Commit | Line | Data |
e41182b5 |
1 | =head1 NAME |
2 | |
3 | perlport - Writing portable Perl |
4 | |
e41182b5 |
5 | =head1 DESCRIPTION |
6 | |
b7df3edc |
7 | Perl runs on numerous operating systems. While most of them share |
8 | much in common, they also have their own unique features. |
e41182b5 |
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 |
12 | you know where the lines are drawn, and you can stay within them. |
13 | |
b7df3edc |
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 |
26 | problem. |
27 | |
28 | Looking at it another way, writing portable code is usually about |
b7df3edc |
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 |
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 |
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 |
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 |
51 | writing portable code, and this document is entirely about those issues. |
52 | |
53 | =back |
54 | |
b7df3edc |
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 |
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 |
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 |
67 | deliberate in your decision. |
68 | |
69 | The material below is separated into three main sections: main issues of |
70 | portability (L<"ISSUES">, platform-specific issues (L<"PLATFORMS">, and |
b7df3edc |
71 | built-in perl functions that behave differently on various ports |
0a47030a |
72 | (L<"FUNCTION IMPLEMENTATIONS">. |
e41182b5 |
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 |
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 |
87 | and S<Mac OS> uses C<\015>. |
88 | |
b7df3edc |
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 |
94 | Unix does the same thing on ttys in canonical mode. C<\015\012> |
95 | is commonly referred to as CRLF. |
96 | |
5b3eff12 |
97 | A common cause of unportable programs is the misuse of chop() to trim |
98 | newlines: |
99 | |
100 | # XXX UNPORTABLE! |
101 | while(<FILE>) { |
102 | chop; |
103 | @array = split(/:/); |
104 | #... |
105 | } |
106 | |
8939ba94 |
107 | You can get away with this on Unix and Mac OS (they have a single |
5b3eff12 |
108 | character end-of-line), but the same program will break under DOSish |
109 | perls because you're only chop()ing half the end-of-line. Instead, |
110 | chomp() should be used to trim newlines. The Dunce::Files module can |
111 | help audit your code for misuses of chop(). |
112 | |
113 | When dealing with binary files (or text files in binary mode) be sure |
114 | to explicitly set $/ to the appropriate value for your file format |
115 | before using chomp(). |
116 | |
b7df3edc |
117 | Because of the "text" mode translation, DOSish perls have limitations |
118 | in using C<seek> and C<tell> on a file accessed in "text" mode. |
119 | Stick to C<seek>-ing to locations you got from C<tell> (and no |
120 | others), and you are usually free to use C<seek> and C<tell> even |
121 | in "text" mode. Using C<seek> or C<tell> or other file operations |
122 | may be non-portable. If you use C<binmode> on a file, however, you |
123 | can usually C<seek> and C<tell> with arbitrary values in safety. |
e41182b5 |
124 | |
125 | A common misconception in socket programming is that C<\n> eq C<\012> |
0a47030a |
126 | everywhere. When using protocols such as common Internet protocols, |
e41182b5 |
127 | C<\012> and C<\015> are called for specifically, and the values of |
128 | the logical C<\n> and C<\r> (carriage return) are not reliable. |
129 | |
130 | print SOCKET "Hi there, client!\r\n"; # WRONG |
131 | print SOCKET "Hi there, client!\015\012"; # RIGHT |
132 | |
0a47030a |
133 | However, using C<\015\012> (or C<\cM\cJ>, or C<\x0D\x0A>) can be tedious |
134 | and unsightly, as well as confusing to those maintaining the code. As |
6ab3f9cb |
135 | such, the Socket module supplies the Right Thing for those who want it. |
e41182b5 |
136 | |
137 | use Socket qw(:DEFAULT :crlf); |
138 | print SOCKET "Hi there, client!$CRLF" # RIGHT |
139 | |
6ab3f9cb |
140 | When reading from a socket, remember that the default input record |
b7df3edc |
141 | separator C<$/> is C<\n>, but robust socket code will recognize as |
142 | either C<\012> or C<\015\012> as end of line: |
e41182b5 |
143 | |
144 | while (<SOCKET>) { |
145 | # ... |
146 | } |
147 | |
b7df3edc |
148 | Because both CRLF and LF end in LF, the input record separator can |
149 | be set to LF and any CR stripped later. Better to write: |
e41182b5 |
150 | |
151 | use Socket qw(:DEFAULT :crlf); |
152 | local($/) = LF; # not needed if $/ is already \012 |
153 | |
154 | while (<SOCKET>) { |
155 | s/$CR?$LF/\n/; # not sure if socket uses LF or CRLF, OK |
156 | # s/\015?\012/\n/; # same thing |
157 | } |
158 | |
b7df3edc |
159 | This example is preferred over the previous one--even for Unix |
160 | platforms--because now any C<\015>'s (C<\cM>'s) are stripped out |
e41182b5 |
161 | (and there was much rejoicing). |
162 | |
6ab3f9cb |
163 | Similarly, functions that return text data--such as a function that |
b7df3edc |
164 | fetches a web page--should sometimes translate newlines before |
165 | returning the data, if they've not yet been translated to the local |
166 | newline representation. A single line of code will often suffice: |
2ee0eb3c |
167 | |
b7df3edc |
168 | $data =~ s/\015?\012/\n/g; |
169 | return $data; |
2ee0eb3c |
170 | |
6ab3f9cb |
171 | Some of this may be confusing. Here's a handy reference to the ASCII CR |
172 | and LF characters. You can print it out and stick it in your wallet. |
173 | |
74555b7a |
174 | LF eq \012 eq \x0A eq \cJ eq chr(10) eq ASCII 10 |
175 | CR eq \015 eq \x0D eq \cM eq chr(13) eq ASCII 13 |
6ab3f9cb |
176 | |
177 | | Unix | DOS | Mac | |
178 | --------------------------- |
179 | \n | LF | LF | CR | |
180 | \r | CR | CR | LF | |
181 | \n * | LF | CRLF | CR | |
182 | \r * | CR | CR | LF | |
183 | --------------------------- |
184 | * text-mode STDIO |
185 | |
b7df3edc |
186 | The Unix column assumes that you are not accessing a serial line |
187 | (like a tty) in canonical mode. If you are, then CR on input becomes |
188 | "\n", and "\n" on output becomes CRLF. |
189 | |
6ab3f9cb |
190 | These are just the most common definitions of C<\n> and C<\r> in Perl. |
522b859a |
191 | There may well be others. For example, on an EBCDIC implementation |
192 | such as z/OS (OS/390) or OS/400 (using the ILE, the PASE is ASCII-based) |
193 | the above material is similar to "Unix" but the code numbers change: |
74555b7a |
194 | |
195 | LF eq \025 eq \x15 eq chr(21) eq CP-1047 21 |
196 | LF eq \045 eq \x25 eq \cU eq chr(37) eq CP-0037 37 |
197 | CR eq \015 eq \x0D eq \cM eq chr(13) eq CP-1047 13 |
198 | CR eq \015 eq \x0D eq \cM eq chr(13) eq CP-0037 13 |
199 | |
200 | | z/OS | OS/400 | |
201 | ---------------------- |
202 | \n | LF | LF | |
203 | \r | CR | CR | |
204 | \n * | LF | LF | |
205 | \r * | CR | CR | |
206 | ---------------------- |
207 | * text-mode STDIO |
6ab3f9cb |
208 | |
322422de |
209 | =head2 Numbers endianness and Width |
210 | |
211 | Different CPUs store integers and floating point numbers in different |
212 | orders (called I<endianness>) and widths (32-bit and 64-bit being the |
b7df3edc |
213 | most common today). This affects your programs when they attempt to transfer |
214 | numbers in binary format from one CPU architecture to another, |
215 | usually either "live" via network connection, or by storing the |
216 | numbers to secondary storage such as a disk file or tape. |
322422de |
217 | |
b7df3edc |
218 | Conflicting storage orders make utter mess out of the numbers. If a |
d1e3b762 |
219 | little-endian host (Intel, VAX) stores 0x12345678 (305419896 in |
b84d4f81 |
220 | decimal), a big-endian host (Motorola, Sparc, PA) reads it as |
221 | 0x78563412 (2018915346 in decimal). Alpha and MIPS can be either: |
222 | Digital/Compaq used/uses them in little-endian mode; SGI/Cray uses |
223 | them in big-endian mode. To avoid this problem in network (socket) |
224 | connections use the C<pack> and C<unpack> formats C<n> and C<N>, the |
225 | "network" orders. These are guaranteed to be portable. |
322422de |
226 | |
d1e3b762 |
227 | You can explore the endianness of your platform by unpacking a |
228 | data structure packed in native format such as: |
229 | |
230 | print unpack("h*", pack("s2", 1, 2)), "\n"; |
231 | # '10002000' on e.g. Intel x86 or Alpha 21064 in little-endian mode |
232 | # '00100020' on e.g. Motorola 68040 |
233 | |
234 | If you need to distinguish between endian architectures you could use |
235 | either of the variables set like so: |
236 | |
237 | $is_big_endian = unpack("h*", pack("s", 1)) =~ /01/; |
4375e838 |
238 | $is_little_endian = unpack("h*", pack("s", 1)) =~ /^1/; |
d1e3b762 |
239 | |
b7df3edc |
240 | Differing widths can cause truncation even between platforms of equal |
241 | endianness. The platform of shorter width loses the upper parts of the |
322422de |
242 | number. There is no good solution for this problem except to avoid |
243 | transferring or storing raw binary numbers. |
244 | |
b7df3edc |
245 | One can circumnavigate both these problems in two ways. Either |
322422de |
246 | transfer and store numbers always in text format, instead of raw |
b7df3edc |
247 | binary, or else consider using modules like Data::Dumper (included in |
83df6a1d |
248 | the standard distribution as of Perl 5.005) and Storable (included as |
249 | of perl 5.8). Keeping all data as text significantly simplifies matters. |
322422de |
250 | |
979699d9 |
251 | The v-strings are portable only up to v2147483647 (0x7FFFFFFF), that's |
252 | how far EBCDIC, or more precisely UTF-EBCDIC will go. |
253 | |
433acd8a |
254 | =head2 Files and Filesystems |
e41182b5 |
255 | |
256 | Most platforms these days structure files in a hierarchical fashion. |
b7df3edc |
257 | So, it is reasonably safe to assume that all platforms support the |
6ab3f9cb |
258 | notion of a "path" to uniquely identify a file on the system. How |
b7df3edc |
259 | that path is really written, though, differs considerably. |
e41182b5 |
260 | |
4375e838 |
261 | Although similar, file path specifications differ between Unix, |
b7df3edc |
262 | Windows, S<Mac OS>, OS/2, VMS, VOS, S<RISC OS>, and probably others. |
263 | Unix, for example, is one of the few OSes that has the elegant idea |
264 | of a single root directory. |
322422de |
265 | |
6ab3f9cb |
266 | DOS, OS/2, VMS, VOS, and Windows can work similarly to Unix with C</> |
267 | as path separator, or in their own idiosyncratic ways (such as having |
268 | several root directories and various "unrooted" device files such NIL: |
269 | and LPT:). |
322422de |
270 | |
271 | S<Mac OS> uses C<:> as a path separator instead of C</>. |
272 | |
6ab3f9cb |
273 | The filesystem may support neither hard links (C<link>) nor |
274 | symbolic links (C<symlink>, C<readlink>, C<lstat>). |
433acd8a |
275 | |
6ab3f9cb |
276 | The filesystem may support neither access timestamp nor change |
433acd8a |
277 | timestamp (meaning that about the only portable timestamp is the |
278 | modification timestamp), or one second granularity of any timestamps |
279 | (e.g. the FAT filesystem limits the time granularity to two seconds). |
280 | |
74555b7a |
281 | The "inode change timestamp" (the C<-C> filetest) may really be the |
95a3fe12 |
282 | "creation timestamp" (which it is not in UNIX). |
283 | |
495c5fdc |
284 | VOS perl can emulate Unix filenames with C</> as path separator. The |
285 | native pathname characters greater-than, less-than, number-sign, and |
286 | percent-sign are always accepted. |
287 | |
6ab3f9cb |
288 | S<RISC OS> perl can emulate Unix filenames with C</> as path |
322422de |
289 | separator, or go native and use C<.> for path separator and C<:> to |
6ab3f9cb |
290 | signal filesystems and disk names. |
e41182b5 |
291 | |
a1667ba3 |
292 | Don't assume UNIX filesystem access semantics: that read, write, |
293 | and execute are all the permissions there are, and even if they exist, |
294 | that their semantics (for example what do r, w, and x mean on |
295 | a directory) are the UNIX ones. The various UNIX/POSIX compatibility |
296 | layers usually try to make interfaces like chmod() work, but sometimes |
297 | there simply is no good mapping. |
298 | |
b7df3edc |
299 | If all this is intimidating, have no (well, maybe only a little) |
300 | fear. There are modules that can help. The File::Spec modules |
301 | provide methods to do the Right Thing on whatever platform happens |
302 | to be running the program. |
e41182b5 |
303 | |
6ab3f9cb |
304 | use File::Spec::Functions; |
305 | chdir(updir()); # go up one directory |
306 | $file = catfile(curdir(), 'temp', 'file.txt'); |
e41182b5 |
307 | # on Unix and Win32, './temp/file.txt' |
308 | # on Mac OS, ':temp:file.txt' |
d1e3b762 |
309 | # on VMS, '[.temp]file.txt' |
e41182b5 |
310 | |
