Commit | Line | Data |
---|---|---|
86530b38 AT |
1 | =head1 NAME |
2 | ||
3 | perlport - Writing portable Perl | |
4 | ||
5 | =head1 DESCRIPTION | |
6 | ||
7 | Perl runs on numerous operating systems. While most of them share | |
8 | much in common, they also have their own unique features. | |
9 | ||
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. | |
13 | ||
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 | |
26 | problem. | |
27 | ||
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. | |
32 | ||
33 | Be aware of two important points: | |
34 | ||
35 | =over 4 | |
36 | ||
37 | =item Not all Perl programs have to be portable | |
38 | ||
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. | |
43 | ||
44 | =item Nearly all of Perl already I<is> portable | |
45 | ||
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. | |
52 | ||
53 | =back | |
54 | ||
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. | |
63 | ||
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. | |
68 | ||
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">. | |
73 | ||
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 | (C<< <IMG SRC="yellow_sign.gif" ALT="Under Construction"> >>). | |
79 | ||
80 | =head1 ISSUES | |
81 | ||
82 | =head2 Newlines | |
83 | ||
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>. | |
88 | ||
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. | |
96 | ||
97 | A common cause of unportable programs is the misuse of chop() to trim | |
98 | newlines: | |
99 | ||
100 | # XXX UNPORTABLE! | |
101 | while(<FILE>) { | |
102 | chop; | |
103 | @array = split(/:/); | |
104 | #... | |
105 | } | |
106 | ||
107 | You can get away with this on Unix and Mac OS (they have a single | |
108 | character end-of-line), but the same program will break under DOSish | |
109 | perls because you're only chop()ing half the end-of-line. Instead, | |
110 | chomp() should be used to trim newlines. The Dunce::Files module can | |
111 | help audit your code for misuses of chop(). | |
112 | ||
113 | When dealing with binary files (or text files in binary mode) be sure | |
114 | to explicitly set $/ to the appropriate value for your file format | |
115 | before using chomp(). | |
116 | ||
117 | Because of the "text" mode translation, DOSish perls have limitations | |
118 | in using C<seek> and C<tell> on a file accessed in "text" mode. | |
119 | Stick to C<seek>-ing to locations you got from C<tell> (and no | |
120 | others), and you are usually free to use C<seek> and C<tell> even | |
121 | in "text" mode. Using C<seek> or C<tell> or other file operations | |
122 | may be non-portable. If you use C<binmode> on a file, however, you | |
123 | can usually C<seek> and C<tell> with arbitrary values in safety. | |
124 | ||
125 | A common misconception in socket programming is that C<\n> eq C<\012> | |
126 | everywhere. When using protocols such as common Internet protocols, | |
127 | C<\012> and C<\015> are called for specifically, and the values of | |
128 | the logical C<\n> and C<\r> (carriage return) are not reliable. | |
129 | ||
130 | print SOCKET "Hi there, client!\r\n"; # WRONG | |
131 | print SOCKET "Hi there, client!\015\012"; # RIGHT | |
132 | ||
133 | However, using C<\015\012> (or C<\cM\cJ>, or C<\x0D\x0A>) can be tedious | |
134 | and unsightly, as well as confusing to those maintaining the code. As | |
135 | such, the Socket module supplies the Right Thing for those who want it. | |
136 | ||
137 | use Socket qw(:DEFAULT :crlf); | |
138 | print SOCKET "Hi there, client!$CRLF" # RIGHT | |
139 | ||
140 | When reading from a socket, remember that the default input record | |
141 | separator C<$/> is C<\n>, but robust socket code will recognize as | |
142 | either C<\012> or C<\015\012> as end of line: | |
143 | ||
144 | while (<SOCKET>) { | |
145 | # ... | |
146 | } | |
147 | ||
148 | Because both CRLF and LF end in LF, the input record separator can | |
149 | be set to LF and any CR stripped later. Better to write: | |
150 | ||
151 | use Socket qw(:DEFAULT :crlf); | |
152 | local($/) = LF; # not needed if $/ is already \012 | |
153 | ||
154 | while (<SOCKET>) { | |
155 | s/$CR?$LF/\n/; # not sure if socket uses LF or CRLF, OK | |
156 | # s/\015?\012/\n/; # same thing | |
157 | } | |
158 | ||
159 | This example is preferred over the previous one--even for Unix | |
160 | platforms--because now any C<\015>'s (C<\cM>'s) are stripped out | |
161 | (and there was much rejoicing). | |
162 | ||
163 | Similarly, functions that return text data--such as a function that | |
164 | fetches a web page--should sometimes translate newlines before | |
165 | returning the data, if they've not yet been translated to the local | |
166 | newline representation. A single line of code will often suffice: | |
167 | ||
168 | $data =~ s/\015?\012/\n/g; | |
169 | return $data; | |
170 | ||
171 | Some of this may be confusing. Here's a handy reference to the ASCII CR | |
172 | and LF characters. You can print it out and stick it in your wallet. | |
173 | ||
174 | LF eq \012 eq \x0A eq \cJ eq chr(10) eq ASCII 10 | |
175 | CR eq \015 eq \x0D eq \cM eq chr(13) eq ASCII 13 | |
176 | ||
177 | | Unix | DOS | Mac | | |
178 | --------------------------- | |
179 | \n | LF | LF | CR | | |
180 | \r | CR | CR | LF | | |
181 | \n * | LF | CRLF | CR | | |
182 | \r * | CR | CR | LF | | |
183 | --------------------------- | |
184 | * text-mode STDIO | |
185 | ||
186 | The Unix column assumes that you are not accessing a serial line | |
187 | (like a tty) in canonical mode. If you are, then CR on input becomes | |
188 | "\n", and "\n" on output becomes CRLF. | |
189 | ||
190 | These are just the most common definitions of C<\n> and C<\r> in Perl. | |
191 | There may well be others. For example, on an EBCDIC implementation such | |
192 | as z/OS or OS/400 the above material is similar to "Unix" but the code | |
193 | numbers change: | |
194 | ||
195 | LF eq \025 eq \x15 eq chr(21) eq CP-1047 21 | |
196 | LF eq \045 eq \x25 eq \cU eq chr(37) eq CP-0037 37 | |
197 | CR eq \015 eq \x0D eq \cM eq chr(13) eq CP-1047 13 | |
198 | CR eq \015 eq \x0D eq \cM eq chr(13) eq CP-0037 13 | |
199 | ||
200 | | z/OS | OS/400 | | |
201 | ---------------------- | |
202 | \n | LF | LF | | |
203 | \r | CR | CR | | |
204 | \n * | LF | LF | | |
205 | \r * | CR | CR | | |
206 | ---------------------- | |
207 | * text-mode STDIO | |
208 | ||
209 | =head2 Numbers endianness and Width | |
210 | ||
211 | Different CPUs store integers and floating point numbers in different | |
212 | orders (called I<endianness>) and widths (32-bit and 64-bit being the | |
213 | most common today). This affects your programs when they attempt to transfer | |
214 | numbers in binary format from one CPU architecture to another, | |
215 | usually either "live" via network connection, or by storing the | |
216 | numbers to secondary storage such as a disk file or tape. | |
217 | ||
218 | Conflicting storage orders make utter mess out of the numbers. If a | |
219 | little-endian host (Intel, VAX) stores 0x12345678 (305419896 in | |
220 | decimal), a big-endian host (Motorola, Sparc, PA) reads it as | |
221 | 0x78563412 (2018915346 in decimal). Alpha and MIPS can be either: | |
222 | Digital/Compaq used/uses them in little-endian mode; SGI/Cray uses | |
223 | them in big-endian mode. To avoid this problem in network (socket) | |
224 | connections use the C<pack> and C<unpack> formats C<n> and C<N>, the | |
225 | "network" orders. These are guaranteed to be portable. | |
226 | ||
227 | You can explore the endianness of your platform by unpacking a | |
228 | data structure packed in native format such as: | |
229 | ||
230 | print unpack("h*", pack("s2", 1, 2)), "\n"; | |
231 | # '10002000' on e.g. Intel x86 or Alpha 21064 in little-endian mode | |
232 | # '00100020' on e.g. Motorola 68040 | |
233 | ||
234 | If you need to distinguish between endian architectures you could use | |
235 | either of the variables set like so: | |
236 | ||
237 | $is_big_endian = unpack("h*", pack("s", 1)) =~ /01/; | |
238 | $is_little_endian = unpack("h*", pack("s", 1)) =~ /^1/; | |
239 | ||
240 | Differing widths can cause truncation even between platforms of equal | |
241 | endianness. The platform of shorter width loses the upper parts of the | |
242 | number. There is no good solution for this problem except to avoid | |
243 | transferring or storing raw binary numbers. | |
244 | ||
245 | One can circumnavigate both these problems in two ways. Either | |
246 | transfer and store numbers always in text format, instead of raw | |
247 | binary, or else consider using modules like Data::Dumper (included in | |
248 | the standard distribution as of Perl 5.005) and Storable (included as | |
249 | of perl 5.8). Keeping all data as text significantly simplifies matters. | |
250 | ||
251 | The v-strings are portable only up to v2147483647 (0x7FFFFFFF), that's | |
252 | how far EBCDIC, or more precisely UTF-EBCDIC will go. | |
253 | ||
254 | =head2 Files and Filesystems | |
255 | ||
256 | Most platforms these days structure files in a hierarchical fashion. | |
257 | So, it is reasonably safe to assume that all platforms support the | |
258 | notion of a "path" to uniquely identify a file on the system. How | |
259 | that path is really written, though, differs considerably. | |
260 | ||
261 | Although similar, file path specifications differ between Unix, | |
262 | Windows, S<Mac OS>, OS/2, VMS, VOS, S<RISC OS>, and probably others. | |
263 | Unix, for example, is one of the few OSes that has the elegant idea | |
264 | of a single root directory. | |
265 | ||
266 | DOS, OS/2, VMS, VOS, and Windows can work similarly to Unix with C</> | |
267 | as path separator, or in their own idiosyncratic ways (such as having | |
268 | several root directories and various "unrooted" device files such NIL: | |
269 | and LPT:). | |
270 | ||
271 | S<Mac OS> uses C<:> as a path separator instead of C</>. | |
272 | ||
273 | The filesystem may support neither hard links (C<link>) nor | |
274 | symbolic links (C<symlink>, C<readlink>, C<lstat>). | |
275 | ||
276 | The filesystem may support neither access timestamp nor change | |
277 | timestamp (meaning that about the only portable timestamp is the | |
278 | modification timestamp), or one second granularity of any timestamps | |
279 | (e.g. the FAT filesystem limits the time granularity to two seconds). | |
280 | ||
281 | The "inode change timestamp" (the C<-C> filetest) may really be the | |
282 | "creation timestamp" (which it is not in UNIX). | |
283 | ||
284 | VOS perl can emulate Unix filenames with C</> as path separator. The | |
285 | native pathname characters greater-than, less-than, number-sign, and | |
286 | percent-sign are always accepted. | |
287 | ||
288 | S<RISC OS> perl can emulate Unix filenames with C</> as path | |
289 | separator, or go native and use C<.> for path separator and C<:> to | |
290 | signal filesystems and disk names. | |
291 | ||
292 | Don't assume UNIX filesystem access semantics: that read, write, | |
293 | and execute are all the permissions there are, and even if they exist, | |
294 | that their semantics (for example what do r, w, and x mean on | |
295 | a directory) are the UNIX ones. The various UNIX/POSIX compatibility | |
296 | layers usually try to make interfaces like chmod() work, but sometimes | |
297 | there simply is no good mapping. | |
298 | ||
299 | If all this is intimidating, have no (well, maybe only a little) | |
300 | fear. There are modules that can help. The File::Spec modules | |
301 | provide methods to do the Right Thing on whatever platform happens | |
302 | to be running the program. | |
303 | ||
304 | use File::Spec::Functions; | |
305 | chdir(updir()); # go up one directory | |
