Commit | Line | Data |
---|---|---|
e41182b5 GS |
1 | =head1 NAME |
2 | ||
3 | perlport - Writing portable Perl | |
4 | ||
e41182b5 GS |
5 | =head1 DESCRIPTION |
6 | ||
b7df3edc GS |
7 | Perl runs on numerous operating systems. While most of them share |
8 | much in common, they also have their own unique features. | |
e41182b5 GS |
9 | |
10 | This document is meant to help you to find out what constitutes portable | |
b7df3edc | 11 | Perl code. That way once you make a decision to write portably, |
e41182b5 GS |
12 | you know where the lines are drawn, and you can stay within them. |
13 | ||
b7df3edc GS |
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 | |
0a47030a GS |
26 | problem. |
27 | ||
28 | Looking at it another way, writing portable code is usually about | |
b7df3edc GS |
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. | |
e41182b5 GS |
32 | |
33 | Be aware of two important points: | |
34 | ||
35 | =over 4 | |
36 | ||
37 | =item Not all Perl programs have to be portable | |
38 | ||
b7df3edc | 39 | There is no reason you should not use Perl as a language to glue Unix |
e41182b5 GS |
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 | ||
b7df3edc | 44 | =item Nearly all of Perl already I<is> portable |
e41182b5 GS |
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 | |
6ab3f9cb | 50 | without modification. But there are some significant issues in |
e41182b5 GS |
51 | writing portable code, and this document is entirely about those issues. |
52 | ||
53 | =back | |
54 | ||
b7df3edc GS |
55 | Here's the general rule: When you approach a task commonly done |
56 | using a whole range of platforms, think about writing portable | |
e41182b5 GS |
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 | ||
b7df3edc GS |
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 | |
0a47030a GS |
67 | deliberate in your decision. |
68 | ||
69 | The material below is separated into three main sections: main issues of | |
7ee27b7c | 70 | portability (L<"ISSUES">), platform-specific issues (L<"PLATFORMS">), and |
b7df3edc | 71 | built-in perl functions that behave differently on various ports |
7ee27b7c | 72 | (L<"FUNCTION IMPLEMENTATIONS">). |
e41182b5 GS |
73 | |
74 | This information should not be considered complete; it includes possibly | |
b8099c3d | 75 | transient information about idiosyncrasies of some of the ports, almost |
b7df3edc | 76 | all of which are in a state of constant evolution. Thus, this material |
e41182b5 | 77 | should be considered a perpetual work in progress |
cc07ed0b | 78 | (C<< <IMG SRC="yellow_sign.gif" ALT="Under Construction"> >>). |
e41182b5 | 79 | |
e41182b5 GS |
80 | =head1 ISSUES |
81 | ||
82 | =head2 Newlines | |
83 | ||
638bc118 | 84 | In most operating systems, lines in files are terminated by newlines. |
e41182b5 | 85 | Just what is used as a newline may vary from OS to OS. Unix |
b7df3edc | 86 | traditionally uses C<\012>, one type of DOSish I/O uses C<\015\012>, |
e41182b5 GS |
87 | and S<Mac OS> uses C<\015>. |
88 | ||
b7df3edc GS |
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 | |
56d7751a | 93 | from) C<\015\012>, depending on whether you're reading or writing. |
b7df3edc GS |
94 | Unix does the same thing on ttys in canonical mode. C<\015\012> |
95 | is commonly referred to as CRLF. | |
96 | ||
31bb23b1 XN |
97 | To trim trailing newlines from text lines use chomp(). With default |
98 | settings that function looks for a trailing C<\n> character and thus | |
99 | trims in a portable way. | |
5b3eff12 MS |
100 | |
101 | When dealing with binary files (or text files in binary mode) be sure | |
102 | to explicitly set $/ to the appropriate value for your file format | |
103 | before using chomp(). | |
104 | ||
b7df3edc GS |
105 | Because of the "text" mode translation, DOSish perls have limitations |
106 | in using C<seek> and C<tell> on a file accessed in "text" mode. | |
107 | Stick to C<seek>-ing to locations you got from C<tell> (and no | |
108 | others), and you are usually free to use C<seek> and C<tell> even | |
109 | in "text" mode. Using C<seek> or C<tell> or other file operations | |
110 | may be non-portable. If you use C<binmode> on a file, however, you | |
111 | can usually C<seek> and C<tell> with arbitrary values in safety. | |
e41182b5 GS |
112 | |
113 | A common misconception in socket programming is that C<\n> eq C<\012> | |
0a47030a | 114 | everywhere. When using protocols such as common Internet protocols, |
e41182b5 GS |
115 | C<\012> and C<\015> are called for specifically, and the values of |
116 | the logical C<\n> and C<\r> (carriage return) are not reliable. | |
117 | ||
118 | print SOCKET "Hi there, client!\r\n"; # WRONG | |
119 | print SOCKET "Hi there, client!\015\012"; # RIGHT | |
120 | ||
0a47030a GS |
121 | However, using C<\015\012> (or C<\cM\cJ>, or C<\x0D\x0A>) can be tedious |
122 | and unsightly, as well as confusing to those maintaining the code. As | |
6ab3f9cb | 123 | such, the Socket module supplies the Right Thing for those who want it. |
e41182b5 GS |
124 | |
125 | use Socket qw(:DEFAULT :crlf); | |
126 | print SOCKET "Hi there, client!$CRLF" # RIGHT | |
127 | ||
6ab3f9cb | 128 | When reading from a socket, remember that the default input record |
b7df3edc GS |
129 | separator C<$/> is C<\n>, but robust socket code will recognize as |
130 | either C<\012> or C<\015\012> as end of line: | |
e41182b5 GS |
131 | |
132 | while (<SOCKET>) { | |
133 | # ... | |
134 | } | |
135 | ||
b7df3edc GS |
136 | Because both CRLF and LF end in LF, the input record separator can |
137 | be set to LF and any CR stripped later. Better to write: | |
e41182b5 GS |
138 | |
139 | use Socket qw(:DEFAULT :crlf); | |
140 | local($/) = LF; # not needed if $/ is already \012 | |
141 | ||
142 | while (<SOCKET>) { | |
143 | s/$CR?$LF/\n/; # not sure if socket uses LF or CRLF, OK | |
144 | # s/\015?\012/\n/; # same thing | |
145 | } | |
146 | ||
b7df3edc GS |
147 | This example is preferred over the previous one--even for Unix |
148 | platforms--because now any C<\015>'s (C<\cM>'s) are stripped out | |
e41182b5 GS |
149 | (and there was much rejoicing). |
150 | ||
6ab3f9cb | 151 | Similarly, functions that return text data--such as a function that |
b7df3edc GS |
152 | fetches a web page--should sometimes translate newlines before |
153 | returning the data, if they've not yet been translated to the local | |
154 | newline representation. A single line of code will often suffice: | |
2ee0eb3c | 155 | |
b7df3edc GS |
156 | $data =~ s/\015?\012/\n/g; |
157 | return $data; | |
2ee0eb3c | 158 | |
6ab3f9cb GS |
159 | Some of this may be confusing. Here's a handy reference to the ASCII CR |
160 | and LF characters. You can print it out and stick it in your wallet. | |
161 | ||
74555b7a PP |
162 | LF eq \012 eq \x0A eq \cJ eq chr(10) eq ASCII 10 |
163 | CR eq \015 eq \x0D eq \cM eq chr(13) eq ASCII 13 | |
6ab3f9cb GS |
164 | |
165 | | Unix | DOS | Mac | | |
166 | --------------------------- | |
167 | \n | LF | LF | CR | | |
168 | \r | CR | CR | LF | | |
169 | \n * | LF | CRLF | CR | | |
170 | \r * | CR | CR | LF | | |
171 | --------------------------- | |
172 | * text-mode STDIO | |
173 | ||
b7df3edc GS |
174 | The Unix column assumes that you are not accessing a serial line |
175 | (like a tty) in canonical mode. If you are, then CR on input becomes | |
176 | "\n", and "\n" on output becomes CRLF. | |
177 | ||
6ab3f9cb | 178 | These are just the most common definitions of C<\n> and C<\r> in Perl. |
522b859a JH |
179 | There may well be others. For example, on an EBCDIC implementation |
180 | such as z/OS (OS/390) or OS/400 (using the ILE, the PASE is ASCII-based) | |
181 | the above material is similar to "Unix" but the code numbers change: | |
74555b7a | 182 | |
d770bc45 TS |
183 | LF eq \025 eq \x15 eq \cU eq chr(21) eq CP-1047 21 |
184 | LF eq \045 eq \x25 eq chr(37) eq CP-0037 37 | |
74555b7a PP |
185 | CR eq \015 eq \x0D eq \cM eq chr(13) eq CP-1047 13 |
186 | CR eq \015 eq \x0D eq \cM eq chr(13) eq CP-0037 13 | |
187 | ||
188 | | z/OS | OS/400 | | |
189 | ---------------------- | |
190 | \n | LF | LF | | |
191 | \r | CR | CR | | |
192 | \n * | LF | LF | | |
193 | \r * | CR | CR | | |
194 | ---------------------- | |
195 | * text-mode STDIO | |
6ab3f9cb | 196 | |
322422de GS |
197 | =head2 Numbers endianness and Width |
198 | ||
199 | Different CPUs store integers and floating point numbers in different | |
200 | orders (called I<endianness>) and widths (32-bit and 64-bit being the | |
b7df3edc GS |
201 | most common today). This affects your programs when they attempt to transfer |
202 | numbers in binary format from one CPU architecture to another, | |
203 | usually either "live" via network connection, or by storing the | |
204 | numbers to secondary storage such as a disk file or tape. | |
322422de | 205 | |
b7df3edc | 206 | Conflicting storage orders make utter mess out of the numbers. If a |
d1e3b762 | 207 | little-endian host (Intel, VAX) stores 0x12345678 (305419896 in |
b84d4f81 JH |
208 | decimal), a big-endian host (Motorola, Sparc, PA) reads it as |
209 | 0x78563412 (2018915346 in decimal). Alpha and MIPS can be either: | |
210 | Digital/Compaq used/uses them in little-endian mode; SGI/Cray uses | |
211 | them in big-endian mode. To avoid this problem in network (socket) | |
212 | connections use the C<pack> and C<unpack> formats C<n> and C<N>, the | |
213 | "network" orders. These are guaranteed to be portable. | |
322422de | 214 | |
7a4d2905 | 215 | As of perl 5.9.2, you can also use the C<E<gt>> and C<E<lt>> modifiers |
1109a392 MHM |
216 | to force big- or little-endian byte-order. This is useful if you want |
217 | to store signed integers or 64-bit integers, for example. | |
218 | ||
d1e3b762 GS |
219 | You can explore the endianness of your platform by unpacking a |
220 | data structure packed in native format such as: | |
221 | ||
222 | print unpack("h*", pack("s2", 1, 2)), "\n"; | |
223 | # '10002000' on e.g. Intel x86 or Alpha 21064 in little-endian mode | |
224 | # '00100020' on e.g. Motorola 68040 | |
225 | ||
226 | If you need to distinguish between endian architectures you could use | |
227 | either of the variables set like so: | |
228 | ||
229 | $is_big_endian = unpack("h*", pack("s", 1)) =~ /01/; | |
4375e838 | 230 | $is_little_endian = unpack("h*", pack("s", 1)) =~ /^1/; |
d1e3b762 | 231 | |
b7df3edc GS |
232 | Differing widths can cause truncation even between platforms of equal |
233 | endianness. The platform of shorter width loses the upper parts of the | |
322422de GS |
234 | number. There is no good solution for this problem except to avoid |
235 | transferring or storing raw binary numbers. | |
236 | ||
b7df3edc | 237 | One can circumnavigate both these problems in two ways. Either |
322422de | 238 | transfer and store numbers always in text format, instead of raw |
b7df3edc | 239 | binary, or else consider using modules like Data::Dumper (included in |
83df6a1d JH |
240 | the standard distribution as of Perl 5.005) and Storable (included as |
241 | of perl 5.8). Keeping all data as text significantly simplifies matters. | |
322422de | 242 | |
979699d9 JH |
243 | The v-strings are portable only up to v2147483647 (0x7FFFFFFF), that's |
244 | how far EBCDIC, or more precisely UTF-EBCDIC will go. | |
245 | ||
433acd8a | 246 | =head2 Files and Filesystems |
e41182b5 GS |
247 | |
248 | Most platforms these days structure files in a hierarchical fashion. | |
b7df3edc | 249 | So, it is reasonably safe to assume that all platforms support the |
6ab3f9cb | 250 | notion of a "path" to uniquely identify a file on the system. How |
b7df3edc | 251 | that path is really written, though, differs considerably. |
e41182b5 | 252 | |
4375e838 | 253 | Although similar, file path specifications differ between Unix, |
b7df3edc GS |
254 | Windows, S<Mac OS>, OS/2, VMS, VOS, S<RISC OS>, and probably others. |
255 | Unix, for example, is one of the few OSes that has the elegant idea | |
256 | of a single root directory. | |
322422de | 257 | |
6ab3f9cb GS |
258 | DOS, OS/2, VMS, VOS, and Windows can work similarly to Unix with C</> |
259 | as path separator, or in their own idiosyncratic ways (such as having | |
260 | several root directories and various "unrooted" device files such NIL: | |
261 | and LPT:). | |
322422de GS |
262 | |
263 | S<Mac OS> uses C<:> as a path separator instead of C</>. | |
264 | ||
6ab3f9cb GS |
265 | The filesystem may support neither hard links (C<link>) nor |
266 | symbolic links (C<symlink>, C<readlink>, C<lstat>). | |
433acd8a | 267 | |
6ab3f9cb | 268 | The filesystem may support neither access timestamp nor change |
433acd8a JH |
269 | timestamp (meaning that about the only portable timestamp is the |
270 | modification timestamp), or one second granularity of any timestamps | |
271 | (e.g. the FAT filesystem limits the time granularity to two seconds). | |
272 | ||
74555b7a | 273 | The "inode change timestamp" (the C<-C> filetest) may really be the |
95a3fe12 MS |
274 | "creation timestamp" (which it is not in UNIX). |
275 | ||
495c5fdc PG |
276 | VOS perl can emulate Unix filenames with C</> as path separator. The |
277 | native pathname characters greater-than, less-than, number-sign, and | |
278 | percent-sign are always accepted. | |
279 | ||
6ab3f9cb | 280 | S<RISC OS> perl can emulate Unix filenames with C</> as path |
322422de | 281 | separator, or go native and use C<.> for path separator and C<:> to |
6ab3f9cb | 282 | signal filesystems and disk names. |
e41182b5 | 283 | |
a1667ba3 JH |
284 | Don't assume UNIX filesystem access semantics: that read, write, |
285 | and execute are all the permissions there are, and even if they exist, | |
286 | that their semantics (for example what do r, w, and x mean on | |
287 | a directory) are the UNIX ones. The various UNIX/POSIX compatibility | |
288 | layers usually try to make interfaces like chmod() work, but sometimes | |
289 | there simply is no good mapping. | |
290 | ||
b7df3edc GS |
291 | If all this is intimidating, have no (well, maybe only a little) |
292 | fear. There are modules that can help. The File::Spec modules | |
293 | provide methods to do the Right Thing on whatever platform happens | |
294 | to be running the program. | |
e41182b5 | 295 | |
6ab3f9cb GS |
296 | use File::Spec::Functions; |
297 | chdir(updir()); # go up one directory | |
