3 perlfaq5 - Files and Formats ($Revision: 7875 $)
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?
11 X<flush> X<buffer> X<unbuffer> X<autoflush>
13 Perl does not support truly unbuffered output (except
14 insofar as you can C<syswrite(OUT, $char, 1)>), although it
15 does support is "command buffering", in which a physical
16 write is performed after every output command.
18 The C standard I/O library (stdio) normally buffers
19 characters sent to devices so that there isn't a system call
20 for each byte. In most stdio implementations, the type of
21 output buffering and the size of the buffer varies according
22 to the type of device. Perl's print() and write() functions
23 normally buffer output, while syswrite() bypasses buffering
26 If you want your output to be sent immediately when you
27 execute print() or write() (for instance, for some network
28 protocols), you must set the handle's autoflush flag. This
29 flag is the Perl variable $| and when it is set to a true
30 value, Perl will flush the handle's buffer after each
31 print() or write(). Setting $| affects buffering only for
32 the currently selected default file handle. You choose this
33 handle with the one argument select() call (see
34 L<perlvar/$E<verbar>> and L<perlfunc/select>).
36 Use select() to choose the desired handle, then set its
37 per-filehandle variables.
39 $old_fh = select(OUTPUT_HANDLE);
43 Some modules offer object-oriented access to handles and their
44 variables, although they may be overkill if this is the only
45 thing you do with them. You can use IO::Handle:
48 open(DEV, ">/dev/printer"); # but is this?
53 use IO::Socket; # this one is kinda a pipe?
54 my $sock = IO::Socket::INET->new( 'www.example.com:80' );
58 =head2 How do I change, delete, or insert a line in a file, or append to the beginning of a file?
61 (contributed by brian d foy)
63 The basic idea of inserting, changing, or deleting a line from a text
64 file involves reading and printing the file to the point you want to
65 make the change, making the change, then reading and printing the rest
66 of the file. Perl doesn't provide random access to lines (especially
67 since the record input separator, C<$/>, is mutable), although modules
68 such as C<Tie::File> can fake it.
70 A Perl program to do these tasks takes the basic form of opening a
71 file, printing its lines, then closing the file:
73 open my $in, '<', $file or die "Can't read old file: $!";
74 open my $out, '>', "$file.new" or die "Can't write new file: $!";
83 Within that basic form, add the parts that you need to insert, change,
86 To prepend lines to the beginning, print those lines before you enter
87 the loop that prints the existing lines.
89 open my $in, '<', $file or die "Can't read old file: $!";
90 open my $out, '>', "$file.new" or die "Can't write new file: $!";
92 print "# Add this line to the top\n"; # <--- HERE'S THE MAGIC
101 To change existing lines, insert the code to modify the lines inside
102 the C<while> loop. In this case, the code finds all lowercased
103 versions of "perl" and uppercases them. The happens for every line, so
104 be sure that you're supposed to do that on every line!
106 open my $in, '<', $file or die "Can't read old file: $!";
107 open my $out, '>', "$file.new" or die "Can't write new file: $!";
109 print "# Add this line to the top\n";
119 To change only a particular line, the input line number, C<$.>, is
120 useful. Use C<next> to skip all lines up to line 5, make a change and
121 print the result, then stop further processing with C<last>.
131 To skip lines, use the looping controls. The C<next> in this example
132 skips comment lines, and the C<last> stops all processing once it
133 encounters either C<__END__> or C<__DATA__>.
137 next if /^\s+#/; # skip comment lines
138 last if /^__(END|DATA)__$/; # stop at end of code marker
142 Do the same sort of thing to delete a particular line by using C<next>
143 to skip the lines you don't want to show up in the output. This
144 example skips every fifth line:
152 If, for some odd reason, you really want to see the whole file at once
153 rather than processing line by line, you can slurp it in (as long as
154 you can fit the whole thing in memory!):
156 open my $in, '<', $file or die "Can't read old file: $!"
157 open my $out, '>', "$file.new" or die "Can't write new file: $!";
159 my @lines = do { local $/; <$in> }; # slurp!
165 Modules such as C<File::Slurp> and C<Tie::File> can help with that
166 too. If you can, however, avoid reading the entire file at once. Perl
167 won't give that memory back to the operating system until the process
170 You can also use Perl one-liners to modify a file in-place. The
171 following changes all 'Fred' to 'Barney' in F<inFile.txt>, overwriting
172 the file with the new contents. With the C<-p> switch, Perl wraps a
173 C<while> loop around the code you specify with C<-e>, and C<-i> turns
174 on in-place editing. The current line is in C<$_>. With C<-p>, Perl
175 automatically prints the value of C<$_> at the end of the loop. See
176 L<perlrun> for more details.
178 perl -pi -e 's/Fred/Barney/' inFile.txt
180 To make a backup of C<inFile.txt>, give C<-i> a file extension to add:
182 perl -pi.bak -e 's/Fred/Barney/' inFile.txt
184 To change only the fifth line, you can add a test checking C<$.>, the
185 input line number, then only perform the operation when the test
188 perl -pi -e 's/Fred/Barney/ if $. == 5' inFile.txt
190 To add lines before a certain line, you can add a line (or lines!)
191 before Perl prints C<$_>:
193 perl -pi -e 'print "Put before third line\n" if $. == 3' inFile.txt
195 You can even add a line to the beginning of a file, since the current
196 line prints at the end of the loop:
198 perl -pi -e 'print "Put before first line\n" if $. == 1' inFile.txt
200 To insert a line after one already in the file, use the C<-n> switch.
201 It's just like C<-p> except that it doesn't print C<$_> at the end of
202 the loop, so you have to do that yourself. In this case, print C<$_>
203 first, then print the line that you want to add.
