3 perldelta - what is new for perl v5.8.0
7 This document describes differences between the 5.6.0 release
10 Many of the bug fixes in 5.8.0 were already seen in the 5.6.1
11 maintenance release since the two releases were kept closely
14 If you are upgrading from Perl 5.005_03, you might also want
15 to read L<perl56delta>.
17 =head1 Highlights In 5.8.0
23 Better Unicode support
27 New Thread Implementation
35 Better Numeric Accuracy
43 More Extensive Regression Testing
47 =head1 Incompatible Changes
49 =head2 64-bit platforms and malloc
51 If your pointers are 64 bits wide, the Perl malloc is no longer being
52 used because it does not work well with 8-byte pointers. Also,
53 usually the system mallocs on such platforms are much better optimized
54 for such large memory models than the Perl malloc. Some memory-hungry
55 Perl applications like the PDL don't work well with Perl's malloc.
56 Finally, other applications than Perl (like modperl) tend to prefer
57 the system malloc. Such platforms include Alpha and 64-bit HPPA,
60 =head2 AIX Dynaloading
62 The AIX dynaloading now uses in AIX releases 4.3 and newer the native
63 dlopen interface of AIX instead of the old emulated interface. This
64 change will probably break backward compatibility with compiled
65 modules. The change was made to make Perl more compliant with other
66 applications like modperl which are using the AIX native interface.
68 =head2 Attributes for C<my> variables now handled at run-time.
70 The C<my EXPR : ATTRS> syntax now applies variable attributes at
71 run-time. (Subroutine and C<our> variables still get attributes applied
72 at compile-time.) See L<attributes> for additional details. In particular,
73 however, this allows variable attributes to be useful for C<tie> interfaces,
74 which was a deficiency of earlier releases. Note that the new semantics
75 doesn't work with the Attribute::Handlers module (as of version 0.76).
77 =head2 Socket Extension Dynamic in VMS
79 The Socket extension is now dynamically loaded instead of being
80 statically built in. This may or may not be a problem with ancient
81 TCP/IP stacks of VMS: we do not know since we weren't able to test
82 Perl in such configurations.
84 =head2 IEEE-format Floating Point Default on OpenVMS Alpha
86 Perl now uses IEEE format (T_FLOAT) as the default internal floating
87 point format on OpenVMS Alpha, potentially breaking binary compatibility
88 with external libraries or existing data. G_FLOAT is still available as
89 a configuration option. The default on VAX (D_FLOAT) has not changed.
91 =head2 New Unicode Properties
93 Unicode I<scripts> are now supported. Scripts are similar to (and superior
94 to) Unicode I<blocks>. The difference between scripts and blocks is that
95 scripts are the glyphs used by a language or a group of languages, while
96 the blocks are more artificial groupings of (mostly) 256 characters based
97 on the Unicode numbering.
99 In general, scripts are more inclusive, but not universally so. For
100 example, while the script C<Latin> includes all the Latin characters and
101 their various diacritic-adorned versions, it does not include the various
102 punctuation or digits (since they are not solely C<Latin>).
104 A number of other properties are now supported, including C<\p{L&}>,
105 C<\p{Any}> C<\p{Assigned}>, C<\p{Unassigned}>, C<\p{Blank}> and
106 C<\p{SpacePerl}> (along with their C<\P{...}> versions, of course).
107 See L<perlunicode> for details, and more additions.
109 The C<In> or C<Is> prefix to names used with the C<\p{...}> and C<\P{...}>
110 are now almost always optional. The only exception is that a C<In> prefix
111 is required to signify a Unicode block when a block name conflicts with a
112 script name. For example, C<\p{Tibetan}> refers to the script, while
113 C<\p{InTibetan}> refers to the block. When there is no name conflict, you
114 can omit the C<In> from the block name (e.g. C<\p{BraillePatterns}>), but
115 to be safe, it's probably best to always use the C<In>).
117 =head2 Perl Parser Stress Tested
119 The Perl parser has been stress tested using both random input and
120 Markov chain input and the few found crashes and lockups have been
123 =head2 REF(...) Instead Of SCALAR(...)
125 A reference to a reference now stringifies as "REF(0x81485ec)" instead
126 of "SCALAR(0x81485ec)" in order to be more consistent with the return
135 The semantics of bless(REF, REF) were unclear and until someone proves
136 it to make some sense, it is forbidden.
140 The obsolete chat2 library that should never have been allowed
141 to escape the laboratory has been decommissioned.
145 The builtin dump() function has probably outlived most of its
146 usefulness. The core-dumping functionality will remain in future
147 available as an explicit call to C<CORE::dump()>, but in future
148 releases the behaviour of an unqualified C<dump()> call may change.
152 The very dusty examples in the eg/ directory have been removed.
153 Suggestions for new shiny examples welcome but the main issue is that
154 the examples need to be documented, tested and (most importantly)
159 The (bogus) escape sequences \8 and \9 now give an optional warning
160 ("Unrecognized escape passed through"). There is no need to \-escape
165 The list of filenames from glob() (or <...>) is now by default sorted
166 alphabetically to be csh-compliant (which is what happened before
167 in most UNIX platforms). (bsd_glob() does still sort platform
168 natively, ASCII or EBCDIC, unless GLOB_ALPHASORT is specified.)
172 Spurious syntax errors generated in certain situations, when glob()
173 caused File::Glob to be loaded for the first time, have been fixed.
177 Although "you shouldn't do that", it was possible to write code that
178 depends on Perl's hashed key order (Data::Dumper does this). The new
179 algorithm "One-at-a-Time" produces a different hashed key order.
180 More details are in L</"Performance Enhancements">.
184 lstat(FILEHANDLE) now gives a warning because the operation makes no sense.
185 In future releases this may become a fatal error.
189 The C<package;> syntax (C<package> without an argument) has been
190 deprecated. Its semantics were never that clear and its
191 implementation even less so. If you have used that feature to
192 disallow all but fully qualified variables, C<use strict;> instead.
196 The unimplemented POSIX regex features [[.cc.]] and [[=c=]] are still
197 recognised but now cause fatal errors. The previous behaviour of
198 ignoring them by default and warning if requested was unacceptable
199 since it, in a way, falsely promised that the features could be used.
203 The current user-visible implementation of pseudo-hashes (the weird
204 use of the first array element) is deprecated starting from Perl 5.8.0
205 and will be removed in Perl 5.10.0, and the feature will be
206 implemented differently. Not only is the current interface rather
207 ugly, but the current implementation slows down normal array and hash
208 use quite noticeably. The C<fields> pragma interface will remain
213 The syntaxes C<< @a->[...] >> and C<< %h->{...} >> have now been deprecated.
217 After years of trying the suidperl is considered to be too complex to
218 ever be considered truly secure. The suidperl functionality is likely
219 to be removed in a future release.
223 The long deprecated uppercase aliases for the string comparison
224 operators (EQ, NE, LT, LE, GE, GT) have now been removed.
228 The tr///C and tr///U features have been removed and will not return;
229 the interface was a mistake. Sorry about that. For similar
230 functionality, see pack('U0', ...) and pack('C0', ...).
234 Earlier Perls treated "sub foo (@bar)" as equivalent to "sub foo (@)".
235 The prototypes are now checked at compile-time for invalid characters.
236 An optional warning is generated ("Illegal character in prototype...")
237 but this may be upgraded to a fatal error in a future release.
241 =head1 Core Enhancements
243 =head2 PerlIO is Now The Default
249 IO is now by default done via PerlIO rather than system's "stdio".
250 PerlIO allows "layers" to be "pushed" onto a file handle to alter the
251 handle's behaviour. Layers can be specified at open time via 3-arg
254 open($fh,'>:crlf :utf8', $path) || ...
256 or on already opened handles via extended C<binmode>:
258 binmode($fh,':encoding(iso-8859-7)');
260 The built-in layers are: unix (low level read/write), stdio (as in
261 previous Perls), perlio (re-implementation of stdio buffering in a
262 portable manner), crlf (does CRLF <=> "\n" translation as on Win32,
263 but available on any platform). A mmap layer may be available if
264 platform supports it (mostly UNIXes).
266 Layers to be applied by default may be specified via the 'open' pragma.
268 See L</"Installation and Configuration Improvements"> for the effects
269 of PerlIO on your architecture name.
273 File handles can be marked as accepting Perl's internal encoding of Unicode
274 (UTF-8 or UTF-EBCDIC depending on platform) by a pseudo layer ":utf8" :
276 open($fh,">:utf8","Uni.txt");
278 Note for EBCDIC users: the pseudo layer ":utf8" is erroneously named
279 for you since it's not UTF-8 what you will be getting but instead
280 UTF-EBCDIC. See L<perlunicode>, L<utf8>, and
281 http://www.unicode.org/unicode/reports/tr16/ for more information.
282 In future releases this naming may change.
286 File handles can translate character encodings from/to Perl's internal
287 Unicode form on read/write via the ":encoding()" layer.
291 File handles can be opened to "in memory" files held in Perl scalars via:
293 open($fh,'>', \$variable) || ...
297 Anonymous temporary files are available without need to
298 'use FileHandle' or other module via
300 open($fh,"+>", undef) || ...
