3 perldelta - what's new for perl v5.6 (as of v5.005_64)
7 This is an unsupported alpha release, meant for intrepid Perl developers
8 only. The included sources may not even build correctly on some platforms.
9 Subscribing to perl5-porters is the best way to monitor and contribute
10 to the progress of development releases (see www.perl.org for info).
12 This document describes differences between the 5.005 release and this one.
14 =head1 Incompatible Changes
16 =head2 Perl Source Incompatibilities
18 Beware that any new warnings that have been added or old ones
19 that have been enhanced are B<not> considered incompatible changes.
21 Since all new warnings must be explicitly requested via the C<-w>
22 switch or the C<warnings> pragma, it is ultimately the programmer's
23 responsibility to ensure that warnings are enabled judiciously.
27 =item STOP is a new keyword
29 In addition to C<BEGIN>, C<INIT>, C<END>, C<DESTROY> and C<AUTOLOAD>,
30 subroutines named C<STOP> are now special. These are queued up during
31 compilation and behave similar to END blocks, except they are called at
32 the end of compilation rather than at the end of execution. They cannot
35 =item Treatment of list slices of undef has changed
37 When taking a slice of a literal list (as opposed to a slice of
38 an array or hash), Perl used to return an empty list if the
39 result happened to be composed of all undef values.
41 The new behavior is to produce an empty list if (and only if)
42 the original list was empty. Consider the following example:
44 @a = (1,undef,undef,2)[2,1,2];
46 The old behavior would have resulted in @a having no elements.
47 The new behavior ensures it has three undefined elements.
49 Note in particular that the behavior of slices of the following
50 cases remains unchanged:
54 @a = (anything_returning_empty_list())[2,1,2];
60 =item Possibly changed pseudo-random number generator
62 In 5.005_0x and earlier, perl's rand() function used the C library
63 rand(3) function. As of 5.005_52, Configure tests for drand48(),
64 random(), and rand() (in that order) and picks the first one it finds.
65 Perl programs that depend on reproducing a specific set of pseudo-random
66 numbers will now likely produce different output. You can use
67 C<sh Configure -Drandfunc=rand> to obtain the old behavior.
69 =item Hashing function for hash keys has changed
71 Perl hashes are not order preserving. The apparently random order
72 encountered when iterating on the contents of a hash is determined
73 by the hashing algorithm used. To improve the distribution of lower
74 bits in the hashed value, the algorithm has changed slightly as of
75 5.005_52. When iterating over hashes, this may yield a random order
76 that is B<different> from that of previous versions.
78 =item C<undef> fails on read only values
80 Using the C<undef> operator on a readonly value (such as $1) has
81 the same effect as assigning C<undef> to the readonly value--it
84 =item Close-on-exec bit may be set on pipe() handles
86 On systems that support a close-on-exec flag on filehandles, the
87 flag will be set for any handles created by pipe(), if that is
88 warranted by the value of $^F that may be in effect. Earlier
89 versions neglected to set the flag for handles created with
90 pipe(). See L<perlfunc/pipe> and L<perlvar/$^F>.
92 =item Writing C<"$$1"> to mean C<"${$}1"> is unsupported
94 Perl 5.004 deprecated the interpretation of C<$$1> and
95 similar within interpolated strings to mean C<$$ . "1">,
98 In Perl 5.6 and later, C<"$$1"> always means C<"${$1}">.
100 =item delete(), values() and C<\(%h)> operate on aliases to values, not copies
102 delete(), each(), values() and hashes in a list context return the actual
103 values in the hash, instead of copies (as they used to in earlier
104 versions). Typical idioms for using these constructs copy the
105 returned values, but this can make a significant difference when
106 creating references to the returned values.
108 Keys in the hash are still returned as copies when iterating on
111 =item vec(EXPR,OFFSET,BITS) enforces powers-of-two BITS
113 vec() generates a run-time error if the BITS argument is not
114 a valid power-of-two integer.
116 =item Text of some diagnostic output has changed
118 Most references to internal Perl operations in diagnostics
119 have been changed to be more descriptive. This may be an
120 issue for programs that may incorrectly rely on the exact
121 text of diagnostics for proper functioning.
123 =item C<%@> has been removed
125 The undocumented special variable C<%@> that used to accumulate
126 "background" errors (such as those that happen in DESTROY())
127 has been removed, because it could potentially result in memory
130 =item Parenthesized not() behaves like a list operator
132 The C<not> operator now falls under the "if it looks like a function,
133 it behaves like a function" rule.
135 As a result, the parenthesized form can be used with C<grep> and C<map>.
136 The following construct used to be a syntax error before, but it works
139 grep not($_), @things;
141 On the other hand, using C<not> with a literal list slice may not
142 work. The following previously allowed construct:
144 print not (1,2,3)[0];
146 needs to be written with additional parentheses now:
148 print not((1,2,3)[0]);
150 The behavior remains unaffected when C<not> is not followed by parentheses.
154 =head2 C Source Incompatibilities
158 =item C<PERL_POLLUTE>
160 Release 5.005 grandfathered old global symbol names by providing preprocessor
161 macros for extension source compatibility. As of release 5.6, these
162 preprocessor definitions are not available by default. You need to explicitly
163 compile perl with C<-DPERL_POLLUTE> to get these definitions. For
164 extensions still using the old symbols, this option can be
165 specified via MakeMaker:
167 perl Makefile.PL POLLUTE=1
169 =item C<PERL_IMPLICIT_CONTEXT>
171 PERL_IMPLICIT_CONTEXT is automatically enabled whenever Perl is built
172 with one of -Dusethreads, -Dusemultiplicity, or both. It is not
173 intended to be enabled by users at this time.
175 This new build option provides a set of macros for all API functions
176 such that an implicit interpreter/thread context argument is passed to
177 every API function. As a result of this, something like C<sv_setsv(foo,bar)>
178 amounts to a macro invocation that actually translates to something like
179 C<Perl_sv_setsv(my_perl,foo,bar)>. While this is generally expected
180 to not have any significant source compatibility issues, the difference
181 between a macro and a real function call will need to be considered.
183 This means that there B<is> a source compatibility issue as a result of
184 this if your extensions attempt to use pointers to any of the Perl API
187 Note that the above issue is not relevant to the default build of
188 Perl, whose interfaces continue to match those of prior versions
189 (but subject to the other options described here).
191 See L<perlguts/"The Perl API"> for detailed information on the
192 ramifications of building Perl using this option.
194 =item C<PERL_POLLUTE_MALLOC>
196 Enabling Perl's malloc in release 5.005 and earlier caused
197 the namespace of system versions of the malloc family of functions to
198 be usurped by the Perl versions, since by default they used the
201 Besides causing problems on platforms that do not allow these functions to
202 be cleanly replaced, this also meant that the system versions could not
203 be called in programs that used Perl's malloc. Previous versions of Perl
204 have allowed this behaviour to be suppressed with the HIDEMYMALLOC and
205 EMBEDMYMALLOC preprocessor definitions.
207 As of release 5.6, Perl's malloc family of functions have default names
208 distinct from the system versions. You need to explicitly compile perl with
209 C<-DPERL_POLLUTE_MALLOC> to get the older behaviour. HIDEMYMALLOC
210 and EMBEDMYMALLOC have no effect, since the behaviour they enabled is now
213 Note that these functions do B<not> constitute Perl's memory allocation API.
214 See L<perlguts/"Memory Allocation"> for further information about that.
218 =head2 Compatible C Source API Changes
222 =item C<PATCHLEVEL> is now C<PERL_VERSION>
224 The cpp macros C<PERL_REVISION>, C<PERL_VERSION>, and C<PERL_SUBVERSION>
225 are now available by default from perl.h, and reflect the base revision,
226 patchlevel, and subversion respectively. C<PERL_REVISION> had no
227 prior equivalent, while C<PERL_VERSION> and C<PERL_SUBVERSION> were
228 previously available as C<PATCHLEVEL> and C<SUBVERSION>.
230 The new names cause less pollution of the B<cpp> namespace and reflect what
231 the numbers have come to stand for in common practice. For compatibility,
232 the old names are still supported when F<patchlevel.h> is explicitly
233 included (as required before), so there is no source incompatibility
236 =item Support for C++ exceptions
238 change#3386, also needs perlguts documentation
239 [TODO - Chip Salzenberg <chip@perlsupport.com>]
243 =head2 Binary Incompatibilities
245 The default build of this release is binary compatible with the 5.005
246 release or its maintenance versions.
