4 perlop - Perl operators and precedence
8 In Perl, the operator determines what operation is performed,
9 independent of the type of the operands. For example C<$a + $b>
10 is always a numeric addition, and if C<$a> or C<$b> do not contain
11 numbers, an attempt is made to convert them to numbers first.
13 This is in contrast to many other dynamic languages, where the
14 operation is determined by the type of the first argument. It also
15 means that Perl has two versions of some operators, one for numeric
16 and one for string comparison. For example C<$a == $b> compares
17 two numbers for equality, and C<$a eq $b> compares two strings.
19 There are a few exceptions though: C<x> can be either string
20 repetition or list repetition, depending on the type of the left
21 operand, and C<&>, C<|> and C<^> can be either string or numeric bit
24 =head2 Operator Precedence and Associativity
25 X<operator, precedence> X<precedence> X<associativity>
27 Operator precedence and associativity work in Perl more or less like
28 they do in mathematics.
30 I<Operator precedence> means some operators are evaluated before
31 others. For example, in C<2 + 4 * 5>, the multiplication has higher
32 precedence so C<4 * 5> is evaluated first yielding C<2 + 20 ==
33 22> and not C<6 * 5 == 30>.
35 I<Operator associativity> defines what happens if a sequence of the
36 same operators is used one after another: whether the evaluator will
37 evaluate the left operations first or the right. For example, in C<8
38 - 4 - 2>, subtraction is left associative so Perl evaluates the
39 expression left to right. C<8 - 4> is evaluated first making the
40 expression C<4 - 2 == 2> and not C<8 - 2 == 6>.
42 Perl operators have the following associativity and precedence,
43 listed from highest precedence to lowest. Operators borrowed from
44 C keep the same precedence relationship with each other, even where
45 C's precedence is slightly screwy. (This makes learning Perl easier
46 for C folks.) With very few exceptions, these all operate on scalar
47 values only, not array values.
49 left terms and list operators (leftward)
53 right ! ~ \ and unary + and -
58 nonassoc named unary operators
59 nonassoc < > <= >= lt gt le ge
60 nonassoc == != <=> eq ne cmp ~~
67 right = += -= *= etc. goto last next redo dump
69 nonassoc list operators (rightward)
74 In the following sections, these operators are covered in precedence order.
76 Many operators can be overloaded for objects. See L<overload>.
78 =head2 Terms and List Operators (Leftward)
79 X<list operator> X<operator, list> X<term>
81 A TERM has the highest precedence in Perl. They include variables,
82 quote and quote-like operators, any expression in parentheses,
83 and any function whose arguments are parenthesized. Actually, there
84 aren't really functions in this sense, just list operators and unary
85 operators behaving as functions because you put parentheses around
86 the arguments. These are all documented in L<perlfunc>.
88 If any list operator (print(), etc.) or any unary operator (chdir(), etc.)
89 is followed by a left parenthesis as the next token, the operator and
90 arguments within parentheses are taken to be of highest precedence,
91 just like a normal function call.
93 In the absence of parentheses, the precedence of list operators such as
94 C<print>, C<sort>, or C<chmod> is either very high or very low depending on
95 whether you are looking at the left side or the right side of the operator.
98 @ary = (1, 3, sort 4, 2);
99 print @ary; # prints 1324
101 the commas on the right of the sort are evaluated before the sort,
102 but the commas on the left are evaluated after. In other words,
103 list operators tend to gobble up all arguments that follow, and
104 then act like a simple TERM with regard to the preceding expression.
105 Be careful with parentheses:
107 # These evaluate exit before doing the print:
108 print($foo, exit); # Obviously not what you want.
109 print $foo, exit; # Nor is this.
111 # These do the print before evaluating exit:
112 (print $foo), exit; # This is what you want.
113 print($foo), exit; # Or this.
114 print ($foo), exit; # Or even this.
118 print ($foo & 255) + 1, "\n";
120 probably doesn't do what you expect at first glance. The parentheses
121 enclose the argument list for C<print> which is evaluated (printing
122 the result of C<$foo & 255>). Then one is added to the return value
123 of C<print> (usually 1). The result is something like this:
125 1 + 1, "\n"; # Obviously not what you meant.
127 To do what you meant properly, you must write:
129 print(($foo & 255) + 1, "\n");
131 See L<Named Unary Operators> for more discussion of this.
133 Also parsed as terms are the C<do {}> and C<eval {}> constructs, as
134 well as subroutine and method calls, and the anonymous
135 constructors C<[]> and C<{}>.
137 See also L<Quote and Quote-like Operators> toward the end of this section,
138 as well as L</"I/O Operators">.
140 =head2 The Arrow Operator
141 X<arrow> X<dereference> X<< -> >>
143 "C<< -> >>" is an infix dereference operator, just as it is in C
144 and C++. If the right side is either a C<[...]>, C<{...}>, or a
145 C<(...)> subscript, then the left side must be either a hard or
146 symbolic reference to an array, a hash, or a subroutine respectively.
147 (Or technically speaking, a location capable of holding a hard
148 reference, if it's an array or hash reference being used for
149 assignment.) See L<perlreftut> and L<perlref>.
151 Otherwise, the right side is a method name or a simple scalar
152 variable containing either the method name or a subroutine reference,
153 and the left side must be either an object (a blessed reference)
154 or a class name (that is, a package name). See L<perlobj>.
156 The dereferencing cases (as opposed to method-calling cases) are
157 somewhat extended by the experimental C<postderef> feature. For the
158 details of that feature, consult L<perlref/Postfix Dereference Syntax>.
160 =head2 Auto-increment and Auto-decrement
161 X<increment> X<auto-increment> X<++> X<decrement> X<auto-decrement> X<-->
163 "++" and "--" work as in C. That is, if placed before a variable,
164 they increment or decrement the variable by one before returning the
165 value, and if placed after, increment or decrement after returning the
169 print $i++; # prints 0
170 print ++$j; # prints 1
172 Note that just as in C, Perl doesn't define B<when> the variable is
173 incremented or decremented. You just know it will be done sometime
174 before or after the value is returned. This also means that modifying
175 a variable twice in the same statement will lead to undefined behavior.
176 Avoid statements like:
181 Perl will not guarantee what the result of the above statements is.
183 The auto-increment operator has a little extra builtin magic to it. If
184 you increment a variable that is numeric, or that has ever been used in
185 a numeric context, you get a normal increment. If, however, the
186 variable has been used in only string contexts since it was set, and
187 has a value that is not the empty string and matches the pattern
188 C</^[a-zA-Z]*[0-9]*\z/>, the increment is done as a string, preserving each
189 character within its range, with carry:
191 print ++($foo = "99"); # prints "100"
192 print ++($foo = "a0"); # prints "a1"
193 print ++($foo = "Az"); # prints "Ba"
194 print ++($foo = "zz"); # prints "aaa"
196 C<undef> is always treated as numeric, and in particular is changed
197 to C<0> before incrementing (so that a post-increment of an undef value
198 will return C<0> rather than C<undef>).
200 The auto-decrement operator is not magical.
202 =head2 Exponentiation
203 X<**> X<exponentiation> X<power>
205 Binary "**" is the exponentiation operator. It binds even more
206 tightly than unary minus, so -2**4 is -(2**4), not (-2)**4. (This is
207 implemented using C's pow(3) function, which actually works on doubles
210 =head2 Symbolic Unary Operators
211 X<unary operator> X<operator, unary>
213 Unary "!" performs logical negation, that is, "not". See also C<not> for a lower
214 precedence version of this.
217 Unary "-" performs arithmetic negation if the operand is numeric,
218 including any string that looks like a number. If the operand is
219 an identifier, a string consisting of a minus sign concatenated
220 with the identifier is returned. Otherwise, if the string starts
221 with a plus or minus, a string starting with the opposite sign is
222 returned. One effect of these rules is that -bareword is equivalent
223 to the string "-bareword". If, however, the string begins with a
224 non-alphabetic character (excluding "+" or "-"), Perl will attempt to convert
225 the string to a numeric and the arithmetic negation is performed. If the
226 string cannot be cleanly converted to a numeric, Perl will give the warning
227 B<Argument "the string" isn't numeric in negation (-) at ...>.
228 X<-> X<negation, arithmetic>
230 Unary "~" performs bitwise negation, that is, 1's complement. For
231 example, C<0666 & ~027> is 0640. (See also L<Integer Arithmetic> and
232 L<Bitwise String Operators>.) Note that the width of the result is
233 platform-dependent: ~0 is 32 bits wide on a 32-bit platform, but 64
234 bits wide on a 64-bit platform, so if you are expecting a certain bit
235 width, remember to use the "&" operator to mask off the excess bits.
236 X<~> X<negation, binary>
238 When complementing strings, if all characters have ordinal values under
239 256, then their complements will, also. But if they do not, all
240 characters will be in either 32- or 64-bit complements, depending on your
241 architecture. So for example, C<~"\x{3B1}"> is C<"\x{FFFF_FC4E}"> on
242 32-bit machines and C<"\x{FFFF_FFFF_FFFF_FC4E}"> on 64-bit machines.
244 Unary "+" has no effect whatsoever, even on strings. It is useful
245 syntactically for separating a function name from a parenthesized expression
246 that would otherwise be interpreted as the complete list of function
247 arguments. (See examples above under L<Terms and List Operators (Leftward)>.)
250 Unary "\" creates a reference to whatever follows it. See L<perlreftut>
251 and L<perlref>. Do not confuse this behavior with the behavior of
252 backslash within a string, although both forms do convey the notion
253 of protecting the next thing from interpolation.
254 X<\> X<reference> X<backslash>
256 =head2 Binding Operators
257 X<binding> X<operator, binding> X<=~> X<!~>
259 Binary "=~" binds a scalar expression to a pattern match. Certain operations
260 search or modify the string $_ by default. This operator makes that kind
261 of operation work on some other string. The right argument is a search
262 pattern, substitution, or transliteration. The left argument is what is
263 supposed to be searched, substituted, or transliterated instead of the default
264 $_. When used in scalar context, the return value generally indicates the
265 success of the operation. The exceptions are substitution (s///)
266 and transliteration (y///) with the C</r> (non-destructive) option,
267 which cause the B<r>eturn value to be the result of the substitution.
268 Behavior in list context depends on the particular operator.
269 See L</"Regexp Quote-Like Operators"> for details and L<perlretut> for
270 examples using these operators.
272 If the right argument is an expression rather than a search pattern,
273 substitution, or transliteration, it is interpreted as a search pattern at run
274 time. Note that this means that its contents will be interpolated twice, so
278 is not ok, as the regex engine will end up trying to compile the
279 pattern C<\>, which it will consider a syntax error.
281 Binary "!~" is just like "=~" except the return value is negated in
284 Binary "!~" with a non-destructive substitution (s///r) or transliteration
285 (y///r) is a syntax error.
287 =head2 Multiplicative Operators
288 X<operator, multiplicative>
290 Binary "*" multiplies two numbers.
293 Binary "/" divides two numbers.
296 Binary "%" is the modulo operator, which computes the division
297 remainder of its first argument with respect to its second argument.
299 operands C<$a> and C<$b>: If C<$b> is positive, then C<$a % $b> is
300 C<$a> minus the largest multiple of C<$b> less than or equal to
301 C<$a>. If C<$b> is negative, then C<$a % $b> is C<$a> minus the
302 smallest multiple of C<$b> that is not less than C<$a> (that is, the
303 result will be less than or equal to zero). If the operands
304 C<$a> and C<$b> are floating point values and the absolute value of
305 C<$b> (that is C<abs($b)>) is less than C<(UV_MAX + 1)>, only
306 the integer portion of C<$a> and C<$b> will be used in the operation
307 (Note: here C<UV_MAX> means the maximum of the unsigned integer type).
308 If the absolute value of the right operand (C<abs($b)>) is greater than
309 or equal to C<(UV_MAX + 1)>, "%" computes the floating-point remainder
310 C<$r> in the equation C<($r = $a - $i*$b)> where C<$i> is a certain
311 integer that makes C<$r> have the same sign as the right operand
312 C<$b> (B<not> as the left operand C<$a> like C function C<fmod()>)
313 and the absolute value less than that of C<$b>.
314 Note that when C<use integer> is in scope, "%" gives you direct access
315 to the modulo operator as implemented by your C compiler. This
316 operator is not as well defined for negative operands, but it will
318 X<%> X<remainder> X<modulo> X<mod>
320 Binary "x" is the repetition operator. In scalar context or if the left
321 operand is not enclosed in parentheses, it returns a string consisting
322 of the left operand repeated the number of times specified by the right
323 operand. In list context, if the left operand is enclosed in
324 parentheses or is a list formed by C<qw/STRING/>, it repeats the list.
325 If the right operand is zero or negative, it returns an empty string
326 or an empty list, depending on the context.
329 print '-' x 80; # print row of dashes
331 print "\t" x ($tab/8), ' ' x ($tab%8); # tab over
333 @ones = (1) x 80; # a list of 80 1's
334 @ones = (5) x @ones; # set all elements to 5
337 =head2 Additive Operators
338 X<operator, additive>
340 Binary C<+> returns the sum of two numbers.
343 Binary C<-> returns the difference of two numbers.
346 Binary C<.> concatenates two strings.
347 X<string, concatenation> X<concatenation>
348 X<cat> X<concat> X<concatenate> X<.>
350 =head2 Shift Operators
351 X<shift operator> X<operator, shift> X<<< << >>>
352 X<<< >> >>> X<right shift> X<left shift> X<bitwise shift>
353 X<shl> X<shr> X<shift, right> X<shift, left>
355 Binary C<<< << >>> returns the value of its left argument shifted left by the
356 number of bits specified by the right argument. Arguments should be
357 integers. (See also L<Integer Arithmetic>.)
359 Binary C<<< >> >>> returns the value of its left argument shifted right by
360 the number of bits specified by the right argument. Arguments should
361 be integers. (See also L<Integer Arithmetic>.)
363 Note that both C<<< << >>> and C<<< >> >>> in Perl are implemented directly using
364 C<<< << >>> and C<<< >> >>> in C. If C<use integer> (see L<Integer Arithmetic>) is
365 in force then signed C integers are used, else unsigned C integers are
366 used. Either way, the implementation isn't going to generate results
367 larger than the size of the integer type Perl was built with (32 bits
370 The result of overflowing the range of the integers is undefined
371 because it is undefined also in C. In other words, using 32-bit
372 integers, C<< 1 << 32 >> is undefined. Shifting by a negative number
373 of bits is also undefined.
375 If you get tired of being subject to your platform's native integers,
376 the C<use bigint> pragma neatly sidesteps the issue altogether:
378 print 20 << 20; # 20971520
379 print 20 << 40; # 5120 on 32-bit machines,
380 # 21990232555520 on 64-bit machines
382 print 20 << 100; # 25353012004564588029934064107520
384 =head2 Named Unary Operators
385 X<operator, named unary>
387 The various named unary operators are treated as functions with one
388 argument, with optional parentheses.
