bash/doc/bashref.texi

\input texinfo.tex @c -*- texinfo -*-
@c %**start of header
@setfilename bashref.info
@settitle Bash Reference Manual

@include version.texi
@c %**end of header

@copying
This text is a brief description of the features that are present in
the Bash shell (version @value{VERSION}, @value{UPDATED}).

This is Edition @value{EDITION}, last updated @value{UPDATED},
of @cite{The GNU Bash Reference Manual},
for @code{Bash}, Version @value{VERSION}.

Copyright @copyright{} 1988--2026 Free Software Foundation, Inc.

@quotation
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License, Version 1.3 or
any later version published by the Free Software Foundation; with no
Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts.
A copy of the license is included in the section entitled
``GNU Free Documentation License''.
@end quotation
@end copying

@defcodeindex bt
@defcodeindex rw
@set BashFeatures

@dircategory Basics
@direntry
* Bash: (bash).                     The GNU Bourne-Again SHell.
@end direntry

@finalout

@titlepage
@title Bash Reference Manual
@subtitle Reference Documentation for Bash
@subtitle Edition @value{EDITION}, for @code{Bash} Version @value{VERSION}.
@subtitle @value{UPDATED-MONTH}
@author Chet Ramey, Case Western Reserve University
@author Brian Fox, Free Software Foundation

@page
@vskip 0pt plus 1filll
@insertcopying

@end titlepage

@contents

@ifnottex
@node Top, Introduction, (dir), (dir)
@top Bash Features

This text is a brief description of the features that are present in
the Bash shell (version @value{VERSION}, @value{UPDATED}).
The Bash home page is @url{http://www.gnu.org/software/bash/}.

This is Edition @value{EDITION}, last updated @value{UPDATED},
of @cite{The GNU Bash Reference Manual},
for @code{Bash}, Version @value{VERSION}.

Bash contains features that appear in other popular shells, and some
features that only appear in Bash.
Some of the shells that Bash has
borrowed concepts from are the Bourne Shell (@file{sh}), the Korn Shell
(@file{ksh}), and the C-shell (@file{csh} and its successor,
@file{tcsh}).
The following menu breaks the features up into
categories, noting which features were inspired by other shells and
which are specific to Bash.

This manual is meant as a brief introduction to features found in Bash.
The Bash manual page should be used as the definitive
reference on shell behavior.

@menu
* Introduction::		An introduction to the shell.
* Definitions::			Some definitions used in the rest of this
				manual.
* Basic Shell Features::	The shell "building blocks".
* Shell Builtin Commands::	Commands that are a part of the shell.
* Shell Variables::		Variables used or set by Bash.
* Bash Features::		Features found only in Bash.
* Job Control::			What job control is and how Bash allows you
				to use it.
* Command Line Editing::	Chapter describing the command line
				editing features.
* Using History Interactively::	Command History Expansion
* Installing Bash::		How to build and install Bash on your system.
* Reporting Bugs::		How to report bugs in Bash.
* Major Differences From The Bourne Shell::	A terse list of the differences
						between Bash and historical
						versions of /bin/sh.
* GNU Free Documentation License::	Copying and sharing this documentation.
* Indexes::			Various indexes for this manual.
@end menu
@end ifnottex

@node Introduction
@chapter Introduction
@menu
* What is Bash?::		A short description of Bash.
* What is a shell?::		A brief introduction to shells.
@end menu

@node What is Bash?
@section What is Bash?

Bash is the shell, or command language interpreter,
for the @sc{gnu} operating system.
The name is an acronym for the @samp{Bourne-Again SHell},
a pun on Stephen Bourne, the author of the direct ancestor of
the current Unix shell @code{sh}, 
which appeared in the Seventh Edition Bell Labs Research version
of Unix.

Bash is largely compatible with @code{sh} and incorporates useful
features from the Korn shell @code{ksh} and the C shell @code{csh}.
It is intended to be a conformant implementation of the @sc{ieee}
@sc{posix} Shell and Tools portion of the @sc{ieee} @sc{posix}
specification (@sc{ieee} Standard 1003.1).
It offers functional improvements over @code{sh} for both interactive and
programming use.

While the @sc{gnu} operating system provides other shells, including
a version of @code{csh}, Bash is the default shell.
Like other @sc{gnu} software, Bash is quite portable.  It currently runs
on nearly every version of Unix and a few other operating systems @minus{}
independently-supported ports exist for Windows and other platforms.

@node What is a shell?
@section What is a shell?

At its base, a shell is simply a macro processor that executes commands.
The term macro processor means functionality where text
and symbols are expanded to create larger expressions.

A Unix shell is both a command interpreter and a programming language.
As a command interpreter, the shell provides the user
interface to the rich set of @sc{gnu} utilities.
The programming language features allow these utilities to be combined.
Users can create files containing commands, and these become
commands themselves. 
These new commands have the same status as
system commands in directories such as @file{/bin}, allowing users
or groups to establish custom environments to automate their common
tasks.

Shells may be used interactively or non-interactively.  In
interactive mode, they accept input typed from the keyboard.
When executing non-interactively, shells execute commands read
from a file or a string.

A shell allows execution of @sc{gnu} commands, both synchronously and
asynchronously.
The shell waits for synchronous commands to complete before accepting
more input; asynchronous commands continue to execute in parallel
with the shell while it reads and executes additional commands.
The @dfn{redirection} constructs permit
fine-grained control of the input and output of those commands.
Moreover, the shell allows control over the contents of commands'
environments.

Shells also provide a small set of built-in
commands (@dfn{builtins}) implementing functionality impossible
or inconvenient to obtain via separate utilities.
For example, @code{cd}, @code{break}, @code{continue}, and
@code{exec} cannot be implemented outside of the shell because
they directly manipulate the shell itself.
The @code{history}, @code{getopts}, @code{kill}, or @code{pwd}
builtins, among others, could be implemented in separate utilities,
but they are more convenient to use as builtin commands.
All of the shell builtins are described in subsequent sections.

While executing commands is essential, most of the power (and
complexity) of shells is due to their embedded programming languages. 
Like any high-level language, the shell provides variables,
flow control constructs, quoting, and functions. 

Shells offer features geared specifically for
interactive use rather than to augment the programming language. 
These interactive features include job control, command line
editing, command history and aliases.
This manual describes how Bash provides all of these features.

@node Definitions
@chapter Definitions
These definitions are used throughout the remainder of this manual.

@table @code

@item POSIX
@cindex POSIX
A family of open system standards based on Unix.  Bash
is primarily concerned with the Shell and Utilities portion of the
@sc{posix} 1003.1 standard. 

@item blank
A space or tab character.

@item whitespace
A character belonging to the @code{space} character class in the
current locale, or for which @code{isspace()} returns true.

@item builtin
@cindex builtin
A command that is implemented internally by the shell itself, rather
than by an executable program somewhere in the file system.

@item control operator
@cindex control operator
A @code{token} that performs a control function.
It is a @code{newline} or one of the following:
@samp{||}, @samp{&&}, @samp{&}, @samp{;}, @samp{;;}, @samp{;&}, @samp{;;&},
@samp{|}, @samp{|&}, @samp{(}, or @samp{)}.

@item exit status
@cindex exit status
The value returned by a command to its caller.
The value is restricted to eight bits, so the maximum value is 255.

@item field
@cindex field
A unit of text that is the result of one of the shell expansions.
After expansion, when executing a command, the resulting fields are
used as the command name and arguments.

@item filename
@cindex filename
A string of characters used to identify a file.

@item job
@cindex job
A set of processes comprising a pipeline, and any processes descended
from it, that are all in the same process group.

@item job control
@cindex job control
A mechanism by which users can selectively stop (suspend) and restart
(resume) execution of processes.

@item metacharacter
@cindex metacharacter
A character that, when unquoted, separates words.
A metacharacter is a @code{space}, @code{tab}, @code{newline},
or one of the following characters:
@samp{|}, @samp{&}, @samp{;}, @samp{(}, @samp{)}, @samp{<}, or
@samp{>}.

@item name
@cindex name
@cindex identifier
A @code{word} consisting solely of letters, numbers, and underscores,
and beginning with a letter or underscore.
@code{Name}s are used as shell variable and function names.
Also referred to as an @code{identifier}.

@item operator
@cindex operator, shell
A @code{control operator} or a @code{redirection operator}.
@xref{Redirections}, for a list of redirection operators.
Operators contain at least one unquoted @code{metacharacter}.

@item process group
@cindex process group
A collection of related processes each having the same process
group @sc{id}.

@item process group ID
@cindex process group ID
A unique identifier that represents a @code{process group}
during its lifetime.

@item reserved word
@cindex reserved word
A @code{word} that has a special meaning to the shell.
Most reserved words introduce shell flow control constructs, such as
@code{for} and @code{while}.

@item return status
@cindex return status
A synonym for @code{exit status}.

@item signal
@cindex signal
A mechanism by which a process may be notified by the kernel
of an event occurring in the system.

@item special builtin
@cindex special builtin
A shell builtin command that has been classified as special by the
@sc{posix} standard.

@item token
@cindex token
A sequence of characters considered a single unit by the shell.
It is either a @code{word} or an @code{operator}.

@item word
@cindex word
A sequence of characters treated as a unit by the shell.
Words may not include unquoted @code{metacharacters}.
@end table

@node Basic Shell Features
@chapter Basic Shell Features
@cindex Bourne shell

Bash is an acronym for @samp{Bourne-Again SHell}.
The Bourne shell is
the traditional Unix shell originally written by Stephen Bourne.
All of the Bourne shell builtin commands are available in Bash, and
the rules for evaluation and quoting are taken from the @sc{posix}
specification for the ``standard'' Unix shell.

This chapter briefly summarizes the shell's ``building blocks'':
commands, control structures, shell functions, shell @i{parameters},
shell expansions,
@i{redirections}, which are a way to direct input and output from
and to named files, and how the shell executes commands.

@menu
* Shell Syntax::		What your input means to the shell.
* Shell Commands::		The types of commands you can use.
* Shell Functions::		Grouping commands by name.
* Shell Parameters::		How the shell stores values.
* Shell Expansions::		How Bash expands parameters and the various
				expansions available.
* Redirections::		A way to control where input and output go.
* Executing Commands::		What happens when you run a command.
* Shell Scripts::		Executing files of shell commands.
@end menu

@node Shell Syntax
@section Shell Syntax
@menu
* Shell Operation::	The basic operation of the shell.
* Quoting::		How to remove the special meaning from characters.
* Comments::		How to specify comments.
@end menu

When the shell reads input, it proceeds through a
sequence of operations.  If the input indicates the beginning of a
comment, the shell ignores the comment symbol (@samp{#}), and the rest
of that line.
                                
Otherwise, roughly speaking,  the shell reads its input and
divides the input into words and operators, employing the quoting rules
to select which meanings to assign various words and characters.

The shell then parses these tokens into commands and other constructs,
removes the special meaning of certain words or characters, expands
others, redirects input and output as needed, executes the specified
command, waits for the command's exit status, and makes that exit status
available for further inspection or processing.

@node Shell Operation
@subsection Shell Operation

The following is a brief description of the shell's operation when it
reads and executes a command.
Basically, the shell does the following:

@enumerate
@item
Reads its input from a file (@pxref{Shell Scripts}), from a string
supplied as an argument to the @option{-c} invocation option
(@pxref{Invoking Bash}), or from the user's terminal.

@item
Breaks the input into words and operators, obeying the quoting rules
described in @ref{Quoting}.
These tokens are separated by @code{metacharacters}.
This step performs alias expansion (@pxref{Aliases}).

@item
Parses the tokens into simple and compound commands
(@pxref{Shell Commands}).

@item
Performs the various shell expansions (@pxref{Shell Expansions}), breaking
the expanded tokens into lists of filenames (@pxref{Filename Expansion})
and commands and arguments.

@item
Performs any necessary redirections (@pxref{Redirections}) and removes
the redirection operators and their operands from the argument list.

@item
Executes the command (@pxref{Executing Commands}).

@item
Optionally waits for the command to complete and collects its exit
status (@pxref{Exit Status}).

@end enumerate

@node Quoting
@subsection Quoting
@cindex quoting
@menu
* Escape Character::	How to remove the special meaning from a single
			character.
* Single Quotes::	How to inhibit all interpretation of a sequence
			of characters.
* Double Quotes::	How to suppress most of the interpretation of a
			sequence of characters.
* ANSI-C Quoting::	How to expand ANSI-C sequences in quoted strings.
* Locale Translation::	How to translate strings into different languages.
@end menu

Quoting is used to remove the special meaning of certain
characters or words to the shell.
Quoting can be used to
disable special treatment for special characters, to prevent
reserved words from being recognized as such, and to prevent
parameter expansion.

Each of the shell metacharacters (@pxref{Definitions})
has special meaning to the shell and must be quoted if it is to
represent itself.

When the command history expansion facilities are being used
(@pxref{History Interaction}), the
@dfn{history expansion} character, usually @samp{!}, must be quoted
to prevent history expansion.
@xref{Bash History Facilities}, for more details concerning history
expansion.

There are four quoting mechanisms: the
@dfn{escape character}, single quotes, double quotes,
and dollar-single quotes.

@node Escape Character
@subsubsection Escape Character
A non-quoted backslash @samp{\} is the Bash escape character.
It preserves the literal value of the next character that follows,
removing any special meaning it has,
with the exception of @code{newline}.
If a @code{\newline} pair appears, and the backslash itself is not quoted,
the @code{\newline} is treated as a line continuation (that is, it is
removed from the input stream and effectively ignored).

@node Single Quotes
@subsubsection Single Quotes

Enclosing characters in single quotes (@samp{'}) preserves the literal value
of each character within the quotes.
A single quote may not occur
between single quotes, even when preceded by a backslash.

@node Double Quotes
@subsubsection Double Quotes

Enclosing characters in double quotes (@samp{"}) preserves the literal value
of all characters within the quotes, with the exception of
@samp{$}, @samp{`}, @samp{\},
and, when history expansion is enabled, @samp{!}.
When the shell is in
@sc{posix} mode (@pxref{Bash POSIX Mode}),
the @samp{!} has no special meaning
within double quotes, even when history expansion is enabled.
The characters @samp{$} and @samp{`}
retain their special meaning within double quotes (@pxref{Shell Expansions}).
The backslash retains its special meaning only when followed by one of
the following characters:
@samp{$}, @samp{`}, @samp{"}, @samp{\}, or @code{newline}.
Within double quotes, backslashes that are followed by one of these
characters are removed.
Backslashes preceding characters without a
special meaning are left unmodified.

A double quote may be quoted within double quotes by preceding it with
a backslash.
If enabled, history expansion will be performed unless an
@samp{!}
appearing in double quotes is escaped using a backslash.
The backslash preceding the @samp{!} is not removed.

The special parameters @samp{*} and @samp{@@} have special meaning
when in double quotes (@pxref{Shell Parameter Expansion}).

@node ANSI-C Quoting
@subsubsection ANSI-C Quoting
@cindex quoting, ANSI
@cindex dollar-single quote quoting

Character sequences of the form @code{$'@var{string}'} are treated as
a special kind of single quotes.
The sequence expands to @var{string}, with backslash-escaped characters
in @var{string} replaced as specified by the ANSI C standard.
Backslash escape sequences, if present, are decoded as follows:

@table @code
@item \a
alert (bell)
@item \b
backspace
@item \e
@itemx \E
An escape character (not in ANSI C).
@item \f
form feed
@item \n
newline
@item \r
carriage return
@item \t
horizontal tab
@item \v
vertical tab
@item \\
backslash
@item \'
single quote
@item \"
double quote
@item \?
question mark
@item \@var{nnn}
The eight-bit character whose value is the octal value @var{nnn}
(one to three octal digits).
@item \x@var{HH}
The eight-bit character whose value is the hexadecimal value @var{HH}
(one or two hex digits).
@item \u@var{HHHH}
The Unicode (ISO/IEC 10646) character whose value is the hexadecimal value
@var{HHHH} (one to four hex digits).
@item \U@var{HHHHHHHH}
The Unicode (ISO/IEC 10646) character whose value is the hexadecimal value
@var{HHHHHHHH} (one to eight hex digits).
@item \c@var{x}
A control-@var{x} character.
@end table

@noindent
The expanded result is single-quoted, as if the dollar sign had not
been present.

@node Locale Translation
@subsubsection Locale-Specific Translation
@cindex localization
@cindex internationalization
@cindex native languages
@cindex translation, native languages
@menu
* Creating Internationalized Scripts::	How to use translations and different
					languages in your scripts.
@end menu

Prefixing a double-quoted string with a dollar sign (@samp{$}), such
as @verb{|$"hello, world"|},
causes the string to be translated according to the current locale.
The @code{gettext} infrastructure performs the lookup and 
translation, using the @code{LC_MESSAGES}, @code{TEXTDOMAINDIR},
and @code{TEXTDOMAIN} shell variables, as explained below.
See the gettext documentation for additional details not covered here.
If the current locale is @code{C} or @code{POSIX},
if there are no translations available,
or if the string is not translated, the dollar sign is ignored,
and the string is treated as double-quoted as described above.
Since this is a form of double quoting, the string remains double-quoted
by default, whether or not it is translated and replaced.
If the @code{noexpand_translation} option is enabled
using the @code{shopt} builtin (@pxref{The Shopt Builtin}),
translated strings are single-quoted instead of double-quoted.

The rest of this section is a brief overview of how you use gettext to
create translations for strings in a shell script named @var{scriptname}.
There are more details in the gettext documentation.

@node Creating Internationalized Scripts
@cindex internationalized scripts
@cindex string translations
Once you've marked the strings in your script
that you want to translate using $"@dots{}",
you create a gettext "template" file using the command

@example
bash --dump-po-strings @var{scriptname} > @var{domain}.pot
@end example

@noindent
The @var{domain} is your @dfn{message domain}.
It's just an arbitrary string that's used to identify the files gettext
needs, like a package or script name.
It needs to be unique among all
the message domains on systems where you install the translations, so
gettext knows which translations correspond to your script.
You'll use the template file to create translations for each target language.
The template file conventionally has the suffix @samp{.pot}.

You copy this template file to a separate file for each target language
you want to support (called "PO" files, which use the suffix @samp{.po}).
PO files use various naming conventions, but
when you are working to translate a template file into a particular
language, you first copy the template file to a file whose name is the
language you want to target, with the @samp{.po} suffix.
For instance, the Spanish translations of your strings would be
in a file named @samp{es.po}, and to get started using a message
domain named "example," you would run

@example
cp example.pot es.po
@end example

@noindent
Ultimately, PO files are often named @var{domain}.po and installed in
directories that contain multiple translation files for a particular language.

Whichever naming convention you choose, you will need to translate the
strings in the PO files into the appropriate languages.
This has to be done manually.

When you have the translations and PO files complete, you'll use the
gettext tools to produce what are called "MO" files, which are compiled
versions of the PO files the gettext tools use to look up translations
efficiently.
MO files are also called "message catalog" files.
You use the @command{msgfmt} program to do this.
For instance, if you had a file with Spanish translations, you could run

@example
msgfmt -o es.mo es.po
@end example

@noindent
to produce the corresponding MO file.

Once you have the MO files, you decide where to install them and use the
@code{TEXTDOMAINDIR} shell variable to tell the gettext tools where they are.
Make sure to use the same message domain to name the MO files 
as you did for the PO files when you install them.

@vindex LANG
@vindex LC_MESSAGES
@vindex TEXTDOMAIN
@vindex TEXTDOMAINDIR
Your users will use the @env{LANG} or @env{LC_MESSAGES} shell variables to
select the desired language.

You set the @env{TEXTDOMAIN} variable to the script's message domain.
As above, you use the message domain to name your translation files.

You, or possibly your users, set the @env{TEXTDOMAINDIR} variable to the
name of a directory where the message catalog files are stored.
If you install the message files into the system's standard message catalog
directory, you don't need to worry about this variable.

The directory where the message catalog files are stored varies between
systems.
Some use the message catalog selected by the @env{LC_MESSAGES}
shell variable.
Others create the name of the message catalog from the value of the
@env{TEXTDOMAIN} shell variable, possibly adding the @samp{.mo} suffix.
If you use the @env{TEXTDOMAIN} variable, you may need to set the
@env{TEXTDOMAINDIR} variable to the location of the message catalog files,
as above.
It's common to use both variables in this fashion:
@env{$TEXTDOMAINDIR}/@env{$LC_MESSAGES}/LC_MESSAGES/@env{$TEXTDOMAIN}.mo.

If you used that last convention, and you wanted to store the message
catalog files with Spanish (es) and Esperanto (eo) translations into a
local directory you use for custom translation files, you could run

@example
TEXTDOMAIN=example
TEXTDOMAINDIR=/usr/local/share/locale

cp es.mo $@{TEXTDOMAINDIR@}/es/LC_MESSAGES/$@{TEXTDOMAIN@}.mo
cp eo.mo $@{TEXTDOMAINDIR@}/eo/LC_MESSAGES/$@{TEXTDOMAIN@}.mo
@end example

When all of this is done, and the message catalog files containing the
compiled translations are installed in the correct location,
your users will be able to see translated strings
in any of the supported languages by setting the @env{LANG} or
@env{LC_MESSAGES} environment variables before running your script.

@node Comments
@subsection Comments
@cindex comments, shell

In a non-interactive shell, or an interactive shell in which the
@code{interactive_comments} option to the @code{shopt}
builtin is enabled (@pxref{The Shopt Builtin}),
a word beginning with @samp{#} introduces a comment.
A word begins at the beginning of a line, after unquoted whitespace, or
after an operator.
The comment causes that word and all remaining characters on that line to
be ignored.
An interactive shell without the @code{interactive_comments}
option enabled does not allow comments.
The @code{interactive_comments}
option is enabled by default in interactive shells.
@xref{Interactive Shells}, for a description of what makes
a shell interactive.

@node Shell Commands
@section Shell Commands
@cindex commands, shell

A simple shell command such as @code{echo a b c} consists of the command
itself followed by arguments, separated by spaces.

More complex shell commands are composed of simple commands arranged together
in a variety of ways: in a pipeline in which the output of one command
becomes the input of a second, in a loop or conditional construct, or in
some other grouping.

@menu
* Reserved Words::		Words that have special meaning to the shell.
* Simple Commands::		The most common type of command.
* Pipelines::			Connecting the input and output of several
				commands.
* Lists::			How to execute commands sequentially.
* Compound Commands::		Shell commands for control flow.
* Coprocesses::			Two-way communication between commands.
* GNU Parallel::		Running commands in parallel.
@end menu

@node Reserved Words
@subsection Reserved Words
@cindex reserved words

Reserved words are words that have special meaning to the shell.
They are used to begin and end the shell's compound commands.

Reserved words are recognized as reserved when unquoted and either

@itemize @bullet
@item
the first word of a command;
@item
the first word following a reserved word
other than @code{case}, @code{for}, @code{select}, or @code{in};
@item
the third word of a @code{case} command
(only @code{in} is valid);
@item
the third word of a @code{for} or @code{select}
command (only @code{in} and @code{do} are valid);
@item
following a control operator.
@end itemize

@noindent
The shell will also recognize reserved words where the syntax of a command
specifically requires the reserved word as the only correct token.

The following are reserved words:

@multitable @columnfractions .1 .1 .1 .1 .12 .1
@item @code{if} @tab @code{then} @tab @code{elif}
@tab @code{else} @tab @code{fi} @tab @code{time}
@item @code{for} @tab @code{in} @tab @code{until}
@tab @code{while} @tab @code{do} @tab @code{done}
@item @code{case} @tab @code{esac} @tab @code{coproc}
@tab @code{select} @tab @code{function}
@item @code{@{} @tab @code{@}} @tab @code{[[} @tab @code{]]} @tab @code{!}
@end multitable

@node Simple Commands
@subsection Simple Commands
@cindex commands, simple

A simple command is the kind of command that's executed most often.
It's just a sequence of words separated by @code{blank}s, terminated
by one of the shell's control operators (@pxref{Definitions}).
The first word generally specifies a command to be executed, with the
rest of the words being that command's arguments.

The return status (@pxref{Exit Status}) of a simple command is
its exit status as provided
by the @sc{posix} 1003.1 @code{waitpid} function, or 128+@var{n} if
the command was terminated by signal @var{n}.

@node Pipelines
@subsection Pipelines
@cindex pipeline
@cindex commands, pipelines

A @code{pipeline} is a sequence of one or more commands separated by
one of the control operators @samp{|} or @samp{|&}.

@rwindex time
@rwindex !
@cindex command timing
The format for a pipeline is
@example
[time [-p]] [!] @var{command1} [ | or |& @var{command2} ] @dots{}
@end example

@noindent
The output of each command in the pipeline is connected via a pipe
to the input of the next command.
That is, each command reads the previous command's output.
This connection is performed before any redirections specified by
@var{command1}.

If @samp{|&} is the pipeline operator,
@var{command1}'s standard error, in addition to
its standard output, is connected to
@var{command2}'s standard input through the pipe;
it is shorthand for @code{2>&1 |}.
This implicit redirection of the standard error to the standard output is
performed after any redirections specified by @var{command1},
consistent with that shorthand.

If the reserved word @code{time} precedes the pipeline,
Bash prints timing statistics for the pipeline once it finishes.
The statistics currently consist of elapsed (wall-clock) time and
user and system time consumed by the command's execution.
The @option{-p} option changes the output format to that specified
by @sc{posix}.
When the shell is in @sc{posix} mode (@pxref{Bash POSIX Mode}),
it does not recognize @code{time} as a reserved word if the next
token begins with a @samp{-}.
The value of the @env{TIMEFORMAT} variable is a format string that
specifies how the timing information should be displayed.
@xref{Bash Variables}, for a description of the available formats.
Providing @code{time} as a reserved word permits the timing of
shell builtins, shell functions, and pipelines.
An external @code{time} command cannot time these easily.

When the shell is in @sc{posix} mode (@pxref{Bash POSIX Mode}),
you can use @code{time} by itself as a simple command.
In this case, the shell displays the
total user and system time consumed by the shell and its children.
The @env{TIMEFORMAT} variable specifies the format of the time information.

If a pipeline is not executed asynchronously (@pxref{Lists}), the
shell waits for all commands in the pipeline to complete.

Each command in a multi-command pipeline,
where pipes are created,
is executed in its own @dfn{subshell}, which is a
separate process (@pxref{Command Execution Environment}).
If the @code{lastpipe} option is enabled using the @code{shopt} builtin
(@pxref{The Shopt Builtin}),
and  job control is not active,
the last element of a pipeline may be run by the shell process.

The exit
status of a pipeline is the exit status of the last command in the
pipeline, unless the @code{pipefail} option is enabled
(@pxref{The Set Builtin}).
If @code{pipefail} is enabled, the pipeline's return status is the
value of the last (rightmost) command to exit with a non-zero status,
or zero if all commands exit successfully.
If the reserved word @samp{!} precedes the pipeline, the
exit status is the logical negation of the exit status as described
above.
If a pipeline is not executed asynchronously (@pxref{Lists}), the
shell waits for all commands in the pipeline to terminate before
returning a value.
The return status of an asynchronous pipeline is 0.

@node Lists
@subsection Lists of Commands
@cindex commands, lists

A @code{list} is a sequence of one or more @sc{and} or @sc{or} lists
separated by one of the operators
@samp{;} or @samp{&}, or a @code{newline},
and optionally terminated by one of those three characters.

@sc{and} and @sc{or} lists are sequences of one or more pipelines
separated by the control operators @samp{&&} and @samp{||},
respectively.
@sc{and} and @sc{or} lists are executed with left associativity.

An @sc{and} list has the form

@example
@var{command1} && @var{command2}
@end example

@noindent
@var{command2} is executed if, and only if, @var{command1}
returns an exit status of zero (success).

An @sc{or} list has the form

@example
@var{command1} || @var{command2}
@end example

@noindent
@var{command2} is executed if, and only if, @var{command1}
returns a non-zero exit status.

The return status of
@sc{and} and @sc{or} lists is the exit status of the last command
executed in the list.

Of these list operators, @samp{&&} and @samp{||}
have equal precedence, followed by @samp{;} and @samp{&},
which have equal precedence.

A sequence of one or more newlines may appear in a @code{list}
to delimit commands, equivalent to a semicolon.

If a command is terminated by the control operator @samp{&},
the shell executes the command asynchronously in a subshell.
This is known as executing the command in the @dfn{background},
and these are referred to as @dfn{asynchronous} commands.
The shell does not wait for the command to finish, and the return
status is 0 (true).
When job control is not active (@pxref{Job Control}),
the standard input for asynchronous commands,
in the absence of any explicit redirections involving the standard input,
is redirected from @file{/dev/null}.

Commands separated or terminated by 
@samp{;} (or equivalent @code{newline})
are executed sequentially; the shell waits for each
command to terminate in turn.

The return status of a list is the exit status of the last command executed.

@node Compound Commands
@subsection Compound Commands
@cindex commands, compound

@menu
* Looping Constructs::		Shell commands for iterative action.
* Conditional Constructs::	Shell commands for conditional execution.
* Command Grouping::		Ways to group commands.
@end menu

Compound commands are the shell programming language constructs.
Each construct begins with a reserved word or control operator and is
terminated by a corresponding reserved word or operator.
Any redirections (@pxref{Redirections}) associated with a compound command
apply to all commands within that compound command unless explicitly overridden.

In most cases a list of commands in a compound command's description may be
separated from the rest of the command by one or more newlines, and may be
followed by a newline in place of a semicolon.

Bash provides looping constructs, conditional commands, and mechanisms
to group commands and execute them as a unit.

@node Looping Constructs
@subsubsection Looping Constructs
@cindex commands, looping

Bash supports the following looping constructs.

Note that wherever a @samp{;} appears in the description of a
command's syntax, it may be replaced with one or more newlines.

@table @code
@item until
@rwindex until
@rwindex do
@rwindex done
The syntax of the @code{until} command is:

@example
until @var{test-commands}; do @var{consequent-commands}; done
@end example

Execute @var{consequent-commands} as long as
@var{test-commands} has an exit status which is not zero.
The return status is the exit status of the last command executed
in @var{consequent-commands}, or zero if none was executed.

@item while
@rwindex while
The syntax of the @code{while} command is:

@example
while @var{test-commands}; do @var{consequent-commands}; done
@end example

Execute @var{consequent-commands} as long as
@var{test-commands} has an exit status of zero.
The return status is the exit status of the last command executed
in @var{consequent-commands}, or zero if none was executed.

@item for
@rwindex for
The syntax of the @code{for} command is:

@example
for @var{name} [ [in @var{words} @dots{}] ; ] do @var{commands}; done
@end example

Expand @var{words} (@pxref{Shell Expansions}), and then
execute @var{commands} once for each word
in the resultant list, with @var{name} bound to the current word.
If @samp{in @var{words}} is not present, the @code{for} command
executes the @var{commands} once for each positional parameter that is
set, as if @samp{in "$@@"} had been specified
(@pxref{Special Parameters}).

The return status is the exit status of the last command that executes.
If there are no items in the expansion of @var{words}, no commands are
executed, and the return status is zero.

There is an alternate form of the @code{for} command which is similar to the
C language:

@example
for (( @var{expr1} ; @var{expr2} ; @var{expr3} )) [;] do @var{commands} ; done
@end example

First, evaluate the arithmetic expression @var{expr1} according
to the rules described below (@pxref{Shell Arithmetic}).
Then, repeatedly evaluate the arithmetic expression @var{expr2}
until it evaluates to zero.   
Each time @var{expr2} evaluates to a non-zero value, execute @var{commands}
and evaluate the arithmetic expression @var{expr3}.
If any expression is omitted, it behaves as if it evaluates to 1.
The return value is the exit status of the last command in @var{commands}
that is executed, or non-zero if any of the expressions is invalid.
@end table

Use the @code{break} and @code{continue} builtins
(@pxref{Bourne Shell Builtins})
to control loop execution.

@node Conditional Constructs
@subsubsection Conditional Constructs
@cindex commands, conditional

@table @code
@item if
@rwindex if
@rwindex then
@rwindex else
@rwindex elif
@rwindex fi
The syntax of the @code{if} command is:

@example
if @var{test-commands}; then
  @var{consequent-commands};
[elif @var{more-test-commands}; then
  @var{more-consequents};]
[else @var{alternate-consequents};]
fi
@end example

The @var{test-commands} list is executed, and if its return status is zero,
the @var{consequent-commands} list is executed.
If @var{test-commands} returns a non-zero status, each @code{elif} list
is executed in turn, and if its exit status is zero,
the corresponding @var{more-consequents} is executed and the   
command completes.
If @samp{else @var{alternate-consequents}} is present, and
the final command in the final @code{if} or @code{elif} clause
has a non-zero exit status, then @var{alternate-consequents} is executed.
The return status is the exit status of the last command executed, or
zero if no condition tested true.

@item case
@rwindex case
@rwindex in
@rwindex esac
The syntax of the @code{case} command is:

@example
case @var{word} in
    [ [(] @var{pattern} [| @var{pattern}]@dots{}) @var{command-list} ;;]@dots{}
esac
@end example

@code{case} will selectively execute the @var{command-list} corresponding to
the first @var{pattern} that matches @var{word},
proceeding from the first pattern to the last.
The match is performed according
to the rules described below in @ref{Pattern Matching}.
The @var{word} undergoes tilde expansion, parameter expansion, command
substitution, process substitution, arithmetic expansion, and quote removal
(@pxref{Shell Parameter Expansion})
before the shell attempts to match the pattern.
Each @var{pattern} examined undergoes tilde expansion, parameter expansion,
command substitution, arithmetic expansion, process substitution, and
quote removal.
If the @code{nocasematch} shell option
(see the description of @code{shopt} in @ref{The Shopt Builtin})
is enabled, the match is performed without regard to the case
of alphabetic characters.

A pattern list is a set of one or more patterns separated by @samp{|},
and terminated by the @samp{)} operator.
A case @var{clause} is a pattern list and an associated @var{command-list},
terminated by @samp{;;}, @samp{;&}, or @samp{;;&}.
The terminator is optional for the last clause preceding @code{esac}.
There may be an arbitrary number of @code{case} clauses.
The first pattern that matches determines the
command-list that is executed.
It's a common idiom to use @samp{*} as the final pattern to define the
default case, since that pattern will always match.

Here is an example using @code{case} in a script that could be used to
describe one interesting feature of an animal:

@example
echo -n "Enter the name of an animal: "
read ANIMAL
echo -n "The $ANIMAL has "
case $ANIMAL in
  horse | dog | cat) echo -n "four";;
  man | kangaroo ) echo -n "two";;
  *) echo -n "an unknown number of";;
esac
echo " legs."
@end example

@noindent

When a match is found, @code{case} executes
the corresponding @var{command-list}.
If the @samp{;;} operator terminates the case clause,
the @code{case} command completes after the first pattern match.
Using the @samp{;&} terminator continues execution with
the @var{command-list} associated with the next clause, if any.
Using the @samp{;;&} terminator causes the shell to test the pattern list
in the next clause, if any, and execute any associated @var{command-list}
if the match succeeds,
continuing the case statement execution as if the pattern list had not matched.

The return status is zero if no @var{pattern} matches. 
Otherwise, the return status is the exit status of the
last @var{command-list} executed.

@item select
@rwindex select

The @code{select} construct allows the easy generation of menus.
It has almost the same syntax as the @code{for} command:

@example
select @var{name} [in @var{words} @dots{}]; do @var{commands}; done
@end example

First, expand the list of words following @code{in}, generating a list
of items, and print the set of expanded words on the standard
error stream, each preceded by a number.
If the @samp{in @var{words}} is omitted, print the positional parameters,
as if @samp{in "$@@"} had been specified.
@code{select} then displays the @env{PS3}
prompt and reads a line from the standard input.
If the line consists of a number corresponding to one of the displayed
words, then @code{select} sets the value of @var{name} to that word.
If the line is empty, @code{select} displays the words and prompt again.
If @code{EOF} is read, @code{select} completes and returns 1.
Any other value read causes @var{name} to be set to null.
The line read is saved in the variable @env{REPLY}.