248 The usethreads or usemultiplicity builds are B<not> binary compatible
249 with the corresponding builds in 5.005.
251 =head1 Installation and Configuration Improvements
253 =head2 New Configure flags
255 The following new flags may be enabled on the Configure command line
256 by running Configure with C<-Dflag>.
264 =head2 -Dusethreads and -Duse64bits now more daring
266 The Configure options enabling the use of threads and the use of
267 64-bitness are now more daring in the sense that they no more have
268 an explicit list of operating systems of known threads/64-bit
269 capabilities. In other words: if your operating system has the
270 necessary APIs, you should be able just to go ahead and use them.
271 See also L<"64-bit support">.
275 Some platforms have "long doubles", floating point numbers of even
276 larger range than ordinary "doubles". To enable using long doubles for
277 Perl's scalars, use -Duselongdouble.
281 You can enable both -Duse64bits and -Dlongdouble by -Dusemorebits.
282 See also L<"64-bit support">.
284 =head2 -Duselargefiles
286 Some platforms support large files, files larger than two gigabytes.
287 See L<"Large file support"> for more information.
289 =head2 installusrbinperl
291 You can use "Configure -Uinstallusrbinperl" which causes installperl
292 to skip installing perl also as /usr/bin/perl. This is useful if you
293 prefer not to modify /usr/bin for some reason or another but harmful
294 because many scripts assume to find Perl in /usr/bin/perl.
298 You can use "Configure -Dusesocks" which causes Perl to probe
299 for the SOCKS (v5, not v4) proxy protocol library,
300 http://www.socks.nec.com/
304 You can "post-edit" the Configure variables using the Configure C<-A>
305 flag. The editing happens immediately after the platform specific
306 hints files have been processed but before the actual configuration
307 process starts. Run C<Configure -h> to find out the full C<-A> syntax.
309 =head2 Enhanced Installation Directories
311 The installation structure has been enriched to improve the support for
312 maintaining multiple versions of perl, to provide locations for
313 vendor-supplied modules and scripts, and to ease maintenance of
314 locally-added modules and scripts. See the section on Installation
315 Directories in the INSTALL file for complete details. For most users
316 building and installing from source, the defaults should be fine.
320 =head2 Unicode and UTF-8 support
322 Perl can optionally use UTF-8 as its internal representation for character
323 strings. The C<utf8> pragma enables this support in the current lexical
324 scope. See L<utf8> for more information.
326 =head2 Interpreter threads
328 WARNING: This is an experimental feature in a pre-alpha state. Use
331 Perl 5.005_63 introduces the beginnings of support for running multiple
332 interpreters concurrently in different threads. In conjunction with
333 the perl_clone() API call, which can be used to selectively duplicate
334 the state of any given interpreter, it is possible to compile a
335 piece of code once in an interpreter, clone that interpreter
336 one or more times, and run all the resulting interpreters in distinct
339 On Windows, this feature is used to emulate fork() at the interpreter
340 level. See L<perlfork>.
342 This feature is still in evolution. It is eventually meant to be used
343 to selectively clone a subroutine and data reachable from that
344 subroutine in a separate interpreter and run the cloned subroutine
345 in a separate thread. Since there is no shared data between the
346 interpreters, little or no locking will be needed (unless parts of
347 the symbol table are explicitly shared). This is obviously intended
348 to be an easy-to-use replacement for the existing threads support.
350 Support for cloning interpreters must currently be manually enabled
351 by defining the cpp macro USE_ITHREADS on non-Windows platforms.
352 (See win32/Makefile for how to enable it on Windows.) The resulting
353 perl executable will be functionally identical to one that was built
354 without USE_ITHREADS, but the perl_clone() API call will only be
355 available in the former.
357 USE_ITHREADS enables Perl source code changes that provide a clear
358 separation between the op tree and the data it operates with. The
359 former is considered immutable, and can therefore be shared between
360 an interpreter and all of its clones, while the latter is considered
361 local to each interpreter, and is therefore copied for each clone.
363 Note that building Perl with the -Dusemultiplicity Configure option
364 is adequate if you wish to run multiple B<independent> interpreters
365 concurrently in different threads. USE_ITHREADS only needs to be
366 enabled if you wish to obtain access to perl_clone() and cloned
369 [XXX TODO - the Compiler backends may be broken when USE_ITHREADS is
372 =head2 Lexically scoped warning categories
374 You can now control the granularity of warnings emitted by perl at a finer
375 level using the C<use warnings> pragma. See L<warnings> and L<perllexwarn>
378 =head2 Lvalue subroutines
380 WARNING: This is an experimental feature.
383 [TODO - Ilya Zakharevich <ilya@math.ohio-state.edu>,
384 Tuomas Lukka <lukka@fas.harvard.edu>)]
386 =head2 "our" declarations
388 An "our" declaration introduces a value that can be best understood
389 as a lexically scoped symbolic alias to a global variable in the
390 current package. This is mostly useful as an alternative to the
391 C<vars> pragma, but also provides the opportunity to introduce
392 typing and other attributes for such variables. See L<perlfunc/our>.
394 =head2 Weak references
396 WARNING: This is an experimental feature.
398 change#3385, also need perlguts documentation
400 [TODO - Tuomas Lukka <lukka@fas.harvard.edu>]
402 =head2 File globbing implemented internally
404 WARNING: This is currently an experimental feature. Interfaces and
405 implementation are likely to change.
407 Perl now uses the File::Glob implementation of the glob() operator
408 automatically. This avoids using an external csh process and the
409 problems associated with it.
411 =head2 Binary numbers supported
413 Binary numbers are now supported as literals, in s?printf formats, and
417 printf "The answer is: %b\n", oct("0b101010");
419 =head2 Some arrows may be omitted in calls through references
421 Perl now allows the arrow to be omitted in many constructs
422 involving subroutine calls through references. For example,
423 C<$foo[10]->('foo')> may now be written C<$foo[10]('foo')>.
424 This is rather similar to how the arrow may be omitted from
425 C<$foo[10]->{'foo'}>. Note however, that the arrow is still
426 required for C<foo(10)->('bar')>.
428 =head2 exists() is supported on subroutine names
430 The exists() builtin now works on subroutine names. A subroutine
431 is considered to exist if it has been declared (even if implicitly).
432 See L<perlfunc/exists> for examples.
434 =head2 exists() and delete() are supported on array elements
436 The exists() and delete() builtins now work on simple arrays as well.
437 The behavior is similar to that on hash elements.
439 exists() can be used to check whether an array element has been
440 initialized without autovivifying it. If the array is tied, the
441 EXISTS() method in the corresponding tied package will be invoked.
443 delete() may be used to remove an element from the array and return
444 it. The array element at that position returns to its unintialized
445 state, so that testing for the same element with exists() will return
446 false. If the element happens to be the one at the end, the size of
447 the array also shrinks by one. If the array is tied, the DELETE() method
448 in the corresponding tied package will be invoked.
450 See L<perlfunc/exists> and L<perlfunc/delete> for examples.
452 =head2 syswrite() ease-of-use
454 The length argument of C<syswrite()> has become optional.
456 =head2 File and directory handles can be autovivified
458 Similar to how constructs such as C<$x->[0]> autovivify a reference,
459 handle constructors (open(), opendir(), pipe(), socketpair(), sysopen(),
460 socket(), and accept()) now autovivify a file or directory handle
461 if the handle passed to them is an uninitialized scalar variable. This
462 allows the constructs such as C<open(my $fh, ...)> and C<open(local $fh,...)>
463 to be used to create filehandles that will conveniently be closed
464 automatically when the scope ends, provided there are no other references
465 to them. This largely eliminates the need for typeglobs when opening
466 filehandles that must be passed around, as in the following example:
470 or die "Can't open '@_': $!";
475 my $f = myopen("</etc/motd");
477 # $f implicitly closed here
480 [TODO - this idiom needs more pod penetration]
482 =head2 64-bit support
484 All platforms that have 64-bit integers either (a) natively as longs
485 or ints (b) via special compiler flags (c) using long long are able to
486 use "quads" (64-integers) as follows:
492 constants (decimal, hexadecimal, octal, binary) in the code
496 arguments to oct() and hex()
500 arguments to print(), printf() and sprintf() (flag prefixes ll, L, q)
508 pack() and unpack() "q" and "Q" formats
512 in basic arithmetics: + - * / %
516 vec() (but see the below note about bit arithmetics)
520 Note that unless you have the case (a) you will have to configure
521 and compile Perl using the -Duse64bits Configure flag.