302 That is a literal undef, not an undefined value.
306 The list form of C<open> is now implemented for pipes (at least on UNIX):
308 open($fh,"-|", 'cat', '/etc/motd')
310 creates a pipe, and runs the equivalent of exec('cat', '/etc/motd') in
317 Perl used to be fragile in that signals arriving at inopportune moments
318 could corrupt Perl's internal state. Now Perl postpones handling of
319 signals until it's safe (between opcodes).
321 This change may have surprising side effects because signals no longer
322 interrupt Perl instantly. Perl will now first finish whatever it was
323 doing, like finishing an internal operation (like sort()) or an
324 external operation (like an I/O operation), and only then look at any
325 arrived signals (and before starting the next operation). No more corrupt
326 internal state since the current operation is always finished first,
327 but the signal may take more time to get heard.
329 =head2 Unicode Overhaul
331 Unicode in general should be now much more usable than in Perl 5.6.0
332 (or even in 5.6.1). Unicode can be used in hash keys, Unicode in
333 regular expressions should work now, Unicode in tr/// should work now,
334 Unicode in I/O should work now.
340 The Unicode Character Database coming with Perl has been upgraded
341 to Unicode 3.1.1. For more information, see http://www.unicode.org/.
345 For developers interested in enhancing Perl's Unicode capabilities:
346 almost all the UCD files are included with the Perl distribution in
347 the F<lib/unicore subdirectory>. The most notable omission, for space
348 considerations, is the Unihan database.
352 The properties \p{Blank} and \p{SpacePerl} have been added. "Blank" is like
353 C isblank(), that is, it contains only "horizontal whitespace" (the space
354 character is, the newline isn't), and the "SpacePerl" is the Unicode
355 equivalent of C<\s> (\p{Space} isn't, since that includes the vertical
356 tabulator character, whereas C<\s> doesn't.)
358 See "New Unicode Properties" earlier in this document for additional
359 information on changes with Unicode properties.
363 =head2 Understanding of Numbers
365 In general a lot of fixing has happened in the area of Perl's
366 understanding of numbers, both integer and floating point. Since in
367 many systems the standard number parsing functions like C<strtoul()>
368 and C<atof()> seem to have bugs, Perl tries to work around their
369 deficiencies. This results hopefully in more accurate numbers.
371 Perl now tries internally to use integer values in numeric conversions
372 and basic arithmetics (+ - * /) if the arguments are integers, and
373 tries also to keep the results stored internally as integers.
374 This change leads to often slightly faster and always less lossy
375 arithmetics. (Previously Perl always preferred floating point numbers
378 =head2 Miscellaneous Changes
384 AUTOLOAD is now lvaluable, meaning that you can add the :lvalue attribute
385 to AUTOLOAD subroutines and you can assign to the AUTOLOAD return value.
389 C<perl -d:Module=arg,arg,arg> now works (previously one couldn't pass
390 in multiple arguments.)
394 The builtin dump() now gives an optional warning
395 C<Ambiguous call resolved as CORE::dump(), qualify as such or use &>
396 meaning that by default C<dump(...)> is resolved as the builtin
397 dump() which dumps core and aborts, not as (possibly) user-defined
398 C<sub dump>. To call the latter, qualify the call as C<&dump(...)>.
399 (The whole dump() feature is to considered deprecated, and possibly
400 removed/changed in future releases.)
404 chomp() and chop() have been demoted back to I<not> being overrideable
405 because they cannot really be overridden-- the problem is that their
406 prototype cannot be expressed and therefore one really cannot write
407 replacements to override these builtins.
411 END blocks are now run even if you exit/die in a BEGIN block.
412 Internally, the execution of END blocks is now controlled by
413 PL_exit_flags & PERL_EXIT_DESTRUCT_END. This enables the new
414 behaviour for Perl embedders. This will default in 5.10. See
419 Formats now support zero-padded decimal fields.
423 Lvalue subroutines can now return C<undef> in list context.
424 However, the lvalue subroutine feature still remains experimental.
428 A lost warning "Can't declare ... dereference in my" has been
429 restored (Perl had it earlier but it became lost in later releases.)
433 A new special regular expression variable has been introduced:
434 C<$^N>, which contains the most-recently closed group (submatch).
438 C<no Module;> now works even if there is no "sub unimport" in the Module.
442 The numerical comparison operators return C<undef> if either operand
443 is a NaN. Previously the behaviour was unspecified.
447 The following builtin functions are now overridable: each(), keys(),
448 pop(), push(), shift(), splice(), unshift().
452 C<pack() / unpack()> now can group template letters with C<()> and then
453 apply repetition/count modifiers on the groups.
457 C<pack() / unpack()> can now process the Perl internal numeric types:
458 IVs, UVs, NVs-- and also long doubles, if supported by the platform.
462 C<pack('U0a*', ...)> can now be used to force a string to UTF8.
466 my __PACKAGE__ $obj now works.
470 The printf() and sprintf() now support parameter reordering using the
471 C<%\d+\$> and C<*\d+\$> syntaxes. For example
473 print "%2\$s %1\$s\n", "foo", "bar";
475 will print "bar foo\n". This feature helps in writing
476 internationalised software, and in general when the order
477 of the parameters can vary.
481 prototype(\&) is now available.
485 prototype(\[$@%&]) is now available to implicitly create references
486 (useful for example if you want to emulate the tie() interface).
490 A new command-line option, C<-t> is available. It is the
491 little brother of C<-T>: instead of dieing on taint violations,
492 lexical warnings are given. B<This is only meant as a temporary
493 debugging aid while securing the code of old legacy applications.
494 This is not a substitute for -T.>
498 In other taint news, the C<exec LIST> and C<system LIST> have now been
499 considered too risky (think C<exec @ARGV>: it can start any program
500 with any arguments), and now the said forms cause a warning.
501 You should carefully launder the arguments to guarantee their
502 validity. In future releases of Perl the forms will become fatal
503 errors so consider starting laundering now.
507 If tr/// is just counting characters, it doesn't attempt to
512 untie() will now call an UNTIE() hook if it exists. See L<perltie>
517 L<utime> now supports C<utime undef, undef, @files> to change the
518 file timestamps to the current time.
522 The rules for allowing underscores (underbars) in numeric constants
523 have been relaxed and simplified: now you can have an underscore
524 simply B<between digits>.
528 =head1 Modules and Pragmata
530 =head2 New Modules and Pragmata
536 C<Attribute::Handlers> allows a class to define attribute handlers.
539 use Attribute::Handlers;
540 sub Wolf :ATTR(SCALAR) { print "howl!\n" }
542 # later, in some package using or inheriting from MyPack...
544 my MyPack $Fluffy : Wolf; # the attribute handler Wolf will be called
546 Both variables and routines can have attribute handlers. Handlers can
547 be specific to type (SCALAR, ARRAY, HASH, or CODE), or specific to the
548 exact compilation phase (BEGIN, CHECK, INIT, or END).
552 B<B::Concise> is a new compiler backend for walking the Perl syntax
553 tree, printing concise info about ops, from Stephen McCamant. The
554 output is highly customisable. See L<B::Concise>.
558 C<Class::ISA> for reporting the search path for a class's ISA tree,
559 by Sean Burke, has been added. See L<Class::ISA>.
563 C<Cwd> has now a split personality: if possible, an XS extension is
564 used, (this will hopefully be faster, more secure, and more robust)
565 but if not possible, the familiar Perl implementation is used.
569 C<Devel::PPPort>, originally from Kenneth Albanowski and now
570 maintained by Paul Marquess, has been added. It is primarily used
571 by C<h2xs> to enhance portability of of XS modules between different
576 C<Digest>, frontend module for calculating digests (checksums), from
577 Gisle Aas, has been added. See L<Digest>.
581 C<Digest::MD5> for calculating MD5 digests (checksums) as defined in
582 RFC 1321, from Gisle Aas, has been added. See L<Digest::MD5>.
584 use Digest::MD5 'md5_hex';
586 $digest = md5_hex("Thirsty Camel");
588 print $digest, "\n"; # 01d19d9d2045e005c3f1b80e8b164de1
590 NOTE: the C<MD5> backward compatibility module is deliberately not
591 included since its further use is discouraged.
595 C<Encode>, by Nick Ing-Simmons, provides a mechanism to translate
596 between different character encodings. Support for Unicode,
597 ISO-8859-*, ASCII, CP*, KOI8-R, and three variants of EBCDIC are
598 compiled in to the module. Several other encodings (like Japanese,
599 Chinese, and MacIntosh encodings) are included and will be loaded at
600 runtime. See L<Encode>.
602 Any encoding supported by Encode module is also available to the
603 ":encoding()" layer if PerlIO is used.
607 C<I18N::Langinfo> can be use to query locale information.
608 See L<I18N::Langinfo>.
612 C<I18N::LangTags> has functions for dealing with RFC3066-style
613 language tags, by Sean Burke. See L<I18N::LangTags>.
617 C<ExtUtils::Constant> is a new tool for extension writers for
618 generating XS code to import C header constants, by Nicholas Clark.
619 See L<ExtUtils::Constant>.
623 C<Filter::Simple> is an easy-to-use frontend to Filter::Util::Call,
624 from Damian Conway. See L<Filter::Simple>.
