrupt the format; e.g.:
2133 *bar=*foo{FORMAT}; # foo format now bad
2137 Attempting to coerce a typeglob to a string or number could cause an
2138 assertion failure. The correct error message is now generated,
2139 C<Can't coerce GLOB to I<$type>>.
2143 Under C<use filetest 'access'>, C<-x> was using the wrong access mode. This
2144 has been fixed [RT #49003].
2148 C<length> on a tied scalar that returned a Unicode value would not be
2149 correct the first time. This has been fixed.
2153 Using an array C<tie> inside in array C<tie> could SEGV. This has been
2158 A race condition inside C<PerlIOStdio_close()> has been identified and
2159 fixed. This used to cause various threading issues, including SEGVs.
2163 In C<unpack>, the use of C<()> groups in scalar context was internally
2164 placing a list on the interpreter's stack, which manifested in various
2165 ways, including SEGVs. This is now fixed [RT #50256].
2169 Magic was called twice in C<substr>, C<\&$x>, C<tie $x, $m> and C<chop>.
2170 These have all been fixed.
2174 A 5.10.0 optimisation to clear the temporary stack within the implicit
2175 loop of C<s///ge> has been reverted, as it turned out to be the cause of
2176 obscure bugs in seemingly unrelated parts of the interpreter [commit
2181 The line numbers for warnings inside C<elsif> are now correct.
2185 The C<..> operator now works correctly with ranges whose ends are at or
2186 close to the values of the smallest and largest integers.
2190 C<binmode STDIN, ':raw'> could lead to segmentation faults on some platforms.
2191 This has been fixed [RT #54828].
2195 An off-by-one error meant that C<index $str, ...> was effectively being
2196 executed as C<index "$str\0", ...>. This has been fixed [RT #53746].
2200 Various leaks associated with named captures in regexes have been fixed
2205 A weak reference to a hash would leak. This was affecting C<DBI>
2210 Using (?|) in a regex could cause a segfault [RT #59734].
2214 Use of a UTF-8 C<tr//> within a closure could cause a segfault [RT #61520].
2218 Calling C<Perl_sv_chop()> or otherwise upgrading an SV could result in an
2219 unaligned 64-bit access on the SPARC architecture [RT #60574].
2223 In the 5.10.0 release, C<inc_version_list> would incorrectly list
2224 C<5.10.*> after C<5.8.*>; this affected the C<@INC> search order
2229 In 5.10.0, C<pack "a*", $tainted_value> returned a non-tainted value
2234 In 5.10.0, C<printf> and C<sprintf> could produce the fatal error
2235 C<panic: utf8_mg_pos_cache_update> when printing UTF-8 strings
2240 In the 5.10.0 release, a dynamically created C<AUTOLOAD> method might be
2241 missed (method cache issue) [RT #60220,60232].
2245 In the 5.10.0 release, a combination of C<use feature> and C<//ee> could
2246 cause a memory leak [RT #63110].
2250 C<-C> on the shebang (C<#!>) line is once more permitted if it is also
2251 specified on the command line. C<-C> on the shebang line used to be a
2252 silent no-op I<if> it was not also on the command line, so perl 5.10.0
2253 disallowed it, which broke some scripts. Now perl checks whether it is
2254 also on the command line and only dies if it is not [RT #67880].
2258 In 5.10.0, certain types of re-entrant regular expression could crash,
2259 or cause the following assertion failure [RT #60508]:
2261 Assertion rx->sublen >= (s - rx->subbeg) + i failed
2265 Previously missing files from Unicode 5.1 Character Database are now included.
2269 C<TMPDIR> is now honored when opening an anonymous temporary file
2273 =head1 Platform Specific Changes
2275 =head2 New Platforms
2281 Patches from the Haiku maintainers have been merged in. Perl should now
2286 Perl should now build on MirOS BSD.
2291 =head2 Discontinued Platforms
2297 Support for Apollo DomainOS was removed in Perl 5.11.0
2301 Support for Tenon Intersystems MachTen Unix layer for MacOS Classic was
2302 removed in Perl 5.11.0
2306 Support for Atari MiNT was removed in Perl 5.11.0.
2310 =head2 Updated Platforms
2314 =item Darwin (Mac OS X)
2320 Skip testing the be_BY.CP1131 locale on Darwin 10 (Mac OS X 10.6),
2321 as it's still buggy.
2325 Correct infelicities in the regexp used to identify buggy locales
2326 on Darwin 8 and 9 (Mac OS X 10.4 and 10.5, respectively).
2336 Fix thread library selection [perl #69686]
2346 Initial support for mingw64 is now available
2350 Various bits of Perl's build infrastructure are no longer converted to
2351 win32 line endings at release time. If this hurts you, please report the
2352 problem with the L<perlbug> program included with perl.
2356 Always add a manifest resource to C<perl.exe> to specify the C<trustInfo>
2357 settings for Windows Vista and later. Without this setting Windows
2358 will treat C<perl.exe> as a legacy application and apply various
2359 heuristics like redirecting access to protected file system areas
2360 (like the "Program Files" folder) to the users "VirtualStore"
2361 instead of generating a proper "permission denied" error.
2363 For VC8 and VC9 this manifest setting is automatically generated by
2364 the compiler/linker (together with the binding information for their
2365 respective runtime libraries); for all other compilers we need to
2366 embed the manifest resource explicitly in the external resource file.
