3 perlop - Perl operators and precedence
7 =head2 Operator Precedence and Associativity
9 Operator precedence and associativity work in Perl more or less like
10 they do in mathematics.
12 I<Operator precedence> means some operators are evaluated before
13 others. For example, in C<2 + 4 * 5>, the multiplication has higher
14 precedence so C<4 * 5> is evaluated first yielding C<2 + 20 ==
15 22> and not C<6 * 5 == 30>.
17 I<Operator associativity> defines what happens if a sequence of the
18 same operators is used one after another: whether the evaluator will
19 evaluate the left operations first or the right. For example, in C<8
20 - 4 - 2>, subtraction is left associative so Perl evaluates the
21 expression left to right. C<8 - 4> is evaluated first making the
22 expression C<4 - 2 == 2> and not C<8 - 2 == 6>.
24 Perl operators have the following associativity and precedence,
25 listed from highest precedence to lowest. Operators borrowed from
26 C keep the same precedence relationship with each other, even where
27 C's precedence is slightly screwy. (This makes learning Perl easier
28 for C folks.) With very few exceptions, these all operate on scalar
29 values only, not array values.
31 left terms and list operators (leftward)
35 right ! ~ \ and unary + and -
40 nonassoc named unary operators
41 nonassoc < > <= >= lt gt le ge
42 nonassoc == != <=> eq ne cmp
51 nonassoc list operators (rightward)
56 In the following sections, these operators are covered in precedence order.
58 Many operators can be overloaded for objects. See L<overload>.
60 =head2 Terms and List Operators (Leftward)
62 A TERM has the highest precedence in Perl. They include variables,
63 quote and quote-like operators, any expression in parentheses,
64 and any function whose arguments are parenthesized. Actually, there
65 aren't really functions in this sense, just list operators and unary
66 operators behaving as functions because you put parentheses around
67 the arguments. These are all documented in L<perlfunc>.
69 If any list operator (print(), etc.) or any unary operator (chdir(), etc.)
70 is followed by a left parenthesis as the next token, the operator and
71 arguments within parentheses are taken to be of highest precedence,
72 just like a normal function call.
74 In the absence of parentheses, the precedence of list operators such as
75 C<print>, C<sort>, or C<chmod> is either very high or very low depending on
76 whether you are looking at the left side or the right side of the operator.
79 @ary = (1, 3, sort 4, 2);
80 print @ary; # prints 1324
82 the commas on the right of the sort are evaluated before the sort,
83 but the commas on the left are evaluated after. In other words,
84 list operators tend to gobble up all arguments that follow, and
85 then act like a simple TERM with regard to the preceding expression.
86 Be careful with parentheses:
88 # These evaluate exit before doing the print:
89 print($foo, exit); # Obviously not what you want.
90 print $foo, exit; # Nor is this.
92 # These do the print before evaluating exit:
93 (print $foo), exit; # This is what you want.
94 print($foo), exit; # Or this.
95 print ($foo), exit; # Or even this.
99 print ($foo & 255) + 1, "\n";
101 probably doesn't do what you expect at first glance. The parentheses
102 enclose the argument list for C<print> which is evaluated (printing
103 the result of C<$foo & 255>). Then one is added to the return value
104 of C<print> (usually 1). The result is something like this:
106 1 + 1, "\n"; # Obviously not what you meant.
108 To do what you meant properly, you must write:
110 print(($foo & 255) + 1, "\n");
112 See L<Named Unary Operators> for more discussion of this.
114 Also parsed as terms are the C<do {}> and C<eval {}> constructs, as
115 well as subroutine and method calls, and the anonymous
116 constructors C<[]> and C<{}>.
118 See also L<Quote and Quote-like Operators> toward the end of this section,
119 as well as L<"I/O Operators">.
121 =head2 The Arrow Operator
123 "C<< -> >>" is an infix dereference operator, just as it is in C
124 and C++. If the right side is either a C<[...]>, C<{...}>, or a
125 C<(...)> subscript, then the left side must be either a hard or
126 symbolic reference to an array, a hash, or a subroutine respectively.
127 (Or technically speaking, a location capable of holding a hard
128 reference, if it's an array or hash reference being used for
129 assignment.) See L<perlreftut> and L<perlref>.
131 Otherwise, the right side is a method name or a simple scalar
132 variable containing either the method name or a subroutine reference,
133 and the left side must be either an object (a blessed reference)
134 or a class name (that is, a package name). See L<perlobj>.
136 =head2 Auto-increment and Auto-decrement
138 "++" and "--" work as in C. That is, if placed before a variable,
139 they increment or decrement the variable by one before returning the
140 value, and if placed after, increment or decrement after returning the
144 print $i++; # prints 0
145 print ++$j; # prints 1
147 Note that just as in C, Perl doesn't define B<when> the variable is
148 incremented or decremented. You just know it will be done sometime
149 before or after the value is returned. This also means that modifying
150 a variable twice in the same statement will lead to undefined behaviour.
151 Avoid statements like:
156 Perl will not guarantee what the result of the above statements is.
158 The auto-increment operator has a little extra builtin magic to it. If
159 you increment a variable that is numeric, or that has ever been used in
160 a numeric context, you get a normal increment. If, however, the
161 variable has been used in only string contexts since it was set, and
162 has a value that is not the empty string and matches the pattern
163 C</^[a-zA-Z]*[0-9]*\z/>, the increment is done as a string, preserving each
164 character within its range, with carry:
166 print ++($foo = '99'); # prints '100'
167 print ++($foo = 'a0'); # prints 'a1'
168 print ++($foo = 'Az'); # prints 'Ba'
169 print ++($foo = 'zz'); # prints 'aaa'
171 C<undef> is always treated as numeric, and in particular is changed
172 to C<0> before incrementing (so that a post-increment of an undef value
173 will return C<0> rather than C<undef>).
175 The auto-decrement operator is not magical.
177 =head2 Exponentiation
179 Binary "**" is the exponentiation operator. It binds even more
180 tightly than unary minus, so -2**4 is -(2**4), not (-2)**4. (This is
181 implemented using C's pow(3) function, which actually works on doubles
184 =head2 Symbolic Unary Operators
186 Unary "!" performs logical negation, i.e., "not". See also C<not> for a lower
187 precedence version of this.
189 Unary "-" performs arithmetic negation if the operand is numeric. If
190 the operand is an identifier, a string consisting of a minus sign
191 concatenated with the identifier is returned. Otherwise, if the string
192 starts with a plus or minus, a string starting with the opposite sign
193 is returned. One effect of these rules is that -bareword is equivalent
196 Unary "~" performs bitwise negation, i.e., 1's complement. For
197 example, C<0666 & ~027> is 0640. (See also L<Integer Arithmetic> and
198 L<Bitwise String Operators>.) Note that the width of the result is
199 platform-dependent: ~0 is 32 bits wide on a 32-bit platform, but 64
200 bits wide on a 64-bit platform, so if you are expecting a certain bit
201 width, remember to use the & operator to mask off the excess bits.
203 Unary "+" has no effect whatsoever, even on strings. It is useful
204 syntactically for separating a function name from a parenthesized expression
205 that would otherwise be interpreted as the complete list of function
206 arguments. (See examples above under L<Terms and List Operators (Leftward)>.)
208 Unary "\" creates a reference to whatever follows it. See L<perlreftut>
209 and L<perlref>. Do not confuse this behavior with the behavior of
210 backslash within a string, although both forms do convey the notion
211 of protecting the next thing from interpolation.
213 =head2 Binding Operators
215 Binary "=~" binds a scalar expression to a pattern match. Certain operations
216 search or modify the string $_ by default. This operator makes that kind
217 of operation work on some other string. The right argument is a search
218 pattern, substitution, or transliteration. The left argument is what is
219 supposed to be searched, substituted, or transliterated instead of the default
220 $_. When used in scalar context, the return value generally indicates the
221 success of the operation. Behavior in list context depends on the particular
222 operator. See L</"Regexp Quote-Like Operators"> for details.
224 If the right argument is an expression rather than a search pattern,
225 substitution, or transliteration, it is interpreted as a search pattern at run
228 Binary "!~" is just like "=~" except the return value is negated in
231 =head2 Multiplicative Operators
233 Binary "*" multiplies two numbers.
235 Binary "/" divides two numbers.
237 Binary "%" computes the modulus of two numbers. Given integer
238 operands C<$a> and C<$b>: If C<$b> is positive, then C<$a % $b> is
239 C<$a> minus the largest multiple of C<$b> that is not greater than
240 C<$a>. If C<$b> is negative, then C<$a % $b> is C<$a> minus the
241 smallest multiple of C<$b> that is not less than C<$a> (i.e. the
242 result will be less than or equal to zero).
243 Note that when C<use integer> is in scope, "%" gives you direct access
244 to the modulus operator as implemented by your C compiler. This
245 operator is not as well defined for negative operands, but it will
248 Binary "x" is the repetition operator. In scalar context or if the left
249 operand is not enclosed in parentheses, it returns a string consisting
250 of the left operand repeated the number of times specified by the right
251 operand. In list context, if the left operand is enclosed in
252 parentheses, it repeats the list. If the right operand is zero or
253 negative, it returns an empty string or an empty list, depending on the
256 print '-' x 80; # print row of dashes
258 print "\t" x ($tab/8), ' ' x ($tab%8); # tab over
260 @ones = (1) x 80; # a list of 80 1's
261 @ones = (5) x @ones; # set all elements to 5
264 =head2 Additive Operators
266 Binary "+" returns the sum of two numbers.
268 Binary "-" returns the difference of two numbers.
270 Binary "." concatenates two strings.
272 =head2 Shift Operators
274 Binary "<<" returns the value of its left argument shifted left by the
275 number of bits specified by the right argument. Arguments should be
276 integers. (See also L<Integer Arithmetic>.)
278 Binary ">>" returns the value of its left argument shifted right by
279 the number of bits specified by the right argument. Arguments should
280 be integers. (See also L<Integer Arithmetic>.)
