This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Retract #8865 and #8869, un?pack C now again agree with Camel 3
[perl5.git] / pod / perlop.pod
CommitLineData
a0d0e21e
LW
1=head1 NAME
2
3perlop - Perl operators and precedence
4
5=head1 SYNOPSIS
6
7Perl operators have the following associativity and precedence,
19799a22
GS
8listed from highest precedence to lowest. Operators borrowed from
9C keep the same precedence relationship with each other, even where
10C's precedence is slightly screwy. (This makes learning Perl easier
11for C folks.) With very few exceptions, these all operate on scalar
12values only, not array values.
a0d0e21e
LW
13
14 left terms and list operators (leftward)
15 left ->
16 nonassoc ++ --
17 right **
18 right ! ~ \ and unary + and -
54310121 19 left =~ !~
a0d0e21e
LW
20 left * / % x
21 left + - .
22 left << >>
23 nonassoc named unary operators
24 nonassoc < > <= >= lt gt le ge
25 nonassoc == != <=> eq ne cmp
26 left &
27 left | ^
28 left &&
29 left ||
137443ea 30 nonassoc .. ...
a0d0e21e
LW
31 right ?:
32 right = += -= *= etc.
33 left , =>
34 nonassoc list operators (rightward)
a5f75d66 35 right not
a0d0e21e
LW
36 left and
37 left or xor
38
39In the following sections, these operators are covered in precedence order.
40
5a964f20
TC
41Many operators can be overloaded for objects. See L<overload>.
42
cb1a09d0 43=head1 DESCRIPTION
a0d0e21e
LW
44
45=head2 Terms and List Operators (Leftward)
46
62c18ce2 47A TERM has the highest precedence in Perl. They include variables,
5f05dabc 48quote and quote-like operators, any expression in parentheses,
a0d0e21e
LW
49and any function whose arguments are parenthesized. Actually, there
50aren't really functions in this sense, just list operators and unary
51operators behaving as functions because you put parentheses around
52the arguments. These are all documented in L<perlfunc>.
53
54If any list operator (print(), etc.) or any unary operator (chdir(), etc.)
55is followed by a left parenthesis as the next token, the operator and
56arguments within parentheses are taken to be of highest precedence,
57just like a normal function call.
58
59In the absence of parentheses, the precedence of list operators such as
60C<print>, C<sort>, or C<chmod> is either very high or very low depending on
54310121 61whether you are looking at the left side or the right side of the operator.
a0d0e21e
LW
62For example, in
63
64 @ary = (1, 3, sort 4, 2);
65 print @ary; # prints 1324
66
19799a22
GS
67the commas on the right of the sort are evaluated before the sort,
68but the commas on the left are evaluated after. In other words,
69list operators tend to gobble up all arguments that follow, and
a0d0e21e 70then act like a simple TERM with regard to the preceding expression.
19799a22 71Be careful with parentheses:
a0d0e21e
LW
72
73 # These evaluate exit before doing the print:
74 print($foo, exit); # Obviously not what you want.
75 print $foo, exit; # Nor is this.
76
77 # These do the print before evaluating exit:
78 (print $foo), exit; # This is what you want.
79 print($foo), exit; # Or this.
80 print ($foo), exit; # Or even this.
81
82Also note that
83
84 print ($foo & 255) + 1, "\n";
85
54310121 86probably doesn't do what you expect at first glance. See
a0d0e21e
LW
87L<Named Unary Operators> for more discussion of this.
88
89Also parsed as terms are the C<do {}> and C<eval {}> constructs, as
54310121 90well as subroutine and method calls, and the anonymous
a0d0e21e
LW
91constructors C<[]> and C<{}>.
92
2ae324a7 93See also L<Quote and Quote-like Operators> toward the end of this section,
c07a80fd 94as well as L<"I/O Operators">.
a0d0e21e
LW
95
96=head2 The Arrow Operator
97
35f2feb0 98"C<< -> >>" is an infix dereference operator, just as it is in C
19799a22
GS
99and C++. If the right side is either a C<[...]>, C<{...}>, or a
100C<(...)> subscript, then the left side must be either a hard or
101symbolic reference to an array, a hash, or a subroutine respectively.
102(Or technically speaking, a location capable of holding a hard
103reference, if it's an array or hash reference being used for
104assignment.) See L<perlreftut> and L<perlref>.
a0d0e21e 105
19799a22
GS
106Otherwise, the right side is a method name or a simple scalar
107variable containing either the method name or a subroutine reference,
108and the left side must be either an object (a blessed reference)
109or a class name (that is, a package name). See L<perlobj>.
a0d0e21e 110
5f05dabc 111=head2 Auto-increment and Auto-decrement
a0d0e21e
LW
112
113"++" and "--" work as in C. That is, if placed before a variable, they
114increment or decrement the variable before returning the value, and if
115placed after, increment or decrement the variable after returning the value.
116
54310121 117The auto-increment operator has a little extra builtin magic to it. If
a0d0e21e
LW
118you increment a variable that is numeric, or that has ever been used in
119a numeric context, you get a normal increment. If, however, the
5f05dabc 120variable has been used in only string contexts since it was set, and
5a964f20 121has a value that is not the empty string and matches the pattern
9c0670e1 122C</^[a-zA-Z]*[0-9]*\z/>, the increment is done as a string, preserving each
a0d0e21e
LW
123character within its range, with carry:
124
125 print ++($foo = '99'); # prints '100'
126 print ++($foo = 'a0'); # prints 'a1'
127 print ++($foo = 'Az'); # prints 'Ba'
128 print ++($foo = 'zz'); # prints 'aaa'
129
5f05dabc 130The auto-decrement operator is not magical.
a0d0e21e
LW
131
132=head2 Exponentiation
133
19799a22 134Binary "**" is the exponentiation operator. It binds even more
cb1a09d0
AD
135tightly than unary minus, so -2**4 is -(2**4), not (-2)**4. (This is
136implemented using C's pow(3) function, which actually works on doubles
137internally.)
a0d0e21e
LW
138
139=head2 Symbolic Unary Operators
140
5f05dabc 141Unary "!" performs logical negation, i.e., "not". See also C<not> for a lower
a0d0e21e
LW
142precedence version of this.
143
144Unary "-" performs arithmetic negation if the operand is numeric. If
145the operand is an identifier, a string consisting of a minus sign
146concatenated with the identifier is returned. Otherwise, if the string
147starts with a plus or minus, a string starting with the opposite sign
148is returned. One effect of these rules is that C<-bareword> is equivalent
149to C<"-bareword">.
150
972b05a9
JH
151Unary "~" performs bitwise negation, i.e., 1's complement. For
152example, C<0666 & ~027> is 0640. (See also L<Integer Arithmetic> and
153L<Bitwise String Operators>.) Note that the width of the result is
154platform-dependent: ~0 is 32 bits wide on a 32-bit platform, but 64
155bits wide on a 64-bit platform, so if you are expecting a certain bit
156width, remember use the & operator to mask off the excess bits.
a0d0e21e
LW
157
158Unary "+" has no effect whatsoever, even on strings. It is useful
159syntactically for separating a function name from a parenthesized expression
160that would otherwise be interpreted as the complete list of function
5ba421f6 161arguments. (See examples above under L<Terms and List Operators (Leftward)>.)
a0d0e21e 162
19799a22
GS
163Unary "\" creates a reference to whatever follows it. See L<perlreftut>
164and L<perlref>. Do not confuse this behavior with the behavior of
165backslash within a string, although both forms do convey the notion
166of protecting the next thing from interpolation.
a0d0e21e
LW
167
168=head2 Binding Operators
169
c07a80fd 170Binary "=~" binds a scalar expression to a pattern match. Certain operations
cb1a09d0
AD
171search or modify the string $_ by default. This operator makes that kind
172of operation work on some other string. The right argument is a search
2c268ad5
TP
173pattern, substitution, or transliteration. The left argument is what is
174supposed to be searched, substituted, or transliterated instead of the default
f8bab1e9
GS
175$_. When used in scalar context, the return value generally indicates the
176success of the operation. Behavior in list context depends on the particular
177operator. See L</"Regexp Quote-Like Operators"> for details.
178
179If the right argument is an expression rather than a search pattern,
2c268ad5 180substitution, or transliteration, it is interpreted as a search pattern at run
2decb4fb
GS
181time. This can be less efficient than an explicit search, because the
182pattern must be compiled every time the expression is evaluated.
a0d0e21e
LW
183
184Binary "!~" is just like "=~" except the return value is negated in
185the logical sense.
186
187=head2 Multiplicative Operators
188
189Binary "*" multiplies two numbers.
190
191Binary "/" divides two numbers.
192
54310121
PP
193Binary "%" computes the modulus of two numbers. Given integer
194operands C<$a> and C<$b>: If C<$b> is positive, then C<$a % $b> is
195C<$a> minus the largest multiple of C<$b> that is not greater than
196C<$a>. If C<$b> is negative, then C<$a % $b> is C<$a> minus the
197smallest multiple of C<$b> that is not less than C<$a> (i.e. the
6bb4e6d4 198result will be less than or equal to zero).
f3798619 199Note than when C<use integer> is in scope, "%" gives you direct access
55d729e4
GS
200to the modulus operator as implemented by your C compiler. This
201operator is not as well defined for negative operands, but it will
202execute faster.
203
62d10b70
GS
204Binary "x" is the repetition operator. In scalar context or if the left
205operand is not enclosed in parentheses, it returns a string consisting
206of the left operand repeated the number of times specified by the right
207operand. In list context, if the left operand is enclosed in
208parentheses, it repeats the list.
a0d0e21e
LW
209
210 print '-' x 80; # print row of dashes
211
212 print "\t" x ($tab/8), ' ' x ($tab%8); # tab over
213
214 @ones = (1) x 80; # a list of 80 1's
215 @ones = (5) x @ones; # set all elements to 5
216
217
218=head2 Additive Operators
219
220Binary "+" returns the sum of two numbers.
221
222Binary "-" returns the difference of two numbers.
223
224Binary "." concatenates two strings.
225
226=head2 Shift Operators
227
55497cff
PP
228Binary "<<" returns the value of its left argument shifted left by the
229number of bits specified by the right argument. Arguments should be
982ce180 230integers. (See also L<Integer Arithmetic>.)
a0d0e21e 231
55497cff
PP
232Binary ">>" returns the value of its left argument shifted right by
233the number of bits specified by the right argument. Arguments should
982ce180 234be integers. (See also L<Integer Arithmetic>.)
a0d0e21e
LW
235
236=head2 Named Unary Operators
237
238The various named unary operators are treated as functions with one
239argument, with optional parentheses. These include the filetest
240operators, like C<-f>, C<-M>, etc. See L<perlfunc>.
241
242If any list operator (print(), etc.) or any unary operator (chdir(), etc.)
