1PERL5110DELTA(1) Perl Programmers Reference Guide PERL5110DELTA(1)
2
6 perl5110delta - what is new for perl v5.11.0
7
9 This document describes differences between the 5.10.0 release and the
10 5.11.0 development release.
11
13 Unicode interpretation of \w, \d, \s, and the POSIX character classes
14 redefined.
15 Previous versions of Perl tried to map POSIX style character class
16 definitions onto Unicode property names so that patterns would "dwim"
17 when matches were made against latin-1 or unicode strings. This proved
18 to be a mistake, breaking character class negation, causing forward
19 compatibility problems (as Unicode keeps updating their property
20 definitions and adding new characters), and other problems.
21 Therefore we have now defined a new set of artificial "unicode"
23 property names which will be used to do unicode matching of patterns
24 using POSIX style character classes and perl short-form escape
25 character classes like \w and \d.
26 The key change here is that \d will no longer match every digit in the
28 unicode standard (there are thousands) nor will \w match every word
29 character in the standard, instead they will match precisely their
30 POSIX or Perl definition.
31 Those needing to match based on Unicode properties can continue to do
33 so by using the \p{} syntax to match whichever property they like,
34 including the new artificial definitions.
35 NOTE: This is a backwards incompatible no-warning change in behaviour.
37 If you are upgrading and you process large volumes of text look for
38 POSIX and Perl style character classes and change them to the relevent
39 property name (by removing the word 'Posix' from the current name).
40 The following table maps the POSIX character class names, the escapes
42 and the old and new Unicode property mappings:
43 POSIX Esc Class New-Property ! Old-Property
45 ----------------------------------------------+-------------
46 alnum [0-9A-Za-z] IsPosixAlnum ! IsAlnum
47 alpha [A-Za-z] IsPosixAlpha ! IsAlpha
48 ascii [\000-\177] IsASCII = IsASCII
49 blank [\011 ] IsPosixBlank !
50 cntrl [\0-\37\177] IsPosixCntrl ! IsCntrl
51 digit \d [0-9] IsPosixDigit ! IsDigit
52 graph [!-~] IsPosixGraph ! IsGraph
53 lower [a-z] IsPosixLower ! IsLower
54 print [ -~] IsPosixPrint ! IsPrint
55 punct [!-/:-@[-`{-~] IsPosixPunct ! IsPunct
56 space [\11-\15 ] IsPosixSpace ! IsSpace
57 \s [\11\12\14\15 ] IsPerlSpace ! IsSpacePerl
58 upper [A-Z] IsPosixUpper ! IsUpper
59 word \w [0-9A-Z_a-z] IsPerlWord ! IsWord
60 xdigit [0-9A-Fa-f] IsXDigit = IsXDigit
61 If you wish to build perl with the old mapping you may do so by setting
63 #define PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS 1
65 in regcomp.h, and then setting
67 PERL_TEST_LEGACY_POSIX_CC
69 to true your enviornment when testing.
71 @INC reorganization
73 In @INC, ARCHLIB and PRIVLIB now occur after after the current
74 version's site_perl and vendor_perl.
75 Switch statement changes
77 The handling of complex expressions by the "given"/"when" switch
78 statement has been enhanced. These enhancements are also available in
79 5.10.1 and subsequent 5.10 releases. There are two new cases where
80 "when" now interprets its argument as a boolean, instead of an
81 expression to be used in a smart match:
82 flip-flop operators
84 The ".." and "..." flip-flop operators are now evaluated in boolean
85 context, following their usual semantics; see "Range Operators" in
86 perlop.
87 Note that, as in perl 5.10.0, "when (1..10)" will not work to test
89 whether a given value is an integer between 1 and 10; you should
90 use "when ([1..10])" instead (note the array reference).
91 However, contrary to 5.10.0, evaluating the flip-flop operators in
93 boolean context ensures it can now be useful in a "when()", notably
94 for implementing bistable conditions, like in:
95 when (/^=begin/ .. /^=end/) {
97 # do something
98 }
99 defined-or operator
101 A compound expression involving the defined-or operator, as in
102 "when (expr1 // expr2)", will be treated as boolean if the first
103 expression is boolean. (This just extends the existing rule that
104 applies to the regular or operator, as in "when (expr1 || expr2)".)
105 The next section details more changes brought to the semantics to the
107 smart match operator, that naturally also modify the behaviour of the
108 switch statements where smart matching is implicitly used. These
109 changers were also made for the 5.10.1 release, and will remain in
110 subsequent 5.10 releases.
111 Smart match changes
113 Changes to type-based dispatch
114 The smart match operator "~~" is no longer commutative. The behaviour
116 of a smart match now depends primarily on the type of its right hand
117 argument. Moreover, its semantics have been adjusted for greater
118 consistency or usefulness in several cases. While the general backwards
119 compatibility is maintained, several changes must be noted:
120 · Code references with an empty prototype are no longer treated
122 specially. They are passed an argument like the other code
123 references (even if they choose to ignore it).
124 · "%hash ~~ sub {}" and "@array ~~ sub {}" now test that the
126 subroutine returns a true value for each key of the hash (or
127 element of the array), instead of passing the whole hash or array
128 as a reference to the subroutine.
129 · Due to the commutativity breakage, code references are no longer
131 treated specially when appearing on the left of the "~~" operator,
132 but like any vulgar scalar.
133 · "undef ~~ %hash" is always false (since "undef" can't be a key in a
135 hash). No implicit conversion to "" is done (as was the case in
136 perl 5.10.0).
137 · "$scalar ~~ @array" now always distributes the smart match across
139 the elements of the array. It's true if one element in @array
140 verifies "$scalar ~~ $element". This is a generalization of the old
141 behaviour that tested whether the array contained the scalar.
