3 perlfaq5 - Files and Formats ($Revision: 1.37 $, $Date: 2005/08/10 15:55:23 $)
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 Perl does not support truly unbuffered output (except
13 insofar as you can C<syswrite(OUT, $char, 1)>), although it
14 does support is "command buffering", in which a physical
15 write is performed after every output command.
17 The C standard I/O library (stdio) normally buffers
18 characters sent to devices so that there isn't a system call
19 for each byte. In most stdio implementations, the type of
20 output buffering and the size of the buffer varies according
21 to the type of device. Perl's print() and write() functions
22 normally buffer output, while syswrite() bypasses buffering
25 If you want your output to be sent immediately when you
26 execute print() or write() (for instance, for some network
27 protocols), you must set the handle's autoflush flag. This
28 flag is the Perl variable $| and when it is set to a true
29 value, Perl will flush the handle's buffer after each
30 print() or write(). Setting $| affects buffering only for
31 the currently selected default file handle. You choose this
32 handle with the one argument select() call (see
33 L<perlvar/$E<verbar>> and L<perlfunc/select>).
35 Use select() to choose the desired handle, then set its
36 per-filehandle variables.
38 $old_fh = select(OUTPUT_HANDLE);
42 Some idioms can handle this in a single statement:
44 select((select(OUTPUT_HANDLE), $| = 1)[0]);
46 $| = 1, select $_ for select OUTPUT_HANDLE;
48 Some modules offer object-oriented access to handles and their
49 variables, although they may be overkill if this is the only
50 thing you do with them. You can use IO::Handle:
53 open(DEV, ">/dev/printer"); # but is this?
58 use IO::Socket; # this one is kinda a pipe?
59 my $sock = IO::Socket::INET->new( 'www.example.com:80' ) ;
63 =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?
65 Use the Tie::File module, which is included in the standard
66 distribution since Perl 5.8.0.
68 =head2 How do I count the number of lines in a file?
70 One fairly efficient way is to count newlines in the file. The
71 following program uses a feature of tr///, as documented in L<perlop>.
72 If your text file doesn't end with a newline, then it's not really a
73 proper text file, so this may report one fewer line than you expect.
76 open(FILE, $filename) or die "Can't open `$filename': $!";
77 while (sysread FILE, $buffer, 4096) {
78 $lines += ($buffer =~ tr/\n//);
82 This assumes no funny games with newline translations.
84 =head2 How can I use Perl's C<-i> option from within a program?
86 C<-i> sets the value of Perl's C<$^I> variable, which in turn affects
87 the behavior of C<< <> >>; see L<perlrun> for more details. By
88 modifying the appropriate variables directly, you can get the same
89 behavior within a larger program. For example:
93 local($^I, @ARGV) = ('.orig', glob("*.c"));
96 print "This line should appear at the top of each file\n";
98 s/\b(p)earl\b/${1}erl/i; # Correct typos, preserving case
100 close ARGV if eof; # Reset $.
103 # $^I and @ARGV return to their old values here
105 This block modifies all the C<.c> files in the current directory,
106 leaving a backup of the original data from each file in a new
109 =head2 How can I copy a file?
111 (contributed by brian d foy)
113 Use the File::Copy module. It comes with Perl and can do a
114 true copy across file systems, and it does its magic in
119 copy( $original, $new_copy ) or die "Copy failed: $!";
121 If you can't use File::Copy, you'll have to do the work yourself:
122 open the original file, open the destination file, then print
123 to the destination file as you read the original.
125 =head2 How do I make a temporary file name?
127 If you don't need to know the name of the file, you can use C<open()>
128 with C<undef> in place of the file name. The C<open()> function
129 creates an anonymous temporary file.
131 open my $tmp, '+>', undef or die $!;
133 Otherwise, you can use the File::Temp module.
135 use File::Temp qw/ tempfile tempdir /;
137 $dir = tempdir( CLEANUP => 1 );
138 ($fh, $filename) = tempfile( DIR => $dir );
140 # or if you don't need to know the filename
142 $fh = tempfile( DIR => $dir );
144 The File::Temp has been a standard module since Perl 5.6.1. If you
145 don't have a modern enough Perl installed, use the C<new_tmpfile>
146 class method from the IO::File module to get a filehandle opened for
147 reading and writing. Use it if you don't need to know the file's name:
150 $fh = IO::File->new_tmpfile()
151 or die "Unable to make new temporary file: $!";
153 If you're committed to creating a temporary file by hand, use the
154 process ID and/or the current time-value. If you need to have many
155 temporary files in one process, use a counter:
159 my $temp_dir = -d '/tmp' ? '/tmp' : $ENV{TMPDIR} || $ENV{TEMP};
160 my $base_name = sprintf("%s/%d-%d-0000", $temp_dir, $$, time());
164 until (defined(fileno(FH)) || $count++ > 100) {
165 $base_name =~ s/-(\d+)$/"-" . (1 + $1)/e;
166 # O_EXCL is required for security reasons.