243is followed by a left parenthesis as the next token, the operator and
244arguments within parentheses are taken to be of highest precedence,
245just like a normal function call. Examples:
246
247 chdir $foo || die; # (chdir $foo) || die
248 chdir($foo) || die; # (chdir $foo) || die
249 chdir ($foo) || die; # (chdir $foo) || die
250 chdir +($foo) || die; # (chdir $foo) || die
251
252but, because * is higher precedence than ||:
253
254 chdir $foo * 20; # chdir ($foo * 20)
255 chdir($foo) * 20; # (chdir $foo) * 20
256 chdir ($foo) * 20; # (chdir $foo) * 20
257 chdir +($foo) * 20; # chdir ($foo * 20)
258
259 rand 10 * 20; # rand (10 * 20)
260 rand(10) * 20; # (rand 10) * 20
261 rand (10) * 20; # (rand 10) * 20
262 rand +(10) * 20; # rand (10 * 20)
263
5ba421f6 264See also L<"Terms and List Operators (Leftward)">.
a0d0e21e
LW
265
266=head2 Relational Operators
267
35f2feb0 268Binary "<" returns true if the left argument is numerically less than
a0d0e21e
LW
269the right argument.
270
35f2feb0 271Binary ">" returns true if the left argument is numerically greater
a0d0e21e
LW
272than the right argument.
273
35f2feb0 274Binary "<=" returns true if the left argument is numerically less than
a0d0e21e
LW
275or equal to the right argument.
276
35f2feb0 277Binary ">=" returns true if the left argument is numerically greater
a0d0e21e
LW
278than or equal to the right argument.
279
280Binary "lt" returns true if the left argument is stringwise less than
281the right argument.
282
283Binary "gt" returns true if the left argument is stringwise greater
284than the right argument.
285
286Binary "le" returns true if the left argument is stringwise less than
287or equal to the right argument.
288
289Binary "ge" returns true if the left argument is stringwise greater
290than or equal to the right argument.
291
292=head2 Equality Operators
293
294Binary "==" returns true if the left argument is numerically equal to
295the right argument.
296
297Binary "!=" returns true if the left argument is numerically not equal
298to the right argument.
299
35f2feb0 300Binary "<=>" returns -1, 0, or 1 depending on whether the left
6ee5d4e7 301argument is numerically less than, equal to, or greater than the right
d4ad863d 302argument. If your platform supports NaNs (not-a-numbers) as numeric
7d3a9d88
NC
303values, using them with "<=>" returns undef. NaN is not "<", "==", ">",
304"<=" or ">=" anything (even NaN), so those 5 return false. NaN != NaN
305returns true, as does NaN != anything else. If your platform doesn't
306support NaNs then NaN is just a string with numeric value 0.
307
308 perl -le '$a = NaN; print "No NaN support here" if $a == $a'
309 perl -le '$a = NaN; print "NaN support here" if $a != $a'
a0d0e21e
LW
310
311Binary "eq" returns true if the left argument is stringwise equal to
312the right argument.
313
314Binary "ne" returns true if the left argument is stringwise not equal
315to the right argument.
316
d4ad863d
JH
317Binary "cmp" returns -1, 0, or 1 depending on whether the left
318argument is stringwise less than, equal to, or greater than the right
319argument.
a0d0e21e 320
a034a98d
DD
321"lt", "le", "ge", "gt" and "cmp" use the collation (sort) order specified
322by the current locale if C<use locale> is in effect. See L<perllocale>.
323
a0d0e21e
LW
324=head2 Bitwise And
325
326Binary "&" returns its operators ANDed together bit by bit.
2c268ad5 327(See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
a0d0e21e
LW
328
329=head2 Bitwise Or and Exclusive Or
330
331Binary "|" returns its operators ORed together bit by bit.
2c268ad5 332(See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
a0d0e21e
LW
333
334Binary "^" returns its operators XORed together bit by bit.
2c268ad5 335(See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
a0d0e21e
LW
336
337=head2 C-style Logical And
338
339Binary "&&" performs a short-circuit logical AND operation. That is,
340if the left operand is false, the right operand is not even evaluated.
341Scalar or list context propagates down to the right operand if it
342is evaluated.
343
344=head2 C-style Logical Or
345
346Binary "||" performs a short-circuit logical OR operation. That is,
347if the left operand is true, the right operand is not even evaluated.
348Scalar or list context propagates down to the right operand if it
349is evaluated.
350
351The C<||> and C<&&> operators differ from C's in that, rather than returning
3520 or 1, they return the last value evaluated. Thus, a reasonably portable
353way to find out the home directory (assuming it's not "0") might be:
354
355 $home = $ENV{'HOME'} || $ENV{'LOGDIR'} ||
356 (getpwuid($<))[7] || die "You're homeless!\n";
357
5a964f20
TC
358In particular, this means that you shouldn't use this
359for selecting between two aggregates for assignment:
360
361 @a = @b || @c; # this is wrong
362 @a = scalar(@b) || @c; # really meant this
363 @a = @b ? @b : @c; # this works fine, though
364
365As more readable alternatives to C<&&> and C<||> when used for
366control flow, Perl provides C<and> and C<or> operators (see below).
367The short-circuit behavior is identical. The precedence of "and" and
368"or" is much lower, however, so that you can safely use them after a
369list operator without the need for parentheses:
a0d0e21e
LW
370
371 unlink "alpha", "beta", "gamma"
372 or gripe(), next LINE;
373
374With the C-style operators that would have been written like this:
375
376 unlink("alpha", "beta", "gamma")
377 || (gripe(), next LINE);
378
eeb6a2c9 379Using "or" for assignment is unlikely to do what you want; see below.
5a964f20
TC
380
381=head2 Range Operators
a0d0e21e
LW
382
383Binary ".." is the range operator, which is really two different
5a964f20 384operators depending on the context. In list context, it returns an
2cdbc966
JD
385array of values counting (up by ones) from the left value to the right
386value. If the left value is greater than the right value then it
387returns the empty array. The range operator is useful for writing
388C<foreach (1..10)> loops and for doing slice operations on arrays. In
389the current implementation, no temporary array is created when the
390range operator is used as the expression in C<foreach> loops, but older
391versions of Perl might burn a lot of memory when you write something
392like this:
a0d0e21e
LW
393
394 for (1 .. 1_000_000) {
395 # code
54310121 396 }
a0d0e21e 397
5a964f20 398In scalar context, ".." returns a boolean value. The operator is
a0d0e21e
LW
399bistable, like a flip-flop, and emulates the line-range (comma) operator
400of B<sed>, B<awk>, and various editors. Each ".." operator maintains its
401own boolean state. It is false as long as its left operand is false.
402Once the left operand is true, the range operator stays true until the
403right operand is true, I<AFTER> which the range operator becomes false
19799a22 404again. It doesn't become false till the next time the range operator is
a0d0e21e
LW
405evaluated. It can test the right operand and become false on the same
406evaluation it became true (as in B<awk>), but it still returns true once.
19799a22
GS
407If you don't want it to test the right operand till the next
408evaluation, as in B<sed>, just use three dots ("...") instead of
409two. In all other regards, "..." behaves just like ".." does.
410
411The right operand is not evaluated while the operator is in the
412"false" state, and the left operand is not evaluated while the
413operator is in the "true" state. The precedence is a little lower
414than || and &&. The value returned is either the empty string for
415false, or a sequence number (beginning with 1) for true. The
416sequence number is reset for each range encountered. The final
417sequence number in a range has the string "E0" appended to it, which
418doesn't affect its numeric value, but gives you something to search
419for if you want to exclude the endpoint. You can exclude the
420beginning point by waiting for the sequence number to be greater
421than 1. If either operand of scalar ".." is a constant expression,
422that operand is implicitly compared to the C<$.> variable, the
423current line number. Examples:
a0d0e21e
LW
424
425As a scalar operator:
426
427 if (101 .. 200) { print; } # print 2nd hundred lines
428 next line if (1 .. /^$/); # skip header lines
429 s/^/> / if (/^$/ .. eof()); # quote body
430
5a964f20
TC
431 # parse mail messages
432 while (<>) {
433 $in_header = 1 .. /^$/;
434 $in_body = /^$/ .. eof();
435 # do something based on those
436 } continue {
437 close ARGV if eof; # reset $. each file
438 }
439
a0d0e21e
LW
440As a list operator:
441
442 for (101 .. 200) { print; } # print $_ 100 times
3e3baf6d 443 @foo = @foo[0 .. $#foo]; # an expensive no-op
a0d0e21e
LW
444 @foo = @foo[$#foo-4 .. $#foo]; # slice last 5 items
445
5a964f20 446The range operator (in list context) makes use of the magical
5f05dabc 447auto-increment algorithm if the operands are strings. You
a0d0e21e
LW
448can say
449
450 @alphabet = ('A' .. 'Z');
451
19799a22 452to get all normal letters of the alphabet, or
a0d0e21e
LW
453
454 $hexdigit = (0 .. 9, 'a' .. 'f')[$num & 15];
455
456to get a hexadecimal digit, or
457
458 @z2 = ('01' .. '31'); print $z2[$mday];
459
460to get dates with leading zeros. If the final value specified is not
461in the sequence that the magical increment would produce, the sequence
462goes until the next value would be longer than the final value
463specified.
464
465=head2 Conditional Operator
466
467Ternary "?:" is the conditional operator, just as in C. It works much
468like an if-then-else. If the argument before the ? is true, the
469argument before the : is returned, otherwise the argument after the :
cb1a09d0
AD
470is returned. For example:
471
54310121 472 printf "I have %d dog%s.\n", $n,
cb1a09d0
AD
473 ($n == 1) ? '' : "s";
474
475Scalar or list context propagates downward into the 2nd
54310121 476or 3rd argument, whichever is selected.
cb1a09d0
AD
477
478 $a = $ok ? $b : $c; # get a scalar
479 @a = $ok ? @b : @c; # get an array
480 $a = $ok ? @b : @c; # oops, that's just a count!
481
482The operator may be assigned to if both the 2nd and 3rd arguments are
483legal lvalues (meaning that you can assign to them):
a0d0e21e
LW
484
485 ($a_or_b ? $a : $b) = $c;
486
5a964f20
TC
487Because this operator produces an assignable result, using assignments
488without parentheses will get you in trouble. For example, this:
489
490 $a % 2 ? $a += 10 : $a += 2
491
492Really means this:
493
494 (($a % 2) ? ($a += 10) : $a) += 2
495
496Rather than this:
497
498 ($a % 2) ? ($a += 10) : ($a += 2)
499
19799a22
GS
500That should probably be written more simply as:
501
502 $a += ($a % 2) ? 10 : 2;
503
4633a7c4 504=head2 Assignment Operators
a0d0e21e
LW
505
506"=" is the ordinary assignment operator.
507
508Assignment operators work as in C. That is,
509
510 $a += 2;
511
512is equivalent to
513
514 $a = $a + 2;
515
516although without duplicating any side effects that dereferencing the lvalue
54310121
PP
517might trigger, such as from tie(). Other assignment operators work similarly.
518The following are recognized:
a0d0e21e
LW
519
520 **= += *= &= <<= &&=
521 -= /= |= >>= ||=
522 .= %= ^=
523 x=
524
19799a22 525Although these are grouped by family, they all have the precedence
a0d0e21e
LW
526of assignment.
527
b350dd2f
GS
528Unlike in C, the scalar assignment operator produces a valid lvalue.
