3 perldelta - what is new for perl v5.8.0
7 This document describes differences between the 5.6.0 release and the
10 =head1 Incompatible Changes
12 =head2 64-bit platforms and malloc
14 If your pointers are 64 bits wide, the Perl malloc is no longer being
15 used because it does not work well with 8-byte pointers. Also,
16 usually the system mallocs on such platforms are much better optimized
17 for such large memory models than the Perl malloc. Some memory-hungry
18 Perl applications like the PDL don't work well with Perl's malloc.
19 Finally, other applications than Perl (like modperl) tend to prefer
20 the system malloc. Such platforms include Alpha and 64-bit HPPA,
23 =head2 AIX Dynaloading
25 The AIX dynaloading now uses in AIX releases 4.3 and newer the native
26 dlopen interface of AIX instead of the old emulated interface. This
27 change will probably break backward compatibility with compiled
28 modules. The change was made to make Perl more compliant with other
29 applications like modperl which are using the AIX native interface.
31 =head2 Socket Extension Dynamic in VMS
33 The Socket extension is now dynamically loaded instead of being
34 statically built in. This may or may not be a problem with ancient
35 TCP/IP stacks of VMS: we do not know since we weren't able to test
36 Perl in such configurations.
38 =head2 Different Definition of the Unicode Character Classes \p{In...}
40 As suggested by the Unicode consortium, the Unicode character classes
41 now prefer I<scripts> as opposed to I<blocks> (as defined by Unicode);
42 in Perl, when the C<\p{In....}> and the C<\p{In....}> regular expression
43 constructs are used. This has changed the definition of some of those
46 The difference between scripts and blocks is that scripts are the
47 glyphs used by a language or a group of languages, while the blocks
48 are more artificial groupings of 256 characters based on the Unicode
51 In general this change results in more inclusive Unicode character
52 classes, but changes to the other direction also do take place:
53 for example while the script C<Latin> includes all the Latin
54 characters and their various diacritic-adorned versions, it
55 does not include the various punctuation or digits (since they
56 are not solely C<Latin>).
58 Changes in the character class semantics may have happened if a script
59 and a block happen to have the same name, for example C<Hebrew>.
60 In such cases the script wins and C<\p{InHebrew}> now means the script
61 definition of Hebrew. The block definition in still available,
62 though, by appending C<Block> to the name: C<\p{InHebrewBlock}> means
63 what C<\p{InHebrew}> meant in perl 5.6.0. For the full list
64 of affected character classes, see L<perlunicode/Blocks>.
66 =head2 Perl Parser Stress Tested
68 The Perl parser has been stress tested using both random input and
69 Markov chain input and the few found crashes and lockups have been
72 =head2 REF(...) Instead Of SCALAR(...)
74 A reference to a reference now stringifies as "REF(0x81485ec)" instead
75 of "SCALAR(0x81485ec)" in order to be more consistent with the return
84 The semantics of bless(REF, REF) were unclear and until someone proves
85 it to make some sense, it is forbidden.
89 The obsolete chat2 library that should never have been allowed
90 to escape the laboratory has been decommissioned.
94 The very dusty examples in the eg/ directory have been removed.
95 Suggestions for new shiny examples welcome but the main issue is that
96 the examples need to be documented, tested and (most importantly)
101 The (bogus) escape sequences \8 and \9 now give an optional warning
102 ("Unrecognized escape passed through"). There is no need to \-escape
107 The list of filenames from glob() (or <...>) is now by default sorted
108 alphabetically to be csh-compliant. (bsd_glob() does still sort platform
109 natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.)
113 Although "you shouldn't do that", it was possible to write code that
114 depends on Perl's hashed key order (Data::Dumper does this). The new
115 algorithm "One-at-a-Time" produces a different hashed key order.
116 More details are in L</"Performance Enhancements">.
120 lstat(FILEHANDLE) now gives a warning because the operation makes no sense.
121 In future releases this may become a fatal error.
125 The C<package;> syntax (C<package> without an argument) has been
126 deprecated. Its semantics were never that clear and its
127 implementation even less so. If you have used that feature to
128 disallow all but fully qualified variables, C<use strict;> instead.
132 The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still
133 recognised but now cause fatal errors. The previous behaviour of
134 ignoring them by default and warning if requested was unacceptable
135 since it, in a way, falsely promised that the features could be used.
139 The current user-visible implementation of pseudo-hashes (the weird
140 use of the first array element) is deprecated starting from Perl 5.8.0
141 and will be removed in Perl 5.10.0, and the feature will be
142 implemented differently. Not only is the current interface rather
143 ugly, but the current implementation slows down normal array and hash
144 use quite noticeably. The C<fields> pragma interface will remain
149 The syntaxes C<< @a->[...] >> and C<< %h->{...}>> have now been deprecated.
153 After years of trying the suidperl is considered to be too complex to
154 ever be considered truly secure. The suidperl functionality is likely
155 to be removed in a future release.
159 The long deprecated uppercase aliases for the string comparison
160 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
164 The tr///C and tr///U features have been removed and will not return;
165 the interface was a mistake. Sorry about that. For similar
166 functionality, see pack('U0', ...) and pack('C0', ...).
170 =head1 Core Enhancements
172 =head2 PerlIO is Now The Default
178 IO is now by default done via PerlIO rather than system's "stdio".
179 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
180 handle's behaviour. Layers can be specified at open time via 3-arg
183 open($fh,'>:crlf :utf8', $path) || ...
185 or on already opened handles via extended C<binmode>:
187 binmode($fh,':encoding(iso-8859-7)');
189 The built-in layers are: unix (low level read/write), stdio (as in
190 previous Perls), perlio (re-implementation of stdio buffering in a
191 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
192 but available on any platform). A mmap layer may be available if
193 platform supports it (mostly UNIXes).
195 Layers to be applied by default may be specified via the 'open' pragma.
197 See L</"Installation and Configuration Improvements"> for the effects
198 of PerlIO on your architecture name.
202 File handles can be marked as accepting Perl's internal encoding of Unicode
203 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
205 open($fh,">:utf8","Uni.txt");
207 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
208 for you since it's not UTF-8 what you will be getting but instead
209 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
210 http://www.unicode.org/unicode/reports/tr16/ for more information.
211 In future releases this naming may change.
215 File handles can translate character encodings from/to Perl's internal
216 Unicode form on read/write via the ":encoding()" layer.
220 File handles can be opened to "in memory" files held in Perl scalars via:
222 open($fh,'>', \$variable) || ...
226 Anonymous temporary files are available without need to
227 'use FileHandle' or other module via
229 open($fh,"+>", undef) || ...
231 That is a literal undef, not an undefined value.
235 The list form of C<open> is now implemented for pipes (at least on UNIX):
237 open($fh,"-|", 'cat', '/etc/motd')
239 creates a pipe, and runs the equivalent of exec('cat', '/etc/motd') in
244 =head2 Signals Are Now Safe
246 Perl used to be fragile in that signals arriving at inopportune moments
247 could corrupt Perl's internal state. Now Perl postpones handling of
248 signals until it's safe.
250 =head2 Unicode Overhaul
252 Unicode in general should be now much more usable than in Perl 5.6.0
253 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
254 regular expressions should work now, Unicode in tr/// should work now,
255 Unicode in I/O should work now.
261 The Unicode Character Database coming with Perl has been upgraded
262 to Unicode 3.1.1. For more information, see http://www.unicode.org/.
