Commit | Line | Data |
a0d0e21e |
1 | =head1 NAME |
2 | |
3 | perlop - Perl operators and precedence |
4 | |
d042e63d |
5 | =head1 DESCRIPTION |
6 | |
7 | =head2 Operator Precedence and Associativity |
8 | |
9 | Operator precedence and associativity work in Perl more or less like |
10 | they do in mathematics. |
11 | |
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>. |
16 | |
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>. |
a0d0e21e |
23 | |
24 | Perl operators have the following associativity and precedence, |
19799a22 |
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. |
a0d0e21e |
30 | |
31 | left terms and list operators (leftward) |
32 | left -> |
33 | nonassoc ++ -- |
34 | right ** |
35 | right ! ~ \ and unary + and - |
54310121 |
36 | left =~ !~ |
a0d0e21e |
37 | left * / % x |
38 | left + - . |
39 | left << >> |
40 | nonassoc named unary operators |
41 | nonassoc < > <= >= lt gt le ge |
42 | nonassoc == != <=> eq ne cmp |
43 | left & |
44 | left | ^ |
45 | left && |
c963b151 |
46 | left || // |
137443ea |
47 | nonassoc .. ... |
a0d0e21e |
48 | right ?: |
49 | right = += -= *= etc. |
50 | left , => |
51 | nonassoc list operators (rightward) |
a5f75d66 |
52 | right not |
a0d0e21e |
53 | left and |
c963b151 |
54 | left or xor err |
a0d0e21e |
55 | |
56 | In the following sections, these operators are covered in precedence order. |
57 | |
5a964f20 |
58 | Many operators can be overloaded for objects. See L<overload>. |
59 | |
a0d0e21e |
60 | =head2 Terms and List Operators (Leftward) |
61 | |
62c18ce2 |
62 | A TERM has the highest precedence in Perl. They include variables, |
5f05dabc |
63 | quote and quote-like operators, any expression in parentheses, |
a0d0e21e |
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>. |
68 | |
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. |
73 | |
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 |
54310121 |
76 | whether you are looking at the left side or the right side of the operator. |
a0d0e21e |
77 | For example, in |
78 | |
79 | @ary = (1, 3, sort 4, 2); |
80 | print @ary; # prints 1324 |
81 | |
19799a22 |
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 |
a0d0e21e |
85 | then act like a simple TERM with regard to the preceding expression. |
19799a22 |
86 | Be careful with parentheses: |
a0d0e21e |
87 | |
88 | # These evaluate exit before doing the print: |
89 | print($foo, exit); # Obviously not what you want. |
90 | print $foo, exit; # Nor is this. |
91 | |
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 ($foo), exit; # Or even this. |
96 | |
97 | Also note that |
98 | |
99 | print ($foo & 255) + 1, "\n"; |
100 | |
d042e63d |
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<$foo & 255>). Then one is added to the return value |
104 | of C<print> (usually 1). The result is something like this: |
105 | |
106 | 1 + 1, "\n"; # Obviously not what you meant. |
107 | |
108 | To do what you meant properly, you must write: |
109 | |
110 | print(($foo & 255) + 1, "\n"); |
111 | |
112 | See L<Named Unary Operators> for more discussion of this. |
a0d0e21e |
113 | |
114 | Also parsed as terms are the C<do {}> and C<eval {}> constructs, as |
54310121 |
115 | well as subroutine and method calls, and the anonymous |
a0d0e21e |
116 | constructors C<[]> and C<{}>. |
117 | |
2ae324a7 |
118 | See also L<Quote and Quote-like Operators> toward the end of this section, |
c07a80fd |
119 | as well as L<"I/O Operators">. |
a0d0e21e |
120 | |
121 | =head2 The Arrow Operator |
122 | |
35f2feb0 |
123 | "C<< -> >>" is an infix dereference operator, just as it is in C |
19799a22 |
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>. |
a0d0e21e |
130 | |
19799a22 |
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>. |
a0d0e21e |
135 | |
5f05dabc |
136 | =head2 Auto-increment and Auto-decrement |
a0d0e21e |
137 | |
d042e63d |
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 |
141 | value. |
142 | |
143 | $i = 0; $j = 0; |
144 | print $i++; # prints 0 |
145 | print ++$j; # prints 1 |
a0d0e21e |
146 | |
54310121 |
147 | The auto-increment operator has a little extra builtin magic to it. If |
a0d0e21e |
148 | you increment a variable that is numeric, or that has ever been used in |
149 | a numeric context, you get a normal increment. If, however, the |
5f05dabc |
150 | variable has been used in only string contexts since it was set, and |
5a964f20 |
151 | has a value that is not the empty string and matches the pattern |
9c0670e1 |
152 | C</^[a-zA-Z]*[0-9]*\z/>, the increment is done as a string, preserving each |
a0d0e21e |
153 | character within its range, with carry: |
154 | |
155 | print ++($foo = '99'); # prints '100' |
156 | print ++($foo = 'a0'); # prints 'a1' |
157 | print ++($foo = 'Az'); # prints 'Ba' |
158 | print ++($foo = 'zz'); # prints 'aaa' |
159 | |
6a61d433 |
160 | C<undef> is always treated as numeric, and in particular is changed |
161 | to C<0> before incrementing (so that a post-increment of an undef value |
162 | will return C<0> rather than C<undef>). |
163 | |
5f05dabc |
164 | The auto-decrement operator is not magical. |
a0d0e21e |
165 | |
166 | =head2 Exponentiation |
167 | |
19799a22 |
168 | Binary "**" is the exponentiation operator. It binds even more |
cb1a09d0 |
169 | tightly than unary minus, so -2**4 is -(2**4), not (-2)**4. (This is |
170 | implemented using C's pow(3) function, which actually works on doubles |
171 | internally.) |
a0d0e21e |
172 | |
173 | =head2 Symbolic Unary Operators |
174 | |
5f05dabc |
175 | Unary "!" performs logical negation, i.e., "not". See also C<not> for a lower |
a0d0e21e |
176 | precedence version of this. |
177 | |
178 | Unary "-" performs arithmetic negation if the operand is numeric. If |
179 | the operand is an identifier, a string consisting of a minus sign |
180 | concatenated with the identifier is returned. Otherwise, if the string |
181 | starts with a plus or minus, a string starting with the opposite sign |
182 | is returned. One effect of these rules is that C<-bareword> is equivalent |
183 | to C<"-bareword">. |
184 | |
972b05a9 |
185 | Unary "~" performs bitwise negation, i.e., 1's complement. For |
186 | example, C<0666 & ~027> is 0640. (See also L<Integer Arithmetic> and |
187 | L<Bitwise String Operators>.) Note that the width of the result is |
188 | platform-dependent: ~0 is 32 bits wide on a 32-bit platform, but 64 |
189 | bits wide on a 64-bit platform, so if you are expecting a certain bit |
d042e63d |
190 | width, remember to use the & operator to mask off the excess bits. |
a0d0e21e |
191 | |
192 | Unary "+" has no effect whatsoever, even on strings. It is useful |
193 | syntactically for separating a function name from a parenthesized expression |
194 | that would otherwise be interpreted as the complete list of function |
5ba421f6 |
195 | arguments. (See examples above under L<Terms and List Operators (Leftward)>.) |
a0d0e21e |
196 | |
19799a22 |
197 | Unary "\" creates a reference to whatever follows it. See L<perlreftut> |
198 | and L<perlref>. Do not confuse this behavior with the behavior of |
199 | backslash within a string, although both forms do convey the notion |
200 | of protecting the next thing from interpolation. |
a0d0e21e |
201 | |
202 | =head2 Binding Operators |
203 | |
c07a80fd |
204 | Binary "=~" binds a scalar expression to a pattern match. Certain operations |
cb1a09d0 |
205 | search or modify the string $_ by default. This operator makes that kind |
206 | of operation work on some other string. The right argument is a search |
2c268ad5 |
207 | pattern, substitution, or transliteration. The left argument is what is |
208 | supposed to be searched, substituted, or transliterated instead of the default |
f8bab1e9 |
209 | $_. When used in scalar context, the return value generally indicates the |
210 | success of the operation. Behavior in list context depends on the particular |
211 | operator. See L</"Regexp Quote-Like Operators"> for details. |
212 | |
213 | If the right argument is an expression rather than a search pattern, |
2c268ad5 |
214 | substitution, or transliteration, it is interpreted as a search pattern at run |
573e01ca |
215 | time. |
a0d0e21e |
216 | |
217 | Binary "!~" is just like "=~" except the return value is negated in |
218 | the logical sense. |
219 | |
220 | =head2 Multiplicative Operators |
221 | |
222 | Binary "*" multiplies two numbers. |
223 | |
224 | Binary "/" divides two numbers. |
225 | |
54310121 |
226 | Binary "%" computes the modulus of two numbers. Given integer |
227 | operands C<$a> and C<$b>: If C<$b> is positive, then C<$a % $b> is |
228 | C<$a> minus the largest multiple of C<$b> that is not greater than |
229 | C<$a>. If C<$b> is negative, then C<$a % $b> is C<$a> minus the |
230 | smallest multiple of C<$b> that is not less than C<$a> (i.e. the |
6bb4e6d4 |
231 | result will be less than or equal to zero). |
0412d526 |
232 | Note that when C<use integer> is in scope, "%" gives you direct access |
55d729e4 |
233 | to the modulus operator as implemented by your C compiler. This |
234 | operator is not as well defined for negative operands, but it will |
235 | execute faster. |
236 | |
62d10b70 |
237 | Binary "x" is the repetition operator. In scalar context or if the left |
238 | operand is not enclosed in parentheses, it returns a string consisting |
239 | of the left operand repeated the number of times specified by the right |
240 | operand. In list context, if the left operand is enclosed in |
58a9d1fc |
241 | parentheses, it repeats the list. If the right operand is zero or |
242 | negative, it returns an empty string or an empty list, depending on the |
243 | context. |
a0d0e21e |
244 | |
245 | print '-' x 80; # print row of dashes |
246 | |
247 | print "\t" x ($tab/8), ' ' x ($tab%8); # tab over |
248 | |
249 | @ones = (1) x 80; # a list of 80 1's |
250 | @ones = (5) x @ones; # set all elements to 5 |
251 | |
252 | |
253 | =head2 Additive Operators |
254 | |
255 | Binary "+" returns the sum of two numbers. |
256 | |
257 | Binary "-" returns the difference of two numbers. |
258 | |
259 | Binary "." concatenates two strings. |
260 | |
261 | =head2 Shift Operators |
262 | |
55497cff |
263 | Binary "<<" returns the value of its left argument shifted left by the |
264 | number of bits specified by the right argument. Arguments should be |
982ce180 |
265 | integers. (See also L<Integer Arithmetic>.) |
a0d0e21e |
266 | |
55497cff |
267 | Binary ">>" returns the value of its left argument shifted right by |
268 | the number of bits specified by the right argument. Arguments should |
982ce180 |
269 | be integers. (See also L<Integer Arithmetic>.) |
a0d0e21e |
270 | |
b16cf6df |
271 | Note that both "<<" and ">>" in Perl are implemented directly using |
272 | "<<" and ">>" in C. If C<use integer> (see L<Integer Arithmetic>) is |
273 | in force then signed C integers are used, else unsigned C integers are |
274 | used. Either way, the implementation isn't going to generate results |
275 | larger than the size of the integer type Perl was built with (32 bits |
276 | or 64 bits). |
277 | |
278 | The result of overflowing the range of the integers is undefined |
279 | because it is undefined also in C. In other words, using 32-bit |
280 | integers, C<< 1 << 32 >> is undefined. Shifting by a negative number |
281 | of bits is also undefined. |
282 | |
a0d0e21e |
283 | =head2 Named Unary Operators |
284 | |
285 | The various named unary operators are treated as functions with one |
568e6d8b |
286 | argument, with optional parentheses. |
a0d0e21e |
287 | |
288 | If any list operator (print(), etc.) or any unary operator (chdir(), etc.) |
289 | is followed by a left parenthesis as the next token, the operator and |
290 | arguments within parentheses are taken to be of highest precedence, |
3981b0eb |
291 | just like a normal function call. For example, |
292 | because named unary operators are higher precedence than ||: |
a0d0e21e |
293 | |
294 | chdir $foo || die; # (chdir $foo) || die |
295 | chdir($foo) || die; # (chdir $foo) || die |
296 | chdir ($foo) || die; # (chdir $foo) || die |
297 | chdir +($foo) || die; # (chdir $foo) || die |
298 | |
3981b0eb |
299 | but, because * is higher precedence than named operators: |
a0d0e21e |
300 | |
301 | chdir $foo * 20; # chdir ($foo * 20) |
302 | chdir($foo) * 20; # (chdir $foo) * 20 |
303 | chdir ($foo) * 20; # (chdir $foo) * 20 |
304 | chdir +($foo) * 20; # chdir ($foo * 20) |
305 | |
306 | rand 10 * 20; # rand (10 * 20) |
307 | rand(10) * 20; # (rand 10) * 20 |
308 | rand (10) * 20; # (rand 10) * 20 |
309 | rand +(10) * 20; # rand (10 * 20) |
310 | |
568e6d8b |
311 | Regarding precedence, the filetest operators, like C<-f>, C<-M>, etc. are |
312 | treated like named unary operators, but they don't follow this functional |
313 | parenthesis rule. That means, for example, that C<-f($file).".bak"> is |
314 | equivalent to C<-f "$file.bak">. |
