1ZSHEXPN(1) General Commands Manual ZSHEXPN(1)
2
6 zshexpn - zsh expansion and substitution
7
9 The following types of expansions are performed in the indicated order
10 in five steps:
11 History Expansion
13 This is performed only in interactive shells.
14 Alias Expansion
16 Aliases are expanded immediately before the command line is
17 parsed as explained under Aliasing in zshmisc(1).
18 Process Substitution
20 Parameter Expansion
21 Command Substitution
22 Arithmetic Expansion
23 Brace Expansion
24 These five are performed in left-to-right fashion. On each
25 argument, any of the five steps that are needed are performed
26 one after the other. Hence, for example, all the parts of
27 parameter expansion are completed before command substitution is
28 started. After these expansions, all unquoted occurrences of
29 the characters `\',`'' and `"' are removed.
30 Filename Expansion
32 If the SH_FILE_EXPANSION option is set, the order of expansion
33 is modified for compatibility with sh and ksh. In that case
34 filename expansion is performed immediately after alias expan‐
35 sion, preceding the set of five expansions mentioned above.
36 Filename Generation
38 This expansion, commonly referred to as globbing, is always done
39 last.
40 The following sections explain the types of expansion in detail.
42
44 History expansion allows you to use words from previous command lines
45 in the command line you are typing. This simplifies spelling correc‐
46 tions and the repetition of complicated commands or arguments.
47 Immediately before execution, each command is saved in the history
49 list, the size of which is controlled by the HISTSIZE parameter. The
50 one most recent command is always retained in any case. Each saved
51 command in the history list is called a history event and is assigned a
52 number, beginning with 1 (one) when the shell starts up. The history
53 number that you may see in your prompt (see EXPANSION OF PROMPT
54 SEQUENCES in zshmisc(1)) is the number that is to be assigned to the
55 next command.
56 Overview
58 A history expansion begins with the first character of the histchars
59 parameter, which is `!' by default, and may occur anywhere on the com‐
60 mand line, including inside double quotes (but not inside single quotes
61 '...' or C-style quotes $'...' nor when escaped with a backslash).
62 The first character is followed by an optional event designator (see
64 the section `Event Designators') and then an optional word designator
65 (the section `Word Designators'); if neither of these designators is
66 present, no history expansion occurs.
67 Input lines containing history expansions are echoed after being
69 expanded, but before any other expansions take place and before the
70 command is executed. It is this expanded form that is recorded as the
71 history event for later references.
72 History expansions do not nest.
74 By default, a history reference with no event designator refers to the
76 same event as any preceding history reference on that command line; if
77 it is the only history reference in a command, it refers to the previ‐
78 ous command. However, if the option CSH_JUNKIE_HISTORY is set, then
79 every history reference with no event specification always refers to
80 the previous command.
81 For example, `!' is the event designator for the previous command, so
83 `!!:1' always refers to the first word of the previous command, and
84 `!!$' always refers to the last word of the previous command. With
85 CSH_JUNKIE_HISTORY set, then `!:1' and `!$' function in the same manner
86 as `!!:1' and `!!$', respectively. Conversely, if CSH_JUNKIE_HISTORY
87 is unset, then `!:1' and `!$' refer to the first and last words,
88 respectively, of the same event referenced by the nearest other history
89 reference preceding them on the current command line, or to the previ‐
90 ous command if there is no preceding reference.
91 The character sequence `^foo^bar' (where `^' is actually the second
93 character of the histchars parameter) repeats the last command, replac‐
94 ing the string foo with bar. More precisely, the sequence `^foo^bar^'
95 is synonymous with `!!:s^foo^bar^', hence other modifiers (see the sec‐
96 tion `Modifiers') may follow the final `^'. In particular,
97 `^foo^bar^:G' performs a global substitution.
98 If the shell encounters the character sequence `!"' in the input, the
100 history mechanism is temporarily disabled until the current list (see
101 zshmisc(1)) is fully parsed. The `!"' is removed from the input, and
102 any subsequent `!' characters have no special significance.
103 A less convenient but more comprehensible form of command history sup‐
105 port is provided by the fc builtin.
106 Event Designators
108 An event designator is a reference to a command-line entry in the his‐
109 tory list. In the list below, remember that the initial `!' in each
110 item may be changed to another character by setting the histchars
111 parameter.
112 ! Start a history expansion, except when followed by a blank, new‐
114 line, `=' or `('. If followed immediately by a word designator
115 (see the section `Word Designators'), this forms a history ref‐
116 erence with no event designator (see the section `Overview').
117 !! Refer to the previous command. By itself, this expansion
119 repeats the previous command.
120 !n Refer to command-line n.
122 !-n Refer to the current command-line minus n.
124 !str Refer to the most recent command starting with str.
126 !?str[?]
128 Refer to the most recent command containing str. The trailing
129 `?' is necessary if this reference is to be followed by a modi‐
130 fier or followed by any text that is not to be considered part
131 of str.
132 !# Refer to the current command line typed in so far. The line is
134 treated as if it were complete up to and including the word
135 before the one with the `!#' reference.
136 !{...} Insulate a history reference from adjacent characters (if neces‐
138 sary).
139 Word Designators
141 A word designator indicates which word or words of a given command line
142 are to be included in a history reference. A `:' usually separates the
143 event specification from the word designator. It may be omitted only
144 if the word designator begins with a `^', `$', `*', `-' or `%'. Word
145 designators include:
146 0 The first input word (command).
148 n The nth argument.
149 ^ The first argument. That is, 1.
150 $ The last argument.
151 % The word matched by (the most recent) ?str search.
152 x-y A range of words; x defaults to 0.
153 * All the arguments, or a null value if there are none.
154 x* Abbreviates `x-$'.
155 x- Like `x*' but omitting word $.
156 Note that a `%' word designator works only when used in one of `!%',
158 `!:%' or `!?str?:%', and only when used after a !? expansion (possibly
159 in an earlier command). Anything else results in an error, although
160 the error may not be the most obvious one.
161 Modifiers
163 After the optional word designator, you can add a sequence of one or
164 more of the following modifiers, each preceded by a `:'. These modi‐
165 fiers also work on the result of filename generation and parameter
166 expansion, except where noted.
167 a Turn a file name into an absolute path: prepends the current
169 directory, if necessary; remove `.' path segments; and remove
170 `..' path segments and the segments that immediately precede
171 them.
172 This transformation is agnostic about what is in the filesystem,
174 i.e. is on the logical, not the physical directory. It takes
175 place in the same manner as when changing directories when nei‐
176 ther of the options CHASE_DOTS or CHASE_LINKS is set. For exam‐
177 ple, `/before/here/../after' is always transformed to
178 `/before/after', regardless of whether `/before/here' exists or
179 what kind of object (dir, file, symlink, etc.) it is.
180 A Turn a file name into an absolute path as the `a' modifier does,
182 and then pass the result through the realpath(3) library func‐
183 tion to resolve symbolic links.
184 Note: on systems that do not have a realpath(3) library func‐
186 tion, symbolic links are not resolved, so on those systems `a'
187 and `A' are equivalent.
188 Note: foo:A and realpath(foo) are different on some inputs. For
190 realpath(foo) semantics, see the `P` modifier.
191 c Resolve a command name into an absolute path by searching the
193 command path given by the PATH variable. This does not work for
194 commands containing directory parts. Note also that this does
195 not usually work as a glob qualifier unless a file of the same
196 name is found in the current directory.
197 e Remove all but the part of the filename extension following the
199 `.'; see the definition of the filename extension in the
200 description of the r modifier below. Note that according to
201 that definition the result will be empty if the string ends with
202 a `.'.
203 h [ digits ]
205 Remove a trailing pathname component, shortening the path by one
206 directory level: this is the `head' of the pathname. This works
207 like `dirname'. If the h is followed immediately (with no spa‐
208 ces or other separator) by any number of decimal digits, and the
209 value of the resulting number is non-zero, that number of lead‐
210 ing components is preserved instead of the final component being
211 removed. In an absolute path the leading `/' is the first com‐
212 ponent, so, for example, if var=/my/path/to/something, then
213 ${var:h3} substitutes /my/path. Consecutive `/'s are treated
214 the same as a single `/'. In parameter substitution, digits may
215 only be used if the expression is in braces, so for example the
216 short form substitution $var:h2 is treated as ${var:h}2, not as
217 ${var:h2}. No restriction applies to the use of digits in his‐
218 tory substitution or globbing qualifiers. If more components
219 are requested than are present, the entire path is substituted
220 (so this does not trigger a `failed modifier' error in history
221 expansion).
222 l Convert the words to all lowercase.
224 p Print the new command but do not execute it. Only works with
226 history expansion.
227 P Turn a file name into an absolute path, like realpath(3). The
229 resulting path will be absolute, have neither `.' nor `..' com‐
230 ponents, and refer to the same directory entry as the input
231 filename.
232 Unlike realpath(3), non-existent trailing components are permit‐
234 ted and preserved.
235 q Quote the substituted words, escaping further substitutions.
237 Works with history expansion and parameter expansion, though for
238 parameters it is only useful if the resulting text is to be
239 re-evaluated such as by eval.
240 Q Remove one level of quotes from the substituted words.
242 r Remove a filename extension leaving the root name. Strings with
244 no filename extension are not altered. A filename extension is
245 a `.' followed by any number of characters (including zero) that
246 are neither `.' nor `/' and that continue to the end of the
247 string. For example, the extension of `foo.orig.c' is `.c', and
248 `dir.c/foo' has no extension.
249 s/l/r[/]
251 Substitute r for l as described below. The substitution is done
252 only for the first string that matches l. For arrays and for
253 filename generation, this applies to each word of the expanded
254 text. See below for further notes on substitutions.
255 The forms `gs/l/r' and `s/l/r/:G' perform global substitution,
257 i.e. substitute every occurrence of r for l. Note that the g or
258 :G must appear in exactly the position shown.
259 See further notes on this form of substitution below.
261 & Repeat the previous s substitution. Like s, may be preceded
263 immediately by a g. In parameter expansion the & must appear
264 inside braces, and in filename generation it must be quoted with
265 a backslash.
266 t [ digits ]
268 Remove all leading pathname components, leaving the final compo‐
269 nent (tail). This works like `basename'. Any trailing slashes
270 are first removed. Decimal digits are handled as described
271 above for (h), but in this case that number of trailing compo‐
272 nents is preserved instead of the default 1; 0 is treated the
273 same as 1.
274 u Convert the words to all uppercase.
276 x Like q, but break into words at whitespace. Does not work with
278 parameter expansion.
279 The s/l/r/ substitution works as follows. By default the left-hand
281 side of substitutions are not patterns, but character strings. Any
282 character can be used as the delimiter in place of `/'. A backslash
283 quotes the delimiter character. The character `&', in the
284 right-hand-side r, is replaced by the text from the left-hand-side l.
285 The `&' can be quoted with a backslash. A null l uses the previous
286 string either from the previous l or from the contextual scan string s
287 from `!?s'. You can omit the rightmost delimiter if a newline immedi‐
288 ately follows r; the rightmost `?' in a context scan can similarly be
289 omitted. Note the same record of the last l and r is maintained across
290 all forms of expansion.
291 Note that if a `&' is used within glob qualifiers an extra backslash is
293 needed as a & is a special character in this case.
294 Also note that the order of expansions affects the interpretation of l
296 and r. When used in a history expansion, which occurs before any other
297 expansions, l and r are treated as literal strings (except as explained
298 for HIST_SUBST_PATTERN below). When used in parameter expansion, the
299 replacement of r into the parameter's value is done first, and then any
300 additional process, parameter, command, arithmetic, or brace references
301 are applied, which may evaluate those substitutions and expansions more
302 than once if l appears more than once in the starting value. When used
303 in a glob qualifier, any substitutions or expansions are performed once
304 at the time the qualifier is parsed, even before the `:s' expression
305 itself is divided into l and r sides.
306 If the option HIST_SUBST_PATTERN is set, l is treated as a pattern of
308 the usual form described in the section FILENAME GENERATION below.
309 This can be used in all the places where modifiers are available; note,
310 however, that in globbing qualifiers parameter substitution has already
311 taken place, so parameters in the replacement string should be quoted
312 to ensure they are replaced at the correct time. Note also that com‐
313 plicated patterns used in globbing qualifiers may need the extended
314 glob qualifier notation (#q:s/.../.../) in order for the shell to rec‐
315 ognize the expression as a glob qualifier. Further, note that bad pat‐
316 terns in the substitution are not subject to the NO_BAD_PATTERN option
317 so will cause an error.
318 When HIST_SUBST_PATTERN is set, l may start with a # to indicate that
320 the pattern must match at the start of the string to be substituted,
321 and a % may appear at the start or after an # to indicate that the pat‐
322 tern must match at the end of the string to be substituted. The % or #
323 may be quoted with two backslashes.
324 For example, the following piece of filename generation code with the
326 EXTENDED_GLOB option:
327 print -r -- *.c(#q:s/#%(#b)s(*).c/'S${match[1]}.C'/)
329 takes the expansion of *.c and applies the glob qualifiers in the
331 (#q...) expression, which consists of a substitution modifier anchored
332 to the start and end of each word (#%). This turns on backreferences
333 ((#b)), so that the parenthesised subexpression is available in the
334 replacement string as ${match[1]}. The replacement string is quoted so
335 that the parameter is not substituted before the start of filename gen‐
336 eration.