266 For developers interested in enhancing Perl's Unicode capabilities:
267 almost all the UCD files are included with the Perl distribution in
268 the lib/unicore subdirectory. The most notable omission, for space
269 considerations, is the Unihan database.
273 The Unicode character classes \p{Blank} and \p{SpacePerl} have been
274 added. "Blank" is like C isblank(), that is, it contains only
275 "horizontal whitespace" (the space character is, the newline isn't),
276 and the "SpacePerl" is the Unicode equivalent of C<\s> (\p{Space}
277 isn't, since that includes the vertical tabulator character, whereas
282 =head2 Understanding of Numbers
284 In general a lot of fixing has happened in the area of Perl's
285 understanding of numbers, both integer and floating point. Since in
286 many systems the standard number parsing functions like C<strtoul()>
287 and C<atof()> seem to have bugs, Perl tries to work around their
288 deficiencies. This results hopefully in more accurate numbers.
290 Perl now tries internally to use integer values in numeric conversions
291 and basic arithmetics (+ - * /) if the arguments are integers, and
292 tries also to keep the results stored internally as integers.
293 This change leads to often slightly faster and always less lossy
294 arithmetics. (Previously Perl always preferred floating point numbers
297 =head2 Miscellaneous Enhancements
303 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
304 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
308 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
309 in multiple arguments.)
313 END blocks are now run even if you exit/die in a BEGIN block.
314 Internally, the execution of END blocks is now controlled by
315 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
316 behaviour for Perl embedders. This will default in 5.10. See
321 Formats now support zero-padded decimal fields.
325 Lvalue subroutines can now return C<undef> in list context.
329 A new special regular expression variable has been introduced:
330 C<$^N>, which contains the most-recently closed group (submatch).
334 C<no Module;> now works even if there is no "sub unimport" in the Module.
338 The numerical comparison operators return C<undef> if either operand
339 is a NaN. Previously the behaviour was unspecified.
343 The following builtin functions are now overridable: each(), keys(),
344 pop(), push(), shift(), splice(), unshift().
348 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
352 my __PACKAGE__ $obj now works.
356 The printf() and sprintf() now support parameter reordering using the
357 C<%\d+\$> and C<*\d+\$> syntaxes. For example
359 print "%2\$s %1\$s\n", "foo", "bar";
361 will print "bar foo\n"; This feature helps in writing
362 internationalised software.
366 prototype(\&) is now available.
370 prototype(\[$@%&]) is now available to implicitly create references
371 (useful for example if you want to emulate the tie() interface).
375 UNTIE method is now recognised.
379 L<utime> now supports C<utime undef, undef, @files> to change the
380 file timestamps to the current time.
384 The rules for allowing underscores (underbars) in numeric constants
385 have been relaxed and simplified: now you can have an underscore
386 simply B<between digits>.
390 =head1 Modules and Pragmata
392 =head2 New Modules and Pragmata
398 C<Attribute::Handlers> allows a class to define attribute handlers.
401 use Attribute::Handlers;
402 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
404 # later, in some package using or inheriting from MyPack...
406 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
408 Both variables and routines can have attribute handlers. Handlers can
409 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
410 exact compilation phase (BEGIN, CHECK, INIT, or END).
414 B<B::Concise> is a new compiler backend for walking the Perl syntax
415 tree, printing concise info about ops, from Stephen McCamant. The
416 output is highly customisable. See L<B::Concise>.
420 C<Class::ISA> for reporting the search path for a class's ISA tree,
421 by Sean Burke, has been added. See L<Class::ISA>.
425 C<Cwd> has now a split personality: if possible, an XS extension is
426 used, (this will hopefully be faster, more secure, and more robust)
427 but if not possible, the familiar Perl implementation is used.
431 C<Devel::PPPort>, originally from Kenneth Albanowski and now
432 maintained by Paul Marquess, has been added. It is primarily used
433 by C<h2xs> to enhance portability of of XS modules between different
438 C<Digest>, frontend module for calculating digests (checksums), from
439 Gisle Aas, has been added. See L<Digest>.
443 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
444 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
446 use Digest::MD5 'md5_hex';
448 $digest = md5_hex("Thirsty Camel");
450 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
452 NOTE: the C<MD5> backward compatibility module is deliberately not
453 included since its further use is discouraged.
457 C<Encode>, by Nick Ing-Simmons, provides a mechanism to translate
458 between different character encodings. Support for Unicode,
459 ISO-8859-*, ASCII, CP*, KOI8-R, and three variants of EBCDIC are
460 compiled in to the module. Several other encodings (like Japanese,
461 Chinese, and MacIntosh encodings) are included and will be loaded at
462 runtime. See L<Encode>.
464 Any encoding supported by Encode module is also available to the
465 ":encoding()" layer if PerlIO is used.
469 C<I18N::Langinfo> can be use to query locale information.
470 See L<I18N::Langinfo>.
474 C<I18N::LangTags> has functions for dealing with RFC3066-style
475 language tags, by Sean Burke. See L<I18N::LangTags>.
479 C<ExtUtils::Constant> is a new tool for extension writers for
480 generating XS code to import C header constants, by Nicholas Clark.
481 See L<ExtUtils::Constant>.
485 C<Filter::Simple> is an easy-to-use frontend to Filter::Util::Call,
486 from Damian Conway. See L<Filter::Simple>.
492 use Filter::Simple sub {
493 while (my ($from, $to) = splice @_, 0, 2) {
502 use MyFilter qr/red/ => 'green';
504 print "red\n"; # this code is filtered, will print "green\n"
505 print "bored\n"; # this code is filtered, will print "bogreen\n"
509 print "red\n"; # this code is not filtered, will print "red\n"
513 C<File::Temp> allows one to create temporary files and directories in
514 an easy, portable, and secure way, by Tim Jenness. See L<File::Temp>.
518 C<Filter::Util::Call> provides you with the framework to write
519 I<Source Filters> in Perl, from Paul Marquess. For most uses the
520 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
524 L<libnet> is a collection of perl5 modules related to network
525 programming, from Graham Barr. See L<Net::FTP>, L<Net::NNTP>,
526 L<Net::Ping>, L<Net::POP3>, L<Net::SMTP>, and L<Net::Time>.
528 Perl installation leaves libnet unconfigured, use F<libnetcfg> to configure.
532 C<List::Util> is a selection of general-utility list subroutines, like
533 sum(), min(), first(), and shuffle(), by Graham Barr. See L<List::Util>.
537 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>, and
538 C<Locale::Language>, from Neil Bowers, have been added. They provide the
539 codes for various locale standards, such as "fr" for France, "usd" for
540 US Dollar, and "jp" for Japanese.
544 $country = code2country('jp'); # $country gets 'Japan'
545 $code = country2code('Norway'); # $code gets 'no'
547 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
548 and L<Locale::Language>.
552 C<Locale::Maketext> is localization framework from Sean Burke. See
553 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
554 article about software localization, originally published in The Perl
555 Journal #13, republished here with kind permission.
559 C<Memoize> can make your functions faster by trading space for time,
560 from Mark-Jason Dominus. See L<Memoize>.
