zsh-manual-mdbook/zsh_manual/book/Expansion.html
2021-05-17 10:18:05 -05:00

3293 lines
168 KiB
HTML
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

<!DOCTYPE HTML>
<html lang="en" class="sidebar-visible no-js light">
<head>
<!-- Book generated using mdBook -->
<meta charset="UTF-8">
<title>Expansion - Zsh Manual</title>
<!-- Custom HTML head -->
<meta content="text/html; charset=utf-8" http-equiv="Content-Type">
<meta name="description" content="">
<meta name="viewport" content="width=device-width, initial-scale=1">
<meta name="theme-color" content="#ffffff" />
<link rel="icon" href="favicon.svg">
<link rel="shortcut icon" href="favicon.png">
<link rel="stylesheet" href="css/variables.css">
<link rel="stylesheet" href="css/general.css">
<link rel="stylesheet" href="css/chrome.css">
<link rel="stylesheet" href="css/print.css" media="print">
<!-- Fonts -->
<link rel="stylesheet" href="FontAwesome/css/font-awesome.css">
<link rel="stylesheet" href="fonts/fonts.css">
<!-- Highlight.js Stylesheets -->
<link rel="stylesheet" href="highlight.css">
<link rel="stylesheet" href="tomorrow-night.css">
<link rel="stylesheet" href="ayu-highlight.css">
<!-- Custom theme stylesheets -->
</head>
<body>
<!-- Provide site root to javascript -->
<script type="text/javascript">
var path_to_root = "";
var default_theme = window.matchMedia("(prefers-color-scheme: dark)").matches ? "navy" : "light";
</script>
<!-- Work around some values being stored in localStorage wrapped in quotes -->
<script type="text/javascript">
try {
var theme = localStorage.getItem('mdbook-theme');
var sidebar = localStorage.getItem('mdbook-sidebar');
if (theme.startsWith('"') && theme.endsWith('"')) {
localStorage.setItem('mdbook-theme', theme.slice(1, theme.length - 1));
}
if (sidebar.startsWith('"') && sidebar.endsWith('"')) {
localStorage.setItem('mdbook-sidebar', sidebar.slice(1, sidebar.length - 1));
}
} catch (e) { }
</script>
<!-- Set the theme before any content is loaded, prevents flash -->
<script type="text/javascript">
var theme;
try { theme = localStorage.getItem('mdbook-theme'); } catch(e) { }
if (theme === null || theme === undefined) { theme = default_theme; }
var html = document.querySelector('html');
html.classList.remove('no-js')
html.classList.remove('light')
html.classList.add(theme);
html.classList.add('js');
</script>
<!-- Hide / unhide sidebar before it is displayed -->
<script type="text/javascript">
var html = document.querySelector('html');
var sidebar = 'hidden';
if (document.body.clientWidth >= 1080) {
try { sidebar = localStorage.getItem('mdbook-sidebar'); } catch(e) { }
sidebar = sidebar || 'visible';
}
html.classList.remove('sidebar-visible');
html.classList.add("sidebar-" + sidebar);
</script>
<nav id="sidebar" class="sidebar" aria-label="Table of contents">
<div class="sidebar-scrollbox">
<ol class="chapter"><li class="chapter-item expanded "><a href="The-Z-Shell-Manual.html"><strong aria-hidden="true">1.</strong> The Z Shell Manual</a></li><li class="chapter-item expanded "><a href="Introduction.html"><strong aria-hidden="true">2.</strong> Introduction</a></li><li class="chapter-item expanded "><a href="Roadmap.html"><strong aria-hidden="true">3.</strong> Roadmap</a></li><li class="chapter-item expanded "><a href="Invocation.html"><strong aria-hidden="true">4.</strong> Invocation</a></li><li class="chapter-item expanded "><a href="Files.html"><strong aria-hidden="true">5.</strong> Files</a></li><li class="chapter-item expanded "><a href="Shell-Grammar.html"><strong aria-hidden="true">6.</strong> Shell Grammar</a></li><li class="chapter-item expanded "><a href="Redirection.html"><strong aria-hidden="true">7.</strong> Redirection</a></li><li class="chapter-item expanded "><a href="Command-Execution.html"><strong aria-hidden="true">8.</strong> Command Execution</a></li><li class="chapter-item expanded "><a href="Functions.html"><strong aria-hidden="true">9.</strong> Functions</a></li><li class="chapter-item expanded "><a href="Jobs-_0026-Signals.html"><strong aria-hidden="true">10.</strong> Jobs &amp; Signals</a></li><li class="chapter-item expanded "><a href="Arithmetic-Evaluation.html"><strong aria-hidden="true">11.</strong> Arithmetic Evaluation</a></li><li class="chapter-item expanded "><a href="Conditional-Expressions.html"><strong aria-hidden="true">12.</strong> Conditional Expressions</a></li><li class="chapter-item expanded "><a href="Prompt-Expansion.html"><strong aria-hidden="true">13.</strong> Prompt Expansion</a></li><li class="chapter-item expanded "><a href="Expansion.html" class="active"><strong aria-hidden="true">14.</strong> Expansion</a></li><li class="chapter-item expanded "><a href="Parameters.html"><strong aria-hidden="true">15.</strong> Parameters</a></li><li class="chapter-item expanded "><a href="Options.html"><strong aria-hidden="true">16.</strong> Options</a></li><li class="chapter-item expanded "><a href="Shell-Builtin-Commands.html"><strong aria-hidden="true">17.</strong> Shell Builtin Commands</a></li><li class="chapter-item expanded "><a href="Zsh-Line-Editor.html"><strong aria-hidden="true">18.</strong> Zsh Line Editor</a></li><li class="chapter-item expanded "><a href="Completion-Widgets.html"><strong aria-hidden="true">19.</strong> Completion Widgets</a></li><li class="chapter-item expanded "><a href="Completion-System.html"><strong aria-hidden="true">20.</strong> Completion System</a></li><li class="chapter-item expanded "><a href="Completion-Using-compctl.html"><strong aria-hidden="true">21.</strong> Completion Using compctl</a></li><li class="chapter-item expanded "><a href="Zsh-Modules.html"><strong aria-hidden="true">22.</strong> Zsh Modules</a></li><li class="chapter-item expanded "><a href="Calendar-Function-System.html"><strong aria-hidden="true">23.</strong> Calendar Function System</a></li><li class="chapter-item expanded "><a href="TCP-Function-System.html"><strong aria-hidden="true">24.</strong> TCP Function System</a></li><li class="chapter-item expanded "><a href="Zftp-Function-System.html"><strong aria-hidden="true">25.</strong> Zftp Function System</a></li><li class="chapter-item expanded "><a href="User-Contributions.html"><strong aria-hidden="true">26.</strong> User Contributions</a></li></ol>
</div>
<div id="sidebar-resize-handle" class="sidebar-resize-handle"></div>
</nav>
<div id="page-wrapper" class="page-wrapper">
<div class="page">
<div id="menu-bar-hover-placeholder"></div>
<div id="menu-bar" class="menu-bar sticky bordered">
<div class="left-buttons">
<button id="sidebar-toggle" class="icon-button" type="button" title="Toggle Table of Contents" aria-label="Toggle Table of Contents" aria-controls="sidebar">
<i class="fa fa-bars"></i>
</button>
<button id="theme-toggle" class="icon-button" type="button" title="Change theme" aria-label="Change theme" aria-haspopup="true" aria-expanded="false" aria-controls="theme-list">
<i class="fa fa-paint-brush"></i>
</button>
<ul id="theme-list" class="theme-popup" aria-label="Themes" role="menu">
<li role="none"><button role="menuitem" class="theme" id="light">Light (default)</button></li>
<li role="none"><button role="menuitem" class="theme" id="rust">Rust</button></li>
<li role="none"><button role="menuitem" class="theme" id="coal">Coal</button></li>
<li role="none"><button role="menuitem" class="theme" id="navy">Navy</button></li>
<li role="none"><button role="menuitem" class="theme" id="ayu">Ayu</button></li>
</ul>
<button id="search-toggle" class="icon-button" type="button" title="Search. (Shortkey: s)" aria-label="Toggle Searchbar" aria-expanded="false" aria-keyshortcuts="S" aria-controls="searchbar">
<i class="fa fa-search"></i>
</button>
</div>
<h1 class="menu-title">Zsh Manual</h1>
<div class="right-buttons">
<a href="print.html" title="Print this book" aria-label="Print this book">
<i id="print-button" class="fa fa-print"></i>
</a>
</div>
</div>
<div id="search-wrapper" class="hidden">
<form id="searchbar-outer" class="searchbar-outer">
<input type="search" id="searchbar" name="searchbar" placeholder="Search this book ..." aria-controls="searchresults-outer" aria-describedby="searchresults-header">
</form>
<div id="searchresults-outer" class="searchresults-outer hidden">
<div id="searchresults-header" class="searchresults-header"></div>
<ul id="searchresults">
</ul>
</div>
</div>
<!-- Apply ARIA attributes after the sidebar and the sidebar toggle button are added to the DOM -->
<script type="text/javascript">
document.getElementById('sidebar-toggle').setAttribute('aria-expanded', sidebar === 'visible');
document.getElementById('sidebar').setAttribute('aria-hidden', sidebar !== 'visible');
Array.from(document.querySelectorAll('#sidebar a')).forEach(function(link) {
link.setAttribute('tabIndex', sidebar === 'visible' ? 0 : -1);
});
</script>
<div id="content" class="content">
<main>
<!-- START doctoc generated TOC please keep comment here to allow auto update -->
<!-- DON'T EDIT THIS SECTION, INSTEAD RE-RUN doctoc TO UPDATE -->
<p><strong>Table of Contents</strong> <em>generated with <a href="https://github.com/thlorenz/doctoc">DocToc</a></em></p>
<ul>
<li><a href="#14-expansion">14 Expansion</a>
<ul>
<li><a href="#141-history-expansion">14.1 History Expansion</a>
<ul>
<li><a href="#1411-overview">14.1.1 Overview</a></li>
<li><a href="#1412-event-designators">14.1.2 Event Designators</a></li>
<li><a href="#1413-word-designators">14.1.3 Word Designators</a></li>
<li><a href="#1414-modifiers">14.1.4 Modifiers</a></li>
</ul>
</li>
<li><a href="#142-process-substitution">14.2 Process Substitution</a></li>
<li><a href="#143-parameter-expansion">14.3 Parameter Expansion</a>
<ul>
<li><a href="#1431-parameter-expansion-flags">14.3.1 Parameter Expansion Flags</a></li>
<li><a href="#1432-rules">14.3.2 Rules</a></li>
<li><a href="#1433-examples">14.3.3 Examples</a></li>
</ul>
</li>
<li><a href="#144-command-substitution">14.4 Command Substitution</a></li>
<li><a href="#145-arithmetic-expansion">14.5 Arithmetic Expansion</a></li>
<li><a href="#146-brace-expansion">14.6 Brace Expansion</a></li>
<li><a href="#147-filename-expansion">14.7 Filename Expansion</a>
<ul>
<li><a href="#1471-dynamic-named-directories">14.7.1 Dynamic named directories</a></li>
<li><a href="#1472-static-named-directories">14.7.2 Static named directories</a></li>
<li><a href="#1473--expansion">14.7.3 = expansion</a></li>
<li><a href="#1474-notes">14.7.4 Notes</a></li>
</ul>
</li>
<li><a href="#148-filename-generation">14.8 Filename Generation</a>
<ul>
<li><a href="#1481-glob-operators">14.8.1 Glob Operators</a></li>
<li><a href="#1482-ksh-like-glob-operators">14.8.2 ksh-like Glob Operators</a></li>
<li><a href="#1483-precedence">14.8.3 Precedence</a></li>
<li><a href="#1484-globbing-flags">14.8.4 Globbing Flags</a></li>
<li><a href="#1485-approximate-matching">14.8.5 Approximate Matching</a></li>
<li><a href="#1486-recursive-globbing">14.8.6 Recursive Globbing</a></li>
<li><a href="#1487-glob-qualifiers">14.8.7 Glob Qualifiers</a></li>
</ul>
</li>
</ul>
</li>
</ul>
<!-- END doctoc generated TOC please keep comment here to allow auto update -->
<p><span id="Expansion"></span> <span id="Expansion-1"></span></p>
<h1 id="14-expansion"><a class="header" href="#14-expansion">14 Expansion</a></h1>
<p><span id="index-expansion"></span></p>
<p>The following types of expansions are performed in the indicated order
in five steps:</p>
<ul>
<li>
<p><em>History Expansion</em><br />
This is performed only in interactive shells.</p>
</li>
<li>
<p><em>Alias Expansion</em><br />
Aliases are expanded immediately before the command line is parsed
as explained in <a href="Shell-Grammar.html#Aliasing">Aliasing</a>.</p>
</li>
<li>
<p><em>Process Substitution</em><br />
<em>Parameter Expansion</em><br />
<em>Command Substitution</em><br />
<em>Arithmetic Expansion</em><br />
<em>Brace Expansion</em><br />
These five are performed in left-to-right fashion. On each argument,
any of the five steps that are needed are performed one after the
other. Hence, for example, all the parts of parameter expansion are
completed before command substitution is started. After these
expansions, all unquoted occurrences of the characters <code>\</code>,<code></code>
and <code>&quot;</code> are removed.</p>
</li>
<li>
<p><em>Filename Expansion</em><br />
If the <code>SH_FILE_EXPANSION</code> option is set, the order of expansion is
modified for compatibility with sh and ksh. In that case <em>filename
expansion</em> is performed immediately after <em>alias expansion</em>,
preceding the set of five expansions mentioned above.</p>
<p><span id="index-globbing"></span></p>
</li>
<li>
<p><em>Filename Generation</em><br />
This expansion, commonly referred to as globbing, is always done
last.</p>
</li>
</ul>
<p>The following sections explain the types of expansion in detail.</p>
<hr />
<p><span id="History-Expansion"></span>
<span id="History-Expansion-1"></span></p>
<h2 id="141-history-expansion"><a class="header" href="#141-history-expansion">14.1 History Expansion</a></h2>
<p><span id="index-history"></span>
<span id="index-history-expansion"></span>
<span id="index-expansion_002c-history"></span></p>
<p>History expansion allows you to use words from previous command lines in
the command line you are typing. This simplifies spelling corrections
and the repetition of complicated commands or arguments.</p>
<p><span id="index-HISTSIZE_002c-use-of"></span></p>
<p>Immediately before execution, each command is saved in the history list,
the size of which is controlled by the <code>HISTSIZE</code> parameter. The one
most recent command is always retained in any case. Each saved command
in the history list is called a history <em>event</em> and is assigned a
number, beginning with 1 (one) when the shell starts up. The history
number that you may see in your prompt (see <a href="Prompt-Expansion.html#Prompt-Expansion">Prompt
Expansion</a>) is the number that
is to be assigned to the <em>next</em> command.</p>
<hr />
<p><span id="Overview"></span> <span id="Overview-2"></span></p>
<h3 id="1411-overview"><a class="header" href="#1411-overview">14.1.1 Overview</a></h3>
<p><span id="index-histchars_002c-use-of-1"></span></p>
<p>A history expansion begins with the first character of the <code>histchars</code>
parameter, which is <code>!</code> by default, and may occur anywhere on the
command line, including inside double quotes (but not inside single
quotes <code>...</code> or C-style quotes <code>$...</code> nor when escaped with a
backslash).</p>
<p>The first character is followed by an optional event designator (<a href="#Event-Designators">Event
Designators</a>) and then an optional word designator
(<a href="#Word-Designators">Word Designators</a>); if neither of these designators
is present, no history expansion occurs.</p>
<p>Input lines containing history expansions are echoed after being
expanded, but before any other expansions take place and before the
command is executed. It is this expanded form that is recorded as the
history event for later references.</p>
<p>History expansions do not nest.</p>
<p>By default, a history reference with no event designator refers to the
same event as any preceding history reference on that command line; if
it is the only history reference in a command, it refers to the previous
command. <span id="index-CSH_005fJUNKIE_005fHISTORY_002c-use-of"></span>
However, if the option <code>CSH_JUNKIE_HISTORY</code> is set, then every history
reference with no event specification <em>always</em> refers to the previous
command.</p>
<p>For example, <code>!</code> is the event designator for the previous command, so
<code>!!:1</code> always refers to the first word of the previous command, and
<code>!!$</code> always refers to the last word of the previous command. With
<code>CSH_JUNKIE_HISTORY</code> set, then <code>!:1</code> and <code>!$</code> function in the same
manner as <code>!!:1</code> and <code>!!$</code>, respectively. Conversely, if
<code>CSH_JUNKIE_HISTORY</code> is unset, then <code>!:1</code> and <code>!$</code> refer to the
first and last words, respectively, of the same event referenced by the
nearest other history reference preceding them on the current command
line, or to the previous command if there is no preceding reference.</p>
<p>The character sequence <code>^``foo``^``bar</code> (where <code>^</code> is actually the
second character of the <code>histchars</code> parameter) repeats the last command,
replacing the string <code>foo</code> with <code>bar</code>. More precisely, the sequence
<code>^``foo``^``bar``^</code> is synonymous with <code>!!:s``^``foo``^``bar``^</code>,
hence other modifiers (see <a href="#Modifiers">Modifiers</a>) may follow the final
<code>^</code>. In particular, <code>^``foo``^``bar``^:G</code> performs a global
substitution.</p>
<p>If the shell encounters the character sequence <code>!&quot;</code> in the input, the
history mechanism is temporarily disabled until the current list (see
<a href="Shell-Grammar.html#Shell-Grammar">Shell Grammar</a>) is fully parsed. The
<code>!&quot;</code> is removed from the input, and any subsequent <code>!</code> characters
have no special significance.</p>
<p><span id="index-fc_002c-use-of"></span></p>
<p>A less convenient but more comprehensible form of command history
support is provided by the <code>fc</code> builtin.</p>
<hr />
<p><span id="Event-Designators"></span>
<span id="Event-Designators-1"></span></p>
<h3 id="1412-event-designators"><a class="header" href="#1412-event-designators">14.1.2 Event Designators</a></h3>
<p><span id="index-history-event-designators"></span>
<span id="index-event-designators_002c-history"></span></p>
<p>An event designator is a reference to a command-line entry in the
history list. In the list below, remember that the initial <code>!</code> in each
item may be changed to another character by setting the <code>histchars</code>
parameter.</p>
<ul>
<li>
<p><code>!</code><br />
Start a history expansion, except when followed by a blank, newline,
<code>=</code> or <code>(</code>. If followed immediately by a word designator (<a href="#Word-Designators">Word
Designators</a>), this forms a history reference
with no event designator (<a href="#Overview">Overview</a>).</p>
</li>
<li>
<p><code>!!</code><br />
Refer to the previous command. By itself, this expansion repeats the
previous command.</p>
</li>
<li>
<p><code>!``n</code><br />
Refer to command-line <code>n</code>.</p>
</li>
<li>
<p><code>!-``n</code><br />
Refer to the current command-line minus <code>n</code>.</p>
</li>
<li>
<p><code>!``str</code><br />
Refer to the most recent command starting with <code>str</code>.</p>
</li>
<li>
<p><code>!?``str</code>[<code>?</code>]<br />
Refer to the most recent command containing <code>str</code>. The trailing
<code>?</code> is necessary if this reference is to be followed by a
modifier or followed by any text that is not to be considered part
of <code>str</code>.</p>
</li>
<li>
<p><code>!#</code><br />
Refer to the current command line typed in so far. The line is
treated as if it were complete up to and including the word before
the one with the <code>!#</code> reference.</p>
</li>
<li>
<p><code>!{</code>...<code>}</code><br />
Insulate a history reference from adjacent characters (if
necessary).</p>
</li>
</ul>
<hr />
<p><span id="Word-Designators"></span>
<span id="Word-Designators-1"></span></p>
<h3 id="1413-word-designators"><a class="header" href="#1413-word-designators">14.1.3 Word Designators</a></h3>
<p><span id="index-history-word-designators"></span>
<span id="index-word-designators_002c-history"></span></p>
<p>A word designator indicates which word or words of a given command line
are to be included in a history reference. A <code>:</code> usually separates the
event specification from the word designator. It may be omitted only if
the word designator begins with a <code>^</code>, <code>$</code>, <code>*</code>, <code>-</code> or <code>%</code>.
