3 perlport - Writing portable Perl
7 Perl runs on numerous operating systems. While most of them share
8 much in common, they also have their own unique features.
10 This document is meant to help you to find out what constitutes portable
11 Perl code. That way once you make a decision to write portably,
12 you know where the lines are drawn, and you can stay within them.
14 There is a tradeoff between taking full advantage of one particular
15 type of computer and taking advantage of a full range of them.
16 Naturally, as you broaden your range and become more diverse, the
17 common factors drop, and you are left with an increasingly smaller
18 area of common ground in which you can operate to accomplish a
19 particular task. Thus, when you begin attacking a problem, it is
20 important to consider under which part of the tradeoff curve you
21 want to operate. Specifically, you must decide whether it is
22 important that the task that you are coding have the full generality
23 of being portable, or whether to just get the job done right now.
24 This is the hardest choice to be made. The rest is easy, because
25 Perl provides many choices, whichever way you want to approach your
28 Looking at it another way, writing portable code is usually about
29 willfully limiting your available choices. Naturally, it takes
30 discipline and sacrifice to do that. The product of portability
31 and convenience may be a constant. You have been warned.
33 Be aware of two important points:
37 =item Not all Perl programs have to be portable
39 There is no reason you should not use Perl as a language to glue Unix
40 tools together, or to prototype a Macintosh application, or to manage the
41 Windows registry. If it makes no sense to aim for portability for one
42 reason or another in a given program, then don't bother.
44 =item Nearly all of Perl already I<is> portable
46 Don't be fooled into thinking that it is hard to create portable Perl
47 code. It isn't. Perl tries its level-best to bridge the gaps between
48 what's available on different platforms, and all the means available to
49 use those features. Thus almost all Perl code runs on any machine
50 without modification. But there are some significant issues in
51 writing portable code, and this document is entirely about those issues.
55 Here's the general rule: When you approach a task commonly done
56 using a whole range of platforms, think about writing portable
57 code. That way, you don't sacrifice much by way of the implementation
58 choices you can avail yourself of, and at the same time you can give
59 your users lots of platform choices. On the other hand, when you have to
60 take advantage of some unique feature of a particular platform, as is
61 often the case with systems programming (whether for Unix, Windows,
62 S<Mac OS>, VMS, etc.), consider writing platform-specific code.
64 When the code will run on only two or three operating systems, you
65 may need to consider only the differences of those particular systems.
66 The important thing is to decide where the code will run and to be
67 deliberate in your decision.
69 The material below is separated into three main sections: main issues of
70 portability (L<"ISSUES">, platform-specific issues (L<"PLATFORMS">, and
71 built-in perl functions that behave differently on various ports
72 (L<"FUNCTION IMPLEMENTATIONS">.
74 This information should not be considered complete; it includes possibly
75 transient information about idiosyncrasies of some of the ports, almost
76 all of which are in a state of constant evolution. Thus, this material
77 should be considered a perpetual work in progress
78 (<IMG SRC="yellow_sign.gif" ALT="Under Construction">).
84 In most operating systems, lines in files are terminated by newlines.
85 Just what is used as a newline may vary from OS to OS. Unix
86 traditionally uses C<\012>, one type of DOSish I/O uses C<\015\012>,
87 and S<Mac OS> uses C<\015>.
89 Perl uses C<\n> to represent the "logical" newline, where what is
90 logical may depend on the platform in use. In MacPerl, C<\n> always
91 means C<\015>. In DOSish perls, C<\n> usually means C<\012>, but
92 when accessing a file in "text" mode, STDIO translates it to (or
93 from) C<\015\012>, depending on whether you're reading or writing.
94 Unix does the same thing on ttys in canonical mode. C<\015\012>
95 is commonly referred to as CRLF.
97 Because of the "text" mode translation, DOSish perls have limitations
98 in using C<seek> and C<tell> on a file accessed in "text" mode.
99 Stick to C<seek>-ing to locations you got from C<tell> (and no
100 others), and you are usually free to use C<seek> and C<tell> even
101 in "text" mode. Using C<seek> or C<tell> or other file operations
102 may be non-portable. If you use C<binmode> on a file, however, you
103 can usually C<seek> and C<tell> with arbitrary values in safety.
105 A common misconception in socket programming is that C<\n> eq C<\012>
106 everywhere. When using protocols such as common Internet protocols,
107 C<\012> and C<\015> are called for specifically, and the values of
108 the logical C<\n> and C<\r> (carriage return) are not reliable.
110 print SOCKET "Hi there, client!\r\n"; # WRONG
111 print SOCKET "Hi there, client!\015\012"; # RIGHT
113 However, using C<\015\012> (or C<\cM\cJ>, or C<\x0D\x0A>) can be tedious
114 and unsightly, as well as confusing to those maintaining the code. As
115 such, the Socket module supplies the Right Thing for those who want it.
117 use Socket qw(:DEFAULT :crlf);
118 print SOCKET "Hi there, client!$CRLF" # RIGHT
120 When reading from a socket, remember that the default input record
121 separator C<$/> is C<\n>, but robust socket code will recognize as
122 either C<\012> or C<\015\012> as end of line:
128 Because both CRLF and LF end in LF, the input record separator can
129 be set to LF and any CR stripped later. Better to write:
131 use Socket qw(:DEFAULT :crlf);
132 local($/) = LF; # not needed if $/ is already \012
135 s/$CR?$LF/\n/; # not sure if socket uses LF or CRLF, OK
136 # s/\015?\012/\n/; # same thing
139 This example is preferred over the previous one--even for Unix
140 platforms--because now any C<\015>'s (C<\cM>'s) are stripped out
141 (and there was much rejoicing).
143 Similarly, functions that return text data--such as a function that
144 fetches a web page--should sometimes translate newlines before
145 returning the data, if they've not yet been translated to the local
146 newline representation. A single line of code will often suffice:
148 $data =~ s/\015?\012/\n/g;
151 Some of this may be confusing. Here's a handy reference to the ASCII CR
152 and LF characters. You can print it out and stick it in your wallet.
154 LF == \012 == \x0A == \cJ == ASCII 10
155 CR == \015 == \x0D == \cM == ASCII 13
158 ---------------------------
161 \n * | LF | CRLF | CR |
162 \r * | CR | CR | LF |
163 ---------------------------
166 The Unix column assumes that you are not accessing a serial line
167 (like a tty) in canonical mode. If you are, then CR on input becomes
168 "\n", and "\n" on output becomes CRLF.
170 These are just the most common definitions of C<\n> and C<\r> in Perl.
171 There may well be others.
173 =head2 Numbers endianness and Width
175 Different CPUs store integers and floating point numbers in different
176 orders (called I<endianness>) and widths (32-bit and 64-bit being the
177 most common today). This affects your programs when they attempt to transfer
178 numbers in binary format from one CPU architecture to another,
179 usually either "live" via network connection, or by storing the
180 numbers to secondary storage such as a disk file or tape.
182 Conflicting storage orders make utter mess out of the numbers. If a
183 little-endian host (Intel, VAX) stores 0x12345678 (305419896 in
184 decimal), a big-endian host (Motorola, MIPS, Sparc, PA) reads it as
185 0x78563412 (2018915346 in decimal). To avoid this problem in network
186 (socket) connections use the C<pack> and C<unpack> formats C<n>
187 and C<N>, the "network" orders. These are guaranteed to be portable.
189 You can explore the endianness of your platform by unpacking a
190 data structure packed in native format such as:
192 print unpack("h*", pack("s2", 1, 2)), "\n";
193 # '10002000' on e.g. Intel x86 or Alpha 21064 in little-endian mode
194 # '00100020' on e.g. Motorola 68040
196 If you need to distinguish between endian architectures you could use
197 either of the variables set like so:
199 $is_big_endian = unpack("h*", pack("s", 1)) =~ /01/;
200 $is_litte_endian = unpack("h*", pack("s", 1)) =~ /^1/;
202 Differing widths can cause truncation even between platforms of equal
203 endianness. The platform of shorter width loses the upper parts of the
204 number. There is no good solution for this problem except to avoid
205 transferring or storing raw binary numbers.
207 One can circumnavigate both these problems in two ways. Either
208 transfer and store numbers always in text format, instead of raw
209 binary, or else consider using modules like Data::Dumper (included in
210 the standard distribution as of Perl 5.005) and Storable. Keeping
211 all data as text significantly simplifies matters.
213 =head2 Files and Filesystems
215 Most platforms these days structure files in a hierarchical fashion.
216 So, it is reasonably safe to assume that all platforms support the
217 notion of a "path" to uniquely identify a file on the system. How
218 that path is really written, though, differs considerably.