b7df3edc |
311 | File::Spec is available in the standard distribution as of version |
d1e3b762 |
312 | 5.004_05. File::Spec::Functions is only in File::Spec 0.7 and later, |
313 | and some versions of perl come with version 0.6. If File::Spec |
314 | is not updated to 0.7 or later, you must use the object-oriented |
315 | interface from File::Spec (or upgrade File::Spec). |
e41182b5 |
316 | |
b7df3edc |
317 | In general, production code should not have file paths hardcoded. |
318 | Making them user-supplied or read from a configuration file is |
319 | better, keeping in mind that file path syntax varies on different |
320 | machines. |
e41182b5 |
321 | |
322 | This is especially noticeable in scripts like Makefiles and test suites, |
323 | which often assume C</> as a path separator for subdirectories. |
324 | |
b7df3edc |
325 | Also of use is File::Basename from the standard distribution, which |
e41182b5 |
326 | splits a pathname into pieces (base filename, full path to directory, |
327 | and file suffix). |
328 | |
19799a22 |
329 | Even when on a single platform (if you can call Unix a single platform), |
b7df3edc |
330 | remember not to count on the existence or the contents of particular |
3c075c7d |
331 | system-specific files or directories, like F</etc/passwd>, |
b7df3edc |
332 | F</etc/sendmail.conf>, F</etc/resolv.conf>, or even F</tmp/>. For |
333 | example, F</etc/passwd> may exist but not contain the encrypted |
334 | passwords, because the system is using some form of enhanced security. |
335 | Or it may not contain all the accounts, because the system is using NIS. |
3c075c7d |
336 | If code does need to rely on such a file, include a description of the |
b7df3edc |
337 | file and its format in the code's documentation, then make it easy for |
3c075c7d |
338 | the user to override the default location of the file. |
339 | |
b7df3edc |
340 | Don't assume a text file will end with a newline. They should, |
341 | but people forget. |
e41182b5 |
342 | |
ec481373 |
343 | Do not have two files or directories of the same name with different |
344 | case, like F<test.pl> and F<Test.pl>, as many platforms have |
345 | case-insensitive (or at least case-forgiving) filenames. Also, try |
346 | not to have non-word characters (except for C<.>) in the names, and |
347 | keep them to the 8.3 convention, for maximum portability, onerous a |
348 | burden though this may appear. |
dd9f0070 |
349 | |
b7df3edc |
350 | Likewise, when using the AutoSplit module, try to keep your functions to |
351 | 8.3 naming and case-insensitive conventions; or, at the least, |
dd9f0070 |
352 | make it so the resulting files have a unique (case-insensitively) |
353 | first 8 characters. |
354 | |
ec481373 |
355 | Whitespace in filenames is tolerated on most systems, but not all, |
356 | and even on systems where it might be tolerated, some utilities |
fe829689 |
357 | might become confused by such whitespace. |
ec481373 |
358 | |
b7df3edc |
359 | Many systems (DOS, VMS) cannot have more than one C<.> in their filenames. |
433acd8a |
360 | |
c47ff5f1 |
361 | Don't assume C<< > >> won't be the first character of a filename. |
fe829689 |
362 | Always use C<< < >> explicitly to open a file for reading, or even |
363 | better, use the three-arg version of open, unless you want the user to |
364 | be able to specify a pipe open. |
0a47030a |
365 | |
fe829689 |
366 | open(FILE, '<', $existing_file) or die $!; |
0a47030a |
367 | |
6ab3f9cb |
368 | If filenames might use strange characters, it is safest to open it |
369 | with C<sysopen> instead of C<open>. C<open> is magic and can |
c47ff5f1 |
370 | translate characters like C<< > >>, C<< < >>, and C<|>, which may |
b7df3edc |
371 | be the wrong thing to do. (Sometimes, though, it's the right thing.) |
fe829689 |
372 | Three-arg open can also help protect against this translation in cases |
373 | where it is undesirable. |
e41182b5 |
374 | |
ec481373 |
375 | Don't use C<:> as a part of a filename since many systems use that for |
8939ba94 |
376 | their own semantics (Mac OS Classic for separating pathname components, |
ec481373 |
377 | many networking schemes and utilities for separating the nodename and |
08fef530 |
378 | the pathname, and so on). For the same reasons, avoid C<@>, C<;> and |
379 | C<|>. |
ec481373 |
380 | |
e1516da7 |
381 | Don't assume that in pathnames you can collapse two leading slashes |
382 | C<//> into one: some networking and clustering filesystems have special |
383 | semantics for that. Let the operating system to sort it out. |
384 | |
ec481373 |
385 | The I<portable filename characters> as defined by ANSI C are |
386 | |
387 | a b c d e f g h i j k l m n o p q r t u v w x y z |
388 | A B C D E F G H I J K L M N O P Q R T U V W X Y Z |
389 | 0 1 2 3 4 5 6 7 8 9 |
390 | . _ - |
391 | |
08fef530 |
392 | and the "-" shouldn't be the first character. If you want to be |
e1516da7 |
393 | hypercorrect, stay case-insensitive and within the 8.3 naming |
394 | convention (all the files and directories have to be unique within one |
395 | directory if their names are lowercased and truncated to eight |
396 | characters before the C<.>, if any, and to three characters after the |
397 | C<.>, if any). (And do not use C<.>s in directory names.) |
ec481373 |
398 | |
e41182b5 |
399 | =head2 System Interaction |
400 | |
b7df3edc |
401 | Not all platforms provide a command line. These are usually platforms |
402 | that rely primarily on a Graphical User Interface (GUI) for user |
403 | interaction. A program requiring a command line interface might |
404 | not work everywhere. This is probably for the user of the program |
405 | to deal with, so don't stay up late worrying about it. |
e41182b5 |
406 | |
c87488a3 |
407 | Some platforms can't delete or rename files held open by the system, |
408 | this limitation may also apply to changing filesystem metainformation |
409 | like file permissions or owners. Remember to C<close> files when you |
410 | are done with them. Don't C<unlink> or C<rename> an open file. Don't |
411 | C<tie> or C<open> a file already tied or opened; C<untie> or C<close> |
412 | it first. |
e41182b5 |
413 | |
0a47030a |
414 | Don't open the same file more than once at a time for writing, as some |
415 | operating systems put mandatory locks on such files. |
416 | |
73e9292c |
417 | Don't assume that write/modify permission on a directory gives the |
418 | right to add or delete files/directories in that directory. That is |
419 | filesystem specific: in some filesystems you need write/modify |
420 | permission also (or even just) in the file/directory itself. In some |
421 | filesystems (AFS, DFS) the permission to add/delete directory entries |
422 | is a completely separate permission. |
423 | |
424 | Don't assume that a single C<unlink> completely gets rid of the file: |
425 | some filesystems (most notably the ones in VMS) have versioned |
426 | filesystems, and unlink() removes only the most recent one (it doesn't |
427 | remove all the versions because by default the native tools on those |
428 | platforms remove just the most recent version, too). The portable |
429 | idiom to remove all the versions of a file is |
430 | |
94bb614c |
431 | 1 while unlink "file"; |
73e9292c |
432 | |
433 | This will terminate if the file is undeleteable for some reason |
434 | (protected, not there, and so on). |
435 | |
e41182b5 |
436 | Don't count on a specific environment variable existing in C<%ENV>. |
0a47030a |
437 | Don't count on C<%ENV> entries being case-sensitive, or even |
4a0d0822 |
438 | case-preserving. Don't try to clear %ENV by saying C<%ENV = ();>, or, |
6be8f7a6 |
439 | if you really have to, make it conditional on C<$^O ne 'VMS'> since in |
4a0d0822 |
440 | VMS the C<%ENV> table is much more than a per-process key-value string |
441 | table. |
e41182b5 |
442 | |
d1e3b762 |
443 | Don't count on signals or C<%SIG> for anything. |
e41182b5 |
444 | |
445 | Don't count on filename globbing. Use C<opendir>, C<readdir>, and |
446 | C<closedir> instead. |
447 | |
b8099c3d |
448 | Don't count on per-program environment variables, or per-program current |
dd9f0070 |
449 | directories. |
b8099c3d |
450 | |
c87488a3 |
451 | Don't count on specific values of C<$!>, neither numeric nor |
452 | especially the strings values-- users may switch their locales causing |
453 | error messages to be translated into their languages. If you can |
454 | trust a POSIXish environment, you can portably use the symbols defined |
455 | by the Errno module, like ENOENT. And don't trust on the values of C<$!> |
456 | at all except immediately after a failed system call. |
3c075c7d |
457 | |
a10d74f3 |
458 | =head2 Command names versus file pathnames |
459 | |
460 | Don't assume that the name used to invoke a command or program with |
461 | C<system> or C<exec> can also be used to test for the existence of the |
462 | file that holds the executable code for that command or program. |
68fb0eb7 |
463 | First, many systems have "internal" commands that are built-in to the |
464 | shell or OS and while these commands can be invoked, there is no |
465 | corresponding file. Second, some operating systems (e.g., Cygwin, |
466 | DJGPP, OS/2, and VOS) have required suffixes for executable files; |
467 | these suffixes are generally permitted on the command name but are not |
a10d74f3 |
468 | required. Thus, a command like "perl" might exist in a file named |
469 | "perl", "perl.exe", or "perl.pm", depending on the operating system. |
470 | The variable "_exe" in the Config module holds the executable suffix, |
68fb0eb7 |
471 | if any. Third, the VMS port carefully sets up $^X and |
472 | $Config{perlpath} so that no further processing is required. This is |
473 | just as well, because the matching regular expression used below would |
474 | then have to deal with a possible trailing version number in the VMS |
475 | file name. |
a10d74f3 |
476 | |
477 | To convert $^X to a file pathname, taking account of the requirements |
478 | of the various operating system possibilities, say: |
479 | use Config; |
a10d74f3 |
480 | $thisperl = $^X; |
68fb0eb7 |
481 | if ($^O ne 'VMS') |
482 | {$thisperl .= $Config{_exe} unless $thisperl =~ m/$Config{_exe}$/i;} |
a10d74f3 |
483 | |
484 | To convert $Config{perlpath} to a file pathname, say: |
a10d74f3 |
485 | use Config; |
68fb0eb7 |
486 | $thisperl = $Config{perlpath}; |
487 | if ($^O ne 'VMS') |
488 | {$thisperl .= $Config{_exe} unless $thisperl =~ m/$Config{_exe}$/i;} |
a10d74f3 |
489 | |
e41182b5 |
490 | =head2 Interprocess Communication (IPC) |
491 | |
b7df3edc |
492 | In general, don't directly access the system in code meant to be |
493 | portable. That means, no C<system>, C<exec>, C<fork>, C<pipe>, |
494 | C<``>, C<qx//>, C<open> with a C<|>, nor any of the other things |
495 | that makes being a perl hacker worth being. |
e41182b5 |
496 | |
497 | Commands that launch external processes are generally supported on |
b7df3edc |
498 | most platforms (though many of them do not support any type of |
499 | forking). The problem with using them arises from what you invoke |
500 | them on. External tools are often named differently on different |
4375e838 |
501 | platforms, may not be available in the same location, might accept |
b7df3edc |
502 | different arguments, can behave differently, and often present their |
503 | results in a platform-dependent way. Thus, you should seldom depend |
504 | on them to produce consistent results. (Then again, if you're calling |
505 | I<netstat -a>, you probably don't expect it to run on both Unix and CP/M.) |
e41182b5 |
506 | |
b7df3edc |
507 | One especially common bit of Perl code is opening a pipe to B<sendmail>: |
e41182b5 |
508 | |
b7df3edc |
509 | open(MAIL, '|/usr/lib/sendmail -t') |
510 | or die "cannot fork sendmail: $!"; |
e41182b5 |
511 | |
512 | This is fine for systems programming when sendmail is known to be |
513 | available. But it is not fine for many non-Unix systems, and even |
514 | some Unix systems that may not have sendmail installed. If a portable |
b7df3edc |
515 | solution is needed, see the various distributions on CPAN that deal |
516 | with it. Mail::Mailer and Mail::Send in the MailTools distribution are |
517 | commonly used, and provide several mailing methods, including mail, |
518 | sendmail, and direct SMTP (via Net::SMTP) if a mail transfer agent is |
519 | not available. Mail::Sendmail is a standalone module that provides |
520 | simple, platform-independent mailing. |
521 | |
522 | The Unix System V IPC (C<msg*(), sem*(), shm*()>) is not available |
523 | even on all Unix platforms. |
e41182b5 |
524 | |
a81e5e2e |
525 | Do not use either the bare result of C<pack("N", 10, 20, 30, 40)> or |
526 | bare v-strings (such as C<v10.20.30.40>) to represent IPv4 addresses: |
527 | both forms just pack the four bytes into network order. That this |