142 The full dispatch table for the smart match operator is given in "Smart
144 matching in detail" in perlsyn.
145 Smart match and overloading
147 According to the rule of dispatch based on the rightmost argument type,
149 when an object overloading "~~" appears on the right side of the
150 operator, the overload routine will always be called (with a 3rd
151 argument set to a true value, see overload.) However, when the object
152 will appear on the left, the overload routine will be called only when
153 the rightmost argument is a simple scalar. This way distributivity of
154 smart match across arrays is not broken, as well as the other
155 behaviours with complex types (coderefs, hashes, regexes). Thus,
156 writers of overloading routines for smart match mostly need to worry
157 only with comparing against a scalar, and possibly with stringification
158 overloading; the other common cases will be automatically handled
159 consistently.
160 "~~" will now refuse to work on objects that do not overload it (in
162 order to avoid relying on the object's underlying structure). (However,
163 if the object overloads the stringification or the numification
164 operators, and if overload fallback is active, it will be used instead,
165 as usual.)
166 Labels can't be keywords
168 Labels used as targets for the "goto", "last", "next" or "redo"
169 statements cannot be keywords anymore. This restriction will prevent
170 potential confusion between the "goto LABEL" and "goto EXPR" syntaxes:
171 for example, a statement like "goto print" would jump to a label whose
172 name would be the return value of "print()", (usually 1), instead of a
173 label named "print". Moreover, the other control flow statements would
174 just ignore any keyword passed to them as a label name. Since such
175 labels cannot be defined anymore, this kind of error will be avoided.
176 Other incompatible changes
178 · The semantics of "use feature :5.10*" have changed slightly. See
179 "Modules and Pragmata" for more information.
180 · It is now a run-time error to use the smart match operator "~~"
182 with an object that has no overload defined for it. (This way "~~"
183 will not break encapsulation by matching against the object's
184 internal representation as a reference.)
185 · The version control system used for the development of the perl
187 interpreter has been switched from Perforce to git. This is mainly
188 an internal issue that only affects people actively working on the
189 perl core; but it may have minor external visibility, for example
190 in some of details of the output of "perl -V". See perlrepository
191 for more information.
192 · The internal structure of the "ext/" directory in the perl source
194 has been reorganised. In general, a module "Foo::Bar" whose source
195 was stored under ext/Foo/Bar/ is now located under ext/Foo-Bar/.
196 Also, nearly all dual-life modules have been moved from lib/ to
197 ext/. This is purely a source tarball change, and should make no
198 difference to the compilation or installation of perl, unless you
199 have a very customised build process that explicitly relies on this
200 structure, or which hard-codes the "nonxs_ext" Configure parameter.
201 Specifically, this change does not by default alter the location of
202 any files in the final installation.
203 · As part of the "Test::Harness" 2.x to 3.x upgrade, the experimental
205 "Test::Harness::Straps" module has been removed. See "Updated
206 Modules" for more details.
207 · As part of the "ExtUtils::MakeMaker" upgrade, the
209 "ExtUtils::MakeMaker::bytes" and "ExtUtils::MakeMaker::vmsish"
210 modules have been removed from this distribution.
211 · "Module::CoreList" no longer contains the %:patchlevel hash.
213 · This one is actually a change introduced in 5.10.0, but it was
215 missed from that release's perldelta, so it is mentioned here
216 instead.
217 A bugfix related to the handling of the "/m" modifier and "qr"
219 resulted in a change of behaviour between 5.8.x and 5.10.0:
220 # matches in 5.8.x, doesn't match in 5.10.0
222 $re = qr/^bar/; "foo\nbar" =~ /$re/m;
223 · "length undef" now returns undef.
225 · Unsupported private C API functions are now declared "static" to
227 prevent leakage to Perl's public API.
228 · To support the bootstrapping process, miniperl no longer builds
230 with UTF-8 support in the regexp engine.
231 This allows a build to complete with PERL_UNICODE set and a UTF-8
233 locale. Without this there's a bootstrapping problem, as miniperl
234 can't load the UTF-8 components of the regexp engine, because
235 they're not yet built.
236 · miniperl's @INC is now restricted to just -I..., the split of
238 $ENV{PERL5LIB}, and "."
239 · A space or a newline is now required after a "#line XXX" directive.
241 · Tied filehandles now have an additional method EOF which provides
243 the EOF type
244 · To better match all other flow control statements, "foreach" may no
246 longer be used as an attribute.
247
249 Unicode Character Database 5.1.0
250 The copy of the Unicode Character Database included in Perl 5.11.0 has
251 been updated to 5.1.0 from 5.0.0. See
252 <http://www.unicode.org/versions/Unicode5.1.0/#Notable_Changes> for the
253 notable changes.
254 A proper interface for pluggable Method Resolution Orders
256 As of Perl 5.11.0 there is a new interface for plugging and using
257 method resolution orders other than the default (linear depth first
258 search). The C3 method resolution order added in 5.10.0 has been re-
259 implemented as a plugin, without changing its Perl-space interface. See
260 perlmroapi for more information.
261 The "overloading" pragma
263 This pragma allows you to lexically disable or enable overloading for
264 some or all operations. (Yuval Kogman)
265 "\N" regex escape
267 A new regex escape has been added, "\N". It will match any character
268 that is not a newline, independently from the presence or absence of
269 the single line match modifier "/s". (If "\N" is followed by an opening
270 brace and by a letter, perl will still assume that a Unicode character
271 name is coming, so compatibility is preserved.) (Rafael Garcia-Suarez)
272 Implicit strictures
274 Using the "use VERSION" syntax with a version number greater or equal
275 to 5.11.0 will also lexically enable strictures just like "use strict"
276 would do (in addition to enabling features.) So, the following:
277 use 5.11.0;
279 will now imply:
281 use strict;
283 use feature ':5.11';
284 Parallel tests
286 The core distribution can now run its regression tests in parallel on
287 Unix-like platforms. Instead of running "make test", set "TEST_JOBS" in
288 your environment to the number of tests to run in parallel, and run
289 "make test_harness". On a Bourne-like shell, this can be done as
290 TEST_JOBS=3 make test_harness # Run 3 tests in parallel
292 An environment variable is used, rather than parallel make itself,
294 because TAP::Harness needs to be able to schedule individual non-
295 conflicting test scripts itself, and there is no standard interface to
296 "make" utilities to interact with their job schedulers.