167 sysopen(FH, $base_name, O_WRONLY|O_EXCL|O_CREAT);
169 if (defined(fileno(FH))
170 return (*FH, $base_name);
177 =head2 How can I manipulate fixed-record-length files?
179 The most efficient way is using L<pack()|perlfunc/"pack"> and
180 L<unpack()|perlfunc/"unpack">. This is faster than using
181 L<substr()|perlfunc/"substr"> when taking many, many strings. It is
182 slower for just a few.
184 Here is a sample chunk of code to break up and put back together again
185 some fixed-format input lines, in this case from the output of a normal,
189 # 15158 p5 T 0:00 perl /home/tchrist/scripts/now-what
190 my $PS_T = 'A6 A4 A7 A5 A*';
191 open my $ps, '-|', 'ps';
193 my @fields = qw( pid tt stat time command );
196 @process{@fields} = unpack($PS_T, $_);
197 for my $field ( @fields ) {
198 print "$field: <$process{$field}>\n";
200 print 'line=', pack($PS_T, @process{@fields} ), "\n";
203 We've used a hash slice in order to easily handle the fields of each row.
204 Storing the keys in an array means it's easy to operate on them as a
205 group or loop over them with for. It also avoids polluting the program
206 with global variables and using symbolic references.
208 =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?
210 As of perl5.6, open() autovivifies file and directory handles
211 as references if you pass it an uninitialized scalar variable.
212 You can then pass these references just like any other scalar,
213 and use them in the place of named handles.
215 open my $fh, $file_name;
217 open local $fh, $file_name;
219 print $fh "Hello World!\n";
223 Before perl5.6, you had to deal with various typeglob idioms
224 which you may see in older code.
226 open FILE, "> $filename";
227 process_typeglob( *FILE );
228 process_reference( \*FILE );
230 sub process_typeglob { local *FH = shift; print FH "Typeglob!" }
231 sub process_reference { local $fh = shift; print $fh "Reference!" }
233 If you want to create many anonymous handles, you should
234 check out the Symbol or IO::Handle modules.
236 =head2 How can I use a filehandle indirectly?
238 An indirect filehandle is using something other than a symbol
239 in a place that a filehandle is expected. Here are ways
240 to get indirect filehandles:
242 $fh = SOME_FH; # bareword is strict-subs hostile
243 $fh = "SOME_FH"; # strict-refs hostile; same package only
244 $fh = *SOME_FH; # typeglob
245 $fh = \*SOME_FH; # ref to typeglob (bless-able)
246 $fh = *SOME_FH{IO}; # blessed IO::Handle from *SOME_FH typeglob
248 Or, you can use the C<new> method from one of the IO::* modules to
249 create an anonymous filehandle, store that in a scalar variable,
250 and use it as though it were a normal filehandle.
252 use IO::Handle; # 5.004 or higher
253 $fh = IO::Handle->new();
255 Then use any of those as you would a normal filehandle. Anywhere that
256 Perl is expecting a filehandle, an indirect filehandle may be used
257 instead. An indirect filehandle is just a scalar variable that contains
258 a filehandle. Functions like C<print>, C<open>, C<seek>, or
259 the C<< <FH> >> diamond operator will accept either a named filehandle
260 or a scalar variable containing one:
262 ($ifh, $ofh, $efh) = (*STDIN, *STDOUT, *STDERR);
263 print $ofh "Type it: ";
265 print $efh "What was that: $got";
267 If you're passing a filehandle to a function, you can write
268 the function in two ways:
272 print $fh "Sending to indirect filehandle\n";
275 Or it can localize a typeglob and use the filehandle directly:
279 print FH "Sending to localized filehandle\n";
282 Both styles work with either objects or typeglobs of real filehandles.
283 (They might also work with strings under some circumstances, but this
289 In the examples above, we assigned the filehandle to a scalar variable
290 before using it. That is because only simple scalar variables, not
291 expressions or subscripts of hashes or arrays, can be used with
292 built-ins like C<print>, C<printf>, or the diamond operator. Using
293 something other than a simple scalar variable as a filehandle is
294 illegal and won't even compile:
296 @fd = (*STDIN, *STDOUT, *STDERR);
297 print $fd[1] "Type it: "; # WRONG
298 $got = <$fd[0]> # WRONG
299 print $fd[2] "What was that: $got"; # WRONG
301 With C<print> and C<printf>, you get around this by using a block and
302 an expression where you would place the filehandle:
304 print { $fd[1] } "funny stuff\n";
305 printf { $fd[1] } "Pity the poor %x.\n", 3_735_928_559;
306 # Pity the poor deadbeef.