306 | $file = catfile(curdir(), 'temp', 'file.txt'); | |
307 | # on Unix and Win32, './temp/file.txt' | |
308 | # on Mac OS, ':temp:file.txt' | |
309 | # on VMS, '[.temp]file.txt' | |
310 | ||
311 | File::Spec is available in the standard distribution as of version | |
312 | 5.004_05. File::Spec::Functions is only in File::Spec 0.7 and later, | |
313 | and some versions of perl come with version 0.6. If File::Spec | |
314 | is not updated to 0.7 or later, you must use the object-oriented | |
315 | interface from File::Spec (or upgrade File::Spec). | |
316 | ||
317 | In general, production code should not have file paths hardcoded. | |
318 | Making them user-supplied or read from a configuration file is | |
319 | better, keeping in mind that file path syntax varies on different | |
320 | machines. | |
321 | ||
322 | This is especially noticeable in scripts like Makefiles and test suites, | |
323 | which often assume C</> as a path separator for subdirectories. | |
324 | ||
325 | Also of use is File::Basename from the standard distribution, which | |
326 | splits a pathname into pieces (base filename, full path to directory, | |
327 | and file suffix). | |
328 | ||
329 | Even when on a single platform (if you can call Unix a single platform), | |
330 | remember not to count on the existence or the contents of particular | |
331 | system-specific files or directories, like F</etc/passwd>, | |
332 | F</etc/sendmail.conf>, F</etc/resolv.conf>, or even F</tmp/>. For | |
333 | example, F</etc/passwd> may exist but not contain the encrypted | |
334 | passwords, because the system is using some form of enhanced security. | |
335 | Or it may not contain all the accounts, because the system is using NIS. | |
336 | If code does need to rely on such a file, include a description of the | |
337 | file and its format in the code's documentation, then make it easy for | |
338 | the user to override the default location of the file. | |
339 | ||
340 | Don't assume a text file will end with a newline. They should, | |
341 | but people forget. | |
342 | ||
343 | Do not have two files or directories of the same name with different | |
344 | case, like F<test.pl> and F<Test.pl>, as many platforms have | |
345 | case-insensitive (or at least case-forgiving) filenames. Also, try | |
346 | not to have non-word characters (except for C<.>) in the names, and | |
347 | keep them to the 8.3 convention, for maximum portability, onerous a | |
348 | burden though this may appear. | |
349 | ||
350 | Likewise, when using the AutoSplit module, try to keep your functions to | |
351 | 8.3 naming and case-insensitive conventions; or, at the least, | |
352 | make it so the resulting files have a unique (case-insensitively) | |
353 | first 8 characters. | |
354 | ||
355 | Whitespace in filenames is tolerated on most systems, but not all, | |
356 | and even on systems where it might be tolerated, some utilities | |
357 | might become confused by such whitespace. | |
358 | ||
359 | Many systems (DOS, VMS) cannot have more than one C<.> in their filenames. | |
360 | ||
361 | Don't assume C<< > >> won't be the first character of a filename. | |
362 | Always use C<< < >> explicitly to open a file for reading, or even | |
363 | better, use the three-arg version of open, unless you want the user to | |
364 | be able to specify a pipe open. | |
365 | ||
366 | open(FILE, '<', $existing_file) or die $!; | |
367 | ||
368 | If filenames might use strange characters, it is safest to open it | |
369 | with C<sysopen> instead of C<open>. C<open> is magic and can | |
370 | translate characters like C<< > >>, C<< < >>, and C<|>, which may | |
371 | be the wrong thing to do. (Sometimes, though, it's the right thing.) | |
372 | Three-arg open can also help protect against this translation in cases | |
373 | where it is undesirable. | |
374 | ||
375 | Don't use C<:> as a part of a filename since many systems use that for | |
376 | their own semantics (Mac OS Classic for separating pathname components, | |
377 | many networking schemes and utilities for separating the nodename and | |
378 | the pathname, and so on). For the same reasons, avoid C<@>, C<;> and | |
379 | C<|>. | |
380 | ||
381 | Don't assume that in pathnames you can collapse two leading slashes | |
382 | C<//> into one: some networking and clustering filesystems have special | |
383 | semantics for that. Let the operating system to sort it out. | |
384 | ||
385 | The I<portable filename characters> as defined by ANSI C are | |
386 | ||
387 | a b c d e f g h i j k l m n o p q r t u v w x y z | |
388 | A B C D E F G H I J K L M N O P Q R T U V W X Y Z | |
389 | 0 1 2 3 4 5 6 7 8 9 | |
390 | . _ - | |
391 | ||
392 | and the "-" shouldn't be the first character. If you want to be | |
393 | hypercorrect, stay case-insensitive and within the 8.3 naming | |
394 | convention (all the files and directories have to be unique within one | |
395 | directory if their names are lowercased and truncated to eight | |
396 | characters before the C<.>, if any, and to three characters after the | |
397 | C<.>, if any). (And do not use C<.>s in directory names.) | |
398 | ||
399 | =head2 System Interaction | |
400 | ||
401 | Not all platforms provide a command line. These are usually platforms | |
402 | that rely primarily on a Graphical User Interface (GUI) for user | |
403 | interaction. A program requiring a command line interface might | |
404 | not work everywhere. This is probably for the user of the program | |
405 | to deal with, so don't stay up late worrying about it. | |
406 | ||
407 | Some platforms can't delete or rename files held open by the system. | |
408 | Remember to C<close> files when you are done with them. Don't | |
409 | C<unlink> or C<rename> an open file. Don't C<tie> or C<open> a | |
410 | file already tied or opened; C<untie> or C<close> it first. | |
411 | ||
412 | Don't open the same file more than once at a time for writing, as some | |
413 | operating systems put mandatory locks on such files. | |
414 | ||
415 | Don't assume that write/modify permission on a directory gives the | |
416 | right to add or delete files/directories in that directory. That is | |
417 | filesystem specific: in some filesystems you need write/modify | |
418 | permission also (or even just) in the file/directory itself. In some | |
419 | filesystems (AFS, DFS) the permission to add/delete directory entries | |
420 | is a completely separate permission. | |
421 | ||
422 | Don't assume that a single C<unlink> completely gets rid of the file: | |
423 | some filesystems (most notably the ones in VMS) have versioned | |
424 | filesystems, and unlink() removes only the most recent one (it doesn't | |
425 | remove all the versions because by default the native tools on those | |
426 | platforms remove just the most recent version, too). The portable | |
427 | idiom to remove all the versions of a file is | |
428 | ||
429 | 1 while unlink "file"; | |
430 | ||
431 | This will terminate if the file is undeleteable for some reason | |
432 | (protected, not there, and so on). | |
433 | ||
434 | Don't count on a specific environment variable existing in C<%ENV>. | |
435 | Don't count on C<%ENV> entries being case-sensitive, or even | |
436 | case-preserving. Don't try to clear %ENV by saying C<%ENV = ();>, or, | |
437 | if you really have to, make it conditional on C<$^O ne 'VMS'> since in | |
438 | VMS the C<%ENV> table is much more than a per-process key-value string | |
439 | table. | |
440 | ||
441 | Don't count on signals or C<%SIG> for anything. | |
442 | ||
443 | Don't count on filename globbing. Use C<opendir>, C<readdir>, and | |
444 | C<closedir> instead. | |
445 | ||
446 | Don't count on per-program environment variables, or per-program current | |
447 | directories. | |
448 | ||
449 | Don't count on specific values of C<$!>. | |
450 | ||
451 | =head2 Command names versus file pathnames | |
452 | ||
453 | Don't assume that the name used to invoke a command or program with | |
454 | C<system> or C<exec> can also be used to test for the existence of the | |
455 | file that holds the executable code for that command or program. | |
456 | First, many systems have "internal" commands that are built-in to the | |
457 | shell or OS and while these commands can be invoked, there is no | |
458 | corresponding file. Second, some operating systems (e.g., Cygwin, | |
459 | DJGPP, OS/2, and VOS) have required suffixes for executable files; | |
460 | these suffixes are generally permitted on the command name but are not | |
461 | required. Thus, a command like "perl" might exist in a file named | |
462 | "perl", "perl.exe", or "perl.pm", depending on the operating system. | |
463 | The variable "_exe" in the Config module holds the executable suffix, | |
464 | if any. Third, the VMS port carefully sets up $^X and | |
465 | $Config{perlpath} so that no further processing is required. This is | |
466 | just as well, because the matching regular expression used below would | |
467 | then have to deal with a possible trailing version number in the VMS | |
468 | file name. | |
469 | ||
470 | To convert $^X to a file pathname, taking account of the requirements | |
471 | of the various operating system possibilities, say: | |
472 | use Config; | |
473 | $thisperl = $^X; | |
474 | if ($^O ne 'VMS') | |
475 | {$thisperl .= $Config{_exe} unless $thisperl =~ m/$Config{_exe}$/i;} | |
476 | ||
477 | To convert $Config{perlpath} to a file pathname, say: | |
478 | use Config; | |
479 | $thisperl = $Config{perlpath}; | |
480 | if ($^O ne 'VMS') | |
481 | {$thisperl .= $Config{_exe} unless $thisperl =~ m/$Config{_exe}$/i;} | |
482 | ||
483 | =head2 Interprocess Communication (IPC) | |
484 | ||
485 | In general, don't directly access the system in code meant to be | |
486 | portable. That means, no C<system>, C<exec>, C<fork>, C<pipe>, | |
487 | C<``>, C<qx//>, C<open> with a C<|>, nor any of the other things | |
488 | that makes being a perl hacker worth being. | |
489 | ||
490 | Commands that launch external processes are generally supported on | |
491 | most platforms (though many of them do not support any type of | |
492 | forking). The problem with using them arises from what you invoke | |
493 | them on. External tools are often named differently on different | |
494 | platforms, may not be available in the same location, might accept | |
495 | different arguments, can behave differently, and often present their | |
496 | results in a platform-dependent way. Thus, you should seldom depend | |
497 | on them to produce consistent results. (Then again, if you're calling | |
498 | I<netstat -a>, you probably don't expect it to run on both Unix and CP/M.) | |
499 | ||
500 | One especially common bit of Perl code is opening a pipe to B<sendmail>: | |
501 | ||
502 | open(MAIL, '|/usr/lib/sendmail -t') | |
503 | or die "cannot fork sendmail: $!"; | |
504 | ||
505 | This is fine for systems programming when sendmail is known to be | |
506 | available. But it is not fine for many non-Unix systems, and even | |
507 | some Unix systems that may not have sendmail installed. If a portable | |
508 | solution is needed, see the various distributions on CPAN that deal | |
509 | with it. Mail::Mailer and Mail::Send in the MailTools distribution are | |
510 | commonly used, and provide several mailing methods, including mail, | |
511 | sendmail, and direct SMTP (via Net::SMTP) if a mail transfer agent is | |
512 | not available. Mail::Sendmail is a standalone module that provides | |
513 | simple, platform-independent mailing. | |
514 | ||
515 | The Unix System V IPC (C<msg*(), sem*(), shm*()>) is not available | |
516 | even on all Unix platforms. | |
517 | ||
518 | Do not use either the bare result of C<pack("N", 10, 20, 30, 40)> or | |