The @var{commands} are executed after each selection until a
@code{break} command is executed, at which
point the @code{select} command completes.

Here is an example that allows the user to pick a filename from the
current directory, and displays the name and index of the file
selected.

@example
select fname in *;
do
	echo you picked $fname \($REPLY\)
	break;
done
@end example

@item ((@dots{}))
@example
(( @var{expression} ))
@end example

The arithmetic @var{expression} is evaluated according to the rules
described below (@pxref{Shell Arithmetic}).
The @var{expression} undergoes the same expansions
as if it were within double quotes,
but unescaped double quote characters
in @var{expression} are not treated
specially and are removed.
Since this can potentially result in empty strings,
this command treats those as expressions that evaluate to 0.
If the value of the expression is non-zero, the return status is 0;
otherwise the return status is 1. 

@item [[@dots{}]]
@rwindex [[
@rwindex ]]
@example
[[ @var{expression} ]]
@end example

Evaluate the conditional expression @var{expression} and
return a status of zero (true) or non-zero (false).
Expressions are composed of the primaries described below in
@ref{Bash Conditional Expressions}.
The words between the @code{[[} and @code{]]} do not undergo word splitting
and filename expansion.
The shell performs tilde expansion, parameter and
variable expansion, arithmetic expansion, command substitution, process
substitution, and quote removal on those words.
Conditional operators such as @samp{-f} must be unquoted to be recognized
as primaries.

When used with @code{[[}, the @samp{<} and @samp{>} operators sort
lexicographically using the current locale.

When the @samp{==} and @samp{!=} operators are used, the string to the
right of the operator is considered a pattern and matched according
to the rules described below in @ref{Pattern Matching},
as if the @code{extglob} shell option were enabled.
The @samp{=} operator is identical to @samp{==}.
If the @code{nocasematch} shell option
(see the description of @code{shopt} in @ref{The Shopt Builtin})
is enabled, the match is performed without regard to the case
of alphabetic characters.
The return value is 0 if the string matches (@samp{==}) or does not
match (@samp{!=}) the pattern, and 1 otherwise.

If you quote any part of the pattern,
using any of the shell's quoting mechanisms,
the quoted portion is matched literally.
This means every character in the quoted portion matches itself,
instead of having any special pattern matching meaning.

An additional binary operator, @samp{=~}, is available, with the same
precedence as @samp{==} and @samp{!=}.
When you use @samp{=~}, the string to the right of the operator is considered
a @sc{posix} extended regular expression pattern and matched accordingly
(using the @sc{posix} @code{regcomp} and @code{regexec} interfaces
usually described in @i{regex}(3)).
The return value is 0 if the string matches the pattern, and 1 if it does not.
If the regular expression is syntactically incorrect, the conditional
expression returns 2.
If the @code{nocasematch} shell option
(see the description of @code{shopt} in @ref{The Shopt Builtin})
is enabled, the match is performed without regard to the case
of alphabetic characters.

You can quote any part of the pattern
to force the quoted portion to be matched literally
instead of as a regular expression (see above).
If the pattern is stored in a shell variable, quoting the variable
expansion forces the entire pattern to be matched literally.

The match succeeds if the pattern matches any part of the string.
If you want to force the pattern to match the entire string,
anchor the pattern using the @samp{^} and @samp{$} regular expression
operators.

For example, the following will match a line
(stored in the shell variable @code{line})
if there is a sequence of characters anywhere in the value consisting of
any number, including zero, of 
characters in the @code{space} character class,
immediately followed by zero or one instances of @samp{a},
then a @samp{b}:

@example
[[ $line =~ [[:space:]]*(a)?b ]]
@end example

@noindent
That means values for @code{line} like
@samp{aab}, @samp{  aaaaaab}, @samp{xaby}, and @samp{ ab}
will all match,
as will a line containing a @samp{b} anywhere in its value.

If you want to match a character that's special to the regular expression
grammar (@samp{^$|[]()\.*+?}), it has to be quoted to remove its special
meaning.
This means that in the pattern @samp{xxx.txt}, the @samp{.} matches any
character in the string (its usual regular expression meaning), but in the
pattern @samp{"xxx.txt"}, it can only match a literal @samp{.}.

Likewise, if you want to include a character in your pattern that has a
special meaning to the regular expression grammar, you must make sure it's
not quoted.
If you want to anchor a pattern at the beginning or end of the string,
for instance, you cannot quote the @samp{^} or @samp{$}
characters using any form of shell quoting.

If you want to match @samp{initial string} at the start of a line,
the following will work:
@example
[[ $line =~ ^"initial string" ]]
@end example
@noindent
but this will not:
@example
[[ $line =~ "^initial string" ]]
@end example
@noindent
because in the second example the @samp{^} is quoted and doesn't have its
usual special meaning.

It is sometimes difficult to specify a regular expression properly
without using quotes, or to keep track of the quoting used by regular
expressions while paying attention to
shell quoting and the shell's quote removal.
Storing the regular expression in a shell variable is often a useful
way to avoid problems with quoting characters that are special to the
shell.
For example, the following is equivalent to the pattern used above:

@example
pattern='[[:space:]]*(a)?b'
[[ $line =~ $pattern ]]
@end example

Shell programmers should take special care with backslashes, since
backslashes are used by both the shell and regular expressions to remove
the special meaning from the following character.
This means that after the shell's word expansions complete
(@pxref{Shell Expansions}),
any backslashes remaining in parts of the pattern
that were originally not quoted can remove the
special meaning of pattern characters.
If any part of the pattern is quoted, the shell does its best to ensure that
the regular expression treats those remaining backslashes as literal,
if they appeared in a quoted portion.

The following two sets of commands are @emph{not} equivalent:

@example
pattern='\.'

[[ . =~ $pattern ]]
[[ . =~ \. ]]

[[ . =~ "$pattern" ]]
[[ . =~ '\.' ]]
@end example

@noindent
The first two matches will succeed, but the second two will not, because
in the second two the backslash will be part of the pattern to be matched.
In the first two examples, the pattern passed to the regular expression
parser is @samp{\.}.
The backslash removes the special meaning from
@samp{.}, so the literal @samp{.} matches.
In the second two examples, the pattern passed to the regular expression
parser has the backslash quoted (e.g., @samp{\\\.}), which will not match
the string, since it does not contain a backslash.
If the string in the first examples were anything other than @samp{.}, say
@samp{a}, the pattern would not match, because the quoted @samp{.} in the
pattern loses its special meaning of matching any single character.

Bracket expressions in regular expressions can be sources of errors as well,
since characters that are normally special in regular expressions
lose their special meanings between brackets.
However, you can use bracket expressions to match special pattern characters
without quoting them, so they are sometimes useful for this purpose.

Though it might seem like a strange way to write it, the following pattern
will match a @samp{.} in the string:

@example
[[ . =~ [.] ]]
@end example

The shell performs any word expansions before passing the pattern
to the regular expression functions,
so you can assume that the shell's quoting takes precedence.
As noted above, the regular expression parser will interpret any
unquoted backslashes remaining in the pattern after shell expansion
according to its own rules.
The intention is to avoid making shell programmers quote things twice
as much as possible, so shell quoting should be sufficient to quote
special pattern characters where that's necessary.

The array variable @code{BASH_REMATCH} records which parts of the string
matched the pattern.
Bash unsets @code{BASH_REMATCH}
before attempting the match, so if there is no match, it remains unset.
The element of @code{BASH_REMATCH} with index 0 contains the portion of
the string matching the entire regular expression.
Substrings matched by parenthesized subexpressions within the regular
expression are saved in the remaining @code{BASH_REMATCH} indices.
The element of @code{BASH_REMATCH} with index @var{n} is the portion of the
string matching the @var{n}th parenthesized subexpression.

Bash sets
@code{BASH_REMATCH}
in the global scope if it is not set;
if it is declared as a local variable before running
@code{[[},
Bash keeps it a local variable.

Expressions may be combined using the following operators, listed
in decreasing order of precedence:

@table @code
@item ( @var{expression} )
Returns the value of @var{expression}.
This may be used to override the normal precedence of operators.

@item ! @var{expression}
True if @var{expression} is false.

@item @var{expression1} && @var{expression2}
True if both @var{expression1} and @var{expression2} are true.

@item @var{expression1} || @var{expression2}
True if either @var{expression1} or @var{expression2} is true.
@end table

@noindent
The @code{&&} and @code{||} operators do not evaluate @var{expression2} if the
value of @var{expression1} is sufficient to determine the return
value of the entire conditional expression.
@end table

@node Command Grouping
@subsubsection Grouping Commands
@cindex commands, grouping

Bash provides two ways to group a list of commands to be executed
as a unit.  When commands are grouped, redirections may be applied
to the entire command list.  For example, the output of all the
commands in the list may be redirected to a single stream.

@table @code
@item ()
@example
( @var{list} )
@end example

Placing a list of commands between parentheses forces the shell to create
a subshell (@pxref{Command Execution Environment}), and each
of the commands in @var{list} is executed in that subshell environment.
Since the @var{list} is executed in a subshell, variable assignments do not
remain in effect after the subshell completes. 

@item @{@}
@rwindex @{
@rwindex @}
@example
@{ @var{list}; @}
@end example

Placing a list of commands between curly braces causes the list to
be executed in the current shell environment.
No subshell is created.
The semicolon (or newline) following @var{list} is required.
@end table

In addition to the creation of a subshell, there is a subtle difference
between these two constructs due to historical reasons.  The braces
are reserved words, so they must be separated from the @var{list}
by @code{blank}s or other shell metacharacters.
The parentheses are operators, and are
recognized as separate tokens by the shell even if they are not separated
from the @var{list} by whitespace.

The exit status of both of these constructs is the exit status of
@var{list}.

@node Coprocesses
@subsection Coprocesses
@cindex coprocess

A @code{coprocess} is a shell command preceded by the @code{coproc}
reserved word.
A coprocess is executed asynchronously in a subshell, as if the command
had been terminated with the @samp{&} control operator, with a two-way pipe
established between the executing shell and the coprocess.

The syntax for a coprocess is:

@example
coproc [@var{NAME}] @var{command} [@var{redirections}]
@end example

@noindent
This creates a coprocess named @var{NAME}.
@var{command} may be either a simple command (@pxref{Simple Commands})
or a compound command (@pxref{Compound Commands}).
@var{NAME} is a shell variable name.
If @var{NAME} is not supplied, the default name is @code{COPROC}.

The recommended form to use for a coprocess is

@example
coproc @var{NAME} @{ @var{command}; @}
@end example

@noindent
This form is preferred because simple commands result in the coprocess
always being named @code{COPROC}, and it is simpler to use and more complete
than the other compound commands.

There are other forms of coprocesses:

@example
coproc @var{NAME} @var{compound-command}
coproc @var{compound-command}
coproc @var{simple-command}
@end example

@noindent
If @var{command} is a compound command, @var{NAME} is optional. The
word following @code{coproc} determines whether that word is interpreted
as a variable name: it is interpreted as @var{NAME} if it is not a
reserved word that introduces a compound command.
If @var{command} is a simple command, @var{NAME} is not allowed; this
is to avoid confusion between @var{NAME} and the first word of the simple
command.

When the coprocess is executed, the shell creates an array variable
(@pxref{Arrays})
named @var{NAME} in the context of the executing shell.
The standard output of @var{command}
is connected via a pipe to a file descriptor in the executing shell,
and that file descriptor is assigned to @var{NAME}[0].
The standard input of @var{command}
is connected via a pipe to a file descriptor in the executing shell,
and that file descriptor is assigned to @var{NAME}[1].
This pipe is established before any redirections specified by the
command (@pxref{Redirections}).
The file descriptors can be utilized as arguments to shell commands
and redirections using standard word expansions.
Other than those created to execute command and process substitutions,
the file descriptors are not available in subshells.

The process ID of the shell spawned to execute the coprocess is
available as the value of the variable @env{@var{NAME}_PID}.
The @code{wait}
builtin may be used to wait for the coprocess to terminate.

Since the coprocess is created as an asynchronous command,
the @code{coproc} command always returns success.
The return status of a coprocess is the exit status of @var{command}.

@node GNU Parallel
@subsection GNU Parallel

There are ways to run commands in parallel that are not built into Bash.
GNU Parallel is a tool to do just that.

GNU Parallel, as its name suggests, can be used to build and run commands
in parallel.  You may run the same command with different arguments, whether
they are filenames, usernames, hostnames, or lines read from files.  GNU
Parallel provides shorthand references to many of the most common operations
(input lines, various portions of the input line, different ways to specify
the input source, and so on).  Parallel can replace @code{xargs} or feed
commands from its input sources to several different instances of Bash.

For a complete description, refer to the GNU Parallel documentation, which
is available at
@url{https://www.gnu.org/software/parallel/parallel_tutorial.html}.

@node Shell Functions
@section Shell Functions
@cindex shell function
@cindex functions, shell

Shell functions are a way to group commands for later execution
using a single name for the group.
They are executed just like a "regular" simple command.
When the name of a shell function is used as a simple command name,
the shell executes
the list of commands associated with that function name.
Shell functions are executed in the current
shell context; there is no new process created to interpret them.

Functions are declared using this syntax:
@rwindex function
@example
@var{fname} () @var{compound-command} [ @var{redirections} ]
@end example

or

@example
function @var{fname} [()] @var{compound-command} [ @var{redirections} ]
@end example

This defines a shell function named @var{fname}.
The reserved word @code{function} is optional.
If the @code{function} reserved word is supplied, the parentheses are optional.
The @dfn{body} of the function is the compound command
@var{compound-command} (@pxref{Compound Commands}).
That command is usually a @var{list} enclosed between @{ and @}, but
may be any compound command listed above.
If the @code{function} reserved word is used, but the
parentheses are not supplied, the braces are recommended.
When the shell is in @sc{posix} mode (@pxref{Bash POSIX Mode}),
@var{fname} must be a valid shell name and
may not be the same as one of the special builtins
(@pxref{Special Builtins}).
When not in @sc{posix}  mode,
a function name can be any unquoted shell word that does
not contain @samp{$}.

Any redirections (@pxref{Redirections}) associated with the shell function
are performed when the function is executed.
Function definitions are deleted using the @option{-f} option to the
@code{unset} builtin (@pxref{Bourne Shell Builtins}).

The exit status of a function definition is zero unless a syntax error
occurs or a readonly function with the same name already exists.
When executed, the exit status of a function is the exit status of the
last command executed in the body.

Note that for historical reasons, in the most common usage the curly braces
that surround the body of the function must be separated from the body by
@code{blank}s or newlines.
This is because the braces are reserved words and are only recognized
as such when they are separated from the command list
by whitespace or another shell metacharacter.
When using the braces, the @var{list} must be terminated by a semicolon,
a @samp{&}, or a newline.

@var{compound-command} is executed whenever @var{fname} is specified
as the name of a simple command.
Functions are executed in the context of the
calling shell; there is no new process created to interpret
them (contrast this with the execution of a shell script).

When a function is executed, the arguments to the
function become the positional parameters
during its execution (@pxref{Positional Parameters}).
The special parameter @samp{#} that expands to the number of
positional parameters
is updated to reflect the new set of positional parameters.
Special parameter @code{0} is unchanged.
The first element of the @env{FUNCNAME} variable is set to the
name of the function while the function is executing.

All other aspects of the shell execution
environment are identical between a function and its caller
with these exceptions:
the @env{DEBUG} and @env{RETURN} traps
are not inherited unless the function has been given the
@code{trace} attribute using the @code{declare} builtin or
the @code{-o functrace} option has been enabled with
the @code{set} builtin,
(in which case all functions inherit the @env{DEBUG} and @env{RETURN} traps),
and the @env{ERR} trap is not inherited unless the @code{-o errtrace}
shell option has been enabled.
@xref{Bourne Shell Builtins}, for the description of the
@code{trap} builtin.

The @env{FUNCNEST} variable, if set to a numeric value greater
than 0, defines a maximum function nesting level.  Function
invocations that exceed the limit cause the entire command to
abort.

If the builtin command @code{return}
is executed in a function, the function completes and
execution resumes with the next command after the function
call.
Any command associated with the @code{RETURN} trap is executed
before execution resumes.
When a function completes, the values of the
positional parameters and the special parameter @samp{#}
are restored to the values they had prior to the function's
execution.
If @code{return} is supplied a numeric argument,
that is the function's return status; otherwise the function's
return status is the exit status of the last command executed
before the @code{return}.

Variables local to the function are declared with the
@code{local} builtin (@dfn{local variables}).
Ordinarily, variables and their values
are shared between a function and its caller.
These variables are visible only to
the function and the commands it invokes.
This is particularly
important when a shell function calls other functions.

In the following description, the @dfn{current scope} is a currently-
executing function.
Previous scopes consist of that function's caller and so on,
back to the "global" scope, where the shell is not executing
any shell function.
A local variable at the current local scope is a variable
declared using the @code{local} or @code{declare} builtins in the
function that is currently executing.

Local variables "shadow"
variables with the same name declared at previous scopes.
For instance, a local variable declared in a function
hides variables with the same name declared at previous scopes,
including global variables: references and assignments
refer to the local variable, leaving the variables
at previous scopes  unmodified.
When the function returns, the global variable is once again visible.

The shell uses @dfn{dynamic scoping} to control a variable's visibility
within functions.
With dynamic scoping, visible variables and their values
are a result of the sequence of function calls that caused execution
to reach the current function.
The value of a variable that a function sees depends
on its value within its caller, if any, whether that caller is
the global scope or another shell function.
This is also the value that a local variable
declaration shadows, and the value that is restored when the function
returns.

For example, if a variable @env{var} is declared as local in function
@code{func1}, and @code{func1} calls another function @code{func2},
references to @env{var} made from within @code{func2} resolve to the
local variable @env{var} from @code{func1}, shadowing any global variable
named @env{var}.

The following script demonstrates this behavior.
When executed, the script displays

@example
In func2, var = func1 local
@end example

@example
func1()
@{
    local var='func1 local'
    func2
@}

func2()
@{
    echo "In func2, var = $var"
@}

var=global
func1
@end example

The @code{unset} builtin also acts using the same dynamic scope: if a
variable is local to the current scope, @code{unset} unsets it;       
otherwise the unset will refer to the variable found in any calling scope
as described above.
If a variable at the current local scope is unset, it remains so
(appearing as unset)
until it is reset in that scope or until the function returns.
Once the function returns, any instance of the variable at a previous
scope becomes visible.
If the unset acts on a variable at a previous scope, any instance of a
variable with that name that had been shadowed becomes visible
(see below how the @code{localvar_unset} shell option changes this behavior).

The @option{-f} option to the @code{declare} (@code{typeset})
builtin command (@pxref{Bash Builtins})
lists function names and definitions.
The @option{-F} option to @code{declare} or @code{typeset}
lists the function names only
(and optionally the source file and line number, if the @code{extdebug}
shell option is enabled).
Functions may be exported so that child shell processes
(those created when executing a separate shell invocation)
automatically have them defined with the
@option{-f} option to the @code{export} builtin
(@pxref{Bourne Shell Builtins}).
The @option{-f} option to
the @code{unset} builtin
(@pxref{Bourne Shell Builtins})
deletes a function definition.

Functions may be recursive.
The @code{FUNCNEST} variable may be used to limit the depth of the
function call stack and restrict the number of function invocations.
By default, Bash places no limit on the number of recursive  calls.

@node Shell Parameters
@section Shell Parameters
@cindex parameters
@cindex variable, shell
@cindex shell variable

@menu
* Positional Parameters::	The shell's command-line arguments.
* Special Parameters::		Parameters denoted by special characters.
@end menu

A @dfn{parameter} is an entity that stores values.
It can be a @code{name}, a number, or one of the special characters
listed below.
A @dfn{variable} is a parameter denoted by a @code{name}.
A variable has a @code{value} and zero or more @code{attributes}.
Attributes are assigned using the @code{declare} builtin command
(see the description of the @code{declare} builtin in @ref{Bash Builtins}).
The @code{export} and @code{readonly} builtins assign specific attributes.

A parameter is set if it has been assigned a value.
The null string is a valid value.
Once a variable is set, it may be unset only by using
the @code{unset} builtin command.

A variable is assigned to using a statement of the form
@example
@var{name}=[@var{value}]
@end example
@noindent
If @var{value} is not given, the variable is assigned the null string.
All @var{value}s undergo tilde expansion, parameter and variable expansion,
command substitution, arithmetic expansion, and quote
removal (@pxref{Shell Parameter Expansion}).
If the variable has its @code{integer}
attribute set, then @var{value} 
is evaluated as an arithmetic expression even if the @code{$((@dots{}))}
expansion is not used (@pxref{Arithmetic Expansion}).
Word splitting and filename expansion are not performed.
Assignment statements may also appear as arguments to the
@code{alias}, 
@code{declare}, @code{typeset}, @code{export}, @code{readonly},
and @code{local} builtin commands (@dfn{declaration commands}).
When in @sc{posix} mode (@pxref{Bash POSIX Mode}), these builtins may appear
in a command after one or more instances of the @code{command} builtin
and retain these assignment statement properties.
For example,
@example
command export var=value
@end example

In the context where an assignment statement is assigning a value  
to a shell variable or array index (@pxref{Arrays}), the
@samp{+=} operator appends to or adds to
the variable's previous value.
This includes arguments to declaration commands such as
@code{declare}
that accept assignment statements.
When @samp{+=}
is applied to a variable for which the
@code{integer} attribute has been set,
the variable's current value and @var{value} are each evaluated as
arithmetic expressions,
and the sum of the results is assigned as the variable's value.
The current value is usually an integer constant, but may be an expression.
When @samp{+=}
is applied to an array variable using compound assignment (@pxref{Arrays}),
the variable's value is not unset
(as it is when using @samp{=}),
and new values are appended to the array
beginning at one greater than the array's maximum index (for indexed arrays),
or added as additional key-value pairs in an associative array.
When applied to a string-valued variable, @var{value} is expanded and
appended to the variable's value.

A variable can be assigned the @code{nameref} attribute using the
@option{-n} option to the @code{declare} or @code{local} builtin commands
(@pxref{Bash Builtins})
to create a @dfn{nameref}, or a reference to another variable.
This allows variables to be manipulated indirectly.
Whenever the nameref variable is referenced, assigned to, unset, or has
its attributes modified (other than using or changing the nameref
attribute itself), the
operation is actually performed on the variable specified by the nameref
variable's value.
A nameref is commonly used within shell functions to refer to a variable
whose name is passed as an argument to the function.
For instance, if a variable name is passed to a shell function as its first
argument, running
@example
declare -n ref=$1
@end example
@noindent
inside the function creates a local nameref variable @env{ref} whose value
is the variable name passed as the first argument.
References and assignments to @env{ref}, and changes to its attributes,
are treated as references, assignments, and attribute modifications
to the variable whose name was passed as @code{$1}.

If the control variable in a @code{for} loop has the nameref attribute,
the list of words can be a list of shell variables, and a name reference
is established for each word in the list, in turn, when the loop is
executed.
Array variables cannot be given the nameref attribute.
However, nameref variables can reference array variables and subscripted
array variables.
Namerefs can be unset using the @option{-n} option to the @code{unset} builtin
(@pxref{Bourne Shell Builtins}).
Otherwise, if @code{unset} is executed with the name of a nameref variable
as an argument, the variable referenced by the nameref variable is unset.

When the shell starts, it reads its environment and creates a shell
variable from each environment variable that has a valid name,
as described below (@pxref{Environment}).

@node Positional Parameters
@subsection Positional Parameters
@cindex parameters, positional

A @dfn{positional parameter} is a parameter denoted by one or more
digits, other than the single digit @code{0}.
Positional parameters are
assigned from the shell's arguments when it is invoked,
and may be reassigned using the @code{set} builtin command.
Positional parameter @code{N} may be referenced as @code{$@{N@}}, or
as @code{$N} when @code{N} consists of a single digit.
Positional parameters may not be assigned to with assignment statements.
The @code{set} and @code{shift} builtins are used to set and
unset them (@pxref{Shell Builtin Commands}).
The positional parameters are
temporarily replaced when a shell function is executed
(@pxref{Shell Functions}).

When a positional parameter consisting of more than a single
digit is expanded, it must be enclosed in braces.
Without braces, a digit following @samp{$} can only refer to
one of the first nine positional parameters ($1\-$9) or the
special parameter $0 (see below).

@node Special Parameters
@subsection Special Parameters
@cindex parameters, special

The shell treats several parameters specially.  These parameters may
only be referenced; assignment to them is not allowed.
Special parameters are denoted by one of the following characters.

@vtable @code

@item *
@vindex $*
($*) Expands to the positional parameters, starting from one.
When the expansion is not within double quotes, each positional parameter
expands to a separate word.
In contexts where word expansions are performed, those words
are subject to further word splitting and filename expansion.
When the expansion occurs within double quotes, it expands to a single word
with the value of each parameter separated by the first character of the
@env{IFS} variable.
That is, @code{"$*"} is equivalent
to @code{"$1@var{c}$2@var{c}@dots{}"}, where @var{c}
is the first character of the value of the @code{IFS}
variable.
If @env{IFS} is unset, the parameters are separated by spaces.
If @env{IFS} is null, the parameters are joined without intervening
separators.

@item @@
@vindex $@@
($@@) Expands to the positional parameters, starting from one.
In contexts where word splitting is performed, this expands each
positional parameter to a separate word; if not within double
quotes, these words are subject to word splitting.
In contexts where word splitting is not performed,
such as the value portion of an assignment statement,
this expands to a single word
with each positional parameter separated by a space.
When the expansion occurs within double quotes,
and word splitting is performed,
each parameter expands to a separate word.
That is, @code{"$@@"} is equivalent to
@code{"$1" "$2" @dots{}}.
If the double-quoted expansion occurs within a word, the expansion of
the first parameter is joined with the expansion of the
beginning part of the original
word, and the expansion of the last parameter is joined with the
expansion of the last part of the original word.
When there are no positional parameters, @code{"$@@"} and
@code{$@@}
expand to nothing (i.e., they are removed).

@item #
@vindex $#
($#) Expands to the number of positional parameters in decimal.

@item ?
@vindex $?
($?) Expands to the exit status of the most recently executed command.

@item -
@vindex $-
($-, a hyphen.)  Expands to the current option flags as specified upon
invocation, by the @code{set}
builtin command, or those set by the shell itself
(such as the @option{-i} option).

@item $
@vindex $$
($$) Expands to the process @sc{id} of the shell.
In a subshell, it expands to the process @sc{id} of the invoking shell,
not the subshell.

@item !
@vindex $!
($!) Expands to the process @sc{id} of the job most recently placed into the
background, whether executed as an asynchronous command or using
the @code{bg} builtin (@pxref{Job Control Builtins}).

@item 0
@vindex $0
($0) Expands to the name of the shell or shell script.
This is set at shell initialization.
If Bash is invoked with a file of commands (@pxref{Shell Scripts}),
@code{$0} is set to the name of that file.
If Bash is started with the @option{-c} option (@pxref{Invoking Bash}),
then @code{$0} is set to the first argument after the string to be
executed, if one is present.
Otherwise, it is set to the filename used to invoke Bash, as given by
argument zero.
@end vtable

@node Shell Expansions
@section Shell Expansions
@cindex expansion

Expansion is performed on the command line after it has been split into
@code{token}s.
Bash performs these expansions:

@itemize @bullet
@item brace expansion
@item tilde expansion
@item parameter and variable expansion
@item command substitution
@item arithmetic expansion
@item word splitting
@item filename expansion
@item quote removal
@end itemize

@menu
* Brace Expansion::		Expansion of expressions within braces.
* Tilde Expansion::		Expansion of the ~ character.
* Shell Parameter Expansion::	How Bash expands variables to their values.
* Command Substitution::	Using the output of a command as an argument.
* Arithmetic Expansion::	How to use arithmetic in shell expansions.
* Process Substitution::	A way to write and read to and from a
				command.
* Word Splitting::	How the results of expansion are split into separate
			arguments.
* Filename Expansion::	A shorthand for specifying filenames matching patterns.
* Quote Removal::	How and when quote characters are removed from
			words.
@end menu

The order of expansions is:
brace expansion;
tilde expansion, parameter and variable expansion, arithmetic expansion,
and command substitution (done in a left-to-right fashion);
word splitting;
filename expansion;
and quote removal.

On systems that can support it, there is an additional expansion
available: @dfn{process substitution}.
This is performed at the
same time as tilde, parameter, variable, and arithmetic expansion and
command substitution.

@dfn{Quote removal} is always performed last.
It removes quote characters present in the original word,
not ones resulting from one of the other expansions,
unless they have been quoted themselves.
@xref{Quote Removal} for more details.

Only brace expansion, word splitting, and filename expansion
can increase the number of words of the expansion; other expansions
expand a single word to a single word.
The only exceptions to this are the expansions of
@code{"$@@"} and @code{$*} (@pxref{Special Parameters}), and
@code{"$@{@var{name}[@@]@}"} and @code{$@{@var{name}[*]@}}
(@pxref{Arrays}).

@node Brace Expansion
@subsection Brace Expansion
@cindex brace expansion
@cindex expansion, brace

Brace expansion is a mechanism to generate arbitrary strings
sharing a common prefix and suffix, either of which can be empty.
This mechanism is similar to
@dfn{filename expansion} (@pxref{Filename Expansion}),
but the filenames generated need not exist.
Patterns to be brace expanded are formed from an optional @var{preamble},
followed by either a series of comma-separated strings or
a sequence expression between a pair of braces,
followed by an optional
@var{postscript}.
The preamble is prefixed to each string contained within the braces,
and the postscript is then appended to each resulting string,
expanding left to right.

Brace expansions may be nested.
The results of each expanded string are not sorted;
brace expansion preserves left to right order.
For example,
@example
bash$ echo a@{d,c,b@}e
ade ace abe
@end example

A sequence expression takes the form
@code{@var{x}..@var{y}[..@var{incr}]},
where @var{x} and @var{y} are either integers or letters,
and @var{incr}, an optional increment, is an integer.
When integers are supplied, the expression expands to each number between
@var{x} and @var{y}, inclusive.
If either @var{x} or @var{y} begins with a zero,
each generated term will contain the same number of digits,
zero-padding where necessary.
When letters are supplied, the expression expands to each character
lexicographically between @var{x} and @var{y}, inclusive,
using the C locale.
Note that both @var{x} and @var{y} must be of the same type
(integer or letter).
When the increment is supplied, it is used as the difference between
each term.
The default increment is 1 or -1 as appropriate.

Brace expansion is performed before any other expansions,
and any characters special to other expansions are preserved
in the result.
It is strictly textual.
Bash does not apply any syntactic interpretation to the context
of the expansion or the text between the braces.

A correctly-formed brace expansion must contain unquoted opening
and closing braces, and at least one unquoted comma or a valid
sequence expression.
Any incorrectly formed brace expansion is left unchanged.

A @samp{@{} or @samp{,}
may be quoted with a backslash to prevent its
being considered part of a brace expression.
To avoid conflicts with parameter expansion, the string @samp{$@{}
is not considered eligible for brace expansion,
and inhibits brace expansion until the closing
@samp{@}}.

This construct is typically used as shorthand when the common
prefix of the strings to be generated is longer than in the
above example:
@example
mkdir /usr/local/src/bash/@{old,new,dist,bugs@}
@end example
or
@example
chown root /usr/@{ucb/@{ex,edit@},lib/@{ex?.?*,how_ex@}@}
@end example

Brace expansion introduces a slight incompatibility with
historical versions of
@command{sh}.
@command{sh}
does not treat opening or closing braces specially when they
appear as part of a word, and preserves them in the output.
Bash
removes braces from words as a consequence of brace
expansion.
For example, a word entered to
@command{sh}
as
@samp{file@{1,2@}}
appears identically in the output.
Bash
outputs that word as
@samp{file1 file2}
after brace expansion.
Start
Bash
with the
@option{+B}
option or disable brace expansion with the
@option{+B}
option to the
@code{set}
command
(@pxref{Shell Builtin Commands})
for strict
@command{sh}
compatibility.

@node Tilde Expansion
@subsection Tilde Expansion
@cindex tilde expansion
@cindex expansion, tilde

If a word begins with an unquoted tilde character (@samp{~}), all of the
characters up to the first unquoted slash (or all characters,
if there is no unquoted slash) are considered a @dfn{tilde-prefix}.
If none of the characters in the tilde-prefix are quoted, the
characters in the tilde-prefix following the tilde are treated as a
possible @dfn{login name}.
If this login name is the null string, the tilde is replaced with the
value of the @env{HOME} shell variable.
If @env{HOME} is unset, the tilde expands to
the home directory of the user executing the shell instead.
Otherwise, the tilde-prefix is replaced with the home directory
associated with the specified login name.

If the tilde-prefix is @samp{~+}, the value of
the shell variable @env{PWD} replaces the tilde-prefix.
If the tilde-prefix is @samp{~-}, the shell substitutes
the value of the shell variable
@env{OLDPWD}, if it is set.

If the characters following the tilde in the tilde-prefix consist of a
number @var{N}, optionally prefixed by a @samp{+} or a @samp{-},
the tilde-prefix is replaced with the
corresponding element from the directory stack, as it would be displayed
by the @code{dirs} builtin invoked with the characters following tilde
in the tilde-prefix as an argument (@pxref{The Directory Stack}).
If the tilde-prefix, sans the tilde, consists of a number without a
leading @samp{+} or @samp{-}, tilde expansion assumes @samp{+}.

The results of tilde expansion are treated as if they were quoted, so
the replacement is not subject to word splitting and filename expansion.

If the login name is invalid, or the tilde expansion fails, the
tilde-prefix is left unchanged.

Bash checks each variable assignment
for unquoted tilde-prefixes immediately
following a @samp{:} or the first @samp{=},
and performs tilde expansion in these cases.
Consequently, one may use filenames with tildes in assignments to
@env{PATH}, @env{MAILPATH}, and @env{CDPATH},
and the shell assigns the expanded value.

The following table shows how Bash treats unquoted tilde-prefixes:

@table @code
@item ~
The value of @code{$HOME}.
@item ~/foo
@file{$HOME/foo}

@item ~fred/foo
The directory or file @code{foo} in the home directory of the user
@code{fred}.

@item ~+/foo
@file{$PWD/foo}

@item ~-/foo
@file{$@{OLDPWD-'~-'@}/foo}

@item ~@var{N}
The string that would be displayed by @samp{dirs +@var{N}}.

@item ~+@var{N}
The string that would be displayed by @samp{dirs +@var{N}}.

@item ~-@var{N}
The string that would be displayed by @samp{dirs -@var{N}}.
@end table

Bash also performs tilde expansion on words satisfying the conditions of
variable assignments (@pxref{Shell Parameters})
when they appear as arguments to simple commands.
Bash does not do this, except for the declaration commands listed
above, when in @sc{posix} mode.

@node Shell Parameter Expansion
@subsection Shell Parameter Expansion
@cindex parameter expansion
@cindex expansion, parameter

The @samp{$} character introduces parameter expansion,
command substitution, or arithmetic expansion.
The parameter name
or symbol to be expanded may be enclosed in braces, which
are optional but serve to protect the variable to be expanded from
characters immediately following it which could be
interpreted as part of the name.
For example, if the first positional parameter has the value @samp{a},
then @code{$@{11@}} expands to the value of the eleventh positional
parameter, while @code{$11} expands to @samp{a1}.

When braces are used, the matching ending brace is the first @samp{@}}
not escaped by a backslash or within a quoted string, and not within an
embedded arithmetic expansion, command substitution, or parameter
expansion.

The basic form of parameter expansion is $@{@var{parameter}@},
which substitutes the value of @var{parameter}.
The @var{parameter} is a shell parameter as described above
(@pxref{Shell Parameters}) or an array reference (@pxref{Arrays}).
The braces are required when @var{parameter}
is a positional parameter with more than one digit,
or when @var{parameter} is followed by a character that is not to be
interpreted as part of its name.