529Modifying an assignment is equivalent to doing the assignment and
530then modifying the variable that was assigned to. This is useful
531for modifying a copy of something, like this:
a0d0e21e
LW
532
533 ($tmp = $global) =~ tr [A-Z] [a-z];
534
535Likewise,
536
537 ($a += 2) *= 3;
538
539is equivalent to
540
541 $a += 2;
542 $a *= 3;
543
b350dd2f
GS
544Similarly, a list assignment in list context produces the list of
545lvalues assigned to, and a list assignment in scalar context returns
546the number of elements produced by the expression on the right hand
547side of the assignment.
548
748a9306 549=head2 Comma Operator
a0d0e21e 550
5a964f20 551Binary "," is the comma operator. In scalar context it evaluates
a0d0e21e
LW
552its left argument, throws that value away, then evaluates its right
553argument and returns that value. This is just like C's comma operator.
554
5a964f20 555In list context, it's just the list argument separator, and inserts
a0d0e21e
LW
556both its arguments into the list.
557
35f2feb0 558The => digraph is mostly just a synonym for the comma operator. It's useful for
cb1a09d0 559documenting arguments that come in pairs. As of release 5.001, it also forces
4633a7c4 560any word to the left of it to be interpreted as a string.
748a9306 561
a0d0e21e
LW
562=head2 List Operators (Rightward)
563
564On the right side of a list operator, it has very low precedence,
565such that it controls all comma-separated expressions found there.
566The only operators with lower precedence are the logical operators
567"and", "or", and "not", which may be used to evaluate calls to list
568operators without the need for extra parentheses:
569
570 open HANDLE, "filename"
571 or die "Can't open: $!\n";
572
5ba421f6 573See also discussion of list operators in L<Terms and List Operators (Leftward)>.
a0d0e21e
LW
574
575=head2 Logical Not
576
577Unary "not" returns the logical negation of the expression to its right.
578It's the equivalent of "!" except for the very low precedence.
579
580=head2 Logical And
581
582Binary "and" returns the logical conjunction of the two surrounding
583expressions. It's equivalent to && except for the very low
5f05dabc 584precedence. This means that it short-circuits: i.e., the right
a0d0e21e
LW
585expression is evaluated only if the left expression is true.
586
587=head2 Logical or and Exclusive Or
588
589Binary "or" returns the logical disjunction of the two surrounding
5a964f20
TC
590expressions. It's equivalent to || except for the very low precedence.
591This makes it useful for control flow
592
593 print FH $data or die "Can't write to FH: $!";
594
595This means that it short-circuits: i.e., the right expression is evaluated
596only if the left expression is false. Due to its precedence, you should
597probably avoid using this for assignment, only for control flow.
598
599 $a = $b or $c; # bug: this is wrong
600 ($a = $b) or $c; # really means this
601 $a = $b || $c; # better written this way
602
19799a22 603However, when it's a list-context assignment and you're trying to use
5a964f20
TC
604"||" for control flow, you probably need "or" so that the assignment
605takes higher precedence.
606
607 @info = stat($file) || die; # oops, scalar sense of stat!
608 @info = stat($file) or die; # better, now @info gets its due
609
19799a22 610Then again, you could always use parentheses.
a0d0e21e
LW
611
612Binary "xor" returns the exclusive-OR of the two surrounding expressions.
613It cannot short circuit, of course.
614
615=head2 C Operators Missing From Perl
616
617Here is what C has that Perl doesn't:
618
619=over 8
620
621=item unary &
622
623Address-of operator. (But see the "\" operator for taking a reference.)
624
625=item unary *
626
54310121 627Dereference-address operator. (Perl's prefix dereferencing
a0d0e21e
LW
628operators are typed: $, @, %, and &.)
629
630=item (TYPE)
631
19799a22 632Type-casting operator.
a0d0e21e
LW
633
634=back
635
5f05dabc 636=head2 Quote and Quote-like Operators
a0d0e21e
LW
637
638While we usually think of quotes as literal values, in Perl they
639function as operators, providing various kinds of interpolating and
640pattern matching capabilities. Perl provides customary quote characters
641for these behaviors, but also provides a way for you to choose your
642quote character for any of them. In the following table, a C<{}> represents
87275199 643any pair of delimiters you choose.
a0d0e21e 644
2c268ad5
TP
645 Customary Generic Meaning Interpolates
646 '' q{} Literal no
647 "" qq{} Literal yes
675d96b5 648 qu{} Literal yes (UTF-8, see below)
01ae956f 649 `` qx{} Command yes (unless '' is delimiter)
2c268ad5 650 qw{} Word list no
f70b4f9c
AB
651 // m{} Pattern match yes (unless '' is delimiter)
652 qr{} Pattern yes (unless '' is delimiter)
653 s{}{} Substitution yes (unless '' is delimiter)
2c268ad5 654 tr{}{} Transliteration no (but see below)
a0d0e21e 655
87275199
GS
656Non-bracketing delimiters use the same character fore and aft, but the four
657sorts of brackets (round, angle, square, curly) will all nest, which means
658that
659
660 q{foo{bar}baz}
35f2feb0 661
87275199
GS
662is the same as
663
664 'foo{bar}baz'
665
666Note, however, that this does not always work for quoting Perl code:
667
668 $s = q{ if($a eq "}") ... }; # WRONG
669
670is a syntax error. The C<Text::Balanced> module on CPAN is able to do this
671properly.
672
19799a22 673There can be whitespace between the operator and the quoting
fb73857a 674characters, except when C<#> is being used as the quoting character.
19799a22
GS
675C<q#foo#> is parsed as the string C<foo>, while C<q #foo#> is the
676operator C<q> followed by a comment. Its argument will be taken
677from the next line. This allows you to write:
fb73857a
PP
678
679 s {foo} # Replace foo
680 {bar} # with bar.
681
19799a22
GS
682For constructs that do interpolate, variables beginning with "C<$>"
683or "C<@>" are interpolated, as are the following escape sequences. Within
a0ed51b3 684a transliteration, the first eleven of these sequences may be used.
a0d0e21e 685
6ee5d4e7 686 \t tab (HT, TAB)
5a964f20 687 \n newline (NL)
6ee5d4e7
PP
688 \r return (CR)
689 \f form feed (FF)
690 \b backspace (BS)
691 \a alarm (bell) (BEL)
692 \e escape (ESC)
a0ed51b3
LW
693 \033 octal char (ESC)
694 \x1b hex char (ESC)
695 \x{263a} wide hex char (SMILEY)
19799a22 696 \c[ control char (ESC)
4a2d328f 697 \N{name} named char
2c268ad5 698
a0d0e21e
LW
699 \l lowercase next char
700 \u uppercase next char
701 \L lowercase till \E
702 \U uppercase till \E
703 \E end case modification
1d2dff63 704 \Q quote non-word characters till \E
a0d0e21e 705
a034a98d 706If C<use locale> is in effect, the case map used by C<\l>, C<\L>, C<\u>
423cee85 707and C<\U> is taken from the current locale. See L<perllocale>. For
4a2d328f 708documentation of C<\N{name}>, see L<charnames>.
a034a98d 709
5a964f20
TC
710All systems use the virtual C<"\n"> to represent a line terminator,
711called a "newline". There is no such thing as an unvarying, physical
19799a22 712newline character. It is only an illusion that the operating system,
5a964f20
TC
713device drivers, C libraries, and Perl all conspire to preserve. Not all
714systems read C<"\r"> as ASCII CR and C<"\n"> as ASCII LF. For example,
715on a Mac, these are reversed, and on systems without line terminator,
716printing C<"\n"> may emit no actual data. In general, use C<"\n"> when
717you mean a "newline" for your system, but use the literal ASCII when you
718need an exact character. For example, most networking protocols expect
2a380090 719and prefer a CR+LF (C<"\015\012"> or C<"\cM\cJ">) for line terminators,
5a964f20
TC
720and although they often accept just C<"\012">, they seldom tolerate just
721C<"\015">. If you get in the habit of using C<"\n"> for networking,
722you may be burned some day.
723
1d2dff63
GS
724You cannot include a literal C<$> or C<@> within a C<\Q> sequence.
725An unescaped C<$> or C<@> interpolates the corresponding variable,
726while escaping will cause the literal string C<\$> to be inserted.
727You'll need to write something like C<m/\Quser\E\@\Qhost/>.
728
a0d0e21e
LW
729Patterns are subject to an additional level of interpretation as a
730regular expression. This is done as a second pass, after variables are
731interpolated, so that regular expressions may be incorporated into the
732pattern from the variables. If this is not what you want, use C<\Q> to
733interpolate a variable literally.
734
19799a22
GS
735Apart from the behavior described above, Perl does not expand
736multiple levels of interpolation. In particular, contrary to the
737expectations of shell programmers, back-quotes do I<NOT> interpolate
738within double quotes, nor do single quotes impede evaluation of
739variables when used within double quotes.
a0d0e21e 740
5f05dabc 741=head2 Regexp Quote-Like Operators
cb1a09d0 742
5f05dabc 743Here are the quote-like operators that apply to pattern
cb1a09d0
AD
744matching and related activities.
745
a0d0e21e
LW
746=over 8
747
748=item ?PATTERN?
749
750This is just like the C</pattern/> search, except that it matches only
751once between calls to the reset() operator. This is a useful
5f05dabc 752optimization when you want to see only the first occurrence of
a0d0e21e
LW
753something in each file of a set of files, for instance. Only C<??>
754patterns local to the current package are reset.
755
5a964f20
TC
756 while (<>) {
757 if (?^$?) {
758 # blank line between header and body
759 }
760 } continue {
761 reset if eof; # clear ?? status for next file
762 }
763
19799a22
GS
764This usage is vaguely depreciated, which means it just might possibly
765be removed in some distant future version of Perl, perhaps somewhere
766around the year 2168.
a0d0e21e 767
fb73857a 768=item m/PATTERN/cgimosx
a0d0e21e 769
fb73857a 770=item /PATTERN/cgimosx
a0d0e21e 771
5a964f20 772Searches a string for a pattern match, and in scalar context returns
19799a22
GS
773true if it succeeds, false if it fails. If no string is specified
774via the C<=~> or C<!~> operator, the $_ string is searched. (The
775string specified with C<=~> need not be an lvalue--it may be the
776result of an expression evaluation, but remember the C<=~> binds
777rather tightly.) See also L<perlre>. See L<perllocale> for
778discussion of additional considerations that apply when C<use locale>
779is in effect.
a0d0e21e
LW
780
781Options are:
782
fb73857a 783 c Do not reset search position on a failed match when /g is in effect.
5f05dabc 784 g Match globally, i.e., find all occurrences.
a0d0e21e
LW
785 i Do case-insensitive pattern matching.
786 m Treat string as multiple lines.
5f05dabc 787 o Compile pattern only once.
a0d0e21e
LW
788 s Treat string as single line.
789 x Use extended regular expressions.
790
791If "/" is the delimiter then the initial C<m> is optional. With the C<m>
01ae956f 792you can use any pair of non-alphanumeric, non-whitespace characters
19799a22
GS
793as delimiters. This is particularly useful for matching path names
794that contain "/", to avoid LTS (leaning toothpick syndrome). If "?" is
7bac28a0 795the delimiter, then the match-only-once rule of C<?PATTERN?> applies.