523 Unfortunately bit arithmetics (&, |, ^, ~, <<, >>) for numbers are not
524 64-bit clean, they are explictly forced to be 32-bit. Bit arithmetics
525 for bit vectors (created by vec()) are not limited in their width.
527 Last but not least: note that due to Perl's habit of always using
528 floating point numbers the quads are still not true integers.
529 When quads overflow their limits (0...18_446_744_073_709_551_615 unsigned,
530 -9_223_372_036_854_775_808...9_223_372_036_854_775_807 signed), they
531 are silently promoted to floating point numbers, after which they will
532 start losing precision (their lower digits).
534 =head2 Large file support
536 If you have filesystems that support "large files" (files larger than
537 2 gigabytes), you may now also be able to create and access them from
538 Perl. You have to use Configure -Duselargefiles. Turning on the
539 large file support turns on also the 64-bit support, for obvious reasons.
541 Note that in addition to requiring a proper file system to do large
542 files you may also need to adjust your per-process (or your
543 per-system, or per-process-group, or per-user-group) maximum filesize
544 limits before running Perl scripts that try to handle large files,
545 especially if you intend to write such files.
547 Finally, in addition to your process/process group maximum filesize
548 limits, you may have quota limits on your filesystems that stop you
549 (your user id or your user group id) from using large files.
551 Adjusting your process/user/group/file system/operating system limits
552 is outside the scope of Perl core language. For process limits, you
553 may try increasing the limits using your shell's limits/limit/ulimit
554 command before running Perl. The BSD::Resource extension (not
555 included with the standard Perl distribution) may also be of use, it
556 offers the getrlimit/setrlimit interface that can be used to adjust
557 process resource usage limits, including the maximum filesize limit.
561 In some systems you may be able to use long doubles to enhance the
562 range of precision of your double precision floating point numbers
563 (that is, Perl's numbers). Use Configure -Duselongdouble to enable
564 this support (if it is available).
568 You can Configure -Dusemorebits to turn on both the 64-bit support
569 and the long double support.
571 =head2 Enhanced support for sort() subroutines
573 Perl subroutines with a prototype of C<($$)> and XSUBs in general can
574 now be used as sort subroutines. In either case, the two elements to
575 be compared are passed as normal parameters in @_. See L<perlfunc/sort>.
577 For unprototyped sort subroutines, the historical behavior of passing
578 the elements to be compared as the global variables $a and $b remains
581 =head2 Better syntax checks on parenthesized unary operators
585 print defined(&foo,&bar,&baz);
586 print uc("foo","bar","baz");
589 used to be accidentally allowed in earlier versions, and produced
590 unpredictable behaviour. Some produced ancillary warnings
591 when used in this way; others silently did the wrong thing.
593 The parenthesized forms of most unary operators that expect a single
594 argument now ensure that they are not called with more than one
595 argument, making the cases shown above syntax errors. The usual
598 print defined &foo, &bar, &baz;
599 print uc "foo", "bar", "baz";
602 remains unchanged. See L<perlop>.
604 =head2 POSIX character class syntax [: :] supported
606 For example to match alphabetic characters use /[[:alpha:]]/.
607 See L<perlre> for details.
609 =head2 Improved C<qw//> operator
611 The C<qw//> operator is now evaluated at compile time into a true list
612 instead of being replaced with a run time call to C<split()>. This
613 removes the confusing misbehaviour of C<qw//> in scalar context, which
614 had inherited that behaviour from split().
618 $foo = ($bar) = qw(a b c); print "$foo|$bar\n";
620 now correctly prints "3|a", instead of "2|a".
622 =head2 pack() format 'Z' supported
624 The new format type 'Z' is useful for packing and unpacking null-terminated
625 strings. See L<perlfunc/"pack">.
627 =head2 pack() format modifier '!' supported
629 The new format type modifier '!' is useful for packing and unpacking
630 native shorts, ints, and longs. See L<perlfunc/"pack">.
632 =head2 pack() and unpack() support counted strings
634 The template character '/' can be used to specify a counted string
635 type to be packed or unpacked. See L<perlfunc/"pack">.
637 =head2 Comments in pack() templates
639 The '#' character in a template introduces a comment up to
640 end of the line. This facilitates documentation of pack()
643 =head2 $^X variables may now have names longer than one character
645 Formerly, $^X was synonymous with ${"\cX"}, but $^XY was a syntax
646 error. Now variable names that begin with a control character may be
647 arbitrarily long. However, for compatibility reasons, these variables
648 I<must> be written with explicit braces, as C<${^XY}> for example.
649 C<${^XYZ}> is synonymous with ${"\cXYZ"}. Variable names with more
650 than one control character, such as C<${^XY^Z}>, are illegal.
652 The old syntax has not changed. As before, `^X' may be either a
653 literal control-X character or the two-character sequence `caret' plus
654 `X'. When braces are omitted, the variable name stops after the
655 control character. Thus C<"$^XYZ"> continues to be synonymous with
656 C<$^X . "YZ"> as before.
658 As before, lexical variables may not have names beginning with control
659 characters. As before, variables whose names begin with a control
660 character are always forced to be in package `main'. All such variables
661 are reserved for future extensions, except those that begin with
662 C<^_>, which may be used by user programs and are guaranteed not to
663 acquire special meaning in any future version of Perl.
665 =head2 C<use attrs> implicit in subroutine attributes
667 Formerly, if you wanted to mark a subroutine as being a method call or
668 as requiring an automatic lock() when it is entered, you had to declare
669 that with a C<use attrs> pragma in the body of the subroutine.
670 That can now be accomplished with a declaration syntax, like this:
672 sub mymethod : locked, method ;
674 sub mymethod : locked, method {
678 F<AutoSplit.pm> and F<SelfLoader.pm> have been updated to keep the attributes
679 with the stubs they provide. See L<attributes>.
681 =head2 Regular expression improvements
683 change#2827,2373,2372,2365,1813,1800,4112,4158,4215,4301
684 [TODO - Ilya Zakharevich <ilya@math.ohio-state.edu>]
686 =head2 Overloading improvements
689 [TODO - Ilya Zakharevich <ilya@math.ohio-state.edu>]
691 =head2 open() with more than two arguments
693 [TODO - Ilya Zakharevich <ilya@math.ohio-state.edu>]
695 =head2 Support for interpolating named characters
698 [TODO - Ilya Zakharevich <ilya@math.ohio-state.edu>]
700 =head2 Experimental support for user-hooks in @INC
702 [TODO - Ken Fox <kfox@ford.com>]
704 =head2 C<require> and C<do> may be overridden
706 C<require> and C<do 'file'> operations may be overridden locally
707 by importing subroutines of the same name into the current package
708 (or globally by importing them into the CORE::GLOBAL:: namespace).
709 Overriding C<require> will also affect C<use>, provided the override
710 is visible at compile-time.
711 See L<perlsub/"Overriding Built-in Functions">.
713 =head2 New variable $^C reflects C<-c> switch
715 C<$^C> has a boolean value that reflects whether perl is being run
716 in compile-only mode (i.e. via the C<-c> switch). Since
717 BEGIN blocks are executed under such conditions, this variable
718 enables perl code to determine whether actions that make sense
719 only during normal running are warranted. See L<perlvar>.
721 =head2 Optional Y2K warnings
723 If Perl is built with the cpp macro C<PERL_Y2KWARN> defined,
724 it emits optional warnings when concatenating the number 19
727 This behavior must be specifically enabled when running Configure.
728 See L<INSTALL> and L<README.Y2K>.
730 =head1 Significant bug fixes
732 =head2 E<lt>HANDLEE<gt> on empty files
734 With C<$/> set to C<undef>, "slurping" an empty file returns a string of
735 zero length (instead of C<undef>, as it used to) the first time the
736 HANDLE is read after C<$/> is set to C<undef>. Further reads yield
739 This means that the following will append "foo" to an empty file (it used
742 perl -0777 -pi -e 's/^/foo/' empty_file
746 perl -pi -e 's/^/foo/' empty_file
748 is unchanged (it continues to leave the file empty).