If the first character of @var{parameter} is an exclamation point (!),
and @var{parameter} is not a nameref,
it introduces a level of indirection.
Bash uses the value formed by expanding the rest of
@var{parameter} as the new @var{parameter};
this new parameter is then
expanded and that value is used
in the rest of the expansion, rather
than the expansion of the original @var{parameter}.
This is known as @code{indirect expansion}.
The value is subject to tilde expansion,
parameter expansion, command substitution, and arithmetic expansion.
If @var{parameter} is a nameref, this expands to the name of the
variable referenced by @var{parameter} instead of performing the
complete indirect expansion, for compatibility.
The exceptions to this are the expansions of $@{!@var{prefix}*@}
and $@{!@var{name}[@@]@}
described below.
The exclamation point must immediately follow the left brace in order to
introduce indirection.

In each of the cases below, @var{word} is subject to tilde expansion,
parameter expansion, command substitution, and arithmetic expansion.

When performing the first four expansions documented below
(@samp{:-}, @samp{:=}, @samp{:?}, and @samp{:+}),
including the colon,
Bash tests for a parameter that is unset or null.
Omitting the colon results in a test only for a parameter that is unset.
Put another way, if the colon is included,
the operator tests for both @var{parameter}'s existence and that its value
is not null; if the colon is omitted, the operator tests only for existence.

@table @code

@item $@{@var{parameter}:@minus{}@var{word}@}
If @var{parameter}
is unset or null,
or unset if the colon is not present,
the expansion of
@var{word} is substituted.
Otherwise, the value of @var{parameter} is substituted.

@example
$ v=123
$ echo $@{v-unset@}
123
$ echo $@{v:-unset-or-null@}
123
$ unset v
$ echo $@{v-unset@}
unset
$ v=
$ echo $@{v-unset@}

$ echo $@{v:-unset-or-null@}
unset-or-null
@end example

@item $@{@var{parameter}:=@var{word}@}
If @var{parameter}
is unset or null,
or unset if the colon is not present,
the expansion of @var{word}
is assigned to @var{parameter},
and the result of the expansion
is the final value of @var{parameter}.
Positional parameters and special parameters may not be assigned
in this way.

@example
$ unset var
$ : $@{var=DEFAULT@}
$ echo $var
DEFAULT
$ var=
$ : $@{var=DEFAULT@}
$ echo $var

$ var=
$ : $@{var:=DEFAULT@}
$ echo $var
DEFAULT
$ unset var
$ : $@{var:=DEFAULT@}
$ echo $var
DEFAULT
@end example

@item $@{@var{parameter}:?@var{word}@}
If @var{parameter}
is unset or null,
or unset if the colon is not present,
the shell writes
the expansion of @var{word} (or a message
to that effect if @var{word}
is not present) to the standard error and, if it
is not interactive, exits with a non-zero status.
An interactive shell does not exit, but does not execute the command
associated with the expansion.
Otherwise, the value of @var{parameter} is substituted.

@example
$ var=
$ : $@{var:?var is unset or null@}
bash: var: var is unset or null
$ echo $@{var?var is unset@}

$ unset var
$ : $@{var?var is unset@}
bash: var: var is unset
$ : $@{var:?var is unset or null@}
bash: var: var is unset or null
$ var=123
$ echo $@{var:?var is unset or null@}
123
@end example

@item $@{@var{parameter}:+@var{word}@}
If @var{parameter}
is unset or null,
or unset if the colon is not present,
nothing is substituted,
otherwise the expansion of
@var{word} is substituted.
The value of @var{parameter} is not used.

@example
$ var=123
$ echo $@{var:+var is set and not null@}
var is set and not null
$ echo $@{var+var is set@}
var is set
$ var=
$ echo $@{var:+var is set and not null@}

$ echo $@{var+var is set@}
var is set
$ unset var
$ echo $@{var+var is set@}

$ echo $@{var:+var is set and not null@}

$ 
@end example

@item $@{@var{parameter}:@var{offset}@}
@itemx $@{@var{parameter}:@var{offset}:@var{length}@}
This is referred to as Substring Expansion.
It expands to up to @var{length} characters of the value of @var{parameter}
starting at the character specified by @var{offset}.
If @var{parameter} is @samp{@@} or @samp{*}, an indexed array subscripted by
@samp{@@} or @samp{*}, or an associative array name, the results differ as
described below.
If :@var{length} is omitted (the first form above), this
expands to the substring of the value of
@var{parameter} starting at the character specified by @var{offset}
and extending to the end of the value.
If @var{offset} is omitted,
it is treated as 0.
If @var{length} is omitted,
but the colon after @var{offset} is present,
it is treated as 0.
@var{length} and @var{offset} are arithmetic expressions
(@pxref{Shell Arithmetic}).

If @var{offset} evaluates to a number less than zero, the value
is used as an offset in characters
from the end of the value of @var{parameter}.
If @var{length} evaluates to a number less than zero,
it is interpreted as an offset in characters
from the end of the value of @var{parameter} rather than
a number of characters, and the expansion is the characters between
@var{offset} and that result.

Note that a negative offset must be separated from the colon by at least
one space to avoid being confused with the @samp{:@minus{}} expansion.

Here are some examples illustrating substring expansion on parameters and
subscripted arrays:

@verbatim
$ string=01234567890abcdefgh
$ echo ${string:7}
7890abcdefgh
$ echo ${string:7:0}

$ echo ${string:7:2}
78
$ echo ${string:7:-2}
7890abcdef
$ echo ${string: -7}
bcdefgh
$ echo ${string: -7:0}

$ echo ${string: -7:2}
bc
$ echo ${string: -7:-2}
bcdef
$ set -- 01234567890abcdefgh
$ echo ${1:7}
7890abcdefgh
$ echo ${1:7:0}

$ echo ${1:7:2}
78
$ echo ${1:7:-2}
7890abcdef
$ echo ${1: -7}
bcdefgh
$ echo ${1: -7:0}

$ echo ${1: -7:2}
bc
$ echo ${1: -7:-2}
bcdef
$ array[0]=01234567890abcdefgh
$ echo ${array[0]:7}
7890abcdefgh
$ echo ${array[0]:7:0}

$ echo ${array[0]:7:2}
78
$ echo ${array[0]:7:-2}
7890abcdef
$ echo ${array[0]: -7}
bcdefgh
$ echo ${array[0]: -7:0}

$ echo ${array[0]: -7:2}
bc
$ echo ${array[0]: -7:-2}
bcdef
@end verbatim

If @var{parameter} is @samp{@@} or @samp{*}, the result is @var{length}
positional parameters beginning at @var{offset}.
A negative @var{offset} is taken relative to one greater than the greatest
positional parameter, so an offset of -1 evaluates to the last positional
parameter (or 0 if there are no positional parameters).
It is an expansion error if @var{length} evaluates to a number less than zero.

The following examples illustrate substring expansion using positional
parameters:

@verbatim
$ set -- 1 2 3 4 5 6 7 8 9 0 a b c d e f g h
$ echo ${@:7}
7 8 9 0 a b c d e f g h
$ echo ${@:7:0}

$ echo ${@:7:2}
7 8
$ echo ${@:7:-2}
bash: -2: substring expression < 0
$ echo ${@: -7:2}
b c
$ echo ${@:0}
./bash 1 2 3 4 5 6 7 8 9 0 a b c d e f g h
$ echo ${@:0:2}
./bash 1
$ echo ${@: -7:0}

@end verbatim

If @var{parameter} is an indexed array name subscripted
by @samp{@@} or @samp{*}, the result is the @var{length}
members of the array beginning with @code{$@{@var{parameter}[@var{offset}]@}}.
A negative @var{offset} is taken relative to one greater than the maximum
index of the specified array.
It is an expansion error if @var{length} evaluates to a number less than zero.

These examples show how you can use substring expansion with indexed
arrays:

@verbatim
$ array=(0 1 2 3 4 5 6 7 8 9 0 a b c d e f g h)
$ echo ${array[@]:7}
7 8 9 0 a b c d e f g h
$ echo ${array[@]:7:2}
7 8
$ echo ${array[@]: -7:2}
b c
$ echo ${array[@]: -7:-2}
bash: -2: substring expression < 0
$ echo ${array[@]:0}
0 1 2 3 4 5 6 7 8 9 0 a b c d e f g h
$ echo ${array[@]:0:2}
0 1
$ echo ${array[@]: -7:0}

@end verbatim

Substring expansion applied to an associative array produces undefined
results.

Substring indexing is zero-based unless the positional parameters
are used, in which case the indexing starts at 1 by default.
If @var{offset} is 0, and the positional parameters are used, @code{$0} is
prefixed to the list.

@item $@{!@var{prefix}*@}
@itemx $@{!@var{prefix}@@@}
Expands to the names of variables whose names begin with @var{prefix},
separated by the first character of the @env{IFS} special variable.
When @samp{@@} is used and the expansion appears within double quotes, each
variable name expands to a separate word.

@item $@{!@var{name}[@@]@}
@itemx $@{!@var{name}[*]@}
If @var{name} is an array variable, expands to the list of array indices
(keys) assigned in @var{name}.
If @var{name} is not an array, expands to 0 if @var{name} is set and null
otherwise.
When @samp{@@} is used and the expansion appears within double quotes, each
key expands to a separate word.

@item $@{#@var{parameter}@}
Substitutes the length in characters of the value of @var{parameter}.
If @var{parameter} is @samp{*} or @samp{@@}, the value substituted
is the number of positional parameters.
If @var{parameter} is an array name subscripted by @samp{*} or @samp{@@}, 
the value substituted is the number of elements in the array.
If @var{parameter}
is an indexed array name subscripted by a negative number, that number is
interpreted as relative to one greater than the maximum index of
@var{parameter}, so negative indices count back from the end of the
array, and an index of -1 references the last element.

@item $@{@var{parameter}#@var{word}@}
@itemx $@{@var{parameter}##@var{word}@}
The @var{word}
is expanded to produce a pattern and matched
against the expanded value of @var{parameter}
according to the rules
described below (@pxref{Pattern Matching}).
If the pattern matches
the beginning of the expanded value of @var{parameter},
then the result of the expansion is the expanded value of @var{parameter}
with the shortest matching pattern (the @samp{#} case) or the
longest matching pattern (the @samp{##} case) deleted.
If @var{parameter} is @samp{@@} or @samp{*},
the pattern removal operation is applied to each positional
parameter in turn, and the expansion is the resultant list.
If @var{parameter} is an array variable subscripted with
@samp{@@} or @samp{*},
the pattern removal operation is applied to each member of the
array in turn, and the expansion is the resultant list.

@item $@{@var{parameter}%@var{word}@}
@itemx $@{@var{parameter}%%@var{word}@}
The @var{word}
is expanded to produce a pattern and matched
against the expanded value of @var{parameter}
according to the rules
described below (@pxref{Pattern Matching}).
If the pattern matches a trailing portion of the expanded value of
@var{parameter}, then the result of the expansion is the value of
@var{parameter} with the shortest matching pattern (the @samp{%} case)
or the longest matching pattern (the @samp{%%} case) deleted.
If @var{parameter} is @samp{@@} or @samp{*},
the pattern removal operation is applied to each positional
parameter in turn, and the expansion is the resultant list.
If @var{parameter}
is an array variable subscripted with @samp{@@} or @samp{*},
the pattern removal operation is applied to each member of the
array in turn, and the expansion is the resultant list.

@item $@{@var{parameter}/@var{pattern}/@var{string}@} 
@itemx $@{@var{parameter}//@var{pattern}/@var{string}@} 
@itemx $@{@var{parameter}/#@var{pattern}/@var{string}@} 
@itemx $@{@var{parameter}/%@var{pattern}/@var{string}@} 
The @var{pattern} is expanded to produce a pattern
and matched against the expanded value of @var{parameter}
as described below (@pxref{Pattern Matching}).
The longest match of @var{pattern}
in the expanded value is replaced with @var{string}.
@var{string} undergoes tilde expansion, parameter and variable expansion,
arithmetic expansion, command and process substitution, and quote removal.

In the first form above, only the first match is replaced.
If there are two slashes separating @var{parameter} and @var{pattern}
(the second form above), all matches of @var{pattern} are
replaced with @var{string}.
If @var{pattern} is preceded by @samp{#} (the third form above),
it must match at the beginning of the expanded value of @var{parameter}.
If @var{pattern} is preceded by @samp{%} (the fourth form above),
it must match at the end of the expanded value of @var{parameter}.

If the expansion of @var{string} is null,
matches of @var{pattern} are deleted
and the @samp{/} following @var{pattern} may be omitted.

If the @code{patsub_replacement} shell option is enabled using @code{shopt}
(@pxref{The Shopt Builtin}),
any unquoted instances of @samp{&} in @var{string} are replaced with the
matching portion of @var{pattern}.
This is intended to duplicate a common @code{sed} idiom.

Quoting any part of @var{string} inhibits replacement in the
expansion of the quoted portion, including replacement strings stored
in shell variables.
Backslash escapes @samp{&} in @var{string}; the backslash is removed
in order to permit a literal @samp{&} in the replacement string.
Users should take care if @var{string} is double-quoted to avoid
unwanted interactions between the backslash and double-quoting, since
backslash has special meaning within double quotes.
Pattern substitution performs the check for unquoted @samp{&} after
expanding @var{string},
so users should ensure to properly quote any occurrences of @samp{&}
they want to be taken literally in the replacement
and ensure any instances of @samp{&} they want to be replaced are unquoted.

For instance,

@example
var=abcdef
rep='& '
echo $@{var/abc/& @}
echo "$@{var/abc/& @}"
echo $@{var/abc/$rep@}
echo "$@{var/abc/$rep@}"
@end example

@noindent
will display four lines of "abc def", while

@example
var=abcdef
rep='& '
echo $@{var/abc/\& @}
echo "$@{var/abc/\& @}"
echo $@{var/abc/"& "@}
echo $@{var/abc/"$rep"@}
@end example

@noindent
will display four lines of "& def".
Like the pattern removal operators, double quotes surrounding the
replacement string quote the expanded characters, while double quotes
enclosing the entire parameter substitution do not, since
the expansion is performed in a
context that doesn't take any enclosing double quotes into account.

Since backslash can escape @samp{&}, it can also escape a backslash in
the replacement string.
This means that @samp{\\} will insert a literal
backslash into the replacement, so these two @code{echo} commands

@example
var=abcdef
rep='\\&xyz'
echo $@{var/abc/\\&xyz@}
echo $@{var/abc/$rep@}
@end example

@noindent
will both output @samp{\abcxyzdef}.

It should rarely be necessary to enclose only @var{string} in double
quotes.

If the @code{nocasematch} shell option   
(see the description of @code{shopt} in @ref{The Shopt Builtin})
is enabled, the match is performed without regard to the case   
of alphabetic characters.

If @var{parameter} is @samp{@@} or @samp{*},
the substitution operation is applied to each positional
parameter in turn, and the expansion is the resultant list.
If @var{parameter}
is an array variable subscripted with @samp{@@} or @samp{*},
the substitution operation is applied to each member of the
array in turn, and the expansion is the resultant list.

@item $@{@var{parameter}^@var{pattern}@}
@itemx $@{@var{parameter}^^@var{pattern}@}
@itemx $@{@var{parameter},@var{pattern}@}
@itemx $@{@var{parameter},,@var{pattern}@}
This expansion modifies the case of alphabetic characters in @var{parameter}.
First, the @var{pattern} is expanded to produce a pattern
as described below in @ref{Pattern Matching}.

@code{Bash}
then examines characters in the expanded value of @var{parameter}
against @var{pattern} as described below.
If a character matches the pattern, its case is converted.
The pattern should not attempt to match more than one character.

Using
@samp{^}
converts lowercase letters matching @var{pattern} to uppercase;
@samp{,}
converts matching uppercase letters to lowercase.
The
@samp{^} and @samp{,} variants
examine the first character in the expanded value
and convert its case if it matches @var{pattern};
the
@samp{^^} and @samp{,,} variants
examine all characters in the expanded value
and convert each one that matches @var{pattern}.
If @var{pattern} is omitted, it is treated like a @samp{?}, which matches
every character.

If @var{parameter} is @samp{@@} or @samp{*},
the case modification operation is applied to each positional
parameter in turn, and the expansion is the resultant list.
If @var{parameter}
is an array variable subscripted with @samp{@@} or @samp{*},
the case modification operation is applied to each member of the
array in turn, and the expansion is the resultant list.

@item $@{@var{parameter}@@@var{operator}@}
The expansion is either a transformation of the value of @var{parameter}
or information about @var{parameter} itself, depending on the value of
@var{operator}.  Each @var{operator} is a single letter:

@table @code
@item U
The expansion is a string that is the value of @var{parameter} with lowercase
alphabetic characters converted to uppercase.
@item u
The expansion is a string that is the value of @var{parameter} with the first
character converted to uppercase, if it is alphabetic.
@item L
The expansion is a string that is the value of @var{parameter} with uppercase
alphabetic characters converted to lowercase.
@item Q
The expansion is a string that is the value of @var{parameter} quoted in a
format that can be reused as input.
@item E
The expansion is a string that is the value of @var{parameter} with backslash
escape sequences expanded as with the @code{$'@dots{}'} quoting mechanism.
@item P
The expansion is a string that is the result of expanding the value of
@var{parameter} as if it were a prompt string (@pxref{Controlling the Prompt}).
@item A
The expansion is a string in the form of
an assignment statement or @code{declare} command that, if
evaluated, recreates @var{parameter} with its attributes and value.
@item K
Produces a possibly-quoted version of the value of @var{parameter},
except that it prints the values of
indexed and associative arrays as a sequence of quoted key-value pairs
(@pxref{Arrays}).
The keys and values are quoted in a format that can be reused as input.
@item a
The expansion is a string consisting of flag values representing
@var{parameter}'s attributes.
@item k
Like the @samp{K} transformation, but expands the keys and values of
indexed and associative arrays to separate words after word splitting.
@end table

If @var{parameter} is @samp{@@} or @samp{*},
the operation is applied to each positional
parameter in turn, and the expansion is the resultant list.
If @var{parameter}
is an array variable subscripted with @samp{@@} or @samp{*},
the operation is applied to each member of the
array in turn, and the expansion is the resultant list.

The result of the expansion is subject to word splitting and filename
expansion as described below.
@end table

@node Command Substitution
@subsection Command Substitution
@cindex command substitution

Command substitution allows the output of a command to replace
the command itself.
The standard form of command substitution occurs when a command is
enclosed as follows:
@example
$(@var{command})
@end example
@noindent
or (deprecated)
@example
`@var{command}`.
@end example

@noindent
Bash performs command substitution by executing @var{command} in a subshell
environment and replacing the command substitution with the standard
output of the command, with any trailing newlines deleted.
Embedded newlines are not deleted, but they may be removed during
word splitting.
The command substitution @code{$(cat @var{file})} can be
replaced by the equivalent but faster @code{$(< @var{file})}.

With the old-style backquote form of substitution,
backslash retains its literal meaning except when followed by
@samp{$}, @samp{`}, or @samp{\}. 
The first backquote not preceded by a backslash terminates the
command substitution.
When using the @code{$(@var{command})} form, all characters between
the parentheses make up the command; none are treated specially.

There is an alternate form of command substitution:

@example
$@{@var{c} @var{command}; @}
@end example

@noindent
which executes @var{command} in the current execution environment
and captures its output, again with trailing newlines removed.

The character @var{c} following the open brace must be a space, tab,
newline, @samp{|}, or @samp{;};
and the close brace must be in a position
where a reserved word may appear (i.e., preceded by a command terminator
such as semicolon).
Bash allows the close brace to be joined to the remaining characters in
the word without being followed by a shell metacharacter as a reserved
word would usually require.

Any side effects of @var{command} take effect immediately
in the current execution environment and persist in the current 
environment after the command completes (e.g., the @code{exit} builtin
exits the shell).

This type of command substitution superficially resembles executing an
unnamed shell function: local variables are created as when a shell
function is executing, and the @code{return} builtin forces
@var{command} to complete;
however, the rest of the execution environment,
including the positional parameters, is shared with the caller.

If the first character following the open brace is a
@samp{;},
the construct behaves like the form above but
preserves any trailing newlines in the output of @var{command}
instead of removing them.
This form is useful when the trailing newlines are significant
and should not be stripped from the command's output.

If the first character following the open brace is a
@samp{|},
the construct expands to the
value of the @code{REPLY} shell variable after @var{command} executes,
without removing any trailing newlines,
and the standard output of @var{command} remains the same as in the
calling shell.
Bash creates @code{REPLY} as an initially-unset local variable when
@var{command} executes, and restores @code{REPLY} to the value it had
before the command substitution after @var{command} completes,
as with any local variable.

For example, this construct expands to @samp{12345}, and leaves the
shell variable @code{X} unchanged in the current execution environment:

@example

$@{ local X=12345 ; echo $X; @}
@end example

@noindent
(not declaring @code{X} as local would modify its value in the current
environment, as with normal shell function execution),
while this construct does not require any output to expand to
@samp{12345}:

@example
$@{| REPLY=12345; @}
@end example

@noindent
and restores @code{REPLY} to the value it had before the command substitution.

Command substitutions may be nested.  To nest when using the backquoted
form, escape the inner backquotes with backslashes.

If the substitution appears within double quotes, Bash does not perform
word splitting and filename expansion on the results.

@node Arithmetic Expansion
@subsection Arithmetic Expansion
@cindex expansion, arithmetic
@cindex arithmetic expansion

Arithmetic expansion evaluates an arithmetic expression
and substitutes the result.
The format for arithmetic expansion is:

@example
$(( @var{expression} ))
@end example

The @var{expression} undergoes the same expansions
as if it were within double quotes,
but unescaped double quote characters
in @var{expression} are not treated
specially and are removed.
All tokens in the expression undergo parameter and variable expansion,
command substitution, and quote removal.
The result is treated as the arithmetic expression to be evaluated.
Since the way Bash handles double quotes
can potentially result in empty strings,
arithmetic expansion treats
those as expressions that evaluate to 0.
Arithmetic expansions may be nested. 

The evaluation is performed according to the rules listed below
(@pxref{Shell Arithmetic}).
If the expression is invalid, Bash prints a message indicating
failure to the standard error,
does not perform the substitution,
and does not
continue to execute the command in which the error occurs.

@node Process Substitution
@subsection Process Substitution
@cindex process substitution

Process substitution allows a process's input or output to be
referred to using a filename.
It takes the form of 
@example
<(@var{list})
@end example
@noindent
or
@example
>(@var{list})
@end example
@noindent
The process @var{list},
as long as it is not a null command without redirections,
is run asynchronously, and its input or output 
appears as a filename.
This filename is
passed as an argument to the current command as the result of the
expansion.

If the
@code{>(@var{list})} form is used, writing to the file
provides input for @var{list}.
If the
@code{<(@var{list})} form is used, reading the file 
obtains the output of @var{list}.
Note that no space may appear between the @code{<} or @code{>}
and the left parenthesis, otherwise the construct would be interpreted
as a redirection.

Process substitution is supported on systems that support named
pipes (@sc{fifo}s) or the @file{/dev/fd} method of naming open files.

When available, process substitution is performed simultaneously with
parameter and variable expansion, command substitution, and arithmetic
expansion.

@node Word Splitting
@subsection Word Splitting
@cindex word splitting

The shell scans the results of parameter expansion, command substitution,
and arithmetic expansion that did not occur within double quotes for
word splitting.
Words that were not expanded are not split.

The shell treats each character of @env{$IFS} as a delimiter,
and splits the results of the other expansions into fields
using these characters as field terminators.

An @dfn{IFS whitespace} character is whitespace as defined above
(@pxref{Definitions}) that appears in the value of @env{IFS}.
Space, tab, and newline are always considered IFS whitespace, even
if they don't appear in the locale's @code{space} category.

If @env{IFS} is unset, word splitting behaves as if its value were
@code{<space><tab><newline>},
and treats these characters as IFS whitespace.
If the value of @env{IFS} is null, no word splitting occurs,
but implicit null arguments (see below) are still removed.

Word splitting begins by removing sequences of IFS whitespace characters
from the beginning and end of the results of the previous expansions,
then splits the remaining words.

If the value of @env{IFS} consists solely of IFS whitespace,
any sequence of IFS whitespace characters delimits a field,
so a field consists of characters that are not unquoted IFS
whitespace, and null fields result only from quoting.

If @env{IFS} contains a non-whitespace character, then any
character in the value of @env{IFS} that is not IFS whitespace,
along with any adjacent IFS whitespace characters, delimits a field.
This means that adjacent non-IFS-whitespace delimiters produce a
null field.
A sequence of IFS whitespace characters also delimits a field.

Explicit null arguments (@code{""} or @code{''}) are retained
and passed to commands as empty strings.
Unquoted implicit null arguments, resulting from the expansion of
parameters that have no values, are removed.
Expanding a parameter with no value within double quotes
produces a null field,
which is retained and passed to a command as an empty string.

When a quoted null argument appears as part of a word whose expansion
is non-null, word splitting removes the null argument portion,
leaving the non-null expansion.
That is, the word
@code{-d''} becomes @code{-d} after word splitting and
null argument removal.

@node Filename Expansion
@subsection Filename Expansion
@menu
* Pattern Matching::	How the shell matches patterns.
@end menu
@cindex expansion, filename
@cindex expansion, pathname
@cindex filename expansion
@cindex pathname expansion

After word splitting, unless the @option{-f} option has been set
(@pxref{The Set Builtin}), Bash scans each word for the characters
@samp{*}, @samp{?}, and @samp{[}.
If one of these characters appears, and is not quoted, then the word is
regarded as a @var{pattern},
and replaced with a sorted list of filenames matching the pattern
(@pxref{Pattern Matching}),
subject to the value of the @code{GLOBSORT} shell variable
(@pxref{Bash Variables}).

If no matching filenames are found,
and the shell option @code{nullglob} is disabled, the word is left
unchanged.
If the @code{nullglob} option is set, and no matches are found, the word
is removed.
If the @code{failglob} shell option is set, and no matches are found,
Bash prints an error message and does not execute the command.
If the shell option @code{nocaseglob} is enabled, the match is performed
without regard to the case of alphabetic characters.

When a pattern is used for filename expansion, the character @samp{.}
at the start of a filename or immediately following a slash
must be matched explicitly, unless the shell option @code{dotglob} is set.
In order to match the filenames @file{.} and @file{..},
the pattern must begin with @samp{.} (for example, @samp{.?}),
even if @code{dotglob} is set.
If the @code{globskipdots} shell option is enabled, the filenames
@file{.} and @file{..} never match, even if the pattern begins
with a @samp{.}.
When not matching filenames, the @samp{.} character is not treated specially.

When matching a filename, the slash character must always be
matched explicitly by a slash in the pattern, but in other matching
contexts it can be matched by a special pattern character as described
below (@pxref{Pattern Matching}).

See the description of @code{shopt} in @ref{The Shopt Builtin},
for a description of the @code{nocaseglob}, @code{nullglob},
@code{globskipdots},
@code{failglob}, and @code{dotglob} options.

The @env{GLOBIGNORE}
shell variable may be used to restrict the set of file names matching a
pattern.
If @env{GLOBIGNORE}
is set, each matching file name that also matches one of the patterns in
@env{GLOBIGNORE} is removed from the list of matches.
If the @code{nocaseglob} option is set, the matching against the patterns in
@env{GLOBIGNORE} is performed without regard to case.
The filenames
@file{.} and @file{..}
are always ignored when @env{GLOBIGNORE}
is set and not null.
However, setting @env{GLOBIGNORE}
to a non-null value has the effect of enabling the
@code{dotglob}
shell option, so all other filenames beginning with a
@samp{.}
match.
To get the old behavior of ignoring filenames beginning with a
@samp{.},
make @samp{.*} one of the patterns in @env{GLOBIGNORE}.
The @code{dotglob} option is disabled when @env{GLOBIGNORE}
is unset.
The @code{GLOBIGNORE}
pattern matching honors the setting of the @code{extglob} shell
option.

The value of the
@env{GLOBSORT}
shell variable controls how the results of pathname expansion are sorted,
as described below (@pxref{Bash Variables}).

@node Pattern Matching
@subsubsection Pattern Matching
@cindex pattern matching
@cindex matching, pattern

Any character that appears in a pattern, other than the special pattern
characters described below, matches itself.
The @sc{nul} character may not occur in a pattern.
A backslash escapes the following character; the
escaping backslash is discarded when matching.
The special pattern characters must be quoted if they are to be matched
literally.

The special pattern characters have the following meanings:
@table @code
@item *
Matches any string, including the null string.
When the @code{globstar} shell option is enabled, and @samp{*} is used in
a filename expansion context, two adjacent @samp{*}s used as a single
pattern match all files and zero or more directories and
subdirectories.
If followed by a @samp{/}, two adjacent @samp{*}s match only
directories and subdirectories.
@item ?
Matches any single character.
@item [@dots{}]
Matches any one of the characters enclosed between the brackets.
This is known as a @dfn{bracket expression}
and matches a single character.
A pair of characters separated by a hyphen denotes a @dfn{range expression};
any character that falls between those two characters, inclusive,
using the current locale's collating sequence and character set, matches.
If the first character following the
@samp{[} is a @samp{!} or a @samp{^}
then any character not within the range matches.
To match a @samp{@minus{}}, include it as the first
or last character in the set.
To match a @samp{]}, include it as the first
character in the set.

The sorting order of characters in range expressions,
and the characters included in the range,
are determined by the collating sequence of the
current locale and the values of the
@env{LC_COLLATE} and @env{LC_ALL} shell variables, if set.

For example, in the default C locale,
@samp{[a-dx-z]}
is equivalent to
@samp{[abcdxyz]}.
Many locales sort characters in dictionary order, and in these locales
@samp{[a-dx-z]}
is typically not equivalent to
@samp{[abcdxyz]};
it might be equivalent to
@samp{[aBbCcDdxYyZz]}
or
@samp{[aAbBcCdxXyYz]}.
To obtain the traditional interpretation of ranges in bracket expressions,
you can force the use of the C locale by setting the
@env{LC_COLLATE} or @env{LC_ALL}
environment variables to the value @samp{C}, or enable the
@code{globasciiranges} shell option.

Within a bracket expression, @dfn{character classes} can be specified
using the syntax
@code{[:}@var{class}@code{:]}, where @var{class} is one of the
following classes defined in the
@sc{posix}
standard:
@example
alnum   alpha   ascii   blank   cntrl   digit   graph   lower
print   punct   space   upper   word    xdigit
@end example
@noindent
A character class matches any character belonging to that class.
The @code{word} character class matches letters, digits, and the character
@samp{_}.

For instance, the following pattern will match any character belonging
to the @code{space} character class in the current locale, then any
upper case letter or @samp{!}, a dot, and finally any lower case letter
or a hyphen.

@example
[[:space:]][[:upper:]!].[-[:lower:]]
@end example

Within a bracket expression, an @dfn{equivalence class} can be
specified using the syntax @code{[=}@var{c}@code{=]}, which
matches all characters with the same collation weight (as defined
by the current locale) as the character @var{c}.

Within a bracket expression, the syntax @code{[.}@var{symbol}@code{.]}
matches the collating symbol @var{symbol}.
@end table

If the @code{extglob} shell option is enabled using the @code{shopt}
builtin, the shell recognizes several extended pattern matching operators.
In the following description, a @var{pattern-list} is a list of one
or more patterns separated by a @samp{|}.
When matching filenames, the @code{dotglob} shell option determines
the set of filenames that are tested, as described above.
Composite patterns may be formed using one or more of the following
sub-patterns:

@table @code
@item ?(@var{pattern-list})
Matches zero or one occurrence of the given patterns.

@item *(@var{pattern-list})
Matches zero or more occurrences of the given patterns.

@item +(@var{pattern-list})
Matches one or more occurrences of the given patterns.

@item @@(@var{pattern-list})
Matches one of the given patterns.

@item !(@var{pattern-list})
Matches anything except one of the given patterns.
@end table

The @code{extglob} option changes the behavior of the parser, since the
parentheses are normally treated as operators with syntactic meaning.
To ensure that extended matching patterns are parsed correctly, make sure
that @code{extglob} is enabled before parsing constructs containing the
patterns, including shell functions and command substitutions.

When matching filenames, the @code{dotglob} shell option determines
the set of filenames that are tested:
when @code{dotglob} is enabled, the set of filenames includes all files
beginning with @samp{.}, but the filenames
@file{.} and @file{..} must be matched by a
pattern or sub-pattern that begins with a dot;
when it is disabled, the set does not
include any filenames beginning with @samp{.} unless the pattern
or sub-pattern begins with a @samp{.}.
If the @code{globskipdots}
shell option is enabled, the filenames
@file{.} and @file{..}
never appear in the set.
As above, @samp{.} only has a special meaning when matching filenames. 

Complicated extended pattern matching against long strings is slow,
especially when the patterns contain alternations and the strings
contain multiple matches.
Using separate matches against shorter strings, or using arrays of
strings instead of a single long string, may be faster.

@node Quote Removal
@subsection Quote Removal

After the preceding expansions, all unquoted occurrences of the
characters @samp{\}, @samp{'}, and @samp{"} that did not
result from one of the above expansions are removed.

@node Redirections
@section Redirections
@cindex redirection

Before a command is executed, its input and output may be
@dfn{redirected}
using a special notation interpreted by the shell.
@dfn{Redirection} allows commands' file handles to be
duplicated, opened, closed, made to refer to different files,
and can change the files the command reads from and writes to.
When used with the @code{exec} builtin,
redirections modify file handles in the current shell execution environment.
The following redirection
operators may precede or appear anywhere within a
simple command or may follow a command.
Redirections are processed in the order they appear, from
left to right.

Each redirection that may be preceded by a file descriptor number
may instead be preceded by a word of the form @{@var{varname}@}.
In this case, for each redirection operator except
@code{>&-}
and
@code{<&-},
the shell allocates a file descriptor greater
than or equal to 10 and assigns it to @{@var{varname}@}.
If @{@var{varname}@} precedes
@code{>&-}
or
@code{<&-},
the value of @var{varname} defines the file descriptor to close.
If @{@var{varname}@} is supplied, the redirection persists beyond
the scope of the command, which allows the shell programmer to
manage the file descriptor's lifetime manually without using
the @code{exec} builtin.
The @code{varredir_close} shell option manages this behavior
(@pxref{The Shopt Builtin}).

In the following descriptions, if the file descriptor number is
omitted, and the first character of the redirection operator is
@samp{<},
the redirection refers to the standard input (file descriptor 0).
If the first character of the redirection operator is
@samp{>},
the redirection refers to the standard output (file descriptor 1).

The @var{word} following the redirection operator in the following
descriptions, unless otherwise noted, is subjected to
brace expansion,
tilde expansion,
parameter and variable expansion,
command substitution,
arithmetic expansion,
quote removal,
filename expansion,
and word splitting.
If it expands to more than one word, Bash reports an error.

The order of redirections is significant.
For example,
the command
@example
ls > @var{dirlist} 2>&1
@end example
@noindent
directs both standard output (file descriptor 1) and standard error
(file descriptor 2) to the file @var{dirlist}, while the command
@example
ls 2>&1 > @var{dirlist}
@end example
@noindent
directs only the standard output to file @var{dirlist},
because the standard error was made a copy of the standard output
before the standard output was redirected to @var{dirlist}.

Bash handles several filenames specially when they are used in
redirections, as described in the following table.
If the operating system on which Bash is running provides these
special files, Bash uses them; otherwise it emulates them
internally with the behavior described below.

@table @code
@item /dev/fd/@var{fd}
If @var{fd} is a valid integer, duplicate file descriptor @var{fd}.

@item /dev/stdin
File descriptor 0 is duplicated.

@item /dev/stdout
File descriptor 1 is duplicated.

@item /dev/stderr
File descriptor 2 is duplicated.

@item /dev/tcp/@var{host}/@var{port}
If @var{host} is a valid hostname or Internet address, and @var{port}
is an integer port number or service name, Bash attempts to open
the corresponding TCP socket.

@item /dev/udp/@var{host}/@var{port}
If @var{host} is a valid hostname or Internet address, and @var{port}
is an integer port number or service name, Bash attempts to open 
the corresponding UDP socket.
@end table

A failure to open or create a file causes the redirection to fail.

Redirections using file descriptors greater than 9 should be used with
care, as they may conflict with file descriptors the shell uses
internally.

