This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
The rest of the changes for Sys-Syslog-0.17 which my new patch no
[perl5.git] / pod / perlop.pod
CommitLineData
a0d0e21e 1=head1 NAME
d74e8afc 2X<operator>
a0d0e21e
LW
3
4perlop - Perl operators and precedence
5
d042e63d
MS
6=head1 DESCRIPTION
7
89d205f2 8=head2 Operator Precedence and Associativity
d74e8afc 9X<operator, precedence> X<precedence> X<associativity>
d042e63d
MS
10
11Operator precedence and associativity work in Perl more or less like
12they do in mathematics.
13
14I<Operator precedence> means some operators are evaluated before
15others. For example, in C<2 + 4 * 5>, the multiplication has higher
16precedence so C<4 * 5> is evaluated first yielding C<2 + 20 ==
1722> and not C<6 * 5 == 30>.
18
19I<Operator associativity> defines what happens if a sequence of the
20same operators is used one after another: whether the evaluator will
21evaluate the left operations first or the right. For example, in C<8
22- 4 - 2>, subtraction is left associative so Perl evaluates the
23expression left to right. C<8 - 4> is evaluated first making the
24expression C<4 - 2 == 2> and not C<8 - 2 == 6>.
a0d0e21e
LW
25
26Perl operators have the following associativity and precedence,
19799a22
GS
27listed from highest precedence to lowest. Operators borrowed from
28C keep the same precedence relationship with each other, even where
29C's precedence is slightly screwy. (This makes learning Perl easier
30for C folks.) With very few exceptions, these all operate on scalar
31values only, not array values.
a0d0e21e
LW
32
33 left terms and list operators (leftward)
34 left ->
35 nonassoc ++ --
36 right **
37 right ! ~ \ and unary + and -
54310121 38 left =~ !~
a0d0e21e
LW
39 left * / % x
40 left + - .
41 left << >>
42 nonassoc named unary operators
43 nonassoc < > <= >= lt gt le ge
0d863452 44 nonassoc == != <=> eq ne cmp ~~
a0d0e21e
LW
45 left &
46 left | ^
47 left &&
c963b151 48 left || //
137443ea 49 nonassoc .. ...
a0d0e21e
LW
50 right ?:
51 right = += -= *= etc.
52 left , =>
53 nonassoc list operators (rightward)
a5f75d66 54 right not
a0d0e21e 55 left and
c963b151 56 left or xor err
a0d0e21e
LW
57
58In the following sections, these operators are covered in precedence order.
59
5a964f20
TC
60Many operators can be overloaded for objects. See L<overload>.
61
a0d0e21e 62=head2 Terms and List Operators (Leftward)
d74e8afc 63X<list operator> X<operator, list> X<term>
a0d0e21e 64
62c18ce2 65A TERM has the highest precedence in Perl. They include variables,
5f05dabc 66quote and quote-like operators, any expression in parentheses,
a0d0e21e
LW
67and any function whose arguments are parenthesized. Actually, there
68aren't really functions in this sense, just list operators and unary
69operators behaving as functions because you put parentheses around
70the arguments. These are all documented in L<perlfunc>.
71
72If any list operator (print(), etc.) or any unary operator (chdir(), etc.)
73is followed by a left parenthesis as the next token, the operator and
74arguments within parentheses are taken to be of highest precedence,
75just like a normal function call.
76
77In the absence of parentheses, the precedence of list operators such as
78C<print>, C<sort>, or C<chmod> is either very high or very low depending on
54310121 79whether you are looking at the left side or the right side of the operator.
a0d0e21e
LW
80For example, in
81
82 @ary = (1, 3, sort 4, 2);
83 print @ary; # prints 1324
84
19799a22
GS
85the commas on the right of the sort are evaluated before the sort,
86but the commas on the left are evaluated after. In other words,
87list operators tend to gobble up all arguments that follow, and
a0d0e21e 88then act like a simple TERM with regard to the preceding expression.
19799a22 89Be careful with parentheses:
a0d0e21e
LW
90
91 # These evaluate exit before doing the print:
92 print($foo, exit); # Obviously not what you want.
93 print $foo, exit; # Nor is this.
94
95 # These do the print before evaluating exit:
96 (print $foo), exit; # This is what you want.
97 print($foo), exit; # Or this.
98 print ($foo), exit; # Or even this.
99
100Also note that
101
102 print ($foo & 255) + 1, "\n";
103
d042e63d
MS
104probably doesn't do what you expect at first glance. The parentheses
105enclose the argument list for C<print> which is evaluated (printing
106the result of C<$foo & 255>). Then one is added to the return value
107of C<print> (usually 1). The result is something like this:
108
109 1 + 1, "\n"; # Obviously not what you meant.
110
111To do what you meant properly, you must write:
112
113 print(($foo & 255) + 1, "\n");
114
115See L<Named Unary Operators> for more discussion of this.
a0d0e21e
LW
116
117Also parsed as terms are the C<do {}> and C<eval {}> constructs, as
54310121 118well as subroutine and method calls, and the anonymous
a0d0e21e
LW
119constructors C<[]> and C<{}>.
120
2ae324a7 121See also L<Quote and Quote-like Operators> toward the end of this section,
da87341d 122as well as L</"I/O Operators">.
a0d0e21e
LW
123
124=head2 The Arrow Operator
d74e8afc 125X<arrow> X<dereference> X<< -> >>
a0d0e21e 126
35f2feb0 127"C<< -> >>" is an infix dereference operator, just as it is in C
19799a22
GS
128and C++. If the right side is either a C<[...]>, C<{...}>, or a
129C<(...)> subscript, then the left side must be either a hard or
130symbolic reference to an array, a hash, or a subroutine respectively.
131(Or technically speaking, a location capable of holding a hard
132reference, if it's an array or hash reference being used for
133assignment.) See L<perlreftut> and L<perlref>.
a0d0e21e 134
19799a22
GS
135Otherwise, the right side is a method name or a simple scalar
136variable containing either the method name or a subroutine reference,
137and the left side must be either an object (a blessed reference)
138or a class name (that is, a package name). See L<perlobj>.
a0d0e21e 139
5f05dabc 140=head2 Auto-increment and Auto-decrement
d74e8afc 141X<increment> X<auto-increment> X<++> X<decrement> X<auto-decrement> X<-->
a0d0e21e 142
d042e63d
MS
143"++" and "--" work as in C. That is, if placed before a variable,
144they increment or decrement the variable by one before returning the
145value, and if placed after, increment or decrement after returning the
146value.
147
148 $i = 0; $j = 0;
149 print $i++; # prints 0
150 print ++$j; # prints 1
a0d0e21e 151
b033823e 152Note that just as in C, Perl doesn't define B<when> the variable is
89d205f2 153incremented or decremented. You just know it will be done sometime
b033823e
A
154before or after the value is returned. This also means that modifying
155a variable twice in the same statement will lead to undefined behaviour.
156Avoid statements like:
157
158 $i = $i ++;
159 print ++ $i + $i ++;
160
161Perl will not guarantee what the result of the above statements is.
162
54310121 163The auto-increment operator has a little extra builtin magic to it. If
a0d0e21e
LW
164you increment a variable that is numeric, or that has ever been used in
165a numeric context, you get a normal increment. If, however, the
5f05dabc 166variable has been used in only string contexts since it was set, and
5a964f20 167has a value that is not the empty string and matches the pattern
9c0670e1 168C</^[a-zA-Z]*[0-9]*\z/>, the increment is done as a string, preserving each
a0d0e21e
LW
169character within its range, with carry:
170
171 print ++($foo = '99'); # prints '100'
172 print ++($foo = 'a0'); # prints 'a1'
173 print ++($foo = 'Az'); # prints 'Ba'
174 print ++($foo = 'zz'); # prints 'aaa'
175
6a61d433
HS
176C<undef> is always treated as numeric, and in particular is changed
177to C<0> before incrementing (so that a post-increment of an undef value
178will return C<0> rather than C<undef>).
179
5f05dabc 180The auto-decrement operator is not magical.
a0d0e21e
LW
181
182=head2 Exponentiation
d74e8afc 183X<**> X<exponentiation> X<power>
a0d0e21e 184
19799a22 185Binary "**" is the exponentiation operator. It binds even more
cb1a09d0
AD
186tightly than unary minus, so -2**4 is -(2**4), not (-2)**4. (This is
187implemented using C's pow(3) function, which actually works on doubles
188internally.)
a0d0e21e
LW
189
190=head2 Symbolic Unary Operators
d74e8afc 191X<unary operator> X<operator, unary>
a0d0e21e 192
5f05dabc 193Unary "!" performs logical negation, i.e., "not". See also C<not> for a lower
a0d0e21e 194precedence version of this.
d74e8afc 195X<!>
a0d0e21e
LW
196
197Unary "-" performs arithmetic negation if the operand is numeric. If
198the operand is an identifier, a string consisting of a minus sign
199concatenated with the identifier is returned. Otherwise, if the string
200starts with a plus or minus, a string starting with the opposite sign
bff5667c 201is returned. One effect of these rules is that -bareword is equivalent
8705167b 202to the string "-bareword". If, however, the string begins with a
06705523
SP
203non-alphabetic character (exluding "+" or "-"), Perl will attempt to convert
204the string to a numeric and the arithmetic negation is performed. If the
205string cannot be cleanly converted to a numeric, Perl will give the warning
206B<Argument "the string" isn't numeric in negation (-) at ...>.
d74e8afc 207X<-> X<negation, arithmetic>
a0d0e21e 208
972b05a9
JH
209Unary "~" performs bitwise negation, i.e., 1's complement. For
210example, C<0666 & ~027> is 0640. (See also L<Integer Arithmetic> and
211L<Bitwise String Operators>.) Note that the width of the result is
212platform-dependent: ~0 is 32 bits wide on a 32-bit platform, but 64
213bits wide on a 64-bit platform, so if you are expecting a certain bit
d042e63d 214width, remember to use the & operator to mask off the excess bits.
d74e8afc 215X<~> X<negation, binary>
a0d0e21e
LW
216
217Unary "+" has no effect whatsoever, even on strings. It is useful
218syntactically for separating a function name from a parenthesized expression
219that would otherwise be interpreted as the complete list of function
5ba421f6 220arguments. (See examples above under L<Terms and List Operators (Leftward)>.)
d74e8afc 221X<+>
a0d0e21e 222
19799a22
GS
223Unary "\" creates a reference to whatever follows it. See L<perlreftut>
224and L<perlref>. Do not confuse this behavior with the behavior of
225backslash within a string, although both forms do convey the notion
226of protecting the next thing from interpolation.
d74e8afc 227X<\> X<reference> X<backslash>
a0d0e21e
LW
228
229=head2 Binding Operators
d74e8afc 230X<binding> X<operator, binding> X<=~> X<!~>
a0d0e21e 231
c07a80fd 232Binary "=~" binds a scalar expression to a pattern match. Certain operations
cb1a09d0
AD
233search or modify the string $_ by default. This operator makes that kind
234of operation work on some other string. The right argument is a search
2c268ad5
TP
235pattern, substitution, or transliteration. The left argument is what is
236supposed to be searched, substituted, or transliterated instead of the default
f8bab1e9
GS
237$_. When used in scalar context, the return value generally indicates the
238success of the operation. Behavior in list context depends on the particular
89d205f2 239operator. See L</"Regexp Quote-Like Operators"> for details and
d7782e69 240L<perlretut> for examples using these operators.
f8bab1e9
GS
241
242If the right argument is an expression rather than a search pattern,
2c268ad5 243substitution, or transliteration, it is interpreted as a search pattern at run
89d205f2
YO
244time. Note that this means that its contents will be interpolated twice, so
245
246 '\\' =~ q'\\';
247
248is not ok, as the regex engine will end up trying to compile the
249pattern C<\>, which it will consider a syntax error.
a0d0e21e
LW
250
251Binary "!~" is just like "=~" except the return value is negated in
252the logical sense.
253
254=head2 Multiplicative Operators
d74e8afc 255X<operator, multiplicative>
a0d0e21e
LW
256
257Binary "*" multiplies two numbers.
d74e8afc 258X<*>
a0d0e21e
LW
259
260Binary "/" divides two numbers.
d74e8afc 261X</> X<slash>
a0d0e21e 262
54310121
PP
263Binary "%" computes the modulus of two numbers. Given integer
264operands C<$a> and C<$b>: If C<$b> is positive, then C<$a % $b> is
265C<$a> minus the largest multiple of C<$b> that is not greater than
266C<$a>. If C<$b> is negative, then C<$a % $b> is C<$a> minus the
267smallest multiple of C<$b> that is not less than C<$a> (i.e. the
89b4f0ad 268result will be less than or equal to zero). If the operands
89d205f2 269C<$a> and C<$b> are floting point values, only the integer portion
89b4f0ad 270of C<$a> and C<$b> will be used in the operation.
0412d526 271Note that when C<use integer> is in scope, "%" gives you direct access
55d729e4
GS
272to the modulus operator as implemented by your C compiler. This
273operator is not as well defined for negative operands, but it will
274execute faster.
d74e8afc 275X<%> X<remainder> X<modulus> X<mod>
55d729e4 276
62d10b70
GS
277Binary "x" is the repetition operator. In scalar context or if the left
278operand is not enclosed in parentheses, it returns a string consisting
279of the left operand repeated the number of times specified by the right
280operand. In list context, if the left operand is enclosed in
3585017f
YST
281parentheses or is a list formed by C<qw/STRING/>, it repeats the list.