220 Atlhough similar, file path specifications differ between Unix,
221 Windows, S<Mac OS>, OS/2, VMS, VOS, S<RISC OS>, and probably others.
222 Unix, for example, is one of the few OSes that has the elegant idea
223 of a single root directory.
225 DOS, OS/2, VMS, VOS, and Windows can work similarly to Unix with C</>
226 as path separator, or in their own idiosyncratic ways (such as having
227 several root directories and various "unrooted" device files such NIL:
230 S<Mac OS> uses C<:> as a path separator instead of C</>.
232 The filesystem may support neither hard links (C<link>) nor
233 symbolic links (C<symlink>, C<readlink>, C<lstat>).
235 The filesystem may support neither access timestamp nor change
236 timestamp (meaning that about the only portable timestamp is the
237 modification timestamp), or one second granularity of any timestamps
238 (e.g. the FAT filesystem limits the time granularity to two seconds).
240 VOS perl can emulate Unix filenames with C</> as path separator. The
241 native pathname characters greater-than, less-than, number-sign, and
242 percent-sign are always accepted.
244 S<RISC OS> perl can emulate Unix filenames with C</> as path
245 separator, or go native and use C<.> for path separator and C<:> to
246 signal filesystems and disk names.
248 If all this is intimidating, have no (well, maybe only a little)
249 fear. There are modules that can help. The File::Spec modules
250 provide methods to do the Right Thing on whatever platform happens
251 to be running the program.
253 use File::Spec::Functions;
254 chdir(updir()); # go up one directory
255 $file = catfile(curdir(), 'temp', 'file.txt');
256 # on Unix and Win32, './temp/file.txt'
257 # on Mac OS, ':temp:file.txt'
258 # on VMS, '[.temp]file.txt'
260 File::Spec is available in the standard distribution as of version
261 5.004_05. File::Spec::Functions is only in File::Spec 0.7 and later,
262 and some versions of perl come with version 0.6. If File::Spec
263 is not updated to 0.7 or later, you must use the object-oriented
264 interface from File::Spec (or upgrade File::Spec).
266 In general, production code should not have file paths hardcoded.
267 Making them user-supplied or read from a configuration file is
268 better, keeping in mind that file path syntax varies on different
271 This is especially noticeable in scripts like Makefiles and test suites,
272 which often assume C</> as a path separator for subdirectories.
274 Also of use is File::Basename from the standard distribution, which
275 splits a pathname into pieces (base filename, full path to directory,
278 Even when on a single platform (if you can call Unix a single platform),
279 remember not to count on the existence or the contents of particular
280 system-specific files or directories, like F</etc/passwd>,
281 F</etc/sendmail.conf>, F</etc/resolv.conf>, or even F</tmp/>. For
282 example, F</etc/passwd> may exist but not contain the encrypted
283 passwords, because the system is using some form of enhanced security.
284 Or it may not contain all the accounts, because the system is using NIS.
285 If code does need to rely on such a file, include a description of the
286 file and its format in the code's documentation, then make it easy for
287 the user to override the default location of the file.
289 Don't assume a text file will end with a newline. They should,
292 Do not have two files of the same name with different case, like
293 F<test.pl> and F<Test.pl>, as many platforms have case-insensitive
294 filenames. Also, try not to have non-word characters (except for C<.>)
295 in the names, and keep them to the 8.3 convention, for maximum
296 portability, onerous a burden though this may appear.
298 Likewise, when using the AutoSplit module, try to keep your functions to
299 8.3 naming and case-insensitive conventions; or, at the least,
300 make it so the resulting files have a unique (case-insensitively)
303 Whitespace in filenames is tolerated on most systems, but not all.
304 Many systems (DOS, VMS) cannot have more than one C<.> in their filenames.
306 Don't assume C<< > >> won't be the first character of a filename.
307 Always use C<< < >> explicitly to open a file for reading,
308 unless you want the user to be able to specify a pipe open.
310 open(FILE, "< $existing_file") or die $!;
312 If filenames might use strange characters, it is safest to open it
313 with C<sysopen> instead of C<open>. C<open> is magic and can
314 translate characters like C<< > >>, C<< < >>, and C<|>, which may
315 be the wrong thing to do. (Sometimes, though, it's the right thing.)
317 =head2 System Interaction
319 Not all platforms provide a command line. These are usually platforms
320 that rely primarily on a Graphical User Interface (GUI) for user
321 interaction. A program requiring a command line interface might
322 not work everywhere. This is probably for the user of the program
323 to deal with, so don't stay up late worrying about it.
325 Some platforms can't delete or rename files held open by the system.
326 Remember to C<close> files when you are done with them. Don't
327 C<unlink> or C<rename> an open file. Don't C<tie> or C<open> a
328 file already tied or opened; C<untie> or C<close> it first.
330 Don't open the same file more than once at a time for writing, as some
331 operating systems put mandatory locks on such files.
333 Don't count on a specific environment variable existing in C<%ENV>.
334 Don't count on C<%ENV> entries being case-sensitive, or even
337 Don't count on signals or C<%SIG> for anything.
339 Don't count on filename globbing. Use C<opendir>, C<readdir>, and
342 Don't count on per-program environment variables, or per-program current
345 Don't count on specific values of C<$!>.
347 =head2 Interprocess Communication (IPC)
349 In general, don't directly access the system in code meant to be
350 portable. That means, no C<system>, C<exec>, C<fork>, C<pipe>,
351 C<``>, C<qx//>, C<open> with a C<|>, nor any of the other things
352 that makes being a perl hacker worth being.
354 Commands that launch external processes are generally supported on
355 most platforms (though many of them do not support any type of
356 forking). The problem with using them arises from what you invoke
357 them on. External tools are often named differently on different
358 platforms, may not be available in the same location, migth accept
359 different arguments, can behave differently, and often present their
360 results in a platform-dependent way. Thus, you should seldom depend
361 on them to produce consistent results. (Then again, if you're calling
362 I<netstat -a>, you probably don't expect it to run on both Unix and CP/M.)
364 One especially common bit of Perl code is opening a pipe to B<sendmail>:
366 open(MAIL, '|/usr/lib/sendmail -t')
367 or die "cannot fork sendmail: $!";
369 This is fine for systems programming when sendmail is known to be
370 available. But it is not fine for many non-Unix systems, and even
371 some Unix systems that may not have sendmail installed. If a portable
372 solution is needed, see the various distributions on CPAN that deal
373 with it. Mail::Mailer and Mail::Send in the MailTools distribution are
374 commonly used, and provide several mailing methods, including mail,
375 sendmail, and direct SMTP (via Net::SMTP) if a mail transfer agent is
376 not available. Mail::Sendmail is a standalone module that provides
377 simple, platform-independent mailing.
379 The Unix System V IPC (C<msg*(), sem*(), shm*()>) is not available
380 even on all Unix platforms.
382 The rule of thumb for portable code is: Do it all in portable Perl, or
383 use a module (that may internally implement it with platform-specific
384 code, but expose a common interface).
386 =head2 External Subroutines (XS)
388 XS code can usually be made to work with any platform, but dependent
389 libraries, header files, etc., might not be readily available or
390 portable, or the XS code itself might be platform-specific, just as Perl
391 code might be. If the libraries and headers are portable, then it is
392 normally reasonable to make sure the XS code is portable, too.
394 A different type of portability issue arises when writing XS code:
395 availability of a C compiler on the end-user's system. C brings
396 with it its own portability issues, and writing XS code will expose
397 you to some of those. Writing purely in Perl is an easier way to
400 =head2 Standard Modules
402 In general, the standard modules work across platforms. Notable
403 exceptions are the CPAN module (which currently makes connections to external
404 programs that may not be available), platform-specific modules (like
405 ExtUtils::MM_VMS), and DBM modules.
407 There is no one DBM module available on all platforms.
408 SDBM_File and the others are generally available on all Unix and DOSish
409 ports, but not in MacPerl, where only NBDM_File and DB_File are
412 The good news is that at least some DBM module should be available, and
413 AnyDBM_File will use whichever module it can find. Of course, then
414 the code needs to be fairly strict, dropping to the greatest common
415 factor (e.g., not exceeding 1K for each record), so that it will
416 work with any DBM module. See L<AnyDBM_File> for more details.