297 Note that currently some test scripts may fail when run in parallel
299 (most notably "ext/IO/t/io_dir.t"). If necessary run just the failing
300 scripts again sequentially and see if the failures go away.
301 The "..." operator
303 A new operator, "...", nicknamed the Yada Yada operator, has been
304 added. It is intended to mark placeholder code, that is not yet
305 implemented. See "Yada Yada Operator" in perlop. (chromatic)
306 DTrace support
308 Some support for DTrace has been added. See "DTrace support" in
309 INSTALL.
310 Support for "configure_requires" in CPAN module metadata
312 Both "CPAN" and "CPANPLUS" now support the "configure_requires" keyword
313 in the META.yml metadata file included in most recent CPAN
314 distributions. This allows distribution authors to specify
315 configuration prerequisites that must be installed before running
316 Makefile.PL or Build.PL.
317 See the documentation for "ExtUtils::MakeMaker" or "Module::Build" for
319 more on how to specify "configure_requires" when creating a
320 distribution for CPAN.
321 "each" is now more flexible
323 The "each" function can now operate on arrays.
324 Y2038 compliance
326 Perl's core time-related functions are now Y2038 compliant. (With 29
327 years to spare!)
328 $, flexibility
330 The variable $, may now be tied.
331 // in where clauses
333 // now behaves like || in when clauses
334 Enabling warnings from your shell environment
336 You can now set "-W" from the "PERL5OPT" environment variable
337 "delete local"
339 "delete local" now allows you to locally delete a hash entry.
340 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()
346 system call.
347
349 Dual-lifed modules moved
350 Dual-lifed modules maintained primarily in the Perl core now live in
351 dist/. Dual-lifed modules maintained primarily on CPAN now live in
352 cpan/
353 In previous releases of Perl, it was customary to enumerate all module
355 changes in this section of the "perldelta" file. From 5.11.0 forward
356 only notable updates (such as new or deprecated modules ) will be
357 listed in this section. For a complete reference to the versions of
358 modules shipped in a given release of perl, please see
359 Module::CoreList.
360 New Modules and Pragmata
362 "autodie"
363 This is a new lexically-scoped alternative for the "Fatal" module.
364 The bundled version is 2.06_01. Note that in this release, using a
365 string eval when "autodie" is in effect can cause the autodie
366 behaviour to leak into the surrounding scope. See "BUGS" in autodie
367 for more details.
368 "Compress::Raw::Bzip2"
370 This has been added to the core (version 2.020).
371 "parent"
373 This pragma establishes an ISA relationship with base classes at
374 compile time. It provides the key feature of "base" without the
375 feature creep.
376 "Parse::CPAN::Meta"
378 This has been added to the core (version 1.39).
379 Pragmata Changes
381 "overloading"
382 See "The "overloading" pragma" above.
383 "attrs"
385 The "attrs" pragma has been removed. It had been marked as
386 deprecated since 5.6.0.
387 "charnames"
389 The Unicode NameAliases.txt database file has been added. This has
390 the effect of adding some extra "\N" character names that formerly
391 wouldn't have been recognised; for example, "\N{LATIN CAPITAL
392 LETTER GHA}".
393 "feature"
395 The meaning of the ":5.10" and ":5.10.X" feature bundles has
396 changed slightly. The last component, if any (i.e. "X") is simply
397 ignored. This is predicated on the assumption that new features
398 will not, in general, be added to maintenance releases. So ":5.10"
399 and ":5.10.X" have identical effect. This is a change to the
400 behaviour documented for 5.10.0.
401 "mro"
403 Upgraded from version 1.00 to 1.01. Performance for single
404 inheritance is 40% faster - see "Performance Enhancements" below.
405 "mro" is now implemented as an XS extension. The documented
407 interface has not changed. Code relying on the implementation
408 detail that some "mro::" methods happened to be available at all
409 times gets to "keep both pieces".
410 Updated Modules
412 "ExtUtils::MakeMaker"
413 Upgraded from version 6.42 to 6.55_02.
414 Note that "ExtUtils::MakeMaker::bytes" and
416 "ExtUtils::MakeMaker::vmsish" have been removed from this
417 distribution.
418 "Test::Harness"
420 Upgraded from version 2.64 to 3.17.
421 Note that one side-effect of the 2.x to 3.x upgrade is that the
423 experimental "Test::Harness::Straps" module (and its supporting
424 "Assert", "Iterator", "Point" and "Results" modules) have been
425 removed. If you still need this, then they are available in the
426 (unmaintained) "Test-Harness-Straps" distribution on CPAN.
427 "UNIVERSAL"
429 Upgraded from version 1.04 to 1.05.
430 "UNIVERSAL->import()" is now deprecated.
432
434 h2ph
435 Now looks in "include-fixed" too, which is a recent addition to
436 gcc's search path.
437 h2xs
439 No longer incorrectly treats enum values like macros (Daniel Burr).
440 Now handles C++ style constants ("//") properly in enums. (A patch
442 from Rainer Weikusat was used; Daniel Burr also proposed a similar
443 fix).
444 perl5db.pl
446 "LVALUE" subroutines now work under the debugger.