315 | |
5ba421f6 |
316 | See also L<"Terms and List Operators (Leftward)">. |
a0d0e21e |
317 | |
318 | =head2 Relational Operators |
319 | |
35f2feb0 |
320 | Binary "<" returns true if the left argument is numerically less than |
a0d0e21e |
321 | the right argument. |
322 | |
35f2feb0 |
323 | Binary ">" returns true if the left argument is numerically greater |
a0d0e21e |
324 | than the right argument. |
325 | |
35f2feb0 |
326 | Binary "<=" returns true if the left argument is numerically less than |
a0d0e21e |
327 | or equal to the right argument. |
328 | |
35f2feb0 |
329 | Binary ">=" returns true if the left argument is numerically greater |
a0d0e21e |
330 | than or equal to the right argument. |
331 | |
332 | Binary "lt" returns true if the left argument is stringwise less than |
333 | the right argument. |
334 | |
335 | Binary "gt" returns true if the left argument is stringwise greater |
336 | than the right argument. |
337 | |
338 | Binary "le" returns true if the left argument is stringwise less than |
339 | or equal to the right argument. |
340 | |
341 | Binary "ge" returns true if the left argument is stringwise greater |
342 | than or equal to the right argument. |
343 | |
344 | =head2 Equality Operators |
345 | |
346 | Binary "==" returns true if the left argument is numerically equal to |
347 | the right argument. |
348 | |
349 | Binary "!=" returns true if the left argument is numerically not equal |
350 | to the right argument. |
351 | |
35f2feb0 |
352 | Binary "<=>" returns -1, 0, or 1 depending on whether the left |
6ee5d4e7 |
353 | argument is numerically less than, equal to, or greater than the right |
d4ad863d |
354 | argument. If your platform supports NaNs (not-a-numbers) as numeric |
7d3a9d88 |
355 | values, using them with "<=>" returns undef. NaN is not "<", "==", ">", |
356 | "<=" or ">=" anything (even NaN), so those 5 return false. NaN != NaN |
357 | returns true, as does NaN != anything else. If your platform doesn't |
358 | support NaNs then NaN is just a string with numeric value 0. |
359 | |
360 | perl -le '$a = NaN; print "No NaN support here" if $a == $a' |
361 | perl -le '$a = NaN; print "NaN support here" if $a != $a' |
a0d0e21e |
362 | |
363 | Binary "eq" returns true if the left argument is stringwise equal to |
364 | the right argument. |
365 | |
366 | Binary "ne" returns true if the left argument is stringwise not equal |
367 | to the right argument. |
368 | |
d4ad863d |
369 | Binary "cmp" returns -1, 0, or 1 depending on whether the left |
370 | argument is stringwise less than, equal to, or greater than the right |
371 | argument. |
a0d0e21e |
372 | |
a034a98d |
373 | "lt", "le", "ge", "gt" and "cmp" use the collation (sort) order specified |
374 | by the current locale if C<use locale> is in effect. See L<perllocale>. |
375 | |
a0d0e21e |
376 | =head2 Bitwise And |
377 | |
2cdc098b |
378 | Binary "&" returns its operands ANDed together bit by bit. |
2c268ad5 |
379 | (See also L<Integer Arithmetic> and L<Bitwise String Operators>.) |
a0d0e21e |
380 | |
2cdc098b |
381 | Note that "&" has lower priority than relational operators, so for example |
382 | the brackets are essential in a test like |
383 | |
384 | print "Even\n" if ($x & 1) == 0; |
385 | |
a0d0e21e |
386 | =head2 Bitwise Or and Exclusive Or |
387 | |
2cdc098b |
388 | Binary "|" returns its operands ORed together bit by bit. |
2c268ad5 |
389 | (See also L<Integer Arithmetic> and L<Bitwise String Operators>.) |
a0d0e21e |
390 | |
2cdc098b |
391 | Binary "^" returns its operands XORed together bit by bit. |
2c268ad5 |
392 | (See also L<Integer Arithmetic> and L<Bitwise String Operators>.) |
a0d0e21e |
393 | |
2cdc098b |
394 | Note that "|" and "^" have lower priority than relational operators, so |
395 | for example the brackets are essential in a test like |
396 | |
397 | print "false\n" if (8 | 2) != 10; |
398 | |
a0d0e21e |
399 | =head2 C-style Logical And |
400 | |
401 | Binary "&&" performs a short-circuit logical AND operation. That is, |
402 | if the left operand is false, the right operand is not even evaluated. |
403 | Scalar or list context propagates down to the right operand if it |
404 | is evaluated. |
405 | |
406 | =head2 C-style Logical Or |
407 | |
408 | Binary "||" performs a short-circuit logical OR operation. That is, |
409 | if the left operand is true, the right operand is not even evaluated. |
410 | Scalar or list context propagates down to the right operand if it |
411 | is evaluated. |
412 | |
c963b151 |
413 | =head2 C-style Logical Defined-Or |
414 | |
415 | Although it has no direct equivalent in C, Perl's C<//> operator is related |
416 | to its C-style or. In fact, it's exactly the same as C<||>, except that it |
417 | tests the left hand side's definedness instead of its truth. Thus, C<$a // $b> |
418 | is similar to C<defined($a) || $b> (except that it returns the value of C<$a> |
419 | rather than the value of C<defined($a)>) and is exactly equivalent to |
420 | C<defined($a) ? $a : $b>. This is very useful for providing default values |
d042e63d |
421 | for variables. If you actually want to test if at least one of C<$a> and |
422 | C<$b> is defined, use C<defined($a // $b)>. |
c963b151 |
423 | |
d042e63d |
424 | The C<||>, C<//> and C<&&> operators return the last value evaluated |
425 | (unlike C's C<||> and C<&&>, which return 0 or 1). Thus, a reasonably |
426 | portable way to find out the home directory might be: |
a0d0e21e |
427 | |
c963b151 |
428 | $home = $ENV{'HOME'} // $ENV{'LOGDIR'} // |
429 | (getpwuid($<))[7] // die "You're homeless!\n"; |
a0d0e21e |
430 | |
5a964f20 |
431 | In particular, this means that you shouldn't use this |
432 | for selecting between two aggregates for assignment: |
433 | |
434 | @a = @b || @c; # this is wrong |
435 | @a = scalar(@b) || @c; # really meant this |
436 | @a = @b ? @b : @c; # this works fine, though |
437 | |
c963b151 |
438 | As more readable alternatives to C<&&>, C<//> and C<||> when used for |
439 | control flow, Perl provides C<and>, C<err> and C<or> operators (see below). |
440 | The short-circuit behavior is identical. The precedence of "and", "err" |
441 | and "or" is much lower, however, so that you can safely use them after a |
5a964f20 |
442 | list operator without the need for parentheses: |
a0d0e21e |
443 | |
444 | unlink "alpha", "beta", "gamma" |
445 | or gripe(), next LINE; |
446 | |
447 | With the C-style operators that would have been written like this: |
448 | |
449 | unlink("alpha", "beta", "gamma") |
450 | || (gripe(), next LINE); |
451 | |
eeb6a2c9 |
452 | Using "or" for assignment is unlikely to do what you want; see below. |
5a964f20 |
453 | |
454 | =head2 Range Operators |
a0d0e21e |
455 | |
456 | Binary ".." is the range operator, which is really two different |
fb53bbb2 |
457 | operators depending on the context. In list context, it returns a |
54ae734e |
458 | list of values counting (up by ones) from the left value to the right |
2cdbc966 |
459 | value. If the left value is greater than the right value then it |
fb53bbb2 |
460 | returns the empty list. The range operator is useful for writing |
54ae734e |
461 | C<foreach (1..10)> loops and for doing slice operations on arrays. In |
2cdbc966 |
462 | the current implementation, no temporary array is created when the |
463 | range operator is used as the expression in C<foreach> loops, but older |
464 | versions of Perl might burn a lot of memory when you write something |
465 | like this: |
a0d0e21e |
466 | |
467 | for (1 .. 1_000_000) { |
468 | # code |
54310121 |
469 | } |
a0d0e21e |
470 | |
54ae734e |
471 | The range operator also works on strings, using the magical auto-increment, |
472 | see below. |
473 | |
5a964f20 |
474 | In scalar context, ".." returns a boolean value. The operator is |
a0d0e21e |
475 | bistable, like a flip-flop, and emulates the line-range (comma) operator |
476 | of B<sed>, B<awk>, and various editors. Each ".." operator maintains its |
477 | own boolean state. It is false as long as its left operand is false. |
478 | Once the left operand is true, the range operator stays true until the |
479 | right operand is true, I<AFTER> which the range operator becomes false |
19799a22 |
480 | again. It doesn't become false till the next time the range operator is |
a0d0e21e |
481 | evaluated. It can test the right operand and become false on the same |
482 | evaluation it became true (as in B<awk>), but it still returns true once. |
19799a22 |
483 | If you don't want it to test the right operand till the next |
484 | evaluation, as in B<sed>, just use three dots ("...") instead of |
485 | two. In all other regards, "..." behaves just like ".." does. |
486 | |
487 | The right operand is not evaluated while the operator is in the |
488 | "false" state, and the left operand is not evaluated while the |
489 | operator is in the "true" state. The precedence is a little lower |
490 | than || and &&. The value returned is either the empty string for |
491 | false, or a sequence number (beginning with 1) for true. The |
492 | sequence number is reset for each range encountered. The final |
493 | sequence number in a range has the string "E0" appended to it, which |
494 | doesn't affect its numeric value, but gives you something to search |
495 | for if you want to exclude the endpoint. You can exclude the |
496 | beginning point by waiting for the sequence number to be greater |
df5f8116 |
497 | than 1. |
498 | |
499 | If either operand of scalar ".." is a constant expression, |
500 | that operand is considered true if it is equal (C<==>) to the current |
501 | input line number (the C<$.> variable). |
502 | |
503 | To be pedantic, the comparison is actually C<int(EXPR) == int(EXPR)>, |
504 | but that is only an issue if you use a floating point expression; when |
505 | implicitly using C<$.> as described in the previous paragraph, the |
506 | comparison is C<int(EXPR) == int($.)> which is only an issue when C<$.> |
507 | is set to a floating point value and you are not reading from a file. |
508 | Furthermore, C<"span" .. "spat"> or C<2.18 .. 3.14> will not do what |
509 | you want in scalar context because each of the operands are evaluated |
510 | using their integer representation. |
511 | |
512 | Examples: |
a0d0e21e |
513 | |
514 | As a scalar operator: |
515 | |
df5f8116 |
516 | if (101 .. 200) { print; } # print 2nd hundred lines, short for |
517 | # if ($. == 101 .. $. == 200) ... |
518 | next line if (1 .. /^$/); # skip header lines, short for |
519 | # ... if ($. == 1 .. /^$/); |
a0d0e21e |
520 | s/^/> / if (/^$/ .. eof()); # quote body |
521 | |
5a964f20 |
522 | # parse mail messages |
523 | while (<>) { |
524 | $in_header = 1 .. /^$/; |
df5f8116 |
525 | $in_body = /^$/ .. eof; |
526 | if ($in_header) { |
527 | # ... |
528 | } else { # in body |
529 | # ... |
530 | } |
5a964f20 |
531 | } continue { |
df5f8116 |
532 | close ARGV if eof; # reset $. each file |
5a964f20 |
533 | } |
534 | |
a0d0e21e |
535 | As a list operator: |
536 | |
537 | for (101 .. 200) { print; } # print $_ 100 times |
3e3baf6d |
538 | @foo = @foo[0 .. $#foo]; # an expensive no-op |
a0d0e21e |
539 | @foo = @foo[$#foo-4 .. $#foo]; # slice last 5 items |
540 | |
5a964f20 |
541 | The range operator (in list context) makes use of the magical |
5f05dabc |
542 | auto-increment algorithm if the operands are strings. You |
a0d0e21e |
543 | can say |
544 | |
545 | @alphabet = ('A' .. 'Z'); |
546 | |
54ae734e |
547 | to get all normal letters of the English alphabet, or |
a0d0e21e |
548 | |
549 | $hexdigit = (0 .. 9, 'a' .. 'f')[$num & 15]; |
550 | |
551 | to get a hexadecimal digit, or |
552 | |
553 | @z2 = ('01' .. '31'); print $z2[$mday]; |
554 | |
555 | to get dates with leading zeros. If the final value specified is not |
556 | in the sequence that the magical increment would produce, the sequence |
557 | goes until the next value would be longer than the final value |
558 | specified. |
559 | |
df5f8116 |
560 | Because each operand is evaluated in integer form, C<2.18 .. 3.14> will |
561 | return two elements in list context. |
562 | |
563 | @list = (2.18 .. 3.14); # same as @list = (2 .. 3); |
564 | |
a0d0e21e |
565 | =head2 Conditional Operator |
566 | |
567 | Ternary "?:" is the conditional operator, just as in C. It works much |
568 | like an if-then-else. If the argument before the ? is true, the |
569 | argument before the : is returned, otherwise the argument after the : |
cb1a09d0 |
570 | is returned. For example: |
571 | |
54310121 |
572 | printf "I have %d dog%s.\n", $n, |
cb1a09d0 |
573 | ($n == 1) ? '' : "s"; |
574 | |
575 | Scalar or list context propagates downward into the 2nd |
54310121 |
576 | or 3rd argument, whichever is selected. |
cb1a09d0 |
577 | |
578 | $a = $ok ? $b : $c; # get a scalar |
579 | @a = $ok ? @b : @c; # get an array |
580 | $a = $ok ? @b : @c; # oops, that's just a count! |
581 | |