519 | bare v-strings (such as C<v10.20.30.40>) to represent IPv4 addresses: | |
520 | both forms just pack the four bytes into network order. That this | |
521 | would be equal to the C language C<in_addr> struct (which is what the | |
522 | socket code internally uses) is not guaranteed. To be portable use | |
523 | the routines of the Socket extension, such as C<inet_aton()>, | |
524 | C<inet_ntoa()>, and C<sockaddr_in()>. | |
525 | ||
526 | The rule of thumb for portable code is: Do it all in portable Perl, or | |
527 | use a module (that may internally implement it with platform-specific | |
528 | code, but expose a common interface). | |
529 | ||
530 | =head2 External Subroutines (XS) | |
531 | ||
532 | XS code can usually be made to work with any platform, but dependent | |
533 | libraries, header files, etc., might not be readily available or | |
534 | portable, or the XS code itself might be platform-specific, just as Perl | |
535 | code might be. If the libraries and headers are portable, then it is | |
536 | normally reasonable to make sure the XS code is portable, too. | |
537 | ||
538 | A different type of portability issue arises when writing XS code: | |
539 | availability of a C compiler on the end-user's system. C brings | |
540 | with it its own portability issues, and writing XS code will expose | |
541 | you to some of those. Writing purely in Perl is an easier way to | |
542 | achieve portability. | |
543 | ||
544 | =head2 Standard Modules | |
545 | ||
546 | In general, the standard modules work across platforms. Notable | |
547 | exceptions are the CPAN module (which currently makes connections to external | |
548 | programs that may not be available), platform-specific modules (like | |
549 | ExtUtils::MM_VMS), and DBM modules. | |
550 | ||
551 | There is no one DBM module available on all platforms. | |
552 | SDBM_File and the others are generally available on all Unix and DOSish | |
553 | ports, but not in MacPerl, where only NBDM_File and DB_File are | |
554 | available. | |
555 | ||
556 | The good news is that at least some DBM module should be available, and | |
557 | AnyDBM_File will use whichever module it can find. Of course, then | |
558 | the code needs to be fairly strict, dropping to the greatest common | |
559 | factor (e.g., not exceeding 1K for each record), so that it will | |
560 | work with any DBM module. See L<AnyDBM_File> for more details. | |
561 | ||
562 | =head2 Time and Date | |
563 | ||
564 | The system's notion of time of day and calendar date is controlled in | |
565 | widely different ways. Don't assume the timezone is stored in C<$ENV{TZ}>, | |
566 | and even if it is, don't assume that you can control the timezone through | |
567 | that variable. | |
568 | ||
569 | Don't assume that the epoch starts at 00:00:00, January 1, 1970, | |
570 | because that is OS- and implementation-specific. It is better to store a date | |
571 | in an unambiguous representation. The ISO-8601 standard defines | |
572 | "YYYY-MM-DD" as the date format. A text representation (like "1987-12-18") | |
573 | can be easily converted into an OS-specific value using a module like | |
574 | Date::Parse. An array of values, such as those returned by | |
575 | C<localtime>, can be converted to an OS-specific representation using | |
576 | Time::Local. | |
577 | ||
578 | When calculating specific times, such as for tests in time or date modules, | |
579 | it may be appropriate to calculate an offset for the epoch. | |
580 | ||
581 | require Time::Local; | |
582 | $offset = Time::Local::timegm(0, 0, 0, 1, 0, 70); | |
583 | ||
584 | The value for C<$offset> in Unix will be C<0>, but in Mac OS will be | |
585 | some large number. C<$offset> can then be added to a Unix time value | |
586 | to get what should be the proper value on any system. | |
587 | ||
588 | =head2 Character sets and character encoding | |
589 | ||
590 | Assume very little about character sets. | |
591 | ||
592 | Assume nothing about numerical values (C<ord>, C<chr>) of characters. | |
593 | Do not use explicit code point ranges (like \xHH-\xHH); use for | |
594 | example symbolic character classes like C<[:print:]>. | |
595 | ||
596 | Do not assume that the alphabetic characters are encoded contiguously | |
597 | (in the numeric sense). There may be gaps. | |
598 | ||
599 | Do not assume anything about the ordering of the characters. | |
600 | The lowercase letters may come before or after the uppercase letters; | |
601 | the lowercase and uppercase may be interlaced so that both `a' and `A' | |
602 | come before `b'; the accented and other international characters may | |
603 | be interlaced so that E<auml> comes before `b'. | |
604 | ||
605 | =head2 Internationalisation | |
606 | ||
607 | If you may assume POSIX (a rather large assumption), you may read | |
608 | more about the POSIX locale system from L<perllocale>. The locale | |
609 | system at least attempts to make things a little bit more portable, | |
610 | or at least more convenient and native-friendly for non-English | |
611 | users. The system affects character sets and encoding, and date | |
612 | and time formatting--amongst other things. | |
613 | ||
614 | =head2 System Resources | |
615 | ||
616 | If your code is destined for systems with severely constrained (or | |
617 | missing!) virtual memory systems then you want to be I<especially> mindful | |
618 | of avoiding wasteful constructs such as: | |
619 | ||
620 | # NOTE: this is no longer "bad" in perl5.005 | |
621 | for (0..10000000) {} # bad | |
622 | for (my $x = 0; $x <= 10000000; ++$x) {} # good | |
623 | ||
624 | @lines = <VERY_LARGE_FILE>; # bad | |
625 | ||
626 | while (<FILE>) {$file .= $_} # sometimes bad | |
627 | $file = join('', <FILE>); # better | |
628 | ||
629 | The last two constructs may appear unintuitive to most people. The | |
630 | first repeatedly grows a string, whereas the second allocates a | |
631 | large chunk of memory in one go. On some systems, the second is | |
632 | more efficient that the first. | |
633 | ||
634 | =head2 Security | |
635 | ||
636 | Most multi-user platforms provide basic levels of security, usually | |
637 | implemented at the filesystem level. Some, however, do | |
638 | not-- unfortunately. Thus the notion of user id, or "home" directory, | |
639 | or even the state of being logged-in, may be unrecognizable on many | |
640 | platforms. If you write programs that are security-conscious, it | |
641 | is usually best to know what type of system you will be running | |
642 | under so that you can write code explicitly for that platform (or | |
643 | class of platforms). | |
644 | ||
645 | Don't assume the UNIX filesystem access semantics: the operating | |
646 | system or the filesystem may be using some ACL systems, which are | |
647 | richer languages than the usual rwx. Even if the rwx exist, | |
648 | their semantics might be different. | |
649 | ||
650 | (From security viewpoint testing for permissions before attempting to | |
651 | do something is silly anyway: if one tries this, there is potential | |
652 | for race conditions-- someone or something might change the | |
653 | permissions between the permissions check and the actual operation. | |
654 | Just try the operation.) | |
655 | ||
656 | Don't assume the UNIX user and group semantics: especially, don't | |
657 | expect the C<< $< >> and C<< $> >> (or the C<$(> and C<$)>) to work | |
658 | for switching identities (or memberships). | |
659 | ||
660 | Don't assume set-uid and set-gid semantics. (And even if you do, | |
661 | think twice: set-uid and set-gid are a known can of security worms.) | |
662 | ||
663 | =head2 Style | |
664 | ||
665 | For those times when it is necessary to have platform-specific code, | |
666 | consider keeping the platform-specific code in one place, making porting | |
667 | to other platforms easier. Use the Config module and the special | |
668 | variable C<$^O> to differentiate platforms, as described in | |
669 | L<"PLATFORMS">. | |
670 | ||
671 | Be careful in the tests you supply with your module or programs. | |
672 | Module code may be fully portable, but its tests might not be. This | |
673 | often happens when tests spawn off other processes or call external | |
674 | programs to aid in the testing, or when (as noted above) the tests | |
675 | assume certain things about the filesystem and paths. Be careful | |
676 | not to depend on a specific output style for errors, such as when | |
677 | checking C<$!> after a system call. Some platforms expect a certain | |
678 | output format, and perl on those platforms may have been adjusted | |
679 | accordingly. Most specifically, don't anchor a regex when testing | |
680 | an error value. | |
681 | ||
682 | =head1 CPAN Testers | |
683 | ||
684 | Modules uploaded to CPAN are tested by a variety of volunteers on | |
685 | different platforms. These CPAN testers are notified by mail of each | |
686 | new upload, and reply to the list with PASS, FAIL, NA (not applicable to | |
687 | this platform), or UNKNOWN (unknown), along with any relevant notations. | |
688 | ||
689 | The purpose of the testing is twofold: one, to help developers fix any | |
690 | problems in their code that crop up because of lack of testing on other | |
691 | platforms; two, to provide users with information about whether | |
692 | a given module works on a given platform. | |
693 | ||
694 | =over 4 | |
695 | ||
696 | =item Mailing list: cpan-testers@perl.org | |
697 | ||
698 | =item Testing results: http://testers.cpan.org/ | |
699 | ||
700 | =back | |
701 | ||
702 | =head1 PLATFORMS | |
703 | ||
704 | As of version 5.002, Perl is built with a C<$^O> variable that | |
705 | indicates the operating system it was built on. This was implemented | |
706 | to help speed up code that would otherwise have to C<use Config> | |
707 | and use the value of C<$Config{osname}>. Of course, to get more | |
708 | detailed information about the system, looking into C<%Config> is | |
709 | certainly recommended. | |
710 | ||
711 | C<%Config> cannot always be trusted, however, because it was built | |
712 | at compile time. If perl was built in one place, then transferred | |
713 | elsewhere, some values may be wrong. The values may even have been | |
714 | edited after the fact. | |
715 | ||
716 | =head2 Unix | |
717 | ||
718 | Perl works on a bewildering variety of Unix and Unix-like platforms (see | |
719 | e.g. most of the files in the F<hints/> directory in the source code kit). | |
720 | On most of these systems, the value of C<$^O> (hence C<$Config{'osname'}>, | |
721 | too) is determined either by lowercasing and stripping punctuation from the | |
722 | first field of the string returned by typing C<uname -a> (or a similar command) | |
723 | at the shell prompt or by testing the file system for the presence of | |
724 | uniquely named files such as a kernel or header file. Here, for example, | |
725 | are a few of the more popular Unix flavors: | |
726 | ||
727 | uname $^O $Config{'archname'} | |
728 | -------------------------------------------- | |
729 | AIX aix aix | |
730 | BSD/OS bsdos i386-bsdos | |
731 | Darwin darwin darwin | |
732 | dgux dgux AViiON-dgux | |
733 | DYNIX/ptx dynixptx i386-dynixptx | |
734 | FreeBSD freebsd freebsd-i386 | |
735 | Linux linux arm-linux | |
736 | Linux linux i386-linux | |
737 | Linux linux i586-linux | |
738 | Linux linux ppc-linux | |
739 | HP-UX hpux PA-RISC1.1 | |
740 | IRIX irix irix | |
741 | Mac OS X darwin darwin | |
742 | MachTen PPC machten powerpc-machten | |
743 | NeXT 3 next next-fat | |
744 | NeXT 4 next OPENSTEP-Mach | |
745 | openbsd openbsd i386-openbsd | |
746 | OSF1 dec_osf alpha-dec_osf | |
747 | reliantunix-n svr4 RM400-svr4 | |