205 perl -ni -e 'print; print "Put after fifth line\n" if $. == 5' inFile.txt
207 To delete lines, only print the ones that you want.
209 perl -ni -e 'print unless /d/' inFile.txt
213 perl -pi -e 'next unless /d/' inFile.txt
215 =head2 How do I count the number of lines in a file?
216 X<file, counting lines> X<lines> X<line>
218 One fairly efficient way is to count newlines in the file. The
219 following program uses a feature of tr///, as documented in L<perlop>.
220 If your text file doesn't end with a newline, then it's not really a
221 proper text file, so this may report one fewer line than you expect.
224 open(FILE, $filename) or die "Can't open `$filename': $!";
225 while (sysread FILE, $buffer, 4096) {
226 $lines += ($buffer =~ tr/\n//);
230 This assumes no funny games with newline translations.
232 =head2 How can I use Perl's C<-i> option from within a program?
235 C<-i> sets the value of Perl's C<$^I> variable, which in turn affects
236 the behavior of C<< <> >>; see L<perlrun> for more details. By
237 modifying the appropriate variables directly, you can get the same
238 behavior within a larger program. For example:
242 local($^I, @ARGV) = ('.orig', glob("*.c"));
245 print "This line should appear at the top of each file\n";
247 s/\b(p)earl\b/${1}erl/i; # Correct typos, preserving case
249 close ARGV if eof; # Reset $.
252 # $^I and @ARGV return to their old values here
254 This block modifies all the C<.c> files in the current directory,
255 leaving a backup of the original data from each file in a new
258 =head2 How can I copy a file?
259 X<copy> X<file, copy>
261 (contributed by brian d foy)
263 Use the File::Copy module. It comes with Perl and can do a
264 true copy across file systems, and it does its magic in
269 copy( $original, $new_copy ) or die "Copy failed: $!";
271 If you can't use File::Copy, you'll have to do the work yourself:
272 open the original file, open the destination file, then print
273 to the destination file as you read the original.
275 =head2 How do I make a temporary file name?
278 If you don't need to know the name of the file, you can use C<open()>
279 with C<undef> in place of the file name. The C<open()> function
280 creates an anonymous temporary file.
282 open my $tmp, '+>', undef or die $!;
284 Otherwise, you can use the File::Temp module.
286 use File::Temp qw/ tempfile tempdir /;
288 $dir = tempdir( CLEANUP => 1 );
289 ($fh, $filename) = tempfile( DIR => $dir );
291 # or if you don't need to know the filename
293 $fh = tempfile( DIR => $dir );
295 The File::Temp has been a standard module since Perl 5.6.1. If you
296 don't have a modern enough Perl installed, use the C<new_tmpfile>
297 class method from the IO::File module to get a filehandle opened for
298 reading and writing. Use it if you don't need to know the file's name:
301 $fh = IO::File->new_tmpfile()
302 or die "Unable to make new temporary file: $!";
304 If you're committed to creating a temporary file by hand, use the
305 process ID and/or the current time-value. If you need to have many
306 temporary files in one process, use a counter:
310 my $temp_dir = -d '/tmp' ? '/tmp' : $ENV{TMPDIR} || $ENV{TEMP};
311 my $base_name = sprintf("%s/%d-%d-0000", $temp_dir, $$, time());
316 until (defined(fileno(FH)) || $count++ > 100) {
317 $base_name =~ s/-(\d+)$/"-" . (1 + $1)/e;
318 # O_EXCL is required for security reasons.
319 sysopen(FH, $base_name, O_WRONLY|O_EXCL|O_CREAT);
322 if (defined(fileno(FH))
323 return (*FH, $base_name);
331 =head2 How can I manipulate fixed-record-length files?
332 X<fixed-length> X<file, fixed-length records>
334 The most efficient way is using L<pack()|perlfunc/"pack"> and
335 L<unpack()|perlfunc/"unpack">. This is faster than using
336 L<substr()|perlfunc/"substr"> when taking many, many strings. It is
337 slower for just a few.
339 Here is a sample chunk of code to break up and put back together again
340 some fixed-format input lines, in this case from the output of a normal,
344 # 15158 p5 T 0:00 perl /home/tchrist/scripts/now-what
345 my $PS_T = 'A6 A4 A7 A5 A*';
346 open my $ps, '-|', 'ps';
348 my @fields = qw( pid tt stat time command );
351 @process{@fields} = unpack($PS_T, $_);
352 for my $field ( @fields ) {
353 print "$field: <$process{$field}>\n";
355 print 'line=', pack($PS_T, @process{@fields} ), "\n";
358 We've used a hash slice in order to easily handle the fields of each row.
359 Storing the keys in an array means it's easy to operate on them as a
360 group or loop over them with for. It also avoids polluting the program
361 with global variables and using symbolic references.
363 =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?
