This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Once more, the disappearing typo!
[perl5.git] / pod / perlvar.pod
CommitLineData
a0d0e21e
LW
1=head1 NAME
2
3perlvar - Perl predefined variables
4
5=head1 DESCRIPTION
6
7=head2 Predefined Names
8
9The following names have special meaning to Perl. Most of the
10punctuational names have reasonable mnemonics, or analogues in one of
11the shells. Nevertheless, if you wish to use the long variable names,
12you just need to say
13
14 use English;
15
16at the top of your program. This will alias all the short names to the
17long names in the current package. Some of them even have medium names,
18generally borrowed from B<awk>.
19
20To go a step further, those variables that depend on the currently
21selected filehandle may instead be set by calling an object method on
22the FileHandle object. (Summary lines below for this contain the word
23HANDLE.) First you must say
24
25 use FileHandle;
26
27after which you may use either
28
29 method HANDLE EXPR
30
31or
32
33 HANDLE->method(EXPR)
34
35Each of the methods returns the old value of the FileHandle attribute.
36The methods each take an optional EXPR, which if supplied specifies the
37new value for the FileHandle attribute in question. If not supplied,
38most of the methods do nothing to the current value, except for
39autoflush(), which will assume a 1 for you, just to be different.
40
748a9306
LW
41A few of these variables are considered "read-only". This means that if
42you try to assign to this variable, either directly or indirectly through
43a reference, you'll raise a run-time exception.
a0d0e21e
LW
44
45=over 8
46
47=item $ARG
48
49=item $_
50
51The default input and pattern-searching space. The following pairs are
52equivalent:
53
54 while (<>) {...} # only equivalent in while!
55 while ($_ = <>) {...}
56
57 /^Subject:/
58 $_ =~ /^Subject:/
59
60 tr/a-z/A-Z/
61 $_ =~ tr/a-z/A-Z/
62
63 chop
64 chop($_)
65
cb1a09d0
AD
66Here are the places where Perl will assume $_ even if you
67don't use it:
68
69=over 3
70
71=item *
72
73Various unary functions, including functions like ord() and int(), as well
74as the all file tests (C<-f>, C<-d>) except for C<-t>, which defaults to
75STDIN.
76
77=item *
78
79Various list functions like print() and unlink().
80
81=item *
82
83The pattern matching operations C<m//>, C<s///>, and C<tr///> when used
84without an C<=~> operator.
85
86=item *
87
88The default iterator variable in a C<foreach> loop if no other
89variable is supplied.
90
91=item *
92
93The implicit iterator variable in the grep() and map() functions.
94
95=item *
96
97The default place to put an input record when a C<E<lt>FHE<gt>>
98operation's result is tested by itself as the sole criterion of a C<while>
99test. Note that outside of a C<while> test, this will not happen.
100
101=back
102
a0d0e21e
LW
103(Mnemonic: underline is understood in certain operations.)
104
105=item $<I<digit>>
106
107Contains the subpattern from the corresponding set of parentheses in
108the last pattern matched, not counting patterns matched in nested
109blocks that have been exited already. (Mnemonic: like \digit.)
110These variables are all read-only.
111
112=item $MATCH
113
114=item $&
115
116The string matched by the last successful pattern match (not counting
117any matches hidden within a BLOCK or eval() enclosed by the current
118BLOCK). (Mnemonic: like & in some editors.) This variable is read-only.
119
120=item $PREMATCH
121
122=item $`
123
124The string preceding whatever was matched by the last successful
125pattern match (not counting any matches hidden within a BLOCK or eval
126enclosed by the current BLOCK). (Mnemonic: ` often precedes a quoted
127string.) This variable is read-only.
128
129=item $POSTMATCH
130
131=item $'
132
133The string following whatever was matched by the last successful
134pattern match (not counting any matches hidden within a BLOCK or eval()
135enclosed by the current BLOCK). (Mnemonic: ' often follows a quoted
136string.) Example:
137
138 $_ = 'abcdefghi';
139 /def/;
140 print "$`:$&:$'\n"; # prints abc:def:ghi
141
142This variable is read-only.
143
144=item $LAST_PAREN_MATCH
145
146=item $+
147
148The last bracket matched by the last search pattern. This is useful if
149you don't know which of a set of alternative patterns matched. For
150example:
151
152 /Version: (.*)|Revision: (.*)/ && ($rev = $+);
153
154(Mnemonic: be positive and forward looking.)
155This variable is read-only.