447 The debugger now correctly handles proxy constant subroutines, and
449 subroutine stubs.
450 perlbug
452 perlbug now uses %Module::CoreList::bug_tracker to print out
453 upstream bug tracker URLs.
454 Where the user names a module that their bug report is about, and
456 we know the URL for its upstream bug tracker, provide a message to
457 the user explaining that the core copies the CPAN version directly,
458 and provide the URL for reporting the bug directly to upstream.
459 perlthanks
461 Perl 5.11.0 added a new utility perlthanks, which is a variant of
462 perlbug, but for sending non-bug-reports to the authors and
463 maintainers of Perl. Getting nothing but bug reports can become a
464 bit demoralising: we'll see if this changes things.
465
467 perlhaiku
468 This contains instructions on how to build perl for the Haiku
469 platform.
470 perlmroapi
472 This describes the new interface for pluggable Method Resolution
473 Orders.
474 perlperf
476 This document, by Richard Foley, provides an introduction to the
477 use of performance and optimization techniques which can be used
478 with particular reference to perl programs.
479 perlrepository
481 This describes how to access the perl source using the git version
482 control system.
483
485 The various large Changes* files (which listed every change made to
486 perl over the last 18 years) have been removed, and replaced by a small
487 file, also called Changes, which just explains how that same
488 information may be extracted from the git version control system.
489 The file Porting/patching.pod has been deleted, as it mainly described
491 interacting with the old Perforce-based repository, which is now
492 obsolete. Information still relevant has been moved to perlrepository.
493 perlapi, perlintern, perlmodlib and perltoc are now all generated at
495 build time, rather than being shipped as part of the release.
496 · Documented -X overloading.
498 · Documented that "when()" treats specially most of the filetest
500 operators
501 · Documented when as a syntax modifier
503 · Eliminated "Old Perl threads tutorial", which describes 5005
505 threads.
506 pod/perlthrtut.pod is the same material reworked for ithreads.
508 · Correct previous documentation: v-strings are not deprecated
510 With version objects, we need them to use MODULE VERSION syntax.
512 This patch removes the deprecation note.
513 · Added security contact information to perlsec
515
517 · A new internal cache means that "isa()" will often be faster.
518 · The implementation of "C3" Method Resolution Order has been
520 optimised - linearisation for classes with single inheritance is
521 40% faster. Performance for multiple inheritance is unchanged.
522 · Under "use locale", the locale-relevant information is now cached
524 on read-only values, such as the list returned by "keys %hash".
525 This makes operations such as "sort keys %hash" in the scope of
526 "use locale" much faster.
527 · Empty "DESTROY" methods are no longer called.
529 · Faster "Perl_sv_utf8_upgrade()"
531 · Speed up "keys" on empty hash
533
535 ext/ reorganisation
536 The layout of directories in ext has been revised. Specifically, all
537 extensions are now flat, and at the top level, with "/" in pathnames
538 replaced by "-", so that ext/Data/Dumper/ is now ext/Data-Dumper/, etc.
539 The names of the extensions as specified to Configure, and as reported
540 by %Config::Config under the keys "dynamic_ext", "known_extensions",
541 "nonxs_ext" and "static_ext" have not changed, and still use "/". Hence
542 this change will not have any affect once perl is installed. "Safe" has
543 been split out from being part of "Opcode", and "mro" is now an
544 extension in its own right.
545 Nearly all dual-life modules have been moved from lib to ext, and will
547 now appear as known "nonxs_ext". This will made no difference to the
548 structure of an installed perl, nor will the modules installed differ,
549 unless you run Configure with options to specify an exact list of
550 extensions to build. In this case, you will rapidly become aware that
551 you need to add to your list, because various modules needed to
552 complete the build, such as "ExtUtils::ParseXS", have now become
553 extensions, and without them the build will fail well before it
554 attempts to run the regression tests.
555 Configuration improvements
557 If "vendorlib" and "vendorarch" are the same, then they are only added
558 to @INC once.
559 $Config{usedevel} and the C-level "PERL_USE_DEVEL" are now defined if
561 perl is built with "-Dusedevel".
562 Configure will enable use of "-fstack-protector", to provide protection
564 against stack-smashing attacks, if the compiler supports it.
565 Configure will now determine the correct prototypes for re-entrant
567 functions, and for "gconvert", if you are using a C++ compiler rather
568 than a C compiler.
569 On Unix, if you build from a tree containing a git repository, the
571 configuration process will note the commit hash you have checked out,
572 for display in the output of "perl -v" and "perl -V". Unpushed local
573 commits are automatically added to the list of local patches displayed
574 by "perl -V".
575 Compilation improvements
577 As part of the flattening of ext, all extensions on all platforms are
578 built by make_ext.pl. This replaces the Unix-specific
579 ext/util/make_ext, VMS-specific make_ext.com and Win32-specific
580 win32/buildext.pl.
581 Platform Specific Changes
583 AIX Removed libbsd for AIX 5L and 6.1. Only "flock()" was used from
584 libbsd.
585 Removed libgdbm for AIX 5L and 6.1. The libgdbm is delivered as an
587 optional package with the AIX Toolbox. Unfortunately the 64 bit
588 version is broken.
589 Hints changes mean that AIX 4.2 should work again.
591 Cygwin
593 On Cygwin we now strip the last number from the DLL. This has been
594 the behaviour in the cygwin.com build for years. The hints files
595 have been updated.
596 DomainOS
598 Support for Apollo DomainOS was removed in Perl 5.11.0
599 FreeBSD
601 The hints files now identify the correct threading libraries on
602 FreeBSD 7 and later.
603 Irix
605 We now work around a bizarre preprocessor bug in the Irix 6.5
606 compiler: "cc -E -" unfortunately goes into K&R mode, but "cc -E
607 file.c" doesn't.