282If the right operand is zero or negative, it returns an empty string
283or an empty list, depending on the context.
d74e8afc 284X<x>
a0d0e21e
LW
285
286 print '-' x 80; # print row of dashes
287
288 print "\t" x ($tab/8), ' ' x ($tab%8); # tab over
289
290 @ones = (1) x 80; # a list of 80 1's
291 @ones = (5) x @ones; # set all elements to 5
292
293
294=head2 Additive Operators
d74e8afc 295X<operator, additive>
a0d0e21e
LW
296
297Binary "+" returns the sum of two numbers.
d74e8afc 298X<+>
a0d0e21e
LW
299
300Binary "-" returns the difference of two numbers.
d74e8afc 301X<->
a0d0e21e
LW
302
303Binary "." concatenates two strings.
d74e8afc
ITB
304X<string, concatenation> X<concatenation>
305X<cat> X<concat> X<concatenate> X<.>
a0d0e21e
LW
306
307=head2 Shift Operators
d74e8afc
ITB
308X<shift operator> X<operator, shift> X<<< << >>>
309X<<< >> >>> X<right shift> X<left shift> X<bitwise shift>
310X<shl> X<shr> X<shift, right> X<shift, left>
a0d0e21e 311
55497cff
PP
312Binary "<<" returns the value of its left argument shifted left by the
313number of bits specified by the right argument. Arguments should be
982ce180 314integers. (See also L<Integer Arithmetic>.)
a0d0e21e 315
55497cff
PP
316Binary ">>" returns the value of its left argument shifted right by
317the number of bits specified by the right argument. Arguments should
982ce180 318be integers. (See also L<Integer Arithmetic>.)
a0d0e21e 319
b16cf6df
JH
320Note that both "<<" and ">>" in Perl are implemented directly using
321"<<" and ">>" in C. If C<use integer> (see L<Integer Arithmetic>) is
322in force then signed C integers are used, else unsigned C integers are
323used. Either way, the implementation isn't going to generate results
324larger than the size of the integer type Perl was built with (32 bits
325or 64 bits).
326
327The result of overflowing the range of the integers is undefined
328because it is undefined also in C. In other words, using 32-bit
329integers, C<< 1 << 32 >> is undefined. Shifting by a negative number
330of bits is also undefined.
331
a0d0e21e 332=head2 Named Unary Operators
d74e8afc 333X<operator, named unary>
a0d0e21e
LW
334
335The various named unary operators are treated as functions with one
568e6d8b 336argument, with optional parentheses.
a0d0e21e
LW
337
338If any list operator (print(), etc.) or any unary operator (chdir(), etc.)
339is followed by a left parenthesis as the next token, the operator and
340arguments within parentheses are taken to be of highest precedence,
3981b0eb 341just like a normal function call. For example,
342because named unary operators are higher precedence than ||:
a0d0e21e
LW
343
344 chdir $foo || die; # (chdir $foo) || die
345 chdir($foo) || die; # (chdir $foo) || die
346 chdir ($foo) || die; # (chdir $foo) || die
347 chdir +($foo) || die; # (chdir $foo) || die
348
3981b0eb 349but, because * is higher precedence than named operators:
a0d0e21e
LW
350
351 chdir $foo * 20; # chdir ($foo * 20)
352 chdir($foo) * 20; # (chdir $foo) * 20
353 chdir ($foo) * 20; # (chdir $foo) * 20
354 chdir +($foo) * 20; # chdir ($foo * 20)
355
356 rand 10 * 20; # rand (10 * 20)
357 rand(10) * 20; # (rand 10) * 20
358 rand (10) * 20; # (rand 10) * 20
359 rand +(10) * 20; # rand (10 * 20)
360
568e6d8b
RGS
361Regarding precedence, the filetest operators, like C<-f>, C<-M>, etc. are
362treated like named unary operators, but they don't follow this functional
363parenthesis rule. That means, for example, that C<-f($file).".bak"> is
364equivalent to C<-f "$file.bak">.
d74e8afc 365X<-X> X<filetest> X<operator, filetest>
568e6d8b 366
5ba421f6 367See also L<"Terms and List Operators (Leftward)">.
a0d0e21e
LW
368
369=head2 Relational Operators
d74e8afc 370X<relational operator> X<operator, relational>
a0d0e21e 371
35f2feb0 372Binary "<" returns true if the left argument is numerically less than
a0d0e21e 373the right argument.
d74e8afc 374X<< < >>
a0d0e21e 375
35f2feb0 376Binary ">" returns true if the left argument is numerically greater
a0d0e21e 377than the right argument.
d74e8afc 378X<< > >>
a0d0e21e 379
35f2feb0 380Binary "<=" returns true if the left argument is numerically less than
a0d0e21e 381or equal to the right argument.
d74e8afc 382X<< <= >>
a0d0e21e 383
35f2feb0 384Binary ">=" returns true if the left argument is numerically greater
a0d0e21e 385than or equal to the right argument.
d74e8afc 386X<< >= >>
a0d0e21e
LW
387
388Binary "lt" returns true if the left argument is stringwise less than
389the right argument.
d74e8afc 390X<< lt >>
a0d0e21e
LW
391
392Binary "gt" returns true if the left argument is stringwise greater
393than the right argument.
d74e8afc 394X<< gt >>
a0d0e21e
LW
395
396Binary "le" returns true if the left argument is stringwise less than
397or equal to the right argument.
d74e8afc 398X<< le >>
a0d0e21e
LW
399
400Binary "ge" returns true if the left argument is stringwise greater
401than or equal to the right argument.
d74e8afc 402X<< ge >>
a0d0e21e
LW
403
404=head2 Equality Operators
d74e8afc 405X<equality> X<equal> X<equals> X<operator, equality>
a0d0e21e
LW
406
407Binary "==" returns true if the left argument is numerically equal to
408the right argument.
d74e8afc 409X<==>
a0d0e21e
LW
410
411Binary "!=" returns true if the left argument is numerically not equal
412to the right argument.
d74e8afc 413X<!=>
a0d0e21e 414
35f2feb0 415Binary "<=>" returns -1, 0, or 1 depending on whether the left
6ee5d4e7 416argument is numerically less than, equal to, or greater than the right
d4ad863d 417argument. If your platform supports NaNs (not-a-numbers) as numeric
7d3a9d88
NC
418values, using them with "<=>" returns undef. NaN is not "<", "==", ">",
419"<=" or ">=" anything (even NaN), so those 5 return false. NaN != NaN
420returns true, as does NaN != anything else. If your platform doesn't
421support NaNs then NaN is just a string with numeric value 0.
d74e8afc 422X<< <=> >> X<spaceship>
7d3a9d88 423
2b54f59f
YST
424 perl -le '$a = "NaN"; print "No NaN support here" if $a == $a'
425 perl -le '$a = "NaN"; print "NaN support here" if $a != $a'
a0d0e21e
LW
426
427Binary "eq" returns true if the left argument is stringwise equal to
428the right argument.
d74e8afc 429X<eq>
a0d0e21e
LW
430
431Binary "ne" returns true if the left argument is stringwise not equal
432to the right argument.
d74e8afc 433X<ne>
a0d0e21e 434
d4ad863d
JH
435Binary "cmp" returns -1, 0, or 1 depending on whether the left
436argument is stringwise less than, equal to, or greater than the right
437argument.
d74e8afc 438X<cmp>
a0d0e21e 439
0d863452
RH
440Binary "~~" does a smart match between its arguments. Smart matching
441is described in L<perlsyn/"Smart Matching in Detail">.
442This operator is only available if you enable the "~~" feature:
443see L<feature> for more information.
444X<~~>
445
a034a98d
DD
446"lt", "le", "ge", "gt" and "cmp" use the collation (sort) order specified
447by the current locale if C<use locale> is in effect. See L<perllocale>.
448
a0d0e21e 449=head2 Bitwise And
d74e8afc 450X<operator, bitwise, and> X<bitwise and> X<&>
a0d0e21e 451
2cdc098b 452Binary "&" returns its operands ANDed together bit by bit.
2c268ad5 453(See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
a0d0e21e 454
2cdc098b
MG
455Note that "&" has lower priority than relational operators, so for example
456the brackets are essential in a test like
457
458 print "Even\n" if ($x & 1) == 0;
459
a0d0e21e 460=head2 Bitwise Or and Exclusive Or
d74e8afc
ITB
461X<operator, bitwise, or> X<bitwise or> X<|> X<operator, bitwise, xor>
462X<bitwise xor> X<^>
a0d0e21e 463
2cdc098b 464Binary "|" returns its operands ORed together bit by bit.
2c268ad5 465(See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
a0d0e21e 466
2cdc098b 467Binary "^" returns its operands XORed together bit by bit.
2c268ad5 468(See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
a0d0e21e 469
2cdc098b
MG
470Note that "|" and "^" have lower priority than relational operators, so
471for example the brackets are essential in a test like
472
473 print "false\n" if (8 | 2) != 10;
474
a0d0e21e 475=head2 C-style Logical And
d74e8afc 476X<&&> X<logical and> X<operator, logical, and>
a0d0e21e
LW
477
478Binary "&&" performs a short-circuit logical AND operation. That is,
479if the left operand is false, the right operand is not even evaluated.
480Scalar or list context propagates down to the right operand if it
481is evaluated.
482
483=head2 C-style Logical Or
d74e8afc 484X<||> X<operator, logical, or>
a0d0e21e
LW
485
486Binary "||" performs a short-circuit logical OR operation. That is,
487if the left operand is true, the right operand is not even evaluated.
488Scalar or list context propagates down to the right operand if it
489is evaluated.
490
c963b151 491=head2 C-style Logical Defined-Or
d74e8afc 492X<//> X<operator, logical, defined-or>
c963b151
BD
493
494Although it has no direct equivalent in C, Perl's C<//> operator is related
89d205f2 495to its C-style or. In fact, it's exactly the same as C<||>, except that it
c963b151 496tests the left hand side's definedness instead of its truth. Thus, C<$a // $b>
89d205f2
YO
497is similar to C<defined($a) || $b> (except that it returns the value of C<$a>
498rather than the value of C<defined($a)>) and is exactly equivalent to
c963b151 499C<defined($a) ? $a : $b>. This is very useful for providing default values
89d205f2 500for variables. If you actually want to test if at least one of C<$a> and
d042e63d 501C<$b> is defined, use C<defined($a // $b)>.
c963b151 502
d042e63d
MS
503The C<||>, C<//> and C<&&> operators return the last value evaluated
504(unlike C's C<||> and C<&&>, which return 0 or 1). Thus, a reasonably
505portable way to find out the home directory might be:
a0d0e21e 506
c963b151
BD
507 $home = $ENV{'HOME'} // $ENV{'LOGDIR'} //
508 (getpwuid($<))[7] // die "You're homeless!\n";
a0d0e21e 509
5a964f20
TC
510In particular, this means that you shouldn't use this
511for selecting between two aggregates for assignment:
512
513 @a = @b || @c; # this is wrong
514 @a = scalar(@b) || @c; # really meant this
515 @a = @b ? @b : @c; # this works fine, though
516
c963b151
BD
517As more readable alternatives to C<&&>, C<//> and C<||> when used for
518control flow, Perl provides C<and>, C<err> and C<or> operators (see below).
89d205f2 519The short-circuit behavior is identical. The precedence of "and", "err"
c963b151 520and "or" is much lower, however, so that you can safely use them after a
5a964f20 521list operator without the need for parentheses:
a0d0e21e
LW
522
523 unlink "alpha", "beta", "gamma"
524 or gripe(), next LINE;
525
526With the C-style operators that would have been written like this:
527
528 unlink("alpha", "beta", "gamma")
529 || (gripe(), next LINE);
530
eeb6a2c9 531Using "or" for assignment is unlikely to do what you want; see below.
5a964f20
TC
532
533=head2 Range Operators
d74e8afc 534X<operator, range> X<range> X<..> X<...>
a0d0e21e
LW
535
536Binary ".." is the range operator, which is really two different
fb53bbb2 537operators depending on the context. In list context, it returns a
54ae734e 538list of values counting (up by ones) from the left value to the right
2cdbc966 539value. If the left value is greater than the right value then it
fb53bbb2 540returns the empty list. The range operator is useful for writing
54ae734e 541C<foreach (1..10)> loops and for doing slice operations on arrays. In
2cdbc966
JD
542the current implementation, no temporary array is created when the
543range operator is used as the expression in C<foreach> loops, but older
544versions of Perl might burn a lot of memory when you write something
545like this:
a0d0e21e
LW
546
547 for (1 .. 1_000_000) {
548 # code
54310121 549 }
a0d0e21e 550
54ae734e
MG
551The range operator also works on strings, using the magical auto-increment,
552see below.
553
5a964f20 554In scalar context, ".." returns a boolean value. The operator is
a0d0e21e
LW
555bistable, like a flip-flop, and emulates the line-range (comma) operator
556of B<sed>, B<awk>, and various editors. Each ".." operator maintains its
557own boolean state. It is false as long as its left operand is false.
558Once the left operand is true, the range operator stays true until the
559right operand is true, I<AFTER> which the range operator becomes false
19799a22 560again. It doesn't become false till the next time the range operator is
a0d0e21e
LW
561evaluated. It can test the right operand and become false on the same
562evaluation it became true (as in B<awk>), but it still returns true once.
19799a22
GS
563If you don't want it to test the right operand till the next
564evaluation, as in B<sed>, just use three dots ("...") instead of
565two. In all other regards, "..." behaves just like ".." does.
566
567The right operand is not evaluated while the operator is in the
568"false" state, and the left operand is not evaluated while the
569operator is in the "true" state. The precedence is a little lower
570than || and &&. The value returned is either the empty string for
571false, or a sequence number (beginning with 1) for true. The
572sequence number is reset for each range encountered. The final
573sequence number in a range has the string "E0" appended to it, which
574doesn't affect its numeric value, but gives you something to search
575for if you want to exclude the endpoint. You can exclude the
576beginning point by waiting for the sequence number to be greater
df5f8116
CW
577than 1.
578
579If either operand of scalar ".." is a constant expression,
580that operand is considered true if it is equal (C<==>) to the current
581input line number (the C<$.> variable).
582
583To be pedantic, the comparison is actually C<int(EXPR) == int(EXPR)>,
584but that is only an issue if you use a floating point expression; when
585implicitly using C<$.> as described in the previous paragraph, the
586comparison is C<int(EXPR) == int($.)> which is only an issue when C<$.>
587is set to a floating point value and you are not reading from a file.
588Furthermore, C<"span" .. "spat"> or C<2.18 .. 3.14> will not do what
589you want in scalar context because each of the operands are evaluated
590using their integer representation.
591
592Examples:
a0d0e21e
LW
593
594As a scalar operator:
595
df5f8116
CW
596 if (101 .. 200) { print; } # print 2nd hundred lines, short for
597 # if ($. == 101 .. $. == 200) ...