364 X<filehandle, local> X<filehandle, passing> X<filehandle, reference>
366 As of perl5.6, open() autovivifies file and directory handles
367 as references if you pass it an uninitialized scalar variable.
368 You can then pass these references just like any other scalar,
369 and use them in the place of named handles.
371 open my $fh, $file_name;
373 open local $fh, $file_name;
375 print $fh "Hello World!\n";
379 If you like, you can store these filehandles in an array or a hash.
380 If you access them directly, they aren't simple scalars and you
381 need to give C<print> a little help by placing the filehandle
382 reference in braces. Perl can only figure it out on its own when
383 the filehandle reference is a simple scalar.
385 my @fhs = ( $fh1, $fh2, $fh3 );
387 for( $i = 0; $i <= $#fhs; $i++ ) {
388 print {$fhs[$i]} "just another Perl answer, \n";
391 Before perl5.6, you had to deal with various typeglob idioms
392 which you may see in older code.
394 open FILE, "> $filename";
395 process_typeglob( *FILE );
396 process_reference( \*FILE );
398 sub process_typeglob { local *FH = shift; print FH "Typeglob!" }
399 sub process_reference { local $fh = shift; print $fh "Reference!" }
401 If you want to create many anonymous handles, you should
402 check out the Symbol or IO::Handle modules.
404 =head2 How can I use a filehandle indirectly?
405 X<filehandle, indirect>
407 An indirect filehandle is using something other than a symbol
408 in a place that a filehandle is expected. Here are ways
409 to get indirect filehandles:
411 $fh = SOME_FH; # bareword is strict-subs hostile
412 $fh = "SOME_FH"; # strict-refs hostile; same package only
413 $fh = *SOME_FH; # typeglob
414 $fh = \*SOME_FH; # ref to typeglob (bless-able)
415 $fh = *SOME_FH{IO}; # blessed IO::Handle from *SOME_FH typeglob
417 Or, you can use the C<new> method from one of the IO::* modules to
418 create an anonymous filehandle, store that in a scalar variable,
419 and use it as though it were a normal filehandle.
421 use IO::Handle; # 5.004 or higher
422 $fh = IO::Handle->new();
424 Then use any of those as you would a normal filehandle. Anywhere that
425 Perl is expecting a filehandle, an indirect filehandle may be used
426 instead. An indirect filehandle is just a scalar variable that contains
427 a filehandle. Functions like C<print>, C<open>, C<seek>, or
428 the C<< <FH> >> diamond operator will accept either a named filehandle
429 or a scalar variable containing one:
431 ($ifh, $ofh, $efh) = (*STDIN, *STDOUT, *STDERR);
432 print $ofh "Type it: ";
434 print $efh "What was that: $got";
436 If you're passing a filehandle to a function, you can write
437 the function in two ways:
441 print $fh "Sending to indirect filehandle\n";
444 Or it can localize a typeglob and use the filehandle directly:
448 print FH "Sending to localized filehandle\n";
451 Both styles work with either objects or typeglobs of real filehandles.
452 (They might also work with strings under some circumstances, but this
458 In the examples above, we assigned the filehandle to a scalar variable
459 before using it. That is because only simple scalar variables, not
460 expressions or subscripts of hashes or arrays, can be used with
461 built-ins like C<print>, C<printf>, or the diamond operator. Using
462 something other than a simple scalar variable as a filehandle is
463 illegal and won't even compile:
465 @fd = (*STDIN, *STDOUT, *STDERR);
466 print $fd[1] "Type it: "; # WRONG
467 $got = <$fd[0]> # WRONG
468 print $fd[2] "What was that: $got"; # WRONG
470 With C<print> and C<printf>, you get around this by using a block and
471 an expression where you would place the filehandle:
473 print { $fd[1] } "funny stuff\n";
474 printf { $fd[1] } "Pity the poor %x.\n", 3_735_928_559;
475 # Pity the poor deadbeef.
477 That block is a proper block like any other, so you can put more
478 complicated code there. This sends the message out to one of two places:
481 print { $ok ? $fd[1] : $fd[2] } "cat stat $ok\n";
482 print { $fd[ 1+ ($ok || 0) ] } "cat stat $ok\n";
484 This approach of treating C<print> and C<printf> like object methods
485 calls doesn't work for the diamond operator. That's because it's a
486 real operator, not just a function with a comma-less argument. Assuming
487 you've been storing typeglobs in your structure as we did above, you
488 can use the built-in function named C<readline> to read a record just
489 as C<< <> >> does. Given the initialization shown above for @fd, this
490 would work, but only because readline() requires a typeglob. It doesn't
491 work with objects or strings, which might be a bug we haven't fixed yet.
493 $got = readline($fd[0]);
495 Let it be noted that the flakiness of indirect filehandles is not
496 related to whether they're strings, typeglobs, objects, or anything else.
497 It's the syntax of the fundamental operators. Playing the object
498 game doesn't help you at all here.
500 =head2 How can I set up a footer format to be used with write()?
503 There's no builtin way to do this, but L<perlform> has a couple of
504 techniques to make it possible for the intrepid hacker.
506 =head2 How can I write() into a string?
507 X<write, into a string>
509 See L<perlform/"Accessing Formatting Internals"> for an swrite() function.
511 =head2 How can I output my numbers with commas added?
