3 perlfaq5 - Files and Formats ($Revision: 1.9 $, $Date: 2002/02/11 19:30:21 $)
7 This section deals with I/O and the "f" issues: filehandles, flushing,
10 =head2 How do I flush/unbuffer an output filehandle? Why must I do this?
12 The C standard I/O library (stdio) normally buffers characters sent to
13 devices. This is done for efficiency reasons so that there isn't a
14 system call for each byte. Any time you use print() or write() in
15 Perl, you go though this buffering. syswrite() circumvents stdio and
18 In most stdio implementations, the type of output buffering and the size of
19 the buffer varies according to the type of device. Disk files are block
20 buffered, often with a buffer size of more than 2k. Pipes and sockets
21 are often buffered with a buffer size between 1/2 and 2k. Serial devices
22 (e.g. modems, terminals) are normally line-buffered, and stdio sends
23 the entire line when it gets the newline.
25 Perl does not support truly unbuffered output (except insofar as you can
26 C<syswrite(OUT, $char, 1)>). What it does instead support is "command
27 buffering", in which a physical write is performed after every output
28 command. This isn't as hard on your system as unbuffering, but does
29 get the output where you want it when you want it.
31 If you expect characters to get to your device when you print them there,
32 you'll want to autoflush its handle.
33 Use select() and the C<$|> variable to control autoflushing
34 (see L<perlvar/$|> and L<perlfunc/select>):
36 $old_fh = select(OUTPUT_HANDLE);
40 Or using the traditional idiom:
42 select((select(OUTPUT_HANDLE), $| = 1)[0]);
44 Or if don't mind slowly loading several thousand lines of module code
45 just because you're afraid of the C<$|> variable:
48 open(DEV, "+</dev/tty"); # ceci n'est pas une pipe
51 or the newer IO::* modules:
54 open(DEV, ">/dev/printer"); # but is this?
59 use IO::Socket; # this one is kinda a pipe?
60 $sock = IO::Socket::INET->new(PeerAddr => 'www.perl.com',
61 PeerPort => 'http(80)',
63 die "$!" unless $sock;
66 print $sock "GET / HTTP/1.0" . "\015\012" x 2;
67 $document = join('', <$sock>);
68 print "DOC IS: $document\n";
70 Note the bizarrely hard coded carriage return and newline in their octal
71 equivalents. This is the ONLY way (currently) to assure a proper flush
72 on all platforms, including Macintosh. That's the way things work in
73 network programming: you really should specify the exact bit pattern
74 on the network line terminator. In practice, C<"\n\n"> often works,
75 but this is not portable.
77 See L<perlfaq9> for other examples of fetching URLs over the web.
79 =head2 How do I change one line in a file/delete a line in a file/insert a line in the middle of a file/append to the beginning of a file?
81 Those are operations of a text editor. Perl is not a text editor.
82 Perl is a programming language. You have to decompose the problem into
83 low-level calls to read, write, open, close, and seek.
85 Although humans have an easy time thinking of a text file as being a
86 sequence of lines that operates much like a stack of playing cards--or
87 punch cards--computers usually see the text file as a sequence of bytes.
88 In general, there's no direct way for Perl to seek to a particular line
89 of a file, insert text into a file, or remove text from a file.
91 (There are exceptions in special circumstances. You can add or remove
92 data at the very end of the file. A sequence of bytes can be replaced
93 with another sequence of the same length. The C<$DB_RECNO> array
94 bindings as documented in L<DB_File> also provide a direct way of
95 modifying a file. Files where all lines are the same length are also
98 The general solution is to create a temporary copy of the text file with
99 the changes you want, then copy that over the original. This assumes
103 $new = "$file.tmp.$$";
106 open(OLD, "< $old") or die "can't open $old: $!";
107 open(NEW, "> $new") or die "can't open $new: $!";
109 # Correct typos, preserving case
111 s/\b(p)earl\b/${1}erl/i;
112 (print NEW $_) or die "can't write to $new: $!";
115 close(OLD) or die "can't close $old: $!";
116 close(NEW) or die "can't close $new: $!";
118 rename($old, $bak) or die "can't rename $old to $bak: $!";
119 rename($new, $old) or die "can't rename $new to $old: $!";
121 Perl can do this sort of thing for you automatically with the C<-i>
122 command-line switch or the closely-related C<$^I> variable (see
123 L<perlrun> for more details). Note that
124 C<-i> may require a suffix on some non-Unix systems; see the
125 platform-specific documentation that came with your port.
127 # Renumber a series of tests from the command line
128 perl -pi -e 's/(^\s+test\s+)\d+/ $1 . ++$count /e' t/op/taint.t
131 local($^I, @ARGV) = ('.orig', glob("*.c"));
134 print "This line should appear at the top of each file\n";
136 s/\b(p)earl\b/${1}erl/i; # Correct typos, preserving case
138 close ARGV if eof; # Reset $.