390 If any list operator (print(), etc.) or any unary operator (chdir(), etc.)
391 is followed by a left parenthesis as the next token, the operator and
392 arguments within parentheses are taken to be of highest precedence,
393 just like a normal function call. For example,
394 because named unary operators are higher precedence than C<||>:
396 chdir $foo || die; # (chdir $foo) || die
397 chdir($foo) || die; # (chdir $foo) || die
398 chdir ($foo) || die; # (chdir $foo) || die
399 chdir +($foo) || die; # (chdir $foo) || die
401 but, because * is higher precedence than named operators:
403 chdir $foo * 20; # chdir ($foo * 20)
404 chdir($foo) * 20; # (chdir $foo) * 20
405 chdir ($foo) * 20; # (chdir $foo) * 20
406 chdir +($foo) * 20; # chdir ($foo * 20)
408 rand 10 * 20; # rand (10 * 20)
409 rand(10) * 20; # (rand 10) * 20
410 rand (10) * 20; # (rand 10) * 20
411 rand +(10) * 20; # rand (10 * 20)
413 Regarding precedence, the filetest operators, like C<-f>, C<-M>, etc. are
414 treated like named unary operators, but they don't follow this functional
415 parenthesis rule. That means, for example, that C<-f($file).".bak"> is
416 equivalent to C<-f "$file.bak">.
417 X<-X> X<filetest> X<operator, filetest>
419 See also L<"Terms and List Operators (Leftward)">.
421 =head2 Relational Operators
422 X<relational operator> X<operator, relational>
424 Perl operators that return true or false generally return values
425 that can be safely used as numbers. For example, the relational
426 operators in this section and the equality operators in the next
427 one return C<1> for true and a special version of the defined empty
428 string, C<"">, which counts as a zero but is exempt from warnings
429 about improper numeric conversions, just as C<"0 but true"> is.
431 Binary "<" returns true if the left argument is numerically less than
435 Binary ">" returns true if the left argument is numerically greater
436 than the right argument.
439 Binary "<=" returns true if the left argument is numerically less than
440 or equal to the right argument.
443 Binary ">=" returns true if the left argument is numerically greater
444 than or equal to the right argument.
447 Binary "lt" returns true if the left argument is stringwise less than
451 Binary "gt" returns true if the left argument is stringwise greater
452 than the right argument.
455 Binary "le" returns true if the left argument is stringwise less than
456 or equal to the right argument.
459 Binary "ge" returns true if the left argument is stringwise greater
460 than or equal to the right argument.
463 =head2 Equality Operators
464 X<equality> X<equal> X<equals> X<operator, equality>
466 Binary "==" returns true if the left argument is numerically equal to
470 Binary "!=" returns true if the left argument is numerically not equal
471 to the right argument.
474 Binary "<=>" returns -1, 0, or 1 depending on whether the left
475 argument is numerically less than, equal to, or greater than the right
476 argument. If your platform supports NaNs (not-a-numbers) as numeric
477 values, using them with "<=>" returns undef. NaN is not "<", "==", ">",
478 "<=" or ">=" anything (even NaN), so those 5 return false. NaN != NaN
479 returns true, as does NaN != anything else. If your platform doesn't
480 support NaNs then NaN is just a string with numeric value 0.
481 X<< <=> >> X<spaceship>
483 $ perl -le '$a = "NaN"; print "No NaN support here" if $a == $a'
484 $ perl -le '$a = "NaN"; print "NaN support here" if $a != $a'
486 (Note that the L<bigint>, L<bigrat>, and L<bignum> pragmas all
489 Binary "eq" returns true if the left argument is stringwise equal to
493 Binary "ne" returns true if the left argument is stringwise not equal
494 to the right argument.
497 Binary "cmp" returns -1, 0, or 1 depending on whether the left
498 argument is stringwise less than, equal to, or greater than the right
502 Binary "~~" does a smartmatch between its arguments. Smart matching
503 is described in the next section.
506 "lt", "le", "ge", "gt" and "cmp" use the collation (sort) order specified
507 by the current locale if a legacy C<use locale> (but not
508 C<use locale ':not_characters'>) is in effect. See
509 L<perllocale>. Do not mix these with Unicode, only with legacy binary
510 encodings. The standard L<Unicode::Collate> and
511 L<Unicode::Collate::Locale> modules offer much more powerful solutions to
514 =head2 Smartmatch Operator
516 First available in Perl 5.10.1 (the 5.10.0 version behaved differently),
517 binary C<~~> does a "smartmatch" between its arguments. This is mostly
518 used implicitly in the C<when> construct described in L<perlsyn>, although
519 not all C<when> clauses call the smartmatch operator. Unique among all of
520 Perl's operators, the smartmatch operator can recurse.
522 It is also unique in that all other Perl operators impose a context
523 (usually string or numeric context) on their operands, autoconverting
524 those operands to those imposed contexts. In contrast, smartmatch
525 I<infers> contexts from the actual types of its operands and uses that
526 type information to select a suitable comparison mechanism.
528 The C<~~> operator compares its operands "polymorphically", determining how
529 to compare them according to their actual types (numeric, string, array,
530 hash, etc.) Like the equality operators with which it shares the same
531 precedence, C<~~> returns 1 for true and C<""> for false. It is often best
532 read aloud as "in", "inside of", or "is contained in", because the left
533 operand is often looked for I<inside> the right operand. That makes the
534 order of the operands to the smartmatch operand often opposite that of
535 the regular match operator. In other words, the "smaller" thing is usually
536 placed in the left operand and the larger one in the right.
538 The behavior of a smartmatch depends on what type of things its arguments
539 are, as determined by the following table. The first row of the table
540 whose types apply determines the smartmatch behavior. Because what
541 actually happens is mostly determined by the type of the second operand,
542 the table is sorted on the right operand instead of on the left.
544 Left Right Description and pseudocode
545 ===============================================================
546 Any undef check whether Any is undefined
549 Any Object invoke ~~ overloading on Object, or die
551 Right operand is an ARRAY:
553 Left Right Description and pseudocode
554 ===============================================================
555 ARRAY1 ARRAY2 recurse on paired elements of ARRAY1 and ARRAY2[2]
556 like: (ARRAY1[0] ~~ ARRAY2[0])
557 && (ARRAY1[1] ~~ ARRAY2[1]) && ...
558 HASH ARRAY any ARRAY elements exist as HASH keys
559 like: grep { exists HASH->{$_} } ARRAY
560 Regexp ARRAY any ARRAY elements pattern match Regexp
561 like: grep { /Regexp/ } ARRAY
562 undef ARRAY undef in ARRAY
563 like: grep { !defined } ARRAY
564 Any ARRAY smartmatch each ARRAY element[3]
565 like: grep { Any ~~ $_ } ARRAY
567 Right operand is a HASH:
569 Left Right Description and pseudocode
570 ===============================================================
571 HASH1 HASH2 all same keys in both HASHes
573 grep { exists HASH2->{$_} } keys HASH1
574 ARRAY HASH any ARRAY elements exist as HASH keys
575 like: grep { exists HASH->{$_} } ARRAY
576 Regexp HASH any HASH keys pattern match Regexp
577 like: grep { /Regexp/ } keys HASH
578 undef HASH always false (undef can't be a key)
580 Any HASH HASH key existence
581 like: exists HASH->{Any}
583 Right operand is CODE:
585 Left Right Description and pseudocode
586 ===============================================================
587 ARRAY CODE sub returns true on all ARRAY elements[1]
588 like: !grep { !CODE->($_) } ARRAY
589 HASH CODE sub returns true on all HASH keys[1]
590 like: !grep { !CODE->($_) } keys HASH
591 Any CODE sub passed Any returns true
594 Right operand is a Regexp:
596 Left Right Description and pseudocode
597 ===============================================================
598 ARRAY Regexp any ARRAY elements match Regexp
599 like: grep { /Regexp/ } ARRAY
600 HASH Regexp any HASH keys match Regexp
601 like: grep { /Regexp/ } keys HASH
602 Any Regexp pattern match
603 like: Any =~ /Regexp/
607 Left Right Description and pseudocode
608 ===============================================================
609 Object Any invoke ~~ overloading on Object,
612 Any Num numeric equality
614 Num nummy[4] numeric equality
616 undef Any check whether undefined
618 Any Any string equality
627 Empty hashes or arrays match.
630 That is, each element smartmatches the element of the same index in the other array.[3]
633 If a circular reference is found, fall back to referential equality.
636 Either an actual number, or a string that looks like one.
640 The smartmatch implicitly dereferences any non-blessed hash or array
641 reference, so the C<I<HASH>> and C<I<ARRAY>> entries apply in those cases.
642 For blessed references, the C<I<Object>> entries apply. Smartmatches
643 involving hashes only consider hash keys, never hash values.
645 The "like" code entry is not always an exact rendition. For example, the
646 smartmatch operator short-circuits whenever possible, but C<grep> does
647 not. Also, C<grep> in scalar context returns the number of matches, but
648 C<~~> returns only true or false.
650 Unlike most operators, the smartmatch operator knows to treat C<undef>
654 @array = (1, 2, 3, undef, 4, 5);
655 say "some elements undefined" if undef ~~ @array;
657 Each operand is considered in a modified scalar context, the modification
658 being that array and hash variables are passed by reference to the
659 operator, which implicitly dereferences them. Both elements
660 of each pair are the same:
664 my %hash = (red => 1, blue => 2, green => 3,
665 orange => 4, yellow => 5, purple => 6,
666 black => 7, grey => 8, white => 9);
668 my @array = qw(red blue green);
670 say "some array elements in hash keys" if @array ~~ %hash;
671 say "some array elements in hash keys" if \@array ~~ \%hash;
673 say "red in array" if "red" ~~ @array;
674 say "red in array" if "red" ~~ \@array;
676 say "some keys end in e" if /e$/ ~~ %hash;
677 say "some keys end in e" if /e$/ ~~ \%hash;
679 Two arrays smartmatch if each element in the first array smartmatches
680 (that is, is "in") the corresponding element in the second array,
684 my @little = qw(red blue green);
685 my @bigger = ("red", "blue", [ "orange", "green" ] );
686 if (@little ~~ @bigger) { # true!
687 say "little is contained in bigger";
690 Because the smartmatch operator recurses on nested arrays, this
691 will still report that "red" is in the array.
694 my @array = qw(red blue green);
695 my $nested_array = [[[[[[[ @array ]]]]]]];
696 say "red in array" if "red" ~~ $nested_array;
698 If two arrays smartmatch each other, then they are deep
699 copies of each others' values, as this example reports:
702 my @a = (0, 1, 2, [3, [4, 5], 6], 7);
703 my @b = (0, 1, 2, [3, [4, 5], 6], 7);
705 if (@a ~~ @b && @b ~~ @a) {
706 say "a and b are deep copies of each other";
709 say "a smartmatches in b";
712 say "b smartmatches in a";
715 say "a and b don't smartmatch each other at all";
719 If you were to set C<$b[3] = 4>, then instead of reporting that "a and b
720 are deep copies of each other", it now reports that "b smartmatches in a".
721 That because the corresponding position in C<@a> contains an array that
722 (eventually) has a 4 in it.
724 Smartmatching one hash against another reports whether both contain the
725 same keys, no more and no less. This could be used to see whether two
726 records have the same field names, without caring what values those fields
727 might have. For example:
731 state $REQUIRED_FIELDS = { name=>1, rank=>1, serial_num=>1 };
733 my ($class, $init_fields) = @_;
735 die "Must supply (only) name, rank, and serial number"
736 unless $init_fields ~~ $REQUIRED_FIELDS;
741 or, if other non-required fields are allowed, use ARRAY ~~ HASH:
745 state $REQUIRED_FIELDS = { name=>1, rank=>1, serial_num=>1 };
747 my ($class, $init_fields) = @_;
749 die "Must supply (at least) name, rank, and serial number"
750 unless [keys %{$init_fields}] ~~ $REQUIRED_FIELDS;
755 The smartmatch operator is most often used as the implicit operator of a
756 C<when> clause. See the section on "Switch Statements" in L<perlsyn>.
758 =head3 Smartmatching of Objects
760 To avoid relying on an object's underlying representation, if the
761 smartmatch's right operand is an object that doesn't overload C<~~>,
762 it raises the exception "C<Smartmatching a non-overloaded object
763 breaks encapsulation>". That's because one has no business digging
764 around to see whether something is "in" an object. These are all
765 illegal on objects without a C<~~> overload:
771 However, you can change the way an object is smartmatched by overloading
772 the C<~~> operator. This is allowed to extend the usual smartmatch semantics.
773 For objects that do have an C<~~> overload, see L<overload>.
775 Using an object as the left operand is allowed, although not very useful.
776 Smartmatching rules take precedence over overloading, so even if the
777 object in the left operand has smartmatch overloading, this will be
778 ignored. A left operand that is a non-overloaded object falls back on a
779 string or numeric comparison of whatever the C<ref> operator returns. That
784 does I<not> invoke the overload method with C<I<X>> as an argument.
785 Instead the above table is consulted as normal, and based on the type of
786 C<I<X>>, overloading may or may not be invoked. For simple strings or
787 numbers, in becomes equivalent to this:
789 $object ~~ $number ref($object) == $number
790 $object ~~ $string ref($object) eq $string
792 For example, this reports that the handle smells IOish
793 (but please don't really do this!):
796 my $fh = IO::Handle->new();
797 if ($fh ~~ /\bIO\b/) {
798 say "handle smells IOish";
801 That's because it treats C<$fh> as a string like
802 C<"IO::Handle=GLOB(0x8039e0)">, then pattern matches against that.
805 X<operator, bitwise, and> X<bitwise and> X<&>
807 Binary "&" returns its operands ANDed together bit by bit.
808 (See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
810 Note that "&" has lower priority than relational operators, so for example
811 the parentheses are essential in a test like
813 print "Even\n" if ($x & 1) == 0;
815 =head2 Bitwise Or and Exclusive Or
816 X<operator, bitwise, or> X<bitwise or> X<|> X<operator, bitwise, xor>
819 Binary "|" returns its operands ORed together bit by bit.
820 (See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
822 Binary "^" returns its operands XORed together bit by bit.
823 (See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
825 Note that "|" and "^" have lower priority than relational operators, so
826 for example the brackets are essential in a test like
828 print "false\n" if (8 | 2) != 10;
830 =head2 C-style Logical And
831 X<&&> X<logical and> X<operator, logical, and>
833 Binary "&&" performs a short-circuit logical AND operation. That is,
834 if the left operand is false, the right operand is not even evaluated.