630 use Filter::Simple sub {
631 while (my ($from, $to) = splice @_, 0, 2) {
640 use MyFilter qr/red/ => 'green';
642 print "red\n"; # this code is filtered, will print "green\n"
643 print "bored\n"; # this code is filtered, will print "bogreen\n"
647 print "red\n"; # this code is not filtered, will print "red\n"
651 C<File::Temp> allows one to create temporary files and directories in
652 an easy, portable, and secure way, by Tim Jenness. See L<File::Temp>.
656 C<Filter::Util::Call> provides you with the framework to write
657 I<Source Filters> in Perl, from Paul Marquess. For most uses the
658 frontend Filter::Simple is to be preferred. See L<Filter::Util::Call>.
662 L<libnet> is a collection of perl5 modules related to network
663 programming, from Graham Barr. See L<Net::FTP>, L<Net::NNTP>,
664 L<Net::Ping>, L<Net::POP3>, L<Net::SMTP>, and L<Net::Time>.
666 Perl installation leaves libnet unconfigured, use F<libnetcfg> to configure.
670 C<List::Util> is a selection of general-utility list subroutines, like
671 sum(), min(), first(), and shuffle(), by Graham Barr. See L<List::Util>.
675 C<Locale::Constants>, C<Locale::Country>, C<Locale::Currency>, and
676 C<Locale::Language>, from Neil Bowers, have been added. They provide the
677 codes for various locale standards, such as "fr" for France, "usd" for
678 US Dollar, and "jp" for Japanese.
682 $country = code2country('jp'); # $country gets 'Japan'
683 $code = country2code('Norway'); # $code gets 'no'
685 See L<Locale::Constants>, L<Locale::Country>, L<Locale::Currency>,
686 and L<Locale::Language>.
690 C<Locale::Maketext> is localization framework from Sean Burke. See
691 L<Locale::Maketext>, and L<Locale::Maketext::TPJ13>. The latter is an
692 article about software localization, originally published in The Perl
693 Journal #13, republished here with kind permission.
697 C<Memoize> can make your functions faster by trading space for time,
698 from Mark-Jason Dominus. See L<Memoize>.
702 C<MIME::Base64> allows you to encode data in base64, from Gisle Aas,
703 as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
708 $encoded = encode_base64('Aladdin:open sesame');
709 $decoded = decode_base64($encoded);
711 print $encoded, "\n"; # "QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
717 C<MIME::QuotedPrint> allows you to encode data in quoted-printable
718 encoding, as defined in RFC 2045 - I<MIME (Multipurpose Internet Mail
719 Extensions)>, from Gisle Aas.
721 use MIME::QuotedPrint;
723 $encoded = encode_qp("Smiley in Unicode: \x{263a}");
724 $decoded = decode_qp($encoded);
726 print $encoded, "\n"; # "Smiley in Unicode: =263A"
728 MIME::QuotedPrint has been enhanced to provide the basic methods
729 necessary to use it with PerlIO::Via as in :
731 use MIME::QuotedPrint;
732 open($fh,">Via(MIME::QuotedPrint)",$path);
734 See L<MIME::QuotedPrint>.
738 C<NEXT> is pseudo-class for method redispatch, from Damian Conway.
743 C<open> is a new pragma for setting the default I/O disciplines
748 C<PerlIO::Scalar> provides the implementation of IO to "in memory"
749 Perl scalars as discussed above, from Nick Ing-Simmons. It also
750 serves as an example of a loadable PerlIO layer. Other future
751 possibilities include PerlIO::Array and PerlIO::Code.
752 See L<PerlIO::Scalar>.
756 C<PerlIO::Via> acts as a PerlIO layer and wraps PerlIO layer
757 functionality provided by a class (typically implemented in perl
758 code), from Nick Ing-Simmons.
760 use MIME::QuotedPrint;
761 open($fh,">Via(MIME::QuotedPrint)",$path);
763 This will automatically convert everything output to C<$fh>
764 to Quoted-Printable. See L<PerlIO::Via>.
768 C<Pod::ParseLink>, by Russ Allbery, has been added,
769 to parse LZ<><> links in pods as described in the new
774 C<Pod::Text::Overstrike>, by Joe Smith, has been added.
775 It converts POD data to formatted overstrike text.
776 See L<Pod::Text::Overstrike>.
780 C<Scalar::Util> is a selection of general-utility scalar subroutines,
781 like blessed(), reftype(), and tainted(). See L<Scalar::Util>.
785 C<sort> is a new pragma for controlling the behaviour of sort().
789 C<Storable> gives persistence to Perl data structures by allowing the
790 storage and retrieval of Perl data to and from files in a fast and
791 compact binary format, from Raphael Manfredi. See L<Storable>.
795 C<Switch>, from Damian Conway, has been added. Just by saying
799 you have C<switch> and C<case> available in Perl.
805 case 1 { print "number 1" }
806 case "a" { print "string a" }
807 case [1..10,42] { print "number in list" }
808 case (@array) { print "number in list" }
809 case /\w+/ { print "pattern" }
810 case qr/\w+/ { print "pattern" }
811 case (%hash) { print "entry in hash" }
812 case (\%hash) { print "entry in hash" }
813 case (\&sub) { print "arg to subroutine" }
814 else { print "previous case not true" }
821 C<Test::More> is yet another framework for writing test scripts,
822 more extensive than Test::Simple, by Michael Schwern. See L<Test::More>.
826 C<Test::Simple> has basic utilities for writing tests, by Michael
827 Schwern. See L<Test::Simple>.
831 C<Text::Balanced> has been added, for extracting delimited text
832 sequences from strings, from Damian Conway.
834 use Text::Balanced 'extract_delimited';
836 ($a, $b) = extract_delimited("'never say never', he never said", "'", '');
838 $a will be "'never say never'", $b will be ', he never said'.
840 In addition to extract_delimited() there are also extract_bracketed(),
841 extract_quotelike(), extract_codeblock(), extract_variable(),
842 extract_tagged(), extract_multiple(), gen_delimited_pat(), and
843 gen_extract_tagged(). With these you can implement rather advanced
844 parsing algorithms. See L<Text::Balanced>.
848 C<threads> is an interface to interpreter threads, by Arthur Bergman.
849 Interpreter threads (ithreads) is the new thread model introduced in
850 Perl 5.6 but only available as an internal interface for extension
851 writers (and for Win32 Perl for C<fork()> emulation). See L<threads>.
855 C<threads::shared> allows data sharing for interpreter threads, from
856 Arthur Bergman. In the ithreads model any data sharing between
857 threads must be explicit, as opposed to the old 5.005 thread model
858 where data sharing was implicit. See L<threads::shared>.
862 C<Tie::RefHash::Nestable>, by Edward Avis, allows storing hash
863 references (unlike the standard Tie::RefHash) The module is contained
864 within Tie::RefHash, see L<Tie::RefHash>.
868 C<Time::HiRes> provides high resolution timing (ualarm, usleep,
869 and gettimeofday), from Douglas E. Wegscheid. See L<Time::HiRes>.
873 C<Unicode::UCD> offers a querying interface to the Unicode Character
874 Database. See L<Unicode::UCD>.
878 C<Unicode::Collate> implements the UCA (Unicode Collation Algorithm)
879 for sorting Unicode strings, by SADAHIRO Tomoyuki. See L<Unicode::Collate>.
883 C<Unicode::Normalize> implements the various Unicode normalization
884 forms, by SADAHIRO Tomoyuki. See L<Unicode::Normalize>.
888 C<XS::Typemap>, by Tim Jenness, is a test extension that exercises XS
889 typemaps. Nothing gets installed but for extension writers the code
894 =head2 Updated And Improved Modules and Pragmata
900 The following independently supported modules have been updated to the
901 newest versions from CPAN: CGI, CPAN, DB_File, File::Spec, File::Temp,
902 Getopt::Long, Math::BigFloat, Math::BigInt, the podlators bundle
903 (Pod::Man, Pod::Text), Pod::LaTeX, Pod::Parser, Storable,
904 Term::ANSIColor, Test, Text-Tabs+Wrap.
908 The attributes::reftype() now works on tied arguments.
912 AutoLoader can now be disabled with C<no AutoLoader;>.
916 B::Deparse has been significantly enhanced. It now can deparse almost
917 all of the standard test suite (so that the tests still succeed).
918 There is a make target "test.deparse" for trying this out.
922 Class::Struct can now define the classes in compile time.
926 Class::Struct now assigns the array/hash element if the accessor
927 is called with an array/hash element as the B<sole> argument.
931 Data::Dumper has now an option to sort hashes.
935 Data::Dumper has now an option to dump code references
940 DB_File now supports newer Berkeley DB versions, among
945 The English module can now be used without the infamous performance
948 use English '-no_performance_hit';
950 (Assuming, of course, that one doesn't need the troublesome variables
951 C<$`>, C<$&>, or C<$'>.) Also, introduced C<@LAST_MATCH_START> and
952 C<@LAST_MATCH_END> English aliases for C<@-> and C<@+>.
956 Fcntl, Socket, and Sys::Syslog have been rewritten to use the
957 new-style constant dispatch section (see L<ExtUtils::Constant>).
958 This means that they will be more robust and hopefully faster.
962 File::Find now chdir()s correctly when chasing symbolic links.
966 File::Find now has pre- and post-processing callbacks. It also
967 correctly changes directories when chasing symbolic links. Callbacks
968 (naughtily) exiting with "next;" instead of "return;" now work.