337 The following f, F, w and W modifiers work only with parameter expan‐
339 sion and filename generation. They are listed here to provide a single
340 point of reference for all modifiers.
341 f Repeats the immediately (without a colon) following modifier
343 until the resulting word doesn't change any more.
344 F:expr:
346 Like f, but repeats only n times if the expression expr evalu‐
347 ates to n. Any character can be used instead of the `:'; if
348 `(', `[', or `{' is used as the opening delimiter, the closing
349 delimiter should be ')', `]', or `}', respectively.
350 w Makes the immediately following modifier work on each word in
352 the string.
353 W:sep: Like w but words are considered to be the parts of the string
355 that are separated by sep. Any character can be used instead of
356 the `:'; opening parentheses are handled specially, see above.
357
359 Each part of a command argument that takes the form `<(list)',
360 `>(list)' or `=(list)' is subject to process substitution. The expres‐
361 sion may be preceded or followed by other strings except that, to pre‐
362 vent clashes with commonly occurring strings and patterns, the last
363 form must occur at the start of a command argument, and the forms are
364 only expanded when first parsing command or assignment arguments.
365 Process substitutions may be used following redirection operators; in
366 this case, the substitution must appear with no trailing string.
367 Note that `<<(list)' is not a special syntax; it is equivalent to `<
369 <(list)', redirecting standard input from the result of process substi‐
370 tution. Hence all the following documentation applies. The second
371 form (with the space) is recommended for clarity.
372 In the case of the < or > forms, the shell runs the commands in list as
374 a subprocess of the job executing the shell command line. If the sys‐
375 tem supports the /dev/fd mechanism, the command argument is the name of
376 the device file corresponding to a file descriptor; otherwise, if the
377 system supports named pipes (FIFOs), the command argument will be a
378 named pipe. If the form with > is selected then writing on this spe‐
379 cial file will provide input for list. If < is used, then the file
380 passed as an argument will be connected to the output of the list
381 process. For example,
382 paste <(cut -f1 file1) <(cut -f3 file2) |
384 tee >(process1) >(process2) >/dev/null
385 cuts fields 1 and 3 from the files file1 and file2 respectively, pastes
387 the results together, and sends it to the processes process1 and
388 process2.
389 If =(...) is used instead of <(...), then the file passed as an argu‐
391 ment will be the name of a temporary file containing the output of the
392 list process. This may be used instead of the < form for a program
393 that expects to lseek (see lseek(2)) on the input file.
394 There is an optimisation for substitutions of the form =(<<<arg), where
396 arg is a single-word argument to the here-string redirection <<<. This
397 form produces a file name containing the value of arg after any substi‐
398 tutions have been performed. This is handled entirely within the cur‐
399 rent shell. This is effectively the reverse of the special form
400 $(<arg) which treats arg as a file name and replaces it with the file's
401 contents.
402 The = form is useful as both the /dev/fd and the named pipe implementa‐
404 tion of <(...) have drawbacks. In the former case, some programmes may
405 automatically close the file descriptor in question before examining
406 the file on the command line, particularly if this is necessary for
407 security reasons such as when the programme is running setuid. In the
408 second case, if the programme does not actually open the file, the sub‐
409 shell attempting to read from or write to the pipe will (in a typical
410 implementation, different operating systems may have different behav‐
411 iour) block for ever and have to be killed explicitly. In both cases,
412 the shell actually supplies the information using a pipe, so that pro‐
413 grammes that expect to lseek (see lseek(2)) on the file will not work.
414 Also note that the previous example can be more compactly and effi‐
416 ciently written (provided the MULTIOS option is set) as:
417 paste <(cut -f1 file1) <(cut -f3 file2) \
419 > >(process1) > >(process2)
420 The shell uses pipes instead of FIFOs to implement the latter two
422 process substitutions in the above example.
423 There is an additional problem with >(process); when this is attached
425 to an external command, the parent shell does not wait for process to
426 finish and hence an immediately following command cannot rely on the
427 results being complete. The problem and solution are the same as
428 described in the section MULTIOS in zshmisc(1). Hence in a simplified
429 version of the example above:
430 paste <(cut -f1 file1) <(cut -f3 file2) > >(process)
432 (note that no MULTIOS are involved), process will be run asynchronously
434 as far as the parent shell is concerned. The workaround is:
435 { paste <(cut -f1 file1) <(cut -f3 file2) } > >(process)
437 The extra processes here are spawned from the parent shell which will
439 wait for their completion.
440 Another problem arises any time a job with a substitution that requires
442 a temporary file is disowned by the shell, including the case where
443 `&!' or `&|' appears at the end of a command containing a substitution.
444 In that case the temporary file will not be cleaned up as the shell no
445 longer has any memory of the job. A workaround is to use a subshell,
446 for example,
447 (mycmd =(myoutput)) &!
449 as the forked subshell will wait for the command to finish then remove
451 the temporary file.
452 A general workaround to ensure a process substitution endures for an
454 appropriate length of time is to pass it as a parameter to an anonymous
455 shell function (a piece of shell code that is run immediately with
456 function scope). For example, this code:
457 () {
459 print File $1:
460 cat $1
461 } =(print This be the verse)
462 outputs something resembling the following
464 File /tmp/zsh6nU0kS:
466 This be the verse
467 The temporary file created by the process substitution will be deleted
469 when the function exits.
470
472 The character `$' is used to introduce parameter expansions. See zsh‐
473 param(1) for a description of parameters, including arrays, associative
474 arrays, and subscript notation to access individual array elements.
475 Note in particular the fact that words of unquoted parameters are not
477 automatically split on whitespace unless the option SH_WORD_SPLIT is
478 set; see references to this option below for more details. This is an
479 important difference from other shells. However, as in other shells,
480 null words are elided from unquoted parameters' expansions.
481 With default options, after the assignments:
483 array=("first word" "" "third word")
485 scalar="only word"
486 then $array substitutes two words, `first word' and `third word', and
488 $scalar substitutes a single word `only word'. Note that second ele‐
489 ment of array was elided. Scalar parameters can be elided too if their
490 value is null (empty). To avoid elision, use quoting as follows:
491 "$scalar" for scalars and "${array[@]}" or "${(@)array}" for arrays.
492 (The last two forms are equivalent.)
493 Parameter expansions can involve flags, as in `${(@kv)aliases}', and
495 other operators, such as `${PREFIX:-"/usr/local"}'. Parameter expan‐
496 sions can also be nested. These topics will be introduced below. The
497 full rules are complicated and are noted at the end.
498 In the expansions discussed below that require a pattern, the form of
500 the pattern is the same as that used for filename generation; see the
501 section `Filename Generation'. Note that these patterns, along with
502 the replacement text of any substitutions, are themselves subject to
503 parameter expansion, command substitution, and arithmetic expansion.
504 In addition to the following operations, the colon modifiers described
505 in the section `Modifiers' in the section `History Expansion' can be
506 applied: for example, ${i:s/foo/bar/} performs string substitution on
507 the expansion of parameter $i.
508 In the following descriptions, `word' refers to a single word substi‐
510 tuted on the command line, not necessarily a space delimited word.
511 ${name}
513 The value, if any, of the parameter name is substituted. The
514 braces are required if the expansion is to be followed by a let‐
515 ter, digit, or underscore that is not to be interpreted as part
516 of name. In addition, more complicated forms of substitution
517 usually require the braces to be present; exceptions, which only
518 apply if the option KSH_ARRAYS is not set, are a single sub‐
519 script or any colon modifiers appearing after the name, or any
520 of the characters `^', `=', `~', `#' or `+' appearing before the
521 name, all of which work with or without braces.
522 If name is an array parameter, and the KSH_ARRAYS option is not
524 set, then the value of each element of name is substituted, one
525 element per word. Otherwise, the expansion results in one word
526 only; with KSH_ARRAYS, this is the first element of an array.
527 No field splitting is done on the result unless the
528 SH_WORD_SPLIT option is set. See also the flags = and
529 s:string:.
530 ${+name}
532 If name is the name of a set parameter `1' is substituted, oth‐
533 erwise `0' is substituted.
534 ${name-word}
536 ${name:-word}
537 If name is set, or in the second form is non-null, then substi‐
538 tute its value; otherwise substitute word. In the second form
539 name may be omitted, in which case word is always substituted.
540 ${name+word}
542 ${name:+word}
543 If name is set, or in the second form is non-null, then substi‐
544 tute word; otherwise substitute nothing.
545 ${name=word}
547 ${name:=word}
548 ${name::=word}
549 In the first form, if name is unset then set it to word; in the
550 second form, if name is unset or null then set it to word; and
551 in the third form, unconditionally set name to word. In all
552 forms, the value of the parameter is then substituted.
553 ${name?word}
555 ${name:?word}
556 In the first form, if name is set, or in the second form if name
557 is both set and non-null, then substitute its value; otherwise,
558 print word and exit from the shell. Interactive shells instead
559 return to the prompt. If word is omitted, then a standard mes‐
560 sage is printed.
561 In any of the above expressions that test a variable and substitute an
563 alternate word, note that you can use standard shell quoting in the
564 word value to selectively override the splitting done by the
565 SH_WORD_SPLIT option and the = flag, but not splitting by the s:string:
566 flag.
567 In the following expressions, when name is an array and the substitu‐
569 tion is not quoted, or if the `(@)' flag or the name[@] syntax is used,
570 matching and replacement is performed on each array element separately.
571 ${name#pattern}
573 ${name##pattern}
574 If the pattern matches the beginning of the value of name, then
575 substitute the value of name with the matched portion deleted;
576 otherwise, just substitute the value of name. In the first
577 form, the smallest matching pattern is preferred; in the second
578 form, the largest matching pattern is preferred.
579 ${name%pattern}
581 ${name%%pattern}
582 If the pattern matches the end of the value of name, then sub‐
583 stitute the value of name with the matched portion deleted; oth‐
584 erwise, just substitute the value of name. In the first form,
585 the smallest matching pattern is preferred; in the second form,
586 the largest matching pattern is preferred.
587 ${name:#pattern}
589 If the pattern matches the value of name, then substitute the
590 empty string; otherwise, just substitute the value of name. If
591 name is an array the matching array elements are removed (use
592 the `(M)' flag to remove the non-matched elements).
593 ${name:|arrayname}
595 If arrayname is the name (N.B., not contents) of an array vari‐
596 able, then any elements contained in arrayname are removed from
597 the substitution of name. If the substitution is scalar, either
598 because name is a scalar variable or the expression is quoted,
599 the elements of arrayname are instead tested against the entire
600 expression.
601 ${name:*arrayname}
603 Similar to the preceding substitution, but in the opposite
604 sense, so that entries present in both the original substitution
605 and as elements of arrayname are retained and others removed.
606 ${name:^arrayname}
608 ${name:^^arrayname}
609 Zips two arrays, such that the output array is twice as long as
610 the shortest (longest for `:^^') of name and arrayname, with the
611 elements alternatingly being picked from them. For `:^', if one
612 of the input arrays is longer, the output will stop when the end
613 of the shorter array is reached. Thus,
614 a=(1 2 3 4); b=(a b); print ${a:^b}
616 will output `1 a 2 b'. For `:^^', then the input is repeated
618 until all of the longer array has been used up and the above
619 will output `1 a 2 b 3 a 4 b'.
620 Either or both inputs may be a scalar, they will be treated as
622 an array of length 1 with the scalar as the only element. If
623 either array is empty, the other array is output with no extra
624 elements inserted.
625 Currently the following code will output `a b' and `1' as two
627 separate elements, which can be unexpected. The second print
628 provides a workaround which should continue to work if this is
629 changed.
630 a=(a b); b=(1 2); print -l "${a:^b}"; print -l "${${a:^b}}"
632 ${name:offset}
634 ${name:offset:length}
635 This syntax gives effects similar to parameter subscripting in
636 the form $name[start,end], but is compatible with other shells;
637 note that both offset and length are interpreted differently
638 from the components of a subscript.
639 If offset is non-negative, then if the variable name is a scalar
641 substitute the contents starting offset characters from the
642 first character of the string, and if name is an array substi‐
643 tute elements starting offset elements from the first element.
644 If length is given, substitute that many characters or elements,
645 otherwise the entire rest of the scalar or array.
646 A positive offset is always treated as the offset of a character
648 or element in name from the first character or element of the
649 array (this is different from native zsh subscript notation).
650 Hence 0 refers to the first character or element regardless of
651 the setting of the option KSH_ARRAYS.
652 A negative offset counts backwards from the end of the scalar or
654 array, so that -1 corresponds to the last character or element,
655 and so on.
656 When positive, length counts from the offset position toward the
658 end of the scalar or array. When negative, length counts back
659 from the end. If this results in a position smaller than off‐
660 set, a diagnostic is printed and nothing is substituted.
661 The option MULTIBYTE is obeyed, i.e. the offset and length count
663 multibyte characters where appropriate.
664 offset and length undergo the same set of shell substitutions as
666 for scalar assignment; in addition, they are then subject to
667 arithmetic evaluation. Hence, for example
668 print ${foo:3}
670 print ${foo: 1 + 2}
671 print ${foo:$(( 1 + 2))}
672 print ${foo:$(echo 1 + 2)}
673 all have the same effect, extracting the string starting at the
675 fourth character of $foo if the substitution would otherwise
676 return a scalar, or the array starting at the fourth element if
677 $foo would return an array. Note that with the option
678 KSH_ARRAYS $foo always returns a scalar (regardless of the use
679 of the offset syntax) and a form such as ${foo[*]:3} is required
680 to extract elements of an array named foo.