514 (contributed by brian d foy and Benjamin Goldberg)
516 You can use L<Number::Format> to separate places in a number.
517 It handles locale information for those of you who want to insert
518 full stops instead (or anything else that they want to use,
521 This subroutine will add commas to your number:
525 1 while s/^([-+]?\d+)(\d{3})/$1,$2/;
529 This regex from Benjamin Goldberg will add commas to numbers:
531 s/(^[-+]?\d+?(?=(?>(?:\d{3})+)(?!\d))|\G\d{3}(?=\d))/$1,/g;
533 It is easier to see with comments:
536 ^[-+]? # beginning of number.
537 \d+? # first digits before first comma
538 (?= # followed by, (but not included in the match) :
539 (?>(?:\d{3})+) # some positive multiple of three digits.
540 (?!\d) # an *exact* multiple, not x * 3 + 1 or whatever.
543 \G\d{3} # after the last group, get three digits
544 (?=\d) # but they have to have more digits after them.
547 =head2 How can I translate tildes (~) in a filename?
548 X<tilde> X<tilde expansion>
550 Use the <> (glob()) operator, documented in L<perlfunc>. Older
551 versions of Perl require that you have a shell installed that groks
552 tildes. Recent perl versions have this feature built in. The
553 File::KGlob module (available from CPAN) gives more portable glob
556 Within Perl, you may use this directly:
559 ^ ~ # find a leading tilde
561 [^/] # a non-slash character
562 * # repeated 0 or more times (0 means me)
567 : ( $ENV{HOME} || $ENV{LOGDIR} )
570 =head2 How come when I open a file read-write it wipes it out?
571 X<clobber> X<read-write> X<clobbering> X<truncate> X<truncating>
573 Because you're using something like this, which truncates the file and
574 I<then> gives you read-write access:
576 open(FH, "+> /path/name"); # WRONG (almost always)
578 Whoops. You should instead use this, which will fail if the file
581 open(FH, "+< /path/name"); # open for update
583 Using ">" always clobbers or creates. Using "<" never does
584 either. The "+" doesn't change this.
586 Here are examples of many kinds of file opens. Those using sysopen()
591 To open file for reading:
593 open(FH, "< $path") || die $!;
594 sysopen(FH, $path, O_RDONLY) || die $!;
596 To open file for writing, create new file if needed or else truncate old file:
598 open(FH, "> $path") || die $!;
599 sysopen(FH, $path, O_WRONLY|O_TRUNC|O_CREAT) || die $!;
600 sysopen(FH, $path, O_WRONLY|O_TRUNC|O_CREAT, 0666) || die $!;
602 To open file for writing, create new file, file must not exist:
604 sysopen(FH, $path, O_WRONLY|O_EXCL|O_CREAT) || die $!;
605 sysopen(FH, $path, O_WRONLY|O_EXCL|O_CREAT, 0666) || die $!;
607 To open file for appending, create if necessary:
609 open(FH, ">> $path") || die $!;
610 sysopen(FH, $path, O_WRONLY|O_APPEND|O_CREAT) || die $!;
611 sysopen(FH, $path, O_WRONLY|O_APPEND|O_CREAT, 0666) || die $!;
613 To open file for appending, file must exist:
615 sysopen(FH, $path, O_WRONLY|O_APPEND) || die $!;
617 To open file for update, file must exist:
619 open(FH, "+< $path") || die $!;
620 sysopen(FH, $path, O_RDWR) || die $!;
622 To open file for update, create file if necessary:
624 sysopen(FH, $path, O_RDWR|O_CREAT) || die $!;
625 sysopen(FH, $path, O_RDWR|O_CREAT, 0666) || die $!;
627 To open file for update, file must not exist:
629 sysopen(FH, $path, O_RDWR|O_EXCL|O_CREAT) || die $!;
630 sysopen(FH, $path, O_RDWR|O_EXCL|O_CREAT, 0666) || die $!;
632 To open a file without blocking, creating if necessary:
634 sysopen(FH, "/foo/somefile", O_WRONLY|O_NDELAY|O_CREAT)
635 or die "can't open /foo/somefile: $!":
637 Be warned that neither creation nor deletion of files is guaranteed to
638 be an atomic operation over NFS. That is, two processes might both
639 successfully create or unlink the same file! Therefore O_EXCL
640 isn't as exclusive as you might wish.
642 See also the new L<perlopentut> if you have it (new for 5.6).
644 =head2 Why do I sometimes get an "Argument list too long" when I use E<lt>*E<gt>?
645 X<argument list too long>
647 The C<< <> >> operator performs a globbing operation (see above).
648 In Perl versions earlier than v5.6.0, the internal glob() operator forks
649 csh(1) to do the actual glob expansion, but
650 csh can't handle more than 127 items and so gives the error message
651 C<Argument list too long>. People who installed tcsh as csh won't
652 have this problem, but their users may be surprised by it.
654 To get around this, either upgrade to Perl v5.6.0 or later, do the glob
655 yourself with readdir() and patterns, or use a module like File::KGlob,
656 one that doesn't use the shell to do globbing.
658 =head2 Is there a leak/bug in glob()?
661 Due to the current implementation on some operating systems, when you
662 use the glob() function or its angle-bracket alias in a scalar
663 context, you may cause a memory leak and/or unpredictable behavior. It's
664 best therefore to use glob() only in list context.