748 | SCO_SV sco_sv i386-sco_sv | |
749 | SINIX-N svr4 RM400-svr4 | |
750 | sn4609 unicos CRAY_C90-unicos | |
751 | sn6521 unicosmk t3e-unicosmk | |
752 | sn9617 unicos CRAY_J90-unicos | |
753 | SunOS solaris sun4-solaris | |
754 | SunOS solaris i86pc-solaris | |
755 | SunOS4 sunos sun4-sunos | |
756 | ||
757 | Because the value of C<$Config{archname}> may depend on the | |
758 | hardware architecture, it can vary more than the value of C<$^O>. | |
759 | ||
760 | =head2 DOS and Derivatives | |
761 | ||
762 | Perl has long been ported to Intel-style microcomputers running under | |
763 | systems like PC-DOS, MS-DOS, OS/2, and most Windows platforms you can | |
764 | bring yourself to mention (except for Windows CE, if you count that). | |
765 | Users familiar with I<COMMAND.COM> or I<CMD.EXE> style shells should | |
766 | be aware that each of these file specifications may have subtle | |
767 | differences: | |
768 | ||
769 | $filespec0 = "c:/foo/bar/file.txt"; | |
770 | $filespec1 = "c:\\foo\\bar\\file.txt"; | |
771 | $filespec2 = 'c:\foo\bar\file.txt'; | |
772 | $filespec3 = 'c:\\foo\\bar\\file.txt'; | |
773 | ||
774 | System calls accept either C</> or C<\> as the path separator. | |
775 | However, many command-line utilities of DOS vintage treat C</> as | |
776 | the option prefix, so may get confused by filenames containing C</>. | |
777 | Aside from calling any external programs, C</> will work just fine, | |
778 | and probably better, as it is more consistent with popular usage, | |
779 | and avoids the problem of remembering what to backwhack and what | |
780 | not to. | |
781 | ||
782 | The DOS FAT filesystem can accommodate only "8.3" style filenames. Under | |
783 | the "case-insensitive, but case-preserving" HPFS (OS/2) and NTFS (NT) | |
784 | filesystems you may have to be careful about case returned with functions | |
785 | like C<readdir> or used with functions like C<open> or C<opendir>. | |
786 | ||
787 | DOS also treats several filenames as special, such as AUX, PRN, | |
788 | NUL, CON, COM1, LPT1, LPT2, etc. Unfortunately, sometimes these | |
789 | filenames won't even work if you include an explicit directory | |
790 | prefix. It is best to avoid such filenames, if you want your code | |
791 | to be portable to DOS and its derivatives. It's hard to know what | |
792 | these all are, unfortunately. | |
793 | ||
794 | Users of these operating systems may also wish to make use of | |
795 | scripts such as I<pl2bat.bat> or I<pl2cmd> to | |
796 | put wrappers around your scripts. | |
797 | ||
798 | Newline (C<\n>) is translated as C<\015\012> by STDIO when reading from | |
799 | and writing to files (see L<"Newlines">). C<binmode(FILEHANDLE)> | |
800 | will keep C<\n> translated as C<\012> for that filehandle. Since it is a | |
801 | no-op on other systems, C<binmode> should be used for cross-platform code | |
802 | that deals with binary data. That's assuming you realize in advance | |
803 | that your data is in binary. General-purpose programs should | |
804 | often assume nothing about their data. | |
805 | ||
806 | The C<$^O> variable and the C<$Config{archname}> values for various | |
807 | DOSish perls are as follows: | |
808 | ||
809 | OS $^O $Config{archname} ID Version | |
810 | -------------------------------------------------------- | |
811 | MS-DOS dos ? | |
812 | PC-DOS dos ? | |
813 | OS/2 os2 ? | |
814 | Windows 3.1 ? ? 0 3 01 | |
815 | Windows 95 MSWin32 MSWin32-x86 1 4 00 | |
816 | Windows 98 MSWin32 MSWin32-x86 1 4 10 | |
817 | Windows ME MSWin32 MSWin32-x86 1 ? | |
818 | Windows NT MSWin32 MSWin32-x86 2 4 xx | |
819 | Windows NT MSWin32 MSWin32-ALPHA 2 4 xx | |
820 | Windows NT MSWin32 MSWin32-ppc 2 4 xx | |
821 | Windows 2000 MSWin32 MSWin32-x86 2 5 xx | |
822 | Windows XP MSWin32 MSWin32-x86 2 ? | |
823 | Windows CE MSWin32 ? 3 | |
824 | Cygwin cygwin ? | |
825 | ||
826 | The various MSWin32 Perl's can distinguish the OS they are running on | |
827 | via the value of the fifth element of the list returned from | |
828 | Win32::GetOSVersion(). For example: | |
829 | ||
830 | if ($^O eq 'MSWin32') { | |
831 | my @os_version_info = Win32::GetOSVersion(); | |
832 | print +('3.1','95','NT')[$os_version_info[4]],"\n"; | |
833 | } | |
834 | ||
835 | There are also Win32::IsWinNT() and Win32::IsWin95(), try C<perldoc Win32>, | |
836 | and as of libwin32 0.19 (not part of the core Perl distribution) | |
837 | Win32::GetOSName(). The very portable POSIX::uname() will work too: | |
838 | ||
839 | c:\> perl -MPOSIX -we "print join '|', uname" | |
840 | Windows NT|moonru|5.0|Build 2195 (Service Pack 2)|x86 | |
841 | ||
842 | Also see: | |
843 | ||
844 | =over 4 | |
845 | ||
846 | =item * | |
847 | ||
848 | The djgpp environment for DOS, http://www.delorie.com/djgpp/ | |
849 | and L<perldos>. | |
850 | ||
851 | =item * | |
852 | ||
853 | The EMX environment for DOS, OS/2, etc. emx@iaehv.nl, | |
854 | http://www.leo.org/pub/comp/os/os2/leo/gnu/emx+gcc/index.html or | |
855 | ftp://hobbes.nmsu.edu/pub/os2/dev/emx/ Also L<perlos2>. | |
856 | ||
857 | =item * | |
858 | ||
859 | Build instructions for Win32 in L<perlwin32>, or under the Cygnus environment | |
860 | in L<perlcygwin>. | |
861 | ||
862 | =item * | |
863 | ||
864 | The C<Win32::*> modules in L<Win32>. | |
865 | ||
866 | =item * | |
867 | ||
868 | The ActiveState Pages, http://www.activestate.com/ | |
869 | ||
870 | =item * | |
871 | ||
872 | The Cygwin environment for Win32; F<README.cygwin> (installed | |
873 | as L<perlcygwin>), http://www.cygwin.com/ | |
874 | ||
875 | =item * | |
876 | ||
877 | The U/WIN environment for Win32, | |
878 | http://www.research.att.com/sw/tools/uwin/ | |
879 | ||
880 | =item * | |
881 | ||
882 | Build instructions for OS/2, L<perlos2> | |
883 | ||
884 | =back | |
885 | ||
886 | =head2 S<Mac OS> | |
887 | ||
888 | Any module requiring XS compilation is right out for most people, because | |
889 | MacPerl is built using non-free (and non-cheap!) compilers. Some XS | |
890 | modules that can work with MacPerl are built and distributed in binary | |
891 | form on CPAN. | |
892 | ||
893 | Directories are specified as: | |
894 | ||
895 | volume:folder:file for absolute pathnames | |
896 | volume:folder: for absolute pathnames | |
897 | :folder:file for relative pathnames | |
898 | :folder: for relative pathnames | |
899 | :file for relative pathnames | |
900 | file for relative pathnames | |
901 | ||
902 | Files are stored in the directory in alphabetical order. Filenames are | |
903 | limited to 31 characters, and may include any character except for | |
904 | null and C<:>, which is reserved as the path separator. | |
905 | ||
906 | Instead of C<flock>, see C<FSpSetFLock> and C<FSpRstFLock> in the | |
907 | Mac::Files module, or C<chmod(0444, ...)> and C<chmod(0666, ...)>. | |
908 | ||
909 | In the MacPerl application, you can't run a program from the command line; | |
910 | programs that expect C<@ARGV> to be populated can be edited with something | |
911 | like the following, which brings up a dialog box asking for the command | |
912 | line arguments. | |
913 | ||
914 | if (!@ARGV) { | |
915 | @ARGV = split /\s+/, MacPerl::Ask('Arguments?'); | |
916 | } | |
917 | ||
918 | A MacPerl script saved as a "droplet" will populate C<@ARGV> with the full | |
919 | pathnames of the files dropped onto the script. | |
920 | ||
921 | Mac users can run programs under a type of command line interface | |
922 | under MPW (Macintosh Programmer's Workshop, a free development | |
923 | environment from Apple). MacPerl was first introduced as an MPW | |
924 | tool, and MPW can be used like a shell: | |
925 | ||
926 | perl myscript.plx some arguments | |
927 | ||
928 | ToolServer is another app from Apple that provides access to MPW tools | |
929 | from MPW and the MacPerl app, which allows MacPerl programs to use | |
930 | C<system>, backticks, and piped C<open>. | |
931 | ||
932 | "S<Mac OS>" is the proper name for the operating system, but the value | |
933 | in C<$^O> is "MacOS". To determine architecture, version, or whether | |
934 | the application or MPW tool version is running, check: | |
935 | ||
936 | $is_app = $MacPerl::Version =~ /App/; | |
937 | $is_tool = $MacPerl::Version =~ /MPW/; | |
938 | ($version) = $MacPerl::Version =~ /^(\S+)/; | |
939 | $is_ppc = $MacPerl::Architecture eq 'MacPPC'; | |
940 | $is_68k = $MacPerl::Architecture eq 'Mac68K'; | |
941 | ||
942 | S<Mac OS X>, based on NeXT's OpenStep OS, runs MacPerl natively, under the | |
943 | "Classic" environment. There is no "Carbon" version of MacPerl to run | |
944 | under the primary Mac OS X environment. S<Mac OS X> and its Open Source | |
945 | version, Darwin, both run Unix perl natively. | |
946 | ||
947 | Also see: | |
948 | ||
949 | =over 4 | |
950 | ||
951 | =item * | |
952 | ||
953 | MacPerl Development, http://dev.macperl.org/ . | |
954 | ||
955 | =item * | |
956 | ||
957 | The MacPerl Pages, http://www.macperl.com/ . | |
958 | ||
959 | =item * | |
960 | ||
961 | The MacPerl mailing lists, http://lists.perl.org/ . | |
962 | ||
963 | =back | |
964 | ||
965 | =head2 VMS | |
966 | ||
967 | Perl on VMS is discussed in L<perlvms> in the perl distribution. | |
968 | Perl on VMS can accept either VMS- or Unix-style file | |
969 | specifications as in either of the following: | |
970 | ||
971 | $ perl -ne "print if /perl_setup/i" SYS$LOGIN:LOGIN.COM | |
972 | $ perl -ne "print if /perl_setup/i" /sys$login/login.com | |
973 | ||
974 | but not a mixture of both as in: | |
975 | ||
976 | $ perl -ne "print if /perl_setup/i" sys$login:/login.com | |
977 | Can't open sys$login:/login.com: file specification syntax error | |
978 | ||
979 | Interacting with Perl from the Digital Command Language (DCL) shell | |
980 | often requires a different set of quotation marks than Unix shells do. | |
981 | For example: | |
982 | ||
983 | $ perl -e "print ""Hello, world.\n""" | |
984 | Hello, world. | |
985 | ||
986 | There are several ways to wrap your perl scripts in DCL F<.COM> files, if | |
987 | you are so inclined. For example: | |
988 | ||
989 | $ write sys$output "Hello from DCL!" | |
990 | $ if p1 .eqs. "" | |
991 | $ then perl -x 'f$environment("PROCEDURE") | |
992 | $ else perl -x - 'p1 'p2 'p3 'p4 'p5 'p6 'p7 'p8 | |
993 | $ deck/dollars="__END__" | |
994 | #!/usr/bin/perl | |
995 | ||
996 | print "Hello from Perl!\n"; | |
997 | ||
998 | __END__ | |
999 | $ endif | |
1000 | ||
1001 | Do take care with C<$ ASSIGN/nolog/user SYS$COMMAND: SYS$INPUT> if your | |
1002 | perl-in-DCL script expects to do things like C<< $read = <STDIN>; >>. | |
1003 | ||
1004 | Filenames are in the format "name.extension;version". The maximum | |
1005 | length for filenames is 39 characters, and the maximum length for | |
1006 | extensions is also 39 characters. Version is a number from 1 to | |
1007 | 32767. Valid characters are C</[A-Z0-9$_-]/>. | |
1008 | ||
1009 | VMS's RMS filesystem is case-insensitive and does not preserve case. | |
1010 | C<readdir> returns lowercased filenames, but specifying a file for | |
1011 | opening remains case-insensitive. Files without extensions have a | |
1012 | trailing period on them, so doing a C<readdir> with a file named F<A.;5> | |
1013 | will return F<a.> (though that file could be opened with | |
1014 | C<open(FH, 'A')>). | |
1015 | ||
1016 | RMS had an eight level limit on directory depths from any rooted logical | |
1017 | (allowing 16 levels overall) prior to VMS 7.2. Hence | |
1018 | C<PERL_ROOT:[LIB.2.3.4.5.6.7.8]> is a valid directory specification but | |
1019 | C<PERL_ROOT:[LIB.2.3.4.5.6.7.8.9]> is not. F<Makefile.PL> authors might | |
1020 | have to take this into account, but at least they can refer to the former | |
1021 | as C</PERL_ROOT/lib/2/3/4/5/6/7/8/>. | |
1022 | ||