528 | would be equal to the C language C<in_addr> struct (which is what the |
529 | socket code internally uses) is not guaranteed. To be portable use |
530 | the routines of the Socket extension, such as C<inet_aton()>, |
531 | C<inet_ntoa()>, and C<sockaddr_in()>. |
6b2463a0 |
532 | |
e41182b5 |
533 | The rule of thumb for portable code is: Do it all in portable Perl, or |
0a47030a |
534 | use a module (that may internally implement it with platform-specific |
535 | code, but expose a common interface). |
e41182b5 |
536 | |
e41182b5 |
537 | =head2 External Subroutines (XS) |
538 | |
b7df3edc |
539 | XS code can usually be made to work with any platform, but dependent |
e41182b5 |
540 | libraries, header files, etc., might not be readily available or |
541 | portable, or the XS code itself might be platform-specific, just as Perl |
542 | code might be. If the libraries and headers are portable, then it is |
543 | normally reasonable to make sure the XS code is portable, too. |
544 | |
b7df3edc |
545 | A different type of portability issue arises when writing XS code: |
546 | availability of a C compiler on the end-user's system. C brings |
547 | with it its own portability issues, and writing XS code will expose |
548 | you to some of those. Writing purely in Perl is an easier way to |
e41182b5 |
549 | achieve portability. |
550 | |
e41182b5 |
551 | =head2 Standard Modules |
552 | |
553 | In general, the standard modules work across platforms. Notable |
6ab3f9cb |
554 | exceptions are the CPAN module (which currently makes connections to external |
e41182b5 |
555 | programs that may not be available), platform-specific modules (like |
6ab3f9cb |
556 | ExtUtils::MM_VMS), and DBM modules. |
e41182b5 |
557 | |
b7df3edc |
558 | There is no one DBM module available on all platforms. |
6ab3f9cb |
559 | SDBM_File and the others are generally available on all Unix and DOSish |
560 | ports, but not in MacPerl, where only NBDM_File and DB_File are |
0a47030a |
561 | available. |
e41182b5 |
562 | |
563 | The good news is that at least some DBM module should be available, and |
6ab3f9cb |
564 | AnyDBM_File will use whichever module it can find. Of course, then |
b7df3edc |
565 | the code needs to be fairly strict, dropping to the greatest common |
566 | factor (e.g., not exceeding 1K for each record), so that it will |
6ab3f9cb |
567 | work with any DBM module. See L<AnyDBM_File> for more details. |
e41182b5 |
568 | |
e41182b5 |
569 | =head2 Time and Date |
570 | |
0a47030a |
571 | The system's notion of time of day and calendar date is controlled in |
b7df3edc |
572 | widely different ways. Don't assume the timezone is stored in C<$ENV{TZ}>, |
0a47030a |
573 | and even if it is, don't assume that you can control the timezone through |
c87488a3 |
574 | that variable. Don't assume anything about the three-letter timezone |
575 | abbreviations (for example that MST would be the Mountain Standard Time, |
576 | it's been known to stand for Moscow Standard Time). If you need to |
577 | use timezones, express them in some unambiguous format like the |
578 | exact number of minutes offset from UTC, or the POSIX timezone |
579 | format. |
e41182b5 |
580 | |
322422de |
581 | Don't assume that the epoch starts at 00:00:00, January 1, 1970, |
c87488a3 |
582 | because that is OS- and implementation-specific. It is better to |
583 | store a date in an unambiguous representation. The ISO 8601 standard |
584 | defines YYYY-MM-DD as the date format, or YYYY-MM-DDTHH-MM-SS |
585 | (that's a literal "T" separating the date from the time). |
586 | Please do use the ISO 8601 instead of making us to guess what |
587 | date 02/03/04 might be. ISO 8601 even sorts nicely as-is. |
588 | A text representation (like "1987-12-18") can be easily converted |
589 | into an OS-specific value using a module like Date::Parse. |
590 | An array of values, such as those returned by C<localtime>, can be |
591 | converted to an OS-specific representation using Time::Local. |
322422de |
592 | |
19799a22 |
593 | When calculating specific times, such as for tests in time or date modules, |
594 | it may be appropriate to calculate an offset for the epoch. |
b7df3edc |
595 | |
19799a22 |
596 | require Time::Local; |
597 | $offset = Time::Local::timegm(0, 0, 0, 1, 0, 70); |
b7df3edc |
598 | |
19799a22 |
599 | The value for C<$offset> in Unix will be C<0>, but in Mac OS will be |
600 | some large number. C<$offset> can then be added to a Unix time value |
601 | to get what should be the proper value on any system. |
322422de |
602 | |
603 | =head2 Character sets and character encoding |
604 | |
ec481373 |
605 | Assume very little about character sets. |
606 | |
607 | Assume nothing about numerical values (C<ord>, C<chr>) of characters. |
608 | Do not use explicit code point ranges (like \xHH-\xHH); use for |
609 | example symbolic character classes like C<[:print:]>. |
610 | |
611 | Do not assume that the alphabetic characters are encoded contiguously |
612 | (in the numeric sense). There may be gaps. |
613 | |
614 | Do not assume anything about the ordering of the characters. |
615 | The lowercase letters may come before or after the uppercase letters; |
616 | the lowercase and uppercase may be interlaced so that both `a' and `A' |
617 | come before `b'; the accented and other international characters may |
618 | be interlaced so that E<auml> comes before `b'. |
322422de |
619 | |
620 | =head2 Internationalisation |
621 | |
b7df3edc |
622 | If you may assume POSIX (a rather large assumption), you may read |
623 | more about the POSIX locale system from L<perllocale>. The locale |
624 | system at least attempts to make things a little bit more portable, |
625 | or at least more convenient and native-friendly for non-English |
626 | users. The system affects character sets and encoding, and date |
627 | and time formatting--amongst other things. |
e41182b5 |
628 | |
c87488a3 |
629 | If you really want to be international, you should consider Unicode. |
630 | See L<perluniintro> and L<perlunicode> for more information. |
631 | |
11264fdb |
632 | If you want to use non-ASCII bytes (outside the bytes 0x00..0x7f) in |
633 | the "source code" of your code, to be portable you have to be explicit |
634 | about what bytes they are. Someone might for example be using your |
635 | code under a UTF-8 locale, in which case random native bytes might be |
636 | illegal ("Malformed UTF-8 ...") This means that for example embedding |
637 | ISO 8859-1 bytes beyond 0x7f into your strings might cause trouble |
638 | later. If the bytes are native 8-bit bytes, you can use the C<bytes> |
639 | pragma. If the bytes are in a string (regular expression being a |
640 | curious string), you can often also use the C<\xHH> notation instead |
641 | of embedding the bytes as-is. If they are in some particular legacy |
642 | encoding (ether single-byte or something more complicated), you can |
643 | use the C<encoding> pragma. (If you want to write your code in UTF-8, |
644 | you can use either the C<utf8> pragma, or the C<encoding> pragma.) |
645 | The C<bytes> and C<utf8> pragmata are available since Perl 5.6.0, and |
646 | the C<encoding> pragma since Perl 5.8.0. |
647 | |
e41182b5 |
648 | =head2 System Resources |
649 | |
0a47030a |
650 | If your code is destined for systems with severely constrained (or |
651 | missing!) virtual memory systems then you want to be I<especially> mindful |
652 | of avoiding wasteful constructs such as: |
e41182b5 |
653 | |
654 | # NOTE: this is no longer "bad" in perl5.005 |
655 | for (0..10000000) {} # bad |
656 | for (my $x = 0; $x <= 10000000; ++$x) {} # good |
657 | |
658 | @lines = <VERY_LARGE_FILE>; # bad |
659 | |
660 | while (<FILE>) {$file .= $_} # sometimes bad |
0a47030a |
661 | $file = join('', <FILE>); # better |
e41182b5 |
662 | |
b7df3edc |
663 | The last two constructs may appear unintuitive to most people. The |
664 | first repeatedly grows a string, whereas the second allocates a |
665 | large chunk of memory in one go. On some systems, the second is |
666 | more efficient that the first. |
0a47030a |
667 | |
e41182b5 |
668 | =head2 Security |
669 | |
b7df3edc |
670 | Most multi-user platforms provide basic levels of security, usually |
671 | implemented at the filesystem level. Some, however, do |
a1667ba3 |
672 | not-- unfortunately. Thus the notion of user id, or "home" directory, |
b7df3edc |
673 | or even the state of being logged-in, may be unrecognizable on many |
674 | platforms. If you write programs that are security-conscious, it |
675 | is usually best to know what type of system you will be running |
676 | under so that you can write code explicitly for that platform (or |
677 | class of platforms). |
0a47030a |
678 | |
a1667ba3 |
679 | Don't assume the UNIX filesystem access semantics: the operating |
680 | system or the filesystem may be using some ACL systems, which are |
681 | richer languages than the usual rwx. Even if the rwx exist, |
682 | their semantics might be different. |
683 | |
684 | (From security viewpoint testing for permissions before attempting to |
685 | do something is silly anyway: if one tries this, there is potential |
686 | for race conditions-- someone or something might change the |
687 | permissions between the permissions check and the actual operation. |
688 | Just try the operation.) |
689 | |
690 | Don't assume the UNIX user and group semantics: especially, don't |
2edcc0d9 |
691 | expect the C<< $< >> and C<< $> >> (or the C<$(> and C<$)>) to work |
a1667ba3 |
692 | for switching identities (or memberships). |
693 | |
694 | Don't assume set-uid and set-gid semantics. (And even if you do, |
695 | think twice: set-uid and set-gid are a known can of security worms.) |
696 | |
e41182b5 |
697 | =head2 Style |
698 | |
699 | For those times when it is necessary to have platform-specific code, |
700 | consider keeping the platform-specific code in one place, making porting |
6ab3f9cb |
701 | to other platforms easier. Use the Config module and the special |
0a47030a |
702 | variable C<$^O> to differentiate platforms, as described in |
703 | L<"PLATFORMS">. |
e41182b5 |
704 | |
b7df3edc |
705 | Be careful in the tests you supply with your module or programs. |
706 | Module code may be fully portable, but its tests might not be. This |
707 | often happens when tests spawn off other processes or call external |
708 | programs to aid in the testing, or when (as noted above) the tests |
c87488a3 |
709 | assume certain things about the filesystem and paths. Be careful not |
710 | to depend on a specific output style for errors, such as when checking |
711 | C<$!> after a failed system call. Using C<$!> for anything else than |
712 | displaying it as output is doubtful (though see the Errno module for |
713 | testing reasonably portably for error value). Some platforms expect |
714 | a certain output format, and Perl on those platforms may have been |
715 | adjusted accordingly. Most specifically, don't anchor a regex when |
716 | testing an error value. |
e41182b5 |
717 | |
0a47030a |
718 | =head1 CPAN Testers |
e41182b5 |
719 | |
0a47030a |
720 | Modules uploaded to CPAN are tested by a variety of volunteers on |
721 | different platforms. These CPAN testers are notified by mail of each |
e41182b5 |
722 | new upload, and reply to the list with PASS, FAIL, NA (not applicable to |
0a47030a |
723 | this platform), or UNKNOWN (unknown), along with any relevant notations. |
e41182b5 |
724 | |
725 | The purpose of the testing is twofold: one, to help developers fix any |
0a47030a |
726 | problems in their code that crop up because of lack of testing on other |
b7df3edc |
727 | platforms; two, to provide users with information about whether |
0a47030a |
728 | a given module works on a given platform. |
e41182b5 |
729 | |
730 | =over 4 |
731 | |
732 | =item Mailing list: cpan-testers@perl.org |
733 | |
c997b287 |
734 | =item Testing results: http://testers.cpan.org/ |
e41182b5 |
735 | |
736 | =back |
737 | |
e41182b5 |
738 | =head1 PLATFORMS |
739 | |
740 | As of version 5.002, Perl is built with a C<$^O> variable that |
741 | indicates the operating system it was built on. This was implemented |
b7df3edc |
742 | to help speed up code that would otherwise have to C<use Config> |
743 | and use the value of C<$Config{osname}>. Of course, to get more |
e41182b5 |
744 | detailed information about the system, looking into C<%Config> is |
745 | certainly recommended. |
746 | |
b7df3edc |
747 | C<%Config> cannot always be trusted, however, because it was built |
748 | at compile time. If perl was built in one place, then transferred |
749 | elsewhere, some values may be wrong. The values may even have been |
750 | edited after the fact. |
6ab3f9cb |
751 | |
e41182b5 |
752 | =head2 Unix |
753 | |