564 C<MIME::Base64> allows you to encode data in base64, from Gisle Aas,
565 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
570 $encoded = encode_base64('Aladdin:open sesame');
571 $decoded = decode_base64($encoded);
573 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
579 C<MIME::QuotedPrint> allows you to encode data in quoted-printable
580 encoding, as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
581 Extensions)>, from Gisle Aas.
583 use MIME::QuotedPrint;
585 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
586 $decoded = decode_qp($encoded);
588 print $encoded, "\n"; # "Smiley in Unicode: =263A"
590 MIME::QuotedPrint has been enhanced to provide the basic methods
591 necessary to use it with PerlIO::Via as in :
593 use MIME::QuotedPrint;
594 open($fh,">Via(MIME::QuotedPrint)",$path);
596 See L<MIME::QuotedPrint>.
600 C<NEXT> is pseudo-class for method redispatch, from Damian Conway.
605 C<open> is a new pragma for setting the default I/O disciplines
610 C<PerlIO::Scalar> provides the implementation of IO to "in memory"
611 Perl scalars as discussed above, from Nick Ing-Simmons. It also
612 serves as an example of a loadable PerlIO layer. Other future
613 possibilities include PerlIO::Array and PerlIO::Code.
614 See L<PerlIO::Scalar>.
618 C<PerlIO::Via> acts as a PerlIO layer and wraps PerlIO layer
619 functionality provided by a class (typically implemented in perl
620 code), from Nick Ing-Simmons.
622 use MIME::QuotedPrint;
623 open($fh,">Via(MIME::QuotedPrint)",$path);
625 This will automatically convert everything output to C<$fh>
626 to Quoted-Printable. See L<PerlIO::Via>.
630 C<Pod::ParseLink>, by Russ Allbery, has been added,
631 to parse L<> links in pods as described in the new
636 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
637 It converts POD data to formatted overstrike text.
638 See L<Pod::Text::Overstrike>.
642 C<Scalar::Util> is a selection of general-utility scalar subroutines,
643 like blessed(), reftype(), and tainted(). See L<Scalar::Util>.
647 C<sort> is a new pragma for controlling the behaviour of sort().
651 C<Storable> gives persistence to Perl data structures by allowing the
652 storage and retrieval of Perl data to and from files in a fast and
653 compact binary format, from Raphael Manfredi. See L<Storable>.
657 C<Switch>, from Damian Conway, has been added. Just by saying
661 you have C<switch> and C<case> available in Perl.
667 case 1 { print "number 1" }
668 case "a" { print "string a" }
669 case [1..10,42] { print "number in list" }
670 case (@array) { print "number in list" }
671 case /\w+/ { print "pattern" }
672 case qr/\w+/ { print "pattern" }
673 case (%hash) { print "entry in hash" }
674 case (\%hash) { print "entry in hash" }
675 case (\&sub) { print "arg to subroutine" }
676 else { print "previous case not true" }
683 C<Test::More> is yet another framework for writing test scripts,
684 more extensive than Test::Simple, by Michael Schwern. See L<Test::More>.
688 C<Test::Simple> has the- basic utilities for writing tests, by Michael
689 Schwern. See L<Test::Simple>.
693 C<Text::Balanced> has been added, for extracting delimited text
694 sequences from strings, from Damian Conway.
696 use Text::Balanced 'extract_delimited';
698 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
700 $a will be "'never say never'", $b will be ', he never said'.
702 In addition to extract_delimited() there are also extract_bracketed(),
703 extract_quotelike(), extract_codeblock(), extract_variable(),
704 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
705 gen_extract_tagged(). With these you can implement rather advanced
706 parsing algorithms. See L<Text::Balanced>.
710 C<threads> is an interface to interpreter threads, by Arthur Bergman.
711 Interpreter threads (ithreads) is the new thread model introduced in
712 Perl 5.6 but only available as an internal interface for extension
713 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>.
717 C<threads::shared> allows data sharing for interpreter threads, from
718 Arthur Bergman. In the ithreads model any data sharing between
719 threads must be explicit, as opposed to the old 5.005 thread model
720 where data sharing was implicit. See L<threads::shared>.
724 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
725 references (unlike the standard Tie::RefHash) The module is contained
726 within Tie::RefHash, see L<Tie::RefHash>.
730 C<Time::HiRes> provides high resolution timing (ualarm, usleep,
731 and gettimeofday), from Douglas E. Wegscheid. See L<Time::HiRes>.
735 C<Unicode::UCD> offers a querying interface to the Unicode Character
736 Database. See L<Unicode::UCD>.
740 C<Unicode::Collate> implements the UCA (Unicode Collation Algorithm)
741 for sorting Unicode strings, by SADAHIRO Tomoyuki. See L<Unicode::Collate>.
745 C<Unicode::Normalize> implements the various Unicode normalization
746 forms, by SADAHIRO Tomoyuki. See L<Unicode::Normalize>.
750 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises XS
751 typemaps. Nothing gets installed but for extension writers the code
756 =head2 Updated And Improved Modules and Pragmata
762 The following independently supported modules have been updated to the
763 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
764 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
765 (Pod::Man, Pod::Text), Pod::LaTeX, Pod::Parser, Storable,
766 Term::ANSIColor, Test, Text-Tabs+Wrap.
770 The attributes::reftype() now works on tied arguments.
774 AutoLoader can now be disabled with C<no AutoLoader;>.
778 B::Deparse has been significantly enhanced. It now can deparse almost
779 all of the standard test suite (so that the tests still succeed).
780 There is a make target "test.deparse" for trying this out.
784 Class::Struct can now define the classes in compile time.
788 Class::Struct now assigns the array/hash element if the accessor
789 is called with an array/hash element as the B<sole> argument.
793 Data::Dumper has now an option to sort hashes.
797 Data::Dumper has now an option to dump code references
802 The English module can now be used without the infamous performance
805 use English '-no_performance_hit';
807 (Assuming, of course, that one doesn't need the troublesome variables
808 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
809 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
813 Fcntl, Socket, and Sys::Syslog have been rewritten to use the
814 new-style constant dispatch section (see L<ExtUtils::Constant>).
815 This means that they will be more robust and hopefully faster.
819 File::Find now has pre- and post-processing callbacks. It also
820 correctly changes directories when chasing symbolic links. Callbacks
821 (naughtily) exiting with "next;" instead of "return;" now work.
825 File::Find is now (again) reentrant. It also has been made
830 File::Glob::glob() renamed to File::Glob::bsd_glob() to avoid
831 prototype mismatch with CORE::glob().
835 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
836 the returned list of filenames.
840 Devel::Peek now has an interface for the Perl memory statistics
841 (this works only if you are using perl's malloc, and if you have
842 compiled with debugging).
846 IPC::Open3 now allows the use of numeric file descriptors.
850 IO::Socket has now atmark() method, which returns true if the socket
851 is positioned at the out-of-band mark. The method is also exportable
852 as a sockatmark() function.
856 IO::Socket::INET has support for ReusePort option (if your platform
857 supports it). The Reuse option now has an alias, ReuseAddr. For clarity
858 you may want to prefer ReuseAddr.
862 IO::Socket::INET now supports C<LocalPort> of zero (usually meaning
863 that the operating system will make one up.)
867 use lib now works identically to @INC. Removing directories
868 with 'no lib' now works.
872 Math::BigFloat and Math::BigInt have undergone a full rewrite.
873 They are now magnitudes faster, and they support various
874 bignum libraries such as GMP and PARI as their backends.
878 Net::Ping has been enhanced. There is now "external" protocol which
879 uses Net::Ping::External module which runs external ping(1) and parses
880 the output. An alpha version of Net::Ping::External is available in
881 CPAN and in 5.7.2 the Net::Ping::External may be integrated to Perl.
885 C<POSIX::sigaction()> is now much more flexible and robust.
886 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
887 handlers, installing new handlers was not atomic.