9f10b797
RGS
598
599 next LINE if (1 .. /^$/); # skip header lines, short for
df5f8116 600 # ... if ($. == 1 .. /^$/);
9f10b797
RGS
601 # (typically in a loop labeled LINE)
602
603 s/^/> / if (/^$/ .. eof()); # quote body
a0d0e21e 604
5a964f20
TC
605 # parse mail messages
606 while (<>) {
607 $in_header = 1 .. /^$/;
df5f8116
CW
608 $in_body = /^$/ .. eof;
609 if ($in_header) {
610 # ...
611 } else { # in body
612 # ...
613 }
5a964f20 614 } continue {
df5f8116 615 close ARGV if eof; # reset $. each file
5a964f20
TC
616 }
617
acf31ca5
SF
618Here's a simple example to illustrate the difference between
619the two range operators:
620
621 @lines = (" - Foo",
622 "01 - Bar",
623 "1 - Baz",
624 " - Quux");
625
9f10b797
RGS
626 foreach (@lines) {
627 if (/0/ .. /1/) {
acf31ca5
SF
628 print "$_\n";
629 }
630 }
631
9f10b797
RGS
632This program will print only the line containing "Bar". If
633the range operator is changed to C<...>, it will also print the
acf31ca5
SF
634"Baz" line.
635
636And now some examples as a list operator:
a0d0e21e
LW
637
638 for (101 .. 200) { print; } # print $_ 100 times
3e3baf6d 639 @foo = @foo[0 .. $#foo]; # an expensive no-op
a0d0e21e
LW
640 @foo = @foo[$#foo-4 .. $#foo]; # slice last 5 items
641
5a964f20 642The range operator (in list context) makes use of the magical
5f05dabc 643auto-increment algorithm if the operands are strings. You
a0d0e21e
LW
644can say
645
646 @alphabet = ('A' .. 'Z');
647
54ae734e 648to get all normal letters of the English alphabet, or
a0d0e21e
LW
649
650 $hexdigit = (0 .. 9, 'a' .. 'f')[$num & 15];
651
652to get a hexadecimal digit, or
653
654 @z2 = ('01' .. '31'); print $z2[$mday];
655
ea4f5703
YST
656to get dates with leading zeros.
657
658If the final value specified is not in the sequence that the magical
659increment would produce, the sequence goes until the next value would
660be longer than the final value specified.
661
662If the initial value specified isn't part of a magical increment
663sequence (that is, a non-empty string matching "/^[a-zA-Z]*[0-9]*\z/"),
664only the initial value will be returned. So the following will only
665return an alpha:
666
667 use charnames 'greek';
668 my @greek_small = ("\N{alpha}" .. "\N{omega}");
669
670To get lower-case greek letters, use this instead:
671
672 my @greek_small = map { chr } ( ord("\N{alpha}") .. ord("\N{omega}") );
a0d0e21e 673
df5f8116
CW
674Because each operand is evaluated in integer form, C<2.18 .. 3.14> will
675return two elements in list context.
676
677 @list = (2.18 .. 3.14); # same as @list = (2 .. 3);
678
a0d0e21e 679=head2 Conditional Operator
d74e8afc 680X<operator, conditional> X<operator, ternary> X<ternary> X<?:>
a0d0e21e
LW
681
682Ternary "?:" is the conditional operator, just as in C. It works much
683like an if-then-else. If the argument before the ? is true, the
684argument before the : is returned, otherwise the argument after the :
cb1a09d0
AD
685is returned. For example:
686
54310121 687 printf "I have %d dog%s.\n", $n,
cb1a09d0
AD
688 ($n == 1) ? '' : "s";
689
690Scalar or list context propagates downward into the 2nd
54310121 691or 3rd argument, whichever is selected.
cb1a09d0
AD
692
693 $a = $ok ? $b : $c; # get a scalar
694 @a = $ok ? @b : @c; # get an array
695 $a = $ok ? @b : @c; # oops, that's just a count!
696
697The operator may be assigned to if both the 2nd and 3rd arguments are
698legal lvalues (meaning that you can assign to them):
a0d0e21e
LW
699
700 ($a_or_b ? $a : $b) = $c;
701
5a964f20
TC
702Because this operator produces an assignable result, using assignments
703without parentheses will get you in trouble. For example, this:
704
705 $a % 2 ? $a += 10 : $a += 2
706
707Really means this:
708
709 (($a % 2) ? ($a += 10) : $a) += 2
710
711Rather than this:
712
713 ($a % 2) ? ($a += 10) : ($a += 2)
714
19799a22
GS
715That should probably be written more simply as:
716
717 $a += ($a % 2) ? 10 : 2;
718
4633a7c4 719=head2 Assignment Operators
d74e8afc 720X<assignment> X<operator, assignment> X<=> X<**=> X<+=> X<*=> X<&=>
5ac3b81c 721X<<< <<= >>> X<&&=> X<-=> X</=> X<|=> X<<< >>= >>> X<||=> X<//=> X<.=>
d74e8afc 722X<%=> X<^=> X<x=>
a0d0e21e
LW
723
724"=" is the ordinary assignment operator.
725
726Assignment operators work as in C. That is,
727
728 $a += 2;
729
730is equivalent to
731
732 $a = $a + 2;
733
734although without duplicating any side effects that dereferencing the lvalue
54310121
PP
735might trigger, such as from tie(). Other assignment operators work similarly.
736The following are recognized:
a0d0e21e
LW
737
738 **= += *= &= <<= &&=
9f10b797
RGS
739 -= /= |= >>= ||=
740 .= %= ^= //=
741 x=
a0d0e21e 742
19799a22 743Although these are grouped by family, they all have the precedence
a0d0e21e
LW
744of assignment.
745
b350dd2f
GS
746Unlike in C, the scalar assignment operator produces a valid lvalue.
747Modifying an assignment is equivalent to doing the assignment and
748then modifying the variable that was assigned to. This is useful
749for modifying a copy of something, like this:
a0d0e21e
LW
750
751 ($tmp = $global) =~ tr [A-Z] [a-z];
752
753Likewise,
754
755 ($a += 2) *= 3;
756
757is equivalent to
758
759 $a += 2;
760 $a *= 3;
761
b350dd2f
GS
762Similarly, a list assignment in list context produces the list of
763lvalues assigned to, and a list assignment in scalar context returns
764the number of elements produced by the expression on the right hand
765side of the assignment.
766
748a9306 767=head2 Comma Operator
d74e8afc 768X<comma> X<operator, comma> X<,>
a0d0e21e 769
5a964f20 770Binary "," is the comma operator. In scalar context it evaluates
a0d0e21e
LW
771its left argument, throws that value away, then evaluates its right
772argument and returns that value. This is just like C's comma operator.
773
5a964f20 774In list context, it's just the list argument separator, and inserts
ed5c6d31
PJ
775both its arguments into the list. These arguments are also evaluated
776from left to right.
a0d0e21e 777
d042e63d 778The C<< => >> operator is a synonym for the comma, but forces any word
719b43e8 779(consisting entirely of word characters) to its left to be interpreted
a44e5664
MS
780as a string (as of 5.001). This includes words that might otherwise be
781considered a constant or function call.
782
783 use constant FOO => "something";
784
785 my %h = ( FOO => 23 );
786
787is equivalent to:
788
789 my %h = ("FOO", 23);
790
791It is I<NOT>:
792
793 my %h = ("something", 23);
794
795If the argument on the left is not a word, it is first interpreted as
796an expression, and then the string value of that is used.
719b43e8
RGS
797
798The C<< => >> operator is helpful in documenting the correspondence
799between keys and values in hashes, and other paired elements in lists.
748a9306 800
a44e5664
MS
801 %hash = ( $key => $value );
802 login( $username => $password );
803
a0d0e21e 804=head2 List Operators (Rightward)
d74e8afc 805X<operator, list, rightward> X<list operator>
a0d0e21e
LW
806
807On the right side of a list operator, it has very low precedence,
808such that it controls all comma-separated expressions found there.
809The only operators with lower precedence are the logical operators
810"and", "or", and "not", which may be used to evaluate calls to list
811operators without the need for extra parentheses:
812
813 open HANDLE, "filename"
814 or die "Can't open: $!\n";
815
5ba421f6 816See also discussion of list operators in L<Terms and List Operators (Leftward)>.
a0d0e21e
LW
817
818=head2 Logical Not
d74e8afc 819X<operator, logical, not> X<not>
a0d0e21e
LW
820
821Unary "not" returns the logical negation of the expression to its right.
822It's the equivalent of "!" except for the very low precedence.
823
824=head2 Logical And
d74e8afc 825X<operator, logical, and> X<and>
a0d0e21e
LW
826
827Binary "and" returns the logical conjunction of the two surrounding
828expressions. It's equivalent to && except for the very low
5f05dabc 829precedence. This means that it short-circuits: i.e., the right
a0d0e21e
LW
830expression is evaluated only if the left expression is true.
831
c963b151 832=head2 Logical or, Defined or, and Exclusive Or
d74e8afc
ITB
833X<operator, logical, or> X<operator, logical, xor> X<operator, logical, err>
834X<operator, logical, defined or> X<operator, logical, exclusive or>
835X<or> X<xor> X<err>
a0d0e21e
LW
836
837Binary "or" returns the logical disjunction of the two surrounding
5a964f20
TC
838expressions. It's equivalent to || except for the very low precedence.
839This makes it useful for control flow
840
841 print FH $data or die "Can't write to FH: $!";
842
843This means that it short-circuits: i.e., the right expression is evaluated
844only if the left expression is false. Due to its precedence, you should
845probably avoid using this for assignment, only for control flow.
846
847 $a = $b or $c; # bug: this is wrong
848 ($a = $b) or $c; # really means this
849 $a = $b || $c; # better written this way
850
19799a22 851However, when it's a list-context assignment and you're trying to use
5a964f20
TC
852"||" for control flow, you probably need "or" so that the assignment
853takes higher precedence.
854
855 @info = stat($file) || die; # oops, scalar sense of stat!
856 @info = stat($file) or die; # better, now @info gets its due
857
c963b151
BD
858Then again, you could always use parentheses.
859
9f10b797
RGS
860Binary "err" is equivalent to C<//>--it's just like binary "or", except it
861tests its left argument's definedness instead of its truth. There are two
862ways to remember "err": either because many functions return C<undef> on
863an B<err>or, or as a sort of correction: C<$a = ($b err 'default')>. This
864keyword is only available when the 'err' feature is enabled: see
865L<feature> for more information.
a0d0e21e
LW
866
867Binary "xor" returns the exclusive-OR of the two surrounding expressions.
868It cannot short circuit, of course.
869
870=head2 C Operators Missing From Perl
d74e8afc
ITB
871X<operator, missing from perl> X<&> X<*>
872X<typecasting> X<(TYPE)>
a0d0e21e
LW
873
874Here is what C has that Perl doesn't:
875
876=over 8
877
878=item unary &
879
880Address-of operator. (But see the "\" operator for taking a reference.)
881
882=item unary *
883
54310121 884Dereference-address operator. (Perl's prefix dereferencing
a0d0e21e
LW
885operators are typed: $, @, %, and &.)
886
887=item (TYPE)
888
19799a22 889Type-casting operator.
a0d0e21e
LW
890
891=back
892
5f05dabc 893=head2 Quote and Quote-like Operators
89d205f2 894X<operator, quote> X<operator, quote-like> X<q> X<qq> X<qx> X<qw> X<m>
d74e8afc
ITB
895X<qr> X<s> X<tr> X<'> X<''> X<"> X<""> X<//> X<`> X<``> X<<< << >>>
896X<escape sequence> X<escape>
897
a0d0e21e
LW
898
899While we usually think of quotes as literal values, in Perl they
900function as operators, providing various kinds of interpolating and
901pattern matching capabilities. Perl provides customary quote characters
902for these behaviors, but also provides a way for you to choose your
903quote character for any of them. In the following table, a C<{}> represents
9f10b797 904any pair of delimiters you choose.
a0d0e21e 905
2c268ad5
TP
906 Customary Generic Meaning Interpolates
907 '' q{} Literal no
908 "" qq{} Literal yes
af9219ee 909 `` qx{} Command yes*
2c268ad5 910 qw{} Word list no
af9219ee
MG
911 // m{} Pattern match yes*
912 qr{} Pattern yes*
913 s{}{} Substitution yes*
2c268ad5 914 tr{}{} Transliteration no (but see below)
7e3b091d 915 <<EOF here-doc yes*
a0d0e21e 916
af9219ee
MG
917 * unless the delimiter is ''.
918
87275199
GS
919Non-bracketing delimiters use the same character fore and aft, but the four
920sorts of brackets (round, angle, square, curly) will all nest, which means
9f10b797 921that
87275199 922
9f10b797 923 q{foo{bar}baz}
35f2feb0 924
9f10b797 925is the same as
87275199
GS
926
927 'foo{bar}baz'
928
929Note, however, that this does not always work for quoting Perl code:
930
931 $s = q{ if($a eq "}") ... }; # WRONG
932
83df6a1d
JH
933is a syntax error. The C<Text::Balanced> module (from CPAN, and
934starting from Perl 5.8 part of the standard distribution) is able
935to do this properly.
87275199 936
19799a22 937There can be whitespace between the operator and the quoting
fb73857a 938characters, except when C<#> is being used as the quoting character.
19799a22
GS
939C<q#foo#> is parsed as the string C<foo>, while C<q #foo#> is the
940operator C<q> followed by a comment. Its argument will be taken
941from the next line. This allows you to write:
fb73857a
PP
942
943 s {foo} # Replace foo
944 {bar} # with bar.
945
904501ec
MG
946The following escape sequences are available in constructs that interpolate
947and in transliterations.
d74e8afc 948X<\t> X<\n> X<\r> X<\f> X<\b> X<\a> X<\e> X<\x> X<\0> X<\c> X<\N>
a0d0e21e 949
6ee5d4e7 950 \t tab (HT, TAB)
5a964f20 951 \n newline (NL)
6ee5d4e7
PP
952 \r return (CR)
953 \f form feed (FF)
954 \b backspace (BS)
955 \a alarm (bell) (BEL)
956 \e escape (ESC)
a0ed51b3
LW
957 \033 octal char (ESC)
958 \x1b hex char (ESC)
959 \x{263a} wide hex char (SMILEY)
19799a22 960 \c[ control char (ESC)
95cc3e0c 961 \N{name} named Unicode character
2c268ad5 962
4c77eaa2
AE
963B<NOTE>: Unlike C and other languages, Perl has no \v escape sequence for
964the vertical tab (VT - ASCII 11).