835 Scalar or list context propagates down to the right operand if it
838 =head2 C-style Logical Or
839 X<||> X<operator, logical, or>
841 Binary "||" performs a short-circuit logical OR operation. That is,
842 if the left operand is true, the right operand is not even evaluated.
843 Scalar or list context propagates down to the right operand if it
846 =head2 Logical Defined-Or
847 X<//> X<operator, logical, defined-or>
849 Although it has no direct equivalent in C, Perl's C<//> operator is related
850 to its C-style or. In fact, it's exactly the same as C<||>, except that it
851 tests the left hand side's definedness instead of its truth. Thus,
852 C<< EXPR1 // EXPR2 >> returns the value of C<< EXPR1 >> if it's defined,
853 otherwise, the value of C<< EXPR2 >> is returned. (C<< EXPR1 >> is evaluated
854 in scalar context, C<< EXPR2 >> in the context of C<< // >> itself). Usually,
855 this is the same result as C<< defined(EXPR1) ? EXPR1 : EXPR2 >> (except that
856 the ternary-operator form can be used as a lvalue, while C<< EXPR1 // EXPR2 >>
857 cannot). This is very useful for
858 providing default values for variables. If you actually want to test if
859 at least one of C<$a> and C<$b> is defined, use C<defined($a // $b)>.
861 The C<||>, C<//> and C<&&> operators return the last value evaluated
862 (unlike C's C<||> and C<&&>, which return 0 or 1). Thus, a reasonably
863 portable way to find out the home directory might be:
868 // die "You're homeless!\n";
870 In particular, this means that you shouldn't use this
871 for selecting between two aggregates for assignment:
873 @a = @b || @c; # this is wrong
874 @a = scalar(@b) || @c; # really meant this
875 @a = @b ? @b : @c; # this works fine, though
877 As alternatives to C<&&> and C<||> when used for
878 control flow, Perl provides the C<and> and C<or> operators (see below).
879 The short-circuit behavior is identical. The precedence of "and"
880 and "or" is much lower, however, so that you can safely use them after a
881 list operator without the need for parentheses:
883 unlink "alpha", "beta", "gamma"
884 or gripe(), next LINE;
886 With the C-style operators that would have been written like this:
888 unlink("alpha", "beta", "gamma")
889 || (gripe(), next LINE);
891 It would be even more readable to write that this way:
893 unless(unlink("alpha", "beta", "gamma")) {
898 Using "or" for assignment is unlikely to do what you want; see below.
900 =head2 Range Operators
901 X<operator, range> X<range> X<..> X<...>
903 Binary ".." is the range operator, which is really two different
904 operators depending on the context. In list context, it returns a
905 list of values counting (up by ones) from the left value to the right
906 value. If the left value is greater than the right value then it
907 returns the empty list. The range operator is useful for writing
908 C<foreach (1..10)> loops and for doing slice operations on arrays. In
909 the current implementation, no temporary array is created when the
910 range operator is used as the expression in C<foreach> loops, but older
911 versions of Perl might burn a lot of memory when you write something
914 for (1 .. 1_000_000) {
918 The range operator also works on strings, using the magical
919 auto-increment, see below.
921 In scalar context, ".." returns a boolean value. The operator is
922 bistable, like a flip-flop, and emulates the line-range (comma)
923 operator of B<sed>, B<awk>, and various editors. Each ".." operator
924 maintains its own boolean state, even across calls to a subroutine
925 that contains it. It is false as long as its left operand is false.
926 Once the left operand is true, the range operator stays true until the
927 right operand is true, I<AFTER> which the range operator becomes false
928 again. It doesn't become false till the next time the range operator
929 is evaluated. It can test the right operand and become false on the
930 same evaluation it became true (as in B<awk>), but it still returns
931 true once. If you don't want it to test the right operand until the
932 next evaluation, as in B<sed>, just use three dots ("...") instead of
933 two. In all other regards, "..." behaves just like ".." does.
935 The right operand is not evaluated while the operator is in the
936 "false" state, and the left operand is not evaluated while the
937 operator is in the "true" state. The precedence is a little lower
938 than || and &&. The value returned is either the empty string for
939 false, or a sequence number (beginning with 1) for true. The sequence
940 number is reset for each range encountered. The final sequence number
941 in a range has the string "E0" appended to it, which doesn't affect
942 its numeric value, but gives you something to search for if you want
943 to exclude the endpoint. You can exclude the beginning point by
944 waiting for the sequence number to be greater than 1.
946 If either operand of scalar ".." is a constant expression,
947 that operand is considered true if it is equal (C<==>) to the current
948 input line number (the C<$.> variable).
950 To be pedantic, the comparison is actually C<int(EXPR) == int(EXPR)>,
951 but that is only an issue if you use a floating point expression; when
952 implicitly using C<$.> as described in the previous paragraph, the
953 comparison is C<int(EXPR) == int($.)> which is only an issue when C<$.>
954 is set to a floating point value and you are not reading from a file.
955 Furthermore, C<"span" .. "spat"> or C<2.18 .. 3.14> will not do what
956 you want in scalar context because each of the operands are evaluated
957 using their integer representation.
961 As a scalar operator:
963 if (101 .. 200) { print; } # print 2nd hundred lines, short for
964 # if ($. == 101 .. $. == 200) { print; }
966 next LINE if (1 .. /^$/); # skip header lines, short for
967 # next LINE if ($. == 1 .. /^$/);
968 # (typically in a loop labeled LINE)
970 s/^/> / if (/^$/ .. eof()); # quote body
972 # parse mail messages
974 $in_header = 1 .. /^$/;
975 $in_body = /^$/ .. eof;
982 close ARGV if eof; # reset $. each file
985 Here's a simple example to illustrate the difference between
986 the two range operators:
999 This program will print only the line containing "Bar". If
1000 the range operator is changed to C<...>, it will also print the
1003 And now some examples as a list operator:
1005 for (101 .. 200) { print } # print $_ 100 times
1006 @foo = @foo[0 .. $#foo]; # an expensive no-op
1007 @foo = @foo[$#foo-4 .. $#foo]; # slice last 5 items
1009 The range operator (in list context) makes use of the magical
1010 auto-increment algorithm if the operands are strings. You
1013 @alphabet = ("A" .. "Z");
1015 to get all normal letters of the English alphabet, or
1017 $hexdigit = (0 .. 9, "a" .. "f")[$num & 15];
1019 to get a hexadecimal digit, or
1021 @z2 = ("01" .. "31");
1024 to get dates with leading zeros.
1026 If the final value specified is not in the sequence that the magical
1027 increment would produce, the sequence goes until the next value would
1028 be longer than the final value specified.
1030 If the initial value specified isn't part of a magical increment
1031 sequence (that is, a non-empty string matching C</^[a-zA-Z]*[0-9]*\z/>),
1032 only the initial value will be returned. So the following will only
1035 use charnames "greek";
1036 my @greek_small = ("\N{alpha}" .. "\N{omega}");
1038 To get the 25 traditional lowercase Greek letters, including both sigmas,
1039 you could use this instead:
1041 use charnames "greek";
1042 my @greek_small = map { chr } ( ord("\N{alpha}")
1047 However, because there are I<many> other lowercase Greek characters than
1048 just those, to match lowercase Greek characters in a regular expression,
1049 you would use the pattern C</(?:(?=\p{Greek})\p{Lower})+/>.
1051 Because each operand is evaluated in integer form, C<2.18 .. 3.14> will
1052 return two elements in list context.
1054 @list = (2.18 .. 3.14); # same as @list = (2 .. 3);
1056 =head2 Conditional Operator
1057 X<operator, conditional> X<operator, ternary> X<ternary> X<?:>
1059 Ternary "?:" is the conditional operator, just as in C. It works much
1060 like an if-then-else. If the argument before the ? is true, the
1061 argument before the : is returned, otherwise the argument after the :
1062 is returned. For example:
1064 printf "I have %d dog%s.\n", $n,
1065 ($n == 1) ? "" : "s";
1067 Scalar or list context propagates downward into the 2nd
1068 or 3rd argument, whichever is selected.
1070 $a = $ok ? $b : $c; # get a scalar
1071 @a = $ok ? @b : @c; # get an array
1072 $a = $ok ? @b : @c; # oops, that's just a count!
1074 The operator may be assigned to if both the 2nd and 3rd arguments are
1075 legal lvalues (meaning that you can assign to them):
1077 ($a_or_b ? $a : $b) = $c;
1079 Because this operator produces an assignable result, using assignments
1080 without parentheses will get you in trouble. For example, this:
1082 $a % 2 ? $a += 10 : $a += 2
1086 (($a % 2) ? ($a += 10) : $a) += 2
1090 ($a % 2) ? ($a += 10) : ($a += 2)
1092 That should probably be written more simply as:
1094 $a += ($a % 2) ? 10 : 2;
1096 =head2 Assignment Operators
1097 X<assignment> X<operator, assignment> X<=> X<**=> X<+=> X<*=> X<&=>
1098 X<<< <<= >>> X<&&=> X<-=> X</=> X<|=> X<<< >>= >>> X<||=> X<//=> X<.=>
1101 "=" is the ordinary assignment operator.
1103 Assignment operators work as in C. That is,
1111 although without duplicating any side effects that dereferencing the lvalue
1112 might trigger, such as from tie(). Other assignment operators work similarly.
1113 The following are recognized:
1115 **= += *= &= <<= &&=
1120 Although these are grouped by family, they all have the precedence
1123 Unlike in C, the scalar assignment operator produces a valid lvalue.
1124 Modifying an assignment is equivalent to doing the assignment and
1125 then modifying the variable that was assigned to. This is useful
1126 for modifying a copy of something, like this:
1128 ($tmp = $global) =~ tr/13579/24680/;
1130 Although as of 5.14, that can be also be accomplished this way:
1133 $tmp = ($global =~ tr/13579/24680/r);
1144 Similarly, a list assignment in list context produces the list of
1145 lvalues assigned to, and a list assignment in scalar context returns
1146 the number of elements produced by the expression on the right hand
1147 side of the assignment.
1149 =head2 Comma Operator
1150 X<comma> X<operator, comma> X<,>
1152 Binary "," is the comma operator. In scalar context it evaluates
1153 its left argument, throws that value away, then evaluates its right
1154 argument and returns that value. This is just like C's comma operator.
1156 In list context, it's just the list argument separator, and inserts
1157 both its arguments into the list. These arguments are also evaluated
1160 The C<< => >> operator is a synonym for the comma except that it causes a
1161 word on its left to be interpreted as a string if it begins with a letter
1162 or underscore and is composed only of letters, digits and underscores.
1163 This includes operands that might otherwise be interpreted as operators,
1164 constants, single number v-strings or function calls. If in doubt about
1165 this behavior, the left operand can be quoted explicitly.
1167 Otherwise, the C<< => >> operator behaves exactly as the comma operator
1168 or list argument separator, according to context.
1172 use constant FOO => "something";
1174 my %h = ( FOO => 23 );
1178 my %h = ("FOO", 23);
1182 my %h = ("something", 23);
1184 The C<< => >> operator is helpful in documenting the correspondence
1185 between keys and values in hashes, and other paired elements in lists.
1187 %hash = ( $key => $value );
1188 login( $username => $password );
1190 The special quoting behavior ignores precedence, and hence may apply to
1191 I<part> of the left operand:
1193 print time.shift => "bbb";
1195 That example prints something like "1314363215shiftbbb", because the
1196 C<< => >> implicitly quotes the C<shift> immediately on its left, ignoring
1197 the fact that C<time.shift> is the entire left operand.
1199 =head2 List Operators (Rightward)
1200 X<operator, list, rightward> X<list operator>
1202 On the right side of a list operator, the comma has very low precedence,
1203 such that it controls all comma-separated expressions found there.
1204 The only operators with lower precedence are the logical operators
1205 "and", "or", and "not", which may be used to evaluate calls to list
1206 operators without the need for parentheses:
1208 open HANDLE, "< :utf8", "filename" or die "Can't open: $!\n";
1210 However, some people find that code harder to read than writing
1211 it with parentheses:
1213 open(HANDLE, "< :utf8", "filename") or die "Can't open: $!\n";
1215 in which case you might as well just use the more customary "||" operator:
1217 open(HANDLE, "< :utf8", "filename") || die "Can't open: $!\n";
1219 See also discussion of list operators in L<Terms and List Operators (Leftward)>.
1222 X<operator, logical, not> X<not>
1224 Unary "not" returns the logical negation of the expression to its right.
1225 It's the equivalent of "!" except for the very low precedence.
1228 X<operator, logical, and> X<and>
1230 Binary "and" returns the logical conjunction of the two surrounding
1231 expressions. It's equivalent to C<&&> except for the very low
1232 precedence. This means that it short-circuits: the right
1233 expression is evaluated only if the left expression is true.
1235 =head2 Logical or and Exclusive Or
1236 X<operator, logical, or> X<operator, logical, xor>
1237 X<operator, logical, exclusive or>
1240 Binary "or" returns the logical disjunction of the two surrounding
1241 expressions. It's equivalent to C<||> except for the very low precedence.
1242 This makes it useful for control flow:
1244 print FH $data or die "Can't write to FH: $!";
1246 This means that it short-circuits: the right expression is evaluated
1247 only if the left expression is false. Due to its precedence, you must
1248 be careful to avoid using it as replacement for the C<||> operator.
1249 It usually works out better for flow control than in assignments:
1251 $a = $b or $c; # bug: this is wrong
1252 ($a = $b) or $c; # really means this
1253 $a = $b || $c; # better written this way
1255 However, when it's a list-context assignment and you're trying to use
1256 C<||> for control flow, you probably need "or" so that the assignment
1257 takes higher precedence.
1259 @info = stat($file) || die; # oops, scalar sense of stat!
1260 @info = stat($file) or die; # better, now @info gets its due
1262 Then again, you could always use parentheses.
1264 Binary C<xor> returns the exclusive-OR of the two surrounding expressions.
1265 It cannot short-circuit (of course).
1267 There is no low precedence operator for defined-OR.
1269 =head2 C Operators Missing From Perl
1270 X<operator, missing from perl> X<&> X<*>
1271 X<typecasting> X<(TYPE)>
1273 Here is what C has that Perl doesn't:
1279 Address-of operator. (But see the "\" operator for taking a reference.)
1283 Dereference-address operator. (Perl's prefix dereferencing
1284 operators are typed: $, @, %, and &.)
1288 Type-casting operator.
1292 =head2 Quote and Quote-like Operators
1293 X<operator, quote> X<operator, quote-like> X<q> X<qq> X<qx> X<qw> X<m>
1294 X<qr> X<s> X<tr> X<'> X<''> X<"> X<""> X<//> X<`> X<``> X<<< << >>>
1295 X<escape sequence> X<escape>
1297 While we usually think of quotes as literal values, in Perl they
1298 function as operators, providing various kinds of interpolating and
1299 pattern matching capabilities. Perl provides customary quote characters
1300 for these behaviors, but also provides a way for you to choose your
1301 quote character for any of them. In the following table, a C<{}> represents
1302 any pair of delimiters you choose.