141 If you need to seek to an arbitrary line of a file that changes
142 infrequently, you could build up an index of byte positions of where
143 the line ends are in the file. If the file is large, an index of
144 every tenth or hundredth line end would allow you to seek and read
145 fairly efficiently. If the file is sorted, try the look.pl library
146 (part of the standard perl distribution).
148 In the unique case of deleting lines at the end of a file, you
149 can use tell() and truncate(). The following code snippet deletes
150 the last line of a file without making a copy or reading the
151 whole file into memory:
153 open (FH, "+< $file");
154 while ( <FH> ) { $addr = tell(FH) unless eof(FH) }
157 Error checking is left as an exercise for the reader.
159 =head2 How do I count the number of lines in a file?
161 One fairly efficient way is to count newlines in the file. The
162 following program uses a feature of tr///, as documented in L<perlop>.
163 If your text file doesn't end with a newline, then it's not really a
164 proper text file, so this may report one fewer line than you expect.
167 open(FILE, $filename) or die "Can't open `$filename': $!";
168 while (sysread FILE, $buffer, 4096) {
169 $lines += ($buffer =~ tr/\n//);
173 This assumes no funny games with newline translations.
175 =head2 How do I make a temporary file name?
177 Use the File::Temp module, see L<File::Temp> for more information.
179 use File::Temp qw/ tempfile tempdir /;
181 $dir = tempdir( CLEANUP => 1 );
182 ($fh, $filename) = tempfile( DIR => $dir );
184 # or if you don't need to know the filename
186 $fh = tempfile( DIR => $dir );
188 The File::Temp has been a standard module since Perl 5.6.1. If you
189 don't have a modern enough Perl installed, use the C<new_tmpfile>
190 class method from the IO::File module to get a filehandle opened for
191 reading and writing. Use it if you don't need to know the file's name:
194 $fh = IO::File->new_tmpfile()
195 or die "Unable to make new temporary file: $!";
197 If you're committed to creating a temporary file by hand, use the
198 process ID and/or the current time-value. If you need to have many
199 temporary files in one process, use a counter:
203 my $temp_dir = -d '/tmp' ? '/tmp' : $ENV{TMPDIR} || $ENV{TEMP};
204 my $base_name = sprintf("%s/%d-%d-0000", $temp_dir, $$, time());
208 until (defined(fileno(FH)) || $count++ > 100) {
209 $base_name =~ s/-(\d+)$/"-" . (1 + $1)/e;
210 sysopen(FH, $base_name, O_WRONLY|O_EXCL|O_CREAT);
212 if (defined(fileno(FH))
213 return (*FH, $base_name);
220 =head2 How can I manipulate fixed-record-length files?
222 The most efficient way is using pack() and unpack(). This is faster than
223 using substr() when taking many, many strings. It is slower for just a few.
225 Here is a sample chunk of code to break up and put back together again
226 some fixed-format input lines, in this case from the output of a normal,
230 # 15158 p5 T 0:00 perl /home/tchrist/scripts/now-what
231 $PS_T = 'A6 A4 A7 A5 A*';
235 ($pid, $tt, $stat, $time, $command) = unpack($PS_T, $_);
236 for $var (qw!pid tt stat time command!) {
237 print "$var: <$$var>\n";
239 print 'line=', pack($PS_T, $pid, $tt, $stat, $time, $command),
243 We've used C<$$var> in a way that forbidden by C<use strict 'refs'>.
244 That is, we've promoted a string to a scalar variable reference using
245 symbolic references. This is okay in small programs, but doesn't scale
246 well. It also only works on global variables, not lexicals.
248 =head2 How can I make a filehandle local to a subroutine? How do I pass filehandles between subroutines? How do I make an array of filehandles?
250 The fastest, simplest, and most direct way is to localize the typeglob
251 of the filehandle in question:
255 Typeglobs are fast (especially compared with the alternatives) and
256 reasonably easy to use, but they also have one subtle drawback. If you
257 had, for example, a function named TmpHandle(), or a variable named
258 %TmpHandle, you just hid it from yourself.
262 open(HostFile, "</etc/hosts") or die "no /etc/hosts: $!";
263 local $_; # <- VERY IMPORTANT
265 print if /\b127\.(0\.0\.)?1\b/;
267 # *HostFile automatically closes/disappears here
270 Here's how to use typeglobs in a loop to open and store a bunch of
271 filehandles. We'll use as values of the hash an ordered
272 pair to make it easy to sort the hash in insertion order.
274 @names = qw(motd termcap passwd hosts);
276 foreach $filename (@names) {
278 open(FH, "/etc/$filename") || die "$filename: $!";
279 $file{$filename} = [ $i++, *FH ];
282 # Using the filehandles in the array
283 foreach $name (sort { $file{$a}[0] <=> $file{$b}[0] } keys %file) {
284 my $fh = $file{$name}[1];
286 print "$name $. $line";
289 For passing filehandles to functions, the easiest way is to
290 preface them with a star, as in func(*STDIN).