681 If offset is negative, the - may not appear immediately after
683 the : as this indicates the ${name:-word} form of substitution.
684 Instead, a space may be inserted before the -. Furthermore,
685 neither offset nor length may begin with an alphabetic character
686 or & as these are used to indicate history-style modifiers. To
687 substitute a value from a variable, the recommended approach is
688 to precede it with a $ as this signifies the intention (parame‐
689 ter substitution can easily be rendered unreadable); however, as
690 arithmetic substitution is performed, the expression ${var:
691 offs} does work, retrieving the offset from $offs.
692 For further compatibility with other shells there is a special
694 case for array offset 0. This usually accesses the first ele‐
695 ment of the array. However, if the substitution refers to the
696 positional parameter array, e.g. $@ or $*, then offset 0 instead
697 refers to $0, offset 1 refers to $1, and so on. In other words,
698 the positional parameter array is effectively extended by
699 prepending $0. Hence ${*:0:1} substitutes $0 and ${*:1:1} sub‐
700 stitutes $1.
701 ${name/pattern/repl}
703 ${name//pattern/repl}
704 ${name:/pattern/repl}
705 Replace the longest possible match of pattern in the expansion
706 of parameter name by string repl. The first form replaces just
707 the first occurrence, the second form all occurrences, and the
708 third form replaces only if pattern matches the entire string.
709 Both pattern and repl are subject to double-quoted substitution,
710 so that expressions like ${name/$opat/$npat} will work, but obey
711 the usual rule that pattern characters in $opat are not treated
712 specially unless either the option GLOB_SUBST is set, or $opat
713 is instead substituted as ${~opat}.
714 The pattern may begin with a `#', in which case the pattern must
716 match at the start of the string, or `%', in which case it must
717 match at the end of the string, or `#%' in which case the pat‐
718 tern must match the entire string. The repl may be an empty
719 string, in which case the final `/' may also be omitted. To
720 quote the final `/' in other cases it should be preceded by a
721 single backslash; this is not necessary if the `/' occurs inside
722 a substituted parameter. Note also that the `#', `%' and `#%
723 are not active if they occur inside a substituted parameter,
724 even at the start.
725 If, after quoting rules apply, ${name} expands to an array, the
727 replacements act on each element individually. Note also the
728 effect of the I and S parameter expansion flags below; however,
729 the flags M, R, B, E and N are not useful.
730 For example,
732 foo="twinkle twinkle little star" sub="t*e" rep="spy"
734 print ${foo//${~sub}/$rep}
735 print ${(S)foo//${~sub}/$rep}
736 Here, the `~' ensures that the text of $sub is treated as a pat‐
738 tern rather than a plain string. In the first case, the longest
739 match for t*e is substituted and the result is `spy star', while
740 in the second case, the shortest matches are taken and the
741 result is `spy spy lispy star'.
742 ${#spec}
744 If spec is one of the above substitutions, substitute the length
745 in characters of the result instead of the result itself. If
746 spec is an array expression, substitute the number of elements
747 of the result. This has the side-effect that joining is skipped
748 even in quoted forms, which may affect other sub-expressions in
749 spec. Note that `^', `=', and `~', below, must appear to the
750 left of `#' when these forms are combined.
751 If the option POSIX_IDENTIFIERS is not set, and spec is a simple
753 name, then the braces are optional; this is true even for spe‐
754 cial parameters so e.g. $#- and $#* take the length of the
755 string $- and the array $* respectively. If POSIX_IDENTIFIERS
756 is set, then braces are required for the # to be treated in this
757 fashion.
758 ${^spec}
760 Turn on the RC_EXPAND_PARAM option for the evaluation of spec;
761 if the `^' is doubled, turn it off. When this option is set,
762 array expansions of the form foo${xx}bar, where the parameter xx
763 is set to (a b c), are substituted with `fooabar foobbar
764 foocbar' instead of the default `fooa b cbar'. Note that an
765 empty array will therefore cause all arguments to be removed.
766 Internally, each such expansion is converted into the equivalent
768 list for brace expansion. E.g., ${^var} becomes
769 {$var[1],$var[2],...}, and is processed as described in the sec‐
770 tion `Brace Expansion' below: note, however, the expansion hap‐
771 pens immediately, with any explicit brace expansion happening
772 later. If word splitting is also in effect the $var[N] may
773 themselves be split into different list elements.
774 ${=spec}
776 Perform word splitting using the rules for SH_WORD_SPLIT during
777 the evaluation of spec, but regardless of whether the parameter
778 appears in double quotes; if the `=' is doubled, turn it off.
779 This forces parameter expansions to be split into separate words
780 before substitution, using IFS as a delimiter. This is done by
781 default in most other shells.
782 Note that splitting is applied to word in the assignment forms
784 of spec before the assignment to name is performed. This
785 affects the result of array assignments with the A flag.
786 ${~spec}
788 Turn on the GLOB_SUBST option for the evaluation of spec; if the
789 `~' is doubled, turn it off. When this option is set, the
790 string resulting from the expansion will be interpreted as a
791 pattern anywhere that is possible, such as in filename expansion
792 and filename generation and pattern-matching contexts like the
793 right hand side of the `=' and `!=' operators in conditions.
794 In nested substitutions, note that the effect of the ~ applies
796 to the result of the current level of substitution. A surround‐
797 ing pattern operation on the result may cancel it. Hence, for
798 example, if the parameter foo is set to *, ${~foo//\*/*.c} is
799 substituted by the pattern *.c, which may be expanded by file‐
800 name generation, but ${${~foo}//\*/*.c} substitutes to the
801 string *.c, which will not be further expanded.
802 If a ${...} type parameter expression or a $(...) type command substi‐
804 tution is used in place of name above, it is expanded first and the
805 result is used as if it were the value of name. Thus it is possible to
806 perform nested operations: ${${foo#head}%tail} substitutes the value
807 of $foo with both `head' and `tail' deleted. The form with $(...) is
808 often useful in combination with the flags described next; see the
809 examples below. Each name or nested ${...} in a parameter expansion
810 may also be followed by a subscript expression as described in Array
811 Parameters in zshparam(1).
812 Note that double quotes may appear around nested expressions, in which
814 case only the part inside is treated as quoted; for example,
815 ${(f)"$(foo)"} quotes the result of $(foo), but the flag `(f)' (see
816 below) is applied using the rules for unquoted expansions. Note fur‐
817 ther that quotes are themselves nested in this context; for example, in
818 "${(@f)"$(foo)"}", there are two sets of quotes, one surrounding the
819 whole expression, the other (redundant) surrounding the $(foo) as
820 before.
821 Parameter Expansion Flags
823 If the opening brace is directly followed by an opening parenthesis,
824 the string up to the matching closing parenthesis will be taken as a
825 list of flags. In cases where repeating a flag is meaningful, the rep‐
826 etitions need not be consecutive; for example, `(q%q%q)' means the same
827 thing as the more readable `(%%qqq)'. The following flags are sup‐
828 ported:
829 # Evaluate the resulting words as numeric expressions and output
831 the characters corresponding to the resulting integer. Note
832 that this form is entirely distinct from use of the # without
833 parentheses.
834 If the MULTIBYTE option is set and the number is greater than
836 127 (i.e. not an ASCII character) it is treated as a Unicode
837 character.
838 % Expand all % escapes in the resulting words in the same way as
840 in prompts (see EXPANSION OF PROMPT SEQUENCES in zshmisc(1)). If
841 this flag is given twice, full prompt expansion is done on the
842 resulting words, depending on the setting of the PROMPT_PERCENT,
843 PROMPT_SUBST and PROMPT_BANG options.
844 @ In double quotes, array elements are put into separate words.
846 E.g., `"${(@)foo}"' is equivalent to `"${foo[@]}"' and
847 `"${(@)foo[1,2]}"' is the same as `"$foo[1]" "$foo[2]"'. This
848 is distinct from field splitting by the f, s or z flags, which
849 still applies within each array element.
850 A Convert the substitution into an array expression, even if it
852 otherwise would be scalar. This has lower precedence than sub‐
853 scripting, so one level of nested expansion is required in order
854 that subscripts apply to array elements. Thus ${${(A)name}[1]}
855 yields the full value of name when name is scalar.
856 This assigns an array parameter with `${...=...}', `${...:=...}'
858 or `${...::=...}'. If this flag is repeated (as in `AA'),
859 assigns an associative array parameter. Assignment is made
860 before sorting or padding; if field splitting is active, the
861 word part is split before assignment. The name part may be a
862 subscripted range for ordinary arrays; when assigning an asso‐
863 ciative array, the word part must be converted to an array, for
864 example by using `${(AA)=name=...}' to activate field splitting.
865 Surrounding context such as additional nesting or use of the
867 value in a scalar assignment may cause the array to be joined
868 back into a single string again.
869 a Sort in array index order; when combined with `O' sort in
871 reverse array index order. Note that `a' is therefore equiva‐
872 lent to the default but `Oa' is useful for obtaining an array's
873 elements in reverse order.
874 b Quote with backslashes only characters that are special to pat‐
876 tern matching. This is useful when the contents of the variable
877 are to be tested using GLOB_SUBST, including the ${~...} switch.
878 Quoting using one of the q family of flags does not work for
880 this purpose since quotes are not stripped from non-pattern
881 characters by GLOB_SUBST. In other words,
882 pattern=${(q)str}
884 [[ $str = ${~pattern} ]]
885 works if $str is `a*b' but not if it is `a b', whereas
887 pattern=${(b)str}
889 [[ $str = ${~pattern} ]]
890 is always true for any possible value of $str.
892 c With ${#name}, count the total number of characters in an array,
894 as if the elements were concatenated with spaces between them.
895 This is not a true join of the array, so other expressions used
896 with this flag may have an effect on the elements of the array
897 before it is counted.
898 C Capitalize the resulting words. `Words' in this case refers to
900 sequences of alphanumeric characters separated by non-alphanu‐
901 merics, not to words that result from field splitting.
902 D Assume the string or array elements contain directories and
904 attempt to substitute the leading part of these by names. The
905 remainder of the path (the whole of it if the leading part was
906 not substituted) is then quoted so that the whole string can be
907 used as a shell argument. This is the reverse of `~' substitu‐
908 tion: see the section FILENAME EXPANSION below.
909 e Perform single word shell expansions, namely parameter expan‐
911 sion, command substitution and arithmetic expansion, on the
912 result. Such expansions can be nested but too deep recursion may
913 have unpredictable effects.
914 f Split the result of the expansion at newlines. This is a short‐
916 hand for `ps:\n:'.
917 F Join the words of arrays together using newline as a separator.
919 This is a shorthand for `pj:\n:'.
920 g:opts:
922 Process escape sequences like the echo builtin when no options
923 are given (g::). With the o option, octal escapes don't take a
924 leading zero. With the c option, sequences like `^X' are also
925 processed. With the e option, processes `\M-t' and similar
926 sequences like the print builtin. With both of the o and e
927 options, behaves like the print builtin except that in none of
928 these modes is `\c' interpreted.
929 i Sort case-insensitively. May be combined with `n' or `O'.
931 k If name refers to an associative array, substitute the keys
933 (element names) rather than the values of the elements. Used
934 with subscripts (including ordinary arrays), force indices or
935 keys to be substituted even if the subscript form refers to val‐
936 ues. However, this flag may not be combined with subscript
937 ranges. With the KSH_ARRAYS option a subscript `[*]' or `[@]'
938 is needed to operate on the whole array, as usual.
939 L Convert all letters in the result to lower case.
941 n Sort decimal integers numerically; if the first differing char‐
943 acters of two test strings are not digits, sorting is lexical.
944 Integers with more initial zeroes are sorted before those with
945 fewer or none. Hence the array `foo1 foo02 foo2 foo3 foo20
946 foo23' is sorted into the order shown. May be combined with `i'
947 or `O'.
948 o Sort the resulting words in ascending order; if this appears on
950 its own the sorting is lexical and case-sensitive (unless the
951 locale renders it case-insensitive). Sorting in ascending order
952 is the default for other forms of sorting, so this is ignored if
953 combined with `a', `i' or `n'.
954 O Sort the resulting words in descending order; `O' without `a',
956 `i' or `n' sorts in reverse lexical order. May be combined with
957 `a', `i' or `n' to reverse the order of sorting.
958 P This forces the value of the parameter name to be interpreted as
960 a further parameter name, whose value will be used where appro‐
961 priate. Note that flags set with one of the typeset family of
962 commands (in particular case transformations) are not applied to
963 the value of name used in this fashion.
964 If used with a nested parameter or command substitution, the
966 result of that will be taken as a parameter name in the same
967 way. For example, if you have `foo=bar' and `bar=baz', the
968 strings ${(P)foo}, ${(P)${foo}}, and ${(P)$(echo bar)} will be
969 expanded to `baz'.
970 Likewise, if the reference is itself nested, the expression with
972 the flag is treated as if it were directly replaced by the
973 parameter name. It is an error if this nested substitution pro‐
974 duces an array with more than one word. For example, if
975 `name=assoc' where the parameter assoc is an associative array,
976 then `${${(P)name}[elt]}' refers to the element of the associa‐
977 tive subscripted `elt'.
978 q Quote characters that are special to the shell in the resulting
980 words with backslashes; unprintable or invalid characters are
981 quoted using the $'\NNN' form, with separate quotes for each
982 octet.