308 That block is a proper block like any other, so you can put more
309 complicated code there. This sends the message out to one of two places:
312 print { $ok ? $fd[1] : $fd[2] } "cat stat $ok\n";
313 print { $fd[ 1+ ($ok || 0) ] } "cat stat $ok\n";
315 This approach of treating C<print> and C<printf> like object methods
316 calls doesn't work for the diamond operator. That's because it's a
317 real operator, not just a function with a comma-less argument. Assuming
318 you've been storing typeglobs in your structure as we did above, you
319 can use the built-in function named C<readline> to read a record just
320 as C<< <> >> does. Given the initialization shown above for @fd, this
321 would work, but only because readline() requires a typeglob. It doesn't
322 work with objects or strings, which might be a bug we haven't fixed yet.
324 $got = readline($fd[0]);
326 Let it be noted that the flakiness of indirect filehandles is not
327 related to whether they're strings, typeglobs, objects, or anything else.
328 It's the syntax of the fundamental operators. Playing the object
329 game doesn't help you at all here.
331 =head2 How can I set up a footer format to be used with write()?
333 There's no builtin way to do this, but L<perlform> has a couple of
334 techniques to make it possible for the intrepid hacker.
336 =head2 How can I write() into a string?
338 See L<perlform/"Accessing Formatting Internals"> for an swrite() function.
340 =head2 How can I output my numbers with commas added?
342 (contributed by brian d foy and Benjamin Goldberg)
344 You can use L<Number::Format> to separate places in a number.
345 It handles locale information for those of you who want to insert
346 full stops instead (or anything else that they want to use,
349 This subroutine will add commas to your number:
353 1 while s/^([-+]?\d+)(\d{3})/$1,$2/;
357 This regex from Benjamin Goldberg will add commas to numbers:
359 s/(^[-+]?\d+?(?=(?>(?:\d{3})+)(?!\d))|\G\d{3}(?=\d))/$1,/g;
361 It is easier to see with comments:
364 ^[-+]? # beginning of number.
365 \d+? # first digits before first comma
366 (?= # followed by, (but not included in the match) :
367 (?>(?:\d{3})+) # some positive multiple of three digits.
368 (?!\d) # an *exact* multiple, not x * 3 + 1 or whatever.
371 \G\d{3} # after the last group, get three digits
372 (?=\d) # but they have to have more digits after them.
375 =head2 How can I translate tildes (~) in a filename?
377 Use the <> (glob()) operator, documented in L<perlfunc>. Older
378 versions of Perl require that you have a shell installed that groks
379 tildes. Recent perl versions have this feature built in. The
380 File::KGlob module (available from CPAN) gives more portable glob
383 Within Perl, you may use this directly:
386 ^ ~ # find a leading tilde
388 [^/] # a non-slash character
389 * # repeated 0 or more times (0 means me)
394 : ( $ENV{HOME} || $ENV{LOGDIR} )
397 =head2 How come when I open a file read-write it wipes it out?
399 Because you're using something like this, which truncates the file and
400 I<then> gives you read-write access:
402 open(FH, "+> /path/name"); # WRONG (almost always)
404 Whoops. You should instead use this, which will fail if the file
407 open(FH, "+< /path/name"); # open for update
409 Using ">" always clobbers or creates. Using "<" never does
410 either. The "+" doesn't change this.