582 | The operator may be assigned to if both the 2nd and 3rd arguments are |
583 | legal lvalues (meaning that you can assign to them): |
a0d0e21e |
584 | |
585 | ($a_or_b ? $a : $b) = $c; |
586 | |
5a964f20 |
587 | Because this operator produces an assignable result, using assignments |
588 | without parentheses will get you in trouble. For example, this: |
589 | |
590 | $a % 2 ? $a += 10 : $a += 2 |
591 | |
592 | Really means this: |
593 | |
594 | (($a % 2) ? ($a += 10) : $a) += 2 |
595 | |
596 | Rather than this: |
597 | |
598 | ($a % 2) ? ($a += 10) : ($a += 2) |
599 | |
19799a22 |
600 | That should probably be written more simply as: |
601 | |
602 | $a += ($a % 2) ? 10 : 2; |
603 | |
4633a7c4 |
604 | =head2 Assignment Operators |
a0d0e21e |
605 | |
606 | "=" is the ordinary assignment operator. |
607 | |
608 | Assignment operators work as in C. That is, |
609 | |
610 | $a += 2; |
611 | |
612 | is equivalent to |
613 | |
614 | $a = $a + 2; |
615 | |
616 | although without duplicating any side effects that dereferencing the lvalue |
54310121 |
617 | might trigger, such as from tie(). Other assignment operators work similarly. |
618 | The following are recognized: |
a0d0e21e |
619 | |
620 | **= += *= &= <<= &&= |
621 | -= /= |= >>= ||= |
622 | .= %= ^= |
623 | x= |
624 | |
19799a22 |
625 | Although these are grouped by family, they all have the precedence |
a0d0e21e |
626 | of assignment. |
627 | |
b350dd2f |
628 | Unlike in C, the scalar assignment operator produces a valid lvalue. |
629 | Modifying an assignment is equivalent to doing the assignment and |
630 | then modifying the variable that was assigned to. This is useful |
631 | for modifying a copy of something, like this: |
a0d0e21e |
632 | |
633 | ($tmp = $global) =~ tr [A-Z] [a-z]; |
634 | |
635 | Likewise, |
636 | |
637 | ($a += 2) *= 3; |
638 | |
639 | is equivalent to |
640 | |
641 | $a += 2; |
642 | $a *= 3; |
643 | |
b350dd2f |
644 | Similarly, a list assignment in list context produces the list of |
645 | lvalues assigned to, and a list assignment in scalar context returns |
646 | the number of elements produced by the expression on the right hand |
647 | side of the assignment. |
648 | |
748a9306 |
649 | =head2 Comma Operator |
a0d0e21e |
650 | |
5a964f20 |
651 | Binary "," is the comma operator. In scalar context it evaluates |
a0d0e21e |
652 | its left argument, throws that value away, then evaluates its right |
653 | argument and returns that value. This is just like C's comma operator. |
654 | |
5a964f20 |
655 | In list context, it's just the list argument separator, and inserts |
a0d0e21e |
656 | both its arguments into the list. |
657 | |
d042e63d |
658 | The C<< => >> operator is a synonym for the comma, but forces any word |
659 | to its left to be interpreted as a string (as of 5.001). It is helpful |
660 | in documenting the correspondence between keys and values in hashes, |
661 | and other paired elements in lists. |
748a9306 |
662 | |
a0d0e21e |
663 | =head2 List Operators (Rightward) |
664 | |
665 | On the right side of a list operator, it has very low precedence, |
666 | such that it controls all comma-separated expressions found there. |
667 | The only operators with lower precedence are the logical operators |
668 | "and", "or", and "not", which may be used to evaluate calls to list |
669 | operators without the need for extra parentheses: |
670 | |
671 | open HANDLE, "filename" |
672 | or die "Can't open: $!\n"; |
673 | |
5ba421f6 |
674 | See also discussion of list operators in L<Terms and List Operators (Leftward)>. |
a0d0e21e |
675 | |
676 | =head2 Logical Not |
677 | |
678 | Unary "not" returns the logical negation of the expression to its right. |
679 | It's the equivalent of "!" except for the very low precedence. |
680 | |
681 | =head2 Logical And |
682 | |
683 | Binary "and" returns the logical conjunction of the two surrounding |
684 | expressions. It's equivalent to && except for the very low |
5f05dabc |
685 | precedence. This means that it short-circuits: i.e., the right |
a0d0e21e |
686 | expression is evaluated only if the left expression is true. |
687 | |
c963b151 |
688 | =head2 Logical or, Defined or, and Exclusive Or |
a0d0e21e |
689 | |
690 | Binary "or" returns the logical disjunction of the two surrounding |
5a964f20 |
691 | expressions. It's equivalent to || except for the very low precedence. |
692 | This makes it useful for control flow |
693 | |
694 | print FH $data or die "Can't write to FH: $!"; |
695 | |
696 | This means that it short-circuits: i.e., the right expression is evaluated |
697 | only if the left expression is false. Due to its precedence, you should |
698 | probably avoid using this for assignment, only for control flow. |
699 | |
700 | $a = $b or $c; # bug: this is wrong |
701 | ($a = $b) or $c; # really means this |
702 | $a = $b || $c; # better written this way |
703 | |
19799a22 |
704 | However, when it's a list-context assignment and you're trying to use |
5a964f20 |
705 | "||" for control flow, you probably need "or" so that the assignment |
706 | takes higher precedence. |
707 | |
708 | @info = stat($file) || die; # oops, scalar sense of stat! |
709 | @info = stat($file) or die; # better, now @info gets its due |
710 | |
c963b151 |
711 | Then again, you could always use parentheses. |
712 | |
713 | Binary "err" is equivalent to C<//>--it's just like binary "or", except it tests |
714 | its left argument's definedness instead of its truth. There are two ways to |
715 | remember "err": either because many functions return C<undef> on an B<err>or, |
716 | or as a sort of correction: C<$a=($b err 'default')> |
a0d0e21e |
717 | |
718 | Binary "xor" returns the exclusive-OR of the two surrounding expressions. |
719 | It cannot short circuit, of course. |
720 | |
721 | =head2 C Operators Missing From Perl |
722 | |
723 | Here is what C has that Perl doesn't: |
724 | |
725 | =over 8 |
726 | |
727 | =item unary & |
728 | |
729 | Address-of operator. (But see the "\" operator for taking a reference.) |
730 | |
731 | =item unary * |
732 | |
54310121 |
733 | Dereference-address operator. (Perl's prefix dereferencing |
a0d0e21e |
734 | operators are typed: $, @, %, and &.) |
735 | |
736 | =item (TYPE) |
737 | |
19799a22 |
738 | Type-casting operator. |
a0d0e21e |
739 | |
740 | =back |
741 | |
5f05dabc |
742 | =head2 Quote and Quote-like Operators |
a0d0e21e |
743 | |
744 | While we usually think of quotes as literal values, in Perl they |
745 | function as operators, providing various kinds of interpolating and |
746 | pattern matching capabilities. Perl provides customary quote characters |
747 | for these behaviors, but also provides a way for you to choose your |
748 | quote character for any of them. In the following table, a C<{}> represents |
87275199 |
749 | any pair of delimiters you choose. |
a0d0e21e |
750 | |
2c268ad5 |
751 | Customary Generic Meaning Interpolates |
752 | '' q{} Literal no |
753 | "" qq{} Literal yes |
af9219ee |
754 | `` qx{} Command yes* |
2c268ad5 |
755 | qw{} Word list no |
af9219ee |
756 | // m{} Pattern match yes* |
757 | qr{} Pattern yes* |
758 | s{}{} Substitution yes* |
2c268ad5 |
759 | tr{}{} Transliteration no (but see below) |
7e3b091d |
760 | <<EOF here-doc yes* |
a0d0e21e |
761 | |
af9219ee |
762 | * unless the delimiter is ''. |
763 | |
87275199 |
764 | Non-bracketing delimiters use the same character fore and aft, but the four |
765 | sorts of brackets (round, angle, square, curly) will all nest, which means |
766 | that |
767 | |
768 | q{foo{bar}baz} |
35f2feb0 |
769 | |
87275199 |
770 | is the same as |
771 | |
772 | 'foo{bar}baz' |
773 | |
774 | Note, however, that this does not always work for quoting Perl code: |
775 | |
776 | $s = q{ if($a eq "}") ... }; # WRONG |
777 | |
83df6a1d |
778 | is a syntax error. The C<Text::Balanced> module (from CPAN, and |
779 | starting from Perl 5.8 part of the standard distribution) is able |
780 | to do this properly. |
87275199 |
781 | |
19799a22 |
782 | There can be whitespace between the operator and the quoting |
fb73857a |
783 | characters, except when C<#> is being used as the quoting character. |
19799a22 |
784 | C<q#foo#> is parsed as the string C<foo>, while C<q #foo#> is the |
785 | operator C<q> followed by a comment. Its argument will be taken |
786 | from the next line. This allows you to write: |
fb73857a |
787 | |
788 | s {foo} # Replace foo |
789 | {bar} # with bar. |
790 | |
904501ec |
791 | The following escape sequences are available in constructs that interpolate |
792 | and in transliterations. |
a0d0e21e |
793 | |
6ee5d4e7 |
794 | \t tab (HT, TAB) |
5a964f20 |
795 | \n newline (NL) |
6ee5d4e7 |
796 | \r return (CR) |
797 | \f form feed (FF) |
798 | \b backspace (BS) |
799 | \a alarm (bell) (BEL) |
800 | \e escape (ESC) |
a0ed51b3 |
801 | \033 octal char (ESC) |
802 | \x1b hex char (ESC) |
803 | \x{263a} wide hex char (SMILEY) |
19799a22 |
804 | \c[ control char (ESC) |
95cc3e0c |
805 | \N{name} named Unicode character |
2c268ad5 |
806 | |
4c77eaa2 |
807 | B<NOTE>: Unlike C and other languages, Perl has no \v escape sequence for |
808 | the vertical tab (VT - ASCII 11). |
809 | |
904501ec |
810 | The following escape sequences are available in constructs that interpolate |
811 | but not in transliterations. |
812 | |
a0d0e21e |
813 | \l lowercase next char |
814 | \u uppercase next char |
815 | \L lowercase till \E |
816 | \U uppercase till \E |
817 | \E end case modification |
1d2dff63 |
818 | \Q quote non-word characters till \E |
a0d0e21e |
819 | |
95cc3e0c |
820 | If C<use locale> is in effect, the case map used by C<\l>, C<\L>, |
821 | C<\u> and C<\U> is taken from the current locale. See L<perllocale>. |
822 | If Unicode (for example, C<\N{}> or wide hex characters of 0x100 or |
823 | beyond) is being used, the case map used by C<\l>, C<\L>, C<\u> and |
824 | C<\U> is as defined by Unicode. For documentation of C<\N{name}>, |
825 | see L<charnames>. |
a034a98d |
826 | |
5a964f20 |
827 | All systems use the virtual C<"\n"> to represent a line terminator, |
828 | called a "newline". There is no such thing as an unvarying, physical |
19799a22 |
829 | newline character. It is only an illusion that the operating system, |
5a964f20 |
830 | device drivers, C libraries, and Perl all conspire to preserve. Not all |
831 | systems read C<"\r"> as ASCII CR and C<"\n"> as ASCII LF. For example, |
832 | on a Mac, these are reversed, and on systems without line terminator, |
833 | printing C<"\n"> may emit no actual data. In general, use C<"\n"> when |
834 | you mean a "newline" for your system, but use the literal ASCII when you |
835 | need an exact character. For example, most networking protocols expect |
2a380090 |
836 | and prefer a CR+LF (C<"\015\012"> or C<"\cM\cJ">) for line terminators, |
5a964f20 |
837 | and although they often accept just C<"\012">, they seldom tolerate just |
838 | C<"\015">. If you get in the habit of using C<"\n"> for networking, |
839 | you may be burned some day. |
840 | |
904501ec |
841 | For constructs that do interpolate, variables beginning with "C<$>" |
842 | or "C<@>" are interpolated. Subscripted variables such as C<$a[3]> or |
ad0f383a |
843 | C<< $href->{key}[0] >> are also interpolated, as are array and hash slices. |
844 | But method calls such as C<< $obj->meth >> are not. |
af9219ee |
845 | |
846 | Interpolating an array or slice interpolates the elements in order, |
847 | separated by the value of C<$">, so is equivalent to interpolating |
904501ec |
848 | C<join $", @array>. "Punctuation" arrays such as C<@+> are only |
849 | interpolated if the name is enclosed in braces C<@{+}>. |
af9219ee |
850 | |
1d2dff63 |
851 | You cannot include a literal C<$> or C<@> within a C<\Q> sequence. |
852 | An unescaped C<$> or C<@> interpolates the corresponding variable, |
853 | while escaping will cause the literal string C<\$> to be inserted. |
854 | You'll need to write something like C<m/\Quser\E\@\Qhost/>. |
855 | |
a0d0e21e |
856 | Patterns are subject to an additional level of interpretation as a |
857 | regular expression. This is done as a second pass, after variables are |
858 | interpolated, so that regular expressions may be incorporated into the |
859 | pattern from the variables. If this is not what you want, use C<\Q> to |
860 | interpolate a variable literally. |
861 | |
19799a22 |
862 | Apart from the behavior described above, Perl does not expand |
863 | multiple levels of interpolation. In particular, contrary to the |
864 | expectations of shell programmers, back-quotes do I<NOT> interpolate |
865 | within double quotes, nor do single quotes impede evaluation of |
866 | variables when used within double quotes. |
a0d0e21e |
867 | |
5f05dabc |
868 | =head2 Regexp Quote-Like Operators |
cb1a09d0 |
869 | |
5f05dabc |
870 | Here are the quote-like operators that apply to pattern |
cb1a09d0 |
871 | matching and related activities. |