@subsection Redirecting Input
Redirecting input opens the file whose name results from
the expansion of @var{word} for reading on file descriptor @code{n},
or the standard input (file descriptor 0) if @code{n}
is not specified.

The general format for redirecting input is:
@example
[@var{n}]<@var{word}
@end example

@subsection Redirecting Output
Redirecting output opens the file whose name results from
the expansion of @var{word}
for writing on file descriptor @var{n},
or the standard output (file descriptor 1) if @var{n}
is not specified.
If the file does not exist it is created;
if it does exist it is truncated to zero size.

The general format for redirecting output is:
@example
[@var{n}]>[|]@var{word}
@end example

If the redirection operator is
@samp{>},
and the
@code{noclobber}
option to the
@code{set}
builtin command has been enabled, the redirection fails if the file
whose name results from the expansion of @var{word} exists and is
a regular file.
If the redirection operator is @samp{>|},
or the redirection operator is @samp{>} and
the @code{noclobber} option to the @code{set}
builtin is not enabled,
Bash attempts the redirection
even if the file named by @var{word} exists.

@subsection Appending Redirected Output
Redirecting output in this fashion opens
the file whose name results from the expansion of
@var{word}
for appending on file descriptor @var{n},
or the standard output (file descriptor 1) if @var{n}
is not specified.
If the file does not exist it is created.

The general format for appending output is:
@example
[@var{n}]>>@var{word}
@end example

@subsection Redirecting Standard Output and Standard Error
This construct redirects both the
standard output (file descriptor 1) and
the standard error output (file descriptor 2)
to the file whose name is the expansion of @var{word}.

There are two formats for redirecting standard output and
standard error:
@example
&>@var{word}
@end example
@noindent
and
@example
>&@var{word}
@end example
@noindent
Of the two forms, the first is preferred.
This is semantically equivalent to
@example
>@var{word} 2>&1
@end example
When using the second form, @var{word} may not expand to a number or
@samp{-}.
If it does, other redirection operators apply
(see Duplicating File Descriptors below) for compatibility reasons.

@subsection Appending Standard Output and Standard Error
This construct appends both the
standard output (file descriptor 1) and
the standard error output (file descriptor 2)
to the file whose name is the expansion of @var{word}.

The format for appending standard output and standard error is:
@example
&>>@var{word}
@end example
@noindent
This is semantically equivalent to
@example
>>@var{word} 2>&1
@end example
(see Duplicating File Descriptors below).

@subsection Here Documents
This type of redirection instructs the shell to read input from the
current source until it reads a line containing only @var{delimiter}
(with no trailing blanks).
All of the lines read up to that point then become the standard
input (or file descriptor @var{n} if @var{n} is specified) for a command.

The format of here-documents is:
@example
[@var{n}]<<[@minus{}]@var{word}
        @var{here-document}
@var{delimiter}
@end example

The shell does not perform
parameter and variable expansion, command substitution,
arithmetic expansion, or filename expansion on
@var{word}.

If any part of @var{word} is quoted, the
@var{delimiter} is the result of quote removal on @var{word},
and the lines in the here-document are not expanded.
If @var{word} is unquoted,
@var{delimiter} is @var{word} itself,
and the here-document text is treated similarly to a double-quoted string:
all lines of the here-document are subjected to
parameter expansion, command substitution, and arithmetic expansion,
the character sequence @code{\newline} is treated as a line continuation,
and @samp{\} must be used to quote the characters
@samp{\}, @samp{$}, and @samp{`};
however, double quote characters have no special meaning.

If the redirection operator is @samp{<<-},
the shell strips leading tab characters from input lines
and the line containing @var{delimiter}.
This allows here-documents within shell scripts to be indented in a
natural fashion.

If the delimiter is not quoted, the shell treats the
@code{\<newline>}
sequence as a line continuation: the two lines are joined
and the backslash-newline is removed.
This happens while reading the here-document, before the check for
the ending delimiter, so joined lines can form the end delimiter.

@subsection Here Strings
A variant of here documents, the format is:
@example
[@var{n}]<<< @var{word}
@end example

The @var{word} undergoes
tilde expansion, parameter and variable expansion,
command substitution, arithmetic expansion, and quote removal.
Filename expansion and word splitting are not performed.
The result is supplied as a single string, with a newline appended,
to the command on its
standard input (or file descriptor @var{n} if @var{n} is specified).

@subsection Duplicating File Descriptors
The redirection operator
@example
[@var{n}]<&@var{word}
@end example
@noindent
is used to duplicate input file descriptors.
If @var{word}
expands to one or more digits, file descriptor @var{n}
is made to be a copy of that file descriptor.
It is a redirection error if the digits in
@var{word} do not specify a file descriptor open for input.
If @var{word}
evaluates to @samp{-}, file descriptor @var{n} is closed.
If @var{n} is not specified, this uses the standard input (file descriptor 0).

The operator
@example
[@var{n}]>&@var{word}
@end example
@noindent
is used similarly to duplicate output file descriptors.
If @var{n} is not specified, this uses the standard output (file descriptor 1).
It is a redirection error if the digits in
@var{word} do not specify a file descriptor open for output.
If @var{word}
evaluates to @samp{-}, file descriptor @var{n} is closed.
As a special case, if @var{n} is omitted, and
@var{word} does not expand to one or more digits or @samp{-},
this redirects the standard output and standard error as described
previously.

@subsection Moving File Descriptors
The redirection operator
@example
[@var{n}]<&@var{digit}-
@end example
@noindent
moves the file descriptor @var{digit} to file descriptor @var{n},
or the standard input (file descriptor 0) if @var{n} is not specified.
@var{digit} is closed after being duplicated to @var{n}.

Similarly, the redirection operator
@example
[@var{n}]>&@var{digit}-
@end example
@noindent
moves the file descriptor @var{digit} to file descriptor @var{n},
or the standard output (file descriptor 1) if @var{n} is not specified.

@subsection Opening File Descriptors for Reading and Writing
The redirection operator
@example
[@var{n}]<>@var{word}
@end example
@noindent
opens the file whose name is the expansion of @var{word}
for both reading and writing on file descriptor
@var{n}, or on file descriptor 0 if @var{n}
is not specified.
If the file does not exist, it is created.

@node Executing Commands
@section Executing Commands

@menu
* Simple Command Expansion::	How Bash expands simple commands before
				executing them.
* Command Search and Execution::	How Bash finds commands and runs them.
* Command Execution Environment::	The environment in which Bash
					executes commands that are not
					shell builtins.
* Environment::		The environment given to a command.
* Exit Status::		The status returned by commands and how Bash
			interprets it.
* Signals::		What happens when Bash or a command it runs
			receives a signal.
@end menu

@node Simple Command Expansion
@subsection Simple Command Expansion
@cindex command expansion

When the shell executes a simple command, it performs the following
expansions, assignments, and redirections, from left to right, in
the following order.

@enumerate
@item
The words that the parser has marked as variable assignments (those
preceding the command name) and redirections are saved for later
processing.

@item
The words that are not variable assignments or redirections are
expanded (@pxref{Shell Expansions}).
If any words remain after expansion, the first word
is taken to be the name of the command and the remaining words are
the arguments.

@item
Redirections are performed as described above (@pxref{Redirections}).

@item
The text after the @samp{=} in each variable assignment undergoes tilde
expansion, parameter expansion, command substitution, arithmetic expansion,
and quote removal before being assigned to the variable.
@end enumerate

If no command name results, the variable assignments affect the current
shell environment.
In the case of such a command (one that consists only of assignment
statements and redirections), assignment statements are performed before
redirections.
Otherwise, the variables are added to the environment
of the executed command and do not affect the current shell environment.
If any of the assignments attempts to assign a value to a readonly variable,
an error occurs, and the command exits with a non-zero status.

If no command name results, redirections are performed, but do not
affect the current shell environment.
A redirection error causes the command to exit with a non-zero status.

If there is a command name left after expansion, execution proceeds as
described below.
Otherwise, the command exits.
If one of the expansions contained a command substitution,
the exit status of the command is the exit status of
the last command substitution performed.
If there were no command substitutions,
the command exits with a zero status.

@node Command Search and Execution
@subsection Command Search and Execution
@cindex command execution
@cindex command search

After a command has been split into words, if it results in a
simple command and an optional list of arguments, the shell performs
the following actions.

@enumerate
@item
If the command name contains no slashes, the shell attempts to
locate it.
If there exists a shell function by that name, that
function is invoked as described in @ref{Shell Functions}.

@item
If the name does not match a function, the shell searches for
it in the list of shell builtins.
If a match is found, that builtin is invoked.

@item
If the name is neither a shell function nor a builtin,
and contains no slashes, Bash searches each element of
@env{$PATH} for a directory containing an executable file
by that name.
Bash uses a hash table to remember the full
pathnames of executable files to avoid multiple @env{PATH} searches
(see the description of @code{hash} in @ref{Bourne Shell Builtins}).
Bash performs a full search of the directories in @env{$PATH}
only if the command is not found in the hash table.
If the search is unsuccessful, the shell searches for a defined shell
function named @code{command_not_found_handle}.
If that function exists, it is invoked in a separate execution environment
with the original command and
the original command's arguments as its arguments, and the function's
exit status becomes the exit status of that subshell.
If that function is not defined, the shell prints an error
message and returns an exit status of 127.

@item
If the search is successful, or if the command name contains
one or more slashes, the shell executes the named program in
a separate execution environment.
Argument 0 is set to the name given, and the remaining arguments
to the command are set to the arguments supplied, if any.

@item
If this execution fails because the file is not in executable
format, and the file is not a directory, it is assumed to be a
@dfn{shell script},
a file containing shell commands,
and the shell executes it as described in @ref{Shell Scripts}.

@item
If the command was not begun asynchronously, the shell waits for
the command to complete and collects its exit status.

@end enumerate

@node Command Execution Environment
@subsection Command Execution Environment
@cindex execution environment

The shell has an @dfn{execution environment}, which consists of the
following:

@itemize @bullet
@item
Open files inherited by the shell at invocation, as modified by
redirections supplied to the @code{exec} builtin.

@item
The current working directory as set by @code{cd}, @code{pushd}, or
@code{popd}, or inherited by the shell at invocation.

@item
The file creation mode mask as set by @code{umask} or inherited from
the shell's parent.

@item
Current traps set by @code{trap}.

@item
Shell parameters that are set by variable assignment or with @code{set}
or inherited from the shell's parent in the environment.

@item
Shell functions defined during execution or inherited from the shell's
parent in the environment.

@item
Options enabled at invocation (either by default or with command-line
arguments) or by @code{set}.

@item
Options enabled by @code{shopt} (@pxref{The Shopt Builtin}).

@item
Shell aliases defined with @code{alias} (@pxref{Aliases}).

@item
Various process @sc{id}s, including those of background jobs
(@pxref{Lists}), the value of @code{$$}, and the value of @env{$PPID}.

@end itemize

When a simple command other than a builtin or shell function
is to be executed, it
is invoked in a separate execution environment that consists of
the following.
Unless otherwise noted, the values are inherited from the shell.

@itemize @bullet
@item
The shell's open files, plus any modifications and additions specified
by redirections to the command.

@item
The current working directory.

@item
The file creation mode mask.

@item
Shell variables and functions marked for export, along with variables
exported for the command, passed in the environment (@pxref{Environment}).

@item
Traps caught by the shell are reset to the values inherited from the
shell's parent, and traps ignored by the shell are ignored.

@end itemize

A command invoked in this separate environment cannot affect the
shell's execution environment.

A @dfn{subshell} is a copy of the shell process.

Command substitution, commands grouped with parentheses,
and asynchronous commands are invoked in a
subshell environment that is a duplicate of the shell environment,
except that traps caught by the shell are reset to the values
that the shell inherited from its parent at invocation.
Builtin commands that are invoked as part of a pipeline,
except possibly in the last element depending on the value of the
@code{lastpipe} shell option (@pxref{The Shopt Builtin}),
are also executed in a subshell environment.
Changes made to the subshell environment
cannot affect the shell's execution environment.

When the shell is in @sc{posix} mode,
subshells spawned to execute command substitutions inherit the value of
the @option{-e} option from the parent shell.
When not in @sc{posix} mode,
Bash clears the @option{-e} option in such subshells
See the description of the @code{inherit_errexit} shell option
(@pxref{Bash Builtins}) for how to control this behavior when not
in @sc{posix} mode.

If a command is followed by a @samp{&} and job control is not active, the
default standard input for the command is the empty file @file{/dev/null},
unless the command has an explicit redirection involving the standard input.
Otherwise, the invoked command inherits the file descriptors of the calling
shell as modified by redirections.

@node Environment
@subsection Environment
@cindex environment

When a program is invoked it is given an array of strings
called the @dfn{environment}.
This is a list of name-value pairs, of the form @code{name=value}.

Bash provides several ways to manipulate the environment.
On invocation, the shell scans its own environment and
creates a parameter for each name found, automatically marking
it for @code{export} to child processes.
Executed commands inherit the environment.
The @code{export}, @samp{declare -x}, and @code{unset}
commands modify the environment by
adding and deleting parameters and functions.
If the value of a parameter in the environment is modified,
the new value automatically becomes part
of the environment, replacing the old.
The environment
inherited by any executed command consists of the shell's
initial environment, whose values may be modified in the shell,
less any pairs removed by the @code{unset} and @samp{export -n}
commands, plus any additions via the @code{export} and
@samp{declare -x} commands.

If any parameter assignment statements,
as described in @ref{Shell Parameters},
appear before a simple command,
the variable assignments are part of that command's environment
for as long as it executes.
These assignment statements affect only the environment seen
by that command.
If these assignments precede a call to a shell function, the variables
are local to the function and exported to that function's children.

If the @option{-k} option is set (@pxref{The Set Builtin}), then all
parameter assignments are placed in the environment for a command,
not just those that precede the command name.

When Bash invokes an external command, the variable @samp{$_}
is set to the full pathname of the command and passed to that
command in its environment.

@node Exit Status
@subsection Exit Status
@cindex exit status

The exit status of an executed command is the value returned by the
@code{waitpid} system call or equivalent function.
Exit statuses fall between 0 and 255, though, as explained below,
the shell may use values above 125 specially.
Exit statuses from shell builtins and compound commands are also limited
to this range.
Under certain circumstances, the shell will use special values to
indicate specific failure modes.

For the shell's purposes, a command which exits with a
zero exit status has succeeded.
So while an exit status of zero indicates success, a non-zero
exit status indicates failure.
This seemingly counter-intuitive scheme is used so there
is one well-defined way to indicate success and a variety of
ways to indicate various failure modes.

When a command terminates on a fatal signal whose number is @var{N},
Bash uses the value 128+@var{N} as the exit status.

If a command is not found, the child process created to
execute it returns a status of 127.
If a command is found but is not executable, the return status is 126.

If a command fails because of an error during expansion or redirection,
the exit status is greater than zero.

The exit status is used by the Bash conditional commands
(@pxref{Conditional Constructs}) and some of the list
constructs (@pxref{Lists}).

All of the Bash builtins return an exit status of zero if they succeed
and a non-zero status on failure, so they may be used by the
conditional and list constructs.
All builtins return an exit status of 2 to indicate incorrect usage,
generally invalid options or missing arguments.

The exit status of the last command is available in the special
parameter $? (@pxref{Special Parameters}).

Bash itself returns the exit status of the last command
executed, unless a syntax error occurs, in which case it exits
with a non-zero value.
See also the @code{exit} builtin command (@pxref{Bourne Shell Builtins}).

@node Signals
@subsection Signals
@cindex signal handling

When Bash is interactive, in the absence of any traps,
it ignores @code{SIGTERM}
(so that @samp{kill 0} does not kill an interactive shell),
and catches and handles @code{SIGINT}
(so that the @code{wait} builtin is interruptible).
When Bash receives a @code{SIGINT},
it breaks out of any executing loops and command lists.
In all cases, Bash ignores @code{SIGQUIT}.
If job control is in effect (@pxref{Job Control}), Bash
ignores @code{SIGTTIN}, @code{SIGTTOU}, and @code{SIGTSTP}.

The @code{trap} builtin modifies the shell's signal handling, as
described below (@pxref{Bourne Shell Builtins}).

Non-builtin commands Bash executes have signal handlers set to the
values inherited by the shell from its parent,
unless @code{trap} sets them to be ignored, in which case the child
process will ignore them as well.
When job control is not in effect, asynchronous commands
ignore @code{SIGINT} and @code{SIGQUIT} in addition to these inherited
handlers.
Commands run as a result of
command substitution ignore the keyboard-generated job control signals
@code{SIGTTIN}, @code{SIGTTOU}, and @code{SIGTSTP}.

The shell exits by default upon receipt of a @code{SIGHUP}.
Before exiting, an interactive shell resends the @code{SIGHUP} to
all jobs, running or stopped.
The shell sends @code{SIGCONT} to stopped jobs to ensure that they
receive the @code{SIGHUP}
(@xref{Job Control}, for more information about running and stopped jobs).
To prevent the shell from sending the @code{SIGHUP} signal to a
particular job, remove it from the jobs table with the @code{disown}
builtin (@pxref{Job Control Builtins}) or mark it
not to receive @code{SIGHUP} using @code{disown -h}.

If the  @code{huponexit} shell option has been set using @code{shopt}
(@pxref{The Shopt Builtin}), Bash sends a @code{SIGHUP} to all jobs when
an interactive login shell exits.

If Bash is waiting for a command to complete and receives a signal
for which a trap has been set,
it will not execute the trap until the command completes.
If Bash is waiting for an asynchronous command via the @code{wait} builtin,
and it receives a signal for which a trap has been set,
the @code{wait} builtin will return immediately with an exit status
greater than 128, immediately after which the shell executes the trap.

When job control is not enabled, and Bash is waiting for a foreground
command to complete, the shell receives keyboard-generated signals
such as @code{SIGINT} (usually generated by @samp{^C}) that users
commonly intend to send to that command.
This happens because the shell and the command are in the same process
group as the terminal, and @samp{^C} sends @code{SIGINT} to all processes
in that process group.
Since Bash does not enable job control by default when the
shell is not interactive,
this scenario is most common in non-interactive shells.

When job control is enabled, and Bash is waiting for a foreground
command to complete, the shell does not receive keyboard-generated
signals, because it is not in the same process group as the terminal.
This scenario is most common in interactive shells, where Bash
attempts to enable job control by default.
See @ref{Job Control}, for a more in-depth discussion of process groups.

When job control is not enabled, and Bash receives @code{SIGINT}
while waiting for a foreground command, it waits until that foreground
command terminates and then decides what to do about the @code{SIGINT}:

@enumerate
@item
If the command terminates due to the @code{SIGINT}, Bash concludes
that the user meant to send the @code{SIGINT} to the shell as well,
and acts on the
@code{SIGINT}
(e.g., by running a @code{SIGINT} trap,
exiting a non-interactive shell,
or returning to the top level to read a new command).

@item
If the command does not terminate due to @code{SIGINT}, the program
handled the @code{SIGINT} itself and did not treat it as a fatal signal.
In that case, Bash does not treat @code{SIGINT} as a fatal signal,
either, instead assuming that the @code{SIGINT} was used as part of the
program's normal operation (e.g., @command{emacs} uses it to abort editing
commands) or deliberately discarded.
However, Bash will run any
trap set on @code{SIGINT}, as it does with any other trapped signal it
receives while it is waiting for the foreground command to
complete, for compatibility.
@end enumerate

When job control is enabled, Bash does not receive keyboard-generated
signals such as @code{SIGINT}
while it is waiting for a foreground command.
An interactive shell does not pay attention to the @code{SIGINT},
even if the foreground command terminates as a result, other than noting
its exit status.
If the shell is not interactive, and
the foreground command terminates due to the @code{SIGINT},
Bash pretends it received the @code{SIGINT}
itself (scenario 1 above), for compatibility.

@node Shell Scripts
@section Shell Scripts
@cindex shell script

A @dfn{shell script} is a text file containing shell commands.
When such a file is used as the first non-option argument when
invoking Bash, and neither the @option{-c} nor @option{-s} option
is supplied (@pxref{Invoking Bash}), 
Bash reads and executes commands from the file, then exits.
This mode of operation creates a non-interactive shell.
If the filename does not contain any slashes, the shell first searches
for the file in the current directory, and looks in the directories in
@env{$PATH} if not found there.

Bash tries to determine whether the file is a text file or a binary,
and will not execute files it determines to be binaries.

When Bash runs
a shell script, it sets the special parameter @code{0} to the name
of the file, rather than the name of the shell, and the positional
parameters are set to the remaining arguments, if any are given.
If no additional arguments are supplied, the positional parameters
are unset.

A shell script may be made executable by using the @code{chmod} command
to turn on the execute bit.
When Bash finds such a file while searching the @env{$PATH} for a command,
it creates a new instance of itself to execute it.
In other words, executing
@example
filename @var{arguments}
@end example
@noindent
is equivalent to executing
@example
bash filename @var{arguments}
@end example

@noindent
if @code{filename} is an executable shell script.
This subshell reinitializes itself, so that the effect is as if a
new shell had been invoked to interpret the script, with the
exception that the locations of commands remembered by the parent
(see the description of @code{hash} in @ref{Bourne Shell Builtins})
are retained by the child.

The @sc{gnu} operating system,
and most versions of Unix,
make this a part of the operating system's command execution mechanism.
If the first line of a script begins with
the two characters @samp{#!}, the remainder of the line specifies
an interpreter for the program and, depending on the operating system, one
or more optional arguments for that interpreter.
Thus, you can specify Bash, @code{awk}, Perl, or some other
interpreter and write the rest of the script file in that language.

The arguments to the interpreter
consist of one or more optional arguments following the interpreter
name on the first line of the script file, followed by the name of
the script file, followed by the rest of the arguments supplied to the
script.
The details of how the interpreter line is split into an interpreter name
and a set of arguments vary across systems.
Bash will perform this action on operating systems that do not handle it
themselves.
Note that some older versions of Unix limit the interpreter
name and a single argument to a maximum of 32 characters, so it's not
portable to assume that using more than one argument will work.

Bash scripts often begin with @code{#! /bin/bash} (assuming that
Bash has been installed in @file{/bin}), since this ensures that
Bash will be used to interpret the script, even if it is executed
under another shell.
It's a common idiom to use @code{env} to find
@code{bash} even if it's been installed in another directory:
@code{#!/usr/bin/env bash} will find the first occurrence of @code{bash}
in @env{$PATH}.

@node Shell Builtin Commands
@chapter Shell Builtin Commands

@menu
* Bourne Shell Builtins::	Builtin commands inherited from the Bourne
				Shell.
* Bash Builtins::		Table of builtins specific to Bash.
* Modifying Shell Behavior::	Builtins to modify shell attributes and
				optional behavior.
* Special Builtins::		Builtin commands classified specially by
				POSIX.
@end menu

Builtin commands are contained within the shell itself.
When the name of a builtin command is used as the first word of
a simple command (@pxref{Simple Commands}), the shell executes
the command directly, without invoking another program.
Builtin commands are necessary to implement functionality impossible
or inconvenient to obtain with separate utilities.

This section briefly describes the builtins which Bash inherits from
the Bourne Shell, as well as the builtin commands which are unique
to or have been extended in Bash.

Several builtin commands are described in other chapters: builtin
commands which provide the Bash interface to the job control
facilities (@pxref{Job Control Builtins}), the directory stack
(@pxref{Directory Stack Builtins}), the command history
(@pxref{Bash History Builtins}), and the programmable completion
facilities (@pxref{Programmable Completion Builtins}).

Many of the builtins have been extended by @sc{posix} or Bash.

Unless otherwise noted, each builtin command documented as accepting
options preceded by @samp{-} accepts @samp{--}
to signify the end of the options.
The @code{:}, @code{true}, @code{false}, and @code{test}/@code{[}
builtins do not accept options and do not treat @samp{--} specially.
The @code{exit}, @code{logout}, @code{return},
@code{break}, @code{continue}, @code{let},
and @code{shift} builtins accept and process arguments beginning
with @samp{-} without requiring @samp{--}.
Other builtins that accept arguments but are not specified as accepting
options interpret arguments beginning with @samp{-} as invalid options and
require @samp{--} to prevent this interpretation.

All builtins except
@code{:}, @code{true}, @code{false}, @code{echo}, and @code{test}/@code{[}
accept @samp{--help} as a special option.
If @samp{--help} is supplied, these builtins output
a help message and exit with a status of 0.  

@node Bourne Shell Builtins
@section Bourne Shell Builtins

The following shell builtin commands are inherited from the Bourne Shell.
These commands are implemented as specified by the @sc{posix} standard.

@table @code
@item :    @r{(a colon)}
@btindex :
@example
: [@var{arguments}]
@end example

Do nothing beyond expanding @var{arguments} and performing redirections.
The return status is zero.

@item .    @r{(a period)}
@btindex .
@example
. [-p @var{path}] @var{filename} [@var{arguments}]
@end example

The @code{.} command reads and execute commands from the @var{filename}
argument in the current shell context.

If @var{filename} does not contain a slash, @code{.} searches for it.
If @option{-p} is supplied, @code{.} treats @var{path}
as a colon-separated list of directories in which to find @var{filename};
otherwise, @code{.} uses the directories in @env{PATH} to find @var{filename}.
@var{filename} does not need to be executable.
When Bash is not in @sc{posix} mode, it searches the current directory
if @var{filename} is not found in @env{$PATH},
but does not search the current directory if @option{-p} is supplied.
If the @code{sourcepath} option (@pxref{The Shopt Builtin}) is turned off,
@code{.} does not search @env{PATH}.

If any @var{arguments} are supplied, they become the positional
parameters when @var{filename} is executed.
Otherwise the positional parameters are unchanged.

If the @option{-T} option is enabled, @code{.} inherits any trap on
@code{DEBUG}; if it is not, any @code{DEBUG} trap string is saved and
restored around the call to @code{.}, and @code{.} unsets the
@code{DEBUG} trap while it executes.
If @option{-T} is not set, and the sourced file changes
the @code{DEBUG} trap, the new value persists after @code{.} completes.
The return status is the exit status of the last command executed
from @var{filename}, or
zero if no commands are executed.
If @var{filename} is not found, or cannot be read,
the return status is non-zero.
This builtin is equivalent to @code{source}.

@item break
@btindex break
@example
break [@var{n}]
@end example

Exit from a @code{for}, @code{while}, @code{until}, or @code{select} loop.
If @var{n} is supplied, @code{break} exits the @var{n}th enclosing loop.
@var{n} must be greater than or equal to 1.
The return status is zero unless @var{n} is not greater than or equal to 1.

@item cd
@btindex cd
@example
cd [-L] [-@@] [@var{directory}]
cd -P [-e] [-@@] [@var{directory}]
@end example

Change the current working directory to @var{directory}.
If @var{directory} is not supplied, the value of the @env{HOME}
shell variable is used as @var{directory}.
If @var{directory} is the empty string, @code{cd} treats it as an error.
If the shell variable
@env{CDPATH} exists,
and @var{directory} does not begin with a slash,
@code{cd} uses it as a search path:
@code{cd} searches each directory name in @env{CDPATH} for
@var{directory}, with alternative directory names in @env{CDPATH}
separated by a colon (@samp{:}).
A null directory name in @env{CDPATH} means the same thing as the
current directory.

The @option{-P} option means not to follow symbolic links: symbolic links
are resolved while @code{cd} is traversing @var{directory} and before
processing an instance of @file{..} in @var{directory}.

By default, or when the @option{-L} option is supplied, symbolic links
in @var{directory} are resolved after @code{cd} processes an instance
of @file{..} in @var{directory}.

If @file{..} appears in @var{directory}, @code{cd} processes it by removing the
immediately preceding pathname component, back to a slash or the beginning
of @var{directory},
and verifying that the portion of @var{directory}
it has processed to
that point is still a valid directory name after removing the pathname
component.
If it is not a valid directory name, @code{cd} returns a non-zero status.

If the @option{-e} option is supplied with @option{-P}
and @code{cd} cannot successfully determine the current working directory 
after a successful directory change, it returns a non-zero status.

On systems that support it, the @option{-@@} option presents the extended
attributes associated with a file as a directory.              

If @var{directory} is @samp{-}, it is converted to @env{$OLDPWD}
before attempting the directory change.

If @code{cd} uses a non-empty directory name from @env{CDPATH}, or if
@samp{-} is the first argument, and the directory change is
successful, @code{cd} writes the absolute pathname of the new
working directory to the standard output.

If the directory change is successful, @code{cd} sets the value of the
@env{PWD} environment variable to the new directory name, and sets the
@env{OLDPWD} environment variable to the value of the current working
directory before the change.

The return status is zero if the directory is successfully changed,
non-zero otherwise.

@item continue
@btindex continue
@example
continue [@var{n}]
@end example

@code{continue} resumes the next iteration of an enclosing @code{for},
@code{while}, @code{until}, or @code{select} loop.
If @var{n} is supplied, Bash resumes the execution of the @var{n}th
enclosing loop.
@var{n} must be greater than or equal to 1.
The return status is zero unless @var{n} is not greater than or equal to 1.

@item eval
@btindex eval
@example
eval [@var{arguments}]
@end example

The @var{arguments} are concatenated together into a single command,
separated by spaces.
Bash then reads and executes this command and returns its exit status
as the exit status of @code{eval}.
If there are no arguments or only empty arguments, the return status is
zero.

@item exec
@btindex exec
@example
exec [-cl] [-a @var{name}] [@var{command} [@var{arguments}]]
@end example

If @var{command}
is supplied, it replaces the shell without creating a new process.
@var{command} cannot be a shell builtin or function.
The @var{arguments} become the arguments to @var{command}
If the @option{-l} option is supplied, the shell places a dash at the
beginning of the zeroth argument passed to @var{command}.
This is what the @code{login} program does.
The @option{-c} option causes @var{command} to be executed with an empty
environment.
If @option{-a} is supplied, the shell passes @var{name} as the zeroth
argument to @var{command}.

If @var{command}
cannot be executed for some reason, a non-interactive shell exits,
unless the @code{execfail} shell option is enabled.
In that case, it returns a non-zero status.
An interactive shell returns a non-zero status if the file cannot be executed.
A subshell exits unconditionally if @code{exec} fails.

If @var{command} is not specified, redirections may be used to affect
the current shell environment.
If there are no redirection errors, the return status is zero;
otherwise the return status is non-zero.

@item exit
@btindex exit
@example
exit [@var{n}]
@end example

Exit the shell, returning a status of @var{n} to the shell's parent.
If @var{n} is omitted, the exit status is that of the last command executed.
Any trap on @code{EXIT} is executed before the shell terminates.

@item export
@btindex export
@example
export [-fn] [-p] [@var{name}[=@var{value}]]
@end example

Mark each @var{name} to be passed to subsequently executed commands in the
environment.
If the @option{-f} option is supplied, the @var{name}s
refer to shell functions; otherwise the names refer to shell variables.

The @option{-n} option means to unexport each name: no longer mark
it for export.
If no @var{name}s are supplied, or if only
the @option{-p} option is given,
@code{export} displays a list of names of all exported
variables on the standard output.
Using @option{-p} and @option{-f} together displays exported functions.
The @option{-p} option displays output in a form that may be reused as input.

@code{export} allows the value of a variable to be set at the same time
it is exported or unexported by following the variable name with
=@var{value}.
This sets the value of the variable is to @var{value} while modifying
the export attribute.

The return status is zero unless an invalid option is supplied, one of
the names is not a valid shell variable name, or @option{-f} is supplied
with a name that is not a shell function.

@item false
@btindex false
@example
false
@end example

Does nothing; returns a non-zero status.

@item getopts
@btindex getopts
@example
getopts @var{optstring} @var{name} [@var{arg} @dots{}]
@end example

@code{getopts} is used by shell scripts or functions to parse positional
parameters and obtain options and their arguments.
@var{optstring} contains the option characters to be recognized; if a
character is followed by a colon, the option is expected to have an
argument, which should be separated from it by whitespace.
The colon (@samp{:}) and question mark (@samp{?}) may not be
used as option characters.

Each time it is invoked, @code{getopts}
places the next option in the shell variable @var{name}, initializing
@var{name} if it does not exist,
and the index of the next argument to be processed into the
variable @env{OPTIND}.
@env{OPTIND} is initialized to 1 each time the shell or a shell script
is invoked.
When an option requires an argument,
@code{getopts} places that argument into the variable @env{OPTARG}.

The shell does not reset @env{OPTIND} automatically; it must be manually
reset between multiple calls to @code{getopts} within the same shell
invocation to use a new set of parameters.

When it reaches the end of options, @code{getopts} exits with a
return value greater than zero.
@env{OPTIND} is set to the index of the first non-option argument,
and @var{name} is set to @samp{?}.

@code{getopts}
normally parses the positional parameters, but if more arguments are
supplied as @var{arg} values, @code{getopts} parses those instead.

@code{getopts} can report errors in two ways.
If the first character of
@var{optstring} is a colon, @code{getopts} uses @emph{silent}
error reporting.
In normal operation, @code{getopts} prints diagnostic messages
when it encounters invalid options or missing option arguments.
If the variable @env{OPTERR}
is set to 0, @code{getopts} does not display any error messages,
even if the first character of @code{optstring} is not a colon.

If @code{getopts} detects an invalid option, it
places @samp{?} into @var{name} and, if not silent,
prints an error message and unsets @env{OPTARG}.
If @code{getopts} is silent, it assigns the option character found
to @env{OPTARG} and does not print a diagnostic message.

If a required argument is not found,
and @code{getopts} is not silent,
it sets the value of @var{name} to a question mark (@samp{?}),
unsets @code{OPTARG}, and prints a diagnostic message.
If @code{getopts} is silent,
it sets the value of @var{name} to a colon (@samp{:}),
and sets @env{OPTARG} to the option character found.

@code{getopts} returns true if an option, specified or unspecified,
is found.
It returns false when it encounters the end of options or if an error occurs.

@item hash
@btindex hash
@example
hash [-r] [-p @var{filename}] [-dt] [@var{name}]
@end example

Each time @code{hash} is invoked, it remembers the full filenames of the
commands specified as @var{name} arguments,
so they need not be searched for on subsequent invocations.
The commands are found by searching through the directories listed in
@env{$PATH}.
Any previously-remembered filename associated with @var{name} is discarded.
The @option{-p} option inhibits the path search, and @code{hash} uses
@var{filename} as the location of @var{name}.

The @option{-r} option causes the shell to forget all remembered locations.
Assigning to the @env{PATH} variable also clears all hashed filenames.
The @option{-d} option causes the shell to forget the remembered location
of each @var{name}.

If the @option{-t} option is supplied, @code{hash} prints the full pathname
corresponding to each @var{name}.
If multiple @var{name} arguments are
supplied with @option{-t}, @code{hash} prints each @var{name}
before the corresponding hashed full path.
The @option{-l} option displays output in a format that may be reused as input.

If no arguments are given, or if only @option{-l} is supplied,
@code{hash} prints information about remembered commands.
The @option{-t}, @option{-d}, and @option{-p} options (the options that
act on the @var{name} arguments) are mutually exclusive.
Only one will be active.
If more than one is supplied, @option{-t} has higher priority than
@option{-p}, and both have higher priority than @option{-d}.

The return status is zero unless a @var{name} is not found or an invalid
option is supplied.

@item pwd
@btindex pwd
@example
pwd [-LP]
@end example

Print the absolute pathname of the current working directory.
If the @option{-P} option is supplied,
or the @option{-o physical} option to the @code{set} builtin
(@pxref{The Set Builtin}) is enabled,
the pathname printed will not
contain symbolic links.
If the @option{-L} option is supplied, the pathname printed may contain
symbolic links.
The return status is zero unless an error is encountered while
determining the name of the current directory or an invalid option
is supplied.

@item readonly
@btindex readonly
@example
readonly [-aAf] [-p] [@var{name}[=@var{value}]] @dots{}
@end example

Mark each @var{name} as readonly.
The values of these names may not be changed by subsequent assignment
or unset.
If the @option{-f} option is supplied, each @var{name} refers to a shell
function.
The @option{-a} option means each @var{name} refers to an indexed
array variable; the @option{-A} option means each @var{name} refers
to an associative array variable.
If both options are supplied, @option{-A} takes precedence.
If no @var{name} arguments are supplied, or if the @option{-p}
option is supplied, print a list of all readonly names.
The other options may be used to restrict the output to a subset of
the set of readonly names.
The @option{-p}
option displays output in a format that may be reused as input.