420 The system's notion of time of day and calendar date is controlled in
421 widely different ways. Don't assume the timezone is stored in C<$ENV{TZ}>,
422 and even if it is, don't assume that you can control the timezone through
425 Don't assume that the epoch starts at 00:00:00, January 1, 1970,
426 because that is OS- and implementation-specific. It is better to store a date
427 in an unambiguous representation. The ISO-8601 standard defines
428 "YYYY-MM-DD" as the date format. A text representation (like "1987-12-18")
429 can be easily converted into an OS-specific value using a module like
430 Date::Parse. An array of values, such as those returned by
431 C<localtime>, can be converted to an OS-specific representation using
434 When calculating specific times, such as for tests in time or date modules,
435 it may be appropriate to calculate an offset for the epoch.
438 $offset = Time::Local::timegm(0, 0, 0, 1, 0, 70);
440 The value for C<$offset> in Unix will be C<0>, but in Mac OS will be
441 some large number. C<$offset> can then be added to a Unix time value
442 to get what should be the proper value on any system.
444 =head2 Character sets and character encoding
446 Assume little about character sets. Assume nothing about
447 numerical values (C<ord>, C<chr>) of characters. Do not
448 assume that the alphabetic characters are encoded contiguously (in
449 the numeric sense). Do not assume anything about the ordering of the
450 characters. The lowercase letters may come before or after the
451 uppercase letters; the lowercase and uppercase may be interlaced so
452 that both `a' and `A' come before `b'; the accented and other
453 international characters may be interlaced so that E<auml> comes
456 =head2 Internationalisation
458 If you may assume POSIX (a rather large assumption), you may read
459 more about the POSIX locale system from L<perllocale>. The locale
460 system at least attempts to make things a little bit more portable,
461 or at least more convenient and native-friendly for non-English
462 users. The system affects character sets and encoding, and date
463 and time formatting--amongst other things.
465 =head2 System Resources
467 If your code is destined for systems with severely constrained (or
468 missing!) virtual memory systems then you want to be I<especially> mindful
469 of avoiding wasteful constructs such as:
471 # NOTE: this is no longer "bad" in perl5.005
472 for (0..10000000) {} # bad
473 for (my $x = 0; $x <= 10000000; ++$x) {} # good
475 @lines = <VERY_LARGE_FILE>; # bad
477 while (<FILE>) {$file .= $_} # sometimes bad
478 $file = join('', <FILE>); # better
480 The last two constructs may appear unintuitive to most people. The
481 first repeatedly grows a string, whereas the second allocates a
482 large chunk of memory in one go. On some systems, the second is
483 more efficient that the first.
487 Most multi-user platforms provide basic levels of security, usually
488 implemented at the filesystem level. Some, however, do
489 not--unfortunately. Thus the notion of user id, or "home" directory,
490 or even the state of being logged-in, may be unrecognizable on many
491 platforms. If you write programs that are security-conscious, it
492 is usually best to know what type of system you will be running
493 under so that you can write code explicitly for that platform (or
498 For those times when it is necessary to have platform-specific code,
499 consider keeping the platform-specific code in one place, making porting
500 to other platforms easier. Use the Config module and the special
501 variable C<$^O> to differentiate platforms, as described in
504 Be careful in the tests you supply with your module or programs.
505 Module code may be fully portable, but its tests might not be. This
506 often happens when tests spawn off other processes or call external
507 programs to aid in the testing, or when (as noted above) the tests
508 assume certain things about the filesystem and paths. Be careful
509 not to depend on a specific output style for errors, such as when
510 checking C<$!> after an system call. Some platforms expect a certain
511 output format, and perl on those platforms may have been adjusted
512 accordingly. Most specifically, don't anchor a regex when testing
517 Modules uploaded to CPAN are tested by a variety of volunteers on
518 different platforms. These CPAN testers are notified by mail of each
519 new upload, and reply to the list with PASS, FAIL, NA (not applicable to
520 this platform), or UNKNOWN (unknown), along with any relevant notations.
522 The purpose of the testing is twofold: one, to help developers fix any
523 problems in their code that crop up because of lack of testing on other
524 platforms; two, to provide users with information about whether
525 a given module works on a given platform.
529 =item Mailing list: cpan-testers@perl.org
531 =item Testing results: http://testers.cpan.org/
537 As of version 5.002, Perl is built with a C<$^O> variable that
538 indicates the operating system it was built on. This was implemented
539 to help speed up code that would otherwise have to C<use Config>
540 and use the value of C<$Config{osname}>. Of course, to get more
541 detailed information about the system, looking into C<%Config> is
542 certainly recommended.
544 C<%Config> cannot always be trusted, however, because it was built
545 at compile time. If perl was built in one place, then transferred
546 elsewhere, some values may be wrong. The values may even have been
547 edited after the fact.
551 Perl works on a bewildering variety of Unix and Unix-like platforms (see
552 e.g. most of the files in the F<hints/> directory in the source code kit).
553 On most of these systems, the value of C<$^O> (hence C<$Config{'osname'}>,
554 too) is determined either by lowercasing and stripping punctuation from the
555 first field of the string returned by typing C<uname -a> (or a similar command)
556 at the shell prompt or by testing the file system for the presence of
557 uniquely named files such as a kernel or header file. Here, for example,
558 are a few of the more popular Unix flavors:
560 uname $^O $Config{'archname'}
561 --------------------------------------------
563 BSD/OS bsdos i386-bsdos
564 dgux dgux AViiON-dgux
565 DYNIX/ptx dynixptx i386-dynixptx
566 FreeBSD freebsd freebsd-i386
567 Linux linux arm-linux
568 Linux linux i386-linux
569 Linux linux i586-linux
570 Linux linux ppc-linux
571 HP-UX hpux PA-RISC1.1
573 Mac OS X rhapsody rhapsody
574 MachTen PPC machten powerpc-machten
576 NeXT 4 next OPENSTEP-Mach
577 openbsd openbsd i386-openbsd
578 OSF1 dec_osf alpha-dec_osf
579 reliantunix-n svr4 RM400-svr4
580 SCO_SV sco_sv i386-sco_sv
581 SINIX-N svr4 RM400-svr4
582 sn4609 unicos CRAY_C90-unicos
583 sn6521 unicosmk t3e-unicosmk
584 sn9617 unicos CRAY_J90-unicos
585 SunOS solaris sun4-solaris
586 SunOS solaris i86pc-solaris
587 SunOS4 sunos sun4-sunos
589 Because the value of C<$Config{archname}> may depend on the
590 hardware architecture, it can vary more than the value of C<$^O>.
592 =head2 DOS and Derivatives
594 Perl has long been ported to Intel-style microcomputers running under
595 systems like PC-DOS, MS-DOS, OS/2, and most Windows platforms you can
596 bring yourself to mention (except for Windows CE, if you count that).
597 Users familiar with I<COMMAND.COM> or I<CMD.EXE> style shells should
598 be aware that each of these file specifications may have subtle
601 $filespec0 = "c:/foo/bar/file.txt";
602 $filespec1 = "c:\\foo\\bar\\file.txt";
603 $filespec2 = 'c:\foo\bar\file.txt';
604 $filespec3 = 'c:\\foo\\bar\\file.txt';
606 System calls accept either C</> or C<\> as the path separator.
607 However, many command-line utilities of DOS vintage treat C</> as
608 the option prefix, so may get confused by filenames containing C</>.
609 Aside from calling any external programs, C</> will work just fine,
610 and probably better, as it is more consistent with popular usage,
611 and avoids the problem of remembering what to backwhack and what
614 The DOS FAT filesystem can accommodate only "8.3" style filenames. Under
615 the "case-insensitive, but case-preserving" HPFS (OS/2) and NTFS (NT)
616 filesystems you may have to be careful about case returned with functions
617 like C<readdir> or used with functions like C<open> or C<opendir>.
619 DOS also treats several filenames as special, such as AUX, PRN,
620 NUL, CON, COM1, LPT1, LPT2, etc. Unfortunately, sometimes these
621 filenames won't even work if you include an explicit directory
622 prefix. It is best to avoid such filenames, if you want your code
623 to be portable to DOS and its derivatives. It's hard to know what
624 these all are, unfortunately.
626 Users of these operating systems may also wish to make use of
627 scripts such as I<pl2bat.bat> or I<pl2cmd> to
628 put wrappers around your scripts.
630 Newline (C<\n>) is translated as C<\015\012> by STDIO when reading from
631 and writing to files (see L<"Newlines">). C<binmode(FILEHANDLE)>
632 will keep C<\n> translated as C<\012> for that filehandle. Since it is a
633 no-op on other systems, C<binmode> should be used for cross-platform code
634 that deals with binary data. That's assuming you realize in advance
635 that your data is in binary. General-purpose programs should
636 often assume nothing about their data.