972 File::Find is now (again) reentrant. It also has been made
977 File::Glob::glob() renamed to File::Glob::bsd_glob() to avoid
978 prototype mismatch with CORE::glob().
982 File::Glob now supports C<GLOB_LIMIT> constant to limit the size of
983 the returned list of filenames.
987 Devel::Peek now has an interface for the Perl memory statistics
988 (this works only if you are using perl's malloc, and if you have
989 compiled with debugging).
993 IPC::Open3 now allows the use of numeric file descriptors.
997 IO::Socket has now atmark() method, which returns true if the socket
998 is positioned at the out-of-band mark. The method is also exportable
999 as a sockatmark() function.
1003 IO::Socket::INET has support for ReusePort option (if your platform
1004 supports it). The Reuse option now has an alias, ReuseAddr. For clarity
1005 you may want to prefer ReuseAddr.
1009 IO::Socket::INET now supports C<LocalPort> of zero (usually meaning
1010 that the operating system will make one up.)
1014 use lib now works identically to @INC. Removing directories
1015 with 'no lib' now works.
1019 ExtUtils::MakeMaker now uses File::Spec internally, which hopefully
1020 leads into better portability.
1024 Math::BigFloat and Math::BigInt have undergone a full rewrite.
1025 They are now magnitudes faster, and they support various
1026 bignum libraries such as GMP and PARI as their backends.
1030 Math::Complex handles inf, NaN etc., better.
1034 Net::Ping has been muchly enhanced. Multihoming is now supported.
1035 There is now "external" protocol which uses Net::Ping::External module
1036 which runs external ping(1) and parses the output. A version of
1037 Net::Ping::External is available in CPAN.
1041 POSIX::sigaction() is now much more flexible and robust.
1042 You can now install coderef handlers, 'DEFAULT', and 'IGNORE'
1043 handlers, installing new handlers was not atomic.
1047 In Safe the C<%INC> now localised in a Safe compartment so that
1052 In SDBM_File on dosish platforms, some keys went missing because of
1053 lack of support for files with "holes". A workaround for the problem
1058 In Search::Dict one can now have a pre-processing hook for the
1059 lines being searched.
1063 The Shell module now has an OO interface.
1067 The Test module has been significantly enhanced.
1071 The vars pragma now supports declaring fully qualified variables.
1072 (Something that C<our()> does not and will not support.)
1076 The C<utf8::> name space (as in the pragma) provides various
1077 Perl-callable functions to provide low level access to Perl's
1078 internal Unicode representation. At the moment only length()
1079 has been implemented.
1083 =head1 Utility Changes
1089 Emacs perl mode (emacs/cperl-mode.el) has been updated to version
1094 F<emacs/e2ctags.pl> is now much faster.
1098 C<h2ph> now supports C trigraphs.
1102 C<h2xs> now produces a template README.
1106 C<h2xs> now uses C<Devel::PPort> for better portability between
1107 different versions of Perl.
1111 C<h2xs> uses the new L<ExtUtils::Constant> module which will affect
1112 newly created extensions that define constants. Since the new code is
1113 more correct (if you have two constants where the first one is a
1114 prefix of the second one, the first constant B<never> gets defined),
1115 less lossy (it uses integers for integer constant, as opposed to the
1116 old code that used floating point numbers even for integer constants),
1117 and slightly faster, you might want to consider regenerating your
1118 extension code (the new scheme makes regenerating easy).
1119 L<h2xs> now also supports C trigraphs.
1123 C<libnetcfg> has been added to configure the libnet.
1127 C<perlbug> is now much more robust. It also sends the bug report to
1128 perl.org, not perl.com.
1132 C<perlcc> has been rewritten and its user interface (that is,
1133 command line) is much more like that of the UNIX C compiler, cc.
1134 (The perlbc tools has been removed. Use C<perlcc -B> instead.)
1138 C<perlivp> is a new Installation Verification Procedure utility
1139 for running any time after installing Perl.
1143 C<pod2html> now allows specifying a cache directory.
1147 C<s2p> has been completely rewritten in Perl. (It is in fact a full
1148 implementation of sed in Perl: you can use the sed functionality by
1149 using the C<psed> utility.)
1153 C<xsubpp> now understands POD documentation embedded in the *.xs files.
1157 C<xsubpp> now supports OUT keyword.
1161 =head1 New Documentation
1167 perl56delta details the changes between the 5.005 release and the
1172 perlclib documents the internal replacements for standard C library
1173 functions. (Interesting only for extension writers and Perl core
1178 perldebtut is a Perl debugging tutorial.
1182 perlebcdic contains considerations for running Perl on EBCDIC platforms.
1186 perlintro is a gentle introduction to Perl.
1190 perliol documents the internals of PerlIO with layers.
1194 perlmodstyle is a style guide for writing modules.
1198 perlnewmod tells about writing and submitting a new module.
1202 perlpacktut is a pack() tutorial.
1206 perlpod has been rewritten to be clearer and to record the best
1207 practices gathered over the years.
1211 perlpodspec is a more formal specification of the pod format,
1212 mainly of interest for writers of pod applications, not to
1213 people writing in pod.
1217 perlretut is a regular expression tutorial.
1221 perlrequick is a regular expressions quick-start guide.
1222 Yes, much quicker than perlretut.
1226 perltodo has been updated.
1230 perltootc has been renamed as perltooc (to not to conflict
1231 with perltoot in filesystems restricted to "8.3" names)
1235 perluniintro is an introduction to using Unicode in Perl.
1236 (perlunicode is more of a detailed reference and background
1241 perlutil explains the command line utilities packaged with the Perl
1246 The following platform-specific documents are available before
1247 the installation as README.I<platform>, and after the installation
1250 perlaix perlamiga perlapollo perlbeos perlbs2000
1251 perlce perlcygwin perldgux perldos perlepoc perlhpux
1252 perlhurd perlmachten perlmacos perlmint perlmpeix
1253 perlnetware perlos2 perlos390 perlplan9 perlqnx perlsolaris
1254 perltru64 perluts perlvmesa perlvms perlvos perlwin32
1260 The documentation for the POSIX-BC platform is called "BS2000", to avoid
1261 confusion with the Perl POSIX module.
1265 The documentation for the WinCE platform is called "CE", to avoid
1266 confusion with the perlwin32 documentation on 8.3-restricted filesystems.
1270 =head1 Performance Enhancements
1276 map() could get pathologically slow when the result list it generates
1277 is larger than the source list. The performance has been improved for
1282 sort() has been changed to use primarily mergesort internally as
1283 opposed to the earlier quicksort. For very small lists this may
1284 result in slightly slower sorting times, but in general the speedup
1285 should be at least 20%. Additional bonuses are that the worst case
1286 behaviour of sort() is now better (in computer science terms it now
1287 runs in time O(N log N), as opposed to quicksort's Theta(N**2)
1288 worst-case run time behaviour), and that sort() is now stable
1289 (meaning that elements with identical keys will stay ordered as they
1290 were before the sort). See the C<sort> pragma for information.
1292 The story in more detail: suppose you want to serve yourself a little
1295 @digits = ( 3,1,4,1,5,9 );
1297 A numerical sort of the digits will yield (1,1,3,4,5,9), as expected.
1298 Which C<1> comes first is hard to know, since one C<1> looks pretty
1299 much like any other. You can regard this as totally trivial,
1300 or somewhat profound. However, if you just want to sort the even
1301 digits ahead of the odd ones, then what will
1303 sort { ($a % 2) <=> ($b % 2) } @digits;
1305 yield? The only even digit, C<4>, will come first. But how about
1306 the odd numbers, which all compare equal? With the quicksort algorithm
1307 used to implement Perl 5.6 and earlier, the order of ties is left up
1308 to the sort. So, as you add more and more digits of Pi, the order
1309 in which the sorted even and odd digits appear will change.
1310 and, for sufficiently large slices of Pi, the quicksort algorithm
1311 in Perl 5.8 won't return the same results even if reinvoked with the
1312 same input. The justification for this rests with quicksort's
1313 worst case behavior. If you run
1315 sort { $a <=> $b } ( 1 .. $N , 1 .. $N );
1317 (something you might approximate if you wanted to merge two sorted
1318 arrays using sort), doubling $N doesn't just double the quicksort time,
1319 it I<quadruples> it. Quicksort has a worst case run time that can
1320 grow like N**2, so-called I<quadratic> behaviour, and it can happen
1321 on patterns that may well arise in normal use. You won't notice this
1322 for small arrays, but you I<will> notice it with larger arrays,
1323 and you may not live long enough for the sort to complete on arrays
1324 of a million elements. So the 5.8 quicksort scrambles large arrays
1325 before sorting them, as a statistical defence against quadratic behaviour.
1326 But that means if you sort the same large array twice, ties may be
1327 broken in different ways.
1329 Because of the unpredictability of tie-breaking order, and the quadratic
1330 worst-case behaviour, quicksort was I<almost> replaced completely with
1331 a stable mergesort. I<Stable> means that ties are broken to preserve
1332 the original order of appearance in the input array. So
1334 sort { ($a % 2) <=> ($b % 2) } (3,1,4,1,5,9);
1336 will yield (4,3,1,1,5,9), guaranteed. The even and odd numbers
1337 appear in the output in the same order they appeared in the input.