891 C<%INC> now localised in a Safe compartment so that use/require work.
895 The Shell module now has an OO interface.
899 The Test module has been significantly enhanced.
903 The C<vars> pragma now supports declaring fully qualified variables.
904 (Something that C<our()> does not and will not support.)
908 The utf8:: name space (as in the pragma) provides various
909 Perl-callable functions to provide low level access to Perl's
910 internal Unicode representation. At the moment only length()
911 has been implemented.
915 =head1 Utility Changes
921 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
926 F<emacs/e2ctags.pl> is now much faster.
930 C<h2ph> now supports C trigraphs.
934 C<h2xs> now produces a template README.
938 C<h2xs> now uses C<Devel::PPort> for better portability between
939 different versions of Perl.
943 C<h2xs> uses the new L<ExtUtils::Constant> module which will affect
944 newly created extensions that define constants. Since the new code is
945 more correct (if you have two constants where the first one is a
946 prefix of the second one, the first constant B<never> gets defined),
947 less lossy (it uses integers for integer constant, as opposed to the
948 old code that used floating point numbers even for integer constants),
949 and slightly faster, you might want to consider regenerating your
950 extension code (the new scheme makes regenerating easy).
951 L<h2xs> now also supports C trigraphs.
955 C<libnetcfg> has been added to configure the libnet.
959 C<perlbug> is now much more robust. It also sends the bug report to
960 perl.org, not perl.com.
964 C<perlcc> has been rewritten and its user interface (that is,
965 command line) is much more like that of the UNIX C compiler, cc.
969 C<perlivp> is a new utility for doing Installation Verification
970 Procedure after installing Perl.
974 C<pod2html> now allows specifying a cache directory.
978 C<s2p> has been completely rewritten in Perl. (It is in fact a full
979 implementation of sed in Perl: you can use the sed functionality by
980 using the C<psed> utility.)
984 C<xsubpp> now understands POD documentation embedded in the *.xs files.
988 C<xsubpp> now supports OUT keyword.
992 =head1 New Documentation
998 perl56delta details the changes between the 5.005 release and the
1003 perlclib documents the internal replacements for standard C library
1004 functions. (Interesting only for extension writers and Perl core
1009 perldebtut is a Perl debugging tutorial.
1013 perlebcdic contains considerations for running Perl on EBCDIC platforms.
1014 Note that unfortunately EBCDIC platforms that used to be supported back
1015 in Perl 5.005 are still unsupported by Perl 5.8.0; the plan, however, is
1016 to bring them back to the fold.
1020 perlintro is a gentle introduction to Perl.
1024 perliol documents the internals of PerlIO with layers.
1028 perlmodstyle is a style guide for writing modules.
1032 perlnewmod tells about writing and submitting a new module.
1036 perlpod has been rewritten to be clearer and to record the best
1037 practices gathered over the years.
1041 perlpodspec is a more formal specification of the pod format,
1042 mainly of interest for writers of pod applications, not to
1043 people writing in pod.
1047 perlposix-bc explains using Perl on the POSIX-BC platform
1048 (an EBCDIC mainframe platform).
1052 perlretut is a regular expression tutorial.
1056 perlrequick is a regular expressions quick-start guide.
1057 Yes, much quicker than perlretut.
1061 perltodo has been updated.
1065 perltootc has been renamed as perltooc (to not to conflict
1066 with perltoot in filesystems restricted to "8.3" names)
1070 perluniintro is an introduction to using Unicode in Perl
1071 (perlunicode is more of a reference)
1075 perlutil explains the command line utilities packaged with the Perl
1080 The following platform-specific documents are available before
1081 the installation as README.I<platform>, and after the installation
1084 perlaix perlamiga perlapollo perlbeos perlbs2000
1085 perlce perlcygwin perldgux perldos perlepoc perlhpux
1086 perlhurd perlmachten perlmacos perlmint perlmpeix
1087 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1088 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1094 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1095 confusion with the Perl POSIX module.
1099 The documentation for the WinCE platform is called "CE", to avoid
1100 confusion with the perlwin32 documentation on 8.3-restricted filesystems.
1104 =head1 Performance Enhancements
1110 map() that changes the size of the list should now work faster.
1114 sort() has been changed to use primarily mergesort internally as
1115 opposed to the earlier quicksort. For very small lists this may
1116 result in slightly slower sorting times, but in general the speedup
1117 should be at least 20%. Additional bonuses are that the worst case
1118 behaviour of sort() is now better (in computer science terms it now
1119 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1120 worst-case run time behaviour), and that sort() is now stable
1121 (meaning that elements with identical keys will stay ordered as they
1122 were before the sort). See the C<sort> pragma for information.
1124 The story in more detail: suppose you want to serve yourself a little
1127 @digits = ( 3,1,4,1,5,9 );
1129 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1130 Which C<1> comes first is hard to know, since one C<1> looks pretty
1131 much like any other. You can regard this as totally trivial,
1132 or somewhat profound. However, if you just want to sort the even
1133 digits ahead of the odd ones, then what will
1135 sort { ($a % 2) <=> ($b % 2) } @digits;
1137 yield? The only even digit, C<4>, will come first. But how about
1138 the odd numbers, which all compare equal? With the quicksort algorithm
1139 used to implement Perl 5.6 and earlier, the order of ties is left up
1140 to the sort. So, as you add more and more digits of Pi, the order
1141 in which the sorted even and odd digits appear will change.
1142 and, for sufficiently large slices of Pi, the quicksort algorithm
1143 in Perl 5.8 won't return the same results even if reinvoked with the
1144 same input. The justification for this rests with quicksort's
1145 worst case behavior. If you run
1147 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1149 (something you might approximate if you wanted to merge two sorted
1150 arrays using sort), doubling $N doesn't just double the quicksort time,
1151 it I<quadruples> it. Quicksort has a worst case run time that can
1152 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1153 on patterns that may well arise in normal use. You won't notice this
1154 for small arrays, but you I<will> notice it with larger arrays,
1155 and you may not live long enough for the sort to complete on arrays
1156 of a million elements. So the 5.8 quicksort scrambles large arrays
1157 before sorting them, as a statistical defence against quadratic behaviour.
1158 But that means if you sort the same large array twice, ties may be
1159 broken in different ways.
1161 Because of the unpredictability of tie-breaking order, and the quadratic
1162 worst-case behaviour, quicksort was I<almost> replaced completely with
1163 a stable mergesort. I<Stable> means that ties are broken to preserve
1164 the original order of appearance in the input array. So
1166 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1168 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1169 appear in the output in the same order they appeared in the input.
1170 Mergesort has worst case O(NlogN) behaviour, the best value
1171 attainable. And, ironically, this mergesort does particularly
1172 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1173 in O(N) time. But quicksort was rescued at the last moment because
1174 it is faster than mergesort on certain inputs and platforms.
1175 For example, if you really I<don't> care about the order of even
1176 and odd digits, quicksort will run in O(N) time; it's very good
1177 at sorting many repetitions of a small number of distinct elements.
1178 The quicksort divide and conquer strategy works well on platforms
1179 with relatively small, very fast, caches. Eventually, the problem gets
1180 whittled down to one that fits in the cache, from which point it
1181 benefits from the increased memory speed.