@code{readonly} allows the value of a variable to be set at the same time
the readonly attribute is changed by following the variable name with
=@var{value}.
This sets the value of the variable is to @var{value} while modifying
the readonly attribute.

The return status is zero unless an invalid option is supplied, one of
the @var{name} arguments is not a valid shell variable or function name,
or the @option{-f} option is supplied with a name that is not a shell function.

@item return
@btindex return
@example
return [@var{n}]
@end example

Stop executing a shell function or sourced file and return the value @var{n}
to its caller.
If @var{n} is not supplied, the return value is the exit status of the
last command executed.
If @code{return} is executed by a trap handler, the last command used to
determine the status is the last command executed before the trap handler.
If @code{return} is executed during a @code{DEBUG} trap, the last command
used to determine the status is the last command executed by the trap
handler before @code{return} was invoked.

When @code{return} is used to terminate execution of a script
being executed with the @code{.} (@code{source}) builtin, it
returns either @var{n} or
the exit status of the last command executed within the script as the exit
status of the script.
If @var{n} is supplied, the return value is its least significant
8 bits.

Any command associated with the @code{RETURN} trap is executed
before execution resumes after the function or script.

The return status is non-zero if @code{return} is supplied a non-numeric
argument or is used outside a function
and not during the execution of a script by @code{.} or @code{source}.

@item shift
@btindex shift
@example
shift [@var{n}]
@end example

Shift the positional parameters to the left by @var{n}:
the positional parameters from @var{n}+1 @dots{} @code{$#} are
renamed to @code{$1} @dots{} @code{$#}-@var{n}.
Parameters represented by the numbers @code{$#} down to @code{$#}-@var{n}+1
are unset.
@var{n} must be a non-negative number less than or equal to @code{$#}.
If @var{n} is not supplied, it is assumed to be 1.
If @var{n} is zero or greater than @code{$#}, the positional parameters
are not changed.
The return status is zero unless @var{n} is greater than @code{$#} or
less than zero, non-zero otherwise.

@item test
@itemx [
@btindex test
@btindex [
@example
test @var{expr}
@end example

Evaluate a conditional expression @var{expr} and return a status of
0 (true) or 1 (false).
Each operator and operand must be a separate argument.
Expressions are composed of the primaries described below in
@ref{Bash Conditional Expressions}.
@code{test} does not accept any options, nor does it accept and
ignore an argument of @option{--} as signifying the end of options.
When using the @code{[} form, the last argument to the command must
be a @code{]}.

Expressions may be combined using the following operators, listed in
decreasing order of precedence.
The evaluation depends on the number of arguments; see below.
@code{test} uses operator precedence when there are five or more arguments.

@table @code
@item ! @var{expr}
True if @var{expr} is false.

@item ( @var{expr} )
Returns the value of @var{expr}.
This may be used to override normal operator precedence.

@item @var{expr1} -a @var{expr2}
True if both @var{expr1} and @var{expr2} are true.

@item @var{expr1} -o @var{expr2}
True if either @var{expr1} or @var{expr2} is true.
@end table

The @code{test} and @code{[} builtins evaluate conditional
expressions using a set of rules based on the number of arguments.

@table @asis
@item 0 arguments
The expression is false.

@item 1 argument
The expression is true if, and only if, the argument is not null.

@item 2 arguments
If the first argument is @samp{!}, the expression is true if and
only if the second argument is null.
If the first argument is one of the unary conditional operators
(@pxref{Bash Conditional Expressions}), the expression
is true if the unary test is true.
If the first argument is not a valid unary operator, the expression is
false.

@item 3 arguments
The following conditions are applied in the order listed.

@enumerate
@item
If the second argument is one of the binary conditional
operators (@pxref{Bash Conditional Expressions}), the
result of the expression is the result of the binary test using the
first and third arguments as operands.
The @samp{-a} and @samp{-o} operators are considered binary operators
when there are three arguments.
@item
If the first argument is @samp{!}, the value is the negation of
the two-argument test using the second and third arguments.
@item
If the first argument is exactly @samp{(} and the third argument is
exactly @samp{)}, the result is the one-argument test of the second
argument.
@item
Otherwise, the expression is false.
@end enumerate

@item 4 arguments
The following conditions are applied in the order listed.

@enumerate
@item
If the first argument is @samp{!}, the result is the negation of
the three-argument expression composed of the remaining arguments.
@item
If the first argument is exactly @samp{(} and the fourth argument is
exactly @samp{)}, the result is the two-argument test of the second
and third arguments.
@item
Otherwise, the expression is parsed and evaluated according to
precedence using the rules listed above.
@end enumerate

@item 5 or more arguments
The expression is parsed and evaluated according to precedence
using the rules listed above.
@end table

If the shell is in @sc{posix} mode, or if the expression is part
of the @code{[[} command,
the @samp{<} and @samp{>} operators sort using the current locale.
If the shell is not in @sc{posix} mode, the @code{test} and @samp{[}
commands sort lexicographically using ASCII ordering.

The historical operator-precedence parsing with 4 or more arguments can
lead to ambiguities when it encounters strings that look like primaries.
The
@sc{posix}
standard has deprecated the @option{-a} and @option{-o}
primaries and enclosing expressions within parentheses.
Scripts should no longer use them.
It's much more reliable to restrict test invocations to a single primary,
and to replace uses of @option{-a} and @option{-o} with the shell's
@code{&&} and @code{||} list operators. For example, use

@example
test -n string1 && test -n string2
@end example

@noindent
instead of

@example
test -n string1 -a -n string2
@end example

@item times
@btindex times
@example
times
@end example

Print out the user and system times used by the shell and its children.
The return status is zero.

@item trap
@btindex trap
@example
trap [-lpP] [@var{action}] [@var{sigspec} @dots{}]
@end example

The @var{action} is a command that is read and executed when the
shell receives any of the signals @var{sigspec}.
If @var{action} is absent (and
there is a single @var{sigspec}) or
equal to @samp{-}, each specified @var{sigspec}'s disposition is reset
to the value it had when the shell was started.
If @var{action} is the null string, then the signal specified by
each @var{sigspec} is ignored by the shell and commands it invokes.

If no arguments are supplied, @code{trap} prints the actions
associated with each trapped signal
as a set of @code{trap} commands that can be reused as shell input to
restore the current signal dispositions.

If @var{action} is not present and @option{-p} has been supplied,
@code{trap} displays the trap commands associated with each @var{sigspec},
or, if no @var{sigspec}s are supplied, for all trapped signals,
as a set of @code{trap} commands that can be reused as shell input to
restore the current signal dispositions.
The @option{-P} option behaves similarly, but displays only the actions
associated with each @var{sigspec} argument.
@option{-P} requires at least one @var{sigspec} argument.
The @option{-P} or @option{-p} options may be
used in a subshell environment (e.g., command substitution) and,
as long as they are used before @code{trap} is used to change a
signal's handling, will display the state of its parent's traps.

The @option{-l} option prints a list of signal names
and their corresponding numbers.
Each @var{sigspec} is either a signal name or a signal number.
Signal names are case insensitive and the @code{SIG} prefix is optional.
If @option{-l} is supplied with no @var{sigspec} arguments, it prints a
list of valid signal names.

If a @var{sigspec}
is @code{0} or @code{EXIT}, @var{action} is executed when the shell exits.
If a @var{sigspec} is @code{DEBUG}, @var{action} is executed
before every simple command, @code{for} command, @code{case} command,
@code{select} command, (( arithmetic command, [[ conditional command,
arithmetic @code{for} command,
and before the first command executes in a shell function.
Refer to the description of the @code{extdebug} shell option
(@pxref{The Shopt Builtin}) for details of its
effect on the @code{DEBUG} trap.
If a @var{sigspec} is @code{RETURN}, @var{action} is executed
each time a shell function or a script executed with the @code{.} or
@code{source} builtins finishes executing.

If a @var{sigspec} is @code{ERR}, @var{action} 
is executed whenever
a pipeline (which may consist of a single simple
command), a list, or a compound command returns a
non-zero exit status,
subject to the following conditions.
The @code{ERR} trap is not executed if the failed command is part of the
command list immediately following an
@code{until} or @code{while} reserved word,
part of the test following the @code{if} or @code{elif} reserved words,
part of a command executed in a @code{&&} or @code{||} list
except the command following the final @code{&&} or @code{||},
any command in a pipeline but the last,
(subject to the state of the @code{pipefail} shell option),
or if the command's return
status is being inverted using @code{!}.
These are the same conditions obeyed by the @code{errexit} (@option{-e})
option.

When the shell is not interactive,
signals ignored upon entry to a non-interactive shell cannot be trapped or
reset.
Interactive shells permit trapping signals ignored on entry.
Trapped signals that are not being ignored are reset to their original
values in a subshell or subshell environment when one is created.

The return status is zero unless a @var{sigspec} does not specify a
valid signal; non-zero otherwise.

@item true
@btindex true
@example
true
@end example

Does nothing, returns a 0 status.

@item umask
@btindex umask
@example
umask [-p] [-S] [@var{mode}]
@end example

Set the shell process's file creation mask to @var{mode}.
If @var{mode} begins with a digit, it is interpreted as an octal number;
if not, it is interpreted as a symbolic mode mask similar
to that accepted by the @code{chmod} command.
If @var{mode} is omitted, @code{umask} prints the current value of the mask. 
If the @option{-S}
option is supplied without a @var{mode} argument, @code{umask}
prints the mask in a symbolic format;
the default output is an octal number.
If the  @option{-p} option is supplied, and @var{mode}
is omitted, the output is in a form that may be reused as input.
The return status is zero if the mode is successfully changed or if
no @var{mode} argument is supplied, and non-zero otherwise.

Note that when the mode is interpreted as an octal number, each number
of the umask is subtracted from @code{7}.  Thus, a umask of @code{022}
results in permissions of @code{755}.

@item unset
@btindex unset
@example
unset [-fnv] [@var{name}]
@end example

Remove each variable or function @var{name}.
If the @option{-v} option is given, each
@var{name} refers to a shell variable and that variable is removed.
If the @option{-f} option is given, the @var{name}s refer to shell
functions, and the function definition is removed.
If the @option{-n} option is supplied, and @var{name} is a variable with
the @code{nameref} attribute, @var{name} will be unset rather than the
variable it references.
@option{-n} has no effect if the @option{-f} option is supplied.
If no options are supplied, each @var{name} refers to a variable; if
there is no variable by that name, a function with that name, if any, is
unset.
Readonly variables and functions may not be unset.
When variables or functions are removed, they are also removed
from the environment passed to subsequent commands.
Some shell variables may not be unset.
Some shell variables lose their special behavior if they are unset; such
behavior is noted in the description of the individual variables.
The return status is zero unless a @var{name} is readonly or may not be unset.
@end table

@node Bash Builtins
@section Bash Builtin Commands

This section describes builtin commands which are unique to
or have been extended in Bash.
Some of these commands are specified in the @sc{posix} standard.

@table @code

@item alias
@btindex alias
@example
alias [-p] [@var{name}[=@var{value}] @dots{}]
@end example

Without arguments or with the @option{-p} option, @code{alias} prints
the list of aliases on the standard output in a form that allows
them to be reused as input.
If arguments are supplied, define an alias for each @var{name}
whose @var{value} is given.
If no @var{value} is given, print the name and value of the alias @var{name}.
A trailing space in @var{value} causes the next word to be
checked for alias substitution when the alias is expanded
during command parsing.
@code{alias} returns true unless a @var{name} is given
(without a corresponding =@var{value})
for which no alias has been defined.
Aliases are described in @ref{Aliases}.

@item bind
@btindex bind
@example
bind [-m @var{keymap}] [-lsvSVX]
bind [-m @var{keymap}] [-q @var{function}] [-u @var{function}] [-r @var{keyseq}]
bind [-m @var{keymap}] -f @var{filename}
bind [-m @var{keymap}] -x @var{keyseq[: ]shell-command}
bind [-m @var{keymap}] @var{keyseq:function-name}
bind [-m @var{keymap}] @var{keyseq:readline-command}
bind [-m @var{keymap}] -p|-P [@var{readline-command}]
bind @var{readline-command-line}
@end example

Display current Readline (@pxref{Command Line Editing})
key and function bindings,
bind a key sequence to a Readline function or macro
or to a shell command,
or set a Readline variable.
Each non-option argument is a key binding or command as it would appear in a
Readline initialization file (@pxref{Readline Init File}),
but each binding or command must be passed as a separate argument;  e.g.,
@samp{"\C-x\C-r":re-read-init-file}.

In the following descriptions, options that display output in a form
available to be re-read format their output
as commands that would appear in a Readline initialization file or
that would be supplied as individual arguments to a @code{bind} command.

Options, if supplied, have the following meanings:

@table @code
@item -m @var{keymap}
Use @var{keymap} as the keymap to be affected by
the subsequent bindings.  Acceptable @var{keymap}
names are
@code{emacs},
@code{emacs-standard},
@code{emacs-meta},
@code{emacs-ctlx},
@code{vi},
@code{vi-move},
@code{vi-command}, and
@code{vi-insert}.
@code{vi} is equivalent to @code{vi-command}
(@code{vi-move} is also a synonym);
@code{emacs} is equivalent to @code{emacs-standard}.

@item -l
List the names of all Readline functions.

@item -p
Display Readline function names and bindings
in such a way that they can be used as an argument to a subsequent
@code{bind} command or in a Readline initialization file.
If arguments remain after option processing, @code{bind} treats
them as readline command names and restricts output to those names.

@item -P
List current Readline function names and bindings.
If arguments remain after option processing, @code{bind} treats
them as readline command names and restricts output to those names.

@item -s
Display Readline key sequences bound to macros and the strings
they output in such a way that they can be used as
an argument to a subsequent @code{bind} command
or in a Readline initialization file.

@item -S
Display Readline key sequences bound to macros and the strings they output.

@item -v
Display Readline variable names and values
in such a way that they can be used as
an argument to a subsequent @code{bind} command
or in a Readline initialization file.

@item -V
List current Readline variable names and values.

@item -f @var{filename}
Read key bindings from @var{filename}.

@item -q @var{function}
Display key sequences that invoke the named Readline @var{function}.

@item -u @var{function}
Unbind all key sequences bound to the named Readline @var{function}.

@item -r @var{keyseq}
Remove any current binding for @var{keyseq}.

@item -x @var{keyseq:shell-command}
Cause @var{shell-command} to be executed whenever @var{keyseq} is
entered.
The separator between @var{keyseq} and @var{shell-command} is either
whitespace or a colon optionally followed by whitespace.
If the separator is whitespace, @var{shell-command}
must be enclosed in double quotes and Readline expands any of its
special backslash-escapes in @var{shell-command} before saving it.
If the separator is a colon, any enclosing double quotes are optional, and
Readline does not expand the command string before saving it.
Since the entire key binding expression must be a single argument, it
should be enclosed in single quotes.
When @var{shell-command} is executed, the shell sets the
@code{READLINE_LINE} variable to the contents of the Readline line
buffer
and the
@code{READLINE_POINT} and @code{READLINE_MARK}
variables to the current location of the insertion point and the saved
insertion point (the @var{mark}), respectively.
The shell assigns any numeric argument the user supplied to the
@code{READLINE_ARGUMENT} variable.
If there was no argument, that variable is not set.
If the executed command changes the value of any of
@code{READLINE_LINE}, @code{READLINE_POINT}, or @code{READLINE_MARK},
those new values will be reflected in the editing state.

@item -X
List all key sequences bound to shell commands and the associated commands
in a format that can be reused as
an argument to a subsequent @code{bind} command.
@end table

@noindent
The return status is zero unless an invalid option is supplied or an
error occurs.

@item builtin
@btindex builtin
@example
builtin [@var{shell-builtin} [@var{args}]]
@end example

Execute the specified shell builtin @var{shell-builtin},
passing it @var{args}, and return its exit status.
This is useful when defining a shell function with the same
name as a shell builtin,
retaining the functionality of the builtin within the function.
The return status is non-zero if @var{shell-builtin} is not a shell
builtin command.

@item caller
@btindex caller
@example
caller [@var{expr}]
@end example

Returns the context of any active subroutine call (a shell function or
a script executed with the @code{.} or @code{source} builtins).

Without @var{expr}, @code{caller} displays the line number and source
filename of the current subroutine call.
If a non-negative integer is supplied as @var{expr}, @code{caller}
displays the line number, subroutine name, and source file corresponding
to that position in the current execution call stack.
This extra information may be used, for example, to print a stack trace.
The current frame is frame 0.

The return value is 0 unless the shell is not executing a subroutine
call or @var{expr} does not correspond to a valid position in the
call stack.

@item command
@btindex command
@example
command [-pVv] @var{command} [@var{arguments} @dots{}]
@end example

The @code{command} builtin runs @var{command} with @var{arguments}
ignoring any shell function named @var{command}.
Only shell builtin commands or commands found by searching the
@env{PATH} are executed.
If there is a shell function named @code{ls}, running @samp{command ls}
within the function will execute the external command @code{ls}
instead of calling the function recursively.
The @option{-p} option means to use a default value for @env{PATH}
that is guaranteed to find all of the standard utilities.
The return status in this case is 127 if @var{command} cannot be
found or an error occurred, and the exit status of @var{command}
otherwise.

If either the @option{-V} or @option{-v} option is supplied, @code{command}
prints a description of @var{command}.
The @option{-v} option
displays a single word indicating the command or file name used to
invoke @var{command};
the @option{-V} option produces a more verbose description.
In this case, the return status is
zero if @var{command} is found, and non-zero if not.

@item declare
@btindex declare
@example
declare [-aAfFgiIlnrtux] [-p] [@var{name}[=@var{value}] @dots{}]
@end example

Declare variables and give them attributes.
If no @var{name}s are given, then display the values of variables or
shell functions instead. 

The @option{-p} option will display the attributes and values of each
@var{name}.
When @option{-p} is used with @var{name} arguments, additional options,
other than @option{-f} and @option{-F}, are ignored.

When @option{-p} is supplied without @var{name} arguments,
@code{declare} will display the attributes and values
of all variables having the attributes specified by the additional options.
If no other options are supplied with @option{-p}, @code{declare} will
display the attributes and values of all shell variables.
The @option{-f} option restricts the display to shell functions.

The @option{-F} option inhibits the display of function definitions;
only the function name and attributes are printed.
If the @code{extdebug} shell option is enabled using @code{shopt}
(@pxref{The Shopt Builtin}), the source file name and line number where
each @var{name} is defined are displayed as well.
@option{-F} implies @option{-f}.

The @option{-g} option forces variables to be created or modified at
the global scope, even when @code{declare} is executed in a shell function.
It is ignored in when @code{declare} is  not executed in a shell function.

The @option{-I} option causes local variables to inherit the attributes
(except the @code{nameref} attribute)
and value of any existing variable with the same
@var{name} at a surrounding scope.
If there is no existing variable, the local variable is initially unset.

The following options can be used to restrict output to variables with
the specified attributes or to give variables attributes:

@table @code
@item -a
Each @var{name} is an indexed array variable (@pxref{Arrays}).

@item -A
Each @var{name} is an associative array variable (@pxref{Arrays}).

@item -f
Each @var{name} refers to a shell function.

@item -i
The variable is to be treated as an integer;
arithmetic evaluation (@pxref{Shell Arithmetic}) is
performed when the variable is assigned a value.

@item -l
When the variable is assigned a value, all upper-case characters are
converted to lower-case.
The upper-case attribute is disabled.

@item -n
Give each @var{name} the @code{nameref} attribute, making
it a name reference to another variable.
That other variable is defined by the value of @var{name}.
All references, assignments, and attribute modifications
to @var{name}, except for those using or changing the
@option{-n} attribute itself, are performed on the variable referenced by
@var{name}'s value.
The nameref attribute cannot be applied to array variables.

@item -r
Make @var{name}s readonly.
These names cannot then be assigned values
by subsequent assignment statements or unset.

@item -t
Give each @var{name} the @code{trace} attribute.
Traced functions inherit the @code{DEBUG} and @code{RETURN} traps from
the calling shell.
The trace attribute has no special meaning for variables.

@item -u
When the variable is assigned a value, all lower-case characters are
converted to upper-case.
The lower-case attribute is disabled.

@item -x
Mark each @var{name} for export to subsequent commands via
the environment.
@end table

Using @samp{+} instead of @samp{-} turns off the specified
attribute instead, with the exceptions that @samp{+a} and @samp{+A}
may not be used to destroy array variables and @samp{+r} will not
remove the readonly attribute.

When used in a function, @code{declare} makes each @var{name} local,
as with the @code{local} command, unless the @option{-g} option is supplied.
If a variable name is followed by =@var{value}, the value of the variable
is set to @var{value}.

When using @option{-a} or @option{-A} and the compound assignment syntax to 
create array variables, additional attributes do not take effect until
subsequent assignments.

The return status is zero unless an invalid option is encountered,
an attempt is made to define a function using @samp{-f foo=bar},
an attempt is made to assign a value to a readonly variable,
an attempt is made to assign a value to an array variable without
using the compound assignment syntax (@pxref{Arrays}),
one of the @var{name}s is not a valid shell variable name,
an attempt is made to turn off readonly status for a readonly variable,
an attempt is made to turn off array status for an array variable,
or an attempt is made to display a non-existent function with @option{-f}.

@item echo
@btindex echo
@example
echo [-neE] [@var{arg} @dots{}]
@end example

Output the @var{arg}s, separated by spaces, terminated with a
newline.
The return status is 0 unless a write error occurs.
If @option{-n} is specified, the trailing newline is not printed.

If the @option{-e} option is given, @code{echo} interprets the following
backslash-escaped characters.
The @option{-E} option disables interpretation of these escape characters,
even on systems where they are interpreted by default.
The @code{xpg_echo} shell option determines
whether or not @code{echo} interprets any options and
expands these escape characters.
@code{echo} does not interpret @option{--} to mean the end of options.

@code{echo} interprets the following escape sequences:
@table @code
@item \a
alert (bell)
@item \b
backspace
@item \c
suppress further output
@item \e
@itemx \E
escape
@item \f
form feed
@item \n
new line
@item \r
carriage return
@item \t 
horizontal tab
@item \v
vertical tab
@item \\
backslash
@item \0@var{nnn}
The eight-bit character whose value is the octal value @var{nnn}
(zero to three octal digits).
@item \x@var{HH}
The eight-bit character whose value is the hexadecimal value @var{HH}
(one or two hex digits).
@item \u@var{HHHH}
The Unicode (ISO/IEC 10646) character whose value is the hexadecimal value
@var{HHHH} (one to four hex digits).
@item \U@var{HHHHHHHH}
The Unicode (ISO/IEC 10646) character whose value is the hexadecimal value
@var{HHHHHHHH} (one to eight hex digits).
@end table

@code{echo} writes any unrecognized backslash-escaped characters unchanged.

@item enable
@btindex enable
@example
enable [-a] [-dnps] [-f @var{filename}] [@var{name} @dots{}]
@end example

Enable and disable builtin shell commands.
Disabling a builtin allows an executable file which has the same name
as a shell builtin to be executed without specifying a full pathname,
even though the shell normally searches for builtins before files.

If @option{-n} is supplied, the @var{name}s are disabled.
Otherwise @var{name}s are enabled.
For example, to use the @code{test} binary
found using @env{$PATH} instead of the shell builtin version, type
@samp{enable -n test}.

If the @option{-p} option is supplied, or no @var{name} arguments are
supplied, print a list of shell builtins.
With no other arguments, the list consists of all enabled shell builtins.
The @option{-n} option means to print only disabled builtins.
The @option{-a} option means to list
each builtin with an indication of whether or not it is enabled. 
The @option{-s} option means to
restrict @code{enable} to the
@sc{posix}
special builtins.

The @option{-f} option means to load the new builtin command @var{name}
from shared object @var{filename}, on systems that support dynamic loading.
If @var{filename} does not contain a slash,
Bash will use the value of the @env{BASH_LOADABLES_PATH} variable as a
colon-separated list of directories in which to search for @var{filename}.
The default for @env{BASH_LOADABLES_PATH} is system-dependent,
and may include "." to force a search of the current directory.
The @option{-d} option will delete a builtin loaded with @option{-f}.
If @option{-s} is used with @option{-f}, the new builtin becomes a
@sc{posix}
special builtin (@pxref{Special Builtins}).

If no options are supplied and a @var{name} is not a shell builtin,
@code{enable} will attempt to load @var{name} from a shared object named
@var{name}, as if the command were
@samp{enable -f @var{name} @var{name}}.

The return status is zero unless a @var{name} is not a shell builtin
or there is an error loading a new builtin from a shared object.

@item help
@btindex help
@example
help [-dms] [@var{pattern}]
@end example

Display helpful information about builtin commands.
If @var{pattern} is specified, @code{help} gives detailed help
on all commands matching @var{pattern}
as described below;
otherwise it displays a list of
all builtins and shell compound commands.

Options, if supplied, have the following meanings:

@table @code
@item -d
Display a short description of each @var{pattern}
@item -m
Display the description of each @var{pattern} in a manpage-like format
@item -s
Display only a short usage synopsis for each @var{pattern}
@end table

If @var{pattern} contains pattern matching characters
(@pxref{Pattern Matching})
it's treated as a shell pattern and
@code{help} prints the description of each
help topic matching @var{pattern}.

If not, and @var{pattern} exactly matches the name of a help topic,
@code{help} prints the description
associated with that topic.
Otherwise, @code{help} performs prefix matching and
prints the descriptions of all matching help topics. 

The return status is zero unless no command matches @var{pattern}.

@item let
@btindex let
@example
let @var{expression} [@var{expression} @dots{}]
@end example

The @code{let} builtin allows arithmetic to be performed on shell variables. 
Each @var{expression} is evaluated as an arithmetic expression
according to the rules given below in @ref{Shell Arithmetic}.
If the last @var{expression} evaluates to 0,
@code{let} returns 1;
otherwise @code{let} returns 0.

@item local
@btindex local
@example
local [@var{option}] @var{name}[=@var{value}] @dots{}
@end example

For each argument, create a local variable named @var{name},
and assign it @var{value}.
The @var{option} can be any of the options accepted by @code{declare}.
@code{local} can only be used within a function; it makes the variable
@var{name} have a visible scope restricted to that function and its
children.
It is an error to use @code{local} when not within a function.

If @var{name} is @samp{-}, it makes the set of shell options
local to the  function in which @code{local} is invoked:
any shell options changed using the @code{set} builtin inside
the function after the call to @code{local} are restored to their
original values when the function returns.
The restore is performed as if a series of @code{set} commands were
executed to restore the values that were in place before the function.

With no operands, @code{local}
writes a list of local variables to the standard output.

The return status is zero unless @code{local} is used outside
a function, an invalid @var{name} is supplied, or @var{name} is a
readonly variable.

@item logout
@btindex logout
@example
logout [@var{n}]
@end example

Exit a login shell, returning a status of @var{n} to the shell's
parent.

@item mapfile
@btindex mapfile
@example
mapfile [-d @var{delim}] [-n @var{count}] [-O @var{origin}] [-s @var{count}]
    [-t] [-u @var{fd}] [-C @var{callback}] [-c @var{quantum}] [@var{array}]
@end example

Read lines from the standard input,
or from file descriptor @var{fd} if the @option{-u} option is supplied,
into the indexed array variable @var{array}.
The variable @code{MAPFILE} is the default @var{array}.
Options, if supplied, have the following meanings:

@table @code

@item -d
Use the first character of @var{delim} to terminate each input line,
rather than newline.
If @var{delim} is the empty string, @code{mapfile} will terminate a line
when it reads a NUL character.
@item -n
Copy at most @var{count} lines.
If @var{count} is 0, copy all lines.
@item -O
Begin assigning to @var{array} at index @var{origin}.
The default index is 0.
@item -s
Discard the first @var{count} lines read.
@item -t
Remove a trailing @var{delim} (default newline) from each line read.
@item -u
Read lines from file descriptor @var{fd} instead of the standard input.
@item -C
Evaluate @var{callback} each time @var{quantum} lines are read.
The @option{-c} option specifies @var{quantum}.
@item -c
Specify the number of lines read between each call to @var{callback}.
@end table

If @option{-C} is specified without @option{-c},
the default quantum is 5000.
When @var{callback}  is evaluated, it is supplied the index of the next
array element to be assigned and the line to be assigned to that element
as additional arguments.
@var{callback} is evaluated after the line is read but before the
array element is assigned.

If not supplied with an explicit origin, @code{mapfile} will clear @var{array}
before assigning to it.

@code{mapfile} returns zero unless an invalid option or option
argument is supplied, @var{array} is invalid or unassignable, or if
@var{array} is not an indexed array.

@item printf
@btindex printf
@example
printf [-v @var{var}] @var{format} [@var{arguments}]
@end example

Write the formatted @var{arguments} to the standard output under the
control of the @var{format}.
The @option{-v} option assigns the output to the variable
@var{var} rather than printing it to the standard output.

The @var{format} is a character string which contains three types of objects:
plain characters, which are simply copied to standard output, character
escape sequences, which are converted and copied to the standard output, and
format specifications, each of which causes printing of the next successive
@var{argument}.
In addition to the standard @code{printf(3)} format characters
@code{cCsSndiouxXeEfFgGaA},
@code{printf} interprets the following additional format specifiers:

@table @code
@item %b
Causes @code{printf} to expand backslash escape sequences in the
corresponding @var{argument} in the same way as @code{echo -e}
(@pxref{Bash Builtins}).
@item %q
Causes @code{printf} to output the
corresponding @var{argument} in a format that can be reused as shell input.
@code{%q} and @code{%Q} use the ANSI-C quoting style (@pxref{ANSI-C Quoting})
if any characters
in the argument string require it, and backslash quoting otherwise.
If the format string uses the @code{printf} @emph{alternate form}, these two
formats quote the argument string using single quotes.

@item %Q
like @code{%q}, but applies any supplied precision to the @var{argument}
before quoting it.

@item %(@var{datefmt})T
Causes @code{printf} to output the date-time string resulting from using
@var{datefmt} as a format string for @code{strftime}(3).
The corresponding @var{argument} is an integer representing the number of
seconds since the epoch.
This format specifier recognizes Two special argument values:
-1 represents the current time,
and -2 represents the time the shell was invoked.
If no argument is specified, conversion behaves as if -1 had been supplied.
This is an exception to the usual @code{printf} behavior.
@end table

@noindent
The %b, %q, and %T format specifiers all use the field width and precision
arguments from the format specification and write that many bytes from    
(or use that wide a field for) the expanded argument, which usually
contains more characters than the original.

The %n format specifier accepts a corresponding argument that is treated
as a shell variable name.

The %s and %c format specifiers accept an l (long) modifier, which forces
them to convert the argument string to a wide-character string and apply
any supplied field width and precision in terms of characters, not bytes.
The %S and %C format specifiers are equivalent to %ls and %lc, respectively.

@ignore
The %s format specifier understands the "altform" flag (#) and behaves   
identically to the %b format specifier if it's supplied.
@end ignore

Arguments to non-string format specifiers are treated as C language constants,
except that a leading plus or minus sign is allowed, and if the leading
character is a single or double quote, the value is the numeric value of
the following character, using the current locale.

Format specifiers may apply to the @var{n}th argument rather than the next
sequential argument.
In this case, the @samp{%} in the format specifier is replaced by the
sequence
@samp{%@var{n}$},
where @var{n} is a decimal integer greater than 0,
giving the argument number to use as the operand.
The format string should not mix numbered and unnumbered argument specifiers,
though this is allowed.
Unnumbered argument specifiers always refer to the next argument following
the last argument consumed by an unnumbered specifier; numbered argument
specifiers refer to absolute positions in the argument list.

The @var{format} is reused as necessary to consume all of the @var{arguments}.
If the @var{format} requires more @var{arguments} than are supplied, the
extra format specifications behave as if a zero value or null string, as
appropriate, had been supplied.
If @var{format} is reused, a numbered argument specifier
@samp{%@var{n}$}
refers to the @var{n}th argument following the highest numbered argument
consumed by the previous use of @var{format}.

The return value is zero on success,
non-zero if an invalid option is supplied or a write or assignment error
occurs.

@item read
@btindex read
@example
read [-Eers] [-a @var{aname}] [-d @var{delim}] [-i @var{text}] [-n @var{nchars}]
    [-N @var{nchars}] [-p @var{prompt}] [-t @var{timeout}] [-u @var{fd}] [@var{name} @dots{}]
@end example

Read one line from the standard input, or from the file descriptor
@var{fd} supplied as an argument to the @option{-u} option,
split it into words as described above in @ref{Word Splitting},
and assign the first word to the first @var{name},
the second word to the second @var{name}, and so on.
If there are more words than names,
the remaining words and their intervening delimiters are assigned
to the last @var{name}.
If there are fewer words read from the input stream than names,
the remaining names are assigned empty values.
The characters in the value of the @env{IFS} variable
are used to split the line into words using the same rules the shell
uses for expansion (described above in @ref{Word Splitting}).
The backslash character @samp{\} removes any special
meaning for the next character read and is used for line continuation.

Options, if supplied, have the following meanings:

@table @code
@item -a @var{aname}
The words are assigned to sequential indices of the array variable
@var{aname},
starting at 0.
All elements are removed from @var{aname} before the assignment.
Other @var{name} arguments are ignored.

@item -d @var{delim}
The first character of @var{delim} terminates the input line,
rather than newline.
If @var{delim} is the empty string, @code{read} will terminate a line
when it reads a NUL character.

@item -e
If the standard input is coming from a terminal,
@code{read} uses
Readline (@pxref{Command Line Editing})
to obtain the line.
Readline uses the current (or default, if line editing was not previously
active) editing settings, but uses Readline's default filename completion.

@item -E
If the standard input is coming from a terminal,
@code{read} uses
Readline (@pxref{Command Line Editing}) to obtain the line.
Readline uses the current (or default, if line editing was not previously
active) editing settings, but uses Bash's default completion, including
programmable completion.

@item -i @var{text}
If Readline is being used to read the line, @code{read} places @var{text} into
the editing buffer before editing begins.

@item -n @var{nchars}
@code{read} returns after reading @var{nchars} characters rather than
waiting for a complete line of input,
unless it encounters EOF or @code{read} times out,
but honors a delimiter if it reads fewer
than @var{nchars} characters before the delimiter.

@item -N @var{nchars}
@code{read} returns after reading exactly @var{nchars} characters rather
than waiting for a complete line of input,
unless it encounters EOF or @code{read} times out.
Delimiter characters in the input are
not treated specially and do not cause @code{read} to return until
it has read @var{nchars} characters.
The result is not split on the characters in @code{IFS}; the intent is
that the variable is assigned exactly the characters read
(with the exception of backslash; see the @option{-r} option below).

@item -p @var{prompt}
Display @var{prompt}, without a trailing newline, before attempting
to read any input, but only if input is coming from a terminal.

@item -r
If this option is given, backslash does not act as an escape character.
The backslash is considered to be part of the line.
In particular, a backslash-newline pair may not then be used as a line
continuation.

@item -s
Silent mode.
If input is coming from a terminal, characters are not echoed.

@item -t @var{timeout}
Cause @code{read} to time out and return failure if it does not read
a complete line of input (or a specified number of characters)
within @var{timeout} seconds.
@var{timeout}  may be a decimal number with a fractional portion following
the decimal point.
This option is only effective if @code{read} is reading input from a
terminal, pipe, or other special file; it has no effect when reading
from regular files.
If @code{read} times out, it saves any partial input read into
the specified variable @var{name}, and returns a status greater than 128.
If @var{timeout} is 0, @code{read} returns immediately, without trying to
read any data.
In this case, the exit status is 0 if input is available on the specified
file descriptor, or the read will return EOF, non-zero otherwise.

@item -u @var{fd}
Read input from file descriptor @var{fd} instead of the standard input.
@end table

Other than the case where @var{delim} is the empty string, @code{read}
ignores any NUL characters in the input.