291 See L<perlfaq7/"Passing Filehandles"> for details.
293 If you want to create many anonymous handles, you should check out the
294 Symbol, FileHandle, or IO::Handle (etc.) modules. Here's the equivalent
295 code with Symbol::gensym, which is reasonably light-weight:
297 foreach $filename (@names) {
300 open($fh, "/etc/$filename") || die "open /etc/$filename: $!";
301 $file{$filename} = [ $i++, $fh ];
304 Here's using the semi-object-oriented FileHandle module, which certainly
309 foreach $filename (@names) {
310 my $fh = FileHandle->new("/etc/$filename") or die "$filename: $!";
311 $file{$filename} = [ $i++, $fh ];
314 Please understand that whether the filehandle happens to be a (probably
315 localized) typeglob or an anonymous handle from one of the modules
316 in no way affects the bizarre rules for managing indirect handles.
317 See the next question.
319 =head2 How can I use a filehandle indirectly?
321 An indirect filehandle is using something other than a symbol
322 in a place that a filehandle is expected. Here are ways
323 to get indirect filehandles:
325 $fh = SOME_FH; # bareword is strict-subs hostile
326 $fh = "SOME_FH"; # strict-refs hostile; same package only
327 $fh = *SOME_FH; # typeglob
328 $fh = \*SOME_FH; # ref to typeglob (bless-able)
329 $fh = *SOME_FH{IO}; # blessed IO::Handle from *SOME_FH typeglob
331 Or, you can use the C<new> method from the FileHandle or IO modules to
332 create an anonymous filehandle, store that in a scalar variable,
333 and use it as though it were a normal filehandle.
336 $fh = FileHandle->new();
338 use IO::Handle; # 5.004 or higher
339 $fh = IO::Handle->new();
341 Then use any of those as you would a normal filehandle. Anywhere that
342 Perl is expecting a filehandle, an indirect filehandle may be used
343 instead. An indirect filehandle is just a scalar variable that contains
344 a filehandle. Functions like C<print>, C<open>, C<seek>, or
345 the C<< <FH> >> diamond operator will accept either a read filehandle
346 or a scalar variable containing one:
348 ($ifh, $ofh, $efh) = (*STDIN, *STDOUT, *STDERR);
349 print $ofh "Type it: ";
351 print $efh "What was that: $got";
353 If you're passing a filehandle to a function, you can write
354 the function in two ways:
358 print $fh "Sending to indirect filehandle\n";
361 Or it can localize a typeglob and use the filehandle directly:
365 print FH "Sending to localized filehandle\n";
368 Both styles work with either objects or typeglobs of real filehandles.
369 (They might also work with strings under some circumstances, but this
375 In the examples above, we assigned the filehandle to a scalar variable
376 before using it. That is because only simple scalar variables, not
377 expressions or subscripts of hashes or arrays, can be used with
378 built-ins like C<print>, C<printf>, or the diamond operator. Using
379 something other than a simple scalar variable as a filehandle is
380 illegal and won't even compile:
382 @fd = (*STDIN, *STDOUT, *STDERR);
383 print $fd[1] "Type it: "; # WRONG
384 $got = <$fd[0]> # WRONG
385 print $fd[2] "What was that: $got"; # WRONG
387 With C<print> and C<printf>, you get around this by using a block and
388 an expression where you would place the filehandle:
390 print { $fd[1] } "funny stuff\n";
391 printf { $fd[1] } "Pity the poor %x.\n", 3_735_928_559;
392 # Pity the poor deadbeef.
394 That block is a proper block like any other, so you can put more
395 complicated code there. This sends the message out to one of two places:
398 print { $ok ? $fd[1] : $fd[2] } "cat stat $ok\n";
399 print { $fd[ 1+ ($ok || 0) ] } "cat stat $ok\n";
401 This approach of treating C<print> and C<printf> like object methods
402 calls doesn't work for the diamond operator. That's because it's a
403 real operator, not just a function with a comma-less argument. Assuming
404 you've been storing typeglobs in your structure as we did above, you
405 can use the built-in function named C<readline> to reads a record just
406 as C<< <> >> does. Given the initialization shown above for @fd, this
407 would work, but only because readline() require a typeglob. It doesn't
408 work with objects or strings, which might be a bug we haven't fixed yet.
410 $got = readline($fd[0]);
412 Let it be noted that the flakiness of indirect filehandles is not
413 related to whether they're strings, typeglobs, objects, or anything else.
414 It's the syntax of the fundamental operators. Playing the object
415 game doesn't help you at all here.
417 =head2 How can I set up a footer format to be used with write()?
419 There's no builtin way to do this, but L<perlform> has a couple of
420 techniques to make it possible for the intrepid hacker.
422 =head2 How can I write() into a string?
424 See L<perlform/"Accessing Formatting Internals"> for an swrite() function.
426 =head2 How can I output my numbers with commas added?