298 | $file = catfile(curdir(), 'temp', 'file.txt'); | |
e41182b5 GS |
299 | # on Unix and Win32, './temp/file.txt' |
300 | # on Mac OS, ':temp:file.txt' | |
d1e3b762 | 301 | # on VMS, '[.temp]file.txt' |
e41182b5 | 302 | |
b7df3edc | 303 | File::Spec is available in the standard distribution as of version |
d1e3b762 GS |
304 | 5.004_05. File::Spec::Functions is only in File::Spec 0.7 and later, |
305 | and some versions of perl come with version 0.6. If File::Spec | |
306 | is not updated to 0.7 or later, you must use the object-oriented | |
307 | interface from File::Spec (or upgrade File::Spec). | |
e41182b5 | 308 | |
b7df3edc GS |
309 | In general, production code should not have file paths hardcoded. |
310 | Making them user-supplied or read from a configuration file is | |
311 | better, keeping in mind that file path syntax varies on different | |
312 | machines. | |
e41182b5 GS |
313 | |
314 | This is especially noticeable in scripts like Makefiles and test suites, | |
315 | which often assume C</> as a path separator for subdirectories. | |
316 | ||
b7df3edc | 317 | Also of use is File::Basename from the standard distribution, which |
e41182b5 GS |
318 | splits a pathname into pieces (base filename, full path to directory, |
319 | and file suffix). | |
320 | ||
19799a22 | 321 | Even when on a single platform (if you can call Unix a single platform), |
b7df3edc | 322 | remember not to count on the existence or the contents of particular |
3c075c7d | 323 | system-specific files or directories, like F</etc/passwd>, |
b7df3edc GS |
324 | F</etc/sendmail.conf>, F</etc/resolv.conf>, or even F</tmp/>. For |
325 | example, F</etc/passwd> may exist but not contain the encrypted | |
326 | passwords, because the system is using some form of enhanced security. | |
327 | Or it may not contain all the accounts, because the system is using NIS. | |
3c075c7d | 328 | If code does need to rely on such a file, include a description of the |
b7df3edc | 329 | file and its format in the code's documentation, then make it easy for |
3c075c7d CN |
330 | the user to override the default location of the file. |
331 | ||
b7df3edc GS |
332 | Don't assume a text file will end with a newline. They should, |
333 | but people forget. | |
e41182b5 | 334 | |
ec481373 JH |
335 | Do not have two files or directories of the same name with different |
336 | case, like F<test.pl> and F<Test.pl>, as many platforms have | |
337 | case-insensitive (or at least case-forgiving) filenames. Also, try | |
338 | not to have non-word characters (except for C<.>) in the names, and | |
339 | keep them to the 8.3 convention, for maximum portability, onerous a | |
340 | burden though this may appear. | |
dd9f0070 | 341 | |
b7df3edc GS |
342 | Likewise, when using the AutoSplit module, try to keep your functions to |
343 | 8.3 naming and case-insensitive conventions; or, at the least, | |
dd9f0070 CN |
344 | make it so the resulting files have a unique (case-insensitively) |
345 | first 8 characters. | |
346 | ||
ec481373 JH |
347 | Whitespace in filenames is tolerated on most systems, but not all, |
348 | and even on systems where it might be tolerated, some utilities | |
fe829689 | 349 | might become confused by such whitespace. |
ec481373 | 350 | |
016930a6 JM |
351 | Many systems (DOS, VMS ODS-2) cannot have more than one C<.> in their |
352 | filenames. | |
433acd8a | 353 | |
c47ff5f1 | 354 | Don't assume C<< > >> won't be the first character of a filename. |
fe829689 PN |
355 | Always use C<< < >> explicitly to open a file for reading, or even |
356 | better, use the three-arg version of open, unless you want the user to | |
357 | be able to specify a pipe open. | |
0a47030a | 358 | |
fe829689 | 359 | open(FILE, '<', $existing_file) or die $!; |
0a47030a | 360 | |
6ab3f9cb GS |
361 | If filenames might use strange characters, it is safest to open it |
362 | with C<sysopen> instead of C<open>. C<open> is magic and can | |
c47ff5f1 | 363 | translate characters like C<< > >>, C<< < >>, and C<|>, which may |
b7df3edc | 364 | be the wrong thing to do. (Sometimes, though, it's the right thing.) |
fe829689 PN |
365 | Three-arg open can also help protect against this translation in cases |
366 | where it is undesirable. | |
e41182b5 | 367 | |
ec481373 | 368 | Don't use C<:> as a part of a filename since many systems use that for |
8939ba94 | 369 | their own semantics (Mac OS Classic for separating pathname components, |
ec481373 | 370 | many networking schemes and utilities for separating the nodename and |
08fef530 JH |
371 | the pathname, and so on). For the same reasons, avoid C<@>, C<;> and |
372 | C<|>. | |
ec481373 | 373 | |
e1516da7 JH |
374 | Don't assume that in pathnames you can collapse two leading slashes |
375 | C<//> into one: some networking and clustering filesystems have special | |
376 | semantics for that. Let the operating system to sort it out. | |
377 | ||
ec481373 JH |
378 | The I<portable filename characters> as defined by ANSI C are |
379 | ||
380 | 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 | |
381 | 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 | |
382 | 0 1 2 3 4 5 6 7 8 9 | |
383 | . _ - | |
384 | ||
08fef530 | 385 | and the "-" shouldn't be the first character. If you want to be |
e1516da7 JH |
386 | hypercorrect, stay case-insensitive and within the 8.3 naming |
387 | convention (all the files and directories have to be unique within one | |
388 | directory if their names are lowercased and truncated to eight | |
389 | characters before the C<.>, if any, and to three characters after the | |
390 | C<.>, if any). (And do not use C<.>s in directory names.) | |
ec481373 | 391 | |
e41182b5 GS |
392 | =head2 System Interaction |
393 | ||
b7df3edc GS |
394 | Not all platforms provide a command line. These are usually platforms |
395 | that rely primarily on a Graphical User Interface (GUI) for user | |
396 | interaction. A program requiring a command line interface might | |
397 | not work everywhere. This is probably for the user of the program | |
398 | to deal with, so don't stay up late worrying about it. | |
e41182b5 | 399 | |
c87488a3 HS |
400 | Some platforms can't delete or rename files held open by the system, |
401 | this limitation may also apply to changing filesystem metainformation | |
402 | like file permissions or owners. Remember to C<close> files when you | |
403 | are done with them. Don't C<unlink> or C<rename> an open file. Don't | |
404 | C<tie> or C<open> a file already tied or opened; C<untie> or C<close> | |
405 | it first. | |
e41182b5 | 406 | |
0a47030a GS |
407 | Don't open the same file more than once at a time for writing, as some |
408 | operating systems put mandatory locks on such files. | |
409 | ||
73e9292c JH |
410 | Don't assume that write/modify permission on a directory gives the |
411 | right to add or delete files/directories in that directory. That is | |
412 | filesystem specific: in some filesystems you need write/modify | |
413 | permission also (or even just) in the file/directory itself. In some | |
414 | filesystems (AFS, DFS) the permission to add/delete directory entries | |
415 | is a completely separate permission. | |
416 | ||
417 | Don't assume that a single C<unlink> completely gets rid of the file: | |
418 | some filesystems (most notably the ones in VMS) have versioned | |
419 | filesystems, and unlink() removes only the most recent one (it doesn't | |
420 | remove all the versions because by default the native tools on those | |
421 | platforms remove just the most recent version, too). The portable | |
422 | idiom to remove all the versions of a file is | |
423 | ||
94bb614c | 424 | 1 while unlink "file"; |
73e9292c JH |
425 | |
426 | This will terminate if the file is undeleteable for some reason | |
427 | (protected, not there, and so on). | |
428 | ||
e41182b5 | 429 | Don't count on a specific environment variable existing in C<%ENV>. |
0a47030a | 430 | Don't count on C<%ENV> entries being case-sensitive, or even |
4a0d0822 | 431 | case-preserving. Don't try to clear %ENV by saying C<%ENV = ();>, or, |
6be8f7a6 | 432 | if you really have to, make it conditional on C<$^O ne 'VMS'> since in |
4a0d0822 JH |
433 | VMS the C<%ENV> table is much more than a per-process key-value string |
434 | table. | |
e41182b5 | 435 | |
c73b03b7 JM |
436 | On VMS, some entries in the %ENV hash are dynamically created when |
437 | their key is used on a read if they did not previously exist. The | |
438 | values for C<$ENV{HOME}>, C<$ENV{TERM}>, C<$ENV{HOME}>, and C<$ENV{USER}>, | |
439 | are known to be dynamically generated. The specific names that are | |
440 | dynamically generated may vary with the version of the C library on VMS, | |
441 | and more may exist than is documented. | |
442 | ||
443 | On VMS by default, changes to the %ENV hash are persistent after the process | |
444 | exits. This can cause unintended issues. | |
445 | ||
d1e3b762 | 446 | Don't count on signals or C<%SIG> for anything. |
e41182b5 GS |
447 | |
448 | Don't count on filename globbing. Use C<opendir>, C<readdir>, and | |
449 | C<closedir> instead. | |
450 | ||
b8099c3d | 451 | Don't count on per-program environment variables, or per-program current |
dd9f0070 | 452 | directories. |
b8099c3d | 453 | |
c87488a3 HS |
454 | Don't count on specific values of C<$!>, neither numeric nor |
455 | especially the strings values-- users may switch their locales causing | |
456 | error messages to be translated into their languages. If you can | |
457 | trust a POSIXish environment, you can portably use the symbols defined | |
458 | by the Errno module, like ENOENT. And don't trust on the values of C<$!> | |
459 | at all except immediately after a failed system call. | |
3c075c7d | 460 | |
a10d74f3 PG |
461 | =head2 Command names versus file pathnames |
462 | ||
463 | Don't assume that the name used to invoke a command or program with | |
464 | C<system> or C<exec> can also be used to test for the existence of the | |
465 | file that holds the executable code for that command or program. | |
68fb0eb7 PG |
466 | First, many systems have "internal" commands that are built-in to the |
467 | shell or OS and while these commands can be invoked, there is no | |
468 | corresponding file. Second, some operating systems (e.g., Cygwin, | |
469 | DJGPP, OS/2, and VOS) have required suffixes for executable files; | |
470 | these suffixes are generally permitted on the command name but are not | |
a10d74f3 PG |
471 | required. Thus, a command like "perl" might exist in a file named |
472 | "perl", "perl.exe", or "perl.pm", depending on the operating system. | |
473 | The variable "_exe" in the Config module holds the executable suffix, | |
68fb0eb7 PG |
474 | if any. Third, the VMS port carefully sets up $^X and |
475 | $Config{perlpath} so that no further processing is required. This is | |
476 | just as well, because the matching regular expression used below would | |
477 | then have to deal with a possible trailing version number in the VMS | |
478 | file name. | |
a10d74f3 PG |
479 | |
480 | To convert $^X to a file pathname, taking account of the requirements | |
481 | of the various operating system possibilities, say: | |
7ee27b7c | 482 | |
a10d74f3 | 483 | use Config; |
a10d74f3 | 484 | $thisperl = $^X; |
68fb0eb7 PG |
485 | if ($^O ne 'VMS') |
486 | {$thisperl .= $Config{_exe} unless $thisperl =~ m/$Config{_exe}$/i;} | |
a10d74f3 PG |
487 | |
488 | To convert $Config{perlpath} to a file pathname, say: | |
7ee27b7c | 489 | |
a10d74f3 | 490 | use Config; |
68fb0eb7 PG |
491 | $thisperl = $Config{perlpath}; |
492 | if ($^O ne 'VMS') | |
493 | {$thisperl .= $Config{_exe} unless $thisperl =~ m/$Config{_exe}$/i;} | |
a10d74f3 | 494 | |
7137b697 JH |
495 | =head2 Networking |
496 | ||
497 | Don't assume that you can reach the public Internet. | |
498 | ||
499 | Don't assume that there is only one way to get through firewalls | |
500 | to the public Internet. | |
501 | ||
932f293e JH |
502 | Don't assume that you can reach outside world through any other port |
503 | than 80, or some web proxy. ftp is blocked by many firewalls. | |
504 | ||
dbc6a9ce JH |
505 | Don't assume that you can send email by connecting to the local SMTP port. |
506 | ||
7137b697 | 507 | Don't assume that you can reach yourself or any node by the name |
dbc6a9ce | 508 | 'localhost'. The same goes for '127.0.0.1'. You will have to try both. |
932f293e | 509 | |
86feb2c5 JH |
510 | Don't assume that the host has only one network card, or that it |
511 | can't bind to many virtual IP addresses. | |
932f293e JH |
512 | |
513 | Don't assume a particular network device name. | |
7137b697 | 514 | |
dbc6a9ce | 515 | Don't assume a particular set of ioctl()s will work. |
7137b697 JH |
516 | |
517 | Don't assume that you can ping hosts and get replies. | |
518 | ||
dbc6a9ce JH |
519 | Don't assume that any particular port (service) will respond. |
520 | ||
7ee27b7c | 521 | Don't assume that Sys::Hostname (or any other API or command) |
dbc6a9ce JH |
522 | returns either a fully qualified hostname or a non-qualified hostname: |
523 | it all depends on how the system had been configured. Also remember | |
524 | things like DHCP and NAT-- the hostname you get back might not be very | |
525 | useful. | |
526 | ||
932f293e JH |
527 | All the above "don't":s may look daunting, and they are -- but the key |
528 | is to degrade gracefully if one cannot reach the particular network | |
529 | service one wants. Croaking or hanging do not look very professional. | |
530 | ||
e41182b5 GS |
531 | =head2 Interprocess Communication (IPC) |
532 | ||
b7df3edc GS |
533 | In general, don't directly access the system in code meant to be |
534 | portable. That means, no C<system>, C<exec>, C<fork>, C<pipe>, | |
535 | C<``>, C<qx//>, C<open> with a C<|>, nor any of the other things | |
536 | that makes being a perl hacker worth being. | |
e41182b5 GS |
537 | |
538 | Commands that launch external processes are generally supported on | |
b7df3edc GS |
539 | most platforms (though many of them do not support any type of |
540 | forking). The problem with using them arises from what you invoke | |
541 | them on. External tools are often named differently on different | |
4375e838 | 542 | platforms, may not be available in the same location, might accept |
b7df3edc GS |