754 | Perl works on a bewildering variety of Unix and Unix-like platforms (see |
755 | e.g. most of the files in the F<hints/> directory in the source code kit). |
756 | On most of these systems, the value of C<$^O> (hence C<$Config{'osname'}>, |
d1e3b762 |
757 | too) is determined either by lowercasing and stripping punctuation from the |
758 | first field of the string returned by typing C<uname -a> (or a similar command) |
759 | at the shell prompt or by testing the file system for the presence of |
760 | uniquely named files such as a kernel or header file. Here, for example, |
761 | are a few of the more popular Unix flavors: |
e41182b5 |
762 | |
b7df3edc |
763 | uname $^O $Config{'archname'} |
6ab3f9cb |
764 | -------------------------------------------- |
b7df3edc |
765 | AIX aix aix |
6ab3f9cb |
766 | BSD/OS bsdos i386-bsdos |
e1516da7 |
767 | Darwin darwin darwin |
6ab3f9cb |
768 | dgux dgux AViiON-dgux |
769 | DYNIX/ptx dynixptx i386-dynixptx |
b7df3edc |
770 | FreeBSD freebsd freebsd-i386 |
d1e3b762 |
771 | Linux linux arm-linux |
b7df3edc |
772 | Linux linux i386-linux |
6ab3f9cb |
773 | Linux linux i586-linux |
774 | Linux linux ppc-linux |
b7df3edc |
775 | HP-UX hpux PA-RISC1.1 |
776 | IRIX irix irix |
b787fad4 |
777 | Mac OS X darwin darwin |
d1e3b762 |
778 | MachTen PPC machten powerpc-machten |
779 | NeXT 3 next next-fat |
780 | NeXT 4 next OPENSTEP-Mach |
6ab3f9cb |
781 | openbsd openbsd i386-openbsd |
b7df3edc |
782 | OSF1 dec_osf alpha-dec_osf |
6ab3f9cb |
783 | reliantunix-n svr4 RM400-svr4 |
784 | SCO_SV sco_sv i386-sco_sv |
785 | SINIX-N svr4 RM400-svr4 |
786 | sn4609 unicos CRAY_C90-unicos |
787 | sn6521 unicosmk t3e-unicosmk |
788 | sn9617 unicos CRAY_J90-unicos |
b7df3edc |
789 | SunOS solaris sun4-solaris |
790 | SunOS solaris i86pc-solaris |
791 | SunOS4 sunos sun4-sunos |
e41182b5 |
792 | |
b7df3edc |
793 | Because the value of C<$Config{archname}> may depend on the |
794 | hardware architecture, it can vary more than the value of C<$^O>. |
6ab3f9cb |
795 | |
e41182b5 |
796 | =head2 DOS and Derivatives |
797 | |
b7df3edc |
798 | Perl has long been ported to Intel-style microcomputers running under |
e41182b5 |
799 | systems like PC-DOS, MS-DOS, OS/2, and most Windows platforms you can |
800 | bring yourself to mention (except for Windows CE, if you count that). |
b7df3edc |
801 | Users familiar with I<COMMAND.COM> or I<CMD.EXE> style shells should |
e41182b5 |
802 | be aware that each of these file specifications may have subtle |
803 | differences: |
804 | |
805 | $filespec0 = "c:/foo/bar/file.txt"; |
806 | $filespec1 = "c:\\foo\\bar\\file.txt"; |
807 | $filespec2 = 'c:\foo\bar\file.txt'; |
808 | $filespec3 = 'c:\\foo\\bar\\file.txt'; |
809 | |
b7df3edc |
810 | System calls accept either C</> or C<\> as the path separator. |
811 | However, many command-line utilities of DOS vintage treat C</> as |
812 | the option prefix, so may get confused by filenames containing C</>. |
813 | Aside from calling any external programs, C</> will work just fine, |
814 | and probably better, as it is more consistent with popular usage, |
815 | and avoids the problem of remembering what to backwhack and what |
816 | not to. |
e41182b5 |
817 | |
b7df3edc |
818 | The DOS FAT filesystem can accommodate only "8.3" style filenames. Under |
819 | the "case-insensitive, but case-preserving" HPFS (OS/2) and NTFS (NT) |
0a47030a |
820 | filesystems you may have to be careful about case returned with functions |
e41182b5 |
821 | like C<readdir> or used with functions like C<open> or C<opendir>. |
822 | |
b7df3edc |
823 | DOS also treats several filenames as special, such as AUX, PRN, |
824 | NUL, CON, COM1, LPT1, LPT2, etc. Unfortunately, sometimes these |
825 | filenames won't even work if you include an explicit directory |
826 | prefix. It is best to avoid such filenames, if you want your code |
827 | to be portable to DOS and its derivatives. It's hard to know what |
828 | these all are, unfortunately. |
e41182b5 |
829 | |
830 | Users of these operating systems may also wish to make use of |
b7df3edc |
831 | scripts such as I<pl2bat.bat> or I<pl2cmd> to |
e41182b5 |
832 | put wrappers around your scripts. |
833 | |
834 | Newline (C<\n>) is translated as C<\015\012> by STDIO when reading from |
6ab3f9cb |
835 | and writing to files (see L<"Newlines">). C<binmode(FILEHANDLE)> |
836 | will keep C<\n> translated as C<\012> for that filehandle. Since it is a |
837 | no-op on other systems, C<binmode> should be used for cross-platform code |
b7df3edc |
838 | that deals with binary data. That's assuming you realize in advance |
839 | that your data is in binary. General-purpose programs should |
840 | often assume nothing about their data. |
e41182b5 |
841 | |
b7df3edc |
842 | The C<$^O> variable and the C<$Config{archname}> values for various |
e41182b5 |
843 | DOSish perls are as follows: |
844 | |
67ac489e |
845 | OS $^O $Config{archname} ID Version |
846 | -------------------------------------------------------- |
847 | MS-DOS dos ? |
848 | PC-DOS dos ? |
849 | OS/2 os2 ? |
850 | Windows 3.1 ? ? 0 3 01 |
851 | Windows 95 MSWin32 MSWin32-x86 1 4 00 |
852 | Windows 98 MSWin32 MSWin32-x86 1 4 10 |
853 | Windows ME MSWin32 MSWin32-x86 1 ? |
854 | Windows NT MSWin32 MSWin32-x86 2 4 xx |
855 | Windows NT MSWin32 MSWin32-ALPHA 2 4 xx |
856 | Windows NT MSWin32 MSWin32-ppc 2 4 xx |
857 | Windows 2000 MSWin32 MSWin32-x86 2 5 xx |
858 | Windows XP MSWin32 MSWin32-x86 2 ? |
859 | Windows CE MSWin32 ? 3 |
860 | Cygwin cygwin ? |
e41182b5 |
861 | |
34aaaa84 |
862 | The various MSWin32 Perl's can distinguish the OS they are running on |
863 | via the value of the fifth element of the list returned from |
864 | Win32::GetOSVersion(). For example: |
865 | |
866 | if ($^O eq 'MSWin32') { |
867 | my @os_version_info = Win32::GetOSVersion(); |
868 | print +('3.1','95','NT')[$os_version_info[4]],"\n"; |
869 | } |
870 | |
7939d86b |
871 | There are also Win32::IsWinNT() and Win32::IsWin95(), try C<perldoc Win32>, |
872 | and as of libwin32 0.19 (not part of the core Perl distribution) |
873 | Win32::GetOSName(). The very portable POSIX::uname() will work too: |
1d65be3a |
874 | |
875 | c:\> perl -MPOSIX -we "print join '|', uname" |
876 | Windows NT|moonru|5.0|Build 2195 (Service Pack 2)|x86 |
d99f392e |
877 | |
e41182b5 |
878 | Also see: |
879 | |
880 | =over 4 |
881 | |
c997b287 |
882 | =item * |
e41182b5 |
883 | |
c997b287 |
884 | The djgpp environment for DOS, http://www.delorie.com/djgpp/ |
885 | and L<perldos>. |
e41182b5 |
886 | |
c997b287 |
887 | =item * |
e41182b5 |
888 | |
c997b287 |
889 | The EMX environment for DOS, OS/2, etc. emx@iaehv.nl, |
890 | http://www.leo.org/pub/comp/os/os2/leo/gnu/emx+gcc/index.html or |
f224927c |
891 | ftp://hobbes.nmsu.edu/pub/os2/dev/emx/ Also L<perlos2>. |
e41182b5 |
892 | |
c997b287 |
893 | =item * |
d1e3b762 |
894 | |
c997b287 |
895 | Build instructions for Win32 in L<perlwin32>, or under the Cygnus environment |
896 | in L<perlcygwin>. |
897 | |
898 | =item * |
899 | |
900 | The C<Win32::*> modules in L<Win32>. |
901 | |
902 | =item * |
903 | |
904 | The ActiveState Pages, http://www.activestate.com/ |
905 | |
906 | =item * |
907 | |
908 | The Cygwin environment for Win32; F<README.cygwin> (installed |
47dafe4d |
909 | as L<perlcygwin>), http://www.cygwin.com/ |
c997b287 |
910 | |
911 | =item * |
912 | |
913 | The U/WIN environment for Win32, |
cea6626f |
914 | http://www.research.att.com/sw/tools/uwin/ |
c997b287 |
915 | |
cea6626f |
916 | =item * |
d1e3b762 |
917 | |
cea6626f |
918 | Build instructions for OS/2, L<perlos2> |
d1e3b762 |
919 | |
e41182b5 |
920 | =back |
921 | |
dd9f0070 |
922 | =head2 S<Mac OS> |
e41182b5 |
923 | |
924 | Any module requiring XS compilation is right out for most people, because |
925 | MacPerl is built using non-free (and non-cheap!) compilers. Some XS |
926 | modules that can work with MacPerl are built and distributed in binary |
6ab3f9cb |
927 | form on CPAN. |
e41182b5 |
928 | |
929 | Directories are specified as: |
930 | |
931 | volume:folder:file for absolute pathnames |
932 | volume:folder: for absolute pathnames |
933 | :folder:file for relative pathnames |
934 | :folder: for relative pathnames |
935 | :file for relative pathnames |
936 | file for relative pathnames |
937 | |
b7df3edc |
938 | Files are stored in the directory in alphabetical order. Filenames are |
6ab3f9cb |
939 | limited to 31 characters, and may include any character except for |
b7df3edc |
940 | null and C<:>, which is reserved as the path separator. |
e41182b5 |
941 | |
0a47030a |
942 | Instead of C<flock>, see C<FSpSetFLock> and C<FSpRstFLock> in the |
6ab3f9cb |
943 | Mac::Files module, or C<chmod(0444, ...)> and C<chmod(0666, ...)>. |
e41182b5 |
944 | |
945 | In the MacPerl application, you can't run a program from the command line; |
946 | programs that expect C<@ARGV> to be populated can be edited with something |
947 | like the following, which brings up a dialog box asking for the command |
948 | line arguments. |
949 | |
950 | if (!@ARGV) { |
951 | @ARGV = split /\s+/, MacPerl::Ask('Arguments?'); |
952 | } |
953 | |
b7df3edc |
954 | A MacPerl script saved as a "droplet" will populate C<@ARGV> with the full |
e41182b5 |
955 | pathnames of the files dropped onto the script. |
956 | |
b7df3edc |
957 | Mac users can run programs under a type of command line interface |
958 | under MPW (Macintosh Programmer's Workshop, a free development |
959 | environment from Apple). MacPerl was first introduced as an MPW |
960 | tool, and MPW can be used like a shell: |
e41182b5 |
961 | |
962 | perl myscript.plx some arguments |
963 | |
964 | ToolServer is another app from Apple that provides access to MPW tools |
0a47030a |
965 | from MPW and the MacPerl app, which allows MacPerl programs to use |
e41182b5 |
966 | C<system>, backticks, and piped C<open>. |
967 | |
968 | "S<Mac OS>" is the proper name for the operating system, but the value |
969 | in C<$^O> is "MacOS". To determine architecture, version, or whether |
970 | the application or MPW tool version is running, check: |
971 | |
972 | $is_app = $MacPerl::Version =~ /App/; |
973 | $is_tool = $MacPerl::Version =~ /MPW/; |
974 | ($version) = $MacPerl::Version =~ /^(\S+)/; |
975 | $is_ppc = $MacPerl::Architecture eq 'MacPPC'; |
976 | $is_68k = $MacPerl::Architecture eq 'Mac68K'; |
977 | |
b787fad4 |
978 | S<Mac OS X>, based on NeXT's OpenStep OS, runs MacPerl natively, under the |
979 | "Classic" environment. There is no "Carbon" version of MacPerl to run |
980 | under the primary Mac OS X environment. S<Mac OS X> and its Open Source |
981 | version, Darwin, both run Unix perl natively. |
6ab3f9cb |
982 | |
e41182b5 |
983 | Also see: |
984 | |
985 | =over 4 |
986 | |
c997b287 |
987 | =item * |
988 | |
862b5365 |
989 | MacPerl Development, http://dev.macperl.org/ . |
c997b287 |
990 | |
991 | =item * |
992 | |
862b5365 |
993 | The MacPerl Pages, http://www.macperl.com/ . |
e41182b5 |
994 | |
c997b287 |
995 | =item * |
6ab3f9cb |
996 | |
862b5365 |
997 | The MacPerl mailing lists, http://lists.perl.org/ . |
e41182b5 |
998 | |
999 | =back |
1000 | |
e41182b5 |
1001 | =head2 VMS |
1002 | |
c997b287 |
1003 | Perl on VMS is discussed in L<perlvms> in the perl distribution. |
b7df3edc |
1004 | Perl on VMS can accept either VMS- or Unix-style file |
e41182b5 |
1005 | specifications as in either of the following: |
1006 | |
1007 | $ perl -ne "print if /perl_setup/i" SYS$LOGIN:LOGIN.COM |
1008 | $ perl -ne "print if /perl_setup/i" /sys$login/login.com |
1009 | |
1010 | but not a mixture of both as in: |
1011 | |
1012 | $ perl -ne "print if /perl_setup/i" sys$login:/login.com |
1013 | Can't open sys$login:/login.com: file specification syntax error |
1014 | |
1015 | Interacting with Perl from the Digital Command Language (DCL) shell |
1016 | often requires a different set of quotation marks than Unix shells do. |
1017 | For example: |
1018 | |
1019 | $ perl -e "print ""Hello, world.\n""" |
1020 | Hello, world. |
1021 | |
b7df3edc |
1022 | There are several ways to wrap your perl scripts in DCL F<.COM> files, if |
e41182b5 |
1023 | you are so inclined. For example: |
1024 | |
1025 | $ write sys$output "Hello from DCL!" |
1026 | $ if p1 .eqs. "" |
1027 | $ then perl -x 'f$environment("PROCEDURE") |
1028 | $ else perl -x - 'p1 'p2 'p3 'p4 'p5 'p6 'p7 'p8 |
1029 | $ deck/dollars="__END__" |
1030 | #!/usr/bin/perl |
1031 | |
1032 | print "Hello from Perl!\n"; |
1033 | |
1034 | __END__ |
1035 | $ endif |
1036 | |
1037 | Do take care with C<$ ASSIGN/nolog/user SYS$COMMAND: SYS$INPUT> if your |
c47ff5f1 |
1038 | perl-in-DCL script expects to do things like C<< $read = <STDIN>; >>. |
e41182b5 |
1039 | |
1040 | Filenames are in the format "name.extension;version". The maximum |
1041 | length for filenames is 39 characters, and the maximum length for |
1042 | extensions is also 39 characters. Version is a number from 1 to |