282 Note that both "<<" and ">>" in Perl are implemented directly using
283 "<<" and ">>" in C. If C<use integer> (see L<Integer Arithmetic>) is
284 in force then signed C integers are used, else unsigned C integers are
285 used. Either way, the implementation isn't going to generate results
286 larger than the size of the integer type Perl was built with (32 bits
289 The result of overflowing the range of the integers is undefined
290 because it is undefined also in C. In other words, using 32-bit
291 integers, C<< 1 << 32 >> is undefined. Shifting by a negative number
292 of bits is also undefined.
294 =head2 Named Unary Operators
296 The various named unary operators are treated as functions with one
297 argument, with optional parentheses.
299 If any list operator (print(), etc.) or any unary operator (chdir(), etc.)
300 is followed by a left parenthesis as the next token, the operator and
301 arguments within parentheses are taken to be of highest precedence,
302 just like a normal function call. For example,
303 because named unary operators are higher precedence than ||:
305 chdir $foo || die; # (chdir $foo) || die
306 chdir($foo) || die; # (chdir $foo) || die
307 chdir ($foo) || die; # (chdir $foo) || die
308 chdir +($foo) || die; # (chdir $foo) || die
310 but, because * is higher precedence than named operators:
312 chdir $foo * 20; # chdir ($foo * 20)
313 chdir($foo) * 20; # (chdir $foo) * 20
314 chdir ($foo) * 20; # (chdir $foo) * 20
315 chdir +($foo) * 20; # chdir ($foo * 20)
317 rand 10 * 20; # rand (10 * 20)
318 rand(10) * 20; # (rand 10) * 20
319 rand (10) * 20; # (rand 10) * 20
320 rand +(10) * 20; # rand (10 * 20)
322 Regarding precedence, the filetest operators, like C<-f>, C<-M>, etc. are
323 treated like named unary operators, but they don't follow this functional
324 parenthesis rule. That means, for example, that C<-f($file).".bak"> is
325 equivalent to C<-f "$file.bak">.
327 See also L<"Terms and List Operators (Leftward)">.
329 =head2 Relational Operators
331 Binary "<" returns true if the left argument is numerically less than
334 Binary ">" returns true if the left argument is numerically greater
335 than the right argument.
337 Binary "<=" returns true if the left argument is numerically less than
338 or equal to the right argument.
340 Binary ">=" returns true if the left argument is numerically greater
341 than or equal to the right argument.
343 Binary "lt" returns true if the left argument is stringwise less than
346 Binary "gt" returns true if the left argument is stringwise greater
347 than the right argument.
349 Binary "le" returns true if the left argument is stringwise less than
350 or equal to the right argument.
352 Binary "ge" returns true if the left argument is stringwise greater
353 than or equal to the right argument.
355 =head2 Equality Operators
357 Binary "==" returns true if the left argument is numerically equal to
360 Binary "!=" returns true if the left argument is numerically not equal
361 to the right argument.
363 Binary "<=>" returns -1, 0, or 1 depending on whether the left
364 argument is numerically less than, equal to, or greater than the right
365 argument. If your platform supports NaNs (not-a-numbers) as numeric
366 values, using them with "<=>" returns undef. NaN is not "<", "==", ">",
367 "<=" or ">=" anything (even NaN), so those 5 return false. NaN != NaN
368 returns true, as does NaN != anything else. If your platform doesn't
369 support NaNs then NaN is just a string with numeric value 0.
371 perl -le '$a = NaN; print "No NaN support here" if $a == $a'
372 perl -le '$a = NaN; print "NaN support here" if $a != $a'
374 Binary "eq" returns true if the left argument is stringwise equal to
377 Binary "ne" returns true if the left argument is stringwise not equal
378 to the right argument.
380 Binary "cmp" returns -1, 0, or 1 depending on whether the left
381 argument is stringwise less than, equal to, or greater than the right
384 "lt", "le", "ge", "gt" and "cmp" use the collation (sort) order specified
385 by the current locale if C<use locale> is in effect. See L<perllocale>.
389 Binary "&" returns its operands ANDed together bit by bit.
390 (See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
392 Note that "&" has lower priority than relational operators, so for example
393 the brackets are essential in a test like
395 print "Even\n" if ($x & 1) == 0;
397 =head2 Bitwise Or and Exclusive Or
399 Binary "|" returns its operands ORed together bit by bit.
400 (See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
402 Binary "^" returns its operands XORed together bit by bit.
403 (See also L<Integer Arithmetic> and L<Bitwise String Operators>.)
405 Note that "|" and "^" have lower priority than relational operators, so
406 for example the brackets are essential in a test like
408 print "false\n" if (8 | 2) != 10;
410 =head2 C-style Logical And
412 Binary "&&" performs a short-circuit logical AND operation. That is,
413 if the left operand is false, the right operand is not even evaluated.
414 Scalar or list context propagates down to the right operand if it
417 =head2 C-style Logical Or
419 Binary "||" performs a short-circuit logical OR operation. That is,
420 if the left operand is true, the right operand is not even evaluated.
421 Scalar or list context propagates down to the right operand if it
424 =head2 C-style Logical Defined-Or
426 Although it has no direct equivalent in C, Perl's C<//> operator is related
427 to its C-style or. In fact, it's exactly the same as C<||>, except that it
428 tests the left hand side's definedness instead of its truth. Thus, C<$a // $b>
429 is similar to C<defined($a) || $b> (except that it returns the value of C<$a>
430 rather than the value of C<defined($a)>) and is exactly equivalent to
431 C<defined($a) ? $a : $b>. This is very useful for providing default values
432 for variables. If you actually want to test if at least one of C<$a> and
433 C<$b> is defined, use C<defined($a // $b)>.
435 The C<||>, C<//> and C<&&> operators return the last value evaluated
436 (unlike C's C<||> and C<&&>, which return 0 or 1). Thus, a reasonably
437 portable way to find out the home directory might be:
439 $home = $ENV{'HOME'} // $ENV{'LOGDIR'} //
440 (getpwuid($<))[7] // die "You're homeless!\n";
442 In particular, this means that you shouldn't use this
443 for selecting between two aggregates for assignment:
445 @a = @b || @c; # this is wrong
446 @a = scalar(@b) || @c; # really meant this
447 @a = @b ? @b : @c; # this works fine, though
449 As more readable alternatives to C<&&>, C<//> and C<||> when used for
450 control flow, Perl provides C<and>, C<err> and C<or> operators (see below).
451 The short-circuit behavior is identical. The precedence of "and", "err"
452 and "or" is much lower, however, so that you can safely use them after a
453 list operator without the need for parentheses:
455 unlink "alpha", "beta", "gamma"
456 or gripe(), next LINE;
458 With the C-style operators that would have been written like this:
460 unlink("alpha", "beta", "gamma")
461 || (gripe(), next LINE);
463 Using "or" for assignment is unlikely to do what you want; see below.
465 =head2 Range Operators
467 Binary ".." is the range operator, which is really two different
468 operators depending on the context. In list context, it returns a
469 list of values counting (up by ones) from the left value to the right
470 value. If the left value is greater than the right value then it
471 returns the empty list. The range operator is useful for writing
472 C<foreach (1..10)> loops and for doing slice operations on arrays. In
473 the current implementation, no temporary array is created when the
474 range operator is used as the expression in C<foreach> loops, but older
475 versions of Perl might burn a lot of memory when you write something
478 for (1 .. 1_000_000) {
482 The range operator also works on strings, using the magical auto-increment,
485 In scalar context, ".." returns a boolean value. The operator is
486 bistable, like a flip-flop, and emulates the line-range (comma) operator
487 of B<sed>, B<awk>, and various editors. Each ".." operator maintains its
488 own boolean state. It is false as long as its left operand is false.
489 Once the left operand is true, the range operator stays true until the
490 right operand is true, I<AFTER> which the range operator becomes false
491 again. It doesn't become false till the next time the range operator is
492 evaluated. It can test the right operand and become false on the same
493 evaluation it became true (as in B<awk>), but it still returns true once.
494 If you don't want it to test the right operand till the next
495 evaluation, as in B<sed>, just use three dots ("...") instead of
496 two. In all other regards, "..." behaves just like ".." does.
498 The right operand is not evaluated while the operator is in the
499 "false" state, and the left operand is not evaluated while the
500 operator is in the "true" state. The precedence is a little lower
501 than || and &&. The value returned is either the empty string for
502 false, or a sequence number (beginning with 1) for true. The
503 sequence number is reset for each range encountered. The final
504 sequence number in a range has the string "E0" appended to it, which
505 doesn't affect its numeric value, but gives you something to search
506 for if you want to exclude the endpoint. You can exclude the
507 beginning point by waiting for the sequence number to be greater
510 If either operand of scalar ".." is a constant expression,
511 that operand is considered true if it is equal (C<==>) to the current
512 input line number (the C<$.> variable).
514 To be pedantic, the comparison is actually C<int(EXPR) == int(EXPR)>,
515 but that is only an issue if you use a floating point expression; when
516 implicitly using C<$.> as described in the previous paragraph, the
517 comparison is C<int(EXPR) == int($.)> which is only an issue when C<$.>
518 is set to a floating point value and you are not reading from a file.
519 Furthermore, C<"span" .. "spat"> or C<2.18 .. 3.14> will not do what
520 you want in scalar context because each of the operands are evaluated
521 using their integer representation.
525 As a scalar operator:
527 if (101 .. 200) { print; } # print 2nd hundred lines, short for
528 # if ($. == 101 .. $. == 200) ...