638 The C<$^O> variable and the C<$Config{archname}> values for various
639 DOSish perls are as follows:
641 OS $^O $Config{'archname'}
642 --------------------------------------------
646 Windows 95 MSWin32 MSWin32-x86
647 Windows 98 MSWin32 MSWin32-x86
648 Windows NT MSWin32 MSWin32-x86
649 Windows NT MSWin32 MSWin32-ALPHA
650 Windows NT MSWin32 MSWin32-ppc
659 The djgpp environment for DOS, http://www.delorie.com/djgpp/
664 The EMX environment for DOS, OS/2, etc. emx@iaehv.nl,
665 http://www.leo.org/pub/comp/os/os2/leo/gnu/emx+gcc/index.html or
666 ftp://hobbes.nmsu.edu/pub/os2/dev/emx. Also L<perlos2>.
670 Build instructions for Win32 in L<perlwin32>, or under the Cygnus environment
675 The C<Win32::*> modules in L<Win32>.
679 The ActiveState Pages, http://www.activestate.com/
683 The Cygwin environment for Win32; F<README.cygwin> (installed
684 as L<perlcygwin>), http://sourceware.cygnus.com/cygwin/
688 The U/WIN environment for Win32,
689 <http://www.research.att.com/sw/tools/uwin/
691 =item Build instructions for OS/2, L<perlos2>
698 Any module requiring XS compilation is right out for most people, because
699 MacPerl is built using non-free (and non-cheap!) compilers. Some XS
700 modules that can work with MacPerl are built and distributed in binary
703 Directories are specified as:
705 volume:folder:file for absolute pathnames
706 volume:folder: for absolute pathnames
707 :folder:file for relative pathnames
708 :folder: for relative pathnames
709 :file for relative pathnames
710 file for relative pathnames
712 Files are stored in the directory in alphabetical order. Filenames are
713 limited to 31 characters, and may include any character except for
714 null and C<:>, which is reserved as the path separator.
716 Instead of C<flock>, see C<FSpSetFLock> and C<FSpRstFLock> in the
717 Mac::Files module, or C<chmod(0444, ...)> and C<chmod(0666, ...)>.
719 In the MacPerl application, you can't run a program from the command line;
720 programs that expect C<@ARGV> to be populated can be edited with something
721 like the following, which brings up a dialog box asking for the command
725 @ARGV = split /\s+/, MacPerl::Ask('Arguments?');
728 A MacPerl script saved as a "droplet" will populate C<@ARGV> with the full
729 pathnames of the files dropped onto the script.
731 Mac users can run programs under a type of command line interface
732 under MPW (Macintosh Programmer's Workshop, a free development
733 environment from Apple). MacPerl was first introduced as an MPW
734 tool, and MPW can be used like a shell:
736 perl myscript.plx some arguments
738 ToolServer is another app from Apple that provides access to MPW tools
739 from MPW and the MacPerl app, which allows MacPerl programs to use
740 C<system>, backticks, and piped C<open>.
742 "S<Mac OS>" is the proper name for the operating system, but the value
743 in C<$^O> is "MacOS". To determine architecture, version, or whether
744 the application or MPW tool version is running, check:
746 $is_app = $MacPerl::Version =~ /App/;
747 $is_tool = $MacPerl::Version =~ /MPW/;
748 ($version) = $MacPerl::Version =~ /^(\S+)/;
749 $is_ppc = $MacPerl::Architecture eq 'MacPPC';
750 $is_68k = $MacPerl::Architecture eq 'Mac68K';
752 S<Mac OS X> and S<Mac OS X Server>, based on NeXT's OpenStep OS, will
753 (in theory) be able to run MacPerl natively, under the "Classic"
754 environment. The new "Cocoa" environment (formerly called the "Yellow Box")
755 may run a slightly modified version of MacPerl, using the Carbon interfaces.
757 S<Mac OS X Server> and its Open Source version, Darwin, both run Unix
758 perl natively (with a few patches). Full support for these
759 is slated for perl 5.6.
767 The MacPerl Pages, http://www.macperl.com/ .
771 The MacPerl mailing lists, http://www.macperl.org/ .
775 MacPerl Module Porters, http://pudge.net/mmp/ .
781 Perl on VMS is discussed in L<perlvms> in the perl distribution.
782 Perl on VMS can accept either VMS- or Unix-style file
783 specifications as in either of the following:
785 $ perl -ne "print if /perl_setup/i" SYS$LOGIN:LOGIN.COM
786 $ perl -ne "print if /perl_setup/i" /sys$login/login.com
788 but not a mixture of both as in:
790 $ perl -ne "print if /perl_setup/i" sys$login:/login.com
791 Can't open sys$login:/login.com: file specification syntax error
793 Interacting with Perl from the Digital Command Language (DCL) shell
794 often requires a different set of quotation marks than Unix shells do.
797 $ perl -e "print ""Hello, world.\n"""
800 There are several ways to wrap your perl scripts in DCL F<.COM> files, if
801 you are so inclined. For example:
803 $ write sys$output "Hello from DCL!"
805 $ then perl -x 'f$environment("PROCEDURE")
806 $ else perl -x - 'p1 'p2 'p3 'p4 'p5 'p6 'p7 'p8
807 $ deck/dollars="__END__"
810 print "Hello from Perl!\n";
815 Do take care with C<$ ASSIGN/nolog/user SYS$COMMAND: SYS$INPUT> if your
816 perl-in-DCL script expects to do things like C<< $read = <STDIN>; >>.
818 Filenames are in the format "name.extension;version". The maximum
819 length for filenames is 39 characters, and the maximum length for
820 extensions is also 39 characters. Version is a number from 1 to
821 32767. Valid characters are C</[A-Z0-9$_-]/>.
823 VMS's RMS filesystem is case-insensitive and does not preserve case.
824 C<readdir> returns lowercased filenames, but specifying a file for
825 opening remains case-insensitive. Files without extensions have a
826 trailing period on them, so doing a C<readdir> with a file named F<A.;5>
827 will return F<a.> (though that file could be opened with
830 RMS had an eight level limit on directory depths from any rooted logical
831 (allowing 16 levels overall) prior to VMS 7.2. Hence
832 C<PERL_ROOT:[LIB.2.3.4.5.6.7.8]> is a valid directory specification but
833 C<PERL_ROOT:[LIB.2.3.4.5.6.7.8.9]> is not. F<Makefile.PL> authors might
834 have to take this into account, but at least they can refer to the former
835 as C</PERL_ROOT/lib/2/3/4/5/6/7/8/>.
837 The VMS::Filespec module, which gets installed as part of the build
838 process on VMS, is a pure Perl module that can easily be installed on
839 non-VMS platforms and can be helpful for conversions to and from RMS
842 What C<\n> represents depends on the type of file opened. It could
843 be C<\015>, C<\012>, C<\015\012>, or nothing. The VMS::Stdio module
844 provides access to the special fopen() requirements of files with unusual
847 TCP/IP stacks are optional on VMS, so socket routines might not be
848 implemented. UDP sockets may not be supported.
850 The value of C<$^O> on OpenVMS is "VMS". To determine the architecture
851 that you are running on without resorting to loading all of C<%Config>
852 you can examine the content of the C<@INC> array like so:
854 if (grep(/VMS_AXP/, @INC)) {
855 print "I'm on Alpha!\n";
857 } elsif (grep(/VMS_VAX/, @INC)) {
858 print "I'm on VAX!\n";
861 print "I'm not so sure about where $^O is...\n";
864 On VMS, perl determines the UTC offset from the C<SYS$TIMEZONE_DIFFERENTIAL>
865 logical name. Although the VMS epoch began at 17-NOV-1858 00:00:00.00,
866 calls to C<localtime> are adjusted to count offsets from
867 01-JAN-1970 00:00:00.00, just like Unix.
875 F<README.vms> (installed as L<README_vms>), L<perlvms>
879 vmsperl list, majordomo@perl.org
881 (Put the words C<subscribe vmsperl> in message body.)