965
904501ec
MG
966The following escape sequences are available in constructs that interpolate
967but not in transliterations.
d74e8afc 968X<\l> X<\u> X<\L> X<\U> X<\E> X<\Q>
904501ec 969
a0d0e21e
LW
970 \l lowercase next char
971 \u uppercase next char
972 \L lowercase till \E
973 \U uppercase till \E
974 \E end case modification
1d2dff63 975 \Q quote non-word characters till \E
a0d0e21e 976
95cc3e0c
JH
977If C<use locale> is in effect, the case map used by C<\l>, C<\L>,
978C<\u> and C<\U> is taken from the current locale. See L<perllocale>.
979If Unicode (for example, C<\N{}> or wide hex characters of 0x100 or
980beyond) is being used, the case map used by C<\l>, C<\L>, C<\u> and
981C<\U> is as defined by Unicode. For documentation of C<\N{name}>,
982see L<charnames>.
a034a98d 983
5a964f20
TC
984All systems use the virtual C<"\n"> to represent a line terminator,
985called a "newline". There is no such thing as an unvarying, physical
19799a22 986newline character. It is only an illusion that the operating system,
5a964f20
TC
987device drivers, C libraries, and Perl all conspire to preserve. Not all
988systems read C<"\r"> as ASCII CR and C<"\n"> as ASCII LF. For example,
989on a Mac, these are reversed, and on systems without line terminator,
990printing C<"\n"> may emit no actual data. In general, use C<"\n"> when
991you mean a "newline" for your system, but use the literal ASCII when you
992need an exact character. For example, most networking protocols expect
2a380090 993and prefer a CR+LF (C<"\015\012"> or C<"\cM\cJ">) for line terminators,
5a964f20
TC
994and although they often accept just C<"\012">, they seldom tolerate just
995C<"\015">. If you get in the habit of using C<"\n"> for networking,
996you may be burned some day.
d74e8afc
ITB
997X<newline> X<line terminator> X<eol> X<end of line>
998X<\n> X<\r> X<\r\n>
5a964f20 999
904501ec
MG
1000For constructs that do interpolate, variables beginning with "C<$>"
1001or "C<@>" are interpolated. Subscripted variables such as C<$a[3]> or
ad0f383a
A
1002C<< $href->{key}[0] >> are also interpolated, as are array and hash slices.
1003But method calls such as C<< $obj->meth >> are not.
af9219ee
MG
1004
1005Interpolating an array or slice interpolates the elements in order,
1006separated by the value of C<$">, so is equivalent to interpolating
904501ec
MG
1007C<join $", @array>. "Punctuation" arrays such as C<@+> are only
1008interpolated if the name is enclosed in braces C<@{+}>.
af9219ee 1009
89d205f2
YO
1010You cannot include a literal C<$> or C<@> within a C<\Q> sequence.
1011An unescaped C<$> or C<@> interpolates the corresponding variable,
1d2dff63 1012while escaping will cause the literal string C<\$> to be inserted.
89d205f2 1013You'll need to write something like C<m/\Quser\E\@\Qhost/>.
1d2dff63 1014
a0d0e21e
LW
1015Patterns are subject to an additional level of interpretation as a
1016regular expression. This is done as a second pass, after variables are
1017interpolated, so that regular expressions may be incorporated into the
1018pattern from the variables. If this is not what you want, use C<\Q> to
1019interpolate a variable literally.
1020
19799a22
GS
1021Apart from the behavior described above, Perl does not expand
1022multiple levels of interpolation. In particular, contrary to the
1023expectations of shell programmers, back-quotes do I<NOT> interpolate
1024within double quotes, nor do single quotes impede evaluation of
1025variables when used within double quotes.
a0d0e21e 1026
5f05dabc 1027=head2 Regexp Quote-Like Operators
d74e8afc 1028X<operator, regexp>
cb1a09d0 1029
5f05dabc 1030Here are the quote-like operators that apply to pattern
cb1a09d0
AD
1031matching and related activities.
1032
a0d0e21e
LW
1033=over 8
1034
1035=item ?PATTERN?
d74e8afc 1036X<?>
a0d0e21e
LW
1037
1038This is just like the C</pattern/> search, except that it matches only
1039once between calls to the reset() operator. This is a useful
5f05dabc 1040optimization when you want to see only the first occurrence of
a0d0e21e
LW
1041something in each file of a set of files, for instance. Only C<??>
1042patterns local to the current package are reset.
1043
5a964f20
TC
1044 while (<>) {
1045 if (?^$?) {
1046 # blank line between header and body
1047 }
1048 } continue {
1049 reset if eof; # clear ?? status for next file
1050 }
1051
483b4840 1052This usage is vaguely deprecated, which means it just might possibly
19799a22
GS
1053be removed in some distant future version of Perl, perhaps somewhere
1054around the year 2168.
a0d0e21e 1055
fb73857a 1056=item m/PATTERN/cgimosx
89d205f2
YO
1057X<m> X<operator, match>
1058X<regexp, options> X<regexp> X<regex, options> X<regex>
d74e8afc 1059X</c> X</i> X</m> X</o> X</s> X</x>
a0d0e21e 1060
fb73857a 1061=item /PATTERN/cgimosx
a0d0e21e 1062
5a964f20 1063Searches a string for a pattern match, and in scalar context returns
19799a22
GS
1064true if it succeeds, false if it fails. If no string is specified
1065via the C<=~> or C<!~> operator, the $_ string is searched. (The
1066string specified with C<=~> need not be an lvalue--it may be the
1067result of an expression evaluation, but remember the C<=~> binds
1068rather tightly.) See also L<perlre>. See L<perllocale> for
1069discussion of additional considerations that apply when C<use locale>
1070is in effect.
a0d0e21e
LW
1071
1072Options are:
1073
fb73857a 1074 c Do not reset search position on a failed match when /g is in effect.
5f05dabc 1075 g Match globally, i.e., find all occurrences.
a0d0e21e
LW
1076 i Do case-insensitive pattern matching.
1077 m Treat string as multiple lines.
5f05dabc 1078 o Compile pattern only once.
a0d0e21e
LW
1079 s Treat string as single line.
1080 x Use extended regular expressions.
1081
1082If "/" is the delimiter then the initial C<m> is optional. With the C<m>
89d205f2 1083you can use any pair of non-alphanumeric, non-whitespace characters
19799a22
GS
1084as delimiters. This is particularly useful for matching path names
1085that contain "/", to avoid LTS (leaning toothpick syndrome). If "?" is
7bac28a0 1086the delimiter, then the match-only-once rule of C<?PATTERN?> applies.
19799a22 1087If "'" is the delimiter, no interpolation is performed on the PATTERN.
a0d0e21e
LW
1088
1089PATTERN may contain variables, which will be interpolated (and the
f70b4f9c 1090pattern recompiled) every time the pattern search is evaluated, except
1f247705
GS
1091for when the delimiter is a single quote. (Note that C<$(>, C<$)>, and
1092C<$|> are not interpolated because they look like end-of-string tests.)
f70b4f9c
AB
1093If you want such a pattern to be compiled only once, add a C</o> after
1094the trailing delimiter. This avoids expensive run-time recompilations,
1095and is useful when the value you are interpolating won't change over
1096the life of the script. However, mentioning C</o> constitutes a promise
1097that you won't change the variables in the pattern. If you change them,
13a2d996 1098Perl won't even notice. See also L<"qr/STRING/imosx">.
a0d0e21e 1099
5a964f20 1100If the PATTERN evaluates to the empty string, the last
d65afb4b
HS
1101I<successfully> matched regular expression is used instead. In this
1102case, only the C<g> and C<c> flags on the empty pattern is honoured -
1103the other flags are taken from the original pattern. If no match has
1104previously succeeded, this will (silently) act instead as a genuine
1105empty pattern (which will always match).
a0d0e21e 1106
89d205f2
YO
1107Note that it's possible to confuse Perl into thinking C<//> (the empty
1108regex) is really C<//> (the defined-or operator). Perl is usually pretty
1109good about this, but some pathological cases might trigger this, such as
1110C<$a///> (is that C<($a) / (//)> or C<$a // />?) and C<print $fh //>
1111(C<print $fh(//> or C<print($fh //>?). In all of these examples, Perl
1112will assume you meant defined-or. If you meant the empty regex, just
1113use parentheses or spaces to disambiguate, or even prefix the empty
c963b151
BD
1114regex with an C<m> (so C<//> becomes C<m//>).
1115
19799a22 1116If the C</g> option is not used, C<m//> in list context returns a
a0d0e21e 1117list consisting of the subexpressions matched by the parentheses in the
f7e33566
GS
1118pattern, i.e., (C<$1>, C<$2>, C<$3>...). (Note that here C<$1> etc. are
1119also set, and that this differs from Perl 4's behavior.) When there are
1120no parentheses in the pattern, the return value is the list C<(1)> for
1121success. With or without parentheses, an empty list is returned upon
1122failure.
a0d0e21e
LW
1123
1124Examples:
1125
1126 open(TTY, '/dev/tty');
1127 <TTY> =~ /^y/i && foo(); # do foo if desired
1128
1129 if (/Version: *([0-9.]*)/) { $version = $1; }
1130
1131 next if m#^/usr/spool/uucp#;
1132
1133 # poor man's grep
1134 $arg = shift;
1135 while (<>) {
1136 print if /$arg/o; # compile only once
1137 }
1138
1139 if (($F1, $F2, $Etc) = ($foo =~ /^(\S+)\s+(\S+)\s*(.*)/))
1140
1141This last example splits $foo into the first two words and the
5f05dabc
PP
1142remainder of the line, and assigns those three fields to $F1, $F2, and
1143$Etc. The conditional is true if any variables were assigned, i.e., if
a0d0e21e
LW
1144the pattern matched.
1145
19799a22
GS
1146The C</g> modifier specifies global pattern matching--that is,
1147matching as many times as possible within the string. How it behaves
1148depends on the context. In list context, it returns a list of the
1149substrings matched by any capturing parentheses in the regular
1150expression. If there are no parentheses, it returns a list of all
1151the matched strings, as if there were parentheses around the whole
1152pattern.
a0d0e21e 1153
7e86de3e 1154In scalar context, each execution of C<m//g> finds the next match,
19799a22 1155returning true if it matches, and false if there is no further match.
7e86de3e
G
1156The position after the last match can be read or set using the pos()
1157function; see L<perlfunc/pos>. A failed match normally resets the
1158search position to the beginning of the string, but you can avoid that
1159by adding the C</c> modifier (e.g. C<m//gc>). Modifying the target
1160string also resets the search position.
c90c0ff4
PP
1161
1162You can intermix C<m//g> matches with C<m/\G.../g>, where C<\G> is a
1163zero-width assertion that matches the exact position where the previous
5d43e42d
DC
1164C<m//g>, if any, left off. Without the C</g> modifier, the C<\G> assertion
1165still anchors at pos(), but the match is of course only attempted once.
1166Using C<\G> without C</g> on a target string that has not previously had a
1167C</g> match applied to it is the same as using the C<\A> assertion to match
fe4b3f22
RGS
1168the beginning of the string. Note also that, currently, C<\G> is only
1169properly supported when anchored at the very beginning of the pattern.
c90c0ff4
PP
1170
1171Examples:
a0d0e21e
LW
1172
1173 # list context
1174 ($one,$five,$fifteen) = (`uptime` =~ /(\d+\.\d+)/g);
1175
1176 # scalar context
5d43e42d 1177 $/ = "";
19799a22
GS
1178 while (defined($paragraph = <>)) {
1179 while ($paragraph =~ /[a-z]['")]*[.!?]+['")]*\s/g) {
1180 $sentences++;
a0d0e21e
LW
1181 }
1182 }
1183 print "$sentences\n";
1184
c90c0ff4 1185 # using m//gc with \G
137443ea 1186 $_ = "ppooqppqq";
44a8e56a
PP
1187 while ($i++ < 2) {
1188 print "1: '";
c90c0ff4 1189 print $1 while /(o)/gc; print "', pos=", pos, "\n";
44a8e56a 1190 print "2: '";
c90c0ff4 1191 print $1 if /\G(q)/gc; print "', pos=", pos, "\n";
44a8e56a 1192 print "3: '";
c90c0ff4 1193 print $1 while /(p)/gc; print "', pos=", pos, "\n";
44a8e56a 1194 }
5d43e42d 1195 print "Final: '$1', pos=",pos,"\n" if /\G(.)/;
44a8e56a
PP
1196
1197The last example should print:
1198
1199 1: 'oo', pos=4
137443ea 1200 2: 'q', pos=5
44a8e56a
PP
1201 3: 'pp', pos=7
1202 1: '', pos=7
137443ea
PP
1203 2: 'q', pos=8
1204 3: '', pos=8
5d43e42d
DC
1205 Final: 'q', pos=8
1206
1207Notice that the final match matched C<q> instead of C<p>, which a match
1208without the C<\G> anchor would have done. Also note that the final match
1209did not update C<pos> -- C<pos> is only updated on a C</g> match. If the
1210final match did indeed match C<p>, it's a good bet that you're running an
1211older (pre-5.6.0) Perl.