529 next line if (1 .. /^$/); # skip header lines, short for
530 # ... if ($. == 1 .. /^$/);
531 s/^/> / if (/^$/ .. eof()); # quote body
533 # parse mail messages
535 $in_header = 1 .. /^$/;
536 $in_body = /^$/ .. eof;
543 close ARGV if eof; # reset $. each file
546 Here's a simple example to illustrate the difference between
547 the two range operators:
562 This program will print only the line containing "Bar". If
563 the range operator is changed to C<...>, it will also print the
566 And now some examples as a list operator:
568 for (101 .. 200) { print; } # print $_ 100 times
569 @foo = @foo[0 .. $#foo]; # an expensive no-op
570 @foo = @foo[$#foo-4 .. $#foo]; # slice last 5 items
572 The range operator (in list context) makes use of the magical
573 auto-increment algorithm if the operands are strings. You
576 @alphabet = ('A' .. 'Z');
578 to get all normal letters of the English alphabet, or
580 $hexdigit = (0 .. 9, 'a' .. 'f')[$num & 15];
582 to get a hexadecimal digit, or
584 @z2 = ('01' .. '31'); print $z2[$mday];
586 to get dates with leading zeros. If the final value specified is not
587 in the sequence that the magical increment would produce, the sequence
588 goes until the next value would be longer than the final value
591 Because each operand is evaluated in integer form, C<2.18 .. 3.14> will
592 return two elements in list context.
594 @list = (2.18 .. 3.14); # same as @list = (2 .. 3);
596 =head2 Conditional Operator
598 Ternary "?:" is the conditional operator, just as in C. It works much
599 like an if-then-else. If the argument before the ? is true, the
600 argument before the : is returned, otherwise the argument after the :
601 is returned. For example:
603 printf "I have %d dog%s.\n", $n,
604 ($n == 1) ? '' : "s";
606 Scalar or list context propagates downward into the 2nd
607 or 3rd argument, whichever is selected.
609 $a = $ok ? $b : $c; # get a scalar
610 @a = $ok ? @b : @c; # get an array
611 $a = $ok ? @b : @c; # oops, that's just a count!
613 The operator may be assigned to if both the 2nd and 3rd arguments are
614 legal lvalues (meaning that you can assign to them):
616 ($a_or_b ? $a : $b) = $c;
618 Because this operator produces an assignable result, using assignments
619 without parentheses will get you in trouble. For example, this:
621 $a % 2 ? $a += 10 : $a += 2
625 (($a % 2) ? ($a += 10) : $a) += 2
629 ($a % 2) ? ($a += 10) : ($a += 2)
631 That should probably be written more simply as:
633 $a += ($a % 2) ? 10 : 2;
635 =head2 Assignment Operators
637 "=" is the ordinary assignment operator.
639 Assignment operators work as in C. That is,
647 although without duplicating any side effects that dereferencing the lvalue
648 might trigger, such as from tie(). Other assignment operators work similarly.
649 The following are recognized:
656 Although these are grouped by family, they all have the precedence
659 Unlike in C, the scalar assignment operator produces a valid lvalue.
660 Modifying an assignment is equivalent to doing the assignment and
661 then modifying the variable that was assigned to. This is useful
662 for modifying a copy of something, like this:
664 ($tmp = $global) =~ tr [A-Z] [a-z];
675 Similarly, a list assignment in list context produces the list of
676 lvalues assigned to, and a list assignment in scalar context returns
677 the number of elements produced by the expression on the right hand
678 side of the assignment.
680 =head2 Comma Operator
682 Binary "," is the comma operator. In scalar context it evaluates
683 its left argument, throws that value away, then evaluates its right
684 argument and returns that value. This is just like C's comma operator.
686 In list context, it's just the list argument separator, and inserts
687 both its arguments into the list.
689 The C<< => >> operator is a synonym for the comma, but forces any word
690 to its left to be interpreted as a string (as of 5.001). It is helpful
691 in documenting the correspondence between keys and values in hashes,
692 and other paired elements in lists.
694 =head2 List Operators (Rightward)
696 On the right side of a list operator, it has very low precedence,
697 such that it controls all comma-separated expressions found there.
698 The only operators with lower precedence are the logical operators
699 "and", "or", and "not", which may be used to evaluate calls to list
700 operators without the need for extra parentheses:
702 open HANDLE, "filename"
703 or die "Can't open: $!\n";
705 See also discussion of list operators in L<Terms and List Operators (Leftward)>.
709 Unary "not" returns the logical negation of the expression to its right.
710 It's the equivalent of "!" except for the very low precedence.
714 Binary "and" returns the logical conjunction of the two surrounding
715 expressions. It's equivalent to && except for the very low
716 precedence. This means that it short-circuits: i.e., the right
717 expression is evaluated only if the left expression is true.
719 =head2 Logical or, Defined or, and Exclusive Or
721 Binary "or" returns the logical disjunction of the two surrounding
722 expressions. It's equivalent to || except for the very low precedence.
723 This makes it useful for control flow
725 print FH $data or die "Can't write to FH: $!";
727 This means that it short-circuits: i.e., the right expression is evaluated
728 only if the left expression is false. Due to its precedence, you should
729 probably avoid using this for assignment, only for control flow.
731 $a = $b or $c; # bug: this is wrong
732 ($a = $b) or $c; # really means this
733 $a = $b || $c; # better written this way
735 However, when it's a list-context assignment and you're trying to use
736 "||" for control flow, you probably need "or" so that the assignment
737 takes higher precedence.
739 @info = stat($file) || die; # oops, scalar sense of stat!
740 @info = stat($file) or die; # better, now @info gets its due
742 Then again, you could always use parentheses.
744 Binary "err" is equivalent to C<//>--it's just like binary "or", except it tests
745 its left argument's definedness instead of its truth. There are two ways to
746 remember "err": either because many functions return C<undef> on an B<err>or,
747 or as a sort of correction: C<$a=($b err 'default')>
749 Binary "xor" returns the exclusive-OR of the two surrounding expressions.
750 It cannot short circuit, of course.
752 =head2 C Operators Missing From Perl
754 Here is what C has that Perl doesn't:
760 Address-of operator. (But see the "\" operator for taking a reference.)
764 Dereference-address operator. (Perl's prefix dereferencing
765 operators are typed: $, @, %, and &.)
769 Type-casting operator.
773 =head2 Quote and Quote-like Operators
775 While we usually think of quotes as literal values, in Perl they
776 function as operators, providing various kinds of interpolating and
777 pattern matching capabilities. Perl provides customary quote characters
778 for these behaviors, but also provides a way for you to choose your
779 quote character for any of them. In the following table, a C<{}> represents
780 any pair of delimiters you choose.
782 Customary Generic Meaning Interpolates
787 // m{} Pattern match yes*
789 s{}{} Substitution yes*
790 tr{}{} Transliteration no (but see below)
793 * unless the delimiter is ''.
795 Non-bracketing delimiters use the same character fore and aft, but the four
796 sorts of brackets (round, angle, square, curly) will all nest, which means
805 Note, however, that this does not always work for quoting Perl code:
807 $s = q{ if($a eq "}") ... }; # WRONG
809 is a syntax error. The C<Text::Balanced> module (from CPAN, and
810 starting from Perl 5.8 part of the standard distribution) is able
813 There can be whitespace between the operator and the quoting
814 characters, except when C<#> is being used as the quoting character.
815 C<q#foo#> is parsed as the string C<foo>, while C<q #foo#> is the
816 operator C<q> followed by a comment. Its argument will be taken
817 from the next line. This allows you to write:
819 s {foo} # Replace foo
822 The following escape sequences are available in constructs that interpolate
823 and in transliterations.
830 \a alarm (bell) (BEL)
832 \033 octal char (ESC)
834 \x{263a} wide hex char (SMILEY)
835 \c[ control char (ESC)
836 \N{name} named Unicode character
838 B<NOTE>: Unlike C and other languages, Perl has no \v escape sequence for
839 the vertical tab (VT - ASCII 11).
841 The following escape sequences are available in constructs that interpolate
842 but not in transliterations.
844 \l lowercase next char
845 \u uppercase next char
848 \E end case modification
849 \Q quote non-word characters till \E
851 If C<use locale> is in effect, the case map used by C<\l>, C<\L>,
852 C<\u> and C<\U> is taken from the current locale. See L<perllocale>.
853 If Unicode (for example, C<\N{}> or wide hex characters of 0x100 or
854 beyond) is being used, the case map used by C<\l>, C<\L>, C<\u> and
855 C<\U> is as defined by Unicode. For documentation of C<\N{name}>,
858 All systems use the virtual C<"\n"> to represent a line terminator,
859 called a "newline". There is no such thing as an unvarying, physical
860 newline character. It is only an illusion that the operating system,
861 device drivers, C libraries, and Perl all conspire to preserve. Not all
862 systems read C<"\r"> as ASCII CR and C<"\n"> as ASCII LF. For example,
863 on a Mac, these are reversed, and on systems without line terminator,
864 printing C<"\n"> may emit no actual data. In general, use C<"\n"> when
865 you mean a "newline" for your system, but use the literal ASCII when you
866 need an exact character. For example, most networking protocols expect
867 and prefer a CR+LF (C<"\015\012"> or C<"\cM\cJ">) for line terminators,
868 and although they often accept just C<"\012">, they seldom tolerate just
869 C<"\015">. If you get in the habit of using C<"\n"> for networking,
870 you may be burned some day.
872 For constructs that do interpolate, variables beginning with "C<$>"
873 or "C<@>" are interpolated. Subscripted variables such as C<$a[3]> or
874 C<< $href->{key}[0] >> are also interpolated, as are array and hash slices.
875 But method calls such as C<< $obj->meth >> are not.
877 Interpolating an array or slice interpolates the elements in order,
878 separated by the value of C<$">, so is equivalent to interpolating
879 C<join $", @array>. "Punctuation" arrays such as C<@+> are only
880 interpolated if the name is enclosed in braces C<@{+}>.
882 You cannot include a literal C<$> or C<@> within a C<\Q> sequence.
883 An unescaped C<$> or C<@> interpolates the corresponding variable,
884 while escaping will cause the literal string C<\$> to be inserted.
885 You'll need to write something like C<m/\Quser\E\@\Qhost/>.
887 Patterns are subject to an additional level of interpretation as a
888 regular expression. This is done as a second pass, after variables are
889 interpolated, so that regular expressions may be incorporated into the
890 pattern from the variables. If this is not what you want, use C<\Q> to
891 interpolate a variable literally.