1338 Mergesort has worst case O(NlogN) behaviour, the best value
1339 attainable. And, ironically, this mergesort does particularly
1340 well where quicksort goes quadratic: mergesort sorts (1..$N, 1..$N)
1341 in O(N) time. But quicksort was rescued at the last moment because
1342 it is faster than mergesort on certain inputs and platforms.
1343 For example, if you really I<don't> care about the order of even
1344 and odd digits, quicksort will run in O(N) time; it's very good
1345 at sorting many repetitions of a small number of distinct elements.
1346 The quicksort divide and conquer strategy works well on platforms
1347 with relatively small, very fast, caches. Eventually, the problem gets
1348 whittled down to one that fits in the cache, from which point it
1349 benefits from the increased memory speed.
1351 Quicksort was rescued by implementing a sort pragma to control aspects
1352 of the sort. The B<stable> subpragma forces stable behaviour,
1353 regardless of algorithm. The B<_quicksort> and B<_mergesort>
1354 subpragmas are heavy-handed ways to select the underlying implementation.
1355 The leading C<_> is a reminder that these subpragmas may not survive
1356 beyond 5.8. More appropriate mechanisms for selecting the implementation
1357 exist, but they wouldn't have arrived in time to save quicksort.
1361 Hashes now use Bob Jenkins "One-at-a-Time" hashing key algorithm
1362 (http://burtleburtle.net/bob/hash/doobs.html). This algorithm is
1363 reasonably fast while producing a much better spread of values than
1364 the old hashing algorithm (originally by Chris Torek, later tweaked by
1365 Ilya Zakharevich). Hash values output from the algorithm on a hash of
1366 all 3-char printable ASCII keys comes much closer to passing the
1367 DIEHARD random number generation tests. According to perlbench, this
1368 change has not affected the overall speed of Perl.
1372 unshift() should now be noticeably faster.
1376 =head1 Installation and Configuration Improvements
1378 =head2 Generic Improvements
1384 INSTALL now explains how you can configure Perl to use 64-bit
1385 integers even on non-64-bit platforms.
1389 Policy.sh policy change: if you are reusing a Policy.sh file
1390 (see INSTALL) and you use Configure -Dprefix=/foo/bar and in the old
1391 Policy $prefix eq $siteprefix and $prefix eq $vendorprefix, all of
1392 them will now be changed to the new prefix, /foo/bar. (Previously
1393 only $prefix changed.) If you do not like this new behaviour,
1394 specify prefix, siteprefix, and vendorprefix explicitly.
1398 A new optional location for Perl libraries, otherlibdirs, is available.
1399 It can be used for example for vendor add-ons without disturbing Perl's
1400 own library directories.
1404 In many platforms the vendor-supplied 'cc' is too stripped-down to
1405 build Perl (basically, 'cc' doesn't do ANSI C). If this seems
1406 to be the case and 'cc' does not seem to be the GNU C compiler
1407 'gcc', an automatic attempt is made to find and use 'gcc' instead.
1411 gcc needs to closely track the operating system release to avoid
1412 build problems. If Configure finds that gcc was built for a different
1413 operating system release than is running, it now gives a clearly visible
1414 warning that there may be trouble ahead.
1418 If binary compatibility with the 5.005 release is not wanted, Configure
1419 no longer suggests including the 5.005 modules in @INC.
1423 Configure C<-S> can now run non-interactively.
1427 Configure support for pdp11-style memory models has been removed due
1432 configure.gnu now works with options with whitespace in them.
1436 installperl now outputs everything to STDERR.
1440 $Config{byteorder} is now computed dynamically (this is more robust
1441 with "fat binaries" where an executable image contains binaries for
1442 more than one binary platform.)
1446 Because PerlIO is now the default on most platforms, "-perlio" doesn't
1447 get appended to the $Config{archname} (also known as $^O) anymore.
1448 Instead, if you explicitly choose not to use perlio (Configure command
1449 line option -Uuseperlio), you will get "-stdio" appended.
1453 Another change related to the architecture name is that "-64all"
1454 (-Duse64bitall, or "maximally 64-bit") is appended only if your
1455 pointers are 64 bits wide. (To be exact, the use64bitall is ignored.)
1459 In AFS installations one can configure the root of the AFS to be
1460 somewhere else than the default F</afs> by using the Configure
1461 parameter C<-Dafsroot=/some/where/else>.
1465 APPLLIB_EXP, a less-know configuration-time definition, has been
1466 documented. It can be used to prepend site-specific directories
1467 to Perl's default search path (@INC), see INSTALL for information.
1471 The version of Berkeley DB used when the Perl (and, presumably, the
1472 DB_File extension) was built is now available as
1473 C<@Config{qw(db_version_major db_version_minor db_version_patch)}>
1474 from Perl and as C<DB_VERSION_MAJOR_CFG DB_VERSION_MINOR_CFG
1475 DB_VERSION_PATCH_CFG> from C.
1479 Building Berkeley DB3 for compatibility modes for DB, NDBM, and ODBM
1480 has been documented in INSTALL.
1484 If you have CPAN access (either network or a local copy such as a
1485 CD-ROM) you can during specify extra modules to Configure to build and
1486 install with Perl using the -Dextras=... option. See INSTALL for
1491 In addition to config.over a new override file, config.arch, is
1492 available. That is supposed to be used by hints file writers for
1493 architecture-wide changes (as opposed to config.over which is for
1498 If your file system supports symbolic links you can build Perl outside
1499 of the source directory by
1501 mkdir /tmp/perl/build/directory
1502 cd /tmp/perl/build/directory
1503 sh /path/to/perl/source/Configure -Dmksymlinks ...
1505 This will create in /tmp/perl/build/directory a tree of symbolic links
1506 pointing to files in /path/to/perl/source. The original files are left
1507 unaffected. After Configure has finished you can just say
1511 and Perl will be built and tested, all in /tmp/perl/build/directory.
1515 For Perl developers several new make targets for profiling
1516 and debugging have been added, see L<perlhack>.
1522 Use of the F<gprof> tool to profile Perl has been documented in
1523 L<perlhack>. There is a make target called "perl.gprof" for
1524 generating a gprofiled Perl executable.
1528 If you have GCC 3, there is a make target called "perl.gcov" for
1529 creating a gcoved Perl executable for coverage analysis. See
1534 If you are on IRIX or Tru64 platforms, new profiling/debugging options
1535 have been added, see L<perlhack> for more information about pixie and
1542 Guidelines of how to construct minimal Perl installations have
1543 been added to INSTALL.
1547 The Thread extension is now not built at all under ithreads
1548 (C<Configure -Duseithreads>) because it wouldn't work anyway (the
1549 Thread extension requires being Configured with C<-Duse5005threads>).
1551 But note that the Thread.pm interface is now shared by both
1556 =head2 New Or Improved Platforms
1558 For the list of platforms known to support Perl,
1559 see L<perlport/"Supported Platforms">.
1565 AIX dynamic loading should be now better supported.
1569 AIX should now work better with gcc, threads, and 64-bitness. Also the
1570 long doubles support in AIX should be better now. See L<perlaix>.
1574 After a long pause, AmigaOS has been verified to be happy with Perl.
1578 AtheOS (http://www.atheos.cx/) is a new platform.
1582 BeOS has been reclaimed.
1586 DG/UX platform now supports the 5.005-style threads. See L<perldgux>.
1590 DYNIX/ptx platform (a.k.a. dynixptx) is supported at or near osvers 4.5.2.
1594 EBCDIC platforms (z/OS, also known as OS/390, POSIX-BC, and VM/ESA)
1595 have been regained. Many test suite tests still fail and the
1596 co-existence of Unicode and EBCDIC isn't quite settled, but the
1597 situation is much better than with Perl 5.6. See L<perlos390>,
1598 L<perlbs2000> (for POSIX-BC), and L<perlvmesa> for more information.
1602 Building perl with -Duseithreads or -Duse5005threads now works under
1603 HP-UX 10.20 (previously it only worked under 10.30 or later). You will
1604 need a thread library package installed. See README.hpux.
1608 MacOS Classic (MacPerl has of course been available since
1609 perl 5.004 but now the source code bases of standard Perl
1610 and MacPerl have been synchronised)
1614 MacOS X (or Darwin) should now be able to build Perl even on HFS+
1615 filesystems. (The case-insensitivity confused the Perl build process.)
1619 NCR MP-RAS is now supported.
1623 All the NetBSD specific patches (except for the installation
1624 specific ones) have been merged back to the main distribution.
1628 NetWare from Novell is now supported. See L<perlnetware>.
1632 NonStop-UX is now supported.
1636 NEC SUPER-UX is now supported.
1640 All the OpenBSD specific patches (except for the installation
1641 specific ones) have been merged back to the main distribution.
1645 Perl has been tested with the GNU pth userlevel thread package
1646 ( http://www.gnu.org/software/pth/pth.html ) . All but one thread
1647 test worked, and that one failure was because of test results arriving
1648 in unexpected order.
1652 Amdahl UTS UNIX mainframe platform is now supported.
1656 WinCE is now supported. See L<perlce>.
1660 z/OS (formerly known as OS/390, formerly known as MVS OE) has now
1661 support for dynamic loading. This is not selected by default,
1662 however, you must specify -Dusedl in the arguments of Configure.