Word designators include:</p>
<ul>
<li>
<p><code>0</code><br />
The first input word (command).</p>
</li>
<li>
<p><code>n</code><br />
The <code>n</code>th argument.</p>
</li>
<li>
<p><code>^</code><br />
The first argument. That is, <code>1</code>.</p>
</li>
<li>
<p><code>$</code><br />
The last argument.</p>
</li>
<li>
<p><code>%</code><br />
The word matched by (the most recent) <code>?``str</code> search.</p>
</li>
<li>
<p><code>x``-``y</code><br />
A range of words; <code>x</code> defaults to <code>0</code>.</p>
</li>
<li>
<p><code>*</code><br />
All the arguments, or a null value if there are none.</p>
</li>
<li>
<p><code>x``*</code><br />
Abbreviates <code>x``-$</code>.</p>
</li>
<li>
<p><code>x``-</code><br />
Like <code>x``*</code> but omitting word <code>$</code>.</p>
</li>
</ul>
<p>Note that a <code>%</code> word designator works only when used in one of <code>!%</code>,
<code>!:%</code> or <code>!?``str``?:%</code>, and only when used after a <code>!?</code> expansion
(possibly in an earlier command). Anything else results in an error,
although the error may not be the most obvious one.</p>
<hr />
<p><span id="Modifiers"></span> <span id="Modifiers-1"></span></p>
<h3 id="1414-modifiers"><a class="header" href="#1414-modifiers">14.1.4 Modifiers</a></h3>
<p><span id="index-modifiers"></span>
<span id="index-colon-modifiers"></span>
<span id="index-history-modifiers"></span>
<span id="index-globbing-modifiers"></span>
<span id="index-parameter-modifiers"></span></p>
<p>After the optional word designator, you can add a sequence of one or
more of the following modifiers, each preceded by a <code>:</code>. These
modifiers also work on the result of <em>filename generation</em> and
<em>parameter expansion</em>, except where noted.</p>
<ul>
<li>
<p><code>a</code><br />
Turn a file name into an absolute path: prepends the current
directory, if necessary; remove <code>.</code> path segments; and remove
<code>..</code> path segments and the segments that immediately precede
them.</p>
<p>This transformation is agnostic about what is in the filesystem,
i.e. is on the logical, not the physical directory. It takes place
in the same manner as when changing directories when neither of the
options <code>CHASE_DOTS</code> or <code>CHASE_LINKS</code> is set. For example,
<code>/before/here/../after</code> is always transformed to
<code>/before/after</code>, regardless of whether <code>/before/here</code> exists
or what kind of object (dir, file, symlink, etc.) it is.</p>
</li>
<li>
<p><code>A</code><br />
Turn a file name into an absolute path as the <code>a</code> modifier does,
and <em>then</em> pass the result through the <code>realpath(3)</code> library
function to resolve symbolic links.</p>
<p>Note: on systems that do not have a <code>realpath(3)</code> library function,
symbolic links are not resolved, so on those systems <code>a</code> and <code>A</code>
are equivalent.</p>
<p>Note: <code>foo:A</code> and <code>realpath(foo)</code> are different on some inputs. For
<code>realpath(foo)</code> semantics, see the <code>P</code> modifier.</p>
</li>
<li>
<p><code>c</code><br />
Resolve a command name into an absolute path by searching the
command path given by the <code>PATH</code> variable. This does not work for
commands containing directory parts. Note also that this does not
usually work as a glob qualifier unless a file of the same name is
found in the current directory.</p>
</li>
<li>
<p><code>e</code><br />
Remove all but the part of the filename extension following the
<code>.</code>; see the definition of the filename extension in the
description of the <code>r</code> modifier below. Note that according to that
definition the result will be empty if the string ends with a <code>.</code>.</p>
</li>
<li>
<p><code>h</code> [ <code>digits</code> ]<br />
Remove a trailing pathname component, shortening the path by one
directory level: this is the head of the pathname. This works like
<code>dirname</code>. If the <code>h</code> is followed immediately (with no spaces or
other separator) by any number of decimal digits, and the value of
the resulting number is non-zero, that number of leading components
is preserved instead of the final component being removed. In an
absolute path the leading <code>/</code> is the first component, so, for
example, if <code>var=/my/path/to/something</code>, then <code>${var:h3}</code>
substitutes <code>/my/path</code>. Consecutive /s are treated the same as a
single /. In parameter substitution, digits may only be used if
the expression is in braces, so for example the short form
substitution <code>$var:h2</code> is treated as <code>${var:h}2</code>, not as
<code>${var:h2}</code>. No restriction applies to the use of digits in history
substitution or globbing qualifiers. If more components are
requested than are present, the entire path is substituted (so this
does not trigger a failed modifier error in history expansion).</p>
</li>
<li>
<p><code>l</code><br />
Convert the words to all lowercase.</p>
</li>
<li>
<p><code>p</code><br />
Print the new command but do not execute it. Only works with history
expansion.</p>
</li>
<li>
<p><code>P</code><br />
Turn a file name into an absolute path, like <code>realpath(3)</code>. The
resulting path will be absolute, have neither <code>.</code> nor <code>..</code>
components, and refer to the same directory entry as the input
filename.</p>
<p>Unlike <code>realpath(3)</code>, non-existent trailing components are permitted
and preserved.</p>
</li>
<li>
<p><code>q</code><br />
Quote the substituted words, escaping further substitutions. Works
with history expansion and parameter expansion, though for
parameters it is only useful if the resulting text is to be
re-evaluated such as by <code>eval</code>.</p>
</li>
<li>
<p><code>Q</code><br />
Remove one level of quotes from the substituted words.</p>
</li>
<li>
<p><code>r</code><br />
Remove a filename extension leaving the root name. Strings with no
filename extension are not altered. A filename extension is a <code>.</code>
followed by any number of characters (including zero) that are
neither <code>.</code> nor <code>/</code> and that continue to the end of the string.
For example, the extension of <code>foo.orig.c</code> is <code>.c</code>, and
<code>dir.c/foo</code> has no extension.</p>
</li>
<li>
<p><code>s/``l``/``r</code>[<code>/</code>]<br />
Substitute <code>r</code> for <code>l</code> as described below. The substitution is done
only for the first string that matches <code>l</code>. For arrays and for
filename generation, this applies to each word of the expanded text.
See below for further notes on substitutions.</p>
<p>The forms <code>gs/``l``/``r</code> and <code>s/``l``/``r``/:G</code> perform global
substitution, i.e. substitute every occurrence of <code>r</code> for <code>l</code>. Note
that the <code>g</code> or <code>:G</code> must appear in exactly the position shown.</p>
<p>See further notes on this form of substitution below.</p>
</li>
<li>
<p><code>&amp;</code><br />
Repeat the previous <code>s</code> substitution. Like <code>s</code>, may be preceded
immediately by a <code>g</code>. In parameter expansion the <code>&amp;</code> must appear
inside braces, and in filename generation it must be quoted with a
backslash.</p>
</li>
<li>
<p><code>t</code> [ <code>digits</code> ]<br />
Remove all leading pathname components, leaving the final component
(tail). This works like <code>basename</code>. Any trailing slashes are first
removed. Decimal digits are handled as described above for (h), but
in this case that number of trailing components is preserved instead
of the default 1; 0 is treated the same as 1.</p>
</li>
<li>
<p><code>u</code><br />
Convert the words to all uppercase.</p>
</li>
<li>
<p><code>x</code><br />
Like <code>q</code>, but break into words at whitespace. Does not work with
parameter expansion.</p>
</li>
</ul>
<p>The <code>s/``l``/``r``/</code> substitution works as follows. By default the
left-hand side of substitutions are not patterns, but character strings.
Any character can be used as the delimiter in place of <code>/</code>. A
backslash quotes the delimiter character. The character <code>&amp;</code>, in the
right-hand-side <code>r</code>, is replaced by the text from the left-hand-side
<code>l</code>. The <code>&amp;</code> can be quoted with a backslash. A null <code>l</code> uses the
previous string either from the previous <code>l</code> or from the contextual scan
string <code>s</code> from <code>!?``s</code>. You can omit the rightmost delimiter if a
newline immediately follows <code>r</code>; the rightmost <code>?</code> in a context scan
can similarly be omitted. Note the same record of the last <code>l</code> and <code>r</code>
is maintained across all forms of expansion.</p>
<p>Note that if a <code>&amp;</code> is used within glob qualifiers an extra backslash
is needed as a <code>&amp;</code> is a special character in this case.</p>
<p>Also note that the order of expansions affects the interpretation of <code>l</code>
and <code>r</code>. When used in a history expansion, which occurs before any other
expansions, <code>l</code> and <code>r</code> are treated as literal strings (except as
explained for <code>HIST_SUBST_PATTERN</code> below). When used in parameter
expansion, the replacement of <code>r</code> into the parameters value is done
first, and then any additional process, parameter, command, arithmetic,
or brace references are applied, which may evaluate those substitutions
and expansions more than once if <code>l</code> appears more than once in the
starting value. When used in a glob qualifier, any substitutions or
expansions are performed once at the time the qualifier is parsed, even
before the <code>:s</code> expression itself is divided into <code>l</code> and <code>r</code> sides.</p>
<p>If the option <code>HIST_SUBST_PATTERN</code> is set, <code>l</code> is treated as a pattern
of the usual form described in <a href="#Filename-Generation">Filename
Generation</a>. This can be used in all the places
where modifiers are available; note, however, that in globbing
qualifiers parameter substitution has already taken place, so parameters
in the replacement string should be quoted to ensure they are replaced
at the correct time. Note also that complicated patterns used in
globbing qualifiers may need the extended glob qualifier notation
<code>(#q:s/``...``/``...``/)</code> in order for the shell to recognize the
expression as a glob qualifier. Further, note that bad patterns in the
substitution are not subject to the <code>NO_BAD_PATTERN</code> option so will
cause an error.</p>
<p>When <code>HIST_SUBST_PATTERN</code> is set, <code>l</code> may start with a <code>#</code> to indicate
that the pattern must match at the start of the string to be
substituted, and a <code>%</code> may appear at the start or after an <code>#</code> to
indicate that the pattern must match at the end of the string to be
substituted. The <code>%</code> or <code>#</code> may be quoted with two backslashes.</p>
<p>For example, the following piece of filename generation code with the
<code>EXTENDED_GLOB</code> option:</p>
<div class="example">
<pre><code class="language-example">print -r -- *.c(#q:s/#%(#b)s(*).c/'S${match[1]}.C'/)
</code></pre>
</div>
<p>takes the expansion of <code>*.c</code> and applies the glob qualifiers in the
<code>(#q``...``)</code> expression, which consists of a substitution modifier
anchored to the start and end of each word (<code>#%</code>). This turns on
backreferences (<code>(#b)</code>), so that the parenthesised subexpression is
available in the replacement string as <code>${match[1]}</code>. The replacement
string is quoted so that the parameter is not substituted before the
start of filename generation.</p>
<p>The following <code>f</code>, <code>F</code>, <code>w</code> and <code>W</code> modifiers work only with parameter
expansion and filename generation. They are listed here to provide a
single point of reference for all modifiers.</p>
<ul>
<li>
<p><code>f</code><br />
Repeats the immediately (without a colon) following modifier until
the resulting word doesnt change any more.</p>
</li>
<li>
<p><code>F:``expr``:</code><br />
Like <code>f</code>, but repeats only <code>n</code> times if the expression <code>expr</code>
evaluates to <code>n</code>. Any character can be used instead of the <code>:</code>; if
<code>(</code>, <code>[</code>, or <code>{</code> is used as the opening delimiter, the closing
delimiter should be <code>)</code>, <code>]</code>, or <code>}</code>, respectively.</p>
</li>
<li>
<p><code>w</code><br />
Makes the immediately following modifier work on each word in the
string.</p>
</li>
<li>
<p><code>W:``sep``:</code><br />
Like <code>w</code> but words are considered to be the parts of the string that
are separated by <code>sep</code>. Any character can be used instead of the
<code>:</code>; opening parentheses are handled specially, see above.</p>
</li>
</ul>
<hr />
<p><span id="Process-Substitution"></span>
<span id="Process-Substitution-1"></span></p>
<h2 id="142-process-substitution"><a class="header" href="#142-process-substitution">14.2 Process Substitution</a></h2>
<p><span id="index-process-substitution"></span>
<span id="index-substitution_002c-process"></span></p>
<p>Each part of a command argument that takes the form <code>&lt;(``list``)</code>,
<code>&gt;(``list``)</code> or <code>=(``list``)</code> is subject to process substitution.
The expression may be preceded or followed by other strings except that,
to prevent clashes with commonly occurring strings and patterns, the
last form must occur at the start of a command argument, and the forms
are only expanded when first parsing command or assignment arguments.
Process substitutions may be used following redirection operators; in
this case, the substitution must appear with no trailing string.</p>
<p>Note that <code>&lt;&lt;(``list``)</code> is not a special syntax; it is equivalent to
<code>&lt; &lt;(``list``)</code>, redirecting standard input from the result of
process substitution. Hence all the following documentation applies. The
second form (with the space) is recommended for clarity.</p>
<p>In the case of the <code>&lt;</code> or <code>&gt;</code> forms, the shell runs the commands in
<code>list</code> as a subprocess of the job executing the shell command line. If
the system supports the <code>/dev/fd</code> mechanism, the command argument is the
name of the device file corresponding to a file descriptor; otherwise,
if the system supports named pipes (FIFOs), the command argument will be
a named pipe. If the form with <code>&gt;</code> is selected then writing on this
special file will provide input for <code>list</code>. If <code>&lt;</code> is used, then the
file passed as an argument will be connected to the output of the <code>list</code>
process. For example,</p>
<div class="example">
<pre><code class="language-example">paste &lt;(cut -f1 file1) &lt;(cut -f3 file2) |
tee &gt;(process1) &gt;(process2) &gt;/dev/null
</code></pre>
</div>
<p>cuts fields 1 and 3 from the files <code>file1</code> and <code>file2</code> respectively,
pastes the results together, and sends it to the processes <code>process1</code>
and <code>process2</code>.</p>
<p>If <code>=(``...``)</code> is used instead of <code>&lt;(``...``)</code>, then the file passed as
an argument will be the name of a temporary file containing the output
of the <code>list</code> process. This may be used instead of the <code>&lt;</code> form for a
program that expects to lseek (see man page lseek(2)) on the input file.</p>
<p>There is an optimisation for substitutions of the form <code>=(&lt;&lt;&lt;``arg``)</code>,
where <code>arg</code> is a single-word argument to the here-string redirection
<code>&lt;&lt;&lt;</code>. This form produces a file name containing the value of <code>arg</code>
after any substitutions have been performed. This is handled entirely
within the current shell. This is effectively the reverse of the special
form <code>$(&lt;``arg``)</code> which treats <code>arg</code> as a file name and replaces it
with the files contents.</p>
<p>The <code>=</code> form is useful as both the <code>/dev/fd</code> and the named pipe
implementation of <code>&lt;(``...``)</code> have drawbacks. In the former case, some
programmes may automatically close the file descriptor in question
before examining the file on the command line, particularly if this is
necessary for security reasons such as when the programme is running
setuid. In the second case, if the programme does not actually open the
file, the subshell attempting to read from or write to the pipe will (in
a typical implementation, different operating systems may have different
behaviour) block for ever and have to be killed explicitly. In both
cases, the shell actually supplies the information using a pipe, so that
programmes that expect to lseek (see man page lseek(2)) on the file will
not work.</p>
<p>Also note that the previous example can be more compactly and
efficiently written (provided the <code>MULTIOS</code> option is set) as:</p>
<div class="example">
<pre><code class="language-example">paste &lt;(cut -f1 file1) &lt;(cut -f3 file2) &gt; &gt;(process1) &gt; &gt;(process2)
</code></pre>
</div>
<p>The shell uses pipes instead of FIFOs to implement the latter two
process substitutions in the above example.</p>
<p>There is an additional problem with <code>&gt;(``process``)</code>; when this is
attached to an external command, the parent shell does not wait for
<code>process</code> to finish and hence an immediately following command cannot
rely on the results being complete. The problem and solution are the
same as described in the section <em>MULTIOS</em> in
<a href="Redirection.html#Redirection">Redirection</a>. Hence in a simplified
version of the example above:</p>
<div class="example">
<pre><code class="language-example">paste &lt;(cut -f1 file1) &lt;(cut -f3 file2) &gt; &gt;(process)
</code></pre>
</div>
<p>(note that no <code>MULTIOS</code> are involved), <code>process</code> will be run
asynchronously as far as the parent shell is concerned. The workaround
is:</p>
<div class="example">
<pre><code class="language-example">{ paste &lt;(cut -f1 file1) &lt;(cut -f3 file2) } &gt; &gt;(process)
</code></pre>
</div>
<p>The extra processes here are spawned from the parent shell which will
wait for their completion.</p>
<p>Another problem arises any time a job with a substitution that requires
a temporary file is disowned by the shell, including the case where
<code>&amp;!</code> or <code>&amp;|</code> appears at the end of a command containing a
substitution. In that case the temporary file will not be cleaned up as
the shell no longer has any memory of the job. A workaround is to use a
subshell, for example,</p>
<div class="example">
<pre><code class="language-example">(mycmd =(myoutput)) &amp;!
</code></pre>
</div>
<p>as the forked subshell will wait for the command to finish then remove
the temporary file.</p>
<p>A general workaround to ensure a process substitution endures for an
appropriate length of time is to pass it as a parameter to an anonymous
shell function (a piece of shell code that is run immediately with
function scope). For example, this code:</p>
<div class="example">
<pre><code class="language-example">() {
print File $1:
cat $1
} =(print This be the verse)
</code></pre>
</div>
<p>outputs something resembling the following</p>
<div class="example">
<pre><code class="language-example">File /tmp/zsh6nU0kS:
This be the verse
</code></pre>
</div>
<p>The temporary file created by the process substitution will be deleted
when the function exits.</p>
<hr />
<p><span id="Parameter-Expansion"></span>
<span id="Parameter-Expansion-1"></span></p>
<h2 id="143-parameter-expansion"><a class="header" href="#143-parameter-expansion">14.3 Parameter Expansion</a></h2>
<p><span id="index-parameter-expansion"></span>
<span id="index-expansion_002c-parameter"></span></p>
<p>The character <code>$</code> is used to introduce parameter expansions. See
<a href="Parameters.html#Parameters">Parameters</a> for a description of
parameters, including arrays, associative arrays, and subscript notation
to access individual array elements.</p>
<p>Note in particular the fact that words of unquoted parameters are not
automatically split on whitespace unless the option <code>SH_WORD_SPLIT</code> is
set; see references to this option below for more details. This is an
important difference from other shells. However, as in other shells,
null words are elided from unquoted parameters expansions.</p>
<p>With default options, after the assignments:</p>
<div class="example">
<pre><code class="language-example">array=(&quot;first word&quot; &quot;&quot; &quot;third word&quot;)
scalar=&quot;only word&quot;
</code></pre>
</div>
<p>then <code>$array</code> substitutes two words, <code>first word</code> and <code>third word</code>,
and <code>$scalar</code> substitutes a single word <code>only word</code>. Note that second
element of <code>array</code> was elided. Scalar parameters can be elided too if
their value is null (empty). To avoid elision, use quoting as follows:
<code>&quot;$scalar&quot;</code> for scalars and <code>&quot;${array[@]}&quot;</code> or <code>&quot;${(@)array}&quot;</code> for
arrays. (The last two forms are equivalent.)</p>
<p>Parameter expansions can involve <em>flags</em>, as in <code>${(@kv)aliases}</code>, and
other operators, such as <code>${PREFIX:-&quot;/usr/local&quot;}</code>. Parameter
expansions can also be nested. These topics will be introduced below.