983 If this flag is given twice, the resulting words are quoted in
985 single quotes and if it is given three times, the words are
986 quoted in double quotes; in these forms no special handling of
987 unprintable or invalid characters is attempted. If the flag is
988 given four times, the words are quoted in single quotes preceded
989 by a $. Note that in all three of these forms quoting is done
990 unconditionally, even if this does not change the way the
991 resulting string would be interpreted by the shell.
992 If a q- is given (only a single q may appear), a minimal form of
994 single quoting is used that only quotes the string if needed to
995 protect special characters. Typically this form gives the most
996 readable output.
997 If a q+ is given, an extended form of minimal quoting is used
999 that causes unprintable characters to be rendered using $'...'.
1000 This quoting is similar to that used by the output of values by
1001 the typeset family of commands.
1002 Q Remove one level of quotes from the resulting words.
1004 t Use a string describing the type of the parameter where the
1006 value of the parameter would usually appear. This string con‐
1007 sists of keywords separated by hyphens (`-'). The first keyword
1008 in the string describes the main type, it can be one of
1009 `scalar', `array', `integer', `float' or `association'. The
1010 other keywords describe the type in more detail:
1011 local for local parameters
1013 left for left justified parameters
1015 right_blanks
1017 for right justified parameters with leading blanks
1018 right_zeros
1020 for right justified parameters with leading zeros
1021 lower for parameters whose value is converted to all lower case
1023 when it is expanded
1024 upper for parameters whose value is converted to all upper case
1026 when it is expanded
1027 readonly
1029 for readonly parameters
1030 tag for tagged parameters
1032 export for exported parameters
1034 unique for arrays which keep only the first occurrence of dupli‐
1036 cated values
1037 hide for parameters with the `hide' flag
1039 hideval
1041 for parameters with the `hideval' flag
1042 special
1044 for special parameters defined by the shell
1045 u Expand only the first occurrence of each unique word.
1047 U Convert all letters in the result to upper case.
1049 v Used with k, substitute (as two consecutive words) both the key
1051 and the value of each associative array element. Used with sub‐
1052 scripts, force values to be substituted even if the subscript
1053 form refers to indices or keys.
1054 V Make any special characters in the resulting words visible.
1056 w With ${#name}, count words in arrays or strings; the s flag may
1058 be used to set a word delimiter.
1059 W Similar to w with the difference that empty words between
1061 repeated delimiters are also counted.
1062 X With this flag, parsing errors occurring with the Q, e and #
1064 flags or the pattern matching forms such as `${name#pattern}'
1065 are reported. Without the flag, errors are silently ignored.
1066 z Split the result of the expansion into words using shell parsing
1068 to find the words, i.e. taking into account any quoting in the
1069 value. Comments are not treated specially but as ordinary
1070 strings, similar to interactive shells with the INTERACTIVE_COM‐
1071 MENTS option unset (however, see the Z flag below for related
1072 options)
1073 Note that this is done very late, even later than the `(s)'
1075 flag. So to access single words in the result use nested expan‐
1076 sions as in `${${(z)foo}[2]}'. Likewise, to remove the quotes in
1077 the resulting words use `${(Q)${(z)foo}}'.
1078 0 Split the result of the expansion on null bytes. This is a
1080 shorthand for `ps:\0:'.
1081 The following flags (except p) are followed by one or more arguments as
1083 shown. Any character, or the matching pairs `(...)', `{...}', `[...]',
1084 or `<...>', may be used in place of a colon as delimiters, but note
1085 that when a flag takes more than one argument, a matched pair of delim‐
1086 iters must surround each argument.
1087 p Recognize the same escape sequences as the print builtin in
1089 string arguments to any of the flags described below that follow
1090 this argument.
1091 Alternatively, with this option string arguments may be in the
1093 form $var in which case the value of the variable is substi‐
1094 tuted. Note this form is strict; the string argument does not
1095 undergo general parameter expansion.
1096 For example,
1098 sep=:
1100 val=a:b:c
1101 print ${(ps.$sep.)val}
1102 splits the variable on a :.
1104 ~ Strings inserted into the expansion by any of the flags below
1106 are to be treated as patterns. This applies to the string argu‐
1107 ments of flags that follow ~ within the same set of parentheses.
1108 Compare with ~ outside parentheses, which forces the entire sub‐
1109 stituted string to be treated as a pattern. Hence, for example,
1110 [[ "?" = ${(~j.|.)array} ]]
1112 treats `|' as a pattern and succeeds if and only if $array con‐
1114 tains the string `?' as an element. The ~ may be repeated to
1115 toggle the behaviour; its effect only lasts to the end of the
1116 parenthesised group.
1117 j:string:
1119 Join the words of arrays together using string as a separator.
1120 Note that this occurs before field splitting by the s:string:
1121 flag or the SH_WORD_SPLIT option.
1122 l:expr::string1::string2:
1124 Pad the resulting words on the left. Each word will be trun‐
1125 cated if required and placed in a field expr characters wide.
1126 The arguments :string1: and :string2: are optional; neither, the
1128 first, or both may be given. Note that the same pairs of delim‐
1129 iters must be used for each of the three arguments. The space
1130 to the left will be filled with string1 (concatenated as often
1131 as needed) or spaces if string1 is not given. If both string1
1132 and string2 are given, string2 is inserted once directly to the
1133 left of each word, truncated if necessary, before string1 is
1134 used to produce any remaining padding.
1135 If either of string1 or string2 is present but empty, i.e. there
1137 are two delimiters together at that point, the first character
1138 of $IFS is used instead.
1139 If the MULTIBYTE option is in effect, the flag m may also be
1141 given, in which case widths will be used for the calculation of
1142 padding; otherwise individual multibyte characters are treated
1143 as occupying one unit of width.
1144 If the MULTIBYTE option is not in effect, each byte in the
1146 string is treated as occupying one unit of width.
1147 Control characters are always assumed to be one unit wide; this
1149 allows the mechanism to be used for generating repetitions of
1150 control characters.
1151 m Only useful together with one of the flags l or r or with the #
1153 length operator when the MULTIBYTE option is in effect. Use the
1154 character width reported by the system in calculating how much
1155 of the string it occupies or the overall length of the string.
1156 Most printable characters have a width of one unit, however cer‐
1157 tain Asian character sets and certain special effects use wider
1158 characters; combining characters have zero width. Non-printable
1159 characters are arbitrarily counted as zero width; how they would
1160 actually be displayed will vary.
1161 If the m is repeated, the character either counts zero (if it
1163 has zero width), else one. For printable character strings this
1164 has the effect of counting the number of glyphs (visibly sepa‐
1165 rate characters), except for the case where combining characters
1166 themselves have non-zero width (true in certain alphabets).
1167 r:expr::string1::string2:
1169 As l, but pad the words on the right and insert string2 immedi‐
1170 ately to the right of the string to be padded.
1171 Left and right padding may be used together. In this case the
1173 strategy is to apply left padding to the first half width of
1174 each of the resulting words, and right padding to the second
1175 half. If the string to be padded has odd width the extra pad‐
1176 ding is applied on the left.
1177 s:string:
1179 Force field splitting at the separator string. Note that a
1180 string of two or more characters means that all of them must
1181 match in sequence; this differs from the treatment of two or
1182 more characters in the IFS parameter. See also the = flag and
1183 the SH_WORD_SPLIT option. An empty string may also be given in
1184 which case every character will be a separate element.
1185 For historical reasons, the usual behaviour that empty array
1187 elements are retained inside double quotes is disabled for
1188 arrays generated by splitting; hence the following:
1189 line="one::three"
1191 print -l "${(s.:.)line}"
1192 produces two lines of output for one and three and elides the
1194 empty field. To override this behaviour, supply the `(@)' flag
1195 as well, i.e. "${(@s.:.)line}".
1196 Z:opts:
1198 As z but takes a combination of option letters between a follow‐
1199 ing pair of delimiter characters. With no options the effect is
1200 identical to z. (Z+c+) causes comments to be parsed as a string
1201 and retained; any field in the resulting array beginning with an
1202 unquoted comment character is a comment. (Z+C+) causes comments
1203 to be parsed and removed. The rule for comments is standard:
1204 anything between a word starting with the third character of
1205 $HISTCHARS, default #, up to the next newline is a comment.
1206 (Z+n+) causes unquoted newlines to be treated as ordinary white‐
1207 space, else they are treated as if they are shell code delim‐
1208 iters and converted to semicolons. Options are combined within
1209 the same set of delimiters, e.g. (Z+Cn+).
1210 _:flags:
1212 The underscore (_) flag is reserved for future use. As of this
1213 revision of zsh, there are no valid flags; anything following an
1214 underscore, other than an empty pair of delimiters, is treated
1215 as an error, and the flag itself has no effect.
1216 The following flags are meaningful with the ${...#...} or ${...%...}
1218 forms. The S and I flags may also be used with the ${.../...} forms.
1219 S With # or ##, search for the match that starts closest to the
1221 start of the string (a `substring match'). Of all matches at a
1222 particular position, # selects the shortest and ## the longest:
1223 % str="aXbXc"
1225 % echo ${(S)str#X*}
1226 abXc
1227 % echo ${(S)str##X*}
1228 a
1229 %
1230 With % or %%, search for the match that starts closest to the
1232 end of the string:
1233 % str="aXbXc"
1235 % echo ${(S)str%X*}
1236 aXbc
1237 % echo ${(S)str%%X*}
1238 aXb
1239 %
1240 (Note that % and %% don't search for the match that ends closest
1242 to the end of the string, as one might expect.)
1243 With substitution via ${.../...} or ${...//...}, specifies
1245 non-greedy matching, i.e. that the shortest instead of the long‐
1246 est match should be replaced:
1247 % str="abab"
1249 % echo ${str/*b/_}
1250 _
1251 % echo ${(S)str/*b/_}
1252 _ab
1253 %
1254 I:expr:
1256 Search the exprth match (where expr evaluates to a number).
1257 This only applies when searching for substrings, either with the
1258 S flag, or with ${.../...} (only the exprth match is substi‐
1259 tuted) or ${...//...} (all matches from the exprth on are sub‐
1260 stituted). The default is to take the first match.
1261 The exprth match is counted such that there is either one or
1263 zero matches from each starting position in the string, although
1264 for global substitution matches overlapping previous replace‐
1265 ments are ignored. With the ${...%...} and ${...%%...} forms,
1266 the starting position for the match moves backwards from the end
1267 as the index increases, while with the other forms it moves for‐
1268 ward from the start.
1269 Hence with the string
1271 which switch is the right switch for Ipswich?
1272 substitutions of the form ${(SI:N:)string#w*ch} as N increases
1273 from 1 will match and remove `which', `witch', `witch' and
1274 `wich'; the form using `##' will match and remove `which switch
1275 is the right switch for Ipswich', `witch is the right switch for
1276 Ipswich', `witch for Ipswich' and `wich'. The form using `%'
1277 will remove the same matches as for `#', but in reverse order,
1278 and the form using `%%' will remove the same matches as for `##'
1279 in reverse order.
1280 B Include the index of the beginning of the match in the result.
1282 E Include the index one character past the end of the match in the
1284 result (note this is inconsistent with other uses of parameter
1285 index).
1286 M Include the matched portion in the result.
1288 N Include the length of the match in the result.
1290 R Include the unmatched portion in the result (the Rest).
1292 Rules
1294 Here is a summary of the rules for substitution; this assumes that
1295 braces are present around the substitution, i.e. ${...}. Some particu‐
1296 lar examples are given below. Note that the Zsh Development Group
1297 accepts no responsibility for any brain damage which may occur during
1298 the reading of the following rules.
1299 1. Nested substitution
1301 If multiple nested ${...} forms are present, substitution is
1302 performed from the inside outwards. At each level, the substi‐
1303 tution takes account of whether the current value is a scalar or
1304 an array, whether the whole substitution is in double quotes,
1305 and what flags are supplied to the current level of substitu‐
1306 tion, just as if the nested substitution were the outermost.
1307 The flags are not propagated up to enclosing substitutions; the
1308 nested substitution will return either a scalar or an array as
1309 determined by the flags, possibly adjusted for quoting. All the
1310 following steps take place where applicable at all levels of
1311 substitution.
1312 Note that, unless the `(P)' flag is present, the flags and any
1314 subscripts apply directly to the value of the nested substitu‐
1315 tion; for example, the expansion ${${foo}} behaves exactly the
1316 same as ${foo}. When the `(P)' flag is present in a nested sub‐
1317 stitution, the other substitution rules are applied to the value
1318 before it is interpreted as a name, so ${${(P)foo}} may differ
1319 from ${(P)foo}.
1320 At each nested level of substitution, the substituted words
1322 undergo all forms of single-word substitution (i.e. not filename
1323 generation), including command substitution, arithmetic expan‐
1324 sion and filename expansion (i.e. leading ~ and =). Thus, for
1325 example, ${${:-=cat}:h} expands to the directory where the cat
1326 program resides. (Explanation: the internal substitution has no
1327 parameter but a default value =cat, which is expanded by file‐
1328 name expansion to a full path; the outer substitution then
1329 applies the modifier :h and takes the directory part of the
1330 path.)
1331 2. Internal parameter flags
1333 Any parameter flags set by one of the typeset family of com‐
1334 mands, in particular the -L, -R, -Z, -u and -l options for pad‐
1335 ding and capitalization, are applied directly to the parameter
1336 value. Note these flags are options to the command, e.g. `type‐
1337 set -Z'; they are not the same as the flags used within parame‐
1338 ter substitutions.