44a8e56a 1212
c90c0ff4 1213A useful idiom for C<lex>-like scanners is C</\G.../gc>. You can
e7ea3e70 1214combine several regexps like this to process a string part-by-part,
c90c0ff4
PP
1215doing different actions depending on which regexp matched. Each
1216regexp tries to match where the previous one leaves off.
e7ea3e70 1217
3fe9a6f1 1218 $_ = <<'EOL';
e7ea3e70 1219 $url = new URI::URL "http://www/"; die if $url eq "xXx";
3fe9a6f1
PP
1220 EOL
1221 LOOP:
e7ea3e70 1222 {
c90c0ff4
PP
1223 print(" digits"), redo LOOP if /\G\d+\b[,.;]?\s*/gc;
1224 print(" lowercase"), redo LOOP if /\G[a-z]+\b[,.;]?\s*/gc;
1225 print(" UPPERCASE"), redo LOOP if /\G[A-Z]+\b[,.;]?\s*/gc;
1226 print(" Capitalized"), redo LOOP if /\G[A-Z][a-z]+\b[,.;]?\s*/gc;
1227 print(" MiXeD"), redo LOOP if /\G[A-Za-z]+\b[,.;]?\s*/gc;
1228 print(" alphanumeric"), redo LOOP if /\G[A-Za-z0-9]+\b[,.;]?\s*/gc;
1229 print(" line-noise"), redo LOOP if /\G[^A-Za-z0-9]+/gc;
e7ea3e70
IZ
1230 print ". That's all!\n";
1231 }
1232
1233Here is the output (split into several lines):
1234
1235 line-noise lowercase line-noise lowercase UPPERCASE line-noise
1236 UPPERCASE line-noise lowercase line-noise lowercase line-noise
1237 lowercase lowercase line-noise lowercase lowercase line-noise
1238 MiXeD line-noise. That's all!
44a8e56a 1239
a0d0e21e 1240=item q/STRING/
5d44bfff 1241X<q> X<quote, single> X<'> X<''>
a0d0e21e 1242
5d44bfff 1243=item 'STRING'
a0d0e21e 1244
19799a22 1245A single-quoted, literal string. A backslash represents a backslash
68dc0745
PP
1246unless followed by the delimiter or another backslash, in which case
1247the delimiter or backslash is interpolated.
a0d0e21e
LW
1248
1249 $foo = q!I said, "You said, 'She said it.'"!;
1250 $bar = q('This is it.');
68dc0745 1251 $baz = '\n'; # a two-character string
a0d0e21e
LW
1252
1253=item qq/STRING/
d74e8afc 1254X<qq> X<quote, double> X<"> X<"">
a0d0e21e
LW
1255
1256=item "STRING"
1257
1258A double-quoted, interpolated string.
1259
1260 $_ .= qq
1261 (*** The previous line contains the naughty word "$1".\n)
19799a22 1262 if /\b(tcl|java|python)\b/i; # :-)
68dc0745 1263 $baz = "\n"; # a one-character string
a0d0e21e 1264
eec2d3df 1265=item qr/STRING/imosx
d74e8afc 1266X<qr> X</i> X</m> X</o> X</s> X</x>
eec2d3df 1267
322edccd 1268This operator quotes (and possibly compiles) its I<STRING> as a regular
19799a22
GS
1269expression. I<STRING> is interpolated the same way as I<PATTERN>
1270in C<m/PATTERN/>. If "'" is used as the delimiter, no interpolation
1271is done. Returns a Perl value which may be used instead of the
1272corresponding C</STRING/imosx> expression.
4b6a7270
IZ
1273
1274For example,
1275
1276 $rex = qr/my.STRING/is;
1277 s/$rex/foo/;
1278
1279is equivalent to
1280
1281 s/my.STRING/foo/is;
1282
1283The result may be used as a subpattern in a match:
eec2d3df
GS
1284
1285 $re = qr/$pattern/;
0a92e3a8
GS
1286 $string =~ /foo${re}bar/; # can be interpolated in other patterns
1287 $string =~ $re; # or used standalone
4b6a7270
IZ
1288 $string =~ /$re/; # or this way
1289
1290Since Perl may compile the pattern at the moment of execution of qr()
19799a22 1291operator, using qr() may have speed advantages in some situations,
4b6a7270
IZ
1292notably if the result of qr() is used standalone:
1293
1294 sub match {
1295 my $patterns = shift;
1296 my @compiled = map qr/$_/i, @$patterns;
1297 grep {
1298 my $success = 0;
a7665c5e 1299 foreach my $pat (@compiled) {
4b6a7270
IZ
1300 $success = 1, last if /$pat/;
1301 }
1302 $success;
1303 } @_;
1304 }
1305
19799a22
GS
1306Precompilation of the pattern into an internal representation at
1307the moment of qr() avoids a need to recompile the pattern every
1308time a match C</$pat/> is attempted. (Perl has many other internal
1309optimizations, but none would be triggered in the above example if
1310we did not use qr() operator.)
eec2d3df
GS
1311
1312Options are:
1313
1314 i Do case-insensitive pattern matching.
1315 m Treat string as multiple lines.
1316 o Compile pattern only once.
1317 s Treat string as single line.
1318 x Use extended regular expressions.
1319
0a92e3a8
GS
1320See L<perlre> for additional information on valid syntax for STRING, and
1321for a detailed look at the semantics of regular expressions.
1322
a0d0e21e 1323=item qx/STRING/
d74e8afc 1324X<qx> X<`> X<``> X<backtick>
a0d0e21e
LW
1325
1326=item `STRING`
1327
43dd4d21
JH
1328A string which is (possibly) interpolated and then executed as a
1329system command with C</bin/sh> or its equivalent. Shell wildcards,
1330pipes, and redirections will be honored. The collected standard
1331output of the command is returned; standard error is unaffected. In
1332scalar context, it comes back as a single (potentially multi-line)
1333string, or undef if the command failed. In list context, returns a
1334list of lines (however you've defined lines with $/ or
1335$INPUT_RECORD_SEPARATOR), or an empty list if the command failed.
5a964f20
TC
1336
1337Because backticks do not affect standard error, use shell file descriptor
1338syntax (assuming the shell supports this) if you care to address this.
1339To capture a command's STDERR and STDOUT together:
a0d0e21e 1340
5a964f20
TC
1341 $output = `cmd 2>&1`;
1342
1343To capture a command's STDOUT but discard its STDERR:
1344
1345 $output = `cmd 2>/dev/null`;
1346
1347To capture a command's STDERR but discard its STDOUT (ordering is
1348important here):
1349
1350 $output = `cmd 2>&1 1>/dev/null`;
1351
1352To exchange a command's STDOUT and STDERR in order to capture the STDERR
1353but leave its STDOUT to come out the old STDERR:
1354
1355 $output = `cmd 3>&1 1>&2 2>&3 3>&-`;
1356
1357To read both a command's STDOUT and its STDERR separately, it's easiest
2359510d
SD
1358to redirect them separately to files, and then read from those files
1359when the program is done:
5a964f20 1360
2359510d 1361 system("program args 1>program.stdout 2>program.stderr");
5a964f20 1362
30398227
SP
1363The STDIN filehandle used by the command is inherited from Perl's STDIN.
1364For example:
1365
1366 open BLAM, "blam" || die "Can't open: $!";
1367 open STDIN, "<&BLAM";
1368 print `sort`;
1369
1370will print the sorted contents of the file "blam".
1371
5a964f20
TC
1372Using single-quote as a delimiter protects the command from Perl's
1373double-quote interpolation, passing it on to the shell instead:
1374
1375 $perl_info = qx(ps $$); # that's Perl's $$
1376 $shell_info = qx'ps $$'; # that's the new shell's $$
1377
19799a22 1378How that string gets evaluated is entirely subject to the command
5a964f20
TC
1379interpreter on your system. On most platforms, you will have to protect
1380shell metacharacters if you want them treated literally. This is in
1381practice difficult to do, as it's unclear how to escape which characters.
1382See L<perlsec> for a clean and safe example of a manual fork() and exec()
1383to emulate backticks safely.
a0d0e21e 1384
bb32b41a
GS
1385On some platforms (notably DOS-like ones), the shell may not be
1386capable of dealing with multiline commands, so putting newlines in
1387the string may not get you what you want. You may be able to evaluate
1388multiple commands in a single line by separating them with the command
1389separator character, if your shell supports that (e.g. C<;> on many Unix
1390shells; C<&> on the Windows NT C<cmd> shell).
1391
0f897271
GS
1392Beginning with v5.6.0, Perl will attempt to flush all files opened for
1393output before starting the child process, but this may not be supported
1394on some platforms (see L<perlport>). To be safe, you may need to set
1395C<$|> ($AUTOFLUSH in English) or call the C<autoflush()> method of
1396C<IO::Handle> on any open handles.
1397
bb32b41a
GS
1398Beware that some command shells may place restrictions on the length
1399of the command line. You must ensure your strings don't exceed this
1400limit after any necessary interpolations. See the platform-specific
1401release notes for more details about your particular environment.
1402
5a964f20
TC
1403Using this operator can lead to programs that are difficult to port,
1404because the shell commands called vary between systems, and may in
1405fact not be present at all. As one example, the C<type> command under
1406the POSIX shell is very different from the C<type> command under DOS.
1407That doesn't mean you should go out of your way to avoid backticks
1408when they're the right way to get something done. Perl was made to be
1409a glue language, and one of the things it glues together is commands.
1410Just understand what you're getting yourself into.
bb32b41a 1411
da87341d 1412See L</"I/O Operators"> for more discussion.
a0d0e21e 1413
945c54fd 1414=item qw/STRING/
d74e8afc 1415X<qw> X<quote, list> X<quote, words>
945c54fd
JH
1416
1417Evaluates to a list of the words extracted out of STRING, using embedded
1418whitespace as the word delimiters. It can be understood as being roughly
1419equivalent to:
1420
1421 split(' ', q/STRING/);
1422
efb1e162
CW
1423the differences being that it generates a real list at compile time, and
1424in scalar context it returns the last element in the list. So
945c54fd
JH
1425this expression:
1426
1427 qw(foo bar baz)
1428
1429is semantically equivalent to the list:
1430
1431 'foo', 'bar', 'baz'
1432
1433Some frequently seen examples:
1434
1435 use POSIX qw( setlocale localeconv )
1436 @EXPORT = qw( foo bar baz );
1437
1438A common mistake is to try to separate the words with comma or to
1439put comments into a multi-line C<qw>-string. For this reason, the
89d205f2 1440C<use warnings> pragma and the B<-w> switch (that is, the C<$^W> variable)
945c54fd
JH
1441produces warnings if the STRING contains the "," or the "#" character.
1442
a0d0e21e 1443=item s/PATTERN/REPLACEMENT/egimosx
d74e8afc
ITB
1444X<substitute> X<substitution> X<replace> X<regexp, replace>
1445X<regexp, substitute> X</e> X</g> X</i> X</m> X</o> X</s> X</x>
a0d0e21e
LW
1446
1447Searches a string for a pattern, and if found, replaces that pattern
1448with the replacement text and returns the number of substitutions
e37d713d 1449made. Otherwise it returns false (specifically, the empty string).
a0d0e21e
LW
1450
1451If no string is specified via the C<=~> or C<!~> operator, the C<$_>
1452variable is searched and modified. (The string specified with C<=~> must
5a964f20 1453be scalar variable, an array element, a hash element, or an assignment
5f05dabc 1454to one of those, i.e., an lvalue.)
a0d0e21e 1455
19799a22 1456If the delimiter chosen is a single quote, no interpolation is
a0d0e21e
LW
1457done on either the PATTERN or the REPLACEMENT. Otherwise, if the
1458PATTERN contains a $ that looks like a variable rather than an
1459end-of-string test, the variable will be interpolated into the pattern
5f05dabc 1460at run-time. If you want the pattern compiled only once the first time
a0d0e21e 1461the variable is interpolated, use the C</o> option. If the pattern
5a964f20 1462evaluates to the empty string, the last successfully executed regular
a0d0e21e 1463expression is used instead. See L<perlre> for further explanation on these.
5a964f20 1464See L<perllocale> for discussion of additional considerations that apply
a034a98d 1465when C<use locale> is in effect.
a0d0e21e
LW
1466
1467Options are:
1468
1469 e Evaluate the right side as an expression.
5f05dabc 1470 g Replace globally, i.e., all occurrences.
a0d0e21e
LW
1471 i Do case-insensitive pattern matching.
1472 m Treat string as multiple lines.
5f05dabc 1473 o Compile pattern only once.
a0d0e21e
LW
1474 s Treat string as single line.
1475 x Use extended regular expressions.
1476
1477Any non-alphanumeric, non-whitespace delimiter may replace the
1478slashes. If single quotes are used, no interpretation is done on the
e37d713d 1479replacement string (the C</e> modifier overrides this, however). Unlike
54310121 1480Perl 4, Perl 5 treats backticks as normal delimiters; the replacement
e37d713d 1481text is not evaluated as a command. If the
a0d0e21e 1482PATTERN is delimited by bracketing quotes, the REPLACEMENT has its own
5f05dabc 1483pair of quotes, which may or may not be bracketing quotes, e.g.,
35f2feb0 1484C<s(foo)(bar)> or C<< s<foo>/bar/ >>. A C</e> will cause the
cec88af6
GS
1485replacement portion to be treated as a full-fledged Perl expression
1486and evaluated right then and there. It is, however, syntax checked at
1487compile-time. A second C<e> modifier will cause the replacement portion
1488to be C<eval>ed before being run as a Perl expression.
a0d0e21e
LW
1489
1490Examples:
1491
1492 s/\bgreen\b/mauve/g; # don't change wintergreen
1493
1494 $path =~ s|/usr/bin|/usr/local/bin|;
1495
1496 s/Login: $foo/Login: $bar/; # run-time pattern
1497
5a964f20 1498 ($foo = $bar) =~ s/this/that/; # copy first, then change
a0d0e21e 1499
5a964f20 1500 $count = ($paragraph =~ s/Mister\b/Mr./g); # get change-count
a0d0e21e
LW
1501
1502 $_ = 'abc123xyz';
1503 s/\d+/$&*2/e; # yields 'abc246xyz'
1504 s/\d+/sprintf("%5d",$&)/e; # yields 'abc 246xyz'
1505 s/\w/$& x 2/eg; # yields 'aabbcc 224466xxyyzz'
1506
1507 s/%(.)/$percent{$1}/g; # change percent escapes; no /e
1508 s/%(.)/$percent{$1} || $&/ge; # expr now, so /e
1509 s/^=(\w+)/&pod($1)/ge; # use function call
1510
5a964f20
TC
1511 # expand variables in $_, but dynamics only, using
1512 # symbolic dereferencing
1513 s/\$(\w+)/${$1}/g;
1514
cec88af6
GS
1515 # Add one to the value of any numbers in the string
1516 s/(\d+)/1 + $1/eg;
1517
1518 # This will expand any embedded scalar variable
1519 # (including lexicals) in $_ : First $1 is interpolated
1520 # to the variable name, and then evaluated
a0d0e21e
LW
1521 s/(\$\w+)/$1/eeg;
1522
5a964f20 1523 # Delete (most) C comments.
a0d0e21e 1524 $program =~ s {
4633a7c4
LW
1525 /\* # Match the opening delimiter.
1526 .*? # Match a minimal number of characters.