893 Apart from the behavior described above, Perl does not expand
894 multiple levels of interpolation. In particular, contrary to the
895 expectations of shell programmers, back-quotes do I<NOT> interpolate
896 within double quotes, nor do single quotes impede evaluation of
897 variables when used within double quotes.
899 =head2 Regexp Quote-Like Operators
901 Here are the quote-like operators that apply to pattern
902 matching and related activities.
908 This is just like the C</pattern/> search, except that it matches only
909 once between calls to the reset() operator. This is a useful
910 optimization when you want to see only the first occurrence of
911 something in each file of a set of files, for instance. Only C<??>
912 patterns local to the current package are reset.
916 # blank line between header and body
919 reset if eof; # clear ?? status for next file
922 This usage is vaguely deprecated, which means it just might possibly
923 be removed in some distant future version of Perl, perhaps somewhere
924 around the year 2168.
926 =item m/PATTERN/cgimosx
928 =item /PATTERN/cgimosx
930 Searches a string for a pattern match, and in scalar context returns
931 true if it succeeds, false if it fails. If no string is specified
932 via the C<=~> or C<!~> operator, the $_ string is searched. (The
933 string specified with C<=~> need not be an lvalue--it may be the
934 result of an expression evaluation, but remember the C<=~> binds
935 rather tightly.) See also L<perlre>. See L<perllocale> for
936 discussion of additional considerations that apply when C<use locale>
941 c Do not reset search position on a failed match when /g is in effect.
942 g Match globally, i.e., find all occurrences.
943 i Do case-insensitive pattern matching.
944 m Treat string as multiple lines.
945 o Compile pattern only once.
946 s Treat string as single line.
947 x Use extended regular expressions.
949 If "/" is the delimiter then the initial C<m> is optional. With the C<m>
950 you can use any pair of non-alphanumeric, non-whitespace characters
951 as delimiters. This is particularly useful for matching path names
952 that contain "/", to avoid LTS (leaning toothpick syndrome). If "?" is
953 the delimiter, then the match-only-once rule of C<?PATTERN?> applies.
954 If "'" is the delimiter, no interpolation is performed on the PATTERN.
956 PATTERN may contain variables, which will be interpolated (and the
957 pattern recompiled) every time the pattern search is evaluated, except
958 for when the delimiter is a single quote. (Note that C<$(>, C<$)>, and
959 C<$|> are not interpolated because they look like end-of-string tests.)
960 If you want such a pattern to be compiled only once, add a C</o> after
961 the trailing delimiter. This avoids expensive run-time recompilations,
962 and is useful when the value you are interpolating won't change over
963 the life of the script. However, mentioning C</o> constitutes a promise
964 that you won't change the variables in the pattern. If you change them,
965 Perl won't even notice. See also L<"qr/STRING/imosx">.
967 If the PATTERN evaluates to the empty string, the last
968 I<successfully> matched regular expression is used instead. In this
969 case, only the C<g> and C<c> flags on the empty pattern is honoured -
970 the other flags are taken from the original pattern. If no match has
971 previously succeeded, this will (silently) act instead as a genuine
972 empty pattern (which will always match).
974 Note that it's possible to confuse Perl into thinking C<//> (the empty
975 regex) is really C<//> (the defined-or operator). Perl is usually pretty
976 good about this, but some pathological cases might trigger this, such as
977 C<$a///> (is that C<($a) / (//)> or C<$a // />?) and C<print $fh //>
978 (C<print $fh(//> or C<print($fh //>?). In all of these examples, Perl
979 will assume you meant defined-or. If you meant the empty regex, just
980 use parentheses or spaces to disambiguate, or even prefix the empty
981 regex with an C<m> (so C<//> becomes C<m//>).
983 If the C</g> option is not used, C<m//> in list context returns a
984 list consisting of the subexpressions matched by the parentheses in the
985 pattern, i.e., (C<$1>, C<$2>, C<$3>...). (Note that here C<$1> etc. are
986 also set, and that this differs from Perl 4's behavior.) When there are
987 no parentheses in the pattern, the return value is the list C<(1)> for
988 success. With or without parentheses, an empty list is returned upon
993 open(TTY, '/dev/tty');
994 <TTY> =~ /^y/i && foo(); # do foo if desired
996 if (/Version: *([0-9.]*)/) { $version = $1; }
998 next if m#^/usr/spool/uucp#;
1003 print if /$arg/o; # compile only once
1006 if (($F1, $F2, $Etc) = ($foo =~ /^(\S+)\s+(\S+)\s*(.*)/))
1008 This last example splits $foo into the first two words and the
1009 remainder of the line, and assigns those three fields to $F1, $F2, and
1010 $Etc. The conditional is true if any variables were assigned, i.e., if
1011 the pattern matched.
1013 The C</g> modifier specifies global pattern matching--that is,
1014 matching as many times as possible within the string. How it behaves
1015 depends on the context. In list context, it returns a list of the
1016 substrings matched by any capturing parentheses in the regular
1017 expression. If there are no parentheses, it returns a list of all
1018 the matched strings, as if there were parentheses around the whole
1021 In scalar context, each execution of C<m//g> finds the next match,
1022 returning true if it matches, and false if there is no further match.
1023 The position after the last match can be read or set using the pos()
1024 function; see L<perlfunc/pos>. A failed match normally resets the
1025 search position to the beginning of the string, but you can avoid that
1026 by adding the C</c> modifier (e.g. C<m//gc>). Modifying the target
1027 string also resets the search position.
1029 You can intermix C<m//g> matches with C<m/\G.../g>, where C<\G> is a
1030 zero-width assertion that matches the exact position where the previous
1031 C<m//g>, if any, left off. Without the C</g> modifier, the C<\G> assertion
1032 still anchors at pos(), but the match is of course only attempted once.
1033 Using C<\G> without C</g> on a target string that has not previously had a
1034 C</g> match applied to it is the same as using the C<\A> assertion to match
1035 the beginning of the string. Note also that, currently, C<\G> is only
1036 properly supported when anchored at the very beginning of the pattern.
1041 ($one,$five,$fifteen) = (`uptime` =~ /(\d+\.\d+)/g);
1045 while (defined($paragraph = <>)) {
1046 while ($paragraph =~ /[a-z]['")]*[.!?]+['")]*\s/g) {
1050 print "$sentences\n";
1052 # using m//gc with \G
1056 print $1 while /(o)/gc; print "', pos=", pos, "\n";
1058 print $1 if /\G(q)/gc; print "', pos=", pos, "\n";
1060 print $1 while /(p)/gc; print "', pos=", pos, "\n";
1062 print "Final: '$1', pos=",pos,"\n" if /\G(.)/;
1064 The last example should print:
1074 Notice that the final match matched C<q> instead of C<p>, which a match
1075 without the C<\G> anchor would have done. Also note that the final match
1076 did not update C<pos> -- C<pos> is only updated on a C</g> match. If the
1077 final match did indeed match C<p>, it's a good bet that you're running an
1078 older (pre-5.6.0) Perl.
1080 A useful idiom for C<lex>-like scanners is C</\G.../gc>. You can
1081 combine several regexps like this to process a string part-by-part,
1082 doing different actions depending on which regexp matched. Each
1083 regexp tries to match where the previous one leaves off.
1086 $url = new URI::URL "http://www/"; die if $url eq "xXx";
1090 print(" digits"), redo LOOP if /\G\d+\b[,.;]?\s*/gc;
1091 print(" lowercase"), redo LOOP if /\G[a-z]+\b[,.;]?\s*/gc;
1092 print(" UPPERCASE"), redo LOOP if /\G[A-Z]+\b[,.;]?\s*/gc;
1093 print(" Capitalized"), redo LOOP if /\G[A-Z][a-z]+\b[,.;]?\s*/gc;
1094 print(" MiXeD"), redo LOOP if /\G[A-Za-z]+\b[,.;]?\s*/gc;
1095 print(" alphanumeric"), redo LOOP if /\G[A-Za-z0-9]+\b[,.;]?\s*/gc;
1096 print(" line-noise"), redo LOOP if /\G[^A-Za-z0-9]+/gc;
1097 print ". That's all!\n";
1100 Here is the output (split into several lines):
1102 line-noise lowercase line-noise lowercase UPPERCASE line-noise
1103 UPPERCASE line-noise lowercase line-noise lowercase line-noise
1104 lowercase lowercase line-noise lowercase lowercase line-noise
1105 MiXeD line-noise. That's all!
1111 A single-quoted, literal string. A backslash represents a backslash
1112 unless followed by the delimiter or another backslash, in which case
1113 the delimiter or backslash is interpolated.
1115 $foo = q!I said, "You said, 'She said it.'"!;
1116 $bar = q('This is it.');
1117 $baz = '\n'; # a two-character string
1123 A double-quoted, interpolated string.
1126 (*** The previous line contains the naughty word "$1".\n)
1127 if /\b(tcl|java|python)\b/i; # :-)
1128 $baz = "\n"; # a one-character string
1130 =item qr/STRING/imosx
1132 This operator quotes (and possibly compiles) its I<STRING> as a regular
1133 expression. I<STRING> is interpolated the same way as I<PATTERN>
1134 in C<m/PATTERN/>. If "'" is used as the delimiter, no interpolation
1135 is done. Returns a Perl value which may be used instead of the
1136 corresponding C</STRING/imosx> expression.
1140 $rex = qr/my.STRING/is;
1147 The result may be used as a subpattern in a match:
1150 $string =~ /foo${re}bar/; # can be interpolated in other patterns
1151 $string =~ $re; # or used standalone
1152 $string =~ /$re/; # or this way
1154 Since Perl may compile the pattern at the moment of execution of qr()
1155 operator, using qr() may have speed advantages in some situations,
1156 notably if the result of qr() is used standalone:
1159 my $patterns = shift;
1160 my @compiled = map qr/$_/i, @$patterns;
1163 foreach my $pat (@compiled) {
1164 $success = 1, last if /$pat/;
1170 Precompilation of the pattern into an internal representation at
1171 the moment of qr() avoids a need to recompile the pattern every
1172 time a match C</$pat/> is attempted. (Perl has many other internal
1173 optimizations, but none would be triggered in the above example if
1174 we did not use qr() operator.)