872 | |
a0d0e21e |
873 | =over 8 |
874 | |
875 | =item ?PATTERN? |
876 | |
877 | This is just like the C</pattern/> search, except that it matches only |
878 | once between calls to the reset() operator. This is a useful |
5f05dabc |
879 | optimization when you want to see only the first occurrence of |
a0d0e21e |
880 | something in each file of a set of files, for instance. Only C<??> |
881 | patterns local to the current package are reset. |
882 | |
5a964f20 |
883 | while (<>) { |
884 | if (?^$?) { |
885 | # blank line between header and body |
886 | } |
887 | } continue { |
888 | reset if eof; # clear ?? status for next file |
889 | } |
890 | |
483b4840 |
891 | This usage is vaguely deprecated, which means it just might possibly |
19799a22 |
892 | be removed in some distant future version of Perl, perhaps somewhere |
893 | around the year 2168. |
a0d0e21e |
894 | |
fb73857a |
895 | =item m/PATTERN/cgimosx |
a0d0e21e |
896 | |
fb73857a |
897 | =item /PATTERN/cgimosx |
a0d0e21e |
898 | |
5a964f20 |
899 | Searches a string for a pattern match, and in scalar context returns |
19799a22 |
900 | true if it succeeds, false if it fails. If no string is specified |
901 | via the C<=~> or C<!~> operator, the $_ string is searched. (The |
902 | string specified with C<=~> need not be an lvalue--it may be the |
903 | result of an expression evaluation, but remember the C<=~> binds |
904 | rather tightly.) See also L<perlre>. See L<perllocale> for |
905 | discussion of additional considerations that apply when C<use locale> |
906 | is in effect. |
a0d0e21e |
907 | |
908 | Options are: |
909 | |
fb73857a |
910 | c Do not reset search position on a failed match when /g is in effect. |
5f05dabc |
911 | g Match globally, i.e., find all occurrences. |
a0d0e21e |
912 | i Do case-insensitive pattern matching. |
913 | m Treat string as multiple lines. |
5f05dabc |
914 | o Compile pattern only once. |
a0d0e21e |
915 | s Treat string as single line. |
916 | x Use extended regular expressions. |
917 | |
918 | If "/" is the delimiter then the initial C<m> is optional. With the C<m> |
01ae956f |
919 | you can use any pair of non-alphanumeric, non-whitespace characters |
19799a22 |
920 | as delimiters. This is particularly useful for matching path names |
921 | that contain "/", to avoid LTS (leaning toothpick syndrome). If "?" is |
7bac28a0 |
922 | the delimiter, then the match-only-once rule of C<?PATTERN?> applies. |
19799a22 |
923 | If "'" is the delimiter, no interpolation is performed on the PATTERN. |
a0d0e21e |
924 | |
925 | PATTERN may contain variables, which will be interpolated (and the |
f70b4f9c |
926 | pattern recompiled) every time the pattern search is evaluated, except |
1f247705 |
927 | for when the delimiter is a single quote. (Note that C<$(>, C<$)>, and |
928 | C<$|> are not interpolated because they look like end-of-string tests.) |
f70b4f9c |
929 | If you want such a pattern to be compiled only once, add a C</o> after |
930 | the trailing delimiter. This avoids expensive run-time recompilations, |
931 | and is useful when the value you are interpolating won't change over |
932 | the life of the script. However, mentioning C</o> constitutes a promise |
933 | that you won't change the variables in the pattern. If you change them, |
13a2d996 |
934 | Perl won't even notice. See also L<"qr/STRING/imosx">. |
a0d0e21e |
935 | |
5a964f20 |
936 | If the PATTERN evaluates to the empty string, the last |
d65afb4b |
937 | I<successfully> matched regular expression is used instead. In this |
938 | case, only the C<g> and C<c> flags on the empty pattern is honoured - |
939 | the other flags are taken from the original pattern. If no match has |
940 | previously succeeded, this will (silently) act instead as a genuine |
941 | empty pattern (which will always match). |
a0d0e21e |
942 | |
c963b151 |
943 | Note that it's possible to confuse Perl into thinking C<//> (the empty |
944 | regex) is really C<//> (the defined-or operator). Perl is usually pretty |
945 | good about this, but some pathological cases might trigger this, such as |
946 | C<$a///> (is that C<($a) / (//)> or C<$a // />?) and C<print $fh //> |
947 | (C<print $fh(//> or C<print($fh //>?). In all of these examples, Perl |
948 | will assume you meant defined-or. If you meant the empty regex, just |
949 | use parentheses or spaces to disambiguate, or even prefix the empty |
950 | regex with an C<m> (so C<//> becomes C<m//>). |
951 | |
19799a22 |
952 | If the C</g> option is not used, C<m//> in list context returns a |
a0d0e21e |
953 | list consisting of the subexpressions matched by the parentheses in the |
f7e33566 |
954 | pattern, i.e., (C<$1>, C<$2>, C<$3>...). (Note that here C<$1> etc. are |
955 | also set, and that this differs from Perl 4's behavior.) When there are |
956 | no parentheses in the pattern, the return value is the list C<(1)> for |
957 | success. With or without parentheses, an empty list is returned upon |
958 | failure. |
a0d0e21e |
959 | |
960 | Examples: |
961 | |
962 | open(TTY, '/dev/tty'); |
963 | <TTY> =~ /^y/i && foo(); # do foo if desired |
964 | |
965 | if (/Version: *([0-9.]*)/) { $version = $1; } |
966 | |
967 | next if m#^/usr/spool/uucp#; |
968 | |
969 | # poor man's grep |
970 | $arg = shift; |
971 | while (<>) { |
972 | print if /$arg/o; # compile only once |
973 | } |
974 | |
975 | if (($F1, $F2, $Etc) = ($foo =~ /^(\S+)\s+(\S+)\s*(.*)/)) |
976 | |
977 | This last example splits $foo into the first two words and the |
5f05dabc |
978 | remainder of the line, and assigns those three fields to $F1, $F2, and |
979 | $Etc. The conditional is true if any variables were assigned, i.e., if |
a0d0e21e |
980 | the pattern matched. |
981 | |
19799a22 |
982 | The C</g> modifier specifies global pattern matching--that is, |
983 | matching as many times as possible within the string. How it behaves |
984 | depends on the context. In list context, it returns a list of the |
985 | substrings matched by any capturing parentheses in the regular |
986 | expression. If there are no parentheses, it returns a list of all |
987 | the matched strings, as if there were parentheses around the whole |
988 | pattern. |
a0d0e21e |
989 | |
7e86de3e |
990 | In scalar context, each execution of C<m//g> finds the next match, |
19799a22 |
991 | returning true if it matches, and false if there is no further match. |
7e86de3e |
992 | The position after the last match can be read or set using the pos() |
993 | function; see L<perlfunc/pos>. A failed match normally resets the |
994 | search position to the beginning of the string, but you can avoid that |
995 | by adding the C</c> modifier (e.g. C<m//gc>). Modifying the target |
996 | string also resets the search position. |
c90c0ff4 |
997 | |
998 | You can intermix C<m//g> matches with C<m/\G.../g>, where C<\G> is a |
999 | zero-width assertion that matches the exact position where the previous |
5d43e42d |
1000 | C<m//g>, if any, left off. Without the C</g> modifier, the C<\G> assertion |
1001 | still anchors at pos(), but the match is of course only attempted once. |
1002 | Using C<\G> without C</g> on a target string that has not previously had a |
1003 | C</g> match applied to it is the same as using the C<\A> assertion to match |
fe4b3f22 |
1004 | the beginning of the string. Note also that, currently, C<\G> is only |
1005 | properly supported when anchored at the very beginning of the pattern. |
c90c0ff4 |
1006 | |
1007 | Examples: |
a0d0e21e |
1008 | |
1009 | # list context |
1010 | ($one,$five,$fifteen) = (`uptime` =~ /(\d+\.\d+)/g); |
1011 | |
1012 | # scalar context |
5d43e42d |
1013 | $/ = ""; |
19799a22 |
1014 | while (defined($paragraph = <>)) { |
1015 | while ($paragraph =~ /[a-z]['")]*[.!?]+['")]*\s/g) { |
1016 | $sentences++; |
a0d0e21e |
1017 | } |
1018 | } |
1019 | print "$sentences\n"; |
1020 | |
c90c0ff4 |
1021 | # using m//gc with \G |
137443ea |
1022 | $_ = "ppooqppqq"; |
44a8e56a |
1023 | while ($i++ < 2) { |
1024 | print "1: '"; |
c90c0ff4 |
1025 | print $1 while /(o)/gc; print "', pos=", pos, "\n"; |
44a8e56a |
1026 | print "2: '"; |
c90c0ff4 |
1027 | print $1 if /\G(q)/gc; print "', pos=", pos, "\n"; |
44a8e56a |
1028 | print "3: '"; |
c90c0ff4 |
1029 | print $1 while /(p)/gc; print "', pos=", pos, "\n"; |
44a8e56a |
1030 | } |
5d43e42d |
1031 | print "Final: '$1', pos=",pos,"\n" if /\G(.)/; |
44a8e56a |
1032 | |
1033 | The last example should print: |
1034 | |
1035 | 1: 'oo', pos=4 |
137443ea |
1036 | 2: 'q', pos=5 |
44a8e56a |
1037 | 3: 'pp', pos=7 |
1038 | 1: '', pos=7 |
137443ea |
1039 | 2: 'q', pos=8 |
1040 | 3: '', pos=8 |
5d43e42d |
1041 | Final: 'q', pos=8 |
1042 | |
1043 | Notice that the final match matched C<q> instead of C<p>, which a match |
1044 | without the C<\G> anchor would have done. Also note that the final match |
1045 | did not update C<pos> -- C<pos> is only updated on a C</g> match. If the |
1046 | final match did indeed match C<p>, it's a good bet that you're running an |
1047 | older (pre-5.6.0) Perl. |
44a8e56a |
1048 | |
c90c0ff4 |
1049 | A useful idiom for C<lex>-like scanners is C</\G.../gc>. You can |
e7ea3e70 |
1050 | combine several regexps like this to process a string part-by-part, |
c90c0ff4 |
1051 | doing different actions depending on which regexp matched. Each |
1052 | regexp tries to match where the previous one leaves off. |
e7ea3e70 |
1053 | |
3fe9a6f1 |
1054 | $_ = <<'EOL'; |
e7ea3e70 |
1055 | $url = new URI::URL "http://www/"; die if $url eq "xXx"; |
3fe9a6f1 |
1056 | EOL |
1057 | LOOP: |
e7ea3e70 |
1058 | { |
c90c0ff4 |
1059 | print(" digits"), redo LOOP if /\G\d+\b[,.;]?\s*/gc; |
1060 | print(" lowercase"), redo LOOP if /\G[a-z]+\b[,.;]?\s*/gc; |
1061 | print(" UPPERCASE"), redo LOOP if /\G[A-Z]+\b[,.;]?\s*/gc; |
1062 | print(" Capitalized"), redo LOOP if /\G[A-Z][a-z]+\b[,.;]?\s*/gc; |
1063 | print(" MiXeD"), redo LOOP if /\G[A-Za-z]+\b[,.;]?\s*/gc; |
1064 | print(" alphanumeric"), redo LOOP if /\G[A-Za-z0-9]+\b[,.;]?\s*/gc; |
1065 | print(" line-noise"), redo LOOP if /\G[^A-Za-z0-9]+/gc; |
e7ea3e70 |
1066 | print ". That's all!\n"; |
1067 | } |
1068 | |
1069 | Here is the output (split into several lines): |
1070 | |
1071 | line-noise lowercase line-noise lowercase UPPERCASE line-noise |
1072 | UPPERCASE line-noise lowercase line-noise lowercase line-noise |
1073 | lowercase lowercase line-noise lowercase lowercase line-noise |
1074 | MiXeD line-noise. That's all! |
44a8e56a |
1075 | |
a0d0e21e |
1076 | =item q/STRING/ |
1077 | |
1078 | =item C<'STRING'> |
1079 | |
19799a22 |
1080 | A single-quoted, literal string. A backslash represents a backslash |
68dc0745 |
1081 | unless followed by the delimiter or another backslash, in which case |
1082 | the delimiter or backslash is interpolated. |
a0d0e21e |
1083 | |
1084 | $foo = q!I said, "You said, 'She said it.'"!; |
1085 | $bar = q('This is it.'); |
68dc0745 |
1086 | $baz = '\n'; # a two-character string |
a0d0e21e |
1087 | |
1088 | =item qq/STRING/ |
1089 | |
1090 | =item "STRING" |
1091 | |
1092 | A double-quoted, interpolated string. |
1093 | |
1094 | $_ .= qq |
1095 | (*** The previous line contains the naughty word "$1".\n) |
19799a22 |
1096 | if /\b(tcl|java|python)\b/i; # :-) |
68dc0745 |
1097 | $baz = "\n"; # a one-character string |
a0d0e21e |
1098 | |
eec2d3df |
1099 | =item qr/STRING/imosx |
1100 | |
322edccd |
1101 | This operator quotes (and possibly compiles) its I<STRING> as a regular |
19799a22 |
1102 | expression. I<STRING> is interpolated the same way as I<PATTERN> |
1103 | in C<m/PATTERN/>. If "'" is used as the delimiter, no interpolation |
1104 | is done. Returns a Perl value which may be used instead of the |
1105 | corresponding C</STRING/imosx> expression. |
4b6a7270 |
1106 | |
1107 | For example, |
1108 | |
1109 | $rex = qr/my.STRING/is; |
1110 | s/$rex/foo/; |
1111 | |
1112 | is equivalent to |
1113 | |
1114 | s/my.STRING/foo/is; |
1115 | |
1116 | The result may be used as a subpattern in a match: |
eec2d3df |
1117 | |
1118 | $re = qr/$pattern/; |
0a92e3a8 |
1119 | $string =~ /foo${re}bar/; # can be interpolated in other patterns |
1120 | $string =~ $re; # or used standalone |
4b6a7270 |
1121 | $string =~ /$re/; # or this way |
1122 | |
1123 | Since Perl may compile the pattern at the moment of execution of qr() |
19799a22 |
1124 | operator, using qr() may have speed advantages in some situations, |
4b6a7270 |
1125 | notably if the result of qr() is used standalone: |
1126 | |
1127 | sub match { |
1128 | my $patterns = shift; |
1129 | my @compiled = map qr/$_/i, @$patterns; |
1130 | grep { |
1131 | my $success = 0; |
a7665c5e |
1132 | foreach my $pat (@compiled) { |
4b6a7270 |
1133 | $success = 1, last if /$pat/; |
1134 | } |
1135 | $success; |
1136 | } @_; |
1137 | } |
1138 | |
19799a22 |
1139 | Precompilation of the pattern into an internal representation at |
1140 | the moment of qr() avoids a need to recompile the pattern every |