412 Here are examples of many kinds of file opens. Those using sysopen()
417 To open file for reading:
419 open(FH, "< $path") || die $!;
420 sysopen(FH, $path, O_RDONLY) || die $!;
422 To open file for writing, create new file if needed or else truncate old file:
424 open(FH, "> $path") || die $!;
425 sysopen(FH, $path, O_WRONLY|O_TRUNC|O_CREAT) || die $!;
426 sysopen(FH, $path, O_WRONLY|O_TRUNC|O_CREAT, 0666) || die $!;
428 To open file for writing, create new file, file must not exist:
430 sysopen(FH, $path, O_WRONLY|O_EXCL|O_CREAT) || die $!;
431 sysopen(FH, $path, O_WRONLY|O_EXCL|O_CREAT, 0666) || die $!;
433 To open file for appending, create if necessary:
435 open(FH, ">> $path") || die $!;
436 sysopen(FH, $path, O_WRONLY|O_APPEND|O_CREAT) || die $!;
437 sysopen(FH, $path, O_WRONLY|O_APPEND|O_CREAT, 0666) || die $!;
439 To open file for appending, file must exist:
441 sysopen(FH, $path, O_WRONLY|O_APPEND) || die $!;
443 To open file for update, file must exist:
445 open(FH, "+< $path") || die $!;
446 sysopen(FH, $path, O_RDWR) || die $!;
448 To open file for update, create file if necessary:
450 sysopen(FH, $path, O_RDWR|O_CREAT) || die $!;
451 sysopen(FH, $path, O_RDWR|O_CREAT, 0666) || die $!;
453 To open file for update, file must not exist:
455 sysopen(FH, $path, O_RDWR|O_EXCL|O_CREAT) || die $!;
456 sysopen(FH, $path, O_RDWR|O_EXCL|O_CREAT, 0666) || die $!;
458 To open a file without blocking, creating if necessary:
460 sysopen(FH, "/foo/somefile", O_WRONLY|O_NDELAY|O_CREAT)
461 or die "can't open /foo/somefile: $!":
463 Be warned that neither creation nor deletion of files is guaranteed to
464 be an atomic operation over NFS. That is, two processes might both
465 successfully create or unlink the same file! Therefore O_EXCL
466 isn't as exclusive as you might wish.
468 See also the new L<perlopentut> if you have it (new for 5.6).
470 =head2 Why do I sometimes get an "Argument list too long" when I use E<lt>*E<gt>?
472 The C<< <> >> operator performs a globbing operation (see above).
473 In Perl versions earlier than v5.6.0, the internal glob() operator forks
474 csh(1) to do the actual glob expansion, but
475 csh can't handle more than 127 items and so gives the error message
476 C<Argument list too long>. People who installed tcsh as csh won't
477 have this problem, but their users may be surprised by it.
479 To get around this, either upgrade to Perl v5.6.0 or later, do the glob
480 yourself with readdir() and patterns, or use a module like File::KGlob,
481 one that doesn't use the shell to do globbing.
483 =head2 Is there a leak/bug in glob()?
485 Due to the current implementation on some operating systems, when you
486 use the glob() function or its angle-bracket alias in a scalar
487 context, you may cause a memory leak and/or unpredictable behavior. It's
488 best therefore to use glob() only in list context.
490 =head2 How can I open a file with a leading ">" or trailing blanks?
492 (contributed by Brian McCauley)
494 The special two argument form of Perl's open() function ignores
495 trailing blanks in filenames and infers the mode from certain leading
496 characters (or a trailing "|"). In older versions of Perl this was the
497 only version of open() and so it is prevalent in old code and books.
499 Unless you have a particular reason to use the two argument form you
500 should use the three argument form of open() which does not treat any
501 charcters in the filename as special.
503 open FILE, "<", " file "; # filename is " file "
504 open FILE, ">", ">file"; # filename is ">file"
506 =head2 How can I reliably rename a file?
508 If your operating system supports a proper mv(1) utility or its
509 functional equivalent, this works:
511 rename($old, $new) or system("mv", $old, $new);
513 It may be more portable to use the File::Copy module instead.
514 You just copy to the new file to the new name (checking return
515 values), then delete the old one. This isn't really the same
516 semantically as a rename(), which preserves meta-information like
517 permissions, timestamps, inode info, etc.
519 Newer versions of File::Copy export a move() function.
521 =head2 How can I lock a file?
523 Perl's builtin flock() function (see L<perlfunc> for details) will call
524 flock(2) if that exists, fcntl(2) if it doesn't (on perl version 5.004 and
525 later), and lockf(3) if neither of the two previous system calls exists.
526 On some systems, it may even use a different form of native locking.
527 Here are some gotchas with Perl's flock():
533 Produces a fatal error if none of the three system calls (or their
534 close equivalent) exists.
538 lockf(3) does not provide shared locking, and requires that the
539 filehandle be open for writing (or appending, or read/writing).
543 Some versions of flock() can't lock files over a network (e.g. on NFS file
544 systems), so you'd need to force the use of fcntl(2) when you build Perl.
545 But even this is dubious at best. See the flock entry of L<perlfunc>
546 and the F<INSTALL> file in the source distribution for information on
547 building Perl to do this.