1304 Customary Generic Meaning Interpolates
1307 `` qx{} Command yes*
1309 // m{} Pattern match yes*
1311 s{}{} Substitution yes*
1312 tr{}{} Transliteration no (but see below)
1313 y{}{} Transliteration no (but see below)
1316 * unless the delimiter is ''.
1318 Non-bracketing delimiters use the same character fore and aft, but the four
1319 sorts of ASCII brackets (round, angle, square, curly) all nest, which means
1328 Note, however, that this does not always work for quoting Perl code:
1330 $s = q{ if($a eq "}") ... }; # WRONG
1332 is a syntax error. The C<Text::Balanced> module (standard as of v5.8,
1333 and from CPAN before then) is able to do this properly.
1335 There can be whitespace between the operator and the quoting
1336 characters, except when C<#> is being used as the quoting character.
1337 C<q#foo#> is parsed as the string C<foo>, while C<q #foo#> is the
1338 operator C<q> followed by a comment. Its argument will be taken
1339 from the next line. This allows you to write:
1341 s {foo} # Replace foo
1344 The following escape sequences are available in constructs that interpolate,
1345 and in transliterations:
1346 X<\t> X<\n> X<\r> X<\f> X<\b> X<\a> X<\e> X<\x> X<\0> X<\c> X<\N> X<\N{}>
1349 Sequence Note Description
1355 \a alarm (bell) (BEL)
1357 \x{263A} [1,8] hex char (example: SMILEY)
1358 \x1b [2,8] restricted range hex char (example: ESC)
1359 \N{name} [3] named Unicode character or character sequence
1360 \N{U+263D} [4,8] Unicode character (example: FIRST QUARTER MOON)
1361 \c[ [5] control char (example: chr(27))
1362 \o{23072} [6,8] octal char (example: SMILEY)
1363 \033 [7,8] restricted range octal char (example: ESC)
1369 The result is the character specified by the hexadecimal number between
1370 the braces. See L</[8]> below for details on which character.
1372 Only hexadecimal digits are valid between the braces. If an invalid
1373 character is encountered, a warning will be issued and the invalid
1374 character and all subsequent characters (valid or invalid) within the
1375 braces will be discarded.
1377 If there are no valid digits between the braces, the generated character is
1378 the NULL character (C<\x{00}>). However, an explicit empty brace (C<\x{}>)
1379 will not cause a warning (currently).
1383 The result is the character specified by the hexadecimal number in the range
1384 0x00 to 0xFF. See L</[8]> below for details on which character.
1386 Only hexadecimal digits are valid following C<\x>. When C<\x> is followed
1387 by fewer than two valid digits, any valid digits will be zero-padded. This
1388 means that C<\x7> will be interpreted as C<\x07>, and a lone <\x> will be
1389 interpreted as C<\x00>. Except at the end of a string, having fewer than
1390 two valid digits will result in a warning. Note that although the warning
1391 says the illegal character is ignored, it is only ignored as part of the
1392 escape and will still be used as the subsequent character in the string.
1395 Original Result Warns?
1403 The result is the Unicode character or character sequence given by I<name>.
1408 C<\N{U+I<hexadecimal number>}> means the Unicode character whose Unicode code
1409 point is I<hexadecimal number>.
1413 The character following C<\c> is mapped to some other character as shown in the
1430 In other words, it's the character whose code point has had 64 xor'd with
1431 its uppercase. C<\c?> is DELETE because C<ord("?") ^ 64> is 127, and
1432 C<\c@> is NULL because the ord of "@" is 64, so xor'ing 64 itself produces 0.
1434 Also, C<\c\I<X>> yields C< chr(28) . "I<X>"> for any I<X>, but cannot come at the
1435 end of a string, because the backslash would be parsed as escaping the end
1438 On ASCII platforms, the resulting characters from the list above are the
1439 complete set of ASCII controls. This isn't the case on EBCDIC platforms; see
1440 L<perlebcdic/OPERATOR DIFFERENCES> for the complete list of what these
1441 sequences mean on both ASCII and EBCDIC platforms.
1443 Use of any other character following the "c" besides those listed above is
1444 discouraged, and some are deprecated with the intention of removing
1445 those in a later Perl version. What happens for any of these
1446 other characters currently though, is that the value is derived by xor'ing
1447 with the seventh bit, which is 64.
1449 To get platform independent controls, you can use C<\N{...}>.
1453 The result is the character specified by the octal number between the braces.
1454 See L</[8]> below for details on which character.
1456 If a character that isn't an octal digit is encountered, a warning is raised,
1457 and the value is based on the octal digits before it, discarding it and all
1458 following characters up to the closing brace. It is a fatal error if there are
1459 no octal digits at all.
1463 The result is the character specified by the three-digit octal number in the
1464 range 000 to 777 (but best to not use above 077, see next paragraph). See
1465 L</[8]> below for details on which character.
1467 Some contexts allow 2 or even 1 digit, but any usage without exactly
1468 three digits, the first being a zero, may give unintended results. (For
1469 example, in a regular expression it may be confused with a backreference;
1470 see L<perlrebackslash/Octal escapes>.) Starting in Perl 5.14, you may
1471 use C<\o{}> instead, which avoids all these problems. Otherwise, it is best to
1472 use this construct only for ordinals C<\077> and below, remembering to pad to
1473 the left with zeros to make three digits. For larger ordinals, either use
1474 C<\o{}>, or convert to something else, such as to hex and use C<\x{}>
1477 Having fewer than 3 digits may lead to a misleading warning message that says
1478 that what follows is ignored. For example, C<"\128"> in the ASCII character set
1479 is equivalent to the two characters C<"\n8">, but the warning C<Illegal octal
1480 digit '8' ignored> will be thrown. If C<"\n8"> is what you want, you can
1481 avoid this warning by padding your octal number with C<0>'s: C<"\0128">.
1485 Several constructs above specify a character by a number. That number
1486 gives the character's position in the character set encoding (indexed from 0).
1487 This is called synonymously its ordinal, code position, or code point. Perl
1488 works on platforms that have a native encoding currently of either ASCII/Latin1
1489 or EBCDIC, each of which allow specification of 256 characters. In general, if
1490 the number is 255 (0xFF, 0377) or below, Perl interprets this in the platform's
1491 native encoding. If the number is 256 (0x100, 0400) or above, Perl interprets
1492 it as a Unicode code point and the result is the corresponding Unicode
1493 character. For example C<\x{50}> and C<\o{120}> both are the number 80 in
1494 decimal, which is less than 256, so the number is interpreted in the native
1495 character set encoding. In ASCII the character in the 80th position (indexed
1496 from 0) is the letter "P", and in EBCDIC it is the ampersand symbol "&".
1497 C<\x{100}> and C<\o{400}> are both 256 in decimal, so the number is interpreted
1498 as a Unicode code point no matter what the native encoding is. The name of the
1499 character in the 256th position (indexed by 0) in Unicode is
1500 C<LATIN CAPITAL LETTER A WITH MACRON>.
1502 There are a couple of exceptions to the above rule. S<C<\N{U+I<hex number>}>> is
1503 always interpreted as a Unicode code point, so that C<\N{U+0050}> is "P" even
1504 on EBCDIC platforms. And if L<C<S<use encoding>>|encoding> is in effect, the
1505 number is considered to be in that encoding, and is translated from that into
1506 the platform's native encoding if there is a corresponding native character;
1507 otherwise to Unicode.
1511 B<NOTE>: Unlike C and other languages, Perl has no C<\v> escape sequence for
1512 the vertical tab (VT, which is 11 in both ASCII and EBCDIC), but you may
1515 does have meaning in regular expression patterns in Perl, see L<perlre>.)
1517 The following escape sequences are available in constructs that interpolate,
1518 but not in transliterations.
1519 X<\l> X<\u> X<\L> X<\U> X<\E> X<\Q> X<\F>
1521 \l lowercase next character only
1522 \u titlecase (not uppercase!) next character only
1523 \L lowercase all characters till \E or end of string
1524 \U uppercase all characters till \E or end of string
1525 \F foldcase all characters till \E or end of string
1526 \Q quote (disable) pattern metacharacters till \E or
1528 \E end either case modification or quoted section
1529 (whichever was last seen)
1531 See L<perlfunc/quotemeta> for the exact definition of characters that
1532 are quoted by C<\Q>.
1534 C<\L>, C<\U>, C<\F>, and C<\Q> can stack, in which case you need one
1535 C<\E> for each. For example:
1537 say"This \Qquoting \ubusiness \Uhere isn't quite\E done yet,\E is it?";
1538 This quoting\ Business\ HERE\ ISN\'T\ QUITE\ done\ yet\, is it?
1540 If C<use locale> is in effect (but not C<use locale ':not_characters'>),
1541 the case map used by C<\l>, C<\L>,
1542 C<\u>, and C<\U> is taken from the current locale. See L<perllocale>.
1543 If Unicode (for example, C<\N{}> or code points of 0x100 or
1544 beyond) is being used, the case map used by C<\l>, C<\L>, C<\u>, and
1545 C<\U> is as defined by Unicode. That means that case-mapping
1546 a single character can sometimes produce several characters.
1547 Under C<use locale>, C<\F> produces the same results as C<\L>.
1549 All systems use the virtual C<"\n"> to represent a line terminator,
1550 called a "newline". There is no such thing as an unvarying, physical
1551 newline character. It is only an illusion that the operating system,
1552 device drivers, C libraries, and Perl all conspire to preserve. Not all
1553 systems read C<"\r"> as ASCII CR and C<"\n"> as ASCII LF. For example,
1554 on the ancient Macs (pre-MacOS X) of yesteryear, these used to be reversed,
1555 and on systems without line terminator,
1556 printing C<"\n"> might emit no actual data. In general, use C<"\n"> when
1557 you mean a "newline" for your system, but use the literal ASCII when you
1558 need an exact character. For example, most networking protocols expect
1559 and prefer a CR+LF (C<"\015\012"> or C<"\cM\cJ">) for line terminators,
1560 and although they often accept just C<"\012">, they seldom tolerate just
1561 C<"\015">. If you get in the habit of using C<"\n"> for networking,
1562 you may be burned some day.
1563 X<newline> X<line terminator> X<eol> X<end of line>
1566 For constructs that do interpolate, variables beginning with "C<$>"
1567 or "C<@>" are interpolated. Subscripted variables such as C<$a[3]> or
1568 C<< $href->{key}[0] >> are also interpolated, as are array and hash slices.
1569 But method calls such as C<< $obj->meth >> are not.
1571 Interpolating an array or slice interpolates the elements in order,
1572 separated by the value of C<$">, so is equivalent to interpolating
1573 C<join $", @array>. "Punctuation" arrays such as C<@*> are usually
1574 interpolated only if the name is enclosed in braces C<@{*}>, but the
1575 arrays C<@_>, C<@+>, and C<@-> are interpolated even without braces.
1577 For double-quoted strings, the quoting from C<\Q> is applied after
1578 interpolation and escapes are processed.
1580 "abc\Qfoo\tbar$s\Exyz"
1584 "abc" . quotemeta("foo\tbar$s") . "xyz"
1586 For the pattern of regex operators (C<qr//>, C<m//> and C<s///>),
1587 the quoting from C<\Q> is applied after interpolation is processed,
1588 but before escapes are processed. This allows the pattern to match
1589 literally (except for C<$> and C<@>). For example, the following matches:
1593 Because C<$> or C<@> trigger interpolation, you'll need to use something
1594 like C</\Quser\E\@\Qhost/> to match them literally.
1596 Patterns are subject to an additional level of interpretation as a
1597 regular expression. This is done as a second pass, after variables are
1598 interpolated, so that regular expressions may be incorporated into the
1599 pattern from the variables. If this is not what you want, use C<\Q> to
1600 interpolate a variable literally.
1602 Apart from the behavior described above, Perl does not expand
1603 multiple levels of interpolation. In particular, contrary to the
1604 expectations of shell programmers, back-quotes do I<NOT> interpolate
1605 within double quotes, nor do single quotes impede evaluation of
1606 variables when used within double quotes.
1608 =head2 Regexp Quote-Like Operators
1611 Here are the quote-like operators that apply to pattern
1612 matching and related activities.
1616 =item qr/STRING/msixpodual
1617 X<qr> X</i> X</m> X</o> X</s> X</x> X</p>
1619 This operator quotes (and possibly compiles) its I<STRING> as a regular
1620 expression. I<STRING> is interpolated the same way as I<PATTERN>
1621 in C<m/PATTERN/>. If "'" is used as the delimiter, no interpolation
1622 is done. Returns a Perl value which may be used instead of the
1623 corresponding C</STRING/msixpodual> expression. The returned value is a
1624 normalized version of the original pattern. It magically differs from
1625 a string containing the same characters: C<ref(qr/x/)> returns "Regexp";
1626 however, dereferencing it is not well defined (you currently get the
1627 normalized version of the original pattern, but this may change).
1632 $rex = qr/my.STRING/is;
1633 print $rex; # prints (?si-xm:my.STRING)
1640 The result may be used as a subpattern in a match:
1643 $string =~ /foo${re}bar/; # can be interpolated in other
1645 $string =~ $re; # or used standalone
1646 $string =~ /$re/; # or this way
1648 Since Perl may compile the pattern at the moment of execution of the qr()
1649 operator, using qr() may have speed advantages in some situations,
1650 notably if the result of qr() is used standalone:
1653 my $patterns = shift;
1654 my @compiled = map qr/$_/i, @$patterns;
1657 foreach my $pat (@compiled) {
1658 $success = 1, last if /$pat/;
1664 Precompilation of the pattern into an internal representation at
1665 the moment of qr() avoids a need to recompile the pattern every
1666 time a match C</$pat/> is attempted. (Perl has many other internal
1667 optimizations, but none would be triggered in the above example if
1668 we did not use qr() operator.)
1670 Options (specified by the following modifiers) are:
1672 m Treat string as multiple lines.
1673 s Treat string as single line. (Make . match a newline)
1674 i Do case-insensitive pattern matching.
1675 x Use extended regular expressions.
1676 p When matching preserve a copy of the matched string so
1677 that ${^PREMATCH}, ${^MATCH}, ${^POSTMATCH} will be
1679 o Compile pattern only once.
1680 a ASCII-restrict: Use ASCII for \d, \s, \w; specifying two
1681 a's further restricts /i matching so that no ASCII
1682 character will match a non-ASCII one.
1684 u Use Unicode rules.
1685 d Use Unicode or native charset, as in 5.12 and earlier.
1687 If a precompiled pattern is embedded in a larger pattern then the effect
1688 of "msixpluad" will be propagated appropriately. The effect the "o"
1689 modifier has is not propagated, being restricted to those patterns
1690 explicitly using it.