156
157=item $MULTILINE_MATCHING
158
159=item $*
160
161Set to 1 to do multiline matching within a string, 0 to tell Perl
162that it can assume that strings contain a single line, for the purpose
163of optimizing pattern matches. Pattern matches on strings containing
164multiple newlines can produce confusing results when "C<$*>" is 0. Default
165is 0. (Mnemonic: * matches multiple things.) Note that this variable
166only influences the interpretation of "C<^>" and "C<$>". A literal newline can
167be searched for even when C<$* == 0>.
168
169Use of "C<$*>" is deprecated in Perl 5.
170
171=item input_line_number HANDLE EXPR
172
173=item $INPUT_LINE_NUMBER
174
175=item $NR
176
177=item $.
178
4633a7c4
LW
179The current input line number of the last filehandle that was read. An
180explicit close on the filehandle resets the line number. Since
181"C<E<lt>E<gt>>" never does an explicit close, line numbers increase
182across ARGV files (but see examples under eof()). Localizing C<$.> has
183the effect of also localizing Perl's notion of "the last read
184filehandle". (Mnemonic: many programs use "." to mean the current line
185number.)
a0d0e21e
LW
186
187=item input_record_separator HANDLE EXPR
188
189=item $INPUT_RECORD_SEPARATOR
190
191=item $RS
192
193=item $/
194
195The input record separator, newline by default. Works like B<awk>'s RS
196variable, including treating blank lines as delimiters if set to the
197null string. You may set it to a multicharacter string to match a
198multi-character delimiter. Note that setting it to C<"\n\n"> means
199something slightly different than setting it to C<"">, if the file
200contains consecutive blank lines. Setting it to C<""> will treat two or
201more consecutive blank lines as a single blank line. Setting it to
202C<"\n\n"> will blindly assume that the next input character belongs to the
203next paragraph, even if it's a newline. (Mnemonic: / is used to
204delimit line boundaries when quoting poetry.)
205
206 undef $/;
207 $_ = <FH>; # whole file now here
208 s/\n[ \t]+/ /g;
209
210=item autoflush HANDLE EXPR
211
212=item $OUTPUT_AUTOFLUSH
213
214=item $|
215
216If set to nonzero, forces a flush after every write or print on the
217currently selected output channel. Default is 0. Note that STDOUT
218will typically be line buffered if output is to the terminal and block
219buffered otherwise. Setting this variable is useful primarily when you
220are outputting to a pipe, such as when you are running a Perl script
cb1a09d0
AD
221under rsh and want to see the output as it's happening. This has no
222effect on input buffering.
223(Mnemonic: when you want your pipes to be piping hot.)
a0d0e21e
LW
224
225=item output_field_separator HANDLE EXPR
226
227=item $OUTPUT_FIELD_SEPARATOR
228
229=item $OFS
230
231=item $,
232
233The output field separator for the print operator. Ordinarily the
234print operator simply prints out the comma separated fields you
235specify. In order to get behavior more like B<awk>, set this variable
236as you would set B<awk>'s OFS variable to specify what is printed
237between fields. (Mnemonic: what is printed when there is a , in your
238print statement.)
239
240=item output_record_separator HANDLE EXPR
241
242=item $OUTPUT_RECORD_SEPARATOR
243
244=item $ORS
245
246=item $\
247
248The output record separator for the print operator. Ordinarily the
249print operator simply prints out the comma separated fields you
250specify, with no trailing newline or record separator assumed. In
251order to get behavior more like B<awk>, set this variable as you would
252set B<awk>'s ORS variable to specify what is printed at the end of the
253print. (Mnemonic: you set "C<$\>" instead of adding \n at the end of the
254print. Also, it's just like /, but it's what you get "back" from
255Perl.)
256
257=item $LIST_SEPARATOR
258
259=item $"
260
261This is like "C<$,>" except that it applies to array values interpolated
262into a double-quoted string (or similar interpreted string). Default
263is a space. (Mnemonic: obvious, I think.)
264
265=item $SUBSCRIPT_SEPARATOR
266
267=item $SUBSEP
268
269=item $;
270
271The subscript separator for multi-dimensional array emulation. If you
272refer to a hash element as
273
274 $foo{$a,$b,$c}
275
276it really means
277
278 $foo{join($;, $a, $b, $c)}
279
280But don't put
281
282 @foo{$a,$b,$c} # a slice--note the @
283
284which means
285
286 ($foo{$a},$foo{$b},$foo{$c})
287
288Default is "\034", the same as SUBSEP in B<awk>. Note that if your
289keys contain binary data there might not be any safe value for "C<$;>".