19799a22 796If "'" is the delimiter, no interpolation is performed on the PATTERN.
a0d0e21e
LW
797
798PATTERN may contain variables, which will be interpolated (and the
f70b4f9c 799pattern recompiled) every time the pattern search is evaluated, except
1f247705
GS
800for when the delimiter is a single quote. (Note that C<$(>, C<$)>, and
801C<$|> are not interpolated because they look like end-of-string tests.)
f70b4f9c
AB
802If you want such a pattern to be compiled only once, add a C</o> after
803the trailing delimiter. This avoids expensive run-time recompilations,
804and is useful when the value you are interpolating won't change over
805the life of the script. However, mentioning C</o> constitutes a promise
806that you won't change the variables in the pattern. If you change them,
13a2d996 807Perl won't even notice. See also L<"qr/STRING/imosx">.
a0d0e21e 808
5a964f20
TC
809If the PATTERN evaluates to the empty string, the last
810I<successfully> matched regular expression is used instead.
a0d0e21e 811
19799a22 812If the C</g> option is not used, C<m//> in list context returns a
a0d0e21e 813list consisting of the subexpressions matched by the parentheses in the
f7e33566
GS
814pattern, i.e., (C<$1>, C<$2>, C<$3>...). (Note that here C<$1> etc. are
815also set, and that this differs from Perl 4's behavior.) When there are
816no parentheses in the pattern, the return value is the list C<(1)> for
817success. With or without parentheses, an empty list is returned upon
818failure.
a0d0e21e
LW
819
820Examples:
821
822 open(TTY, '/dev/tty');
823 <TTY> =~ /^y/i && foo(); # do foo if desired
824
825 if (/Version: *([0-9.]*)/) { $version = $1; }
826
827 next if m#^/usr/spool/uucp#;
828
829 # poor man's grep
830 $arg = shift;
831 while (<>) {
832 print if /$arg/o; # compile only once
833 }
834
835 if (($F1, $F2, $Etc) = ($foo =~ /^(\S+)\s+(\S+)\s*(.*)/))
836
837This last example splits $foo into the first two words and the
5f05dabc
PP
838remainder of the line, and assigns those three fields to $F1, $F2, and
839$Etc. The conditional is true if any variables were assigned, i.e., if
a0d0e21e
LW
840the pattern matched.
841
19799a22
GS
842The C</g> modifier specifies global pattern matching--that is,
843matching as many times as possible within the string. How it behaves
844depends on the context. In list context, it returns a list of the
845substrings matched by any capturing parentheses in the regular
846expression. If there are no parentheses, it returns a list of all
847the matched strings, as if there were parentheses around the whole
848pattern.
a0d0e21e 849
7e86de3e 850In scalar context, each execution of C<m//g> finds the next match,
19799a22 851returning true if it matches, and false if there is no further match.
7e86de3e
G
852The position after the last match can be read or set using the pos()
853function; see L<perlfunc/pos>. A failed match normally resets the
854search position to the beginning of the string, but you can avoid that
855by adding the C</c> modifier (e.g. C<m//gc>). Modifying the target
856string also resets the search position.
c90c0ff4
PP
857
858You can intermix C<m//g> matches with C<m/\G.../g>, where C<\G> is a
859zero-width assertion that matches the exact position where the previous
5d43e42d
DC
860C<m//g>, if any, left off. Without the C</g> modifier, the C<\G> assertion
861still anchors at pos(), but the match is of course only attempted once.
862Using C<\G> without C</g> on a target string that has not previously had a
863C</g> match applied to it is the same as using the C<\A> assertion to match
864the beginning of the string.
c90c0ff4
PP
865
866Examples:
a0d0e21e
LW
867
868 # list context
869 ($one,$five,$fifteen) = (`uptime` =~ /(\d+\.\d+)/g);
870
871 # scalar context
5d43e42d 872 $/ = "";
19799a22
GS
873 while (defined($paragraph = <>)) {
874 while ($paragraph =~ /[a-z]['")]*[.!?]+['")]*\s/g) {
875 $sentences++;
a0d0e21e
LW
876 }
877 }
878 print "$sentences\n";
879
c90c0ff4 880 # using m//gc with \G
137443ea 881 $_ = "ppooqppqq";
44a8e56a
PP
882 while ($i++ < 2) {
883 print "1: '";
c90c0ff4 884 print $1 while /(o)/gc; print "', pos=", pos, "\n";
44a8e56a 885 print "2: '";
c90c0ff4 886 print $1 if /\G(q)/gc; print "', pos=", pos, "\n";
44a8e56a 887 print "3: '";
c90c0ff4 888 print $1 while /(p)/gc; print "', pos=", pos, "\n";
44a8e56a 889 }
5d43e42d 890 print "Final: '$1', pos=",pos,"\n" if /\G(.)/;
44a8e56a
PP
891
892The last example should print:
893
894 1: 'oo', pos=4
137443ea 895 2: 'q', pos=5
44a8e56a
PP
896 3: 'pp', pos=7
897 1: '', pos=7
137443ea
PP
898 2: 'q', pos=8
899 3: '', pos=8
5d43e42d
DC
900 Final: 'q', pos=8
901
902Notice that the final match matched C<q> instead of C<p>, which a match
903without the C<\G> anchor would have done. Also note that the final match
904did not update C<pos> -- C<pos> is only updated on a C</g> match. If the
905final match did indeed match C<p>, it's a good bet that you're running an
906older (pre-5.6.0) Perl.
44a8e56a 907
c90c0ff4 908A useful idiom for C<lex>-like scanners is C</\G.../gc>. You can
e7ea3e70 909combine several regexps like this to process a string part-by-part,
c90c0ff4
PP
910doing different actions depending on which regexp matched. Each
911regexp tries to match where the previous one leaves off.
e7ea3e70 912
3fe9a6f1 913 $_ = <<'EOL';
e7ea3e70 914 $url = new URI::URL "http://www/"; die if $url eq "xXx";
3fe9a6f1
PP
915 EOL
916 LOOP:
e7ea3e70 917 {
c90c0ff4
PP
918 print(" digits"), redo LOOP if /\G\d+\b[,.;]?\s*/gc;
919 print(" lowercase"), redo LOOP if /\G[a-z]+\b[,.;]?\s*/gc;
920 print(" UPPERCASE"), redo LOOP if /\G[A-Z]+\b[,.;]?\s*/gc;
921 print(" Capitalized"), redo LOOP if /\G[A-Z][a-z]+\b[,.;]?\s*/gc;
922 print(" MiXeD"), redo LOOP if /\G[A-Za-z]+\b[,.;]?\s*/gc;
923 print(" alphanumeric"), redo LOOP if /\G[A-Za-z0-9]+\b[,.;]?\s*/gc;
924 print(" line-noise"), redo LOOP if /\G[^A-Za-z0-9]+/gc;
e7ea3e70
IZ
925 print ". That's all!\n";
926 }
927
928Here is the output (split into several lines):
929
930 line-noise lowercase line-noise lowercase UPPERCASE line-noise
931 UPPERCASE line-noise lowercase line-noise lowercase line-noise
932 lowercase lowercase line-noise lowercase lowercase line-noise
933 MiXeD line-noise. That's all!
44a8e56a 934
a0d0e21e
LW
935=item q/STRING/
936
937=item C<'STRING'>
938
19799a22 939A single-quoted, literal string. A backslash represents a backslash
68dc0745
PP
940unless followed by the delimiter or another backslash, in which case
941the delimiter or backslash is interpolated.
a0d0e21e
LW
942
943 $foo = q!I said, "You said, 'She said it.'"!;
944 $bar = q('This is it.');
68dc0745 945 $baz = '\n'; # a two-character string
a0d0e21e
LW
946
947=item qq/STRING/
948
949=item "STRING"
950
951A double-quoted, interpolated string.
952
953 $_ .= qq
954 (*** The previous line contains the naughty word "$1".\n)
19799a22 955 if /\b(tcl|java|python)\b/i; # :-)
68dc0745 956 $baz = "\n"; # a one-character string
a0d0e21e 957
eec2d3df
GS
958=item qr/STRING/imosx
959
322edccd 960This operator quotes (and possibly compiles) its I<STRING> as a regular
19799a22
GS
961expression. I<STRING> is interpolated the same way as I<PATTERN>
962in C<m/PATTERN/>. If "'" is used as the delimiter, no interpolation
963is done. Returns a Perl value which may be used instead of the
964corresponding C</STRING/imosx> expression.
4b6a7270
IZ
965
966For example,
967
968 $rex = qr/my.STRING/is;
969 s/$rex/foo/;
970
971is equivalent to
972
973 s/my.STRING/foo/is;
974
975The result may be used as a subpattern in a match:
eec2d3df
GS
976
977 $re = qr/$pattern/;
0a92e3a8
GS
978 $string =~ /foo${re}bar/; # can be interpolated in other patterns
979 $string =~ $re; # or used standalone
4b6a7270
IZ
980 $string =~ /$re/; # or this way
981
982Since Perl may compile the pattern at the moment of execution of qr()
19799a22 983operator, using qr() may have speed advantages in some situations,
4b6a7270
IZ
984notably if the result of qr() is used standalone:
985
986 sub match {
987 my $patterns = shift;
988 my @compiled = map qr/$_/i, @$patterns;
989 grep {
990 my $success = 0;
a7665c5e 991 foreach my $pat (@compiled) {
4b6a7270
IZ
992 $success = 1, last if /$pat/;
993 }
994 $success;
995 } @_;
996 }
997
19799a22
GS
998Precompilation of the pattern into an internal representation at
999the moment of qr() avoids a need to recompile the pattern every
1000time a match C</$pat/> is attempted. (Perl has many other internal
1001optimizations, but none would be triggered in the above example if
1002we did not use qr() operator.)
eec2d3df
GS
1003
1004Options are:
1005
1006 i Do case-insensitive pattern matching.
1007 m Treat string as multiple lines.
1008 o Compile pattern only once.
1009 s Treat string as single line.
1010 x Use extended regular expressions.
1011
0a92e3a8
GS
1012See L<perlre> for additional information on valid syntax for STRING, and
1013for a detailed look at the semantics of regular expressions.
1014
49cb94c6
JH
1015=item qw/STRING/
1016
1017Evaluates to a list of the words extracted out of STRING, using embedded
1018whitespace as the word delimiters. It can be understood as being roughly
1019equivalent to:
1020
1021 split(' ', q/STRING/);
1022
1023the difference being that it generates a real list at compile time. So
1024this expression:
1025
1026 qw(foo bar baz)
1027
1028is semantically equivalent to the list:
1029
1030 'foo', 'bar', 'baz'
1031
1032Some frequently seen examples:
1033
1034 use POSIX qw( setlocale localeconv )
1035 @EXPORT = qw( foo bar baz );
1036
1037A common mistake is to try to separate the words with comma or to
1038put comments into a multi-line C<qw>-string. For this reason, the
1039C<use warnings> pragma and the B<-w> switch (that is, the C<$^W> variable)
1040produces warnings if the STRING contains the "," or the "#" character.