549 Two potentially non-obvious but traditional flock semantics are that
550 it waits indefinitely until the lock is granted, and that its locks are
551 I<merely advisory>. Such discretionary locks are more flexible, but
552 offer fewer guarantees. This means that files locked with flock() may
553 be modified by programs that do not also use flock(). Cars that stop
554 for red lights get on well with each other, but not with cars that don't
555 stop for red lights. See the perlport manpage, your port's specific
556 documentation, or your system-specific local manpages for details. It's
557 best to assume traditional behavior if you're writing portable programs.
558 (If you're not, you should as always feel perfectly free to write
559 for your own system's idiosyncrasies (sometimes called "features").
560 Slavish adherence to portability concerns shouldn't get in the way of
561 your getting your job done.)
563 For more information on file locking, see also
564 L<perlopentut/"File Locking"> if you have it (new for 5.6).
568 =head2 Why can't I just open(FH, "E<gt>file.lock")?
570 A common bit of code B<NOT TO USE> is this:
572 sleep(3) while -e "file.lock"; # PLEASE DO NOT USE
573 open(LCK, "> file.lock"); # THIS BROKEN CODE
575 This is a classic race condition: you take two steps to do something
576 which must be done in one. That's why computer hardware provides an
577 atomic test-and-set instruction. In theory, this "ought" to work:
579 sysopen(FH, "file.lock", O_WRONLY|O_EXCL|O_CREAT)
580 or die "can't open file.lock: $!";
582 except that lamentably, file creation (and deletion) is not atomic
583 over NFS, so this won't work (at least, not every time) over the net.
584 Various schemes involving link() have been suggested, but
585 these tend to involve busy-wait, which is also subdesirable.
587 =head2 I still don't get locking. I just want to increment the number in the file. How can I do this?
589 Didn't anyone ever tell you web-page hit counters were useless?
590 They don't count number of hits, they're a waste of time, and they serve
591 only to stroke the writer's vanity. It's better to pick a random number;
592 they're more realistic.
594 Anyway, this is what you can do if you can't help yourself.
596 use Fcntl qw(:DEFAULT :flock);
597 sysopen(FH, "numfile", O_RDWR|O_CREAT) or die "can't open numfile: $!";
598 flock(FH, LOCK_EX) or die "can't flock numfile: $!";
600 seek(FH, 0, 0) or die "can't rewind numfile: $!";
601 truncate(FH, 0) or die "can't truncate numfile: $!";
602 (print FH $num+1, "\n") or die "can't write numfile: $!";
603 close FH or die "can't close numfile: $!";
605 Here's a much better web-page hit counter:
607 $hits = int( (time() - 850_000_000) / rand(1_000) );
609 If the count doesn't impress your friends, then the code might. :-)
611 =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?
613 If you are on a system that correctly implements flock() and you use the
614 example appending code from "perldoc -f flock" everything will be OK
615 even if the OS you are on doesn't implement append mode correctly (if
616 such a system exists.) So if you are happy to restrict yourself to OSs
617 that implement flock() (and that's not really much of a restriction)
618 then that is what you should do.
620 If you know you are only going to use a system that does correctly
621 implement appending (i.e. not Win32) then you can omit the seek() from
624 If you know you are only writing code to run on an OS and filesystem that
625 does implement append mode correctly (a local filesystem on a modern
626 Unix for example), and you keep the file in block-buffered mode and you
627 write less than one buffer-full of output between each manual flushing
628 of the buffer then each bufferload is almost guaranteed to be written to
629 the end of the file in one chunk without getting intermingled with
630 anyone else's output. You can also use the syswrite() function which is
631 simply a wrapper around your systems write(2) system call.
633 There is still a small theoretical chance that a signal will interrupt
634 the system level write() operation before completion. There is also a
635 possibility that some STDIO implementations may call multiple system
636 level write()s even if the buffer was empty to start. There may be some
637 systems where this probability is reduced to zero.
639 =head2 How do I randomly update a binary file?
641 If you're just trying to patch a binary, in many cases something as
642 simple as this works:
644 perl -i -pe 's{window manager}{window mangler}g' /usr/bin/emacs
646 However, if you have fixed sized records, then you might do something more
649 $RECSIZE = 220; # size of record, in bytes
650 $recno = 37; # which record to update
651 open(FH, "+<somewhere") || die "can't update somewhere: $!";
652 seek(FH, $recno * $RECSIZE, 0);
653 read(FH, $record, $RECSIZE) == $RECSIZE || die "can't read record $recno: $!";
655 seek(FH, -$RECSIZE, 1);
659 Locking and error checking are left as an exercise for the reader.
660 Don't forget them or you'll be quite sorry.
662 =head2 How do I get a file's timestamp in perl?