If no @var{name}s are supplied, @code{read} assigns the line read,
without the ending delimiter but otherwise unmodified,
to the variable @env{REPLY}.

The exit status is zero, unless end-of-file is encountered, @code{read}
times out (in which case the status is greater than 128),
a variable assignment error (such as assigning to a readonly variable) occurs,
or an invalid file descriptor is supplied as the argument to @option{-u}.

@item readarray
@btindex readarray
@example
readarray [-d @var{delim}] [-n @var{count}] [-O @var{origin}] [-s @var{count}]
    [-t] [-u @var{fd}] [-C @var{callback}] [-c @var{quantum}] [@var{array}]
@end example

Read lines from the standard input into the indexed array variable @var{array},
or from file descriptor @var{fd}
if the @option{-u} option is supplied.

A synonym for @code{mapfile}.

@item source
@btindex source
@example
source [-p @var{path}] @var{filename} [@var{arguments}]
@end example

A synonym for @code{.} (@pxref{Bourne Shell Builtins}).

@item type
@btindex type
@example
type [-afptP] [@var{name} @dots{}]
@end example

Indicate how each @var{name} would be interpreted if used as a command
name.

If the @option{-t} option is used, @code{type} prints a single word
which is one of @samp{alias}, @samp{keyword}, @samp{function},
@samp{builtin}, or @samp{file},
if @var{name} is an alias, shell reserved word, shell function,
shell builtin, or executable file, respectively.
If the @var{name} is not found, @code{type} prints nothing and
returns a failure status.

If the @option{-p} option is used, @code{type} either returns the name
of the executable file that would be found by searching @code{$PATH}
for @code{name},
or nothing if @option{-t} would not return @samp{file}.

The @option{-P} option forces a path search for each @var{name}, even if
@option{-t} would not return @samp{file}.

If a @var{name} is present in the table of hashed commands,
options @option{-p} and @option{-P} print the hashed value, which is not
necessarily the file that appears first in @code{$PATH}.

If the @option{-a} option is used, @code{type} returns all of the places
that contain a command named @var{name}.
This includes aliases, reserved words, functions, and builtins,
but the path search options (@option{-p} and @option{-P}) can be supplied
to restrict the output to executable files.
If @option{-a} is supplied with @option{-p}, @code{type} does not look
in the table of hashed commands, and only performs a @code{PATH}
search for @var{name}.

If the @option{-f} option is used, @code{type} does not attempt to find
shell functions, as with the @code{command} builtin.

The return status is zero if all of the @var{name}s are found, non-zero
if any are not found.

@item typeset
@btindex typeset
@example
typeset [-afFgrxilnrtux] [-p] [@var{name}[=@var{value}] @dots{}]
@end example

The @code{typeset} command is supplied for compatibility with the Korn
shell.
It is a synonym for the @code{declare} builtin command.

@item ulimit
@btindex ulimit
@example
ulimit [-HS] -a
ulimit [-HS] [-bcdefiklmnpqrstuvxPRT] [@var{limit}]
@end example

@code{ulimit} provides control over the resources available to the
shell and to processes it starts, on systems that allow such control.
If an option is given, it is interpreted as follows:

@table @code
@item -S
Change and report the soft limit associated with a resource.

@item -H
Change and report the hard limit associated with a resource.

@item -a
Report all current limits; no limits are set.

@item -b
The maximum socket buffer size.

@item -c
The maximum size of core files created.

@item -d
The maximum size of a process's data segment.

@item -e
The maximum scheduling priority ("nice").

@item -f
The maximum size of files written by the shell and its children.

@item -i
The maximum number of pending signals.

@item -k
The maximum number of kqueues that may be allocated.

@item -l
The maximum size that may be locked into memory.

@item -m
The maximum resident set size (many systems do not honor this limit).

@item -n
The maximum number of open file descriptors (most systems do not
allow this value to be set).

@item -p
The pipe buffer size.

@item -q
The maximum number of bytes in
@sc{posix}
message queues.

@item -r
The maximum real-time scheduling priority.

@item -s
The maximum stack size.

@item -t
The maximum amount of cpu time in seconds.

@item -u
The maximum number of processes available to a single user.

@item -v
The maximum amount of virtual memory available to the shell, and, on
some systems, to its children.

@item -x
The maximum number of file locks.

@item -P
The maximum number of pseudoterminals.

@item -R
The maximum time a real-time process can run before blocking, in microseconds.

@item -T
The maximum number of threads.
@end table

If @var{limit} is supplied, and the @option{-a} option is not used,
@var{limit} is the new value of the specified resource.
The special @var{limit} values @code{hard}, @code{soft}, and
@code{unlimited} stand for the current hard limit, the current soft limit,
and no limit, respectively.
A hard limit cannot be increased by a non-root user once it is set;
a soft limit may be increased up to the value of the hard limit.
Otherwise, @code{ulimit} prints the current value of the soft limit
for the specified resource, unless the @option{-H} option is supplied.
When more than one
resource is specified, the limit name and unit, if appropriate,
are printed before the value.
When setting new limits, if neither @option{-H} nor @option{-S} is supplied,
@code{ulimit} sets both the hard and soft limits.
If no option is supplied, then @option{-f} is assumed.

Values are in 1024-byte increments, except for
@option{-t}, which is in seconds;
@option{-R}, which is in microseconds;
@option{-p}, which is in units of 512-byte blocks;
@option{-P},
@option{-T},
@option{-b},
@option{-k},
@option{-n} and @option{-u}, which are unscaled values;
and, when in @sc{posix} mode (@pxref{Bash POSIX Mode}),
@option{-c} and @option{-f}, which are in 512-byte increments.

The return status is zero unless an invalid option or argument is supplied,
or an error occurs while setting a new limit.

@item unalias
@btindex unalias
@example
unalias [-a] [@var{name} @dots{} ]
@end example

Remove each @var{name} from the list of aliases.
If @option{-a} is supplied, remove all aliases.
The return value is true unless a supplied @var{name} is not a defined alias.
Aliases are described in @ref{Aliases}.
@end table

@node Modifying Shell Behavior
@section Modifying Shell Behavior
 
@menu
* The Set Builtin::		Change the values of shell attributes and
				positional parameters.
* The Shopt Builtin::		Modify shell optional behavior.
@end menu

@node The Set Builtin
@subsection The Set Builtin

This builtin is so complicated that it deserves its own section.  @code{set}
allows you to change the values of shell options and set the positional
parameters, or to display the names and values of shell variables.

@table @code
@item set
@btindex set
@example
set [-abefhkmnptuvxBCEHPT] [-o @var{option-name}] [--] [-] [@var{argument} @dots{}]
set [+abefhkmnptuvxBCEHPT] [+o @var{option-name}] [--] [-] [@var{argument} @dots{}]
set -o
set +o
@end example

If no options or arguments are supplied, @code{set} displays the names
and values of all shell variables and functions, sorted according to the
current locale, in a format that may be reused as input
for setting or resetting the currently-set variables.
Read-only variables cannot be reset.
In @sc{posix} mode, only shell variables are listed.

When options are supplied, they set or unset shell attributes.
Any arguments remaining after option processing replace the
positional parameters.

Options, if specified, have the following meanings:

@table @code
@item -a
Each variable or function that is created or modified is given the
export attribute and marked for export to the environment of
subsequent commands.

@item -b
Cause the status of terminated background jobs to be reported
immediately, rather than before printing the next primary prompt
or, under some circumstances,  when a foreground command exits.
This is effective only when job control is enabled.

@item -e
Exit immediately if
a pipeline (@pxref{Pipelines}), which may consist of a single simple command
(@pxref{Simple Commands}),
a list (@pxref{Lists}),
or a compound command (@pxref{Compound Commands})
returns a non-zero status.
The shell does not exit if the command that fails is part of the
command list immediately following a
@code{while} or @code{until} reserved word,
part of the test in an @code{if} statement,
part of any command executed in a @code{&&} or @code{||} list except
the command following the final @code{&&} or @code{||},
any command in a pipeline but the last
(subject to the state of the @code{pipefail} shell option),
or if the command's return status is being inverted with @code{!}.
If a compound command other than a subshell
returns a non-zero status because a command failed
while @option{-e} was being ignored, the shell does not exit.
A trap on @code{ERR}, if set, is executed before the shell exits.

This option applies to the shell environment and each subshell environment
separately (@pxref{Command Execution Environment}), and may cause
subshells to exit before executing all the commands in the subshell.

If a compound command or shell function executes in a context where
@option{-e} is being ignored,
none of the commands executed within the compound command or function body
will be affected by the @option{-e} setting, even if @option{-e} is set
and a command returns a failure status.
If a compound command or shell function sets @option{-e} while executing in
a context where @option{-e} is ignored, that setting will not have any
effect until the compound command or the command containing the function
call completes.

@item -f
Disable filename expansion (globbing).

@item -h
Locate and remember (hash) commands as they are looked up for execution.
This option is enabled by default.

@item -k
All arguments in the form of assignment statements are placed
in the environment for a command, not just those that precede
the command name.

@item -m
Job control is enabled (@pxref{Job Control}).
All processes run in a separate process group.
When a background job completes, the shell prints a line
containing its exit status.

@item -n
Read commands but do not execute them.
This may be used to check a script for syntax errors.
This option is ignored by interactive shells.

@item -o @var{option-name}

Set the option corresponding to @var{option-name}.
If @option{-o} is supplied with no @var{option-name},
@code{set} prints the current shell options settings.
If @option{+o} is supplied with no @var{option-name},
@code{set} prints a series of
@code{set}
commands to recreate the current option settings
on the standard output.
Valid option names are:

@table @code
@item allexport
Same as @code{-a}.

@item braceexpand
Same as @code{-B}.

@item emacs
Use an @code{emacs}-style line editing interface (@pxref{Command Line Editing}).
This also affects the editing interface used for @code{read -e}.

@item errexit
Same as @code{-e}.

@item errtrace
Same as @code{-E}.

@item functrace
Same as @code{-T}.

@item hashall
Same as @code{-h}.

@item histexpand
Same as @code{-H}.

@item history
Enable command history, as described in @ref{Bash History Facilities}.
This option is on by default in interactive shells.

@item ignoreeof
An interactive shell will not exit upon reading EOF.

@item keyword
Same as @code{-k}.

@item monitor
Same as @code{-m}.

@item noclobber
Same as @code{-C}.

@item noexec
Same as @code{-n}.

@item noglob
Same as @code{-f}.

@item nolog
Currently ignored.

@item notify
Same as @code{-b}.

@item nounset
Same as @code{-u}.

@item onecmd
Same as @code{-t}.

@item physical
Same as @code{-P}.

@item pipefail
If set, the return value of a pipeline is the value of the last
(rightmost) command to exit with a non-zero status, or zero if all
commands in the pipeline exit successfully.
This option is disabled by default.

@item posix
Enable @sc{posix} mode;
change the behavior of Bash where the default operation differs
from the @sc{posix} standard to match the standard
(@pxref{Bash POSIX Mode}).
This is intended to make Bash behave as a strict superset of that
standard.

@item privileged
Same as @code{-p}.

@item verbose
Same as @code{-v}.

@item vi
Use a @code{vi}-style line editing interface.
This also affects the editing interface used for @code{read -e}.

@item xtrace
Same as @code{-x}.
@end table

@item -p
Turn on privileged mode.
In this mode, the @env{$BASH_ENV} and @env{$ENV} files are not
processed, shell functions are not inherited from the environment,
and the @env{SHELLOPTS}, @env{BASHOPTS}, @env{CDPATH} and @env{GLOBIGNORE}
variables, if they appear in the environment, are ignored.
If the shell is started with the effective user (group) id not equal to the
real user (group) id, and the @option{-p} option is not supplied, these actions
are taken and the effective user id is set to the real user id.
If the @option{-p} option is supplied at startup, the effective user id is
not reset.
Turning this option off causes the effective user
and group ids to be set to the real user and group ids.

@item -r
Enable restricted shell mode (@pxref{The Restricted Shell}).
This option cannot be unset once it has been set.

@item -t
Exit after reading and executing one command.

@item -u
Treat unset variables and parameters other than the special parameters
@samp{@@} or @samp{*},
or array variables subscripted with @samp{@@} or @samp{*},
as an error when performing parameter expansion.
An error message will be written to the standard error, and a non-interactive
shell will exit.

@item -v
Print shell input lines to standard error as they are read.

@item -x
Print a trace of simple commands, @code{for} commands, @code{case}
commands, @code{select} commands, and arithmetic @code{for} commands
and their arguments or associated word lists to the standard error
after they are expanded and before they are executed.
The shell prints the expanded value of the @env{PS4} variable before
the command and its expanded arguments.

@item -B
The shell will perform brace expansion (@pxref{Brace Expansion}).
This option is on by default.

@item -C
Prevent output redirection using @samp{>}, @samp{>&}, and @samp{<>}
from overwriting existing files.
Using the redirection operator @samp{>|} instead of @samp{>}
will override this and force the creation of an output file.

@item -E
If set, any trap on @code{ERR} is inherited by shell functions, command
substitutions, and commands executed in a subshell environment.
The @code{ERR} trap is normally not inherited in such cases.

@item -H
Enable @samp{!} style history substitution (@pxref{History Interaction}).
This option is on by default for interactive shells.

@item -P
If set, Bash does not resolve symbolic links when executing commands
such as @code{cd} which change the current directory.
It uses the physical directory structure instead.
By default, Bash follows
the logical chain of directories when performing commands
which change the current directory.

For example, if @file{/usr/sys} is a symbolic link to @file{/usr/local/sys}
then:
@example
$ cd /usr/sys; echo $PWD
/usr/sys
$ cd ..; pwd
/usr
@end example

@noindent
If @code{set -P} is on, then:
@example
$ cd /usr/sys; echo $PWD
/usr/local/sys
$ cd ..; pwd
/usr/local
@end example

@item -T
If set, any traps on @code{DEBUG} and @code{RETURN} are inherited by
shell functions, command substitutions, and commands executed
in a subshell environment.
The @code{DEBUG} and @code{RETURN} traps are normally not inherited
in such cases.

@item --
If no arguments follow this option, unset the positional parameters.
Otherwise, the positional parameters are set to the
@var{arguments}, even if some of them begin with a @samp{-}.

@item -
Signal the end of options, and assign all remaining @var{arguments}
to the positional parameters.
The @option{-x} and @option{-v} options are turned off.
If there are no arguments, the positional parameters remain unchanged.
@end table

Using @samp{+} rather than @samp{-} causes these options to be
turned off.
The options can also be used upon invocation of the shell.
The current set of options may be found in @code{$-}.

The remaining N @var{arguments} are positional parameters and are
assigned, in order, to @code{$1}, @code{$2}, @dots{}  @code{$N}.
The special parameter @code{#} is set to N.

The return status is always zero unless an invalid option is supplied.
@end table

@node The Shopt Builtin
@subsection The Shopt Builtin

This builtin allows you to change additional optional shell behavior.

@table @code

@item shopt
@btindex shopt
@example
shopt [-pqsu] [-o] [@var{optname} @dots{}]
@end example

Toggle the values of settings controlling optional shell behavior.
The settings can be either those listed below, or, if the
@option{-o} option is used, those available with the @option{-o}
option to the @code{set} builtin command (@pxref{The Set Builtin}).

With no options, or with the @option{-p} option, display a list of all
settable options, with an indication of whether or not each is set;
if any @var{optname}s are supplied, the output is restricted to those options.
The @option{-p} option displays output in a form that
may be reused as input.

Other options have the following meanings:

@table @code
@item -s
Enable (set) each @var{optname}.

@item -u
Disable (unset) each @var{optname}.

@item -q
Suppresses normal output; the return status
indicates whether the @var{optname} is set or unset.
If multiple @var{optname} arguments are supplied with @option{-q},
the return status is zero if all @var{optname}s are enabled;
non-zero otherwise.

@item -o
Restricts the values of
@var{optname} to be those defined for the @option{-o} option to the
@code{set} builtin (@pxref{The Set Builtin}).
@end table

If either @option{-s} or @option{-u}
is used with no @var{optname} arguments, @code{shopt} shows only
those options which are set or unset, respectively.

Unless otherwise noted, the @code{shopt} options are disabled (off)
by default.

The return status when listing options is zero if all @var{optname}s
are enabled, non-zero otherwise.  When setting or unsetting options,
the return status is zero unless an @var{optname} is not a valid shell
option.

The list of @code{shopt} options is:
@table @code

@item array_expand_once
If set, the shell suppresses multiple evaluation of
associative and indexed array subscripts
during arithmetic expression evaluation, while executing
builtins that can perform variable assignments,
and while executing builtins that perform array dereferencing.

@item assoc_expand_once
Deprecated; a synonym for @code{array_expand_once}. 

@item autocd
If set, a command name that is the name of a directory is executed as if
it were the argument to the @code{cd} command.
This option is only used by interactive shells.

@item bash_source_fullpath
If set, filenames added to the @code{BASH_SOURCE} array variable are
converted to full pathnames (@pxref{Bash Variables}).

@item cdable_vars
If this is set, an argument to the @code{cd} builtin command that
is not a directory is assumed to be the name of a variable whose
value is the directory to change to.

@item cdspell
If set, the @code{cd} command
attempts to correct
minor errors in the spelling of a directory component.
Minor errors include transposed characters,
a missing character, and one extra character.
If @code{cd} corrects the directory name, it prints the corrected filename,
and the command proceeds.
This option is only used by interactive shells.

@item checkhash
If this is set, Bash checks that a command found in the hash
table exists before trying to execute it.
If a hashed command no longer exists, Bash performs a normal path search.

@item checkjobs
If set, Bash lists the status of any stopped and running jobs before
exiting an interactive shell.
If any jobs are running, Bash defers
the exit until a second exit is attempted without an
intervening command (@pxref{Job Control}).
The shell always postpones exiting if any jobs are stopped.

@item checkwinsize
If set, Bash checks the window size after each external (non-builtin)
command and, if necessary, updates the values of    
@env{LINES} and @env{COLUMNS},
using the file descriptor associated with stderr if it is a terminal.
This option is enabled by default.

@item cmdhist
If set, Bash
attempts to save all lines of a multiple-line
command in the same history entry.
This allows easy re-editing of multi-line commands.
This option is enabled by default, but only has an effect if command
history is enabled (@pxref{Bash History Facilities}).

@item compat31
@itemx compat32
@itemx compat40
@itemx compat41
@itemx compat42
@itemx compat43
@itemx compat44
These control aspects of the shell's compatibility mode
(@pxref{Shell Compatibility Mode}).

@item complete_fullquote
If set, Bash
quotes all shell metacharacters in filenames and directory names when
performing completion.
If not set, Bash
removes metacharacters such as the dollar sign from the set of
characters that will be quoted in completed filenames
when these metacharacters appear in shell variable references in words to be
completed.
This means that dollar signs in variable names that expand to directories
will not be quoted;
however, any dollar signs appearing in filenames will not be quoted, either.
This is active only when Bash is using backslashes to quote completed
filenames.
This variable is set by default, which is the default Bash behavior in
versions through 4.2.

@item direxpand
If set, Bash
replaces directory names with the results of word expansion when performing
filename completion.
This changes the contents of the Readline editing buffer.
If not set, Bash attempts to preserve what the user typed.

@item dirspell
If set, Bash
attempts spelling correction on directory names during word completion 
if the directory name initially supplied does not exist.

@item dotglob
If set, Bash includes filenames beginning with a @samp{.} in
the results of filename expansion.
The filenames @file{.} and @file{..} must always be matched explicitly,
even if @code{dotglob} is set.

@item execfail
If this is set, a non-interactive shell will not exit if
it cannot execute the file specified as an argument to the @code{exec}
builtin.
An interactive shell does not exit if @code{exec} fails.

@item expand_aliases
If set, aliases are expanded as described below under Aliases,
@ref{Aliases}.
This option is enabled by default for interactive shells.

@item extdebug
If set at shell invocation, or in a shell startup file,
arrange to execute the debugger profile
before the shell starts, identical to the @option{--debugger} option.
If set after invocation, behavior intended for use by debuggers is enabled:

@enumerate
@item
The @option{-F} option to the @code{declare} builtin (@pxref{Bash Builtins})
displays the source file name and line number corresponding to each function
name supplied as an argument.

@item
If the command run by the @code{DEBUG} trap returns a non-zero value, the
next command is skipped and not executed.

@item
If the command run by the @code{DEBUG} trap returns a value of 2, and the
shell is executing in a subroutine (a shell function or a shell script
executed by the @code{.} or @code{source} builtins), the shell simulates
a call to @code{return}.

@item
@code{BASH_ARGC} and @code{BASH_ARGV} are updated as described in their
descriptions (@pxref{Bash Variables}).

@item
Function tracing is enabled: command substitution, shell functions, and
subshells invoked with @code{( @var{command} )} inherit the
@code{DEBUG} and @code{RETURN} traps.

@item
Error tracing is enabled: command substitution, shell functions, and
subshells invoked with @code{( @var{command} )} inherit the
@code{ERR} trap.
@end enumerate

@item extglob
If set, enable the extended pattern matching features described above
(@pxref{Pattern Matching}).

@item extquote
If set, @code{$'@var{string}'} and @code{$"@var{string}"} quoting is  
performed within @code{$@{@var{parameter}@}} expansions                     
enclosed in double quotes.
This option is enabled by default. 

@item failglob
If set, patterns which fail to match filenames during filename expansion
result in an expansion error.

@item force_fignore
If set, the suffixes specified by the @env{FIGNORE} shell variable
cause words to be ignored when performing word completion even if
the ignored words are the only possible completions.
@xref{Bash Variables}, for a description of @env{FIGNORE}.
This option is enabled by default.

@item globasciiranges
If set, range expressions used in pattern matching bracket expressions
(@pxref{Pattern Matching})
behave as if in the traditional C locale when performing
comparisons.
That is, pattern matching does not take
the current locale's collating sequence into account,
so @samp{b} will not collate between @samp{A} and @samp{B},
and upper-case and lower-case ASCII characters will collate together.   

@item globskipdots
If set, filename expansion will never match the filenames
@file{.} and @file{..}, even if the pattern begins with a @samp{.}.
This option is enabled by default.

@item globstar
If set, the pattern @samp{**} used in a filename expansion context will
match all files and zero or more directories and subdirectories.
If the pattern is followed by a @samp{/}, only directories and
subdirectories match.

@item gnu_errfmt
If set, shell error messages are written in the standard @sc{gnu} error
message format.

@item histappend
If set, the history list is appended to the file named by the value
of the @env{HISTFILE}
variable when the shell exits, rather than overwriting the file.

@item histreedit
If set, and Readline is being used,
the user is given the opportunity to re-edit a failed history substitution.

@item histverify
If set, and Readline
is being used, the results of history substitution are not immediately
passed to the shell parser.
Instead, the resulting line is loaded into the Readline editing buffer,
allowing further modification.

@item hostcomplete
If set, and Readline is being used, Bash will attempt to perform
hostname completion when a word containing a @samp{@@} is being
completed (@pxref{Commands For Completion}).
This option is enabled by default.

@item huponexit
If set, Bash will send @code{SIGHUP} to all jobs when an interactive
login shell exits (@pxref{Signals}).

@item inherit_errexit
If set, command substitution inherits the value of the @code{errexit} option,
instead of unsetting it in the subshell environment.
This option is enabled when @sc{posix} mode is enabled.

@item interactive_comments
In an interactive shell, a word beginning with @samp{#}
causes that word and all remaining characters on that
line to be ignored, as in a non-interactive shell.
This option is enabled by default.

@item lastpipe
If set, and job control is not active, the shell runs the last command of
a pipeline not executed in the background in the current shell environment.

@item lithist
If enabled, and the @code{cmdhist}
option is enabled, multi-line commands are saved to the history with
embedded newlines rather than using semicolon separators where possible.

@item localvar_inherit
If set, local variables inherit the value and attributes of a variable of
the same name that exists at a previous scope before any new value is
assigned.  The @code{nameref} attribute is not inherited.

@item localvar_unset
If set, calling @code{unset} on local variables in previous function scopes
marks them so subsequent lookups find them unset until that function
returns.
This is identical to the behavior of unsetting local variables at the
current function scope.

@item login_shell
The shell sets this option if it is started as a login shell
(@pxref{Invoking Bash}).
The value may not be changed.

@item mailwarn
If set, and a file that Bash is checking for mail has been
accessed since the last time it was checked, Bash displays the message
@code{"The mail in @var{mailfile} has been read"}.

@item no_empty_cmd_completion
If set, and Readline is being used, Bash does not search
the @env{PATH}
for possible completions when completion is attempted on an empty line.

@item nocaseglob
If set, Bash matches filenames in a case-insensitive fashion when
performing filename expansion.

@item nocasematch
If set, Bash matches patterns in a case-insensitive fashion when
performing matching while executing @code{case} or @code{[[}
conditional commands (@pxref{Conditional Constructs}),
when performing pattern substitution word expansions,
or when filtering possible completions as part of programmable completion.

@item noexpand_translation
If set, Bash
encloses the translated results of $"@dots{}" quoting in single quotes
instead of double quotes.
If the string is not translated, this has no effect.

@item nullglob
If set, filename expansion patterns which match no files
(@pxref{Filename Expansion})
expand to nothing and are removed, rather than expanding to themselves.

@item patsub_replacement
If set, Bash
expands occurrences of @samp{&} in the replacement string of pattern
substitution to the text matched by the pattern, as described
above (@pxref{Shell Parameter Expansion}).
This option is enabled by default.

@item progcomp
If set, enable the programmable completion facilities
(@pxref{Programmable Completion}).
This option is enabled by default.

@item progcomp_alias
If set, and programmable completion is enabled, Bash treats a command
name that doesn't have any completions as a possible alias and attempts
alias expansion.
If it has an alias, Bash attempts programmable
completion using the command word resulting from the expanded alias.

@item promptvars
If set, prompt strings undergo
parameter expansion, command substitution, arithmetic
expansion, and quote removal after being expanded
as described below (@pxref{Controlling the Prompt}).
This option is enabled by default.

@item restricted_shell
The shell sets this option if it is started in restricted mode
(@pxref{The Restricted Shell}).
The value may not be changed.
This is not reset when the startup files are executed, allowing
the startup files to discover whether or not a shell is restricted.

@item shift_verbose
If this is set, the @code{shift}
builtin prints an error message when the shift count exceeds the
number of positional parameters.

@item sourcepath
If set, the @code{.} (@code{source}) builtin uses the value of @env{PATH}
to find the directory containing the file supplied as an argument
when the @option{-p} option is not supplied.
This option is enabled by default.

@item varredir_close
If set, the shell automatically closes file descriptors assigned using the
@code{@{varname@}} redirection syntax (@pxref{Redirections}) instead of
leaving them open when the command completes.

@item xpg_echo
If set, the @code{echo} builtin expands backslash-escape sequences
by default.
If the @code{posix} shell option (@pxref{The Set Builtin}) is also enabled,
@code{echo} does not interpret any options.

@end table
@end table

@node Special Builtins
@section Special Builtins
@cindex special builtin

For historical reasons, the @sc{posix} standard has classified
several builtin commands as @emph{special}.
When Bash is executing in @sc{posix} mode, the special builtins
differ from other builtin commands in three respects:

@enumerate
@item
Special builtins are found before shell functions during command lookup.

@item
If a special builtin returns an error status, a non-interactive shell exits.

@item
Assignment statements preceding the command stay in effect in the shell
environment after the command completes.
@end enumerate

When Bash is not executing in @sc{posix} mode, these builtins behave no
differently than the rest of the Bash builtin commands.
The Bash @sc{posix} mode is described in @ref{Bash POSIX Mode}. 

These are the @sc{posix} special builtins:
@example
@w{break : . source continue eval exec exit export readonly return set}
@w{shift times trap unset}
@end example

@node Shell Variables
@chapter Shell Variables

@menu
* Bourne Shell Variables::	Variables which Bash uses in the same way
				as the Bourne Shell.
* Bash Variables::		List of variables that exist in Bash.
@end menu

This chapter describes the shell variables that Bash uses.
Bash automatically assigns default values to a number of variables.

@node Bourne Shell Variables
@section Bourne Shell Variables

Bash uses certain shell variables in the same way as the Bourne shell.
In some cases, Bash assigns a default value to the variable.

@vtable @code

@item CDPATH
A colon-separated list of directories used as a search path for
the @code{cd} builtin command.

@item HOME
The current user's home directory; the default for the @code{cd} builtin
command.
The value of this variable is also used by tilde expansion
(@pxref{Tilde Expansion}).

@item IFS
A list of characters that separate fields; used when the shell splits
words as part of expansion and by the @code{read} builtin to split
lines into words.
@xref{Word Splitting}, for a description of word splitting.

@item MAIL
If the value is set to a filename or directory name
and the @env{MAILPATH} variable
is not set, Bash informs the user of the arrival of mail in
the specified file or Maildir-format directory.

@item MAILPATH
A colon-separated list of filenames which the shell periodically checks
for new mail.
Each list entry can specify the message that is printed when new mail
arrives in the mail file by separating the filename from the message with
a @samp{?}.
When used in the text of the message, @code{$_} expands to the name of
the current mail file.

@item OPTARG
The value of the last option argument processed by the @code{getopts} builtin.

@item OPTIND
The index of the next argument to be processed by the @code{getopts} builtin.

@item PATH
A colon-separated list of directories in which the shell looks for
commands.
A zero-length (null) directory name in the value of @code{PATH} indicates the
current directory.
A null directory name may appear as two adjacent colons, or as an initial
or trailing colon.
The default path is system-dependent, and is set by the administrator
who installs @code{bash}.
A common value is
"/usr/local/bin:/usr/local/sbin:/usr/bin:/usr/sbin:/bin:/sbin".

@item PS1
The primary prompt string.
The default value is @samp{\s-\v\$ }.
@xref{Controlling the Prompt}, for the complete list of escape
sequences that are expanded before @env{PS1} is displayed.

@item PS2
The secondary prompt string.
The default value is @samp{> }.
@env{PS2} is expanded in the same way as @env{PS1} before being
displayed.

@end vtable

@node Bash Variables
@section Bash Variables

These variables are set or used by Bash, but other shells
do not normally treat them specially.

A few variables used by Bash are described in different chapters:
variables for controlling the job control facilities
(@pxref{Job Control Variables}).

@vtable @code

@item _
@vindex $_
($_, an underscore.)
This has a number of meanings depending on context.
At shell startup, $_ set to the pathname used to invoke the
shell or shell script being executed as passed in the environment
or argument list.
Subsequently, it expands to the last argument to the previous simple
command executed in the foreground, after expansion. 
It is also set to the full pathname used to invoke each command executed
and placed in the environment exported to that command.
When checking mail, $_ expands to the name of the mail file.

@item BASH
The full pathname used to execute the current instance of Bash.

@item BASHOPTS
A colon-separated list of enabled shell options.
Each word in the list is a valid argument for the @option{-s} option
to the @code{shopt} builtin command (@pxref{The Shopt Builtin}).
The options appearing in @env{BASHOPTS} are those reported
as @samp{on} by @samp{shopt}.
If this variable is in the environment when Bash
starts up, the shell enables each option in the list before
reading any startup files.
If this variable is exported, child shells will enable each option
in the list.
This variable is readonly.

@item BASHPID
Expands to the process @sc{id} of the current Bash process.
This differs from @code{$$} under certain circumstances, such as subshells
that do not require Bash to be re-initialized.
Assignments to @env{BASHPID} have no effect.
If @env{BASHPID}
is unset, it loses its special properties, even if it is
subsequently reset.

@item BASH_ALIASES
An associative array variable whose members correspond to the internal
list of aliases as maintained by the @code{alias} builtin.
(@pxref{Bourne Shell Builtins}).
Elements added to this array appear in the alias list; however,
unsetting array elements currently does not cause aliases to be removed
from the alias list.
If @env{BASH_ALIASES}
is unset, it loses its special properties, even if it is
subsequently reset.

@item BASH_ARGC
An array variable whose values are the number of parameters in each
frame of the current Bash execution call stack.
The number of parameters to the current subroutine (shell function or
script executed with @code{.} or @code{source}) is at the top of the stack.
When a
subroutine is executed, the number of parameters passed is pushed onto
@code{BASH_ARGC}.
The shell sets @code{BASH_ARGC} only when in extended debugging mode
(see @ref{The Shopt Builtin}
for a description of the @code{extdebug} option to the @code{shopt}
builtin).
Setting @code{extdebug} after the shell has started to execute a subroutine,
or referencing this variable when @code{extdebug} is not set,
may result in inconsistent values.
Assignments to @env{BASH_ARGC} have no effect, and it may not be unset.

@item BASH_ARGV
An array variable containing all of the parameters in the current Bash
execution call stack.
The final parameter of the last subroutine call is at the top of the stack;
the first parameter of the initial call is at the bottom.
When a subroutine is executed, the shell pushes the supplied parameters
onto @code{BASH_ARGV}.
The shell sets @code{BASH_ARGV} only when in extended debugging mode
(see @ref{The Shopt Builtin}
for a description of the @code{extdebug} option to the @code{shopt}
builtin).
Setting @code{extdebug} after the shell has started to execute a script,
or referencing this variable when @code{extdebug} is not set,
may result in inconsistent values.
Assignments to @env{BASH_ARGV} have no effect, and it may not be unset.

@item BASH_ARGV0
When referenced, this variable expands to the name of the shell or shell
script (identical to @code{$0}; @xref{Special Parameters},
for the description of special parameter 0).
Assigning a value to
@code{BASH_ARGV0}
sets @code{$0} to the same value.
If
@env{BASH_ARGV0}
is unset, it loses its special properties, even if it is
subsequently reset.

@item BASH_CMDS
An associative array variable whose members correspond to the internal
hash table of commands as maintained by the @code{hash} builtin
(@pxref{Bourne Shell Builtins}).
Adding elements to this array makes them appear in the hash table; however,
unsetting array elements currently does not remove command names
from the hash table.
If @env{BASH_CMDS}
is unset, it loses its special properties, even if it is
subsequently reset.

@item BASH_COMMAND
Expands to the command currently being executed or about to be executed,
unless the shell is executing a command as the result of a trap,
in which case it is the command executing at the time of the trap.
If @env{BASH_COMMAND}
is unset, it loses its special properties, even if it is
subsequently reset.

@item BASH_COMPAT
The value is used to set the shell's compatibility level.
@xref{Shell Compatibility Mode}, for a description of the various
compatibility levels and their effects.
The value may be a decimal number (e.g., 4.2) or an integer (e.g., 42)
corresponding to the desired compatibility level.
If @env{BASH_COMPAT} is unset or set to the empty string, the compatibility
level is set to the default for the current version.
If @env{BASH_COMPAT} is set to a value that is not one of the valid
compatibility levels, the shell prints an error message and sets the
compatibility level to the default for the current version.
A subset of the valid values correspond to the compatibility levels
described below (@pxref{Shell Compatibility Mode}).
For example, 4.2 and 42 are valid values that correspond
to the @code{compat42} @code{shopt} option
and set the compatibility level to 42.
The current version is also a valid value.

@item BASH_ENV
If this variable is set when Bash is invoked to execute a shell
script, its value is expanded and used as the name of a startup file
to read before executing the script.
Bash does not use @env{PATH} to search for the resultant filename.
@xref{Bash Startup Files}.

@item BASH_EXECUTION_STRING
The command argument to the @option{-c} invocation option.

@item BASH_LINENO
An array variable whose members are the line numbers in source files
where each corresponding member of @env{FUNCNAME} was invoked.
@code{$@{BASH_LINENO[$i]@}} is the line number in the source file
(@code{$@{BASH_SOURCE[$i+1]@}}) where
@code{$@{FUNCNAME[$i]@}} was called (or @code{$@{BASH_LINENO[$i-1]@}} if
referenced within another shell function). 
Use @code{LINENO} to obtain the current line number.
Assignments to @env{BASH_LINENO} have no effect, and it may not be unset.

@item BASH_LOADABLES_PATH
A colon-separated list of directories in which the @code{enable}
command looks for dynamically loadable builtins.