1178 i Do case-insensitive pattern matching.
1179 m Treat string as multiple lines.
1180 o Compile pattern only once.
1181 s Treat string as single line.
1182 x Use extended regular expressions.
1184 See L<perlre> for additional information on valid syntax for STRING, and
1185 for a detailed look at the semantics of regular expressions.
1191 A string which is (possibly) interpolated and then executed as a
1192 system command with C</bin/sh> or its equivalent. Shell wildcards,
1193 pipes, and redirections will be honored. The collected standard
1194 output of the command is returned; standard error is unaffected. In
1195 scalar context, it comes back as a single (potentially multi-line)
1196 string, or undef if the command failed. In list context, returns a
1197 list of lines (however you've defined lines with $/ or
1198 $INPUT_RECORD_SEPARATOR), or an empty list if the command failed.
1200 Because backticks do not affect standard error, use shell file descriptor
1201 syntax (assuming the shell supports this) if you care to address this.
1202 To capture a command's STDERR and STDOUT together:
1204 $output = `cmd 2>&1`;
1206 To capture a command's STDOUT but discard its STDERR:
1208 $output = `cmd 2>/dev/null`;
1210 To capture a command's STDERR but discard its STDOUT (ordering is
1213 $output = `cmd 2>&1 1>/dev/null`;
1215 To exchange a command's STDOUT and STDERR in order to capture the STDERR
1216 but leave its STDOUT to come out the old STDERR:
1218 $output = `cmd 3>&1 1>&2 2>&3 3>&-`;
1220 To read both a command's STDOUT and its STDERR separately, it's easiest
1221 to redirect them separately to files, and then read from those files
1222 when the program is done:
1224 system("program args 1>program.stdout 2>program.stderr");
1226 Using single-quote as a delimiter protects the command from Perl's
1227 double-quote interpolation, passing it on to the shell instead:
1229 $perl_info = qx(ps $$); # that's Perl's $$
1230 $shell_info = qx'ps $$'; # that's the new shell's $$
1232 How that string gets evaluated is entirely subject to the command
1233 interpreter on your system. On most platforms, you will have to protect
1234 shell metacharacters if you want them treated literally. This is in
1235 practice difficult to do, as it's unclear how to escape which characters.
1236 See L<perlsec> for a clean and safe example of a manual fork() and exec()
1237 to emulate backticks safely.
1239 On some platforms (notably DOS-like ones), the shell may not be
1240 capable of dealing with multiline commands, so putting newlines in
1241 the string may not get you what you want. You may be able to evaluate
1242 multiple commands in a single line by separating them with the command
1243 separator character, if your shell supports that (e.g. C<;> on many Unix
1244 shells; C<&> on the Windows NT C<cmd> shell).
1246 Beginning with v5.6.0, Perl will attempt to flush all files opened for
1247 output before starting the child process, but this may not be supported
1248 on some platforms (see L<perlport>). To be safe, you may need to set
1249 C<$|> ($AUTOFLUSH in English) or call the C<autoflush()> method of
1250 C<IO::Handle> on any open handles.
1252 Beware that some command shells may place restrictions on the length
1253 of the command line. You must ensure your strings don't exceed this
1254 limit after any necessary interpolations. See the platform-specific
1255 release notes for more details about your particular environment.
1257 Using this operator can lead to programs that are difficult to port,
1258 because the shell commands called vary between systems, and may in
1259 fact not be present at all. As one example, the C<type> command under
1260 the POSIX shell is very different from the C<type> command under DOS.
1261 That doesn't mean you should go out of your way to avoid backticks
1262 when they're the right way to get something done. Perl was made to be
1263 a glue language, and one of the things it glues together is commands.
1264 Just understand what you're getting yourself into.
1266 See L<"I/O Operators"> for more discussion.
1270 Evaluates to a list of the words extracted out of STRING, using embedded
1271 whitespace as the word delimiters. It can be understood as being roughly
1274 split(' ', q/STRING/);
1276 the differences being that it generates a real list at compile time, and
1277 in scalar context it returns the last element in the list. So
1282 is semantically equivalent to the list:
1286 Some frequently seen examples:
1288 use POSIX qw( setlocale localeconv )
1289 @EXPORT = qw( foo bar baz );
1291 A common mistake is to try to separate the words with comma or to
1292 put comments into a multi-line C<qw>-string. For this reason, the
1293 C<use warnings> pragma and the B<-w> switch (that is, the C<$^W> variable)
1294 produces warnings if the STRING contains the "," or the "#" character.
1296 =item s/PATTERN/REPLACEMENT/egimosx
1298 Searches a string for a pattern, and if found, replaces that pattern
1299 with the replacement text and returns the number of substitutions
1300 made. Otherwise it returns false (specifically, the empty string).
1302 If no string is specified via the C<=~> or C<!~> operator, the C<$_>
1303 variable is searched and modified. (The string specified with C<=~> must
1304 be scalar variable, an array element, a hash element, or an assignment
1305 to one of those, i.e., an lvalue.)
1307 If the delimiter chosen is a single quote, no interpolation is
1308 done on either the PATTERN or the REPLACEMENT. Otherwise, if the
1309 PATTERN contains a $ that looks like a variable rather than an
1310 end-of-string test, the variable will be interpolated into the pattern
1311 at run-time. If you want the pattern compiled only once the first time
1312 the variable is interpolated, use the C</o> option. If the pattern
1313 evaluates to the empty string, the last successfully executed regular
1314 expression is used instead. See L<perlre> for further explanation on these.
1315 See L<perllocale> for discussion of additional considerations that apply
1316 when C<use locale> is in effect.
1320 e Evaluate the right side as an expression.
1321 g Replace globally, i.e., all occurrences.
1322 i Do case-insensitive pattern matching.
1323 m Treat string as multiple lines.
1324 o Compile pattern only once.
1325 s Treat string as single line.
1326 x Use extended regular expressions.
1328 Any non-alphanumeric, non-whitespace delimiter may replace the
1329 slashes. If single quotes are used, no interpretation is done on the
1330 replacement string (the C</e> modifier overrides this, however). Unlike
1331 Perl 4, Perl 5 treats backticks as normal delimiters; the replacement
1332 text is not evaluated as a command. If the
1333 PATTERN is delimited by bracketing quotes, the REPLACEMENT has its own
1334 pair of quotes, which may or may not be bracketing quotes, e.g.,
1335 C<s(foo)(bar)> or C<< s<foo>/bar/ >>. A C</e> will cause the
1336 replacement portion to be treated as a full-fledged Perl expression
1337 and evaluated right then and there. It is, however, syntax checked at
1338 compile-time. A second C<e> modifier will cause the replacement portion
1339 to be C<eval>ed before being run as a Perl expression.
1343 s/\bgreen\b/mauve/g; # don't change wintergreen
1345 $path =~ s|/usr/bin|/usr/local/bin|;
1347 s/Login: $foo/Login: $bar/; # run-time pattern
1349 ($foo = $bar) =~ s/this/that/; # copy first, then change
1351 $count = ($paragraph =~ s/Mister\b/Mr./g); # get change-count
1354 s/\d+/$&*2/e; # yields 'abc246xyz'
1355 s/\d+/sprintf("%5d",$&)/e; # yields 'abc 246xyz'
1356 s/\w/$& x 2/eg; # yields 'aabbcc 224466xxyyzz'
1358 s/%(.)/$percent{$1}/g; # change percent escapes; no /e
1359 s/%(.)/$percent{$1} || $&/ge; # expr now, so /e
1360 s/^=(\w+)/&pod($1)/ge; # use function call
1362 # expand variables in $_, but dynamics only, using
1363 # symbolic dereferencing
1366 # Add one to the value of any numbers in the string
1369 # This will expand any embedded scalar variable
1370 # (including lexicals) in $_ : First $1 is interpolated
1371 # to the variable name, and then evaluated
1374 # Delete (most) C comments.
1376 /\* # Match the opening delimiter.
1377 .*? # Match a minimal number of characters.
1378 \*/ # Match the closing delimiter.
1381 s/^\s*(.*?)\s*$/$1/; # trim white space in $_, expensively
1383 for ($variable) { # trim white space in $variable, cheap
1388 s/([^ ]*) *([^ ]*)/$2 $1/; # reverse 1st two fields
1390 Note the use of $ instead of \ in the last example. Unlike
1391 B<sed>, we use the \<I<digit>> form in only the left hand side.
1392 Anywhere else it's $<I<digit>>.
1394 Occasionally, you can't use just a C</g> to get all the changes
1395 to occur that you might want. Here are two common cases:
1397 # put commas in the right places in an integer
1398 1 while s/(\d)(\d\d\d)(?!\d)/$1,$2/g;
1400 # expand tabs to 8-column spacing
1401 1 while s/\t+/' ' x (length($&)*8 - length($`)%8)/e;
1403 =item tr/SEARCHLIST/REPLACEMENTLIST/cds
1405 =item y/SEARCHLIST/REPLACEMENTLIST/cds
1407 Transliterates all occurrences of the characters found in the search list
1408 with the corresponding character in the replacement list. It returns
1409 the number of characters replaced or deleted. If no string is
1410 specified via the =~ or !~ operator, the $_ string is transliterated. (The
1411 string specified with =~ must be a scalar variable, an array element, a
1412 hash element, or an assignment to one of those, i.e., an lvalue.)
1414 A character range may be specified with a hyphen, so C<tr/A-J/0-9/>
1415 does the same replacement as C<tr/ACEGIBDFHJ/0246813579/>.