664 If you want to retrieve the time at which the file was last
665 read, written, or had its meta-data (owner, etc) changed,
666 you use the B<-M>, B<-A>, or B<-C> file test operations as
667 documented in L<perlfunc>. These retrieve the age of the
668 file (measured against the start-time of your program) in
669 days as a floating point number. Some platforms may not have
670 all of these times. See L<perlport> for details. To
671 retrieve the "raw" time in seconds since the epoch, you
672 would call the stat function, then use localtime(),
673 gmtime(), or POSIX::strftime() to convert this into
678 $write_secs = (stat($file))[9];
679 printf "file %s updated at %s\n", $file,
680 scalar localtime($write_secs);
682 If you prefer something more legible, use the File::stat module
683 (part of the standard distribution in version 5.004 and later):
685 # error checking left as an exercise for reader.
688 $date_string = ctime(stat($file)->mtime);
689 print "file $file updated at $date_string\n";
691 The POSIX::strftime() approach has the benefit of being,
692 in theory, independent of the current locale. See L<perllocale>
695 =head2 How do I set a file's timestamp in perl?
697 You use the utime() function documented in L<perlfunc/utime>.
698 By way of example, here's a little program that copies the
699 read and write times from its first argument to all the rest
703 die "usage: cptimes timestamp_file other_files ...\n";
706 ($atime, $mtime) = (stat($timestamp))[8,9];
707 utime $atime, $mtime, @ARGV;
709 Error checking is, as usual, left as an exercise for the reader.
711 The perldoc for utime also has an example that has the same
712 effect as touch(1) on files that I<already exist>.
714 Certain file systems have a limited ability to store the times
715 on a file at the expected level of precision. For example, the
716 FAT and HPFS filesystem are unable to create dates on files with
717 a finer granularity than two seconds. This is a limitation of
718 the filesystems, not of utime().
720 =head2 How do I print to more than one file at once?
722 To connect one filehandle to several output filehandles,
723 you can use the IO::Tee or Tie::FileHandle::Multiplex modules.
725 If you only have to do this once, you can print individually
728 for $fh (FH1, FH2, FH3) { print $fh "whatever\n" }
730 =head2 How can I read in an entire file all at once?
732 You can use the File::Slurp module to do it in one step.
736 $all_of_it = read_file($filename); # entire file in scalar
737 @all_lines = read_file($filename); # one line perl element
739 The customary Perl approach for processing all the lines in a file is to
740 do so one line at a time:
742 open (INPUT, $file) || die "can't open $file: $!";
745 # do something with $_
747 close(INPUT) || die "can't close $file: $!";
749 This is tremendously more efficient than reading the entire file into
750 memory as an array of lines and then processing it one element at a time,
751 which is often--if not almost always--the wrong approach. Whenever
752 you see someone do this:
756 you should think long and hard about why you need everything loaded at
757 once. It's just not a scalable solution. You might also find it more
758 fun to use the standard Tie::File module, or the DB_File module's
759 $DB_RECNO bindings, which allow you to tie an array to a file so that
760 accessing an element the array actually accesses the corresponding
763 You can read the entire filehandle contents into a scalar.
767 open (INPUT, $file) || die "can't open $file: $!";
771 That temporarily undefs your record separator, and will automatically
772 close the file at block exit. If the file is already open, just use this:
774 $var = do { local $/; <INPUT> };
776 For ordinary files you can also use the read function.
778 read( INPUT, $var, -s INPUT );
780 The third argument tests the byte size of the data on the INPUT filehandle
781 and reads that many bytes into the buffer $var.
783 =head2 How can I read in a file by paragraphs?
785 Use the C<$/> variable (see L<perlvar> for details). You can either
786 set it to C<""> to eliminate empty paragraphs (C<"abc\n\n\n\ndef">,
787 for instance, gets treated as two paragraphs and not three), or
788 C<"\n\n"> to accept empty paragraphs.
790 Note that a blank line must have no blanks in it. Thus
791 S<C<"fred\n \nstuff\n\n">> is one paragraph, but C<"fred\n\nstuff\n\n"> is two.
793 =head2 How can I read a single character from a file? From the keyboard?