The full rules are complicated and are noted at the end.</p>
<p>In the expansions discussed below that require a pattern, the form of
the pattern is the same as that used for filename generation; see
<a href="#Filename-Generation">Filename Generation</a>. Note that these patterns,
along with the replacement text of any substitutions, are themselves
subject to parameter expansion, command substitution, and arithmetic
expansion. In addition to the following operations, the colon modifiers
described in <a href="#Modifiers">Modifiers</a> in <a href="#History-Expansion">History
Expansion</a> can be applied: for example,
<code>${i:s/foo/bar/}</code> performs string substitution on the expansion of
parameter <code>$i</code>.</p>
<p>In the following descriptions, <code>word</code> refers to a single word
substituted on the command line, not necessarily a space delimited word.</p>
<ul>
<li>
<p><code>${``name``}</code><br />
The value, if any, of the parameter <code>name</code> is substituted. The
braces are required if the expansion is to be followed by a letter,
digit, or underscore that is not to be interpreted as part of
<code>name</code>. In addition, more complicated forms of substitution usually
require the braces to be present; exceptions, which only apply if
the option <code>KSH_ARRAYS</code> is not set, are a single subscript or any
colon modifiers appearing after the name, or any of the characters
<code>^</code>, <code>=</code>, <code>~</code>, <code>#</code> or <code>+</code> appearing before the name, all
of which work with or without braces.</p>
<p>If <code>name</code> is an array parameter, and the <code>KSH_ARRAYS</code> option is not
set, then the value of each element of <code>name</code> is substituted, one
element per word. Otherwise, the expansion results in one word only;
with <code>KSH_ARRAYS</code>, this is the first element of an array. No field
splitting is done on the result unless the <code>SH_WORD_SPLIT</code> option is
set. See also the flags <code>=</code> and <code>s:``string``:</code>.</p>
</li>
<li>
<p><code>${+``name``}</code><br />
If <code>name</code> is the name of a set parameter <code>1</code> is substituted,
otherwise <code>0</code> is substituted.</p>
</li>
<li>
<p><code>${``name``-``word``}</code><br />
<code>${``name``:-``word``}</code><br />
If <code>name</code> is set, or in the second form is non-null, then substitute
its value; otherwise substitute <code>word</code>. In the second form <code>name</code>
may be omitted, in which case <code>word</code> is always substituted.</p>
</li>
<li>
<p><code>${``name``+``word``}</code><br />
<code>${``name``:+``word``}</code><br />
If <code>name</code> is set, or in the second form is non-null, then substitute
<code>word</code>; otherwise substitute nothing.</p>
</li>
<li>
<p><code>${``name``=``word``}</code><br />
<code>${``name``:=``word``}</code><br />
<code>${``name``::=``word``}</code><br />
In the first form, if <code>name</code> is unset then set it to <code>word</code>; in the
second form, if <code>name</code> is unset or null then set it to <code>word</code>; and
in the third form, unconditionally set <code>name</code> to <code>word</code>. In all
forms, the value of the parameter is then substituted.</p>
</li>
<li>
<p><code>${``name``?``word``}</code><br />
<code>${``name``:?``word``}</code><br />
In the first form, if <code>name</code> is set, or in the second form if <code>name</code>
is both set and non-null, then substitute its value; otherwise,
print <code>word</code> and exit from the shell. Interactive shells instead
return to the prompt. If <code>word</code> is omitted, then a standard message
is printed.</p>
</li>
</ul>
<p>In any of the above expressions that test a variable and substitute an
alternate <code>word</code>, note that you can use standard shell quoting in the
<code>word</code> value to selectively override the splitting done by the
<code>SH_WORD_SPLIT</code> option and the <code>=</code> flag, but not splitting by the
<code>s:``string``:</code> flag.</p>
<p>In the following expressions, when <code>name</code> is an array and the
substitution is not quoted, or if the <code>(@)</code> flag or the <code>name``[@]</code>
syntax is used, matching and replacement is performed on each array
element separately.</p>
<ul>
<li>
<p><code>${``name``#``pattern``}</code><br />
<code>${``name``##``pattern``}</code><br />
If the <code>pattern</code> matches the beginning of the value of <code>name</code>, then
substitute the value of <code>name</code> with the matched portion deleted;
otherwise, just substitute the value of <code>name</code>. In the first form,
the smallest matching pattern is preferred; in the second form, the
largest matching pattern is preferred.</p>
</li>
<li>
<p><code>${``name``%``pattern``}</code><br />
<code>${``name``%%``pattern``}</code><br />
If the <code>pattern</code> matches the end of the value of <code>name</code>, then
substitute the value of <code>name</code> with the matched portion deleted;
otherwise, just substitute the value of <code>name</code>. In the first form,
the smallest matching pattern is preferred; in the second form, the
largest matching pattern is preferred.</p>
</li>
<li>
<p><code>${``name``:#``pattern``}</code><br />
If the <code>pattern</code> matches the value of <code>name</code>, then substitute the
empty string; otherwise, just substitute the value of <code>name</code>. If
<code>name</code> is an array the matching array elements are removed (use the
<code>(M)</code> flag to remove the non-matched elements).</p>
</li>
<li>
<p><code>${``name``:|``arrayname``}</code><br />
If <code>arrayname</code> is the name (N.B., not contents) of an array
variable, then any elements contained in <code>arrayname</code> are removed
from the substitution of <code>name</code>. If the substitution is scalar,
either because <code>name</code> is a scalar variable or the expression is
quoted, the elements of <code>arrayname</code> are instead tested against the
entire expression.</p>
</li>
<li>
<p><code>${``name``:*``arrayname``}</code><br />
Similar to the preceding substitution, but in the opposite sense, so
that entries present in both the original substitution and as
elements of <code>arrayname</code> are retained and others removed.</p>
</li>
<li>
<p><code>${``name``:^``arrayname``}</code><br />
<code>${``name``:^^``arrayname``}</code><br />
Zips two arrays, such that the output array is twice as long as the
shortest (longest for <code>:^^</code>) of <code>name</code> and <code>arrayname</code>, with the
elements alternatingly being picked from them. For <code>:^</code>, if one of
the input arrays is longer, the output will stop when the end of the
shorter array is reached. Thus,</p>
<div class="example">
<pre><code class="language-example">a=(1 2 3 4); b=(a b); print ${a:^b}
</code></pre>
</div>
<p>will output <code>1 a 2 b</code>. For <code>:^^</code>, then the input is repeated
until all of the longer array has been used up and the above will
output <code>1 a 2 b 3 a 4 b</code>.</p>
<p>Either or both inputs may be a scalar, they will be treated as an
array of length 1 with the scalar as the only element. If either
array is empty, the other array is output with no extra elements
inserted.</p>
<p>Currently the following code will output <code>a b</code> and <code>1</code> as two
separate elements, which can be unexpected. The second print
provides a workaround which should continue to work if this is
changed.</p>
<div class="example">
<pre><code class="language-example">a=(a b); b=(1 2); print -l &quot;${a:^b}&quot;; print -l &quot;${${a:^b}}&quot;
</code></pre>
</div>
</li>
<li>
<p><code>${``name``:``offset``}</code><br />
<code>${``name``:``offset``:``length``}</code><br />
This syntax gives effects similar to parameter subscripting in the
form <code>$``name``[``start``,``end``]</code>, but is compatible with other
shells; note that both <code>offset</code> and <code>length</code> are interpreted
differently from the components of a subscript.</p>
<p>If <code>offset</code> is non-negative, then if the variable <code>name</code> is a scalar
substitute the contents starting <code>offset</code> characters from the first
character of the string, and if <code>name</code> is an array substitute
elements starting <code>offset</code> elements from the first element. If
<code>length</code> is given, substitute that many characters or elements,
otherwise the entire rest of the scalar or array.</p>
<p>A positive <code>offset</code> is always treated as the offset of a character
or element in <code>name</code> from the first character or element of the
array (this is different from native zsh subscript notation). Hence
0 refers to the first character or element regardless of the setting
of the option <code>KSH_ARRAYS</code>.</p>
<p>A negative offset counts backwards from the end of the scalar or
array, so that -1 corresponds to the last character or element, and
so on.</p>
<p>When positive, <code>length</code> counts from the <code>offset</code> position toward the
end of the scalar or array. When negative, <code>length</code> counts back from
the end. If this results in a position smaller than <code>offset</code>, a
diagnostic is printed and nothing is substituted.</p>
<p>The option <code>MULTIBYTE</code> is obeyed, i.e. the offset and length count
multibyte characters where appropriate.</p>
<p><code>offset</code> and <code>length</code> undergo the same set of shell substitutions as
for scalar assignment; in addition, they are then subject to
arithmetic evaluation. Hence, for example</p>
<div class="example">
<pre><code class="language-example">print ${foo:3}
print ${foo: 1 + 2}
print ${foo:$(( 1 + 2))}
print ${foo:$(echo 1 + 2)}
</code></pre>
</div>
<p>all have the same effect, extracting the string starting at the
fourth character of <code>$foo</code> if the substitution would otherwise
return a scalar, or the array starting at the fourth element if
<code>$foo</code> would return an array. Note that with the option <code>KSH_ARRAYS</code>
<code>$foo</code> always returns a scalar (regardless of the use of the offset
syntax) and a form such as <code>${foo[*]:3}</code> is required to extract
elements of an array named <code>foo</code>.</p>
<p>If <code>offset</code> is negative, the <code>-</code> may not appear immediately after
the <code>:</code> as this indicates the <code>${``name``:-``word``}</code> form of
substitution. Instead, a space may be inserted before the <code>-</code>.
Furthermore, neither <code>offset</code> nor <code>length</code> may begin with an
alphabetic character or <code>&amp;</code> as these are used to indicate
history-style modifiers. To substitute a value from a variable, the
recommended approach is to precede it with a <code>$</code> as this signifies
the intention (parameter substitution can easily be rendered
unreadable); however, as arithmetic substitution is performed, the
expression <code>${var: offs}</code> does work, retrieving the offset from
<code>$offs</code>.</p>
<p>For further compatibility with other shells there is a special case
for array offset 0. This usually accesses the first element of the
array. However, if the substitution refers to the positional
parameter array, e.g. <code>$@</code> or <code>$*</code>, then offset 0 instead refers to
<code>$0</code>, offset 1 refers to <code>$1</code>, and so on. In other words, the
positional parameter array is effectively extended by prepending
<code>$0</code>. Hence <code>${*:0:1}</code> substitutes <code>$0</code> and <code>${*:1:1}</code> substitutes
<code>$1</code>.</p>
</li>
<li>
<p><code>${``name``/``pattern``/``repl``}</code><br />
<code>${``name``//``pattern``/``repl``}</code><br />
<code>${``name``:/``pattern``/``repl``}</code><br />
Replace the longest possible match of <code>pattern</code> in the expansion of
parameter <code>name</code> by string <code>repl</code>. The first form replaces just the
first occurrence, the second form all occurrences, and the third
form replaces only if <code>pattern</code> matches the entire string. Both
<code>pattern</code> and <code>repl</code> are subject to double-quoted substitution, so
that expressions like <code>${name/$opat/$npat}</code> will work, but obey the
usual rule that pattern characters in <code>$opat</code> are not treated
specially unless either the option <code>GLOB_SUBST</code> is set, or <code>$opat</code>
is instead substituted as <code>${~opat}</code>.</p>
<p>The <code>pattern</code> may begin with a <code>#</code>, in which case the <code>pattern</code>
must match at the start of the string, or <code>%</code>, in which case it
must match at the end of the string, or <code>#%</code> in which case the
<code>pattern</code> must match the entire string. The <code>repl</code> may be an empty
string, in which case the final <code>/</code> may also be omitted. To quote
the final <code>/</code> in other cases it should be preceded by a single
backslash; this is not necessary if the <code>/</code> occurs inside a
substituted parameter. Note also that the <code>#</code>, <code>%</code> and <code>#%</code> are
not active if they occur inside a substituted parameter, even at the
start.</p>
<p>If, after quoting rules apply, <code>${``name``}</code> expands to an array,
the replacements act on each element individually. Note also the
effect of the <code>I</code> and <code>S</code> parameter expansion flags below; however,
the flags <code>M</code>, <code>R</code>, <code>B</code>, <code>E</code> and <code>N</code> are not useful.</p>
<p>For example,</p>
<div class="example">
<pre><code class="language-example">foo=&quot;twinkle twinkle little star&quot; sub=&quot;t*e&quot; rep=&quot;spy&quot;
print ${foo//${~sub}/$rep}
print ${(S)foo//${~sub}/$rep}
</code></pre>
</div>
<p>Here, the <code>~</code> ensures that the text of <code>$sub</code> is treated as a
pattern rather than a plain string. In the first case, the longest
match for <code>t*e</code> is substituted and the result is <code>spy star</code>, while
in the second case, the shortest matches are taken and the result is
<code>spy spy lispy star</code>.</p>
</li>
<li>
<p><code>${#``spec``}</code><br />
If <code>spec</code> is one of the above substitutions, substitute the length
in characters of the result instead of the result itself. If <code>spec</code>
is an array expression, substitute the number of elements of the
result. This has the side-effect that joining is skipped even in
quoted forms, which may affect other sub-expressions in <code>spec</code>. Note
that <code>^</code>, <code>=</code>, and <code>~</code>, below, must appear to the left of
<code>#</code> when these forms are combined.</p>
<p>If the option <code>POSIX_IDENTIFIERS</code> is not set, and <code>spec</code> is a simple
name, then the braces are optional; this is true even for special
parameters so e.g. <code>$#-</code> and <code>$#*</code> take the length of the string
<code>$-</code> and the array <code>$*</code> respectively. If <code>POSIX_IDENTIFIERS</code> is set,
then braces are required for the <code>#</code> to be treated in this fashion.</p>
</li>
<li>
<p><code>${^``spec``}</code><br />
<span id="index-RC_005fEXPAND_005fPARAM_002c-toggle"></span>
<span id="index-array-expansion-style_002c-rc"></span>
<span id="index-rc_002c-array-expansion-style"></span></p>
<p>Turn on the <code>RC_EXPAND_PARAM</code> option for the evaluation of <code>spec</code>;
if the <code>^</code> is doubled, turn it off. When this option is set, array
expansions of the form <code>foo``${``xx``}``bar</code>, where the parameter
<code>xx</code> is set to <code>(``a b c``)</code>, are substituted with <code>fooabar foobbar foocbar</code> instead of the default <code>fooa b cbar</code>. Note that an empty
array will therefore cause all arguments to be removed.</p>
<p>Internally, each such expansion is converted into the equivalent
list for brace expansion. E.g., <code>${^var}</code> becomes
<code>{$var[1],$var[2],</code>...<code>}</code>, and is processed as described in <a href="#Brace-Expansion">Brace
Expansion</a> below: note, however, the expansion
happens immediately, with any explicit brace expansion happening
later. If word splitting is also in effect the <code>$var[``N``]</code> may
themselves be split into different list elements.</p>
</li>
<li>
<p><code>${=``spec``}</code><br />
<span id="index-SH_005fWORD_005fSPLIT_002c-toggle"></span>
<span id="index-field-splitting_002c-sh-style_002c-parameter"></span>
<span id="index-sh_002c-field-splitting-style_002c-parameter"></span></p>
<p>Perform word splitting using the rules for <code>SH_WORD_SPLIT</code> during
the evaluation of <code>spec</code>, but regardless of whether the parameter
appears in double quotes; if the <code>=</code> is doubled, turn it off.
<span id="index-IFS_002c-use-of"></span> This forces parameter
expansions to be split into separate words before substitution,
using <code>IFS</code> as a delimiter. This is done by default in most other
shells.</p>
<p>Note that splitting is applied to <code>word</code> in the assignment forms of
<code>spec</code> <em>before</em> the assignment to <code>name</code> is performed. This affects
the result of array assignments with the <code>A</code> flag.</p>
</li>
<li>
<p><code>${~``spec``}</code><br />
<span id="index-GLOB_005fSUBST_002c-toggle"></span></p>
<p>Turn on the <code>GLOB_SUBST</code> option for the evaluation of <code>spec</code>; if the
<code>~</code> is doubled, turn it off. When this option is set, the string
resulting from the expansion will be interpreted as a pattern
anywhere that is possible, such as in filename expansion and
filename generation and pattern-matching contexts like the right
hand side of the <code>=</code> and <code>!=</code> operators in conditions.</p>
<p>In nested substitutions, note that the effect of the <code>~</code> applies to
the result of the current level of substitution. A surrounding
pattern operation on the result may cancel it. Hence, for example,
if the parameter <code>foo</code> is set to <code>*</code>, <code>${~foo//\*/*.c}</code> is
substituted by the pattern <code>*.c</code>, which may be expanded by filename
generation, but <code>${${~foo}//\*/*.c}</code> substitutes to the string
<code>*.c</code>, which will not be further expanded.</p>
</li>
</ul>
<p>If a <code>${</code>...<code>}</code> type parameter expression or a <code>$(</code>...<code>)</code> type command
substitution is used in place of <code>name</code> above, it is expanded first and
the result is used as if it were the value of <code>name</code>. Thus it is
possible to perform nested operations: <code>${${foo#head}%tail}</code> substitutes
the value of <code>$foo</code> with both <code>head</code> and <code>tail</code> deleted. The form
with <code>$(</code>...<code>)</code> is often useful in combination with the flags described
next; see the examples below. Each <code>name</code> or nested <code>${</code>...<code>}</code> in a
parameter expansion may also be followed by a subscript expression as
described in <a href="Parameters.html#Array-Parameters">Array Parameters</a>.</p>
<p>Note that double quotes may appear around nested expressions, in which
case only the part inside is treated as quoted; for example,
<code>${(f)&quot;$(foo)&quot;}</code> quotes the result of <code>$(foo)</code>, but the flag <code>(f)</code>
(see below) is applied using the rules for unquoted expansions. Note
further that quotes are themselves nested in this context; for example,
in <code>&quot;${(@f)&quot;$(foo)&quot;}&quot;</code>, there are two sets of quotes, one surrounding
the whole expression, the other (redundant) surrounding the <code>$(foo)</code> as
before.</p>
<hr />
<p><span id="Parameter-Expansion-Flags"></span></p>
<h3 id="1431-parameter-expansion-flags"><a class="header" href="#1431-parameter-expansion-flags">14.3.1 Parameter Expansion Flags</a></h3>
<p><span id="index-parameter-expansion-flags"></span>
<span id="index-flags_002c-parameter-expansion"></span>
<span id="index-substitution_002c-parameter_002c-flags"></span></p>
<p>If the opening brace is directly followed by an opening parenthesis, the
string up to the matching closing parenthesis will be taken as a list of
flags. In cases where repeating a flag is meaningful, the repetitions
need not be consecutive; for example, (<code>q%q%q</code>) means the same thing
as the more readable (<code>%%qqq</code>). The following flags are supported:</p>
<ul>
<li>
<p><code>#</code><br />
Evaluate the resulting words as numeric expressions and output the
characters corresponding to the resulting integer. Note that this
form is entirely distinct from use of the <code>#</code> without parentheses.</p>
<p>If the <code>MULTIBYTE</code> option is set and the number is greater than 127
(i.e. not an ASCII character) it is treated as a Unicode character.</p>
</li>
<li>
<p><code>%</code><br />
Expand all <code>%</code> escapes in the resulting words in the same way as in
prompts (see <a href="Prompt-Expansion.html#Prompt-Expansion">Prompt
Expansion</a>). If this flag is
given twice, full prompt expansion is done on the resulting words,
depending on the setting of the <code>PROMPT_PERCENT</code>, <code>PROMPT_SUBST</code> and
<code>PROMPT_BANG</code> options.</p>
</li>
<li>
<p><code>@</code><br />
In double quotes, array elements are put into separate words. E.g.,
<code>&quot;${(@)foo}&quot;</code> is equivalent to <code>&quot;${foo[@]}&quot;</code> and
<code>&quot;${(@)foo[1,2]}&quot;</code> is the same as <code>&quot;$foo[1]&quot; &quot;$foo[2]&quot;</code>. This
is distinct from <em>field splitting</em> by the <code>f</code>, <code>s</code> or <code>z</code> flags,
which still applies within each array element.</p>
</li>
<li>
<p><code>A</code><br />
Convert the substitution into an array expression, even if it
otherwise would be scalar. This has lower precedence than
subscripting, so one level of nested expansion is required in order
that subscripts apply to array elements. Thus
<code>${${(A``)``name``}[1]}</code> yields the full value of <code>name</code> when <code>name</code>
is scalar.</p>
<p>This assigns an array parameter with <code>${</code>...<code>=</code>...<code>}</code>,
<code>${</code>...<code>:=</code>...<code>}</code> or <code>${</code>...<code>::=</code>...<code>}</code>. If this flag is
repeated (as in <code>AA</code>), assigns an associative array parameter.