1141 | time a match C</$pat/> is attempted. (Perl has many other internal |
1142 | optimizations, but none would be triggered in the above example if |
1143 | we did not use qr() operator.) |
eec2d3df |
1144 | |
1145 | Options are: |
1146 | |
1147 | i Do case-insensitive pattern matching. |
1148 | m Treat string as multiple lines. |
1149 | o Compile pattern only once. |
1150 | s Treat string as single line. |
1151 | x Use extended regular expressions. |
1152 | |
0a92e3a8 |
1153 | See L<perlre> for additional information on valid syntax for STRING, and |
1154 | for a detailed look at the semantics of regular expressions. |
1155 | |
a0d0e21e |
1156 | =item qx/STRING/ |
1157 | |
1158 | =item `STRING` |
1159 | |
43dd4d21 |
1160 | A string which is (possibly) interpolated and then executed as a |
1161 | system command with C</bin/sh> or its equivalent. Shell wildcards, |
1162 | pipes, and redirections will be honored. The collected standard |
1163 | output of the command is returned; standard error is unaffected. In |
1164 | scalar context, it comes back as a single (potentially multi-line) |
1165 | string, or undef if the command failed. In list context, returns a |
1166 | list of lines (however you've defined lines with $/ or |
1167 | $INPUT_RECORD_SEPARATOR), or an empty list if the command failed. |
5a964f20 |
1168 | |
1169 | Because backticks do not affect standard error, use shell file descriptor |
1170 | syntax (assuming the shell supports this) if you care to address this. |
1171 | To capture a command's STDERR and STDOUT together: |
a0d0e21e |
1172 | |
5a964f20 |
1173 | $output = `cmd 2>&1`; |
1174 | |
1175 | To capture a command's STDOUT but discard its STDERR: |
1176 | |
1177 | $output = `cmd 2>/dev/null`; |
1178 | |
1179 | To capture a command's STDERR but discard its STDOUT (ordering is |
1180 | important here): |
1181 | |
1182 | $output = `cmd 2>&1 1>/dev/null`; |
1183 | |
1184 | To exchange a command's STDOUT and STDERR in order to capture the STDERR |
1185 | but leave its STDOUT to come out the old STDERR: |
1186 | |
1187 | $output = `cmd 3>&1 1>&2 2>&3 3>&-`; |
1188 | |
1189 | To read both a command's STDOUT and its STDERR separately, it's easiest |
2359510d |
1190 | to redirect them separately to files, and then read from those files |
1191 | when the program is done: |
5a964f20 |
1192 | |
2359510d |
1193 | system("program args 1>program.stdout 2>program.stderr"); |
5a964f20 |
1194 | |
1195 | Using single-quote as a delimiter protects the command from Perl's |
1196 | double-quote interpolation, passing it on to the shell instead: |
1197 | |
1198 | $perl_info = qx(ps $$); # that's Perl's $$ |
1199 | $shell_info = qx'ps $$'; # that's the new shell's $$ |
1200 | |
19799a22 |
1201 | How that string gets evaluated is entirely subject to the command |
5a964f20 |
1202 | interpreter on your system. On most platforms, you will have to protect |
1203 | shell metacharacters if you want them treated literally. This is in |
1204 | practice difficult to do, as it's unclear how to escape which characters. |
1205 | See L<perlsec> for a clean and safe example of a manual fork() and exec() |
1206 | to emulate backticks safely. |
a0d0e21e |
1207 | |
bb32b41a |
1208 | On some platforms (notably DOS-like ones), the shell may not be |
1209 | capable of dealing with multiline commands, so putting newlines in |
1210 | the string may not get you what you want. You may be able to evaluate |
1211 | multiple commands in a single line by separating them with the command |
1212 | separator character, if your shell supports that (e.g. C<;> on many Unix |
1213 | shells; C<&> on the Windows NT C<cmd> shell). |
1214 | |
0f897271 |
1215 | Beginning with v5.6.0, Perl will attempt to flush all files opened for |
1216 | output before starting the child process, but this may not be supported |
1217 | on some platforms (see L<perlport>). To be safe, you may need to set |
1218 | C<$|> ($AUTOFLUSH in English) or call the C<autoflush()> method of |
1219 | C<IO::Handle> on any open handles. |
1220 | |
bb32b41a |
1221 | Beware that some command shells may place restrictions on the length |
1222 | of the command line. You must ensure your strings don't exceed this |
1223 | limit after any necessary interpolations. See the platform-specific |
1224 | release notes for more details about your particular environment. |
1225 | |
5a964f20 |
1226 | Using this operator can lead to programs that are difficult to port, |
1227 | because the shell commands called vary between systems, and may in |
1228 | fact not be present at all. As one example, the C<type> command under |
1229 | the POSIX shell is very different from the C<type> command under DOS. |
1230 | That doesn't mean you should go out of your way to avoid backticks |
1231 | when they're the right way to get something done. Perl was made to be |
1232 | a glue language, and one of the things it glues together is commands. |
1233 | Just understand what you're getting yourself into. |
bb32b41a |
1234 | |
dc848c6f |
1235 | See L<"I/O Operators"> for more discussion. |
a0d0e21e |
1236 | |
945c54fd |
1237 | =item qw/STRING/ |
1238 | |
1239 | Evaluates to a list of the words extracted out of STRING, using embedded |
1240 | whitespace as the word delimiters. It can be understood as being roughly |
1241 | equivalent to: |
1242 | |
1243 | split(' ', q/STRING/); |
1244 | |
efb1e162 |
1245 | the differences being that it generates a real list at compile time, and |
1246 | in scalar context it returns the last element in the list. So |
945c54fd |
1247 | this expression: |
1248 | |
1249 | qw(foo bar baz) |
1250 | |
1251 | is semantically equivalent to the list: |
1252 | |
1253 | 'foo', 'bar', 'baz' |
1254 | |
1255 | Some frequently seen examples: |
1256 | |
1257 | use POSIX qw( setlocale localeconv ) |
1258 | @EXPORT = qw( foo bar baz ); |
1259 | |
1260 | A common mistake is to try to separate the words with comma or to |
1261 | put comments into a multi-line C<qw>-string. For this reason, the |
1262 | C<use warnings> pragma and the B<-w> switch (that is, the C<$^W> variable) |
1263 | produces warnings if the STRING contains the "," or the "#" character. |
1264 | |
a0d0e21e |
1265 | =item s/PATTERN/REPLACEMENT/egimosx |
1266 | |
1267 | Searches a string for a pattern, and if found, replaces that pattern |
1268 | with the replacement text and returns the number of substitutions |
e37d713d |
1269 | made. Otherwise it returns false (specifically, the empty string). |
a0d0e21e |
1270 | |
1271 | If no string is specified via the C<=~> or C<!~> operator, the C<$_> |
1272 | variable is searched and modified. (The string specified with C<=~> must |
5a964f20 |
1273 | be scalar variable, an array element, a hash element, or an assignment |
5f05dabc |
1274 | to one of those, i.e., an lvalue.) |
a0d0e21e |
1275 | |
19799a22 |
1276 | If the delimiter chosen is a single quote, no interpolation is |
a0d0e21e |
1277 | done on either the PATTERN or the REPLACEMENT. Otherwise, if the |
1278 | PATTERN contains a $ that looks like a variable rather than an |
1279 | end-of-string test, the variable will be interpolated into the pattern |
5f05dabc |
1280 | at run-time. If you want the pattern compiled only once the first time |
a0d0e21e |
1281 | the variable is interpolated, use the C</o> option. If the pattern |
5a964f20 |
1282 | evaluates to the empty string, the last successfully executed regular |
a0d0e21e |
1283 | expression is used instead. See L<perlre> for further explanation on these. |
5a964f20 |
1284 | See L<perllocale> for discussion of additional considerations that apply |
a034a98d |
1285 | when C<use locale> is in effect. |
a0d0e21e |
1286 | |
1287 | Options are: |
1288 | |
1289 | e Evaluate the right side as an expression. |
5f05dabc |
1290 | g Replace globally, i.e., all occurrences. |
a0d0e21e |
1291 | i Do case-insensitive pattern matching. |
1292 | m Treat string as multiple lines. |
5f05dabc |
1293 | o Compile pattern only once. |
a0d0e21e |
1294 | s Treat string as single line. |
1295 | x Use extended regular expressions. |
1296 | |
1297 | Any non-alphanumeric, non-whitespace delimiter may replace the |
1298 | slashes. If single quotes are used, no interpretation is done on the |
e37d713d |
1299 | replacement string (the C</e> modifier overrides this, however). Unlike |
54310121 |
1300 | Perl 4, Perl 5 treats backticks as normal delimiters; the replacement |
e37d713d |
1301 | text is not evaluated as a command. If the |
a0d0e21e |
1302 | PATTERN is delimited by bracketing quotes, the REPLACEMENT has its own |
5f05dabc |
1303 | pair of quotes, which may or may not be bracketing quotes, e.g., |
35f2feb0 |
1304 | C<s(foo)(bar)> or C<< s<foo>/bar/ >>. A C</e> will cause the |
cec88af6 |
1305 | replacement portion to be treated as a full-fledged Perl expression |
1306 | and evaluated right then and there. It is, however, syntax checked at |
1307 | compile-time. A second C<e> modifier will cause the replacement portion |
1308 | to be C<eval>ed before being run as a Perl expression. |
a0d0e21e |
1309 | |
1310 | Examples: |
1311 | |
1312 | s/\bgreen\b/mauve/g; # don't change wintergreen |
1313 | |
1314 | $path =~ s|/usr/bin|/usr/local/bin|; |
1315 | |
1316 | s/Login: $foo/Login: $bar/; # run-time pattern |
1317 | |
5a964f20 |
1318 | ($foo = $bar) =~ s/this/that/; # copy first, then change |
a0d0e21e |
1319 | |
5a964f20 |
1320 | $count = ($paragraph =~ s/Mister\b/Mr./g); # get change-count |
a0d0e21e |
1321 | |
1322 | $_ = 'abc123xyz'; |
1323 | s/\d+/$&*2/e; # yields 'abc246xyz' |
1324 | s/\d+/sprintf("%5d",$&)/e; # yields 'abc 246xyz' |
1325 | s/\w/$& x 2/eg; # yields 'aabbcc 224466xxyyzz' |
1326 | |
1327 | s/%(.)/$percent{$1}/g; # change percent escapes; no /e |
1328 | s/%(.)/$percent{$1} || $&/ge; # expr now, so /e |
1329 | s/^=(\w+)/&pod($1)/ge; # use function call |
1330 | |
5a964f20 |
1331 | # expand variables in $_, but dynamics only, using |
1332 | # symbolic dereferencing |
1333 | s/\$(\w+)/${$1}/g; |
1334 | |
cec88af6 |
1335 | # Add one to the value of any numbers in the string |
1336 | s/(\d+)/1 + $1/eg; |
1337 | |
1338 | # This will expand any embedded scalar variable |
1339 | # (including lexicals) in $_ : First $1 is interpolated |
1340 | # to the variable name, and then evaluated |
a0d0e21e |
1341 | s/(\$\w+)/$1/eeg; |
1342 | |
5a964f20 |
1343 | # Delete (most) C comments. |
a0d0e21e |
1344 | $program =~ s { |
4633a7c4 |
1345 | /\* # Match the opening delimiter. |
1346 | .*? # Match a minimal number of characters. |
1347 | \*/ # Match the closing delimiter. |
a0d0e21e |
1348 | } []gsx; |
1349 | |
5a964f20 |
1350 | s/^\s*(.*?)\s*$/$1/; # trim white space in $_, expensively |
1351 | |
1352 | for ($variable) { # trim white space in $variable, cheap |
1353 | s/^\s+//; |
1354 | s/\s+$//; |
1355 | } |
a0d0e21e |
1356 | |
1357 | s/([^ ]*) *([^ ]*)/$2 $1/; # reverse 1st two fields |
1358 | |
54310121 |
1359 | Note the use of $ instead of \ in the last example. Unlike |
35f2feb0 |
1360 | B<sed>, we use the \<I<digit>> form in only the left hand side. |
1361 | Anywhere else it's $<I<digit>>. |
a0d0e21e |
1362 | |
5f05dabc |
1363 | Occasionally, you can't use just a C</g> to get all the changes |
19799a22 |
1364 | to occur that you might want. Here are two common cases: |
a0d0e21e |
1365 | |
1366 | # put commas in the right places in an integer |
19799a22 |
1367 | 1 while s/(\d)(\d\d\d)(?!\d)/$1,$2/g; |
a0d0e21e |
1368 | |
1369 | # expand tabs to 8-column spacing |
1370 | 1 while s/\t+/' ' x (length($&)*8 - length($`)%8)/e; |
1371 | |
6940069f |
1372 | =item tr/SEARCHLIST/REPLACEMENTLIST/cds |
a0d0e21e |
1373 | |
6940069f |
1374 | =item y/SEARCHLIST/REPLACEMENTLIST/cds |
a0d0e21e |
1375 | |
2c268ad5 |
1376 | Transliterates all occurrences of the characters found in the search list |
a0d0e21e |
1377 | with the corresponding character in the replacement list. It returns |
1378 | the number of characters replaced or deleted. If no string is |
2c268ad5 |
1379 | specified via the =~ or !~ operator, the $_ string is transliterated. (The |
54310121 |
1380 | string specified with =~ must be a scalar variable, an array element, a |
1381 | hash element, or an assignment to one of those, i.e., an lvalue.) |
8ada0baa |
1382 | |
2c268ad5 |
1383 | A character range may be specified with a hyphen, so C<tr/A-J/0-9/> |
1384 | does the same replacement as C<tr/ACEGIBDFHJ/0246813579/>. |
54310121 |
1385 | For B<sed> devotees, C<y> is provided as a synonym for C<tr>. If the |
1386 | SEARCHLIST is delimited by bracketing quotes, the REPLACEMENTLIST has |
1387 | its own pair of quotes, which may or may not be bracketing quotes, |
2c268ad5 |
1388 | e.g., C<tr[A-Z][a-z]> or C<tr(+\-*/)/ABCD/>. |
a0d0e21e |
1389 | |
cc255d5f |
1390 | Note that C<tr> does B<not> do regular expression character classes |
1391 | such as C<\d> or C<[:lower:]>. The <tr> operator is not equivalent to |
1392 | the tr(1) utility. If you want to map strings between lower/upper |