1339 At the outermost level of substitution, the `(P)' flag (rule 4.)
1341 ignores these transformations and uses the unmodified value of
1342 the parameter as the name to be replaced. This is usually the
1343 desired behavior because padding may make the value syntacti‐
1344 cally illegal as a parameter name, but if capitalization changes
1345 are desired, use the ${${(P)foo}} form (rule 25.).
1346 3. Parameter subscripting
1348 If the value is a raw parameter reference with a subscript, such
1349 as ${var[3]}, the effect of subscripting is applied directly to
1350 the parameter. Subscripts are evaluated left to right; subse‐
1351 quent subscripts apply to the scalar or array value yielded by
1352 the previous subscript. Thus if var is an array, ${var[1][2]}
1353 is the second character of the first word, but ${var[2,4][2]} is
1354 the entire third word (the second word of the range of words two
1355 through four of the original array). Any number of subscripts
1356 may appear. Flags such as `(k)' and `(v)' which alter the
1357 result of subscripting are applied.
1358 4. Parameter name replacement
1360 At the outermost level of nesting only, the `(P)' flag is
1361 applied. This treats the value so far as a parameter name
1362 (which may include a subscript expression) and replaces that
1363 with the corresponding value. This replacement occurs later if
1364 the `(P)' flag appears in a nested substitution.
1365 If the value so far names a parameter that has internal flags
1367 (rule 2.), those internal flags are applied to the new value
1368 after replacement.
1369 5. Double-quoted joining
1371 If the value after this process is an array, and the substitu‐
1372 tion appears in double quotes, and neither an `(@)' flag nor a
1373 `#' length operator is present at the current level, then words
1374 of the value are joined with the first character of the parame‐
1375 ter $IFS, by default a space, between each word (single word
1376 arrays are not modified). If the `(j)' flag is present, that is
1377 used for joining instead of $IFS.
1378 6. Nested subscripting
1380 Any remaining subscripts (i.e. of a nested substitution) are
1381 evaluated at this point, based on whether the value is an array
1382 or a scalar. As with 3., multiple subscripts can appear. Note
1383 that ${foo[2,4][2]} is thus equivalent to ${${foo[2,4]}[2]} and
1384 also to "${${(@)foo[2,4]}[2]}" (the nested substitution returns
1385 an array in both cases), but not to "${${foo[2,4]}[2]}" (the
1386 nested substitution returns a scalar because of the quotes).
1387 7. Modifiers
1389 Any modifiers, as specified by a trailing `#', `%', `/' (possi‐
1390 bly doubled) or by a set of modifiers of the form `:...' (see
1391 the section `Modifiers' in the section `History Expansion'), are
1392 applied to the words of the value at this level.
1393 8. Character evaluation
1395 Any `(#)' flag is applied, evaluating the result so far numeri‐
1396 cally as a character.
1397 9. Length
1399 Any initial `#' modifier, i.e. in the form ${#var}, is used to
1400 evaluate the length of the expression so far.
1401 10. Forced joining
1403 If the `(j)' flag is present, or no `(j)' flag is present but
1404 the string is to be split as given by rule 11., and joining did
1405 not take place at rule 5., any words in the value are joined
1406 together using the given string or the first character of $IFS
1407 if none. Note that the `(F)' flag implicitly supplies a string
1408 for joining in this manner.
1409 11. Simple word splitting
1411 If one of the `(s)' or `(f)' flags are present, or the `=' spec‐
1412 ifier was present (e.g. ${=var}), the word is split on occur‐
1413 rences of the specified string, or (for = with neither of the
1414 two flags present) any of the characters in $IFS.
1415 If no `(s)', `(f)' or `=' was given, but the word is not quoted
1417 and the option SH_WORD_SPLIT is set, the word is split on occur‐
1418 rences of any of the characters in $IFS. Note this step, too,
1419 takes place at all levels of a nested substitution.
1420 12. Case modification
1422 Any case modification from one of the flags `(L)', `(U)' or
1423 `(C)' is applied.
1424 13. Escape sequence replacement
1426 First any replacements from the `(g)' flag are performed, then
1427 any prompt-style formatting from the `(%)' family of flags is
1428 applied.
1429 14. Quote application
1431 Any quoting or unquoting using `(q)' and `(Q)' and related flags
1432 is applied.
1433 15. Directory naming
1435 Any directory name substitution using `(D)' flag is applied.
1436 16. Visibility enhancement
1438 Any modifications to make characters visible using the `(V)'
1439 flag are applied.
1440 17. Lexical word splitting
1442 If the '(z)' flag or one of the forms of the '(Z)' flag is
1443 present, the word is split as if it were a shell command line,
1444 so that quotation marks and other metacharacters are used to
1445 decide what constitutes a word. Note this form of splitting is
1446 entirely distinct from that described by rule 11.: it does not
1447 use $IFS, and does not cause forced joining.
1448 18. Uniqueness
1450 If the result is an array and the `(u)' flag was present, dupli‐
1451 cate elements are removed from the array.
1452 19. Ordering
1454 If the result is still an array and one of the `(o)' or `(O)'
1455 flags was present, the array is reordered.
1456 20. RC_EXPAND_PARAM
1458 At this point the decision is made whether any resulting array
1459 elements are to be combined element by element with surrounding
1460 text, as given by either the RC_EXPAND_PARAM option or the `^'
1461 flag.
1462 21. Re-evaluation
1464 Any `(e)' flag is applied to the value, forcing it to be
1465 re-examined for new parameter substitutions, but also for com‐
1466 mand and arithmetic substitutions.
1467 22. Padding
1469 Any padding of the value by the `(l.fill.)' or `(r.fill.)' flags
1470 is applied.
1471 23. Semantic joining
1473 In contexts where expansion semantics requires a single word to
1474 result, all words are rejoined with the first character of IFS
1475 between. So in `${(P)${(f)lines}}' the value of ${lines} is
1476 split at newlines, but then must be joined again before the
1477 `(P)' flag can be applied.
1478 If a single word is not required, this rule is skipped.
1480 24. Empty argument removal
1482 If the substitution does not appear in double quotes, any
1483 resulting zero-length argument, whether from a scalar or an ele‐
1484 ment of an array, is elided from the list of arguments inserted
1485 into the command line.
1486 Strictly speaking, the removal happens later as the same happens
1488 with other forms of substitution; the point to note here is sim‐
1489 ply that it occurs after any of the above parameter operations.
1490 25. Nested parameter name replacement
1492 If the `(P)' flag is present and rule 4. has not applied, the
1493 value so far is treated as a parameter name (which may include a
1494 subscript expression) and replaced with the corresponding value,
1495 with internal flags (rule 2.) applied to the new value.
1496 Examples
1498 The flag f is useful to split a double-quoted substitution line by
1499 line. For example, ${(f)"$(<file)"} substitutes the contents of file
1500 divided so that each line is an element of the resulting array. Com‐
1501 pare this with the effect of $(<file) alone, which divides the file up
1502 by words, or the same inside double quotes, which makes the entire con‐
1503 tent of the file a single string.
1504 The following illustrates the rules for nested parameter expansions.
1506 Suppose that $foo contains the array (bar baz):
1507 "${(@)${foo}[1]}"
1509 This produces the result b. First, the inner substitution
1510 "${foo}", which has no array (@) flag, produces a single word
1511 result "bar baz". The outer substitution "${(@)...[1]}" detects
1512 that this is a scalar, so that (despite the `(@)' flag) the sub‐
1513 script picks the first character.
1514 "${${(@)foo}[1]}"
1516 This produces the result `bar'. In this case, the inner substi‐
1517 tution "${(@)foo}" produces the array `(bar baz)'. The outer
1518 substitution "${...[1]}" detects that this is an array and picks
1519 the first word. This is similar to the simple case "${foo[1]}".
1520 As an example of the rules for word splitting and joining, suppose $foo
1522 contains the array `(ax1 bx1)'. Then
1523 ${(s/x/)foo}
1525 produces the words `a', `1 b' and `1'.
1526 ${(j/x/s/x/)foo}
1528 produces `a', `1', `b' and `1'.
1529 ${(s/x/)foo%%1*}
1531 produces `a' and ` b' (note the extra space). As substitution
1532 occurs before either joining or splitting, the operation first
1533 generates the modified array (ax bx), which is joined to give
1534 "ax bx", and then split to give `a', ` b' and `'. The final
1535 empty string will then be elided, as it is not in double quotes.
1536
1538 A command enclosed in parentheses preceded by a dollar sign, like
1539 `$(...)', or quoted with grave accents, like ``...`', is replaced with
1540 its standard output, with any trailing newlines deleted. If the sub‐
1541 stitution is not enclosed in double quotes, the output is broken into
1542 words using the IFS parameter.
1543 The substitution `$(cat foo)' may be replaced by the faster `$(<foo)'.
1545 In this case foo undergoes single word shell expansions (parameter
1546 expansion, command substitution and arithmetic expansion), but not
1547 filename generation.
1548 If the option GLOB_SUBST is set, the result of any unquoted command
1550 substitution, including the special form just mentioned, is eligible
1551 for filename generation.
1552
1554 A string of the form `$[exp]' or `$((exp))' is substituted with the
1555 value of the arithmetic expression exp. exp is subjected to parameter
1556 expansion, command substitution and arithmetic expansion before it is
1557 evaluated. See the section `Arithmetic Evaluation'.
1558
1560 A string of the form `foo{xx,yy,zz}bar' is expanded to the individual
1561 words `fooxxbar', `fooyybar' and `foozzbar'. Left-to-right order is
1562 preserved. This construct may be nested. Commas may be quoted in
1563 order to include them literally in a word.
1564 An expression of the form `{n1..n2}', where n1 and n2 are integers, is
1566 expanded to every number between n1 and n2 inclusive. If either number
1567 begins with a zero, all the resulting numbers will be padded with lead‐
1568 ing zeroes to that minimum width, but for negative numbers the - char‐
1569 acter is also included in the width. If the numbers are in decreasing
1570 order the resulting sequence will also be in decreasing order.
1571 An expression of the form `{n1..n2..n3}', where n1, n2, and n3 are
1573 integers, is expanded as above, but only every n3th number starting
1574 from n1 is output. If n3 is negative the numbers are output in reverse
1575 order, this is slightly different from simply swapping n1 and n2 in the
1576 case that the step n3 doesn't evenly divide the range. Zero padding
1577 can be specified in any of the three numbers, specifying it in the
1578 third can be useful to pad for example `{-99..100..01}' which is not
1579 possible to specify by putting a 0 on either of the first two numbers
1580 (i.e. pad to two characters).
1581 An expression of the form `{c1..c2}', where c1 and c2 are single char‐
1583 acters (which may be multibyte characters), is expanded to every char‐
1584 acter in the range from c1 to c2 in whatever character sequence is used
1585 internally. For characters with code points below 128 this is US ASCII
1586 (this is the only case most users will need). If any intervening char‐
1587 acter is not printable, appropriate quotation is used to render it
1588 printable. If the character sequence is reversed, the output is in
1589 reverse order, e.g. `{d..a}' is substituted as `d c b a'.
1590 If a brace expression matches none of the above forms, it is left
1592 unchanged, unless the option BRACE_CCL (an abbreviation for `brace
1593 character class') is set. In that case, it is expanded to a list of
1594 the individual characters between the braces sorted into the order of
1595 the characters in the ASCII character set (multibyte characters are not
1596 currently handled). The syntax is similar to a [...] expression in
1597 filename generation: `-' is treated specially to denote a range of
1598 characters, but `^' or `!' as the first character is treated normally.
1599 For example, `{abcdef0-9}' expands to 16 words 0 1 2 3 4 5 6 7 8 9 a b
1600 c d e f.
1601 Note that brace expansion is not part of filename generation (glob‐
1603 bing); an expression such as */{foo,bar} is split into two separate
1604 words */foo and */bar before filename generation takes place. In par‐
1605 ticular, note that this is liable to produce a `no match' error if
1606 either of the two expressions does not match; this is to be contrasted
1607 with */(foo|bar), which is treated as a single pattern but otherwise
1608 has similar effects.
1609 To combine brace expansion with array expansion, see the ${^spec} form
1611 described in the section Parameter Expansion above.
1612
1614 Each word is checked to see if it begins with an unquoted `~'. If it
1615 does, then the word up to a `/', or the end of the word if there is no
1616 `/', is checked to see if it can be substituted in one of the ways
1617 described here. If so, then the `~' and the checked portion are
1618 replaced with the appropriate substitute value.
1619 A `~' by itself is replaced by the value of $HOME. A `~' followed by a
1621 `+' or a `-' is replaced by current or previous working directory,
1622 respectively.
1623 A `~' followed by a number is replaced by the directory at that posi‐
1625 tion in the directory stack. `~0' is equivalent to `~+', and `~1' is
1626 the top of the stack. `~+' followed by a number is replaced by the
1627 directory at that position in the directory stack. `~+0' is equivalent
1628 to `~+', and `~+1' is the top of the stack. `~-' followed by a number
1629 is replaced by the directory that many positions from the bottom of the
1630 stack. `~-0' is the bottom of the stack. The PUSHD_MINUS option
1631 exchanges the effects of `~+' and `~-' where they are followed by a
1632 number.
1633 Dynamic named directories
1635 If the function zsh_directory_name exists, or the shell variable
1636 zsh_directory_name_functions exists and contains an array of function
1637 names, then the functions are used to implement dynamic directory nam‐
1638 ing. The functions are tried in order until one returns status zero,
1639 so it is important that functions test whether they can handle the case
1640 in question and return an appropriate status.