1666 =head1 Selected Bug Fixes
1668 Numerous memory leaks and uninitialized memory accesses have been
1669 hunted down. Most importantly anonymous subs used to leak quite
1676 The autouse pragma didn't work for Multi::Part::Function::Names.
1680 caller() could cause core dumps in certain situations. Carp was sometimes
1681 affected by this problem.
1685 chop(@list) in list context returned the characters chopped in
1686 reverse order. This has been reversed to be in the right order.
1690 Configure no longer includes the DBM libraries (dbm, gdbm, db, ndbm)
1691 when building the Perl binary. The only exception to this is SunOS 4.x,
1696 The behaviour of non-decimal but numeric string constants such as
1697 "0x23" was platform-dependent: in some platforms that was seen as 35,
1698 in some as 0, in some as a floating point number (don't ask). This
1699 was caused by Perl using the operating system libraries in a situation
1700 where the result of the string to number conversion is undefined: now
1701 Perl consistently handles such strings as zero in numeric contexts.
1705 The order of DESTROYs has been made more predictable.
1709 Several debugger fixes: exit code now reflects the script exit code,
1710 condition C<"0"> now treated correctly, the C<d> command now checks
1711 line number, the C<$.> no longer gets corrupted, all debugger output
1712 now goes correctly to the socket if RemotePort is set.
1716 Perl 5.6.0 could emit spurious warnings about redefinition of dl_error()
1717 when statically building extensions into perl. This has been corrected.
1721 L<dprofpp> -R didn't work.
1725 C<*foo{FORMAT}> now works.
1728 Infinity is now recognized as a number.
1732 UNIVERSAL::isa no longer caches methods incorrectly. (This broke
1733 the Tk extension with 5.6.0.)
1737 Lexicals I: lexicals outside an eval "" weren't resolved
1738 correctly inside a subroutine definition inside the eval "" if they
1739 were not already referenced in the top level of the eval""ed code.
1743 Lexicals II: lexicals leaked at file scope into subroutines that
1744 were declared before the lexicals.
1748 Lexical warnings now propagating correctly between scopes
1749 and into C<eval "...">.
1753 C<use warnings qw(FATAL all)> did not work as intended. This has been
1758 warnings::enabled() now reports the state of $^W correctly if the caller
1759 isn't using lexical warnings.
1763 Line renumbering with eval and C<#line> now works.
1767 Fixed numerous memory leaks, especially in eval "".
1771 mkdir() now ignores trailing slashes in the directory name,
1772 as mandated by POSIX.
1776 Some versions of glibc have a broken modfl(). This affects builds
1777 with C<-Duselongdouble>. This version of Perl detects this brokenness
1778 and has a workaround for it. The glibc release 2.2.2 is known to have
1779 fixed the modfl() bug.
1783 Modulus of unsigned numbers now works (4063328477 % 65535 used to
1784 return 27406, instead of 27047).
1788 Some "not a number" warnings introduced in 5.6.0 eliminated to be
1789 more compatible with 5.005. Infinity is now recognised as a number.
1793 Numeric conversions did not recognize changes in the string value
1794 properly in certain circumstances.
1798 Attributes (like :shared) didn't work with our().
1802 our() variables will not cause "will not stay shared" warnings.
1806 "our" variables of the same name declared in two sibling blocks
1807 resulted in bogus warnings about "redeclaration" of the variables.
1808 The problem has been corrected.
1812 pack "Z" now correctly terminates the string with "\0".
1816 Fix password routines which in some shadow password platforms
1817 (e.g. HP-UX) caused getpwent() to return every other entry.
1821 The PERL5OPT environment variable (for passing command line arguments
1822 to Perl) didn't work for more than a single group of options.
1826 PERL5OPT with embedded spaces didn't work.
1830 printf() no longer resets the numeric locale to "C".
1834 C<qw(a\\b)> now parses correctly as C<'a\\b'>.
1838 pos() did not return the correct value within s///ge in earlier
1839 versions. This is now handled correctly.
1843 Printing quads (64-bit integers) with printf/sprintf now works
1844 without the q L ll prefixes (assuming you are on a quad-capable platform).
1848 Regular expressions on references and overloaded scalars now work.
1852 Right-hand side magic (GMAGIC) could in many cases such as string
1853 concatenation be invoked too many times.
1857 scalar() now forces scalar context even when used in void context.
1861 SOCKS support is now much more robust.
1865 sort() arguments are now compiled in the right wantarray context
1866 (they were accidentally using the context of the sort() itself).
1867 The comparison block is now run in scalar context, and the arguments
1868 to be sorted are always provided list context.
1872 Changed the POSIX character class C<[[:space:]]> to include the (very
1873 rarely used) vertical tab character. Added a new POSIX-ish character
1874 class C<[[:blank:]]> which stands for horizontal whitespace
1875 (currently, the space and the tab).
1879 The tainting behaviour of sprintf() has been rationalized. It does
1880 not taint the result of floating point formats anymore, making the
1881 behaviour consistent with that of string interpolation.
1885 Some cases of inconsistent taint propagation (such as within hash
1886 values) have been fixed.
1890 The RE engine found in Perl 5.6.0 accidentally pessimised certain kinds
1891 of simple pattern matches. These are now handled better.
1895 Regular expression debug output (whether through C<use re 'debug'>
1896 or via C<-Dr>) now looks better.
1900 Multi-line matches like C<"a\nxb\n" =~ /(?!\A)x/m> were flawed. The
1905 Use of $& could trigger a core dump under some situations. This
1910 The regular expression captured submatches ($1, $2, ...) are now
1911 more consistently unset if the match fails, instead of leaving false
1912 data lying around in them.
1916 readline() on files opened in "slurp" mode could return an extra "" at
1917 the end in certain situations. This has been corrected.
1921 Autovivification of symbolic references of special variables described
1922 in L<perlvar> (as in C<${$num}>) was accidentally disabled. This works
1927 Sys::Syslog ignored the C<LOG_AUTH> constant.
1931 All but the first argument of the IO syswrite() method are now optional.
1935 $AUTOLOAD, sort(), lock(), and spawning subprocesses
1936 in multiple threads simultaneously are now thread-safe.
1940 Tie::ARRAY SPLICE method was broken.
1944 Allow read-only string on left hand side of non-modifying tr///.
1948 Several Unicode fixes.
1954 BOMs (byte order marks) in the beginning of Perl files
1955 (scripts, modules) should now be transparently skipped.
1956 UTF-16 (UCS-2) encoded Perl files should now be read correctly.
1960 The character tables have been updated to Unicode 3.1.1.
1964 Comparing with utf8 data does not magically upgrade non-utf8 data
1965 into utf8. (This was a problem for example if you were mixing data
1966 from I/O and Unicode data: your output might have got magically encoded
1971 Generating illegal Unicode code points like U+FFFE, or the UTF-16
1972 surrogates, now also generates an optional warning.
1976 C<IsAlnum>, C<IsAlpha>, and C<IsWord> now match titlecase.
1980 Concatenation with the C<.> operator or via variable interpolation,
1981 C<eq>, C<substr>, C<reverse>, C<quotemeta>, the C<x> operator,
1982 substitution with C<s///>, single-quoted UTF8, should now work.
1986 The C<tr///> operator now works. Note that the C<tr///CU>
1987 functionality has been removed (but see pack('U0', ...)).
1991 C<eval "v200"> now works.
1995 Perl 5.6.0 parsed m/\x{ab}/ incorrectly, leading to spurious warnings.
1996 This has been corrected.
2000 Zero entries were missing from the Unicode classes like C<IsDigit>.
2006 Large unsigned numbers (those above 2**31) could sometimes lose their
2007 unsignedness, causing bogus results in arithmetic operations.
2011 =head2 Platform Specific Changes and Fixes
2019 Perl now works on post-4.0 BSD/OSes.
2025 Setting C<$0> now works (as much as possible; see L<perlvar> for details).
2031 Numerous updates; currently synchronised with Cygwin 1.1.4.
2035 Previously DYNIX/ptx had problems in its Configure probe for non-blocking I/O.
2041 EPOC update after Perl 5.6.0. See README.epoc.
2047 Perl now works on post-3.0 FreeBSDs.
2053 README.hpux updated; C<Configure -Duse64bitall> now almost works.
2059 Numerous compilation flag and hint enhancements; accidental mixing
2060 of 32-bit and 64-bit libraries (a doomed attempt) made much harder.
2070 Long doubles should now work (see INSTALL).
2074 Linux previously had problems related to sockaddrlen when using
2075 accept(), revcfrom() (in Perl: recv()), getpeername(), and getsockname().
2083 Compilation of the standard Perl distribution in MacOS Classic should
2084 now work if you have the Metrowerks development environment and
2085 the missing Mac-specific toolkit bits. Contact the macperl mailing
2092 MPE/iX update after Perl 5.6.0. See README.mpeix.
2098 Perl now works on NetBSD/sparc.
2104 Now works with usethreads (see INSTALL).
2110 64-bitness using the Sun Workshop compiler now works.