1041
1042=item qu/STRING/
1043
675d96b5
JH
1044Like L<qq> but explicitly generates UTF-8 from the \0ooo, \xHH, and
1045\x{HH} constructs if the code point is in the 0x80..0xff range (and
1046of course for the 0x100.. range).
49cb94c6 1047
675d96b5
JH
1048Normally you do not need to use this because whether characters are
1049internally encoded in UTF-8 should be transparent, and you can just
1050just use qq, also known as "".
1051
1052(In qq/STRING/ the \0ooo, \xHH, and the \x{HHH...} constructs
1053generate bytes for the 0x80..0xff range. For the whole 0x00..0xff
1054range the generated bytes are host-dependent: in ISO 8859-1 they will
1055be ISO 8859-1, in EBCDIC they will EBCDIC, and so on.)
49cb94c6 1056
a0d0e21e
LW
1057=item qx/STRING/
1058
1059=item `STRING`
1060
43dd4d21
JH
1061A string which is (possibly) interpolated and then executed as a
1062system command with C</bin/sh> or its equivalent. Shell wildcards,
1063pipes, and redirections will be honored. The collected standard
1064output of the command is returned; standard error is unaffected. In
1065scalar context, it comes back as a single (potentially multi-line)
1066string, or undef if the command failed. In list context, returns a
1067list of lines (however you've defined lines with $/ or
1068$INPUT_RECORD_SEPARATOR), or an empty list if the command failed.
5a964f20
TC
1069
1070Because backticks do not affect standard error, use shell file descriptor
1071syntax (assuming the shell supports this) if you care to address this.
1072To capture a command's STDERR and STDOUT together:
a0d0e21e 1073
5a964f20
TC
1074 $output = `cmd 2>&1`;
1075
1076To capture a command's STDOUT but discard its STDERR:
1077
1078 $output = `cmd 2>/dev/null`;
1079
1080To capture a command's STDERR but discard its STDOUT (ordering is
1081important here):
1082
1083 $output = `cmd 2>&1 1>/dev/null`;
1084
1085To exchange a command's STDOUT and STDERR in order to capture the STDERR
1086but leave its STDOUT to come out the old STDERR:
1087
1088 $output = `cmd 3>&1 1>&2 2>&3 3>&-`;
1089
1090To read both a command's STDOUT and its STDERR separately, it's easiest
1091and safest to redirect them separately to files, and then read from those
1092files when the program is done:
1093
1094 system("program args 1>/tmp/program.stdout 2>/tmp/program.stderr");
1095
1096Using single-quote as a delimiter protects the command from Perl's
1097double-quote interpolation, passing it on to the shell instead:
1098
1099 $perl_info = qx(ps $$); # that's Perl's $$
1100 $shell_info = qx'ps $$'; # that's the new shell's $$
1101
19799a22 1102How that string gets evaluated is entirely subject to the command
5a964f20
TC
1103interpreter on your system. On most platforms, you will have to protect
1104shell metacharacters if you want them treated literally. This is in
1105practice difficult to do, as it's unclear how to escape which characters.
1106See L<perlsec> for a clean and safe example of a manual fork() and exec()
1107to emulate backticks safely.
a0d0e21e 1108
bb32b41a
GS
1109On some platforms (notably DOS-like ones), the shell may not be
1110capable of dealing with multiline commands, so putting newlines in
1111the string may not get you what you want. You may be able to evaluate
1112multiple commands in a single line by separating them with the command
1113separator character, if your shell supports that (e.g. C<;> on many Unix
1114shells; C<&> on the Windows NT C<cmd> shell).
1115
0f897271
GS
1116Beginning with v5.6.0, Perl will attempt to flush all files opened for
1117output before starting the child process, but this may not be supported
1118on some platforms (see L<perlport>). To be safe, you may need to set
1119C<$|> ($AUTOFLUSH in English) or call the C<autoflush()> method of
1120C<IO::Handle> on any open handles.
1121
bb32b41a
GS
1122Beware that some command shells may place restrictions on the length
1123of the command line. You must ensure your strings don't exceed this
1124limit after any necessary interpolations. See the platform-specific
1125release notes for more details about your particular environment.
1126
5a964f20
TC
1127Using this operator can lead to programs that are difficult to port,
1128because the shell commands called vary between systems, and may in
1129fact not be present at all. As one example, the C<type> command under
1130the POSIX shell is very different from the C<type> command under DOS.
1131That doesn't mean you should go out of your way to avoid backticks
1132when they're the right way to get something done. Perl was made to be
1133a glue language, and one of the things it glues together is commands.
1134Just understand what you're getting yourself into.
bb32b41a 1135
dc848c6f 1136See L<"I/O Operators"> for more discussion.
a0d0e21e 1137
a0d0e21e
LW
1138=item s/PATTERN/REPLACEMENT/egimosx
1139
1140Searches a string for a pattern, and if found, replaces that pattern
1141with the replacement text and returns the number of substitutions
e37d713d 1142made. Otherwise it returns false (specifically, the empty string).
a0d0e21e
LW
1143
1144If no string is specified via the C<=~> or C<!~> operator, the C<$_>
1145variable is searched and modified. (The string specified with C<=~> must
5a964f20 1146be scalar variable, an array element, a hash element, or an assignment
5f05dabc 1147to one of those, i.e., an lvalue.)
a0d0e21e 1148
19799a22 1149If the delimiter chosen is a single quote, no interpolation is
a0d0e21e
LW
1150done on either the PATTERN or the REPLACEMENT. Otherwise, if the
1151PATTERN contains a $ that looks like a variable rather than an
1152end-of-string test, the variable will be interpolated into the pattern
5f05dabc 1153at run-time. If you want the pattern compiled only once the first time
a0d0e21e 1154the variable is interpolated, use the C</o> option. If the pattern
5a964f20 1155evaluates to the empty string, the last successfully executed regular
a0d0e21e 1156expression is used instead. See L<perlre> for further explanation on these.
5a964f20 1157See L<perllocale> for discussion of additional considerations that apply
a034a98d 1158when C<use locale> is in effect.
a0d0e21e
LW
1159
1160Options are:
1161
1162 e Evaluate the right side as an expression.
5f05dabc 1163 g Replace globally, i.e., all occurrences.
a0d0e21e
LW
1164 i Do case-insensitive pattern matching.
1165 m Treat string as multiple lines.
5f05dabc 1166 o Compile pattern only once.
a0d0e21e
LW
1167 s Treat string as single line.
1168 x Use extended regular expressions.
1169
1170Any non-alphanumeric, non-whitespace delimiter may replace the
1171slashes. If single quotes are used, no interpretation is done on the
e37d713d 1172replacement string (the C</e> modifier overrides this, however). Unlike
54310121 1173Perl 4, Perl 5 treats backticks as normal delimiters; the replacement
e37d713d 1174text is not evaluated as a command. If the
a0d0e21e 1175PATTERN is delimited by bracketing quotes, the REPLACEMENT has its own
5f05dabc 1176pair of quotes, which may or may not be bracketing quotes, e.g.,
35f2feb0 1177C<s(foo)(bar)> or C<< s<foo>/bar/ >>. A C</e> will cause the
cec88af6
GS
1178replacement portion to be treated as a full-fledged Perl expression
1179and evaluated right then and there. It is, however, syntax checked at
1180compile-time. A second C<e> modifier will cause the replacement portion
1181to be C<eval>ed before being run as a Perl expression.
a0d0e21e
LW
1182
1183Examples:
1184
1185 s/\bgreen\b/mauve/g; # don't change wintergreen
1186
1187 $path =~ s|/usr/bin|/usr/local/bin|;
1188
1189 s/Login: $foo/Login: $bar/; # run-time pattern
1190
5a964f20 1191 ($foo = $bar) =~ s/this/that/; # copy first, then change
a0d0e21e 1192
5a964f20 1193 $count = ($paragraph =~ s/Mister\b/Mr./g); # get change-count
a0d0e21e
LW
1194
1195 $_ = 'abc123xyz';
1196 s/\d+/$&*2/e; # yields 'abc246xyz'
1197 s/\d+/sprintf("%5d",$&)/e; # yields 'abc 246xyz'
1198 s/\w/$& x 2/eg; # yields 'aabbcc 224466xxyyzz'
1199
1200 s/%(.)/$percent{$1}/g; # change percent escapes; no /e
1201 s/%(.)/$percent{$1} || $&/ge; # expr now, so /e
1202 s/^=(\w+)/&pod($1)/ge; # use function call
1203
5a964f20
TC
1204 # expand variables in $_, but dynamics only, using
1205 # symbolic dereferencing
1206 s/\$(\w+)/${$1}/g;
1207
cec88af6
GS
1208 # Add one to the value of any numbers in the string
1209 s/(\d+)/1 + $1/eg;
1210
1211 # This will expand any embedded scalar variable
1212 # (including lexicals) in $_ : First $1 is interpolated
1213 # to the variable name, and then evaluated
a0d0e21e
LW
1214 s/(\$\w+)/$1/eeg;
1215
5a964f20 1216 # Delete (most) C comments.
a0d0e21e 1217 $program =~ s {
4633a7c4
LW
1218 /\* # Match the opening delimiter.
1219 .*? # Match a minimal number of characters.
1220 \*/ # Match the closing delimiter.
a0d0e21e
LW
1221 } []gsx;
1222
5a964f20
TC
1223 s/^\s*(.*?)\s*$/$1/; # trim white space in $_, expensively
1224
1225 for ($variable) { # trim white space in $variable, cheap
1226 s/^\s+//;
1227 s/\s+$//;
1228 }
a0d0e21e
LW
1229
1230 s/([^ ]*) *([^ ]*)/$2 $1/; # reverse 1st two fields
1231
54310121 1232Note the use of $ instead of \ in the last example. Unlike
35f2feb0
GS
1233B<sed>, we use the \<I<digit>> form in only the left hand side.
1234Anywhere else it's $<I<digit>>.
a0d0e21e 1235
5f05dabc 1236Occasionally, you can't use just a C</g> to get all the changes
19799a22 1237to occur that you might want. Here are two common cases:
a0d0e21e
LW
1238
1239 # put commas in the right places in an integer
19799a22 1240 1 while s/(\d)(\d\d\d)(?!\d)/$1,$2/g;
a0d0e21e
LW
1241
1242 # expand tabs to 8-column spacing
1243 1 while s/\t+/' ' x (length($&)*8 - length($`)%8)/e;
1244
6940069f 1245=item tr/SEARCHLIST/REPLACEMENTLIST/cds
a0d0e21e 1246
6940069f 1247=item y/SEARCHLIST/REPLACEMENTLIST/cds
a0d0e21e 1248
2c268ad5 1249Transliterates all occurrences of the characters found in the search list
a0d0e21e
LW
1250with the corresponding character in the replacement list. It returns
1251the number of characters replaced or deleted. If no string is
2c268ad5 1252specified via the =~ or !~ operator, the $_ string is transliterated. (The
54310121
PP
1253string specified with =~ must be a scalar variable, an array element, a
1254hash element, or an assignment to one of those, i.e., an lvalue.)