1043 | 32767. Valid characters are C</[A-Z0-9$_-]/>. |
1044 | |
b7df3edc |
1045 | VMS's RMS filesystem is case-insensitive and does not preserve case. |
e41182b5 |
1046 | C<readdir> returns lowercased filenames, but specifying a file for |
b7df3edc |
1047 | opening remains case-insensitive. Files without extensions have a |
e41182b5 |
1048 | trailing period on them, so doing a C<readdir> with a file named F<A.;5> |
0a47030a |
1049 | will return F<a.> (though that file could be opened with |
1050 | C<open(FH, 'A')>). |
e41182b5 |
1051 | |
f34d0673 |
1052 | RMS had an eight level limit on directory depths from any rooted logical |
dd9f0070 |
1053 | (allowing 16 levels overall) prior to VMS 7.2. Hence |
1054 | C<PERL_ROOT:[LIB.2.3.4.5.6.7.8]> is a valid directory specification but |
1055 | C<PERL_ROOT:[LIB.2.3.4.5.6.7.8.9]> is not. F<Makefile.PL> authors might |
1056 | have to take this into account, but at least they can refer to the former |
f34d0673 |
1057 | as C</PERL_ROOT/lib/2/3/4/5/6/7/8/>. |
e41182b5 |
1058 | |
6ab3f9cb |
1059 | The VMS::Filespec module, which gets installed as part of the build |
0a47030a |
1060 | process on VMS, is a pure Perl module that can easily be installed on |
1061 | non-VMS platforms and can be helpful for conversions to and from RMS |
1062 | native formats. |
e41182b5 |
1063 | |
5e12dbfa |
1064 | What C<\n> represents depends on the type of file opened. It usually |
1065 | represents C<\012> but it could also be C<\015>, C<\012>, C<\015\012>, |
1066 | C<\000>, C<\040>, or nothing depending on the file organiztion and |
1067 | record format. The VMS::Stdio module provides access to the |
1068 | special fopen() requirements of files with unusual attributes on VMS. |
e41182b5 |
1069 | |
1070 | TCP/IP stacks are optional on VMS, so socket routines might not be |
1071 | implemented. UDP sockets may not be supported. |
1072 | |
1073 | The value of C<$^O> on OpenVMS is "VMS". To determine the architecture |
1074 | that you are running on without resorting to loading all of C<%Config> |
1075 | you can examine the content of the C<@INC> array like so: |
1076 | |
1077 | if (grep(/VMS_AXP/, @INC)) { |
1078 | print "I'm on Alpha!\n"; |
6ab3f9cb |
1079 | |
e41182b5 |
1080 | } elsif (grep(/VMS_VAX/, @INC)) { |
1081 | print "I'm on VAX!\n"; |
6ab3f9cb |
1082 | |
e41182b5 |
1083 | } else { |
1084 | print "I'm not so sure about where $^O is...\n"; |
1085 | } |
1086 | |
b7df3edc |
1087 | On VMS, perl determines the UTC offset from the C<SYS$TIMEZONE_DIFFERENTIAL> |
1088 | logical name. Although the VMS epoch began at 17-NOV-1858 00:00:00.00, |
6ab3f9cb |
1089 | calls to C<localtime> are adjusted to count offsets from |
b7df3edc |
1090 | 01-JAN-1970 00:00:00.00, just like Unix. |
6ab3f9cb |
1091 | |
e41182b5 |
1092 | Also see: |
1093 | |
1094 | =over 4 |
1095 | |
c997b287 |
1096 | =item * |
1097 | |
1098 | F<README.vms> (installed as L<README_vms>), L<perlvms> |
1099 | |
1100 | =item * |
1101 | |
1102 | vmsperl list, majordomo@perl.org |
e41182b5 |
1103 | |
c997b287 |
1104 | (Put the words C<subscribe vmsperl> in message body.) |
e41182b5 |
1105 | |
c997b287 |
1106 | =item * |
e41182b5 |
1107 | |
c997b287 |
1108 | vmsperl on the web, http://www.sidhe.org/vmsperl/index.html |
e41182b5 |
1109 | |
1110 | =back |
1111 | |
495c5fdc |
1112 | =head2 VOS |
1113 | |
9a997319 |
1114 | Perl on VOS is discussed in F<README.vos> in the perl distribution |
1115 | (installed as L<perlvos>). Perl on VOS can accept either VOS- or |
1116 | Unix-style file specifications as in either of the following: |
495c5fdc |
1117 | |
cc07ed0b |
1118 | C<< $ perl -ne "print if /perl_setup/i" >system>notices >> |
1119 | C<< $ perl -ne "print if /perl_setup/i" /system/notices >> |
495c5fdc |
1120 | |
1121 | or even a mixture of both as in: |
1122 | |
cc07ed0b |
1123 | C<< $ perl -ne "print if /perl_setup/i" >system/notices >> |
495c5fdc |
1124 | |
b7df3edc |
1125 | Even though VOS allows the slash character to appear in object |
495c5fdc |
1126 | names, because the VOS port of Perl interprets it as a pathname |
1127 | delimiting character, VOS files, directories, or links whose names |
1128 | contain a slash character cannot be processed. Such files must be |
a3dfe201 |
1129 | renamed before they can be processed by Perl. Note that VOS limits |
1130 | file names to 32 or fewer characters. |
495c5fdc |
1131 | |
495c5fdc |
1132 | The value of C<$^O> on VOS is "VOS". To determine the architecture that |
1133 | you are running on without resorting to loading all of C<%Config> you |
c997b287 |
1134 | can examine the content of the @INC array like so: |
495c5fdc |
1135 | |
24e8e380 |
1136 | if ($^O =~ /VOS/) { |
495c5fdc |
1137 | print "I'm on a Stratus box!\n"; |
1138 | } else { |
1139 | print "I'm not on a Stratus box!\n"; |
1140 | die; |
1141 | } |
1142 | |
495c5fdc |
1143 | Also see: |
1144 | |
1145 | =over 4 |
1146 | |
c997b287 |
1147 | =item * |
495c5fdc |
1148 | |
cc07ed0b |
1149 | F<README.vos> (installed as L<perlvos>) |
c997b287 |
1150 | |
1151 | =item * |
1152 | |
1153 | The VOS mailing list. |
495c5fdc |
1154 | |
1155 | There is no specific mailing list for Perl on VOS. You can post |
1156 | comments to the comp.sys.stratus newsgroup, or subscribe to the general |
cc07ed0b |
1157 | Stratus mailing list. Send a letter with "subscribe Info-Stratus" in |
495c5fdc |
1158 | the message body to majordomo@list.stratagy.com. |
1159 | |
c997b287 |
1160 | =item * |
1161 | |
cc07ed0b |
1162 | VOS Perl on the web at http://ftp.stratus.com/pub/vos/posix/posix.html |
495c5fdc |
1163 | |
1164 | =back |
1165 | |
e41182b5 |
1166 | =head2 EBCDIC Platforms |
1167 | |
1168 | Recent versions of Perl have been ported to platforms such as OS/400 on |
d1e3b762 |
1169 | AS/400 minicomputers as well as OS/390, VM/ESA, and BS2000 for S/390 |
1170 | Mainframes. Such computers use EBCDIC character sets internally (usually |
0cc436d0 |
1171 | Character Code Set ID 0037 for OS/400 and either 1047 or POSIX-BC for S/390 |
1172 | systems). On the mainframe perl currently works under the "Unix system |
1173 | services for OS/390" (formerly known as OpenEdition), VM/ESA OpenEdition, or |
1174 | the BS200 POSIX-BC system (BS2000 is supported in perl 5.6 and greater). |
522b859a |
1175 | See L<perlos390> for details. Note that for OS/400 there is also a port of |
1176 | Perl 5.8.1/5.9.0 or later to the PASE which is ASCII-based (as opposed to |
1177 | ILE which is EBCDIC-based), see L<perlos400>. |
e41182b5 |
1178 | |
7c5ffed3 |
1179 | As of R2.5 of USS for OS/390 and Version 2.3 of VM/ESA these Unix |
1180 | sub-systems do not support the C<#!> shebang trick for script invocation. |
1181 | Hence, on OS/390 and VM/ESA perl scripts can be executed with a header |
1182 | similar to the following simple script: |
e41182b5 |
1183 | |
1184 | : # use perl |
1185 | eval 'exec /usr/local/bin/perl -S $0 ${1+"$@"}' |
1186 | if 0; |
1187 | #!/usr/local/bin/perl # just a comment really |
1188 | |
1189 | print "Hello from perl!\n"; |
1190 | |
d1e3b762 |
1191 | OS/390 will support the C<#!> shebang trick in release 2.8 and beyond. |
1192 | Calls to C<system> and backticks can use POSIX shell syntax on all |
1193 | S/390 systems. |
1194 | |
b7df3edc |
1195 | On the AS/400, if PERL5 is in your library list, you may need |
6ab3f9cb |
1196 | to wrap your perl scripts in a CL procedure to invoke them like so: |
1197 | |
1198 | BEGIN |
1199 | CALL PGM(PERL5/PERL) PARM('/QOpenSys/hello.pl') |
1200 | ENDPGM |
1201 | |
1202 | This will invoke the perl script F<hello.pl> in the root of the |
1203 | QOpenSys file system. On the AS/400 calls to C<system> or backticks |
1204 | must use CL syntax. |
1205 | |
e41182b5 |
1206 | On these platforms, bear in mind that the EBCDIC character set may have |
0a47030a |
1207 | an effect on what happens with some perl functions (such as C<chr>, |
1208 | C<pack>, C<print>, C<printf>, C<ord>, C<sort>, C<sprintf>, C<unpack>), as |
1209 | well as bit-fiddling with ASCII constants using operators like C<^>, C<&> |
1210 | and C<|>, not to mention dealing with socket interfaces to ASCII computers |
6ab3f9cb |
1211 | (see L<"Newlines">). |
e41182b5 |
1212 | |
b7df3edc |
1213 | Fortunately, most web servers for the mainframe will correctly |
1214 | translate the C<\n> in the following statement to its ASCII equivalent |
1215 | (C<\r> is the same under both Unix and OS/390 & VM/ESA): |
e41182b5 |
1216 | |
1217 | print "Content-type: text/html\r\n\r\n"; |
1218 | |
d1e3b762 |
1219 | The values of C<$^O> on some of these platforms includes: |
e41182b5 |
1220 | |
d1e3b762 |
1221 | uname $^O $Config{'archname'} |
1222 | -------------------------------------------- |
1223 | OS/390 os390 os390 |
1224 | OS400 os400 os400 |
1225 | POSIX-BC posix-bc BS2000-posix-bc |
1226 | VM/ESA vmesa vmesa |
3c075c7d |
1227 | |
e41182b5 |
1228 | Some simple tricks for determining if you are running on an EBCDIC |
1229 | platform could include any of the following (perhaps all): |
1230 | |
1231 | if ("\t" eq "\05") { print "EBCDIC may be spoken here!\n"; } |
1232 | |
1233 | if (ord('A') == 193) { print "EBCDIC may be spoken here!\n"; } |
1234 | |
1235 | if (chr(169) eq 'z') { print "EBCDIC may be spoken here!\n"; } |
1236 | |
b7df3edc |
1237 | One thing you may not want to rely on is the EBCDIC encoding |
0a47030a |
1238 | of punctuation characters since these may differ from code page to code |
1239 | page (and once your module or script is rumoured to work with EBCDIC, |
1240 | folks will want it to work with all EBCDIC character sets). |
e41182b5 |
1241 | |
1242 | Also see: |
1243 | |
1244 | =over 4 |
1245 | |
c997b287 |
1246 | =item * |
1247 | |
1248 | * |
d1e3b762 |
1249 | |
dc5c060f |
1250 | L<perlos390>, F<README.os390>, F<perlbs2000>, F<README.vmesa>, |
bb462878 |
1251 | L<perlebcdic>. |
c997b287 |
1252 | |
1253 | =item * |
e41182b5 |
1254 | |
1255 | The perl-mvs@perl.org list is for discussion of porting issues as well as |
1256 | general usage issues for all EBCDIC Perls. Send a message body of |
1257 | "subscribe perl-mvs" to majordomo@perl.org. |
1258 | |
c997b287 |
1259 | =item * |
1260 | |
1261 | AS/400 Perl information at |
b1866b2d |
1262 | http://as400.rochester.ibm.com/ |
d1e3b762 |
1263 | as well as on CPAN in the F<ports/> directory. |
e41182b5 |
1264 | |
1265 | =back |
1266 | |
b8099c3d |
1267 | =head2 Acorn RISC OS |
1268 | |
b7df3edc |
1269 | Because Acorns use ASCII with newlines (C<\n>) in text files as C<\012> like |
1270 | Unix, and because Unix filename emulation is turned on by default, |
1271 | most simple scripts will probably work "out of the box". The native |
6ab3f9cb |
1272 | filesystem is modular, and individual filesystems are free to be |
0a47030a |
1273 | case-sensitive or insensitive, and are usually case-preserving. Some |
b7df3edc |
1274 | native filesystems have name length limits, which file and directory |
6ab3f9cb |
1275 | names are silently truncated to fit. Scripts should be aware that the |
1276 | standard filesystem currently has a name length limit of B<10> |
1277 | characters, with up to 77 items in a directory, but other filesystems |
0a47030a |
1278 | may not impose such limitations. |
b8099c3d |
1279 | |
1280 | Native filenames are of the form |
1281 | |
6ab3f9cb |
1282 | Filesystem#Special_Field::DiskName.$.Directory.Directory.File |
dd9f0070 |
1283 | |
b8099c3d |
1284 | where |
1285 | |
1286 | Special_Field is not usually present, but may contain . and $ . |
1287 | Filesystem =~ m|[A-Za-z0-9_]| |
1288 | DsicName =~ m|[A-Za-z0-9_/]| |
1289 | $ represents the root directory |
1290 | . is the path separator |
1291 | @ is the current directory (per filesystem but machine global) |
1292 | ^ is the parent directory |
1293 | Directory and File =~ m|[^\0- "\.\$\%\&:\@\\^\|\177]+| |
1294 | |
1295 | The default filename translation is roughly C<tr|/.|./|;> |
1296 | |
6ab3f9cb |
1297 | Note that C<"ADFS::HardDisk.$.File" ne 'ADFS::HardDisk.$.File'> and that |
0a47030a |
1298 | the second stage of C<$> interpolation in regular expressions will fall |
1299 | foul of the C<$.> if scripts are not careful. |
1300 | |
1301 | Logical paths specified by system variables containing comma-separated |
b7df3edc |
1302 | search lists are also allowed; hence C<System:Modules> is a valid |
0a47030a |
1303 | filename, and the filesystem will prefix C<Modules> with each section of |
6ab3f9cb |
1304 | C<System$Path> until a name is made that points to an object on disk. |
b7df3edc |
1305 | Writing to a new file C<System:Modules> would be allowed only if |
0a47030a |
1306 | C<System$Path> contains a single item list. The filesystem will also |
1307 | expand system variables in filenames if enclosed in angle brackets, so |
c47ff5f1 |
1308 | C<< <System$Dir>.Modules >> would look for the file |
0a47030a |
1309 | S<C<$ENV{'System$Dir'} . 'Modules'>>. The obvious implication of this is |