608 Haiku
610 Patches from the Haiku maintainers have been merged in. Perl should
611 now build on Haiku.
612 MachTen
614 Support for Tenon Intersystems MachTen Unix layer for MacOS Classic
615 was removed in Perl 5.11.0
616 MiNT
618 Support for Atari MiNT was removed in Perl 5.11.0.
619 MirOS BSD
621 Perl should now build on MirOS BSD.
622 NetBSD
624 Hints now supports versions 5.*.
625 Stratus VOS
627 Various changes from Stratus have been merged in.
628 Symbian
630 There is now support for Symbian S60 3.2 SDK and S60 5.0 SDK.
631 Win32
633 Improved message window handling means that "alarm" and "kill"
634 messages will no longer be dropped under race conditions.
635 VMS Reads from the in-memory temporary files of "PerlIO::scalar" used
637 to fail if $/ was set to a numeric reference (to indicate record-
638 style reads). This is now fixed.
639 VMS now supports "getgrgid".
641 Many improvements and cleanups have been made to the VMS file name
643 handling and conversion code.
644 Enabling the "PERL_VMS_POSIX_EXIT" logical name now encodes a POSIX
646 exit status in a VMS condition value for better interaction with
647 GNV's bash shell and other utilities that depend on POSIX exit
648 values. See "$?" in perlvms for details.
649 "File::Copy" now detects Unix compatibility mode on VMS.
651
653 · "-I" on shebang line now adds directories in front of @INC as
654 documented, and as does "-I" when specified on the command-line.
655 · "kill" is now fatal when called on non-numeric process identifiers.
657 Previously, an 'undef' process identifier would be interpreted as a
658 request to kill process "0", which would terminate the current
659 process group on POSIX systems. Since process identifiers are
660 always integers, killing a non-numeric process is now fatal.
661 · 5.10.0 inadvertently disabled an optimisation, which caused a
663 measurable performance drop in list assignment, such as is often
664 used to assign function parameters from @_. The optimisation has
665 been re-instated, and the performance regression fixed.
666 · Fixed memory leak on "while (1) { map 1, 1 }" [RT #53038].
668 · Some potential coredumps in PerlIO fixed [RT #57322,54828].
670 · The debugger now works with lvalue subroutines.
672 · The debugger's "m" command was broken on modules that defined
674 constants [RT #61222].
675 · "crypt" and string complement could return tainted values for
677 untainted arguments [RT #59998].
678 · The "-i".suffix command-line switch now recreates the file using
680 restricted permissions, before changing its mode to match the
681 original file. This eliminates a potential race condition [RT
682 #60904].
683 · On some Unix systems, the value in $? would not have the top bit
685 set ("$? & 128") even if the child core dumped.
686 · Under some circumstances, $^R could incorrectly become undefined
688 [RT #57042].
689 · In the XS API, various hash functions, when passed a pre-computed
691 hash where the key is UTF-8, might result in an incorrect lookup.
692 · XS code including XSUB.h before perl.h gave a compile-time error
694 [RT #57176].
695 · "$object->isa('Foo')" would report false if the package "Foo"
697 didn't exist, even if the object's @ISA contained "Foo".
698 · Various bugs in the new-to 5.10.0 mro code, triggered by
700 manipulating @ISA, have been found and fixed.
701 · Bitwise operations on references could crash the interpreter, e.g.
703 "$x=\$y; $x |= "foo"" [RT #54956].
704 · Patterns including alternation might be sensitive to the internal
706 UTF-8 representation, e.g.
707 my $byte = chr(192);
709 my $utf8 = chr(192); utf8::upgrade($utf8);
710 $utf8 =~ /$byte|X}/i; # failed in 5.10.0
711 · Within UTF8-encoded Perl source files (i.e. where "use utf8" is in
713 effect), double-quoted literal strings could be corrupted where a
714 "\xNN", "\0NNN" or "\N{}" is followed by a literal character with
715 ordinal value greater than 255 [RT #59908].
716 · "B::Deparse" failed to correctly deparse various constructs:
718 "readpipe STRING" [RT #62428], "CORE::require(STRING)" [RT #62488],
719 "sub foo(_)" [RT #62484].
720 · Using "setpgrp" with no arguments could corrupt the perl stack.
722 · The block form of "eval" is now specifically trappable by "Safe"
724 and "ops". Previously it was erroneously treated like string
725 "eval".
726 · In 5.10.0, the two characters "[~" were sometimes parsed as the
728 smart match operator ("~~") [RT #63854].
729 · In 5.10.0, the "*" quantifier in patterns was sometimes treated as
731 "{0,32767}" [RT #60034, #60464]. For example, this match would
732 fail:
733 ("ab" x 32768) =~ /^(ab)*$/
735 · "shmget" was limited to a 32 bit segment size on a 64 bit OS [RT
737 #63924].
738 · Using "next" or "last" to exit a "given" block no longer produces a
740 spurious warning like the following:
741 Exiting given via last at foo.pl line 123
743 · On Windows, '.\foo' and '..\foo' were treated differently than
745 './foo' and '../foo' by "do" and "require" [RT #63492].
746 · Assigning a format to a glob could corrupt the format; e.g.:
748 *bar=*foo{FORMAT}; # foo format now bad
750 · Attempting to coerce a typeglob to a string or number could cause
752 an assertion failure. The correct error message is now generated,
753 "Can't coerce GLOB to $type".
754 · Under "use filetest 'access'", "-x" was using the wrong access
756 mode. This has been fixed [RT #49003].
757 · "length" on a tied scalar that returned a Unicode value would not
759 be correct the first time. This has been fixed.