428 This one will do it for you:
432 1 while ($number =~ s/^([-+]?\d+)(\d{3})/$1,$2/);
436 $n = 23659019423.2331;
437 print "GOT: ", commify($n), "\n";
439 GOT: 23,659,019,423.2331
443 s/^([-+]?\d+)(\d{3})/$1,$2/g;
445 because you have to put the comma in and then recalculate your
448 Alternatively, this code commifies all numbers in a line regardless of
449 whether they have decimal portions, are preceded by + or -, or
452 # from Andrew Johnson <ajohnson@gpu.srv.ualberta.ca>
455 $input = reverse $input;
456 $input =~ s<(\d\d\d)(?=\d)(?!\d*\.)><$1,>g;
457 return scalar reverse $input;
460 =head2 How can I translate tildes (~) in a filename?
462 Use the <> (glob()) operator, documented in L<perlfunc>. Older
463 versions of Perl require that you have a shell installed that groks
464 tildes. Recent perl versions have this feature built in. The
465 File::KGlob module (available from CPAN) gives more portable glob
468 Within Perl, you may use this directly:
471 ^ ~ # find a leading tilde
473 [^/] # a non-slash character
474 * # repeated 0 or more times (0 means me)
479 : ( $ENV{HOME} || $ENV{LOGDIR} )
482 =head2 How come when I open a file read-write it wipes it out?
484 Because you're using something like this, which truncates the file and
485 I<then> gives you read-write access:
487 open(FH, "+> /path/name"); # WRONG (almost always)
489 Whoops. You should instead use this, which will fail if the file
492 open(FH, "+< /path/name"); # open for update
494 Using ">" always clobbers or creates. Using "<" never does
495 either. The "+" doesn't change this.
497 Here are examples of many kinds of file opens. Those using sysopen()
502 To open file for reading:
504 open(FH, "< $path") || die $!;
505 sysopen(FH, $path, O_RDONLY) || die $!;
507 To open file for writing, create new file if needed or else truncate old file:
509 open(FH, "> $path") || die $!;
510 sysopen(FH, $path, O_WRONLY|O_TRUNC|O_CREAT) || die $!;
511 sysopen(FH, $path, O_WRONLY|O_TRUNC|O_CREAT, 0666) || die $!;
513 To open file for writing, create new file, file must not exist:
515 sysopen(FH, $path, O_WRONLY|O_EXCL|O_CREAT) || die $!;
516 sysopen(FH, $path, O_WRONLY|O_EXCL|O_CREAT, 0666) || die $!;
518 To open file for appending, create if necessary:
520 open(FH, ">> $path") || die $!;
521 sysopen(FH, $path, O_WRONLY|O_APPEND|O_CREAT) || die $!;
522 sysopen(FH, $path, O_WRONLY|O_APPEND|O_CREAT, 0666) || die $!;
524 To open file for appending, file must exist:
526 sysopen(FH, $path, O_WRONLY|O_APPEND) || die $!;
528 To open file for update, file must exist:
530 open(FH, "+< $path") || die $!;
531 sysopen(FH, $path, O_RDWR) || die $!;
533 To open file for update, create file if necessary:
535 sysopen(FH, $path, O_RDWR|O_CREAT) || die $!;
536 sysopen(FH, $path, O_RDWR|O_CREAT, 0666) || die $!;
538 To open file for update, file must not exist:
540 sysopen(FH, $path, O_RDWR|O_EXCL|O_CREAT) || die $!;
541 sysopen(FH, $path, O_RDWR|O_EXCL|O_CREAT, 0666) || die $!;
543 To open a file without blocking, creating if necessary:
545 sysopen(FH, "/tmp/somefile", O_WRONLY|O_NDELAY|O_CREAT)
546 or die "can't open /tmp/somefile: $!":
548 Be warned that neither creation nor deletion of files is guaranteed to
549 be an atomic operation over NFS. That is, two processes might both
550 successfully create or unlink the same file! Therefore O_EXCL
551 isn't as exclusive as you might wish.
553 See also the new L<perlopentut> if you have it (new for 5.6).
555 =head2 Why do I sometimes get an "Argument list too long" when I use <*>?
557 The C<< <> >> operator performs a globbing operation (see above).
558 In Perl versions earlier than v5.6.0, the internal glob() operator forks
559 csh(1) to do the actual glob expansion, but
560 csh can't handle more than 127 items and so gives the error message
561 C<Argument list too long>. People who installed tcsh as csh won't
562 have this problem, but their users may be surprised by it.
564 To get around this, either upgrade to Perl v5.6.0 or later, do the glob
565 yourself with readdir() and patterns, or use a module like File::KGlob,
566 one that doesn't use the shell to do globbing.
568 =head2 Is there a leak/bug in glob()?
570 Due to the current implementation on some operating systems, when you
571 use the glob() function or its angle-bracket alias in a scalar
572 context, you may cause a memory leak and/or unpredictable behavior. It's
573 best therefore to use glob() only in list context.
575 =head2 How can I open a file with a leading ">" or trailing blanks?
577 Normally perl ignores trailing blanks in filenames, and interprets
578 certain leading characters (or a trailing "|") to mean something
579 special. To avoid this, you might want to use a routine like the one below.