8ada0baa 1255
2c268ad5
TP
1256A character range may be specified with a hyphen, so C<tr/A-J/0-9/>
1257does the same replacement as C<tr/ACEGIBDFHJ/0246813579/>.
54310121
PP
1258For B<sed> devotees, C<y> is provided as a synonym for C<tr>. If the
1259SEARCHLIST is delimited by bracketing quotes, the REPLACEMENTLIST has
1260its own pair of quotes, which may or may not be bracketing quotes,
2c268ad5 1261e.g., C<tr[A-Z][a-z]> or C<tr(+\-*/)/ABCD/>.
a0d0e21e 1262
cc255d5f
JH
1263Note that C<tr> does B<not> do regular expression character classes
1264such as C<\d> or C<[:lower:]>. The <tr> operator is not equivalent to
1265the tr(1) utility. If you want to map strings between lower/upper
1266cases, see L<perlfunc/lc> and L<perlfunc/uc>, and in general consider
1267using the C<s> operator if you need regular expressions.
1268
8ada0baa
JH
1269Note also that the whole range idea is rather unportable between
1270character sets--and even within character sets they may cause results
1271you probably didn't expect. A sound principle is to use only ranges
1272that begin from and end at either alphabets of equal case (a-e, A-E),
1273or digits (0-4). Anything else is unsafe. If in doubt, spell out the
1274character sets in full.
1275
a0d0e21e
LW
1276Options:
1277
1278 c Complement the SEARCHLIST.
1279 d Delete found but unreplaced characters.
1280 s Squash duplicate replaced characters.
1281
19799a22
GS
1282If the C</c> modifier is specified, the SEARCHLIST character set
1283is complemented. If the C</d> modifier is specified, any characters
1284specified by SEARCHLIST not found in REPLACEMENTLIST are deleted.
1285(Note that this is slightly more flexible than the behavior of some
1286B<tr> programs, which delete anything they find in the SEARCHLIST,
1287period.) If the C</s> modifier is specified, sequences of characters
1288that were transliterated to the same character are squashed down
1289to a single instance of the character.
a0d0e21e
LW
1290
1291If the C</d> modifier is used, the REPLACEMENTLIST is always interpreted
1292exactly as specified. Otherwise, if the REPLACEMENTLIST is shorter
1293than the SEARCHLIST, the final character is replicated till it is long
5a964f20 1294enough. If the REPLACEMENTLIST is empty, the SEARCHLIST is replicated.
a0d0e21e
LW
1295This latter is useful for counting characters in a class or for
1296squashing character sequences in a class.
1297
1298Examples:
1299
1300 $ARGV[1] =~ tr/A-Z/a-z/; # canonicalize to lower case
1301
1302 $cnt = tr/*/*/; # count the stars in $_
1303
1304 $cnt = $sky =~ tr/*/*/; # count the stars in $sky
1305
1306 $cnt = tr/0-9//; # count the digits in $_
1307
1308 tr/a-zA-Z//s; # bookkeeper -> bokeper
1309
1310 ($HOST = $host) =~ tr/a-z/A-Z/;
1311
1312 tr/a-zA-Z/ /cs; # change non-alphas to single space
1313
1314 tr [\200-\377]
1315 [\000-\177]; # delete 8th bit
1316
19799a22
GS
1317If multiple transliterations are given for a character, only the
1318first one is used:
748a9306
LW
1319
1320 tr/AAA/XYZ/
1321
2c268ad5 1322will transliterate any A to X.
748a9306 1323
19799a22 1324Because the transliteration table is built at compile time, neither
a0d0e21e 1325the SEARCHLIST nor the REPLACEMENTLIST are subjected to double quote
19799a22
GS
1326interpolation. That means that if you want to use variables, you
1327must use an eval():
a0d0e21e
LW
1328
1329 eval "tr/$oldlist/$newlist/";
1330 die $@ if $@;
1331
1332 eval "tr/$oldlist/$newlist/, 1" or die $@;
1333
1334=back
1335
75e14d17
IZ
1336=head2 Gory details of parsing quoted constructs
1337
19799a22
GS
1338When presented with something that might have several different
1339interpretations, Perl uses the B<DWIM> (that's "Do What I Mean")
1340principle to pick the most probable interpretation. This strategy
1341is so successful that Perl programmers often do not suspect the
1342ambivalence of what they write. But from time to time, Perl's
1343notions differ substantially from what the author honestly meant.
1344
1345This section hopes to clarify how Perl handles quoted constructs.
1346Although the most common reason to learn this is to unravel labyrinthine
1347regular expressions, because the initial steps of parsing are the
1348same for all quoting operators, they are all discussed together.
1349
1350The most important Perl parsing rule is the first one discussed
1351below: when processing a quoted construct, Perl first finds the end
1352of that construct, then interprets its contents. If you understand
1353this rule, you may skip the rest of this section on the first
1354reading. The other rules are likely to contradict the user's
1355expectations much less frequently than this first one.
1356
1357Some passes discussed below are performed concurrently, but because
1358their results are the same, we consider them individually. For different
1359quoting constructs, Perl performs different numbers of passes, from
1360one to five, but these passes are always performed in the same order.
75e14d17 1361
13a2d996 1362=over 4
75e14d17
IZ
1363
1364=item Finding the end
1365
19799a22
GS
1366The first pass is finding the end of the quoted construct, whether
1367it be a multicharacter delimiter C<"\nEOF\n"> in the C<<<EOF>
1368construct, a C</> that terminates a C<qq//> construct, a C<]> which
35f2feb0
GS
1369terminates C<qq[]> construct, or a C<< > >> which terminates a
1370fileglob started with C<< < >>.
75e14d17 1371
19799a22
GS
1372When searching for single-character non-pairing delimiters, such
1373as C</>, combinations of C<\\> and C<\/> are skipped. However,
1374when searching for single-character pairing delimiter like C<[>,
1375combinations of C<\\>, C<\]>, and C<\[> are all skipped, and nested
1376C<[>, C<]> are skipped as well. When searching for multicharacter
1377delimiters, nothing is skipped.
75e14d17 1378
19799a22
GS
1379For constructs with three-part delimiters (C<s///>, C<y///>, and
1380C<tr///>), the search is repeated once more.
75e14d17 1381
19799a22
GS
1382During this search no attention is paid to the semantics of the construct.
1383Thus:
75e14d17
IZ
1384
1385 "$hash{"$foo/$bar"}"
1386
2a94b7ce 1387or:
75e14d17
IZ
1388
1389 m/
2a94b7ce 1390 bar # NOT a comment, this slash / terminated m//!
75e14d17
IZ
1391 /x
1392
19799a22
GS
1393do not form legal quoted expressions. The quoted part ends on the
1394first C<"> and C</>, and the rest happens to be a syntax error.
1395Because the slash that terminated C<m//> was followed by a C<SPACE>,
1396the example above is not C<m//x>, but rather C<m//> with no C</x>
1397modifier. So the embedded C<#> is interpreted as a literal C<#>.
75e14d17
IZ
1398
1399=item Removal of backslashes before delimiters
1400
19799a22
GS
1401During the second pass, text between the starting and ending
1402delimiters is copied to a safe location, and the C<\> is removed
1403from combinations consisting of C<\> and delimiter--or delimiters,
1404meaning both starting and ending delimiters will should these differ.
1405This removal does not happen for multi-character delimiters.
1406Note that the combination C<\\> is left intact, just as it was.
75e14d17 1407
19799a22
GS
1408Starting from this step no information about the delimiters is
1409used in parsing.
75e14d17
IZ
1410
1411=item Interpolation
1412
19799a22
GS
1413The next step is interpolation in the text obtained, which is now
1414delimiter-independent. There are four different cases.
75e14d17 1415
13a2d996 1416=over 4
75e14d17
IZ
1417
1418=item C<<<'EOF'>, C<m''>, C<s'''>, C<tr///>, C<y///>
1419
1420No interpolation is performed.
1421
1422=item C<''>, C<q//>
1423
1424The only interpolation is removal of C<\> from pairs C<\\>.
1425
35f2feb0 1426=item C<"">, C<``>, C<qq//>, C<qx//>, C<< <file*glob> >>
75e14d17 1427
19799a22
GS
1428C<\Q>, C<\U>, C<\u>, C<\L>, C<\l> (possibly paired with C<\E>) are
1429converted to corresponding Perl constructs. Thus, C<"$foo\Qbaz$bar">
1430is converted to C<$foo . (quotemeta("baz" . $bar))> internally.
1431The other combinations are replaced with appropriate expansions.
2a94b7ce 1432
19799a22
GS
1433Let it be stressed that I<whatever falls between C<\Q> and C<\E>>
1434is interpolated in the usual way. Something like C<"\Q\\E"> has
1435no C<\E> inside. instead, it has C<\Q>, C<\\>, and C<E>, so the
1436result is the same as for C<"\\\\E">. As a general rule, backslashes
1437between C<\Q> and C<\E> may lead to counterintuitive results. So,
1438C<"\Q\t\E"> is converted to C<quotemeta("\t")>, which is the same
1439as C<"\\\t"> (since TAB is not alphanumeric). Note also that:
2a94b7ce
IZ
1440
1441 $str = '\t';
1442 return "\Q$str";
1443
1444may be closer to the conjectural I<intention> of the writer of C<"\Q\t\E">.
1445
19799a22 1446Interpolated scalars and arrays are converted internally to the C<join> and
92d29cee 1447C<.> catenation operations. Thus, C<"$foo XXX '@arr'"> becomes:
75e14d17 1448
19799a22 1449 $foo . " XXX '" . (join $", @arr) . "'";
75e14d17 1450
19799a22 1451All operations above are performed simultaneously, left to right.
75e14d17 1452
19799a22
GS
1453Because the result of C<"\Q STRING \E"> has all metacharacters
1454quoted, there is no way to insert a literal C<$> or C<@> inside a
1455C<\Q\E> pair. If protected by C<\>, C<$> will be quoted to became
1456C<"\\\$">; if not, it is interpreted as the start of an interpolated
1457scalar.
75e14d17 1458
19799a22
GS
1459Note also that the interpolation code needs to make a decision on
1460where the interpolated scalar ends. For instance, whether
35f2feb0 1461C<< "a $b -> {c}" >> really means:
75e14d17
IZ
1462
1463 "a " . $b . " -> {c}";
1464
2a94b7ce 1465or:
75e14d17
IZ
1466
1467 "a " . $b -> {c};
1468
19799a22
GS
1469Most of the time, the longest possible text that does not include
1470spaces between components and which contains matching braces or
1471brackets. because the outcome may be determined by voting based
1472on heuristic estimators, the result is not strictly predictable.
1473Fortunately, it's usually correct for ambiguous cases.
75e14d17
IZ
1474
1475=item C<?RE?>, C</RE/>, C<m/RE/>, C<s/RE/foo/>,
1476
19799a22
GS
1477Processing of C<\Q>, C<\U>, C<\u>, C<\L>, C<\l>, and interpolation
1478happens (almost) as with C<qq//> constructs, but the substitution
1479of C<\> followed by RE-special chars (including C<\>) is not
1480performed. Moreover, inside C<(?{BLOCK})>, C<(?# comment )>, and
1481a C<#>-comment in a C<//x>-regular expression, no processing is
1482performed whatsoever. This is the first step at which the presence
1483of the C<//x> modifier is relevant.