885 vmsperl on the web, http://www.sidhe.org/vmsperl/index.html
891 Perl on VOS is discussed in F<README.vos> in the perl distribution.
892 Perl on VOS can accept either VOS- or Unix-style file
893 specifications as in either of the following:
895 $ perl -ne "print if /perl_setup/i" >system>notices
896 $ perl -ne "print if /perl_setup/i" /system/notices
898 or even a mixture of both as in:
900 $ perl -ne "print if /perl_setup/i" >system/notices
902 Even though VOS allows the slash character to appear in object
903 names, because the VOS port of Perl interprets it as a pathname
904 delimiting character, VOS files, directories, or links whose names
905 contain a slash character cannot be processed. Such files must be
906 renamed before they can be processed by Perl. Note that VOS limits
907 file names to 32 or fewer characters.
909 The following C functions are unimplemented on VOS, and any attempt by
910 Perl to use them will result in a fatal error message and an immediate
911 exit from Perl: dup, do_aspawn, do_spawn, fork, waitpid. Once these
912 functions become available in the VOS POSIX.1 implementation, you can
913 either recompile and rebind Perl, or you can download a newer port from
916 The value of C<$^O> on VOS is "VOS". To determine the architecture that
917 you are running on without resorting to loading all of C<%Config> you
918 can examine the content of the @INC array like so:
921 print "I'm on a Stratus box!\n";
923 print "I'm not on a Stratus box!\n";
927 if (grep(/860/, @INC)) {
928 print "This box is a Stratus XA/R!\n";
930 } elsif (grep(/7100/, @INC)) {
931 print "This box is a Stratus HP 7100 or 8xxx!\n";
933 } elsif (grep(/8000/, @INC)) {
934 print "This box is a Stratus HP 8xxx!\n";
937 print "This box is a Stratus 68K!\n";
950 The VOS mailing list.
952 There is no specific mailing list for Perl on VOS. You can post
953 comments to the comp.sys.stratus newsgroup, or subscribe to the general
954 Stratus mailing list. Send a letter with "Subscribe Info-Stratus" in
955 the message body to majordomo@list.stratagy.com.
959 VOS Perl on the web at http://ftp.stratus.com/pub/vos/vos.html
963 =head2 EBCDIC Platforms
965 Recent versions of Perl have been ported to platforms such as OS/400 on
966 AS/400 minicomputers as well as OS/390, VM/ESA, and BS2000 for S/390
967 Mainframes. Such computers use EBCDIC character sets internally (usually
968 Character Code Set ID 0037 for OS/400 and either 1047 or POSIX-BC for S/390
969 systems). On the mainframe perl currently works under the "Unix system
970 services for OS/390" (formerly known as OpenEdition), VM/ESA OpenEdition, or
971 the BS200 POSIX-BC system (BS2000 is supported in perl 5.6 and greater).
972 See L<perlos390> for details.
974 As of R2.5 of USS for OS/390 and Version 2.3 of VM/ESA these Unix
975 sub-systems do not support the C<#!> shebang trick for script invocation.
976 Hence, on OS/390 and VM/ESA perl scripts can be executed with a header
977 similar to the following simple script:
980 eval 'exec /usr/local/bin/perl -S $0 ${1+"$@"}'
982 #!/usr/local/bin/perl # just a comment really
984 print "Hello from perl!\n";
986 OS/390 will support the C<#!> shebang trick in release 2.8 and beyond.
987 Calls to C<system> and backticks can use POSIX shell syntax on all
990 On the AS/400, if PERL5 is in your library list, you may need
991 to wrap your perl scripts in a CL procedure to invoke them like so:
994 CALL PGM(PERL5/PERL) PARM('/QOpenSys/hello.pl')
997 This will invoke the perl script F<hello.pl> in the root of the
998 QOpenSys file system. On the AS/400 calls to C<system> or backticks
1001 On these platforms, bear in mind that the EBCDIC character set may have
1002 an effect on what happens with some perl functions (such as C<chr>,
1003 C<pack>, C<print>, C<printf>, C<ord>, C<sort>, C<sprintf>, C<unpack>), as
1004 well as bit-fiddling with ASCII constants using operators like C<^>, C<&>
1005 and C<|>, not to mention dealing with socket interfaces to ASCII computers
1006 (see L<"Newlines">).
1008 Fortunately, most web servers for the mainframe will correctly
1009 translate the C<\n> in the following statement to its ASCII equivalent
1010 (C<\r> is the same under both Unix and OS/390 & VM/ESA):
1012 print "Content-type: text/html\r\n\r\n";
1014 The values of C<$^O> on some of these platforms includes:
1016 uname $^O $Config{'archname'}
1017 --------------------------------------------
1020 POSIX-BC posix-bc BS2000-posix-bc
1023 Some simple tricks for determining if you are running on an EBCDIC
1024 platform could include any of the following (perhaps all):
1026 if ("\t" eq "\05") { print "EBCDIC may be spoken here!\n"; }
1028 if (ord('A') == 193) { print "EBCDIC may be spoken here!\n"; }
1030 if (chr(169) eq 'z') { print "EBCDIC may be spoken here!\n"; }
1032 One thing you may not want to rely on is the EBCDIC encoding
1033 of punctuation characters since these may differ from code page to code
1034 page (and once your module or script is rumoured to work with EBCDIC,
1035 folks will want it to work with all EBCDIC character sets).
1045 L<perlos390>, F<README.os390>, F<README.posix-bc>, F<README.vmesa>
1049 The perl-mvs@perl.org list is for discussion of porting issues as well as
1050 general usage issues for all EBCDIC Perls. Send a message body of
1051 "subscribe perl-mvs" to majordomo@perl.org.
1055 AS/400 Perl information at
1056 ttp://as400.rochester.ibm.com/
1057 as well as on CPAN in the F<ports/> directory.
1061 =head2 Acorn RISC OS
1063 Because Acorns use ASCII with newlines (C<\n>) in text files as C<\012> like
1064 Unix, and because Unix filename emulation is turned on by default,
1065 most simple scripts will probably work "out of the box". The native
1066 filesystem is modular, and individual filesystems are free to be
1067 case-sensitive or insensitive, and are usually case-preserving. Some
1068 native filesystems have name length limits, which file and directory
1069 names are silently truncated to fit. Scripts should be aware that the
1070 standard filesystem currently has a name length limit of B<10>
1071 characters, with up to 77 items in a directory, but other filesystems
1072 may not impose such limitations.
1074 Native filenames are of the form
1076 Filesystem#Special_Field::DiskName.$.Directory.Directory.File
1080 Special_Field is not usually present, but may contain . and $ .
1081 Filesystem =~ m|[A-Za-z0-9_]|
1082 DsicName =~ m|[A-Za-z0-9_/]|
1083 $ represents the root directory
1084 . is the path separator
1085 @ is the current directory (per filesystem but machine global)
1086 ^ is the parent directory
1087 Directory and File =~ m|[^\0- "\.\$\%\&:\@\\^\|\177]+|
1089 The default filename translation is roughly C<tr|/.|./|;>
1091 Note that C<"ADFS::HardDisk.$.File" ne 'ADFS::HardDisk.$.File'> and that
1092 the second stage of C<$> interpolation in regular expressions will fall
1093 foul of the C<$.> if scripts are not careful.
1095 Logical paths specified by system variables containing comma-separated
1096 search lists are also allowed; hence C<System:Modules> is a valid
1097 filename, and the filesystem will prefix C<Modules> with each section of
1098 C<System$Path> until a name is made that points to an object on disk.
1099 Writing to a new file C<System:Modules> would be allowed only if
1100 C<System$Path> contains a single item list. The filesystem will also
1101 expand system variables in filenames if enclosed in angle brackets, so
1102 C<< <System$Dir>.Modules >> would look for the file
1103 S<C<$ENV{'System$Dir'} . 'Modules'>>. The obvious implication of this is
1104 that B<fully qualified filenames can start with C<< <> >>> and should
1105 be protected when C<open> is used for input.
1107 Because C<.> was in use as a directory separator and filenames could not
1108 be assumed to be unique after 10 characters, Acorn implemented the C
1109 compiler to strip the trailing C<.c> C<.h> C<.s> and C<.o> suffix from
1110 filenames specified in source code and store the respective files in
1111 subdirectories named after the suffix. Hence files are translated:
1114 C:foo.h C:h.foo (logical path variable)
1115 sys/os.h sys.h.os (C compiler groks Unix-speak)
1116 10charname.c c.10charname
1117 10charname.o o.10charname
1118 11charname_.c c.11charname (assuming filesystem truncates at 10)
1120 The Unix emulation library's translation of filenames to native assumes
1121 that this sort of translation is required, and it allows a user-defined list
1122 of known suffixes that it will transpose in this fashion. This may
1123 seem transparent, but consider that with these rules C<foo/bar/baz.h>
1124 and C<foo/bar/h/baz> both map to C<foo.bar.h.baz>, and that C<readdir> and
1125 C<glob> cannot and do not attempt to emulate the reverse mapping. Other
1126 C<.>'s in filenames are translated to C</>.
1128 As implied above, the environment accessed through C<%ENV> is global, and
1129 the convention is that program specific environment variables are of the
1130 form C<Program$Name>. Each filesystem maintains a current directory,
1131 and the current filesystem's current directory is the B<global> current
1132 directory. Consequently, sociable programs don't change the current
1133 directory but rely on full pathnames, and programs (and Makefiles) cannot
1134 assume that they can spawn a child process which can change the current
1135 directory without affecting its parent (and everyone else for that
1138 Because native operating system filehandles are global and are currently
1139 allocated down from 255, with 0 being a reserved value, the Unix emulation
1140 library emulates Unix filehandles. Consequently, you can't rely on
1141 passing C<STDIN>, C<STDOUT>, or C<STDERR> to your children.