580 It turns incomplete pathnames into explicit relative ones, and tacks a
581 trailing null byte on the name to make perl leave it alone:
590 $badpath = "<<<something really wicked ";
591 $fn = safe_filename($badpath");
592 open(FH, "> $fn") or "couldn't open $badpath: $!";
594 This assumes that you are using POSIX (portable operating systems
595 interface) paths. If you are on a closed, non-portable, proprietary
596 system, you may have to adjust the C<"./"> above.
598 It would be a lot clearer to use sysopen(), though:
601 $badpath = "<<<something really wicked ";
602 sysopen (FH, $badpath, O_WRONLY | O_CREAT | O_TRUNC)
603 or die "can't open $badpath: $!";
605 For more information, see also the new L<perlopentut> if you have it
608 =head2 How can I reliably rename a file?
610 If your operating system supports a proper mv(1) utility or its functional
611 equivalent, this works:
613 rename($old, $new) or system("mv", $old, $new);
615 It may be more portable to use the File::Copy module instead.
616 You just copy to the new file to the new name (checking return
617 values), then delete the old one. This isn't really the same
618 semantically as a rename(), which preserves meta-information like
619 permissions, timestamps, inode info, etc.
621 Newer versions of File::Copy export a move() function.
623 =head2 How can I lock a file?
625 Perl's builtin flock() function (see L<perlfunc> for details) will call
626 flock(2) if that exists, fcntl(2) if it doesn't (on perl version 5.004 and
627 later), and lockf(3) if neither of the two previous system calls exists.
628 On some systems, it may even use a different form of native locking.
629 Here are some gotchas with Perl's flock():
635 Produces a fatal error if none of the three system calls (or their
636 close equivalent) exists.
640 lockf(3) does not provide shared locking, and requires that the
641 filehandle be open for writing (or appending, or read/writing).
645 Some versions of flock() can't lock files over a network (e.g. on NFS file
646 systems), so you'd need to force the use of fcntl(2) when you build Perl.
647 But even this is dubious at best. See the flock entry of L<perlfunc>
648 and the F<INSTALL> file in the source distribution for information on
649 building Perl to do this.
651 Two potentially non-obvious but traditional flock semantics are that
652 it waits indefinitely until the lock is granted, and that its locks are
653 I<merely advisory>. Such discretionary locks are more flexible, but
654 offer fewer guarantees. This means that files locked with flock() may
655 be modified by programs that do not also use flock(). Cars that stop
656 for red lights get on well with each other, but not with cars that don't
657 stop for red lights. See the perlport manpage, your port's specific
658 documentation, or your system-specific local manpages for details. It's
659 best to assume traditional behavior if you're writing portable programs.
660 (If you're not, you should as always feel perfectly free to write
661 for your own system's idiosyncrasies (sometimes called "features").
662 Slavish adherence to portability concerns shouldn't get in the way of
663 your getting your job done.)
665 For more information on file locking, see also
666 L<perlopentut/"File Locking"> if you have it (new for 5.6).
670 =head2 Why can't I just open(FH, ">file.lock")?
672 A common bit of code B<NOT TO USE> is this:
674 sleep(3) while -e "file.lock"; # PLEASE DO NOT USE
675 open(LCK, "> file.lock"); # THIS BROKEN CODE
677 This is a classic race condition: you take two steps to do something
678 which must be done in one. That's why computer hardware provides an
679 atomic test-and-set instruction. In theory, this "ought" to work:
681 sysopen(FH, "file.lock", O_WRONLY|O_EXCL|O_CREAT)
682 or die "can't open file.lock: $!":
684 except that lamentably, file creation (and deletion) is not atomic
685 over NFS, so this won't work (at least, not every time) over the net.
686 Various schemes involving link() have been suggested, but
687 these tend to involve busy-wait, which is also subdesirable.
689 =head2 I still don't get locking. I just want to increment the number in the file. How can I do this?
691 Didn't anyone ever tell you web-page hit counters were useless?
692 They don't count number of hits, they're a waste of time, and they serve
693 only to stroke the writer's vanity. It's better to pick a random number;
694 they're more realistic.
696 Anyway, this is what you can do if you can't help yourself.
698 use Fcntl qw(:DEFAULT :flock);
699 sysopen(FH, "numfile", O_RDWR|O_CREAT) or die "can't open numfile: $!";
700 flock(FH, LOCK_EX) or die "can't flock numfile: $!";
702 seek(FH, 0, 0) or die "can't rewind numfile: $!";
703 truncate(FH, 0) or die "can't truncate numfile: $!";
704 (print FH $num+1, "\n") or die "can't write numfile: $!";
705 close FH or die "can't close numfile: $!";
707 Here's a much better web-page hit counter:
709 $hits = int( (time() - 850_000_000) / rand(1_000) );
711 If the count doesn't impress your friends, then the code might. :-)
713 =head2 All I want to do is append a small amount of text to the end of a file. Do I still have to use locking?
715 If you are on a system that correctly implements flock() and you use the
716 example appending code from "perldoc -f flock" everything will be OK
717 even if the OS you are on doesn't implement append mode correctly (if
718 such a system exists.) So if you are happy to restrict yourself to OSs
719 that implement flock() (and that's not really much of a restriction)
720 then that is what you should do.