2368 This change also requests the Microsoft Common-Controls version 6.0
2369 (themed controls introduced in Windows XP) via the dependency list
2370 in the assembly manifest. For VC8 and VC9 this is specified using the
2371 C</manifestdependency> linker commandline option instead.
2375 Improved message window handling means that C<alarm> and C<kill> messages
2376 will no longer be dropped under race conditions.
2386 Enable IPv6 support on cygwin 1.7 and newer
2396 Make -UDEBUGGING the default on VMS for 5.12.0.
2398 Like it has been everywhere else for ages and ages. Also make
2399 command-line selection of -UDEBUGGING and -DDEBUGGING work in
2400 configure.com; before the only way to turn it off was by saying
2401 no in answer to the interactive question.
2405 The default pipe buffer size on VMS has been updated to 8192 on 64-bit
2410 Reads from the in-memory temporary files of C<PerlIO::scalar> used to fail
2411 if C<$/> was set to a numeric reference (to indicate record-style reads).
2416 VMS now supports C<getgrgid>.
2420 Many improvements and cleanups have been made to the VMS file name handling
2421 and conversion code.
2425 Enabling the C<PERL_VMS_POSIX_EXIT> logical name now encodes a POSIX exit
2426 status in a VMS condition value for better interaction with GNV's bash
2427 shell and other utilities that depend on POSIX exit values. See
2428 L<perlvms/"$?"> for details.
2432 C<File::Copy> now detects Unix compatibility mode on VMS.
2438 Removed F<libbsd> for AIX 5L and 6.1. Only C<flock()> was used from F<libbsd>.
2440 Removed F<libgdbm> for AIX 5L and 6.1. The F<libgdbm> is delivered as an
2441 optional package with the AIX Toolbox. Unfortunately the 64 bit version
2444 Hints changes mean that AIX 4.2 should work again.
2448 On Cygwin we now strip the last number from the DLL. This has been the
2449 behaviour in the cygwin.com build for years. The hints files have been
2455 The hints files now identify the correct threading libraries on FreeBSD 7
2460 We now work around a bizarre preprocessor bug in the Irix 6.5 compiler:
2461 C<cc -E -> unfortunately goes into K&R mode, but C<cc -E file.c> doesn't.
2465 Hints now supports versions 5.*.
2469 Various changes from Stratus have been merged in.
2473 There is now support for Symbian S60 3.2 SDK and S60 5.0 SDK.
2477 =head1 Known Problems
2479 This is a list of some significant unfixed bugs, which are regressions
2480 from either 5.10.0 or 5.8.x.
2486 C<List::Util::first> misbehaves in the presence of a lexical C<$_>
2487 (typically introduced by C<my $_> or implicitly by C<given>). The variable
2488 which gets set for each iteration is the package variable C<$_>, not the
2489 lexical C<$_> [RT #67694].
2491 A similar issue may occur in other modules that provide functions which
2492 take a block as their first argument, like
2494 foo { ... $_ ...} list
2498 Some regexes may run much more slowly when run in a child thread compared
2499 with the thread the pattern was compiled into [RT #55600].
2503 Things like C<"\N{LATIN SMALL LIGATURE FF}" =~ /\N{LATIN SMALL LETTER F}+/>
2504 will appear to hang as they get into a very long running loop [RT #72998].
2508 Untriaged test crashes on Windows 2000
2510 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.
2514 Known test failures on VMS
2516 Perl 5.11.1 fails a small set of core and CPAN tests as of this release.
2517 With luck, that'll be sorted out for 5.11.2
2521 Known test failures on VMS
2523 Perl 5.11.2 fails a small set of core and CPAN tests as of this
2524 release. With luck, that'll be sorted out for 5.11.3.
2528 =head1 Acknowledgements
2530 Perl 5.12.0 represents approximately two years of development since
2531 Perl 5.10.0 and contains __ lines of changes across ___ files
2532 from __ authors and committers:
2536 Many of the changes included in this version originated in the CPAN
2537 modules included in Perl's core. We're grateful to the entire CPAN
2538 community for helping Perl to flourish.
2540 =head1 Reporting Bugs
2542 If you find what you think is a bug, you might check the articles
2543 recently posted to the comp.lang.perl.misc newsgroup and the perl
2544 bug database at L<http://rt.perl.org/perlbug/>. There may also be
2545 information at L<http://www.perl.org/>, the Perl Home Page.
2547 If you believe you have an unreported bug, please run the B<perlbug>
2548 program included with your release. Be sure to trim your bug down
2549 to a tiny but sufficient test case. Your bug report, along with the
2550 output of C<perl -V>, will be sent off to perlbug@perl.org to be
2551 analyzed by the Perl porting team.
2553 If the bug you are reporting has security implications, which make it
2554 inappropriate to send to a publicly archived mailing list, then please send
2555 it to perl5-security-report@perl.org. This points to a closed subscription
2556 unarchived mailing list, which includes all the core committers, who be able
2557 to help assess the impact of issues, figure out a resolution, and help
2558 co-ordinate the release of patches to mitigate or fix the problem across all
2559 platforms on which Perl is supported. Please only use this address for
2560 security issues in the Perl core, not for modules independently
2561 distributed on CPAN.
2565 The F<Changes> file for an explanation of how to view exhaustive details
2568 The F<INSTALL> file for how to build Perl.
2570 The F<README> file for general stuff.
2572 The F<Artistic> and F<Copying> files for copyright information.