1023 | The VMS::Filespec module, which gets installed as part of the build | |
1024 | process on VMS, is a pure Perl module that can easily be installed on | |
1025 | non-VMS platforms and can be helpful for conversions to and from RMS | |
1026 | native formats. | |
1027 | ||
1028 | What C<\n> represents depends on the type of file opened. It usually | |
1029 | represents C<\012> but it could also be C<\015>, C<\012>, C<\015\012>, | |
1030 | C<\000>, C<\040>, or nothing depending on the file organiztion and | |
1031 | record format. The VMS::Stdio module provides access to the | |
1032 | special fopen() requirements of files with unusual attributes on VMS. | |
1033 | ||
1034 | TCP/IP stacks are optional on VMS, so socket routines might not be | |
1035 | implemented. UDP sockets may not be supported. | |
1036 | ||
1037 | The value of C<$^O> on OpenVMS is "VMS". To determine the architecture | |
1038 | that you are running on without resorting to loading all of C<%Config> | |
1039 | you can examine the content of the C<@INC> array like so: | |
1040 | ||
1041 | if (grep(/VMS_AXP/, @INC)) { | |
1042 | print "I'm on Alpha!\n"; | |
1043 | ||
1044 | } elsif (grep(/VMS_VAX/, @INC)) { | |
1045 | print "I'm on VAX!\n"; | |
1046 | ||
1047 | } else { | |
1048 | print "I'm not so sure about where $^O is...\n"; | |
1049 | } | |
1050 | ||
1051 | On VMS, perl determines the UTC offset from the C<SYS$TIMEZONE_DIFFERENTIAL> | |
1052 | logical name. Although the VMS epoch began at 17-NOV-1858 00:00:00.00, | |
1053 | calls to C<localtime> are adjusted to count offsets from | |
1054 | 01-JAN-1970 00:00:00.00, just like Unix. | |
1055 | ||
1056 | Also see: | |
1057 | ||
1058 | =over 4 | |
1059 | ||
1060 | =item * | |
1061 | ||
1062 | F<README.vms> (installed as L<README_vms>), L<perlvms> | |
1063 | ||
1064 | =item * | |
1065 | ||
1066 | vmsperl list, majordomo@perl.org | |
1067 | ||
1068 | (Put the words C<subscribe vmsperl> in message body.) | |
1069 | ||
1070 | =item * | |
1071 | ||
1072 | vmsperl on the web, http://www.sidhe.org/vmsperl/index.html | |
1073 | ||
1074 | =back | |
1075 | ||
1076 | =head2 VOS | |
1077 | ||
1078 | Perl on VOS is discussed in F<README.vos> in the perl distribution | |
1079 | (installed as L<perlvos>). Perl on VOS can accept either VOS- or | |
1080 | Unix-style file specifications as in either of the following: | |
1081 | ||
1082 | C<< $ perl -ne "print if /perl_setup/i" >system>notices >> | |
1083 | C<< $ perl -ne "print if /perl_setup/i" /system/notices >> | |
1084 | ||
1085 | or even a mixture of both as in: | |
1086 | ||
1087 | C<< $ perl -ne "print if /perl_setup/i" >system/notices >> | |
1088 | ||
1089 | Even though VOS allows the slash character to appear in object | |
1090 | names, because the VOS port of Perl interprets it as a pathname | |
1091 | delimiting character, VOS files, directories, or links whose names | |
1092 | contain a slash character cannot be processed. Such files must be | |
1093 | renamed before they can be processed by Perl. Note that VOS limits | |
1094 | file names to 32 or fewer characters. | |
1095 | ||
1096 | Perl on VOS can be built using two different compilers and two different | |
1097 | versions of the POSIX runtime. The recommended method for building full | |
1098 | Perl is with the GNU C compiler and the generally-available version of | |
1099 | VOS POSIX support. See F<README.vos> (installed as L<perlvos>) for | |
1100 | restrictions that apply when Perl is built using the VOS Standard C | |
1101 | compiler or the alpha version of VOS POSIX support. | |
1102 | ||
1103 | The value of C<$^O> on VOS is "VOS". To determine the architecture that | |
1104 | you are running on without resorting to loading all of C<%Config> you | |
1105 | can examine the content of the @INC array like so: | |
1106 | ||
1107 | if ($^O =~ /VOS/) { | |
1108 | print "I'm on a Stratus box!\n"; | |
1109 | } else { | |
1110 | print "I'm not on a Stratus box!\n"; | |
1111 | die; | |
1112 | } | |
1113 | ||
1114 | if (grep(/860/, @INC)) { | |
1115 | print "This box is a Stratus XA/R!\n"; | |
1116 | ||
1117 | } elsif (grep(/7100/, @INC)) { | |
1118 | print "This box is a Stratus HP 7100 or 8xxx!\n"; | |
1119 | ||
1120 | } elsif (grep(/8000/, @INC)) { | |
1121 | print "This box is a Stratus HP 8xxx!\n"; | |
1122 | ||
1123 | } else { | |
1124 | print "This box is a Stratus 68K!\n"; | |
1125 | } | |
1126 | ||
1127 | Also see: | |
1128 | ||
1129 | =over 4 | |
1130 | ||
1131 | =item * | |
1132 | ||
1133 | F<README.vos> (installed as L<perlvos>) | |
1134 | ||
1135 | =item * | |
1136 | ||
1137 | The VOS mailing list. | |
1138 | ||
1139 | There is no specific mailing list for Perl on VOS. You can post | |
1140 | comments to the comp.sys.stratus newsgroup, or subscribe to the general | |
1141 | Stratus mailing list. Send a letter with "subscribe Info-Stratus" in | |
1142 | the message body to majordomo@list.stratagy.com. | |
1143 | ||
1144 | =item * | |
1145 | ||
1146 | VOS Perl on the web at http://ftp.stratus.com/pub/vos/posix/posix.html | |
1147 | ||
1148 | =back | |
1149 | ||
1150 | =head2 EBCDIC Platforms | |
1151 | ||
1152 | Recent versions of Perl have been ported to platforms such as OS/400 on | |
1153 | AS/400 minicomputers as well as OS/390, VM/ESA, and BS2000 for S/390 | |
1154 | Mainframes. Such computers use EBCDIC character sets internally (usually | |
1155 | Character Code Set ID 0037 for OS/400 and either 1047 or POSIX-BC for S/390 | |
1156 | systems). On the mainframe perl currently works under the "Unix system | |
1157 | services for OS/390" (formerly known as OpenEdition), VM/ESA OpenEdition, or | |
1158 | the BS200 POSIX-BC system (BS2000 is supported in perl 5.6 and greater). | |
1159 | See L<perlos390> for details. | |
1160 | ||
1161 | As of R2.5 of USS for OS/390 and Version 2.3 of VM/ESA these Unix | |
1162 | sub-systems do not support the C<#!> shebang trick for script invocation. | |
1163 | Hence, on OS/390 and VM/ESA perl scripts can be executed with a header | |
1164 | similar to the following simple script: | |
1165 | ||
1166 | : # use perl | |
1167 | eval 'exec /usr/local/bin/perl -S $0 ${1+"$@"}' | |
1168 | if 0; | |
1169 | #!/usr/local/bin/perl # just a comment really | |
1170 | ||
1171 | print "Hello from perl!\n"; | |
1172 | ||
1173 | OS/390 will support the C<#!> shebang trick in release 2.8 and beyond. | |
1174 | Calls to C<system> and backticks can use POSIX shell syntax on all | |
1175 | S/390 systems. | |
1176 | ||
1177 | On the AS/400, if PERL5 is in your library list, you may need | |
1178 | to wrap your perl scripts in a CL procedure to invoke them like so: | |
1179 | ||
1180 | BEGIN | |
1181 | CALL PGM(PERL5/PERL) PARM('/QOpenSys/hello.pl') | |
1182 | ENDPGM | |
1183 | ||
1184 | This will invoke the perl script F<hello.pl> in the root of the | |
1185 | QOpenSys file system. On the AS/400 calls to C<system> or backticks | |
1186 | must use CL syntax. | |
1187 | ||
1188 | On these platforms, bear in mind that the EBCDIC character set may have | |
1189 | an effect on what happens with some perl functions (such as C<chr>, | |
1190 | C<pack>, C<print>, C<printf>, C<ord>, C<sort>, C<sprintf>, C<unpack>), as | |
1191 | well as bit-fiddling with ASCII constants using operators like C<^>, C<&> | |
1192 | and C<|>, not to mention dealing with socket interfaces to ASCII computers | |
1193 | (see L<"Newlines">). | |
1194 | ||
1195 | Fortunately, most web servers for the mainframe will correctly | |
1196 | translate the C<\n> in the following statement to its ASCII equivalent | |
1197 | (C<\r> is the same under both Unix and OS/390 & VM/ESA): | |
1198 | ||
1199 | print "Content-type: text/html\r\n\r\n"; | |
1200 | ||
1201 | The values of C<$^O> on some of these platforms includes: | |
1202 | ||
1203 | uname $^O $Config{'archname'} | |
1204 | -------------------------------------------- | |
1205 | OS/390 os390 os390 | |
1206 | OS400 os400 os400 | |
1207 | POSIX-BC posix-bc BS2000-posix-bc | |
1208 | VM/ESA vmesa vmesa | |
1209 | ||
1210 | Some simple tricks for determining if you are running on an EBCDIC | |
1211 | platform could include any of the following (perhaps all): | |
1212 | ||
1213 | if ("\t" eq "\05") { print "EBCDIC may be spoken here!\n"; } | |
1214 | ||
1215 | if (ord('A') == 193) { print "EBCDIC may be spoken here!\n"; } | |
1216 | ||
1217 | if (chr(169) eq 'z') { print "EBCDIC may be spoken here!\n"; } | |
1218 | ||
1219 | One thing you may not want to rely on is the EBCDIC encoding | |
1220 | of punctuation characters since these may differ from code page to code | |
1221 | page (and once your module or script is rumoured to work with EBCDIC, | |
1222 | folks will want it to work with all EBCDIC character sets). | |
1223 | ||
1224 | Also see: | |
1225 | ||
1226 | =over 4 | |
1227 | ||
1228 | =item * | |
1229 | ||
1230 | * | |
1231 | ||
1232 | L<perlos390>, F<README.os390>, F<perlbs2000>, F<README.vmesa>, | |
1233 | L<perlebcdic>. | |
1234 | ||
1235 | =item * | |
1236 | ||
1237 | The perl-mvs@perl.org list is for discussion of porting issues as well as | |
1238 | general usage issues for all EBCDIC Perls. Send a message body of | |
1239 | "subscribe perl-mvs" to majordomo@perl.org. | |
1240 | ||
1241 | =item * | |
1242 | ||
1243 | AS/400 Perl information at | |
1244 | http://as400.rochester.ibm.com/ | |
1245 | as well as on CPAN in the F<ports/> directory. | |
1246 | ||
1247 | =back | |
1248 | ||
1249 | =head2 Acorn RISC OS | |
1250 | ||
1251 | Because Acorns use ASCII with newlines (C<\n>) in text files as C<\012> like | |
1252 | Unix, and because Unix filename emulation is turned on by default, | |
1253 | most simple scripts will probably work "out of the box". The native | |
1254 | filesystem is modular, and individual filesystems are free to be | |
1255 | case-sensitive or insensitive, and are usually case-preserving. Some | |
1256 | native filesystems have name length limits, which file and directory | |
1257 | names are silently truncated to fit. Scripts should be aware that the | |
1258 | standard filesystem currently has a name length limit of B<10> | |
1259 | characters, with up to 77 items in a directory, but other filesystems | |
1260 | may not impose such limitations. | |
1261 | ||
1262 | Native filenames are of the form | |
1263 | ||
1264 | Filesystem#Special_Field::DiskName.$.Directory.Directory.File | |
1265 | ||
1266 | where | |
1267 | ||
1268 | Special_Field is not usually present, but may contain . and $ . | |
1269 | Filesystem =~ m|[A-Za-z0-9_]| | |
1270 | DsicName =~ m|[A-Za-z0-9_/]| | |
1271 | $ represents the root directory | |
1272 | . is the path separator | |
1273 | @ is the current directory (per filesystem but machine global) | |
1274 | ^ is the parent directory | |
1275 | Directory and File =~ m|[^\0- "\.\$\%\&:\@\\^\|\177]+| | |
1276 | ||
1277 | The default filename translation is roughly C<tr|/.|./|;> | |
1278 | ||
1279 | Note that C<"ADFS::HardDisk.$.File" ne 'ADFS::HardDisk.$.File'> and that | |
1280 | the second stage of C<$> interpolation in regular expressions will fall | |
1281 | foul of the C<$.> if scripts are not careful. | |
1282 | ||
1283 | Logical paths specified by system variables containing comma-separated | |
1284 | search lists are also allowed; hence C<System:Modules> is a valid | |
1285 | filename, and the filesystem will prefix C<Modules> with each section of | |
1286 | C<System$Path> until a name is made that points to an object on disk. | |
1287 | Writing to a new file C<System:Modules> would be allowed only if | |
1288 | C<System$Path> contains a single item list. The filesystem will also | |
1289 | expand system variables in filenames if enclosed in angle brackets, so | |
1290 | C<< <System$Dir>.Modules >> would look for the file | |
1291 | S<C<$ENV{'System$Dir'} . 'Modules'>>. The obvious implication of this is | |