666 =head2 How can I open a file with a leading ">" or trailing blanks?
667 X<filename, special characters>
669 (contributed by Brian McCauley)
671 The special two argument form of Perl's open() function ignores
672 trailing blanks in filenames and infers the mode from certain leading
673 characters (or a trailing "|"). In older versions of Perl this was the
674 only version of open() and so it is prevalent in old code and books.
676 Unless you have a particular reason to use the two argument form you
677 should use the three argument form of open() which does not treat any
678 charcters in the filename as special.
680 open FILE, "<", " file "; # filename is " file "
681 open FILE, ">", ">file"; # filename is ">file"
683 =head2 How can I reliably rename a file?
684 X<rename> X<mv> X<move> X<file, rename> X<ren>
686 If your operating system supports a proper mv(1) utility or its
687 functional equivalent, this works:
689 rename($old, $new) or system("mv", $old, $new);
691 It may be more portable to use the File::Copy module instead.
692 You just copy to the new file to the new name (checking return
693 values), then delete the old one. This isn't really the same
694 semantically as a rename(), which preserves meta-information like
695 permissions, timestamps, inode info, etc.
697 Newer versions of File::Copy export a move() function.
699 =head2 How can I lock a file?
700 X<lock> X<file, lock> X<flock>
702 Perl's builtin flock() function (see L<perlfunc> for details) will call
703 flock(2) if that exists, fcntl(2) if it doesn't (on perl version 5.004 and
704 later), and lockf(3) if neither of the two previous system calls exists.
705 On some systems, it may even use a different form of native locking.
706 Here are some gotchas with Perl's flock():
712 Produces a fatal error if none of the three system calls (or their
713 close equivalent) exists.
717 lockf(3) does not provide shared locking, and requires that the
718 filehandle be open for writing (or appending, or read/writing).
722 Some versions of flock() can't lock files over a network (e.g. on NFS file
723 systems), so you'd need to force the use of fcntl(2) when you build Perl.
724 But even this is dubious at best. See the flock entry of L<perlfunc>
725 and the F<INSTALL> file in the source distribution for information on
726 building Perl to do this.
728 Two potentially non-obvious but traditional flock semantics are that
729 it waits indefinitely until the lock is granted, and that its locks are
730 I<merely advisory>. Such discretionary locks are more flexible, but
731 offer fewer guarantees. This means that files locked with flock() may
732 be modified by programs that do not also use flock(). Cars that stop
733 for red lights get on well with each other, but not with cars that don't
734 stop for red lights. See the perlport manpage, your port's specific
735 documentation, or your system-specific local manpages for details. It's
736 best to assume traditional behavior if you're writing portable programs.
737 (If you're not, you should as always feel perfectly free to write
738 for your own system's idiosyncrasies (sometimes called "features").
739 Slavish adherence to portability concerns shouldn't get in the way of
740 your getting your job done.)
742 For more information on file locking, see also
743 L<perlopentut/"File Locking"> if you have it (new for 5.6).
747 =head2 Why can't I just open(FH, "E<gt>file.lock")?
748 X<lock, lockfile race condition>
750 A common bit of code B<NOT TO USE> is this:
752 sleep(3) while -e "file.lock"; # PLEASE DO NOT USE
753 open(LCK, "> file.lock"); # THIS BROKEN CODE
755 This is a classic race condition: you take two steps to do something
756 which must be done in one. That's why computer hardware provides an
757 atomic test-and-set instruction. In theory, this "ought" to work:
759 sysopen(FH, "file.lock", O_WRONLY|O_EXCL|O_CREAT)
760 or die "can't open file.lock: $!";
762 except that lamentably, file creation (and deletion) is not atomic
763 over NFS, so this won't work (at least, not every time) over the net.
764 Various schemes involving link() have been suggested, but
765 these tend to involve busy-wait, which is also subdesirable.
767 =head2 I still don't get locking. I just want to increment the number in the file. How can I do this?
768 X<counter> X<file, counter>
770 Didn't anyone ever tell you web-page hit counters were useless?
771 They don't count number of hits, they're a waste of time, and they serve
772 only to stroke the writer's vanity. It's better to pick a random number;
773 they're more realistic.
775 Anyway, this is what you can do if you can't help yourself.
777 use Fcntl qw(:DEFAULT :flock);
778 sysopen(FH, "numfile", O_RDWR|O_CREAT) or die "can't open numfile: $!";
779 flock(FH, LOCK_EX) or die "can't flock numfile: $!";
781 seek(FH, 0, 0) or die "can't rewind numfile: $!";
782 truncate(FH, 0) or die "can't truncate numfile: $!";
783 (print FH $num+1, "\n") or die "can't write numfile: $!";
784 close FH or die "can't close numfile: $!";
786 Here's a much better web-page hit counter:
788 $hits = int( (time() - 850_000_000) / rand(1_000) );
790 If the count doesn't impress your friends, then the code might. :-)
792 =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?
793 X<append> X<file, append>
795 If you are on a system that correctly implements flock() and you use the
796 example appending code from "perldoc -f flock" everything will be OK
797 even if the OS you are on doesn't implement append mode correctly (if
798 such a system exists.) So if you are happy to restrict yourself to OSs
799 that implement flock() (and that's not really much of a restriction)
800 then that is what you should do.