Assignment is made before sorting or padding; if field splitting is
active, the <code>word</code> part is split before assignment. The <code>name</code> part
may be a subscripted range for ordinary arrays; when assigning an
associative array, the <code>word</code> part <em>must</em> be converted to an array,
for example by using <code>${(AA)=``name``=</code>...<code>}</code> to activate field
splitting.</p>
<p>Surrounding context such as additional nesting or use of the value
in a scalar assignment may cause the array to be joined back into a
single string again.</p>
</li>
<li>
<p><code>a</code><br />
Sort in array index order; when combined with <code>O</code> sort in reverse
array index order. Note that <code>a</code> is therefore equivalent to the
default but <code>Oa</code> is useful for obtaining an arrays elements in
reverse order.</p>
</li>
<li>
<p><code>b</code><br />
Quote with backslashes only characters that are special to pattern
matching. This is useful when the contents of the variable are to be
tested using <code>GLOB_SUBST</code>, including the <code>${~``...``}</code> switch.</p>
<p>Quoting using one of the <code>q</code> family of flags does not work for this
purpose since quotes are not stripped from non-pattern characters by
<code>GLOB_SUBST</code>. In other words,</p>
<div class="example">
<pre><code class="language-example">pattern=${(q)str}
[[ $str = ${~pattern} ]]
</code></pre>
</div>
<p>works if <code>$str</code> is <code>a*b</code> but not if it is <code>a b</code>, whereas</p>
<div class="example">
<pre><code class="language-example">pattern=${(b)str}
[[ $str = ${~pattern} ]]
</code></pre>
</div>
<p>is always true for any possible value of <code>$str</code>.</p>
</li>
<li>
<p><code>c</code><br />
With <code>${#``name``}</code>, count the total number of characters in an
array, as if the elements were concatenated with spaces between
them. This is not a true join of the array, so other expressions
used with this flag may have an effect on the elements of the array
before it is counted.</p>
</li>
<li>
<p><code>C</code><br />
Capitalize the resulting words. Words in this case refers to
sequences of alphanumeric characters separated by non-alphanumerics,
<em>not</em> to words that result from field splitting.</p>
</li>
<li>
<p><code>D</code><br />
Assume the string or array elements contain directories and attempt
to substitute the leading part of these by names. The remainder of
the path (the whole of it if the leading part was not substituted)
is then quoted so that the whole string can be used as a shell
argument. This is the reverse of <code>~</code> substitution: see <a href="#Filename-Expansion">Filename
Expansion</a>.</p>
</li>
<li>
<p><code>e</code><br />
Perform single word shell expansions, namely <em>parameter expansion</em>,
<em>command substitution</em> and <em>arithmetic expansion</em>, on the result.
Such expansions can be nested but too deep recursion may have</p>
</li>
<li>
<p><code>f</code><br />
Split the result of the expansion at newlines. This is a shorthand
for <code>ps:\n:</code>.</p>
</li>
<li>
<p><code>F</code><br />
Join the words of arrays together using newline as a separator. This
is a shorthand for <code>pj:\n:</code>.</p>
</li>
<li>
<p><code>g:``opts``:</code><br />
Process escape sequences like the echo builtin when no options are
given (<code>g::</code>). With the <code>o</code> option, octal escapes dont take a
leading zero. With the <code>c</code> option, sequences like <code>^X</code> are also
processed. With the <code>e</code> option, processes <code>\M-t</code> and similar
sequences like the print builtin. With both of the <code>o</code> and <code>e</code>
options, behaves like the print builtin except that in none of these
modes is <code>\c</code> interpreted.</p>
</li>
<li>
<p><code>i</code><br />
Sort case-insensitively. May be combined with <code>n</code> or <code>O</code>.</p>
</li>
<li>
<p><code>k</code><br />
If <code>name</code> refers to an associative array, substitute the <em>keys</em>
(element names) rather than the values of the elements. Used with
subscripts (including ordinary arrays), force indices or keys to be
substituted even if the subscript form refers to values. However,
this flag may not be combined with subscript ranges. With the
<code>KSH_ARRAYS</code> option a subscript <code>[*]</code> or <code>[@]</code> is needed to
operate on the whole array, as usual.</p>
</li>
<li>
<p><code>L</code><br />
Convert all letters in the result to lower case.</p>
</li>
<li>
<p><code>n</code><br />
Sort decimal integers numerically; if the first differing characters
of two test strings are not digits, sorting is lexical. Integers
with more initial zeroes are sorted before those with fewer or none.
Hence the array <code>foo1 foo02 foo2 foo3 foo20 foo23</code> is sorted into
the order shown. May be combined with <code>i</code> or <code>O</code>.</p>
</li>
<li>
<p><code>o</code><br />
Sort the resulting words in ascending order; if this appears on its
own the sorting is lexical and case-sensitive (unless the locale
renders it case-insensitive). Sorting in ascending order is the
default for other forms of sorting, so this is ignored if combined
with <code>a</code>, <code>i</code> or <code>n</code>.</p>
</li>
<li>
<p><code>O</code><br />
Sort the resulting words in descending order; <code>O</code> without <code>a</code>,
<code>i</code> or <code>n</code> sorts in reverse lexical order. May be combined
with <code>a</code>, <code>i</code> or <code>n</code> to reverse the order of sorting.</p>
</li>
<li>
<p><code>P</code><br />
This forces the value of the parameter <code>name</code> to be interpreted as a
further parameter name, whose value will be used where appropriate.
Note that flags set with one of the <code>typeset</code> family of commands (in
particular case transformations) are not applied to the value of
<code>name</code> used in this fashion.</p>
<p>If used with a nested parameter or command substitution, the result
of that will be taken as a parameter name in the same way. For
example, if you have <code>foo=bar</code> and <code>bar=baz</code>, the strings
<code>${(P)foo}</code>, <code>${(P)${foo}}</code>, and <code>${(P)$(echo bar)}</code> will be
expanded to <code>baz</code>.</p>
<p>Likewise, if the reference is itself nested, the expression with the
flag is treated as if it were directly replaced by the parameter
name. It is an error if this nested substitution produces an array
with more than one word. For example, if <code>name=assoc</code> where the
parameter <code>assoc</code> is an associative array, then
<code>${${(P)name}[elt]}</code> refers to the element of the associative
subscripted <code>elt</code>.</p>
</li>
<li>
<p><code>q</code><br />
Quote characters that are special to the shell in the resulting
words with <code>$\``NNN``</code> form, with separate quotes for each octet.</p>
<p>If this flag is given twice, the resulting words are quoted in
single quotes and if it is given three times, the words are quoted
in double characters is attempted. If the flag is given four times,
the words are quoted in single quotes preceded by a <code>$</code>. Note that
in all three of these forms quoting is done unconditionally, even if
this does not change the way the resulting string would be
interpreted by the shell.</p>
<p>If a <code>q-</code> is given (only a single <code>q</code> may appear), a minimal form of
single quoting is used that only quotes the string if needed to
protect special characters. Typically this form gives the most
readable output.</p>
<p>If a <code>q+</code> is given, an extended form of minimal quoting is used that
This quoting is similar to that used by the output of values by the
<code>typeset</code> family of commands.</p>
</li>
<li>
<p><code>Q</code><br />
Remove one level of quotes from the resulting words.</p>
</li>
<li>
<p><code>t</code><br />
Use a string describing the type of the parameter where the value of
the parameter would usually appear. This string consists of keywords
separated by hyphens (<code>-</code>). The first keyword in the string
describes the main type, it can be one of <code>scalar</code>, <code>array</code>,
<code>integer</code>, <code>float</code> or <code>association</code>. The other keywords
describe the type in more detail:</p>
<ul>
<li>
<p><code>local</code><br />
for local parameters</p>
</li>
<li>
<p><code>left</code><br />
for left justified parameters</p>
</li>
<li>
<p><code>right_blanks</code><br />
for right justified parameters with leading blanks</p>
</li>
<li>
<p><code>right_zeros</code><br />
for right justified parameters with leading zeros</p>
</li>
<li>
<p><code>lower</code><br />
for parameters whose value is converted to all lower case when
it is expanded</p>
</li>
<li>
<p><code>upper</code><br />
for parameters whose value is converted to all upper case when
it is expanded</p>
</li>
<li>
<p><code>readonly</code><br />
for readonly parameters</p>
</li>
<li>
<p><code>tag</code><br />
for tagged parameters</p>
</li>
<li>
<p><code>export</code><br />
for exported parameters</p>
</li>
<li>
<p><code>unique</code><br />
for arrays which keep only the first occurrence of duplicated
values</p>
</li>
<li>
<p><code>hide</code><br />
for parameters with the hide flag</p>
</li>
<li>
<p><code>hideval</code><br />
for parameters with the hideval flag</p>
</li>
<li>
<p><code>special</code><br />
for special parameters defined by the shell</p>
</li>
</ul>
</li>
<li>
<p><code>u</code><br />
Expand only the first occurrence of each unique word.</p>
</li>
<li>
<p><code>U</code><br />
Convert all letters in the result to upper case.</p>
</li>
<li>
<p><code>v</code><br />
Used with <code>k</code>, substitute (as two consecutive words) both the key
and the value of each associative array element. Used with
subscripts, force values to be substituted even if the subscript
form refers to indices or keys.</p>
</li>
<li>
<p><code>V</code><br />
Make any special characters in the resulting words visible.</p>
</li>
<li>
<p><code>w</code><br />
With <code>${#``name``}</code>, count words in arrays or strings; the <code>s</code> flag
may be used to set a word delimiter.</p>
</li>
<li>
<p><code>W</code><br />
Similar to <code>w</code> with the difference that empty words between repeated
delimiters are also counted.</p>
</li>
<li>
<p><code>X</code><br />
With this flag, parsing errors occurring with the <code>Q</code>, <code>e</code> and <code>#</code>
flags or the pattern matching forms such as
<code>${``name``#``pattern``}</code> are reported. Without the flag, errors
are silently ignored.</p>
</li>
<li>
<p><code>z</code><br />
Split the result of the expansion into words using shell parsing to
find the words, i.e. taking into account any quoting in the value.
Comments are not treated specially but as ordinary strings, similar
to interactive shells with the <code>INTERACTIVE_COMMENTS</code> option unset
(however, see the <code>Z</code> flag below for related options)</p>
<p>Note that this is done very late, even later than the <code>(s)</code> flag.
So to access single words in the result use nested expansions as in
<code>${${(z)foo}[2]}</code>. Likewise, to remove the quotes in the
resulting words use <code>${(Q)${(z)foo}}</code>.</p>
</li>
<li>
<p><code>0</code><br />
Split the result of the expansion on null bytes. This is a shorthand
for <code>ps:\0:</code>.</p>
</li>
</ul>
<p>The following flags (except <code>p</code>) are followed by one or more arguments
as shown. Any character, or the matching pairs <code>(</code>...<code>)</code>, <code>{</code>...<code>}</code>,
<code>[</code>...<code>]</code>, or <code>&lt;</code>...<code>&gt;</code>, may be used in place of a colon as
delimiters, but note that when a flag takes more than one argument, a
matched pair of delimiters must surround each argument.</p>
<ul>
<li>
<p><code>p</code><br />
Recognize the same escape sequences as the <code>print</code> builtin in string
arguments to any of the flags described below that follow this
argument.</p>
<p>Alternatively, with this option string arguments may be in the form
<code>$``var</code> in which case the value of the variable is substituted.
Note this form is strict; the string argument does not undergo
general parameter expansion.</p>
<p>For example,</p>
<div class="example">
<pre><code class="language-example">sep=:
val=a:b:c
print ${(ps.$sep.)val}
</code></pre>
</div>
<p>splits the variable on a <code>:</code>.</p>
</li>
<li>
<p><code>~</code><br />
Strings inserted into the expansion by any of the flags below are to
be treated as patterns. This applies to the string arguments of
flags that follow <code>~</code> within the same set of parentheses. Compare
with <code>~</code> outside parentheses, which forces the entire substituted
string to be treated as a pattern. Hence, for example,</p>
<div class="example">
<pre><code class="language-example">[[ &quot;?&quot; = ${(~j.|.)array} ]]
</code></pre>
</div>
<p>treats <code>|</code> as a pattern and succeeds if and only if <code>$array</code>
contains the string <code>?</code> as an element. The <code>~</code> may be repeated to
toggle the behaviour; its effect only lasts to the end of the
parenthesised group.</p>
</li>
<li>
<p><code>j:``string``:</code><br />
Join the words of arrays together using <code>string</code> as a separator.
<span id="index-SH_005fWORD_005fSPLIT_002c-use-of"></span> Note that
this occurs before field splitting by the <code>s:``string``:</code> flag or
the <code>SH_WORD_SPLIT</code> option.</p>
</li>
<li>
<p><code>l:``expr``::``string1``::``string2``:</code><br />
Pad the resulting words on the left. Each word will be truncated if
required and placed in a field <code>expr</code> characters wide.</p>
<p>The arguments <code>:``string1``:</code> and <code>:``string2``:</code> are optional;
neither, the first, or both may be given. Note that the same pairs
of delimiters must be used for each of the three arguments. The
space to the left will be filled with <code>string1</code> (concatenated as
often as needed) or spaces if <code>string1</code> is not given. If both
<code>string1</code> and <code>string2</code> are given, <code>string2</code> is inserted once
directly to the left of each word, truncated if necessary, before
<code>string1</code> is used to produce any remaining padding.</p>
<p>If either of <code>string1</code> or <code>string2</code> is present but empty, i.e. there
are two delimiters together at that point, the first character of
<code>$IFS</code> is used instead.</p>
<p>If the <code>MULTIBYTE</code> option is in effect, the flag <code>m</code> may also be
given, in which case widths will be used for the calculation of
padding; otherwise individual multibyte characters are treated as
occupying one unit of width.</p>
<p>If the <code>MULTIBYTE</code> option is not in effect, each byte in the string
is treated as occupying one unit of width.</p>
<p>Control characters are always assumed to be one unit wide; this
allows the mechanism to be used for generating repetitions of
control characters.</p>
</li>
<li>
<p><code>m</code><br />
Only useful together with one of the flags <code>l</code> or <code>r</code> or with the
<code>#</code> length operator when the <code>MULTIBYTE</code> option is in effect. Use
the character width reported by the system in calculating how much
of the string it occupies or the overall unit, however certain Asian
character sets and certain special effects use wider characters;
combining characters have zero width. would actually be displayed
will vary.</p>
<p>If the <code>m</code> is repeated, the character either counts zero (if it has
effect of counting the number of glyphs (visibly separate
characters), except for the case where combining characters
themselves have non-zero width (true in certain alphabets).</p>
</li>
<li>
<p><code>r:``expr``::``string1``::``string2``:</code><br />
As <code>l</code>, but pad the words on the right and insert <code>string2</code>
immediately to the right of the string to be padded.</p>
<p>Left and right padding may be used together. In this case the
strategy is to apply left padding to the first half width of each of
the resulting words, and right padding to the second half. If the
string to be padded has odd width the extra padding is applied on
the left.</p>
</li>
<li>
<p><code>s:``string``:</code><br />
Force field splitting at the separator <code>string</code>. Note that a
<code>string</code> of two or more characters means that all of them must match
in sequence; this differs from the treatment of two or more
characters in the <code>IFS</code> parameter. See also the <code>=</code> flag and the
<code>SH_WORD_SPLIT</code> option. An empty string may also be given in which
case every character will be a separate element.</p>
<p>For historical reasons, the usual behaviour that empty array
elements are retained inside double quotes is disabled for arrays
generated by splitting; hence the following:</p>
<div class="example">
<pre><code class="language-example">line=&quot;one::three&quot;
print -l &quot;${(s.:.)line}&quot;
</code></pre>
</div>
<p>produces two lines of output for <code>one</code> and <code>three</code> and elides the
empty field. To override this behaviour, supply the <code>(@)</code> flag as
well, i.e. <code>&quot;${(@s.:.)line}&quot;</code>.</p>
</li>
<li>
<p><code>Z:``opts``:</code><br />
As <code>z</code> but takes a combination of option letters between a following
pair of delimiter characters. With no options the effect is
identical to <code>z</code>. <code>(Z+c+)</code> causes comments to be parsed as a string
and retained; any field in the resulting array beginning with an
unquoted comment character is a comment. <code>(Z+C+)</code> causes comments to
be parsed and removed. The rule for comments is standard: anything
between a word starting with the third character of <code>$HISTCHARS</code>,
default <code>#</code>, up to the next newline is a comment. <code>(Z+n+)</code> causes
unquoted newlines to be treated as ordinary whitespace, else they
are treated as if they are shell code delimiters and converted to
semicolons. Options are combined within the same set of delimiters,
e.g. <code>(Z+Cn+)</code>.</p>
</li>
<li>
<p><code>_:``flags``:</code><br />
The underscore (<code>_</code>) flag is reserved for future use. As of this
revision of zsh, there are no valid <code>flags</code>; anything following an
underscore, other than an empty pair of delimiters, is treated as an
error, and the flag itself has no effect.</p>
</li>
</ul>
<p>The following flags are meaningful with the <code>${</code>...<code>#</code>...<code>}</code> or
<code>${</code>...<code>%</code>...<code>}</code> forms. The <code>S</code> and <code>I</code> flags may also be used with the
<code>${</code>...<code>/</code>...<code>}</code> forms.</p>
<ul>
<li>
<p><code>S</code><br />
With <code>#</code> or <code>##</code>, search for the match that starts closest to the
start of the string (a substring match). Of all matches at a
particular position, <code>#</code> selects the shortest and <code>##</code> the longest:</p>
<div class="example">
<pre><code class="language-example">% str=&quot;aXbXc&quot;
% echo ${(S)str#X*}
abXc
% echo ${(S)str##X*}
a
%
</code></pre>
</div>
<p>With <code>%</code> or <code>%%</code>, search for the match that starts closest to the
end of the string:</p>
<div class="example">
<pre><code class="language-example">% str=&quot;aXbXc&quot;
% echo ${(S)str%X*}
aXbc
% echo ${(S)str%%X*}
aXb
%
</code></pre>
</div>
<p>(Note that <code>%</code> and <code>%%</code> dont search for the match that ends closest
to the end of the string, as one might expect.)</p>
<p>With substitution via <code>${</code>...<code>/</code>...<code>}</code> or <code>${</code>...<code>//</code>...<code>}</code>,
specifies non-greedy matching, i.e. that the shortest instead of the
longest match should be replaced:</p>
<div class="example">
<pre><code class="language-example">% str=&quot;abab&quot;
% echo ${str/*b/_}
_
% echo ${(S)str/*b/_}
_ab
%
</code></pre>
</div>
</li>
<li>
<p><code>I:``expr``:</code><br />
Search the <code>expr</code>th match (where <code>expr</code> evaluates to a number). This
only applies when searching for substrings, either with the <code>S</code>
flag, or with <code>${</code>...<code>/</code>...<code>}</code> (only the <code>expr</code>th match is
substituted) or <code>${</code>...<code>//</code>...<code>}</code> (all matches from the <code>expr</code>th on
are substituted). The default is to take the first match.</p>
<p>The <code>expr</code>th match is counted such that there is either one or zero
matches from each starting position in the string, although for
global substitution matches overlapping previous replacements are
ignored. With the <code>${</code>...<code>%</code>...<code>}</code> and <code>${</code>...<code>%%</code>...<code>}</code> forms, the
starting position for the match moves backwards from the end as the
index increases, while with the other forms it moves forward from
the start.</p>
<p>Hence with the string</p>
<div class="example">
<pre><code class="language-example">which switch is the right switch for Ipswich?