1641 A `~' followed by a string namstr in unquoted square brackets is
1643 treated specially as a dynamic directory name. Note that the first
1644 unquoted closing square bracket always terminates namstr. The shell
1645 function is passed two arguments: the string n (for name) and namstr.
1646 It should either set the array reply to a single element which is the
1647 directory corresponding to the name and return status zero (executing
1648 an assignment as the last statement is usually sufficient), or it
1649 should return status non-zero. In the former case the element of reply
1650 is used as the directory; in the latter case the substitution is deemed
1651 to have failed. If all functions fail and the option NOMATCH is set,
1652 an error results.
1653 The functions defined as above are also used to see if a directory can
1655 be turned into a name, for example when printing the directory stack or
1656 when expanding %~ in prompts. In this case each function is passed two
1657 arguments: the string d (for directory) and the candidate for dynamic
1658 naming. The function should either return non-zero status, if the
1659 directory cannot be named by the function, or it should set the array
1660 reply to consist of two elements: the first is the dynamic name for the
1661 directory (as would appear within `~[...]'), and the second is the pre‐
1662 fix length of the directory to be replaced. For example, if the trial
1663 directory is /home/myname/src/zsh and the dynamic name for
1664 /home/myname/src (which has 16 characters) is s, then the function sets
1665 reply=(s 16)
1667 The directory name so returned is compared with possible static names
1669 for parts of the directory path, as described below; it is used if the
1670 prefix length matched (16 in the example) is longer than that matched
1671 by any static name.
1672 It is not a requirement that a function implements both n and d calls;
1674 for example, it might be appropriate for certain dynamic forms of
1675 expansion not to be contracted to names. In that case any call with
1676 the first argument d should cause a non-zero status to be returned.
1677 The completion system calls `zsh_directory_name c' followed by equiva‐
1679 lent calls to elements of the array zsh_directory_name_functions, if it
1680 exists, in order to complete dynamic names for directories. The code
1681 for this should be as for any other completion function as described in
1682 zshcompsys(1).
1683 As a working example, here is a function that expands any dynamic names
1685 beginning with the string p: to directories below /home/pws/perforce.
1686 In this simple case a static name for the directory would be just as
1687 effective.
1688 zsh_directory_name() {
1690 emulate -L zsh
1691 setopt extendedglob
1692 local -a match mbegin mend
1693 if [[ $1 = d ]]; then
1694 # turn the directory into a name
1695 if [[ $2 = (#b)(/home/pws/perforce/)([^/]##)* ]]; then
1696 typeset -ga reply
1697 reply=(p:$match[2] $(( ${#match[1]} + ${#match[2]} )) )
1698 else
1699 return 1
1700 fi
1701 elif [[ $1 = n ]]; then
1702 # turn the name into a directory
1703 [[ $2 != (#b)p:(?*) ]] && return 1
1704 typeset -ga reply
1705 reply=(/home/pws/perforce/$match[1])
1706 elif [[ $1 = c ]]; then
1707 # complete names
1708 local expl
1709 local -a dirs
1710 dirs=(/home/pws/perforce/*(/:t))
1711 dirs=(p:${^dirs})
1712 _wanted dynamic-dirs expl 'dynamic directory' compadd -S\] -a dirs
1713 return
1714 else
1715 return 1
1716 fi
1717 return 0
1718 }
1719 Static named directories
1721 A `~' followed by anything not already covered consisting of any number
1722 of alphanumeric characters or underscore (`_'), hyphen (`-'), or dot
1723 (`.') is looked up as a named directory, and replaced by the value of
1724 that named directory if found. Named directories are typically home
1725 directories for users on the system. They may also be defined if the
1726 text after the `~' is the name of a string shell parameter whose value
1727 begins with a `/'. Note that trailing slashes will be removed from the
1728 path to the directory (though the original parameter is not modified).
1729 It is also possible to define directory names using the -d option to
1731 the hash builtin.
1732 When the shell prints a path (e.g. when expanding %~ in prompts or when
1734 printing the directory stack), the path is checked to see if it has a
1735 named directory as its prefix. If so, then the prefix portion is
1736 replaced with a `~' followed by the name of the directory. The shorter
1737 of the two ways of referring to the directory is used, i.e. either the
1738 directory name or the full path; the name is used if they are the same
1739 length. The parameters $PWD and $OLDPWD are never abbreviated in this
1740 fashion.
1741 `=' expansion
1743 If a word begins with an unquoted `=' and the EQUALS option is set, the
1744 remainder of the word is taken as the name of a command. If a command
1745 exists by that name, the word is replaced by the full pathname of the
1746 command.
1747 Notes
1749 Filename expansion is performed on the right hand side of a parameter
1750 assignment, including those appearing after commands of the typeset
1751 family. In this case, the right hand side will be treated as a
1752 colon-separated list in the manner of the PATH parameter, so that a `~'
1753 or an `=' following a `:' is eligible for expansion. All such behav‐
1754 iour can be disabled by quoting the `~', the `=', or the whole expres‐
1755 sion (but not simply the colon); the EQUALS option is also respected.
1756 If the option MAGIC_EQUAL_SUBST is set, any unquoted shell argument in
1758 the form `identifier=expression' becomes eligible for file expansion as
1759 described in the previous paragraph. Quoting the first `=' also
1760 inhibits this.
1761
1763 If a word contains an unquoted instance of one of the characters `*',
1764 `(', `|', `<', `[', or `?', it is regarded as a pattern for filename
1765 generation, unless the GLOB option is unset. If the EXTENDED_GLOB
1766 option is set, the `^' and `#' characters also denote a pattern; other‐
1767 wise they are not treated specially by the shell.
1768 The word is replaced with a list of sorted filenames that match the
1770 pattern. If no matching pattern is found, the shell gives an error
1771 message, unless the NULL_GLOB option is set, in which case the word is
1772 deleted; or unless the NOMATCH option is unset, in which case the word
1773 is left unchanged.
1774 In filename generation, the character `/' must be matched explicitly;
1776 also, a `.' must be matched explicitly at the beginning of a pattern or
1777 after a `/', unless the GLOB_DOTS option is set. No filename genera‐
1778 tion pattern matches the files `.' or `..'. In other instances of pat‐
1779 tern matching, the `/' and `.' are not treated specially.
1780 Glob Operators
1782 * Matches any string, including the null string.
1783 ? Matches any character.
1785 [...] Matches any of the enclosed characters. Ranges of characters
1787 can be specified by separating two characters by a `-'. A `-'
1788 or `]' may be matched by including it as the first character in
1789 the list. There are also several named classes of characters,
1790 in the form `[:name:]' with the following meanings. The first
1791 set use the macros provided by the operating system to test for
1792 the given character combinations, including any modifications
1793 due to local language settings, see ctype(3):
1794 [:alnum:]
1796 The character is alphanumeric
1797 [:alpha:]
1799 The character is alphabetic
1800 [:ascii:]
1802 The character is 7-bit, i.e. is a single-byte character
1803 without the top bit set.
1804 [:blank:]
1806 The character is a blank character
1807 [:cntrl:]
1809 The character is a control character
1810 [:digit:]
1812 The character is a decimal digit
1813 [:graph:]
1815 The character is a printable character other than white‐
1816 space
1817 [:lower:]
1819 The character is a lowercase letter
1820 [:print:]
1822 The character is printable
1823 [:punct:]
1825 The character is printable but neither alphanumeric nor
1826 whitespace
1827 [:space:]
1829 The character is whitespace
1830 [:upper:]
1832 The character is an uppercase letter
1833 [:xdigit:]
1835 The character is a hexadecimal digit
1836 Another set of named classes is handled internally by the shell
1838 and is not sensitive to the locale:
1839 [:IDENT:]
1841 The character is allowed to form part of a shell identi‐
1842 fier, such as a parameter name
1843 [:IFS:]
1845 The character is used as an input field separator, i.e.
1846 is contained in the IFS parameter
1847 [:IFSSPACE:]
1849 The character is an IFS white space character; see the
1850 documentation for IFS in the zshparam(1) manual page.
1851 [:INCOMPLETE:]
1853 Matches a byte that starts an incomplete multibyte char‐
1854 acter. Note that there may be a sequence of more than
1855 one bytes that taken together form the prefix of a multi‐
1856 byte character. To test for a potentially incomplete
1857 byte sequence, use the pattern `[[:INCOMPLETE:]]*'. This
1858 will never match a sequence starting with a valid multi‐
1859 byte character.
1860 [:INVALID:]
1862 Matches a byte that does not start a valid multibyte
1863 character. Note this may be a continuation byte of an
1864 incomplete multibyte character as any part of a multibyte
1865 string consisting of invalid and incomplete multibyte
1866 characters is treated as single bytes.
1867 [:WORD:]
1869 The character is treated as part of a word; this test is
1870 sensitive to the value of the WORDCHARS parameter
1871 Note that the square brackets are additional to those enclosing
1873 the whole set of characters, so to test for a single alphanu‐
1874 meric character you need `[[:alnum:]]'. Named character sets
1875 can be used alongside other types, e.g. `[[:alpha:]0-9]'.
1876 [^...]
1878 [!...] Like [...], except that it matches any character which is not in
1879 the given set.
1880 <[x]-[y]>
1882 Matches any number in the range x to y, inclusive. Either of
1883 the numbers may be omitted to make the range open-ended; hence
1884 `<->' matches any number. To match individual digits, the [...]
1885 form is more efficient.
1886 Be careful when using other wildcards adjacent to patterns of
1888 this form; for example, <0-9>* will actually match any number
1889 whatsoever at the start of the string, since the `<0-9>' will
1890 match the first digit, and the `*' will match any others. This
1891 is a trap for the unwary, but is in fact an inevitable conse‐
1892 quence of the rule that the longest possible match always suc‐
1893 ceeds. Expressions such as `<0-9>[^[:digit:]]*' can be used
1894 instead.
1895 (...) Matches the enclosed pattern. This is used for grouping. If
1897 the KSH_GLOB option is set, then a `@', `*', `+', `?' or `!'
1898 immediately preceding the `(' is treated specially, as detailed
1899 below. The option SH_GLOB prevents bare parentheses from being
1900 used in this way, though the KSH_GLOB option is still available.
1901 Note that grouping cannot extend over multiple directories: it
1903 is an error to have a `/' within a group (this only applies for
1904 patterns used in filename generation). There is one exception:
1905 a group of the form (pat/)# appearing as a complete path segment
1906 can match a sequence of directories. For example, foo/(a*/)#bar
1907 matches foo/bar, foo/any/bar, foo/any/anyother/bar, and so on.
1908 x|y Matches either x or y. This operator has lower precedence than
1910 any other. The `|' character must be within parentheses, to
1911 avoid interpretation as a pipeline. The alternatives are tried
1912 in order from left to right.
1913 ^x (Requires EXTENDED_GLOB to be set.) Matches anything except the
1915 pattern x. This has a higher precedence than `/', so `^foo/bar'
1916 will search directories in `.' except `./foo' for a file named
1917 `bar'.
1918 x~y (Requires EXTENDED_GLOB to be set.) Match anything that matches
1920 the pattern x but does not match y. This has lower precedence
1921 than any operator except `|', so `*/*~foo/bar' will search for
1922 all files in all directories in `.' and then exclude `foo/bar'
1923 if there was such a match. Multiple patterns can be excluded by
1924 `foo~bar~baz'. In the exclusion pattern (y), `/' and `.' are
1925 not treated specially the way they usually are in globbing.
1926 x# (Requires EXTENDED_GLOB to be set.) Matches zero or more occur‐
1928 rences of the pattern x. This operator has high precedence;
1929 `12#' is equivalent to `1(2#)', rather than `[1m(12)#'. It is an
1930 error for an unquoted `#' to follow something which cannot be
1931 repeated; this includes an empty string, a pattern already fol‐
1932 lowed by `##', or parentheses when part of a KSH_GLOB pattern
1933 (for example, `!(foo)#' is invalid and must be replaced by
1934 `*(!(foo))').
1935 x## (Requires EXTENDED_GLOB to be set.) Matches one or more occur‐
1937 rences of the pattern x. This operator has high precedence;
1938 `12##' is equivalent to `1(2##)', rather than `[1m(12)##'. No more
1939 than two active `#' characters may appear together. (Note the
1940 potential clash with glob qualifiers in the form `1(2##)' which
1941 should therefore be avoided.)
1942 ksh-like Glob Operators
1944 If the KSH_GLOB option is set, the effects of parentheses can be modi‐
1945 fied by a preceding `@', `*', `+', `?' or `!'. This character need not
1946 be unquoted to have special effects, but the `(' must be.
1947 @(...) Match the pattern in the parentheses. (Like `(...)'.)
1949 *(...) Match any number of occurrences. (Like `(...)#', except that
1951 recursive directory searching is not supported.)
1952 +(...) Match at least one occurrence. (Like `(...)##', except that
1954 recursive directory searching is not supported.)
1955 ?(...) Match zero or one occurrence. (Like `(|...)'.)
1957 !(...) Match anything but the expression in parentheses. (Like
1959 `(^(...))'.)
1960 Precedence
1962 The precedence of the operators given above is (highest) `^', `/', `~',
1963 `|' (lowest); the remaining operators are simply treated from left to
1964 right as part of a string, with `#' and `##' applying to the shortest
1965 possible preceding unit (i.e. a character, `?', `[...]', `<...>', or a
1966 parenthesised expression). As mentioned above, a `/' used as a direc‐
1967 tory separator may not appear inside parentheses, while a `|' must do
1968 so; in patterns used in other contexts than filename generation (for
1969 example, in case statements and tests within `[[...]]'), a `/' is not
1970 special; and `/' is also not special after a `~' appearing outside
1971 parentheses in a filename pattern.