802 If you know you are only going to use a system that does correctly
803 implement appending (i.e. not Win32) then you can omit the seek() from
806 If you know you are only writing code to run on an OS and filesystem that
807 does implement append mode correctly (a local filesystem on a modern
808 Unix for example), and you keep the file in block-buffered mode and you
809 write less than one buffer-full of output between each manual flushing
810 of the buffer then each bufferload is almost guaranteed to be written to
811 the end of the file in one chunk without getting intermingled with
812 anyone else's output. You can also use the syswrite() function which is
813 simply a wrapper around your systems write(2) system call.
815 There is still a small theoretical chance that a signal will interrupt
816 the system level write() operation before completion. There is also a
817 possibility that some STDIO implementations may call multiple system
818 level write()s even if the buffer was empty to start. There may be some
819 systems where this probability is reduced to zero.
821 =head2 How do I randomly update a binary file?
822 X<file, binary patch>
824 If you're just trying to patch a binary, in many cases something as
825 simple as this works:
827 perl -i -pe 's{window manager}{window mangler}g' /usr/bin/emacs
829 However, if you have fixed sized records, then you might do something more
832 $RECSIZE = 220; # size of record, in bytes
833 $recno = 37; # which record to update
834 open(FH, "+<somewhere") || die "can't update somewhere: $!";
835 seek(FH, $recno * $RECSIZE, 0);
836 read(FH, $record, $RECSIZE) == $RECSIZE || die "can't read record $recno: $!";
838 seek(FH, -$RECSIZE, 1);
842 Locking and error checking are left as an exercise for the reader.
843 Don't forget them or you'll be quite sorry.
845 =head2 How do I get a file's timestamp in perl?
846 X<timestamp> X<file, timestamp>
848 If you want to retrieve the time at which the file was last
849 read, written, or had its meta-data (owner, etc) changed,
850 you use the B<-A>, B<-M>, or B<-C> file test operations as
851 documented in L<perlfunc>. These retrieve the age of the
852 file (measured against the start-time of your program) in
853 days as a floating point number. Some platforms may not have
854 all of these times. See L<perlport> for details. To
855 retrieve the "raw" time in seconds since the epoch, you
856 would call the stat function, then use localtime(),
857 gmtime(), or POSIX::strftime() to convert this into
862 $write_secs = (stat($file))[9];
863 printf "file %s updated at %s\n", $file,
864 scalar localtime($write_secs);
866 If you prefer something more legible, use the File::stat module
867 (part of the standard distribution in version 5.004 and later):
869 # error checking left as an exercise for reader.
872 $date_string = ctime(stat($file)->mtime);
873 print "file $file updated at $date_string\n";
875 The POSIX::strftime() approach has the benefit of being,
876 in theory, independent of the current locale. See L<perllocale>
879 =head2 How do I set a file's timestamp in perl?
880 X<timestamp> X<file, timestamp>
882 You use the utime() function documented in L<perlfunc/utime>.
883 By way of example, here's a little program that copies the
884 read and write times from its first argument to all the rest
888 die "usage: cptimes timestamp_file other_files ...\n";
891 ($atime, $mtime) = (stat($timestamp))[8,9];
892 utime $atime, $mtime, @ARGV;
894 Error checking is, as usual, left as an exercise for the reader.
896 The perldoc for utime also has an example that has the same
897 effect as touch(1) on files that I<already exist>.
899 Certain file systems have a limited ability to store the times
900 on a file at the expected level of precision. For example, the
901 FAT and HPFS filesystem are unable to create dates on files with
902 a finer granularity than two seconds. This is a limitation of
903 the filesystems, not of utime().
905 =head2 How do I print to more than one file at once?
906 X<print, to multiple files>
908 To connect one filehandle to several output filehandles,
909 you can use the IO::Tee or Tie::FileHandle::Multiplex modules.
911 If you only have to do this once, you can print individually
914 for $fh (FH1, FH2, FH3) { print $fh "whatever\n" }
916 =head2 How can I read in an entire file all at once?
917 X<slurp> X<file, slurping>
919 You can use the File::Slurp module to do it in one step.
923 $all_of_it = read_file($filename); # entire file in scalar
924 @all_lines = read_file($filename); # one line perl element
926 The customary Perl approach for processing all the lines in a file is to
927 do so one line at a time:
929 open (INPUT, $file) || die "can't open $file: $!";
932 # do something with $_
934 close(INPUT) || die "can't close $file: $!";
936 This is tremendously more efficient than reading the entire file into
937 memory as an array of lines and then processing it one element at a time,
938 which is often--if not almost always--the wrong approach. Whenever
939 you see someone do this:
943 you should think long and hard about why you need everything loaded at
944 once. It's just not a scalable solution. You might also find it more
945 fun to use the standard Tie::File module, or the DB_File module's
946 $DB_RECNO bindings, which allow you to tie an array to a file so that
947 accessing an element the array actually accesses the corresponding
950 You can read the entire filehandle contents into a scalar.
954 open (INPUT, $file) || die "can't open $file: $!";
958 That temporarily undefs your record separator, and will automatically
959 close the file at block exit. If the file is already open, just use this:
961 $var = do { local $/; <INPUT> };
963 For ordinary files you can also use the read function.
965 read( INPUT, $var, -s INPUT );
967 The third argument tests the byte size of the data on the INPUT filehandle
968 and reads that many bytes into the buffer $var.