290(Mnemonic: comma (the syntactic subscript separator) is a
291semi-semicolon. Yeah, I know, it's pretty lame, but "C<$,>" is already
292taken for something more important.)
293
294Consider using "real" multi-dimensional arrays in Perl 5.
295
296=item $OFMT
297
298=item $#
299
300The output format for printed numbers. This variable is a half-hearted
301attempt to emulate B<awk>'s OFMT variable. There are times, however,
302when B<awk> and Perl have differing notions of what is in fact
303numeric. Also, the initial value is %.20g rather than %.6g, so you
304need to set "C<$#>" explicitly to get B<awk>'s value. (Mnemonic: # is the
305number sign.)
306
307Use of "C<$#>" is deprecated in Perl 5.
308
309=item format_page_number HANDLE EXPR
310
311=item $FORMAT_PAGE_NUMBER
312
313=item $%
314
315The current page number of the currently selected output channel.
316(Mnemonic: % is page number in B<nroff>.)
317
318=item format_lines_per_page HANDLE EXPR
319
320=item $FORMAT_LINES_PER_PAGE
321
322=item $=
323
324The current page length (printable lines) of the currently selected
325output channel. Default is 60. (Mnemonic: = has horizontal lines.)
326
327=item format_lines_left HANDLE EXPR
328
329=item $FORMAT_LINES_LEFT
330
331=item $-
332
333The number of lines left on the page of the currently selected output
334channel. (Mnemonic: lines_on_page - lines_printed.)
335
336=item format_name HANDLE EXPR
337
338=item $FORMAT_NAME
339
340=item $~
341
342The name of the current report format for the currently selected output
343channel. Default is name of the filehandle. (Mnemonic: brother to
344"C<$^>".)
345
346=item format_top_name HANDLE EXPR
347
348=item $FORMAT_TOP_NAME
349
350=item $^
351
352The name of the current top-of-page format for the currently selected
353output channel. Default is name of the filehandle with _TOP
354appended. (Mnemonic: points to top of page.)
355
356=item format_line_break_characters HANDLE EXPR
357
358=item $FORMAT_LINE_BREAK_CHARACTERS
359
360=item $:
361
362The current set of characters after which a string may be broken to
363fill continuation fields (starting with ^) in a format. Default is
364S<" \n-">, to break on whitespace or hyphens. (Mnemonic: a "colon" in
365poetry is a part of a line.)
366
367=item format_formfeed HANDLE EXPR
368
369=item $FORMAT_FORMFEED
370
371=item $^L
372
373What formats output to perform a formfeed. Default is \f.
374
375=item $ACCUMULATOR
376
377=item $^A
378
379The current value of the write() accumulator for format() lines. A format
380contains formline() commands that put their result into C<$^A>. After
381calling its format, write() prints out the contents of C<$^A> and empties.
382So you never actually see the contents of C<$^A> unless you call
383formline() yourself and then look at it. See L<perlform> and
384L<perlfunc/formline()>.
385
386=item $CHILD_ERROR
387
388=item $?
389
390The status returned by the last pipe close, backtick (C<``>) command,
391or system() operator. Note that this is the status word returned by
392the wait() system call, so the exit value of the subprocess is actually
393(C<$? E<gt>E<gt> 8>). Thus on many systems, C<$? & 255> gives which signal,
394if any, the process died from, and whether there was a core dump.
395(Mnemonic: similar to B<sh> and B<ksh>.)
396
397=item $OS_ERROR
398
399=item $ERRNO
400
401=item $!
402
403If used in a numeric context, yields the current value of errno, with
404all the usual caveats. (This means that you shouldn't depend on the
405value of "C<$!>" to be anything in particular unless you've gotten a
406specific error return indicating a system error.) If used in a string
407context, yields the corresponding system error string. You can assign
408to "C<$!>" in order to set I<errno> if, for instance, you want "C<$!>" to return the
409string for error I<n>, or you want to set the exit value for the die()
410operator. (Mnemonic: What just went bang?)
411
412=item $EVAL_ERROR
413
414=item $@
415
416The Perl syntax error message from the last eval() command. If null, the
417last eval() parsed and executed correctly (although the operations you
418invoked may have failed in the normal fashion). (Mnemonic: Where was
419the syntax error "at"?)
420
748a9306
LW
421Note that warning messages are not collected in this variable. You can,
422however, set up a routine to process warnings by setting $SIG{__WARN__} below.