1292 | that B<fully qualified filenames can start with C<< <> >>> and should | |
1293 | be protected when C<open> is used for input. | |
1294 | ||
1295 | Because C<.> was in use as a directory separator and filenames could not | |
1296 | be assumed to be unique after 10 characters, Acorn implemented the C | |
1297 | compiler to strip the trailing C<.c> C<.h> C<.s> and C<.o> suffix from | |
1298 | filenames specified in source code and store the respective files in | |
1299 | subdirectories named after the suffix. Hence files are translated: | |
1300 | ||
1301 | foo.h h.foo | |
1302 | C:foo.h C:h.foo (logical path variable) | |
1303 | sys/os.h sys.h.os (C compiler groks Unix-speak) | |
1304 | 10charname.c c.10charname | |
1305 | 10charname.o o.10charname | |
1306 | 11charname_.c c.11charname (assuming filesystem truncates at 10) | |
1307 | ||
1308 | The Unix emulation library's translation of filenames to native assumes | |
1309 | that this sort of translation is required, and it allows a user-defined list | |
1310 | of known suffixes that it will transpose in this fashion. This may | |
1311 | seem transparent, but consider that with these rules C<foo/bar/baz.h> | |
1312 | and C<foo/bar/h/baz> both map to C<foo.bar.h.baz>, and that C<readdir> and | |
1313 | C<glob> cannot and do not attempt to emulate the reverse mapping. Other | |
1314 | C<.>'s in filenames are translated to C</>. | |
1315 | ||
1316 | As implied above, the environment accessed through C<%ENV> is global, and | |
1317 | the convention is that program specific environment variables are of the | |
1318 | form C<Program$Name>. Each filesystem maintains a current directory, | |
1319 | and the current filesystem's current directory is the B<global> current | |
1320 | directory. Consequently, sociable programs don't change the current | |
1321 | directory but rely on full pathnames, and programs (and Makefiles) cannot | |
1322 | assume that they can spawn a child process which can change the current | |
1323 | directory without affecting its parent (and everyone else for that | |
1324 | matter). | |
1325 | ||
1326 | Because native operating system filehandles are global and are currently | |
1327 | allocated down from 255, with 0 being a reserved value, the Unix emulation | |
1328 | library emulates Unix filehandles. Consequently, you can't rely on | |
1329 | passing C<STDIN>, C<STDOUT>, or C<STDERR> to your children. | |
1330 | ||
1331 | The desire of users to express filenames of the form | |
1332 | C<< <Foo$Dir>.Bar >> on the command line unquoted causes problems, | |
1333 | too: C<``> command output capture has to perform a guessing game. It | |
1334 | assumes that a string C<< <[^<>]+\$[^<>]> >> is a | |
1335 | reference to an environment variable, whereas anything else involving | |
1336 | C<< < >> or C<< > >> is redirection, and generally manages to be 99% | |
1337 | right. Of course, the problem remains that scripts cannot rely on any | |
1338 | Unix tools being available, or that any tools found have Unix-like command | |
1339 | line arguments. | |
1340 | ||
1341 | Extensions and XS are, in theory, buildable by anyone using free | |
1342 | tools. In practice, many don't, as users of the Acorn platform are | |
1343 | used to binary distributions. MakeMaker does run, but no available | |
1344 | make currently copes with MakeMaker's makefiles; even if and when | |
1345 | this should be fixed, the lack of a Unix-like shell will cause | |
1346 | problems with makefile rules, especially lines of the form C<cd | |
1347 | sdbm && make all>, and anything using quoting. | |
1348 | ||
1349 | "S<RISC OS>" is the proper name for the operating system, but the value | |
1350 | in C<$^O> is "riscos" (because we don't like shouting). | |
1351 | ||
1352 | =head2 Other perls | |
1353 | ||
1354 | Perl has been ported to many platforms that do not fit into any of | |
1355 | the categories listed above. Some, such as AmigaOS, Atari MiNT, | |
1356 | BeOS, HP MPE/iX, QNX, Plan 9, and VOS, have been well-integrated | |
1357 | into the standard Perl source code kit. You may need to see the | |
1358 | F<ports/> directory on CPAN for information, and possibly binaries, | |
1359 | for the likes of: aos, Atari ST, lynxos, riscos, Novell Netware, | |
1360 | Tandem Guardian, I<etc.> (Yes, we know that some of these OSes may | |
1361 | fall under the Unix category, but we are not a standards body.) | |
1362 | ||
1363 | Some approximate operating system names and their C<$^O> values | |
1364 | in the "OTHER" category include: | |
1365 | ||
1366 | OS $^O $Config{'archname'} | |
1367 | ------------------------------------------ | |
1368 | Amiga DOS amigaos m68k-amigos | |
1369 | BeOS beos | |
1370 | MPE/iX mpeix PA-RISC1.1 | |
1371 | ||
1372 | See also: | |
1373 | ||
1374 | =over 4 | |
1375 | ||
1376 | =item * | |
1377 | ||
1378 | Amiga, F<README.amiga> (installed as L<perlamiga>). | |
1379 | ||
1380 | =item * | |
1381 | ||
1382 | Atari, F<README.mint> and Guido Flohr's web page | |
1383 | http://stud.uni-sb.de/~gufl0000/ | |
1384 | ||
1385 | =item * | |
1386 | ||
1387 | Be OS, F<README.beos> | |
1388 | ||
1389 | =item * | |
1390 | ||
1391 | HP 300 MPE/iX, F<README.mpeix> and Mark Bixby's web page | |
1392 | http://www.bixby.org/mark/perlix.html | |
1393 | ||
1394 | =item * | |
1395 | ||
1396 | A free perl5-based PERL.NLM for Novell Netware is available in | |
1397 | precompiled binary and source code form from http://www.novell.com/ | |
1398 | as well as from CPAN. | |
1399 | ||
1400 | =item * | |
1401 | ||
1402 | S<Plan 9>, F<README.plan9> | |
1403 | ||
1404 | =back | |
1405 | ||
1406 | =head1 FUNCTION IMPLEMENTATIONS | |
1407 | ||
1408 | Listed below are functions that are either completely unimplemented | |
1409 | or else have been implemented differently on various platforms. | |
1410 | Following each description will be, in parentheses, a list of | |
1411 | platforms that the description applies to. | |
1412 | ||
1413 | The list may well be incomplete, or even wrong in some places. When | |
1414 | in doubt, consult the platform-specific README files in the Perl | |
1415 | source distribution, and any other documentation resources accompanying | |
1416 | a given port. | |
1417 | ||
1418 | Be aware, moreover, that even among Unix-ish systems there are variations. | |
1419 | ||
1420 | For many functions, you can also query C<%Config>, exported by | |
1421 | default from the Config module. For example, to check whether the | |
1422 | platform has the C<lstat> call, check C<$Config{d_lstat}>. See | |
1423 | L<Config> for a full description of available variables. | |
1424 | ||
1425 | =head2 Alphabetical Listing of Perl Functions | |
1426 | ||
1427 | =over 8 | |
1428 | ||
1429 | =item -X FILEHANDLE | |
1430 | ||
1431 | =item -X EXPR | |
1432 | ||
1433 | =item -X | |
1434 | ||
1435 | C<-r>, C<-w>, and C<-x> have a limited meaning only; directories | |
1436 | and applications are executable, and there are no uid/gid | |
1437 | considerations. C<-o> is not supported. (S<Mac OS>) | |
1438 | ||
1439 | C<-r>, C<-w>, C<-x>, and C<-o> tell whether the file is accessible, | |
1440 | which may not reflect UIC-based file protections. (VMS) | |
1441 | ||
1442 | C<-s> returns the size of the data fork, not the total size of data fork | |
1443 | plus resource fork. (S<Mac OS>). | |
1444 | ||
1445 | C<-s> by name on an open file will return the space reserved on disk, | |
1446 | rather than the current extent. C<-s> on an open filehandle returns the | |
1447 | current size. (S<RISC OS>) | |
1448 | ||
1449 | C<-R>, C<-W>, C<-X>, C<-O> are indistinguishable from C<-r>, C<-w>, | |
1450 | C<-x>, C<-o>. (S<Mac OS>, Win32, VMS, S<RISC OS>) | |
1451 | ||
1452 | C<-b>, C<-c>, C<-k>, C<-g>, C<-p>, C<-u>, C<-A> are not implemented. | |
1453 | (S<Mac OS>) | |
1454 | ||
1455 | C<-g>, C<-k>, C<-l>, C<-p>, C<-u>, C<-A> are not particularly meaningful. | |
1456 | (Win32, VMS, S<RISC OS>) | |
1457 | ||
1458 | C<-d> is true if passed a device spec without an explicit directory. | |
1459 | (VMS) | |
1460 | ||
1461 | C<-T> and C<-B> are implemented, but might misclassify Mac text files | |
1462 | with foreign characters; this is the case will all platforms, but may | |
1463 | affect S<Mac OS> often. (S<Mac OS>) | |
1464 | ||
1465 | C<-x> (or C<-X>) determine if a file ends in one of the executable | |
1466 | suffixes. C<-S> is meaningless. (Win32) | |
1467 | ||
1468 | C<-x> (or C<-X>) determine if a file has an executable file type. | |
1469 | (S<RISC OS>) | |
1470 | ||
1471 | =item alarm SECONDS | |
1472 | ||
1473 | =item alarm | |
1474 | ||
1475 | Not implemented. (Win32) | |
1476 | ||
1477 | =item binmode FILEHANDLE | |
1478 | ||
1479 | Meaningless. (S<Mac OS>, S<RISC OS>) | |
1480 | ||
1481 | Reopens file and restores pointer; if function fails, underlying | |
1482 | filehandle may be closed, or pointer may be in a different position. | |
1483 | (VMS) | |
1484 | ||
1485 | The value returned by C<tell> may be affected after the call, and | |
1486 | the filehandle may be flushed. (Win32) | |
1487 | ||
1488 | =item chmod LIST | |
1489 | ||
1490 | Only limited meaning. Disabling/enabling write permission is mapped to | |
1491 | locking/unlocking the file. (S<Mac OS>) | |
1492 | ||
1493 | Only good for changing "owner" read-write access, "group", and "other" | |
1494 | bits are meaningless. (Win32) | |
1495 | ||
1496 | Only good for changing "owner" and "other" read-write access. (S<RISC OS>) | |
1497 | ||
1498 | Access permissions are mapped onto VOS access-control list changes. (VOS) | |
1499 | ||
1500 | The actual permissions set depend on the value of the C<CYGWIN> | |
1501 | in the SYSTEM environment settings. (Cygwin) | |
1502 | ||
1503 | =item chown LIST | |
1504 | ||
1505 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>, VOS) | |
1506 | ||
1507 | Does nothing, but won't fail. (Win32) | |
1508 | ||
1509 | =item chroot FILENAME | |
1510 | ||
1511 | =item chroot | |
1512 | ||
1513 | Not implemented. (S<Mac OS>, Win32, VMS, S<Plan 9>, S<RISC OS>, VOS, VM/ESA) | |
1514 | ||
1515 | =item crypt PLAINTEXT,SALT | |
1516 | ||
1517 | May not be available if library or source was not provided when building | |
1518 | perl. (Win32) | |
1519 | ||
1520 | Not implemented. (VOS) | |
1521 | ||
1522 | =item dbmclose HASH | |
1523 | ||
1524 | Not implemented. (VMS, S<Plan 9>, VOS) | |
1525 | ||
1526 | =item dbmopen HASH,DBNAME,MODE | |
1527 | ||
1528 | Not implemented. (VMS, S<Plan 9>, VOS) | |
1529 | ||
1530 | =item dump LABEL | |
1531 | ||
1532 | Not useful. (S<Mac OS>, S<RISC OS>) | |
1533 | ||
1534 | Not implemented. (Win32) | |
1535 | ||
1536 | Invokes VMS debugger. (VMS) | |
1537 | ||
1538 | =item exec LIST | |
1539 | ||
1540 | Not implemented. (S<Mac OS>) | |
1541 | ||
1542 | Implemented via Spawn. (VM/ESA) | |
1543 | ||
1544 | Does not automatically flush output handles on some platforms. | |
1545 | (SunOS, Solaris, HP-UX) | |
1546 | ||
1547 | =item exit EXPR | |
1548 | ||
1549 | =item exit | |
1550 | ||
1551 | Emulates UNIX exit() (which considers C<exit 1> to indicate an error) by | |
1552 | mapping the C<1> to SS$_ABORT (C<44>). This behavior may be overridden | |
1553 | with the pragma C<use vmsish 'exit'>. As with the CRTL's exit() | |
1554 | function, C<exit 0> is also mapped to an exit status of SS$_NORMAL | |
1555 | (C<1>); this mapping cannot be overridden. Any other argument to exit() | |
1556 | is used directly as Perl's exit status. (VMS) | |
1557 | ||
1558 | =item fcntl FILEHANDLE,FUNCTION,SCALAR | |
1559 | ||
1560 | Not implemented. (Win32, VMS) | |
1561 | ||
1562 | =item flock FILEHANDLE,OPERATION | |
1563 | ||
1564 | Not implemented (S<Mac OS>, VMS, S<RISC OS>, VOS). | |
1565 | ||