1183 Quicksort was rescued by implementing a sort pragma to control aspects
1184 of the sort. The B<stable> subpragma forces stable behaviour,
1185 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1186 subpragmas are heavy-handed ways to select the underlying implementation.
1187 The leading C<_> is a reminder that these subpragmas may not survive
1188 beyond 5.8. More appropriate mechanisms for selecting the implementation
1189 exist, but they wouldn't have arrived in time to save quicksort.
1193 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1194 (http://burtleburtle.net/bob/hash/doobs.html). This algorithm is
1195 reasonably fast while producing a much better spread of values than
1196 the old hashing algorithm (originally by Chris Torek, later tweaked by
1197 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1198 all 3-char printable ASCII keys comes much closer to passing the
1199 DIEHARD random number generation tests. According to perlbench, this
1200 change has not affected the overall speed of Perl.
1204 unshift() should now be noticeably faster.
1208 =head1 Installation and Configuration Improvements
1210 =head2 Generic Improvements
1216 INSTALL now explains how you can configure Perl to use 64-bit
1217 integers even on non-64-bit platforms.
1221 Policy.sh policy change: if you are reusing a Policy.sh file
1222 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1223 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1224 them will now be changed to the new prefix, /foo/bar. (Previously
1225 only $prefix changed.) If you do not like this new behaviour,
1226 specify prefix, siteprefix, and vendorprefix explicitly.
1230 A new optional location for Perl libraries, otherlibdirs, is available.
1231 It can be used for example for vendor add-ons without disturbing Perl's
1232 own library directories.
1236 In many platforms the vendor-supplied 'cc' is too stripped-down to
1237 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1238 to be the case and 'cc' does not seem to be the GNU C compiler
1239 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1243 gcc needs to closely track the operating system release to avoid
1244 build problems. If Configure finds that gcc was built for a different
1245 operating system release than is running, it now gives a clearly visible
1246 warning that there may be trouble ahead.
1250 If binary compatibility with the 5.005 release is not wanted, Configure
1251 no longer suggests including the 5.005 modules in @INC.
1255 Configure C<-S> can now run non-interactively.
1259 configure.gnu now works with options with whitespace in them.
1263 installperl now outputs everything to STDERR.
1267 $Config{byteorder} is now computed dynamically (this is more robust
1268 with "fat binaries" where an executable image contains binaries for
1269 more than one binary platform.)
1273 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1274 get appended to the $Config{archname} (also known as $^O) anymore.
1275 Instead, if you explicitly choose not to use perlio (Configure command
1276 line option -Uuseperlio), you will get "-stdio" appended.
1280 Another change related to the architecture name is that "-64all"
1281 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1282 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1286 In AFS installations one can configure the root of the AFS to be
1287 somewhere else than the default F</afs> by using the Configure
1288 parameter C<-Dafsroot=/some/where/else>.
1292 APPLLIB_EXP, a less-know configuration-time definition, has been
1293 documented. It can be used to prepend site-specific directories
1294 to Perl's default search path (@INC), see INSTALL for information.
1298 The version of Berkeley DB used when the Perl (and, presumably, the
1299 DB_File extension) was built is now available as
1300 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1301 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1302 DB_VERSION_PATCH_CFG> from C.
1306 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1307 has been documented in INSTALL.
1311 If you have CPAN access (either network or a local copy such as a
1312 CD-ROM) you can during specify extra modules to Configure to build and
1313 install with Perl using the -Dextras=... option. See INSTALL for
1318 In addition to config.over a new override file, config.arch, is
1319 available. That is supposed to be used by hints file writers for
1320 architecture-wide changes (as opposed to config.over which is for
1325 If your file system supports symbolic links you can build Perl outside
1326 of the source directory by
1328 mkdir /tmp/perl/build/directory
1329 cd /tmp/perl/build/directory
1330 sh /path/to/perl/source/Configure -Dmksymlinks ...
1332 This will create in /tmp/perl/build/directory a tree of symbolic links
1333 pointing to files in /path/to/perl/source. The original files are left
1334 unaffected. After Configure has finished you can just say
1338 and Perl will be built and tested, all in /tmp/perl/build/directory.
1342 For Perl developers several new make targets for profiling
1343 and debugging have been added, see L<perlhack>.
1349 Use of the F<gprof> tool to profile Perl has been documented in
1350 L<perlhack>. There is a make target called "perl.gprof" for
1351 generating a gprofiled Perl executable.
1355 If you have GCC 3, there is a make target called "perl.gcov" for
1356 creating a gcoved Perl executable for coverage analysis. See
1361 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1362 have been added, see L<perlhack> for more information about pixie and
1369 Guidelines of how to construct minimal Perl installations have
1370 been added to INSTALL.
1374 The Thread extension is now not built at all under ithreads
1375 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1376 Thread extension requires being Configured with C<-Duse5005threads>).
1378 But note that the Thread.pm interface is now shared by both
1383 =head2 New Or Improved Platforms
1385 For the list of platforms known to support Perl,
1386 see L<perlport/"Supported Platforms">.
1392 AIX dynamic loading should be now better supported.
1396 AIX should now work better with gcc, threads, and 64-bitness. Also the
1397 long doubles support in AIX should be better now. See L<perlaix>.
1401 After a long pause, AmigaOS has been verified to be happy with Perl.
1405 AtheOS (http://www.atheos.cx/) is a new platform.
1409 DG/UX platform now supports the 5.005-style threads. See L<perldgux>.
1413 DYNIX/ptx platform (a.k.a. dynixptx) is supported at or near osvers 4.5.2.
1417 EBCDIC platforms (z/OS, also known as OS/390, POSIX-BC, and VM/ESA)
1418 have been regained. Many test suite tests still fail and the
1419 co-existence of Unicode and EBCDIC isn't quite settled, but the
1420 situation is much better than with Perl 5.6. See L<perlos390>,
1421 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
1425 Building perl with -Duseithreads or -Duse5005threads now works under
1426 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
1427 need a thread library package installed. See README.hpux.
1431 MacOS Classic (MacPerl has of course been available since
1432 perl 5.004 but now the source code bases of standard Perl
1433 and MacPerl have been synchronised)
1437 MacOS X (or Darwin) should now be able to build Perl even on HFS+
1438 filesystems. (The case-insensitivity confused the Perl build process.)
1442 NCR MP-RAS is now supported.
1446 NetWare from Novell is now supported. See L<perlnetware>.
1450 NonStop-UX is now supported.
1454 Amdahl UTS UNIX mainframe platform is now supported.
1458 WinCE is now supported. See L<perlce>.
1462 z/OS (formerly known as OS/390, formerly known as MVS OE) has now
1463 support for dynamic loading. This is not selected by default,
1464 however, you must specify -Dusedl in the arguments of Configure.
1468 =head1 Selected Bug Fixes
1470 Numerous memory leaks and uninitialized memory accesses have been
1471 hunted down. Most importantly anonymous subs used to leak quite
1478 The autouse pragma didn't work for Multi::Part::Function::Names.
1482 chop(@list) in list context returned the characters chopped in
1483 reverse order. This has been reversed to be in the right order.