@item BASH_MONOSECONDS
Each time this variable is referenced, it expands to the value returned
by the system's monotonic clock, if one is available.
If there is no monotonic clock, this is equivalent to @env{EPOCHSECONDS}.
If @env{BASH_MONOSECONDS}
is unset, it loses its special properties, even if it is
subsequently reset.

@item BASH_REMATCH
An array variable whose members are assigned by the @samp{=~} binary
operator to the @code{[[} conditional command
(@pxref{Conditional Constructs}).
The element with index 0 is the portion of the string
matching the entire regular expression.
The element with index @var{n} is the portion of the
string matching the @var{n}th parenthesized subexpression.

@item BASH_SOURCE
An array variable whose members are the source filenames where the
corresponding shell function names in the @code{FUNCNAME} array
variable are defined.
The shell function @code{$@{FUNCNAME[$i]@}} is defined in the file
@code{$@{BASH_SOURCE[$i]@}} and called from @code{$@{BASH_SOURCE[$i+1]@}}
Assignments to @env{BASH_SOURCE} have no effect, and it may not be unset.

@item BASH_SUBSHELL
Incremented by one within each subshell or subshell environment when
the shell begins executing in that environment.
The initial value is 0.
If @env{BASH_SUBSHELL}
is unset, it loses its special properties, even if it is
subsequently reset.

@item BASH_TRAPSIG
Set to the signal number corresponding to the trap action being executed
during its execution.
See the description of @code{trap}
(@pxref{Bourne Shell Builtins})
for information about signal numbers and trap execution.

@item BASH_VERSINFO
A readonly array variable (@pxref{Arrays})
whose members hold version information for this instance of Bash.
The values assigned to the array members are as follows:

@table @code

@item BASH_VERSINFO[0]
The major version number (the @dfn{release}).

@item BASH_VERSINFO[1]
The minor version number (the @dfn{version}).

@item BASH_VERSINFO[2]
The patch level.

@item BASH_VERSINFO[3]
The build version.

@item BASH_VERSINFO[4]
The release status (e.g., @code{beta}).

@item BASH_VERSINFO[5]
The value of @env{MACHTYPE}.
@end table

@item BASH_VERSION
Expands to a string describing the version of this  instance  of
Bash (e.g., 5.2.37(3)-release).

@item BASH_XTRACEFD
If set to an integer corresponding to a valid file descriptor,
Bash writes the trace output generated when
@samp{set -x}
is enabled to that file descriptor,
instead of the standard error.
This allows tracing output to be separated from diagnostic and error
messages.
The file descriptor is closed when
@code{BASH_XTRACEFD}
is unset or assigned a new value.
Unsetting
@code{BASH_XTRACEFD}
or assigning it the empty string causes the
trace output to be sent to the standard error.
Note that setting
@code{BASH_XTRACEFD}
to 2 (the standard error file
descriptor) and then unsetting it will result in the standard error
being closed.

@item CHILD_MAX
Set the number of exited child status values for the shell to remember.
Bash will not allow this value to be decreased below a
@sc{posix}-mandated minimum,
and there is a maximum value (currently 8192) that this may not exceed.
The minimum value is system-dependent.

@item COLUMNS
Used by the @code{select} command to determine the terminal width
when printing selection lists.
Automatically set if the @code{checkwinsize} option is enabled
(@pxref{The Shopt Builtin}), or in an interactive shell upon receipt of a
@code{SIGWINCH}.

@item COMP_CWORD
An index into @env{$@{COMP_WORDS@}} of the word containing the current
cursor position.
This variable is available only in shell functions invoked by the
programmable completion facilities (@pxref{Programmable Completion}).

@item COMP_KEY
The key (or final key of a key sequence) used to invoke the current
completion function.
This variable is available only in shell functions and external
commands invoked by the
programmable completion facilities (@pxref{Programmable Completion}).

@item COMP_LINE
The current command line.
This variable is available only in shell functions and external
commands invoked by the
programmable completion facilities (@pxref{Programmable Completion}).

@item COMP_POINT
The index of the current cursor position relative to the beginning of
the current command.
If the current cursor position is at the end of the current command,
the value of this variable is equal to @code{$@{#COMP_LINE@}}.
This variable is available only in shell functions and external
commands invoked by the
programmable completion facilities (@pxref{Programmable Completion}).

@item COMP_TYPE
Set to an integer value corresponding to the type of attempted completion
that caused a completion function to be called:
@key{TAB}, for normal completion,
@samp{?}, for listing completions after successive tabs,
@samp{!}, for listing alternatives on partial word completion,
@samp{@@}, to list completions if the word is not unmodified,
or
@samp{%}, for menu completion.
This variable is available only in shell functions and external
commands invoked by the
programmable completion facilities (@pxref{Programmable Completion}).

@item COMP_WORDBREAKS
The set of characters that the Readline library treats as word
separators when performing word completion.
If @env{COMP_WORDBREAKS}
is unset, it loses its special properties,
even if it is subsequently reset.

@item COMP_WORDS
An array variable consisting of the individual
words in the current command line.
The line is split into words as Readline would split it, using
@code{COMP_WORDBREAKS} as described above.
This variable is available only in shell functions invoked by the
programmable completion facilities (@pxref{Programmable Completion}).

@item COMPREPLY
An array variable from which Bash reads the possible completions
generated by a shell function invoked by the programmable completion
facility (@pxref{Programmable Completion}).
Each array element contains one possible completion.

@item COPROC
An array variable created to hold the file descriptors
for output from and input to an unnamed coprocess (@pxref{Coprocesses}).

@item DIRSTACK
An array variable containing the current contents of the directory stack.
Directories appear in the stack in the order they are displayed by the
@code{dirs} builtin.
Assigning to members of this array variable may be used to modify
directories already in the stack, but the @code{pushd} and @code{popd}
builtins must be used to add and remove directories.
Assigning to this variable does not change the current directory.
If @env{DIRSTACK}
is unset, it loses its special properties, even if
it is subsequently reset.

@item EMACS
If Bash finds this variable in the environment when the shell
starts, and its value is @samp{t}, Bash assumes that the shell is
running in an Emacs shell buffer and disables line editing.

@item ENV
Expanded and executed similarly to @code{BASH_ENV}
(@pxref{Bash Startup Files})
when an interactive shell is invoked in
@sc{posix} mode (@pxref{Bash POSIX Mode}).

@item EPOCHREALTIME
Each time this parameter is referenced, it expands to the number of seconds
since the Unix Epoch as a floating-point value with micro-second granularity
(see the documentation for the C library function @code{time} for the
definition of Epoch).
Assignments to @env{EPOCHREALTIME} are ignored.
If @env{EPOCHREALTIME}
is unset, it loses its special properties, even if
it is subsequently reset.

@item EPOCHSECONDS
Each time this parameter is referenced, it expands to the number of seconds
since the Unix Epoch (see the documentation for the C library function
@code{time} for the definition of Epoch).
Assignments to @env{EPOCHSECONDS} are ignored.
If @env{EPOCHSECONDS}
is unset, it loses its special properties, even if
it is subsequently reset.

@item EUID
The numeric effective user id of the current user.
This variable is readonly.

@item EXECIGNORE
A colon-separated list of shell patterns (@pxref{Pattern Matching})
defining the set of filenames to be ignored by command search using
@code{PATH}.
Files whose full pathnames match one of these patterns are not considered
executable files for the purposes of completion and command execution
via @code{PATH} lookup.
This does not affect the behavior of the @code{[}, @code{test}, and @code{[[}
commands.
Full pathnames in the command hash table are not subject to @code{EXECIGNORE}.
Use this variable to ignore shared library files that have the executable
bit set, but are not executable files.
The pattern matching honors the setting of the @code{extglob} shell 
option.

@item FCEDIT
The editor used as a default by the @code{fc} builtin command.

@item FIGNORE
A colon-separated list of suffixes to ignore when performing
filename completion.
A filename whose suffix matches one of the entries in 
@env{FIGNORE}
is excluded from the list of matched filenames.
A sample value is
@samp{.o:~}.
Since tilde expansion takes place after
@samp{:}
in assignment statements,
make sure to quote assignments appropriately to avoid it as appropriate.

@item FUNCNAME
An array variable containing the names of all shell functions
currently in the execution call stack.
The element with index 0 is the name of any currently-executing
shell function.
The bottom-most element (the one with the highest index)
is @code{"main"}.
This variable exists only when a shell function is executing.
Assignments to @env{FUNCNAME} have no effect.
If @env{FUNCNAME}
is unset, it loses its special properties, even if
it is subsequently reset.

This variable can be used with @code{BASH_LINENO} and @code{BASH_SOURCE}.
Each element of @code{FUNCNAME} has corresponding elements in
@code{BASH_LINENO} and @code{BASH_SOURCE} to describe the call stack.
For instance,
@code{$@{FUNCNAME[$i]@}}
was called from the file
@code{$@{BASH_SOURCE[$i+1]@}}
at line number
@code{$@{BASH_LINENO[$i]@}}.
The @code{caller} builtin displays the current call stack using this
information.

@item FUNCNEST
A numeric value greater than 0 defines a maximum function nesting level.
Function invocations that exceed this nesting level
cause the current command to abort.

@item GLOBIGNORE
A colon-separated list of patterns defining the set of file names to
be ignored by filename expansion.
If a file name matched by a filename expansion pattern also matches one
of the patterns in @env{GLOBIGNORE}, it is removed from the list
of matches.
The pattern matching honors the setting of the @code{extglob} shell
option.

@item GLOBSORT
Controls how the results of filename expansion are sorted.
The value of this variable specifies the sort criteria and sort order for
the results of filename expansion.
If this variable is unset or set to the null string, filename expansion
uses the historical behavior of sorting by name,
in ascending lexicographic order as determined by the
@code{LC_COLLATE} shell variable.

If set, a valid value begins with an optional @samp{+}, which is ignored,
or @samp{-}, which reverses the sort order from ascending to descending,
followed by a sort specifier.
The valid sort specifiers are
@samp{name},
@samp{numeric},
@samp{size},
@samp{mtime},
@samp{atime},
@samp{ctime},
and
@samp{blocks},
which sort the files on name,
names in numeric rather than lexicographic order,
file size, modification time, access time,
inode change time, and number of blocks, respectively.
If any of the non-name keys compare as equal (e.g., if two files are
the same size), sorting uses the name as a secondary sort key.

For example, a value of @code{-mtime} sorts the results in descending
order by modification time (newest first).

The @samp{numeric} specifier treats names consisting solely of digits as
numbers and sorts them using their numeric value
(so ``2'' sorts before ``10'', for example).
When using @samp{numeric}, names containing non-digits sort after all
the all-digit names and are sorted by name using the traditional behavior.

A sort specifier of @samp{nosort} disables sorting completely;
Bash returns the results
in the order they are read from the file system,
ignoring any leading @samp{-}.

If the sort specifier is missing, it defaults to @var{name},
so a value of @samp{+} is equivalent to the null string,
and a value of @samp{-} sorts by name in descending order.

Any invalid value restores the historical sorting behavior.

@item GROUPS
An array variable containing the list of groups of which the current    
user is a member.
Assignments to @env{GROUPS} have no effect.
If @env{GROUPS}
is unset, it loses its special properties, even if it is
subsequently reset.

@item histchars
The two or three characters which control history expansion,
quick substitution, and tokenization
(@pxref{History Interaction}).
The first character is the @dfn{history expansion} character,
the character which begins a history expansion, normally
@samp{!}.
The second character is the
@dfn{quick substitution}
character, normally
@samp{^}.
When it appears as the first character on the line,
history substitution repeats the previous command,
replacing one string with another.
The optional third character is the
@dfn{history comment}
character, normally @samp{#},
which indicates
that the remainder of the line is a comment
when it appears as the first character of a word.
The history comment character disables history substitution
for the remaining words on the line.
It does not necessarily cause the shell parser to treat the rest of the
line as a comment.

@item HISTCMD
The history number, or index in the history list, of the current
command.
Assignments to @env{HISTCMD}
have no effect.
If @env{HISTCMD}
is unset, it loses its special properties,
even if it is subsequently reset.

@item HISTCONTROL
A colon-separated list of values controlling how commands are saved on
the history list.
If the list of values includes @samp{ignorespace}, lines which begin
with a space character are not saved in the history list.
A value of @samp{ignoredups} causes lines which match the previous
history entry not to be saved.
A value of @samp{ignoreboth} is shorthand for
@samp{ignorespace} and @samp{ignoredups}.
A value of @samp{erasedups} causes all previous lines matching the
current line to be removed from the history list before that line
is saved.
Any value not in the above list is ignored.
If @env{HISTCONTROL} is unset, or does not include a valid value,
Bash saves all lines read by the shell parser on the history list, 
subject to the value of @env{HISTIGNORE}.
If the first line of a multi-line compound command was saved,
the second and subsequent lines are not tested,
and are added to the history regardless of the value of
@env{HISTCONTROL}.
If the first line was not saved, the second and subsequent lines of
the command are not saved either.

@item HISTFILE
The name of the file to which the command history is saved.
Bash assigns a default value of @file{~/.bash_history}.
If @env{HISTFILE} is unset or null,
the shell does not save the command history when it exits.

@item HISTFILESIZE
The maximum number of lines contained in the history file.
When this variable is assigned a value, the history file is truncated,
if necessary, to contain no more than
the number of history entries
that total no more than that number of lines
by removing the oldest entries.
If the history list contains multi-line entries,
the history file may contain more lines than this maximum
to avoid leaving partial history entries.
The history file is also truncated to this size after
writing it when a shell exits or by the @code{history} builtin.
If the value is 0, the history file is truncated to zero size.
Non-numeric values and numeric values less than zero inhibit truncation.
The shell sets the default value to the value of @env{HISTSIZE}
after reading any startup files.

@item HISTIGNORE
A colon-separated list of patterns used to decide which command
lines should be saved on the history list.
If a command line matches one of the patterns in the value of
@code{HISTIGNORE},
it is not saved on the history list.
Each pattern is anchored at the
beginning of the line and must match the complete line
(Bash does not implicitly append a
@samp{*}).
Each pattern is tested against the line
after the checks specified by
@env{HISTCONTROL}
are applied.
In addition to the normal shell pattern matching characters,
@samp{&}
matches the previous history line.
A backslash escapes the
@samp{&};
the backslash is removed before attempting a match.
If the first line of a multi-line compound command was saved,
the second and subsequent lines are not tested,
and are added to the history regardless of the value of
@env{HISTIGNORE}.
If the first line was not saved, the second and subsequent lines of
the command are not saved either.
The pattern matching honors the setting of the @code{extglob} shell
option.

@env{HISTIGNORE} subsumes some of the function of @env{HISTCONTROL}.
A pattern of @samp{&} is identical to @code{ignoredups}, and a
pattern of @samp{[ ]*} is identical to @code{ignorespace}. 
Combining these two patterns, separating them with a colon,
provides the functionality of @code{ignoreboth}. 

@item HISTSIZE
The maximum number of commands to remember on the history list.
If the value is 0, commands are not saved in the history list.
Numeric values less than zero result in every command being saved
on the history list (there is no limit).
The shell sets the default value to 500 after reading any startup files.

@item HISTTIMEFORMAT
If this variable is set and not null, its value is used as a format string
for @code{strftime}(3) to print the time stamp associated with each history
entry displayed by the @code{history} builtin.
If this variable is set, the shell writes time stamps to the history file so
they may be preserved across shell sessions.
This uses the history comment character to distinguish timestamps from
other history lines.

@item HOSTFILE
Contains the name of a file in the same format as @file{/etc/hosts} that
should be read when the shell needs to complete a hostname.
The list of possible hostname completions may be changed while the shell
is running;
the next time hostname completion is attempted after the
value is changed, Bash adds the contents of the new file to the
existing list.
If @env{HOSTFILE}
is set, but has no value, or does not name a readable file,
Bash attempts to read 
@file{/etc/hosts}
to obtain the list of possible hostname completions.
When @env{HOSTFILE} is unset,
Bash clears the hostname list.

@item HOSTNAME
The name of the current host.

@item HOSTTYPE
A string describing the machine Bash is running on.

@item IGNOREEOF
Controls the action of the shell on receipt of an @code{EOF} character
as the sole input.
If set, the value is the number
of consecutive @code{EOF} characters that can be read as the
first character on an input line before Bash exits.
If the variable is set but does not have a numeric value,
or the value is null, then the default is 10.
If the variable is unset, then @code{EOF} signifies the end of 
input to the shell.
This is only in effect for interactive shells.

@item INPUTRC
The name of the Readline initialization file, overriding the default
of @file{~/.inputrc}.

@item INSIDE_EMACS
If Bash finds this variable in the environment when the shell
starts, it assumes that the shell is running in an Emacs shell buffer
and may disable line editing depending on the value of @env{TERM}.

@item LANG  
Used to determine the locale category for any category not specifically
selected with a variable starting with @code{LC_}.

@item LC_ALL
This variable overrides the value of @env{LANG} and any other
@code{LC_} variable specifying a locale category.

@item LC_COLLATE
This variable determines the collation order used when sorting the
results of filename expansion, and
determines the behavior of range expressions, equivalence classes,
and collating sequences within filename expansion and pattern matching
(@pxref{Filename Expansion}).

@item LC_CTYPE
This variable determines the interpretation of characters and the
behavior of character classes within filename expansion and pattern
matching (@pxref{Filename Expansion}).

@item LC_MESSAGES
This variable determines the locale used to translate double-quoted
strings preceded by a @samp{$} (@pxref{Locale Translation}).

@item LC_NUMERIC
This variable determines the locale category used for number formatting.

@item LC_TIME
This variable determines the locale category used for data and time
formatting.

@item LINENO
The line number in the script or shell function currently executing.
Line numbers start with 1.
When not in a script or function, the value is not
guaranteed to be meaningful.
If @env{LINENO}
is unset, it loses its special properties, even if it is
subsequently reset.

@item LINES
Used by the @code{select} command to determine the column length
for printing selection lists.
Automatically set if the @code{checkwinsize} option is enabled
(@pxref{The Shopt Builtin}), or in an interactive shell upon receipt of a
@code{SIGWINCH}.

@item MACHTYPE
A string that fully describes the system type on which Bash
is executing, in the standard @sc{gnu} @var{cpu-company-system} format.

@item MAILCHECK
How often (in seconds) that the shell should check for mail in the
files specified in the @env{MAILPATH} or @env{MAIL} variables.
The default is 60 seconds.  When it is time to check
for mail, the shell does so before displaying the primary prompt.
If this variable is unset, or set to a value that is not a number
greater than or equal to zero, the shell disables mail checking.

@item MAPFILE
An array variable created to hold the text read by the
@code{mapfile} builtin when no variable name is supplied.

@item OLDPWD
The previous working directory as set by the @code{cd} builtin.

@item OPTERR
If set to the value 1, Bash displays error messages
generated by the @code{getopts} builtin command.
@env{OPTERR}
is initialized to 1 each time the shell is invoked.

@item OSTYPE
A string describing the operating system Bash is running on.

@item PIPESTATUS
An array variable (@pxref{Arrays})
containing a list of exit status values from the commands
in the most-recently-executed foreground pipeline, which may
consist of only a simple command
(@pxref{Shell Commands}).
Bash sets
@code{PIPESTATUS}
after executing multi-element pipelines,
timed and negated pipelines,
simple commands,
subshells created with the @samp{(} operator,
the
@code{[[}
and
@code{((}
compound commands, and
after error conditions that result in the
shell aborting command execution.

@item POSIXLY_CORRECT
If this variable is in the environment when Bash starts, the shell
enters @sc{posix} mode (@pxref{Bash POSIX Mode}) before reading the
startup files, as if the @option{--posix} invocation option had been supplied.
If it is set while the shell is running, Bash enables @sc{posix} mode,
as if the command
@example
@code{set -o posix}
@end example
@noindent
had been executed.
When the shell enters @sc{posix} mode, it sets this variable if it was
not already set.

@item PPID
The process @sc{id} of the shell's parent process.
This variable is readonly.

@item PROMPT_COMMAND
If this variable is set, and is an array,
the value of each set element is interpreted as a command to execute
before printing the primary prompt (@env{$PS1}).
If this is set but not an array variable,
its value is used as a command to execute instead.

@item PROMPT_DIRTRIM
If set to a number greater than zero, the value is used as the number of
trailing directory components to retain when expanding the @code{\w} and
@code{\W} prompt string escapes (@pxref{Controlling the Prompt}).
Characters removed are replaced with an ellipsis.

@item PS0
The value of this parameter is expanded like @env{PS1}
and displayed by interactive shells after reading a command
and before the command is executed.

@item PS3
The value of this variable is used as the prompt for the
@code{select} command.  If this variable is not set, the
@code{select} command prompts with @samp{#? }

@item PS4
The value of this parameter is expanded like @env{PS1}
and the expanded value is the prompt printed before the command line
is echoed when the @option{-x} option is set (@pxref{The Set Builtin}).
The first character of the expanded value is replicated multiple times,
as necessary, to indicate multiple levels of indirection.
The default is @samp{+ }.

@item PWD
The current working directory as set by the @code{cd} builtin.

@item RANDOM
Each time this parameter is referenced, it expands to a random integer
between 0 and 32767.
Assigning a value to
@env{RANDOM}
initializes (seeds) the sequence of random numbers.
Seeding the random number generator with the same constant value
produces the same sequence of values.
If @env{RANDOM}
is unset, it loses its special properties, even if it is
subsequently reset.

@item READLINE_ARGUMENT
Any numeric argument given to a Readline
command that was defined using
@samp{bind -x} (@pxref{Bash Builtins})
when it was invoked.

@item READLINE_LINE
The contents of the Readline
line buffer, for use with
@samp{bind -x} (@pxref{Bash Builtins}).

@item READLINE_MARK
The position of the @dfn{mark} (saved insertion point) in the Readline
line buffer, for use
with @samp{bind -x} (@pxref{Bash Builtins}).
The characters between the insertion point and the mark are often
called the @dfn{region}.

@item READLINE_POINT
The position of the insertion point in the Readline
line buffer, for use with
@samp{bind -x} (@pxref{Bash Builtins}).

@item REPLY
The default variable for the @code{read} builtin;
set to the line read when @code{read} is not supplied a variable name
argument.

@item SECONDS
This variable expands to the number of seconds since the shell was started. 
Assignment to this variable resets the count to the value assigned, and the
expanded value becomes the value assigned plus the number of seconds
since the assignment.
The number of seconds at shell invocation and the current time are always
determined by querying the system clock at one-second resolution.
If @env{SECONDS}
is unset, it loses its special properties,
even if it is subsequently reset.

@item SHELL
This environment variable expands to the full pathname to the shell.
If it is not set when the shell starts,
Bash assigns to it the full pathname of the current user's login shell.

@item SHELLOPTS
A colon-separated list of enabled shell options.
Each word in the list is a valid argument for the @option{-o} option
to the @code{set} builtin command (@pxref{The Set Builtin}).
The options appearing in @env{SHELLOPTS} are those reported
as @samp{on} by @samp{set -o}.
If this variable is in the environment when Bash
starts up, the shell enables each option in the list before
reading any startup files.
If this variable is exported, child shells will enable each option
in the list.
This variable is readonly.

@item SHLVL
Incremented by one each time a new instance of Bash is started.
This is intended to be a count of how deeply your Bash shells are nested.

@item SRANDOM
This variable expands to a 32-bit pseudo-random number each time it is
referenced.
The random number generator is not linear on systems that
support @file{/dev/urandom} or @code{arc4random}, so each returned number
has no relationship to the numbers preceding it.
The random number generator cannot be seeded, so assignments to this
variable have no effect.
If @env{SRANDOM}
is unset, it loses its special properties,
even if it is subsequently reset.

@item TIMEFORMAT
The value of this parameter is used as a format string specifying
how the timing information for pipelines prefixed with the @code{time}
reserved word should be displayed.
The @samp{%} character introduces an
escape sequence that is expanded to a time value or other
information.
The escape sequences and their meanings are as
follows; the brackets denote optional portions. 

@table @code

@item %%
A literal @samp{%}.

@item %[@var{p}][l]R
The elapsed time in seconds. 

@item %[@var{p}][l]U
The number of CPU seconds spent in user mode. 

@item %[@var{p}][l]S
The number of CPU seconds spent in system mode. 

@item %P
The CPU percentage, computed as (%U + %S) / %R. 
@end table

The optional @var{p} is a digit specifying the precision, the number of
fractional digits after a decimal point.
A value of 0 causes no decimal point or fraction to be output.
@code{time} prints at most six digits after the decimal point;
values of @var{p} greater than 6 are changed to 6.
If @var{p} is not specified,
@code{time} prints three digits after the decimal point.

The optional @code{l} specifies a longer format, including minutes, of
the form @var{MM}m@var{SS}.@var{FF}s.
The value of @var{p} determines whether or not the fraction is included. 

If this variable is not set, Bash acts as if it had the value
@example
@code{$'\nreal\t%3lR\nuser\t%3lU\nsys\t%3lS'}
@end example
If the value is null, Bash does not display any timing information.
A trailing newline is added when the format string is displayed.

@item TMOUT
If set to a value greater than zero, the @code{read} builtin uses the
value as its
default timeout (@pxref{Bash Builtins}).
The @code{select} command (@pxref{Conditional Constructs}) terminates
if input does not arrive after @code{TMOUT} seconds when input is coming
from a terminal.

In an interactive shell, the value is interpreted as
the number of seconds to wait for a line of input after issuing
the primary prompt.
Bash terminates after waiting for that number of seconds if a
complete line of input does not arrive.

@item TMPDIR
If set, Bash uses its value as the name of a directory in which
Bash creates temporary files for the shell's use.

@item UID
The numeric real user id of the current user.  This variable is readonly.

@end vtable

@node Bash Features
@chapter Bash Features

This chapter describes features unique to Bash.

@menu
* Invoking Bash::		Command line options that you can give
				to Bash.
* Bash Startup Files::		When and how Bash executes scripts.
* Interactive Shells::		What an interactive shell is.
* Bash Conditional Expressions::	Primitives used in composing expressions for
				the @code{test} builtin.
* Shell Arithmetic::		Arithmetic on shell variables.
* Aliases::			Substituting one command for another.
* Arrays::			Array Variables.
* The Directory Stack::		History of visited directories.
* Controlling the Prompt::	Customizing the various prompt strings.
* The Restricted Shell::	A more controlled mode of shell execution.
* Bash POSIX Mode::		Making Bash behave more closely to what
				the POSIX standard specifies.
* Shell Compatibility Mode::	How Bash supports behavior that was present
				in earlier versions and has changed.
@end menu

@node Invoking Bash
@section Invoking Bash

@example
bash [long-opt] [-ir] [-abefhkmnptuvxdBCDHP] [-o @var{option}]
    [-O @var{shopt_option}] [@var{argument} @dots{}]
bash [long-opt] [-abefhkmnptuvxdBCDHP] [-o @var{option}]
    [-O @var{shopt_option}] -c @var{string} [@var{argument} @dots{}]
bash [long-opt] -s [-abefhkmnptuvxdBCDHP] [-o @var{option}]
    [-O @var{shopt_option}] [@var{argument} @dots{}]
@end example

All of the single-character options used with the @code{set} builtin
(@pxref{The Set Builtin}) can be used as options when the shell is invoked.
In addition, there are several multi-character
options that you can use.
These options must appear on the command
line before the single-character options to be recognized. 

@table @code
@item --debugger
Arrange for the debugger profile to be executed before the shell starts.
Turns on extended debugging mode (see @ref{The Shopt Builtin}
for a description of the @code{extdebug} option to the @code{shopt}
builtin).

@item --dump-po-strings
Print a list of all double-quoted strings preceded by @samp{$}
on the standard output
in the @sc{gnu} @code{gettext} PO (portable object) file format.
Equivalent to @option{-D} except for the output format.

@item --dump-strings
Equivalent to @option{-D}.

@item --help
Display a usage message on standard output and exit successfully.

@item --init-file @var{filename}
@itemx --rcfile @var{filename}
Execute commands from @var{filename} (instead of @file{~/.bashrc})
in an interactive shell.

@item --login
Equivalent to @option{-l}.

@item --noediting
Do not use the @sc{gnu} Readline library (@pxref{Command Line Editing})
to read  command lines when the shell is interactive.

@item --noprofile
Don't load the system-wide startup file @file{/etc/profile}
or any of the personal initialization files
@file{~/.bash_profile}, @file{~/.bash_login}, or @file{~/.profile}
when Bash is invoked as a login shell.

@item --norc
Don't read the @file{~/.bashrc} initialization file in an
interactive shell.
This is on by default if the shell is invoked as @code{sh}.

@item --posix
Enable @sc{posix} mode;
change the behavior of Bash
where the default operation differs from the
@sc{posix}
standard to match the standard.
This is intended to make Bash behave as a strict superset of that
standard.
@xref{Bash POSIX Mode}, for a description of the Bash @sc{posix} mode.

@item --restricted
Equivalent to @option{-r}.
Make the shell a restricted shell (@pxref{The Restricted Shell}).

@item --verbose
Equivalent to @option{-v}.
Print shell input lines as they're read.

@item --version
Show version information for this instance of
Bash on the standard output and exit successfully.
@end table

There are several single-character options that may be supplied at
invocation which are not available with the @code{set} builtin.

@table @code
@item -c
Read and execute commands from the first non-option argument
@var{command_string}, then exit. 
If there are arguments after the @var{command_string},
the first argument is assigned to @code{$0}
and any remaining arguments are assigned to the positional parameters.
The assignment to @code{$0} sets the name of the shell, which is used
in warning and error messages.

@item -i
Force the shell to run interactively.
Interactive shells are described in @ref{Interactive Shells}.

@item -l
Make this shell act as if it had been directly invoked by login.
When the shell is interactive, this is equivalent to starting a
login shell with @samp{exec -l bash}.
When the shell is not interactive, it will read and execute the login
shell startup files.
@samp{exec bash -l} or @samp{exec bash --login}
will replace the current shell with a Bash login shell.
@xref{Bash Startup Files}, for a description of the special behavior
of a login shell.

@item -r
Make the shell a restricted shell (@pxref{The Restricted Shell}).

@item -s
If this option is present, or if no arguments remain after option
processing, then Bash reads commands from the standard input.
This option allows the positional parameters to be set
when invoking an interactive shell or when reading input
through a pipe.

@item -D
Print a list of all double-quoted strings preceded by @samp{$}
on the standard output.
These are the strings that
are subject to language translation when the current locale
is not @code{C} or @code{POSIX} (@pxref{Locale Translation}).
This implies the @option{-n} option; no commands will be executed.

@item [-+]O [@var{shopt_option}]
@var{shopt_option} is one of the shell options accepted by the
@code{shopt} builtin (@pxref{The Shopt Builtin}).
If @var{shopt_option} is present, @option{-O} sets the value of that option;
@option{+O} unsets it.  
If @var{shopt_option} is not supplied, Bash
prints the names and values of the shell
options accepted by @code{shopt} on the standard output.
If the invocation option is @option{+O}, the output is displayed in a format
that may be reused as input.

@item --
A @code{--} signals the end of options and disables further option
processing.
Any arguments after the @code{--}
are treated as a shell script filename (@pxref{Shell Scripts})
and arguments passed to that script.

@item -
Equivalent to @code{--}.

@end table

@cindex login shell
A @dfn{login shell} is one whose first character of argument zero is
@samp{-}, or one invoked with the @option{--login} option.

@cindex interactive shell
An @dfn{interactive shell} is one started without non-option arguments,
unless @option{-s} is specified,
without specifying the @option{-c} option,
and whose standard input and standard error are both
connected to terminals (as determined by
@i{isatty(3)}),
or one started with the @option{-i} option.
@xref{Interactive Shells}, for more information.

A @dfn{non-interactive shell} is, not surprisingly, one that does not
satisfy the tests for interactivity given above.
A non-interactive shell is usually started to run commands from a script
file supplied as an argument (@pxref{Shell Scripts})
or from a string supplied with the
@option{-c}
option.

If arguments remain after option processing, and neither the
@option{-c} nor the @option{-s}
option has been supplied, the first argument is treated as
the name of a file containing shell commands (@pxref{Shell Scripts}).
When Bash is invoked in this fashion, @code{$0}
is set to the name of the file, and the positional parameters
are set to the remaining arguments.
Bash reads and executes commands from this file, then exits.   
Bash's exit status is the exit status of the last command executed
in the script.
If no commands are executed, the exit status is 0.
Bash
first attempts to open the file in the current directory,
and, if no file is found, searches the directories in
@env{PATH}
for the script.

For example, the following command starts an interactive shell:

@example
bash
@end example

@noindent
whereas this command will read and execute commands from @file{filename}:

@example
bash filename
@end example

@noindent
This command will execute @samp{echo hello}:

@example
bash -c 'echo hello'
@end example

@noindent
and this will execute the same command
read from the standard input:

@example
echo echo hello | bash
@end example

@node Bash Startup Files
@section Bash Startup Files
@cindex startup files

This section describes how Bash executes its startup files.
If any of the files exist but cannot be read, Bash reports an error.
Tildes in filenames are expanded as described above under
Tilde Expansion (@pxref{Tilde Expansion}).

Interactive shells are described in @ref{Interactive Shells}.

@subsubheading Invoked as an interactive login shell, or with @option{--login}

When Bash is invoked as an interactive login shell, or as a
non-interactive shell with the @option{--login} option, it first reads and
executes commands from the file @file{/etc/profile}, if that file exists.
After reading that file, it looks for @file{~/.bash_profile},
@file{~/.bash_login}, and @file{~/.profile}, in that order, and reads
and executes commands from the first one that exists and is readable.
The
@option{--noprofile}
option inhibits this behavior.

When an interactive login shell exits,
or a non-interactive login shell executes the @code{exit} builtin command,
Bash reads and executes commands from
the file @file{~/.bash_logout}, if it exists.

@subsubheading Invoked as an interactive non-login shell

When Bash runs as an interactive shell that is not a login shell, it
reads and executes commands from @file{~/.bashrc}, if that file exists.
The @option{--norc} option inhibits this behavior.
The @option{--rcfile @var{file}} option
causes Bash to
use @var{file} instead of @file{~/.bashrc}.

So, typically, your @file{~/.bash_profile} contains the line
@example
@code{if [ -f ~/.bashrc ]; then . ~/.bashrc; fi}
@end example
@noindent
after (or before) any login-specific initializations.

@subsubheading Invoked non-interactively

When Bash is started non-interactively, to run a shell script,
for example, it looks for the variable @env{BASH_ENV} in the environment,
expands its value if it appears there, and uses the expanded value as
the name of a file to read and execute.
Bash behaves as if the following command were executed:
@example
@code{if [ -n "$BASH_ENV" ]; then . "$BASH_ENV"; fi}
@end example
@noindent
but does not the value of the @env{PATH} variable to search for the
filename.

As noted above, if a non-interactive shell is invoked with the
@option{--login} option,
Bash attempts to read and execute commands from the
login shell startup files. 

@subsubheading Invoked with name @code{sh}

If Bash is invoked with the name @command{sh},
it tries to mimic the startup behavior of historical versions of
@command{sh} as closely as possible,
while conforming to the
@sc{posix}
standard as well.

When invoked as an interactive login shell, or as a non-interactive
shell with the @option{--login} option, it first attempts to read
and execute commands from
@file{/etc/profile} and @file{~/.profile},
in that order.
The @option{--noprofile} option inhibits this behavior.

When invoked as an interactive shell with the name @command{sh}, Bash
looks for the variable @env{ENV},
expands its value if it is defined, and uses the
expanded value as the name of a file to read and execute.
Since a shell invoked as @command{sh}
does not attempt to read and execute commands from any other startup
files, the
@option{--rcfile}
option has no effect.

A non-interactive shell invoked with the name @command{sh}
does not attempt to read any other startup files.

When invoked as @command{sh}, Bash enters @sc{posix} mode after reading
the startup files.

@subsubheading Invoked in @sc{posix} mode

When Bash is started in @sc{posix} mode, as with the
@option{--posix} command line option, it follows the
@sc{posix}
standard for startup files.
In this mode, interactive shells expand the @env{ENV} variable
and read and execute commands from the file whose name is the
expanded value.
A posix-mode shell does not attempt to read any other startup files,
even when invoked as a login shell.