1527 \*/ # Match the closing delimiter.
a0d0e21e
LW
1528 } []gsx;
1529
6b0ac556 1530 s/^\s*(.*?)\s*$/$1/; # trim whitespace in $_, expensively
5a964f20 1531
6b0ac556 1532 for ($variable) { # trim whitespace in $variable, cheap
5a964f20
TC
1533 s/^\s+//;
1534 s/\s+$//;
1535 }
a0d0e21e
LW
1536
1537 s/([^ ]*) *([^ ]*)/$2 $1/; # reverse 1st two fields
1538
54310121 1539Note the use of $ instead of \ in the last example. Unlike
35f2feb0
GS
1540B<sed>, we use the \<I<digit>> form in only the left hand side.
1541Anywhere else it's $<I<digit>>.
a0d0e21e 1542
5f05dabc 1543Occasionally, you can't use just a C</g> to get all the changes
19799a22 1544to occur that you might want. Here are two common cases:
a0d0e21e
LW
1545
1546 # put commas in the right places in an integer
89d205f2 1547 1 while s/(\d)(\d\d\d)(?!\d)/$1,$2/g;
a0d0e21e
LW
1548
1549 # expand tabs to 8-column spacing
1550 1 while s/\t+/' ' x (length($&)*8 - length($`)%8)/e;
1551
6940069f 1552=item tr/SEARCHLIST/REPLACEMENTLIST/cds
d74e8afc 1553X<tr> X<y> X<transliterate> X</c> X</d> X</s>
a0d0e21e 1554
6940069f 1555=item y/SEARCHLIST/REPLACEMENTLIST/cds
a0d0e21e 1556
2c268ad5 1557Transliterates all occurrences of the characters found in the search list
a0d0e21e
LW
1558with the corresponding character in the replacement list. It returns
1559the number of characters replaced or deleted. If no string is
2c268ad5 1560specified via the =~ or !~ operator, the $_ string is transliterated. (The
54310121
PP
1561string specified with =~ must be a scalar variable, an array element, a
1562hash element, or an assignment to one of those, i.e., an lvalue.)
8ada0baa 1563
89d205f2 1564A character range may be specified with a hyphen, so C<tr/A-J/0-9/>
2c268ad5 1565does the same replacement as C<tr/ACEGIBDFHJ/0246813579/>.
54310121
PP
1566For B<sed> devotees, C<y> is provided as a synonym for C<tr>. If the
1567SEARCHLIST is delimited by bracketing quotes, the REPLACEMENTLIST has
1568its own pair of quotes, which may or may not be bracketing quotes,
2c268ad5 1569e.g., C<tr[A-Z][a-z]> or C<tr(+\-*/)/ABCD/>.
a0d0e21e 1570
cc255d5f
JH
1571Note that C<tr> does B<not> do regular expression character classes
1572such as C<\d> or C<[:lower:]>. The <tr> operator is not equivalent to
1573the tr(1) utility. If you want to map strings between lower/upper
1574cases, see L<perlfunc/lc> and L<perlfunc/uc>, and in general consider
1575using the C<s> operator if you need regular expressions.
1576
8ada0baa
JH
1577Note also that the whole range idea is rather unportable between
1578character sets--and even within character sets they may cause results
1579you probably didn't expect. A sound principle is to use only ranges
1580that begin from and end at either alphabets of equal case (a-e, A-E),
1581or digits (0-4). Anything else is unsafe. If in doubt, spell out the
1582character sets in full.
1583
a0d0e21e
LW
1584Options:
1585
1586 c Complement the SEARCHLIST.
1587 d Delete found but unreplaced characters.
1588 s Squash duplicate replaced characters.
1589
19799a22
GS
1590If the C</c> modifier is specified, the SEARCHLIST character set
1591is complemented. If the C</d> modifier is specified, any characters
1592specified by SEARCHLIST not found in REPLACEMENTLIST are deleted.
1593(Note that this is slightly more flexible than the behavior of some
1594B<tr> programs, which delete anything they find in the SEARCHLIST,
1595period.) If the C</s> modifier is specified, sequences of characters
1596that were transliterated to the same character are squashed down
1597to a single instance of the character.
a0d0e21e
LW
1598
1599If the C</d> modifier is used, the REPLACEMENTLIST is always interpreted
1600exactly as specified. Otherwise, if the REPLACEMENTLIST is shorter
1601than the SEARCHLIST, the final character is replicated till it is long
5a964f20 1602enough. If the REPLACEMENTLIST is empty, the SEARCHLIST is replicated.
a0d0e21e
LW
1603This latter is useful for counting characters in a class or for
1604squashing character sequences in a class.
1605
1606Examples:
1607
1608 $ARGV[1] =~ tr/A-Z/a-z/; # canonicalize to lower case
1609
1610 $cnt = tr/*/*/; # count the stars in $_
1611
1612 $cnt = $sky =~ tr/*/*/; # count the stars in $sky
1613
1614 $cnt = tr/0-9//; # count the digits in $_
1615
1616 tr/a-zA-Z//s; # bookkeeper -> bokeper
1617
1618 ($HOST = $host) =~ tr/a-z/A-Z/;
1619
1620 tr/a-zA-Z/ /cs; # change non-alphas to single space
1621
1622 tr [\200-\377]
1623 [\000-\177]; # delete 8th bit
1624
19799a22
GS
1625If multiple transliterations are given for a character, only the
1626first one is used:
748a9306
LW
1627
1628 tr/AAA/XYZ/
1629
2c268ad5 1630will transliterate any A to X.
748a9306 1631
19799a22 1632Because the transliteration table is built at compile time, neither
a0d0e21e 1633the SEARCHLIST nor the REPLACEMENTLIST are subjected to double quote
19799a22
GS
1634interpolation. That means that if you want to use variables, you
1635must use an eval():
a0d0e21e
LW
1636
1637 eval "tr/$oldlist/$newlist/";
1638 die $@ if $@;
1639
1640 eval "tr/$oldlist/$newlist/, 1" or die $@;
1641
7e3b091d 1642=item <<EOF
d74e8afc 1643X<here-doc> X<heredoc> X<here-document> X<<< << >>>
7e3b091d
DA
1644
1645A line-oriented form of quoting is based on the shell "here-document"
1646syntax. Following a C<< << >> you specify a string to terminate
1647the quoted material, and all lines following the current line down to
89d205f2
YO
1648the terminating string are the value of the item.
1649
1650The terminating string may be either an identifier (a word), or some
1651quoted text. An unquoted identifier works like double quotes.
1652There may not be a space between the C<< << >> and the identifier,
1653unless the identifier is explicitly quoted. (If you put a space it
1654will be treated as a null identifier, which is valid, and matches the
1655first empty line.) The terminating string must appear by itself
1656(unquoted and with no surrounding whitespace) on the terminating line.
1657
1658If the terminating string is quoted, the type of quotes used determine
1659the treatment of the text.
1660
1661=over 4
1662
1663=item Double Quotes
1664
1665Double quotes indicate that the text will be interpolated using exactly
1666the same rules as normal double quoted strings.
7e3b091d
DA
1667
1668 print <<EOF;
1669 The price is $Price.
1670 EOF
1671
1672 print << "EOF"; # same as above
1673 The price is $Price.
1674 EOF
1675
89d205f2
YO
1676
1677=item Single Quotes
1678
1679Single quotes indicate the text is to be treated literally with no
1680interpolation of its content. This is similar to single quoted
1681strings except that backslashes have no special meaning, with C<\\>
1682being treated as two backslashes and not one as they would in every
1683other quoting construct.
1684
1685This is the only form of quoting in perl where there is no need
1686to worry about escaping content, something that code generators
1687can and do make good use of.
1688
1689=item Backticks
1690
1691The content of the here doc is treated just as it would be if the
1692string were embedded in backticks. Thus the content is interpolated
1693as though it were double quoted and then executed via the shell, with
1694the results of the execution returned.
1695
1696 print << `EOC`; # execute command and get results
7e3b091d 1697 echo hi there
7e3b091d
DA
1698 EOC
1699
89d205f2
YO
1700=back
1701
1702It is possible to stack multiple here-docs in a row:
1703
7e3b091d
DA
1704 print <<"foo", <<"bar"; # you can stack them
1705 I said foo.
1706 foo
1707 I said bar.
1708 bar
1709
1710 myfunc(<< "THIS", 23, <<'THAT');
1711 Here's a line
1712 or two.
1713 THIS
1714 and here's another.
1715 THAT
1716
1717Just don't forget that you have to put a semicolon on the end
1718to finish the statement, as Perl doesn't know you're not going to
1719try to do this:
1720
1721 print <<ABC
1722 179231
1723 ABC
1724 + 20;
1725
872d7e53
ST
1726If you want to remove the line terminator from your here-docs,
1727use C<chomp()>.
1728
1729 chomp($string = <<'END');
1730 This is a string.
1731 END
1732
1733If you want your here-docs to be indented with the rest of the code,
1734you'll need to remove leading whitespace from each line manually:
7e3b091d
DA
1735
1736 ($quote = <<'FINIS') =~ s/^\s+//gm;
89d205f2 1737 The Road goes ever on and on,
7e3b091d
DA
1738 down from the door where it began.
1739 FINIS
1740
1741If you use a here-doc within a delimited construct, such as in C<s///eg>,
1742the quoted material must come on the lines following the final delimiter.
1743So instead of
1744
1745 s/this/<<E . 'that'
1746 the other
1747 E
1748 . 'more '/eg;
1749
1750you have to write
1751
89d205f2
YO
1752 s/this/<<E . 'that'
1753 . 'more '/eg;
1754 the other
1755 E
7e3b091d
DA
1756
1757If the terminating identifier is on the last line of the program, you
1758must be sure there is a newline after it; otherwise, Perl will give the
1759warning B<Can't find string terminator "END" anywhere before EOF...>.
1760
89d205f2
YO
1761Additionally, the quoting rules for the end of string identifier are not
1762related to Perl's quoting rules -- C<q()>, C<qq()>, and the like are not
1763supported in place of C<''> and C<"">, and the only interpolation is for
1764backslashing the quoting character:
7e3b091d
DA
1765
1766 print << "abc\"def";
1767 testing...
1768 abc"def
1769
1770Finally, quoted strings cannot span multiple lines. The general rule is
1771that the identifier must be a string literal. Stick with that, and you
1772should be safe.
1773
a0d0e21e
LW
1774=back
1775
75e14d17 1776=head2 Gory details of parsing quoted constructs
d74e8afc 1777X<quote, gory details>
75e14d17 1778
19799a22
GS
1779When presented with something that might have several different
1780interpretations, Perl uses the B<DWIM> (that's "Do What I Mean")
1781principle to pick the most probable interpretation. This strategy
1782is so successful that Perl programmers often do not suspect the
1783ambivalence of what they write. But from time to time, Perl's
1784notions differ substantially from what the author honestly meant.
1785
1786This section hopes to clarify how Perl handles quoted constructs.
1787Although the most common reason to learn this is to unravel labyrinthine
1788regular expressions, because the initial steps of parsing are the
1789same for all quoting operators, they are all discussed together.
1790
1791The most important Perl parsing rule is the first one discussed
1792below: when processing a quoted construct, Perl first finds the end
1793of that construct, then interprets its contents. If you understand
1794this rule, you may skip the rest of this section on the first
1795reading. The other rules are likely to contradict the user's
1796expectations much less frequently than this first one.
1797
1798Some passes discussed below are performed concurrently, but because
1799their results are the same, we consider them individually. For different
1800quoting constructs, Perl performs different numbers of passes, from
1801one to five, but these passes are always performed in the same order.
75e14d17 1802
13a2d996 1803=over 4
75e14d17
IZ
1804
1805=item Finding the end
1806
19799a22 1807The first pass is finding the end of the quoted construct, whether
872d7e53 1808it be a multicharacter delimiter C<"EOF\n"> in the C<<<EOF>
19799a22 1809construct, a C</> that terminates a C<qq//> construct, a C<]> which
35f2feb0
GS
1810terminates C<qq[]> construct, or a C<< > >> which terminates a
1811fileglob started with C<< < >>.
75e14d17 1812
19799a22
GS
1813When searching for single-character non-pairing delimiters, such
1814as C</>, combinations of C<\\> and C<\/> are skipped. However,
1815when searching for single-character pairing delimiter like C<[>,
1816combinations of C<\\>, C<\]>, and C<\[> are all skipped, and nested
1817C<[>, C<]> are skipped as well. When searching for multicharacter
872d7e53
ST
1818delimiters like C<"EOF\n">, nothing is skipped, though the delimiter
1819must start from the first column of the terminating line.
75e14d17 1820
19799a22
GS
1821For constructs with three-part delimiters (C<s///>, C<y///>, and
1822C<tr///>), the search is repeated once more.
75e14d17 1823
19799a22
GS
1824During this search no attention is paid to the semantics of the construct.
1825Thus:
75e14d17
IZ
1826
1827 "$hash{"$foo/$bar"}"
1828
2a94b7ce 1829or:
75e14d17 1830
89d205f2 1831 m/
2a94b7ce 1832 bar # NOT a comment, this slash / terminated m//!
75e14d17
IZ
1833 /x
1834
19799a22
GS
1835do not form legal quoted expressions. The quoted part ends on the
1836first C<"> and C</>, and the rest happens to be a syntax error.
1837Because the slash that terminated C<m//> was followed by a C<SPACE>,
1838the example above is not C<m//x>, but rather C<m//> with no C</x>
1839modifier. So the embedded C<#> is interpreted as a literal C<#>.
75e14d17 1840
89d205f2
YO
1841Also no attention is paid to C<\c\> (multichar control char syntax) during
1842this search. Thus the second C<\> in C<qq/\c\/> is interpreted as a part
1843of C<\/>, and the following C</> is not recognized as a delimiter.
0d594e51
ST
1844Instead, use C<\034> or C<\x1c> at the end of quoted constructs.
1845
75e14d17
IZ
1846=item Removal of backslashes before delimiters
1847
19799a22
GS
1848During the second pass, text between the starting and ending
1849delimiters is copied to a safe location, and the C<\> is removed
1850from combinations consisting of C<\> and delimiter--or delimiters,
89d205f2
YO
1851meaning both starting and ending delimiters will be handled,
1852should these differ. This removal does not happen for multi-character
1853delimiters. Note that the combination C<\\> is left intact.
75e14d17 1854
19799a22
GS
1855Starting from this step no information about the delimiters is
1856used in parsing.