970 =head2 How can I read in a file by paragraphs?
971 X<file, reading by paragraphs>
973 Use the C<$/> variable (see L<perlvar> for details). You can either
974 set it to C<""> to eliminate empty paragraphs (C<"abc\n\n\n\ndef">,
975 for instance, gets treated as two paragraphs and not three), or
976 C<"\n\n"> to accept empty paragraphs.
978 Note that a blank line must have no blanks in it. Thus
979 S<C<"fred\n \nstuff\n\n">> is one paragraph, but C<"fred\n\nstuff\n\n"> is two.
981 =head2 How can I read a single character from a file? From the keyboard?
982 X<getc> X<file, reading one character at a time>
984 You can use the builtin C<getc()> function for most filehandles, but
985 it won't (easily) work on a terminal device. For STDIN, either use
986 the Term::ReadKey module from CPAN or use the sample code in
989 If your system supports the portable operating system programming
990 interface (POSIX), you can use the following code, which you'll note
991 turns off echo processing as well.
1005 use POSIX qw(:termios_h);
1007 my ($term, $oterm, $echo, $noecho, $fd_stdin);
1009 $fd_stdin = fileno(STDIN);
1011 $term = POSIX::Termios->new();
1012 $term->getattr($fd_stdin);
1013 $oterm = $term->getlflag();
1015 $echo = ECHO | ECHOK | ICANON;
1016 $noecho = $oterm & ~$echo;
1019 $term->setlflag($noecho);
1020 $term->setcc(VTIME, 1);
1021 $term->setattr($fd_stdin, TCSANOW);
1025 $term->setlflag($oterm);
1026 $term->setcc(VTIME, 0);
1027 $term->setattr($fd_stdin, TCSANOW);
1033 sysread(STDIN, $key, 1);
1042 The Term::ReadKey module from CPAN may be easier to use. Recent versions
1043 include also support for non-portable systems as well.
1046 open(TTY, "</dev/tty");
1047 print "Gimme a char: ";
1049 $key = ReadKey 0, *TTY;
1051 printf "\nYou said %s, char number %03d\n",
1054 =head2 How can I tell whether there's a character waiting on a filehandle?
1056 The very first thing you should do is look into getting the Term::ReadKey
1057 extension from CPAN. As we mentioned earlier, it now even has limited
1058 support for non-portable (read: not open systems, closed, proprietary,
1059 not POSIX, not Unix, etc) systems.
1061 You should also check out the Frequently Asked Questions list in
1062 comp.unix.* for things like this: the answer is essentially the same.
1063 It's very system dependent. Here's one solution that works on BSD
1068 vec($rin, fileno(STDIN), 1) = 1;
1069 return $nfd = select($rin,undef,undef,0);
1072 If you want to find out how many characters are waiting, there's
1073 also the FIONREAD ioctl call to be looked at. The I<h2ph> tool that
1074 comes with Perl tries to convert C include files to Perl code, which
1075 can be C<require>d. FIONREAD ends up defined as a function in the
1076 I<sys/ioctl.ph> file:
1078 require 'sys/ioctl.ph';
1080 $size = pack("L", 0);
1081 ioctl(FH, FIONREAD(), $size) or die "Couldn't call ioctl: $!\n";
1082 $size = unpack("L", $size);
1084 If I<h2ph> wasn't installed or doesn't work for you, you can
1085 I<grep> the include files by hand:
1087 % grep FIONREAD /usr/include/*/*
1088 /usr/include/asm/ioctls.h:#define FIONREAD 0x541B
1090 Or write a small C program using the editor of champions:
1093 #include <sys/ioctl.h>
1095 printf("%#08x\n", FIONREAD);
1098 % cc -o fionread fionread.c
1102 And then hard code it, leaving porting as an exercise to your successor.
1104 $FIONREAD = 0x4004667f; # XXX: opsys dependent
1106 $size = pack("L", 0);
1107 ioctl(FH, $FIONREAD, $size) or die "Couldn't call ioctl: $!\n";
1108 $size = unpack("L", $size);
1110 FIONREAD requires a filehandle connected to a stream, meaning that sockets,
1111 pipes, and tty devices work, but I<not> files.
1113 =head2 How do I do a C<tail -f> in perl?
1114 X<tail> X<IO::Handle> X<File::Tail> X<clearerr>
1120 The statement C<seek(GWFILE, 0, 1)> doesn't change the current position,
1121 but it does clear the end-of-file condition on the handle, so that the
1122 next C<< <GWFILE> >> makes Perl try again to read something.
1124 If that doesn't work (it relies on features of your stdio implementation),
1125 then you need something more like this:
1128 for ($curpos = tell(GWFILE); <GWFILE>; $curpos = tell(GWFILE)) {
1129 # search for some stuff and put it into files
1132 seek(GWFILE, $curpos, 0); # seek to where we had been
1135 If this still doesn't work, look into the C<clearerr> method
1136 from C<IO::Handle>, which resets the error and end-of-file states
1139 There's also a C<File::Tail> module from CPAN.
1141 =head2 How do I dup() a filehandle in Perl?