543 | different arguments, can behave differently, and often present their |
544 | results in a platform-dependent way. Thus, you should seldom depend | |
545 | on them to produce consistent results. (Then again, if you're calling | |
546 | I<netstat -a>, you probably don't expect it to run on both Unix and CP/M.) | |
e41182b5 | 547 | |
b7df3edc | 548 | One especially common bit of Perl code is opening a pipe to B<sendmail>: |
e41182b5 | 549 | |
b7df3edc GS |
550 | open(MAIL, '|/usr/lib/sendmail -t') |
551 | or die "cannot fork sendmail: $!"; | |
e41182b5 GS |
552 | |
553 | This is fine for systems programming when sendmail is known to be | |
554 | available. But it is not fine for many non-Unix systems, and even | |
555 | some Unix systems that may not have sendmail installed. If a portable | |
b7df3edc GS |
556 | solution is needed, see the various distributions on CPAN that deal |
557 | with it. Mail::Mailer and Mail::Send in the MailTools distribution are | |
558 | commonly used, and provide several mailing methods, including mail, | |
559 | sendmail, and direct SMTP (via Net::SMTP) if a mail transfer agent is | |
560 | not available. Mail::Sendmail is a standalone module that provides | |
561 | simple, platform-independent mailing. | |
562 | ||
563 | The Unix System V IPC (C<msg*(), sem*(), shm*()>) is not available | |
564 | even on all Unix platforms. | |
e41182b5 | 565 | |
a81e5e2e A |
566 | Do not use either the bare result of C<pack("N", 10, 20, 30, 40)> or |
567 | bare v-strings (such as C<v10.20.30.40>) to represent IPv4 addresses: | |
568 | both forms just pack the four bytes into network order. That this | |
569 | would be equal to the C language C<in_addr> struct (which is what the | |
570 | socket code internally uses) is not guaranteed. To be portable use | |
571 | the routines of the Socket extension, such as C<inet_aton()>, | |
572 | C<inet_ntoa()>, and C<sockaddr_in()>. | |
6b2463a0 | 573 | |
e41182b5 | 574 | The rule of thumb for portable code is: Do it all in portable Perl, or |
0a47030a GS |
575 | use a module (that may internally implement it with platform-specific |
576 | code, but expose a common interface). | |
e41182b5 | 577 | |
e41182b5 GS |
578 | =head2 External Subroutines (XS) |
579 | ||
b7df3edc | 580 | XS code can usually be made to work with any platform, but dependent |
e41182b5 GS |
581 | libraries, header files, etc., might not be readily available or |
582 | portable, or the XS code itself might be platform-specific, just as Perl | |
583 | code might be. If the libraries and headers are portable, then it is | |
584 | normally reasonable to make sure the XS code is portable, too. | |
585 | ||
b7df3edc GS |
586 | A different type of portability issue arises when writing XS code: |
587 | availability of a C compiler on the end-user's system. C brings | |
588 | with it its own portability issues, and writing XS code will expose | |
589 | you to some of those. Writing purely in Perl is an easier way to | |
e41182b5 GS |
590 | achieve portability. |
591 | ||
e41182b5 GS |
592 | =head2 Standard Modules |
593 | ||
594 | In general, the standard modules work across platforms. Notable | |
6ab3f9cb | 595 | exceptions are the CPAN module (which currently makes connections to external |
e41182b5 | 596 | programs that may not be available), platform-specific modules (like |
6ab3f9cb | 597 | ExtUtils::MM_VMS), and DBM modules. |
e41182b5 | 598 | |
b7df3edc | 599 | There is no one DBM module available on all platforms. |
6ab3f9cb GS |
600 | SDBM_File and the others are generally available on all Unix and DOSish |
601 | ports, but not in MacPerl, where only NBDM_File and DB_File are | |
0a47030a | 602 | available. |
e41182b5 GS |
603 | |
604 | The good news is that at least some DBM module should be available, and | |
6ab3f9cb | 605 | AnyDBM_File will use whichever module it can find. Of course, then |
b7df3edc GS |
606 | the code needs to be fairly strict, dropping to the greatest common |
607 | factor (e.g., not exceeding 1K for each record), so that it will | |
6ab3f9cb | 608 | work with any DBM module. See L<AnyDBM_File> for more details. |
e41182b5 | 609 | |
e41182b5 GS |
610 | =head2 Time and Date |
611 | ||
0a47030a | 612 | The system's notion of time of day and calendar date is controlled in |
b7df3edc | 613 | widely different ways. Don't assume the timezone is stored in C<$ENV{TZ}>, |
0a47030a | 614 | and even if it is, don't assume that you can control the timezone through |
c87488a3 HS |
615 | that variable. Don't assume anything about the three-letter timezone |
616 | abbreviations (for example that MST would be the Mountain Standard Time, | |
617 | it's been known to stand for Moscow Standard Time). If you need to | |
618 | use timezones, express them in some unambiguous format like the | |
619 | exact number of minutes offset from UTC, or the POSIX timezone | |
620 | format. | |
e41182b5 | 621 | |
322422de | 622 | Don't assume that the epoch starts at 00:00:00, January 1, 1970, |
c87488a3 HS |
623 | because that is OS- and implementation-specific. It is better to |
624 | store a date in an unambiguous representation. The ISO 8601 standard | |
625 | defines YYYY-MM-DD as the date format, or YYYY-MM-DDTHH-MM-SS | |
626 | (that's a literal "T" separating the date from the time). | |
627 | Please do use the ISO 8601 instead of making us to guess what | |
628 | date 02/03/04 might be. ISO 8601 even sorts nicely as-is. | |
629 | A text representation (like "1987-12-18") can be easily converted | |
630 | into an OS-specific value using a module like Date::Parse. | |
631 | An array of values, such as those returned by C<localtime>, can be | |
632 | converted to an OS-specific representation using Time::Local. | |
322422de | 633 | |
19799a22 GS |
634 | When calculating specific times, such as for tests in time or date modules, |
635 | it may be appropriate to calculate an offset for the epoch. | |
b7df3edc | 636 | |
19799a22 GS |
637 | require Time::Local; |
638 | $offset = Time::Local::timegm(0, 0, 0, 1, 0, 70); | |
b7df3edc | 639 | |
19799a22 GS |
640 | The value for C<$offset> in Unix will be C<0>, but in Mac OS will be |
641 | some large number. C<$offset> can then be added to a Unix time value | |
642 | to get what should be the proper value on any system. | |
322422de GS |
643 | |
644 | =head2 Character sets and character encoding | |
645 | ||
ec481373 JH |
646 | Assume very little about character sets. |
647 | ||
648 | Assume nothing about numerical values (C<ord>, C<chr>) of characters. | |
649 | Do not use explicit code point ranges (like \xHH-\xHH); use for | |
650 | example symbolic character classes like C<[:print:]>. | |
651 | ||
652 | Do not assume that the alphabetic characters are encoded contiguously | |
653 | (in the numeric sense). There may be gaps. | |
654 | ||
655 | Do not assume anything about the ordering of the characters. | |
656 | The lowercase letters may come before or after the uppercase letters; | |
b432a672 AL |
657 | the lowercase and uppercase may be interlaced so that both "a" and "A" |
658 | come before "b"; the accented and other international characters may | |
659 | be interlaced so that E<auml> comes before "b". | |
322422de GS |
660 | |
661 | =head2 Internationalisation | |
662 | ||
b7df3edc GS |
663 | If you may assume POSIX (a rather large assumption), you may read |
664 | more about the POSIX locale system from L<perllocale>. The locale | |
665 | system at least attempts to make things a little bit more portable, | |
666 | or at least more convenient and native-friendly for non-English | |
667 | users. The system affects character sets and encoding, and date | |
668 | and time formatting--amongst other things. | |
e41182b5 | 669 | |
c87488a3 HS |
670 | If you really want to be international, you should consider Unicode. |
671 | See L<perluniintro> and L<perlunicode> for more information. | |
672 | ||
11264fdb JH |
673 | If you want to use non-ASCII bytes (outside the bytes 0x00..0x7f) in |
674 | the "source code" of your code, to be portable you have to be explicit | |
675 | about what bytes they are. Someone might for example be using your | |
676 | code under a UTF-8 locale, in which case random native bytes might be | |
677 | illegal ("Malformed UTF-8 ...") This means that for example embedding | |
678 | ISO 8859-1 bytes beyond 0x7f into your strings might cause trouble | |
679 | later. If the bytes are native 8-bit bytes, you can use the C<bytes> | |
680 | pragma. If the bytes are in a string (regular expression being a | |
681 | curious string), you can often also use the C<\xHH> notation instead | |
2575c402 JW |
682 | of embedding the bytes as-is. (If you want to write your code in UTF-8, |
683 | you can use the C<utf8>.) The C<bytes> and C<utf8> pragmata are | |
684 | available since Perl 5.6.0. | |
11264fdb | 685 | |
e41182b5 GS |
686 | =head2 System Resources |
687 | ||
0a47030a GS |
688 | If your code is destined for systems with severely constrained (or |
689 | missing!) virtual memory systems then you want to be I<especially> mindful | |
690 | of avoiding wasteful constructs such as: | |
e41182b5 GS |
691 | |
692 | # NOTE: this is no longer "bad" in perl5.005 | |
693 | for (0..10000000) {} # bad | |
694 | for (my $x = 0; $x <= 10000000; ++$x) {} # good | |
695 | ||
696 | @lines = <VERY_LARGE_FILE>; # bad | |
697 | ||
698 | while (<FILE>) {$file .= $_} # sometimes bad | |
0a47030a | 699 | $file = join('', <FILE>); # better |
e41182b5 | 700 | |
b7df3edc GS |
701 | The last two constructs may appear unintuitive to most people. The |
702 | first repeatedly grows a string, whereas the second allocates a | |
703 | large chunk of memory in one go. On some systems, the second is | |
704 | more efficient that the first. | |
0a47030a | 705 | |
e41182b5 GS |
706 | =head2 Security |
707 | ||
b7df3edc GS |
708 | Most multi-user platforms provide basic levels of security, usually |
709 | implemented at the filesystem level. Some, however, do | |
a1667ba3 | 710 | not-- unfortunately. Thus the notion of user id, or "home" directory, |
b7df3edc GS |
711 | or even the state of being logged-in, may be unrecognizable on many |
712 | platforms. If you write programs that are security-conscious, it | |
713 | is usually best to know what type of system you will be running | |
714 | under so that you can write code explicitly for that platform (or | |
715 | class of platforms). | |
0a47030a | 716 | |
a1667ba3 JH |
717 | Don't assume the UNIX filesystem access semantics: the operating |
718 | system or the filesystem may be using some ACL systems, which are | |
719 | richer languages than the usual rwx. Even if the rwx exist, | |
720 | their semantics might be different. | |
721 | ||
722 | (From security viewpoint testing for permissions before attempting to | |
723 | do something is silly anyway: if one tries this, there is potential | |
724 | for race conditions-- someone or something might change the | |
725 | permissions between the permissions check and the actual operation. | |
726 | Just try the operation.) | |
727 | ||
728 | Don't assume the UNIX user and group semantics: especially, don't | |
2edcc0d9 | 729 | expect the C<< $< >> and C<< $> >> (or the C<$(> and C<$)>) to work |
a1667ba3 JH |
730 | for switching identities (or memberships). |
731 | ||
732 | Don't assume set-uid and set-gid semantics. (And even if you do, | |
733 | think twice: set-uid and set-gid are a known can of security worms.) | |
734 | ||
e41182b5 GS |
735 | =head2 Style |
736 | ||
737 | For those times when it is necessary to have platform-specific code, | |
738 | consider keeping the platform-specific code in one place, making porting | |
6ab3f9cb | 739 | to other platforms easier. Use the Config module and the special |
0a47030a GS |
740 | variable C<$^O> to differentiate platforms, as described in |
741 | L<"PLATFORMS">. | |
e41182b5 | 742 | |
b7df3edc GS |
743 | Be careful in the tests you supply with your module or programs. |
744 | Module code may be fully portable, but its tests might not be. This | |
745 | often happens when tests spawn off other processes or call external | |
746 | programs to aid in the testing, or when (as noted above) the tests | |
c87488a3 HS |
747 | assume certain things about the filesystem and paths. Be careful not |
748 | to depend on a specific output style for errors, such as when checking | |
749 | C<$!> after a failed system call. Using C<$!> for anything else than | |
750 | displaying it as output is doubtful (though see the Errno module for | |
751 | testing reasonably portably for error value). Some platforms expect | |
752 | a certain output format, and Perl on those platforms may have been | |
753 | adjusted accordingly. Most specifically, don't anchor a regex when | |
754 | testing an error value. | |
e41182b5 | 755 | |
0a47030a | 756 | =head1 CPAN Testers |
e41182b5 | 757 | |
0a47030a GS |
758 | Modules uploaded to CPAN are tested by a variety of volunteers on |
759 | different platforms. These CPAN testers are notified by mail of each | |
e41182b5 | 760 | new upload, and reply to the list with PASS, FAIL, NA (not applicable to |
0a47030a | 761 | this platform), or UNKNOWN (unknown), along with any relevant notations. |
e41182b5 GS |
762 | |
763 | The purpose of the testing is twofold: one, to help developers fix any | |
0a47030a | 764 | problems in their code that crop up because of lack of testing on other |
b7df3edc | 765 | platforms; two, to provide users with information about whether |
0a47030a | 766 | a given module works on a given platform. |
e41182b5 | 767 | |
7ee27b7c AT |
768 | Also see: |
769 | ||
e41182b5 GS |
770 | =over 4 |
771 | ||
7ee27b7c AT |
772 | =item * |
773 | ||
774 | Mailing list: cpan-testers@perl.org | |
775 | ||
776 | =item * | |
e41182b5 | 777 | |
7ee27b7c | 778 | Testing results: http://testers.cpan.org/ |
e41182b5 GS |
779 | |
780 | =back | |
781 | ||
e41182b5 GS |
782 | =head1 PLATFORMS |
783 | ||
784 | As of version 5.002, Perl is built with a C<$^O> variable that | |
785 | indicates the operating system it was built on. This was implemented | |
b7df3edc GS |
786 | to help speed up code that would otherwise have to C<use Config> |
787 | and use the value of C<$Config{osname}>. Of course, to get more | |
e41182b5 GS |
788 | detailed information about the system, looking into C<%Config> is |