750 =head2 C<eval '...'> improvements
752 Line numbers (as reflected by caller() and most diagnostics) within
753 C<eval '...'> were often incorrect when here documents were involved.
754 This has been corrected.
756 Lexical lookups for variables appearing in C<eval '...'> within
757 functions that were themselves called within an C<eval '...'> were
758 searching the wrong place for lexicals. The lexical search now
759 correctly ends at the subroutine's block boundary.
761 Parsing of here documents used to be flawed when they appeared as
762 the replacement expression in C<eval 's/.../.../e'>. This has
765 =head2 All compilation errors are true errors
767 Some "errors" encountered at compile time were by neccessity
768 generated as warnings followed by eventual termination of the
769 program. This enabled more such errors to be reported in a
770 single run, rather than causing a hard stop at the first error
771 that was encountered.
773 The mechanism for reporting such errors has been reimplemented
774 to queue compile-time errors and report them at the end of the
775 compilation as true errors rather than as warnings. This fixes
776 cases where error messages leaked through in the form of warnings
777 when code was compiled at run time using C<eval STRING>, and
778 also allows such errors to be reliably trapped using __DIE__ hooks.
780 =head2 Automatic flushing of output buffers
782 fork(), exec(), system(), qx//, and pipe open()s now flush buffers
783 of all files opened for output when the operation
784 was attempted. This mostly eliminates confusing
785 buffering mishaps suffered by users unaware of how Perl internally
788 =head2 Better diagnostics on meaningless filehandle operations
790 Constructs such as C<open(E<lt>FHE<gt>)> and C<close(E<lt>FHE<gt>)>
791 are compile time errors. Attempting to read from filehandles that
792 were opened only for writing will now produce warnings (just as
793 writing to read-only filehandles does).
795 =head2 Where possible, buffered data discarded from duped input filehandle
797 C<open(NEW, "E<lt>&OLD")> now attempts to discard any data that
798 was previously read and buffered in C<OLD> before duping the handle.
799 On platforms where doing this is allowed, the next read operation
800 on C<NEW> will return the same data as the corresponding operation
801 on C<OLD>. Formerly, it would have returned the data from the start
802 of the following disk block instead.
804 =head2 eof() has the same old magic as <>
806 C<eof()> would return true if no attempt to read from C<E<lt>E<gt>> had
807 yet been made. C<eof()> has been changed to have a little magic of its
808 own, it now opens the C<E<lt>E<gt>> files.
810 =head2 system(), backticks and pipe open now reflect exec() failure
812 On Unix and similar platforms, system(), qx() and open(FOO, "cmd |")
813 etc., are implemented via fork() and exec(). When the underlying
814 exec() fails, earlier versions did not report the error properly,
815 since the exec() happened to be in a different process.
817 The child process now communicates with the parent about the
818 error in launching the external command, which allows these
819 constructs to return with their usual error value and set $!.
821 =head2 Implicitly closed filehandles are safer
823 Sometimes implicitly closed filehandles (as when they are localized,
824 and Perl automatically closes them on exiting the scope) could
825 inadvertently set $? or $!. This has been corrected.
827 =head2 C<(\$)> prototype and C<$foo{a}>
829 An scalar reference prototype now correctly allows a hash or
830 array element in that slot.
832 =head2 Pseudo-hashes work better
834 Dereferencing some types of reference values in a pseudo-hash,
835 such as C<$ph->{foo}[1]>, was accidentally disallowed. This has
838 When applied to a pseudo-hash element, exists() now reports whether
839 the specified value exists, not merely if the key is valid.
841 delete() now works on pseudo-hashes. When given a pseudo-hash element
842 or slice it deletes the values corresponding to the keys (but not the keys
843 themselves). See L<perlref/"Pseudo-hashes: Using an array as a hash">.
845 =head2 C<goto &sub> and AUTOLOAD
847 The C<goto &sub> construct works correctly when C<&sub> happens
850 =head2 C<-bareword> allowed under C<use integer>
852 The autoquoting of barewords preceded by C<-> did not work
853 in prior versions when the C<integer> pragma was enabled.
856 =head2 Boolean assignment operators are legal lvalues
858 Constructs such as C<($a ||= 2) += 1> are now allowed.
860 =head2 C<sort $coderef @foo> allowed
862 sort() did not accept a subroutine reference as the comparison
863 function in earlier versions. This is now permitted.
865 =head2 Failures in DESTROY()
867 When code in a destructor threw an exception, it went unnoticed
868 in earlier versions of Perl, unless someone happened to be
869 looking in $@ just after the point the destructor happened to
870 run. Such failures are now visible as warnings when warnings are
873 =head2 Locale bugs fixed
875 printf() and sprintf() previously reset the numeric locale
876 back to the default "C" locale. This has been fixed.
878 Numbers formatted according to the local numeric locale
879 (such as using a decimal comma instead of a decimal dot) caused
880 "isn't numeric" warnings, even while the operations accessing
881 those numbers produced correct results. The warnings are gone.
885 The C<eval 'return sub {...}'> construct could sometimes leak
886 memory. This has been fixed.
888 Operations that aren't filehandle constructors used to leak memory
889 when used on invalid filehandles. This has been fixed.
891 Constructs that modified C<@_> could fail to deallocate values
892 in C<@_> and thus leak memory. This has been corrected.
894 =head2 Spurious subroutine stubs after failed subroutine calls
896 Perl could sometimes create empty subroutine stubs when a
897 subroutine was not found in the package. Such cases stopped
898 later method lookups from progressing into base packages.
899 This has been corrected.
901 =head2 Consistent numeric conversions
904 [TODO - Ilya Zakharevich <ilya@math.ohio-state.edu>]
906 =head2 Taint failures under C<-U>
908 When running in unsafe mode, taint violations could sometimes
909 cause silent failures. This has been fixed.
911 =head2 END blocks and the C<-c> switch
913 Prior versions used to run BEGIN B<and> END blocks when Perl was
914 run in compile-only mode. Since this is typically not the expected
915 behavior, END blocks are not executed anymore when the C<-c> switch
918 See L<STOP blocks> for how to run things when the compile phase ends.
920 =head2 Potential to leak DATA filehandles
922 Using the C<__DATA__> token creates an implicit filehandle to
923 the file that contains the token. It is the program's
924 responsibility to close it when it is done reading from it.
926 This caveat is now better explained in the documentation.
929 =head2 Diagnostics follow STDERR
931 Diagnostic output now goes to whichever file the C<STDERR> handle
932 is pointing at, instead of always going to the underlying C runtime
935 =head2 Other fixes for better diagnostics
937 Line numbers are no longer suppressed (under most likely circumstances)
938 during the global destruction phase.
940 Diagnostics emitted from code running in threads other than the main
941 thread are now accompanied by the thread ID.
943 Embedded null characters in diagnostics now actually show up. They
944 used to truncate the message in prior versions.
946 $foo::a and $foo::b are now exempt from "possible typo" warnings only
947 if sort() is encountered in package foo.
949 Unrecognized alphabetic escapes encountered when parsing quote
950 constructs now generate a warning, since they may take on new
951 semantics in later versions of Perl.
953 =head1 Performance enhancements
955 =head2 Simple sort() using { $a <=> $b } and the like are optimized
957 Many common sort() operations using a simple inlined block are now
958 optimized for faster performance.
960 =head2 Optimized assignments to lexical variables
962 Certain operations in the RHS of assignment statements have been
963 optimized to directly set the lexical variable on the LHS,
964 eliminating redundant copying overheads.
966 =head2 Method lookups optimized
968 [TODO - Chip Salzenberg <chip@perlsupport.com>]
970 =head2 Faster mechanism to invoke XSUBs
973 [TODO - Ilya Zakharevich <ilya@math.ohio-state.edu>]
975 =head2 Perl_malloc() improvements
978 [TODO - Ilya Zakharevich <ilya@math.ohio-state.edu>]
980 =head2 Faster subroutine calls
982 Minor changes in how subroutine calls are handled internally
983 provide marginal improvements in performance.