760 · Using an array "tie" inside in array "tie" could SEGV. This has
762 been fixed. [RT #51636]
763 · A race condition inside "PerlIOStdio_close()" has been identified
765 and fixed. This used to cause various threading issues, including
766 SEGVs.
767 · In "unpack", the use of "()" groups in scalar context was
769 internally placing a list on the interpreter's stack, which
770 manifested in various ways, including SEGVs. This is now fixed [RT
771 #50256].
772 · Magic was called twice in "substr", "\&$x", "tie $x, $m" and
774 "chop". These have all been fixed.
775 · A 5.10.0 optimisation to clear the temporary stack within the
777 implicit loop of "s///ge" has been reverted, as it turned out to be
778 the cause of obscure bugs in seemingly unrelated parts of the
779 interpreter [commit ef0d4e17921ee3de].
780 · The line numbers for warnings inside "elsif" are now correct.
782 · The ".." operator now works correctly with ranges whose ends are at
784 or close to the values of the smallest and largest integers.
785 · "binmode STDIN, ':raw'" could lead to segmentation faults on some
787 platforms. This has been fixed [RT #54828].
788 · An off-by-one error meant that "index $str, ..." was effectively
790 being executed as "index "$str\0", ...". This has been fixed [RT
791 #53746].
792 · Various leaks associated with named captures in regexes have been
794 fixed [RT #57024].
795 · A weak reference to a hash would leak. This was affecting "DBI" [RT
797 #56908].
798 · Using (?|) in a regex could cause a segfault [RT #59734].
800 · Use of a UTF-8 "tr//" within a closure could cause a segfault [RT
802 #61520].
803 · Calling "Perl_sv_chop()" or otherwise upgrading an SV could result
805 in an unaligned 64-bit access on the SPARC architecture [RT
806 #60574].
807 · In the 5.10.0 release, "inc_version_list" would incorrectly list
809 "5.10.*" after "5.8.*"; this affected the @INC search order [RT
810 #67628].
811 · In 5.10.0, "pack "a*", $tainted_value" returned a non-tainted value
813 [RT #52552].
814 · In 5.10.0, "printf" and "sprintf" could produce the fatal error
816 "panic: utf8_mg_pos_cache_update" when printing UTF-8 strings [RT
817 #62666].
818 · In the 5.10.0 release, a dynamically created "AUTOLOAD" method
820 might be missed (method cache issue) [RT #60220,60232].
821 · In the 5.10.0 release, a combination of "use feature" and "//ee"
823 could cause a memory leak [RT #63110].
824 · "-C" on the shebang ("#!") line is once more permitted if it is
826 also specified on the command line. "-C" on the shebang line used
827 to be a silent no-op if it was not also on the command line, so
828 perl 5.10.0 disallowed it, which broke some scripts. Now perl
829 checks whether it is also on the command line and only dies if it
830 is not [RT #67880].
831 · In 5.10.0, certain types of re-entrant regular expression could
833 crash, or cause the following assertion failure [RT #60508]:
834 Assertion rx->sublen >= (s - rx->subbeg) + i failed
836 · Previously missing files from Unicode 5.1 Character Database are
838 now included.
839 · "TMPDIR" is now honored when opening an anonymous temporary file
841
843 "panic: sv_chop %s"
844 This new fatal error occurs when the C routine "Perl_sv_chop()" was
845 passed a position that is not within the scalar's string buffer.
846 This could be caused by buggy XS code, and at this point recovery
847 is not possible.
848 "Can't locate package %s for the parents of %s"
850 This warning has been removed. In general, it only got produced in
851 conjunction with other warnings, and removing it allowed an ISA
852 lookup optimisation to be added.
853 "v-string in use/require is non-portable"
855 This warning has been removed.
856 "Deep recursion on subroutine "%s""
858 It is now possible to change the depth threshold for this warning
859 from the default of 100, by recompiling the perl binary, setting
860 the C pre-processor macro "PERL_SUB_DEPTH_WARN" to the desired
861 value.
862
864 · TODO: "SVt_RV" is gone. RVs are now stored in IVs
865 · TODO: REGEXPs are first class
867 · TODO: OOK is reworked, such that an OOKed scalar is PV not PVIV
869 · The J.R.R. Tolkien quotes at the head of C source file have been
871 checked and proper citations added, thanks to a patch from Tom
872 Christiansen.
873 · "Perl_vcroak()" now accepts a null first argument. In addition, a
875 full audit was made of the "not NULL" compiler annotations, and
876 those for several other internal functions were corrected.
877 · New macros "dSAVEDERRNO", "dSAVE_ERRNO", "SAVE_ERRNO",
879 "RESTORE_ERRNO" have been added to formalise the temporary saving
880 of the "errno" variable.
881 · The function "Perl_sv_insert_flags" has been added to augment
883 "Perl_sv_insert".
884 · The function "Perl_newSV_type(type)" has been added, equivalent to
886 "Perl_newSV()" followed by "Perl_sv_upgrade(type)".
887 · The function "Perl_newSVpvn_flags()" has been added, equivalent to
889 "Perl_newSVpvn()" and then performing the action relevant to the
890 flag.
891 Two flag bits are currently supported.
893 "SVf_UTF8"
895 This will call "SvUTF8_on()" for you. (Note that this does not
896 convert an sequence of ISO 8859-1 characters to UTF-8). A
897 wrapper, "newSVpvn_utf8()" is available for this.
898 "SVs_TEMP"
900 Call "Perl_sv_2mortal()" on the new SV.
901 There is also a wrapper that takes constant strings,
903 "newSVpvs_flags()".
904 · The function "Perl_croak_xs_usage" has been added as a wrapper to
906 "Perl_croak".
907 · The functions "PerlIO_find_layer" and "PerlIO_list_alloc" are now
909 exported.