</code></pre>
</div>
<p>substitutions of the form <code>${</code>(<code>SI:``N``:</code>)<code>string#w*ch}</code> as <code>N</code>
increases from 1 will match and remove <code>which</code>, <code>witch</code>,
<code>witch</code> and <code>wich</code>; the form using <code>##</code> will match and
remove <code>which switch is the right switch for Ipswich</code>, <code>witch is the right switch for Ipswich</code>, <code>witch for Ipswich</code> and <code>wich</code>.
The form using <code>%</code> will remove the same matches as for <code>#</code>, but
in reverse order, and the form using <code>%%</code> will remove the same
matches as for <code>##</code> in reverse order.</p>
</li>
<li>
<p><code>B</code><br />
Include the index of the beginning of the match in the result.</p>
</li>
<li>
<p><code>E</code><br />
Include the index one character past the end of the match in the
result (note this is inconsistent with other uses of parameter
index).</p>
</li>
<li>
<p><code>M</code><br />
Include the matched portion in the result.</p>
</li>
<li>
<p><code>N</code><br />
Include the length of the match in the result.</p>
</li>
<li>
<p><code>R</code><br />
Include the unmatched portion in the result (the <em>R</em>est).</p>
</li>
</ul>
<hr />
<p><span id="Rules"></span></p>
<h3 id="1432-rules"><a class="header" href="#1432-rules">14.3.2 Rules</a></h3>
<p>Here is a summary of the rules for substitution; this assumes that
braces are present around the substitution, i.e. <code>${``...``}</code>. Some
particular examples are given below. Note that the Zsh Development Group
accepts <em>no responsibility</em> for any brain damage which may occur during
the reading of the following rules.</p>
<ul>
<li>
<p><code>1.</code> <em>Nested substitution</em><br />
If multiple nested <code>${``...``}</code> forms are present, substitution is
performed from the inside outwards. At each level, the substitution
takes account of whether the current value is a scalar or an array,
whether the whole substitution is in double quotes, and what flags
are supplied to the current level of substitution, just as if the
nested substitution were the outermost. The flags are not propagated
up to enclosing substitutions; the nested substitution will return
either a scalar or an array as determined by the flags, possibly
adjusted for quoting. All the following steps take place where
applicable at all levels of substitution.</p>
<p>Note that, unless the <code>(P)</code> flag is present, the flags and any
subscripts apply directly to the value of the nested substitution;
for example, the expansion <code>${${foo}}</code> behaves exactly the same as
<code>${foo}</code>. When the <code>(P)</code> flag is present in a nested substitution,
the other substitution rules are applied to the value <em>before</em> it is
interpreted as a name, so <code>${${(P)foo}}</code> may differ from
<code>${(P)foo}</code>.</p>
<p>At each nested level of substitution, the substituted words undergo
all forms of single-word substitution (i.e. not filename
generation), including command substitution, arithmetic expansion
and filename expansion (i.e. leading <code>~</code> and <code>=</code>). Thus, for
example, <code>${${:-=cat}:h}</code> expands to the directory where the <code>cat</code>
program resides. (Explanation: the internal substitution has no
parameter but a default value <code>=cat</code>, which is expanded by filename
expansion to a full path; the outer substitution then applies the
modifier <code>:h</code> and takes the directory part of the path.)</p>
</li>
<li>
<p><code>2.</code> <em>Internal parameter flags</em><br />
Any parameter flags set by one of the <code>typeset</code> family of commands,
in particular the <code>-L</code>, <code>-R</code>, <code>-Z</code>, <code>-u</code> and <code>-l</code> options for
padding and capitalization, are applied directly to the parameter
value. Note these flags are options to the command, e.g. <code>typeset -Z</code>; they are not the same as the flags used within parameter
substitutions.</p>
<p>At the outermost level of substitution, the <code>(P)</code> flag (rule <code>4.</code>)
ignores these transformations and uses the unmodified value of the
parameter as the name to be replaced. This is usually the desired
behavior because padding may make the value syntactically illegal as
a parameter name, but if capitalization changes are desired, use the
<code>${${(P)foo}}</code> form (rule <code>25.</code>).</p>
</li>
<li>
<p><code>3.</code> <em>Parameter subscripting</em><br />
If the value is a raw parameter reference with a subscript, such as
<code>${``var``[3]}</code>, the effect of subscripting is applied directly to
the parameter. Subscripts are evaluated left to right; subsequent
subscripts apply to the scalar or array value yielded by the
previous subscript. Thus if <code>var</code> is an array, <code>${var[1][2]}</code> is the
second character of the first word, but <code>${var[2,4][2]}</code> is the
entire third word (the second word of the range of words two through
four of the original array). Any number of subscripts may appear.
Flags such as <code>(k)</code> and <code>(v)</code> which alter the result of
subscripting are applied.</p>
</li>
<li>
<p><code>4.</code> <em>Parameter name replacement</em><br />
At the outermost level of nesting only, the <code>(P)</code> flag is applied.
This treats the value so far as a parameter name (which may include
a subscript expression) and replaces that with the corresponding
value. This replacement occurs later if the <code>(P)</code> flag appears in
a nested substitution.</p>
<p>If the value so far names a parameter that has internal flags (rule
<code>2.</code>), those internal flags are applied to the new value after
replacement.</p>
</li>
<li>
<p><code>5.</code> <em>Double-quoted joining</em><br />
If the value after this process is an array, and the substitution
appears in double quotes, and neither an <code>(@)</code> flag nor a <code>#</code>
length operator is present at the current level, then words of the
value are joined with the first character of the parameter <code>$IFS</code>,
by default a space, between each word (single word arrays are not
modified). If the <code>(j)</code> flag is present, that is used for joining
instead of <code>$IFS</code>.</p>
</li>
<li>
<p><code>6.</code> <em>Nested subscripting</em><br />
Any remaining subscripts (i.e. of a nested substitution) are
evaluated at this point, based on whether the value is an array or a
scalar. As with <code>3.</code>, multiple subscripts can appear. Note that
<code>${foo[2,4][2]}</code> is thus equivalent to <code>${${foo[2,4]}[2]}</code> and also
to <code>&quot;${${(@)foo[2,4]}[2]}&quot;</code> (the nested substitution returns an
array in both cases), but not to <code>&quot;${${foo[2,4]}[2]}&quot;</code> (the nested
substitution returns a scalar because of the quotes).</p>
</li>
<li>
<p><code>7.</code> <em>Modifiers</em><br />
Any modifiers, as specified by a trailing <code>#</code>, <code>%</code>, <code>/</code>
(possibly doubled) or by a set of modifiers of the form <code>:...</code>
(see <a href="#Modifiers">Modifiers</a> in <a href="#History-Expansion">History
Expansion</a>), are applied to the words of the
value at this level.</p>
</li>
<li>
<p><code>8.</code> <em>Character evaluation</em><br />
Any <code>(#)</code> flag is applied, evaluating the result so far
numerically as a character.</p>
</li>
<li>
<p><code>9.</code> <em>Length</em><br />
Any initial <code>#</code> modifier, i.e. in the form <code>${#``var``}</code>, is used
to evaluate the length of the expression so far.</p>
</li>
<li>
<p><code>10.</code> <em>Forced joining</em><br />
If the <code>(j)</code> flag is present, or no <code>(j)</code> flag is present but
the string is to be split as given by rule <code>11.</code>, and joining did
not take place at rule <code>5.</code>, any words in the value are joined
together using the given string or the first character of <code>$IFS</code> if
none. Note that the <code>(F)</code> flag implicitly supplies a string for
joining in this manner.</p>
</li>
<li>
<p><code>11.</code> <em>Simple word splitting</em><br />
If one of the <code>(s)</code> or <code>(f)</code> flags are present, or the <code>=</code>
specifier was present (e.g. <code>${=``var``}</code>), the word is split on
occurrences of the specified string, or (for <code>=</code> with neither of the
two flags present) any of the characters in <code>$IFS</code>.</p>
<p>If no <code>(s)</code>, <code>(f)</code> or <code>=</code> was given, but the word is not
quoted and the option <code>SH_WORD_SPLIT</code> is set, the word is split on
occurrences of any of the characters in <code>$IFS</code>. Note this step, too,
takes place at all levels of a nested substitution.</p>
</li>
<li>
<p><code>12.</code> <em>Case modification</em><br />
Any case modification from one of the flags <code>(L)</code>, <code>(U)</code> or
<code>(C)</code> is applied.</p>
</li>
<li>
<p><code>13.</code> <em>Escape sequence replacement</em><br />
First any replacements from the <code>(g)</code> flag are performed, then any
prompt-style formatting from the <code>(%)</code> family of flags is applied.</p>
</li>
<li>
<p><code>14.</code> <em>Quote application</em><br />
Any quoting or unquoting using <code>(q)</code> and <code>(Q)</code> and related flags
is applied.</p>
</li>
<li>
<p><code>15.</code> <em>Directory naming</em><br />
Any directory name substitution using <code>(D)</code> flag is applied.</p>
</li>
<li>
<p><code>16.</code> <em>Visibility enhancement</em><br />
Any modifications to make characters visible using the <code>(V)</code> flag
are applied.</p>
</li>
<li>
<p><code>17.</code> <em>Lexical word splitting</em><br />
If the <code>(z)</code> flag or one of the forms of the <code>(Z)</code> flag is
present, the word is split as if it were a shell command line, so
that quotation marks and other metacharacters are used to decide
what constitutes a word. Note this form of splitting is entirely
distinct from that described by rule <code>11.</code>: it does not use <code>$IFS</code>,
and does not cause forced joining.</p>
</li>
<li>
<p><code>18.</code> <em>Uniqueness</em><br />
If the result is an array and the <code>(u)</code> flag was present,
duplicate elements are removed from the array.</p>
</li>
<li>
<p><code>19.</code> <em>Ordering</em><br />
If the result is still an array and one of the <code>(o)</code> or <code>(O)</code>
flags was present, the array is reordered.</p>
</li>
<li>
<p><code>20.</code> <code>RC_EXPAND_PARAM</code><br />
At this point the decision is made whether any resulting array
elements are to be combined element by element with surrounding
text, as given by either the <code>RC_EXPAND_PARAM</code> option or the <code>^</code>
flag.</p>
</li>
<li>
<p><code>21.</code> <em>Re-evaluation</em><br />
Any <code>(e)</code> flag is applied to the value, forcing it to be
re-examined for new parameter substitutions, but also for command
and arithmetic substitutions.</p>
</li>
<li>
<p><code>22.</code> <em>Padding</em><br />
Any padding of the value by the <code>(l.``fill``.)</code> or
<code>(r.``fill``.)</code> flags is applied.</p>
</li>
<li>
<p><code>23.</code> <em>Semantic joining</em><br />
In contexts where expansion semantics requires a single word to
result, all words are rejoined with the first character of <code>IFS</code>
between. So in <code>${(P``)${(f``)lines}}</code> the value of <code>${lines}</code> is
split at newlines, but then must be joined again before the <code>(P)</code>
flag can be applied.</p>
<p>If a single word is not required, this rule is skipped.</p>
</li>
<li>
<p><code>24.</code> <em>Empty argument removal</em><br />
If the substitution does not appear in double quotes, any resulting
zero-length argument, whether from a scalar or an element of an
array, is elided from the list of arguments inserted into the
command line.</p>
<p>Strictly speaking, the removal happens later as the same happens
with other forms of substitution; the point to note here is simply
that it occurs after any of the above parameter operations.</p>
</li>
<li>
<p><code>25.</code> <em>Nested parameter name replacement</em><br />
If the <code>(P)</code> flag is present and rule <code>4.</code> has not applied, the
value so far is treated as a parameter name (which may include a
subscript expression) and replaced with the corresponding value,
with internal flags (rule <code>2.</code>) applied to the new value.</p>
</li>
</ul>
<hr />
<p><span id="Examples-1"></span></p>
<h3 id="1433-examples"><a class="header" href="#1433-examples">14.3.3 Examples</a></h3>
<p>The flag <code>f</code> is useful to split a double-quoted substitution line by
line. For example, <code>${(f)&quot;$(&lt;``file``)&quot;}</code> substitutes the contents of
<code>file</code> divided so that each line is an element of the resulting array.
Compare this with the effect of <code>$``(&lt;``file``)</code> alone, which divides
the file up by words, or the same inside double quotes, which makes the
entire content of the file a single string.</p>
<p>The following illustrates the rules for nested parameter expansions.
Suppose that <code>$foo</code> contains the array <code>(bar baz``)</code>:</p>
<ul>
<li>
<p><code>&quot;${(@)${foo}[1]}&quot;</code><br />
This produces the result <code>b</code>. First, the inner substitution
<code>&quot;${foo}&quot;</code>, which has no array (<code>@</code>) flag, produces a single word
result <code>&quot;bar baz&quot;</code>. The outer substitution <code>&quot;${(@)...[1]}&quot;</code> detects
that this is a scalar, so that (despite the <code>(@)</code> flag) the
subscript picks the first character.</p>
</li>
<li>
<p><code>&quot;${${(@)foo}[1]}&quot;</code><br />
This produces the result <code>bar</code>. In this case, the inner
substitution <code>&quot;${(@)foo}&quot;</code> produces the array <code>(bar baz``)</code>. The
outer substitution <code>&quot;${...[1]}&quot;</code> detects that this is an array and
picks the first word. This is similar to the simple case
<code>&quot;${foo[1]}&quot;</code>.</p>
</li>
</ul>
<p>As an example of the rules for word splitting and joining, suppose
<code>$foo</code> contains the array <code>(ax1 bx1``)</code>. Then</p>
<ul>
<li>
<p><code>${(s/x/)foo}</code><br />
produces the words <code>a</code>, <code>1 b</code> and <code>1</code>.</p>
</li>
<li>
<p><code>${(j/x/s/x/)foo}</code><br />
produces <code>a</code>, <code>1</code>, <code>b</code> and <code>1</code>.</p>
</li>
<li>
<p><code>${(s/x/)foo%%1*}</code><br />
produces <code>a</code> and <code> b</code> (note the extra space). As substitution
occurs before either joining or splitting, the operation first
generates the modified array <code>(ax bx``)</code>, which is joined to give
<code>&quot;ax bx&quot;</code>, and then split to give <code>a</code>, <code> b</code> and . The final
empty string will then be elided, as it is not in double quotes.</p>
</li>
</ul>
<hr />
<p><span id="Command-Substitution"></span>
<span id="Command-Substitution-1"></span></p>
<h2 id="144-command-substitution"><a class="header" href="#144-command-substitution">14.4 Command Substitution</a></h2>
<p><span id="index-command-substitution"></span>
<span id="index-substitution_002c-command"></span></p>
<p>A command enclosed in parentheses preceded by a dollar sign, like
<code>$(</code>...<code>)</code>, or quoted with grave accents, like <code></code>...<code></code>, is
replaced with its standard output, with any trailing newlines deleted.
If the substitution is not enclosed in double quotes, the output is
broken into words using the <code>IFS</code> parameter.
<span id="index-IFS_002c-use-of-1"></span></p>
<p>The substitution <code>$(cat</code> <code>foo``)</code> may be replaced by the faster
<code>$(&lt;``foo``)</code>. In this case <code>foo</code> undergoes single word shell
expansions (<em>parameter expansion</em>, <em>command substitution</em> and
<em>arithmetic expansion</em>), but not filename generation.</p>
<p>If the option <code>GLOB_SUBST</code> is set, the result of any unquoted command
substitution, including the special form just mentioned, is eligible for
filename generation.</p>
<hr />
<p><span id="Arithmetic-Expansion"></span>
<span id="Arithmetic-Expansion-1"></span></p>
<h2 id="145-arithmetic-expansion"><a class="header" href="#145-arithmetic-expansion">14.5 Arithmetic Expansion</a></h2>
<p><span id="index-arithmetic-expansion"></span>
<span id="index-expansion_002c-arithmetic"></span></p>
<p>A string of the form <code>$[``exp``]</code> or <code>$((``exp``))</code> is substituted
with the value of the arithmetic expression <code>exp</code>. <code>exp</code> is subjected to
<em>parameter expansion</em>, <em>command substitution</em> and <em>arithmetic expansion</em>
before it is evaluated. See <a href="Arithmetic-Evaluation.html#Arithmetic-Evaluation">Arithmetic
Evaluation</a>.</p>
<hr />
<p><span id="Brace-Expansion"></span> <span id="Brace-Expansion-1"></span></p>
<h2 id="146-brace-expansion"><a class="header" href="#146-brace-expansion">14.6 Brace Expansion</a></h2>
<p><span id="index-brace-expansion"></span>
<span id="index-expansion_002c-brace"></span></p>
<p>A string of the form <code>foo``{``xx``,``yy``,``zz``}``bar</code> is expanded to
the individual words <code>fooxxbar</code>, <code>fooyybar</code> and <code>foozzbar</code>.
Left-to-right order is preserved. This construct may be nested. Commas
may be quoted in order to include them literally in a word.</p>
<p>An expression of the form <code>{``n1``..``n2``}</code>, where <code>n1</code> and <code>n2</code> are
integers, is expanded to every number between <code>n1</code> and <code>n2</code> inclusive.