1416 For B<sed> devotees, C<y> is provided as a synonym for C<tr>. If the
1417 SEARCHLIST is delimited by bracketing quotes, the REPLACEMENTLIST has
1418 its own pair of quotes, which may or may not be bracketing quotes,
1419 e.g., C<tr[A-Z][a-z]> or C<tr(+\-*/)/ABCD/>.
1421 Note that C<tr> does B<not> do regular expression character classes
1422 such as C<\d> or C<[:lower:]>. The <tr> operator is not equivalent to
1423 the tr(1) utility. If you want to map strings between lower/upper
1424 cases, see L<perlfunc/lc> and L<perlfunc/uc>, and in general consider
1425 using the C<s> operator if you need regular expressions.
1427 Note also that the whole range idea is rather unportable between
1428 character sets--and even within character sets they may cause results
1429 you probably didn't expect. A sound principle is to use only ranges
1430 that begin from and end at either alphabets of equal case (a-e, A-E),
1431 or digits (0-4). Anything else is unsafe. If in doubt, spell out the
1432 character sets in full.
1436 c Complement the SEARCHLIST.
1437 d Delete found but unreplaced characters.
1438 s Squash duplicate replaced characters.
1440 If the C</c> modifier is specified, the SEARCHLIST character set
1441 is complemented. If the C</d> modifier is specified, any characters
1442 specified by SEARCHLIST not found in REPLACEMENTLIST are deleted.
1443 (Note that this is slightly more flexible than the behavior of some
1444 B<tr> programs, which delete anything they find in the SEARCHLIST,
1445 period.) If the C</s> modifier is specified, sequences of characters
1446 that were transliterated to the same character are squashed down
1447 to a single instance of the character.
1449 If the C</d> modifier is used, the REPLACEMENTLIST is always interpreted
1450 exactly as specified. Otherwise, if the REPLACEMENTLIST is shorter
1451 than the SEARCHLIST, the final character is replicated till it is long
1452 enough. If the REPLACEMENTLIST is empty, the SEARCHLIST is replicated.
1453 This latter is useful for counting characters in a class or for
1454 squashing character sequences in a class.
1458 $ARGV[1] =~ tr/A-Z/a-z/; # canonicalize to lower case
1460 $cnt = tr/*/*/; # count the stars in $_
1462 $cnt = $sky =~ tr/*/*/; # count the stars in $sky
1464 $cnt = tr/0-9//; # count the digits in $_
1466 tr/a-zA-Z//s; # bookkeeper -> bokeper
1468 ($HOST = $host) =~ tr/a-z/A-Z/;
1470 tr/a-zA-Z/ /cs; # change non-alphas to single space
1473 [\000-\177]; # delete 8th bit
1475 If multiple transliterations are given for a character, only the
1480 will transliterate any A to X.
1482 Because the transliteration table is built at compile time, neither
1483 the SEARCHLIST nor the REPLACEMENTLIST are subjected to double quote
1484 interpolation. That means that if you want to use variables, you
1487 eval "tr/$oldlist/$newlist/";
1490 eval "tr/$oldlist/$newlist/, 1" or die $@;
1494 A line-oriented form of quoting is based on the shell "here-document"
1495 syntax. Following a C<< << >> you specify a string to terminate
1496 the quoted material, and all lines following the current line down to
1497 the terminating string are the value of the item. The terminating
1498 string may be either an identifier (a word), or some quoted text. If
1499 quoted, the type of quotes you use determines the treatment of the
1500 text, just as in regular quoting. An unquoted identifier works like
1501 double quotes. There must be no space between the C<< << >> and
1502 the identifier, unless the identifier is quoted. (If you put a space it
1503 will be treated as a null identifier, which is valid, and matches the first
1504 empty line.) The terminating string must appear by itself (unquoted and
1505 with no surrounding whitespace) on the terminating line.
1508 The price is $Price.
1511 print << "EOF"; # same as above
1512 The price is $Price.
1515 print << `EOC`; # execute commands
1520 print <<"foo", <<"bar"; # you can stack them
1526 myfunc(<< "THIS", 23, <<'THAT');
1533 Just don't forget that you have to put a semicolon on the end
1534 to finish the statement, as Perl doesn't know you're not going to
1542 If you want your here-docs to be indented with the
1543 rest of the code, you'll need to remove leading whitespace
1544 from each line manually:
1546 ($quote = <<'FINIS') =~ s/^\s+//gm;
1547 The Road goes ever on and on,
1548 down from the door where it began.
1551 If you use a here-doc within a delimited construct, such as in C<s///eg>,
1552 the quoted material must come on the lines following the final delimiter.
1567 If the terminating identifier is on the last line of the program, you
1568 must be sure there is a newline after it; otherwise, Perl will give the
1569 warning B<Can't find string terminator "END" anywhere before EOF...>.
1571 Additionally, the quoting rules for the identifier are not related to
1572 Perl's quoting rules -- C<q()>, C<qq()>, and the like are not supported
1573 in place of C<''> and C<"">, and the only interpolation is for backslashing
1574 the quoting character:
1576 print << "abc\"def";
1580 Finally, quoted strings cannot span multiple lines. The general rule is
1581 that the identifier must be a string literal. Stick with that, and you
1586 =head2 Gory details of parsing quoted constructs
1588 When presented with something that might have several different
1589 interpretations, Perl uses the B<DWIM> (that's "Do What I Mean")
1590 principle to pick the most probable interpretation. This strategy
1591 is so successful that Perl programmers often do not suspect the
1592 ambivalence of what they write. But from time to time, Perl's
1593 notions differ substantially from what the author honestly meant.
1595 This section hopes to clarify how Perl handles quoted constructs.
1596 Although the most common reason to learn this is to unravel labyrinthine
1597 regular expressions, because the initial steps of parsing are the
1598 same for all quoting operators, they are all discussed together.
1600 The most important Perl parsing rule is the first one discussed
1601 below: when processing a quoted construct, Perl first finds the end
1602 of that construct, then interprets its contents. If you understand
1603 this rule, you may skip the rest of this section on the first
1604 reading. The other rules are likely to contradict the user's
1605 expectations much less frequently than this first one.
1607 Some passes discussed below are performed concurrently, but because
1608 their results are the same, we consider them individually. For different
1609 quoting constructs, Perl performs different numbers of passes, from
1610 one to five, but these passes are always performed in the same order.
1614 =item Finding the end
1616 The first pass is finding the end of the quoted construct, whether
1617 it be a multicharacter delimiter C<"\nEOF\n"> in the C<<<EOF>
1618 construct, a C</> that terminates a C<qq//> construct, a C<]> which
1619 terminates C<qq[]> construct, or a C<< > >> which terminates a
1620 fileglob started with C<< < >>.
1622 When searching for single-character non-pairing delimiters, such
1623 as C</>, combinations of C<\\> and C<\/> are skipped. However,
1624 when searching for single-character pairing delimiter like C<[>,
1625 combinations of C<\\>, C<\]>, and C<\[> are all skipped, and nested
1626 C<[>, C<]> are skipped as well. When searching for multicharacter
1627 delimiters, nothing is skipped.
1629 For constructs with three-part delimiters (C<s///>, C<y///>, and
1630 C<tr///>), the search is repeated once more.
1632 During this search no attention is paid to the semantics of the construct.
1635 "$hash{"$foo/$bar"}"
1640 bar # NOT a comment, this slash / terminated m//!
1643 do not form legal quoted expressions. The quoted part ends on the
1644 first C<"> and C</>, and the rest happens to be a syntax error.
1645 Because the slash that terminated C<m//> was followed by a C<SPACE>,
1646 the example above is not C<m//x>, but rather C<m//> with no C</x>
1647 modifier. So the embedded C<#> is interpreted as a literal C<#>.
1649 =item Removal of backslashes before delimiters
1651 During the second pass, text between the starting and ending
1652 delimiters is copied to a safe location, and the C<\> is removed
1653 from combinations consisting of C<\> and delimiter--or delimiters,
1654 meaning both starting and ending delimiters will should these differ.
1655 This removal does not happen for multi-character delimiters.
1656 Note that the combination C<\\> is left intact, just as it was.
1658 Starting from this step no information about the delimiters is
1663 The next step is interpolation in the text obtained, which is now
1664 delimiter-independent. There are four different cases.
1668 =item C<<<'EOF'>, C<m''>, C<s'''>, C<tr///>, C<y///>
1670 No interpolation is performed.
1674 The only interpolation is removal of C<\> from pairs C<\\>.
1676 =item C<"">, C<``>, C<qq//>, C<qx//>, C<< <file*glob> >>
1678 C<\Q>, C<\U>, C<\u>, C<\L>, C<\l> (possibly paired with C<\E>) are
1679 converted to corresponding Perl constructs. Thus, C<"$foo\Qbaz$bar">
1680 is converted to C<$foo . (quotemeta("baz" . $bar))> internally.
1681 The other combinations are replaced with appropriate expansions.
1683 Let it be stressed that I<whatever falls between C<\Q> and C<\E>>
1684 is interpolated in the usual way. Something like C<"\Q\\E"> has
1685 no C<\E> inside. instead, it has C<\Q>, C<\\>, and C<E>, so the
1686 result is the same as for C<"\\\\E">. As a general rule, backslashes
1687 between C<\Q> and C<\E> may lead to counterintuitive results. So,
1688 C<"\Q\t\E"> is converted to C<quotemeta("\t")>, which is the same
1689 as C<"\\\t"> (since TAB is not alphanumeric). Note also that:
1694 may be closer to the conjectural I<intention> of the writer of C<"\Q\t\E">.