1972 Globbing Flags
1974 There are various flags which affect any text to their right up to the
1975 end of the enclosing group or to the end of the pattern; they require
1976 the EXTENDED_GLOB option. All take the form (#X) where X may have one
1977 of the following forms:
1978 i Case insensitive: upper or lower case characters in the pattern
1980 match upper or lower case characters.
1981 l Lower case characters in the pattern match upper or lower case
1983 characters; upper case characters in the pattern still only
1984 match upper case characters.
1985 I Case sensitive: locally negates the effect of i or l from that
1987 point on.
1988 b Activate backreferences for parenthesised groups in the pattern;
1990 this does not work in filename generation. When a pattern with
1991 a set of active parentheses is matched, the strings matched by
1992 the groups are stored in the array $match, the indices of the
1993 beginning of the matched parentheses in the array $mbegin, and
1994 the indices of the end in the array $mend, with the first ele‐
1995 ment of each array corresponding to the first parenthesised
1996 group, and so on. These arrays are not otherwise special to the
1997 shell. The indices use the same convention as does parameter
1998 substitution, so that elements of $mend and $mbegin may be used
1999 in subscripts; the KSH_ARRAYS option is respected. Sets of
2000 globbing flags are not considered parenthesised groups; only the
2001 first nine active parentheses can be referenced.
2002 For example,
2004 foo="a_string_with_a_message"
2006 if [[ $foo = (a|an)_(#b)(*) ]]; then
2007 print ${foo[$mbegin[1],$mend[1]]}
2008 fi
2009 prints `string_with_a_message'. Note that the first set of
2011 parentheses is before the (#b) and does not create a backrefer‐
2012 ence.
2013 Backreferences work with all forms of pattern matching other
2015 than filename generation, but note that when performing matches
2016 on an entire array, such as ${array#pattern}, or a global sub‐
2017 stitution, such as ${param//pat/repl}, only the data for the
2018 last match remains available. In the case of global replace‐
2019 ments this may still be useful. See the example for the m flag
2020 below.
2021 The numbering of backreferences strictly follows the order of
2023 the opening parentheses from left to right in the pattern
2024 string, although sets of parentheses may be nested. There are
2025 special rules for parentheses followed by `#' or `##'. Only the
2026 last match of the parenthesis is remembered: for example, in `[[
2027 abab = (#b)([ab])# ]]', only the final `b' is stored in
2028 match[1]. Thus extra parentheses may be necessary to match the
2029 complete segment: for example, use `X((ab|cd)#)Y' to match a
2030 whole string of either `ab' or `cd' between `X' and `Y', using
2031 the value of $match[1] rather than $match[2].
2032 If the match fails none of the parameters is altered, so in some
2034 cases it may be necessary to initialise them beforehand. If
2035 some of the backreferences fail to match -- which happens if
2036 they are in an alternate branch which fails to match, or if they
2037 are followed by # and matched zero times -- then the matched
2038 string is set to the empty string, and the start and end indices
2039 are set to -1.
2040 Pattern matching with backreferences is slightly slower than
2042 without.
2043 B Deactivate backreferences, negating the effect of the b flag
2045 from that point on.
2046 cN,M The flag (#cN,M) can be used anywhere that the # or ## operators
2048 can be used except in the expressions `(*/)#' and `(*/)##' in
2049 filename generation, where `/' has special meaning; it cannot be
2050 combined with other globbing flags and a bad pattern error
2051 occurs if it is misplaced. It is equivalent to the form {N,M}
2052 in regular expressions. The previous character or group is
2053 required to match between N and M times, inclusive. The form
2054 (#cN) requires exactly N matches; (#c,M) is equivalent to speci‐
2055 fying N as 0; (#cN,) specifies that there is no maximum limit on
2056 the number of matches.
2057 m Set references to the match data for the entire string matched;
2059 this is similar to backreferencing and does not work in filename
2060 generation. The flag must be in effect at the end of the pat‐
2061 tern, i.e. not local to a group. The parameters $MATCH, $MBEGIN
2062 and $MEND will be set to the string matched and to the indices
2063 of the beginning and end of the string, respectively. This is
2064 most useful in parameter substitutions, as otherwise the string
2065 matched is obvious.
2066 For example,
2068 arr=(veldt jynx grimps waqf zho buck)
2070 print ${arr//(#m)[aeiou]/${(U)MATCH}}
2071 forces all the matches (i.e. all vowels) into uppercase, print‐
2073 ing `vEldt jynx grImps wAqf zhO bUck'.
2074 Unlike backreferences, there is no speed penalty for using match
2076 references, other than the extra substitutions required for the
2077 replacement strings in cases such as the example shown.
2078 M Deactivate the m flag, hence no references to match data will be
2080 created.
2081 anum Approximate matching: num errors are allowed in the string
2083 matched by the pattern. The rules for this are described in the
2084 next subsection.
2085 s, e Unlike the other flags, these have only a local effect, and each
2087 must appear on its own: `(#s)' and `(#e)' are the only valid
2088 forms. The `(#s)' flag succeeds only at the start of the test
2089 string, and the `(#e)' flag succeeds only at the end of the test
2090 string; they correspond to `^' and `$' in standard regular
2091 expressions. They are useful for matching path segments in pat‐
2092 terns other than those in filename generation (where path seg‐
2093 ments are in any case treated separately). For example,
2094 `*((#s)|/)test((#e)|/)*' matches a path segment `test' in any of
2095 the following strings: test, test/at/start, at/end/test,
2096 in/test/middle.
2097 Another use is in parameter substitution; for example
2099 `${array/(#s)A*Z(#e)}' will remove only elements of an array
2100 which match the complete pattern `A*Z'. There are other ways of
2101 performing many operations of this type, however the combination
2102 of the substitution operations `/' and `//' with the `(#s)' and
2103 `(#e)' flags provides a single simple and memorable method.
2104 Note that assertions of the form `(^(#s))' also work, i.e. match
2106 anywhere except at the start of the string, although this actu‐
2107 ally means `anything except a zero-length portion at the start
2108 of the string'; you need to use `(""~(#s))' to match a
2109 zero-length portion of the string not at the start.
2110 q A `q' and everything up to the closing parenthesis of the glob‐
2112 bing flags are ignored by the pattern matching code. This is
2113 intended to support the use of glob qualifiers, see below. The
2114 result is that the pattern `(#b)(*).c(#q.)' can be used both for
2115 globbing and for matching against a string. In the former case,
2116 the `(#q.)' will be treated as a glob qualifier and the `(#b)'
2117 will not be useful, while in the latter case the `(#b)' is use‐
2118 ful for backreferences and the `(#q.)' will be ignored. Note
2119 that colon modifiers in the glob qualifiers are also not applied
2120 in ordinary pattern matching.
2121 u Respect the current locale in determining the presence of multi‐
2123 byte characters in a pattern, provided the shell was compiled
2124 with MULTIBYTE_SUPPORT. This overrides the MULTIBYTE option;
2125 the default behaviour is taken from the option. Compare U.
2126 (Mnemonic: typically multibyte characters are from Unicode in
2127 the UTF-8 encoding, although any extension of ASCII supported by
2128 the system library may be used.)
2129 U All characters are considered to be a single byte long. The
2131 opposite of u. This overrides the MULTIBYTE option.
2132 For example, the test string fooxx can be matched by the pattern
2134 (#i)FOOXX, but not by (#l)FOOXX, (#i)FOO(#I)XX or ((#i)FOOX)X. The
2135 string (#ia2)readme specifies case-insensitive matching of readme with
2136 up to two errors.
2137 When using the ksh syntax for grouping both KSH_GLOB and EXTENDED_GLOB
2139 must be set and the left parenthesis should be preceded by @. Note
2140 also that the flags do not affect letters inside [...] groups, in other
2141 words (#i)[a-z] still matches only lowercase letters. Finally, note
2142 that when examining whole paths case-insensitively every directory must
2143 be searched for all files which match, so that a pattern of the form
2144 (#i)/foo/bar/... is potentially slow.
2145 Approximate Matching
2147 When matching approximately, the shell keeps a count of the errors
2148 found, which cannot exceed the number specified in the (#anum) flags.
2149 Four types of error are recognised:
2150 1. Different characters, as in fooxbar and fooybar.
2152 2. Transposition of characters, as in banana and abnana.
2154 3. A character missing in the target string, as with the pattern
2156 road and target string rod.
2157 4. An extra character appearing in the target string, as with stove
2159 and strove.
2160 Thus, the pattern (#a3)abcd matches dcba, with the errors occurring by
2162 using the first rule twice and the second once, grouping the string as
2163 [d][cb][a] and [a][bc][d].
2164 Non-literal parts of the pattern must match exactly, including charac‐
2166 ters in character ranges: hence (#a1)??? matches strings of length
2167 four, by applying rule 4 to an empty part of the pattern, but not
2168 strings of length two, since all the ? must match. Other characters
2169 which must match exactly are initial dots in filenames (unless the
2170 GLOB_DOTS option is set), and all slashes in filenames, so that a/bc is
2171 two errors from ab/c (the slash cannot be transposed with another char‐
2172 acter). Similarly, errors are counted separately for non-contiguous
2173 strings in the pattern, so that (ab|cd)ef is two errors from aebf.
2174 When using exclusion via the ~ operator, approximate matching is
2176 treated entirely separately for the excluded part and must be activated
2177 separately. Thus, (#a1)README~READ_ME matches READ.ME but not READ_ME,
2178 as the trailing READ_ME is matched without approximation. However,
2179 (#a1)README~(#a1)READ_ME does not match any pattern of the form READ?ME
2180 as all such forms are now excluded.
2181 Apart from exclusions, there is only one overall error count; however,
2183 the maximum errors allowed may be altered locally, and this can be
2184 delimited by grouping. For example, (#a1)cat((#a0)dog)fox allows one
2185 error in total, which may not occur in the dog section, and the pattern
2186 (#a1)cat(#a0)dog(#a1)fox is equivalent. Note that the point at which
2187 an error is first found is the crucial one for establishing whether to
2188 use approximation; for example, (#a1)abc(#a0)xyz will not match
2189 abcdxyz, because the error occurs at the `x', where approximation is
2190 turned off.
2191 Entire path segments may be matched approximately, so that
2193 `(#a1)/foo/d/is/available/at/the/bar' allows one error in any path seg‐
2194 ment. This is much less efficient than without the (#a1), however,
2195 since every directory in the path must be scanned for a possible
2196 approximate match. It is best to place the (#a1) after any path seg‐
2197 ments which are known to be correct.
2198 Recursive Globbing
2200 A pathname component of the form `(foo/)#' matches a path consisting of
2201 zero or more directories matching the pattern foo.
2202 As a shorthand, `**/' is equivalent to `(*/)#'; note that this there‐
2204 fore matches files in the current directory as well as subdirectories.
2205 Thus:
2206 ls -ld -- (*/)#bar
2208 or
2210 ls -ld -- **/bar
2212 does a recursive directory search for files named `bar' (potentially
2214 including the file `bar' in the current directory). This form does not
2215 follow symbolic links; the alternative form `***/' does, but is other‐
2216 wise identical. Neither of these can be combined with other forms of
2217 globbing within the same path segment; in that case, the `*' operators
2218 revert to their usual effect.
2219 Even shorter forms are available when the option GLOB_STAR_SHORT is
2221 set. In that case if no / immediately follows a ** or *** they are
2222 treated as if both a / plus a further * are present. Hence:
2223 setopt GLOBSTARSHORT
2225 ls -ld -- **.c
2226 is equivalent to
2228 ls -ld -- **/*.c
2230 Glob Qualifiers
2232 Patterns used for filename generation may end in a list of qualifiers
2233 enclosed in parentheses. The qualifiers specify which filenames that
2234 otherwise match the given pattern will be inserted in the argument
2235 list.
2236 If the option BARE_GLOB_QUAL is set, then a trailing set of parentheses
2238 containing no `|' or `(' characters (or `~' if it is special) is taken
2239 as a set of glob qualifiers. A glob subexpression that would normally
2240 be taken as glob qualifiers, for example `(^x)', can be forced to be
2241 treated as part of the glob pattern by doubling the parentheses, in
2242 this case producing `((^x))'.
2243 If the option EXTENDED_GLOB is set, a different syntax for glob quali‐
2245 fiers is available, namely `(#qx)' where x is any of the same glob
2246 qualifiers used in the other format. The qualifiers must still appear
2247 at the end of the pattern. However, with this syntax multiple glob
2248 qualifiers may be chained together. They are treated as a logical AND
2249 of the individual sets of flags. Also, as the syntax is unambiguous,
2250 the expression will be treated as glob qualifiers just as long any
2251 parentheses contained within it are balanced; appearance of `|', `(' or
2252 `~' does not negate the effect. Note that qualifiers will be recog‐
2253 nised in this form even if a bare glob qualifier exists at the end of
2254 the pattern, for example `*(#q*)(.)' will recognise executable regular
2255 files if both options are set; however, mixed syntax should probably be
2256 avoided for the sake of clarity. Note that within conditions using the
2257 `[[' form the presence of a parenthesised expression (#q...) at the end
2258 of a string indicates that globbing should be performed; the expression
2259 may include glob qualifiers, but it is also valid if it is simply (#q).