1144 If you check L<perlfunc/open>, you'll see that several of the ways
1145 to call open() should do the trick. For example:
1147 open(LOG, ">>/foo/logfile");
1148 open(STDERR, ">&LOG");
1150 Or even with a literal numeric descriptor:
1152 $fd = $ENV{MHCONTEXTFD};
1153 open(MHCONTEXT, "<&=$fd"); # like fdopen(3S)
1155 Note that "<&STDIN" makes a copy, but "<&=STDIN" make
1156 an alias. That means if you close an aliased handle, all
1157 aliases become inaccessible. This is not true with
1160 Error checking, as always, has been left as an exercise for the reader.
1162 =head2 How do I close a file descriptor by number?
1163 X<file, closing file descriptors>
1165 This should rarely be necessary, as the Perl close() function is to be
1166 used for things that Perl opened itself, even if it was a dup of a
1167 numeric descriptor as with MHCONTEXT above. But if you really have
1168 to, you may be able to do this:
1170 require 'sys/syscall.ph';
1171 $rc = syscall(&SYS_close, $fd + 0); # must force numeric
1172 die "can't sysclose $fd: $!" unless $rc == -1;
1174 Or, just use the fdopen(3S) feature of open():
1178 open F, "<&=$fd" or die "Cannot reopen fd=$fd: $!";
1182 =head2 Why can't I use "C:\temp\foo" in DOS paths? Why doesn't `C:\temp\foo.exe` work?
1183 X<filename, DOS issues>
1185 Whoops! You just put a tab and a formfeed into that filename!
1186 Remember that within double quoted strings ("like\this"), the
1187 backslash is an escape character. The full list of these is in
1188 L<perlop/Quote and Quote-like Operators>. Unsurprisingly, you don't
1189 have a file called "c:(tab)emp(formfeed)oo" or
1190 "c:(tab)emp(formfeed)oo.exe" on your legacy DOS filesystem.
1192 Either single-quote your strings, or (preferably) use forward slashes.
1193 Since all DOS and Windows versions since something like MS-DOS 2.0 or so
1194 have treated C</> and C<\> the same in a path, you might as well use the
1195 one that doesn't clash with Perl--or the POSIX shell, ANSI C and C++,
1196 awk, Tcl, Java, or Python, just to mention a few. POSIX paths
1197 are more portable, too.
1199 =head2 Why doesn't glob("*.*") get all the files?
1202 Because even on non-Unix ports, Perl's glob function follows standard
1203 Unix globbing semantics. You'll need C<glob("*")> to get all (non-hidden)
1204 files. This makes glob() portable even to legacy systems. Your
1205 port may include proprietary globbing functions as well. Check its
1206 documentation for details.
1208 =head2 Why does Perl let me delete read-only files? Why does C<-i> clobber protected files? Isn't this a bug in Perl?
1210 This is elaborately and painstakingly described in the
1211 F<file-dir-perms> article in the "Far More Than You Ever Wanted To
1212 Know" collection in http://www.cpan.org/misc/olddoc/FMTEYEWTK.tgz .
1214 The executive summary: learn how your filesystem works. The
1215 permissions on a file say what can happen to the data in that file.
1216 The permissions on a directory say what can happen to the list of
1217 files in that directory. If you delete a file, you're removing its
1218 name from the directory (so the operation depends on the permissions
1219 of the directory, not of the file). If you try to write to the file,
1220 the permissions of the file govern whether you're allowed to.
1222 =head2 How do I select a random line from a file?
1223 X<file, selecting a random line>
1225 Here's an algorithm from the Camel Book:
1228 rand($.) < 1 && ($line = $_) while <>;
1230 This has a significant advantage in space over reading the whole file
1231 in. You can find a proof of this method in I<The Art of Computer
1232 Programming>, Volume 2, Section 3.4.2, by Donald E. Knuth.
1234 You can use the File::Random module which provides a function
1237 use File::Random qw/random_line/;
1238 my $line = random_line($filename);
1240 Another way is to use the Tie::File module, which treats the entire
1241 file as an array. Simply access a random array element.
1243 =head2 Why do I get weird spaces when I print an array of lines?
1249 joins together the elements of C<@lines> with a space between them.
1250 If C<@lines> were C<("little", "fluffy", "clouds")> then the above
1251 statement would print
1253 little fluffy clouds
1255 but if each element of C<@lines> was a line of text, ending a newline
1256 character C<("little\n", "fluffy\n", "clouds\n")> then it would print:
1262 If your array contains lines, just print them:
1268 Revision: $Revision: 7875 $
1270 Date: $Date: 2006-10-04 22:39:26 +0200 (mer, 04 oct 2006) $
1272 See L<perlfaq> for source control details and availability.
1274 =head1 AUTHOR AND COPYRIGHT
1276 Copyright (c) 1997-2006 Tom Christiansen, Nathan Torkington, and
1277 other authors as noted. All rights reserved.
1279 This documentation is free; you can redistribute it and/or modify it
1280 under the same terms as Perl itself.
1282 Irrespective of its distribution, all code examples here are in the public
1283 domain. You are permitted and encouraged to use this code and any
1284 derivatives thereof in your own programs for fun or for profit as you
1285 see fit. A simple comment in the code giving credit to the FAQ would
1286 be courteous but is not required.