985 =head1 Platform specific changes
987 =head2 Additional supported platforms
993 VM/ESA is now supported.
997 Siemens BS2000 is now supported under the POSIX Shell.
1001 The Mach CThreads (NEXTSTEP, OPENSTEP) are now supported by the Thread
1006 GNU/Hurd is now supported.
1010 Rhapsody is now supported.
1014 EPOC is is now supported (on Psion 5).
1024 Perl now works with djgpp 2.02 (and 2.03 alpha).
1028 Environment variable names are not converted to uppercase any more.
1032 Wrong exit code from backticks now fixed.
1036 This port is still using its own builtin globbing.
1042 [TODO - Ilya Zakharevich <ilya@math.ohio-state.edu>]
1046 [TODO - Charles Bailey <bailey@newman.upenn.edu>]
1050 Site library searches failed to look for ".../site/5.XXX/lib"
1051 if ".../site/5.XXXYY/lib" wasn't found. This has been corrected.
1053 When given a pathname that consists only of a drivename, such
1054 as C<A:>, opendir() and stat() now use the current working
1055 directory for the drive rather than the drive root.
1057 The builtin XSUB functions in the Win32:: namespace are
1058 documented. See L<Win32>.
1060 $^X now contains the full path name of the running executable.
1062 A Win32::GetLongPathName() function is provided to complement
1063 Win32::GetFullPathName() and Win32::GetShortPathName(). See L<Win32>.
1065 POSIX::uname() is supported.
1067 system(1,...) now returns true process IDs rather than process
1068 handles. kill() accepts any real process id, rather than strictly
1069 return values from system(1,...).
1071 The C<Shell> module is supported.
1073 Rudimentary support for building under command.com in Windows 95
1076 Scripts are read in binary mode by default to allow ByteLoader (and
1077 the filter mechanism in general) to work properly. For compatibility,
1078 the DATA filehandle will be set to text mode if a carriage return is
1079 detected at the end of the line containing the __END__ or __DATA__
1080 token; if not, the DATA filehandle will be left open in binary mode.
1081 Earlier versions always opened the DATA filehandle in text mode.
1091 Compatibility tests for C<sub : attrs> vs the older C<use attrs>.
1095 IO constants (SEEK_*, _IO*).
1099 Directory-related IO methods (new, read, close, rewind, tied delete).
1101 =item lib/io_multihomed
1103 INET sockets with multi-homed hosts.
1115 Regression tests for C<my ($x,@y,%z) : attrs> and <sub : attrs>.
1119 File test operators.
1123 Verify operations that access pad objects (lexicals and temporaries).
1127 Verify C<exists &sub> operations.
1131 =head1 Modules and Pragmata
1139 While used internally by Perl as a pragma, this module also
1140 provides a way to fetch subroutine and variable attributes.
1145 The Perl Compiler suite has been extensively reworked for this
1148 [TODO - Vishal Bhatia <vishal@gol.com>,
1149 Nick Ing-Simmons <nick@ni-s.u-net.com>]
1153 The ByteLoader is a dedicated extension to generate and run
1154 Perl bytecode. See L<ByteLoader>.
1158 References can now be used.
1160 The new version also allows a leading underscore in constant names, but
1161 disallows a double leading underscore (as in "__LINE__"). Some other names
1162 are disallowed or warned against, including BEGIN, END, etc. Some names
1163 which were forced into main:: used to fail silently in some cases; now they're
1164 fatal (outside of main::) and an optional warning (inside of main::).
1165 The ability to detect whether a constant had been set with a given name has
1173 [TODO - Ilya Zakharevich <ilya@math.ohio-state.edu>]
1177 A C<Maxdepth> setting can be specified to avoid venturing
1178 too deeply into deep data structures. See L<Data::Dumper>.
1180 Dumping C<qr//> objects works correctly.
1184 C<DB> is an experimental module that exposes a clean abstraction
1185 to Perl's debugging API.
1189 DB_File can now be built with Berkeley DB versions 1, 2 or 3.
1190 See C<ext/DB_File/Changes>.
1194 Devel::DProf, a Perl source code profiler has been added. See
1195 L<Devel::DProf> and L<dprofpp>.
1199 The Dumpvalue module provides screen dumps of Perl data.
1203 Overall, Benchmark results exhibit lower average error and better timing
1206 You can now run tests for I<n> seconds instead of guessing the right
1207 number of tests to run: e.g. timethese(-5, ...) will run each
1208 code for at least 5 CPU seconds. Zero as the "number of repetitions"
1209 means "for at least 3 CPU seconds". The output format has also
1210 changed. For example:
1212 use Benchmark;$x=3;timethese(-5,{a=>sub{$x*$x},b=>sub{$x**2}})
1214 will now output something like this:
1216 Benchmark: running a, b, each for at least 5 CPU seconds...
1217 a: 5 wallclock secs ( 5.77 usr + 0.00 sys = 5.77 CPU) @ 200551.91/s (n=1156516)
1218 b: 4 wallclock secs ( 5.00 usr + 0.02 sys = 5.02 CPU) @ 159605.18/s (n=800686)
1220 New features: "each for at least N CPU seconds...", "wallclock secs",
1221 and the "@ operations/CPU second (n=operations)".
1223 timethese() now returns a reference to a hash of Benchmark objects containing
1224 the test results, keyed on the names of the tests.
1226 timethis() now returns the iterations field in the Benchmark result object
1229 timethese(), timethis(), and the new cmpthese() (see below) can also take
1230 a format specifier of 'none' to suppress output.
1232 A new function countit() is just like timeit() except that it takes a
1233 TIME instead of a COUNT.
1235 A new function cmpthese() prints a chart comparing the results of each test
1236 returned from a timethese() call. For each possible pair of tests, the
1237 percentage speed difference (iters/sec or seconds/iter) is shown.
1239 For other details, see L<Benchmark>.
1243 The Devel::Peek module provides access to the internal representation
1244 of Perl variables and data. It is a data debugging tool for the XS programmer.
1246 =item ExtUtils::MakeMaker
1248 change#4135, also needs docs in module pod
1249 [TODO - Ilya Zakharevich <ilya@math.ohio-state.edu>]
1253 More Fcntl constants added: F_SETLK64, F_SETLKW64, O_LARGEFILE for
1254 large (more than 4G) file access (64-bit support is not yet
1255 working, though, so no need to get overly excited), Free/Net/OpenBSD
1256 locking behaviour flags F_FLOCK, F_POSIX, Linux F_SHLCK, and
1257 O_ACCMODE: the mask of O_RDONLY, O_WRONLY, and O_RDWR.
1261 A compare_text() function has been added, which allows custom
1262 comparison functions. See L<File::Compare>.
1266 File::Find now works correctly when the wanted() function is either
1267 autoloaded or is a symbolic reference.
1269 A bug that caused File::Find to lose track of the working directory
1270 when pruning top-level directories has been fixed.
1272 File::Find now also supports several other options to control its
1273 behavior. It can follow symbolic links if the C<follow> option is
1274 specified. Enabling the C<no_chdir> option will make File::Find skip
1275 changing the current directory when walking directories. The C<untaint>
1276 flag can be useful when running with taint checks enabled.
1282 This extension implements BSD-style file globbing. By default,
1283 it will also be used for the internal implementation of the glob()
1284 operator. See L<File::Glob>.
1288 New methods have been added to the File::Spec module: devnull() returns
1289 the name of the null device (/dev/null on Unix) and tmpdir() the name of
1290 the temp directory (normally /tmp on Unix). There are now also methods
1291 to convert between absolute and relative filenames: abs2rel() and
1292 rel2abs(). For compatibility with operating systems that specify volume
1293 names in file paths, the splitpath(), splitdir(), and catdir() methods
1296 =item File::Spec::Functions
1298 The new File::Spec::Functions modules provides a function interface
1299 to the File::Spec module. Allows shorthand
1301 $fullname = catfile($dir1, $dir2, $file);
1305 $fullname = File::Spec->catfile($dir1, $dir2, $file);
1309 Getopt::Long licensing has changed to allow the Perl Artistic License
1310 as well as the GPL. It used to be GPL only, which got in the way of
1311 non-GPL applications that wanted to use Getopt::Long.
1313 Getopt::Long encourages the use of Pod::Usage to produce help
1314 messages. For example:
1320 GetOptions('help|?' => \$help, man => \$man) or pod2usage(2);
1321 pod2usage(1) if $help;
1322 pod2usage(-exitstatus => 0, -verbose => 2) if $man;
1328 sample - Using GetOpt::Long and Pod::Usage
1332 sample [options] [file ...]
1335 -help brief help message
1336 -man full documentation
1344 Print a brief help message and exits.
1348 Prints the manual page and exits.
1354 B<This program> will read the given input file(s) and do someting
1355 useful with the contents thereof.