789 | certainly recommended. | |
790 | ||
b7df3edc GS |
791 | C<%Config> cannot always be trusted, however, because it was built |
792 | at compile time. If perl was built in one place, then transferred | |
793 | elsewhere, some values may be wrong. The values may even have been | |
794 | edited after the fact. | |
6ab3f9cb | 795 | |
e41182b5 GS |
796 | =head2 Unix |
797 | ||
798 | Perl works on a bewildering variety of Unix and Unix-like platforms (see | |
799 | e.g. most of the files in the F<hints/> directory in the source code kit). | |
800 | On most of these systems, the value of C<$^O> (hence C<$Config{'osname'}>, | |
d1e3b762 GS |
801 | too) is determined either by lowercasing and stripping punctuation from the |
802 | first field of the string returned by typing C<uname -a> (or a similar command) | |
803 | at the shell prompt or by testing the file system for the presence of | |
804 | uniquely named files such as a kernel or header file. Here, for example, | |
805 | are a few of the more popular Unix flavors: | |
e41182b5 | 806 | |
b7df3edc | 807 | uname $^O $Config{'archname'} |
6ab3f9cb | 808 | -------------------------------------------- |
b7df3edc | 809 | AIX aix aix |
6ab3f9cb | 810 | BSD/OS bsdos i386-bsdos |
e1516da7 | 811 | Darwin darwin darwin |
6ab3f9cb GS |
812 | dgux dgux AViiON-dgux |
813 | DYNIX/ptx dynixptx i386-dynixptx | |
b7df3edc | 814 | FreeBSD freebsd freebsd-i386 |
df00ff3b | 815 | Haiku haiku BePC-haiku |
d1e3b762 | 816 | Linux linux arm-linux |
b7df3edc | 817 | Linux linux i386-linux |
6ab3f9cb GS |
818 | Linux linux i586-linux |
819 | Linux linux ppc-linux | |
b7df3edc GS |
820 | HP-UX hpux PA-RISC1.1 |
821 | IRIX irix irix | |
b787fad4 | 822 | Mac OS X darwin darwin |
d1e3b762 GS |
823 | MachTen PPC machten powerpc-machten |
824 | NeXT 3 next next-fat | |
825 | NeXT 4 next OPENSTEP-Mach | |
6ab3f9cb | 826 | openbsd openbsd i386-openbsd |
b7df3edc | 827 | OSF1 dec_osf alpha-dec_osf |
6ab3f9cb GS |
828 | reliantunix-n svr4 RM400-svr4 |
829 | SCO_SV sco_sv i386-sco_sv | |
830 | SINIX-N svr4 RM400-svr4 | |
831 | sn4609 unicos CRAY_C90-unicos | |
832 | sn6521 unicosmk t3e-unicosmk | |
833 | sn9617 unicos CRAY_J90-unicos | |
b7df3edc GS |
834 | SunOS solaris sun4-solaris |
835 | SunOS solaris i86pc-solaris | |
836 | SunOS4 sunos sun4-sunos | |
e41182b5 | 837 | |
b7df3edc GS |
838 | Because the value of C<$Config{archname}> may depend on the |
839 | hardware architecture, it can vary more than the value of C<$^O>. | |
6ab3f9cb | 840 | |
e41182b5 GS |
841 | =head2 DOS and Derivatives |
842 | ||
b7df3edc | 843 | Perl has long been ported to Intel-style microcomputers running under |
e41182b5 GS |
844 | systems like PC-DOS, MS-DOS, OS/2, and most Windows platforms you can |
845 | bring yourself to mention (except for Windows CE, if you count that). | |
b7df3edc | 846 | Users familiar with I<COMMAND.COM> or I<CMD.EXE> style shells should |
e41182b5 GS |
847 | be aware that each of these file specifications may have subtle |
848 | differences: | |
849 | ||
850 | $filespec0 = "c:/foo/bar/file.txt"; | |
851 | $filespec1 = "c:\\foo\\bar\\file.txt"; | |
852 | $filespec2 = 'c:\foo\bar\file.txt'; | |
853 | $filespec3 = 'c:\\foo\\bar\\file.txt'; | |
854 | ||
b7df3edc GS |
855 | System calls accept either C</> or C<\> as the path separator. |
856 | However, many command-line utilities of DOS vintage treat C</> as | |
857 | the option prefix, so may get confused by filenames containing C</>. | |
858 | Aside from calling any external programs, C</> will work just fine, | |
859 | and probably better, as it is more consistent with popular usage, | |
860 | and avoids the problem of remembering what to backwhack and what | |
861 | not to. | |
e41182b5 | 862 | |
b7df3edc GS |
863 | The DOS FAT filesystem can accommodate only "8.3" style filenames. Under |
864 | the "case-insensitive, but case-preserving" HPFS (OS/2) and NTFS (NT) | |
0a47030a | 865 | filesystems you may have to be careful about case returned with functions |
e41182b5 GS |
866 | like C<readdir> or used with functions like C<open> or C<opendir>. |
867 | ||
b7df3edc GS |
868 | DOS also treats several filenames as special, such as AUX, PRN, |
869 | NUL, CON, COM1, LPT1, LPT2, etc. Unfortunately, sometimes these | |
870 | filenames won't even work if you include an explicit directory | |
871 | prefix. It is best to avoid such filenames, if you want your code | |
872 | to be portable to DOS and its derivatives. It's hard to know what | |
873 | these all are, unfortunately. | |
e41182b5 GS |
874 | |
875 | Users of these operating systems may also wish to make use of | |
b7df3edc | 876 | scripts such as I<pl2bat.bat> or I<pl2cmd> to |
e41182b5 GS |
877 | put wrappers around your scripts. |
878 | ||
879 | Newline (C<\n>) is translated as C<\015\012> by STDIO when reading from | |
6ab3f9cb GS |
880 | and writing to files (see L<"Newlines">). C<binmode(FILEHANDLE)> |
881 | will keep C<\n> translated as C<\012> for that filehandle. Since it is a | |
882 | no-op on other systems, C<binmode> should be used for cross-platform code | |
b7df3edc GS |
883 | that deals with binary data. That's assuming you realize in advance |
884 | that your data is in binary. General-purpose programs should | |
885 | often assume nothing about their data. | |
e41182b5 | 886 | |
b7df3edc | 887 | The C<$^O> variable and the C<$Config{archname}> values for various |
e41182b5 GS |
888 | DOSish perls are as follows: |
889 | ||
67ac489e MS |
890 | OS $^O $Config{archname} ID Version |
891 | -------------------------------------------------------- | |
892 | MS-DOS dos ? | |
893 | PC-DOS dos ? | |
894 | OS/2 os2 ? | |
895 | Windows 3.1 ? ? 0 3 01 | |
896 | Windows 95 MSWin32 MSWin32-x86 1 4 00 | |
897 | Windows 98 MSWin32 MSWin32-x86 1 4 10 | |
898 | Windows ME MSWin32 MSWin32-x86 1 ? | |
899 | Windows NT MSWin32 MSWin32-x86 2 4 xx | |
900 | Windows NT MSWin32 MSWin32-ALPHA 2 4 xx | |
901 | Windows NT MSWin32 MSWin32-ppc 2 4 xx | |
7ee27b7c AT |
902 | Windows 2000 MSWin32 MSWin32-x86 2 5 00 |
903 | Windows XP MSWin32 MSWin32-x86 2 5 01 | |
904 | Windows 2003 MSWin32 MSWin32-x86 2 5 02 | |
67ac489e | 905 | Windows CE MSWin32 ? 3 |
7ee27b7c | 906 | Cygwin cygwin cygwin |
e41182b5 | 907 | |
34aaaa84 PP |
908 | The various MSWin32 Perl's can distinguish the OS they are running on |
909 | via the value of the fifth element of the list returned from | |
910 | Win32::GetOSVersion(). For example: | |
911 | ||
912 | if ($^O eq 'MSWin32') { | |
913 | my @os_version_info = Win32::GetOSVersion(); | |
914 | print +('3.1','95','NT')[$os_version_info[4]],"\n"; | |
915 | } | |
916 | ||
7939d86b JH |
917 | There are also Win32::IsWinNT() and Win32::IsWin95(), try C<perldoc Win32>, |
918 | and as of libwin32 0.19 (not part of the core Perl distribution) | |
919 | Win32::GetOSName(). The very portable POSIX::uname() will work too: | |
1d65be3a JH |
920 | |
921 | c:\> perl -MPOSIX -we "print join '|', uname" | |
922 | Windows NT|moonru|5.0|Build 2195 (Service Pack 2)|x86 | |
d99f392e | 923 | |
e41182b5 GS |
924 | Also see: |
925 | ||
926 | =over 4 | |
927 | ||
c997b287 | 928 | =item * |
e41182b5 | 929 | |
c997b287 GS |
930 | The djgpp environment for DOS, http://www.delorie.com/djgpp/ |
931 | and L<perldos>. | |
e41182b5 | 932 | |
c997b287 | 933 | =item * |
e41182b5 | 934 | |
c997b287 GS |
935 | The EMX environment for DOS, OS/2, etc. emx@iaehv.nl, |
936 | http://www.leo.org/pub/comp/os/os2/leo/gnu/emx+gcc/index.html or | |
f224927c | 937 | ftp://hobbes.nmsu.edu/pub/os2/dev/emx/ Also L<perlos2>. |
e41182b5 | 938 | |
c997b287 | 939 | =item * |
d1e3b762 | 940 | |
c997b287 GS |
941 | Build instructions for Win32 in L<perlwin32>, or under the Cygnus environment |
942 | in L<perlcygwin>. | |
943 | ||
944 | =item * | |
945 | ||
946 | The C<Win32::*> modules in L<Win32>. | |
947 | ||
948 | =item * | |
949 | ||
950 | The ActiveState Pages, http://www.activestate.com/ | |
951 | ||
952 | =item * | |
953 | ||
954 | The Cygwin environment for Win32; F<README.cygwin> (installed | |
47dafe4d | 955 | as L<perlcygwin>), http://www.cygwin.com/ |
c997b287 GS |
956 | |
957 | =item * | |
958 | ||
959 | The U/WIN environment for Win32, | |
cea6626f | 960 | http://www.research.att.com/sw/tools/uwin/ |
c997b287 | 961 | |
cea6626f | 962 | =item * |
d1e3b762 | 963 | |
cea6626f | 964 | Build instructions for OS/2, L<perlos2> |
d1e3b762 | 965 | |
e41182b5 GS |
966 | =back |
967 | ||
dd9f0070 | 968 | =head2 S<Mac OS> |
e41182b5 GS |
969 | |
970 | Any module requiring XS compilation is right out for most people, because | |
971 | MacPerl is built using non-free (and non-cheap!) compilers. Some XS | |
972 | modules that can work with MacPerl are built and distributed in binary | |
6ab3f9cb | 973 | form on CPAN. |
e41182b5 GS |
974 | |
975 | Directories are specified as: | |
976 | ||
977 | volume:folder:file for absolute pathnames | |
978 | volume:folder: for absolute pathnames | |
979 | :folder:file for relative pathnames | |
980 | :folder: for relative pathnames | |
981 | :file for relative pathnames | |
982 | file for relative pathnames | |
983 | ||
b7df3edc | 984 | Files are stored in the directory in alphabetical order. Filenames are |
6ab3f9cb | 985 | limited to 31 characters, and may include any character except for |
b7df3edc | 986 | null and C<:>, which is reserved as the path separator. |
e41182b5 | 987 | |
0a47030a | 988 | Instead of C<flock>, see C<FSpSetFLock> and C<FSpRstFLock> in the |
6ab3f9cb | 989 | Mac::Files module, or C<chmod(0444, ...)> and C<chmod(0666, ...)>. |
e41182b5 GS |
990 | |
991 | In the MacPerl application, you can't run a program from the command line; | |
992 | programs that expect C<@ARGV> to be populated can be edited with something | |
993 | like the following, which brings up a dialog box asking for the command | |
994 | line arguments. | |
995 | ||
996 | if (!@ARGV) { | |
997 | @ARGV = split /\s+/, MacPerl::Ask('Arguments?'); | |
998 | } | |
999 | ||
b7df3edc | 1000 | A MacPerl script saved as a "droplet" will populate C<@ARGV> with the full |
e41182b5 GS |
1001 | pathnames of the files dropped onto the script. |
1002 | ||
b7df3edc GS |
1003 | Mac users can run programs under a type of command line interface |
1004 | under MPW (Macintosh Programmer's Workshop, a free development | |
1005 | environment from Apple). MacPerl was first introduced as an MPW | |
1006 | tool, and MPW can be used like a shell: | |
e41182b5 GS |
1007 | |
1008 | perl myscript.plx some arguments | |
1009 | ||
1010 | ToolServer is another app from Apple that provides access to MPW tools | |
0a47030a | 1011 | from MPW and the MacPerl app, which allows MacPerl programs to use |
e41182b5 GS |
1012 | C<system>, backticks, and piped C<open>. |
1013 | ||
1014 | "S<Mac OS>" is the proper name for the operating system, but the value | |
1015 | in C<$^O> is "MacOS". To determine architecture, version, or whether | |
1016 | the application or MPW tool version is running, check: | |
1017 | ||
1018 | $is_app = $MacPerl::Version =~ /App/; | |
1019 | $is_tool = $MacPerl::Version =~ /MPW/; | |
1020 | ($version) = $MacPerl::Version =~ /^(\S+)/; | |
1021 | $is_ppc = $MacPerl::Architecture eq 'MacPPC'; | |
1022 | $is_68k = $MacPerl::Architecture eq 'Mac68K'; | |
1023 | ||
b787fad4 JH |
1024 | S<Mac OS X>, based on NeXT's OpenStep OS, runs MacPerl natively, under the |
1025 | "Classic" environment. There is no "Carbon" version of MacPerl to run | |
1026 | under the primary Mac OS X environment. S<Mac OS X> and its Open Source | |
1027 | version, Darwin, both run Unix perl natively. | |
6ab3f9cb | 1028 | |
e41182b5 GS |
1029 | Also see: |
1030 | ||
1031 | =over 4 | |
1032 | ||
c997b287 GS |
1033 | =item * |
1034 | ||
862b5365 | 1035 | MacPerl Development, http://dev.macperl.org/ . |
c997b287 GS |
1036 | |
1037 | =item * | |
1038 | ||
862b5365 | 1039 | The MacPerl Pages, http://www.macperl.com/ . |
e41182b5 | 1040 | |
c997b287 | 1041 | =item * |
6ab3f9cb | 1042 | |
862b5365 | 1043 | The MacPerl mailing lists, http://lists.perl.org/ . |
e41182b5 | 1044 | |
7ee27b7c AT |
1045 | =item * |
1046 | ||
1047 | MPW, ftp://ftp.apple.com/developer/Tool_Chest/Core_Mac_OS_Tools/ | |
1048 | ||
e41182b5 GS |
1049 | =back |
1050 | ||
e41182b5 GS |
1051 | =head2 VMS |
1052 | ||
c997b287 | 1053 | Perl on VMS is discussed in L<perlvms> in the perl distribution. |
016930a6 JM |
1054 | |
1055 | The official name of VMS as of this writing is OpenVMS. | |
1056 | ||
b7df3edc | 1057 | Perl on VMS can accept either VMS- or Unix-style file |
e41182b5 GS |
1058 | specifications as in either of the following: |
1059 | ||
1060 | $ perl -ne "print if /perl_setup/i" SYS$LOGIN:LOGIN.COM | |
1061 | $ perl -ne "print if /perl_setup/i" /sys$login/login.com | |
1062 | ||
1063 | but not a mixture of both as in: | |
1064 | ||
1065 | $ perl -ne "print if /perl_setup/i" sys$login:/login.com | |
1066 | Can't open sys$login:/login.com: file specification syntax error | |
1067 | ||
1068 | Interacting with Perl from the Digital Command Language (DCL) shell | |
1069 | often requires a different set of quotation marks than Unix shells do. | |
1070 | For example: | |
1071 | ||
1072 | $ perl -e "print ""Hello, world.\n""" | |
1073 | Hello, world. | |
1074 | ||
b7df3edc | 1075 | There are several ways to wrap your perl scripts in DCL F<.COM> files, if |
e41182b5 GS |
1076 | you are so inclined. For example: |
1077 | ||
1078 | $ write sys$output "Hello from DCL!" | |
1079 | $ if p1 .eqs. "" | |
1080 | $ then perl -x 'f$environment("PROCEDURE") | |
1081 | $ else perl -x - 'p1 'p2 'p3 'p4 'p5 'p6 'p7 'p8 | |
1082 | $ deck/dollars="__END__" | |
1083 | #!/usr/bin/perl | |
1084 | ||
1085 | print "Hello from Perl!\n"; | |
1086 | ||
1087 | __END__ | |
1088 | $ endif | |
1089 | ||
1090 | Do take care with C<$ ASSIGN/nolog/user SYS$COMMAND: SYS$INPUT> if your | |
c47ff5f1 | 1091 | perl-in-DCL script expects to do things like C<< $read = <STDIN>; >>. |
e41182b5 | 1092 | |
016930a6 JM |
1093 | The VMS operating system has two filesystems, known as ODS-2 and ODS-5. |
1094 | ||
1095 | For ODS-2, filenames are in the format "name.extension;version". The | |
1096 | maximum length for filenames is 39 characters, and the maximum length for | |
e41182b5 GS |
1097 | extensions is also 39 characters. Version is a number from 1 to |
1098 | 32767. Valid characters are C</[A-Z0-9$_-]/>. | |
1099 | ||
016930a6 JM |