910 · "PL_na" has been exterminated from the core code, replaced by local
912 STRLEN temporaries, or "*_nolen()" calls. Either approach is faster
913 than "PL_na", which is a pointer deference into the interpreter
914 structure under ithreads, and a global variable otherwise.
915 · "Perl_mg_free()" used to leave freed memory accessible via
917 "SvMAGIC()" on the scalar. It now updates the linked list to remove
918 each piece of magic as it is freed.
919 · Under ithreads, the regex in "PL_reg_curpm" is now reference
921 counted. This eliminates a lot of hackish workarounds to cope with
922 it not being reference counted.
923 · "Perl_mg_magical()" would sometimes incorrectly turn on
925 "SvRMAGICAL()". This has been fixed.
926 · The public IV and NV flags are now not set if the string value has
928 trailing "garbage". This behaviour is consistent with not setting
929 the public IV or NV flags if the value is out of range for the
930 type.
931 · SV allocation tracing has been added to the diagnostics enabled by
933 "-Dm". The tracing can alternatively output via the "PERL_MEM_LOG"
934 mechanism, if that was enabled when the perl binary was compiled.
935 · Smartmatch resolution tracing has been added as a new diagnostic.
937 Use "-DM" to enable it.
938 · A new debugging flag "-DB" now dumps subroutine definitions,
940 leaving "-Dx" for its original purpose of dumping syntax trees.
941 · Uses of "Nullav", "Nullcv", "Nullhv", "Nullop", "Nullsv" etc have
943 been replaced by "NULL" in the core code, and non-dual-life
944 modules, as "NULL" is clearer to those unfamiliar with the core
945 code.
946 · A macro MUTABLE_PTR(p) has been added, which on (non-pedantic) gcc
948 will not cast away "const", returning a "void *". Macros
949 "MUTABLE_SV(av)", "MUTABLE_SV(cv)" etc build on this, casting to
950 "AV *" etc without casting away "const". This allows proper
951 compile-time auditing of "const" correctness in the core, and
952 helped picked up some errors (now fixed).
953 · Macros "mPUSHs()" and "mXPUSHs()" have been added, for pushing SVs
955 on the stack and mortalizing them.
956 · Use of the private structure "mro_meta" has changed slightly.
958 Nothing outside the core should be accessing this directly anyway.
959 · A new tool, Porting/expand-macro.pl has been added, that allows you
961 to view how a C preprocessor macro would be expanded when compiled.
962 This is handy when trying to decode the macro hell that is the perl
963 guts.
964
966 Many modules updated from CPAN incorporate new tests.
967 Several tests that have the potential to hang forever if they fail now
969 incorporate a "watchdog" functionality that will kill them after a
970 timeout, which helps ensure that "make test" and "make test_harness"
971 run to completion automatically. (Jerry Hedden).
972 Some core-specific tests have been added:
974 t/comp/retainedlines.t
976 Check that the debugger can retain source lines from "eval".
977 t/io/perlio_fail.t
979 Check that bad layers fail.
980 t/io/perlio_leaks.t
982 Check that PerlIO layers are not leaking.
983 t/io/perlio_open.t
985 Check that certain special forms of open work.
986 t/io/perlio.t
988 General PerlIO tests.
989 t/io/pvbm.t
991 Check that there is no unexpected interaction between the internal
992 types "PVBM" and "PVGV".
993 t/mro/package_aliases.t
995 Check that mro works properly in the presence of aliased packages.
996 t/op/dbm.t
998 Tests for "dbmopen" and "dbmclose".
999 t/op/index_thr.t
1001 Tests for the interaction of "index" and threads.
1002 t/op/pat_thr.t
1004 Tests for the interaction of esoteric patterns and threads.
1005 t/op/qr_gc.t
1007 Test that "qr" doesn't leak.
1008 t/op/reg_email_thr.t
1010 Tests for the interaction of regex recursion and threads.
1011 t/op/regexp_qr_embed_thr.t
1013 Tests for the interaction of patterns with embedded "qr//" and
1014 threads.
1015 t/op/regexp_unicode_prop.t
1017 Tests for Unicode properties in regular expressions.
1018 t/op/regexp_unicode_prop_thr.t
1020 Tests for the interaction of Unicode properties and threads.
1021 t/op/reg_nc_tie.t
1023 Test the tied methods of "Tie::Hash::NamedCapture".
1024 t/op/reg_posixcc.t
1026 Check that POSIX character classes behave consistently.
1027 t/op/re.t
1029 Check that exportable "re" functions in universal.c work.
1030 t/op/setpgrpstack.t
1032 Check that "setpgrp" works.
1033 t/op/substr_thr.t
1035 Tests for the interaction of "substr" and threads.
1036 t/op/upgrade.t
1038 Check that upgrading and assigning scalars works.
1039 t/uni/lex_utf8.t
1041 Check that Unicode in the lexer works.
1042 t/uni/tie.t
1044 Check that Unicode and "tie" work.
1045
1047 This is a list of some significant unfixed bugs, which are regressions
1048 from either 5.10.0 or 5.8.x.
1049 · "List::Util::first" misbehaves in the presence of a lexical $_
1051 (typically introduced by "my $_" or implicitly by "given"). The
1052 variable which gets set for each iteration is the package variable
1053 $_, not the lexical $_ [RT #67694].
1054 A similar issue may occur in other modules that provide functions
1056 which take a block as their first argument, like
1057 foo { ... $_ ...} list
1059 · The "charnames" pragma may generate a run-time error when a regex
1061 is interpolated [RT #56444]:
1062 use charnames ':full';
1064 my $r1 = qr/\N{THAI CHARACTER SARA I}/;
1065 "foo" =~ $r1; # okay
1066 "foo" =~ /$r1+/; # runtime error
1067 A workaround is to generate the character outside of the regex:
1069 my $a = "\N{THAI CHARACTER SARA I}";
1071 my $r1 = qr/$a/;
1072 · Some regexes may run much more slowly when run in a child thread
1074 compared with the thread the pattern was compiled into [RT #55600].