c47ff5f1 |
1310 | that B<fully qualified filenames can start with C<< <> >>> and should |
0a47030a |
1311 | be protected when C<open> is used for input. |
b8099c3d |
1312 | |
1313 | Because C<.> was in use as a directory separator and filenames could not |
1314 | be assumed to be unique after 10 characters, Acorn implemented the C |
1315 | compiler to strip the trailing C<.c> C<.h> C<.s> and C<.o> suffix from |
1316 | filenames specified in source code and store the respective files in |
b7df3edc |
1317 | subdirectories named after the suffix. Hence files are translated: |
b8099c3d |
1318 | |
1319 | foo.h h.foo |
1320 | C:foo.h C:h.foo (logical path variable) |
1321 | sys/os.h sys.h.os (C compiler groks Unix-speak) |
1322 | 10charname.c c.10charname |
1323 | 10charname.o o.10charname |
1324 | 11charname_.c c.11charname (assuming filesystem truncates at 10) |
1325 | |
1326 | The Unix emulation library's translation of filenames to native assumes |
b7df3edc |
1327 | that this sort of translation is required, and it allows a user-defined list |
1328 | of known suffixes that it will transpose in this fashion. This may |
1329 | seem transparent, but consider that with these rules C<foo/bar/baz.h> |
0a47030a |
1330 | and C<foo/bar/h/baz> both map to C<foo.bar.h.baz>, and that C<readdir> and |
1331 | C<glob> cannot and do not attempt to emulate the reverse mapping. Other |
6ab3f9cb |
1332 | C<.>'s in filenames are translated to C</>. |
0a47030a |
1333 | |
b7df3edc |
1334 | As implied above, the environment accessed through C<%ENV> is global, and |
0a47030a |
1335 | the convention is that program specific environment variables are of the |
6ab3f9cb |
1336 | form C<Program$Name>. Each filesystem maintains a current directory, |
1337 | and the current filesystem's current directory is the B<global> current |
b7df3edc |
1338 | directory. Consequently, sociable programs don't change the current |
1339 | directory but rely on full pathnames, and programs (and Makefiles) cannot |
0a47030a |
1340 | assume that they can spawn a child process which can change the current |
1341 | directory without affecting its parent (and everyone else for that |
1342 | matter). |
1343 | |
b7df3edc |
1344 | Because native operating system filehandles are global and are currently |
1345 | allocated down from 255, with 0 being a reserved value, the Unix emulation |
0a47030a |
1346 | library emulates Unix filehandles. Consequently, you can't rely on |
1347 | passing C<STDIN>, C<STDOUT>, or C<STDERR> to your children. |
1348 | |
1349 | The desire of users to express filenames of the form |
c47ff5f1 |
1350 | C<< <Foo$Dir>.Bar >> on the command line unquoted causes problems, |
0a47030a |
1351 | too: C<``> command output capture has to perform a guessing game. It |
c47ff5f1 |
1352 | assumes that a string C<< <[^<>]+\$[^<>]> >> is a |
0a47030a |
1353 | reference to an environment variable, whereas anything else involving |
c47ff5f1 |
1354 | C<< < >> or C<< > >> is redirection, and generally manages to be 99% |
0a47030a |
1355 | right. Of course, the problem remains that scripts cannot rely on any |
1356 | Unix tools being available, or that any tools found have Unix-like command |
1357 | line arguments. |
1358 | |
b7df3edc |
1359 | Extensions and XS are, in theory, buildable by anyone using free |
1360 | tools. In practice, many don't, as users of the Acorn platform are |
1361 | used to binary distributions. MakeMaker does run, but no available |
1362 | make currently copes with MakeMaker's makefiles; even if and when |
1363 | this should be fixed, the lack of a Unix-like shell will cause |
1364 | problems with makefile rules, especially lines of the form C<cd |
1365 | sdbm && make all>, and anything using quoting. |
b8099c3d |
1366 | |
1367 | "S<RISC OS>" is the proper name for the operating system, but the value |
1368 | in C<$^O> is "riscos" (because we don't like shouting). |
1369 | |
e41182b5 |
1370 | =head2 Other perls |
1371 | |
b7df3edc |
1372 | Perl has been ported to many platforms that do not fit into any of |
1373 | the categories listed above. Some, such as AmigaOS, Atari MiNT, |
1374 | BeOS, HP MPE/iX, QNX, Plan 9, and VOS, have been well-integrated |
1375 | into the standard Perl source code kit. You may need to see the |
1376 | F<ports/> directory on CPAN for information, and possibly binaries, |
1377 | for the likes of: aos, Atari ST, lynxos, riscos, Novell Netware, |
1378 | Tandem Guardian, I<etc.> (Yes, we know that some of these OSes may |
1379 | fall under the Unix category, but we are not a standards body.) |
e41182b5 |
1380 | |
d1e3b762 |
1381 | Some approximate operating system names and their C<$^O> values |
1382 | in the "OTHER" category include: |
1383 | |
1384 | OS $^O $Config{'archname'} |
1385 | ------------------------------------------ |
1386 | Amiga DOS amigaos m68k-amigos |
cec2c193 |
1387 | BeOS beos |
d1e3b762 |
1388 | MPE/iX mpeix PA-RISC1.1 |
1389 | |
e41182b5 |
1390 | See also: |
1391 | |
1392 | =over 4 |
1393 | |
c997b287 |
1394 | =item * |
1395 | |
1396 | Amiga, F<README.amiga> (installed as L<perlamiga>). |
1397 | |
1398 | =item * |
d1e3b762 |
1399 | |
c997b287 |
1400 | Atari, F<README.mint> and Guido Flohr's web page |
1401 | http://stud.uni-sb.de/~gufl0000/ |
e41182b5 |
1402 | |
c997b287 |
1403 | =item * |
d1e3b762 |
1404 | |
c997b287 |
1405 | Be OS, F<README.beos> |
e41182b5 |
1406 | |
c997b287 |
1407 | =item * |
1408 | |
1409 | HP 300 MPE/iX, F<README.mpeix> and Mark Bixby's web page |
34aaaa84 |
1410 | http://www.bixby.org/mark/perlix.html |
c997b287 |
1411 | |
1412 | =item * |
e41182b5 |
1413 | |
6ab3f9cb |
1414 | A free perl5-based PERL.NLM for Novell Netware is available in |
c997b287 |
1415 | precompiled binary and source code form from http://www.novell.com/ |
6ab3f9cb |
1416 | as well as from CPAN. |
e41182b5 |
1417 | |
13a2d996 |
1418 | =item * |
c997b287 |
1419 | |
e6f03d26 |
1420 | S<Plan 9>, F<README.plan9> |
d1e3b762 |
1421 | |
e41182b5 |
1422 | =back |
1423 | |
e41182b5 |
1424 | =head1 FUNCTION IMPLEMENTATIONS |
1425 | |
b7df3edc |
1426 | Listed below are functions that are either completely unimplemented |
1427 | or else have been implemented differently on various platforms. |
1428 | Following each description will be, in parentheses, a list of |
1429 | platforms that the description applies to. |
e41182b5 |
1430 | |
b7df3edc |
1431 | The list may well be incomplete, or even wrong in some places. When |
1432 | in doubt, consult the platform-specific README files in the Perl |
1433 | source distribution, and any other documentation resources accompanying |
1434 | a given port. |
e41182b5 |
1435 | |
0a47030a |
1436 | Be aware, moreover, that even among Unix-ish systems there are variations. |
e41182b5 |
1437 | |
b7df3edc |
1438 | For many functions, you can also query C<%Config>, exported by |
1439 | default from the Config module. For example, to check whether the |
1440 | platform has the C<lstat> call, check C<$Config{d_lstat}>. See |
1441 | L<Config> for a full description of available variables. |
e41182b5 |
1442 | |
1443 | =head2 Alphabetical Listing of Perl Functions |
1444 | |
1445 | =over 8 |
1446 | |
1447 | =item -X FILEHANDLE |
1448 | |
1449 | =item -X EXPR |
1450 | |
1451 | =item -X |
1452 | |
b7df3edc |
1453 | C<-r>, C<-w>, and C<-x> have a limited meaning only; directories |
e41182b5 |
1454 | and applications are executable, and there are no uid/gid |
b7df3edc |
1455 | considerations. C<-o> is not supported. (S<Mac OS>) |
e41182b5 |
1456 | |
b7df3edc |
1457 | C<-r>, C<-w>, C<-x>, and C<-o> tell whether the file is accessible, |
1458 | which may not reflect UIC-based file protections. (VMS) |
e41182b5 |
1459 | |
b8099c3d |
1460 | C<-s> returns the size of the data fork, not the total size of data fork |
1461 | plus resource fork. (S<Mac OS>). |
1462 | |
1463 | C<-s> by name on an open file will return the space reserved on disk, |
1464 | rather than the current extent. C<-s> on an open filehandle returns the |
b7df3edc |
1465 | current size. (S<RISC OS>) |
b8099c3d |
1466 | |
e41182b5 |
1467 | C<-R>, C<-W>, C<-X>, C<-O> are indistinguishable from C<-r>, C<-w>, |
b8099c3d |
1468 | C<-x>, C<-o>. (S<Mac OS>, Win32, VMS, S<RISC OS>) |
e41182b5 |
1469 | |
1470 | C<-b>, C<-c>, C<-k>, C<-g>, C<-p>, C<-u>, C<-A> are not implemented. |
1471 | (S<Mac OS>) |
1472 | |
1473 | C<-g>, C<-k>, C<-l>, C<-p>, C<-u>, C<-A> are not particularly meaningful. |
b8099c3d |
1474 | (Win32, VMS, S<RISC OS>) |
e41182b5 |
1475 | |
1476 | C<-d> is true if passed a device spec without an explicit directory. |
1477 | (VMS) |
1478 | |
1479 | C<-T> and C<-B> are implemented, but might misclassify Mac text files |
0a47030a |
1480 | with foreign characters; this is the case will all platforms, but may |
b7df3edc |
1481 | affect S<Mac OS> often. (S<Mac OS>) |
e41182b5 |
1482 | |
1483 | C<-x> (or C<-X>) determine if a file ends in one of the executable |
b7df3edc |
1484 | suffixes. C<-S> is meaningless. (Win32) |
e41182b5 |
1485 | |
b8099c3d |
1486 | C<-x> (or C<-X>) determine if a file has an executable file type. |
1487 | (S<RISC OS>) |
1488 | |
63f87e49 |
1489 | =item alarm SECONDS |
1490 | |
1491 | =item alarm |
1492 | |
1493 | Not implemented. (Win32) |
1494 | |
e41182b5 |
1495 | =item binmode FILEHANDLE |
1496 | |
b7df3edc |
1497 | Meaningless. (S<Mac OS>, S<RISC OS>) |
e41182b5 |
1498 | |
1499 | Reopens file and restores pointer; if function fails, underlying |
1500 | filehandle may be closed, or pointer may be in a different position. |
1501 | (VMS) |
1502 | |
1503 | The value returned by C<tell> may be affected after the call, and |
1504 | the filehandle may be flushed. (Win32) |
1505 | |
1506 | =item chmod LIST |
1507 | |
b7df3edc |
1508 | Only limited meaning. Disabling/enabling write permission is mapped to |
e41182b5 |
1509 | locking/unlocking the file. (S<Mac OS>) |
1510 | |
1511 | Only good for changing "owner" read-write access, "group", and "other" |
1512 | bits are meaningless. (Win32) |
1513 | |
b8099c3d |
1514 | Only good for changing "owner" and "other" read-write access. (S<RISC OS>) |
1515 | |
495c5fdc |
1516 | Access permissions are mapped onto VOS access-control list changes. (VOS) |
1517 | |
4e51f8e4 |
1518 | The actual permissions set depend on the value of the C<CYGWIN> |
789f0d36 |
1519 | in the SYSTEM environment settings. (Cygwin) |
4e51f8e4 |
1520 | |
e41182b5 |
1521 | =item chown LIST |
1522 | |
3fd80bd6 |
1523 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>) |
e41182b5 |
1524 | |
1525 | Does nothing, but won't fail. (Win32) |
1526 | |
3fd80bd6 |
1527 | A little funky, because VOS's notion of ownership is a little funky (VOS). |
1528 | |
e41182b5 |
1529 | =item chroot FILENAME |
1530 | |
1531 | =item chroot |
1532 | |
e6f03d26 |
1533 | Not implemented. (S<Mac OS>, Win32, VMS, S<Plan 9>, S<RISC OS>, VOS, VM/ESA) |
e41182b5 |
1534 | |
1535 | =item crypt PLAINTEXT,SALT |
1536 | |
1537 | May not be available if library or source was not provided when building |
b8099c3d |
1538 | perl. (Win32) |
e41182b5 |
1539 | |
1540 | =item dbmclose HASH |
1541 | |
e6f03d26 |
1542 | Not implemented. (VMS, S<Plan 9>, VOS) |
e41182b5 |
1543 | |
1544 | =item dbmopen HASH,DBNAME,MODE |
1545 | |
e6f03d26 |
1546 | Not implemented. (VMS, S<Plan 9>, VOS) |
e41182b5 |
1547 | |
1548 | =item dump LABEL |
1549 | |
b8099c3d |
1550 | Not useful. (S<Mac OS>, S<RISC OS>) |
e41182b5 |
1551 | |
1552 | Not implemented. (Win32) |
1553 | |
b8099c3d |
1554 | Invokes VMS debugger. (VMS) |
e41182b5 |
1555 | |
1556 | =item exec LIST |
1557 | |
1558 | Not implemented. (S<Mac OS>) |
1559 | |
7c5ffed3 |
1560 | Implemented via Spawn. (VM/ESA) |
3c075c7d |
1561 | |
0f897271 |
1562 | Does not automatically flush output handles on some platforms. |
1563 | (SunOS, Solaris, HP-UX) |
1564 | |
fe12c0e8 |
1565 | =item exit EXPR |
1566 | |
1567 | =item exit |
1568 | |
1569 | Emulates UNIX exit() (which considers C<exit 1> to indicate an error) by |
1570 | mapping the C<1> to SS$_ABORT (C<44>). This behavior may be overridden |
1571 | with the pragma C<use vmsish 'exit'>. As with the CRTL's exit() |
1572 | function, C<exit 0> is also mapped to an exit status of SS$_NORMAL |
1573 | (C<1>); this mapping cannot be overridden. Any other argument to exit() |
1574 | is used directly as Perl's exit status. (VMS) |
1575 | |
e41182b5 |
1576 | =item fcntl FILEHANDLE,FUNCTION,SCALAR |
1577 | |
1578 | Not implemented. (Win32, VMS) |
1579 | |
1580 | =item flock FILEHANDLE,OPERATION |
1581 | |
495c5fdc |
1582 | Not implemented (S<Mac OS>, VMS, S<RISC OS>, VOS). |
e41182b5 |
1583 | |
1584 | Available only on Windows NT (not on Windows 95). (Win32) |
1585 | |
1586 | =item fork |
1587 | |
3fd80bd6 |
1588 | Not implemented. (S<Mac OS>, AmigaOS, S<RISC OS>, VM/ESA, VMS) |
0f897271 |
1589 | |
1590 | Emulated using multiple interpreters. See L<perlfork>. (Win32) |
1591 | |