1566 | Available only on Windows NT (not on Windows 95). (Win32) | |
1567 | ||
1568 | =item fork | |
1569 | ||
1570 | Not implemented. (S<Mac OS>, AmigaOS, S<RISC OS>, VOS, VM/ESA, VMS) | |
1571 | ||
1572 | Emulated using multiple interpreters. See L<perlfork>. (Win32) | |
1573 | ||
1574 | Does not automatically flush output handles on some platforms. | |
1575 | (SunOS, Solaris, HP-UX) | |
1576 | ||
1577 | =item getlogin | |
1578 | ||
1579 | Not implemented. (S<Mac OS>, S<RISC OS>) | |
1580 | ||
1581 | =item getpgrp PID | |
1582 | ||
1583 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS) | |
1584 | ||
1585 | =item getppid | |
1586 | ||
1587 | Not implemented. (S<Mac OS>, Win32, S<RISC OS>) | |
1588 | ||
1589 | =item getpriority WHICH,WHO | |
1590 | ||
1591 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA) | |
1592 | ||
1593 | =item getpwnam NAME | |
1594 | ||
1595 | Not implemented. (S<Mac OS>, Win32) | |
1596 | ||
1597 | Not useful. (S<RISC OS>) | |
1598 | ||
1599 | =item getgrnam NAME | |
1600 | ||
1601 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>) | |
1602 | ||
1603 | =item getnetbyname NAME | |
1604 | ||
1605 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) | |
1606 | ||
1607 | =item getpwuid UID | |
1608 | ||
1609 | Not implemented. (S<Mac OS>, Win32) | |
1610 | ||
1611 | Not useful. (S<RISC OS>) | |
1612 | ||
1613 | =item getgrgid GID | |
1614 | ||
1615 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>) | |
1616 | ||
1617 | =item getnetbyaddr ADDR,ADDRTYPE | |
1618 | ||
1619 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) | |
1620 | ||
1621 | =item getprotobynumber NUMBER | |
1622 | ||
1623 | Not implemented. (S<Mac OS>) | |
1624 | ||
1625 | =item getservbyport PORT,PROTO | |
1626 | ||
1627 | Not implemented. (S<Mac OS>) | |
1628 | ||
1629 | =item getpwent | |
1630 | ||
1631 | Not implemented. (S<Mac OS>, Win32, VM/ESA) | |
1632 | ||
1633 | =item getgrent | |
1634 | ||
1635 | Not implemented. (S<Mac OS>, Win32, VMS, VM/ESA) | |
1636 | ||
1637 | =item gethostent | |
1638 | ||
1639 | Not implemented. (S<Mac OS>, Win32) | |
1640 | ||
1641 | =item getnetent | |
1642 | ||
1643 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) | |
1644 | ||
1645 | =item getprotoent | |
1646 | ||
1647 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) | |
1648 | ||
1649 | =item getservent | |
1650 | ||
1651 | Not implemented. (Win32, S<Plan 9>) | |
1652 | ||
1653 | =item sethostent STAYOPEN | |
1654 | ||
1655 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>) | |
1656 | ||
1657 | =item setnetent STAYOPEN | |
1658 | ||
1659 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>) | |
1660 | ||
1661 | =item setprotoent STAYOPEN | |
1662 | ||
1663 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>) | |
1664 | ||
1665 | =item setservent STAYOPEN | |
1666 | ||
1667 | Not implemented. (S<Plan 9>, Win32, S<RISC OS>) | |
1668 | ||
1669 | =item endpwent | |
1670 | ||
1671 | Not implemented. (S<Mac OS>, MPE/iX, VM/ESA, Win32) | |
1672 | ||
1673 | =item endgrent | |
1674 | ||
1675 | Not implemented. (S<Mac OS>, MPE/iX, S<RISC OS>, VM/ESA, VMS, Win32) | |
1676 | ||
1677 | =item endhostent | |
1678 | ||
1679 | Not implemented. (S<Mac OS>, Win32) | |
1680 | ||
1681 | =item endnetent | |
1682 | ||
1683 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) | |
1684 | ||
1685 | =item endprotoent | |
1686 | ||
1687 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) | |
1688 | ||
1689 | =item endservent | |
1690 | ||
1691 | Not implemented. (S<Plan 9>, Win32) | |
1692 | ||
1693 | =item getsockopt SOCKET,LEVEL,OPTNAME | |
1694 | ||
1695 | Not implemented. (S<Plan 9>) | |
1696 | ||
1697 | =item glob EXPR | |
1698 | ||
1699 | =item glob | |
1700 | ||
1701 | This operator is implemented via the File::Glob extension on most | |
1702 | platforms. See L<File::Glob> for portability information. | |
1703 | ||
1704 | =item ioctl FILEHANDLE,FUNCTION,SCALAR | |
1705 | ||
1706 | Not implemented. (VMS) | |
1707 | ||
1708 | Available only for socket handles, and it does what the ioctlsocket() call | |
1709 | in the Winsock API does. (Win32) | |
1710 | ||
1711 | Available only for socket handles. (S<RISC OS>) | |
1712 | ||
1713 | =item kill SIGNAL, LIST | |
1714 | ||
1715 | C<kill(0, LIST)> is implemented for the sake of taint checking; | |
1716 | use with other signals is unimplemented. (S<Mac OS>) | |
1717 | ||
1718 | Not implemented, hence not useful for taint checking. (S<RISC OS>) | |
1719 | ||
1720 | C<kill()> doesn't have the semantics of C<raise()>, i.e. it doesn't send | |
1721 | a signal to the identified process like it does on Unix platforms. | |
1722 | Instead C<kill($sig, $pid)> terminates the process identified by $pid, | |
1723 | and makes it exit immediately with exit status $sig. As in Unix, if | |
1724 | $sig is 0 and the specified process exists, it returns true without | |
1725 | actually terminating it. (Win32) | |
1726 | ||
1727 | =item link OLDFILE,NEWFILE | |
1728 | ||
1729 | Not implemented. (S<Mac OS>, MPE/iX, VMS, S<RISC OS>) | |
1730 | ||
1731 | Link count not updated because hard links are not quite that hard | |
1732 | (They are sort of half-way between hard and soft links). (AmigaOS) | |
1733 | ||
1734 | Hard links are implemented on Win32 (Windows NT and Windows 2000) | |
1735 | under NTFS only. | |
1736 | ||
1737 | =item lstat FILEHANDLE | |
1738 | ||
1739 | =item lstat EXPR | |
1740 | ||
1741 | =item lstat | |
1742 | ||
1743 | Not implemented. (VMS, S<RISC OS>) | |
1744 | ||
1745 | Return values (especially for device and inode) may be bogus. (Win32) | |
1746 | ||
1747 | =item msgctl ID,CMD,ARG | |
1748 | ||
1749 | =item msgget KEY,FLAGS | |
1750 | ||
1751 | =item msgsnd ID,MSG,FLAGS | |
1752 | ||
1753 | =item msgrcv ID,VAR,SIZE,TYPE,FLAGS | |
1754 | ||
1755 | Not implemented. (S<Mac OS>, Win32, VMS, S<Plan 9>, S<RISC OS>, VOS) | |
1756 | ||
1757 | =item open FILEHANDLE,EXPR | |
1758 | ||
1759 | =item open FILEHANDLE | |
1760 | ||
1761 | The C<|> variants are supported only if ToolServer is installed. | |
1762 | (S<Mac OS>) | |
1763 | ||
1764 | open to C<|-> and C<-|> are unsupported. (S<Mac OS>, Win32, S<RISC OS>) | |
1765 | ||
1766 | Opening a process does not automatically flush output handles on some | |
1767 | platforms. (SunOS, Solaris, HP-UX) | |
1768 | ||
1769 | =item pipe READHANDLE,WRITEHANDLE | |
1770 | ||
1771 | Very limited functionality. (MiNT) | |
1772 | ||
1773 | =item readlink EXPR | |
1774 | ||
1775 | =item readlink | |
1776 | ||
1777 | Not implemented. (Win32, VMS, S<RISC OS>) | |
1778 | ||
1779 | =item select RBITS,WBITS,EBITS,TIMEOUT | |
1780 | ||
1781 | Only implemented on sockets. (Win32, VMS) | |
1782 | ||
1783 | Only reliable on sockets. (S<RISC OS>) | |
1784 | ||
1785 | Note that the C<select FILEHANDLE> form is generally portable. | |
1786 | ||
1787 | =item semctl ID,SEMNUM,CMD,ARG | |
1788 | ||
1789 | =item semget KEY,NSEMS,FLAGS | |
1790 | ||
1791 | =item semop KEY,OPSTRING | |
1792 | ||
1793 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS) | |
1794 | ||
1795 | =item setgrent | |
1796 | ||
1797 | Not implemented. (S<Mac OS>, MPE/iX, VMS, Win32, S<RISC OS>) | |
1798 | ||
1799 | =item setpgrp PID,PGRP | |
1800 | ||
1801 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS) | |
1802 | ||
1803 | =item setpriority WHICH,WHO,PRIORITY | |
1804 | ||
1805 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS) | |
1806 | ||
1807 | =item setpwent | |
1808 | ||
1809 | Not implemented. (S<Mac OS>, MPE/iX, Win32, S<RISC OS>) | |
1810 | ||
1811 | =item setsockopt SOCKET,LEVEL,OPTNAME,OPTVAL | |
1812 | ||
1813 | Not implemented. (S<Plan 9>) | |
1814 | ||
1815 | =item shmctl ID,CMD,ARG | |
1816 | ||
1817 | =item shmget KEY,SIZE,FLAGS | |
1818 | ||
1819 | =item shmread ID,VAR,POS,SIZE | |
1820 | ||
1821 | =item shmwrite ID,STRING,POS,SIZE | |
1822 | ||
1823 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS) | |
1824 | ||
1825 | =item sockatmark SOCKET | |
1826 | ||
1827 | A relatively recent addition to socket functions, may not | |
1828 | be implemented even in UNIX platforms. | |
1829 | ||
1830 | =item socketpair SOCKET1,SOCKET2,DOMAIN,TYPE,PROTOCOL | |
1831 | ||
1832 | Not implemented. (Win32, VMS, S<RISC OS>, VOS, VM/ESA) | |
1833 | ||
1834 | =item stat FILEHANDLE | |
1835 | ||
1836 | =item stat EXPR | |
1837 | ||
1838 | =item stat | |
1839 | ||
1840 | Platforms that do not have rdev, blksize, or blocks will return these | |
1841 | as '', so numeric comparison or manipulation of these fields may cause | |
1842 | 'not numeric' warnings. | |
1843 | ||
1844 | mtime and atime are the same thing, and ctime is creation time instead of | |
1845 | inode change time. (S<Mac OS>). | |
1846 | ||
1847 | ctime not supported on UFS (S<Mac OS X>). | |
1848 | ||
1849 | ctime is creation time instead of inode change time (Win32). | |
1850 | ||
1851 | device and inode are not meaningful. (Win32) | |
1852 | ||
1853 | device and inode are not necessarily reliable. (VMS) | |
1854 | ||
1855 | mtime, atime and ctime all return the last modification time. Device and | |
1856 | inode are not necessarily reliable. (S<RISC OS>) | |
1857 | ||
1858 | dev, rdev, blksize, and blocks are not available. inode is not | |
1859 | meaningful and will differ between stat calls on the same file. (os2) | |
1860 | ||
1861 | some versions of cygwin when doing a stat("foo") and if not finding it | |
1862 | may then attempt to stat("foo.exe") (Cygwin) | |
1863 | ||
1864 | =item symlink OLDFILE,NEWFILE | |
1865 | ||
1866 | Not implemented. (Win32, VMS, S<RISC OS>) | |
1867 | ||
1868 | =item syscall LIST | |
1869 | ||
1870 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA) | |
1871 | ||
1872 | =item sysopen FILEHANDLE,FILENAME,MODE,PERMS | |
1873 | ||
1874 | The traditional "0", "1", and "2" MODEs are implemented with different | |
1875 | numeric values on some systems. The flags exported by C<Fcntl> | |
1876 | (O_RDONLY, O_WRONLY, O_RDWR) should work everywhere though. (S<Mac | |
1877 | OS>, OS/390, VM/ESA) | |
1878 | ||
1879 | =item system LIST | |
1880 | ||
1881 | In general, do not assume the UNIX/POSIX semantics that you can shift | |
1882 | C<$?> right by eight to get the exit value, or that C<$? & 127> | |
1883 | would give you the number of the signal that terminated the program, | |
1884 | or that C<$? & 128> would test true if the program was terminated by a | |
1885 | coredump. Instead, use the POSIX W*() interfaces: for example, use | |
1886 | WIFEXITED($?) and WEXITVALUE($?) to test for a normal exit and the exit | |
1887 | value, WIFSIGNALED($?) and WTERMSIG($?) for a signal exit and the | |
1888 | signal. Core dumping is not a portable concept, so there's no portable | |
1889 | way to test for that. | |
1890 | ||
1891 | Only implemented if ToolServer is installed. (S<Mac OS>) | |
1892 | ||
1893 | As an optimization, may not call the command shell specified in | |
1894 | C<$ENV{PERL5SHELL}>. C<system(1, @args)> spawns an external | |
1895 | process and immediately returns its process designator, without | |
1896 | waiting for it to terminate. Return value may be used subsequently | |
1897 | in C<wait> or C<waitpid>. Failure to spawn() a subprocess is indicated | |
1898 | by setting $? to "255 << 8". C<$?> is set in a way compatible with | |
1899 | Unix (i.e. the exitstatus of the subprocess is obtained by "$? >> 8", | |
1900 | as described in the documentation). (Win32) | |
1901 | ||
1902 | There is no shell to process metacharacters, and the native standard is | |
1903 | to pass a command line terminated by "\n" "\r" or "\0" to the spawned | |
1904 | program. Redirection such as C<< > foo >> is performed (if at all) by | |
1905 | the run time library of the spawned program. C<system> I<list> will call | |
1906 | the Unix emulation library's C<exec> emulation, which attempts to provide | |