If either number begins with a zero, all the resulting numbers will be
padded with leading zeroes to that minimum width, but for negative
numbers the <code>-</code> character is also included in the width. If the numbers
are in decreasing order the resulting sequence will also be in
decreasing order.</p>
<p>An expression of the form <code>{``n1``..``n2``..``n3``}</code>, where <code>n1</code>,
<code>n2</code>, and <code>n3</code> are integers, is expanded as above, but only every <code>n3</code>th
number starting from <code>n1</code> is output. If <code>n3</code> is negative the numbers are
output in reverse order, this is slightly different from simply swapping
<code>n1</code> and <code>n2</code> in the case that the step <code>n3</code> doesnt evenly divide the
range. Zero padding can be specified in any of the three numbers,
specifying it in the third can be useful to pad for example
<code>{-99..100..01}</code> which is not possible to specify by putting a 0 on
either of the first two numbers (i.e. pad to two characters).</p>
<p>An expression of the form <code>{``c1``..``c2``}</code>, where <code>c1</code> and <code>c2</code> are
single characters (which may be multibyte characters), is expanded to
every character in the range from <code>c1</code> to <code>c2</code> in whatever character
sequence is used internally. For characters with code points below 128
this is US ASCII (this is the only case most users will need). If any
intervening character is not If the character sequence is reversed, the
output is in reverse order, e.g. <code>{d..a}</code> is substituted as <code>d c b a</code>.</p>
<p>If a brace expression matches none of the above forms, it is left
unchanged, unless the option <code>BRACE_CCL</code> (an abbreviation for brace
character class) is set.
<span id="index-BRACE_005fCCL_002c-use-of"></span> In that case, it is
expanded to a list of the individual characters between the braces
sorted into the order of the characters in the ASCII character set
(multibyte characters are not currently handled). The syntax is similar
to a <code>[</code>...<code>]</code> expression in filename generation: <code>-</code> is treated
specially to denote a range of characters, but <code>^</code> or <code>!</code> as the
first character is treated normally. For example, <code>{abcdef0-9}</code>
expands to 16 words <code>0 1 2 3 4 5 6 7 8 9 a b c d e f</code>.</p>
<p>Note that brace expansion is not part of filename generation (globbing);
an expression such as <code>*/{foo,bar}</code> is split into two separate words
<code>*/foo</code> and <code>*/bar</code> before filename generation takes place. In
particular, note that this is liable to produce a no match error if
<em>either</em> of the two expressions does not match; this is to be contrasted
with <code>*/(foo|bar)</code>, which is treated as a single pattern but otherwise
has similar effects.</p>
<p>To combine brace expansion with array expansion, see the <code>${^``spec``}</code>
form described in <a href="#Parameter-Expansion">Parameter Expansion</a> above.</p>
<hr />
<p><span id="Filename-Expansion"></span>
<span id="Filename-Expansion-1"></span></p>
<h2 id="147-filename-expansion"><a class="header" href="#147-filename-expansion">14.7 Filename Expansion</a></h2>
<p><span id="index-filename-expansion"></span>
<span id="index-expansion_002c-filename"></span></p>
<p>Each word is checked to see if it begins with an unquoted <code>~</code>. If it
does, then the word up to a <code>/</code>, or the end of the word if there is no
<code>/</code>, is checked to see if it can be substituted in one of the ways
described here. If so, then the <code>~</code> and the checked portion are
replaced with the appropriate substitute value.</p>
<p>A <code>~</code> by itself is replaced by the value of <code>$HOME</code>. A <code>~</code> followed
by a <code>+</code> or a <code>-</code> is replaced by current or previous working
directory, respectively.</p>
<p>A <code>~</code> followed by a number is replaced by the directory at that
position in the directory stack. <code>~0</code> is equivalent to <code>~+</code>, and
<code>~1</code> is the top of the stack. <code>~+</code> followed by a number is
replaced by the directory at that position in the directory stack.
<code>~+0</code> is equivalent to <code>~+</code>, and <code>~+1</code> is the top of the stack.
<code>~-</code> followed by a number is replaced by the directory that many
positions from the bottom of the stack. <code>~-0</code> is the bottom of the
stack. <span id="index-PUSHD_005fMINUS_002c-use-of"></span> The
<code>PUSHD_MINUS</code> option exchanges the effects of <code>~+</code> and <code>~-</code> where
they are followed by a number.</p>
<hr />
<p><span id="Dynamic-named-directories"></span></p>
<h3 id="1471-dynamic-named-directories"><a class="header" href="#1471-dynamic-named-directories">14.7.1 Dynamic named directories</a></h3>
<p><span id="index-directories_002c-named_002c-dynamic"></span>
<span id="index-named-directories_002c-dynamic"></span>
<span id="index-dynamic-named-directories"></span></p>
<p>If the function <code>zsh_directory_name</code> exists, or the shell variable
<code>zsh_directory_name_functions</code> exists and contains an array of function
names, then the functions are used to implement dynamic directory
naming. The functions are tried in order until one returns status zero,
so it is important that functions test whether they can handle the case
in question and return an appropriate status.</p>
<p>A <code>~</code> followed by a string <code>namstr</code> in unquoted square brackets is
treated specially as a dynamic directory name. Note that the first
unquoted closing square bracket always terminates <code>namstr</code>. The shell
function is passed two arguments: the string <code>n</code> (for name) and
<code>namstr</code>. It should either set the array <code>reply</code> to a single element
which is the directory corresponding to the name and return status zero
(executing an assignment as the last statement is usually sufficient),
or it should return status non-zero. In the former case the element of
reply is used as the directory; in the latter case the substitution is
deemed to have failed. If all functions fail and the option <code>NOMATCH</code> is
set, an error results.</p>
<p>The functions defined as above are also used to see if a directory can
be turned into a name, for example when printing the directory stack or
when expanding <code>%~</code> in prompts. In this case each function is passed two
arguments: the string <code>d</code> (for directory) and the candidate for dynamic
naming. The function should either return non-zero status, if the
directory cannot be named by the function, or it should set the array
reply to consist of two elements: the first is the dynamic name for the
directory (as would appear within <code>~[``...``]</code>), and the second is the
prefix length of the directory to be replaced. For example, if the trial
directory is <code>/home/myname/src/zsh</code> and the dynamic name for
<code>/home/myname/src</code> (which has 16 characters) is <code>s</code>, then the function
sets</p>
<div class="example">
<pre><code class="language-example">reply=(s 16)
</code></pre>
</div>
<p>The directory name so returned is compared with possible static names
for parts of the directory path, as described below; it is used if the
prefix length matched (16 in the example) is longer than that matched by
any static name.</p>
<p>It is not a requirement that a function implements both <code>n</code> and <code>d</code>
calls; for example, it might be appropriate for certain dynamic forms of
expansion not to be contracted to names. In that case any call with the
first argument <code>d</code> should cause a non-zero status to be returned.</p>
<p>The completion system calls <code>zsh_directory_name c</code> followed by
equivalent calls to elements of the array
<code>zsh_directory_name_functions</code>, if it exists, in order to complete
dynamic names for directories. The code for this should be as for any
other completion function as described in <a href="Completion-System.html#Completion-System">Completion
System</a>.</p>
<p>As a working example, here is a function that expands any dynamic names
beginning with the string <code>p:</code> to directories below
<code>/home/pws/perforce</code>. In this simple case a static name for the
directory would be just as effective.</p>
<div class="example">
<pre><code class="language-example">zsh_directory_name() {
emulate -L zsh
setopt extendedglob
local -a match mbegin mend
if [[ $1 = d ]]; then
# turn the directory into a name
if [[ $2 = (#b)(/home/pws/perforce/)([^/]##)* ]]; then
typeset -ga reply
reply=(p:$match[2] $(( ${#match[1]} + ${#match[2]} )) )
else
return 1
fi
elif [[ $1 = n ]]; then
# turn the name into a directory
[[ $2 != (#b)p:(?*) ]] &amp;&amp; return 1
typeset -ga reply
reply=(/home/pws/perforce/$match[1])
elif [[ $1 = c ]]; then
# complete names
local expl
local -a dirs
dirs=(/home/pws/perforce/*(/:t))
dirs=(p:${^dirs})
_wanted dynamic-dirs expl 'dynamic directory' compadd -S\] -a dirs
return
else
return 1
fi
return 0
}
</code></pre>
</div>
<hr />
<p><span id="Static-named-directories"></span></p>
<h3 id="1472-static-named-directories"><a class="header" href="#1472-static-named-directories">14.7.2 Static named directories</a></h3>
<p><span id="index-directories_002c-named_002c-static"></span>
<span id="index-named-directories_002c-static"></span>
<span id="index-static-named-directories"></span></p>
<p>A <code>~</code> followed by anything not already covered consisting of any
number of alphanumeric characters or underscore (<code>_</code>), hyphen (<code>-</code>),
or dot (<code>.</code>) is looked up as a named directory, and replaced by the
value of that named directory if found. Named directories are typically
home directories for users on the system. They may also be defined if
the text after the <code>~</code> is the name of a string shell parameter whose
value begins with a <code>/</code>. Note that trailing slashes will be removed
from the path to the directory (though the original parameter is not
modified).</p>
<p>It is also possible to define directory names using the <code>-d</code> option to
the <code>hash</code> builtin.</p>
<p>When the shell prints a path (e.g. when expanding <code>%~</code> in prompts or
when printing the directory stack), the path is checked to see if it has
a named directory as its prefix. If so, then the prefix portion is
replaced with a <code>~</code> followed by the name of the directory. The shorter
of the two ways of referring to the directory is used, i.e. either the
directory name or the full path; the name is used if they are the same
length. The parameters <code>$PWD</code> and <code>$OLDPWD</code> are never abbreviated in
this fashion.</p>
<hr />
<p><span id="g_t_0060_003d_0027-expansion"></span></p>
<h3 id="1473--expansion"><a class="header" href="#1473--expansion">14.7.3 = expansion</a></h3>
<p>If a word begins with an unquoted <code>=</code> and the <code>EQUALS</code> option is set,
the remainder of the word is taken as the name of a command. If a
command exists by that name, the word is replaced by the full pathname
of the command.</p>
<hr />
<p><span id="Notes"></span></p>
<h3 id="1474-notes"><a class="header" href="#1474-notes">14.7.4 Notes</a></h3>
<p>Filename expansion is performed on the right hand side of a parameter
assignment, including those appearing after commands of the <code>typeset</code>
family. In this case, the right hand side will be treated as a
colon-separated list in the manner of the <code>PATH</code> parameter, so that a
<code>~</code> or an <code>=</code> following a <code>:</code> is eligible for expansion. All
such behaviour can be disabled by quoting the <code>~</code>, the <code>=</code>, or the
whole expression (but not simply the colon); the <code>EQUALS</code> option is also
respected.</p>
<p>If the option <code>MAGIC_EQUAL_SUBST</code> is set, any unquoted shell argument in
the form <code>identifier``=``expression</code> becomes eligible for file
expansion as described in the previous paragraph. Quoting the first
<code>=</code> also inhibits this.</p>
<hr />
<p><span id="Filename-Generation"></span>
<span id="Filename-Generation-1"></span></p>
<h2 id="148-filename-generation"><a class="header" href="#148-filename-generation">14.8 Filename Generation</a></h2>
<p><span id="index-filename-generation"></span></p>
<p>If a word contains an unquoted instance of one of the characters <code>*</code>,
<code>(</code>, <code>|</code>, <code>&lt;</code>, <code>[</code>, or <code>?</code>, it is regarded as a pattern for
filename generation, unless the <code>GLOB</code> option is unset.
<span id="index-GLOB_002c-use-of"></span> If the <code>EXTENDED_GLOB</code> option
is set, <span id="index-EXTENDED_005fGLOB_002c-use-of"></span> the <code>^</code>
and <code>#</code> characters also denote a pattern; otherwise they are not
treated specially by the shell.</p>
<p>The word is replaced with a list of sorted filenames that match the
pattern. If no matching pattern is found, the shell gives an error
message, unless the <code>NULL_GLOB</code> option is set,
<span id="index-NULL_005fGLOB_002c-use-of"></span> in which case the
word is deleted; or unless the <code>NOMATCH</code> option is unset, in which case
the word is left unchanged. <span id="index-NOMATCH_002c-use-of"></span></p>
<p>In filename generation, the character <code>/</code> must be matched explicitly;
also, a <code>.</code> must be matched explicitly at the beginning of a pattern
or after a <code>/</code>, unless the <code>GLOB_DOTS</code> option is set.
<span id="index-GLOB_005fDOTS_002c-use-of"></span> No filename
generation pattern matches the files <code>.</code> or <code>..</code>. In other instances
of pattern matching, the <code>/</code> and <code>.</code> are not treated specially.</p>
<hr />
<p><span id="Glob-Operators"></span></p>
<h3 id="1481-glob-operators"><a class="header" href="#1481-glob-operators">14.8.1 Glob Operators</a></h3>
<ul>
<li>
<p><code>*</code><br />
Matches any string, including the null string.</p>
</li>
<li>
<p><code>?</code><br />
Matches any character.</p>
</li>
<li>
<p><code>[</code>...<code>]</code><br />
Matches any of the enclosed characters. Ranges of characters can be
specified by separating two characters by a <code>-</code>. A <code>-</code> or <code>]</code>
may be matched by including it as the first character in the list.
<span id="index-character-classes"></span> There are also several
named classes of characters, in the form <code>[:``name``:]</code> with the
following meanings. The first set use the macros provided by the
operating system to test for the given character combinations,
including any modifications due to local language settings, see man
page ctype(3):</p>
<ul>
<li>
<p><code>[:alnum:]</code><br />
The character is alphanumeric</p>
</li>
<li>
<p><code>[:alpha:]</code><br />
The character is alphabetic</p>
</li>
<li>
<p><code>[:ascii:]</code><br />
The character is 7-bit, i.e. is a single-byte character without
the top bit set.</p>
</li>
<li>
<p><code>[:blank:]</code><br />
The character is a blank character</p>
</li>
<li>
<p><code>[:cntrl:]</code><br />
The character is a control character</p>
</li>
<li>
<p><code>[:digit:]</code><br />
The character is a decimal digit</p>
</li>
<li>
<p><code>[:graph:]</code><br />
<code>[:lower:]</code><br />
The character is a lowercase letter</p>
</li>
<li>
<p><code>[:print:]</code><br />
<code>[:punct:]</code><br />
<code>[:space:]</code><br />
The character is whitespace</p>
</li>
<li>
<p><code>[:upper:]</code><br />
The character is an uppercase letter</p>
</li>
<li>
<p><code>[:xdigit:]</code><br />
The character is a hexadecimal digit</p>
</li>
</ul>
<p>Another set of named classes is handled internally by the shell and
is not sensitive to the locale:</p>
<ul>
<li>
<p><code>[:IDENT:]</code><br />
The character is allowed to form part of a shell identifier,
such as a parameter name</p>
</li>
<li>
<p><code>[:IFS:]</code><br />
The character is used as an input field separator, i.e. is
contained in the <code>IFS</code> parameter</p>
</li>
<li>
<p><code>[:IFSSPACE:]</code><br />
The character is an IFS white space character; see the
documentation for <code>IFS</code> in <a href="Parameters.html#Parameters-Used-By-The-Shell">Parameters Used By The
Shell</a>.</p>
</li>
<li>
<p><code>[:INCOMPLETE:]</code><br />
Matches a byte that starts an incomplete multibyte character.
Note that there may be a sequence of more than one bytes that
taken together form the prefix of a multibyte character. To test
for a potentially incomplete byte sequence, use the pattern
<code>[[:INCOMPLETE:]]*</code>. This will never match a sequence
starting with a valid multibyte character.</p>
</li>
<li>
<p><code>[:INVALID:]</code><br />
Matches a byte that does not start a valid multibyte character.