2260 This does not apply to the right hand side of pattern match operators
2261 as the syntax already has special significance.
2262 A qualifier may be any one of the following:
2264 / directories
2266 F `full' (i.e. non-empty) directories. Note that the opposite
2268 sense (^F) expands to empty directories and all non-directories.
2269 Use (/^F) for empty directories.
2270 . plain files
2272 @ symbolic links
2274 = sockets
2276 p named pipes (FIFOs)
2278 * executable plain files (0100 or 0010 or 0001)
2280 % device files (character or block special)
2282 %b block special files
2284 %c character special files
2286 r owner-readable files (0400)
2288 w owner-writable files (0200)
2290 x owner-executable files (0100)
2292 A group-readable files (0040)
2294 I group-writable files (0020)
2296 E group-executable files (0010)
2298 R world-readable files (0004)
2300 W world-writable files (0002)
2302 X world-executable files (0001)
2304 s setuid files (04000)
2306 S setgid files (02000)
2308 t files with the sticky bit (01000)
2310 fspec files with access rights matching spec. This spec may be a octal
2312 number optionally preceded by a `=', a `+', or a `-'. If none of
2313 these characters is given, the behavior is the same as for `='.
2314 The octal number describes the mode bits to be expected, if com‐
2315 bined with a `=', the value given must match the file-modes
2316 exactly, with a `+', at least the bits in the given number must
2317 be set in the file-modes, and with a `-', the bits in the number
2318 must not be set. Giving a `?' instead of a octal digit anywhere
2319 in the number ensures that the corresponding bits in the
2320 file-modes are not checked, this is only useful in combination
2321 with `='.
2322 If the qualifier `f' is followed by any other character anything
2324 up to the next matching character (`[', `{', and `<' match `]',
2325 `}', and `>' respectively, any other character matches itself)
2326 is taken as a list of comma-separated sub-specs. Each sub-spec
2327 may be either an octal number as described above or a list of
2328 any of the characters `u', `g', `o', and `a', followed by a `=',
2329 a `+', or a `-', followed by a list of any of the characters
2330 `r', `w', `x', `s', and `t', or an octal digit. The first list
2331 of characters specify which access rights are to be checked. If
2332 a `u' is given, those for the owner of the file are used, if a
2333 `g' is given, those of the group are checked, a `o' means to
2334 test those of other users, and the `a' says to test all three
2335 groups. The `=', `+', and `-' again says how the modes are to be
2336 checked and have the same meaning as described for the first
2337 form above. The second list of characters finally says which
2338 access rights are to be expected: `r' for read access, `w' for
2339 write access, `x' for the right to execute the file (or to
2340 search a directory), `s' for the setuid and setgid bits, and `t'
2341 for the sticky bit.
2342 Thus, `*(f70?)' gives the files for which the owner has read,
2344 write, and execute permission, and for which other group members
2345 have no rights, independent of the permissions for other users.
2346 The pattern `*(f-100)' gives all files for which the owner does
2347 not have execute permission, and `*(f:gu+w,o-rx:)' gives the
2348 files for which the owner and the other members of the group
2349 have at least write permission, and for which other users don't
2350 have read or execute permission.
2351 estring
2353 +cmd The string will be executed as shell code. The filename will be
2354 included in the list if and only if the code returns a zero sta‐
2355 tus (usually the status of the last command).
2356 In the first form, the first character after the `e' will be
2358 used as a separator and anything up to the next matching separa‐
2359 tor will be taken as the string; `[', `{', and `<' match `]',
2360 `}', and `>', respectively, while any other character matches
2361 itself. Note that expansions must be quoted in the string to
2362 prevent them from being expanded before globbing is done.
2363 string is then executed as shell code. The string globqual is
2364 appended to the array zsh_eval_context the duration of execu‐
2365 tion.
2366 During the execution of string the filename currently being
2368 tested is available in the parameter REPLY; the parameter may be
2369 altered to a string to be inserted into the list instead of the
2370 original filename. In addition, the parameter reply may be set
2371 to an array or a string, which overrides the value of REPLY. If
2372 set to an array, the latter is inserted into the command line
2373 word by word.
2374 For example, suppose a directory contains a single file
2376 `lonely'. Then the expression `*(e:'reply=(${REPLY}{1,2})':)'
2377 will cause the words `lonely1' and `lonely2' to be inserted into
2378 the command line. Note the quoting of string.
2379 The form +cmd has the same effect, but no delimiters appear
2381 around cmd. Instead, cmd is taken as the longest sequence of
2382 characters following the + that are alphanumeric or underscore.
2383 Typically cmd will be the name of a shell function that contains
2384 the appropriate test. For example,
2385 nt() { [[ $REPLY -nt $NTREF ]] }
2387 NTREF=reffile
2388 ls -ld -- *(+nt)
2389 lists all files in the directory that have been modified more
2391 recently than reffile.
2392 ddev files on the device dev
2394 l[-|+]ct
2396 files having a link count less than ct (-), greater than ct (+),
2397 or equal to ct
2398 U files owned by the effective user ID
2400 G files owned by the effective group ID
2402 uid files owned by user ID id if that is a number. Otherwise, id
2404 specifies a user name: the character after the `u' will be taken
2405 as a separator and the string between it and the next matching
2406 separator will be taken as a user name. The starting separators
2407 `[', `{', and `<' match the final separators `]', `}', and `>',
2408 respectively; any other character matches itself. The selected
2409 files are those owned by this user. For example, `u:foo:' or
2410 `u[foo]' selects files owned by user `foo'.
2411 gid like uid but with group IDs or names
2413 a[Mwhms][-|+]n
2415 files accessed exactly n days ago. Files accessed within the
2416 last n days are selected using a negative value for n (-n).
2417 Files accessed more than n days ago are selected by a positive n
2418 value (+n). Optional unit specifiers `M', `w', `h', `m' or `s'
2419 (e.g. `ah5') cause the check to be performed with months (of 30
2420 days), weeks, hours, minutes or seconds instead of days, respec‐
2421 tively. An explicit `d' for days is also allowed.
2422 Any fractional part of the difference between the access time
2424 and the current part in the appropriate units is ignored in the
2425 comparison. For instance, `echo *(ah-5)' would echo files
2426 accessed within the last five hours, while `echo *(ah+5)' would
2427 echo files accessed at least six hours ago, as times strictly
2428 between five and six hours are treated as five hours.
2429 m[Mwhms][-|+]n
2431 like the file access qualifier, except that it uses the file
2432 modification time.
2433 c[Mwhms][-|+]n
2435 like the file access qualifier, except that it uses the file
2436 inode change time.
2437 L[+|-]n
2439 files less than n bytes (-), more than n bytes (+), or exactly n
2440 bytes in length.
2441 If this flag is directly followed by a size specifier `k' (`K'),
2443 `m' (`M'), or `p' (`P') (e.g. `Lk-50') the check is performed
2444 with kilobytes, megabytes, or blocks (of 512 bytes) instead.
2445 (On some systems additional specifiers are available for giga‐
2446 bytes, `g' or `G', and terabytes, `t' or `T'.) If a size speci‐
2447 fier is used a file is regarded as "exactly" the size if the
2448 file size rounded up to the next unit is equal to the test size.
2449 Hence `*(Lm1)' matches files from 1 byte up to 1 Megabyte inclu‐
2450 sive. Note also that the set of files "less than" the test size
2451 only includes files that would not match the equality test;
2452 hence `*(Lm-1)' only matches files of zero size.
2453 ^ negates all qualifiers following it
2455 - toggles between making the qualifiers work on symbolic links
2457 (the default) and the files they point to
2458 M sets the MARK_DIRS option for the current pattern
2460 T appends a trailing qualifier mark to the filenames, analogous to
2462 the LIST_TYPES option, for the current pattern (overrides M)
2463 N sets the NULL_GLOB option for the current pattern
2465 D sets the GLOB_DOTS option for the current pattern
2467 n sets the NUMERIC_GLOB_SORT option for the current pattern
2469 Yn enables short-circuit mode: the pattern will expand to at most n
2471 filenames. If more than n matches exist, only the first n
2472 matches in directory traversal order will be considered.
2473 Implies oN when no oc qualifier is used.
2475 oc specifies how the names of the files should be sorted. If c is n
2477 they are sorted by name; if it is L they are sorted depending on
2478 the size (length) of the files; if l they are sorted by the num‐
2479 ber of links; if a, m, or c they are sorted by the time of the
2480 last access, modification, or inode change respectively; if d,
2481 files in subdirectories appear before those in the current
2482 directory at each level of the search -- this is best combined
2483 with other criteria, for example `odon' to sort on names for
2484 files within the same directory; if N, no sorting is performed.
2485 Note that a, m, and c compare the age against the current time,
2486 hence the first name in the list is the youngest file. Also note
2487 that the modifiers ^ and - are used, so `*(^-oL)' gives a list
2488 of all files sorted by file size in descending order, following
2489 any symbolic links. Unless oN is used, multiple order speci‐
2490 fiers may occur to resolve ties.
2491 The default sorting is n (by name) unless the Y glob qualifier
2493 is used, in which case it is N (unsorted).
2494 oe and o+ are special cases; they are each followed by shell
2496 code, delimited as for the e glob qualifier and the + glob qual‐
2497 ifier respectively (see above). The code is executed for each
2498 matched file with the parameter REPLY set to the name of the
2499 file on entry and globsort appended to zsh_eval_context. The
2500 code should modify the parameter REPLY in some fashion. On
2501 return, the value of the parameter is used instead of the file
2502 name as the string on which to sort. Unlike other sort opera‐
2503 tors, oe and o+ may be repeated, but note that the maximum num‐
2504 ber of sort operators of any kind that may appear in any glob
2505 expression is 12.
2506 Oc like `o', but sorts in descending order; i.e. `*(^oc)' is the
2508 same as `*(Oc)' and `*(^Oc)' is the same as `*(oc)'; `Od' puts
2509 files in the current directory before those in subdirectories at
2510 each level of the search.
2511 [beg[,end]]
2513 specifies which of the matched filenames should be included in
2514 the returned list. The syntax is the same as for array sub‐
2515 scripts. beg and the optional end may be mathematical expres‐
2516 sions. As in parameter subscripting they may be negative to make
2517 them count from the last match backward. E.g.: `*(-OL[1,3])'
2518 gives a list of the names of the three largest files.
2519 Pstring
2521 The string will be prepended to each glob match as a separate
2522 word. string is delimited in the same way as arguments to the e
2523 glob qualifier described above. The qualifier can be repeated;
2524 the words are prepended separately so that the resulting command
2525 line contains the words in the same order they were given in the
2526 list of glob qualifiers.
2527 A typical use for this is to prepend an option before all occur‐
2529 rences of a file name; for example, the pattern `*(P:-f:)' pro‐
2530 duces the command line arguments `-f file1 -f file2 ...'
2531 If the modifier ^ is active, then string will be appended
2533 instead of prepended. Prepending and appending is done indepen‐
2534 dently so both can be used on the same glob expression; for
2535 example by writing `*(P:foo:^P:bar:^P:baz:)' which produces the
2536 command line arguments `foo baz file1 bar ...'
2537 More than one of these lists can be combined, separated by commas. The
2539 whole list matches if at least one of the sublists matches (they are
2540 `or'ed, the qualifiers in the sublists are `and'ed). Some qualifiers,
2541 however, affect all matches generated, independent of the sublist in
2542 which they are given. These are the qualifiers `M', `T', `N', `D',
2543 `n', `o', `O' and the subscripts given in brackets (`[...]').
2544 If a `:' appears in a qualifier list, the remainder of the expression
2546 in parenthesis is interpreted as a modifier (see the section `Modi‐
2547 fiers' in the section `History Expansion'). Each modifier must be
2548 introduced by a separate `:'. Note also that the result after modifi‐
2549 cation does not have to be an existing file. The name of any existing
2550 file can be followed by a modifier of the form `(:...)' even if no
2551 actual filename generation is performed, although note that the pres‐
2552 ence of the parentheses causes the entire expression to be subjected to
2553 any global pattern matching options such as NULL_GLOB. Thus:
2554 ls -ld -- *(-/)
2556 lists all directories and symbolic links that point to directories, and
2558 ls -ld -- *(-@)
2560 lists all broken symbolic links, and
2562 ls -ld -- *(%W)
2564 lists all world-writable device files in the current directory, and
2566 ls -ld -- *(W,X)
2568 lists all files in the current directory that are world-writable or
2570 world-executable, and
2571 print -rC1 /tmp/foo*(u0^@:t)
2573 outputs the basename of all root-owned files beginning with the string
2575 `foo' in /tmp, ignoring symlinks, and
2576 ls -ld -- *.*~(lex|parse).[ch](^D^l1)
2578 lists all files having a link count of one whose names contain a dot
2580 (but not those starting with a dot, since GLOB_DOTS is explicitly
2581 switched off) except for lex.c, lex.h, parse.c and parse.h.
2582 print -rC1 b*.pro(#q:s/pro/shmo/)(#q.:s/builtin/shmiltin/)
2584 demonstrates how colon modifiers and other qualifiers may be chained
2586 together. The ordinary qualifier `.' is applied first, then the colon
2587 modifiers in order from left to right. So if EXTENDED_GLOB is set and
2588 the base pattern matches the regular file builtin.pro, the shell will
2589 print `shmiltin.shmo'.
2590zsh 5.8 February 14, 2020 ZSHEXPN(1)