1100 | The ODS-2 filesystem is case-insensitive and does not preserve case. |
1101 | Perl simulates this by converting all filenames to lowercase internally. | |
1102 | ||
1103 | For ODS-5, filenames may have almost any character in them and can include | |
1104 | Unicode characters. Characters that could be misinterpreted by the DCL | |
1105 | shell or file parsing utilities need to be prefixed with the C<^> | |
1106 | character, or replaced with hexadecimal characters prefixed with the | |
1107 | C<^> character. Such prefixing is only needed with the pathnames are | |
1108 | in VMS format in applications. Programs that can accept the UNIX format | |
1109 | of pathnames do not need the escape characters. The maximum length for | |
1110 | filenames is 255 characters. The ODS-5 file system can handle both | |
1111 | a case preserved and a case sensitive mode. | |
1112 | ||
1113 | ODS-5 is only available on the OpenVMS for 64 bit platforms. | |
1114 | ||
1115 | Support for the extended file specifications is being done as optional | |
1116 | settings to preserve backward compatibility with Perl scripts that | |
1117 | assume the previous VMS limitations. | |
1118 | ||
1119 | In general routines on VMS that get a UNIX format file specification | |
1120 | should return it in a UNIX format, and when they get a VMS format | |
1121 | specification they should return a VMS format unless they are documented | |
1122 | to do a conversion. | |
1123 | ||
1124 | For routines that generate return a file specification, VMS allows setting | |
1125 | if the C library which Perl is built on if it will be returned in VMS | |
1126 | format or in UNIX format. | |
1127 | ||
1128 | With the ODS-2 file system, there is not much difference in syntax of | |
1129 | filenames without paths for VMS or UNIX. With the extended character | |
1130 | set available with ODS-5 there can be a significant difference. | |
1131 | ||
1132 | Because of this, existing Perl scripts written for VMS were sometimes | |
1133 | treating VMS and UNIX filenames interchangeably. Without the extended | |
1134 | character set enabled, this behavior will mostly be maintained for | |
1135 | backwards compatibility. | |
1136 | ||
1137 | When extended characters are enabled with ODS-5, the handling of | |
1138 | UNIX formatted file specifications is to that of a UNIX system. | |
1139 | ||
1140 | VMS file specifications without extensions have a trailing dot. An | |
1141 | equivalent UNIX file specification should not show the trailing dot. | |
1142 | ||
1143 | The result of all of this, is that for VMS, for portable scripts, you | |
1144 | can not depend on Perl to present the filenames in lowercase, to be | |
1145 | case sensitive, and that the filenames could be returned in either | |
1146 | UNIX or VMS format. | |
1147 | ||
1148 | And if a routine returns a file specification, unless it is intended to | |
1149 | convert it, it should return it in the same format as it found it. | |
1150 | ||
1151 | C<readdir> by default has traditionally returned lowercased filenames. | |
1152 | When the ODS-5 support is enabled, it will return the exact case of the | |
1153 | filename on the disk. | |
1154 | ||
1155 | Files without extensions have a trailing period on them, so doing a | |
1156 | C<readdir> in the default mode with a file named F<A.;5> will | |
1157 | return F<a.> when VMS is (though that file could be opened with | |
0a47030a | 1158 | C<open(FH, 'A')>). |
e41182b5 | 1159 | |
016930a6 JM |
1160 | With support for extended file specifications and if C<opendir> was |
1161 | given a UNIX format directory, a file named F<A.;5> will return F<a> | |
1162 | and optionally in the exact case on the disk. When C<opendir> is given | |
1163 | a VMS format directory, then C<readdir> should return F<a.>, and | |
1164 | again with the optionally the exact case. | |
1165 | ||
f34d0673 | 1166 | RMS had an eight level limit on directory depths from any rooted logical |
1089a9e3 CB |
1167 | (allowing 16 levels overall) prior to VMS 7.2, and even with versions of |
1168 | VMS on VAX up through 7.3. Hence C<PERL_ROOT:[LIB.2.3.4.5.6.7.8]> is a | |
1169 | valid directory specification but C<PERL_ROOT:[LIB.2.3.4.5.6.7.8.9]> is | |
1170 | not. F<Makefile.PL> authors might have to take this into account, but at | |
1171 | least they can refer to the former as C</PERL_ROOT/lib/2/3/4/5/6/7/8/>. | |
1172 | ||
1173 | Pumpkings and module integrators can easily see whether files with too many | |
1174 | directory levels have snuck into the core by running the following in the | |
1175 | top-level source directory: | |
1176 | ||
1177 | $ perl -ne "$_=~s/\s+.*//; print if scalar(split /\//) > 8;" < MANIFEST | |
1178 | ||
e41182b5 | 1179 | |
6ab3f9cb | 1180 | The VMS::Filespec module, which gets installed as part of the build |
0a47030a GS |
1181 | process on VMS, is a pure Perl module that can easily be installed on |
1182 | non-VMS platforms and can be helpful for conversions to and from RMS | |
016930a6 JM |
1183 | native formats. It is also now the only way that you should check to |
1184 | see if VMS is in a case sensitive mode. | |
e41182b5 | 1185 | |
5e12dbfa PP |
1186 | What C<\n> represents depends on the type of file opened. It usually |
1187 | represents C<\012> but it could also be C<\015>, C<\012>, C<\015\012>, | |
fa11829f | 1188 | C<\000>, C<\040>, or nothing depending on the file organization and |
5e12dbfa PP |
1189 | record format. The VMS::Stdio module provides access to the |
1190 | special fopen() requirements of files with unusual attributes on VMS. | |
e41182b5 GS |
1191 | |
1192 | TCP/IP stacks are optional on VMS, so socket routines might not be | |
1193 | implemented. UDP sockets may not be supported. | |
1194 | ||
016930a6 JM |
1195 | The TCP/IP library support for all current versions of VMS is dynamically |
1196 | loaded if present, so even if the routines are configured, they may | |
1197 | return a status indicating that they are not implemented. | |
1198 | ||
e41182b5 GS |
1199 | The value of C<$^O> on OpenVMS is "VMS". To determine the architecture |
1200 | that you are running on without resorting to loading all of C<%Config> | |
1201 | you can examine the content of the C<@INC> array like so: | |
1202 | ||
1203 | if (grep(/VMS_AXP/, @INC)) { | |
1204 | print "I'm on Alpha!\n"; | |
6ab3f9cb | 1205 | |
e41182b5 GS |
1206 | } elsif (grep(/VMS_VAX/, @INC)) { |
1207 | print "I'm on VAX!\n"; | |
6ab3f9cb | 1208 | |
016930a6 JM |
1209 | } elsif (grep(/VMS_IA64/, @INC)) { |
1210 | print "I'm on IA64!\n"; | |
1211 | ||
e41182b5 GS |
1212 | } else { |
1213 | print "I'm not so sure about where $^O is...\n"; | |
1214 | } | |
1215 | ||
016930a6 JM |
1216 | In general, the significant differences should only be if Perl is running |
1217 | on VMS_VAX or one of the 64 bit OpenVMS platforms. | |
1218 | ||
b7df3edc GS |
1219 | On VMS, perl determines the UTC offset from the C<SYS$TIMEZONE_DIFFERENTIAL> |
1220 | logical name. Although the VMS epoch began at 17-NOV-1858 00:00:00.00, | |
6ab3f9cb | 1221 | calls to C<localtime> are adjusted to count offsets from |
b7df3edc | 1222 | 01-JAN-1970 00:00:00.00, just like Unix. |
6ab3f9cb | 1223 | |
e41182b5 GS |
1224 | Also see: |
1225 | ||
1226 | =over 4 | |
1227 | ||
c997b287 GS |
1228 | =item * |
1229 | ||
1230 | F<README.vms> (installed as L<README_vms>), L<perlvms> | |
1231 | ||
1232 | =item * | |
1233 | ||
1089a9e3 | 1234 | vmsperl list, vmsperl-subscribe@perl.org |
e41182b5 | 1235 | |
c997b287 | 1236 | =item * |
e41182b5 | 1237 | |
c997b287 | 1238 | vmsperl on the web, http://www.sidhe.org/vmsperl/index.html |
e41182b5 GS |
1239 | |
1240 | =back | |
1241 | ||
495c5fdc PG |
1242 | =head2 VOS |
1243 | ||
9a997319 JH |
1244 | Perl on VOS is discussed in F<README.vos> in the perl distribution |
1245 | (installed as L<perlvos>). Perl on VOS can accept either VOS- or | |
1246 | Unix-style file specifications as in either of the following: | |
495c5fdc | 1247 | |
ea8b8ad2 VP |
1248 | $ perl -ne "print if /perl_setup/i" >system>notices |
1249 | $ perl -ne "print if /perl_setup/i" /system/notices | |
495c5fdc PG |
1250 | |
1251 | or even a mixture of both as in: | |
1252 | ||
ea8b8ad2 | 1253 | $ perl -ne "print if /perl_setup/i" >system/notices |
495c5fdc | 1254 | |
b7df3edc | 1255 | Even though VOS allows the slash character to appear in object |
495c5fdc PG |
1256 | names, because the VOS port of Perl interprets it as a pathname |
1257 | delimiting character, VOS files, directories, or links whose names | |
1258 | contain a slash character cannot be processed. Such files must be | |
a3dfe201 | 1259 | renamed before they can be processed by Perl. Note that VOS limits |
b449fc5b NC |
1260 | file names to 32 or fewer characters, file names cannot start with a |
1261 | C<-> character, or contain any character matching C<< tr/ !%&'()*+;<>?// >> | |
495c5fdc | 1262 | |
495c5fdc PG |
1263 | The value of C<$^O> on VOS is "VOS". To determine the architecture that |
1264 | you are running on without resorting to loading all of C<%Config> you | |
c997b287 | 1265 | can examine the content of the @INC array like so: |
495c5fdc | 1266 | |
24e8e380 | 1267 | if ($^O =~ /VOS/) { |
495c5fdc PG |
1268 | print "I'm on a Stratus box!\n"; |
1269 | } else { | |
1270 | print "I'm not on a Stratus box!\n"; | |
1271 | die; | |
1272 | } | |
1273 | ||
495c5fdc PG |
1274 | Also see: |
1275 | ||
1276 | =over 4 | |
1277 | ||
c997b287 | 1278 | =item * |
495c5fdc | 1279 | |
cc07ed0b | 1280 | F<README.vos> (installed as L<perlvos>) |
c997b287 GS |
1281 | |
1282 | =item * | |
1283 | ||
1284 | The VOS mailing list. | |
495c5fdc PG |
1285 | |
1286 | There is no specific mailing list for Perl on VOS. You can post | |
1287 | comments to the comp.sys.stratus newsgroup, or subscribe to the general | |
cc07ed0b | 1288 | Stratus mailing list. Send a letter with "subscribe Info-Stratus" in |
495c5fdc PG |
1289 | the message body to majordomo@list.stratagy.com. |
1290 | ||
c997b287 GS |
1291 | =item * |
1292 | ||
cc07ed0b | 1293 | VOS Perl on the web at http://ftp.stratus.com/pub/vos/posix/posix.html |
495c5fdc PG |
1294 | |
1295 | =back | |
1296 | ||
e41182b5 GS |
1297 | =head2 EBCDIC Platforms |
1298 | ||
1299 | Recent versions of Perl have been ported to platforms such as OS/400 on | |
d1e3b762 GS |
1300 | AS/400 minicomputers as well as OS/390, VM/ESA, and BS2000 for S/390 |
1301 | Mainframes. Such computers use EBCDIC character sets internally (usually | |
0cc436d0 GS |
1302 | Character Code Set ID 0037 for OS/400 and either 1047 or POSIX-BC for S/390 |
1303 | systems). On the mainframe perl currently works under the "Unix system | |
1304 | services for OS/390" (formerly known as OpenEdition), VM/ESA OpenEdition, or | |
1305 | the BS200 POSIX-BC system (BS2000 is supported in perl 5.6 and greater). | |
522b859a JH |
1306 | See L<perlos390> for details. Note that for OS/400 there is also a port of |
1307 | Perl 5.8.1/5.9.0 or later to the PASE which is ASCII-based (as opposed to | |
1308 | ILE which is EBCDIC-based), see L<perlos400>. | |
e41182b5 | 1309 | |
7c5ffed3 JH |
1310 | As of R2.5 of USS for OS/390 and Version 2.3 of VM/ESA these Unix |
1311 | sub-systems do not support the C<#!> shebang trick for script invocation. | |
1312 | Hence, on OS/390 and VM/ESA perl scripts can be executed with a header | |
1313 | similar to the following simple script: | |
e41182b5 GS |
1314 | |
1315 | : # use perl | |
1316 | eval 'exec /usr/local/bin/perl -S $0 ${1+"$@"}' | |
1317 | if 0; | |
1318 | #!/usr/local/bin/perl # just a comment really | |
1319 | ||
1320 | print "Hello from perl!\n"; | |
1321 | ||
d1e3b762 GS |
1322 | OS/390 will support the C<#!> shebang trick in release 2.8 and beyond. |
1323 | Calls to C<system> and backticks can use POSIX shell syntax on all | |
1324 | S/390 systems. | |
1325 | ||
b7df3edc | 1326 | On the AS/400, if PERL5 is in your library list, you may need |
6ab3f9cb GS |
1327 | to wrap your perl scripts in a CL procedure to invoke them like so: |
1328 | ||
1329 | BEGIN | |
1330 | CALL PGM(PERL5/PERL) PARM('/QOpenSys/hello.pl') | |
1331 | ENDPGM | |
1332 | ||
1333 | This will invoke the perl script F<hello.pl> in the root of the | |
1334 | QOpenSys file system. On the AS/400 calls to C<system> or backticks | |
1335 | must use CL syntax. | |
1336 | ||
e41182b5 | 1337 | On these platforms, bear in mind that the EBCDIC character set may have |
0a47030a GS |
1338 | an effect on what happens with some perl functions (such as C<chr>, |
1339 | C<pack>, C<print>, C<printf>, C<ord>, C<sort>, C<sprintf>, C<unpack>), as | |
1340 | well as bit-fiddling with ASCII constants using operators like C<^>, C<&> | |
1341 | and C<|>, not to mention dealing with socket interfaces to ASCII computers | |
6ab3f9cb | 1342 | (see L<"Newlines">). |
e41182b5 | 1343 | |
b7df3edc GS |
1344 | Fortunately, most web servers for the mainframe will correctly |
1345 | translate the C<\n> in the following statement to its ASCII equivalent | |
1346 | (C<\r> is the same under both Unix and OS/390 & VM/ESA): | |
e41182b5 GS |
1347 | |
1348 | print "Content-type: text/html\r\n\r\n"; | |
1349 | ||
d1e3b762 | 1350 | The values of C<$^O> on some of these platforms includes: |
e41182b5 | 1351 | |
d1e3b762 GS |
1352 | uname $^O $Config{'archname'} |
1353 | -------------------------------------------- | |
1354 | OS/390 os390 os390 | |
1355 | OS400 os400 os400 | |
1356 | POSIX-BC posix-bc BS2000-posix-bc | |
1357 | VM/ESA vmesa vmesa | |
3c075c7d | 1358 | |
e41182b5 GS |
1359 | Some simple tricks for determining if you are running on an EBCDIC |
1360 | platform could include any of the following (perhaps all): | |
1361 | ||
1362 | if ("\t" eq "\05") { print "EBCDIC may be spoken here!\n"; } | |
1363 | ||
1364 | if (ord('A') == 193) { print "EBCDIC may be spoken here!\n"; } | |
1365 | ||
1366 | if (chr(169) eq 'z') { print "EBCDIC may be spoken here!\n"; } | |
1367 | ||
b7df3edc | 1368 | One thing you may not want to rely on is the EBCDIC encoding |
0a47030a GS |
1369 | of punctuation characters since these may differ from code page to code |
1370 | page (and once your module or script is rumoured to work with EBCDIC, | |
1371 | folks will want it to work with all EBCDIC character sets). | |
e41182b5 GS |
1372 | |
1373 | Also see: | |
1374 | ||
1375 | =over 4 | |
1376 | ||
c997b287 GS |
1377 | =item * |
1378 | ||