795 You can use the builtin C<getc()> function for most filehandles, but
796 it won't (easily) work on a terminal device. For STDIN, either use
797 the Term::ReadKey module from CPAN or use the sample code in
800 If your system supports the portable operating system programming
801 interface (POSIX), you can use the following code, which you'll note
802 turns off echo processing as well.
816 use POSIX qw(:termios_h);
818 my ($term, $oterm, $echo, $noecho, $fd_stdin);
820 $fd_stdin = fileno(STDIN);
822 $term = POSIX::Termios->new();
823 $term->getattr($fd_stdin);
824 $oterm = $term->getlflag();
826 $echo = ECHO | ECHOK | ICANON;
827 $noecho = $oterm & ~$echo;
830 $term->setlflag($noecho);
831 $term->setcc(VTIME, 1);
832 $term->setattr($fd_stdin, TCSANOW);
836 $term->setlflag($oterm);
837 $term->setcc(VTIME, 0);
838 $term->setattr($fd_stdin, TCSANOW);
844 sysread(STDIN, $key, 1);
853 The Term::ReadKey module from CPAN may be easier to use. Recent versions
854 include also support for non-portable systems as well.
857 open(TTY, "</dev/tty");
858 print "Gimme a char: ";
860 $key = ReadKey 0, *TTY;
862 printf "\nYou said %s, char number %03d\n",
865 =head2 How can I tell whether there's a character waiting on a filehandle?
867 The very first thing you should do is look into getting the Term::ReadKey
868 extension from CPAN. As we mentioned earlier, it now even has limited
869 support for non-portable (read: not open systems, closed, proprietary,
870 not POSIX, not Unix, etc) systems.
872 You should also check out the Frequently Asked Questions list in
873 comp.unix.* for things like this: the answer is essentially the same.
874 It's very system dependent. Here's one solution that works on BSD
879 vec($rin, fileno(STDIN), 1) = 1;
880 return $nfd = select($rin,undef,undef,0);
883 If you want to find out how many characters are waiting, there's
884 also the FIONREAD ioctl call to be looked at. The I<h2ph> tool that
885 comes with Perl tries to convert C include files to Perl code, which
886 can be C<require>d. FIONREAD ends up defined as a function in the
887 I<sys/ioctl.ph> file:
889 require 'sys/ioctl.ph';
891 $size = pack("L", 0);
892 ioctl(FH, FIONREAD(), $size) or die "Couldn't call ioctl: $!\n";
893 $size = unpack("L", $size);
895 If I<h2ph> wasn't installed or doesn't work for you, you can
896 I<grep> the include files by hand:
898 % grep FIONREAD /usr/include/*/*
899 /usr/include/asm/ioctls.h:#define FIONREAD 0x541B
901 Or write a small C program using the editor of champions:
904 #include <sys/ioctl.h>
906 printf("%#08x\n", FIONREAD);
909 % cc -o fionread fionread.c
913 And then hard code it, leaving porting as an exercise to your successor.
915 $FIONREAD = 0x4004667f; # XXX: opsys dependent
917 $size = pack("L", 0);
918 ioctl(FH, $FIONREAD, $size) or die "Couldn't call ioctl: $!\n";
919 $size = unpack("L", $size);
921 FIONREAD requires a filehandle connected to a stream, meaning that sockets,
922 pipes, and tty devices work, but I<not> files.
924 =head2 How do I do a C<tail -f> in perl?
930 The statement C<seek(GWFILE, 0, 1)> doesn't change the current position,
931 but it does clear the end-of-file condition on the handle, so that the
932 next <GWFILE> makes Perl try again to read something.
934 If that doesn't work (it relies on features of your stdio implementation),
935 then you need something more like this:
938 for ($curpos = tell(GWFILE); <GWFILE>; $curpos = tell(GWFILE)) {
939 # search for some stuff and put it into files
942 seek(GWFILE, $curpos, 0); # seek to where we had been
945 If this still doesn't work, look into the POSIX module. POSIX defines
946 the clearerr() method, which can remove the end of file condition on a
947 filehandle. The method: read until end of file, clearerr(), read some
948 more. Lather, rinse, repeat.
950 There's also a File::Tail module from CPAN.
952 =head2 How do I dup() a filehandle in Perl?
954 If you check L<perlfunc/open>, you'll see that several of the ways
955 to call open() should do the trick. For example:
957 open(LOG, ">>/foo/logfile");
958 open(STDERR, ">&LOG");
960 Or even with a literal numeric descriptor:
962 $fd = $ENV{MHCONTEXTFD};
963 open(MHCONTEXT, "<&=$fd"); # like fdopen(3S)
965 Note that "<&STDIN" makes a copy, but "<&=STDIN" make
966 an alias. That means if you close an aliased handle, all
967 aliases become inaccessible. This is not true with
970 Error checking, as always, has been left as an exercise for the reader.
972 =head2 How do I close a file descriptor by number?