423
a0d0e21e
LW
424=item $PROCESS_ID
425
426=item $PID
427
428=item $$
429
430The process number of the Perl running this script. (Mnemonic: same
431as shells.)
432
433=item $REAL_USER_ID
434
435=item $UID
436
437=item $<
438
439The real uid of this process. (Mnemonic: it's the uid you came I<FROM>,
440if you're running setuid.)
441
442=item $EFFECTIVE_USER_ID
443
444=item $EUID
445
446=item $>
447
448The effective uid of this process. Example:
449
450 $< = $>; # set real to effective uid
451 ($<,$>) = ($>,$<); # swap real and effective uid
452
453(Mnemonic: it's the uid you went I<TO>, if you're running setuid.) Note:
454"C<$E<lt>>" and "C<$E<gt>>" can only be swapped on machines supporting setreuid().
455
456=item $REAL_GROUP_ID
457
458=item $GID
459
460=item $(
461
462The real gid of this process. If you are on a machine that supports
463membership in multiple groups simultaneously, gives a space separated
464list of groups you are in. The first number is the one returned by
465getgid(), and the subsequent ones by getgroups(), one of which may be
466the same as the first number. (Mnemonic: parentheses are used to I<GROUP>
467things. The real gid is the group you I<LEFT>, if you're running setgid.)
468
469=item $EFFECTIVE_GROUP_ID
470
471=item $EGID
472
473=item $)
474
475The effective gid of this process. If you are on a machine that
476supports membership in multiple groups simultaneously, gives a space
477separated list of groups you are in. The first number is the one
478returned by getegid(), and the subsequent ones by getgroups(), one of
479which may be the same as the first number. (Mnemonic: parentheses are
480used to I<GROUP> things. The effective gid is the group that's I<RIGHT> for
481you, if you're running setgid.)
482
483Note: "C<$E<lt>>", "C<$E<gt>>", "C<$(>" and "C<$)>" can only be set on machines
484that support the corresponding I<set[re][ug]id()> routine. "C<$(>" and "C<$)>"
cb1a09d0
AD
485can only be swapped on machines supporting setregid(). Because Perl doesn't
486currently use initgroups(), you can't set your group vector to multiple groups.
a0d0e21e
LW
487
488=item $PROGRAM_NAME
489
490=item $0
491
492Contains the name of the file containing the Perl script being
493executed. Assigning to "C<$0>" modifies the argument area that the ps(1)
494program sees. This is more useful as a way of indicating the
495current program state than it is for hiding the program you're running.
496(Mnemonic: same as B<sh> and B<ksh>.)
497
498=item $[
499
500The index of the first element in an array, and of the first character
501in a substring. Default is 0, but you could set it to 1 to make
502Perl behave more like B<awk> (or Fortran) when subscripting and when
503evaluating the index() and substr() functions. (Mnemonic: [ begins
504subscripts.)
505
506As of Perl 5, assignment to "C<$[>" is treated as a compiler directive,
507and cannot influence the behavior of any other file. Its use is
508discouraged.
509
510=item $PERL_VERSION
511
512=item $]
513
cb1a09d0
AD
514The string printed out when you say C<perl -v>.
515(This is currently I<BROKEN>).
516It can be used to
a0d0e21e
LW
517determine at the beginning of a script whether the perl interpreter
518executing the script is in the right range of versions. If used in a
519numeric context, returns the version + patchlevel / 1000. Example:
520
521 # see if getc is available
522 ($version,$patchlevel) =
523 $] =~ /(\d+\.\d+).*\nPatch level: (\d+)/;
524 print STDERR "(No filename completion available.)\n"
525 if $version * 1000 + $patchlevel < 2016;
526
527or, used numerically,
528
529 warn "No checksumming!\n" if $] < 3.019;
530
531(Mnemonic: Is this version of perl in the right bracket?)
532
533=item $DEBUGGING
534
535=item $^D
536
537The current value of the debugging flags. (Mnemonic: value of B<-D>
538switch.)
539
540=item $SYSTEM_FD_MAX
541
542=item $^F
543
544The maximum system file descriptor, ordinarily 2. System file
545descriptors are passed to exec()ed processes, while higher file
546descriptors are not. Also, during an open(), system file descriptors are
547preserved even if the open() fails. (Ordinary file descriptors are
548closed before the open() is attempted.) Note that the close-on-exec
549status of a file descriptor will be decided according to the value of
550C<$^F> at the time of the open, not the time of the exec.