1696 Interpolated scalars and arrays are converted internally to the C<join> and
1697 C<.> catenation operations. Thus, C<"$foo XXX '@arr'"> becomes:
1699 $foo . " XXX '" . (join $", @arr) . "'";
1701 All operations above are performed simultaneously, left to right.
1703 Because the result of C<"\Q STRING \E"> has all metacharacters
1704 quoted, there is no way to insert a literal C<$> or C<@> inside a
1705 C<\Q\E> pair. If protected by C<\>, C<$> will be quoted to became
1706 C<"\\\$">; if not, it is interpreted as the start of an interpolated
1709 Note also that the interpolation code needs to make a decision on
1710 where the interpolated scalar ends. For instance, whether
1711 C<< "a $b -> {c}" >> really means:
1713 "a " . $b . " -> {c}";
1719 Most of the time, the longest possible text that does not include
1720 spaces between components and which contains matching braces or
1721 brackets. because the outcome may be determined by voting based
1722 on heuristic estimators, the result is not strictly predictable.
1723 Fortunately, it's usually correct for ambiguous cases.
1725 =item C<?RE?>, C</RE/>, C<m/RE/>, C<s/RE/foo/>,
1727 Processing of C<\Q>, C<\U>, C<\u>, C<\L>, C<\l>, and interpolation
1728 happens (almost) as with C<qq//> constructs, but the substitution
1729 of C<\> followed by RE-special chars (including C<\>) is not
1730 performed. Moreover, inside C<(?{BLOCK})>, C<(?# comment )>, and
1731 a C<#>-comment in a C<//x>-regular expression, no processing is
1732 performed whatsoever. This is the first step at which the presence
1733 of the C<//x> modifier is relevant.
1735 Interpolation has several quirks: C<$|>, C<$(>, and C<$)> are not
1736 interpolated, and constructs C<$var[SOMETHING]> are voted (by several
1737 different estimators) to be either an array element or C<$var>
1738 followed by an RE alternative. This is where the notation
1739 C<${arr[$bar]}> comes handy: C</${arr[0-9]}/> is interpreted as
1740 array element C<-9>, not as a regular expression from the variable
1741 C<$arr> followed by a digit, which would be the interpretation of
1742 C</$arr[0-9]/>. Since voting among different estimators may occur,
1743 the result is not predictable.
1745 It is at this step that C<\1> is begrudgingly converted to C<$1> in
1746 the replacement text of C<s///> to correct the incorrigible
1747 I<sed> hackers who haven't picked up the saner idiom yet. A warning
1748 is emitted if the C<use warnings> pragma or the B<-w> command-line flag
1749 (that is, the C<$^W> variable) was set.
1751 The lack of processing of C<\\> creates specific restrictions on
1752 the post-processed text. If the delimiter is C</>, one cannot get
1753 the combination C<\/> into the result of this step. C</> will
1754 finish the regular expression, C<\/> will be stripped to C</> on
1755 the previous step, and C<\\/> will be left as is. Because C</> is
1756 equivalent to C<\/> inside a regular expression, this does not
1757 matter unless the delimiter happens to be character special to the
1758 RE engine, such as in C<s*foo*bar*>, C<m[foo]>, or C<?foo?>; or an
1759 alphanumeric char, as in:
1763 In the RE above, which is intentionally obfuscated for illustration, the
1764 delimiter is C<m>, the modifier is C<mx>, and after backslash-removal the
1765 RE is the same as for C<m/ ^ a \s* b /mx>. There's more than one
1766 reason you're encouraged to restrict your delimiters to non-alphanumeric,
1767 non-whitespace choices.
1771 This step is the last one for all constructs except regular expressions,
1772 which are processed further.
1774 =item Interpolation of regular expressions
1776 Previous steps were performed during the compilation of Perl code,
1777 but this one happens at run time--although it may be optimized to
1778 be calculated at compile time if appropriate. After preprocessing
1779 described above, and possibly after evaluation if catenation,
1780 joining, casing translation, or metaquoting are involved, the
1781 resulting I<string> is passed to the RE engine for compilation.
1783 Whatever happens in the RE engine might be better discussed in L<perlre>,
1784 but for the sake of continuity, we shall do so here.
1786 This is another step where the presence of the C<//x> modifier is
1787 relevant. The RE engine scans the string from left to right and
1788 converts it to a finite automaton.
1790 Backslashed characters are either replaced with corresponding
1791 literal strings (as with C<\{>), or else they generate special nodes
1792 in the finite automaton (as with C<\b>). Characters special to the
1793 RE engine (such as C<|>) generate corresponding nodes or groups of
1794 nodes. C<(?#...)> comments are ignored. All the rest is either
1795 converted to literal strings to match, or else is ignored (as is
1796 whitespace and C<#>-style comments if C<//x> is present).
1798 Parsing of the bracketed character class construct, C<[...]>, is
1799 rather different than the rule used for the rest of the pattern.
1800 The terminator of this construct is found using the same rules as
1801 for finding the terminator of a C<{}>-delimited construct, the only
1802 exception being that C<]> immediately following C<[> is treated as
1803 though preceded by a backslash. Similarly, the terminator of
1804 C<(?{...})> is found using the same rules as for finding the
1805 terminator of a C<{}>-delimited construct.
1807 It is possible to inspect both the string given to RE engine and the
1808 resulting finite automaton. See the arguments C<debug>/C<debugcolor>
1809 in the C<use L<re>> pragma, as well as Perl's B<-Dr> command-line
1810 switch documented in L<perlrun/"Command Switches">.
1812 =item Optimization of regular expressions
1814 This step is listed for completeness only. Since it does not change
1815 semantics, details of this step are not documented and are subject
1816 to change without notice. This step is performed over the finite
1817 automaton that was generated during the previous pass.
1819 It is at this stage that C<split()> silently optimizes C</^/> to
1824 =head2 I/O Operators
1826 There are several I/O operators you should know about.
1828 A string enclosed by backticks (grave accents) first undergoes
1829 double-quote interpolation. It is then interpreted as an external
1830 command, and the output of that command is the value of the
1831 backtick string, like in a shell. In scalar context, a single string
1832 consisting of all output is returned. In list context, a list of
1833 values is returned, one per line of output. (You can set C<$/> to use
1834 a different line terminator.) The command is executed each time the
1835 pseudo-literal is evaluated. The status value of the command is
1836 returned in C<$?> (see L<perlvar> for the interpretation of C<$?>).
1837 Unlike in B<csh>, no translation is done on the return data--newlines
1838 remain newlines. Unlike in any of the shells, single quotes do not
1839 hide variable names in the command from interpretation. To pass a
1840 literal dollar-sign through to the shell you need to hide it with a
1841 backslash. The generalized form of backticks is C<qx//>. (Because
1842 backticks always undergo shell expansion as well, see L<perlsec> for
1845 In scalar context, evaluating a filehandle in angle brackets yields
1846 the next line from that file (the newline, if any, included), or
1847 C<undef> at end-of-file or on error. When C<$/> is set to C<undef>
1848 (sometimes known as file-slurp mode) and the file is empty, it
1849 returns C<''> the first time, followed by C<undef> subsequently.
1851 Ordinarily you must assign the returned value to a variable, but
1852 there is one situation where an automatic assignment happens. If
1853 and only if the input symbol is the only thing inside the conditional
1854 of a C<while> statement (even if disguised as a C<for(;;)> loop),
1855 the value is automatically assigned to the global variable $_,
1856 destroying whatever was there previously. (This may seem like an
1857 odd thing to you, but you'll use the construct in almost every Perl
1858 script you write.) The $_ variable is not implicitly localized.
1859 You'll have to put a C<local $_;> before the loop if you want that
1862 The following lines are equivalent:
1864 while (defined($_ = <STDIN>)) { print; }
1865 while ($_ = <STDIN>) { print; }
1866 while (<STDIN>) { print; }
1867 for (;<STDIN>;) { print; }
1868 print while defined($_ = <STDIN>);
1869 print while ($_ = <STDIN>);
1870 print while <STDIN>;
1872 This also behaves similarly, but avoids $_ :
1874 while (my $line = <STDIN>) { print $line }
1876 In these loop constructs, the assigned value (whether assignment
1877 is automatic or explicit) is then tested to see whether it is
1878 defined. The defined test avoids problems where line has a string
1879 value that would be treated as false by Perl, for example a "" or
1880 a "0" with no trailing newline. If you really mean for such values
1881 to terminate the loop, they should be tested for explicitly:
1883 while (($_ = <STDIN>) ne '0') { ... }
1884 while (<STDIN>) { last unless $_; ... }
1886 In other boolean contexts, C<< <I<filehandle>> >> without an
1887 explicit C<defined> test or comparison elicit a warning if the
1888 C<use warnings> pragma or the B<-w>
1889 command-line switch (the C<$^W> variable) is in effect.
1891 The filehandles STDIN, STDOUT, and STDERR are predefined. (The
1892 filehandles C<stdin>, C<stdout>, and C<stderr> will also work except
1893 in packages, where they would be interpreted as local identifiers
1894 rather than global.) Additional filehandles may be created with
1895 the open() function, amongst others. See L<perlopentut> and
1896 L<perlfunc/open> for details on this.
1898 If a <FILEHANDLE> is used in a context that is looking for
1899 a list, a list comprising all input lines is returned, one line per
1900 list element. It's easy to grow to a rather large data space this
1901 way, so use with care.
1903 <FILEHANDLE> may also be spelled C<readline(*FILEHANDLE)>.
1904 See L<perlfunc/readline>.
1906 The null filehandle <> is special: it can be used to emulate the
1907 behavior of B<sed> and B<awk>. Input from <> comes either from
1908 standard input, or from each file listed on the command line. Here's
1909 how it works: the first time <> is evaluated, the @ARGV array is
1910 checked, and if it is empty, C<$ARGV[0]> is set to "-", which when opened
1911 gives you standard input. The @ARGV array is then processed as a list
1912 of filenames. The loop
1915 ... # code for each line
1918 is equivalent to the following Perl-like pseudo code:
1920 unshift(@ARGV, '-') unless @ARGV;
1921 while ($ARGV = shift) {
1924 ... # code for each line
1928 except that it isn't so cumbersome to say, and will actually work.
1929 It really does shift the @ARGV array and put the current filename
1930 into the $ARGV variable. It also uses filehandle I<ARGV>
1931 internally--<> is just a synonym for <ARGV>, which
1932 is magical. (The pseudo code above doesn't work because it treats
1933 <ARGV> as non-magical.)