75e14d17
IZ
1857
1858=item Interpolation
d74e8afc 1859X<interpolation>
75e14d17 1860
19799a22 1861The next step is interpolation in the text obtained, which is now
89d205f2 1862delimiter-independent. There are multiple cases.
75e14d17 1863
13a2d996 1864=over 4
75e14d17 1865
89d205f2 1866=item C<<<'EOF'>
75e14d17
IZ
1867
1868No interpolation is performed.
1869
89d205f2
YO
1870=item C<m''>, C<s'''>
1871
1872No interpolation is performed at this stage, see
1873L</"Interpolation of regular expressions"> for comments on later
1874processing of their contents.
1875
75e14d17
IZ
1876=item C<''>, C<q//>
1877
89d205f2
YO
1878The only interpolation is removal of C<\> from pairs of C<\\>.
1879
1880=item C<tr///>, C<y///>
1881
1882No variable interpolation occurs. Escape sequences such as \200
1883and the common escapes such as \t for tab are converted to literals.
1884The character C<-> is treated specially and therefore C<\-> is treated
1885as a literal C<->.
75e14d17 1886
89d205f2 1887=item C<"">, C<``>, C<qq//>, C<qx//>, C<< <file*glob> >>, C<<<"EOF">
75e14d17 1888
19799a22
GS
1889C<\Q>, C<\U>, C<\u>, C<\L>, C<\l> (possibly paired with C<\E>) are
1890converted to corresponding Perl constructs. Thus, C<"$foo\Qbaz$bar">
1891is converted to C<$foo . (quotemeta("baz" . $bar))> internally.
1892The other combinations are replaced with appropriate expansions.
2a94b7ce 1893
19799a22
GS
1894Let it be stressed that I<whatever falls between C<\Q> and C<\E>>
1895is interpolated in the usual way. Something like C<"\Q\\E"> has
1896no C<\E> inside. instead, it has C<\Q>, C<\\>, and C<E>, so the
1897result is the same as for C<"\\\\E">. As a general rule, backslashes
1898between C<\Q> and C<\E> may lead to counterintuitive results. So,
1899C<"\Q\t\E"> is converted to C<quotemeta("\t")>, which is the same
1900as C<"\\\t"> (since TAB is not alphanumeric). Note also that:
2a94b7ce
IZ
1901
1902 $str = '\t';
1903 return "\Q$str";
1904
1905may be closer to the conjectural I<intention> of the writer of C<"\Q\t\E">.
1906
19799a22 1907Interpolated scalars and arrays are converted internally to the C<join> and
92d29cee 1908C<.> catenation operations. Thus, C<"$foo XXX '@arr'"> becomes:
75e14d17 1909
19799a22 1910 $foo . " XXX '" . (join $", @arr) . "'";
75e14d17 1911
19799a22 1912All operations above are performed simultaneously, left to right.
75e14d17 1913
19799a22
GS
1914Because the result of C<"\Q STRING \E"> has all metacharacters
1915quoted, there is no way to insert a literal C<$> or C<@> inside a
1916C<\Q\E> pair. If protected by C<\>, C<$> will be quoted to became
1917C<"\\\$">; if not, it is interpreted as the start of an interpolated
1918scalar.
75e14d17 1919
19799a22 1920Note also that the interpolation code needs to make a decision on
89d205f2 1921where the interpolated scalar ends. For instance, whether
35f2feb0 1922C<< "a $b -> {c}" >> really means:
75e14d17
IZ
1923
1924 "a " . $b . " -> {c}";
1925
2a94b7ce 1926or:
75e14d17
IZ
1927
1928 "a " . $b -> {c};
1929
19799a22
GS
1930Most of the time, the longest possible text that does not include
1931spaces between components and which contains matching braces or
1932brackets. because the outcome may be determined by voting based
1933on heuristic estimators, the result is not strictly predictable.
1934Fortunately, it's usually correct for ambiguous cases.
75e14d17 1935
89d205f2 1936=item C<?RE?>, C</RE/>, C<m/RE/>, C<s/RE/foo/>,
75e14d17 1937
19799a22
GS
1938Processing of C<\Q>, C<\U>, C<\u>, C<\L>, C<\l>, and interpolation
1939happens (almost) as with C<qq//> constructs, but the substitution
1940of C<\> followed by RE-special chars (including C<\>) is not
1941performed. Moreover, inside C<(?{BLOCK})>, C<(?# comment )>, and
1942a C<#>-comment in a C<//x>-regular expression, no processing is
1943performed whatsoever. This is the first step at which the presence
1944of the C<//x> modifier is relevant.
1945
1946Interpolation has several quirks: C<$|>, C<$(>, and C<$)> are not
1947interpolated, and constructs C<$var[SOMETHING]> are voted (by several
1948different estimators) to be either an array element or C<$var>
1949followed by an RE alternative. This is where the notation
1950C<${arr[$bar]}> comes handy: C</${arr[0-9]}/> is interpreted as
1951array element C<-9>, not as a regular expression from the variable
1952C<$arr> followed by a digit, which would be the interpretation of
1953C</$arr[0-9]/>. Since voting among different estimators may occur,
1954the result is not predictable.
1955
1956It is at this step that C<\1> is begrudgingly converted to C<$1> in
1957the replacement text of C<s///> to correct the incorrigible
1958I<sed> hackers who haven't picked up the saner idiom yet. A warning
9f1b1f2d
GS
1959is emitted if the C<use warnings> pragma or the B<-w> command-line flag
1960(that is, the C<$^W> variable) was set.
19799a22
GS
1961
1962The lack of processing of C<\\> creates specific restrictions on
1963the post-processed text. If the delimiter is C</>, one cannot get
1964the combination C<\/> into the result of this step. C</> will
1965finish the regular expression, C<\/> will be stripped to C</> on
1966the previous step, and C<\\/> will be left as is. Because C</> is
1967equivalent to C<\/> inside a regular expression, this does not
1968matter unless the delimiter happens to be character special to the
1969RE engine, such as in C<s*foo*bar*>, C<m[foo]>, or C<?foo?>; or an
1970alphanumeric char, as in:
2a94b7ce
IZ
1971
1972 m m ^ a \s* b mmx;
1973
19799a22 1974In the RE above, which is intentionally obfuscated for illustration, the
2a94b7ce 1975delimiter is C<m>, the modifier is C<mx>, and after backslash-removal the
89d205f2 1976RE is the same as for C<m/ ^ a \s* b /mx>. There's more than one
19799a22
GS
1977reason you're encouraged to restrict your delimiters to non-alphanumeric,
1978non-whitespace choices.
75e14d17
IZ
1979
1980=back
1981
19799a22 1982This step is the last one for all constructs except regular expressions,
75e14d17
IZ
1983which are processed further.
1984
1985=item Interpolation of regular expressions
d74e8afc 1986X<regexp, interpolation>
75e14d17 1987
19799a22
GS
1988Previous steps were performed during the compilation of Perl code,
1989but this one happens at run time--although it may be optimized to
1990be calculated at compile time if appropriate. After preprocessing
1991described above, and possibly after evaluation if catenation,
1992joining, casing translation, or metaquoting are involved, the
1993resulting I<string> is passed to the RE engine for compilation.
1994
1995Whatever happens in the RE engine might be better discussed in L<perlre>,
1996but for the sake of continuity, we shall do so here.
1997
1998This is another step where the presence of the C<//x> modifier is
1999relevant. The RE engine scans the string from left to right and
2000converts it to a finite automaton.
2001
2002Backslashed characters are either replaced with corresponding
2003literal strings (as with C<\{>), or else they generate special nodes
2004in the finite automaton (as with C<\b>). Characters special to the
2005RE engine (such as C<|>) generate corresponding nodes or groups of
2006nodes. C<(?#...)> comments are ignored. All the rest is either
2007converted to literal strings to match, or else is ignored (as is
2008whitespace and C<#>-style comments if C<//x> is present).
2009
2010Parsing of the bracketed character class construct, C<[...]>, is
2011rather different than the rule used for the rest of the pattern.
2012The terminator of this construct is found using the same rules as
2013for finding the terminator of a C<{}>-delimited construct, the only
2014exception being that C<]> immediately following C<[> is treated as
2015though preceded by a backslash. Similarly, the terminator of
2016C<(?{...})> is found using the same rules as for finding the
2017terminator of a C<{}>-delimited construct.
2018
2019It is possible to inspect both the string given to RE engine and the
2020resulting finite automaton. See the arguments C<debug>/C<debugcolor>
2021in the C<use L<re>> pragma, as well as Perl's B<-Dr> command-line
4a4eefd0 2022switch documented in L<perlrun/"Command Switches">.
75e14d17
IZ
2023
2024=item Optimization of regular expressions
d74e8afc 2025X<regexp, optimization>
75e14d17 2026
7522fed5 2027This step is listed for completeness only. Since it does not change
75e14d17 2028semantics, details of this step are not documented and are subject
19799a22
GS
2029to change without notice. This step is performed over the finite
2030automaton that was generated during the previous pass.
2a94b7ce 2031
19799a22
GS
2032It is at this stage that C<split()> silently optimizes C</^/> to
2033mean C</^/m>.
75e14d17
IZ
2034
2035=back
2036
a0d0e21e 2037=head2 I/O Operators
d74e8afc
ITB
2038X<operator, i/o> X<operator, io> X<io> X<while> X<filehandle>
2039X<< <> >> X<@ARGV>
a0d0e21e 2040
54310121 2041There are several I/O operators you should know about.
fbad3eb5 2042
7b8d334a 2043A string enclosed by backticks (grave accents) first undergoes
19799a22
GS
2044double-quote interpolation. It is then interpreted as an external
2045command, and the output of that command is the value of the
e9c56f9b
JH
2046backtick string, like in a shell. In scalar context, a single string
2047consisting of all output is returned. In list context, a list of
2048values is returned, one per line of output. (You can set C<$/> to use
2049a different line terminator.) The command is executed each time the
2050pseudo-literal is evaluated. The status value of the command is
2051returned in C<$?> (see L<perlvar> for the interpretation of C<$?>).
2052Unlike in B<csh>, no translation is done on the return data--newlines
2053remain newlines. Unlike in any of the shells, single quotes do not
2054hide variable names in the command from interpretation. To pass a
2055literal dollar-sign through to the shell you need to hide it with a
2056backslash. The generalized form of backticks is C<qx//>. (Because
2057backticks always undergo shell expansion as well, see L<perlsec> for
2058security concerns.)
d74e8afc 2059X<qx> X<`> X<``> X<backtick> X<glob>
19799a22
GS
2060
2061In scalar context, evaluating a filehandle in angle brackets yields
2062the next line from that file (the newline, if any, included), or
2063C<undef> at end-of-file or on error. When C<$/> is set to C<undef>
2064(sometimes known as file-slurp mode) and the file is empty, it
2065returns C<''> the first time, followed by C<undef> subsequently.
2066
2067Ordinarily you must assign the returned value to a variable, but
2068there is one situation where an automatic assignment happens. If
2069and only if the input symbol is the only thing inside the conditional
2070of a C<while> statement (even if disguised as a C<for(;;)> loop),
2071the value is automatically assigned to the global variable $_,
2072destroying whatever was there previously. (This may seem like an
2073odd thing to you, but you'll use the construct in almost every Perl
17b829fa 2074script you write.) The $_ variable is not implicitly localized.
19799a22
GS
2075You'll have to put a C<local $_;> before the loop if you want that
2076to happen.
2077
2078The following lines are equivalent:
a0d0e21e 2079
748a9306 2080 while (defined($_ = <STDIN>)) { print; }
7b8d334a 2081 while ($_ = <STDIN>) { print; }
a0d0e21e
LW
2082 while (<STDIN>) { print; }
2083 for (;<STDIN>;) { print; }
748a9306 2084 print while defined($_ = <STDIN>);
7b8d334a 2085 print while ($_ = <STDIN>);
a0d0e21e
LW
2086 print while <STDIN>;
2087
19799a22 2088This also behaves similarly, but avoids $_ :
7b8d334a 2089
89d205f2 2090 while (my $line = <STDIN>) { print $line }
7b8d334a 2091
19799a22
GS
2092In these loop constructs, the assigned value (whether assignment
2093is automatic or explicit) is then tested to see whether it is
2094defined. The defined test avoids problems where line has a string
2095value that would be treated as false by Perl, for example a "" or
2096a "0" with no trailing newline. If you really mean for such values
2097to terminate the loop, they should be tested for explicitly:
7b8d334a
GS
2098
2099 while (($_ = <STDIN>) ne '0') { ... }
2100 while (<STDIN>) { last unless $_; ... }
2101
35f2feb0 2102In other boolean contexts, C<< <I<filehandle>> >> without an
89d205f2 2103explicit C<defined> test or comparison elicit a warning if the
9f1b1f2d 2104C<use warnings> pragma or the B<-w>
19799a22 2105command-line switch (the C<$^W> variable) is in effect.
7b8d334a 2106
5f05dabc 2107The filehandles STDIN, STDOUT, and STDERR are predefined. (The
19799a22
GS
2108filehandles C<stdin>, C<stdout>, and C<stderr> will also work except
2109in packages, where they would be interpreted as local identifiers
2110rather than global.) Additional filehandles may be created with
2111the open() function, amongst others. See L<perlopentut> and
2112L<perlfunc/open> for details on this.
d74e8afc 2113X<stdin> X<stdout> X<sterr>
a0d0e21e 2114
35f2feb0 2115If a <FILEHANDLE> is used in a context that is looking for
19799a22
GS
2116a list, a list comprising all input lines is returned, one line per
2117list element. It's easy to grow to a rather large data space this
2118way, so use with care.
a0d0e21e 2119
35f2feb0 2120<FILEHANDLE> may also be spelled C<readline(*FILEHANDLE)>.
19799a22 2121See L<perlfunc/readline>.
fbad3eb5 2122
35f2feb0
GS
2123The null filehandle <> is special: it can be used to emulate the
2124behavior of B<sed> and B<awk>. Input from <> comes either from
a0d0e21e 2125standard input, or from each file listed on the command line. Here's
35f2feb0 2126how it works: the first time <> is evaluated, the @ARGV array is
5a964f20 2127checked, and if it is empty, C<$ARGV[0]> is set to "-", which when opened
a0d0e21e
LW
2128gives you standard input. The @ARGV array is then processed as a list
2129of filenames. The loop
2130
2131 while (<>) {
2132 ... # code for each line
2133 }
2134
2135is equivalent to the following Perl-like pseudo code:
2136
3e3baf6d 2137 unshift(@ARGV, '-') unless @ARGV;
a0d0e21e
LW
2138 while ($ARGV = shift) {
2139 open(ARGV, $ARGV);
2140 while (<ARGV>) {
2141 ... # code for each line
2142 }
2143 }
2144
19799a22
GS
2145except that it isn't so cumbersome to say, and will actually work.