1075
1077 The following items are now deprecated.
1078 · "Switch" is buggy and should be avoided. From perl 5.11.0 onwards,
1080 it is intended that any use of the core version of this module will
1081 emit a warning, and that the module will eventually be removed from
1082 the core (probably in perl 5.14.0). See "Switch statements" in
1083 perlsyn for its replacement.
1084 · The following modules will be removed from the core distribution in
1086 a future release, and should be installed from CPAN instead.
1087 Distributions on CPAN which require these should add them to their
1088 prerequisites. The core versions of these modules warnings will
1089 issue a deprecation warning.
1090 · "Class::ISA"
1092 · "Pod::Plainer"
1094 · "Shell"
1096 Currently support to install from CPAN without a force is "TODO" in
1098 CPAN and CPANPLUS. This will be addressed before 5.12.0 ships.
1099 · "suidperl" has been removed. It used to provide a mechanism to
1101 emulate setuid permission bits on systems that don't support it
1102 properly.
1103 · Deprecate assignment to $[
1105 · Remove attrs, which has been deprecated since 1999/10/02.
1107 · Deprecate use of the attribute :locked on subroutines.
1109 · Deprecate using "locked" with the attributes pragma.
1111 · Deprecate using "unique" with the attributes pragma.
1113 · warn if ++ or -- are unable to change the value because it's beyond
1115 the limit of representation
1116 This uses a new warnings category: "imprecision".
1118 · Make lc/uc/lcfirst/ucfirst warn when passed undef.
1120 · Show constant in "Useless use of a constant in void context"
1122 · Make the new warning report undef constants as undef
1124 · Add a new warning, "Prototype after '%s'"
1126 · Tweak the "Illegal character in prototype" warning so it's more
1128 precise when reporting illegal characters after _
1129 · Unintended interpolation of $\ in regex
1131 · Make overflow warnings in gmtime/localtime only occur when warnings
1133 are on
1134 · Improve mro merging error messages.
1136 They are now very similar to those produced by Algorithm::C3.
1138 · Amelioration of the error message "Unrecognized character %s in
1140 column %d"
1141 Changes the error message to "Unrecognized character %s; marked by
1143 <-- HERE after %s<-- HERE near column %d". This should make it a
1144 little simpler to spot and correct the suspicious character.
1145 · Explicitely point to $. when it causes an uninitialized warning for
1147 ranges in scalar context
1148 · Deprecated numerous Perl 4-era libraries:
1150 termcap.pl, tainted.pl, stat.pl, shellwords.pl, pwd.pl, open3.pl,
1152 open2.pl, newgetopt.pl, look.pl, find.pl, finddepth.pl,
1153 importenv.pl, hostname.pl, getopts.pl, getopt.pl, getcwd.pl,
1154 flush.pl, fastcwd.pl, exceptions.pl, ctime.pl, complete.pl,
1155 cacheout.pl, bigrat.pl, bigint.pl, bigfloat.pl, assert.pl,
1156 abbrev.pl, dotsh.pl, and timelocal.pl are all now deprecated. Using
1157 them will incur a warning.
1158
1160 Some of the work in this release was funded by a TPF grant funded by
1161 Dijkmat BV, The Netherlands.
1162 Steffen Mueller and David Golden in particular helped getting CPAN
1164 modules polished and synchronised with their in-core equivalents.
1165 Craig Berry was tireless in getting maint to run under VMS, no matter
1167 how many times we broke it for him.
1168 The other core committers contributed most of the changes, and applied
1170 most of the patches sent in by the hundreds of contributors listed in
1171 AUTHORS.
1172 Much of the work of categorizing changes in this perldelta file was
1174 contributed by the following porters using
1175 changelogger.bestpractical.com:
1176 Nicholas Clark, leon, shawn, alexm, rjbs, rafl, Pedro Melo, brunorc,
1178 anonymous, X, Tom Hukins, anonymous, Jesse, dagolden, Moritz Onken,
1179 Mark Fowler, chorny, anonymous, tmtm
1180 Finally, thanks to Larry Wall, without whom none of this would be
1182 necessary.
1183
1185 If you find what you think is a bug, you might check the articles
1186 recently posted to the comp.lang.perl.misc newsgroup and the perl bug
1187 database at http://rt.perl.org/perlbug/ . There may also be
1188 information at http://www.perl.org/ , the Perl Home Page.
1189 If you believe you have an unreported bug, please run the perlbug
1191 program included with your release. Be sure to trim your bug down to a
1192 tiny but sufficient test case. Your bug report, along with the output
1193 of "perl -V", will be sent off to perlbug@perl.org to be analysed by
1194 the Perl porting team.
1195 If the bug you are reporting has security implications, which make it
1197 inappropriate to send to a publicly archived mailing list, then please
1198 send it to perl5-security-report@perl.org. This points to a closed
1199 subscription unarchived mailing list, which includes all the core
1200 committers, who be able to help assess the impact of issues, figure out
1201 a resolution, and help co-ordinate the release of patches to mitigate
1202 or fix the problem across all platforms on which Perl is supported.
1203 Please only use this address for security issues in the Perl core, not
1204 for modules independently distributed on CPAN.
1205
1207 The Changes file for an explanation of how to view exhaustive details
1208 on what changed.
1209 The INSTALL file for how to build Perl.
1211 The README file for general stuff.
1213 The Artistic and Copying files for copyright information.
1215perl v5.12.4 2011-06-07 PERL5110DELTA(1)