1907 | emulation of the stdin, stdout, stderr in force in the parent, providing | |
1908 | the child program uses a compatible version of the emulation library. | |
1909 | I<scalar> will call the native command line direct and no such emulation | |
1910 | of a child Unix program will exists. Mileage B<will> vary. (S<RISC OS>) | |
1911 | ||
1912 | Far from being POSIX compliant. Because there may be no underlying | |
1913 | /bin/sh tries to work around the problem by forking and execing the | |
1914 | first token in its argument string. Handles basic redirection | |
1915 | ("<" or ">") on its own behalf. (MiNT) | |
1916 | ||
1917 | Does not automatically flush output handles on some platforms. | |
1918 | (SunOS, Solaris, HP-UX) | |
1919 | ||
1920 | The return value is POSIX-like (shifted up by 8 bits), which only allows | |
1921 | room for a made-up value derived from the severity bits of the native | |
1922 | 32-bit condition code (unless overridden by C<use vmsish 'status'>). | |
1923 | For more details see L<perlvms/$?>. (VMS) | |
1924 | ||
1925 | =item times | |
1926 | ||
1927 | Only the first entry returned is nonzero. (S<Mac OS>) | |
1928 | ||
1929 | "cumulative" times will be bogus. On anything other than Windows NT | |
1930 | or Windows 2000, "system" time will be bogus, and "user" time is | |
1931 | actually the time returned by the clock() function in the C runtime | |
1932 | library. (Win32) | |
1933 | ||
1934 | Not useful. (S<RISC OS>) | |
1935 | ||
1936 | =item truncate FILEHANDLE,LENGTH | |
1937 | ||
1938 | =item truncate EXPR,LENGTH | |
1939 | ||
1940 | Not implemented. (Older versions of VMS) | |
1941 | ||
1942 | Truncation to zero-length only. (VOS) | |
1943 | ||
1944 | If a FILEHANDLE is supplied, it must be writable and opened in append | |
1945 | mode (i.e., use C<<< open(FH, '>>filename') >>> | |
1946 | or C<sysopen(FH,...,O_APPEND|O_RDWR)>. If a filename is supplied, it | |
1947 | should not be held open elsewhere. (Win32) | |
1948 | ||
1949 | =item umask EXPR | |
1950 | ||
1951 | =item umask | |
1952 | ||
1953 | Returns undef where unavailable, as of version 5.005. | |
1954 | ||
1955 | C<umask> works but the correct permissions are set only when the file | |
1956 | is finally closed. (AmigaOS) | |
1957 | ||
1958 | =item utime LIST | |
1959 | ||
1960 | Only the modification time is updated. (S<BeOS>, S<Mac OS>, VMS, S<RISC OS>) | |
1961 | ||
1962 | May not behave as expected. Behavior depends on the C runtime | |
1963 | library's implementation of utime(), and the filesystem being | |
1964 | used. The FAT filesystem typically does not support an "access | |
1965 | time" field, and it may limit timestamps to a granularity of | |
1966 | two seconds. (Win32) | |
1967 | ||
1968 | =item wait | |
1969 | ||
1970 | =item waitpid PID,FLAGS | |
1971 | ||
1972 | Not implemented. (S<Mac OS>, VOS) | |
1973 | ||
1974 | Can only be applied to process handles returned for processes spawned | |
1975 | using C<system(1, ...)> or pseudo processes created with C<fork()>. (Win32) | |
1976 | ||
1977 | Not useful. (S<RISC OS>) | |
1978 | ||
1979 | =back | |
1980 | ||
1981 | =head1 CHANGES | |
1982 | ||
1983 | =over 4 | |
1984 | ||
1985 | =item v1.48, 02 February 2001 | |
1986 | ||
1987 | Various updates from perl5-porters over the past year, supported | |
1988 | platforms update from Jarkko Hietaniemi. | |
1989 | ||
1990 | =item v1.47, 22 March 2000 | |
1991 | ||
1992 | Various cleanups from Tom Christiansen, including migration of | |
1993 | long platform listings from L<perl>. | |
1994 | ||
1995 | =item v1.46, 12 February 2000 | |
1996 | ||
1997 | Updates for VOS and MPE/iX. (Peter Prymmer) Other small changes. | |
1998 | ||
1999 | =item v1.45, 20 December 1999 | |
2000 | ||
2001 | Small changes from 5.005_63 distribution, more changes to EBCDIC info. | |
2002 | ||
2003 | =item v1.44, 19 July 1999 | |
2004 | ||
2005 | A bunch of updates from Peter Prymmer for C<$^O> values, | |
2006 | endianness, File::Spec, VMS, BS2000, OS/400. | |
2007 | ||
2008 | =item v1.43, 24 May 1999 | |
2009 | ||
2010 | Added a lot of cleaning up from Tom Christiansen. | |
2011 | ||
2012 | =item v1.42, 22 May 1999 | |
2013 | ||
2014 | Added notes about tests, sprintf/printf, and epoch offsets. | |
2015 | ||
2016 | =item v1.41, 19 May 1999 | |
2017 | ||
2018 | Lots more little changes to formatting and content. | |
2019 | ||
2020 | Added a bunch of C<$^O> and related values | |
2021 | for various platforms; fixed mail and web addresses, and added | |
2022 | and changed miscellaneous notes. (Peter Prymmer) | |
2023 | ||
2024 | =item v1.40, 11 April 1999 | |
2025 | ||
2026 | Miscellaneous changes. | |
2027 | ||
2028 | =item v1.39, 11 February 1999 | |
2029 | ||
2030 | Changes from Jarkko and EMX URL fixes Michael Schwern. Additional | |
2031 | note about newlines added. | |
2032 | ||
2033 | =item v1.38, 31 December 1998 | |
2034 | ||
2035 | More changes from Jarkko. | |
2036 | ||
2037 | =item v1.37, 19 December 1998 | |
2038 | ||
2039 | More minor changes. Merge two separate version 1.35 documents. | |
2040 | ||
2041 | =item v1.36, 9 September 1998 | |
2042 | ||
2043 | Updated for Stratus VOS. Also known as version 1.35. | |
2044 | ||
2045 | =item v1.35, 13 August 1998 | |
2046 | ||
2047 | Integrate more minor changes, plus addition of new sections under | |
2048 | L<"ISSUES">: L<"Numbers endianness and Width">, | |
2049 | L<"Character sets and character encoding">, | |
2050 | L<"Internationalisation">. | |
2051 | ||
2052 | =item v1.33, 06 August 1998 | |
2053 | ||
2054 | Integrate more minor changes. | |
2055 | ||
2056 | =item v1.32, 05 August 1998 | |
2057 | ||
2058 | Integrate more minor changes. | |
2059 | ||
2060 | =item v1.30, 03 August 1998 | |
2061 | ||
2062 | Major update for RISC OS, other minor changes. | |
2063 | ||
2064 | =item v1.23, 10 July 1998 | |
2065 | ||
2066 | First public release with perl5.005. | |
2067 | ||
2068 | =back | |
2069 | ||
2070 | =head1 Supported Platforms | |
2071 | ||
2072 | As of June 2002 (the Perl release 5.8.0), the following platforms are | |
2073 | able to build Perl from the standard source code distribution | |
2074 | available at http://www.cpan.org/src/index.html | |
2075 | ||
2076 | AIX | |
2077 | BeOS | |
2078 | Cygwin | |
2079 | DG/UX | |
2080 | DOS DJGPP 1) | |
2081 | DYNIX/ptx | |
2082 | EPOC R5 | |
2083 | FreeBSD | |
2084 | HP-UX | |
2085 | IRIX | |
2086 | Linux | |
2087 | Mac OS Classic | |
2088 | Mac OS X (Darwin) | |
2089 | MPE/iX | |
2090 | NetBSD | |
2091 | NetWare | |
2092 | NonStop-UX | |
2093 | ReliantUNIX (SINIX) | |
2094 | OpenBSD | |
2095 | OpenVMS (VMS) | |
2096 | OS/2 | |
2097 | PowerUX | |
2098 | POSIX-BC (BS2000) | |
2099 | QNX | |
2100 | Solaris | |
2101 | SunOS 4 | |
2102 | SUPER-UX | |
2103 | Tru64 UNIX (DEC OSF/1, Digital UNIX) | |
2104 | UNICOS | |
2105 | UNICOS/mk | |
2106 | UTS | |
2107 | VOS | |
2108 | Win95/98/ME/2K/XP 2) | |
2109 | WinCE | |
2110 | z/OS (OS/390) | |
2111 | VM/ESA | |
2112 | ||
2113 | 1) in DOS mode either the DOS or OS/2 ports can be used | |
2114 | 2) compilers: Borland, MinGW (GCC), VC6 | |
2115 | ||
2116 | The following platforms worked with the previous releases (5.6 and | |
2117 | 5.7), but we did not manage either to fix or to test these in time | |
2118 | for the 5.8.0 release. There is a very good chance that many of these | |
2119 | will work fine with the 5.8.0. | |
2120 | ||
2121 | BSD/OS | |
2122 | DomainOS | |
2123 | Hurd | |
2124 | LynxOS | |
2125 | MachTen | |
2126 | PowerMAX | |
2127 | SCO SV | |
2128 | SVR4 | |
2129 | Unixware | |
2130 | Windows 3.1 | |
2131 | ||
2132 | Known to be broken for 5.8.0 (but 5.6.1 and 5.7.2 can be used): | |
2133 | ||
2134 | AmigaOS | |
2135 | ||
2136 | The following platforms have been known to build Perl from source in | |
2137 | the past (5.005_03 and earlier), but we haven't been able to verify | |
2138 | their status for the current release, either because the | |
2139 | hardware/software platforms are rare or because we don't have an | |
2140 | active champion on these platforms--or both. They used to work, | |
2141 | though, so go ahead and try compiling them, and let perlbug@perl.org | |
2142 | of any trouble. | |
2143 | ||
2144 | 3b1 | |
2145 | A/UX | |
2146 | ConvexOS | |
2147 | CX/UX | |
2148 | DC/OSx | |
2149 | DDE SMES | |
2150 | DOS EMX | |
2151 | Dynix | |
2152 | EP/IX | |
2153 | ESIX | |
2154 | FPS | |
2155 | GENIX | |
2156 | Greenhills | |
2157 | ISC | |
2158 | MachTen 68k | |
2159 | MiNT | |
2160 | MPC | |
2161 | NEWS-OS | |
2162 | NextSTEP | |
2163 | OpenSTEP | |
2164 | Opus | |
2165 | Plan 9 | |
2166 | RISC/os | |
2167 | SCO ODT/OSR | |
2168 | Stellar | |
2169 | SVR2 | |
2170 | TI1500 | |
2171 | TitanOS | |
2172 | Ultrix | |
2173 | Unisys Dynix | |
2174 | ||
2175 | The following platforms have their own source code distributions and | |
2176 | binaries available via http://www.cpan.org/ports/ | |
2177 | ||
2178 | Perl release | |
2179 | ||
2180 | OS/400 5.005_02 | |
2181 | Tandem Guardian 5.004 | |
2182 | ||
2183 | The following platforms have only binaries available via | |
2184 | http://www.cpan.org/ports/index.html : | |
2185 | ||
2186 | Perl release | |
2187 | ||
2188 | Acorn RISCOS 5.005_02 | |
2189 | AOS 5.002 | |
2190 | LynxOS 5.004_02 | |
2191 | ||
2192 | Although we do suggest that you always build your own Perl from | |
2193 | the source code, both for maximal configurability and for security, | |
2194 | in case you are in a hurry you can check | |
2195 | http://www.cpan.org/ports/index.html for binary distributions. | |
2196 | ||
2197 | =head1 SEE ALSO | |
2198 | ||
2199 | L<perlaix>, L<perlamiga>, L<perlapollo>, L<perlbeos>, L<perlbs2000>, | |
2200 | L<perlce>, L<perlcygwin>, L<perldgux>, L<perldos>, L<perlepoc>, | |
2201 | L<perlebcdic>, L<perlfreebsd>, L<perlhurd>, L<perlhpux>, L<perlirix>, | |
2202 | L<perlmachten>, L<perlmacos>, L<perlmint>, L<perlmpeix>, | |
2203 | L<perlnetware>, L<perlos2>, L<perlos390>, L<perlplan9>, L<perlqnx>, | |
2204 | L<perlsolaris>, L<perltru64>, L<perlunicode>, L<perlvmesa>, | |
2205 | L<perlvms>, L<perlvos>, L<perlwin32>, and L<Win32>. | |
2206 | ||
2207 | =head1 AUTHORS / CONTRIBUTORS | |
2208 | ||
2209 | Abigail <abigail@foad.org>, | |
2210 | Charles Bailey <bailey@newman.upenn.edu>, | |
2211 | Graham Barr <gbarr@pobox.com>, | |
2212 | Tom Christiansen <tchrist@perl.com>, | |
2213 | Nicholas Clark <nick@ccl4.org>, | |
2214 | Thomas Dorner <Thomas.Dorner@start.de>, | |
2215 | Andy Dougherty <doughera@lafayette.edu>, | |
2216 | Dominic Dunlop <domo@computer.org>, | |
2217 | Neale Ferguson <neale@vma.tabnsw.com.au>, | |
2218 | David J. Fiander <davidf@mks.com>, | |
2219 | Paul Green <Paul_Green@stratus.com>, | |
2220 | M.J.T. Guy <mjtg@cam.ac.uk>, | |
2221 | Jarkko Hietaniemi <jhi@iki.fi>, | |
2222 | Luther Huffman <lutherh@stratcom.com>, | |
2223 | Nick Ing-Simmons <nick@ing-simmons.net>, | |
2224 | Andreas J. KE<ouml>nig <a.koenig@mind.de>, | |
2225 | Markus Laker <mlaker@contax.co.uk>, | |
2226 | Andrew M. Langmead <aml@world.std.com>, | |
2227 | Larry Moore <ljmoore@freespace.net>, | |
2228 | Paul Moore <Paul.Moore@uk.origin-it.com>, | |
2229 | Chris Nandor <pudge@pobox.com>, | |
2230 | Matthias Neeracher <neeracher@mac.com>, | |
2231 | Philip Newton <pne@cpan.org>, | |
2232 | Gary Ng <71564.1743@CompuServe.COM>, | |
2233 | Tom Phoenix <rootbeer@teleport.com>, | |
2234 | AndrE<eacute> Pirard <A.Pirard@ulg.ac.be>, | |
2235 | Peter Prymmer <pvhp@forte.com>, | |
2236 | Hugo van der Sanden <hv@crypt0.demon.co.uk>, | |
2237 | Gurusamy Sarathy <gsar@activestate.com>, | |
2238 | Paul J. Schinder <schinder@pobox.com>, | |
2239 | Michael G Schwern <schwern@pobox.com>, | |
2240 | Dan Sugalski <dan@sidhe.org>, | |
2241 | Nathan Torkington <gnat@frii.com>. | |
2242 |