2146It really does shift the @ARGV array and put the current filename
2147into the $ARGV variable. It also uses filehandle I<ARGV>
35f2feb0 2148internally--<> is just a synonym for <ARGV>, which
19799a22 2149is magical. (The pseudo code above doesn't work because it treats
35f2feb0 2150<ARGV> as non-magical.)
a0d0e21e 2151
35f2feb0 2152You can modify @ARGV before the first <> as long as the array ends up
a0d0e21e 2153containing the list of filenames you really want. Line numbers (C<$.>)
19799a22
GS
2154continue as though the input were one big happy file. See the example
2155in L<perlfunc/eof> for how to reset line numbers on each file.
5a964f20 2156
89d205f2 2157If you want to set @ARGV to your own list of files, go right ahead.
5a964f20
TC
2158This sets @ARGV to all plain text files if no @ARGV was given:
2159
2160 @ARGV = grep { -f && -T } glob('*') unless @ARGV;
a0d0e21e 2161
5a964f20
TC
2162You can even set them to pipe commands. For example, this automatically
2163filters compressed arguments through B<gzip>:
2164
2165 @ARGV = map { /\.(gz|Z)$/ ? "gzip -dc < $_ |" : $_ } @ARGV;
2166
2167If you want to pass switches into your script, you can use one of the
a0d0e21e
LW
2168Getopts modules or put a loop on the front like this:
2169
2170 while ($_ = $ARGV[0], /^-/) {
2171 shift;
2172 last if /^--$/;
2173 if (/^-D(.*)/) { $debug = $1 }
2174 if (/^-v/) { $verbose++ }
5a964f20 2175 # ... # other switches
a0d0e21e 2176 }
5a964f20 2177
a0d0e21e 2178 while (<>) {
5a964f20 2179 # ... # code for each line
a0d0e21e
LW
2180 }
2181
89d205f2
YO
2182The <> symbol will return C<undef> for end-of-file only once.
2183If you call it again after this, it will assume you are processing another
19799a22 2184@ARGV list, and if you haven't set @ARGV, will read input from STDIN.
a0d0e21e 2185
b159ebd3 2186If what the angle brackets contain is a simple scalar variable (e.g.,
35f2feb0 2187<$foo>), then that variable contains the name of the
19799a22
GS
2188filehandle to input from, or its typeglob, or a reference to the
2189same. For example:
cb1a09d0
AD
2190
2191 $fh = \*STDIN;
2192 $line = <$fh>;
a0d0e21e 2193
5a964f20
TC
2194If what's within the angle brackets is neither a filehandle nor a simple
2195scalar variable containing a filehandle name, typeglob, or typeglob
2196reference, it is interpreted as a filename pattern to be globbed, and
2197either a list of filenames or the next filename in the list is returned,
19799a22 2198depending on context. This distinction is determined on syntactic
35f2feb0
GS
2199grounds alone. That means C<< <$x> >> is always a readline() from
2200an indirect handle, but C<< <$hash{key}> >> is always a glob().
5a964f20 2201That's because $x is a simple scalar variable, but C<$hash{key}> is
ef191992
YST
2202not--it's a hash element. Even C<< <$x > >> (note the extra space)
2203is treated as C<glob("$x ")>, not C<readline($x)>.
5a964f20
TC
2204
2205One level of double-quote interpretation is done first, but you can't
35f2feb0 2206say C<< <$foo> >> because that's an indirect filehandle as explained
5a964f20
TC
2207in the previous paragraph. (In older versions of Perl, programmers
2208would insert curly brackets to force interpretation as a filename glob:
35f2feb0 2209C<< <${foo}> >>. These days, it's considered cleaner to call the
5a964f20 2210internal function directly as C<glob($foo)>, which is probably the right
19799a22 2211way to have done it in the first place.) For example:
a0d0e21e
LW
2212
2213 while (<*.c>) {
2214 chmod 0644, $_;
2215 }
2216
3a4b19e4 2217is roughly equivalent to:
a0d0e21e
LW
2218
2219 open(FOO, "echo *.c | tr -s ' \t\r\f' '\\012\\012\\012\\012'|");
2220 while (<FOO>) {
5b3eff12 2221 chomp;
a0d0e21e
LW
2222 chmod 0644, $_;
2223 }
2224
3a4b19e4
GS
2225except that the globbing is actually done internally using the standard
2226C<File::Glob> extension. Of course, the shortest way to do the above is:
a0d0e21e
LW
2227
2228 chmod 0644, <*.c>;
2229
19799a22
GS
2230A (file)glob evaluates its (embedded) argument only when it is
2231starting a new list. All values must be read before it will start
2232over. In list context, this isn't important because you automatically
2233get them all anyway. However, in scalar context the operator returns
069e01df 2234the next value each time it's called, or C<undef> when the list has
19799a22
GS
2235run out. As with filehandle reads, an automatic C<defined> is
2236generated when the glob occurs in the test part of a C<while>,
2237because legal glob returns (e.g. a file called F<0>) would otherwise
2238terminate the loop. Again, C<undef> is returned only once. So if
2239you're expecting a single value from a glob, it is much better to
2240say
4633a7c4
LW
2241
2242 ($file) = <blurch*>;
2243
2244than
2245
2246 $file = <blurch*>;
2247
2248because the latter will alternate between returning a filename and
19799a22 2249returning false.
4633a7c4 2250
b159ebd3 2251If you're trying to do variable interpolation, it's definitely better
4633a7c4 2252to use the glob() function, because the older notation can cause people
e37d713d 2253to become confused with the indirect filehandle notation.
4633a7c4
LW
2254
2255 @files = glob("$dir/*.[ch]");
2256 @files = glob($files[$i]);
2257
a0d0e21e 2258=head2 Constant Folding
d74e8afc 2259X<constant folding> X<folding>
a0d0e21e
LW
2260
2261Like C, Perl does a certain amount of expression evaluation at
19799a22 2262compile time whenever it determines that all arguments to an
a0d0e21e
LW
2263operator are static and have no side effects. In particular, string
2264concatenation happens at compile time between literals that don't do
19799a22 2265variable substitution. Backslash interpolation also happens at
a0d0e21e
LW
2266compile time. You can say
2267
2268 'Now is the time for all' . "\n" .
2269 'good men to come to.'
2270
54310121 2271and this all reduces to one string internally. Likewise, if
a0d0e21e
LW
2272you say
2273
2274 foreach $file (@filenames) {
5a964f20 2275 if (-s $file > 5 + 100 * 2**16) { }
54310121 2276 }
a0d0e21e 2277
19799a22
GS
2278the compiler will precompute the number which that expression
2279represents so that the interpreter won't have to.
a0d0e21e 2280
fd1abbef 2281=head2 No-ops
d74e8afc 2282X<no-op> X<nop>
fd1abbef
DN
2283
2284Perl doesn't officially have a no-op operator, but the bare constants
2285C<0> and C<1> are special-cased to not produce a warning in a void
2286context, so you can for example safely do
2287
2288 1 while foo();
2289
2c268ad5 2290=head2 Bitwise String Operators
d74e8afc 2291X<operator, bitwise, string>
2c268ad5
TP
2292
2293Bitstrings of any size may be manipulated by the bitwise operators
2294(C<~ | & ^>).
2295
19799a22
GS
2296If the operands to a binary bitwise op are strings of different
2297sizes, B<|> and B<^> ops act as though the shorter operand had
2298additional zero bits on the right, while the B<&> op acts as though
2299the longer operand were truncated to the length of the shorter.
2300The granularity for such extension or truncation is one or more
2301bytes.
2c268ad5 2302
89d205f2 2303 # ASCII-based examples
2c268ad5
TP
2304 print "j p \n" ^ " a h"; # prints "JAPH\n"
2305 print "JA" | " ph\n"; # prints "japh\n"
2306 print "japh\nJunk" & '_____'; # prints "JAPH\n";
2307 print 'p N$' ^ " E<H\n"; # prints "Perl\n";
2308
19799a22 2309If you are intending to manipulate bitstrings, be certain that
2c268ad5 2310you're supplying bitstrings: If an operand is a number, that will imply
19799a22 2311a B<numeric> bitwise operation. You may explicitly show which type of
2c268ad5
TP
2312operation you intend by using C<""> or C<0+>, as in the examples below.
2313
4358a253
SS
2314 $foo = 150 | 105; # yields 255 (0x96 | 0x69 is 0xFF)
2315 $foo = '150' | 105; # yields 255
2c268ad5
TP
2316 $foo = 150 | '105'; # yields 255
2317 $foo = '150' | '105'; # yields string '155' (under ASCII)
2318
2319 $baz = 0+$foo & 0+$bar; # both ops explicitly numeric
2320 $biz = "$foo" ^ "$bar"; # both ops explicitly stringy
a0d0e21e 2321
1ae175c8
GS
2322See L<perlfunc/vec> for information on how to manipulate individual bits
2323in a bit vector.
2324
55497cff 2325=head2 Integer Arithmetic
d74e8afc 2326X<integer>
a0d0e21e 2327
19799a22 2328By default, Perl assumes that it must do most of its arithmetic in
a0d0e21e
LW
2329floating point. But by saying
2330
2331 use integer;
2332
2333you may tell the compiler that it's okay to use integer operations
19799a22
GS
2334(if it feels like it) from here to the end of the enclosing BLOCK.
2335An inner BLOCK may countermand this by saying
a0d0e21e
LW
2336
2337 no integer;
2338
19799a22
GS
2339which lasts until the end of that BLOCK. Note that this doesn't
2340mean everything is only an integer, merely that Perl may use integer
2341operations if it is so inclined. For example, even under C<use
2342integer>, if you take the C<sqrt(2)>, you'll still get C<1.4142135623731>
2343or so.
2344
2345Used on numbers, the bitwise operators ("&", "|", "^", "~", "<<",
89d205f2 2346and ">>") always produce integral results. (But see also
13a2d996 2347L<Bitwise String Operators>.) However, C<use integer> still has meaning for
19799a22
GS
2348them. By default, their results are interpreted as unsigned integers, but
2349if C<use integer> is in effect, their results are interpreted
2350as signed integers. For example, C<~0> usually evaluates to a large
0be96356 2351integral value. However, C<use integer; ~0> is C<-1> on two's-complement
19799a22 2352machines.
68dc0745
PP
2353
2354=head2 Floating-point Arithmetic
d74e8afc 2355X<floating-point> X<floating point> X<float> X<real>
68dc0745
PP
2356
2357While C<use integer> provides integer-only arithmetic, there is no
19799a22
GS
2358analogous mechanism to provide automatic rounding or truncation to a
2359certain number of decimal places. For rounding to a certain number
2360of digits, sprintf() or printf() is usually the easiest route.
2361See L<perlfaq4>.
68dc0745 2362
5a964f20
TC
2363Floating-point numbers are only approximations to what a mathematician
2364would call real numbers. There are infinitely more reals than floats,
2365so some corners must be cut. For example:
2366
2367 printf "%.20g\n", 123456789123456789;
2368 # produces 123456789123456784
2369
2370Testing for exact equality of floating-point equality or inequality is
2371not a good idea. Here's a (relatively expensive) work-around to compare
2372whether two floating-point numbers are equal to a particular number of
2373decimal places. See Knuth, volume II, for a more robust treatment of
2374this topic.
2375
2376 sub fp_equal {
2377 my ($X, $Y, $POINTS) = @_;
2378 my ($tX, $tY);
2379 $tX = sprintf("%.${POINTS}g", $X);
2380 $tY = sprintf("%.${POINTS}g", $Y);
2381 return $tX eq $tY;
2382 }
2383
68dc0745 2384The POSIX module (part of the standard perl distribution) implements
19799a22
GS
2385ceil(), floor(), and other mathematical and trigonometric functions.
2386The Math::Complex module (part of the standard perl distribution)
2387defines mathematical functions that work on both the reals and the
2388imaginary numbers. Math::Complex not as efficient as POSIX, but
68dc0745
PP
2389POSIX can't work with complex numbers.
2390
2391Rounding in financial applications can have serious implications, and
2392the rounding method used should be specified precisely. In these
2393cases, it probably pays not to trust whichever system rounding is
2394being used by Perl, but to instead implement the rounding function you
2395need yourself.
5a964f20
TC
2396
2397=head2 Bigger Numbers
d74e8afc 2398X<number, arbitrary precision>
5a964f20
TC
2399
2400The standard Math::BigInt and Math::BigFloat modules provide
19799a22 2401variable-precision arithmetic and overloaded operators, although
cd5c4fce 2402they're currently pretty slow. At the cost of some space and
19799a22
GS
2403considerable speed, they avoid the normal pitfalls associated with
2404limited-precision representations.
5a964f20
TC
2405
2406 use Math::BigInt;
2407 $x = Math::BigInt->new('123456789123456789');
2408 print $x * $x;
2409
2410 # prints +15241578780673678515622620750190521
19799a22 2411
cd5c4fce
T
2412There are several modules that let you calculate with (bound only by
2413memory and cpu-time) unlimited or fixed precision. There are also
2414some non-standard modules that provide faster implementations via
2415external C libraries.
2416
2417Here is a short, but incomplete summary:
2418
2419 Math::Fraction big, unlimited fractions like 9973 / 12967
2420 Math::String treat string sequences like numbers
2421 Math::FixedPrecision calculate with a fixed precision
2422 Math::Currency for currency calculations
2423 Bit::Vector manipulate bit vectors fast (uses C)
2424 Math::BigIntFast Bit::Vector wrapper for big numbers
2425 Math::Pari provides access to the Pari C library
2426 Math::BigInteger uses an external C library
2427 Math::Cephes uses external Cephes C library (no big numbers)
2428 Math::Cephes::Fraction fractions via the Cephes library
2429 Math::GMP another one using an external C library
2430
2431Choose wisely.
16070b82
GS
2432
2433=cut