1592 | Does not automatically flush output handles on some platforms. |
1593 | (SunOS, Solaris, HP-UX) |
e41182b5 |
1594 | |
1595 | =item getlogin |
1596 | |
b8099c3d |
1597 | Not implemented. (S<Mac OS>, S<RISC OS>) |
e41182b5 |
1598 | |
1599 | =item getpgrp PID |
1600 | |
3fd80bd6 |
1601 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>) |
e41182b5 |
1602 | |
1603 | =item getppid |
1604 | |
41cbbefa |
1605 | Not implemented. (S<Mac OS>, Win32, S<RISC OS>) |
e41182b5 |
1606 | |
1607 | =item getpriority WHICH,WHO |
1608 | |
7c5ffed3 |
1609 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA) |
e41182b5 |
1610 | |
1611 | =item getpwnam NAME |
1612 | |
1613 | Not implemented. (S<Mac OS>, Win32) |
1614 | |
b8099c3d |
1615 | Not useful. (S<RISC OS>) |
1616 | |
e41182b5 |
1617 | =item getgrnam NAME |
1618 | |
b8099c3d |
1619 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>) |
e41182b5 |
1620 | |
1621 | =item getnetbyname NAME |
1622 | |
e6f03d26 |
1623 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) |
e41182b5 |
1624 | |
1625 | =item getpwuid UID |
1626 | |
1627 | Not implemented. (S<Mac OS>, Win32) |
1628 | |
b8099c3d |
1629 | Not useful. (S<RISC OS>) |
1630 | |
e41182b5 |
1631 | =item getgrgid GID |
1632 | |
b8099c3d |
1633 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>) |
e41182b5 |
1634 | |
1635 | =item getnetbyaddr ADDR,ADDRTYPE |
1636 | |
e6f03d26 |
1637 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) |
e41182b5 |
1638 | |
1639 | =item getprotobynumber NUMBER |
1640 | |
1641 | Not implemented. (S<Mac OS>) |
1642 | |
1643 | =item getservbyport PORT,PROTO |
1644 | |
1645 | Not implemented. (S<Mac OS>) |
1646 | |
1647 | =item getpwent |
1648 | |
7c5ffed3 |
1649 | Not implemented. (S<Mac OS>, Win32, VM/ESA) |
e41182b5 |
1650 | |
1651 | =item getgrent |
1652 | |
7c5ffed3 |
1653 | Not implemented. (S<Mac OS>, Win32, VMS, VM/ESA) |
e41182b5 |
1654 | |
ef5a6dd7 |
1655 | =item gethostbyname |
1656 | |
1657 | C<gethostbyname('localhost')> does not work everywhere: you may have |
1658 | to use C<gethostbyname('127.0.0.1')>. (S<Mac OS>, S<Irix 5>) |
1659 | |
e41182b5 |
1660 | =item gethostent |
1661 | |
1662 | Not implemented. (S<Mac OS>, Win32) |
1663 | |
1664 | =item getnetent |
1665 | |
e6f03d26 |
1666 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) |
e41182b5 |
1667 | |
1668 | =item getprotoent |
1669 | |
e6f03d26 |
1670 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) |
e41182b5 |
1671 | |
1672 | =item getservent |
1673 | |
e6f03d26 |
1674 | Not implemented. (Win32, S<Plan 9>) |
e41182b5 |
1675 | |
e41182b5 |
1676 | =item sethostent STAYOPEN |
1677 | |
e6f03d26 |
1678 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>) |
e41182b5 |
1679 | |
1680 | =item setnetent STAYOPEN |
1681 | |
e6f03d26 |
1682 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>) |
e41182b5 |
1683 | |
1684 | =item setprotoent STAYOPEN |
1685 | |
e6f03d26 |
1686 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>, S<RISC OS>) |
e41182b5 |
1687 | |
1688 | =item setservent STAYOPEN |
1689 | |
e6f03d26 |
1690 | Not implemented. (S<Plan 9>, Win32, S<RISC OS>) |
e41182b5 |
1691 | |
1692 | =item endpwent |
1693 | |
a3dfe201 |
1694 | Not implemented. (S<Mac OS>, MPE/iX, VM/ESA, Win32) |
e41182b5 |
1695 | |
1696 | =item endgrent |
1697 | |
a3dfe201 |
1698 | Not implemented. (S<Mac OS>, MPE/iX, S<RISC OS>, VM/ESA, VMS, Win32) |
e41182b5 |
1699 | |
1700 | =item endhostent |
1701 | |
1702 | Not implemented. (S<Mac OS>, Win32) |
1703 | |
1704 | =item endnetent |
1705 | |
e6f03d26 |
1706 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) |
e41182b5 |
1707 | |
1708 | =item endprotoent |
1709 | |
e6f03d26 |
1710 | Not implemented. (S<Mac OS>, Win32, S<Plan 9>) |
e41182b5 |
1711 | |
1712 | =item endservent |
1713 | |
e6f03d26 |
1714 | Not implemented. (S<Plan 9>, Win32) |
e41182b5 |
1715 | |
1716 | =item getsockopt SOCKET,LEVEL,OPTNAME |
1717 | |
e6f03d26 |
1718 | Not implemented. (S<Plan 9>) |
e41182b5 |
1719 | |
1720 | =item glob EXPR |
1721 | |
1722 | =item glob |
1723 | |
63f87e49 |
1724 | This operator is implemented via the File::Glob extension on most |
1725 | platforms. See L<File::Glob> for portability information. |
b8099c3d |
1726 | |
e41182b5 |
1727 | =item ioctl FILEHANDLE,FUNCTION,SCALAR |
1728 | |
1729 | Not implemented. (VMS) |
1730 | |
1731 | Available only for socket handles, and it does what the ioctlsocket() call |
1732 | in the Winsock API does. (Win32) |
1733 | |
b8099c3d |
1734 | Available only for socket handles. (S<RISC OS>) |
1735 | |
b350dd2f |
1736 | =item kill SIGNAL, LIST |
e41182b5 |
1737 | |
862b5365 |
1738 | C<kill(0, LIST)> is implemented for the sake of taint checking; |
1739 | use with other signals is unimplemented. (S<Mac OS>) |
1740 | |
1741 | Not implemented, hence not useful for taint checking. (S<RISC OS>) |
e41182b5 |
1742 | |
63f87e49 |
1743 | C<kill()> doesn't have the semantics of C<raise()>, i.e. it doesn't send |
1744 | a signal to the identified process like it does on Unix platforms. |
1745 | Instead C<kill($sig, $pid)> terminates the process identified by $pid, |
1746 | and makes it exit immediately with exit status $sig. As in Unix, if |
1747 | $sig is 0 and the specified process exists, it returns true without |
1748 | actually terminating it. (Win32) |
e41182b5 |
1749 | |
1750 | =item link OLDFILE,NEWFILE |
1751 | |
a3dfe201 |
1752 | Not implemented. (S<Mac OS>, MPE/iX, VMS, S<RISC OS>) |
e41182b5 |
1753 | |
433acd8a |
1754 | Link count not updated because hard links are not quite that hard |
1755 | (They are sort of half-way between hard and soft links). (AmigaOS) |
1756 | |
a3dfe201 |
1757 | Hard links are implemented on Win32 (Windows NT and Windows 2000) |
1758 | under NTFS only. |
1759 | |
e41182b5 |
1760 | =item lstat FILEHANDLE |
1761 | |
1762 | =item lstat EXPR |
1763 | |
1764 | =item lstat |
1765 | |
b8099c3d |
1766 | Not implemented. (VMS, S<RISC OS>) |
e41182b5 |
1767 | |
63f87e49 |
1768 | Return values (especially for device and inode) may be bogus. (Win32) |
e41182b5 |
1769 | |
1770 | =item msgctl ID,CMD,ARG |
1771 | |
1772 | =item msgget KEY,FLAGS |
1773 | |
1774 | =item msgsnd ID,MSG,FLAGS |
1775 | |
1776 | =item msgrcv ID,VAR,SIZE,TYPE,FLAGS |
1777 | |
e6f03d26 |
1778 | Not implemented. (S<Mac OS>, Win32, VMS, S<Plan 9>, S<RISC OS>, VOS) |
e41182b5 |
1779 | |
1780 | =item open FILEHANDLE,EXPR |
1781 | |
1782 | =item open FILEHANDLE |
1783 | |
b7df3edc |
1784 | The C<|> variants are supported only if ToolServer is installed. |
e41182b5 |
1785 | (S<Mac OS>) |
1786 | |
c47ff5f1 |
1787 | open to C<|-> and C<-|> are unsupported. (S<Mac OS>, Win32, S<RISC OS>) |
e41182b5 |
1788 | |
0f897271 |
1789 | Opening a process does not automatically flush output handles on some |
1790 | platforms. (SunOS, Solaris, HP-UX) |
1791 | |
e41182b5 |
1792 | =item pipe READHANDLE,WRITEHANDLE |
1793 | |
433acd8a |
1794 | Very limited functionality. (MiNT) |
1795 | |
e41182b5 |
1796 | =item readlink EXPR |
1797 | |
1798 | =item readlink |
1799 | |
b8099c3d |
1800 | Not implemented. (Win32, VMS, S<RISC OS>) |
e41182b5 |
1801 | |
1802 | =item select RBITS,WBITS,EBITS,TIMEOUT |
1803 | |
689c5c24 |
1804 | Only implemented on sockets. (Win32, VMS) |
e41182b5 |
1805 | |
b8099c3d |
1806 | Only reliable on sockets. (S<RISC OS>) |
1807 | |
76e05f0b |
1808 | Note that the C<select FILEHANDLE> form is generally portable. |
63f87e49 |
1809 | |
e41182b5 |
1810 | =item semctl ID,SEMNUM,CMD,ARG |
1811 | |
1812 | =item semget KEY,NSEMS,FLAGS |
1813 | |
1814 | =item semop KEY,OPSTRING |
1815 | |
495c5fdc |
1816 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS) |
e41182b5 |
1817 | |
a3dfe201 |
1818 | =item setgrent |
1819 | |
3fd80bd6 |
1820 | Not implemented. (S<Mac OS>, MPE/iX, VMS, Win32, S<RISC OS>, VOS) |
a3dfe201 |
1821 | |
e41182b5 |
1822 | =item setpgrp PID,PGRP |
1823 | |
495c5fdc |
1824 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS) |
e41182b5 |
1825 | |
1826 | =item setpriority WHICH,WHO,PRIORITY |
1827 | |
495c5fdc |
1828 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS) |
e41182b5 |
1829 | |
a3dfe201 |
1830 | =item setpwent |
1831 | |
3fd80bd6 |
1832 | Not implemented. (S<Mac OS>, MPE/iX, Win32, S<RISC OS>, VOS) |
a3dfe201 |
1833 | |
e41182b5 |
1834 | =item setsockopt SOCKET,LEVEL,OPTNAME,OPTVAL |
1835 | |
e6f03d26 |
1836 | Not implemented. (S<Plan 9>) |
e41182b5 |
1837 | |
1838 | =item shmctl ID,CMD,ARG |
1839 | |
1840 | =item shmget KEY,SIZE,FLAGS |
1841 | |
1842 | =item shmread ID,VAR,POS,SIZE |
1843 | |
1844 | =item shmwrite ID,STRING,POS,SIZE |
1845 | |
495c5fdc |
1846 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS) |
e41182b5 |
1847 | |
80cbd5ad |
1848 | =item sockatmark SOCKET |
1849 | |
1850 | A relatively recent addition to socket functions, may not |
1851 | be implemented even in UNIX platforms. |
1852 | |
e41182b5 |
1853 | =item socketpair SOCKET1,SOCKET2,DOMAIN,TYPE,PROTOCOL |
1854 | |
862b5365 |
1855 | Not implemented. (Win32, VMS, S<RISC OS>, VOS, VM/ESA) |
e41182b5 |
1856 | |
1857 | =item stat FILEHANDLE |
1858 | |
1859 | =item stat EXPR |
1860 | |
1861 | =item stat |
1862 | |
d62e1b7f |
1863 | Platforms that do not have rdev, blksize, or blocks will return these |
1864 | as '', so numeric comparison or manipulation of these fields may cause |
1865 | 'not numeric' warnings. |
1866 | |
e41182b5 |
1867 | mtime and atime are the same thing, and ctime is creation time instead of |
3f1f789b |
1868 | inode change time. (S<Mac OS>). |
1869 | |
1870 | ctime not supported on UFS (S<Mac OS X>). |
e41182b5 |
1871 | |
95a3fe12 |
1872 | ctime is creation time instead of inode change time (Win32). |
1873 | |
e41182b5 |
1874 | device and inode are not meaningful. (Win32) |
1875 | |
1876 | device and inode are not necessarily reliable. (VMS) |
1877 | |
b8099c3d |
1878 | mtime, atime and ctime all return the last modification time. Device and |
1879 | inode are not necessarily reliable. (S<RISC OS>) |
1880 | |
d62e1b7f |
1881 | dev, rdev, blksize, and blocks are not available. inode is not |
1882 | meaningful and will differ between stat calls on the same file. (os2) |
1883 | |
73e9292c |
1884 | some versions of cygwin when doing a stat("foo") and if not finding it |
1885 | may then attempt to stat("foo.exe") (Cygwin) |
1886 | |
e41182b5 |
1887 | =item symlink OLDFILE,NEWFILE |
1888 | |
b8099c3d |
1889 | Not implemented. (Win32, VMS, S<RISC OS>) |
e41182b5 |
1890 | |
1891 | =item syscall LIST |
1892 | |
7c5ffed3 |
1893 | Not implemented. (S<Mac OS>, Win32, VMS, S<RISC OS>, VOS, VM/ESA) |
e41182b5 |
1894 | |
f34d0673 |
1895 | =item sysopen FILEHANDLE,FILENAME,MODE,PERMS |
1896 | |
dd9f0070 |
1897 | The traditional "0", "1", and "2" MODEs are implemented with different |
322422de |
1898 | numeric values on some systems. The flags exported by C<Fcntl> |
1899 | (O_RDONLY, O_WRONLY, O_RDWR) should work everywhere though. (S<Mac |
7c5ffed3 |
1900 | OS>, OS/390, VM/ESA) |
f34d0673 |
1901 | |
e41182b5 |
1902 | =item system LIST |
1903 | |
9d6eb86e |
1904 | In general, do not assume the UNIX/POSIX semantics that you can shift |
7717d0e7 |
1905 | C<$?> right by eight to get the exit value, or that C<$? & 127> |
9d6eb86e |
1906 | would give you the number of the signal that terminated the program, |
1907 | or that C<$? & 128> would test true if the program was terminated by a |
1908 | coredump. Instead, use the POSIX W*() interfaces: for example, use |
74555b7a |
1909 | WIFEXITED($?) and WEXITVALUE($?) to test for a normal exit and the exit |
1910 | value, WIFSIGNALED($?) and WTERMSIG($?) for a signal exit and the |
7717d0e7 |
1911 | signal. Core dumping is not a portable concept, so there's no portable |
9d6eb86e |
1912 | way to test for that. |
1913 | |
e41182b5 |
1914 | Only implemented if ToolServer is installed. (S<Mac OS>) |
1915 | |
1916 | As an optimization, may not call the command shell specified in |
b7df3edc |
1917 | C<$ENV{PERL5SHELL}>. C<system(1, @args)> spawns an external |
e41182b5 |
1918 | process and immediately returns its process designator, without |
1919 | waiting for it to terminate. Return value may be used subsequently |
63f87e49 |
1920 | in C<wait> or C<waitpid>. Failure to spawn() a subprocess is indicated |
1921 | by setting $? to "255 << 8". C<$?> is set in a way compatible with |
1922 | Unix (i.e. the exitstatus of the subprocess is obtained by "$? >> 8", |
1923 | as described in the documentation). (Win32) |
e41182b5 |
1924 | |
b8099c3d |
1925 | There is no shell to process metacharacters, and the native standard is |
1926 | to pass a command line terminated by "\n" "\r" or "\0" to the spawned |
c47ff5f1 |
1927 | program. Redirection such as C<< > foo >> is performed (if at all) by |
b8099c3d |
1928 | the run time library of the spawned program. C<system> I<list> will call |
1929 | the Unix emulation library's C<exec> emulation, which attempts to provide |