974 This should rarely be necessary, as the Perl close() function is to be
975 used for things that Perl opened itself, even if it was a dup of a
976 numeric descriptor as with MHCONTEXT above. But if you really have
977 to, you may be able to do this:
979 require 'sys/syscall.ph';
980 $rc = syscall(&SYS_close, $fd + 0); # must force numeric
981 die "can't sysclose $fd: $!" unless $rc == -1;
983 Or, just use the fdopen(3S) feature of open():
987 open F, "<&=$fd" or die "Cannot reopen fd=$fd: $!";
991 =head2 Why can't I use "C:\temp\foo" in DOS paths? Why doesn't `C:\temp\foo.exe` work?
993 Whoops! You just put a tab and a formfeed into that filename!
994 Remember that within double quoted strings ("like\this"), the
995 backslash is an escape character. The full list of these is in
996 L<perlop/Quote and Quote-like Operators>. Unsurprisingly, you don't
997 have a file called "c:(tab)emp(formfeed)oo" or
998 "c:(tab)emp(formfeed)oo.exe" on your legacy DOS filesystem.
1000 Either single-quote your strings, or (preferably) use forward slashes.
1001 Since all DOS and Windows versions since something like MS-DOS 2.0 or so
1002 have treated C</> and C<\> the same in a path, you might as well use the
1003 one that doesn't clash with Perl--or the POSIX shell, ANSI C and C++,
1004 awk, Tcl, Java, or Python, just to mention a few. POSIX paths
1005 are more portable, too.
1007 =head2 Why doesn't glob("*.*") get all the files?
1009 Because even on non-Unix ports, Perl's glob function follows standard
1010 Unix globbing semantics. You'll need C<glob("*")> to get all (non-hidden)
1011 files. This makes glob() portable even to legacy systems. Your
1012 port may include proprietary globbing functions as well. Check its
1013 documentation for details.
1015 =head2 Why does Perl let me delete read-only files? Why does C<-i> clobber protected files? Isn't this a bug in Perl?
1017 This is elaborately and painstakingly described in the
1018 F<file-dir-perms> article in the "Far More Than You Ever Wanted To
1019 Know" collection in http://www.cpan.org/misc/olddoc/FMTEYEWTK.tgz .
1021 The executive summary: learn how your filesystem works. The
1022 permissions on a file say what can happen to the data in that file.
1023 The permissions on a directory say what can happen to the list of
1024 files in that directory. If you delete a file, you're removing its
1025 name from the directory (so the operation depends on the permissions
1026 of the directory, not of the file). If you try to write to the file,
1027 the permissions of the file govern whether you're allowed to.
1029 =head2 How do I select a random line from a file?
1031 Here's an algorithm from the Camel Book:
1034 rand($.) < 1 && ($line = $_) while <>;
1036 This has a significant advantage in space over reading the whole file
1037 in. You can find a proof of this method in I<The Art of Computer
1038 Programming>, Volume 2, Section 3.4.2, by Donald E. Knuth.
1040 You can use the File::Random module which provides a function
1043 use File::Random qw/random_line/;
1044 my $line = random_line($filename);
1046 Another way is to use the Tie::File module, which treats the entire
1047 file as an array. Simply access a random array element.
1049 =head2 Why do I get weird spaces when I print an array of lines?
1055 joins together the elements of C<@lines> with a space between them.
1056 If C<@lines> were C<("little", "fluffy", "clouds")> then the above
1057 statement would print
1059 little fluffy clouds
1061 but if each element of C<@lines> was a line of text, ending a newline
1062 character C<("little\n", "fluffy\n", "clouds\n")> then it would print:
1068 If your array contains lines, just print them:
1072 =head1 AUTHOR AND COPYRIGHT
1074 Copyright (c) 1997-2005 Tom Christiansen, Nathan Torkington, and
1075 other authors as noted. All rights reserved.
1077 This documentation is free; you can redistribute it and/or modify it
1078 under the same terms as Perl itself.
1080 Irrespective of its distribution, all code examples here are in the public
1081 domain. You are permitted and encouraged to use this code and any
1082 derivatives thereof in your own programs for fun or for profit as you
1083 see fit. A simple comment in the code giving credit to the FAQ would
1084 be courteous but is not required.