1487 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
1488 when building the Perl binary. The only exception to this is SunOS 4.x,
1493 The behaviour of non-decimal but numeric string constants such as
1494 "0x23" was platform-dependent: in some platforms that was seen as 35,
1495 in some as 0, in some as a floating point number (don't ask). This
1496 was caused by Perl using the operating system libraries in a situation
1497 where the result of the string to number conversion is undefined: now
1498 Perl consistently handles such strings as zero in numeric contexts.
1502 The order of DESTROYs has been made more predictable.
1506 Several debugger fixes: exit code now reflects the script exit code,
1507 condition C<"0"> now treated correctly, the C<d> command now checks
1508 line number, the C<$.> no longer gets corrupted, all debugger output now
1509 goes correctly to the socket if RemotePort is set.
1513 L<dprofpp> -R didn't work.
1517 C<*foo{FORMAT}> now works.
1521 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
1522 the Tk extension with 5.6.0.)
1526 Lexicals I: lexicals outside an eval "" weren't resolved
1527 correctly inside a subroutine definition inside the eval "" if they
1528 were not already referenced in the top level of the eval""ed code.
1532 Lexicals II: lexicals leaked at file scope into subroutines that
1533 were declared before the lexicals.
1537 Lexical warnings now propagating correctly between scopes.
1541 Line renumbering with eval and C<#line> now works.
1545 Fixed numerous memory leaks, especially in eval "".
1549 mkdir() now ignores trailing slashes in the directory name,
1550 as mandated by POSIX.
1554 Some versions of glibc have a broken modfl(). This affects builds
1555 with C<-Duselongdouble>. This version of Perl detects this brokenness
1556 and has a workaround for it. The glibc release 2.2.2 is known to have
1557 fixed the modfl() bug.
1561 Modulus of unsigned numbers now works (4063328477 % 65535 used to
1562 return 27406, instead of 27047).
1566 Some "not a number" warnings introduced in 5.6.0 eliminated to be
1567 more compatible with 5.005. Infinity is now recognised as a number.
1571 Attributes (like :shared) didn't work with our().
1575 our() variables will not cause "will not stay shared" warnings.
1579 pack "Z" now correctly terminates the string with "\0".
1583 Fix password routines which in some shadow password platforms
1584 (e.g. HP-UX) caused getpwent() to return every other entry.
1588 The PERL5OPT environment variable (for passing command line arguments
1589 to Perl) didn't work for more than a single group of options.
1593 PERL5OPT with embedded spaces didn't work.
1597 printf() no longer resets the numeric locale to "C".
1601 C<q(a\\b)> now parses correctly as C<'a\\b'>.
1605 Printing quads (64-bit integers) with printf/sprintf now works
1606 without the q L ll prefixes (assuming you are on a quad-capable platform).
1610 Regular expressions on references and overloaded scalars now work.
1614 Right-hand side magic (GMAGIC) could in many cases such as string
1615 concatenation be invoked too many times.
1619 scalar() now forces scalar context even when used in void context.
1623 SOCKS support is now much more robust.
1627 sort() arguments are now compiled in the right wantarray context
1628 (they were accidentally using the context of the sort() itself).
1632 Changed the POSIX character class C<[[:space:]]> to include the (very
1633 rarely used) vertical tab character. Added a new POSIX-ish character
1634 class C<[[:blank:]]> which stands for horizontal whitespace
1635 (currently, the space and the tab).
1639 The tainting behaviour of sprintf() has been rationalized. It does
1640 not taint the result of floating point formats anymore, making the
1641 behaviour consistent with that of string interpolation.
1645 The regular expression captured submatches ($1, $2, ...) are now
1646 more consistently unset if the match fails, instead of leaving false
1647 data lying around in them.
1651 C<Sys::Syslog> ignored the C<LOG_AUTH> constant.
1655 All but the first argument of the IO syswrite() method are now optional.
1659 $AUTOLOAD, sort(), lock(), and spawning subprocesses
1660 in multiple threads simultaneously are now thread-safe.
1664 Tie::ARRAY SPLICE method was broken.
1668 Allow read-only string on left hand side of non-modifying tr///.
1672 Several Unicode fixes.
1678 BOMs (byte order marks) in the beginning of Perl files
1679 (scripts, modules) should now be transparently skipped.
1680 UTF-16 (UCS-2) encoded Perl files should now be read correctly.
1684 The character tables have been updated to Unicode 3.1.1.
1688 Comparing with utf8 data does not magically upgrade non-utf8 data
1693 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
1697 Concatenation with the C<.> operator or via variable interpolation,
1698 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
1699 substitution with C<s///>, single-quoted UTF8, should now work.
1703 The C<tr///> operator now works. Note that the C<tr///CU>
1704 functionality has been removed (but see pack('U0', ...)).
1708 C<eval "v200"> now works.
1712 Zero entries were missing from the Unicode classes like C<IsDigit>.
1718 =head2 Platform Specific Changes and Fixes
1726 Perl now works on post-4.0 BSD/OSes.
1732 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
1738 Numerous updates; currently synchronised with Cygwin 1.1.4.
1742 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
1748 EPOC update after Perl 5.6.0. See README.epoc.
1754 Perl now works on post-3.0 FreeBSDs.
1760 README.hpux updated; C<Configure -Duse64bitall> now almost works.
1766 Numerous compilation flag and hint enhancements; accidental mixing
1767 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
1777 Long doubles should now work (see INSTALL).
1781 Linux previously had problems related to sockaddrlen when using
1782 accept(), revcfrom() (in Perl: recv()), getpeername(), and getsockname().
1790 Compilation of the standard Perl distribution in MacOS Classic should
1791 now work if you have the Metrowerks development environment and
1792 the missing Mac-specific toolkit bits. Contact the macperl mailing
1799 MPE/iX update after Perl 5.6.0. See README.mpeix.
1805 Perl now works on NetBSD/sparc.
1811 Now works with usethreads (see INSTALL).
1817 64-bitness using the Sun Workshop compiler now works.
1821 Tru64 (aka Digital UNIX, aka DEC OSF/1)
1823 The operating system version letter now recorded in $Config{osvers}.
1824 Allow compiling with gcc (previously explicitly forbidden). Compiling
1825 with gcc still not recommended because buggy code results, even with
1832 Fixed various alignment problems that lead into core dumps either
1833 during build or later; no longer dies on math errors at runtime;
1834 now using full quad integers (64 bits), previously was using
1835 only 46 bit integers for speed.
1841 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
1842 (see INSTALL); now works with Perl's malloc.
1852 accept() no longer leaks memory.
1856 Borland C++ v5.5 is now a supported compiler that can build Perl.
1857 However, the generated binaries continue to be incompatible with those
1858 generated by the other supported compilers (GCC and Visual C++).
1862 Better chdir() return value for a non-existent directory.
1866 Duping socket handles with open(F, ">&MYSOCK") now works under Windows 9x.
1870 New %ENV entries now propagate to subprocesses.
1874 $ENV{LIB} now used to search for libs under Visual C.
1878 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
1879 Other bugs in chdir() and Cwd::cwd() have also been fixed.
1883 A failed (pseudo)fork now returns undef and sets errno to EAGAIN.
1887 HTML files will be installed in c:\perl\html instead of c:\perl\lib\pod\html
1891 The makefiles now provide a single switch to bulk-enable all the features
1892 enabled in ActiveState ActivePerl (a popular Win32 binary distribution).
1896 Allow REG_EXPAND_SZ keys in the registry.
1900 Can now send() from all threads, not just the first one.
1904 Fake signal handling reenabled, bugs and all.
1908 Less stack reserved per thread so that more threads can run
1909 concurrently. (Still 16M per thread.)