1393 | cases, see L<perlfunc/lc> and L<perlfunc/uc>, and in general consider |
1394 | using the C<s> operator if you need regular expressions. |
1395 | |
8ada0baa |
1396 | Note also that the whole range idea is rather unportable between |
1397 | character sets--and even within character sets they may cause results |
1398 | you probably didn't expect. A sound principle is to use only ranges |
1399 | that begin from and end at either alphabets of equal case (a-e, A-E), |
1400 | or digits (0-4). Anything else is unsafe. If in doubt, spell out the |
1401 | character sets in full. |
1402 | |
a0d0e21e |
1403 | Options: |
1404 | |
1405 | c Complement the SEARCHLIST. |
1406 | d Delete found but unreplaced characters. |
1407 | s Squash duplicate replaced characters. |
1408 | |
19799a22 |
1409 | If the C</c> modifier is specified, the SEARCHLIST character set |
1410 | is complemented. If the C</d> modifier is specified, any characters |
1411 | specified by SEARCHLIST not found in REPLACEMENTLIST are deleted. |
1412 | (Note that this is slightly more flexible than the behavior of some |
1413 | B<tr> programs, which delete anything they find in the SEARCHLIST, |
1414 | period.) If the C</s> modifier is specified, sequences of characters |
1415 | that were transliterated to the same character are squashed down |
1416 | to a single instance of the character. |
a0d0e21e |
1417 | |
1418 | If the C</d> modifier is used, the REPLACEMENTLIST is always interpreted |
1419 | exactly as specified. Otherwise, if the REPLACEMENTLIST is shorter |
1420 | than the SEARCHLIST, the final character is replicated till it is long |
5a964f20 |
1421 | enough. If the REPLACEMENTLIST is empty, the SEARCHLIST is replicated. |
a0d0e21e |
1422 | This latter is useful for counting characters in a class or for |
1423 | squashing character sequences in a class. |
1424 | |
1425 | Examples: |
1426 | |
1427 | $ARGV[1] =~ tr/A-Z/a-z/; # canonicalize to lower case |
1428 | |
1429 | $cnt = tr/*/*/; # count the stars in $_ |
1430 | |
1431 | $cnt = $sky =~ tr/*/*/; # count the stars in $sky |
1432 | |
1433 | $cnt = tr/0-9//; # count the digits in $_ |
1434 | |
1435 | tr/a-zA-Z//s; # bookkeeper -> bokeper |
1436 | |
1437 | ($HOST = $host) =~ tr/a-z/A-Z/; |
1438 | |
1439 | tr/a-zA-Z/ /cs; # change non-alphas to single space |
1440 | |
1441 | tr [\200-\377] |
1442 | [\000-\177]; # delete 8th bit |
1443 | |
19799a22 |
1444 | If multiple transliterations are given for a character, only the |
1445 | first one is used: |
748a9306 |
1446 | |
1447 | tr/AAA/XYZ/ |
1448 | |
2c268ad5 |
1449 | will transliterate any A to X. |
748a9306 |
1450 | |
19799a22 |
1451 | Because the transliteration table is built at compile time, neither |
a0d0e21e |
1452 | the SEARCHLIST nor the REPLACEMENTLIST are subjected to double quote |
19799a22 |
1453 | interpolation. That means that if you want to use variables, you |
1454 | must use an eval(): |
a0d0e21e |
1455 | |
1456 | eval "tr/$oldlist/$newlist/"; |
1457 | die $@ if $@; |
1458 | |
1459 | eval "tr/$oldlist/$newlist/, 1" or die $@; |
1460 | |
7e3b091d |
1461 | =item <<EOF |
1462 | |
1463 | A line-oriented form of quoting is based on the shell "here-document" |
1464 | syntax. Following a C<< << >> you specify a string to terminate |
1465 | the quoted material, and all lines following the current line down to |
1466 | the terminating string are the value of the item. The terminating |
1467 | string may be either an identifier (a word), or some quoted text. If |
1468 | quoted, the type of quotes you use determines the treatment of the |
1469 | text, just as in regular quoting. An unquoted identifier works like |
1470 | double quotes. There must be no space between the C<< << >> and |
1471 | the identifier, unless the identifier is quoted. (If you put a space it |
1472 | will be treated as a null identifier, which is valid, and matches the first |
1473 | empty line.) The terminating string must appear by itself (unquoted and |
1474 | with no surrounding whitespace) on the terminating line. |
1475 | |
1476 | print <<EOF; |
1477 | The price is $Price. |
1478 | EOF |
1479 | |
1480 | print << "EOF"; # same as above |
1481 | The price is $Price. |
1482 | EOF |
1483 | |
1484 | print << `EOC`; # execute commands |
1485 | echo hi there |
1486 | echo lo there |
1487 | EOC |
1488 | |
1489 | print <<"foo", <<"bar"; # you can stack them |
1490 | I said foo. |
1491 | foo |
1492 | I said bar. |
1493 | bar |
1494 | |
1495 | myfunc(<< "THIS", 23, <<'THAT'); |
1496 | Here's a line |
1497 | or two. |
1498 | THIS |
1499 | and here's another. |
1500 | THAT |
1501 | |
1502 | Just don't forget that you have to put a semicolon on the end |
1503 | to finish the statement, as Perl doesn't know you're not going to |
1504 | try to do this: |
1505 | |
1506 | print <<ABC |
1507 | 179231 |
1508 | ABC |
1509 | + 20; |
1510 | |
1511 | If you want your here-docs to be indented with the |
1512 | rest of the code, you'll need to remove leading whitespace |
1513 | from each line manually: |
1514 | |
1515 | ($quote = <<'FINIS') =~ s/^\s+//gm; |
1516 | The Road goes ever on and on, |
1517 | down from the door where it began. |
1518 | FINIS |
1519 | |
1520 | If you use a here-doc within a delimited construct, such as in C<s///eg>, |
1521 | the quoted material must come on the lines following the final delimiter. |
1522 | So instead of |
1523 | |
1524 | s/this/<<E . 'that' |
1525 | the other |
1526 | E |
1527 | . 'more '/eg; |
1528 | |
1529 | you have to write |
1530 | |
1531 | s/this/<<E . 'that' |
1532 | . 'more '/eg; |
1533 | the other |
1534 | E |
1535 | |
1536 | If the terminating identifier is on the last line of the program, you |
1537 | must be sure there is a newline after it; otherwise, Perl will give the |
1538 | warning B<Can't find string terminator "END" anywhere before EOF...>. |
1539 | |
1540 | Additionally, the quoting rules for the identifier are not related to |
1541 | Perl's quoting rules -- C<q()>, C<qq()>, and the like are not supported |
1542 | in place of C<''> and C<"">, and the only interpolation is for backslashing |
1543 | the quoting character: |
1544 | |
1545 | print << "abc\"def"; |
1546 | testing... |
1547 | abc"def |
1548 | |
1549 | Finally, quoted strings cannot span multiple lines. The general rule is |
1550 | that the identifier must be a string literal. Stick with that, and you |
1551 | should be safe. |
1552 | |
a0d0e21e |
1553 | =back |
1554 | |
75e14d17 |
1555 | =head2 Gory details of parsing quoted constructs |
1556 | |
19799a22 |
1557 | When presented with something that might have several different |
1558 | interpretations, Perl uses the B<DWIM> (that's "Do What I Mean") |
1559 | principle to pick the most probable interpretation. This strategy |
1560 | is so successful that Perl programmers often do not suspect the |
1561 | ambivalence of what they write. But from time to time, Perl's |
1562 | notions differ substantially from what the author honestly meant. |
1563 | |
1564 | This section hopes to clarify how Perl handles quoted constructs. |
1565 | Although the most common reason to learn this is to unravel labyrinthine |
1566 | regular expressions, because the initial steps of parsing are the |
1567 | same for all quoting operators, they are all discussed together. |
1568 | |
1569 | The most important Perl parsing rule is the first one discussed |
1570 | below: when processing a quoted construct, Perl first finds the end |
1571 | of that construct, then interprets its contents. If you understand |
1572 | this rule, you may skip the rest of this section on the first |
1573 | reading. The other rules are likely to contradict the user's |
1574 | expectations much less frequently than this first one. |
1575 | |
1576 | Some passes discussed below are performed concurrently, but because |
1577 | their results are the same, we consider them individually. For different |
1578 | quoting constructs, Perl performs different numbers of passes, from |
1579 | one to five, but these passes are always performed in the same order. |
75e14d17 |
1580 | |
13a2d996 |
1581 | =over 4 |
75e14d17 |
1582 | |
1583 | =item Finding the end |
1584 | |
19799a22 |
1585 | The first pass is finding the end of the quoted construct, whether |
1586 | it be a multicharacter delimiter C<"\nEOF\n"> in the C<<<EOF> |
1587 | construct, a C</> that terminates a C<qq//> construct, a C<]> which |
35f2feb0 |
1588 | terminates C<qq[]> construct, or a C<< > >> which terminates a |
1589 | fileglob started with C<< < >>. |
75e14d17 |
1590 | |
19799a22 |
1591 | When searching for single-character non-pairing delimiters, such |
1592 | as C</>, combinations of C<\\> and C<\/> are skipped. However, |
1593 | when searching for single-character pairing delimiter like C<[>, |
1594 | combinations of C<\\>, C<\]>, and C<\[> are all skipped, and nested |
1595 | C<[>, C<]> are skipped as well. When searching for multicharacter |
1596 | delimiters, nothing is skipped. |
75e14d17 |
1597 | |
19799a22 |
1598 | For constructs with three-part delimiters (C<s///>, C<y///>, and |
1599 | C<tr///>), the search is repeated once more. |
75e14d17 |
1600 | |
19799a22 |
1601 | During this search no attention is paid to the semantics of the construct. |
1602 | Thus: |
75e14d17 |
1603 | |
1604 | "$hash{"$foo/$bar"}" |
1605 | |
2a94b7ce |
1606 | or: |
75e14d17 |
1607 | |
1608 | m/ |
2a94b7ce |
1609 | bar # NOT a comment, this slash / terminated m//! |
75e14d17 |
1610 | /x |
1611 | |
19799a22 |
1612 | do not form legal quoted expressions. The quoted part ends on the |
1613 | first C<"> and C</>, and the rest happens to be a syntax error. |
1614 | Because the slash that terminated C<m//> was followed by a C<SPACE>, |
1615 | the example above is not C<m//x>, but rather C<m//> with no C</x> |
1616 | modifier. So the embedded C<#> is interpreted as a literal C<#>. |
75e14d17 |
1617 | |
1618 | =item Removal of backslashes before delimiters |
1619 | |
19799a22 |
1620 | During the second pass, text between the starting and ending |
1621 | delimiters is copied to a safe location, and the C<\> is removed |
1622 | from combinations consisting of C<\> and delimiter--or delimiters, |
1623 | meaning both starting and ending delimiters will should these differ. |
1624 | This removal does not happen for multi-character delimiters. |
1625 | Note that the combination C<\\> is left intact, just as it was. |
75e14d17 |
1626 | |
19799a22 |
1627 | Starting from this step no information about the delimiters is |
1628 | used in parsing. |
75e14d17 |
1629 | |
1630 | =item Interpolation |
1631 | |
19799a22 |
1632 | The next step is interpolation in the text obtained, which is now |
1633 | delimiter-independent. There are four different cases. |
75e14d17 |
1634 | |
13a2d996 |
1635 | =over 4 |
75e14d17 |
1636 | |
1637 | =item C<<<'EOF'>, C<m''>, C<s'''>, C<tr///>, C<y///> |
1638 | |
1639 | No interpolation is performed. |
1640 | |
1641 | =item C<''>, C<q//> |
1642 | |
1643 | The only interpolation is removal of C<\> from pairs C<\\>. |
1644 | |
35f2feb0 |
1645 | =item C<"">, C<``>, C<qq//>, C<qx//>, C<< <file*glob> >> |
75e14d17 |
1646 | |
19799a22 |
1647 | C<\Q>, C<\U>, C<\u>, C<\L>, C<\l> (possibly paired with C<\E>) are |
1648 | converted to corresponding Perl constructs. Thus, C<"$foo\Qbaz$bar"> |
1649 | is converted to C<$foo . (quotemeta("baz" . $bar))> internally. |
1650 | The other combinations are replaced with appropriate expansions. |
2a94b7ce |
1651 | |
19799a22 |
1652 | Let it be stressed that I<whatever falls between C<\Q> and C<\E>> |
1653 | is interpolated in the usual way. Something like C<"\Q\\E"> has |
1654 | no C<\E> inside. instead, it has C<\Q>, C<\\>, and C<E>, so the |
1655 | result is the same as for C<"\\\\E">. As a general rule, backslashes |
1656 | between C<\Q> and C<\E> may lead to counterintuitive results. So, |
1657 | C<"\Q\t\E"> is converted to C<quotemeta("\t")>, which is the same |
1658 | as C<"\\\t"> (since TAB is not alphanumeric). Note also that: |
2a94b7ce |
1659 | |
1660 | $str = '\t'; |
1661 | return "\Q$str"; |
1662 | |
1663 | may be closer to the conjectural I<intention> of the writer of C<"\Q\t\E">. |
1664 | |
19799a22 |
1665 | Interpolated scalars and arrays are converted internally to the C<join> and |
92d29cee |
1666 | C<.> catenation operations. Thus, C<"$foo XXX '@arr'"> becomes: |
75e14d17 |
1667 | |
19799a22 |
1668 | $foo . " XXX '" . (join $", @arr) . "'"; |
75e14d17 |
1669 | |
19799a22 |
1670 | All operations above are performed simultaneously, left to right. |
75e14d17 |
1671 | |
19799a22 |
1672 | Because the result of C<"\Q STRING \E"> has all metacharacters |
1673 | quoted, there is no way to insert a literal C<$> or C<@> inside a |
1674 | C<\Q\E> pair. If protected by C<\>, C<$> will be quoted to became |
1675 | C<"\\\$">; if not, it is interpreted as the start of an interpolated |
1676 | scalar. |
75e14d17 |
1677 | |
19799a22 |