1692 The last four modifiers listed above, added in Perl 5.14,
1693 control the character set semantics, but C</a> is the only one you are likely
1694 to want to specify explicitly; the other three are selected
1695 automatically by various pragmas.
1697 See L<perlre> for additional information on valid syntax for STRING, and
1698 for a detailed look at the semantics of regular expressions. In
1699 particular, all modifiers except the largely obsolete C</o> are further
1700 explained in L<perlre/Modifiers>. C</o> is described in the next section.
1702 =item m/PATTERN/msixpodualgc
1703 X<m> X<operator, match>
1704 X<regexp, options> X<regexp> X<regex, options> X<regex>
1705 X</m> X</s> X</i> X</x> X</p> X</o> X</g> X</c>
1707 =item /PATTERN/msixpodualgc
1709 Searches a string for a pattern match, and in scalar context returns
1710 true if it succeeds, false if it fails. If no string is specified
1711 via the C<=~> or C<!~> operator, the $_ string is searched. (The
1712 string specified with C<=~> need not be an lvalue--it may be the
1713 result of an expression evaluation, but remember the C<=~> binds
1714 rather tightly.) See also L<perlre>.
1716 Options are as described in C<qr//> above; in addition, the following match
1717 process modifiers are available:
1719 g Match globally, i.e., find all occurrences.
1720 c Do not reset search position on a failed match when /g is
1723 If "/" is the delimiter then the initial C<m> is optional. With the C<m>
1724 you can use any pair of non-whitespace (ASCII) characters
1725 as delimiters. This is particularly useful for matching path names
1726 that contain "/", to avoid LTS (leaning toothpick syndrome). If "?" is
1727 the delimiter, then a match-only-once rule applies,
1728 described in C<m?PATTERN?> below. If "'" (single quote) is the delimiter,
1729 no interpolation is performed on the PATTERN.
1730 When using a character valid in an identifier, whitespace is required
1733 PATTERN may contain variables, which will be interpolated
1734 every time the pattern search is evaluated, except
1735 for when the delimiter is a single quote. (Note that C<$(>, C<$)>, and
1736 C<$|> are not interpolated because they look like end-of-string tests.)
1737 Perl will not recompile the pattern unless an interpolated
1738 variable that it contains changes. You can force Perl to skip the
1739 test and never recompile by adding a C</o> (which stands for "once")
1740 after the trailing delimiter.
1741 Once upon a time, Perl would recompile regular expressions
1742 unnecessarily, and this modifier was useful to tell it not to do so, in the
1743 interests of speed. But now, the only reasons to use C</o> are either:
1749 The variables are thousands of characters long and you know that they
1750 don't change, and you need to wring out the last little bit of speed by
1751 having Perl skip testing for that. (There is a maintenance penalty for
1752 doing this, as mentioning C</o> constitutes a promise that you won't
1753 change the variables in the pattern. If you do change them, Perl won't
1758 you want the pattern to use the initial values of the variables
1759 regardless of whether they change or not. (But there are saner ways
1760 of accomplishing this than using C</o>.)
1764 If the pattern contains embedded code, such as
1767 $code = 'foo(?{ $x })';
1770 then perl will recompile each time, even though the pattern string hasn't
1771 changed, to ensure that the current value of C<$x> is seen each time.
1772 Use C</o> if you want to avoid this.
1776 The bottom line is that using C</o> is almost never a good idea.
1778 =item The empty pattern //
1780 If the PATTERN evaluates to the empty string, the last
1781 I<successfully> matched regular expression is used instead. In this
1782 case, only the C<g> and C<c> flags on the empty pattern are honored;
1783 the other flags are taken from the original pattern. If no match has
1784 previously succeeded, this will (silently) act instead as a genuine
1785 empty pattern (which will always match).
1787 Note that it's possible to confuse Perl into thinking C<//> (the empty
1788 regex) is really C<//> (the defined-or operator). Perl is usually pretty
1789 good about this, but some pathological cases might trigger this, such as
1790 C<$a///> (is that C<($a) / (//)> or C<$a // />?) and C<print $fh //>
1791 (C<print $fh(//> or C<print($fh //>?). In all of these examples, Perl
1792 will assume you meant defined-or. If you meant the empty regex, just
1793 use parentheses or spaces to disambiguate, or even prefix the empty
1794 regex with an C<m> (so C<//> becomes C<m//>).
1796 =item Matching in list context
1798 If the C</g> option is not used, C<m//> in list context returns a
1799 list consisting of the subexpressions matched by the parentheses in the
1800 pattern, that is, (C<$1>, C<$2>, C<$3>...) (Note that here C<$1> etc. are
1801 also set). When there are no parentheses in the pattern, the return
1802 value is the list C<(1)> for success.
1803 With or without parentheses, an empty list is returned upon failure.
1807 open(TTY, "+</dev/tty")
1808 || die "can't access /dev/tty: $!";
1810 <TTY> =~ /^y/i && foo(); # do foo if desired
1812 if (/Version: *([0-9.]*)/) { $version = $1; }
1814 next if m#^/usr/spool/uucp#;
1819 print if /$arg/o; # compile only once (no longer needed!)
1822 if (($F1, $F2, $Etc) = ($foo =~ /^(\S+)\s+(\S+)\s*(.*)/))
1824 This last example splits $foo into the first two words and the
1825 remainder of the line, and assigns those three fields to $F1, $F2, and
1826 $Etc. The conditional is true if any variables were assigned; that is,
1827 if the pattern matched.
1829 The C</g> modifier specifies global pattern matching--that is,
1830 matching as many times as possible within the string. How it behaves
1831 depends on the context. In list context, it returns a list of the
1832 substrings matched by any capturing parentheses in the regular
1833 expression. If there are no parentheses, it returns a list of all
1834 the matched strings, as if there were parentheses around the whole
1837 In scalar context, each execution of C<m//g> finds the next match,
1838 returning true if it matches, and false if there is no further match.
1839 The position after the last match can be read or set using the C<pos()>
1840 function; see L<perlfunc/pos>. A failed match normally resets the
1841 search position to the beginning of the string, but you can avoid that
1842 by adding the C</c> modifier (for example, C<m//gc>). Modifying the target
1843 string also resets the search position.
1847 You can intermix C<m//g> matches with C<m/\G.../g>, where C<\G> is a
1848 zero-width assertion that matches the exact position where the
1849 previous C<m//g>, if any, left off. Without the C</g> modifier, the
1850 C<\G> assertion still anchors at C<pos()> as it was at the start of
1851 the operation (see L<perlfunc/pos>), but the match is of course only
1852 attempted once. Using C<\G> without C</g> on a target string that has
1853 not previously had a C</g> match applied to it is the same as using
1854 the C<\A> assertion to match the beginning of the string. Note also
1855 that, currently, C<\G> is only properly supported when anchored at the
1856 very beginning of the pattern.
1861 ($one,$five,$fifteen) = (`uptime` =~ /(\d+\.\d+)/g);
1865 while ($paragraph = <>) {
1866 while ($paragraph =~ /\p{Ll}['")]*[.!?]+['")]*\s/g) {
1872 Here's another way to check for sentences in a paragraph:
1874 my $sentence_rx = qr{
1875 (?: (?<= ^ ) | (?<= \s ) ) # after start-of-string or
1877 \p{Lu} # capital letter
1878 .*? # a bunch of anything
1879 (?<= \S ) # that ends in non-
1881 (?<! \b [DMS]r ) # but isn't a common abbr.
1885 [.?!] # followed by a sentence
1887 (?= $ | \s ) # in front of end-of-string
1891 while (my $paragraph = <>) {
1892 say "NEW PARAGRAPH";
1894 while ($paragraph =~ /($sentence_rx)/g) {
1895 printf "\tgot sentence %d: <%s>\n", ++$count, $1;
1899 Here's how to use C<m//gc> with C<\G>:
1904 print $1 while /(o)/gc; print "', pos=", pos, "\n";
1906 print $1 if /\G(q)/gc; print "', pos=", pos, "\n";
1908 print $1 while /(p)/gc; print "', pos=", pos, "\n";
1910 print "Final: '$1', pos=",pos,"\n" if /\G(.)/;
1912 The last example should print:
1922 Notice that the final match matched C<q> instead of C<p>, which a match
1923 without the C<\G> anchor would have done. Also note that the final match
1924 did not update C<pos>. C<pos> is only updated on a C</g> match. If the
1925 final match did indeed match C<p>, it's a good bet that you're running a
1926 very old (pre-5.6.0) version of Perl.
1928 A useful idiom for C<lex>-like scanners is C</\G.../gc>. You can
1929 combine several regexps like this to process a string part-by-part,
1930 doing different actions depending on which regexp matched. Each
1931 regexp tries to match where the previous one leaves off.
1934 $url = URI::URL->new( "http://example.com/" );
1935 die if $url eq "xXx";
1939 print(" digits"), redo LOOP if /\G\d+\b[,.;]?\s*/gc;
1940 print(" lowercase"), redo LOOP
1941 if /\G\p{Ll}+\b[,.;]?\s*/gc;
1942 print(" UPPERCASE"), redo LOOP
1943 if /\G\p{Lu}+\b[,.;]?\s*/gc;
1944 print(" Capitalized"), redo LOOP
1945 if /\G\p{Lu}\p{Ll}+\b[,.;]?\s*/gc;
1946 print(" MiXeD"), redo LOOP if /\G\pL+\b[,.;]?\s*/gc;
1947 print(" alphanumeric"), redo LOOP
1948 if /\G[\p{Alpha}\pN]+\b[,.;]?\s*/gc;
1949 print(" line-noise"), redo LOOP if /\G\W+/gc;
1950 print ". That's all!\n";
1953 Here is the output (split into several lines):
1955 line-noise lowercase line-noise UPPERCASE line-noise UPPERCASE
1956 line-noise lowercase line-noise lowercase line-noise lowercase
1957 lowercase line-noise lowercase lowercase line-noise lowercase
1958 lowercase line-noise MiXeD line-noise. That's all!
1960 =item m?PATTERN?msixpodualgc
1961 X<?> X<operator, match-once>
1963 =item ?PATTERN?msixpodualgc
1965 This is just like the C<m/PATTERN/> search, except that it matches
1966 only once between calls to the reset() operator. This is a useful
1967 optimization when you want to see only the first occurrence of
1968 something in each file of a set of files, for instance. Only C<m??>
1969 patterns local to the current package are reset.
1973 # blank line between header and body
1976 reset if eof; # clear m?? status for next file
1979 Another example switched the first "latin1" encoding it finds
1980 to "utf8" in a pod file:
1982 s//utf8/ if m? ^ =encoding \h+ \K latin1 ?x;
1984 The match-once behavior is controlled by the match delimiter being
1985 C<?>; with any other delimiter this is the normal C<m//> operator.
1987 For historical reasons, the leading C<m> in C<m?PATTERN?> is optional,
1988 but the resulting C<?PATTERN?> syntax is deprecated, will warn on
1989 usage and might be removed from a future stable release of Perl (without
1992 =item s/PATTERN/REPLACEMENT/msixpodualgcer
1993 X<substitute> X<substitution> X<replace> X<regexp, replace>
1994 X<regexp, substitute> X</m> X</s> X</i> X</x> X</p> X</o> X</g> X</c> X</e> X</r>
1996 Searches a string for a pattern, and if found, replaces that pattern
1997 with the replacement text and returns the number of substitutions
1998 made. Otherwise it returns false (specifically, the empty string).
2000 If the C</r> (non-destructive) option is used then it runs the
2001 substitution on a copy of the string and instead of returning the
2002 number of substitutions, it returns the copy whether or not a
2003 substitution occurred. The original string is never changed when
2004 C</r> is used. The copy will always be a plain string, even if the
2005 input is an object or a tied variable.
2007 If no string is specified via the C<=~> or C<!~> operator, the C<$_>
2008 variable is searched and modified. Unless the C</r> option is used,
2009 the string specified must be a scalar variable, an array element, a
2010 hash element, or an assignment to one of those; that is, some sort of
2013 If the delimiter chosen is a single quote, no interpolation is
2014 done on either the PATTERN or the REPLACEMENT. Otherwise, if the
2015 PATTERN contains a $ that looks like a variable rather than an
2016 end-of-string test, the variable will be interpolated into the pattern
2017 at run-time. If you want the pattern compiled only once the first time
2018 the variable is interpolated, use the C</o> option. If the pattern
2019 evaluates to the empty string, the last successfully executed regular
2020 expression is used instead. See L<perlre> for further explanation on these.
2022 Options are as with m// with the addition of the following replacement
2025 e Evaluate the right side as an expression.
2026 ee Evaluate the right side as a string then eval the
2028 r Return substitution and leave the original string
2031 Any non-whitespace delimiter may replace the slashes. Add space after
2032 the C<s> when using a character allowed in identifiers. If single quotes
2033 are used, no interpretation is done on the replacement string (the C</e>
2034 modifier overrides this, however). Note that Perl treats backticks
2035 as normal delimiters; the replacement text is not evaluated as a command.
2036 If the PATTERN is delimited by bracketing quotes, the REPLACEMENT has
2037 its own pair of quotes, which may or may not be bracketing quotes, for example,
2038 C<s(foo)(bar)> or C<< s<foo>/bar/ >>. A C</e> will cause the
2039 replacement portion to be treated as a full-fledged Perl expression
2040 and evaluated right then and there. It is, however, syntax checked at
2041 compile-time. A second C<e> modifier will cause the replacement portion
2042 to be C<eval>ed before being run as a Perl expression.
2046 s/\bgreen\b/mauve/g; # don't change wintergreen
2048 $path =~ s|/usr/bin|/usr/local/bin|;
2050 s/Login: $foo/Login: $bar/; # run-time pattern
2052 ($foo = $bar) =~ s/this/that/; # copy first, then
2054 ($foo = "$bar") =~ s/this/that/; # convert to string,
2056 $foo = $bar =~ s/this/that/r; # Same as above using /r
2057 $foo = $bar =~ s/this/that/r
2058 =~ s/that/the other/r; # Chained substitutes
2060 @foo = map { s/this/that/r } @bar # /r is very useful in
2063 $count = ($paragraph =~ s/Mister\b/Mr./g); # get change-cnt
2066 s/\d+/$&*2/e; # yields 'abc246xyz'
2067 s/\d+/sprintf("%5d",$&)/e; # yields 'abc 246xyz'
2068 s/\w/$& x 2/eg; # yields 'aabbcc 224466xxyyzz'
2070 s/%(.)/$percent{$1}/g; # change percent escapes; no /e
2071 s/%(.)/$percent{$1} || $&/ge; # expr now, so /e
2072 s/^=(\w+)/pod($1)/ge; # use function call
2075 $a = s/abc/def/r; # $a is 'def123xyz' and
2076 # $_ remains 'abc123xyz'.