1359 See L<Pod::Usage> for details.
1361 A bug that prevented the non-option call-back E<lt>E<gt> from being
1362 specified as the first argument has been fixed.
1364 To specify the characters E<lt> and E<gt> as option starters, use
1365 E<gt>E<lt>. Note, however, that changing option starters is strongly
1370 write() and syswrite() will now accept a single-argument
1371 form of the call, for consistency with Perl's syswrite().
1373 You can now create a TCP-based IO::Socket::INET without forcing
1374 a connect attempt. This allows you to configure its options
1375 (like making it non-blocking) and then call connect() manually.
1377 A bug that prevented the IO::Socket::protocol() accessor
1378 from ever returning the correct value has been corrected.
1382 Java Perl Lingo is now distributed with Perl. See jpl/README
1383 for more information.
1387 C<use lib> now weeds out any trailing duplicate entries.
1388 C<no lib> removes all named entries.
1392 The bitwise operations C<E<lt>E<lt>>, C<E<gt>E<gt>>, C<&>, C<|>,
1393 and C<~> are now supported on bigints.
1397 The accessor methods Re, Im, arg, abs, rho, and theta can now also
1398 act as mutators (accessor $z->Re(), mutator $z->Re(3)).
1402 A little bit of radial trigonometry (cylindrical and spherical),
1403 radial coordinate conversions, and the great circle distance were added.
1407 [TODO - Brad Appleton <bradapp@enteract.com>]
1409 =item Pod::Text and Pod::Man
1411 [TODO - Russ Allbery <rra@stanford.edu>]
1415 An EXISTS method has been added to this module (and sdbm_exists() has
1416 been added to the underlying sdbm library), so one can now call exists
1417 on an SDBM_File tied hash and get the correct result, rather than a
1420 A bug that may have caused data loss when more than one disk block
1421 happens to be read from the database in a single FETCH() has been
1426 The timelocal() and timegm() functions used to silently return bogus
1427 results when the date fell outside the machine's integer range. They
1428 now consistently croak() if the date falls in an unsupported range.
1432 The error return value in list context has been changed for all functions
1433 that return a list of values. Previously these functions returned a list
1434 with a single element C<undef> if an error occurred. Now these functions
1435 return the empty list in these situations. This applies to the following
1441 The remaining functions are unchanged and continue to return C<undef> on
1442 error even in list context.
1444 The Win32::SetLastError(ERROR) function has been added as a complement
1445 to the Win32::GetLastError() function.
1447 The new Win32::GetFullPathName(FILENAME) returns the full absolute
1448 pathname for FILENAME in scalar context. In list context it returns
1449 a two-element list containing the fully qualified directory name and
1450 the filename. See L<Win32>.
1454 A new feature called "DBM Filters" has been added to all the
1455 DBM modules--DB_File, GDBM_File, NDBM_File, ODBM_File, and SDBM_File.
1456 DBM Filters add four new methods to each DBM module:
1463 These can be used to filter key-value pairs before the pairs are
1464 written to the database or just after they are read from the database.
1465 See L<perldbmfilter> for further information.
1471 C<use attrs> is now obsolete, and is only provided for
1472 backward-compatibility. It's been replaced by the C<sub : attributes>
1473 syntax. See L<perlsub/"Subroutine Attributes"> and L<attributes>.
1475 C<use utf8> to enable UTF-8 and Unicode support.
1477 C<use caller 'encoding'> allows modules to inherit pragmatic attributes
1478 from the caller's context. C<encoding> is currently the only supported
1481 Lexical warnings pragma, C<use warnings;>, to control optional warnings.
1484 C<use filetest> to control the behaviour of filetests (C<-r> C<-w>
1485 ...). Currently only one subpragma implemented, "use filetest
1486 'access';", that uses access(2) or equivalent to check permissions
1487 instead of using stat(2) as usual. This matters in filesystems
1488 where there are ACLs (access control lists): the stat(2) might lie,
1489 but access(2) knows better.
1491 =head1 Utility Changes
1495 [TODO - Kurt Starsinic <kstar@chapin.edu>]
1499 C<perlcc> now supports the C and Bytecode backends. By default,
1500 it generates output from the simple C backend rather than the
1501 optimized C backend.
1503 Support for non-Unix platforms has been improved.
1508 [TODO - Ilya Zakharevich <ilya@math.ohio-state.edu>]
1510 =head1 Documentation Changes
1514 =item perlcompile.pod
1516 An introduction to using the Perl Compiler suite.
1518 =item perlfilter.pod
1520 An introduction to writing Perl source filters.
1524 Some guidelines for hacking the Perl source code.
1526 =item perlopentut.pod
1528 A tutorial on using open() effectively.
1530 =item perlreftut.pod
1532 A tutorial that introduces the essentials of references.
1536 A tutorial on managing class data for object modules.
1540 =head1 New or Changed Diagnostics
1544 =item "my sub" not yet implemented
1546 (F) Lexically scoped subroutines are not yet implemented. Don't try that
1549 =item '!' allowed only after types %s
1551 (F) The '!' is allowed in pack() and unpack() only after certain types.
1552 See L<perlfunc/pack>.
1554 =item / cannot take a count
1556 (F) You had an unpack template indicating a counted-length string,
1557 but you have also specified an explicit size for the string.
1558 See L<perlfunc/pack>.
1560 =item / must be followed by a, A or Z
1562 (F) You had an unpack template indicating a counted-length string,
1563 which must be followed by one of the letters a, A or Z
1564 to indicate what sort of string is to be unpacked.
1565 See L<perlfunc/pack>.
1567 =item / must be followed by a*, A* or Z*
1569 (F) You had a pack template indicating a counted-length string,
1570 Currently the only things that can have their length counted are a*, A* or Z*.
1571 See L<perlfunc/pack>.
1573 =item / must follow a numeric type
1575 (F) You had an unpack template that contained a '#',
1576 but this did not follow some numeric unpack specification.
1577 See L<perlfunc/pack>.
1579 =item /%s/: Unrecognized escape \\%c passed through
1581 (W) You used a backslash-character combination which is not recognized
1582 by Perl. This combination appears in an interpolated variable or a
1583 C<'>-delimited regular expression. The character was understood literally.
1585 =item /%s/: Unrecognized escape \\%c in character class passed through
1587 (W) You used a backslash-character combination which is not recognized
1588 by Perl inside character classes. The character was understood literally.
1590 =item /%s/ should probably be written as "%s"
1592 (W) You have used a pattern where Perl expected to find a string,
1593 as in the first argument to C<join>. Perl will treat the true
1594 or false result of matching the pattern against $_ as the string,
1595 which is probably not what you had in mind.
1597 =item %s() called too early to check prototype
1599 (W) You've called a function that has a prototype before the parser saw a
1600 definition or declaration for it, and Perl could not check that the call
1601 conforms to the prototype. You need to either add an early prototype
1602 declaration for the subroutine in question, or move the subroutine
1603 definition ahead of the call to get proper prototype checking. Alternatively,
1604 if you are certain that you're calling the function correctly, you may put
1605 an ampersand before the name to avoid the warning. See L<perlsub>.
1607 =item %s argument is not a subroutine name
1609 (F) The argument to exists() for C<exists &sub> must be a subroutine
1610 name, and not a subroutine call. C<exists &sub()> will generate this error.
1612 =item %s package attribute may clash with future reserved word: %s
1614 (W) A lowercase attribute name was used that had a package-specific handler.
1615 That name might have a meaning to Perl itself some day, even though it
1616 doesn't yet. Perhaps you should use a mixed-case attribute name, instead.
1619 =item (in cleanup) %s
1621 (W) This prefix usually indicates that a DESTROY() method raised
1622 the indicated exception. Since destructors are usually called by
1623 the system at arbitrary points during execution, and often a vast
1624 number of times, the warning is issued only once for any number
1625 of failures that would otherwise result in the same message being
1628 Failure of user callbacks dispatched using the C<G_KEEPERR> flag
1629 could also result in this warning. See L<perlcall/G_KEEPERR>.
1631 =item <> should be quotes
1633 (F) You wrote C<require E<lt>fileE<gt>> when you should have written
1636 =item Attempt to join self
1638 (F) You tried to join a thread from within itself, which is an
1639 impossible task. You may be joining the wrong thread, or you may
1640 need to move the join() to some other thread.