dc5c060f | 1379 | L<perlos390>, F<README.os390>, F<perlbs2000>, F<README.vmesa>, |
bb462878 | 1380 | L<perlebcdic>. |
c997b287 GS |
1381 | |
1382 | =item * | |
e41182b5 GS |
1383 | |
1384 | The perl-mvs@perl.org list is for discussion of porting issues as well as | |
1385 | general usage issues for all EBCDIC Perls. Send a message body of | |
1386 | "subscribe perl-mvs" to majordomo@perl.org. | |
1387 | ||
7ee27b7c | 1388 | =item * |
c997b287 GS |
1389 | |
1390 | AS/400 Perl information at | |
b1866b2d | 1391 | http://as400.rochester.ibm.com/ |
d1e3b762 | 1392 | as well as on CPAN in the F<ports/> directory. |
e41182b5 GS |
1393 | |
1394 | =back | |
1395 | ||
b8099c3d CN |
1396 | =head2 Acorn RISC OS |
1397 | ||
b7df3edc GS |
1398 | Because Acorns use ASCII with newlines (C<\n>) in text files as C<\012> like |
1399 | Unix, and because Unix filename emulation is turned on by default, | |
1400 | most simple scripts will probably work "out of the box". The native | |
6ab3f9cb | 1401 | filesystem is modular, and individual filesystems are free to be |
0a47030a | 1402 | case-sensitive or insensitive, and are usually case-preserving. Some |
b7df3edc | 1403 | native filesystems have name length limits, which file and directory |
6ab3f9cb GS |
1404 | names are silently truncated to fit. Scripts should be aware that the |
1405 | standard filesystem currently has a name length limit of B<10> | |
1406 | characters, with up to 77 items in a directory, but other filesystems | |
0a47030a | 1407 | may not impose such limitations. |
b8099c3d CN |
1408 | |
1409 | Native filenames are of the form | |
1410 | ||
6ab3f9cb | 1411 | Filesystem#Special_Field::DiskName.$.Directory.Directory.File |
dd9f0070 | 1412 | |
b8099c3d CN |
1413 | where |
1414 | ||
1415 | Special_Field is not usually present, but may contain . and $ . | |
1416 | Filesystem =~ m|[A-Za-z0-9_]| | |
1417 | DsicName =~ m|[A-Za-z0-9_/]| | |
1418 | $ represents the root directory | |
1419 | . is the path separator | |
1420 | @ is the current directory (per filesystem but machine global) | |
1421 | ^ is the parent directory | |
1422 | Directory and File =~ m|[^\0- "\.\$\%\&:\@\\^\|\177]+| | |
1423 | ||
1424 | The default filename translation is roughly C<tr|/.|./|;> | |
1425 | ||
6ab3f9cb | 1426 | Note that C<"ADFS::HardDisk.$.File" ne 'ADFS::HardDisk.$.File'> and that |
0a47030a GS |
1427 | the second stage of C<$> interpolation in regular expressions will fall |
1428 | foul of the C<$.> if scripts are not careful. | |
1429 | ||
1430 | Logical paths specified by system variables containing comma-separated | |
b7df3edc | 1431 | search lists are also allowed; hence C<System:Modules> is a valid |
0a47030a | 1432 | filename, and the filesystem will prefix C<Modules> with each section of |
6ab3f9cb | 1433 | C<System$Path> until a name is made that points to an object on disk. |
b7df3edc | 1434 | Writing to a new file C<System:Modules> would be allowed only if |
0a47030a GS |
1435 | C<System$Path> contains a single item list. The filesystem will also |
1436 | expand system variables in filenames if enclosed in angle brackets, so | |
c47ff5f1 | 1437 | C<< <System$Dir>.Modules >> would look for the file |
0a47030a | 1438 | S<C<$ENV{'System$Dir'} . 'Modules'>>. The obvious implication of this is |
c47ff5f1 | 1439 | that B<fully qualified filenames can start with C<< <> >>> and should |
0a47030a | 1440 | be protected when C<open> is used for input. |
b8099c3d CN |
1441 | |
1442 | Because C<.> was in use as a directory separator and filenames could not | |
1443 | be assumed to be unique after 10 characters, Acorn implemented the C | |
1444 | compiler to strip the trailing C<.c> C<.h> C<.s> and C<.o> suffix from | |
1445 | filenames specified in source code and store the respective files in | |
b7df3edc | 1446 | subdirectories named after the suffix. Hence files are translated: |
b8099c3d CN |
1447 | |
1448 | foo.h h.foo | |
1449 | C:foo.h C:h.foo (logical path variable) | |
1450 | sys/os.h sys.h.os (C compiler groks Unix-speak) | |
1451 | 10charname.c c.10charname | |
1452 | 10charname.o o.10charname | |
1453 | 11charname_.c c.11charname (assuming filesystem truncates at 10) | |
1454 | ||
1455 | The Unix emulation library's translation of filenames to native assumes | |
b7df3edc GS |
1456 | that this sort of translation is required, and it allows a user-defined list |
1457 | of known suffixes that it will transpose in this fashion. This may | |
1458 | seem transparent, but consider that with these rules C<foo/bar/baz.h> | |
0a47030a GS |
1459 | and C<foo/bar/h/baz> both map to C<foo.bar.h.baz>, and that C<readdir> and |
1460 | C<glob> cannot and do not attempt to emulate the reverse mapping. Other | |
6ab3f9cb | 1461 | C<.>'s in filenames are translated to C</>. |
0a47030a | 1462 | |
b7df3edc | 1463 | As implied above, the environment accessed through C<%ENV> is global, and |
0a47030a | 1464 | the convention is that program specific environment variables are of the |
6ab3f9cb GS |
1465 | form C<Program$Name>. Each filesystem maintains a current directory, |
1466 | and the current filesystem's current directory is the B<global> current | |
b7df3edc GS |
1467 | directory. Consequently, sociable programs don't change the current |
1468 | directory but rely on full pathnames, and programs (and Makefiles) cannot | |
0a47030a GS |
1469 | assume that they can spawn a child process which can change the current |
1470 | directory without affecting its parent (and everyone else for that | |
1471 | matter). | |
1472 | ||
b7df3edc GS |
1473 | Because native operating system filehandles are global and are currently |
1474 | allocated down from 255, with 0 being a reserved value, the Unix emulation | |
0a47030a GS |
1475 | library emulates Unix filehandles. Consequently, you can't rely on |
1476 | passing C<STDIN>, C<STDOUT>, or C<STDERR> to your children. | |
1477 | ||
1478 | The desire of users to express filenames of the form | |
c47ff5f1 | 1479 | C<< <Foo$Dir>.Bar >> on the command line unquoted causes problems, |
0a47030a | 1480 | too: C<``> command output capture has to perform a guessing game. It |
c47ff5f1 | 1481 | assumes that a string C<< <[^<>]+\$[^<>]> >> is a |
0a47030a | 1482 | reference to an environment variable, whereas anything else involving |
c47ff5f1 | 1483 | C<< < >> or C<< > >> is redirection, and generally manages to be 99% |
0a47030a GS |
1484 | right. Of course, the problem remains that scripts cannot rely on any |
1485 | Unix tools being available, or that any tools found have Unix-like command | |
1486 | line arguments. | |
1487 | ||
b7df3edc GS |
1488 | Extensions and XS are, in theory, buildable by anyone using free |
1489 | tools. In practice, many don't, as users of the Acorn platform are | |
1490 | used to binary distributions. MakeMaker does run, but no available | |
1491 | make currently copes with MakeMaker's makefiles; even if and when | |
1492 | this should be fixed, the lack of a Unix-like shell will cause | |
1493 | problems with makefile rules, especially lines of the form C<cd | |
1494 | sdbm && make all>, and anything using quoting. | |
b8099c3d CN |
1495 | |
1496 | "S<RISC OS>" is the proper name for the operating system, but the value | |
1497 | in C<$^O> is "riscos" (because we don't like shouting). | |
1498 | ||
e41182b5 GS |
1499 | =head2 Other perls |
1500 | ||
b7df3edc GS |
1501 | Perl has been ported to many platforms that do not fit into any of |
1502 | the categories listed above. Some, such as AmigaOS, Atari MiNT, | |
1503 | BeOS, HP MPE/iX, QNX, Plan 9, and VOS, have been well-integrated | |
1504 | into the standard Perl source code kit. You may need to see the | |
1505 | F<ports/> directory on CPAN for information, and possibly binaries, | |
1506 | for the likes of: aos, Atari ST, lynxos, riscos, Novell Netware, | |
1507 | Tandem Guardian, I<etc.> (Yes, we know that some of these OSes may | |
1508 | fall under the Unix category, but we are not a standards body.) | |
e41182b5 | 1509 | |
d1e3b762 GS |
1510 | Some approximate operating system names and their C<$^O> values |
1511 | in the "OTHER" category include: | |
1512 | ||
1513 | OS $^O $Config{'archname'} | |
1514 | ------------------------------------------ | |
1515 | Amiga DOS amigaos m68k-amigos | |
cec2c193 | 1516 | BeOS beos |
d1e3b762 GS |
1517 | MPE/iX mpeix PA-RISC1.1 |
1518 | ||
e41182b5 GS |
1519 | See also: |
1520 | ||
1521 | =over 4 | |
1522 | ||
c997b287 GS |
1523 | =item * |
1524 | ||
1525 | Amiga, F<README.amiga> (installed as L<perlamiga>). | |
1526 | ||
1527 | =item * | |
d1e3b762 | 1528 | |
c997b287 GS |
1529 | Atari, F<README.mint> and Guido Flohr's web page |
1530 | http://stud.uni-sb.de/~gufl0000/ | |
e41182b5 | 1531 | |
c997b287 | 1532 | =item * |
d1e3b762 | 1533 | |
c997b287 | 1534 | Be OS, F<README.beos> |
e41182b5 | 1535 | |
c997b287 GS |
1536 | =item * |
1537 | ||
1538 | HP 300 MPE/iX, F<README.mpeix> and Mark Bixby's web page | |
34aaaa84 | 1539 | http://www.bixby.org/mark/perlix.html |
c997b287 GS |
1540 | |
1541 | =item * | |
e41182b5 | 1542 | |
6ab3f9cb | 1543 | A free perl5-based PERL.NLM for Novell Netware is available in |
c997b287 | 1544 | precompiled binary and source code form from http://www.novell.com/ |
6ab3f9cb | 1545 | as well as from CPAN. |
e41182b5 | 1546 | |
13a2d996 | 1547 | =item * |
c997b287 | 1548 | |
e6f03d26 | 1549 | S<Plan 9>, F<README.plan9> |
d1e3b762 | 1550 | |
e41182b5 GS |
1551 | =back |
1552 | ||
e41182b5 GS |
1553 | =head1 FUNCTION IMPLEMENTATIONS |
1554 | ||
b7df3edc GS |
1555 | Listed below are functions that are either completely unimplemented |
1556 | or else have been implemented differently on various platforms. | |
1557 | Following each description will be, in parentheses, a list of | |
1558 | platforms that the description applies to. | |
e41182b5 | 1559 | |
b7df3edc GS |
1560 | The list may well be incomplete, or even wrong in some places. When |
1561 | in doubt, consult the platform-specific README files in the Perl | |
1562 | source distribution, and any other documentation resources accompanying | |
1563 | a given port. | |
e41182b5 | 1564 | |
0a47030a | 1565 | Be aware, moreover, that even among Unix-ish systems there are variations. |
e41182b5 | 1566 | |
b7df3edc GS |
1567 | For many functions, you can also query C<%Config>, exported by |
1568 | default from the Config module. For example, to check whether the | |
1569 | platform has the C<lstat> call, check C<$Config{d_lstat}>. See | |
1570 | L<Config> for a full description of available variables. | |
e41182b5 GS |
1571 | |
1572 | =head2 Alphabetical Listing of Perl Functions | |
1573 | ||
1574 | =over 8 | |
1575 | ||
e41182b5 GS |
1576 | =item -X |
1577 | ||
b7df3edc | 1578 | C<-r>, C<-w>, and C<-x> have a limited meaning only; directories |
e41182b5 | 1579 | and applications are executable, and there are no uid/gid |
b7df3edc | 1580 | considerations. C<-o> is not supported. (S<Mac OS>) |
e41182b5 | 1581 | |
038ae9a4 SH |
1582 | C<-w> only inspects the read-only file attribute (FILE_ATTRIBUTE_READONLY), |
1583 | which determines whether the directory can be deleted, not whether it can | |
1584 | be written to. Directories always have read and write access unless denied | |
1585 | by discretionary access control lists (DACLs). (S<Win32>) | |
1586 | ||
b7df3edc GS |
1587 | C<-r>, C<-w>, C<-x>, and C<-o> tell whether the file is accessible, |
1588 | which may not reflect UIC-based file protections. (VMS) | |
e41182b5 | 1589 | |
b8099c3d CN |
1590 | C<-s> returns the size of the data fork, not the total size of data fork |
1591 | plus resource fork. (S<Mac OS>). | |
1592 | ||
1593 | C<-s> by name on an open file will return the space reserved on disk, | |
1594 | rather than the current extent. C<-s> on an open filehandle returns the | |
b7df3edc | 1595 | current size. (S<RISC OS>) |
b8099c3d | 1596 | |
e41182b5 | 1597 | C<-R>, C<-W>, C<-X>, C<-O> are indistinguishable from C<-r>, C<-w>, |
b8099c3d | 1598 | C<-x>, C<-o>. (S<Mac OS>, Win32, VMS, S<RISC OS>) |
e41182b5 GS |
1599 | |
1600 | C<-b>, C<-c>, C<-k>, C<-g>, C<-p>, C<-u>, C<-A> are not implemented. | |
1601 | (S<Mac OS>) | |
1602 | ||
287a962e | 1603 | C<-g>, C<-k>, C<-l>, C<-u>, C<-A> are not particularly meaningful. |
b8099c3d | 1604 | (Win32, VMS, S<RISC OS>) |
e41182b5 | 1605 | |
287a962e JD |
1606 | C<-p> is not particularly meaningful. (VMS, S<RISC OS>) |
1607 | ||
e41182b5 GS |
1608 | C<-d> is true if passed a device spec without an explicit directory. |
1609 | (VMS) | |
1610 | ||
1611 | C<-T> and C<-B> are implemented, but might misclassify Mac text files | |
0a47030a | 1612 | with foreign characters; this is the case will all platforms, but may |
b7df3edc | 1613 | affect S<Mac OS> often. (S<Mac OS>) |
e41182b5 GS |
1614 | |
1615 | C<-x> (or C<-X>) determine if a file ends in one of the executable | |
b7df3edc | 1616 | suffixes. C<-S> is meaningless. (Win32) |
e41182b5 | 1617 | |
b8099c3d CN |
1618 | C<-x> (or C<-X>) determine if a file has an executable file type. |
1619 | (S<RISC OS>) | |
1620 | ||
47cd99a4 | 1621 | =item atan2 |
519bc777 RGS |
1622 | |
1623 | Due to issues with various CPUs, math libraries, compilers, and standards, | |
1624 | results for C<atan2()> may vary depending on any combination of the above. | |
1625 | Perl attempts to conform to the Open Group/IEEE standards for the results | |
1626 | returned from C<atan2()>, but cannot force the issue if the system Perl is | |
1627 | run on does not allow it. (Tru64, HP-UX 10.20) | |
1628 | ||
1629 | The current version of the standards for C<atan2()> is available at | |
1630 | L<http://www.opengroup.org/onlinepubs/009695399/functions/atan2.html>. | |
1631 | ||
47cd99a4 | 1632 | =item binmode |
e41182b5 | 1633 | |
b7df3edc | 1634 | Meaningless. (S<Mac OS>, S<RISC OS>) |
e41182b5 GS |
1635 | |
1636 | Reopens file and restores pointer; if function fails, underlying | |
1637 | filehandle may be closed, or pointer may be in a different position. | |
1638 | (VMS) | |
1639 | ||
1640 | The value returned by C<tell> may be affected after the call, and | |
1641 | the filehandle may be flushed. (Win32) | |
1642 | ||
47cd99a4 | 1643 | =item chmod |
e41182b5 | 1644 | |
b7df3edc | 1645 | Only limited meaning. Disabling/enabling write permission is mapped to |
e41182b5 GS |
1646 | locking/unlocking the file. (S<Mac OS>) |
1647 | ||