722 If you know you are only going to use a system that does correctly
723 implement appending (i.e. not Win32) then you can omit the seek() from
726 If you know you are only writing code to run on an OS and filesystem that
727 does implement append mode correctly (a local filesystem on a modern
728 Unix for example), and you keep the file in block-buffered mode and you
729 write less than one buffer-full of output between each manual flushing
730 of the buffer then each bufferload is almost guaranteed to be written to
731 the end of the file in one chunk without getting intermingled with
732 anyone else's output. You can also use the syswrite() function which is
733 simply a wrapper around your systems write(2) system call.
735 There is still a small theoretical chance that a signal will interrupt
736 the system level write() operation before completion. There is also a
737 possibility that some STDIO implementations may call multiple system
738 level write()s even if the buffer was empty to start. There may be some
739 systems where this probability is reduced to zero.
741 =head2 How do I randomly update a binary file?
743 If you're just trying to patch a binary, in many cases something as
744 simple as this works:
746 perl -i -pe 's{window manager}{window mangler}g' /usr/bin/emacs
748 However, if you have fixed sized records, then you might do something more
751 $RECSIZE = 220; # size of record, in bytes
752 $recno = 37; # which record to update
753 open(FH, "+<somewhere") || die "can't update somewhere: $!";
754 seek(FH, $recno * $RECSIZE, 0);
755 read(FH, $record, $RECSIZE) == $RECSIZE || die "can't read record $recno: $!";
757 seek(FH, -$RECSIZE, 1);
761 Locking and error checking are left as an exercise for the reader.
762 Don't forget them or you'll be quite sorry.
764 =head2 How do I get a file's timestamp in perl?
766 If you want to retrieve the time at which the file was last read,
767 written, or had its meta-data (owner, etc) changed, you use the B<-M>,
768 B<-A>, or B<-C> file test operations as documented in L<perlfunc>. These
769 retrieve the age of the file (measured against the start-time of your
770 program) in days as a floating point number. To retrieve the "raw"
771 time in seconds since the epoch, you would call the stat function,
772 then use localtime(), gmtime(), or POSIX::strftime() to convert this
773 into human-readable form.
777 $write_secs = (stat($file))[9];
778 printf "file %s updated at %s\n", $file,
779 scalar localtime($write_secs);
781 If you prefer something more legible, use the File::stat module
782 (part of the standard distribution in version 5.004 and later):
784 # error checking left as an exercise for reader.
787 $date_string = ctime(stat($file)->mtime);
788 print "file $file updated at $date_string\n";
790 The POSIX::strftime() approach has the benefit of being,
791 in theory, independent of the current locale. See L<perllocale>
794 =head2 How do I set a file's timestamp in perl?
796 You use the utime() function documented in L<perlfunc/utime>.
797 By way of example, here's a little program that copies the
798 read and write times from its first argument to all the rest
802 die "usage: cptimes timestamp_file other_files ...\n";
805 ($atime, $mtime) = (stat($timestamp))[8,9];
806 utime $atime, $mtime, @ARGV;
808 Error checking is, as usual, left as an exercise for the reader.
810 Note that utime() currently doesn't work correctly with Win95/NT
811 ports. A bug has been reported. Check it carefully before using
812 utime() on those platforms.
814 =head2 How do I print to more than one file at once?
816 If you only have to do this once, you can do this:
818 for $fh (FH1, FH2, FH3) { print $fh "whatever\n" }
820 To connect up to one filehandle to several output filehandles, it's
821 easiest to use the tee(1) program if you have it, and let it take care
824 open (FH, "| tee file1 file2 file3");
828 # make STDOUT go to three files, plus original STDOUT
829 open (STDOUT, "| tee file1 file2 file3") or die "Teeing off: $!\n";
830 print "whatever\n" or die "Writing: $!\n";
831 close(STDOUT) or die "Closing: $!\n";
833 Otherwise you'll have to write your own multiplexing print
834 function--or your own tee program--or use Tom Christiansen's,
835 at http://www.cpan.org/authors/id/TOMC/scripts/tct.gz , which is
836 written in Perl and offers much greater functionality
837 than the stock version.
839 =head2 How can I read in an entire file all at once?
841 The customary Perl approach for processing all the lines in a file is to
842 do so one line at a time:
844 open (INPUT, $file) || die "can't open $file: $!";
847 # do something with $_
849 close(INPUT) || die "can't close $file: $!";
851 This is tremendously more efficient than reading the entire file into
852 memory as an array of lines and then processing it one element at a time,
853 which is often--if not almost always--the wrong approach. Whenever
854 you see someone do this:
858 you should think long and hard about why you need everything loaded
859 at once. It's just not a scalable solution. You might also find it
860 more fun to use the standard DB_File module's $DB_RECNO bindings,
861 which allow you to tie an array to a file so that accessing an element
862 the array actually accesses the corresponding line in the file.