1935 You can modify @ARGV before the first <> as long as the array ends up
1936 containing the list of filenames you really want. Line numbers (C<$.>)
1937 continue as though the input were one big happy file. See the example
1938 in L<perlfunc/eof> for how to reset line numbers on each file.
1940 If you want to set @ARGV to your own list of files, go right ahead.
1941 This sets @ARGV to all plain text files if no @ARGV was given:
1943 @ARGV = grep { -f && -T } glob('*') unless @ARGV;
1945 You can even set them to pipe commands. For example, this automatically
1946 filters compressed arguments through B<gzip>:
1948 @ARGV = map { /\.(gz|Z)$/ ? "gzip -dc < $_ |" : $_ } @ARGV;
1950 If you want to pass switches into your script, you can use one of the
1951 Getopts modules or put a loop on the front like this:
1953 while ($_ = $ARGV[0], /^-/) {
1956 if (/^-D(.*)/) { $debug = $1 }
1957 if (/^-v/) { $verbose++ }
1958 # ... # other switches
1962 # ... # code for each line
1965 The <> symbol will return C<undef> for end-of-file only once.
1966 If you call it again after this, it will assume you are processing another
1967 @ARGV list, and if you haven't set @ARGV, will read input from STDIN.
1969 If what the angle brackets contain is a simple scalar variable (e.g.,
1970 <$foo>), then that variable contains the name of the
1971 filehandle to input from, or its typeglob, or a reference to the
1977 If what's within the angle brackets is neither a filehandle nor a simple
1978 scalar variable containing a filehandle name, typeglob, or typeglob
1979 reference, it is interpreted as a filename pattern to be globbed, and
1980 either a list of filenames or the next filename in the list is returned,
1981 depending on context. This distinction is determined on syntactic
1982 grounds alone. That means C<< <$x> >> is always a readline() from
1983 an indirect handle, but C<< <$hash{key}> >> is always a glob().
1984 That's because $x is a simple scalar variable, but C<$hash{key}> is
1985 not--it's a hash element.
1987 One level of double-quote interpretation is done first, but you can't
1988 say C<< <$foo> >> because that's an indirect filehandle as explained
1989 in the previous paragraph. (In older versions of Perl, programmers
1990 would insert curly brackets to force interpretation as a filename glob:
1991 C<< <${foo}> >>. These days, it's considered cleaner to call the
1992 internal function directly as C<glob($foo)>, which is probably the right
1993 way to have done it in the first place.) For example:
1999 is roughly equivalent to:
2001 open(FOO, "echo *.c | tr -s ' \t\r\f' '\\012\\012\\012\\012'|");
2007 except that the globbing is actually done internally using the standard
2008 C<File::Glob> extension. Of course, the shortest way to do the above is:
2012 A (file)glob evaluates its (embedded) argument only when it is
2013 starting a new list. All values must be read before it will start
2014 over. In list context, this isn't important because you automatically
2015 get them all anyway. However, in scalar context the operator returns
2016 the next value each time it's called, or C<undef> when the list has
2017 run out. As with filehandle reads, an automatic C<defined> is
2018 generated when the glob occurs in the test part of a C<while>,
2019 because legal glob returns (e.g. a file called F<0>) would otherwise
2020 terminate the loop. Again, C<undef> is returned only once. So if
2021 you're expecting a single value from a glob, it is much better to
2024 ($file) = <blurch*>;
2030 because the latter will alternate between returning a filename and
2033 If you're trying to do variable interpolation, it's definitely better
2034 to use the glob() function, because the older notation can cause people
2035 to become confused with the indirect filehandle notation.
2037 @files = glob("$dir/*.[ch]");
2038 @files = glob($files[$i]);
2040 =head2 Constant Folding
2042 Like C, Perl does a certain amount of expression evaluation at
2043 compile time whenever it determines that all arguments to an
2044 operator are static and have no side effects. In particular, string
2045 concatenation happens at compile time between literals that don't do
2046 variable substitution. Backslash interpolation also happens at
2047 compile time. You can say
2049 'Now is the time for all' . "\n" .
2050 'good men to come to.'
2052 and this all reduces to one string internally. Likewise, if
2055 foreach $file (@filenames) {
2056 if (-s $file > 5 + 100 * 2**16) { }
2059 the compiler will precompute the number which that expression
2060 represents so that the interpreter won't have to.
2064 Perl doesn't officially have a no-op operator, but the bare constants
2065 C<0> and C<1> are special-cased to not produce a warning in a void
2066 context, so you can for example safely do
2070 =head2 Bitwise String Operators
2072 Bitstrings of any size may be manipulated by the bitwise operators
2075 If the operands to a binary bitwise op are strings of different
2076 sizes, B<|> and B<^> ops act as though the shorter operand had
2077 additional zero bits on the right, while the B<&> op acts as though
2078 the longer operand were truncated to the length of the shorter.
2079 The granularity for such extension or truncation is one or more
2082 # ASCII-based examples
2083 print "j p \n" ^ " a h"; # prints "JAPH\n"
2084 print "JA" | " ph\n"; # prints "japh\n"
2085 print "japh\nJunk" & '_____'; # prints "JAPH\n";
2086 print 'p N$' ^ " E<H\n"; # prints "Perl\n";
2088 If you are intending to manipulate bitstrings, be certain that
2089 you're supplying bitstrings: If an operand is a number, that will imply
2090 a B<numeric> bitwise operation. You may explicitly show which type of
2091 operation you intend by using C<""> or C<0+>, as in the examples below.
2093 $foo = 150 | 105 ; # yields 255 (0x96 | 0x69 is 0xFF)
2094 $foo = '150' | 105 ; # yields 255
2095 $foo = 150 | '105'; # yields 255
2096 $foo = '150' | '105'; # yields string '155' (under ASCII)
2098 $baz = 0+$foo & 0+$bar; # both ops explicitly numeric
2099 $biz = "$foo" ^ "$bar"; # both ops explicitly stringy
2101 See L<perlfunc/vec> for information on how to manipulate individual bits
2104 =head2 Integer Arithmetic
2106 By default, Perl assumes that it must do most of its arithmetic in
2107 floating point. But by saying
2111 you may tell the compiler that it's okay to use integer operations
2112 (if it feels like it) from here to the end of the enclosing BLOCK.
2113 An inner BLOCK may countermand this by saying
2117 which lasts until the end of that BLOCK. Note that this doesn't
2118 mean everything is only an integer, merely that Perl may use integer
2119 operations if it is so inclined. For example, even under C<use
2120 integer>, if you take the C<sqrt(2)>, you'll still get C<1.4142135623731>
2123 Used on numbers, the bitwise operators ("&", "|", "^", "~", "<<",
2124 and ">>") always produce integral results. (But see also
2125 L<Bitwise String Operators>.) However, C<use integer> still has meaning for
2126 them. By default, their results are interpreted as unsigned integers, but
2127 if C<use integer> is in effect, their results are interpreted
2128 as signed integers. For example, C<~0> usually evaluates to a large
2129 integral value. However, C<use integer; ~0> is C<-1> on twos-complement
2132 =head2 Floating-point Arithmetic
2134 While C<use integer> provides integer-only arithmetic, there is no
2135 analogous mechanism to provide automatic rounding or truncation to a
2136 certain number of decimal places. For rounding to a certain number
2137 of digits, sprintf() or printf() is usually the easiest route.
2140 Floating-point numbers are only approximations to what a mathematician
2141 would call real numbers. There are infinitely more reals than floats,
2142 so some corners must be cut. For example:
2144 printf "%.20g\n", 123456789123456789;
2145 # produces 123456789123456784
2147 Testing for exact equality of floating-point equality or inequality is
2148 not a good idea. Here's a (relatively expensive) work-around to compare
2149 whether two floating-point numbers are equal to a particular number of
2150 decimal places. See Knuth, volume II, for a more robust treatment of
2154 my ($X, $Y, $POINTS) = @_;
2156 $tX = sprintf("%.${POINTS}g", $X);
2157 $tY = sprintf("%.${POINTS}g", $Y);
2161 The POSIX module (part of the standard perl distribution) implements
2162 ceil(), floor(), and other mathematical and trigonometric functions.
2163 The Math::Complex module (part of the standard perl distribution)
2164 defines mathematical functions that work on both the reals and the
2165 imaginary numbers. Math::Complex not as efficient as POSIX, but
2166 POSIX can't work with complex numbers.
2168 Rounding in financial applications can have serious implications, and
2169 the rounding method used should be specified precisely. In these
2170 cases, it probably pays not to trust whichever system rounding is
2171 being used by Perl, but to instead implement the rounding function you
2174 =head2 Bigger Numbers
2176 The standard Math::BigInt and Math::BigFloat modules provide
2177 variable-precision arithmetic and overloaded operators, although
2178 they're currently pretty slow. At the cost of some space and
2179 considerable speed, they avoid the normal pitfalls associated with
2180 limited-precision representations.
2183 $x = Math::BigInt->new('123456789123456789');
2186 # prints +15241578780673678515622620750190521
2188 There are several modules that let you calculate with (bound only by
2189 memory and cpu-time) unlimited or fixed precision. There are also
2190 some non-standard modules that provide faster implementations via
2191 external C libraries.
2193 Here is a short, but incomplete summary:
2195 Math::Fraction big, unlimited fractions like 9973 / 12967
2196 Math::String treat string sequences like numbers
2197 Math::FixedPrecision calculate with a fixed precision
2198 Math::Currency for currency calculations
2199 Bit::Vector manipulate bit vectors fast (uses C)
2200 Math::BigIntFast Bit::Vector wrapper for big numbers
2201 Math::Pari provides access to the Pari C library
2202 Math::BigInteger uses an external C library
2203 Math::Cephes uses external Cephes C library (no big numbers)
2204 Math::Cephes::Fraction fractions via the Cephes library
2205 Math::GMP another one using an external C library