1913 C<File::Spec->tmpdir()> now prefers C:/temp over /tmp
1914 (works better when perl is running as service).
1918 Better UNC path handling under ithreads.
1922 wait() and waitpid() now work much better.
1926 winsock handle leak fixed.
1932 =head1 New or Changed Diagnostics
1938 All regular expression compilation error messages are now hopefully
1939 easier to understand both because the error message now comes before
1940 the failed regex and because the point of failure is now clearly
1941 marked by a C<E<lt>-- HERE> marker.
1945 The various "opened only for", "on closed", "never opened" warnings
1946 drop the C<main::> prefix for filehandles in the C<main> package,
1947 for example C<STDIN> instead of C<main::STDIN>.
1951 The "Unrecognized escape" warning has been extended to include C<\8>,
1952 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
1956 Two new debugging options have been added: if you have compiled your
1957 Perl with debugging, you can use the -DT and -DR options to trace
1958 tokenising and to add reference counts to displaying variables,
1963 If an attempt to use a (non-blessed) reference as an array index
1964 is made, a warning is given.
1968 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
1969 now give a warning. This may be a problem for generated and evaled
1974 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
1975 using the C<"C"> format you will get an optional warning. Similarly
1976 for the C<"c"> format and a number less than -128 or more than 127.
1980 Certain regex modifiers such as C<(?o)> make sense only if applied to
1981 the entire regex. You will an optional warning if you try to do otherwise.
1985 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
1986 has been deprecated for a while. Now you will get an optional warning.
1990 =head1 Changed Internals
1996 perlapi.pod (a companion to perlguts) now attempts to document the
2001 You can now build a really minimal perl called microperl.
2002 Building microperl does not require even running Configure;
2003 C<make -f Makefile.micro> should be enough. Beware: microperl makes
2004 many assumptions, some of which may be too bold; the resulting
2005 executable may crash or otherwise misbehave in wondrous ways.
2006 For careful hackers only.
2010 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
2011 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
2012 interfaces to the publicised API. For the full list of the available
2013 APIs see L<perlapi>.
2017 Made possible to propagate customised exceptions via croak()ing.
2021 Now xsubs can have attributes just like subs.
2025 dTHR and djSP have been obsoleted; the former removed (because it's
2026 a no-op) and the latter replaced with dSP.
2030 PERL_OBJECT has been completely removed.
2034 The MAGIC constants (e.g. C<'P'>) have been macrofied
2035 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
2036 and maintainability.
2040 The regex compiler now maintains a structure that identifies nodes in
2041 the compiled bytecode with the corresponding syntactic features of the
2042 original regex expression. The information is attached to the new
2043 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
2044 complete information.
2048 The C code has been made much more C<gcc -Wall> clean. Some warning
2049 messages still remain in some platforms, so if you are compiling with
2050 gcc you may see some warnings about dubious practices. The warnings
2051 are being worked on.
2055 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
2059 Documentation on how to use the Perl source repository has been added
2060 to F<Porting/repository.pod>.
2064 There are now several profiling make targets.
2068 =head1 Security Vulnerability Closed
2070 (This change was already made in 5.7.0 but bears repeating here.)
2072 A potential security vulnerability in the optional suidperl component
2073 of Perl was identified in August 2000. suidperl is neither built nor
2074 installed by default. As of November 2001 the only known vulnerable
2075 platform is Linux, most likely all Linux distributions. CERT and
2076 various vendors and distributors have been alerted about the vulnerability.
2077 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
2078 for more information.
2080 The problem was caused by Perl trying to report a suspected security
2081 exploit attempt using an external program, /bin/mail. On Linux
2082 platforms the /bin/mail program had an undocumented feature which
2083 when combined with suidperl gave access to a root shell, resulting in
2084 a serious compromise instead of reporting the exploit attempt. If you
2085 don't have /bin/mail, or if you have 'safe setuid scripts', or if
2086 suidperl is not installed, you are safe.
2088 The exploit attempt reporting feature has been completely removed from
2089 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
2090 from all the Perl 5.7 releases), so that particular vulnerability
2091 isn't there anymore. However, further security vulnerabilities are,
2092 unfortunately, always possible. The suidperl functionality is most
2093 probably going to be removed in Perl 5.10. In any case, suidperl
2094 should only be used by security experts who know exactly what they are
2095 doing and why they are using suidperl instead of some other solution
2096 such as sudo (see http://www.courtesan.com/sudo/).
2100 Several new tests have been added, especially for the F<lib> subsection.
2102 The tests are now reported in a different order than in earlier Perls.
2103 (This happens because the test scripts from under t/lib have been moved
2104 to be closer to the library/extension they are testing.)
2106 =head1 Known Problems
2108 Note that unlike other sections in this document (which describe
2109 changes since 5.7.0) this section is cumulative containing known
2110 problems for all the 5.7 releases.
2118 In AIX 4.2 Perl extensions that use C++ functions that use statics
2119 may have problems in that the statics are not getting initialized.
2120 In newer AIX releases this has been solved by linking Perl with
2121 the libC_r library, but unfortunately in AIX 4.2 the said library
2122 has an obscure bug where the various functions related to time
2123 (such as time() and gettimeofday()) return broken values, and
2124 therefore in AIX 4.2 Perl is not linked against the libC_r.
2128 vac 5.0.0.0 May Produce Buggy Code For Perl
2130 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
2131 resulting in few random tests failing, but when the failing tests
2132 are run by hand, they succeed. We suggest upgrading to at least
2133 vac version 5.0.1.0, that has been known to compile Perl correctly.
2134 "lslpp -L|grep vac.C" will tell you the vac version.
2138 =head2 Amiga Perl Invoking Mystery
2140 One cannot call Perl using the C<volume:> syntax, that is, C<perl -v>
2141 works, but for example C<bin:perl -v> doesn't. The exact reason isn't
2142 known but the current suspect is the F<ixemul> library.
2144 =head2 lib/ftmp-security tests warn 'system possibly insecure'
2146 Don't panic. Read INSTALL 'make test' section instead.
2148 =head2 Cygwin intermittent failures of lib/Memoize/t/expire_file 11 and 12
2150 The subtests 11 and 12 sometimes fail and sometimes work.
2152 =head2 HP-UX lib/io_multihomed Fails When LP64-Configured
2154 The lib/io_multihomed test may hang in HP-UX if Perl has been
2155 configured to be 64-bit. Because other 64-bit platforms do not hang in
2156 this test, HP-UX is suspect. All other tests pass in 64-bit HP-UX. The
2157 test attempts to create and connect to "multihomed" sockets (sockets
2158 which have multiple IP addresses).
2160 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
2162 If perl is configured with -Duse64bitall, the successful result of the
2163 subtest 10 of lib/posix may arrive before the successful result of the
2164 subtest 9, which confuses the test harness so much that it thinks the
2167 =head2 Linux With Sfio Fails op/misc Test 48
2173 OS/390 has rather many test failures but the situation is actually
2174 better than it was in 5.6.0, it's just that so many new modules and
2175 tests have been added.