864 On very rare occasion, you may have an algorithm that demands that
865 the entire file be in memory at once as one scalar. The simplest solution
870 Being in scalar context, you get the whole thing. In list context,
871 you'd get a list of all the lines:
873 @lines = `cat $file`;
875 This tiny but expedient solution is neat, clean, and portable to
876 all systems on which decent tools have been installed. For those
877 who prefer not to use the toolbox, you can of course read the file
878 manually, although this makes for more complicated code.
882 open (INPUT, $file) || die "can't open $file: $!";
886 That temporarily undefs your record separator, and will automatically
887 close the file at block exit. If the file is already open, just use this:
889 $var = do { local $/; <INPUT> };
891 =head2 How can I read in a file by paragraphs?
893 Use the C<$/> variable (see L<perlvar> for details). You can either
894 set it to C<""> to eliminate empty paragraphs (C<"abc\n\n\n\ndef">,
895 for instance, gets treated as two paragraphs and not three), or
896 C<"\n\n"> to accept empty paragraphs.
898 Note that a blank line must have no blanks in it. Thus
899 S<C<"fred\n \nstuff\n\n">> is one paragraph, but C<"fred\n\nstuff\n\n"> is two.
901 =head2 How can I read a single character from a file? From the keyboard?
903 You can use the builtin C<getc()> function for most filehandles, but
904 it won't (easily) work on a terminal device. For STDIN, either use
905 the Term::ReadKey module from CPAN or use the sample code in
908 If your system supports the portable operating system programming
909 interface (POSIX), you can use the following code, which you'll note
910 turns off echo processing as well.
924 use POSIX qw(:termios_h);
926 my ($term, $oterm, $echo, $noecho, $fd_stdin);
928 $fd_stdin = fileno(STDIN);
930 $term = POSIX::Termios->new();
931 $term->getattr($fd_stdin);
932 $oterm = $term->getlflag();
934 $echo = ECHO | ECHOK | ICANON;
935 $noecho = $oterm & ~$echo;
938 $term->setlflag($noecho);
939 $term->setcc(VTIME, 1);
940 $term->setattr($fd_stdin, TCSANOW);
944 $term->setlflag($oterm);
945 $term->setcc(VTIME, 0);
946 $term->setattr($fd_stdin, TCSANOW);
952 sysread(STDIN, $key, 1);
961 The Term::ReadKey module from CPAN may be easier to use. Recent versions
962 include also support for non-portable systems as well.
965 open(TTY, "</dev/tty");
966 print "Gimme a char: ";
968 $key = ReadKey 0, *TTY;
970 printf "\nYou said %s, char number %03d\n",
973 =head2 How can I tell whether there's a character waiting on a filehandle?
975 The very first thing you should do is look into getting the Term::ReadKey
976 extension from CPAN. As we mentioned earlier, it now even has limited
977 support for non-portable (read: not open systems, closed, proprietary,
978 not POSIX, not Unix, etc) systems.
980 You should also check out the Frequently Asked Questions list in
981 comp.unix.* for things like this: the answer is essentially the same.
982 It's very system dependent. Here's one solution that works on BSD
987 vec($rin, fileno(STDIN), 1) = 1;
988 return $nfd = select($rin,undef,undef,0);
991 If you want to find out how many characters are waiting, there's
992 also the FIONREAD ioctl call to be looked at. The I<h2ph> tool that
993 comes with Perl tries to convert C include files to Perl code, which
994 can be C<require>d. FIONREAD ends up defined as a function in the
995 I<sys/ioctl.ph> file:
997 require 'sys/ioctl.ph';
999 $size = pack("L", 0);
1000 ioctl(FH, FIONREAD(), $size) or die "Couldn't call ioctl: $!\n";
1001 $size = unpack("L", $size);
1003 If I<h2ph> wasn't installed or doesn't work for you, you can
1004 I<grep> the include files by hand:
1006 % grep FIONREAD /usr/include/*/*
1007 /usr/include/asm/ioctls.h:#define FIONREAD 0x541B
1009 Or write a small C program using the editor of champions:
1012 #include <sys/ioctl.h>
1014 printf("%#08x\n", FIONREAD);
1017 % cc -o fionread fionread.c
1021 And then hard code it, leaving porting as an exercise to your successor.
1023 $FIONREAD = 0x4004667f; # XXX: opsys dependent
1025 $size = pack("L", 0);
1026 ioctl(FH, $FIONREAD, $size) or die "Couldn't call ioctl: $!\n";
1027 $size = unpack("L", $size);
1029 FIONREAD requires a filehandle connected to a stream, meaning that sockets,
1030 pipes, and tty devices work, but I<not> files.
1032 =head2 How do I do a C<tail -f> in perl?
1038 The statement C<seek(GWFILE, 0, 1)> doesn't change the current position,
1039 but it does clear the end-of-file condition on the handle, so that the
1040 next <GWFILE> makes Perl try again to read something.