1484
1485Interpolation has several quirks: C<$|>, C<$(>, and C<$)> are not
1486interpolated, and constructs C<$var[SOMETHING]> are voted (by several
1487different estimators) to be either an array element or C<$var>
1488followed by an RE alternative. This is where the notation
1489C<${arr[$bar]}> comes handy: C</${arr[0-9]}/> is interpreted as
1490array element C<-9>, not as a regular expression from the variable
1491C<$arr> followed by a digit, which would be the interpretation of
1492C</$arr[0-9]/>. Since voting among different estimators may occur,
1493the result is not predictable.
1494
1495It is at this step that C<\1> is begrudgingly converted to C<$1> in
1496the replacement text of C<s///> to correct the incorrigible
1497I<sed> hackers who haven't picked up the saner idiom yet. A warning
9f1b1f2d
GS
1498is emitted if the C<use warnings> pragma or the B<-w> command-line flag
1499(that is, the C<$^W> variable) was set.
19799a22
GS
1500
1501The lack of processing of C<\\> creates specific restrictions on
1502the post-processed text. If the delimiter is C</>, one cannot get
1503the combination C<\/> into the result of this step. C</> will
1504finish the regular expression, C<\/> will be stripped to C</> on
1505the previous step, and C<\\/> will be left as is. Because C</> is
1506equivalent to C<\/> inside a regular expression, this does not
1507matter unless the delimiter happens to be character special to the
1508RE engine, such as in C<s*foo*bar*>, C<m[foo]>, or C<?foo?>; or an
1509alphanumeric char, as in:
2a94b7ce
IZ
1510
1511 m m ^ a \s* b mmx;
1512
19799a22 1513In the RE above, which is intentionally obfuscated for illustration, the
2a94b7ce 1514delimiter is C<m>, the modifier is C<mx>, and after backslash-removal the
19799a22
GS
1515RE is the same as for C<m/ ^ a s* b /mx>). There's more than one
1516reason you're encouraged to restrict your delimiters to non-alphanumeric,
1517non-whitespace choices.
75e14d17
IZ
1518
1519=back
1520
19799a22 1521This step is the last one for all constructs except regular expressions,
75e14d17
IZ
1522which are processed further.
1523
1524=item Interpolation of regular expressions
1525
19799a22
GS
1526Previous steps were performed during the compilation of Perl code,
1527but this one happens at run time--although it may be optimized to
1528be calculated at compile time if appropriate. After preprocessing
1529described above, and possibly after evaluation if catenation,
1530joining, casing translation, or metaquoting are involved, the
1531resulting I<string> is passed to the RE engine for compilation.
1532
1533Whatever happens in the RE engine might be better discussed in L<perlre>,
1534but for the sake of continuity, we shall do so here.
1535
1536This is another step where the presence of the C<//x> modifier is
1537relevant. The RE engine scans the string from left to right and
1538converts it to a finite automaton.
1539
1540Backslashed characters are either replaced with corresponding
1541literal strings (as with C<\{>), or else they generate special nodes
1542in the finite automaton (as with C<\b>). Characters special to the
1543RE engine (such as C<|>) generate corresponding nodes or groups of
1544nodes. C<(?#...)> comments are ignored. All the rest is either
1545converted to literal strings to match, or else is ignored (as is
1546whitespace and C<#>-style comments if C<//x> is present).
1547
1548Parsing of the bracketed character class construct, C<[...]>, is
1549rather different than the rule used for the rest of the pattern.
1550The terminator of this construct is found using the same rules as
1551for finding the terminator of a C<{}>-delimited construct, the only
1552exception being that C<]> immediately following C<[> is treated as
1553though preceded by a backslash. Similarly, the terminator of
1554C<(?{...})> is found using the same rules as for finding the
1555terminator of a C<{}>-delimited construct.
1556
1557It is possible to inspect both the string given to RE engine and the
1558resulting finite automaton. See the arguments C<debug>/C<debugcolor>
1559in the C<use L<re>> pragma, as well as Perl's B<-Dr> command-line
4a4eefd0 1560switch documented in L<perlrun/"Command Switches">.
75e14d17
IZ
1561
1562=item Optimization of regular expressions
1563
7522fed5 1564This step is listed for completeness only. Since it does not change
75e14d17 1565semantics, details of this step are not documented and are subject
19799a22
GS
1566to change without notice. This step is performed over the finite
1567automaton that was generated during the previous pass.
2a94b7ce 1568
19799a22
GS
1569It is at this stage that C<split()> silently optimizes C</^/> to
1570mean C</^/m>.
75e14d17
IZ
1571
1572=back
1573
a0d0e21e
LW
1574=head2 I/O Operators
1575
54310121 1576There are several I/O operators you should know about.
fbad3eb5 1577
7b8d334a 1578A string enclosed by backticks (grave accents) first undergoes
19799a22
GS
1579double-quote interpolation. It is then interpreted as an external
1580command, and the output of that command is the value of the
e9c56f9b
JH
1581backtick string, like in a shell. In scalar context, a single string
1582consisting of all output is returned. In list context, a list of
1583values is returned, one per line of output. (You can set C<$/> to use
1584a different line terminator.) The command is executed each time the
1585pseudo-literal is evaluated. The status value of the command is
1586returned in C<$?> (see L<perlvar> for the interpretation of C<$?>).
1587Unlike in B<csh>, no translation is done on the return data--newlines
1588remain newlines. Unlike in any of the shells, single quotes do not
1589hide variable names in the command from interpretation. To pass a
1590literal dollar-sign through to the shell you need to hide it with a
1591backslash. The generalized form of backticks is C<qx//>. (Because
1592backticks always undergo shell expansion as well, see L<perlsec> for
1593security concerns.)
19799a22
GS
1594
1595In scalar context, evaluating a filehandle in angle brackets yields
1596the next line from that file (the newline, if any, included), or
1597C<undef> at end-of-file or on error. When C<$/> is set to C<undef>
1598(sometimes known as file-slurp mode) and the file is empty, it
1599returns C<''> the first time, followed by C<undef> subsequently.
1600
1601Ordinarily you must assign the returned value to a variable, but
1602there is one situation where an automatic assignment happens. If
1603and only if the input symbol is the only thing inside the conditional
1604of a C<while> statement (even if disguised as a C<for(;;)> loop),
1605the value is automatically assigned to the global variable $_,
1606destroying whatever was there previously. (This may seem like an
1607odd thing to you, but you'll use the construct in almost every Perl
17b829fa 1608script you write.) The $_ variable is not implicitly localized.
19799a22
GS
1609You'll have to put a C<local $_;> before the loop if you want that
1610to happen.
1611
1612The following lines are equivalent:
a0d0e21e 1613
748a9306 1614 while (defined($_ = <STDIN>)) { print; }
7b8d334a 1615 while ($_ = <STDIN>) { print; }
a0d0e21e
LW
1616 while (<STDIN>) { print; }
1617 for (;<STDIN>;) { print; }
748a9306 1618 print while defined($_ = <STDIN>);
7b8d334a 1619 print while ($_ = <STDIN>);
a0d0e21e
LW
1620 print while <STDIN>;
1621
19799a22 1622This also behaves similarly, but avoids $_ :
7b8d334a
GS
1623
1624 while (my $line = <STDIN>) { print $line }
1625
19799a22
GS
1626In these loop constructs, the assigned value (whether assignment
1627is automatic or explicit) is then tested to see whether it is
1628defined. The defined test avoids problems where line has a string
1629value that would be treated as false by Perl, for example a "" or
1630a "0" with no trailing newline. If you really mean for such values
1631to terminate the loop, they should be tested for explicitly:
7b8d334a
GS
1632
1633 while (($_ = <STDIN>) ne '0') { ... }
1634 while (<STDIN>) { last unless $_; ... }
1635
35f2feb0 1636In other boolean contexts, C<< <I<filehandle>> >> without an
9f1b1f2d
GS
1637explicit C<defined> test or comparison elicit a warning if the
1638C<use warnings> pragma or the B<-w>
19799a22 1639command-line switch (the C<$^W> variable) is in effect.
7b8d334a 1640
5f05dabc 1641The filehandles STDIN, STDOUT, and STDERR are predefined. (The
19799a22
GS
1642filehandles C<stdin>, C<stdout>, and C<stderr> will also work except
1643in packages, where they would be interpreted as local identifiers
1644rather than global.) Additional filehandles may be created with
1645the open() function, amongst others. See L<perlopentut> and
1646L<perlfunc/open> for details on this.
a0d0e21e 1647
35f2feb0 1648If a <FILEHANDLE> is used in a context that is looking for
19799a22
GS
1649a list, a list comprising all input lines is returned, one line per
1650list element. It's easy to grow to a rather large data space this
1651way, so use with care.
a0d0e21e 1652
35f2feb0 1653<FILEHANDLE> may also be spelled C<readline(*FILEHANDLE)>.
19799a22 1654See L<perlfunc/readline>.
fbad3eb5 1655
35f2feb0
GS
1656The null filehandle <> is special: it can be used to emulate the
1657behavior of B<sed> and B<awk>. Input from <> comes either from
a0d0e21e 1658standard input, or from each file listed on the command line. Here's
35f2feb0 1659how it works: the first time <> is evaluated, the @ARGV array is
5a964f20 1660checked, and if it is empty, C<$ARGV[0]> is set to "-", which when opened
a0d0e21e
LW
1661gives you standard input. The @ARGV array is then processed as a list
1662of filenames. The loop
1663
1664 while (<>) {
1665 ... # code for each line
1666 }
1667
1668is equivalent to the following Perl-like pseudo code:
1669
3e3baf6d 1670 unshift(@ARGV, '-') unless @ARGV;
a0d0e21e
LW
1671 while ($ARGV = shift) {
1672 open(ARGV, $ARGV);
1673 while (<ARGV>) {
1674 ... # code for each line
1675 }
1676 }
1677
19799a22
GS
1678except that it isn't so cumbersome to say, and will actually work.
1679It really does shift the @ARGV array and put the current filename
1680into the $ARGV variable. It also uses filehandle I<ARGV>
35f2feb0 1681internally--<> is just a synonym for <ARGV>, which
19799a22 1682is magical. (The pseudo code above doesn't work because it treats
35f2feb0 1683<ARGV> as non-magical.)
a0d0e21e 1684
35f2feb0 1685You can modify @ARGV before the first <> as long as the array ends up
a0d0e21e 1686containing the list of filenames you really want. Line numbers (C<$.>)
19799a22
GS
1687continue as though the input were one big happy file. See the example
1688in L<perlfunc/eof> for how to reset line numbers on each file.
5a964f20
TC
1689
1690If you want to set @ARGV to your own list of files, go right ahead.