Note this may be a continuation byte of an incomplete multibyte
character as any part of a multibyte string consisting of
invalid and incomplete multibyte characters is treated as single
bytes.</p>
</li>
<li>
<p><code>[:WORD:]</code><br />
The character is treated as part of a word; this test is
sensitive to the value of the <code>WORDCHARS</code> parameter</p>
</li>
</ul>
<p>Note that the square brackets are additional to those enclosing the
whole set of characters, so to test for a single alphanumeric
character you need <code>[[:alnum:]]</code>. Named character sets can be used
alongside other types, e.g. <code>[[:alpha:]0-9]</code>.</p>
</li>
<li>
<p><code>[^</code>...<code>]</code><br />
<code>[!</code>...<code>]</code><br />
Like <code>[</code>...<code>]</code>, except that it matches any character which is not in
the given set.</p>
</li>
<li>
<p><code>&lt;</code>[<code>x</code>]<code>-</code>[<code>y</code>]<code>&gt;</code><br />
Matches any number in the range <code>x</code> to <code>y</code>, inclusive. Either of the
numbers may be omitted to make the range open-ended; hence <code>&lt;-&gt;</code>
matches any number. To match individual digits, the <code>[</code>...<code>]</code> form
is more efficient.</p>
<p>Be careful when using other wildcards adjacent to patterns of this
form; for example, <code>&lt;0-9&gt;*</code> will actually match any number
whatsoever at the start of the string, since the <code>&lt;0-9&gt;</code> will
match the first digit, and the <code>*</code> will match any others. This is
a trap for the unwary, but is match always succeeds. Expressions
such as <code>&lt;0-9&gt;[^[:digit:]]*</code> can be used instead.</p>
</li>
<li>
<p><code>(</code>...<code>)</code><br />
Matches the enclosed pattern. This is used for grouping. If the
<code>KSH_GLOB</code> option is set, then a <code>@</code>, <code>*</code>, <code>+</code>, <code>?</code> or <code>!</code>
immediately preceding the <code>(</code> is treated specially, as detailed
below. The option <code>SH_GLOB</code> prevents bare parentheses from being
used in this way, though the <code>KSH_GLOB</code> option is still available.</p>
<p>Note that grouping cannot extend over multiple directories: it is an
error to have a <code>/</code> within a group (this only applies for patterns
used in filename generation). There is one exception: a group of the
form <code>(``pat``/)#</code> appearing as a complete path segment can match a
sequence of directories. For example, <code>foo/(a*/)#bar</code> matches
<code>foo/bar</code>, <code>foo/any/bar</code>, <code>foo/any/anyother/bar</code>, and so on.</p>
</li>
<li>
<p><code>x``|``y</code><br />
Matches either <code>x</code> or <code>y</code>. This operator has lower precedence than
any other. The <code>|</code> character must be within parentheses, to avoid
interpretation as a pipeline. The alternatives are tried in order
from left to right.</p>
</li>
<li>
<p><code>^``x</code><br />
(Requires <code>EXTENDED_GLOB</code> to be set.) Matches anything except the
pattern <code>x</code>. This has a higher precedence than <code>/</code>, so
<code>^foo/bar</code> will search directories in <code>.</code> except <code>./foo</code>
for a file named <code>bar</code>.</p>
</li>
<li>
<p><code>x``~``y</code><br />
(Requires <code>EXTENDED_GLOB</code> to be set.) Match anything that matches
the pattern <code>x</code> but does not match <code>y</code>. This has lower precedence
than any operator except <code>|</code>, so <code>*/*~foo/bar</code> will search for
all files in all directories in <code>.</code> and then exclude <code>foo/bar</code>
if there was such a match. Multiple patterns can be excluded by
<code>foo``~``bar``~``baz</code>. In the exclusion pattern (<code>y</code>), <code>/</code> and
<code>.</code> are not treated specially the way they usually are in
globbing.</p>
</li>
<li>
<p><code>x``#</code><br />
(Requires <code>EXTENDED_GLOB</code> to be set.) Matches zero or more
occurrences of the pattern <code>x</code>. This operator has high precedence;
<code>12#</code> is equivalent to <code>1(2#)</code>, rather than <code>(12)#</code>. It is
an error for an unquoted <code>#</code> to follow something which cannot be
repeated; this includes an empty string, a pattern already followed
by <code>##</code>, or parentheses when part of a <code>KSH_GLOB</code> pattern (for
example, <code>!(``foo``)#</code> is invalid and must be replaced by
<code>*(!(``foo``))</code>).</p>
</li>
<li>
<p><code>x``##</code><br />
(Requires <code>EXTENDED_GLOB</code> to be set.) Matches one or more
occurrences of the pattern <code>x</code>. This operator has high precedence;
<code>12##</code> is equivalent to <code>1(2##)</code>, rather than <code>(12)##</code>. No
more than two active <code>#</code> characters may appear together. (Note the
potential clash with glob qualifiers in the form <code>1(2##)</code> which
should therefore be avoided.)</p>
</li>
</ul>
<hr />
<p><span id="ksh_002dlike-Glob-Operators"></span></p>
<h3 id="1482-ksh-like-glob-operators"><a class="header" href="#1482-ksh-like-glob-operators">14.8.2 ksh-like Glob Operators</a></h3>
<p><span id="index-KSH_005fGLOB_002c-use-of"></span></p>
<p>If the <code>KSH_GLOB</code> option is set, the effects of parentheses can be
modified by a preceding <code>@</code>, <code>*</code>, <code>+</code>, <code>?</code> or <code>!</code>. This
character need not be unquoted to have special effects, but the <code>(</code>
must be.</p>
<ul>
<li>
<p><code>@(</code>...<code>)</code><br />
Match the pattern in the parentheses. (Like <code>(</code>...<code>)</code>.)</p>
</li>
<li>
<p><code>*(</code>...<code>)</code><br />
Match any number of occurrences. (Like <code>(</code>...<code>)#</code>, except that
recursive directory searching is not supported.)</p>
</li>
<li>
<p><code>+(</code>...<code>)</code><br />
Match at least one occurrence. (Like <code>(</code>...<code>)##</code>, except that
recursive directory searching is not supported.)</p>
</li>
<li>
<p><code>?(</code>...<code>)</code><br />
Match zero or one occurrence. (Like <code>(|</code>...<code>)</code>.)</p>
</li>
<li>
<p><code>!(</code>...<code>)</code><br />
Match anything but the expression in parentheses. (Like
<code>(^(</code>...<code>))</code>.)</p>
</li>
</ul>
<hr />
<p><span id="Precedence"></span></p>
<h3 id="1483-precedence"><a class="header" href="#1483-precedence">14.8.3 Precedence</a></h3>
<p><span id="index-precedence-of-glob-operators"></span></p>
<p>The precedence of the operators given above is (highest) <code>^</code>, <code>/</code>,
<code>~</code>, <code>|</code> (lowest); the remaining operators are simply treated from
left to right as part of a string, with <code>#</code> and <code>##</code> applying to the
shortest possible preceding unit (i.e. a character, <code>?</code>, <code>[</code>...<code>]</code>,
<code>&lt;</code>...<code>&gt;</code>, or a parenthesised expression). As mentioned above, a
<code>/</code> used as a directory separator may not appear inside parentheses,
while a <code>|</code> must do so; in patterns used in other contexts than
filename generation (for example, in <code>case</code> statements and tests within
<code>[[</code>...<code>]]</code>), a <code>/</code> is not special; and <code>/</code> is also not special
after a <code>~</code> appearing outside parentheses in a filename pattern.</p>
<hr />
<p><span id="Globbing-Flags"></span></p>
<h3 id="1484-globbing-flags"><a class="header" href="#1484-globbing-flags">14.8.4 Globbing Flags</a></h3>
<p>There are various flags which affect any text to their right up to the
end of the enclosing group or to the end of the pattern; they require
the <code>EXTENDED_GLOB</code> option. All take the form <code>(#``X``)</code> where <code>X</code> may
have one of the following forms:</p>
<ul>
<li>
<p><code>i</code><br />
Case insensitive: upper or lower case characters in the pattern
match upper or lower case characters.</p>
</li>
<li>
<p><code>l</code><br />
Lower case characters in the pattern match upper or lower case
characters; upper case characters in the pattern still only match
upper case characters.</p>
</li>
<li>
<p><code>I</code><br />
Case sensitive: locally negates the effect of <code>i</code> or <code>l</code> from that
point on.</p>
<p><span id="index-match"></span> <span id="index-mbegin"></span>
<span id="index-mend"></span></p>
</li>
<li>
<p><code>b</code><br />
Activate backreferences for parenthesised groups in the pattern;
this does not work in filename generation. When a pattern with a set
of active parentheses is matched, the strings matched by the groups
are stored in the array <code>$match</code>, the indices of the beginning of
the matched parentheses in the array <code>$mbegin</code>, and the indices of
the end in the array <code>$mend</code>, with the first element of each array
corresponding to the first parenthesised group, and so on. These
arrays are not otherwise special to the shell. The indices use the
same convention as does parameter substitution, so that elements of
<code>$mend</code> and <code>$mbegin</code> may be used in subscripts; the <code>KSH_ARRAYS</code>
option is respected. Sets of globbing flags are not considered
parenthesised groups; only the first nine active parentheses can be
referenced.</p>
<p>For example,</p>
<div class="example">
<pre><code class="language-example">foo=&quot;a_string_with_a_message&quot;
if [[ $foo = (a|an)_(#b)(*) ]]; then
print ${foo[$mbegin[1],$mend[1]]}
fi
</code></pre>
</div>
<p>prints <code>string_with_a_message</code>. Note that the first set of
parentheses is before the <code>(#b)</code> and does not create a
backreference.</p>
<p>Backreferences work with all forms of pattern matching other than
filename generation, but note that when performing matches on an
entire array, such as <code>${``array``#``pattern``}</code>, or a global
substitution, such as <code>${``param``//``pat``/``repl``}</code>, only the
data for the last match remains available. In the case of global
replacements this may still be useful. See the example for the <code>m</code>
flag below.</p>
<p>The numbering of backreferences strictly follows the order of the
opening parentheses from left to right in the pattern string,
although sets of parentheses may be nested. There are special rules
for parentheses followed by <code>#</code> or <code>##</code>. Only the last match of
the parenthesis is remembered: for example, in <code>[[ abab = (#b)([ab])# ]]</code>, only the final <code>b</code> is stored in <code>match[1]</code>. Thus
extra parentheses may be necessary to match the complete segment:
for example, use <code>X((ab|cd)#)Y</code> to match a whole string of either
<code>ab</code> or <code>cd</code> between <code>X</code> and <code>Y</code>, using the value of
<code>$match[1]</code> rather than <code>$match[2]</code>.</p>
<p>If the match fails none of the parameters is altered, so in some
cases it may be necessary to initialise them beforehand. If some of
the backreferences fail to match — which happens if they are in an
alternate branch which fails to match, or if they are followed by
<code>#</code> and matched zero times — then the matched string is set to the
empty string, and the start and end indices are set to -1.</p>
<p>Pattern matching with backreferences is slightly slower than
without.</p>
</li>
<li>
<p><code>B</code><br />
Deactivate backreferences, negating the effect of the <code>b</code> flag from
that point on.</p>
</li>
<li>
<p><code>c``N``,``M</code><br />
The flag <code>(#c``N``,``M``)</code> can be used anywhere that the <code>#</code> or <code>##</code>
operators can be used except in the expressions <code>(*/)#</code> and
<code>(*/)##</code> in filename generation, where <code>/</code> has special
meaning; it cannot be combined with other globbing flags and a bad
pattern error occurs if it is misplaced. It is equivalent to the
form <code>{``N``,``M``}</code> in regular expressions. The previous character
or group is required to match between <code>N</code> and <code>M</code> times, inclusive.
The form <code>(#c``N``)</code> requires exactly <code>N</code> matches; <code>(#c,``M``)</code> is
equivalent to specifying <code>N</code> as 0; <code>(#c``N``,)</code> specifies that there
is no maximum limit on the number of matches.</p>
<p><span id="index-MATCH"></span> <span id="index-MBEGIN"></span>
<span id="index-MEND"></span></p>
</li>
<li>
<p><code>m</code><br />
Set references to the match data for the entire string matched; this
is similar to backreferencing and does not work in filename
generation. The flag must be in effect at the end of the pattern,
i.e. not local to a group. The parameters <code>$MATCH</code>, <code>$MBEGIN</code> and
<code>$MEND</code> will be set to the string matched and to the indices of the
beginning and end of the string, respectively. This is most useful
in parameter substitutions, as otherwise the string matched is
obvious.</p>
<p>For example,</p>
<div class="example">
<pre><code class="language-example">arr=(veldt jynx grimps waqf zho buck)
print ${arr//(#m)[aeiou]/${(U)MATCH}}
</code></pre>
</div>
<p>forces all the matches (i.e. all vowels) into uppercase, printing
<code>vEldt jynx grImps wAqf zhO bUck</code>.</p>
<p>Unlike backreferences, there is no speed penalty for using match
references, other than the extra substitutions required for the
replacement strings in cases such as the example shown.</p>
</li>
<li>
<p><code>M</code><br />
Deactivate the <code>m</code> flag, hence no references to match data will be
created.</p>
</li>
<li>
<p><code>a``num</code><br />
Approximate matching: <code>num</code> errors are allowed in the string matched
by the pattern. The rules for this are described in the next
subsection.</p>
</li>
<li>
<p><code>s</code>, <code>e</code><br />
Unlike the other flags, these have only a local effect, and each
must appear on its own: <code>(#s)</code> and <code>(#e)</code> are the only valid
forms. The <code>(#s)</code> flag succeeds only at the start of the test
string, and the <code>(#e)</code> flag succeeds only at the end of the test
string; they correspond to <code>^</code> and <code>$</code> in standard regular
expressions. They are useful for matching path segments in patterns
other than those in filename generation (where path segments are in
any case treated separately). For example,
<code>*((#s)|/)test((#e)|/)*</code> matches a path segment <code>test</code> in
any of the following strings: <code>test</code>, <code>test/at/start</code>,
<code>at/end/test</code>, <code>in/test/middle</code>.</p>
<p>Another use is in parameter substitution; for example
<code>${array/(#s)A*Z(#e)}</code> will remove only elements of an array
which match the complete pattern <code>A*Z</code>. There are other ways of
performing many operations of this type, however the combination of
the substitution operations <code>/</code> and <code>//</code> with the <code>(#s)</code> and
<code>(#e)</code> flags provides a single simple and memorable method.</p>
<p>Note that assertions of the form <code>(^(#s))</code> also work, i.e. match
anywhere except at the start of the string, although this actually
means anything except a zero-length portion at the start of the
string; you need to use <code>(&quot;&quot;~(#s))</code> to match a zero-length
portion of the string not at the start.</p>
</li>
<li>
<p><code>q</code><br />
A <code>q</code> and everything up to the closing parenthesis of the globbing
flags are ignored by the pattern matching code. This is intended to
support the use of glob qualifiers, see below. The result is that
the pattern <code>(#b)(*).c(#q.)</code> can be used both for globbing and for
matching against a string. In the former case, the <code>(#q.)</code> will be
treated as a glob qualifier and the <code>(#b)</code> will not be useful,
while in the latter case the <code>(#b)</code> is useful for backreferences
and the <code>(#q.)</code> will be ignored. Note that colon modifiers in the
glob qualifiers are also not applied in ordinary pattern matching.</p>
</li>
<li>
<p><code>u</code><br />
Respect the current locale in determining the presence of multibyte
characters in a pattern, provided the shell was compiled with
<code>MULTIBYTE_SUPPORT</code>. This overrides the <code>MULTIBYTE</code> option; the
default behaviour is taken from the option. Compare <code>U</code>. (Mnemonic:
typically multibyte characters are from Unicode in the UTF-8
encoding, although any extension of ASCII supported by the system
library may be used.)</p>
</li>
<li>
<p><code>U</code><br />
All characters are considered to be a single byte long. The opposite
of <code>u</code>. This overrides the <code>MULTIBYTE</code> option.</p>
</li>
</ul>
<p>For example, the test string <code>fooxx</code> can be matched by the pattern
<code>(#i``)FOOXX</code>, but not by <code>(#l``)FOOXX</code>, <code>(#i``)FOO``(#I``)XX</code> or
<code>((#i``)FOOX``)X</code>. The string <code>(#ia2``)readme</code> specifies
case-insensitive matching of <code>readme</code> with up to two errors.</p>
<p>When using the ksh syntax for grouping both <code>KSH_GLOB</code> and
<code>EXTENDED_GLOB</code> must be set and the left parenthesis should be preceded
by <code>@</code>. Note also that the flags do not affect letters inside <code>[</code>...<code>]</code>
groups, in other words <code>(#i``)[a-z]</code> still matches only lowercase
letters. Finally, note that when examining whole paths
case-insensitively every directory must be searched for all files which
match, so that a pattern of the form <code>(#i``)/foo/bar/...</code> is potentially
slow.</p>
<hr />
<p><span id="Approximate-Matching"></span></p>
<h3 id="1485-approximate-matching"><a class="header" href="#1485-approximate-matching">14.8.5 Approximate Matching</a></h3>
<p>When matching approximately, the shell keeps a count of the errors
found, which cannot exceed the number specified in the <code>(#a``num``)</code>
flags. Four types of error are recognised:</p>
<ul>
<li>
<p>1.<br />
Different characters, as in <code>fooxbar</code> and <code>fooybar</code>.</p>
</li>
<li>
<p>2.<br />
Transposition of characters, as in <code>banana</code> and <code>abnana</code>.</p>
</li>
<li>
<p>3.<br />
A character missing in the target string, as with the pattern <code>road</code>
and target string <code>rod</code>.</p>
</li>
<li>
<p>4.<br />
An extra character appearing in the target string, as with <code>stove</code>
and <code>strove</code>.</p>
</li>
</ul>
<p>Thus, the pattern <code>(#a3``)abcd</code> matches <code>dcba</code>, with the errors
occurring by using the first rule twice and the second once, grouping
the string as <code>[d][cb][a]</code> and <code>[a][bc][d]</code>.</p>
<p>Non-literal parts of the pattern must match exactly, including
characters in character ranges: hence <code>(#a1``)???</code> matches strings of
length four, by applying rule 4 to an empty part of the pattern, but not
strings of length two, since all the <code>?</code> must match. Other characters
which must match exactly are initial dots in filenames (unless the
<code>GLOB_DOTS</code> option is set), and all slashes in filenames, so that <code>a/bc</code>
is two errors from <code>ab/c</code> (the slash cannot be transposed with another
character). Similarly, errors are counted separately for non-contiguous
strings in the pattern, so that <code>(ab|cd``)ef</code> is two errors from <code>aebf</code>.</p>
<p>When using exclusion via the <code>~</code> operator, approximate matching is
treated entirely separately for the excluded part and must be activated
separately. Thus, <code>(#a1``)README~READ_ME</code> matches <code>READ.ME</code> but not
<code>READ_ME</code>, as the trailing <code>READ_ME</code> is matched without approximation.
However, <code>(#a1``)README~(#a1``)READ_ME</code> does not match any pattern of
the form <code>READ``?``ME</code> as all such forms are now excluded.</p>
<p>Apart from exclusions, there is only one overall error count; however,
the maximum errors allowed may be altered locally, and this can be
delimited by grouping. For example, <code>(#a1``)cat``((#a0``)dog``)fox</code>
allows one error in total, which may not occur in the <code>dog</code> section, and
the pattern <code>(#a1``)cat``(#a0``)dog``(#a1``)fox</code> is equivalent. Note
that the point at which an error is first found is the crucial one for
establishing whether to use approximation; for example,
<code>(#a1)abc(#a0)xyz</code> will not match <code>abcdxyz</code>, because the error occurs at
the <code>x</code>, where approximation is turned off.</p>
<p>Entire path segments may be matched approximately, so that
<code>(#a1)/foo/d/is/available/at/the/bar</code> allows one error in any path
segment. This is much less efficient than without the <code>(#a1)</code>, however,
since every directory in the path must be scanned for a possible
approximate match. It is best to place the <code>(#a1)</code> after any path
segments which are known to be correct.</p>
<hr />
<p><span id="Recursive-Globbing"></span></p>
<h3 id="1486-recursive-globbing"><a class="header" href="#1486-recursive-globbing">14.8.6 Recursive Globbing</a></h3>
<p>A pathname component of the form <code>(``foo``/)#</code> matches a path
consisting of zero or more directories matching the pattern <code>foo</code>.</p>
<p>As a shorthand, <code>**/</code> is equivalent to <code>(*/)#</code>; note that this
therefore matches files in the current directory as well as
subdirectories. Thus:</p>
<div class="example">
<pre><code class="language-example">ls -ld -- (*/)#bar
</code></pre>
</div>
<p>or</p>
<div class="example">
<pre><code class="language-example">ls -ld -- **/bar
</code></pre>
</div>
<p>does a recursive directory search for files named <code>bar</code> (potentially
including the file <code>bar</code> in the current directory). This form does not
follow symbolic links; the alternative form <code>***/</code> does, but is
otherwise identical. Neither of these can be combined with other forms
of globbing within the same path segment; in that case, the <code>*</code>
operators revert to their usual effect.</p>
<p>Even shorter forms are available when the option <code>GLOB_STAR_SHORT</code> is
set. In that case if no <code>/</code> immediately follows a <code>**</code> or <code>***</code> they are
treated as if both a <code>/</code> plus a further <code>*</code> are present. Hence:</p>
<div class="example">
<pre><code class="language-example">setopt GLOBSTARSHORT
ls -ld -- **.c
</code></pre>
</div>
<p>is equivalent to</p>
<div class="example">
<pre><code class="language-example">ls -ld -- **/*.c
</code></pre>
</div>
<hr />
<p><span id="Glob-Qualifiers"></span></p>
<h3 id="1487-glob-qualifiers"><a class="header" href="#1487-glob-qualifiers">14.8.7 Glob Qualifiers</a></h3>
<p><span id="index-globbing_002c-qualifiers"></span>
<span id="index-qualifiers_002c-globbing"></span></p>
<p>Patterns used for filename generation may end in a list of qualifiers
enclosed in parentheses. The qualifiers specify which filenames that
otherwise match the given pattern will be inserted in the argument list.</p>
<p><span id="index-BARE_005fGLOB_005fQUAL_002c-use-of"></span></p>
<p>If the option <code>BARE_GLOB_QUAL</code> is set, then a trailing set of
parentheses containing no <code>|</code> or <code>(</code> characters (or <code>~</code> if it is
special) is taken as a set of glob qualifiers. A glob subexpression that
would normally be taken as glob qualifiers, for example <code>(^x)</code>, can be
forced to be treated as part of the glob pattern by doubling the
parentheses, in this case producing <code>((^x))</code>.</p>
<p>If the option <code>EXTENDED_GLOB</code> is set, a different syntax for glob
qualifiers is available, namely <code>(#q``x``)</code> where <code>x</code> is any of the
same glob qualifiers used in the other format. The qualifiers must still
appear at the end of the pattern. However, with this syntax multiple
glob qualifiers may be chained together. They are treated as a logical
AND of the individual sets of flags. Also, as the syntax is unambiguous,
the expression will be treated as glob qualifiers just as long any
parentheses contained within it are balanced; appearance of <code>|</code>, <code>(</code>
or <code>~</code> does not negate the effect. Note that qualifiers will be
recognised in this form even if a bare glob qualifier exists at the end
of files if both options are set; however, mixed syntax should probably
be avoided for the sake of clarity. Note that within conditions using
the <code>[[</code> form the presence of a parenthesised expression <code>(#q``...``)</code>
at the end of a string indicates that globbing should be performed; the
expression may include glob qualifiers, but it is also valid if it is
simply <code>(#q)</code>. This does not apply to the right hand side of pattern
match operators as the syntax already has special significance.</p>
<p>A qualifier may be any one of the following:</p>
<ul>
<li>
<p><code>/</code><br />
directories</p>
</li>
<li>
<p><code>F</code><br />
full (i.e. non-empty) directories. Note that the opposite sense
<code>(^F``)</code> expands to empty directories and all non-directories. Use
<code>(/^F``)</code> for empty directories.</p>
</li>
<li>
<p><code>.</code><br />
plain files</p>
</li>
<li>
<p><code>@</code><br />
symbolic links</p>
</li>
<li>
<p><code>=</code><br />
sockets</p>
</li>
<li>
<p><code>p</code><br />
named pipes (FIFOs)</p>
</li>
<li>
<p><code>*</code><br />
<code>%</code><br />
device files (character or block special)</p>
</li>
<li>
<p><code>%b</code><br />
block special files</p>
</li>
<li>
<p><code>%c</code><br />
character special files</p>
</li>
<li>
<p><code>r</code><br />
owner-readable files (0400)</p>
</li>
<li>
<p><code>w</code><br />
<code>x</code><br />
<code>A</code><br />
group-readable files (0040)</p>
</li>
<li>
<p><code>I</code><br />
<code>E</code><br />
<code>R</code><br />
world-readable files (0004)</p>
</li>
<li>
<p><code>W</code><br />
<code>X</code><br />
<code>s</code><br />
setuid files (04000)</p>
</li>
<li>
<p><code>S</code><br />
setgid files (02000)</p>
</li>
<li>
<p><code>t</code><br />
files with the sticky bit (01000)</p>
</li>
<li>
<p><code>f``spec</code><br />
files with access rights matching <code>spec</code>. This <code>spec</code> may be a octal
number optionally preceded by a <code>=</code>, a <code>+</code>, or a <code>-</code>. If none
of these characters is given, the behavior is the same as for <code>=</code>.