2177 Failed Test Stat Wstat Total Fail Failed List of Failed
2178 -----------------------------------------------------------------------------
2179 ../ext/B/Deparse.t 14 1 7.14% 14
2180 ../ext/B/Showlex.t 1 1 100.00% 1
2181 ../ext/Encode/Encode/Tcl.t 610 13 2.13% 592 594 596 598
2183 ../ext/IO/lib/IO/t/io_unix.t 113 28928 5 3 60.00% 3-5
2184 ../ext/POSIX/POSIX.t 29 1 3.45% 14
2185 ../ext/Storable/t/lock.t 255 65280 5 3 60.00% 3-5
2186 ../lib/locale.t 129 33024 117 19 16.24% 99-117
2187 ../lib/warnings.t 434 1 0.23% 75
2188 ../lib/ExtUtils.t 27 1 3.70% 25
2189 ../lib/Math/BigInt/t/bigintpm.t 1190 1 0.08% 1145
2190 ../lib/Unicode/UCD.t 81 48 59.26% 1-16 49-64 66-81
2191 ../lib/User/pwent.t 9 1 11.11% 4
2192 op/pat.t 660 6 0.91% 242-243 424-425
2194 op/split.t 0 9 ?? ?? % ??
2195 op/taint.t 174 3 1.72% 156 162 168
2196 op/tr.t 70 3 4.29% 50 58-59
2197 Failed 16/422 test scripts, 96.21% okay. 105/23251 subtests failed, 99.55% okay.
2199 =head2 op/sprintf tests 129 and 130
2201 The op/sprintf tests 129 and 130 are known to fail on some platforms.
2202 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
2203 The failing platforms do not comply with the ANSI C Standard, line
2204 19ff on page 134 of ANSI X3.159 1989 to be exact. (They produce
2205 something other than "1" and "-1" when formatting 0.6 and -0.6 using
2206 the printf format "%.0f", most often they produce "0" and "-0".)
2208 =head2 Failure of Thread tests
2210 B<Note that support for 5.005-style threading remains experimental.>
2212 The following tests are known to fail due to fundamental problems in
2213 the 5.005 threading implementation. These are not new failures--Perl
2214 5.005_0x has the same bugs, but didn't have these tests.
2217 t/lib/thr5005.t 19-20
2225 ext/POSIX/sigaction subtests 6 and 13 may fail.
2229 lib/ExtUtils may spuriously claim that subtest 28 failed,
2230 which is interesting since the test only has 27 tests.
2234 Numerous numerical test failures
2236 op/numconvert 209,210,217,218
2238 ext/Time/HiRes/HiRes 9
2239 lib/Math/BigInt/t/bigintpm 1145
2242 These tests fail because of yet unresolved floating point inaccuracies.
2248 There are a few known test failures, see L<perluts>.
2252 Rather a lot of tests are failing in VMS, but actually more tests
2253 succeed in VMS than they used to; it's just that there are many,
2254 many more tests than there used to be.
2256 Here are the known failures from some compiler/platform combinations.
2258 DEC C V5.3-006 on OpenVMS VAX V6.2
2260 [-.ext.list.util.t]tainted..............FAILED on test 3
2261 [-.ext.posix]sigaction..................FAILED on test 7
2262 [-.ext.time.hires]hires.................FAILED on test 14
2263 [-.lib.file.find]taint..................FAILED on test 17
2264 [-.lib.math.bigint.t]bigintpm...........FAILED on test 1183
2265 [-.lib.test.simple.t]exit...............FAILED on test 1
2266 [.lib]vmsish............................FAILED on test 13
2267 [.op]sprintf............................FAILED on test 12
2268 Failed 8/399 tests, 91.23% okay.
2270 DEC C V6.0-001 on OpenVMS Alpha V7.2-1 and
2271 Compaq C V6.2-008 on OpenVMS Alpha V7.1
2273 [-.ext.list.util.t]tainted..............FAILED on test 3
2274 [-.lib.file.find]taint..................FAILED on test 17
2275 [-.lib.test.simple.t]exit...............FAILED on test 1
2276 [.lib]vmsish............................FAILED on test 13
2277 Failed 4/399 tests, 92.48% okay.
2279 Compaq C V6.4-005 on OpenVMS Alpha 7.2.1
2281 [-.ext.b]showlex........................FAILED on test 1
2282 [-.ext.list.util.t]tainted..............FAILED on test 3
2283 [-.lib.file.find]taint..................FAILED on test 17
2284 [-.lib.test.simple.t]exit...............FAILED on test 1
2285 [.lib]vmsish............................FAILED on test 13
2286 [.op]misc...............................FAILED on test 49
2287 Failed 6/401 tests, 92.77% okay.
2291 In multi-CPU boxes there are some problems with the I/O buffering:
2292 some output may appear twice.
2294 =head2 Localising a Tied Variable Leaks Memory
2297 tie my %tie_hash => 'Tie::StdHash';
2301 local($tie_hash{Foo}) = 1; # leaks
2303 Code like the above is known to leak memory every time the local()
2306 =head2 Self-tying of Arrays and Hashes Is Forbidden
2308 Self-tying of arrays and hashes is broken in rather deep and
2309 hard-to-fix ways. As a stop-gap measure to avoid people from getting
2310 frustrated at the mysterious results (core dumps, most often) it is
2311 for now forbidden (you will get a fatal error even from an attempt).
2313 =head2 Variable Attributes are not Currently Usable for Tieing
2315 This limitation will hopefully be fixed in future. (Subroutine
2316 attributes work fine for tieing, see L<Attribute::Handlers>).
2318 =head2 Building Extensions Can Fail Because Of Largefiles
2320 Some extensions like mod_perl are known to have issues with
2321 `largefiles', a change brought by Perl 5.6.0 in which file offsets
2322 default to 64 bits wide, where supported. Modules may fail to compile
2323 at all or compile and work incorrectly. Currently there is no good
2324 solution for the problem, but Configure now provides appropriate
2325 non-largefile ccflags, ldflags, libswanted, and libs in the %Config
2326 hash (e.g., $Config{ccflags_nolargefiles}) so the extensions that are
2327 having problems can try configuring themselves without the
2328 largefileness. This is admittedly not a clean solution, and the
2329 solution may not even work at all. One potential failure is whether
2330 one can (or, if one can, whether it's a good idea) link together at
2331 all binaries with different ideas about file offsets, all this is
2334 =head2 The Compiler Suite Is Still Experimental
2336 The compiler suite is slowly getting better but is nowhere near
2339 =head2 The Long Double Support is Still Experimental
2341 The ability to configure Perl's numbers to use "long doubles",
2342 floating point numbers of hopefully better accuracy, is still
2343 experimental. The implementations of long doubles are not yet
2344 widespread and the existing implementations are not quite mature
2345 or standardised, therefore trying to support them is a rare
2346 and moving target. The gain of more precision may also be offset
2347 by slowdown in computations (more bits to move around, and the
2348 operations are more likely to be executed by less optimised
2351 =head1 Reporting Bugs
2353 If you find what you think is a bug, you might check the articles
2354 recently posted to the comp.lang.perl.misc newsgroup and the perl
2355 bug database at http://bugs.perl.org. There may also be
2356 information at http://www.perl.com/perl/, the Perl Home Page.
2358 If you believe you have an unreported bug, please run the B<perlbug>
2359 program included with your release. Be sure to trim your bug down
2360 to a tiny but sufficient test case. Your bug report, along with the
2361 output of C<perl -V>, will be sent off to perlbug@perl.org to be
2362 analysed by the Perl porting team.
2366 The F<Changes> file for exhaustive details on what changed.
2368 The F<INSTALL> file for how to build Perl.
2370 The F<README> file for general stuff.
2372 The F<Artistic> and F<Copying> files for copyright information.
2376 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.