1143 The desire of users to express filenames of the form
1144 C<< <Foo$Dir>.Bar >> on the command line unquoted causes problems,
1145 too: C<``> command output capture has to perform a guessing game. It
1146 assumes that a string C<< <[^<>]+\$[^<>]> >> is a
1147 reference to an environment variable, whereas anything else involving
1148 C<< < >> or C<< > >> is redirection, and generally manages to be 99%
1149 right. Of course, the problem remains that scripts cannot rely on any
1150 Unix tools being available, or that any tools found have Unix-like command
1153 Extensions and XS are, in theory, buildable by anyone using free
1154 tools. In practice, many don't, as users of the Acorn platform are
1155 used to binary distributions. MakeMaker does run, but no available
1156 make currently copes with MakeMaker's makefiles; even if and when
1157 this should be fixed, the lack of a Unix-like shell will cause
1158 problems with makefile rules, especially lines of the form C<cd
1159 sdbm && make all>, and anything using quoting.
1161 "S<RISC OS>" is the proper name for the operating system, but the value
1162 in C<$^O> is "riscos" (because we don't like shouting).
1166 Perl has been ported to many platforms that do not fit into any of
1167 the categories listed above. Some, such as AmigaOS, Atari MiNT,
1168 BeOS, HP MPE/iX, QNX, Plan 9, and VOS, have been well-integrated
1169 into the standard Perl source code kit. You may need to see the
1170 F<ports/> directory on CPAN for information, and possibly binaries,
1171 for the likes of: aos, Atari ST, lynxos, riscos, Novell Netware,
1172 Tandem Guardian, I<etc.> (Yes, we know that some of these OSes may
1173 fall under the Unix category, but we are not a standards body.)
1175 Some approximate operating system names and their C<$^O> values
1176 in the "OTHER" category include:
1178 OS $^O $Config{'archname'}
1179 ------------------------------------------
1180 Amiga DOS amigaos m68k-amigos
1181 MPE/iX mpeix PA-RISC1.1
1189 Amiga, F<README.amiga> (installed as L<perlamiga>).
1193 Atari, F<README.mint> and Guido Flohr's web page
1194 http://stud.uni-sb.de/~gufl0000/
1198 Be OS, F<README.beos>
1202 HP 300 MPE/iX, F<README.mpeix> and Mark Bixby's web page
1203 http://www.cccd.edu/~markb/perlix.html
1207 A free perl5-based PERL.NLM for Novell Netware is available in
1208 precompiled binary and source code form from http://www.novell.com/
1209 as well as from CPAN.
1213 Plan 9, F<README.plan9>
1217 =head1 FUNCTION IMPLEMENTATIONS
1219 Listed below are functions that are either completely unimplemented
1220 or else have been implemented differently on various platforms.
1221 Following each description will be, in parentheses, a list of
1222 platforms that the description applies to.
1224 The list may well be incomplete, or even wrong in some places. When
1225 in doubt, consult the platform-specific README files in the Perl
1226 source distribution, and any other documentation resources accompanying
1229 Be aware, moreover, that even among Unix-ish systems there are variations.
1231 For many functions, you can also query C<%Config>, exported by
1232 default from the Config module. For example, to check whether the
1233 platform has the C<lstat> call, check C<$Config{d_lstat}>. See
1234 L<Config> for a full description of available variables.
1236 =head2 Alphabetical Listing of Perl Functions
1246 C<-r>, C<-w>, and C<-x> have a limited meaning only; directories
1247 and applications are executable, and there are no uid/gid
1248 considerations. C<-o> is not supported. (S<Mac OS>)
1250 C<-r>, C<-w>, C<-x>, and C<-o> tell whether the file is accessible,
1251 which may not reflect UIC-based file protections. (VMS)
1253 C<-s> returns the size of the data fork, not the total size of data fork
1254 plus resource fork. (S<Mac OS>).
1256 C<-s> by name on an open file will return the space reserved on disk,
1257 rather than the current extent. C<-s> on an open filehandle returns the
1258 current size. (S<RISC OS>)
1260 C<-R>, C<-W>, C<-X>, C<-O> are indistinguishable from C<-r>, C<-w>,
1261 C<-x>, C<-o>. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1263 C<-b>, C<-c>, C<-k>, C<-g>, C<-p>, C<-u>, C<-A> are not implemented.
1266 C<-g>, C<-k>, C<-l>, C<-p>, C<-u>, C<-A> are not particularly meaningful.
1267 (Win32, VMS, S<RISC OS>)
1269 C<-d> is true if passed a device spec without an explicit directory.
1272 C<-T> and C<-B> are implemented, but might misclassify Mac text files
1273 with foreign characters; this is the case will all platforms, but may
1274 affect S<Mac OS> often. (S<Mac OS>)
1276 C<-x> (or C<-X>) determine if a file ends in one of the executable
1277 suffixes. C<-S> is meaningless. (Win32)
1279 C<-x> (or C<-X>) determine if a file has an executable file type.
1286 Not implemented. (Win32)
1288 =item binmode FILEHANDLE
1290 Meaningless. (S<Mac OS>, S<RISC OS>)
1292 Reopens file and restores pointer; if function fails, underlying
1293 filehandle may be closed, or pointer may be in a different position.
1296 The value returned by C<tell> may be affected after the call, and
1297 the filehandle may be flushed. (Win32)
1301 Only limited meaning. Disabling/enabling write permission is mapped to
1302 locking/unlocking the file. (S<Mac OS>)
1304 Only good for changing "owner" read-write access, "group", and "other"
1305 bits are meaningless. (Win32)
1307 Only good for changing "owner" and "other" read-write access. (S<RISC OS>)
1309 Access permissions are mapped onto VOS access-control list changes. (VOS)
1313 Not implemented. (S<Mac OS>, Win32, Plan9, S<RISC OS>, VOS)
1315 Does nothing, but won't fail. (Win32)
1317 =item chroot FILENAME
1321 Not implemented. (S<Mac OS>, Win32, VMS, Plan9, S<RISC OS>, VOS, VM/ESA)
1323 =item crypt PLAINTEXT,SALT
1325 May not be available if library or source was not provided when building
1328 Not implemented. (VOS)
1332 Not implemented. (VMS, Plan9, VOS)
1334 =item dbmopen HASH,DBNAME,MODE
1336 Not implemented. (VMS, Plan9, VOS)
1340 Not useful. (S<Mac OS>, S<RISC OS>)
1342 Not implemented. (Win32)
1344 Invokes VMS debugger. (VMS)
1348 Not implemented. (S<Mac OS>)
1350 Implemented via Spawn. (VM/ESA)
1352 Does not automatically flush output handles on some platforms.
1353 (SunOS, Solaris, HP-UX)