The octal number describes the mode bits to be expected, if combined
with a <code>=</code>, the value given must match the file-modes exactly,
with a <code>+</code>, at least the bits in the given number must be set in
the file-modes, and with a <code>-</code>, the bits in the number must not be
set. Giving a <code>?</code> instead of a octal digit anywhere in the number
ensures that the corresponding bits in the file-modes are not
checked, this is only useful in combination with <code>=</code>.</p>
<p>If the qualifier <code>f</code> is followed by any other character anything
up to the next matching character (<code>[</code>, <code>{</code>, and <code>&lt;</code> match
<code>]</code>, <code>}</code>, and <code>&gt;</code> respectively, any other character matches
itself) is taken as a list of comma-separated <code>sub-spec</code>s. Each
<code>sub-spec</code> may be either an octal number as described above or a
list of any of the characters <code>u</code>, <code>g</code>, <code>o</code>, and <code>a</code>,
followed by a <code>=</code>, a <code>+</code>, or a <code>-</code>, followed by a list of any
of the characters <code>r</code>, <code>w</code>, <code>x</code>, <code>s</code>, and <code>t</code>, or an octal
digit. The first list of characters specify which access rights are
to be checked. If a <code>u</code> is given, those for the owner of the file
are used, if a <code>g</code> is given, those of the group are checked, a
<code>o</code> means to test those of other users, and the <code>a</code> says to
test all three groups. The <code>=</code>, <code>+</code>, and <code>-</code> again says how
the modes are to be checked and have the same meaning as described
for the first form above. The second list of characters finally says
which access rights are to be expected: <code>r</code> for read access, <code>w</code>
for write access, <code>x</code> for the right to execute the file (or to
search a directory), <code>s</code> for the setuid and setgid bits, and <code>t</code>
for the sticky bit.</p>
<p>Thus, <code>*(f70?)</code> gives the files for which the owner has read,
write, and execute permission, and for which other group members
have no rights, independent of the permissions for other users. The
pattern <code>*(f-100)</code> gives all files for which the owner does not
have execute permission, and <code>*(f:gu+w,o-rx:)</code> gives the files for
which the owner and the other members of the group have at least
write permission, and for which other users dont have read or
execute permission.</p>
</li>
<li>
<p><code>e``string</code><br />
<code>+``cmd</code><br />
The <code>string</code> will be executed as shell code. The filename will be
included in the list if and only if the code returns a zero status
(usually the status of the last command).</p>
<p>In the first form, the first character after the <code>e</code> will be used
as a separator and anything up to the next matching separator will
be taken as the <code>string</code>; <code>[</code>, <code>{</code>, and <code>&lt;</code> match <code>]</code>,
<code>}</code>, and <code>&gt;</code>, respectively, while any other character
matches itself. Note that expansions must be quoted in the <code>string</code>
to prevent them from being expanded before globbing is done.
<code>string</code> is then executed as shell code. The string <code>globqual</code> is
appended to the array <code>zsh_eval_context</code> the duration of execution.</p>
<p><span id="index-REPLY_002c-use-of-1"></span>
<span id="index-reply_002c-use-of"></span></p>
<p>During the execution of <code>string</code> the filename currently being tested
is available in the parameter <code>REPLY</code>; the parameter may be altered
to a string to be inserted into the list instead of the original
filename. In addition, the parameter <code>reply</code> may be set to an array
or a string, which overrides the value of <code>REPLY</code>. If set to an
array, the latter is inserted into the command line word by word.</p>
<p>For example, suppose a directory contains a single file <code>lonely</code>.
Then the expression <code>*(e:reply=(${REPLY}{1,2}):)</code> will cause the
words <code>lonely1</code> and <code>lonely2</code> to be inserted into the command
line. Note the quoting of <code>string</code>.</p>
<p>The form <code>+``cmd</code> has the same effect, but no delimiters appear
around <code>cmd</code>. Instead, <code>cmd</code> is taken as the longest sequence of
characters following the <code>+</code> that are alphanumeric or underscore.
Typically <code>cmd</code> will be the name of a shell function that contains
the appropriate test. For example,</p>
<div class="example">
<pre><code class="language-example">nt() { [[ $REPLY -nt $NTREF ]] }
NTREF=reffile
ls -ld -- *(+nt)
</code></pre>
</div>
<p>lists all files in the directory that have been modified more
recently than <code>reffile</code>.</p>
</li>
<li>
<p><code>d``dev</code><br />
files on the device <code>dev</code></p>
</li>
<li>
<p><code>l</code>[<code>-</code>|<code>+</code>]<code>ct</code><br />
files having a link count less than <code>ct</code> (<code>-</code>), greater than <code>ct</code>
(<code>+</code>), or equal to <code>ct</code></p>
</li>
<li>
<p><code>U</code><br />
files owned by the effective user ID</p>
</li>
<li>
<p><code>G</code><br />
files owned by the effective group ID</p>
</li>
<li>
<p><code>u``id</code><br />
files owned by user ID <code>id</code> if that is a number. Otherwise, <code>id</code>
specifies a user name: the character after the <code>u</code> will be taken
as a separator and the string between it and the next matching
separator will be taken as a user name. The starting separators
<code>[</code>, <code>{</code>, and <code>&lt;</code> match the final separators <code>]</code>, <code>}</code>,
and <code>&gt;</code>, respectively; any other character matches itself. The
selected files are those owned by this user. For example, <code>u:foo:</code>
or <code>u[foo]</code> selects files owned by user <code>foo</code>.</p>
</li>
<li>
<p><code>g``id</code><br />
like <code>u``id</code> but with group IDs or names</p>
</li>
<li>
<p><code>a</code>[<code>Mwhms</code>][<code>-</code>|<code>+</code>]<code>n</code><br />
files accessed exactly <code>n</code> days ago. Files accessed within the last
<code>n</code> days are selected using a negative value for <code>n</code> (<code>-``n</code>). Files
accessed more than <code>n</code> days ago are selected by a positive <code>n</code> value
(<code>+``n</code>). Optional unit specifiers <code>M</code>, <code>w</code>, <code>h</code>, <code>m</code> or
<code>s</code> (e.g. <code>ah5</code>) cause the check to be performed with months
(of 30 days), weeks, hours, minutes or seconds instead of days,
respectively. An explicit <code>d</code> for days is also allowed.</p>
<p>Any fractional part of the difference between the access time and
the current part in the appropriate units is ignored in the
comparison. For instance, <code>echo *(ah-5)</code> would echo files accessed
within the last five hours, while <code>echo *(ah+5)</code> would echo files
accessed at least six hours ago, as times strictly between five and
six hours are treated as five hours.</p>
</li>
<li>
<p><code>m</code>[<code>Mwhms</code>][<code>-</code>|<code>+</code>]<code>n</code><br />
like the file access qualifier, except that it uses the file
modification time.</p>
</li>
<li>
<p><code>c</code>[<code>Mwhms</code>][<code>-</code>|<code>+</code>]<code>n</code><br />
like the file access qualifier, except that it uses the file inode
change time.</p>
</li>
<li>
<p><code>L</code>[<code>+</code>|<code>-</code>]<code>n</code><br />
files less than <code>n</code> bytes (<code>-</code>), more than <code>n</code> bytes (<code>+</code>), or
exactly <code>n</code> bytes in length.</p>
<p>If this flag is directly followed by a <em>size specifier</em> <code>k</code>
(<code>K</code>), <code>m</code> (<code>M</code>), or <code>p</code> (<code>P</code>) (e.g. <code>Lk-50</code>) the
check is performed with kilobytes, megabytes, or blocks (of 512
bytes) instead. (On some systems additional specifiers are available
for gigabytes, <code>g</code> or <code>G</code>, and terabytes, <code>t</code> or <code>T</code>.) If a
size specifier is used a file is regarded as &quot;exactly&quot; the size if
the file size rounded up to the next unit is equal to the test size.
Hence <code>*(Lm1)</code> matches files from 1 byte up to 1 Megabyte
inclusive. Note also that the set of files &quot;less than&quot; the test size
only includes files that would not match the equality test; hence
<code>*(Lm-1)</code> only matches files of zero size.</p>
</li>
<li>
<p><code>^</code><br />
negates all qualifiers following it</p>
</li>
<li>
<p><code>-</code><br />
toggles between making the qualifiers work on symbolic links (the
default) and the files they point to</p>
</li>
<li>
<p><code>M</code><br />
sets the <code>MARK_DIRS</code> option for the current pattern
<span id="index-MARK_005fDIRS_002c-setting-in-pattern"></span></p>
</li>
<li>
<p><code>T</code><br />
appends a trailing qualifier mark to the filenames, analogous to the
<code>LIST_TYPES</code> option, for the current pattern (overrides <code>M</code>)</p>
</li>
<li>
<p><code>N</code><br />
sets the <code>NULL_GLOB</code> option for the current pattern
<span id="index-NULL_005fGLOB_002c-setting-in-pattern"></span></p>
</li>
<li>
<p><code>D</code><br />
sets the <code>GLOB_DOTS</code> option for the current pattern
<span id="index-GLOB_005fDOTS_002c-setting-in-pattern"></span></p>
</li>
<li>
<p><code>n</code><br />
sets the <code>NUMERIC_GLOB_SORT</code> option for the current pattern
<span id="index-NUMERIC_005fGLOB_005fSORT_002c-setting-in-pattern"></span></p>
</li>
<li>
<p><code>Y``n</code><br />
enables short-circuit mode: the pattern will expand to at most <code>n</code>
filenames. If more than <code>n</code> matches exist, only the first <code>n</code>
matches in directory traversal order will be considered.</p>
<p>Implies <code>oN</code> when no <code>o``c</code> qualifier is used.</p>
</li>
<li>
<p><code>o``c</code><br />
specifies how the names of the files should be sorted. If <code>c</code> is <code>n</code>
they are sorted by name; if it is <code>L</code> they are sorted depending on
the size (length) of the files; if <code>l</code> they are sorted by the number
of links; if <code>a</code>, <code>m</code>, or <code>c</code> they are sorted by the time of the
last access, modification, or inode change respectively; if <code>d</code>,
files in subdirectories appear before those in the current directory
at each level of the search — this is best combined with other
criteria, for example <code>odon</code> to sort on names for files within the
same directory; if <code>N</code>, no sorting is performed. Note that <code>a</code>, <code>m</code>,
and <code>c</code> compare the age against the current time, hence the first
name in the list is the youngest file. Also note that the modifiers
<code>^</code> and <code>-</code> are used, so <code>*(^-oL)</code> gives a list of all files
sorted by file size in descending order, following any symbolic
links. Unless <code>oN</code> is used, multiple order specifiers may occur to
resolve ties.</p>
<p>The default sorting is <code>n</code> (by name) unless the <code>Y</code> glob qualifier
is used, in which case it is <code>N</code> (unsorted).</p>
<p><code>oe</code> and <code>o+</code> are special cases; they are each followed by shell
code, delimited as for the <code>e</code> glob qualifier and the <code>+</code> glob
qualifier respectively (see above). The code is executed for each
matched file with the parameter <code>REPLY</code> set to the name of the file
on entry and <code>globsort</code> appended to <code>zsh_eval_context</code>. The code
should modify the parameter <code>REPLY</code> in some fashion. On return, the
value of the parameter is used instead of the file name as the
string on which to sort. Unlike other sort operators, <code>oe</code> and <code>o+</code>
may be repeated, but note that the maximum number of sort operators
of any kind that may appear in any glob expression is 12.</p>
</li>
<li>
<p><code>O``c</code><br />
like <code>o</code>, but sorts in descending order; i.e. <code>*(^oc)</code> is the
same as <code>*(Oc)</code> and <code>*(^Oc)</code> is the same as <code>*(oc)</code>; <code>Od</code>
puts files in the current directory before those in subdirectories
at each level of the search.</p>
</li>
<li>
<p><code>[``beg</code>[<code>,``end</code>]<code>]</code><br />
specifies which of the matched filenames should be included in the
returned list. The syntax is the same as for array subscripts. <code>beg</code>
and the optional <code>end</code> may be mathematical expressions. As in
parameter subscripting they may be negative to make them count from
the last match backward. E.g.: <code>*(-OL[1,3])</code> gives a list of the
names of the three largest files.</p>
</li>
<li>
<p><code>P``string</code><br />
The <code>string</code> will be prepended to each glob match as a separate
word. <code>string</code> is delimited in the same way as arguments to the <code>e</code>
glob qualifier described above. The qualifier can be repeated; the
words are prepended separately so that the resulting command line
contains the words in the same order they were given in the list of
glob qualifiers.</p>
<p>A typical use for this is to prepend an option before all
occurrences of a file name; for example, the pattern <code>*(P:-f:)</code>
produces the command line arguments <code>-f</code> <code>file1</code> <code>-f</code> <code>file2</code> ...</p>
<p>If the modifier <code>^</code> is active, then <code>string</code> will be appended
instead of prepended. Prepending and appending is done independently
so both can be used on the same glob expression; for example by
writing <code>*(P:foo:^P:bar:^P:baz:)</code> which produces the command line
arguments <code>foo</code> <code>baz</code> <code>file1</code> <code>bar</code> ...</p>
</li>
</ul>
<p>More than one of these lists can be combined, separated by commas. The
whole list matches if at least one of the sublists matches (they are
ored, the qualifiers in the sublists are anded). Some
qualifiers, however, affect all matches generated, independent of the
sublist in which they are given. These are the qualifiers <code>M</code>, <code>T</code>,
<code>N</code>, <code>D</code>, <code>n</code>, <code>o</code>, <code>O</code> and the subscripts given in brackets
(<code>[``...``]</code>).</p>
<p>If a <code>:</code> appears in a qualifier list, the remainder of the expression
in parenthesis is interpreted as a modifier (see <a href="#Modifiers">Modifiers</a>
in <a href="#History-Expansion">History Expansion</a>). Each modifier must be
introduced by a separate <code>:</code>. Note also that the result after
modification does not have to be an existing file. The name of any
existing file can be followed by a modifier of the form <code>(:``...``)</code>
even if no actual filename generation is performed, although note that
the presence of the parentheses causes the entire expression to be
subjected to any global pattern matching options such as <code>NULL_GLOB</code>.
Thus:</p>
<div class="example">
<pre><code class="language-example">ls -ld -- *(-/)
</code></pre>
</div>
<p>lists all directories and symbolic links that point to directories, and</p>
<div class="example">
<pre><code class="language-example">ls -ld -- *(-@)
</code></pre>
</div>
<p>lists all broken symbolic links, and</p>
<div class="example">
<pre><code class="language-example">ls -ld -- *(%W)
</code></pre>
</div>
<div class="example">
<pre><code class="language-example">ls -ld -- *(W,X)
</code></pre>
</div>
<p>lists all files in the current directory that are</p>
<div class="example">
<pre><code class="language-example">print -rC1 /tmp/foo*(u0^@:t)
</code></pre>
</div>
<p>outputs the basename of all root-owned files beginning with the string
<code>foo</code> in <code>/tmp</code>, ignoring symlinks, and</p>
<div class="example">
<pre><code class="language-example">ls -ld -- *.*~(lex|parse).[ch](^D^l1)
</code></pre>
</div>
<p>lists all files having a link count of one whose names contain a dot
(but not those starting with a dot, since <code>GLOB_DOTS</code> is explicitly
switched off) except for <code>lex.c</code>, <code>lex.h</code>, <code>parse.c</code> and <code>parse.h</code>.</p>
<div class="example">
<pre><code class="language-example">print -rC1 b*.pro(#q:s/pro/shmo/)(#q.:s/builtin/shmiltin/)
</code></pre>
</div>
<p>demonstrates how colon modifiers and other qualifiers may be chained
together. The ordinary qualifier <code>.</code> is applied first, then the colon
modifiers in order from left to right. So if <code>EXTENDED_GLOB</code> is set and
the base pattern matches the regular file <code>builtin.pro</code>, the shell will
print <code>shmiltin.shmo</code>.</p>
<hr />
<p>This document was generated on <em>February 15, 2020</em> using
<a href="http://www.nongnu.org/texi2html/"><em>texi2html 5.0</em></a>.<br />
Zsh version 5.8, released on February 14, 2020.</p>
</main>
<nav class="nav-wrapper" aria-label="Page navigation">
<!-- Mobile navigation buttons -->
<a rel="prev" href="Prompt-Expansion.html" class="mobile-nav-chapters previous" title="Previous chapter" aria-label="Previous chapter" aria-keyshortcuts="Left">
<i class="fa fa-angle-left"></i>
</a>
<a rel="next" href="Parameters.html" class="mobile-nav-chapters next" title="Next chapter" aria-label="Next chapter" aria-keyshortcuts="Right">
<i class="fa fa-angle-right"></i>
</a>
<div style="clear: both"></div>
</nav>
</div>
</div>
<nav class="nav-wide-wrapper" aria-label="Page navigation">
<a rel="prev" href="Prompt-Expansion.html" class="nav-chapters previous" title="Previous chapter" aria-label="Previous chapter" aria-keyshortcuts="Left">
<i class="fa fa-angle-left"></i>
</a>
<a rel="next" href="Parameters.html" class="nav-chapters next" title="Next chapter" aria-label="Next chapter" aria-keyshortcuts="Right">
<i class="fa fa-angle-right"></i>
</a>
</nav>
</div>
<!-- Livereload script (if served using the cli tool) -->
<script type="text/javascript">
var socket = new WebSocket("ws://localhost:3000/__livereload");
socket.onmessage = function (event) {
if (event.data === "reload") {
socket.close();
location.reload();
}
};
window.onbeforeunload = function() {
socket.close();
}
</script>
<script type="text/javascript">
window.playground_copyable = true;
</script>
<script src="elasticlunr.min.js" type="text/javascript" charset="utf-8"></script>
<script src="mark.min.js" type="text/javascript" charset="utf-8"></script>
<script src="searcher.js" type="text/javascript" charset="utf-8"></script>
<script src="clipboard.min.js" type="text/javascript" charset="utf-8"></script>
<script src="highlight.js" type="text/javascript" charset="utf-8"></script>
<script src="book.js" type="text/javascript" charset="utf-8"></script>
<!-- Custom JS scripts -->
</body>
</html>