2114 Tru64 (aka Digital UNIX, aka DEC OSF/1)
2116 The operating system version letter now recorded in $Config{osvers}.
2117 Allow compiling with gcc (previously explicitly forbidden). Compiling
2118 with gcc still not recommended because buggy code results, even with
2125 Fixed various alignment problems that lead into core dumps either
2126 during build or later; no longer dies on math errors at runtime;
2127 now using full quad integers (64 bits), previously was using
2128 only 46 bit integers for speed.
2134 chdir() now works better despite a CRT bug; now works with MULTIPLICITY
2135 (see INSTALL); now works with Perl's malloc.
2137 The tainting of C<%ENV> elements via C<keys> or C<values> was previously
2138 unimplemented. It now works as documented.
2140 The C<waitpid> emulation has been improved. The worst bug (now fixed)
2141 was that a pid of -1 would cause a wildcard search of all processes on
2142 the system. The most significant enhancement is that we can now
2143 usually get the completion status of a terminated process.
2145 POSIX-style signals are now emulated much better on VMS versions prior
2148 The C<system> function and backticks operator have improved
2149 functionality and better error handling.
2159 accept() no longer leaks memory.
2163 Borland C++ v5.5 is now a supported compiler that can build Perl.
2164 However, the generated binaries continue to be incompatible with those
2165 generated by the other supported compilers (GCC and Visual C++).
2169 Better chdir() return value for a non-existent directory.
2173 Duping socket handles with open(F, ">&MYSOCK") now works under Windows 9x.
2177 New %ENV entries now propagate to subprocesses.
2181 Current directory entries in %ENV are now correctly propagated to child
2186 $ENV{LIB} now used to search for libs under Visual C.
2190 fork() emulation has been improved in various ways, but still continues
2191 to be experimental. See L<perlfork> for known bugs and caveats.
2195 A failed (pseudo)fork now returns undef and sets errno to EAGAIN.
2199 Win32::GetCwd() correctly returns C:\ instead of C: when at the drive root.
2200 Other bugs in chdir() and Cwd::cwd() have also been fixed.
2204 HTML files will be installed in c:\perl\html instead of c:\perl\lib\pod\html
2208 The makefiles now provide a single switch to bulk-enable all the features
2209 enabled in ActiveState ActivePerl (a popular Win32 binary distribution).
2213 Allow REG_EXPAND_SZ keys in the registry.
2217 Can now send() from all threads, not just the first one.
2221 Fake signal handling reenabled, bugs and all.
2225 %SIG has been enabled under USE_ITHREADS, but its use is completely
2226 unsupported under all configurations.
2230 Less stack reserved per thread so that more threads can run
2231 concurrently. (Still 16M per thread.)
2235 C<File::Spec->tmpdir()> now prefers C:/temp over /tmp
2236 (works better when perl is running as service).
2240 Better UNC path handling under ithreads.
2244 wait(), waitpid() and backticks now return the correct exit status under
2249 winsock handle leak fixed.
2255 =head1 New or Changed Diagnostics
2261 All regular expression compilation error messages are now hopefully
2262 easier to understand both because the error message now comes before
2263 the failed regex and because the point of failure is now clearly
2264 marked by a C<E<lt>-- HERE> marker.
2268 The various "opened only for", "on closed", "never opened" warnings
2269 drop the C<main::> prefix for filehandles in the C<main> package,
2270 for example C<STDIN> instead of C<main::STDIN>.
2274 The "Unrecognized escape" warning has been extended to include C<\8>,
2275 C<\9>, and C<\_>. There is no need to escape any of the C<\w> characters.
2279 Two new debugging options have been added: if you have compiled your
2280 Perl with debugging, you can use the -DT and -DR options to trace
2281 tokenising and to add reference counts to displaying variables,
2286 perl5db.pl has been modified to present a more consistent commands
2287 interface, via (CommandSet=580). perl5db.t was also added to test the
2288 changes, and as a placeholder for further tests.
2294 If an attempt to use a (non-blessed) reference as an array index
2295 is made, a warning is given.
2299 C<push @a;> and C<unshift @a;> (with no values to push or unshift)
2300 now give a warning. This may be a problem for generated and evaled
2305 If you try to L<perlfunc/pack> a number less than 0 or larger than 255
2306 using the C<"C"> format you will get an optional warning. Similarly
2307 for the C<"c"> format and a number less than -128 or more than 127.
2311 Certain regex modifiers such as C<(?o)> make sense only if applied to
2312 the entire regex. You will an optional warning if you try to do otherwise.
2316 Using arrays or hashes as references (e.g. C<< %foo->{bar} >>
2317 has been deprecated for a while. Now you will get an optional warning.
2321 =head1 Changed Internals
2327 perlapi.pod (a companion to perlguts) now attempts to document the
2332 You can now build a really minimal perl called microperl.
2333 Building microperl does not require even running Configure;
2334 C<make -f Makefile.micro> should be enough. Beware: microperl makes
2335 many assumptions, some of which may be too bold; the resulting
2336 executable may crash or otherwise misbehave in wondrous ways.
2337 For careful hackers only.
2341 Added rsignal(), whichsig(), do_join(), op_clear, op_null,
2342 ptr_table_clear(), ptr_table_free(), sv_setref_uv(), and several UTF-8
2343 interfaces to the publicised API. For the full list of the available
2344 APIs see L<perlapi>.
2348 Made possible to propagate customised exceptions via croak()ing.
2352 Now xsubs can have attributes just like subs. (Well, at least the
2353 built-in attributes.)
2357 dTHR and djSP have been obsoleted; the former removed (because it's
2358 a no-op) and the latter replaced with dSP.
2362 PERL_OBJECT has been completely removed.
2366 The MAGIC constants (e.g. C<'P'>) have been macrofied
2367 (e.g. C<PERL_MAGIC_TIED>) for better source code readability
2368 and maintainability.
2372 The regex compiler now maintains a structure that identifies nodes in
2373 the compiled bytecode with the corresponding syntactic features of the
2374 original regex expression. The information is attached to the new
2375 C<offsets> member of the C<struct regexp>. See L<perldebguts> for more
2376 complete information.
2380 The C code has been made much more C<gcc -Wall> clean. Some warning
2381 messages still remain in some platforms, so if you are compiling with
2382 gcc you may see some warnings about dubious practices. The warnings
2383 are being worked on.
2387 F<perly.c>, F<sv.c>, and F<sv.h> have now been extensively commented.
2391 Documentation on how to use the Perl source repository has been added
2392 to F<Porting/repository.pod>.
2396 There are now several profiling make targets.
2400 =head1 Security Vulnerability Closed
2402 (This change was already made in 5.7.0 but bears repeating here.)
2404 A potential security vulnerability in the optional suidperl component
2405 of Perl was identified in August 2000. suidperl is neither built nor
2406 installed by default. As of November 2001 the only known vulnerable
2407 platform is Linux, most likely all Linux distributions. CERT and
2408 various vendors and distributors have been alerted about the vulnerability.
2409 See http://www.cpan.org/src/5.0/sperl-2000-08-05/sperl-2000-08-05.txt
2410 for more information.
2412 The problem was caused by Perl trying to report a suspected security
2413 exploit attempt using an external program, /bin/mail. On Linux
2414 platforms the /bin/mail program had an undocumented feature which
2415 when combined with suidperl gave access to a root shell, resulting in
2416 a serious compromise instead of reporting the exploit attempt. If you
2417 don't have /bin/mail, or if you have 'safe setuid scripts', or if
2418 suidperl is not installed, you are safe.
2420 The exploit attempt reporting feature has been completely removed from
2421 Perl 5.8.0 (and the maintenance release 5.6.1, and it was removed also
2422 from all the Perl 5.7 releases), so that particular vulnerability
2423 isn't there anymore. However, further security vulnerabilities are,
2424 unfortunately, always possible. The suidperl functionality is most
2425 probably going to be removed in Perl 5.10. In any case, suidperl
2426 should only be used by security experts who know exactly what they are
2427 doing and why they are using suidperl instead of some other solution
2428 such as sudo (see http://www.courtesan.com/sudo/).
2432 Several new tests have been added, especially for the F<lib>
2433 subsection. There are now about 34 000 individual tests (spread over
2434 about 530 test scripts), in the regression suite (5.6.1 has about
2435 11700 tests, in 258 test scripts) Many of the new tests are introduced
2436 by the new modules, but still in general Perl is now more thoroughly
2439 Because of the large number of tests, running the regression suite
2440 will take considerably longer time than it used to: expect the suite
2441 to take up to 4-5 times longer to run than in perl 5.6. In a really
2442 fast machine you can hope to finish the suite in about 5 minutes
2445 The tests are now reported in a different order than in earlier Perls.
2446 (This happens because the test scripts from under t/lib have been moved
2447 to be closer to the library/extension they are testing.)
2449 =head1 Known Problems
2457 In AIX 4.2 Perl extensions that use C++ functions that use statics
2458 may have problems in that the statics are not getting initialized.
2459 In newer AIX releases this has been solved by linking Perl with
2460 the libC_r library, but unfortunately in AIX 4.2 the said library
2461 has an obscure bug where the various functions related to time
2462 (such as time() and gettimeofday()) return broken values, and
2463 therefore in AIX 4.2 Perl is not linked against the libC_r.