1648 | Only good for changing "owner" read-write access, "group", and "other" | |
1649 | bits are meaningless. (Win32) | |
1650 | ||
b8099c3d CN |
1651 | Only good for changing "owner" and "other" read-write access. (S<RISC OS>) |
1652 | ||
495c5fdc PG |
1653 | Access permissions are mapped onto VOS access-control list changes. (VOS) |
1654 | ||
4e51f8e4 | 1655 | The actual permissions set depend on the value of the C<CYGWIN> |
789f0d36 | 1656 | in the SYSTEM environment settings. (Cygwin) |
4e51f8e4 | 1657 | |
47cd99a4 | 1658 | =item chown |
e41182b5 | 1659 | |
3fd80bd6 | 1660 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>) |
e41182b5 GS |
1661 | |
1662 | Does nothing, but won't fail. (Win32) | |
1663 | ||
3fd80bd6 PG |
1664 | A little funky, because VOS's notion of ownership is a little funky (VOS). |
1665 | ||
e41182b5 GS |
1666 | =item chroot |
1667 | ||
e6f03d26 | 1668 | Not implemented. (S<Mac OS>, Win32, VMS, S<Plan 9>, S<RISC OS>, VOS, VM/ESA) |
e41182b5 | 1669 | |
47cd99a4 | 1670 | =item crypt |
e41182b5 GS |
1671 | |
1672 | May not be available if library or source was not provided when building | |
b8099c3d | 1673 | perl. (Win32) |
e41182b5 | 1674 | |
47cd99a4 | 1675 | =item dbmclose |
e41182b5 | 1676 | |
e6f03d26 | 1677 | Not implemented. (VMS, S<Plan 9>, VOS) |
e41182b5 | 1678 | |
47cd99a4 | 1679 | =item dbmopen |
e41182b5 | 1680 | |
e6f03d26 | 1681 | Not implemented. (VMS, S<Plan 9>, VOS) |
e41182b5 | 1682 | |
47cd99a4 | 1683 | =item dump |
e41182b5 | 1684 | |
b8099c3d | 1685 | Not useful. (S<Mac OS>, S<RISC OS>) |
e41182b5 | 1686 | |
84d78eb7 | 1687 | Not supported. (Cygwin, Win32) |
e41182b5 | 1688 | |
b8099c3d | 1689 | Invokes VMS debugger. (VMS) |
e41182b5 | 1690 | |
47cd99a4 | 1691 | =item exec |
e41182b5 GS |
1692 | |
1693 | Not implemented. (S<Mac OS>) | |
1694 | ||
7c5ffed3 | 1695 | Implemented via Spawn. (VM/ESA) |
3c075c7d | 1696 | |
0f897271 GS |
1697 | Does not automatically flush output handles on some platforms. |
1698 | (SunOS, Solaris, HP-UX) | |
1699 | ||
fe12c0e8 MS |
1700 | =item exit |
1701 | ||
1702 | Emulates UNIX exit() (which considers C<exit 1> to indicate an error) by | |
1703 | mapping the C<1> to SS$_ABORT (C<44>). This behavior may be overridden | |
1704 | with the pragma C<use vmsish 'exit'>. As with the CRTL's exit() | |
1705 | function, C<exit 0> is also mapped to an exit status of SS$_NORMAL | |
1706 | (C<1>); this mapping cannot be overridden. Any other argument to exit() | |
016930a6 JM |
1707 | is used directly as Perl's exit status. On VMS, unless the future |
1708 | POSIX_EXIT mode is enabled, the exit code should always be a valid | |
1709 | VMS exit code and not a generic number. When the POSIX_EXIT mode is | |
1710 | enabled, a generic number will be encoded in a method compatible with | |
1711 | the C library _POSIX_EXIT macro so that it can be decoded by other | |
1712 | programs, particularly ones written in C, like the GNV package. (VMS) | |
fe12c0e8 | 1713 | |
47cd99a4 | 1714 | =item fcntl |
e41182b5 | 1715 | |
016930a6 JM |
1716 | Not implemented. (Win32) |
1717 | Some functions available based on the version of VMS. (VMS) | |
e41182b5 | 1718 | |
47cd99a4 | 1719 | =item flock |
e41182b5 | 1720 | |
495c5fdc | 1721 | Not implemented (S<Mac OS>, VMS, S<RISC OS>, VOS). |
e41182b5 GS |
1722 | |
1723 | Available only on Windows NT (not on Windows 95). (Win32) | |
1724 | ||
1725 | =item fork | |
1726 | ||
3fd80bd6 | 1727 | Not implemented. (S<Mac OS>, AmigaOS, S<RISC OS>, VM/ESA, VMS) |
0f897271 GS |
1728 | |
1729 | Emulated using multiple interpreters. See L<perlfork>. (Win32) | |
1730 | ||
1731 | Does not automatically flush output handles on some platforms. | |
1732 | (SunOS, Solaris, HP-UX) | |
e41182b5 GS |
1733 | |
1734 | =item getlogin | |
1735 | ||
b8099c3d | 1736 | Not implemented. (S<Mac OS>, S<RISC OS>) |
e41182b5 | 1737 | |
47cd99a4 | 1738 | =item getpgrp |
e41182b5 | 1739 | |
3fd80bd6 | 1740 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>) |
e41182b5 GS |
1741 | |
1742 | =item getppid | |
1743 | ||
41cbbefa | 1744 | Not implemented. (S<Mac OS>, Win32, S<RISC OS>) |
e41182b5 | 1745 | |
47cd99a4 | 1746 | =item getpriority |
e41182b5 | 1747 | |
7c5ffed3 | 1748 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA) |
e41182b5 | 1749 | |
47cd99a4 | 1750 | =item getpwnam |
e41182b5 GS |
1751 | |
1752 | Not implemented. (S<Mac OS>, Win32) | |
1753 | ||
b8099c3d CN |
1754 | Not useful. (S<RISC OS>) |
1755 | ||
47cd99a4 | 1756 | =item getgrnam |
e41182b5 | 1757 | |
b8099c3d | 1758 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>) |
e41182b5 | 1759 | |
47cd99a4 | 1760 | =item getnetbyname |
e41182b5 | 1761 | |
e6f03d26 | 1762 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) |
e41182b5 | 1763 | |
47cd99a4 | 1764 | =item getpwuid |
e41182b5 GS |
1765 | |
1766 | Not implemented. (S<Mac OS>, Win32) | |
1767 | ||
b8099c3d CN |
1768 | Not useful. (S<RISC OS>) |
1769 | ||
47cd99a4 | 1770 | =item getgrgid |
e41182b5 | 1771 | |
b8099c3d | 1772 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>) |
e41182b5 | 1773 | |
47cd99a4 | 1774 | =item getnetbyaddr |
e41182b5 | 1775 | |
e6f03d26 | 1776 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) |
e41182b5 | 1777 | |
47cd99a4 | 1778 | =item getprotobynumber |
e41182b5 GS |
1779 | |
1780 | Not implemented. (S<Mac OS>) | |
1781 | ||
47cd99a4 | 1782 | =item getservbyport |
e41182b5 GS |
1783 | |
1784 | Not implemented. (S<Mac OS>) | |
1785 | ||
1786 | =item getpwent | |
1787 | ||
7c5ffed3 | 1788 | Not implemented. (S<Mac OS>, Win32, VM/ESA) |
e41182b5 GS |
1789 | |
1790 | =item getgrent | |
1791 | ||
7c5ffed3 | 1792 | Not implemented. (S<Mac OS>, Win32, VMS, VM/ESA) |
e41182b5 | 1793 | |
ef5a6dd7 JH |
1794 | =item gethostbyname |
1795 | ||
1796 | C<gethostbyname('localhost')> does not work everywhere: you may have | |
1797 | to use C<gethostbyname('127.0.0.1')>. (S<Mac OS>, S<Irix 5>) | |
1798 | ||
e41182b5 GS |
1799 | =item gethostent |
1800 | ||
1801 | Not implemented. (S<Mac OS>, Win32) | |
1802 | ||
1803 | =item getnetent | |
1804 | ||
e6f03d26 | 1805 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) |
e41182b5 GS |
1806 | |
1807 | =item getprotoent | |
1808 | ||
e6f03d26 | 1809 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) |
e41182b5 GS |
1810 | |
1811 | =item getservent | |
1812 | ||
e6f03d26 | 1813 | Not implemented. (Win32, S<Plan 9>) |
e41182b5 | 1814 | |
47cd99a4 | 1815 | =item sethostent |
e41182b5 | 1816 | |
e6f03d26 | 1817 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>) |
e41182b5 | 1818 | |
47cd99a4 | 1819 | =item setnetent |
e41182b5 | 1820 | |
e6f03d26 | 1821 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>) |
e41182b5 | 1822 | |
47cd99a4 | 1823 | =item setprotoent |
e41182b5 | 1824 | |
e6f03d26 | 1825 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>) |
e41182b5 | 1826 | |
47cd99a4 | 1827 | =item setservent |
e41182b5 | 1828 | |
e6f03d26 | 1829 | Not implemented. (S<Plan 9>, Win32, S<RISC OS>) |
e41182b5 GS |
1830 | |
1831 | =item endpwent | |
1832 | ||
a3dfe201 | 1833 | Not implemented. (S<Mac OS>, MPE/iX, VM/ESA, Win32) |
e41182b5 GS |
1834 | |
1835 | =item endgrent | |
1836 | ||
a3dfe201 | 1837 | Not implemented. (S<Mac OS>, MPE/iX, S<RISC OS>, VM/ESA, VMS, Win32) |
e41182b5 GS |
1838 | |
1839 | =item endhostent | |
1840 | ||
1841 | Not implemented. (S<Mac OS>, Win32) | |
1842 | ||
1843 | =item endnetent | |
1844 | ||
e6f03d26 | 1845 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) |
e41182b5 GS |
1846 | |
1847 | =item endprotoent | |
1848 | ||
e6f03d26 | 1849 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) |
e41182b5 GS |
1850 | |
1851 | =item endservent | |
1852 | ||
e6f03d26 | 1853 | Not implemented. (S<Plan 9>, Win32) |
e41182b5 GS |
1854 | |
1855 | =item getsockopt SOCKET,LEVEL,OPTNAME | |
1856 | ||
e6f03d26 | 1857 | Not implemented. (S<Plan 9>) |
e41182b5 | 1858 | |
e41182b5 GS |
1859 | =item glob |
1860 | ||
63f87e49 GS |
1861 | This operator is implemented via the File::Glob extension on most |
1862 | platforms. See L<File::Glob> for portability information. | |
b8099c3d | 1863 | |
62aa5637 MS |
1864 | =item gmtime |
1865 | ||
461d5a49 MS |
1866 | In theory, gmtime() is reliable from -2**63 to 2**63-1. However, |
1867 | because work arounds in the implementation use floating point numbers, | |
1868 | it will become inaccurate as the time gets larger. This is a bug and | |
1869 | will be fixed in the future. | |
62aa5637 | 1870 | |
e41182b5 GS |
1871 | =item ioctl FILEHANDLE,FUNCTION,SCALAR |
1872 | ||
1873 | Not implemented. (VMS) | |
1874 | ||
1875 | Available only for socket handles, and it does what the ioctlsocket() call | |
1876 | in the Winsock API does. (Win32) | |
1877 | ||
b8099c3d CN |
1878 | Available only for socket handles. (S<RISC OS>) |
1879 | ||
47cd99a4 | 1880 | =item kill |
e41182b5 | 1881 | |
862b5365 JH |
1882 | C<kill(0, LIST)> is implemented for the sake of taint checking; |
1883 | use with other signals is unimplemented. (S<Mac OS>) | |
1884 | ||
1885 | Not implemented, hence not useful for taint checking. (S<RISC OS>) | |
e41182b5 | 1886 | |
63f87e49 GS |
1887 | C<kill()> doesn't have the semantics of C<raise()>, i.e. it doesn't send |
1888 | a signal to the identified process like it does on Unix platforms. | |
1889 | Instead C<kill($sig, $pid)> terminates the process identified by $pid, | |
1890 | and makes it exit immediately with exit status $sig. As in Unix, if | |
1891 | $sig is 0 and the specified process exists, it returns true without | |
1892 | actually terminating it. (Win32) | |
e41182b5 | 1893 | |
d0302514 JD |
1894 | C<kill(-9, $pid)> will terminate the process specified by $pid and |
1895 | recursively all child processes owned by it. This is different from | |
1896 | the Unix semantics, where the signal will be delivered to all | |
1897 | processes in the same process group as the process specified by | |
1898 | $pid. (Win32) | |
1899 | ||
016930a6 JM |
1900 | Is not supported for process identification number of 0 or negative |
1901 | numbers. (VMS) | |
1902 | ||
47cd99a4 | 1903 | =item link |
e41182b5 | 1904 | |
016930a6 | 1905 | Not implemented. (S<Mac OS>, MPE/iX, S<RISC OS>) |
e41182b5 | 1906 | |
433acd8a JH |
1907 | Link count not updated because hard links are not quite that hard |
1908 | (They are sort of half-way between hard and soft links). (AmigaOS) | |
1909 | ||
63d6c08b JD |
1910 | Hard links are implemented on Win32 under NTFS only. They are |
1911 | natively supported on Windows 2000 and later. On Windows NT they | |
1912 | are implemented using the Windows POSIX subsystem support and the | |
1913 | Perl process will need Administrator or Backup Operator privileges | |
1914 | to create hard links. | |
a3dfe201 | 1915 | |
016930a6 JM |
1916 | Available on 64 bit OpenVMS 8.2 and later. (VMS) |
1917 | ||
62aa5637 MS |
1918 | =item localtime |
1919 | ||
dc164757 MS |
1920 | localtime() has the same range as L<gmtime>, but because time zone |
1921 | rules change its accuracy for historical and future times may degrade | |
1922 | but usually by no more than an hour. | |
62aa5637 | 1923 | |
e41182b5 GS |
1924 | =item lstat |
1925 | ||
016930a6 | 1926 | Not implemented. (S<RISC OS>) |
e41182b5 | 1927 | |
63f87e49 | 1928 | Return values (especially for device and inode) may be bogus. (Win32) |
e41182b5 | 1929 | |
47cd99a4 | 1930 | =item msgctl |
e41182b5 | 1931 | |
47cd99a4 | 1932 | =item msgget |
e41182b5 | 1933 | |
47cd99a4 | 1934 | =item msgsnd |
e41182b5 | 1935 | |
47cd99a4 | 1936 | =item msgrcv |
e41182b5 | 1937 | |
e6f03d26 | 1938 | Not implemented. (S<Mac OS>, Win32, VMS, S<Plan 9>, S<RISC OS>, VOS) |
e41182b5 | 1939 | |
47cd99a4 | 1940 | =item open |
e41182b5 | 1941 | |
b7df3edc | 1942 | The C<|> variants are supported only if ToolServer is installed. |
e41182b5 GS |
1943 | (S<Mac OS>) |
1944 | ||
c47ff5f1 | 1945 | open to C<|-> and C<-|> are unsupported. (S<Mac OS>, Win32, S<RISC OS>) |
e41182b5 | 1946 | |
0f897271 GS |
1947 | Opening a process does not automatically flush output handles on some |
1948 | platforms. (SunOS, Solaris, HP-UX) | |
1949 | ||
47cd99a4 | 1950 | =item pipe |
e41182b5 | 1951 | |
433acd8a JH |
1952 | Very limited functionality. (MiNT) |
1953 | ||
e41182b5 GS |
1954 | =item readlink |
1955 | ||
b8099c3d | 1956 | Not implemented. (Win32, VMS, S<RISC OS>) |
e41182b5 | 1957 | |
47cd99a4 | 1958 | =item rename |
c9b2b9d4 SS |
1959 | |
1960 | Can't move directories between directories on different logical volumes. (Win32) | |
1961 | ||
47cd99a4 | 1962 | =item select |
e41182b5 | 1963 | |
689c5c24 | 1964 | Only implemented on sockets. (Win32, VMS) |
e41182b5 | 1965 | |
b8099c3d CN |
1966 | Only reliable on sockets. (S<RISC OS>) |
1967 | ||
76e05f0b | 1968 | Note that the C<select FILEHANDLE> form is generally portable. |
63f87e49 | 1969 | |
47cd99a4 | 1970 | =item semctl |
e41182b5 | 1971 | |
47cd99a4 | 1972 | =item semget |
e41182b5 | 1973 | |
47cd99a4 | 1974 | =item semop |
e41182b5 | 1975 | |
495c5fdc | 1976 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS) |
e41182b5 | 1977 | |
a3dfe201 GS |
1978 | =item setgrent |
1979 | ||
3fd80bd6 | 1980 | Not implemented. (S<Mac OS>, MPE/iX, VMS, Win32, S<RISC OS>, VOS) |
a3dfe201 | 1981 | |
47cd99a4 | 1982 | =item setpgrp |
e41182b5 | 1983 | |
495c5fdc | 1984 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS) |
e41182b5 | 1985 | |
47cd99a4 | 1986 | =item setpriority |
e41182b5 | 1987 | |
495c5fdc | 1988 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS) |
e41182b5 | 1989 | |
a3dfe201 GS |
1990 | =item setpwent |
1991 | ||
3fd80bd6 | 1992 | Not implemented. (S<Mac OS>, MPE/iX, Win32, S<RISC OS>, VOS) |
a3dfe201 | 1993 | |
47cd99a4 | 1994 | =item setsockopt |
e41182b5 | 1995 | |
e6f03d26 | 1996 | Not implemented. (S<Plan 9>) |
e41182b5 | 1997 | |
47cd99a4 | 1998 | =item shmctl |
e41182b5 | 1999 | |
47cd99a4 | 2000 | =item shmget |
e41182b5 | 2001 | |
47cd99a4 | 2002 | =item shmread |
e41182b5 | 2003 | |
47cd99a4 | 2004 | =item shmwrite |
e41182b5 | 2005 | |
495c5fdc | 2006 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS) |
e41182b5 | 2007 | |
47cd99a4 | 2008 | =item sockatmark |
80cbd5ad JH |
2009 | |
2010 | A relatively recent addition to socket functions, may not | |
2011 | be implemented even in UNIX platforms. | |
2012 | ||
47cd99a4 | 2013 | =item socketpair |
e41182b5 | 2014 | |