1355 =item fcntl FILEHANDLE,FUNCTION,SCALAR
1357 Not implemented. (Win32, VMS)
1359 =item flock FILEHANDLE,OPERATION
1361 Not implemented (S<Mac OS>, VMS, S<RISC OS>, VOS).
1363 Available only on Windows NT (not on Windows 95). (Win32)
1367 Not implemented. (S<Mac OS>, AmigaOS, S<RISC OS>, VOS, VM/ESA)
1369 Emulated using multiple interpreters. See L<perlfork>. (Win32)
1371 Does not automatically flush output handles on some platforms.
1372 (SunOS, Solaris, HP-UX)
1376 Not implemented. (S<Mac OS>, S<RISC OS>)
1380 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1384 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1386 =item getpriority WHICH,WHO
1388 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1392 Not implemented. (S<Mac OS>, Win32)
1394 Not useful. (S<RISC OS>)
1398 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1400 =item getnetbyname NAME
1402 Not implemented. (S<Mac OS>, Win32, Plan9)
1406 Not implemented. (S<Mac OS>, Win32)
1408 Not useful. (S<RISC OS>)
1412 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1414 =item getnetbyaddr ADDR,ADDRTYPE
1416 Not implemented. (S<Mac OS>, Win32, Plan9)
1418 =item getprotobynumber NUMBER
1420 Not implemented. (S<Mac OS>)
1422 =item getservbyport PORT,PROTO
1424 Not implemented. (S<Mac OS>)
1428 Not implemented. (S<Mac OS>, Win32, VM/ESA)
1432 Not implemented. (S<Mac OS>, Win32, VMS, VM/ESA)
1436 Not implemented. (S<Mac OS>, Win32)
1440 Not implemented. (S<Mac OS>, Win32, Plan9)
1444 Not implemented. (S<Mac OS>, Win32, Plan9)
1448 Not implemented. (Win32, Plan9)
1452 Not implemented. (S<Mac OS>, Win32, S<RISC OS>)
1456 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>)
1458 =item sethostent STAYOPEN
1460 Not implemented. (S<Mac OS>, Win32, Plan9, S<RISC OS>)
1462 =item setnetent STAYOPEN
1464 Not implemented. (S<Mac OS>, Win32, Plan9, S<RISC OS>)
1466 =item setprotoent STAYOPEN
1468 Not implemented. (S<Mac OS>, Win32, Plan9, S<RISC OS>)
1470 =item setservent STAYOPEN
1472 Not implemented. (Plan9, Win32, S<RISC OS>)
1476 Not implemented. (S<Mac OS>, MPE/iX, VM/ESA, Win32)
1480 Not implemented. (S<Mac OS>, MPE/iX, S<RISC OS>, VM/ESA, VMS, Win32)
1484 Not implemented. (S<Mac OS>, Win32)
1488 Not implemented. (S<Mac OS>, Win32, Plan9)
1492 Not implemented. (S<Mac OS>, Win32, Plan9)
1496 Not implemented. (Plan9, Win32)
1498 =item getsockopt SOCKET,LEVEL,OPTNAME
1500 Not implemented. (S<Mac OS>, Plan9)
1506 Globbing built-in, but only C<*> and C<?> metacharacters are supported.
1509 This operator is implemented via the File::Glob extension on most
1510 platforms. See L<File::Glob> for portability information.
1512 =item ioctl FILEHANDLE,FUNCTION,SCALAR
1514 Not implemented. (VMS)
1516 Available only for socket handles, and it does what the ioctlsocket() call
1517 in the Winsock API does. (Win32)
1519 Available only for socket handles. (S<RISC OS>)
1521 =item kill SIGNAL, LIST
1523 Not implemented, hence not useful for taint checking. (S<Mac OS>,
1526 C<kill()> doesn't have the semantics of C<raise()>, i.e. it doesn't send
1527 a signal to the identified process like it does on Unix platforms.
1528 Instead C<kill($sig, $pid)> terminates the process identified by $pid,
1529 and makes it exit immediately with exit status $sig. As in Unix, if
1530 $sig is 0 and the specified process exists, it returns true without
1531 actually terminating it. (Win32)
1533 =item link OLDFILE,NEWFILE
1535 Not implemented. (S<Mac OS>, MPE/iX, VMS, S<RISC OS>)
1537 Link count not updated because hard links are not quite that hard
1538 (They are sort of half-way between hard and soft links). (AmigaOS)
1540 Hard links are implemented on Win32 (Windows NT and Windows 2000)
1543 =item lstat FILEHANDLE
1549 Not implemented. (VMS, S<RISC OS>)
1551 Return values (especially for device and inode) may be bogus. (Win32)
1553 =item msgctl ID,CMD,ARG
1555 =item msgget KEY,FLAGS
1557 =item msgsnd ID,MSG,FLAGS
1559 =item msgrcv ID,VAR,SIZE,TYPE,FLAGS
1561 Not implemented. (S<Mac OS>, Win32, VMS, Plan9, S<RISC OS>, VOS)
1563 =item open FILEHANDLE,EXPR
1565 =item open FILEHANDLE
1567 The C<|> variants are supported only if ToolServer is installed.
1570 open to C<|-> and C<-|> are unsupported. (S<Mac OS>, Win32, S<RISC OS>)
1572 Opening a process does not automatically flush output handles on some
1573 platforms. (SunOS, Solaris, HP-UX)
1575 =item pipe READHANDLE,WRITEHANDLE
1577 Not implemented. (S<Mac OS>)
1579 Very limited functionality. (MiNT)
1585 Not implemented. (Win32, VMS, S<RISC OS>)
1587 =item select RBITS,WBITS,EBITS,TIMEOUT
1589 Only implemented on sockets. (Win32)
1591 Only reliable on sockets. (S<RISC OS>)
1593 Note that the C<socket FILEHANDLE> form is generally portable.
1595 =item semctl ID,SEMNUM,CMD,ARG
1597 =item semget KEY,NSEMS,FLAGS
1599 =item semop KEY,OPSTRING
1601 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1605 Not implemented. (MPE/iX, Win32)
1607 =item setpgrp PID,PGRP
1609 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1611 =item setpriority WHICH,WHO,PRIORITY
1613 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1617 Not implemented. (MPE/iX, Win32)
1619 =item setsockopt SOCKET,LEVEL,OPTNAME,OPTVAL
1621 Not implemented. (S<Mac OS>, Plan9)
1623 =item shmctl ID,CMD,ARG
1625 =item shmget KEY,SIZE,FLAGS
1627 =item shmread ID,VAR,POS,SIZE
1629 =item shmwrite ID,STRING,POS,SIZE
1631 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS)
1633 =item socketpair SOCKET1,SOCKET2,DOMAIN,TYPE,PROTOCOL
1635 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1637 =item stat FILEHANDLE
1643 mtime and atime are the same thing, and ctime is creation time instead of
1644 inode change time. (S<Mac OS>)
1646 device and inode are not meaningful. (Win32)
1648 device and inode are not necessarily reliable. (VMS)
1650 mtime, atime and ctime all return the last modification time. Device and
1651 inode are not necessarily reliable. (S<RISC OS>)
1653 =item symlink OLDFILE,NEWFILE
1655 Not implemented. (Win32, VMS, S<RISC OS>)
1659 Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA)
1661 =item sysopen FILEHANDLE,FILENAME,MODE,PERMS
1663 The traditional "0", "1", and "2" MODEs are implemented with different
1664 numeric values on some systems. The flags exported by C<Fcntl>
1665 (O_RDONLY, O_WRONLY, O_RDWR) should work everywhere though. (S<Mac
1666 OS>, OS/390, VM/ESA)
1670 Only implemented if ToolServer is installed. (S<Mac OS>)
1672 As an optimization, may not call the command shell specified in
1673 C<$ENV{PERL5SHELL}>. C<system(1, @args)> spawns an external
1674 process and immediately returns its process designator, without
1675 waiting for it to terminate. Return value may be used subsequently
1676 in C<wait> or C<waitpid>. Failure to spawn() a subprocess is indicated
1677 by setting $? to "255 << 8". C<$?> is set in a way compatible with
1678 Unix (i.e. the exitstatus of the subprocess is obtained by "$? >> 8",
1679 as described in the documentation). (Win32)
1681 There is no shell to process metacharacters, and the native standard is
1682 to pass a command line terminated by "\n" "\r" or "\0" to the spawned
1683 program. Redirection such as C<< > foo >> is performed (if at all) by
1684 the run time library of the spawned program. C<system> I<list> will call
1685 the Unix emulation library's C<exec> emulation, which attempts to provide
1686 emulation of the stdin, stdout, stderr in force in the parent, providing
1687 the child program uses a compatible version of the emulation library.