2467 vac 5.0.0.0 May Produce Buggy Code For Perl
2469 The AIX C compiler vac version 5.0.0.0 may produce buggy code,
2470 resulting in few random tests failing, but when the failing tests
2471 are run by hand, they succeed. We suggest upgrading to at least
2472 vac version 5.0.1.0, that has been known to compile Perl correctly.
2473 "lslpp -L|grep vac.C" will tell you the vac version.
2477 =head2 Amiga Perl Invoking Mystery
2479 One cannot call Perl using the C<volume:> syntax, that is, C<perl -v>
2480 works, but for example C<bin:perl -v> doesn't. The exact reason isn't
2481 known but the current suspect is the F<ixemul> library.
2483 =head2 lib/ftmp-security tests warn 'system possibly insecure'
2485 Don't panic. Read INSTALL 'make test' section instead.
2487 =head2 Cygwin intermittent failures of lib/Memoize/t/expire_file 11 and 12
2489 The subtests 11 and 12 sometimes fail and sometimes work.
2491 =head2 HP-UX lib/io_multihomed Fails When LP64-Configured
2493 The lib/io_multihomed test may hang in HP-UX if Perl has been
2494 configured to be 64-bit. Because other 64-bit platforms do not hang in
2495 this test, HP-UX is suspect. All other tests pass in 64-bit HP-UX. The
2496 test attempts to create and connect to "multihomed" sockets (sockets
2497 which have multiple IP addresses).
2499 =head2 HP-UX lib/posix Subtest 9 Fails When LP64-Configured
2501 If perl is configured with -Duse64bitall, the successful result of the
2502 subtest 10 of lib/posix may arrive before the successful result of the
2503 subtest 9, which confuses the test harness so much that it thinks the
2506 =head2 Linux With Sfio Fails op/misc Test 48
2512 The following tests are known to fail:
2514 Failed Test Stat Wstat Total Fail Failed List of Failed
2515 -------------------------------------------------------------------------
2516 ../ext/DB_File/t/db-btree.t 0 11 ?? ?? % ??
2517 ../ext/DB_File/t/db-recno.t 149 3 2.01% 61 63 65
2518 ../ext/POSIX/t/posix.t 31 1 3.23% 10
2519 ../lib/warnings.t 450 1 0.22% 316
2523 OS/390 has rather many test failures but the situation is actually
2524 better than it was in 5.6.0, it's just that so many new modules and
2525 tests have been added.
2527 Failed Test Stat Wstat Total Fail Failed List of Failed
2528 -----------------------------------------------------------------------------
2529 ../ext/B/Deparse.t 14 1 7.14% 14
2530 ../ext/B/Showlex.t 1 1 100.00% 1
2531 ../ext/Encode/Encode/Tcl.t 610 13 2.13% 592 594 596 598
2533 ../ext/IO/lib/IO/t/io_unix.t 113 28928 5 3 60.00% 3-5
2534 ../ext/POSIX/POSIX.t 29 1 3.45% 14
2535 ../ext/Storable/t/lock.t 255 65280 5 3 60.00% 3-5
2536 ../lib/locale.t 129 33024 117 19 16.24% 99-117
2537 ../lib/warnings.t 434 1 0.23% 75
2538 ../lib/ExtUtils.t 27 1 3.70% 25
2539 ../lib/Math/BigInt/t/bigintpm.t 1190 1 0.08% 1145
2540 ../lib/Unicode/UCD.t 81 48 59.26% 1-16 49-64 66-81
2541 ../lib/User/pwent.t 9 1 11.11% 4
2542 op/pat.t 660 6 0.91% 242-243 424-425
2544 op/split.t 0 9 ?? ?? % ??
2545 op/taint.t 174 3 1.72% 156 162 168
2546 op/tr.t 70 3 4.29% 50 58-59
2547 Failed 16/422 test scripts, 96.21% okay. 105/23251 subtests failed, 99.55% okay.
2549 =head2 op/sprintf tests 129 and 130
2551 The op/sprintf tests 129 and 130 are known to fail on some platforms.
2552 Examples include any platform using sfio, and Compaq/Tandem's NonStop-UX.
2553 The failing platforms do not comply with the ANSI C Standard, line
2554 19ff on page 134 of ANSI X3.159 1989 to be exact. (They produce
2555 something other than "1" and "-1" when formatting 0.6 and -0.6 using
2556 the printf format "%.0f", most often they produce "0" and "-0".)
2558 =head2 Failure of Thread tests
2560 B<Note that support for 5.005-style threading remains experimental
2561 and practically unsupported.>
2563 The following tests are known to fail due to fundamental problems in
2564 the 5.005 threading implementation. These are not new failures--Perl
2565 5.005_0x has the same bugs, but didn't have these tests.
2567 ext/List/Util/t/first 2
2569 ext/Thread/thr5005 19-20
2571 These failures are unlikely to get fixed.
2579 ext/POSIX/sigaction subtests 6 and 13 may fail.
2583 lib/ExtUtils may spuriously claim that subtest 28 failed,
2584 which is interesting since the test only has 27 tests.
2588 Numerous numerical test failures
2590 op/numconvert 209,210,217,218
2592 ext/Time/HiRes/HiRes 9
2593 lib/Math/BigInt/t/bigintpm 1145
2596 These tests fail because of yet unresolved floating point inaccuracies.
2600 =head2 UNICOS and UNICOS/mk
2602 The io/fs test #31 is failing because in UNICOS and UNICOS/mk
2603 truncate() cannot be used to grow the size of filehandles, only
2604 to reduce the size. The workaround is to truncate files instead
2609 There are a few known test failures, see L<perluts>.
2613 There is one known test failure with a default configuration:
2615 [.run]switches..........................FAILED on test 1
2619 In multi-CPU boxes there are some problems with the I/O buffering:
2620 some output may appear twice.
2622 =head2 Localising a Tied Variable Leaks Memory
2625 tie my %tie_hash => 'Tie::StdHash';
2629 local($tie_hash{Foo}) = 1; # leaks
2631 Code like the above is known to leak memory every time the local()
2634 =head2 Localising Tied Arrays and Hashes Is Broken
2638 doesn't work as one would expect: the old value is restored
2641 =head2 Self-tying of Arrays and Hashes Is Forbidden
2643 Self-tying of arrays and hashes is broken in rather deep and
2644 hard-to-fix ways. As a stop-gap measure to avoid people from getting
2645 frustrated at the mysterious results (core dumps, most often) it is
2646 for now forbidden (you will get a fatal error even from an attempt).
2648 =head2 Building Extensions Can Fail Because Of Largefiles
2650 Some extensions like mod_perl are known to have issues with
2651 `largefiles', a change brought by Perl 5.6.0 in which file offsets
2652 default to 64 bits wide, where supported. Modules may fail to compile
2653 at all or compile and work incorrectly. Currently there is no good
2654 solution for the problem, but Configure now provides appropriate
2655 non-largefile ccflags, ldflags, libswanted, and libs in the %Config
2656 hash (e.g., $Config{ccflags_nolargefiles}) so the extensions that are
2657 having problems can try configuring themselves without the
2658 largefileness. This is admittedly not a clean solution, and the
2659 solution may not even work at all. One potential failure is whether
2660 one can (or, if one can, whether it's a good idea) link together at
2661 all binaries with different ideas about file offsets, all this is
2664 =head2 Unicode Support on EBCDIC Still Spotty
2666 Though mostly working, Unicode support still has problem spots on
2667 EBCDIC platforms. One such known spot are the C<\p{}> and C<\P{}>
2668 regular expression constructs for code points less than 256: the
2669 pP are testing for Unicode code points, not knowing about EBCDIC.
2671 =head2 The Compiler Suite Is Still Experimental
2673 The compiler suite is slowly getting better but it continues to be
2674 highly experimental. Use in production environments is discouraged.
2676 =head2 The Long Double Support Is Still Experimental
2678 The ability to configure Perl's numbers to use "long doubles",
2679 floating point numbers of hopefully better accuracy, is still
2680 experimental. The implementations of long doubles are not yet
2681 widespread and the existing implementations are not quite mature
2682 or standardised, therefore trying to support them is a rare
2683 and moving target. The gain of more precision may also be offset
2684 by slowdown in computations (more bits to move around, and the
2685 operations are more likely to be executed by less optimised
2688 =head2 Seen In Perl 5.7 But Gone Now
2690 C<Time::Piece> (previously known as C<Time::Object>) was removed
2691 because it was felt that it didn't have enough value in it to be a
2692 core module. It is still a useful module, though, and is available
2695 =head1 Reporting Bugs
2697 If you find what you think is a bug, you might check the articles
2698 recently posted to the comp.lang.perl.misc newsgroup and the perl
2699 bug database at http://bugs.perl.org. There may also be
2700 information at http://www.perl.com/, the Perl Home Page.
2702 If you believe you have an unreported bug, please run the B<perlbug>
2703 program included with your release. Be sure to trim your bug down
2704 to a tiny but sufficient test case. Your bug report, along with the
2705 output of C<perl -V>, will be sent off to perlbug@perl.org to be
2706 analysed by the Perl porting team.
2710 The F<Changes> file for exhaustive details on what changed.
2712 The F<INSTALL> file for how to build Perl.
2714 The F<README> file for general stuff.
2716 The F<Artistic> and F<Copying> files for copyright information.
2720 Written by Jarkko Hietaniemi <F<jhi@iki.fi>>.