1678 | Note also that the interpolation code needs to make a decision on |
1679 | where the interpolated scalar ends. For instance, whether |
35f2feb0 |
1680 | C<< "a $b -> {c}" >> really means: |
75e14d17 |
1681 | |
1682 | "a " . $b . " -> {c}"; |
1683 | |
2a94b7ce |
1684 | or: |
75e14d17 |
1685 | |
1686 | "a " . $b -> {c}; |
1687 | |
19799a22 |
1688 | Most of the time, the longest possible text that does not include |
1689 | spaces between components and which contains matching braces or |
1690 | brackets. because the outcome may be determined by voting based |
1691 | on heuristic estimators, the result is not strictly predictable. |
1692 | Fortunately, it's usually correct for ambiguous cases. |
75e14d17 |
1693 | |
1694 | =item C<?RE?>, C</RE/>, C<m/RE/>, C<s/RE/foo/>, |
1695 | |
19799a22 |
1696 | Processing of C<\Q>, C<\U>, C<\u>, C<\L>, C<\l>, and interpolation |
1697 | happens (almost) as with C<qq//> constructs, but the substitution |
1698 | of C<\> followed by RE-special chars (including C<\>) is not |
1699 | performed. Moreover, inside C<(?{BLOCK})>, C<(?# comment )>, and |
1700 | a C<#>-comment in a C<//x>-regular expression, no processing is |
1701 | performed whatsoever. This is the first step at which the presence |
1702 | of the C<//x> modifier is relevant. |
1703 | |
1704 | Interpolation has several quirks: C<$|>, C<$(>, and C<$)> are not |
1705 | interpolated, and constructs C<$var[SOMETHING]> are voted (by several |
1706 | different estimators) to be either an array element or C<$var> |
1707 | followed by an RE alternative. This is where the notation |
1708 | C<${arr[$bar]}> comes handy: C</${arr[0-9]}/> is interpreted as |
1709 | array element C<-9>, not as a regular expression from the variable |
1710 | C<$arr> followed by a digit, which would be the interpretation of |
1711 | C</$arr[0-9]/>. Since voting among different estimators may occur, |
1712 | the result is not predictable. |
1713 | |
1714 | It is at this step that C<\1> is begrudgingly converted to C<$1> in |
1715 | the replacement text of C<s///> to correct the incorrigible |
1716 | I<sed> hackers who haven't picked up the saner idiom yet. A warning |
9f1b1f2d |
1717 | is emitted if the C<use warnings> pragma or the B<-w> command-line flag |
1718 | (that is, the C<$^W> variable) was set. |
19799a22 |
1719 | |
1720 | The lack of processing of C<\\> creates specific restrictions on |
1721 | the post-processed text. If the delimiter is C</>, one cannot get |
1722 | the combination C<\/> into the result of this step. C</> will |
1723 | finish the regular expression, C<\/> will be stripped to C</> on |
1724 | the previous step, and C<\\/> will be left as is. Because C</> is |
1725 | equivalent to C<\/> inside a regular expression, this does not |
1726 | matter unless the delimiter happens to be character special to the |
1727 | RE engine, such as in C<s*foo*bar*>, C<m[foo]>, or C<?foo?>; or an |
1728 | alphanumeric char, as in: |
2a94b7ce |
1729 | |
1730 | m m ^ a \s* b mmx; |
1731 | |
19799a22 |
1732 | In the RE above, which is intentionally obfuscated for illustration, the |
2a94b7ce |
1733 | delimiter is C<m>, the modifier is C<mx>, and after backslash-removal the |
aa863641 |
1734 | RE is the same as for C<m/ ^ a \s* b /mx>. There's more than one |
19799a22 |
1735 | reason you're encouraged to restrict your delimiters to non-alphanumeric, |
1736 | non-whitespace choices. |
75e14d17 |
1737 | |
1738 | =back |
1739 | |
19799a22 |
1740 | This step is the last one for all constructs except regular expressions, |
75e14d17 |
1741 | which are processed further. |
1742 | |
1743 | =item Interpolation of regular expressions |
1744 | |
19799a22 |
1745 | Previous steps were performed during the compilation of Perl code, |
1746 | but this one happens at run time--although it may be optimized to |
1747 | be calculated at compile time if appropriate. After preprocessing |
1748 | described above, and possibly after evaluation if catenation, |
1749 | joining, casing translation, or metaquoting are involved, the |
1750 | resulting I<string> is passed to the RE engine for compilation. |
1751 | |
1752 | Whatever happens in the RE engine might be better discussed in L<perlre>, |
1753 | but for the sake of continuity, we shall do so here. |
1754 | |
1755 | This is another step where the presence of the C<//x> modifier is |
1756 | relevant. The RE engine scans the string from left to right and |
1757 | converts it to a finite automaton. |
1758 | |
1759 | Backslashed characters are either replaced with corresponding |
1760 | literal strings (as with C<\{>), or else they generate special nodes |
1761 | in the finite automaton (as with C<\b>). Characters special to the |
1762 | RE engine (such as C<|>) generate corresponding nodes or groups of |
1763 | nodes. C<(?#...)> comments are ignored. All the rest is either |
1764 | converted to literal strings to match, or else is ignored (as is |
1765 | whitespace and C<#>-style comments if C<//x> is present). |
1766 | |
1767 | Parsing of the bracketed character class construct, C<[...]>, is |
1768 | rather different than the rule used for the rest of the pattern. |
1769 | The terminator of this construct is found using the same rules as |
1770 | for finding the terminator of a C<{}>-delimited construct, the only |
1771 | exception being that C<]> immediately following C<[> is treated as |
1772 | though preceded by a backslash. Similarly, the terminator of |
1773 | C<(?{...})> is found using the same rules as for finding the |
1774 | terminator of a C<{}>-delimited construct. |
1775 | |
1776 | It is possible to inspect both the string given to RE engine and the |
1777 | resulting finite automaton. See the arguments C<debug>/C<debugcolor> |
1778 | in the C<use L<re>> pragma, as well as Perl's B<-Dr> command-line |
4a4eefd0 |
1779 | switch documented in L<perlrun/"Command Switches">. |
75e14d17 |
1780 | |
1781 | =item Optimization of regular expressions |
1782 | |
7522fed5 |
1783 | This step is listed for completeness only. Since it does not change |
75e14d17 |
1784 | semantics, details of this step are not documented and are subject |
19799a22 |
1785 | to change without notice. This step is performed over the finite |
1786 | automaton that was generated during the previous pass. |
2a94b7ce |
1787 | |
19799a22 |
1788 | It is at this stage that C<split()> silently optimizes C</^/> to |
1789 | mean C</^/m>. |
75e14d17 |
1790 | |
1791 | =back |
1792 | |
a0d0e21e |
1793 | =head2 I/O Operators |
1794 | |
54310121 |
1795 | There are several I/O operators you should know about. |
fbad3eb5 |
1796 | |
7b8d334a |
1797 | A string enclosed by backticks (grave accents) first undergoes |
19799a22 |
1798 | double-quote interpolation. It is then interpreted as an external |
1799 | command, and the output of that command is the value of the |
e9c56f9b |
1800 | backtick string, like in a shell. In scalar context, a single string |
1801 | consisting of all output is returned. In list context, a list of |
1802 | values is returned, one per line of output. (You can set C<$/> to use |
1803 | a different line terminator.) The command is executed each time the |
1804 | pseudo-literal is evaluated. The status value of the command is |
1805 | returned in C<$?> (see L<perlvar> for the interpretation of C<$?>). |
1806 | Unlike in B<csh>, no translation is done on the return data--newlines |
1807 | remain newlines. Unlike in any of the shells, single quotes do not |
1808 | hide variable names in the command from interpretation. To pass a |
1809 | literal dollar-sign through to the shell you need to hide it with a |
1810 | backslash. The generalized form of backticks is C<qx//>. (Because |
1811 | backticks always undergo shell expansion as well, see L<perlsec> for |
1812 | security concerns.) |
19799a22 |
1813 | |
1814 | In scalar context, evaluating a filehandle in angle brackets yields |
1815 | the next line from that file (the newline, if any, included), or |
1816 | C<undef> at end-of-file or on error. When C<$/> is set to C<undef> |
1817 | (sometimes known as file-slurp mode) and the file is empty, it |
1818 | returns C<''> the first time, followed by C<undef> subsequently. |
1819 | |
1820 | Ordinarily you must assign the returned value to a variable, but |
1821 | there is one situation where an automatic assignment happens. If |
1822 | and only if the input symbol is the only thing inside the conditional |
1823 | of a C<while> statement (even if disguised as a C<for(;;)> loop), |
1824 | the value is automatically assigned to the global variable $_, |
1825 | destroying whatever was there previously. (This may seem like an |
1826 | odd thing to you, but you'll use the construct in almost every Perl |
17b829fa |
1827 | script you write.) The $_ variable is not implicitly localized. |
19799a22 |
1828 | You'll have to put a C<local $_;> before the loop if you want that |
1829 | to happen. |
1830 | |
1831 | The following lines are equivalent: |
a0d0e21e |
1832 | |
748a9306 |
1833 | while (defined($_ = <STDIN>)) { print; } |
7b8d334a |
1834 | while ($_ = <STDIN>) { print; } |
a0d0e21e |
1835 | while (<STDIN>) { print; } |
1836 | for (;<STDIN>;) { print; } |
748a9306 |
1837 | print while defined($_ = <STDIN>); |
7b8d334a |
1838 | print while ($_ = <STDIN>); |
a0d0e21e |
1839 | print while <STDIN>; |
1840 | |
19799a22 |
1841 | This also behaves similarly, but avoids $_ : |
7b8d334a |
1842 | |
1843 | while (my $line = <STDIN>) { print $line } |
1844 | |
19799a22 |
1845 | In these loop constructs, the assigned value (whether assignment |
1846 | is automatic or explicit) is then tested to see whether it is |
1847 | defined. The defined test avoids problems where line has a string |
1848 | value that would be treated as false by Perl, for example a "" or |
1849 | a "0" with no trailing newline. If you really mean for such values |
1850 | to terminate the loop, they should be tested for explicitly: |
7b8d334a |
1851 | |
1852 | while (($_ = <STDIN>) ne '0') { ... } |
1853 | while (<STDIN>) { last unless $_; ... } |
1854 | |
35f2feb0 |
1855 | In other boolean contexts, C<< <I<filehandle>> >> without an |
9f1b1f2d |
1856 | explicit C<defined> test or comparison elicit a warning if the |
1857 | C<use warnings> pragma or the B<-w> |
19799a22 |
1858 | command-line switch (the C<$^W> variable) is in effect. |
7b8d334a |
1859 | |
5f05dabc |
1860 | The filehandles STDIN, STDOUT, and STDERR are predefined. (The |
19799a22 |
1861 | filehandles C<stdin>, C<stdout>, and C<stderr> will also work except |
1862 | in packages, where they would be interpreted as local identifiers |
1863 | rather than global.) Additional filehandles may be created with |
1864 | the open() function, amongst others. See L<perlopentut> and |
1865 | L<perlfunc/open> for details on this. |
a0d0e21e |
1866 | |
35f2feb0 |
1867 | If a <FILEHANDLE> is used in a context that is looking for |
19799a22 |
1868 | a list, a list comprising all input lines is returned, one line per |
1869 | list element. It's easy to grow to a rather large data space this |
1870 | way, so use with care. |
a0d0e21e |
1871 | |
35f2feb0 |
1872 | <FILEHANDLE> may also be spelled C<readline(*FILEHANDLE)>. |
19799a22 |
1873 | See L<perlfunc/readline>. |
fbad3eb5 |
1874 | |
35f2feb0 |
1875 | The null filehandle <> is special: it can be used to emulate the |
1876 | behavior of B<sed> and B<awk>. Input from <> comes either from |
a0d0e21e |
1877 | standard input, or from each file listed on the command line. Here's |
35f2feb0 |
1878 | how it works: the first time <> is evaluated, the @ARGV array is |
5a964f20 |
1879 | checked, and if it is empty, C<$ARGV[0]> is set to "-", which when opened |
a0d0e21e |
1880 | gives you standard input. The @ARGV array is then processed as a list |
1881 | of filenames. The loop |
1882 | |
1883 | while (<>) { |
1884 | ... # code for each line |
1885 | } |
1886 | |
1887 | is equivalent to the following Perl-like pseudo code: |
1888 | |
3e3baf6d |
1889 | unshift(@ARGV, '-') unless @ARGV; |
a0d0e21e |
1890 | while ($ARGV = shift) { |
1891 | open(ARGV, $ARGV); |
1892 | while (<ARGV>) { |
1893 | ... # code for each line |
1894 | } |
1895 | } |
1896 | |
19799a22 |
1897 | except that it isn't so cumbersome to say, and will actually work. |
1898 | It really does shift the @ARGV array and put the current filename |
1899 | into the $ARGV variable. It also uses filehandle I<ARGV> |
35f2feb0 |
1900 | internally--<> is just a synonym for <ARGV>, which |
19799a22 |
1901 | is magical. (The pseudo code above doesn't work because it treats |
35f2feb0 |
1902 | <ARGV> as non-magical.) |
a0d0e21e |
1903 | |
35f2feb0 |
1904 | You can modify @ARGV before the first <> as long as the array ends up |
a0d0e21e |
1905 | containing the list of filenames you really want. Line numbers (C<$.>) |
19799a22 |
1906 | continue as though the input were one big happy file. See the example |