2078 # expand variables in $_, but dynamics only, using
2079 # symbolic dereferencing
2082 # Add one to the value of any numbers in the string
2085 # Titlecase words in the last 30 characters only
2086 substr($str, -30) =~ s/\b(\p{Alpha}+)\b/\u\L$1/g;
2088 # This will expand any embedded scalar variable
2089 # (including lexicals) in $_ : First $1 is interpolated
2090 # to the variable name, and then evaluated
2093 # Delete (most) C comments.
2095 /\* # Match the opening delimiter.
2096 .*? # Match a minimal number of characters.
2097 \*/ # Match the closing delimiter.
2100 s/^\s*(.*?)\s*$/$1/; # trim whitespace in $_,
2103 for ($variable) { # trim whitespace in $variable,
2109 s/([^ ]*) *([^ ]*)/$2 $1/; # reverse 1st two fields
2111 Note the use of $ instead of \ in the last example. Unlike
2112 B<sed>, we use the \<I<digit>> form in only the left hand side.
2113 Anywhere else it's $<I<digit>>.
2115 Occasionally, you can't use just a C</g> to get all the changes
2116 to occur that you might want. Here are two common cases:
2118 # put commas in the right places in an integer
2119 1 while s/(\d)(\d\d\d)(?!\d)/$1,$2/g;
2121 # expand tabs to 8-column spacing
2122 1 while s/\t+/' ' x (length($&)*8 - length($`)%8)/e;
2126 =head2 Quote-Like Operators
2127 X<operator, quote-like>
2132 X<q> X<quote, single> X<'> X<''>
2136 A single-quoted, literal string. A backslash represents a backslash
2137 unless followed by the delimiter or another backslash, in which case
2138 the delimiter or backslash is interpolated.
2140 $foo = q!I said, "You said, 'She said it.'"!;
2141 $bar = q('This is it.');
2142 $baz = '\n'; # a two-character string
2145 X<qq> X<quote, double> X<"> X<"">
2149 A double-quoted, interpolated string.
2152 (*** The previous line contains the naughty word "$1".\n)
2153 if /\b(tcl|java|python)\b/i; # :-)
2154 $baz = "\n"; # a one-character string
2157 X<qx> X<`> X<``> X<backtick>
2161 A string which is (possibly) interpolated and then executed as a
2162 system command with F</bin/sh> or its equivalent. Shell wildcards,
2163 pipes, and redirections will be honored. The collected standard
2164 output of the command is returned; standard error is unaffected. In
2165 scalar context, it comes back as a single (potentially multi-line)
2166 string, or undef if the command failed. In list context, returns a
2167 list of lines (however you've defined lines with $/ or
2168 $INPUT_RECORD_SEPARATOR), or an empty list if the command failed.
2170 Because backticks do not affect standard error, use shell file descriptor
2171 syntax (assuming the shell supports this) if you care to address this.
2172 To capture a command's STDERR and STDOUT together:
2174 $output = `cmd 2>&1`;
2176 To capture a command's STDOUT but discard its STDERR:
2178 $output = `cmd 2>/dev/null`;
2180 To capture a command's STDERR but discard its STDOUT (ordering is
2183 $output = `cmd 2>&1 1>/dev/null`;
2185 To exchange a command's STDOUT and STDERR in order to capture the STDERR
2186 but leave its STDOUT to come out the old STDERR:
2188 $output = `cmd 3>&1 1>&2 2>&3 3>&-`;
2190 To read both a command's STDOUT and its STDERR separately, it's easiest
2191 to redirect them separately to files, and then read from those files
2192 when the program is done:
2194 system("program args 1>program.stdout 2>program.stderr");
2196 The STDIN filehandle used by the command is inherited from Perl's STDIN.
2199 open(SPLAT, "stuff") || die "can't open stuff: $!";
2200 open(STDIN, "<&SPLAT") || die "can't dupe SPLAT: $!";
2201 print STDOUT `sort`;
2203 will print the sorted contents of the file named F<"stuff">.
2205 Using single-quote as a delimiter protects the command from Perl's
2206 double-quote interpolation, passing it on to the shell instead:
2208 $perl_info = qx(ps $$); # that's Perl's $$
2209 $shell_info = qx'ps $$'; # that's the new shell's $$
2211 How that string gets evaluated is entirely subject to the command
2212 interpreter on your system. On most platforms, you will have to protect
2213 shell metacharacters if you want them treated literally. This is in
2214 practice difficult to do, as it's unclear how to escape which characters.
2215 See L<perlsec> for a clean and safe example of a manual fork() and exec()
2216 to emulate backticks safely.
2218 On some platforms (notably DOS-like ones), the shell may not be
2219 capable of dealing with multiline commands, so putting newlines in
2220 the string may not get you what you want. You may be able to evaluate
2221 multiple commands in a single line by separating them with the command
2222 separator character, if your shell supports that (for example, C<;> on
2223 many Unix shells and C<&> on the Windows NT C<cmd> shell).
2225 Perl will attempt to flush all files opened for
2226 output before starting the child process, but this may not be supported
2227 on some platforms (see L<perlport>). To be safe, you may need to set
2228 C<$|> ($AUTOFLUSH in English) or call the C<autoflush()> method of
2229 C<IO::Handle> on any open handles.
2231 Beware that some command shells may place restrictions on the length
2232 of the command line. You must ensure your strings don't exceed this
2233 limit after any necessary interpolations. See the platform-specific
2234 release notes for more details about your particular environment.
2236 Using this operator can lead to programs that are difficult to port,
2237 because the shell commands called vary between systems, and may in
2238 fact not be present at all. As one example, the C<type> command under
2239 the POSIX shell is very different from the C<type> command under DOS.
2240 That doesn't mean you should go out of your way to avoid backticks
2241 when they're the right way to get something done. Perl was made to be
2242 a glue language, and one of the things it glues together is commands.
2243 Just understand what you're getting yourself into.
2245 See L</"I/O Operators"> for more discussion.
2248 X<qw> X<quote, list> X<quote, words>
2250 Evaluates to a list of the words extracted out of STRING, using embedded
2251 whitespace as the word delimiters. It can be understood as being roughly
2254 split(" ", q/STRING/);
2256 the differences being that it generates a real list at compile time, and
2257 in scalar context it returns the last element in the list. So
2262 is semantically equivalent to the list:
2266 Some frequently seen examples:
2268 use POSIX qw( setlocale localeconv )
2269 @EXPORT = qw( foo bar baz );
2271 A common mistake is to try to separate the words with comma or to
2272 put comments into a multi-line C<qw>-string. For this reason, the
2273 C<use warnings> pragma and the B<-w> switch (that is, the C<$^W> variable)
2274 produces warnings if the STRING contains the "," or the "#" character.
2276 =item tr/SEARCHLIST/REPLACEMENTLIST/cdsr
2277 X<tr> X<y> X<transliterate> X</c> X</d> X</s>
2279 =item y/SEARCHLIST/REPLACEMENTLIST/cdsr
2281 Transliterates all occurrences of the characters found in the search list
2282 with the corresponding character in the replacement list. It returns
2283 the number of characters replaced or deleted. If no string is
2284 specified via the C<=~> or C<!~> operator, the $_ string is transliterated.
2286 If the C</r> (non-destructive) option is present, a new copy of the string
2287 is made and its characters transliterated, and this copy is returned no
2288 matter whether it was modified or not: the original string is always
2289 left unchanged. The new copy is always a plain string, even if the input
2290 string is an object or a tied variable.
2292 Unless the C</r> option is used, the string specified with C<=~> must be a
2293 scalar variable, an array element, a hash element, or an assignment to one
2294 of those; in other words, an lvalue.
2296 A character range may be specified with a hyphen, so C<tr/A-J/0-9/>
2297 does the same replacement as C<tr/ACEGIBDFHJ/0246813579/>.
2298 For B<sed> devotees, C<y> is provided as a synonym for C<tr>. If the
2299 SEARCHLIST is delimited by bracketing quotes, the REPLACEMENTLIST has
2300 its own pair of quotes, which may or may not be bracketing quotes;
2301 for example, C<tr[aeiouy][yuoiea]> or C<tr(+\-*/)/ABCD/>.
2303 Note that C<tr> does B<not> do regular expression character classes such as
2304 C<\d> or C<\pL>. The C<tr> operator is not equivalent to the tr(1)
2305 utility. If you want to map strings between lower/upper cases, see
2306 L<perlfunc/lc> and L<perlfunc/uc>, and in general consider using the C<s>
2307 operator if you need regular expressions. The C<\U>, C<\u>, C<\L>, and
2308 C<\l> string-interpolation escapes on the right side of a substitution
2309 operator will perform correct case-mappings, but C<tr[a-z][A-Z]> will not
2310 (except sometimes on legacy 7-bit data).
2312 Note also that the whole range idea is rather unportable between
2313 character sets--and even within character sets they may cause results
2314 you probably didn't expect. A sound principle is to use only ranges
2315 that begin from and end at either alphabets of equal case (a-e, A-E),
2316 or digits (0-4). Anything else is unsafe. If in doubt, spell out the
2317 character sets in full.
2321 c Complement the SEARCHLIST.
2322 d Delete found but unreplaced characters.
2323 s Squash duplicate replaced characters.
2324 r Return the modified string and leave the original string
2327 If the C</c> modifier is specified, the SEARCHLIST character set
2328 is complemented. If the C</d> modifier is specified, any characters
2329 specified by SEARCHLIST not found in REPLACEMENTLIST are deleted.
2330 (Note that this is slightly more flexible than the behavior of some
2331 B<tr> programs, which delete anything they find in the SEARCHLIST,
2332 period.) If the C</s> modifier is specified, sequences of characters
2333 that were transliterated to the same character are squashed down
2334 to a single instance of the character.
2336 If the C</d> modifier is used, the REPLACEMENTLIST is always interpreted
2337 exactly as specified. Otherwise, if the REPLACEMENTLIST is shorter
2338 than the SEARCHLIST, the final character is replicated till it is long
2339 enough. If the REPLACEMENTLIST is empty, the SEARCHLIST is replicated.
2340 This latter is useful for counting characters in a class or for
2341 squashing character sequences in a class.
2345 $ARGV[1] =~ tr/A-Z/a-z/; # canonicalize to lower case ASCII
2347 $cnt = tr/*/*/; # count the stars in $_
2349 $cnt = $sky =~ tr/*/*/; # count the stars in $sky
2351 $cnt = tr/0-9//; # count the digits in $_
2353 tr/a-zA-Z//s; # bookkeeper -> bokeper
2355 ($HOST = $host) =~ tr/a-z/A-Z/;
2356 $HOST = $host =~ tr/a-z/A-Z/r; # same thing
2358 $HOST = $host =~ tr/a-z/A-Z/r # chained with s///r
2361 tr/a-zA-Z/ /cs; # change non-alphas to single space
2363 @stripped = map tr/a-zA-Z/ /csr, @original;
2367 [\000-\177]; # wickedly delete 8th bit
2369 If multiple transliterations are given for a character, only the
2374 will transliterate any A to X.
2376 Because the transliteration table is built at compile time, neither
2377 the SEARCHLIST nor the REPLACEMENTLIST are subjected to double quote
2378 interpolation. That means that if you want to use variables, you
2381 eval "tr/$oldlist/$newlist/";
2384 eval "tr/$oldlist/$newlist/, 1" or die $@;
2387 X<here-doc> X<heredoc> X<here-document> X<<< << >>>
2389 A line-oriented form of quoting is based on the shell "here-document"
2390 syntax. Following a C<< << >> you specify a string to terminate
2391 the quoted material, and all lines following the current line down to
2392 the terminating string are the value of the item.
2394 The terminating string may be either an identifier (a word), or some
2395 quoted text. An unquoted identifier works like double quotes.
2396 There may not be a space between the C<< << >> and the identifier,
2397 unless the identifier is explicitly quoted. (If you put a space it
2398 will be treated as a null identifier, which is valid, and matches the
2399 first empty line.) The terminating string must appear by itself
2400 (unquoted and with no surrounding whitespace) on the terminating line.
2402 If the terminating string is quoted, the type of quotes used determine
2403 the treatment of the text.
2409 Double quotes indicate that the text will be interpolated using exactly
2410 the same rules as normal double quoted strings.
2413 The price is $Price.
2416 print << "EOF"; # same as above
2417 The price is $Price.
2423 Single quotes indicate the text is to be treated literally with no
2424 interpolation of its content. This is similar to single quoted
2425 strings except that backslashes have no special meaning, with C<\\>
2426 being treated as two backslashes and not one as they would in every
2427 other quoting construct.
2429 Just as in the shell, a backslashed bareword following the C<<< << >>>
2430 means the same thing as a single-quoted string does:
2432 $cost = <<'VISTA'; # hasta la ...
2433 That'll be $10 please, ma'am.
2436 $cost = <<\VISTA; # Same thing!
2437 That'll be $10 please, ma'am.
2440 This is the only form of quoting in perl where there is no need
2441 to worry about escaping content, something that code generators
2442 can and do make good use of.
2446 The content of the here doc is treated just as it would be if the
2447 string were embedded in backticks. Thus the content is interpolated
2448 as though it were double quoted and then executed via the shell, with
2449 the results of the execution returned.
2451 print << `EOC`; # execute command and get results
2457 It is possible to stack multiple here-docs in a row:
2459 print <<"foo", <<"bar"; # you can stack them
2465 myfunc(<< "THIS", 23, <<'THAT');
2472 Just don't forget that you have to put a semicolon on the end
2473 to finish the statement, as Perl doesn't know you're not going to
2481 If you want to remove the line terminator from your here-docs,
2484 chomp($string = <<'END');
2488 If you want your here-docs to be indented with the rest of the code,
2489 you'll need to remove leading whitespace from each line manually:
2491 ($quote = <<'FINIS') =~ s/^\s+//gm;
2492 The Road goes ever on and on,
2493 down from the door where it began.
2496 If you use a here-doc within a delimited construct, such as in C<s///eg>,
2497 the quoted material must still come on the line following the
2498 C<<< <<FOO >>> marker, which means it may be inside the delimited
2506 It works this way as of Perl 5.18. Historically, it was inconsistent, and
2507 you would have to write
2514 outside of string evals.
2516 Additionally, quoting rules for the end-of-string identifier are
2517 unrelated to Perl's quoting rules. C<q()>, C<qq()>, and the like are not
2518 supported in place of C<''> and C<"">, and the only interpolation is for
2519 backslashing the quoting character:
2521 print << "abc\"def";
2525 Finally, quoted strings cannot span multiple lines. The general rule is
2526 that the identifier must be a string literal. Stick with that, and you
2531 =head2 Gory details of parsing quoted constructs
2532 X<quote, gory details>
2534 When presented with something that might have several different
2535 interpretations, Perl uses the B<DWIM> (that's "Do What I Mean")
2536 principle to pick the most probable interpretation. This strategy
2537 is so successful that Perl programmers often do not suspect the
2538 ambivalence of what they write. But from time to time, Perl's
2539 notions differ substantially from what the author honestly meant.