1688 I<scalar> will call the native command line direct and no such emulation
1689 of a child Unix program will exists. Mileage B<will> vary. (S<RISC OS>)
1691 Far from being POSIX compliant. Because there may be no underlying
1692 /bin/sh tries to work around the problem by forking and execing the
1693 first token in its argument string. Handles basic redirection
1694 ("<" or ">") on its own behalf. (MiNT)
1696 Does not automatically flush output handles on some platforms.
1697 (SunOS, Solaris, HP-UX)
1701 Only the first entry returned is nonzero. (S<Mac OS>)
1703 "cumulative" times will be bogus. On anything other than Windows NT
1704 or Windows 2000, "system" time will be bogus, and "user" time is
1705 actually the time returned by the clock() function in the C runtime
1708 Not useful. (S<RISC OS>)
1710 =item truncate FILEHANDLE,LENGTH
1712 =item truncate EXPR,LENGTH
1714 Not implemented. (VMS)
1716 Truncation to zero-length only. (VOS)
1718 If a FILEHANDLE is supplied, it must be writable and opened in append
1719 mode (i.e., use C<open(FH, '>>filename')>
1720 or C<sysopen(FH,...,O_APPEND|O_RDWR)>. If a filename is supplied, it
1721 should not be held open elsewhere. (Win32)
1727 Returns undef where unavailable, as of version 5.005.
1729 C<umask> works but the correct permissions are set only when the file
1730 is finally closed. (AmigaOS)
1734 Only the modification time is updated. (S<Mac OS>, VMS, S<RISC OS>)
1736 May not behave as expected. Behavior depends on the C runtime
1737 library's implementation of utime(), and the filesystem being
1738 used. The FAT filesystem typically does not support an "access
1739 time" field, and it may limit timestamps to a granularity of
1740 two seconds. (Win32)
1744 =item waitpid PID,FLAGS
1746 Not implemented. (S<Mac OS>, VOS)
1748 Can only be applied to process handles returned for processes spawned
1749 using C<system(1, ...)>. (Win32)
1751 Not useful. (S<RISC OS>)
1759 =item v1.47, 22 March 2000
1761 Various cleanups from Tom Christiansen, including migration of
1762 long platform listings from L<perl>.
1764 =item v1.46, 12 February 2000
1766 Updates for VOS and MPE/iX. (Peter Prymmer) Other small changes.
1768 =item v1.45, 20 December 1999
1770 Small changes from 5.005_63 distribution, more changes to EBCDIC info.
1772 =item v1.44, 19 July 1999
1774 A bunch of updates from Peter Prymmer for C<$^O> values,
1775 endianness, File::Spec, VMS, BS2000, OS/400.
1777 =item v1.43, 24 May 1999
1779 Added a lot of cleaning up from Tom Christiansen.
1781 =item v1.42, 22 May 1999
1783 Added notes about tests, sprintf/printf, and epoch offsets.
1785 =item v1.41, 19 May 1999
1787 Lots more little changes to formatting and content.
1789 Added a bunch of C<$^O> and related values
1790 for various platforms; fixed mail and web addresses, and added
1791 and changed miscellaneous notes. (Peter Prymmer)
1793 =item v1.40, 11 April 1999
1795 Miscellaneous changes.
1797 =item v1.39, 11 February 1999
1799 Changes from Jarkko and EMX URL fixes Michael Schwern. Additional
1800 note about newlines added.
1802 =item v1.38, 31 December 1998
1804 More changes from Jarkko.
1806 =item v1.37, 19 December 1998
1808 More minor changes. Merge two separate version 1.35 documents.
1810 =item v1.36, 9 September 1998
1812 Updated for Stratus VOS. Also known as version 1.35.
1814 =item v1.35, 13 August 1998
1816 Integrate more minor changes, plus addition of new sections under
1817 L<"ISSUES">: L<"Numbers endianness and Width">,
1818 L<"Character sets and character encoding">,
1819 L<"Internationalisation">.
1821 =item v1.33, 06 August 1998
1823 Integrate more minor changes.
1825 =item v1.32, 05 August 1998
1827 Integrate more minor changes.
1829 =item v1.30, 03 August 1998
1831 Major update for RISC OS, other minor changes.
1833 =item v1.23, 10 July 1998
1835 First public release with perl5.005.
1839 =head1 Supported Platforms
1841 As of early March 2000 (the Perl release 5.6.0), the following
1842 platforms are able to build Perl from the standard source code
1843 distribution available at http://www.perl.com/CPAN/src/index.html
1875 1) in DOS mode either the DOS or OS/2 ports can be used
1876 2) new in 5.6.0: the BSD/NeXT-based UNIX of Mac OS X
1877 3) formerly known as Digital UNIX and before that DEC OSF/1
1878 4) compilers: Borland, Cygwin, Mingw32 EGCS/GCC, VC++
1880 The following platforms worked for the previous major release
1881 (5.005_03 being the latest maintenance release of that, as of early
1882 March 2000), but be did not manage to test these in time for the 5.6.0
1883 release of Perl. There is a very good chance that these will work
1884 just fine with 5.6.0.
1900 The following platform worked for the previous major release (5.005_03
1901 being the latest maintenance release of that, as of early March 2000).
1902 However, standardization on UTF-8 as the internal string representation
1903 in 5.6.0 has introduced incompatibilities in this EBCDIC platform.
1904 Support for this platform may be enabled in a future release:
1908 1) Previously known as MVS, or OpenEdition MVS.
1910 Strongly related to the OS390 platform by also being EBCDIC-based
1911 mainframe platforms are the following platforms:
1916 These are also not expected to work under 5.6.0 for the same reasons
1917 as OS390. Contact the mailing list perl-mvs@perl.org for more details.
1919 MacOS (Classic, pre-X) is almost 5.6.0-ready; building from the source
1920 does work with 5.6.0, but additional MacOS specific source code is needed
1921 for a complete port. Contact the mailing list macperl-porters@macperl.org
1922 for more information.
1924 The following platforms have been known to build Perl from source in
1925 the past, but we haven't been able to verify their status for the
1926 current release, either because the hardware/software platforms are
1927 rare or because we don't have an active champion on these
1959 Support for the following platform is planned for a future Perl release:
1963 The following platforms have their own source code distributions and
1964 binaries available via http://www.perl.com/CPAN/ports/index.html:
1970 Tandem Guardian 5.004
1972 The following platforms have only binaries available via
1973 http://www.perl.com/CPAN/ports/index.html :
1977 Acorn RISCOS 5.005_02
1981 Although we do suggest that you always build your own Perl from
1982 the source code, both for maximal configurability and for security,
1983 in case you are in a hurry you can check
1984 http://www.perl.com/CPAN/ports/index.html for binary distributions.
1988 L<perlamiga>, L<perlcygwin>, L<perldos>, L<perlhpux>, L<perlos2>,
1989 L<perlos390>, L<perlwin32>, L<perlvms>, and L<Win32>.
1991 =head1 AUTHORS / CONTRIBUTORS
1993 Abigail <abigail@fnx.com>,
1994 Charles Bailey <bailey@newman.upenn.edu>,
1995 Graham Barr <gbarr@pobox.com>,
1996 Tom Christiansen <tchrist@perl.com>,
1997 Nicholas Clark <Nicholas.Clark@liverpool.ac.uk>,
1998 Thomas Dorner <Thomas.Dorner@start.de>,
1999 Andy Dougherty <doughera@lafcol.lafayette.edu>,
2000 Dominic Dunlop <domo@vo.lu>,
2001 Neale Ferguson <neale@mailbox.tabnsw.com.au>,
2002 David J. Fiander <davidf@mks.com>,
2003 Paul Green <Paul_Green@stratus.com>,
2004 M.J.T. Guy <mjtg@cus.cam.ac.uk>,
2005 Jarkko Hietaniemi <jhi@iki.fi<gt>,
2006 Luther Huffman <lutherh@stratcom.com>,
2007 Nick Ing-Simmons <nick@ni-s.u-net.com>,
2008 Andreas J. KE<ouml>nig <koenig@kulturbox.de>,
2009 Markus Laker <mlaker@contax.co.uk>,
2010 Andrew M. Langmead <aml@world.std.com>,
2011 Larry Moore <ljmoore@freespace.net>,
2012 Paul Moore <Paul.Moore@uk.origin-it.com>,
2013 Chris Nandor <pudge@pobox.com>,
2014 Matthias Neeracher <neeri@iis.ee.ethz.ch>,
2015 Gary Ng <71564.1743@CompuServe.COM>,
2016 Tom Phoenix <rootbeer@teleport.com>,
2017 AndrE<eacute> Pirard <A.Pirard@ulg.ac.be>,
2018 Peter Prymmer <pvhp@forte.com>,
2019 Hugo van der Sanden <hv@crypt0.demon.co.uk>,
2020 Gurusamy Sarathy <gsar@activestate.com>,
2021 Paul J. Schinder <schinder@pobox.com>,
2022 Michael G Schwern <schwern@pobox.com>,
2023 Dan Sugalski <sugalskd@ous.edu>,
2024 Nathan Torkington <gnat@frii.com>.
2026 This document is maintained by Chris Nandor
2031 Version 1.47, last modified 22 March 2000