1642 =item Bad evalled substitution pattern
1644 (F) You've used the /e switch to evaluate the replacement for a
1645 substitution, but perl found a syntax error in the code to evaluate,
1646 most likely an unexpected right brace '}'.
1648 =item Bad realloc() ignored
1650 (S) An internal routine called realloc() on something that had never been
1651 malloc()ed in the first place. Mandatory, but can be disabled by
1652 setting environment variable C<PERL_BADFREE> to 1.
1654 =item Binary number > 0b11111111111111111111111111111111 non-portable
1656 (W) The binary number you specified is larger than 2**32-1
1657 (4294967295) and therefore non-portable between systems. See
1658 L<perlport> for more on portability concerns.
1660 =item Bit vector size > 32 non-portable
1662 (W) Using bit vector sizes larger than 32 is non-portable.
1664 =item Buffer overflow in prime_env_iter: %s
1666 (W) A warning peculiar to VMS. While Perl was preparing to iterate over
1667 %ENV, it encountered a logical name or symbol definition which was too long,
1668 so it was truncated to the string shown.
1670 =item Can't check filesystem of script "%s"
1672 (P) For some reason you can't check the filesystem of the script for nosuid.
1674 =item Can't ignore signal CHLD, forcing to default
1676 (W) Perl has detected that it is being run with the SIGCHLD signal
1677 (sometimes known as SIGCLD) disabled. Since disabling this signal
1678 will interfere with proper determination of exit status of child
1679 processes, Perl has reset the signal to its default value.
1680 This situation typically indicates that the parent program under
1681 which Perl may be running (e.g. cron) is being very careless.
1683 =item Can't modify non-lvalue subroutine call
1685 (F) Subroutines meant to be used in lvalue context should be declared as
1686 such, see L<perlsub/"Lvalue subroutines">.
1688 =item Can't read CRTL environ
1690 (S) A warning peculiar to VMS. Perl tried to read an element of %ENV
1691 from the CRTL's internal environment array and discovered the array was
1692 missing. You need to figure out where your CRTL misplaced its environ
1693 or define F<PERL_ENV_TABLES> (see L<perlvms>) so that environ is not searched.
1695 =item Can't remove %s: %s, skipping file
1697 (S) You requested an inplace edit without creating a backup file. Perl
1698 was unable to remove the original file to replace it with the modified
1699 file. The file was left unmodified.
1701 =item Can't return %s from lvalue subroutine
1703 (F) Perl detected an attempt to return illegal lvalues (such
1704 as temporary or readonly values) from a subroutine used as an lvalue.
1705 This is not allowed.
1707 =item Can't weaken a nonreference
1709 (F) You attempted to weaken something that was not a reference. Only
1710 references can be weakened.
1712 =item Character class [:%s:] unknown
1714 (F) The class in the character class [: :] syntax is unknown.
1717 =item Character class syntax [%s] belongs inside character classes
1719 (W) The character class constructs [: :], [= =], and [. .] go
1720 I<inside> character classes, the [] are part of the construct,
1721 for example: /[012[:alpha:]345]/. Note that [= =] and [. .]
1722 are not currently implemented; they are simply placeholders for
1725 =item Constant is not %s reference
1727 (F) A constant value (perhaps declared using the C<use constant> pragma)
1728 is being dereferenced, but it amounts to the wrong type of reference. The
1729 message indicates the type of reference that was expected. This usually
1730 indicates a syntax error in dereferencing the constant value.
1731 See L<perlsub/"Constant Functions"> and L<constant>.
1733 =item constant(%s): %%^H is not localized
1735 (F) When setting compile-time-lexicalized hash %^H one should set the
1736 corresponding bit of $^H as well.
1738 =item constant(%s): %s
1740 (F) Compile-time-substitutions (such as overloaded constants and
1741 character names) were not correctly set up.
1743 =item defined(@array) is deprecated
1745 (D) defined() is not usually useful on arrays because it checks for an
1746 undefined I<scalar> value. If you want to see if the array is empty,
1747 just use C<if (@array) { # not empty }> for example.
1749 =item defined(%hash) is deprecated
1751 (D) defined() is not usually useful on hashes because it checks for an
1752 undefined I<scalar> value. If you want to see if the hash is empty,
1753 just use C<if (%hash) { # not empty }> for example.
1755 =item Did not produce a valid header
1759 =item Document contains no data
1763 =item entering effective %s failed
1765 (F) While under the C<use filetest> pragma, switching the real and
1766 effective uids or gids failed.
1768 =item false [] range "%s" in regexp
1770 (W) A character class range must start and end at a literal character, not
1771 another character class like C<\d> or C<[:alpha:]>. The "-" in your false
1772 range is interpreted as a literal "-". Consider quoting the "-", "\-".
1775 =item Filehandle %s opened only for output
1777 (W) You tried to read from a filehandle opened only for writing. If you
1778 intended it to be a read/write filehandle, you needed to open it with
1779 "+E<lt>" or "+E<gt>" or "+E<gt>E<gt>" instead of with "E<lt>" or nothing. If
1780 you intended only to read from the file, use "E<lt>". See
1783 =item Hexadecimal number > 0xffffffff non-portable
1785 (W) The hexadecimal number you specified is larger than 2**32-1
1786 (4294967295) and therefore non-portable between systems. See
1787 L<perlport> for more on portability concerns.
1789 =item Ill-formed CRTL environ value "%s"
1791 (W) A warning peculiar to VMS. Perl tried to read the CRTL's internal
1792 environ array, and encountered an element without the C<=> delimiter
1793 used to spearate keys from values. The element is ignored.
1795 =item Ill-formed message in prime_env_iter: |%s|
1797 (W) A warning peculiar to VMS. Perl tried to read a logical name
1798 or CLI symbol definition when preparing to iterate over %ENV, and
1799 didn't see the expected delimiter between key and value, so the
1802 =item Illegal binary digit %s
1804 (F) You used a digit other than 0 or 1 in a binary number.
1806 =item Illegal binary digit %s ignored
1808 (W) You may have tried to use a digit other than 0 or 1 in a binary number.
1809 Interpretation of the binary number stopped before the offending digit.
1811 =item Illegal number of bits in vec
1813 (F) The number of bits in vec() (the third argument) must be a power of
1814 two from 1 to 32 (or 64, if your platform supports that).
1816 =item Integer overflow in %s number
1818 (W) The hexadecimal, octal or binary number you have specified either
1819 as a literal or as an argument to hex() or oct() is too big for your
1820 architecture, and has been converted to a floating point number. On a
1821 32-bit architecture the largest hexadecimal, octal or binary number
1822 representable without overflow is 0xFFFFFFFF, 037777777777, or
1823 0b11111111111111111111111111111111 respectively. Note that Perl
1824 transparently promotes all numbers to a floating point representation
1825 internally--subject to loss of precision errors in subsequent
1828 =item Invalid %s attribute: %s
1830 The indicated attribute for a subroutine or variable was not recognized
1831 by Perl or by a user-supplied handler. See L<attributes>.
1833 =item Invalid %s attributes: %s
1835 The indicated attributes for a subroutine or variable were not recognized
1836 by Perl or by a user-supplied handler. See L<attributes>.
1838 =item invalid [] range "%s" in regexp
1840 The offending range is now explicitly displayed.
1842 =item Invalid separator character %s in attribute list
1844 (F) Something other than a comma or whitespace was seen between the
1845 elements of an attribute list. If the previous attribute
1846 had a parenthesised parameter list, perhaps that list was terminated
1847 too soon. See L<attributes>.
1849 =item Invalid separator character %s in subroutine attribute list
1851 (F) Something other than a comma or whitespace was seen between the
1852 elements of a subroutine attribute list. If the previous attribute
1853 had a parenthesised parameter list, perhaps that list was terminated
1856 =item leaving effective %s failed
1858 (F) While under the C<use filetest> pragma, switching the real and
1859 effective uids or gids failed.
1861 =item Lvalue subs returning %s not implemented yet
1863 (F) Due to limitations in the current implementation, array and hash
1864 values cannot be returned in subroutines used in lvalue context.
1865 See L<perlsub/"Lvalue subroutines">.
1867 =item Method %s not permitted
1871 =item Missing %sbrace%s on \N{}
1873 (F) Wrong syntax of character name literal C<\N{charname}> within
1874 double-quotish context.