1042 If that doesn't work (it relies on features of your stdio implementation),
1043 then you need something more like this:
1046 for ($curpos = tell(GWFILE); <GWFILE>; $curpos = tell(GWFILE)) {
1047 # search for some stuff and put it into files
1050 seek(GWFILE, $curpos, 0); # seek to where we had been
1053 If this still doesn't work, look into the POSIX module. POSIX defines
1054 the clearerr() method, which can remove the end of file condition on a
1055 filehandle. The method: read until end of file, clearerr(), read some
1056 more. Lather, rinse, repeat.
1058 There's also a File::Tail module from CPAN.
1060 =head2 How do I dup() a filehandle in Perl?
1062 If you check L<perlfunc/open>, you'll see that several of the ways
1063 to call open() should do the trick. For example:
1065 open(LOG, ">>/tmp/logfile");
1066 open(STDERR, ">&LOG");
1068 Or even with a literal numeric descriptor:
1070 $fd = $ENV{MHCONTEXTFD};
1071 open(MHCONTEXT, "<&=$fd"); # like fdopen(3S)
1073 Note that "<&STDIN" makes a copy, but "<&=STDIN" make
1074 an alias. That means if you close an aliased handle, all
1075 aliases become inaccessible. This is not true with
1078 Error checking, as always, has been left as an exercise for the reader.
1080 =head2 How do I close a file descriptor by number?
1082 This should rarely be necessary, as the Perl close() function is to be
1083 used for things that Perl opened itself, even if it was a dup of a
1084 numeric descriptor as with MHCONTEXT above. But if you really have
1085 to, you may be able to do this:
1087 require 'sys/syscall.ph';
1088 $rc = syscall(&SYS_close, $fd + 0); # must force numeric
1089 die "can't sysclose $fd: $!" unless $rc == -1;
1091 Or, just use the fdopen(3S) feature of open():
1095 open F, "<&=$fd" or die "Cannot reopen fd=$fd: $!";
1099 =head2 Why can't I use "C:\temp\foo" in DOS paths? What doesn't `C:\temp\foo.exe` work?
1101 Whoops! You just put a tab and a formfeed into that filename!
1102 Remember that within double quoted strings ("like\this"), the
1103 backslash is an escape character. The full list of these is in
1104 L<perlop/Quote and Quote-like Operators>. Unsurprisingly, you don't
1105 have a file called "c:(tab)emp(formfeed)oo" or
1106 "c:(tab)emp(formfeed)oo.exe" on your legacy DOS filesystem.
1108 Either single-quote your strings, or (preferably) use forward slashes.
1109 Since all DOS and Windows versions since something like MS-DOS 2.0 or so
1110 have treated C</> and C<\> the same in a path, you might as well use the
1111 one that doesn't clash with Perl--or the POSIX shell, ANSI C and C++,
1112 awk, Tcl, Java, or Python, just to mention a few. POSIX paths
1113 are more portable, too.
1115 =head2 Why doesn't glob("*.*") get all the files?
1117 Because even on non-Unix ports, Perl's glob function follows standard
1118 Unix globbing semantics. You'll need C<glob("*")> to get all (non-hidden)
1119 files. This makes glob() portable even to legacy systems. Your
1120 port may include proprietary globbing functions as well. Check its
1121 documentation for details.
1123 =head2 Why does Perl let me delete read-only files? Why does C<-i> clobber protected files? Isn't this a bug in Perl?
1125 This is elaborately and painstakingly described in the
1126 F<file-dir-perms> article in the "Far More Than You Ever Wanted To
1127 Know" collection in http://www.cpan.org/olddoc/FMTEYEWTK.tgz .
1129 The executive summary: learn how your filesystem works. The
1130 permissions on a file say what can happen to the data in that file.
1131 The permissions on a directory say what can happen to the list of
1132 files in that directory. If you delete a file, you're removing its
1133 name from the directory (so the operation depends on the permissions
1134 of the directory, not of the file). If you try to write to the file,
1135 the permissions of the file govern whether you're allowed to.
1137 =head2 How do I select a random line from a file?
1139 Here's an algorithm from the Camel Book:
1142 rand($.) < 1 && ($line = $_) while <>;
1144 This has a significant advantage in space over reading the whole
1145 file in. A simple proof by induction is available upon
1146 request if you doubt the algorithm's correctness.
1148 =head2 Why do I get weird spaces when I print an array of lines?
1154 joins together the elements of C<@lines> with a space between them.
1155 If C<@lines> were C<("little", "fluffy", "clouds")> then the above
1156 statement would print
1158 little fluffy clouds
1160 but if each element of C<@lines> was a line of text, ending a newline
1161 character C<("little\n", "fluffy\n", "clouds\n")> then it would print:
1167 If your array contains lines, just print them:
1171 =head1 AUTHOR AND COPYRIGHT
1173 Copyright (c) 1997-2002 Tom Christiansen and Nathan Torkington.
1174 All rights reserved.
1176 This documentation is free; you can redistribute it and/or modify it
1177 under the same terms as Perl itself.
1179 Irrespective of its distribution, all code examples here are in the public
1180 domain. You are permitted and encouraged to use this code and any
1181 derivatives thereof in your own programs for fun or for profit as you
1182 see fit. A simple comment in the code giving credit to the FAQ would
1183 be courteous but is not required.