551
552=item $INPLACE_EDIT
553
554=item $^I
555
556The current value of the inplace-edit extension. Use C<undef> to disable
557inplace editing. (Mnemonic: value of B<-i> switch.)
558
559=item $PERLDB
560
561=item $^P
562
563The internal flag that the debugger clears so that it doesn't debug
564itself. You could conceivable disable debugging yourself by clearing
565it.
566
567=item $BASETIME
568
569=item $^T
570
571The time at which the script began running, in seconds since the
572epoch (beginning of 1970). The values returned by the B<-M>, B<-A>
573and B<-C> filetests are
574based on this value.
575
576=item $WARNING
577
578=item $^W
579
580The current value of the warning switch, either TRUE or FALSE. (Mnemonic: related to the
581B<-w> switch.)
582
583=item $EXECUTABLE_NAME
584
585=item $^X
586
587The name that the Perl binary itself was executed as, from C's C<argv[0]>.
588
589=item $ARGV
590
591contains the name of the current file when reading from <>.
592
593=item @ARGV
594
595The array @ARGV contains the command line arguments intended for the
596script. Note that C<$#ARGV> is the generally number of arguments minus
597one, since C<$ARGV[0]> is the first argument, I<NOT> the command name. See
598"C<$0>" for the command name.
599
600=item @INC
601
602The array @INC contains the list of places to look for Perl scripts to
603be evaluated by the C<do EXPR>, C<require>, or C<use> constructs. It
604initially consists of the arguments to any B<-I> command line switches,
605followed by the default Perl library, probably "/usr/local/lib/perl",
cb1a09d0
AD
606followed by ".", to represent the current directory. If you need to
607modify this at runtime, you should use the C<use lib> pragma in order
608to also get the machine-dependent library properly loaded:
a0d0e21e 609
cb1a09d0
AD
610 use lib '/mypath/libdir/';
611 use SomeMod;
612
a0d0e21e
LW
613=item %INC
614
615The hash %INC contains entries for each filename that has
616been included via C<do> or C<require>. The key is the filename you
617specified, and the value is the location of the file actually found.
618The C<require> command uses this array to determine whether a given file
619has already been included.
620
621=item $ENV{expr}
622
623The hash %ENV contains your current environment. Setting a
624value in C<ENV> changes the environment for child processes.
625
626=item $SIG{expr}
627
628The hash %SIG is used to set signal handlers for various
629signals. Example:
630
631 sub handler { # 1st argument is signal name
632 local($sig) = @_;
633 print "Caught a SIG$sig--shutting down\n";
634 close(LOG);
635 exit(0);
636 }
637
638 $SIG{'INT'} = 'handler';
639 $SIG{'QUIT'} = 'handler';
640 ...
641 $SIG{'INT'} = 'DEFAULT'; # restore default action
642 $SIG{'QUIT'} = 'IGNORE'; # ignore SIGQUIT
643
644The %SIG array only contains values for the signals actually set within
645the Perl script. Here are some other examples:
646
647 $SIG{PIPE} = Plumber; # SCARY!!
648 $SIG{"PIPE"} = "Plumber"; # just fine, assumes main::Plumber
649 $SIG{"PIPE"} = \&Plumber; # just fine; assume current Plumber
650 $SIG{"PIPE"} = Plumber(); # oops, what did Plumber() return??
651
652The one marked scary is problematic because it's a bareword, which means
653sometimes it's a string representing the function, and sometimes it's
654going to call the subroutine call right then and there! Best to be sure
748a9306
LW
655and quote it or take a reference to it. *Plumber works too. See L<perlsubs>.
656
657Certain internal hooks can be also set using the %SIG hash. The
658routine indicated by $SIG{__WARN__} is called when a warning message is
659about to be printed. The warning message is passed as the first
660argument. The presence of a __WARN__ hook causes the ordinary printing
661of warnings to STDERR to be suppressed. You can use this to save warnings
662in a variable, or turn warnings into fatal errors, like this:
663
664 local $SIG{__WARN__} = sub { die $_[0] };
665 eval $proggie;
666
667The routine indicated by $SIG{__DIE__} is called when a fatal exception
668is about to be thrown. The error message is passed as the first
669argument. When a __DIE__ hook routine returns, the exception
670processing continues as it would have in the absence of the hook,
cb1a09d0
AD
671unless the hook routine itself exits via a C<goto>, a loop exit, or a die().
672The __DIE__ handler is explicitly disabled during the call, so that you
673can die from a __DIE__ handler. Similarly for __WARN__.
a0d0e21e
LW
674
675=back
676