1691This sets @ARGV to all plain text files if no @ARGV was given:
1692
1693 @ARGV = grep { -f && -T } glob('*') unless @ARGV;
a0d0e21e 1694
5a964f20
TC
1695You can even set them to pipe commands. For example, this automatically
1696filters compressed arguments through B<gzip>:
1697
1698 @ARGV = map { /\.(gz|Z)$/ ? "gzip -dc < $_ |" : $_ } @ARGV;
1699
1700If you want to pass switches into your script, you can use one of the
a0d0e21e
LW
1701Getopts modules or put a loop on the front like this:
1702
1703 while ($_ = $ARGV[0], /^-/) {
1704 shift;
1705 last if /^--$/;
1706 if (/^-D(.*)/) { $debug = $1 }
1707 if (/^-v/) { $verbose++ }
5a964f20 1708 # ... # other switches
a0d0e21e 1709 }
5a964f20 1710
a0d0e21e 1711 while (<>) {
5a964f20 1712 # ... # code for each line
a0d0e21e
LW
1713 }
1714
35f2feb0 1715The <> symbol will return C<undef> for end-of-file only once.
19799a22
GS
1716If you call it again after this, it will assume you are processing another
1717@ARGV list, and if you haven't set @ARGV, will read input from STDIN.
a0d0e21e 1718
19799a22 1719If angle brackets contain is a simple scalar variable (e.g.,
35f2feb0 1720<$foo>), then that variable contains the name of the
19799a22
GS
1721filehandle to input from, or its typeglob, or a reference to the
1722same. For example:
cb1a09d0
AD
1723
1724 $fh = \*STDIN;
1725 $line = <$fh>;
a0d0e21e 1726
5a964f20
TC
1727If what's within the angle brackets is neither a filehandle nor a simple
1728scalar variable containing a filehandle name, typeglob, or typeglob
1729reference, it is interpreted as a filename pattern to be globbed, and
1730either a list of filenames or the next filename in the list is returned,
19799a22 1731depending on context. This distinction is determined on syntactic
35f2feb0
GS
1732grounds alone. That means C<< <$x> >> is always a readline() from
1733an indirect handle, but C<< <$hash{key}> >> is always a glob().
5a964f20
TC
1734That's because $x is a simple scalar variable, but C<$hash{key}> is
1735not--it's a hash element.
1736
1737One level of double-quote interpretation is done first, but you can't
35f2feb0 1738say C<< <$foo> >> because that's an indirect filehandle as explained
5a964f20
TC
1739in the previous paragraph. (In older versions of Perl, programmers
1740would insert curly brackets to force interpretation as a filename glob:
35f2feb0 1741C<< <${foo}> >>. These days, it's considered cleaner to call the
5a964f20 1742internal function directly as C<glob($foo)>, which is probably the right
19799a22 1743way to have done it in the first place.) For example:
a0d0e21e
LW
1744
1745 while (<*.c>) {
1746 chmod 0644, $_;
1747 }
1748
3a4b19e4 1749is roughly equivalent to:
a0d0e21e
LW
1750
1751 open(FOO, "echo *.c | tr -s ' \t\r\f' '\\012\\012\\012\\012'|");
1752 while (<FOO>) {
5b3eff12 1753 chomp;
a0d0e21e
LW
1754 chmod 0644, $_;
1755 }
1756
3a4b19e4
GS
1757except that the globbing is actually done internally using the standard
1758C<File::Glob> extension. Of course, the shortest way to do the above is:
a0d0e21e
LW
1759
1760 chmod 0644, <*.c>;
1761
19799a22
GS
1762A (file)glob evaluates its (embedded) argument only when it is
1763starting a new list. All values must be read before it will start
1764over. In list context, this isn't important because you automatically
1765get them all anyway. However, in scalar context the operator returns
069e01df 1766the next value each time it's called, or C<undef> when the list has
19799a22
GS
1767run out. As with filehandle reads, an automatic C<defined> is
1768generated when the glob occurs in the test part of a C<while>,
1769because legal glob returns (e.g. a file called F<0>) would otherwise
1770terminate the loop. Again, C<undef> is returned only once. So if
1771you're expecting a single value from a glob, it is much better to
1772say
4633a7c4
LW
1773
1774 ($file) = <blurch*>;
1775
1776than
1777
1778 $file = <blurch*>;
1779
1780because the latter will alternate between returning a filename and
19799a22 1781returning false.
4633a7c4
LW
1782
1783It you're trying to do variable interpolation, it's definitely better
1784to use the glob() function, because the older notation can cause people
e37d713d 1785to become confused with the indirect filehandle notation.
4633a7c4
LW
1786
1787 @files = glob("$dir/*.[ch]");
1788 @files = glob($files[$i]);
1789
a0d0e21e
LW
1790=head2 Constant Folding
1791
1792Like C, Perl does a certain amount of expression evaluation at
19799a22 1793compile time whenever it determines that all arguments to an
a0d0e21e
LW
1794operator are static and have no side effects. In particular, string
1795concatenation happens at compile time between literals that don't do
19799a22 1796variable substitution. Backslash interpolation also happens at
a0d0e21e
LW
1797compile time. You can say
1798
1799 'Now is the time for all' . "\n" .
1800 'good men to come to.'
1801
54310121 1802and this all reduces to one string internally. Likewise, if
a0d0e21e
LW
1803you say
1804
1805 foreach $file (@filenames) {
5a964f20 1806 if (-s $file > 5 + 100 * 2**16) { }
54310121 1807 }
a0d0e21e 1808
19799a22
GS
1809the compiler will precompute the number which that expression
1810represents so that the interpreter won't have to.
a0d0e21e 1811
2c268ad5
TP
1812=head2 Bitwise String Operators
1813
1814Bitstrings of any size may be manipulated by the bitwise operators
1815(C<~ | & ^>).
1816
19799a22
GS
1817If the operands to a binary bitwise op are strings of different
1818sizes, B<|> and B<^> ops act as though the shorter operand had
1819additional zero bits on the right, while the B<&> op acts as though
1820the longer operand were truncated to the length of the shorter.
1821The granularity for such extension or truncation is one or more
1822bytes.
2c268ad5
TP
1823
1824 # ASCII-based examples
1825 print "j p \n" ^ " a h"; # prints "JAPH\n"
1826 print "JA" | " ph\n"; # prints "japh\n"
1827 print "japh\nJunk" & '_____'; # prints "JAPH\n";
1828 print 'p N$' ^ " E<H\n"; # prints "Perl\n";
1829
19799a22 1830If you are intending to manipulate bitstrings, be certain that
2c268ad5 1831you're supplying bitstrings: If an operand is a number, that will imply
19799a22 1832a B<numeric> bitwise operation. You may explicitly show which type of
2c268ad5
TP
1833operation you intend by using C<""> or C<0+>, as in the examples below.
1834
1835 $foo = 150 | 105 ; # yields 255 (0x96 | 0x69 is 0xFF)
1836 $foo = '150' | 105 ; # yields 255
1837 $foo = 150 | '105'; # yields 255
1838 $foo = '150' | '105'; # yields string '155' (under ASCII)
1839
1840 $baz = 0+$foo & 0+$bar; # both ops explicitly numeric
1841 $biz = "$foo" ^ "$bar"; # both ops explicitly stringy
a0d0e21e 1842
1ae175c8
GS
1843See L<perlfunc/vec> for information on how to manipulate individual bits
1844in a bit vector.
1845
55497cff 1846=head2 Integer Arithmetic
a0d0e21e 1847
19799a22 1848By default, Perl assumes that it must do most of its arithmetic in
a0d0e21e
LW
1849floating point. But by saying
1850
1851 use integer;
1852
1853you may tell the compiler that it's okay to use integer operations
19799a22
GS
1854(if it feels like it) from here to the end of the enclosing BLOCK.
1855An inner BLOCK may countermand this by saying
a0d0e21e
LW
1856
1857 no integer;
1858
19799a22
GS
1859which lasts until the end of that BLOCK. Note that this doesn't
1860mean everything is only an integer, merely that Perl may use integer
1861operations if it is so inclined. For example, even under C<use
1862integer>, if you take the C<sqrt(2)>, you'll still get C<1.4142135623731>
1863or so.
1864
1865Used on numbers, the bitwise operators ("&", "|", "^", "~", "<<",
13a2d996
SP
1866and ">>") always produce integral results. (But see also
1867L<Bitwise String Operators>.) However, C<use integer> still has meaning for
19799a22
GS
1868them. By default, their results are interpreted as unsigned integers, but
1869if C<use integer> is in effect, their results are interpreted
1870as signed integers. For example, C<~0> usually evaluates to a large
1871integral value. However, C<use integer; ~0> is C<-1> on twos-complement
1872machines.
68dc0745
PP
1873
1874=head2 Floating-point Arithmetic
1875
1876While C<use integer> provides integer-only arithmetic, there is no
19799a22
GS
1877analogous mechanism to provide automatic rounding or truncation to a
1878certain number of decimal places. For rounding to a certain number
1879of digits, sprintf() or printf() is usually the easiest route.
1880See L<perlfaq4>.
68dc0745 1881
5a964f20
TC
1882Floating-point numbers are only approximations to what a mathematician
1883would call real numbers. There are infinitely more reals than floats,
1884so some corners must be cut. For example:
1885
1886 printf "%.20g\n", 123456789123456789;
1887 # produces 123456789123456784
1888
1889Testing for exact equality of floating-point equality or inequality is
1890not a good idea. Here's a (relatively expensive) work-around to compare
1891whether two floating-point numbers are equal to a particular number of
1892decimal places. See Knuth, volume II, for a more robust treatment of
1893this topic.
1894
1895 sub fp_equal {
1896 my ($X, $Y, $POINTS) = @_;
1897 my ($tX, $tY);
1898 $tX = sprintf("%.${POINTS}g", $X);
1899 $tY = sprintf("%.${POINTS}g", $Y);
1900 return $tX eq $tY;
1901 }
1902
68dc0745 1903The POSIX module (part of the standard perl distribution) implements
19799a22
GS
1904ceil(), floor(), and other mathematical and trigonometric functions.
1905The Math::Complex module (part of the standard perl distribution)
1906defines mathematical functions that work on both the reals and the
1907imaginary numbers. Math::Complex not as efficient as POSIX, but
68dc0745
PP
1908POSIX can't work with complex numbers.
1909
1910Rounding in financial applications can have serious implications, and
1911the rounding method used should be specified precisely. In these
1912cases, it probably pays not to trust whichever system rounding is
1913being used by Perl, but to instead implement the rounding function you
1914need yourself.
5a964f20
TC
1915
1916=head2 Bigger Numbers
1917
1918The standard Math::BigInt and Math::BigFloat modules provide
19799a22
GS
1919variable-precision arithmetic and overloaded operators, although
1920they're currently pretty slow. At the cost of some space and
1921considerable speed, they avoid the normal pitfalls associated with
1922limited-precision representations.
5a964f20
TC
1923
1924 use Math::BigInt;
1925 $x = Math::BigInt->new('123456789123456789');
1926 print $x * $x;
1927
1928 # prints +15241578780673678515622620750190521
19799a22
GS
1929
1930The non-standard modules SSLeay::BN and Math::Pari provide
1931equivalent functionality (and much more) with a substantial
1932performance savings.
16070b82
GS
1933
1934=cut