2541 This section hopes to clarify how Perl handles quoted constructs.
2542 Although the most common reason to learn this is to unravel labyrinthine
2543 regular expressions, because the initial steps of parsing are the
2544 same for all quoting operators, they are all discussed together.
2546 The most important Perl parsing rule is the first one discussed
2547 below: when processing a quoted construct, Perl first finds the end
2548 of that construct, then interprets its contents. If you understand
2549 this rule, you may skip the rest of this section on the first
2550 reading. The other rules are likely to contradict the user's
2551 expectations much less frequently than this first one.
2553 Some passes discussed below are performed concurrently, but because
2554 their results are the same, we consider them individually. For different
2555 quoting constructs, Perl performs different numbers of passes, from
2556 one to four, but these passes are always performed in the same order.
2560 =item Finding the end
2562 The first pass is finding the end of the quoted construct, where
2563 the information about the delimiters is used in parsing.
2564 During this search, text between the starting and ending delimiters
2565 is copied to a safe location. The text copied gets delimiter-independent.
2567 If the construct is a here-doc, the ending delimiter is a line
2568 that has a terminating string as the content. Therefore C<<<EOF> is
2569 terminated by C<EOF> immediately followed by C<"\n"> and starting
2570 from the first column of the terminating line.
2571 When searching for the terminating line of a here-doc, nothing
2572 is skipped. In other words, lines after the here-doc syntax
2573 are compared with the terminating string line by line.
2575 For the constructs except here-docs, single characters are used as starting
2576 and ending delimiters. If the starting delimiter is an opening punctuation
2577 (that is C<(>, C<[>, C<{>, or C<< < >>), the ending delimiter is the
2578 corresponding closing punctuation (that is C<)>, C<]>, C<}>, or C<< > >>).
2579 If the starting delimiter is an unpaired character like C</> or a closing
2580 punctuation, the ending delimiter is same as the starting delimiter.
2581 Therefore a C</> terminates a C<qq//> construct, while a C<]> terminates
2582 C<qq[]> and C<qq]]> constructs.
2584 When searching for single-character delimiters, escaped delimiters
2585 and C<\\> are skipped. For example, while searching for terminating C</>,
2586 combinations of C<\\> and C<\/> are skipped. If the delimiters are
2587 bracketing, nested pairs are also skipped. For example, while searching
2588 for closing C<]> paired with the opening C<[>, combinations of C<\\>, C<\]>,
2589 and C<\[> are all skipped, and nested C<[> and C<]> are skipped as well.
2590 However, when backslashes are used as the delimiters (like C<qq\\> and
2591 C<tr\\\>), nothing is skipped.
2592 During the search for the end, backslashes that escape delimiters or
2593 other backslashes are removed (exactly speaking, they are not copied to the
2596 For constructs with three-part delimiters (C<s///>, C<y///>, and
2597 C<tr///>), the search is repeated once more.
2598 If the first delimiter is not an opening punctuation, three delimiters must
2599 be same such as C<s!!!> and C<tr)))>, in which case the second delimiter
2600 terminates the left part and starts the right part at once.
2601 If the left part is delimited by bracketing punctuation (that is C<()>,
2602 C<[]>, C<{}>, or C<< <> >>), the right part needs another pair of
2603 delimiters such as C<s(){}> and C<tr[]//>. In these cases, whitespace
2604 and comments are allowed between both parts, though the comment must follow
2605 at least one whitespace character; otherwise a character expected as the
2606 start of the comment may be regarded as the starting delimiter of the right part.
2608 During this search no attention is paid to the semantics of the construct.
2611 "$hash{"$foo/$bar"}"
2616 bar # NOT a comment, this slash / terminated m//!
2619 do not form legal quoted expressions. The quoted part ends on the
2620 first C<"> and C</>, and the rest happens to be a syntax error.
2621 Because the slash that terminated C<m//> was followed by a C<SPACE>,
2622 the example above is not C<m//x>, but rather C<m//> with no C</x>
2623 modifier. So the embedded C<#> is interpreted as a literal C<#>.
2625 Also no attention is paid to C<\c\> (multichar control char syntax) during
2626 this search. Thus the second C<\> in C<qq/\c\/> is interpreted as a part
2627 of C<\/>, and the following C</> is not recognized as a delimiter.
2628 Instead, use C<\034> or C<\x1c> at the end of quoted constructs.
2633 The next step is interpolation in the text obtained, which is now
2634 delimiter-independent. There are multiple cases.
2640 No interpolation is performed.
2641 Note that the combination C<\\> is left intact, since escaped delimiters
2642 are not available for here-docs.
2644 =item C<m''>, the pattern of C<s'''>
2646 No interpolation is performed at this stage.
2647 Any backslashed sequences including C<\\> are treated at the stage
2648 to L</"parsing regular expressions">.
2650 =item C<''>, C<q//>, C<tr'''>, C<y'''>, the replacement of C<s'''>
2652 The only interpolation is removal of C<\> from pairs of C<\\>.
2653 Therefore C<-> in C<tr'''> and C<y'''> is treated literally
2654 as a hyphen and no character range is available.
2655 C<\1> in the replacement of C<s'''> does not work as C<$1>.
2657 =item C<tr///>, C<y///>
2659 No variable interpolation occurs. String modifying combinations for
2660 case and quoting such as C<\Q>, C<\U>, and C<\E> are not recognized.
2661 The other escape sequences such as C<\200> and C<\t> and backslashed
2662 characters such as C<\\> and C<\-> are converted to appropriate literals.
2663 The character C<-> is treated specially and therefore C<\-> is treated
2666 =item C<"">, C<``>, C<qq//>, C<qx//>, C<< <file*glob> >>, C<<<"EOF">
2668 C<\Q>, C<\U>, C<\u>, C<\L>, C<\l>, C<\F> (possibly paired with C<\E>) are
2669 converted to corresponding Perl constructs. Thus, C<"$foo\Qbaz$bar">
2670 is converted to C<$foo . (quotemeta("baz" . $bar))> internally.
2671 The other escape sequences such as C<\200> and C<\t> and backslashed
2672 characters such as C<\\> and C<\-> are replaced with appropriate
2675 Let it be stressed that I<whatever falls between C<\Q> and C<\E>>
2676 is interpolated in the usual way. Something like C<"\Q\\E"> has
2677 no C<\E> inside. Instead, it has C<\Q>, C<\\>, and C<E>, so the
2678 result is the same as for C<"\\\\E">. As a general rule, backslashes
2679 between C<\Q> and C<\E> may lead to counterintuitive results. So,
2680 C<"\Q\t\E"> is converted to C<quotemeta("\t")>, which is the same
2681 as C<"\\\t"> (since TAB is not alphanumeric). Note also that:
2686 may be closer to the conjectural I<intention> of the writer of C<"\Q\t\E">.
2688 Interpolated scalars and arrays are converted internally to the C<join> and
2689 C<.> catenation operations. Thus, C<"$foo XXX '@arr'"> becomes:
2691 $foo . " XXX '" . (join $", @arr) . "'";
2693 All operations above are performed simultaneously, left to right.
2695 Because the result of C<"\Q STRING \E"> has all metacharacters
2696 quoted, there is no way to insert a literal C<$> or C<@> inside a
2697 C<\Q\E> pair. If protected by C<\>, C<$> will be quoted to became
2698 C<"\\\$">; if not, it is interpreted as the start of an interpolated
2701 Note also that the interpolation code needs to make a decision on
2702 where the interpolated scalar ends. For instance, whether
2703 C<< "a $b -> {c}" >> really means:
2705 "a " . $b . " -> {c}";
2711 Most of the time, the longest possible text that does not include
2712 spaces between components and which contains matching braces or
2713 brackets. because the outcome may be determined by voting based
2714 on heuristic estimators, the result is not strictly predictable.
2715 Fortunately, it's usually correct for ambiguous cases.
2717 =item the replacement of C<s///>
2719 Processing of C<\Q>, C<\U>, C<\u>, C<\L>, C<\l>, C<\F> and interpolation
2720 happens as with C<qq//> constructs.
2722 It is at this step that C<\1> is begrudgingly converted to C<$1> in
2723 the replacement text of C<s///>, in order to correct the incorrigible
2724 I<sed> hackers who haven't picked up the saner idiom yet. A warning
2725 is emitted if the C<use warnings> pragma or the B<-w> command-line flag
2726 (that is, the C<$^W> variable) was set.
2728 =item C<RE> in C<?RE?>, C</RE/>, C<m/RE/>, C<s/RE/foo/>,
2730 Processing of C<\Q>, C<\U>, C<\u>, C<\L>, C<\l>, C<\F>, C<\E>,
2731 and interpolation happens (almost) as with C<qq//> constructs.
2733 Processing of C<\N{...}> is also done here, and compiled into an intermediate
2734 form for the regex compiler. (This is because, as mentioned below, the regex
2735 compilation may be done at execution time, and C<\N{...}> is a compile-time
2738 However any other combinations of C<\> followed by a character
2739 are not substituted but only skipped, in order to parse them
2740 as regular expressions at the following step.
2741 As C<\c> is skipped at this step, C<@> of C<\c@> in RE is possibly
2742 treated as an array symbol (for example C<@foo>),
2743 even though the same text in C<qq//> gives interpolation of C<\c@>.
2745 Code blocks such as C<(?{BLOCK})> are handled by temporarily passing control
2746 back to the perl parser, in a similar way that an interpolated array
2747 subscript expression such as C<"foo$array[1+f("[xyz")]bar"> would be.
2749 Moreover, inside C<(?{BLOCK})>, C<(?# comment )>, and
2750 a C<#>-comment in a C<//x>-regular expression, no processing is
2751 performed whatsoever. This is the first step at which the presence
2752 of the C<//x> modifier is relevant.
2754 Interpolation in patterns has several quirks: C<$|>, C<$(>, C<$)>, C<@+>
2755 and C<@-> are not interpolated, and constructs C<$var[SOMETHING]> are
2756 voted (by several different estimators) to be either an array element
2757 or C<$var> followed by an RE alternative. This is where the notation
2758 C<${arr[$bar]}> comes handy: C</${arr[0-9]}/> is interpreted as
2759 array element C<-9>, not as a regular expression from the variable
2760 C<$arr> followed by a digit, which would be the interpretation of
2761 C</$arr[0-9]/>. Since voting among different estimators may occur,
2762 the result is not predictable.
2764 The lack of processing of C<\\> creates specific restrictions on
2765 the post-processed text. If the delimiter is C</>, one cannot get
2766 the combination C<\/> into the result of this step. C</> will
2767 finish the regular expression, C<\/> will be stripped to C</> on
2768 the previous step, and C<\\/> will be left as is. Because C</> is
2769 equivalent to C<\/> inside a regular expression, this does not
2770 matter unless the delimiter happens to be character special to the
2771 RE engine, such as in C<s*foo*bar*>, C<m[foo]>, or C<?foo?>; or an
2772 alphanumeric char, as in:
2776 In the RE above, which is intentionally obfuscated for illustration, the
2777 delimiter is C<m>, the modifier is C<mx>, and after delimiter-removal the
2778 RE is the same as for C<m/ ^ a \s* b /mx>. There's more than one
2779 reason you're encouraged to restrict your delimiters to non-alphanumeric,
2780 non-whitespace choices.
2784 This step is the last one for all constructs except regular expressions,
2785 which are processed further.
2787 =item parsing regular expressions
2790 Previous steps were performed during the compilation of Perl code,
2791 but this one happens at run time, although it may be optimized to
2792 be calculated at compile time if appropriate. After preprocessing
2793 described above, and possibly after evaluation if concatenation,
2794 joining, casing translation, or metaquoting are involved, the
2795 resulting I<string> is passed to the RE engine for compilation.
2797 Whatever happens in the RE engine might be better discussed in L<perlre>,
2798 but for the sake of continuity, we shall do so here.
2800 This is another step where the presence of the C<//x> modifier is
2801 relevant. The RE engine scans the string from left to right and
2802 converts it to a finite automaton.
2804 Backslashed characters are either replaced with corresponding
2805 literal strings (as with C<\{>), or else they generate special nodes
2806 in the finite automaton (as with C<\b>). Characters special to the
2807 RE engine (such as C<|>) generate corresponding nodes or groups of
2808 nodes. C<(?#...)> comments are ignored. All the rest is either
2809 converted to literal strings to match, or else is ignored (as is
2810 whitespace and C<#>-style comments if C<//x> is present).
2812 Parsing of the bracketed character class construct, C<[...]>, is
2813 rather different than the rule used for the rest of the pattern.
2814 The terminator of this construct is found using the same rules as
2815 for finding the terminator of a C<{}>-delimited construct, the only
2816 exception being that C<]> immediately following C<[> is treated as
2817 though preceded by a backslash.
2819 The terminator of runtime C<(?{...})> is found by temporarily switching
2820 control to the perl parser, which should stop at the point where the
2821 logically balancing terminating C<}> is found.
2823 It is possible to inspect both the string given to RE engine and the
2824 resulting finite automaton. See the arguments C<debug>/C<debugcolor>
2825 in the C<use L<re>> pragma, as well as Perl's B<-Dr> command-line
2826 switch documented in L<perlrun/"Command Switches">.
2828 =item Optimization of regular expressions
2829 X<regexp, optimization>
2831 This step is listed for completeness only. Since it does not change
2832 semantics, details of this step are not documented and are subject
2833 to change without notice. This step is performed over the finite
2834 automaton that was generated during the previous pass.
2836 It is at this stage that C<split()> silently optimizes C</^/> to
2841 =head2 I/O Operators
2842 X<operator, i/o> X<operator, io> X<io> X<while> X<filehandle>
2845 There are several I/O operators you should know about.
2847 A string enclosed by backticks (grave accents) first undergoes
2848 double-quote interpolation. It is then interpreted as an external
2849 command, and the output of that command is the value of the
2850 backtick string, like in a shell. In scalar context, a single string
2851 consisting of all output is returned. In list context, a list of
2852 values is returned, one per line of output. (You can set C<$/> to use
2853 a different line terminator.) The command is executed each time the
2854 pseudo-literal is evaluated. The status value of the command is
2855 returned in C<$?> (see L<perlvar> for the interpretation of C<$?>).