1907 | in L<perlfunc/eof> for how to reset line numbers on each file. |
5a964f20 |
1908 | |
1909 | If you want to set @ARGV to your own list of files, go right ahead. |
1910 | This sets @ARGV to all plain text files if no @ARGV was given: |
1911 | |
1912 | @ARGV = grep { -f && -T } glob('*') unless @ARGV; |
a0d0e21e |
1913 | |
5a964f20 |
1914 | You can even set them to pipe commands. For example, this automatically |
1915 | filters compressed arguments through B<gzip>: |
1916 | |
1917 | @ARGV = map { /\.(gz|Z)$/ ? "gzip -dc < $_ |" : $_ } @ARGV; |
1918 | |
1919 | If you want to pass switches into your script, you can use one of the |
a0d0e21e |
1920 | Getopts modules or put a loop on the front like this: |
1921 | |
1922 | while ($_ = $ARGV[0], /^-/) { |
1923 | shift; |
1924 | last if /^--$/; |
1925 | if (/^-D(.*)/) { $debug = $1 } |
1926 | if (/^-v/) { $verbose++ } |
5a964f20 |
1927 | # ... # other switches |
a0d0e21e |
1928 | } |
5a964f20 |
1929 | |
a0d0e21e |
1930 | while (<>) { |
5a964f20 |
1931 | # ... # code for each line |
a0d0e21e |
1932 | } |
1933 | |
35f2feb0 |
1934 | The <> symbol will return C<undef> for end-of-file only once. |
19799a22 |
1935 | If you call it again after this, it will assume you are processing another |
1936 | @ARGV list, and if you haven't set @ARGV, will read input from STDIN. |
a0d0e21e |
1937 | |
b159ebd3 |
1938 | If what the angle brackets contain is a simple scalar variable (e.g., |
35f2feb0 |
1939 | <$foo>), then that variable contains the name of the |
19799a22 |
1940 | filehandle to input from, or its typeglob, or a reference to the |
1941 | same. For example: |
cb1a09d0 |
1942 | |
1943 | $fh = \*STDIN; |
1944 | $line = <$fh>; |
a0d0e21e |
1945 | |
5a964f20 |
1946 | If what's within the angle brackets is neither a filehandle nor a simple |
1947 | scalar variable containing a filehandle name, typeglob, or typeglob |
1948 | reference, it is interpreted as a filename pattern to be globbed, and |
1949 | either a list of filenames or the next filename in the list is returned, |
19799a22 |
1950 | depending on context. This distinction is determined on syntactic |
35f2feb0 |
1951 | grounds alone. That means C<< <$x> >> is always a readline() from |
1952 | an indirect handle, but C<< <$hash{key}> >> is always a glob(). |
5a964f20 |
1953 | That's because $x is a simple scalar variable, but C<$hash{key}> is |
1954 | not--it's a hash element. |
1955 | |
1956 | One level of double-quote interpretation is done first, but you can't |
35f2feb0 |
1957 | say C<< <$foo> >> because that's an indirect filehandle as explained |
5a964f20 |
1958 | in the previous paragraph. (In older versions of Perl, programmers |
1959 | would insert curly brackets to force interpretation as a filename glob: |
35f2feb0 |
1960 | C<< <${foo}> >>. These days, it's considered cleaner to call the |
5a964f20 |
1961 | internal function directly as C<glob($foo)>, which is probably the right |
19799a22 |
1962 | way to have done it in the first place.) For example: |
a0d0e21e |
1963 | |
1964 | while (<*.c>) { |
1965 | chmod 0644, $_; |
1966 | } |
1967 | |
3a4b19e4 |
1968 | is roughly equivalent to: |
a0d0e21e |
1969 | |
1970 | open(FOO, "echo *.c | tr -s ' \t\r\f' '\\012\\012\\012\\012'|"); |
1971 | while (<FOO>) { |
5b3eff12 |
1972 | chomp; |
a0d0e21e |
1973 | chmod 0644, $_; |
1974 | } |
1975 | |
3a4b19e4 |
1976 | except that the globbing is actually done internally using the standard |
1977 | C<File::Glob> extension. Of course, the shortest way to do the above is: |
a0d0e21e |
1978 | |
1979 | chmod 0644, <*.c>; |
1980 | |
19799a22 |
1981 | A (file)glob evaluates its (embedded) argument only when it is |
1982 | starting a new list. All values must be read before it will start |
1983 | over. In list context, this isn't important because you automatically |
1984 | get them all anyway. However, in scalar context the operator returns |
069e01df |
1985 | the next value each time it's called, or C<undef> when the list has |
19799a22 |
1986 | run out. As with filehandle reads, an automatic C<defined> is |
1987 | generated when the glob occurs in the test part of a C<while>, |
1988 | because legal glob returns (e.g. a file called F<0>) would otherwise |
1989 | terminate the loop. Again, C<undef> is returned only once. So if |
1990 | you're expecting a single value from a glob, it is much better to |
1991 | say |
4633a7c4 |
1992 | |
1993 | ($file) = <blurch*>; |
1994 | |
1995 | than |
1996 | |
1997 | $file = <blurch*>; |
1998 | |
1999 | because the latter will alternate between returning a filename and |
19799a22 |
2000 | returning false. |
4633a7c4 |
2001 | |
b159ebd3 |
2002 | If you're trying to do variable interpolation, it's definitely better |
4633a7c4 |
2003 | to use the glob() function, because the older notation can cause people |
e37d713d |
2004 | to become confused with the indirect filehandle notation. |
4633a7c4 |
2005 | |
2006 | @files = glob("$dir/*.[ch]"); |
2007 | @files = glob($files[$i]); |
2008 | |
a0d0e21e |
2009 | =head2 Constant Folding |
2010 | |
2011 | Like C, Perl does a certain amount of expression evaluation at |
19799a22 |
2012 | compile time whenever it determines that all arguments to an |
a0d0e21e |
2013 | operator are static and have no side effects. In particular, string |
2014 | concatenation happens at compile time between literals that don't do |
19799a22 |
2015 | variable substitution. Backslash interpolation also happens at |
a0d0e21e |
2016 | compile time. You can say |
2017 | |
2018 | 'Now is the time for all' . "\n" . |
2019 | 'good men to come to.' |
2020 | |
54310121 |
2021 | and this all reduces to one string internally. Likewise, if |
a0d0e21e |
2022 | you say |
2023 | |
2024 | foreach $file (@filenames) { |
5a964f20 |
2025 | if (-s $file > 5 + 100 * 2**16) { } |
54310121 |
2026 | } |
a0d0e21e |
2027 | |
19799a22 |
2028 | the compiler will precompute the number which that expression |
2029 | represents so that the interpreter won't have to. |
a0d0e21e |
2030 | |
2c268ad5 |
2031 | =head2 Bitwise String Operators |
2032 | |
2033 | Bitstrings of any size may be manipulated by the bitwise operators |
2034 | (C<~ | & ^>). |
2035 | |
19799a22 |
2036 | If the operands to a binary bitwise op are strings of different |
2037 | sizes, B<|> and B<^> ops act as though the shorter operand had |
2038 | additional zero bits on the right, while the B<&> op acts as though |
2039 | the longer operand were truncated to the length of the shorter. |
2040 | The granularity for such extension or truncation is one or more |
2041 | bytes. |
2c268ad5 |
2042 | |
2043 | # ASCII-based examples |
2044 | print "j p \n" ^ " a h"; # prints "JAPH\n" |
2045 | print "JA" | " ph\n"; # prints "japh\n" |
2046 | print "japh\nJunk" & '_____'; # prints "JAPH\n"; |
2047 | print 'p N$' ^ " E<H\n"; # prints "Perl\n"; |
2048 | |
19799a22 |
2049 | If you are intending to manipulate bitstrings, be certain that |
2c268ad5 |
2050 | you're supplying bitstrings: If an operand is a number, that will imply |
19799a22 |
2051 | a B<numeric> bitwise operation. You may explicitly show which type of |
2c268ad5 |
2052 | operation you intend by using C<""> or C<0+>, as in the examples below. |
2053 | |
2054 | $foo = 150 | 105 ; # yields 255 (0x96 | 0x69 is 0xFF) |
2055 | $foo = '150' | 105 ; # yields 255 |
2056 | $foo = 150 | '105'; # yields 255 |
2057 | $foo = '150' | '105'; # yields string '155' (under ASCII) |
2058 | |
2059 | $baz = 0+$foo & 0+$bar; # both ops explicitly numeric |
2060 | $biz = "$foo" ^ "$bar"; # both ops explicitly stringy |
a0d0e21e |
2061 | |
1ae175c8 |
2062 | See L<perlfunc/vec> for information on how to manipulate individual bits |
2063 | in a bit vector. |
2064 | |
55497cff |
2065 | =head2 Integer Arithmetic |
a0d0e21e |
2066 | |
19799a22 |
2067 | By default, Perl assumes that it must do most of its arithmetic in |
a0d0e21e |
2068 | floating point. But by saying |
2069 | |
2070 | use integer; |
2071 | |
2072 | you may tell the compiler that it's okay to use integer operations |
19799a22 |
2073 | (if it feels like it) from here to the end of the enclosing BLOCK. |
2074 | An inner BLOCK may countermand this by saying |
a0d0e21e |
2075 | |
2076 | no integer; |
2077 | |
19799a22 |
2078 | which lasts until the end of that BLOCK. Note that this doesn't |
2079 | mean everything is only an integer, merely that Perl may use integer |
2080 | operations if it is so inclined. For example, even under C<use |
2081 | integer>, if you take the C<sqrt(2)>, you'll still get C<1.4142135623731> |
2082 | or so. |
2083 | |
2084 | Used on numbers, the bitwise operators ("&", "|", "^", "~", "<<", |
13a2d996 |
2085 | and ">>") always produce integral results. (But see also |
2086 | L<Bitwise String Operators>.) However, C<use integer> still has meaning for |
19799a22 |
2087 | them. By default, their results are interpreted as unsigned integers, but |
2088 | if C<use integer> is in effect, their results are interpreted |
2089 | as signed integers. For example, C<~0> usually evaluates to a large |
2090 | integral value. However, C<use integer; ~0> is C<-1> on twos-complement |
2091 | machines. |
68dc0745 |
2092 | |
2093 | =head2 Floating-point Arithmetic |
2094 | |
2095 | While C<use integer> provides integer-only arithmetic, there is no |
19799a22 |
2096 | analogous mechanism to provide automatic rounding or truncation to a |
2097 | certain number of decimal places. For rounding to a certain number |
2098 | of digits, sprintf() or printf() is usually the easiest route. |
2099 | See L<perlfaq4>. |
68dc0745 |
2100 | |
5a964f20 |
2101 | Floating-point numbers are only approximations to what a mathematician |
2102 | would call real numbers. There are infinitely more reals than floats, |
2103 | so some corners must be cut. For example: |
2104 | |
2105 | printf "%.20g\n", 123456789123456789; |
2106 | # produces 123456789123456784 |
2107 | |
2108 | Testing for exact equality of floating-point equality or inequality is |
2109 | not a good idea. Here's a (relatively expensive) work-around to compare |
2110 | whether two floating-point numbers are equal to a particular number of |
2111 | decimal places. See Knuth, volume II, for a more robust treatment of |
2112 | this topic. |
2113 | |
2114 | sub fp_equal { |
2115 | my ($X, $Y, $POINTS) = @_; |
2116 | my ($tX, $tY); |
2117 | $tX = sprintf("%.${POINTS}g", $X); |
2118 | $tY = sprintf("%.${POINTS}g", $Y); |
2119 | return $tX eq $tY; |
2120 | } |
2121 | |
68dc0745 |
2122 | The POSIX module (part of the standard perl distribution) implements |
19799a22 |
2123 | ceil(), floor(), and other mathematical and trigonometric functions. |
2124 | The Math::Complex module (part of the standard perl distribution) |
2125 | defines mathematical functions that work on both the reals and the |
2126 | imaginary numbers. Math::Complex not as efficient as POSIX, but |
68dc0745 |
2127 | POSIX can't work with complex numbers. |
2128 | |
2129 | Rounding in financial applications can have serious implications, and |
2130 | the rounding method used should be specified precisely. In these |
2131 | cases, it probably pays not to trust whichever system rounding is |
2132 | being used by Perl, but to instead implement the rounding function you |
2133 | need yourself. |
5a964f20 |
2134 | |
2135 | =head2 Bigger Numbers |
2136 | |
2137 | The standard Math::BigInt and Math::BigFloat modules provide |
19799a22 |
2138 | variable-precision arithmetic and overloaded operators, although |
cd5c4fce |
2139 | they're currently pretty slow. At the cost of some space and |
19799a22 |
2140 | considerable speed, they avoid the normal pitfalls associated with |
2141 | limited-precision representations. |
5a964f20 |
2142 | |
2143 | use Math::BigInt; |
2144 | $x = Math::BigInt->new('123456789123456789'); |
2145 | print $x * $x; |
2146 | |
2147 | # prints +15241578780673678515622620750190521 |
19799a22 |
2148 | |
cd5c4fce |
2149 | There are several modules that let you calculate with (bound only by |
2150 | memory and cpu-time) unlimited or fixed precision. There are also |
2151 | some non-standard modules that provide faster implementations via |
2152 | external C libraries. |
2153 | |
2154 | Here is a short, but incomplete summary: |
2155 | |
2156 | Math::Fraction big, unlimited fractions like 9973 / 12967 |
2157 | Math::String treat string sequences like numbers |
2158 | Math::FixedPrecision calculate with a fixed precision |
2159 | Math::Currency for currency calculations |
2160 | Bit::Vector manipulate bit vectors fast (uses C) |
2161 | Math::BigIntFast Bit::Vector wrapper for big numbers |
2162 | Math::Pari provides access to the Pari C library |
2163 | Math::BigInteger uses an external C library |
2164 | Math::Cephes uses external Cephes C library (no big numbers) |
2165 | Math::Cephes::Fraction fractions via the Cephes library |
2166 | Math::GMP another one using an external C library |
2167 | |
2168 | Choose wisely. |
16070b82 |
2169 | |
2170 | =cut |