1876 =item Missing command in piped open
1878 (W) You used the C<open(FH, "| command")> or C<open(FH, "command |")>
1879 construction, but the command was missing or blank.
1881 =item Missing name in "my sub"
1883 (F) The reserved syntax for lexically scoped subroutines requires that they
1884 have a name with which they can be found.
1886 =item no UTC offset information; assuming local time is UTC
1888 (S) A warning peculiar to VMS. Perl was unable to find the local
1889 timezone offset, so it's assuming that local system time is equivalent
1890 to UTC. If it's not, define the logical name F<SYS$TIMEZONE_DIFFERENTIAL>
1891 to translate to the number of seconds which need to be added to UTC to
1894 =item Octal number > 037777777777 non-portable
1896 (W) The octal number you specified is larger than 2**32-1 (4294967295)
1897 and therefore non-portable between systems. See L<perlport> for more
1898 on portability concerns.
1900 See also L<perlport> for writing portable code.
1902 =item panic: del_backref
1904 (P) Failed an internal consistency check while trying to reset a weak
1907 =item panic: kid popen errno read
1909 (F) forked child returned an incomprehensible message about its errno.
1911 =item panic: magic_killbackrefs
1913 (P) Failed an internal consistency check while trying to reset all weak
1914 references to an object.
1916 =item Possible Y2K bug: %s
1918 (W) You are concatenating the number 19 with another number, which
1919 could be a potential Year 2000 problem.
1921 =item Premature end of script headers
1925 =item Repeat count in pack overflows
1927 (F) You can't specify a repeat count so large that it overflows
1928 your signed integers. See L<perlfunc/pack>.
1930 =item Repeat count in unpack overflows
1932 (F) You can't specify a repeat count so large that it overflows
1933 your signed integers. See L<perlfunc/unpack>.
1935 =item realloc() of freed memory ignored
1937 (S) An internal routine called realloc() on something that had already
1940 =item Reference is already weak
1942 (W) You have attempted to weaken a reference that is already weak.
1943 Doing so has no effect.
1945 =item setpgrp can't take arguments
1947 (F) Your system has the setpgrp() from BSD 4.2, which takes no arguments,
1948 unlike POSIX setpgid(), which takes a process ID and process group ID.
1950 =item Strange *+?{} on zero-length expression
1952 (W) You applied a regular expression quantifier in a place where it
1953 makes no sense, such as on a zero-width assertion.
1954 Try putting the quantifier inside the assertion instead. For example,
1955 the way to match "abc" provided that it is followed by three
1956 repetitions of "xyz" is C</abc(?=(?:xyz){3})/>, not C</abc(?=xyz){3}/>.
1958 =item switching effective %s is not implemented
1960 (F) While under the C<use filetest> pragma, we cannot switch the
1961 real and effective uids or gids.
1963 =item This Perl can't reset CRTL environ elements (%s)
1965 =item This Perl can't set CRTL environ elements (%s=%s)
1967 (W) Warnings peculiar to VMS. You tried to change or delete an element
1968 of the CRTL's internal environ array, but your copy of Perl wasn't
1969 built with a CRTL that contained the setenv() function. You'll need to
1970 rebuild Perl with a CRTL that does, or redefine F<PERL_ENV_TABLES> (see
1971 L<perlvms>) so that the environ array isn't the target of the change to
1972 %ENV which produced the warning.
1974 =item Unknown open() mode '%s'
1976 (F) The second argument of 3-argument open() is not among the list
1977 of valid modes: C<L<lt>>, C<L<gt>>, C<E<gt>E<gt>>, C<+L<lt>>,
1978 C<+L<gt>>, C<+E<gt>E<gt>>, C<-|>, C<|->.
1980 =item Unknown process %x sent message to prime_env_iter: %s
1982 (P) An error peculiar to VMS. Perl was reading values for %ENV before
1983 iterating over it, and someone else stuck a message in the stream of
1984 data Perl expected. Someone's very confused, or perhaps trying to
1985 subvert Perl's population of %ENV for nefarious purposes.
1987 =item Unrecognized escape \\%c passed through
1989 (W) You used a backslash-character combination which is not recognized
1990 by Perl. The character was understood literally.
1992 =item Unterminated attribute parameter in attribute list
1994 (F) The lexer saw an opening (left) parenthesis character while parsing an
1995 attribute list, but the matching closing (right) parenthesis
1996 character was not found. You may need to add (or remove) a backslash
1997 character to get your parentheses to balance. See L<attributes>.
1999 =item Unterminated attribute list
2001 (F) The lexer found something other than a simple identifier at the start
2002 of an attribute, and it wasn't a semicolon or the start of a
2003 block. Perhaps you terminated the parameter list of the previous attribute
2004 too soon. See L<attributes>.
2006 =item Unterminated attribute parameter in subroutine attribute list
2008 (F) The lexer saw an opening (left) parenthesis character while parsing a
2009 subroutine attribute list, but the matching closing (right) parenthesis
2010 character was not found. You may need to add (or remove) a backslash
2011 character to get your parentheses to balance.
2013 =item Unterminated subroutine attribute list
2015 (F) The lexer found something other than a simple identifier at the start
2016 of a subroutine attribute, and it wasn't a semicolon or the start of a
2017 block. Perhaps you terminated the parameter list of the previous attribute
2020 =item Value of CLI symbol "%s" too long
2022 (W) A warning peculiar to VMS. Perl tried to read the value of an %ENV
2023 element from a CLI symbol table, and found a resultant string longer
2024 than 1024 characters. The return value has been truncated to 1024
2027 =item Version number must be a constant number
2029 (P) The attempt to translate a C<use Module n.n LIST> statement into
2030 its equivalent C<BEGIN> block found an internal inconsistency with
2035 =head1 Obsolete Diagnostics
2039 =item Character class syntax [: :] is reserved for future extensions
2041 (W) Within regular expression character classes ([]) the syntax beginning
2042 with "[:" and ending with ":]" is reserved for future extensions.
2043 If you need to represent those character sequences inside a regular
2044 expression character class, just quote the square brackets with the
2045 backslash: "\[:" and ":\]".
2047 =item Ill-formed logical name |%s| in prime_env_iter
2049 (W) A warning peculiar to VMS. A logical name was encountered when preparing
2050 to iterate over %ENV which violates the syntactic rules governing logical
2051 names. Because it cannot be translated normally, it is skipped, and will not
2052 appear in %ENV. This may be a benign occurrence, as some software packages
2053 might directly modify logical name tables and introduce nonstandard names,
2054 or it may indicate that a logical name table has been corrupted.
2056 =item regexp too big
2058 (F) The current implementation of regular expressions uses shorts as
2059 address offsets within a string. Unfortunately this means that if
2060 the regular expression compiles to longer than 32767, it'll blow up.
2061 Usually when you want a regular expression this big, there is a better
2062 way to do it with multiple statements. See L<perlre>.
2064 =item Use of "$$<digit>" to mean "${$}<digit>" is deprecated
2066 (D) Perl versions before 5.004 misinterpreted any type marker followed
2067 by "$" and a digit. For example, "$$0" was incorrectly taken to mean
2068 "${$}0" instead of "${$0}". This bug is (mostly) fixed in Perl 5.004.
2070 However, the developers of Perl 5.004 could not fix this bug completely,
2071 because at least two widely-used modules depend on the old meaning of
2072 "$$0" in a string. So Perl 5.004 still interprets "$$<digit>" in the
2073 old (broken) way inside strings; but it generates this message as a
2074 warning. And in Perl 5.005, this special treatment will cease.
2080 If you find what you think is a bug, you might check the
2081 articles recently posted to the comp.lang.perl.misc newsgroup.
2082 There may also be information at http://www.perl.com/perl/, the Perl
2085 If you believe you have an unreported bug, please run the B<perlbug>
2086 program included with your release. Make sure to trim your bug down
2087 to a tiny but sufficient test case. Your bug report, along with the
2088 output of C<perl -V>, will be sent off to perlbug@perl.com to be
2089 analysed by the Perl porting team.
2093 The F<Changes> file for exhaustive details on what changed.
2095 The F<INSTALL> file for how to build Perl.
2097 The F<README> file for general stuff.
2099 The F<Artistic> and F<Copying> files for copyright information.
2103 Written by Gurusamy Sarathy <F<gsar@activestate.com>>, with many
2104 contributions from The Perl Porters.
2106 Send omissions or corrections to <F<perlbug@perl.com>>.