2856 Unlike in B<csh>, no translation is done on the return data--newlines
2857 remain newlines. Unlike in any of the shells, single quotes do not
2858 hide variable names in the command from interpretation. To pass a
2859 literal dollar-sign through to the shell you need to hide it with a
2860 backslash. The generalized form of backticks is C<qx//>. (Because
2861 backticks always undergo shell expansion as well, see L<perlsec> for
2863 X<qx> X<`> X<``> X<backtick> X<glob>
2865 In scalar context, evaluating a filehandle in angle brackets yields
2866 the next line from that file (the newline, if any, included), or
2867 C<undef> at end-of-file or on error. When C<$/> is set to C<undef>
2868 (sometimes known as file-slurp mode) and the file is empty, it
2869 returns C<''> the first time, followed by C<undef> subsequently.
2871 Ordinarily you must assign the returned value to a variable, but
2872 there is one situation where an automatic assignment happens. If
2873 and only if the input symbol is the only thing inside the conditional
2874 of a C<while> statement (even if disguised as a C<for(;;)> loop),
2875 the value is automatically assigned to the global variable $_,
2876 destroying whatever was there previously. (This may seem like an
2877 odd thing to you, but you'll use the construct in almost every Perl
2878 script you write.) The $_ variable is not implicitly localized.
2879 You'll have to put a C<local $_;> before the loop if you want that
2882 The following lines are equivalent:
2884 while (defined($_ = <STDIN>)) { print; }
2885 while ($_ = <STDIN>) { print; }
2886 while (<STDIN>) { print; }
2887 for (;<STDIN>;) { print; }
2888 print while defined($_ = <STDIN>);
2889 print while ($_ = <STDIN>);
2890 print while <STDIN>;
2892 This also behaves similarly, but assigns to a lexical variable
2893 instead of to C<$_>:
2895 while (my $line = <STDIN>) { print $line }
2897 In these loop constructs, the assigned value (whether assignment
2898 is automatic or explicit) is then tested to see whether it is
2899 defined. The defined test avoids problems where the line has a string
2900 value that would be treated as false by Perl; for example a "" or
2901 a "0" with no trailing newline. If you really mean for such values
2902 to terminate the loop, they should be tested for explicitly:
2904 while (($_ = <STDIN>) ne '0') { ... }
2905 while (<STDIN>) { last unless $_; ... }
2907 In other boolean contexts, C<< <FILEHANDLE> >> without an
2908 explicit C<defined> test or comparison elicits a warning if the
2909 C<use warnings> pragma or the B<-w>
2910 command-line switch (the C<$^W> variable) is in effect.
2912 The filehandles STDIN, STDOUT, and STDERR are predefined. (The
2913 filehandles C<stdin>, C<stdout>, and C<stderr> will also work except
2914 in packages, where they would be interpreted as local identifiers
2915 rather than global.) Additional filehandles may be created with
2916 the open() function, amongst others. See L<perlopentut> and
2917 L<perlfunc/open> for details on this.
2918 X<stdin> X<stdout> X<sterr>
2920 If a <FILEHANDLE> is used in a context that is looking for
2921 a list, a list comprising all input lines is returned, one line per
2922 list element. It's easy to grow to a rather large data space this
2923 way, so use with care.
2925 <FILEHANDLE> may also be spelled C<readline(*FILEHANDLE)>.
2926 See L<perlfunc/readline>.
2928 The null filehandle <> is special: it can be used to emulate the
2929 behavior of B<sed> and B<awk>, and any other Unix filter program
2930 that takes a list of filenames, doing the same to each line
2931 of input from all of them. Input from <> comes either from
2932 standard input, or from each file listed on the command line. Here's
2933 how it works: the first time <> is evaluated, the @ARGV array is
2934 checked, and if it is empty, C<$ARGV[0]> is set to "-", which when opened
2935 gives you standard input. The @ARGV array is then processed as a list
2936 of filenames. The loop
2939 ... # code for each line
2942 is equivalent to the following Perl-like pseudo code:
2944 unshift(@ARGV, '-') unless @ARGV;
2945 while ($ARGV = shift) {
2948 ... # code for each line
2952 except that it isn't so cumbersome to say, and will actually work.
2953 It really does shift the @ARGV array and put the current filename
2954 into the $ARGV variable. It also uses filehandle I<ARGV>
2955 internally. <> is just a synonym for <ARGV>, which
2956 is magical. (The pseudo code above doesn't work because it treats
2957 <ARGV> as non-magical.)
2959 Since the null filehandle uses the two argument form of L<perlfunc/open>
2960 it interprets special characters, so if you have a script like this:
2966 and call it with C<perl dangerous.pl 'rm -rfv *|'>, it actually opens a
2967 pipe, executes the C<rm> command and reads C<rm>'s output from that pipe.
2968 If you want all items in C<@ARGV> to be interpreted as file names, you
2969 can use the module C<ARGV::readonly> from CPAN.
2971 You can modify @ARGV before the first <> as long as the array ends up
2972 containing the list of filenames you really want. Line numbers (C<$.>)
2973 continue as though the input were one big happy file. See the example
2974 in L<perlfunc/eof> for how to reset line numbers on each file.
2976 If you want to set @ARGV to your own list of files, go right ahead.
2977 This sets @ARGV to all plain text files if no @ARGV was given:
2979 @ARGV = grep { -f && -T } glob('*') unless @ARGV;
2981 You can even set them to pipe commands. For example, this automatically
2982 filters compressed arguments through B<gzip>:
2984 @ARGV = map { /\.(gz|Z)$/ ? "gzip -dc < $_ |" : $_ } @ARGV;
2986 If you want to pass switches into your script, you can use one of the
2987 Getopts modules or put a loop on the front like this:
2989 while ($_ = $ARGV[0], /^-/) {
2992 if (/^-D(.*)/) { $debug = $1 }
2993 if (/^-v/) { $verbose++ }
2994 # ... # other switches
2998 # ... # code for each line
3001 The <> symbol will return C<undef> for end-of-file only once.
3002 If you call it again after this, it will assume you are processing another
3003 @ARGV list, and if you haven't set @ARGV, will read input from STDIN.
3005 If what the angle brackets contain is a simple scalar variable (for example,
3006 <$foo>), then that variable contains the name of the
3007 filehandle to input from, or its typeglob, or a reference to the
3013 If what's within the angle brackets is neither a filehandle nor a simple
3014 scalar variable containing a filehandle name, typeglob, or typeglob
3015 reference, it is interpreted as a filename pattern to be globbed, and
3016 either a list of filenames or the next filename in the list is returned,
3017 depending on context. This distinction is determined on syntactic
3018 grounds alone. That means C<< <$x> >> is always a readline() from
3019 an indirect handle, but C<< <$hash{key}> >> is always a glob().
3020 That's because $x is a simple scalar variable, but C<$hash{key}> is
3021 not--it's a hash element. Even C<< <$x > >> (note the extra space)
3022 is treated as C<glob("$x ")>, not C<readline($x)>.
3024 One level of double-quote interpretation is done first, but you can't
3025 say C<< <$foo> >> because that's an indirect filehandle as explained
3026 in the previous paragraph. (In older versions of Perl, programmers
3027 would insert curly brackets to force interpretation as a filename glob:
3028 C<< <${foo}> >>. These days, it's considered cleaner to call the
3029 internal function directly as C<glob($foo)>, which is probably the right
3030 way to have done it in the first place.) For example:
3036 is roughly equivalent to:
3038 open(FOO, "echo *.c | tr -s ' \t\r\f' '\\012\\012\\012\\012'|");
3044 except that the globbing is actually done internally using the standard
3045 C<File::Glob> extension. Of course, the shortest way to do the above is:
3049 A (file)glob evaluates its (embedded) argument only when it is
3050 starting a new list. All values must be read before it will start
3051 over. In list context, this isn't important because you automatically
3052 get them all anyway. However, in scalar context the operator returns
3053 the next value each time it's called, or C<undef> when the list has
3054 run out. As with filehandle reads, an automatic C<defined> is
3055 generated when the glob occurs in the test part of a C<while>,
3056 because legal glob returns (for example,
3057 a file called F<0>) would otherwise
3058 terminate the loop. Again, C<undef> is returned only once. So if
3059 you're expecting a single value from a glob, it is much better to
3062 ($file) = <blurch*>;
3068 because the latter will alternate between returning a filename and
3071 If you're trying to do variable interpolation, it's definitely better
3072 to use the glob() function, because the older notation can cause people
3073 to become confused with the indirect filehandle notation.
3075 @files = glob("$dir/*.[ch]");
3076 @files = glob($files[$i]);
3078 =head2 Constant Folding
3079 X<constant folding> X<folding>
3081 Like C, Perl does a certain amount of expression evaluation at
3082 compile time whenever it determines that all arguments to an
3083 operator are static and have no side effects. In particular, string
3084 concatenation happens at compile time between literals that don't do
3085 variable substitution. Backslash interpolation also happens at
3086 compile time. You can say
3088 'Now is the time for all'
3090 . 'good men to come to.'
3092 and this all reduces to one string internally. Likewise, if
3095 foreach $file (@filenames) {
3096 if (-s $file > 5 + 100 * 2**16) { }
3099 the compiler precomputes the number which that expression
3100 represents so that the interpreter won't have to.
3105 Perl doesn't officially have a no-op operator, but the bare constants
3106 C<0> and C<1> are special-cased not to produce a warning in void
3107 context, so you can for example safely do
3111 =head2 Bitwise String Operators
3112 X<operator, bitwise, string>
3114 Bitstrings of any size may be manipulated by the bitwise operators
3117 If the operands to a binary bitwise op are strings of different
3118 sizes, B<|> and B<^> ops act as though the shorter operand had
3119 additional zero bits on the right, while the B<&> op acts as though
3120 the longer operand were truncated to the length of the shorter.
3121 The granularity for such extension or truncation is one or more
3124 # ASCII-based examples
3125 print "j p \n" ^ " a h"; # prints "JAPH\n"
3126 print "JA" | " ph\n"; # prints "japh\n"
3127 print "japh\nJunk" & '_____'; # prints "JAPH\n";
3128 print 'p N$' ^ " E<H\n"; # prints "Perl\n";
3130 If you are intending to manipulate bitstrings, be certain that
3131 you're supplying bitstrings: If an operand is a number, that will imply
3132 a B<numeric> bitwise operation. You may explicitly show which type of
3133 operation you intend by using C<""> or C<0+>, as in the examples below.
3135 $foo = 150 | 105; # yields 255 (0x96 | 0x69 is 0xFF)
3136 $foo = '150' | 105; # yields 255
3137 $foo = 150 | '105'; # yields 255
3138 $foo = '150' | '105'; # yields string '155' (under ASCII)
3140 $baz = 0+$foo & 0+$bar; # both ops explicitly numeric
3141 $biz = "$foo" ^ "$bar"; # both ops explicitly stringy
3143 See L<perlfunc/vec> for information on how to manipulate individual bits
3146 =head2 Integer Arithmetic
3149 By default, Perl assumes that it must do most of its arithmetic in
3150 floating point. But by saying
3154 you may tell the compiler to use integer operations
3155 (see L<integer> for a detailed explanation) from here to the end of
3156 the enclosing BLOCK. An inner BLOCK may countermand this by saying
3160 which lasts until the end of that BLOCK. Note that this doesn't
3161 mean everything is an integer, merely that Perl will use integer
3162 operations for arithmetic, comparison, and bitwise operators. For
3163 example, even under C<use integer>, if you take the C<sqrt(2)>, you'll
3164 still get C<1.4142135623731> or so.
3166 Used on numbers, the bitwise operators ("&", "|", "^", "~", "<<",
3167 and ">>") always produce integral results. (But see also
3168 L<Bitwise String Operators>.) However, C<use integer> still has meaning for
3169 them. By default, their results are interpreted as unsigned integers, but
3170 if C<use integer> is in effect, their results are interpreted
3171 as signed integers. For example, C<~0> usually evaluates to a large
3172 integral value. However, C<use integer; ~0> is C<-1> on two's-complement
3175 =head2 Floating-point Arithmetic
3177 X<floating-point> X<floating point> X<float> X<real>
3179 While C<use integer> provides integer-only arithmetic, there is no
3180 analogous mechanism to provide automatic rounding or truncation to a
3181 certain number of decimal places. For rounding to a certain number
3182 of digits, sprintf() or printf() is usually the easiest route.
3185 Floating-point numbers are only approximations to what a mathematician
3186 would call real numbers. There are infinitely more reals than floats,
3187 so some corners must be cut. For example:
3189 printf "%.20g\n", 123456789123456789;
3190 # produces 123456789123456784
3192 Testing for exact floating-point equality or inequality is not a
3193 good idea. Here's a (relatively expensive) work-around to compare
3194 whether two floating-point numbers are equal to a particular number of
3195 decimal places. See Knuth, volume II, for a more robust treatment of
3199 my ($X, $Y, $POINTS) = @_;
3201 $tX = sprintf("%.${POINTS}g", $X);
3202 $tY = sprintf("%.${POINTS}g", $Y);
3206 The POSIX module (part of the standard perl distribution) implements
3207 ceil(), floor(), and other mathematical and trigonometric functions.
3208 The Math::Complex module (part of the standard perl distribution)
3209 defines mathematical functions that work on both the reals and the
3210 imaginary numbers. Math::Complex not as efficient as POSIX, but
3211 POSIX can't work with complex numbers.
3213 Rounding in financial applications can have serious implications, and
3214 the rounding method used should be specified precisely. In these
3215 cases, it probably pays not to trust whichever system rounding is
3216 being used by Perl, but to instead implement the rounding function you
3219 =head2 Bigger Numbers
3220 X<number, arbitrary precision>
3222 The standard C<Math::BigInt>, C<Math::BigRat>, and C<Math::BigFloat> modules,
3223 along with the C<bignum>, C<bigint>, and C<bigrat> pragmas, provide
3224 variable-precision arithmetic and overloaded operators, although
3225 they're currently pretty slow. At the cost of some space and
3226 considerable speed, they avoid the normal pitfalls associated with
3227 limited-precision representations.
3230 use bigint; # easy interface to Math::BigInt
3231 $x = 123456789123456789;
3233 +15241578780673678515622620750190521
3241 say "a/b is ", $a/$b;
3242 say "a*b is ", $a*$b;
3246 Several modules let you calculate with (bound only by memory and CPU time)
3247 unlimited or fixed precision. There are also some non-standard modules that
3248 provide faster implementations via external C libraries.
3250 Here is a short, but incomplete summary:
3252 Math::String treat string sequences like numbers
3253 Math::FixedPrecision calculate with a fixed precision
3254 Math::Currency for currency calculations
3255 Bit::Vector manipulate bit vectors fast (uses C)
3256 Math::BigIntFast Bit::Vector wrapper for big numbers
3257 Math::Pari provides access to the Pari C library
3258 Math::Cephes uses the external Cephes C library (no
3260 Math::Cephes::Fraction fractions via the Cephes library
3261 Math::GMP another one using an external C library
3262 Math::GMPz an alternative interface to libgmp's big ints
3263 Math::GMPq an interface to libgmp's fraction numbers
3264 Math::GMPf an interface to libgmp's floating point numbers