1930 | emulation of the stdin, stdout, stderr in force in the parent, providing |
1931 | the child program uses a compatible version of the emulation library. |
1932 | I<scalar> will call the native command line direct and no such emulation |
1933 | of a child Unix program will exists. Mileage B<will> vary. (S<RISC OS>) |
1934 | |
433acd8a |
1935 | Far from being POSIX compliant. Because there may be no underlying |
1936 | /bin/sh tries to work around the problem by forking and execing the |
9b63e9ec |
1937 | first token in its argument string. Handles basic redirection |
c47ff5f1 |
1938 | ("<" or ">") on its own behalf. (MiNT) |
433acd8a |
1939 | |
0f897271 |
1940 | Does not automatically flush output handles on some platforms. |
1941 | (SunOS, Solaris, HP-UX) |
1942 | |
9bc98430 |
1943 | The return value is POSIX-like (shifted up by 8 bits), which only allows |
1944 | room for a made-up value derived from the severity bits of the native |
1945 | 32-bit condition code (unless overridden by C<use vmsish 'status'>). |
1946 | For more details see L<perlvms/$?>. (VMS) |
1947 | |
e41182b5 |
1948 | =item times |
1949 | |
1950 | Only the first entry returned is nonzero. (S<Mac OS>) |
1951 | |
63f87e49 |
1952 | "cumulative" times will be bogus. On anything other than Windows NT |
1953 | or Windows 2000, "system" time will be bogus, and "user" time is |
1954 | actually the time returned by the clock() function in the C runtime |
1955 | library. (Win32) |
e41182b5 |
1956 | |
b8099c3d |
1957 | Not useful. (S<RISC OS>) |
1958 | |
e41182b5 |
1959 | =item truncate FILEHANDLE,LENGTH |
1960 | |
1961 | =item truncate EXPR,LENGTH |
1962 | |
6d738113 |
1963 | Not implemented. (Older versions of VMS) |
e41182b5 |
1964 | |
3fd80bd6 |
1965 | Truncation to same-or-shorter lengths only. (VOS) |
495c5fdc |
1966 | |
4cfdb94f |
1967 | If a FILEHANDLE is supplied, it must be writable and opened in append |
e71a7dc8 |
1968 | mode (i.e., use C<<< open(FH, '>>filename') >>> |
4cfdb94f |
1969 | or C<sysopen(FH,...,O_APPEND|O_RDWR)>. If a filename is supplied, it |
1970 | should not be held open elsewhere. (Win32) |
1971 | |
e41182b5 |
1972 | =item umask EXPR |
1973 | |
1974 | =item umask |
1975 | |
1976 | Returns undef where unavailable, as of version 5.005. |
1977 | |
b7df3edc |
1978 | C<umask> works but the correct permissions are set only when the file |
1979 | is finally closed. (AmigaOS) |
433acd8a |
1980 | |
e41182b5 |
1981 | =item utime LIST |
1982 | |
15c65113 |
1983 | Only the modification time is updated. (S<BeOS>, S<Mac OS>, VMS, S<RISC OS>) |
e41182b5 |
1984 | |
322422de |
1985 | May not behave as expected. Behavior depends on the C runtime |
1986 | library's implementation of utime(), and the filesystem being |
1987 | used. The FAT filesystem typically does not support an "access |
1988 | time" field, and it may limit timestamps to a granularity of |
1989 | two seconds. (Win32) |
e41182b5 |
1990 | |
1991 | =item wait |
1992 | |
1993 | =item waitpid PID,FLAGS |
1994 | |
3fd80bd6 |
1995 | Not implemented. (S<Mac OS>) |
e41182b5 |
1996 | |
1997 | Can only be applied to process handles returned for processes spawned |
a6f858fb |
1998 | using C<system(1, ...)> or pseudo processes created with C<fork()>. (Win32) |
e41182b5 |
1999 | |
b8099c3d |
2000 | Not useful. (S<RISC OS>) |
2001 | |
e41182b5 |
2002 | =back |
2003 | |
b8099c3d |
2004 | =head1 CHANGES |
2005 | |
2006 | =over 4 |
2007 | |
3fd80bd6 |
2008 | =item v1.49, 12 August 2002 |
2009 | |
2010 | Updates for VOS from Paul Green. |
2011 | |
fd46a41b |
2012 | =item v1.48, 02 February 2001 |
2013 | |
2014 | Various updates from perl5-porters over the past year, supported |
2015 | platforms update from Jarkko Hietaniemi. |
2016 | |
c997b287 |
2017 | =item v1.47, 22 March 2000 |
2018 | |
2019 | Various cleanups from Tom Christiansen, including migration of |
2020 | long platform listings from L<perl>. |
2021 | |
56d7751a |
2022 | =item v1.46, 12 February 2000 |
2023 | |
2024 | Updates for VOS and MPE/iX. (Peter Prymmer) Other small changes. |
2025 | |
0cc436d0 |
2026 | =item v1.45, 20 December 1999 |
2027 | |
2028 | Small changes from 5.005_63 distribution, more changes to EBCDIC info. |
2029 | |
d1e3b762 |
2030 | =item v1.44, 19 July 1999 |
2031 | |
2032 | A bunch of updates from Peter Prymmer for C<$^O> values, |
2033 | endianness, File::Spec, VMS, BS2000, OS/400. |
2034 | |
b7df3edc |
2035 | =item v1.43, 24 May 1999 |
2036 | |
2037 | Added a lot of cleaning up from Tom Christiansen. |
2038 | |
19799a22 |
2039 | =item v1.42, 22 May 1999 |
b7df3edc |
2040 | |
19799a22 |
2041 | Added notes about tests, sprintf/printf, and epoch offsets. |
b7df3edc |
2042 | |
6ab3f9cb |
2043 | =item v1.41, 19 May 1999 |
2044 | |
2045 | Lots more little changes to formatting and content. |
2046 | |
d1e3b762 |
2047 | Added a bunch of C<$^O> and related values |
6ab3f9cb |
2048 | for various platforms; fixed mail and web addresses, and added |
2049 | and changed miscellaneous notes. (Peter Prymmer) |
2050 | |
2051 | =item v1.40, 11 April 1999 |
2052 | |
2053 | Miscellaneous changes. |
2054 | |
2055 | =item v1.39, 11 February 1999 |
2ee0eb3c |
2056 | |
2057 | Changes from Jarkko and EMX URL fixes Michael Schwern. Additional |
2058 | note about newlines added. |
2059 | |
9b63e9ec |
2060 | =item v1.38, 31 December 1998 |
2061 | |
2062 | More changes from Jarkko. |
2063 | |
3c075c7d |
2064 | =item v1.37, 19 December 1998 |
2065 | |
2066 | More minor changes. Merge two separate version 1.35 documents. |
2067 | |
2068 | =item v1.36, 9 September 1998 |
2069 | |
2070 | Updated for Stratus VOS. Also known as version 1.35. |
2071 | |
2072 | =item v1.35, 13 August 1998 |
495c5fdc |
2073 | |
3c075c7d |
2074 | Integrate more minor changes, plus addition of new sections under |
2075 | L<"ISSUES">: L<"Numbers endianness and Width">, |
2076 | L<"Character sets and character encoding">, |
2077 | L<"Internationalisation">. |
495c5fdc |
2078 | |
3c075c7d |
2079 | =item v1.33, 06 August 1998 |
0a47030a |
2080 | |
2081 | Integrate more minor changes. |
2082 | |
3c075c7d |
2083 | =item v1.32, 05 August 1998 |
dd9f0070 |
2084 | |
2085 | Integrate more minor changes. |
2086 | |
3c075c7d |
2087 | =item v1.30, 03 August 1998 |
b8099c3d |
2088 | |
2089 | Major update for RISC OS, other minor changes. |
2090 | |
3c075c7d |
2091 | =item v1.23, 10 July 1998 |
b8099c3d |
2092 | |
2093 | First public release with perl5.005. |
2094 | |
2095 | =back |
e41182b5 |
2096 | |
ba58ab26 |
2097 | =head1 Supported Platforms |
2098 | |
522b859a |
2099 | As of July 2002 (the Perl release 5.8.0), the following platforms are |
cec2c193 |
2100 | able to build Perl from the standard source code distribution |
2101 | available at http://www.cpan.org/src/index.html |
2102 | |
2103 | AIX |
2104 | BeOS |
2105 | Cygwin |
2106 | DG/UX |
811b48f2 |
2107 | DOS DJGPP 1) |
cec2c193 |
2108 | DYNIX/ptx |
2109 | EPOC R5 |
2110 | FreeBSD |
2111 | HP-UX |
2112 | IRIX |
2113 | Linux |
8939ba94 |
2114 | Mac OS Classic |
2115 | Mac OS X (Darwin) |
cec2c193 |
2116 | MPE/iX |
2117 | NetBSD |
2118 | NetWare |
2119 | NonStop-UX |
2120 | ReliantUNIX (SINIX) |
2121 | OpenBSD |
2122 | OpenVMS (VMS) |
2123 | OS/2 |
522b859a |
2124 | OS/400 (using the PASE) (since Perl 5.8.1/5.9.0) |
70de81db |
2125 | PowerUX |
cec2c193 |
2126 | POSIX-BC (BS2000) |
2127 | QNX |
2128 | Solaris |
70de81db |
2129 | SunOS 4 |
bb5ad0af |
2130 | SUPER-UX |
cec2c193 |
2131 | Tru64 UNIX (DEC OSF/1, Digital UNIX) |
2132 | UNICOS |
2133 | UNICOS/mk |
2134 | UTS |
2135 | VOS |
811b48f2 |
2136 | Win95/98/ME/2K/XP 2) |
c40b5d1d |
2137 | WinCE |
cec2c193 |
2138 | z/OS (OS/390) |
2139 | VM/ESA |
ba58ab26 |
2140 | |
811b48f2 |
2141 | 1) in DOS mode either the DOS or OS/2 ports can be used |
2142 | 2) compilers: Borland, MinGW (GCC), VC6 |
cec2c193 |
2143 | |
c40b5d1d |
2144 | The following platforms worked with the previous releases (5.6 and |
cec2c193 |
2145 | 5.7), but we did not manage either to fix or to test these in time |
2146 | for the 5.8.0 release. There is a very good chance that many of these |
70de81db |
2147 | will work fine with the 5.8.0. |
cec2c193 |
2148 | |
8da2b1be |
2149 | BSD/OS |
cec2c193 |
2150 | DomainOS |
2151 | Hurd |
2152 | LynxOS |
2153 | MachTen |
2154 | PowerMAX |
2155 | SCO SV |
cec2c193 |
2156 | SVR4 |
2157 | Unixware |
2158 | Windows 3.1 |
ba58ab26 |
2159 | |
70de81db |
2160 | Known to be broken for 5.8.0 (but 5.6.1 and 5.7.2 can be used): |
2161 | |
2162 | AmigaOS |
2163 | |
ba58ab26 |
2164 | The following platforms have been known to build Perl from source in |
fd46a41b |
2165 | the past (5.005_03 and earlier), but we haven't been able to verify |
2166 | their status for the current release, either because the |
2167 | hardware/software platforms are rare or because we don't have an |
2168 | active champion on these platforms--or both. They used to work, |
2169 | though, so go ahead and try compiling them, and let perlbug@perl.org |
2170 | of any trouble. |
ba58ab26 |
2171 | |
cec2c193 |
2172 | 3b1 |
2173 | A/UX |
cec2c193 |
2174 | ConvexOS |
2175 | CX/UX |
2176 | DC/OSx |
2177 | DDE SMES |
2178 | DOS EMX |
2179 | Dynix |
2180 | EP/IX |
2181 | ESIX |
2182 | FPS |
2183 | GENIX |
2184 | Greenhills |
2185 | ISC |
2186 | MachTen 68k |
2187 | MiNT |
2188 | MPC |
2189 | NEWS-OS |
2190 | NextSTEP |
2191 | OpenSTEP |
2192 | Opus |
2193 | Plan 9 |
cec2c193 |
2194 | RISC/os |
8da2b1be |
2195 | SCO ODT/OSR |
cec2c193 |
2196 | Stellar |
2197 | SVR2 |
2198 | TI1500 |
2199 | TitanOS |
2200 | Ultrix |
2201 | Unisys Dynix |
ba58ab26 |
2202 | |
2203 | The following platforms have their own source code distributions and |
1577cd80 |
2204 | binaries available via http://www.cpan.org/ports/ |
ba58ab26 |
2205 | |
cec2c193 |
2206 | Perl release |
ba58ab26 |
2207 | |
522b859a |
2208 | OS/400 (ILE) 5.005_02 |
cec2c193 |
2209 | Tandem Guardian 5.004 |
ba58ab26 |
2210 | |
2211 | The following platforms have only binaries available via |
a93751fa |
2212 | http://www.cpan.org/ports/index.html : |
ba58ab26 |
2213 | |
cec2c193 |
2214 | Perl release |
ba58ab26 |
2215 | |
cec2c193 |
2216 | Acorn RISCOS 5.005_02 |
2217 | AOS 5.002 |
2218 | LynxOS 5.004_02 |
ba58ab26 |
2219 | |
2220 | Although we do suggest that you always build your own Perl from |
2221 | the source code, both for maximal configurability and for security, |
2222 | in case you are in a hurry you can check |
a93751fa |
2223 | http://www.cpan.org/ports/index.html for binary distributions. |
ba58ab26 |
2224 | |
c997b287 |
2225 | =head1 SEE ALSO |
2226 | |
cec2c193 |
2227 | L<perlaix>, L<perlamiga>, L<perlapollo>, L<perlbeos>, L<perlbs2000>, |
18a271bd |
2228 | L<perlce>, L<perlcygwin>, L<perldgux>, L<perldos>, L<perlepoc>, |
469e7be4 |
2229 | L<perlebcdic>, L<perlfreebsd>, L<perlhurd>, L<perlhpux>, L<perlirix>, |
18a271bd |
2230 | L<perlmachten>, L<perlmacos>, L<perlmint>, L<perlmpeix>, |
522b859a |
2231 | L<perlnetware>, L<perlos2>, L<perlos390>, L<perlos400>, |
2232 | L<perlplan9>, L<perlqnx>, L<perlsolaris>, L<perltru64>, |
2233 | L<perlunicode>, L<perlvmesa>, L<perlvms>, L<perlvos>, |
2234 | L<perlwin32>, and L<Win32>. |
c997b287 |
2235 | |
e41182b5 |
2236 | =head1 AUTHORS / CONTRIBUTORS |
2237 | |
06e9666b |
2238 | Abigail <abigail@foad.org>, |
c47ff5f1 |
2239 | Charles Bailey <bailey@newman.upenn.edu>, |
2240 | Graham Barr <gbarr@pobox.com>, |
2241 | Tom Christiansen <tchrist@perl.com>, |
06e9666b |
2242 | Nicholas Clark <nick@ccl4.org>, |
c47ff5f1 |
2243 | Thomas Dorner <Thomas.Dorner@start.de>, |
06e9666b |
2244 | Andy Dougherty <doughera@lafayette.edu>, |
2245 | Dominic Dunlop <domo@computer.org>, |
2246 | Neale Ferguson <neale@vma.tabnsw.com.au>, |
c47ff5f1 |
2247 | David J. Fiander <davidf@mks.com>, |
3fd80bd6 |
2248 | Paul Green <Paul.Green@stratus.com>, |
06e9666b |
2249 | M.J.T. Guy <mjtg@cam.ac.uk>, |
61f30a5e |
2250 | Jarkko Hietaniemi <jhi@iki.fi>, |
c47ff5f1 |
2251 | Luther Huffman <lutherh@stratcom.com>, |
06e9666b |
2252 | Nick Ing-Simmons <nick@ing-simmons.net>, |
2253 | Andreas J. KE<ouml>nig <a.koenig@mind.de>, |
c47ff5f1 |
2254 | Markus Laker <mlaker@contax.co.uk>, |
2255 | Andrew M. Langmead <aml@world.std.com>, |
2256 | Larry Moore <ljmoore@freespace.net>, |
2257 | Paul Moore <Paul.Moore@uk.origin-it.com>, |
2258 | Chris Nandor <pudge@pobox.com>, |
1afc07ec |
2259 | Matthias Neeracher <neeracher@mac.com>, |
e71a7dc8 |
2260 | Philip Newton <pne@cpan.org>, |
c47ff5f1 |
2261 | Gary Ng <71564.1743@CompuServe.COM>, |
2262 | Tom Phoenix <rootbeer@teleport.com>, |
2263 | AndrE<eacute> Pirard <A.Pirard@ulg.ac.be>, |
2264 | Peter Prymmer <pvhp@forte.com>, |
2265 | Hugo van der Sanden <hv@crypt0.demon.co.uk>, |
2266 | Gurusamy Sarathy <gsar@activestate.com>, |
2267 | Paul J. Schinder <schinder@pobox.com>, |
2268 | Michael G Schwern <schwern@pobox.com>, |
06e9666b |
2269 | Dan Sugalski <dan@sidhe.org>, |
c47ff5f1 |
2270 | Nathan Torkington <gnat@frii.com>. |
e41182b5 |
2271 | |