f38e12df | 2015 | Not implemented. (S<RISC OS>, VOS, VM/ESA) |
016930a6 JM |
2016 | |
2017 | Available on 64 bit OpenVMS 8.2 and later. (VMS) | |
e41182b5 | 2018 | |
e41182b5 GS |
2019 | =item stat |
2020 | ||
d62e1b7f JH |
2021 | Platforms that do not have rdev, blksize, or blocks will return these |
2022 | as '', so numeric comparison or manipulation of these fields may cause | |
2023 | 'not numeric' warnings. | |
2024 | ||
e41182b5 | 2025 | mtime and atime are the same thing, and ctime is creation time instead of |
3f1f789b JH |
2026 | inode change time. (S<Mac OS>). |
2027 | ||
2028 | ctime not supported on UFS (S<Mac OS X>). | |
e41182b5 | 2029 | |
95a3fe12 MS |
2030 | ctime is creation time instead of inode change time (Win32). |
2031 | ||
e41182b5 GS |
2032 | device and inode are not meaningful. (Win32) |
2033 | ||
2034 | device and inode are not necessarily reliable. (VMS) | |
2035 | ||
b8099c3d CN |
2036 | mtime, atime and ctime all return the last modification time. Device and |
2037 | inode are not necessarily reliable. (S<RISC OS>) | |
2038 | ||
d62e1b7f JH |
2039 | dev, rdev, blksize, and blocks are not available. inode is not |
2040 | meaningful and will differ between stat calls on the same file. (os2) | |
2041 | ||
73e9292c JH |
2042 | some versions of cygwin when doing a stat("foo") and if not finding it |
2043 | may then attempt to stat("foo.exe") (Cygwin) | |
2044 | ||
1fafdf34 JD |
2045 | On Win32 stat() needs to open the file to determine the link count |
2046 | and update attributes that may have been changed through hard links. | |
2047 | Setting ${^WIN32_SLOPPY_STAT} to a true value speeds up stat() by | |
2048 | not performing this operation. (Win32) | |
2049 | ||
47cd99a4 | 2050 | =item symlink |
e41182b5 | 2051 | |
c73b03b7 JM |
2052 | Not implemented. (Win32, S<RISC OS>) |
2053 | ||
2054 | Implemented on 64 bit VMS 8.3. VMS requires the symbolic link to be in Unix | |
2055 | syntax if it is intended to resolve to a valid path. | |
e41182b5 | 2056 | |
47cd99a4 | 2057 | =item syscall |
e41182b5 | 2058 | |
7c5ffed3 | 2059 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA) |
e41182b5 | 2060 | |
47cd99a4 | 2061 | =item sysopen |
f34d0673 | 2062 | |
dd9f0070 | 2063 | The traditional "0", "1", and "2" MODEs are implemented with different |
322422de GS |
2064 | numeric values on some systems. The flags exported by C<Fcntl> |
2065 | (O_RDONLY, O_WRONLY, O_RDWR) should work everywhere though. (S<Mac | |
7c5ffed3 | 2066 | OS>, OS/390, VM/ESA) |
f34d0673 | 2067 | |
47cd99a4 | 2068 | =item system |
e41182b5 GS |
2069 | |
2070 | Only implemented if ToolServer is installed. (S<Mac OS>) | |
2071 | ||
2072 | As an optimization, may not call the command shell specified in | |
b7df3edc | 2073 | C<$ENV{PERL5SHELL}>. C<system(1, @args)> spawns an external |
e41182b5 GS |
2074 | process and immediately returns its process designator, without |
2075 | waiting for it to terminate. Return value may be used subsequently | |
63f87e49 GS |
2076 | in C<wait> or C<waitpid>. Failure to spawn() a subprocess is indicated |
2077 | by setting $? to "255 << 8". C<$?> is set in a way compatible with | |
2078 | Unix (i.e. the exitstatus of the subprocess is obtained by "$? >> 8", | |
2079 | as described in the documentation). (Win32) | |
e41182b5 | 2080 | |
b8099c3d CN |
2081 | There is no shell to process metacharacters, and the native standard is |
2082 | to pass a command line terminated by "\n" "\r" or "\0" to the spawned | |
c47ff5f1 | 2083 | program. Redirection such as C<< > foo >> is performed (if at all) by |
b8099c3d CN |
2084 | the run time library of the spawned program. C<system> I<list> will call |
2085 | the Unix emulation library's C<exec> emulation, which attempts to provide | |
2086 | emulation of the stdin, stdout, stderr in force in the parent, providing | |
2087 | the child program uses a compatible version of the emulation library. | |
2088 | I<scalar> will call the native command line direct and no such emulation | |
2089 | of a child Unix program will exists. Mileage B<will> vary. (S<RISC OS>) | |
2090 | ||
433acd8a JH |
2091 | Far from being POSIX compliant. Because there may be no underlying |
2092 | /bin/sh tries to work around the problem by forking and execing the | |
9b63e9ec | 2093 | first token in its argument string. Handles basic redirection |
c47ff5f1 | 2094 | ("<" or ">") on its own behalf. (MiNT) |
433acd8a | 2095 | |
0f897271 GS |
2096 | Does not automatically flush output handles on some platforms. |
2097 | (SunOS, Solaris, HP-UX) | |
2098 | ||
9bc98430 CB |
2099 | The return value is POSIX-like (shifted up by 8 bits), which only allows |
2100 | room for a made-up value derived from the severity bits of the native | |
2101 | 32-bit condition code (unless overridden by C<use vmsish 'status'>). | |
016930a6 JM |
2102 | If the native condition code is one that has a POSIX value encoded, the |
2103 | POSIX value will be decoded to extract the expected exit value. | |
9bc98430 CB |
2104 | For more details see L<perlvms/$?>. (VMS) |
2105 | ||
e41182b5 GS |
2106 | =item times |
2107 | ||
2108 | Only the first entry returned is nonzero. (S<Mac OS>) | |
2109 | ||
63f87e49 GS |
2110 | "cumulative" times will be bogus. On anything other than Windows NT |
2111 | or Windows 2000, "system" time will be bogus, and "user" time is | |
2112 | actually the time returned by the clock() function in the C runtime | |
2113 | library. (Win32) | |
e41182b5 | 2114 | |
b8099c3d CN |
2115 | Not useful. (S<RISC OS>) |
2116 | ||
47cd99a4 | 2117 | =item truncate |
e41182b5 | 2118 | |
6d738113 | 2119 | Not implemented. (Older versions of VMS) |
e41182b5 | 2120 | |
3fd80bd6 | 2121 | Truncation to same-or-shorter lengths only. (VOS) |
495c5fdc | 2122 | |
4cfdb94f | 2123 | If a FILEHANDLE is supplied, it must be writable and opened in append |
e71a7dc8 | 2124 | mode (i.e., use C<<< open(FH, '>>filename') >>> |
4cfdb94f GS |
2125 | or C<sysopen(FH,...,O_APPEND|O_RDWR)>. If a filename is supplied, it |
2126 | should not be held open elsewhere. (Win32) | |
2127 | ||
e41182b5 GS |
2128 | =item umask |
2129 | ||
2130 | Returns undef where unavailable, as of version 5.005. | |
2131 | ||
b7df3edc GS |
2132 | C<umask> works but the correct permissions are set only when the file |
2133 | is finally closed. (AmigaOS) | |
433acd8a | 2134 | |
47cd99a4 | 2135 | =item utime |
e41182b5 | 2136 | |
15c65113 | 2137 | Only the modification time is updated. (S<BeOS>, S<Mac OS>, VMS, S<RISC OS>) |
e41182b5 | 2138 | |
322422de GS |
2139 | May not behave as expected. Behavior depends on the C runtime |
2140 | library's implementation of utime(), and the filesystem being | |
2141 | used. The FAT filesystem typically does not support an "access | |
2142 | time" field, and it may limit timestamps to a granularity of | |
2143 | two seconds. (Win32) | |
e41182b5 GS |
2144 | |
2145 | =item wait | |
2146 | ||
47cd99a4 | 2147 | =item waitpid |
e41182b5 | 2148 | |
3fd80bd6 | 2149 | Not implemented. (S<Mac OS>) |
e41182b5 GS |
2150 | |
2151 | Can only be applied to process handles returned for processes spawned | |
a6f858fb | 2152 | using C<system(1, ...)> or pseudo processes created with C<fork()>. (Win32) |
e41182b5 | 2153 | |
b8099c3d CN |
2154 | Not useful. (S<RISC OS>) |
2155 | ||
e41182b5 GS |
2156 | =back |
2157 | ||
2158 | ||
ba58ab26 JH |
2159 | =head1 Supported Platforms |
2160 | ||
522b859a | 2161 | As of July 2002 (the Perl release 5.8.0), the following platforms are |
cec2c193 JH |
2162 | able to build Perl from the standard source code distribution |
2163 | available at http://www.cpan.org/src/index.html | |
2164 | ||
2165 | AIX | |
2166 | BeOS | |
6f683aa2 | 2167 | BSD/OS (BSDi) |
cec2c193 JH |
2168 | Cygwin |
2169 | DG/UX | |
811b48f2 | 2170 | DOS DJGPP 1) |
cec2c193 JH |
2171 | DYNIX/ptx |
2172 | EPOC R5 | |
2173 | FreeBSD | |
6f683aa2 | 2174 | HI-UXMPP (Hitachi) (5.8.0 worked but we didn't know it) |
cec2c193 JH |
2175 | HP-UX |
2176 | IRIX | |
2177 | Linux | |
8939ba94 | 2178 | Mac OS Classic |
6f683aa2 | 2179 | Mac OS X (Darwin) |
cec2c193 JH |
2180 | MPE/iX |
2181 | NetBSD | |
2182 | NetWare | |
2183 | NonStop-UX | |
6f683aa2 | 2184 | ReliantUNIX (formerly SINIX) |
cec2c193 | 2185 | OpenBSD |
6f683aa2 | 2186 | OpenVMS (formerly VMS) |
3ebac25b | 2187 | Open UNIX (Unixware) (since Perl 5.8.1/5.9.0) |
cec2c193 | 2188 | OS/2 |
522b859a | 2189 | OS/400 (using the PASE) (since Perl 5.8.1/5.9.0) |
70de81db | 2190 | PowerUX |
6f683aa2 | 2191 | POSIX-BC (formerly BS2000) |
cec2c193 JH |
2192 | QNX |
2193 | Solaris | |
70de81db | 2194 | SunOS 4 |
6f683aa2 JH |
2195 | SUPER-UX (NEC) |
2196 | Tru64 UNIX (formerly DEC OSF/1, Digital UNIX) | |
cec2c193 JH |
2197 | UNICOS |
2198 | UNICOS/mk | |
2199 | UTS | |
2200 | VOS | |
811b48f2 | 2201 | Win95/98/ME/2K/XP 2) |
c40b5d1d | 2202 | WinCE |
6f683aa2 | 2203 | z/OS (formerly OS/390) |
cec2c193 | 2204 | VM/ESA |
ba58ab26 | 2205 | |
811b48f2 JH |
2206 | 1) in DOS mode either the DOS or OS/2 ports can be used |
2207 | 2) compilers: Borland, MinGW (GCC), VC6 | |
cec2c193 | 2208 | |
c40b5d1d | 2209 | The following platforms worked with the previous releases (5.6 and |
cec2c193 JH |
2210 | 5.7), but we did not manage either to fix or to test these in time |
2211 | for the 5.8.0 release. There is a very good chance that many of these | |
70de81db | 2212 | will work fine with the 5.8.0. |
cec2c193 | 2213 | |
8da2b1be | 2214 | BSD/OS |
cec2c193 JH |
2215 | DomainOS |
2216 | Hurd | |
2217 | LynxOS | |
2218 | MachTen | |
2219 | PowerMAX | |
2220 | SCO SV | |
cec2c193 JH |
2221 | SVR4 |
2222 | Unixware | |
2223 | Windows 3.1 | |
ba58ab26 | 2224 | |
70de81db JH |
2225 | Known to be broken for 5.8.0 (but 5.6.1 and 5.7.2 can be used): |
2226 | ||
2227 | AmigaOS | |
2228 | ||
ba58ab26 | 2229 | The following platforms have been known to build Perl from source in |
fd46a41b JH |
2230 | the past (5.005_03 and earlier), but we haven't been able to verify |
2231 | their status for the current release, either because the | |
2232 | hardware/software platforms are rare or because we don't have an | |
2233 | active champion on these platforms--or both. They used to work, | |
2234 | though, so go ahead and try compiling them, and let perlbug@perl.org | |
2235 | of any trouble. | |
ba58ab26 | 2236 | |
cec2c193 JH |
2237 | 3b1 |
2238 | A/UX | |
cec2c193 JH |
2239 | ConvexOS |
2240 | CX/UX | |
2241 | DC/OSx | |
2242 | DDE SMES | |
2243 | DOS EMX | |
2244 | Dynix | |
2245 | EP/IX | |
2246 | ESIX | |
2247 | FPS | |
2248 | GENIX | |
2249 | Greenhills | |
2250 | ISC | |
2251 | MachTen 68k | |
2252 | MiNT | |
2253 | MPC | |
2254 | NEWS-OS | |
2255 | NextSTEP | |
2256 | OpenSTEP | |
2257 | Opus | |
2258 | Plan 9 | |
cec2c193 | 2259 | RISC/os |
8da2b1be | 2260 | SCO ODT/OSR |
cec2c193 JH |
2261 | Stellar |
2262 | SVR2 | |
2263 | TI1500 | |
2264 | TitanOS | |
2265 | Ultrix | |
2266 | Unisys Dynix | |
ba58ab26 JH |
2267 | |
2268 | The following platforms have their own source code distributions and | |
1577cd80 | 2269 | binaries available via http://www.cpan.org/ports/ |
ba58ab26 | 2270 | |
cec2c193 | 2271 | Perl release |
ba58ab26 | 2272 | |
522b859a | 2273 | OS/400 (ILE) 5.005_02 |
cec2c193 | 2274 | Tandem Guardian 5.004 |
ba58ab26 JH |
2275 | |
2276 | The following platforms have only binaries available via | |
a93751fa | 2277 | http://www.cpan.org/ports/index.html : |
ba58ab26 | 2278 | |
cec2c193 | 2279 | Perl release |
ba58ab26 | 2280 | |
cec2c193 JH |
2281 | Acorn RISCOS 5.005_02 |
2282 | AOS 5.002 | |
2283 | LynxOS 5.004_02 | |
ba58ab26 JH |
2284 | |
2285 | Although we do suggest that you always build your own Perl from | |
2286 | the source code, both for maximal configurability and for security, | |
2287 | in case you are in a hurry you can check | |
a93751fa | 2288 | http://www.cpan.org/ports/index.html for binary distributions. |
ba58ab26 | 2289 | |
c997b287 GS |
2290 | =head1 SEE ALSO |
2291 | ||
cec2c193 | 2292 | L<perlaix>, L<perlamiga>, L<perlapollo>, L<perlbeos>, L<perlbs2000>, |
18a271bd | 2293 | L<perlce>, L<perlcygwin>, L<perldgux>, L<perldos>, L<perlepoc>, |
469e7be4 | 2294 | L<perlebcdic>, L<perlfreebsd>, L<perlhurd>, L<perlhpux>, L<perlirix>, |
0d53b15f | 2295 | L<perlmachten>, L<perlmacos>, L<perlmacosx>, L<perlmint>, L<perlmpeix>, |
522b859a JH |
2296 | L<perlnetware>, L<perlos2>, L<perlos390>, L<perlos400>, |
2297 | L<perlplan9>, L<perlqnx>, L<perlsolaris>, L<perltru64>, | |
2298 | L<perlunicode>, L<perlvmesa>, L<perlvms>, L<perlvos>, | |
2299 | L<perlwin32>, and L<Win32>. | |
c997b287 | 2300 | |
e41182b5 GS |
2301 | =head1 AUTHORS / CONTRIBUTORS |
2302 | ||
06e9666b | 2303 | Abigail <abigail@foad.org>, |
c47ff5f1 GS |
2304 | Charles Bailey <bailey@newman.upenn.edu>, |
2305 | Graham Barr <gbarr@pobox.com>, | |
2306 | Tom Christiansen <tchrist@perl.com>, | |
06e9666b | 2307 | Nicholas Clark <nick@ccl4.org>, |
c47ff5f1 | 2308 | Thomas Dorner <Thomas.Dorner@start.de>, |
06e9666b A |
2309 | Andy Dougherty <doughera@lafayette.edu>, |
2310 | Dominic Dunlop <domo@computer.org>, | |
2311 | Neale Ferguson <neale@vma.tabnsw.com.au>, | |
c47ff5f1 | 2312 | David J. Fiander <davidf@mks.com>, |
3fd80bd6 | 2313 | Paul Green <Paul.Green@stratus.com>, |
06e9666b | 2314 | M.J.T. Guy <mjtg@cam.ac.uk>, |
61f30a5e | 2315 | Jarkko Hietaniemi <jhi@iki.fi>, |
c47ff5f1 | 2316 | Luther Huffman <lutherh@stratcom.com>, |
06e9666b A |
2317 | Nick Ing-Simmons <nick@ing-simmons.net>, |
2318 | Andreas J. KE<ouml>nig <a.koenig@mind.de>, | |
c47ff5f1 GS |
2319 | Markus Laker <mlaker@contax.co.uk>, |
2320 | Andrew M. Langmead <aml@world.std.com>, | |
2321 | Larry Moore <ljmoore@freespace.net>, | |
2322 | Paul Moore <Paul.Moore@uk.origin-it.com>, | |
2323 | Chris Nandor <pudge@pobox.com>, | |
1afc07ec | 2324 | Matthias Neeracher <neeracher@mac.com>, |
e71a7dc8 | 2325 | Philip Newton <pne@cpan.org>, |
c47ff5f1 GS |
2326 | Gary Ng <71564.1743@CompuServe.COM>, |
2327 | Tom Phoenix <rootbeer@teleport.com>, | |
2328 | AndrE<eacute> Pirard <A.Pirard@ulg.ac.be>, | |
2329 | Peter Prymmer <pvhp@forte.com>, | |
2330 | Hugo van der Sanden <hv@crypt0.demon.co.uk>, | |
2331 | Gurusamy Sarathy <gsar@activestate.com>, | |
2332 | Paul J. Schinder <schinder@pobox.com>, | |
2333 | Michael G Schwern <schwern@pobox.com>, | |
06e9666b | 2334 | Dan Sugalski <dan@sidhe.org>, |
c47ff5f1 | 2335 | Nathan Torkington <gnat@frii.com>. |
016930a6 | 2336 | John Malmberg <wb8tyw@qsl.net> |