@subsubheading Invoked by remote shell daemon

Bash
attempts to determine when it is being run with its standard input
connected to a network connection, as when executed by
the secure shell daemon @code{sshd}.
or the historical and rarely-seen remote shell daemon, usually @code{rshd}.
If
Bash
determines it is being run non-interactively in this fashion,
it reads and executes commands from
@file{~/.bashrc},
if that file exists and is readable.
Bash does not read this file if invoked as @code{sh}.
The
@option{--norc}
option inhibits this behavior, and the
@option{--rcfile}
option makes Bash use a different file instead of
@file{~/.bashrc},
but neither
@command{sshd} nor @command{rshd} generally invoke the shell with those
options or allow them to be specified.

@subsubheading Invoked with unequal effective and real @sc{uid/gid}s

If Bash is started with the effective user (group) id not equal to the
real user (group) id, and the @option{-p} option is not supplied, no startup
files are read, shell functions are not inherited from the environment,
the @env{SHELLOPTS}, @env{BASHOPTS}, @env{CDPATH}, and @env{GLOBIGNORE}
variables, if they appear in the environment, are ignored, and the effective
user id is set to the real user id.
If the @option{-p} option is supplied at invocation, the startup behavior is
the same, but Bash does not reset the effective user id.

@node Interactive Shells
@section Interactive Shells
@cindex interactive shell
@cindex shell, interactive

@menu
* What is an Interactive Shell?::	What determines whether a shell is Interactive.
* Is this Shell Interactive?::	How to tell if a shell is interactive.
* Interactive Shell Behavior::	What changes in an interactive shell?
@end menu

@node What is an Interactive Shell?
@subsection What is an Interactive Shell?

An interactive shell
is one started without non-option arguments
(unless @option{-s} is specified)
and without specifying the @option{-c} option,
whose input and error output are both
connected to terminals (as determined by @code{isatty(3)}),
or one started with the @option{-i} option.

An interactive shell generally reads from and writes to a user's
terminal.

The @option{-s} invocation option may be used to set the positional parameters
when an interactive shell starts.

@node Is this Shell Interactive?
@subsection Is this Shell Interactive?

To determine within a startup script whether or not Bash is
running interactively,
test the value of the @samp{-} special parameter.
It contains @code{i} when the shell is interactive.
For example:

@example
case "$-" in
*i*)	echo This shell is interactive ;;
*)	echo This shell is not interactive ;;
esac
@end example

Alternatively, startup scripts may examine the variable
@env{PS1}; it is unset in non-interactive shells, and set in
interactive shells.
Thus:

@example
if [ -z "$PS1" ]; then
        echo This shell is not interactive
else
        echo This shell is interactive
fi
@end example

@node Interactive Shell Behavior
@subsection Interactive Shell Behavior

When the shell is running interactively, it changes its behavior in
several ways.

@enumerate
@item
Bash reads and executes startup files as described in @ref{Bash Startup Files}.

@item
Job Control (@pxref{Job Control}) is enabled by default.
When job control is in effect, Bash ignores the keyboard-generated
job control signals @code{SIGTTIN}, @code{SIGTTOU}, and @code{SIGTSTP}.

@item
Bash executes the values of the set elements of the @env{PROMPT_COMMAND}
array variable as commands before printing the primary prompt, @env{$PS1}
(@pxref{Bash Variables}).

@item
Bash expands and displays @env{PS1} before reading the first line
of a command, and expands and displays @env{PS2} before reading the
second and subsequent lines of a multi-line command.
Bash expands and displays @env{PS0} after it reads a command but before
executing it.
See @ref{Controlling the Prompt}, for a complete list of prompt
string escape sequences.

@item
Bash uses Readline (@pxref{Command Line Editing}) to read commands from
the user's terminal.

@item
Bash inspects the value of the @code{ignoreeof} option to @code{set -o}
instead of exiting immediately when it receives an @code{EOF} on its
standard input when reading a command (@pxref{The Set Builtin}).

@item
Bash enables Command history (@pxref{Bash History Facilities})
and history expansion (@pxref{History Interaction})
by default.
When a shell with history enabled exits,
Bash saves the command history to the file named by @env{$HISTFILE}.

@item
Alias expansion (@pxref{Aliases}) is performed by default.

@item
In the absence of any traps, Bash ignores @code{SIGTERM}
(@pxref{Signals}).

@item
In the absence of any traps, @code{SIGINT} is caught and handled
(@pxref{Signals}).
@code{SIGINT} will interrupt some shell builtins.

@item
An interactive login shell sends a @code{SIGHUP} to all jobs on exit
if the @code{huponexit} shell option has been enabled (@pxref{Signals}).

@item
The @option{-n} option has no effect, whether at invocation or when
using @samp{set -n} (@pxref{The Set Builtin}).

@item
Bash will check for mail periodically, depending on the values of the
@env{MAIL}, @env{MAILPATH}, and @env{MAILCHECK} shell variables
(@pxref{Bash Variables}).

@item
The shell will not exit on
expansion errors due to references to unbound shell variables after
@samp{set -u} has been enabled (@pxref{The Set Builtin}).

@item
The shell will not exit on
expansion errors caused by @var{var} being unset or null in
@code{$@{@var{var}:?@var{word}@}} expansions
(@pxref{Shell Parameter Expansion}).

@item
Redirection errors encountered by shell builtins will not cause the
shell to exit.

@item
When running in @sc{posix} mode, a special builtin returning an error
status will not cause the shell to exit (@pxref{Bash POSIX Mode}).

@item
A failed @code{exec} will not cause the shell to exit
(@pxref{Bourne Shell Builtins}).

@item
Parser syntax errors will not cause the shell to exit.

@item
If the @code{cdspell} shell option is enabled, the shell will attempt
simple spelling correction for directory arguments to the @code{cd}
builtin (see the description of the @code{cdspell}
option to the @code{shopt} builtin in @ref{The Shopt Builtin}).
The @code{cdspell} option is only effective in interactive shells.

@item
The shell will check the value of the @env{TMOUT} variable and exit
if a command is not read within the specified number of seconds after
printing @env{$PS1} (@pxref{Bash Variables}).

@end enumerate

@node Bash Conditional Expressions
@section Bash Conditional Expressions
@cindex expressions, conditional

Conditional expressions are used by the @code{[[} compound command
(@pxref{Conditional Constructs})
and the @code{test} and @code{[} builtin commands
(@pxref{Bourne Shell Builtins}).
The @code{test}
and @code{[} commands determine their behavior based on the number
of arguments; see the descriptions of those commands for any other
command-specific actions.

Expressions may be unary or binary,
and are formed from the primaries listed below.
Unary expressions are often used to examine the status of a file
or shell variable.
Binary operators are used for string, numeric, and file attribute
comparisons.

Bash handles several filenames specially when they are used in
expressions.
If the operating system on which Bash is running provides these
special files, Bash uses them; otherwise it emulates them
internally with this behavior:
If the @var{file} argument to one of the primaries is of the form
@file{/dev/fd/@var{N}}, then Bash checks file descriptor @var{N}.
If the @var{file} argument to one of the primaries is one of
@file{/dev/stdin}, @file{/dev/stdout}, or @file{/dev/stderr},
Bash checks file descriptor 0, 1, or 2, respectively.

When used with @code{[[}, the @samp{<} and @samp{>} operators sort
lexicographically using the current locale.
The @code{test} command uses ASCII ordering.

Unless otherwise specified, primaries that operate on files follow symbolic
links and operate on the target of the link, rather than the link itself.

@table @code
@item -a @var{file}
True if @var{file} exists.

@item -b @var{file}
True if @var{file} exists and is a block special file.

@item -c @var{file}
True if @var{file} exists and is a character special file.

@item -d @var{file}
True if @var{file} exists and is a directory.

@item -e @var{file}
True if @var{file} exists.

@item -f @var{file}
True if @var{file} exists and is a regular file.

@item -g @var{file}
True if @var{file} exists and its set-group-id bit is set.

@item -h @var{file}
True if @var{file} exists and is a symbolic link.

@item -k @var{file}
True if @var{file} exists and its "sticky" bit is set.

@item -p @var{file}
True if @var{file} exists and is a named pipe (FIFO).

@item -r @var{file}
True if @var{file} exists and is readable.

@item -s @var{file}
True if @var{file} exists and has a size greater than zero.

@item -t @var{fd}
True if file descriptor @var{fd} is open and refers to a terminal.

@item -u @var{file}
True if @var{file} exists and its set-user-id bit is set.

@item -w @var{file}
True if @var{file} exists and is writable.

@item -x @var{file}
True if @var{file} exists and is executable.

@item -G @var{file}
True if @var{file} exists and is owned by the effective group id.

@item -L @var{file}
True if @var{file} exists and is a symbolic link.

@item -N @var{file}
True if @var{file} exists and has been modified since it was last
accessed.

@item -O @var{file}
True if @var{file} exists and is owned by the effective user id.

@item -S @var{file}
True if @var{file} exists and is a socket.

@item @var{file1} -ef @var{file2}
True if @var{file1} and @var{file2} refer to the same device and
inode numbers.

@item @var{file1} -nt @var{file2}
True if @var{file1} is newer (according to modification date)
than @var{file2}, or if @var{file1} exists and @var{file2} does not.

@item @var{file1} -ot @var{file2}
True if @var{file1} is older than @var{file2},
or if @var{file2} exists and @var{file1} does not.

@item -o @var{optname}
True if the shell option @var{optname} is enabled.
The list of options appears in the description of the @option{-o}
option to the @code{set} builtin (@pxref{The Set Builtin}).

@item -v @var{varname}
True if the shell variable @var{varname} is set (has been assigned a value).
If @var{varname} is an indexed
array variable name subscripted by @samp{@@} or @samp{*},
this returns true if the array has any set elements.
If @var{varname} is an associative
array variable name subscripted by @samp{@@} or @samp{*},
this returns true if an element with that key is set.

@item -R @var{varname}
True if the shell variable @var{varname} is set and is a name reference.

@item -z @var{string}
True if the length of @var{string} is zero.

@item -n @var{string}
@itemx @var{string}
True if the length of @var{string} is non-zero.

@item @var{string1} == @var{string2}
@itemx @var{string1} = @var{string2}
True if the strings are equal.
When used with the @code{[[} command, this performs pattern matching as
described above (@pxref{Conditional Constructs}).

@samp{=} should be used with the @code{test} command for
@sc{posix}
conformance.

@item @var{string1} != @var{string2}
True if the strings are not equal.

@item @var{string1} < @var{string2}
True if @var{string1} sorts before @var{string2} lexicographically.

@item @var{string1} > @var{string2}
True if @var{string1} sorts after @var{string2} lexicographically.

@item @var{arg1} OP @var{arg2}
@code{OP} is one of 
@samp{-eq}, @samp{-ne}, @samp{-lt}, @samp{-le}, @samp{-gt}, or @samp{-ge}.
These arithmetic binary operators return true if @var{arg1}
is equal to, not equal to, less than, less than or equal to,
greater than, or greater than or equal to @var{arg2},
respectively.  @var{Arg1} and @var{arg2}
may be positive or negative integers.
When used with the @code{[[} command, @var{arg1} and @var{arg2}
are evaluated as arithmetic expressions (@pxref{Shell Arithmetic}).
Since the expansions the @code{[[} command performs on
@var{arg1} and @var{arg2}
can potentially result in empty strings,
arithmetic expression evaluation treats
those as expressions that evaluate to 0.
@end table

@node Shell Arithmetic
@section Shell Arithmetic
@cindex arithmetic, shell
@cindex shell arithmetic
@cindex expressions, arithmetic
@cindex evaluation, arithmetic
@cindex arithmetic evaluation

The shell allows arithmetic expressions to be evaluated, as one of
the shell expansions or by using the @code{((} compound command,
the @code{let} and @code{declare} builtins,
the arithmetic @code{for} command,
the @code{[[} conditional command,
or the @option{-i} option to the @code{declare} builtin.

Evaluation is done in the largest fixed-width integers available,
with no check for overflow,
though division by 0 is trapped and flagged as an error.
The operators and their precedence, associativity, and values
are the same as in the C language.
The following list of operators is grouped into levels of
equal-precedence operators.
The levels are listed in order of decreasing precedence. 

@table @code
@cindex arithmetic operators
@cindex unary arithmetic operators
@cindex binary arithmetic operators
@cindex conditional arithmetic operator
@cindex bitwise arithmetic operators

@item @var{id}++ @var{id}--
variable post-increment and post-decrement 

@item ++@var{id} --@var{id}
variable pre-increment and pre-decrement

@item - +
unary minus and plus

@item ! ~
logical and bitwise negation

@item **
exponentiation

@item * / %
multiplication, division, remainder

@item + -
addition, subtraction

@item << >>
left and right bitwise shifts

@item <= >= < >
comparison

@item == !=
equality and inequality

@item &
bitwise AND

@item ^
bitwise exclusive OR

@item |
bitwise OR

@item &&
logical AND

@item ||
logical OR

@item expr ? if-true-expr : if-false-expr
conditional operator

@item = *= /= %= += -= <<= >>= &= ^= |=
assignment

@item expr1 , expr2
comma
@end table

Shell variables are allowed as operands; parameter expansion is
performed before the expression is evaluated. 
Within an expression, shell variables may also be referenced by name
without using the parameter expansion syntax.
This means you can use @var{x},
where @var{x} is a shell variable name, in an arithmetic expression,
and the shell will evaluate its value as an expression and use the
result.
A shell variable that is unset or null evaluates to 0 when referenced
by name in an expression.

The value of a variable is evaluated as an arithmetic expression
when it is referenced, or when a variable which has been given the  
@code{integer} attribute using @samp{declare -i} is assigned a value.
A null value evaluates to 0.
A shell variable need not have its @code{integer} attribute enabled
to be used in an expression.

Integer constants follow the C language definition, without suffixes or
character constants.
Constants with a leading 0 are interpreted as octal numbers.
A leading @samp{0x} or @samp{0X} denotes hexadecimal.  Otherwise,
numbers take the form [@var{base}@code{#}]@var{n}, where the optional @var{base}
is a decimal number between 2 and 64 representing the arithmetic
base, and @var{n} is a number in that base.
If @var{base}@code{#} is omitted, then base 10 is used.
When specifying @var{n},
if a non-digit is required,
the digits greater than 9 are represented by the lowercase letters,
the uppercase letters, @samp{@@}, and @samp{_}, in that order.
If @var{base} is less than or equal to 36, lowercase and uppercase
letters may be used interchangeably to represent numbers between 10
and 35.

Operators are evaluated in precedence order.
Sub-expressions in parentheses are evaluated first
and may override the precedence rules above.

@node Aliases
@section Aliases
@cindex alias expansion

@dfn{Aliases} allow a string to be substituted for a word that is in
a position in the input where it can be the first word of a simple
command.
Aliases have names and corresponding values that are set
and unset using the @code{alias} and @code{unalias} builtin commands
(@pxref{Shell Builtin Commands}).

If the shell reads an unquoted word in the right position, it checks
the word to see if it matches an alias name.
If it matches, the shell
replaces the word with the alias value, and reads that value as if it
had been read instead of the word.
The shell doesn't look at any characters following the word before
attempting alias substitution. 

The characters @samp{/}, @samp{$}, @samp{`}, @samp{=} and any of the
shell metacharacters or quoting characters listed above may not appear
in an alias name.
The replacement text may contain any valid
shell input, including shell metacharacters.
The first word of the replacement text is tested for
aliases, but a word that is identical to an alias being expanded
is not expanded a second time.
This means that one may alias @code{ls} to @code{"ls -F"},
for instance, and Bash does not try to recursively expand the
replacement text.

If the last character of the alias value is a
@code{blank}, then the shell checks
the next command word following the alias for alias expansion.

Aliases are created and listed with the @code{alias}
command, and removed with the @code{unalias} command.

There is no mechanism for using arguments in the replacement text,
as in @code{csh}.
If arguments are needed, use a shell function
(@pxref{Shell Functions}) instead.

Aliases are not expanded when the shell is not interactive,
unless the @code{expand_aliases} shell option is set using
@code{shopt} (@pxref{The Shopt Builtin}).

The rules concerning the definition and use of aliases are
somewhat confusing.
Bash always reads at least one complete line of input,
and all lines that make up a compound command,
before executing any of the commands on that line or the compound command.
Aliases are expanded when a command is read, not when it is executed.
Therefore, an
alias definition appearing on the same line as another
command does not take effect until the shell reads the next line of input,
and an alias definition in a compound command does not take
effect until the shell parses and executes the entire compound command.
The commands following the alias definition
on that line,
or in the rest of a compound command,
are not affected by the new alias.
This behavior is also an issue when functions are executed.
Aliases are expanded when a function definition is read,
not when the function is executed, because a function definition
is itself a command.
As a consequence, aliases
defined in a function are not available until after that
function is executed.
To be safe, always put
alias definitions on a separate line, and do not use @code{alias}
in compound commands.

For almost every purpose, shell functions are preferable to aliases.

@node Arrays
@section Arrays
@cindex arrays

Bash
provides one-dimensional indexed and associative array variables.
Any variable may be used as an indexed array; the
@code{declare}
builtin explicitly declares an array.
There is no maximum
limit on the size of an array, nor any requirement that members
be indexed or assigned contiguously.
Indexed arrays are referenced using arithmetic expressions
that must expand to an integer (@pxref{Shell Arithmetic})
and are zero-based;
associative arrays use arbitrary strings.
Unless otherwise noted, indexed array indices must be non-negative integers.

The shell performs
parameter and variable expansion, arithmetic expansion,
command substitution, and quote removal
on indexed array subscripts.
Since this
can potentially result in empty strings,
subscript indexing treats
those as expressions that evaluate to 0.

The shell performs
tilde expansion,
parameter and variable expansion, arithmetic expansion,
command substitution, and quote removal
on associative array subscripts.
Empty strings cannot be used as associative array keys.

Bash automatically creates an indexed array
if any variable is assigned to using the syntax
@example
@var{name}[@var{subscript}]=@var{value}
@end example

@noindent
The @var{subscript}
is treated as an arithmetic expression that must evaluate to a number
greater than or equal to zero.
To explicitly declare an indexed array, use
@example
declare -a @var{name}
@end example
@noindent
(@pxref{Bash Builtins}).
The syntax
@example
declare -a @var{name}[@var{subscript}]
@end example
@noindent
is also accepted; the @var{subscript} is ignored.

@noindent
Associative arrays are created using
@example
declare -A @var{name}
@end example

Attributes may be specified for an array variable using the
@code{declare} and @code{readonly} builtins.
Each attribute applies to all members of an array.

Arrays are assigned using compound assignments of the form
@example
@var{name}=(@var{value1} @var{value2} @dots{} )
@end example
@noindent
where each
@var{value} may be of the form @code{[@var{subscript}]=}@var{string}.
Indexed array assignments do not require anything but @var{string}.

Each @var{value} in the list undergoes the shell expansions
described above (@pxref{Shell Expansions}),
but @var{value}s that are valid variable assignments
including the brackets and subscript do not undergo
brace expansion and word splitting, as with individual
variable assignments.

When assigning to indexed arrays, if
the optional subscript is supplied, that index is assigned to;
otherwise the index of the element assigned is the last index assigned
to by the statement plus one.
Indexing starts at zero.

When assigning to an associative array, the words in a compound assignment
may be either assignment statements, for which the subscript is required,
or a list of words that is interpreted as a sequence of alternating keys
and values:
@var{name}=(@var{key1} @var{value1} @var{key2} @var{value2} @dots{} ).
These are treated identically to
@var{name}=( [@var{key1}]=@var{value1} [@var{key2}]=@var{value2} @dots{} ).
The first word in the list determines how the remaining words
are interpreted; all assignments in a list must be of the same type.
When using key/value pairs, the keys may not be missing or empty;
a final missing value is treated like the empty string.

This syntax is also accepted by the @code{declare}
builtin.
Individual array elements may be assigned to using the
@code{@var{name}[@var{subscript}]=@var{value}} syntax introduced above.

When assigning to an indexed array, if @var{name}
is subscripted by a negative number, that number is
interpreted as relative to one greater than the maximum index of
@var{name}, so negative indices count back from the end of the
array, and an index of -1 references the last element.

The @samp{+=}
operator appends to an array variable when assigning
using the compound assignment syntax; see
@ref{Shell Parameters} above.

If one of the word expansions in a compound array assignment unsets the
variable, the results are unspecified.

An array element is referenced using
@code{$@{@var{name}[@var{subscript}]@}}.
The braces are required to avoid
conflicts with the shell's filename expansion operators.
If the @var{subscript} is @samp{@@} or @samp{*},
the word expands to all members
of the array @var{name}, unless otherwise noted in the description of a
builtin or word expansion.
These subscripts differ only when the word
appears within double quotes.
If the word is double-quoted,
@code{$@{@var{name}[*]@}} expands to a single word with
the value of each array member separated by the first character of the
@env{IFS} variable, and @code{$@{@var{name}[@@]@}} expands each element of
@var{name} to a separate word.
When there are no array members, @code{$@{@var{name}[@@]@}} expands to
nothing.
If the double-quoted expansion occurs within a word,
the expansion of the first parameter is joined with the beginning part of the
expansion of the original word,
and the expansion of the last parameter is joined with the last part of the
expansion of the original word.
This is analogous to the
expansion of the special parameters @samp{@@} and @samp{*}. 

@code{$@{#@var{name}[@var{subscript}]@}} expands to the length of
@code{$@{@var{name}[@var{subscript}]@}}.
If @var{subscript} is @samp{@@} or @samp{*},
the expansion is the number of elements in the array. 

If the @var{subscript}
used to reference an element of an indexed array
evaluates to a number less than zero, it is
interpreted as relative to one greater than the maximum index of the array,
so negative indices count back from the end of the array,
and an index of -1 refers to the last element.

Referencing an array variable without a subscript is equivalent to
referencing with a subscript of 0.
Any reference to a variable using a valid subscript is valid;
Bash creates an array if necessary.

An array variable is considered set if a subscript has been assigned a
value.
The null string is a valid value.

It is possible to obtain the keys (indices) of an array as well as the values.
$@{!@var{name}[@@]@} and $@{!@var{name}[*]@} expand to the indices
assigned in array variable @var{name}.
The treatment when in double quotes is similar to the expansion of the
special parameters @samp{@@} and @samp{*} within double quotes.

The @code{unset} builtin is used to destroy arrays.
@code{unset @var{name}[@var{subscript}]}
unsets the array element at index @var{subscript}.
Negative subscripts to indexed arrays are interpreted as described above.
Unsetting the last element of an array variable does not unset the variable.
@code{unset @var{name}}, where @var{name} is an array, removes the
entire array.
@code{unset @var{name}[@var{subscript}]} behaves differently
depending on the array type when
@var{subscript} is @samp{*} or @samp{@@}.
When @var{name} is an associative array, it removes the element with key
@samp{*} or @samp{@@}.
If @var{name} is an indexed array, @code{unset} removes all of the elements,
but does not remove the array itself.

When using a variable name with a subscript as an argument to a command,
such as with @code{unset}, without using the word expansion syntax
described above (e.g., unset a[4]),
the argument is subject to the shell's filename expansion.
Quote the argument if pathname expansion is not desired
(e.g., unset 'a[4]').

The @code{declare}, @code{local}, and @code{readonly}
builtins each accept a @option{-a} option to specify an indexed
array and a @option{-A} option to specify an associative array.
If both options are supplied, @option{-A} takes precedence.
The @code{read} builtin accepts a @option{-a}
option to assign a list of words read from the standard input
to an array, and can read values from the standard input into
individual array elements.  The @code{set} and @code{declare}
builtins display array values in a way that allows them to be
reused as input.
Other builtins accept array name arguments as well
(e.g., @code{mapfile}); see the descriptions
of individual builtins for details.
The shell provides a number of builtin array variables.

@node The Directory Stack
@section The Directory Stack
@cindex directory stack

@menu
* Directory Stack Builtins::		Bash builtin commands to manipulate
					the directory stack.
@end menu

The directory stack is a list of recently-visited directories.  The
@code{pushd} builtin adds directories to the stack as it changes
the current directory, and the @code{popd} builtin removes specified
directories from the stack and changes the current directory to
the directory removed.  The @code{dirs} builtin displays the contents
of the directory stack.  The current directory is always the "top"
of the directory stack.

The contents of the directory stack are also visible
as the value of the @env{DIRSTACK} shell variable.

@node Directory Stack Builtins
@subsection Directory Stack Builtins

@table @code

@item dirs
@btindex dirs
@example
dirs [-clpv] [+@var{N} | -@var{N}]
@end example

Without options, display the list of currently remembered directories.
Directories are added to the list with the @code{pushd} command;
the @code{popd} command removes directories from the list.
The current directory is always the first directory in the stack.

Options, if supplied, have the following meanings:

@table @code
@item -c
Clears the directory stack by deleting all of the elements.
@item -l
Produces a listing using full pathnames;
the default listing format uses a tilde to denote the home directory.
@item -p
Causes @code{dirs} to print the directory stack with one entry per
line.
@item -v
Causes @code{dirs} to print the directory stack with one entry per
line, prefixing each entry with its index in the stack.
@item +@var{N}
Displays the @var{N}th directory (counting from the left of the
list printed by @code{dirs} when invoked without options), starting
with zero.
@item -@var{N}
Displays the @var{N}th directory (counting from the right of the
list printed by @code{dirs} when invoked without options), starting
with zero.
@end table

@item popd
@btindex popd
@example
popd [-n] [+@var{N} | -@var{N}]
@end example

Remove elements from the directory stack.
The elements are numbered from 0 starting at the first directory
listed by @code{dirs};
that is, @code{popd} is equivalent to @code{popd +0}.

When no arguments are given, @code{popd} removes the top directory
from the stack and changes to the new top directory.

Arguments, if supplied, have the following meanings:

@table @code
@item -n
Suppress the normal change of directory when removing directories
from the stack, only manipulate the stack.
@item +@var{N}
Remove the @var{N}th directory (counting from the left of the
list printed by @code{dirs}), starting with zero, from the stack.
@item -@var{N}
Remove the @var{N}th directory (counting from the right of the
list printed by @code{dirs}), starting with zero, from the stack.
@end table

If the top element of the directory stack is modified,
and the @option{-n} option was not supplied,
@code{popd} uses the @code{cd}
builtin to change to the directory at the top of the stack.
If the @code{cd} fails, @code{popd} returns a non-zero value.

Otherwise, @code{popd} returns an unsuccessful status if
an invalid option is specified, the directory stack
is empty, or @var{N} specifies a non-existent directory stack entry.

If the @code{popd} command is successful,
Bash runs @code{dirs} to show the final contents of the directory stack,
and the return status is 0.

@item pushd
@btindex pushd
@example
pushd [-n] [@var{+N} | @var{-N} | @var{dir}]
@end example

Add a directory to the top of the directory stack, or rotate
the stack, making the new top of the stack the current working
directory.
With no arguments, @code{pushd} exchanges the top two elements
of the directory stack.

Arguments, if supplied, have the following meanings:

@table @code
@item -n
Suppress the normal change of directory when rotating or
adding directories to the stack, only manipulate the stack.
@item +@var{N}
Rotate the stack so that
the @var{N}th directory (counting from the left of the
list printed by @code{dirs}, starting with zero) is at the top.
@item -@var{N}
Rotate the stack so that
the @var{N}th directory (counting from the right of the
list printed by @code{dirs}, starting with zero) is at the top.
@item @var{dir}
Make @var{dir} be the top of the stack.
@end table

After the stack has been modified, if the @option{-n} option was not
supplied, @code{pushd} uses the @code{cd} builtin to change to the
directory at the top of the stack.
If the @code{cd} fails, @code{pushd} returns a non-zero value.

Otherwise, if no arguments are supplied, @code{pushd} returns zero
unless the directory stack is empty.
When rotating the directory stack, @code{pushd} returns zero unless
the directory stack is empty or @var{N} specifies a non-existent
directory stack element.

If the @code{pushd} command is successful,
Bash runs @code{dirs} to show the final contents of the directory stack.

@end table

@node Controlling the Prompt
@section Controlling the Prompt
@cindex prompting

In addition, the following table describes the special characters which
can appear in the prompt variables @env{PS0}, @env{PS1}, @env{PS2}, and
@env{PS4}:

@table @code
@item \a
A bell character.
@item \d
The date, in "Weekday Month Date" format (e.g., "Tue May 26").
@item \D@{@var{format}@}
The @var{format} is passed to @code{strftime}(3) and the result is inserted
into the prompt string; an empty @var{format} results in a locale-specific
time representation.
The braces are required.
@item \e
An escape character.
@item \h
The hostname, up to the first @samp{.}.
@item \H
The hostname.
@item \j
The number of jobs currently managed by the shell.
@item \l
The basename of the shell's terminal device name (e.g., "ttys0").
@item \n
A newline.
@item \r
A carriage return.
@item \s
The name of the shell: the basename of @code{$0} (the portion
following the final slash).
@item \t
The time, in 24-hour HH:MM:SS format.
@item \T
The time, in 12-hour HH:MM:SS format.
@item \@@
The time, in 12-hour am/pm format.
@item \A
The time, in 24-hour HH:MM format.
@item \u
The username of the current user.
@item \v
The Bash version (e.g., 2.00).
@item \V
The Bash release, version + patchlevel (e.g., 2.00.0).
@item \w
The value of the @code{PWD} shell variable (@env{$PWD}),
with @env{$HOME} abbreviated with a tilde
(uses the @env{$PROMPT_DIRTRIM} variable).
@item \W
The basename of @env{$PWD}, with @env{$HOME} abbreviated with a tilde.
@item \!
The history number of this command.
@item \#
The command number of this command.
@item \$
If the effective uid is 0, @code{#}, otherwise @code{$}.
@item \@var{nnn}
The character whose ASCII code is the octal value @var{nnn}.
@item \\
A backslash.
@item \[
Begin a sequence of non-printing characters.
This could be used to
embed a terminal control sequence into the prompt.
This escape is only useful when the prompt will be supplied to
Readline, so it shouldn't be used in @env{PS0} or @env{PS4},
or when line editing is not enabled.

@item \]
End a sequence of non-printing characters begun with @samp{\[}
@end table

The command number and the history number are usually different:
the history number of a command is its position in the history
list, which may include commands restored from the history file
(@pxref{Bash History Facilities}), while the command number is
the position in the sequence of commands executed during the current
shell session.

After the string is decoded, it is expanded via
parameter expansion, command substitution, arithmetic
expansion, and quote removal, subject to the value of the
@code{promptvars} shell option (@pxref{The Shopt Builtin}).
This can have unwanted side effects if escaped portions of the string
appear within command substitution or contain characters special to
word expansion.

@node The Restricted Shell
@section The Restricted Shell
@cindex restricted shell

If Bash is started with the name @code{rbash}, or the
@option{--restricted}
or
@option{-r}
option is supplied at invocation, the shell becomes @var{restricted}.
A restricted shell is used to
set up an environment more controlled than the standard shell.
A restricted shell behaves identically to @code{bash}
with the exception that the following are disallowed or not performed:

@itemize @bullet
@item
Changing directories with the @code{cd} builtin.
@item
Setting or unsetting the values of the @env{SHELL}, @env{PATH},
@env{HISTFILE},
@env{ENV}, or @env{BASH_ENV} variables.
@item
Specifying command names containing slashes.
@item
Specifying a filename containing a slash as an argument to the @code{.}
builtin command.
@item
Using the @option{-p} option to the @code{.} builtin command
to specify a search path.
@item
Specifying a filename containing a slash as an argument to the @code{history}
builtin command.
@item
Specifying a filename containing a slash as an argument to the @option{-p}
option to the @code{hash} builtin command.
@item
Importing function definitions from the shell environment at startup.
@item
Parsing the value of @env{SHELLOPTS} from the shell environment at startup.
@item
Redirecting output using the @samp{>}, @samp{>|}, @samp{<>}, @samp{>&},
@samp{&>}, and @samp{>>} redirection operators.
@item
Using the @code{exec} builtin to replace the shell with another command.
@item
Adding or deleting builtin commands with the
@option{-f} and @option{-d} options to the @code{enable} builtin.
@item
Using the @code{enable} builtin command to enable disabled shell builtins.
@item
Specifying the @option{-p} option to the @code{command} builtin.
@item
Turning off restricted mode with @samp{set +r} or @samp{shopt -u restricted_shell}.
@end itemize

These restrictions are enforced after any startup files are read.

When a command that is found to be a shell script is executed
(@pxref{Shell Scripts}), @code{rbash} turns off any restrictions in
the shell spawned to execute the script.

The restricted shell mode is only one component of a useful restricted
environment. It should be accompanied by setting @env{PATH} to a value
that allows execution of only a few verified commands (commands that
allow shell escapes are particularly vulnerable), changing the current
directory to a non-writable directory other than @env{$HOME} after login,
not allowing the restricted shell to execute shell scripts, and cleaning
the environment of variables that cause some commands to modify their
behavior (e.g., @env{VISUAL} or @env{PAGER}).
When setting up a restricted environment like this, it's important not
to install or allow symbolic links in the new current directory, since
those could be used to circumvent restrictions on writing to files.

Modern systems provide more secure ways to implement a restricted environment,
such as @code{jails}, @code{zones}, or @code{containers}.


@node Bash POSIX Mode
@section Bash and POSIX

@subsection What is POSIX?
@cindex POSIX description

@sc{posix}
is the name for a family of standards based on Unix.
A number of Unix services, tools, and functions are part of the standard,
ranging from the basic system calls and C library functions to common
applications and tools to system administration and management.

The @sc{posix} Shell and Utilities standard was originally developed by
IEEE Working Group 1003.2 (POSIX.2).
The first edition of the 1003.2 standard was published in 1992.
It was merged with the original IEEE 1003.1 Working Group and is
currently maintained by the Austin Group (a joint working group of the
IEEE, The Open Group and ISO/IEC SC22/WG15).
Today the Shell and Utilities are a volume within the set of documents that
make up IEEE Std 1003.1-2024, and thus the former POSIX.2 (from 1992)
is now part of the current unified @sc{posix} standard.

The Shell and Utilities volume concentrates on the command
interpreter interface and utility programs commonly executed from
the command line or by other programs.
The standard is freely available on the web at
@url{https://pubs.opengroup.org/onlinepubs/9799919799/utilities/contents.html}.

Bash is concerned with the aspects of the shell's behavior defined
by the @sc{posix} Shell and Utilities volume.
The shell command
language has of course been standardized, including the basic flow
control and program execution constructs, I/O redirection and
pipelines, argument handling, variable expansion, and quoting.

The @i{special} builtins, which must be implemented as part of the
shell to provide the desired functionality, are specified as
being part of the shell; examples of these are @code{eval} and
@code{export}.
Other utilities appear in the sections of @sc{posix} not
devoted to the shell which are commonly (and in some cases must
be) implemented as builtin commands, such as
@code{read} and @code{test}.
@sc{posix} also specifies aspects of the shell's interactive
behavior, including job control and command
line editing.
Only vi-style line editing commands have been standardized;
emacs editing commands were left out due to objections.

@subsection Bash POSIX Mode
@cindex POSIX Mode

Although Bash is an implementation of the @sc{posix} shell
specification, there are areas where the Bash default behavior
differs from the specification.
The Bash @dfn{posix mode} changes the Bash
behavior in these areas so that it conforms more strictly
to the standard.

Starting Bash with the @option{--posix} or @option{-o posix}
command-line option or executing
@samp{set -o posix} while Bash is running will cause Bash to conform more
closely to the @sc{posix} standard by changing the behavior to
match that specified by @sc{posix} in areas where the Bash default differs.

When invoked as @code{sh}, Bash enters @sc{posix} mode after reading the
startup files.

The following list is what's changed when @sc{posix} mode is in effect:

@enumerate
@item
Bash ensures that the @env{POSIXLY_CORRECT} variable is set.

@item
Bash reads and executes the @sc{posix} startup files
(@env{$ENV}) rather than
the normal Bash files (@pxref{Bash Startup Files}).

@item
Alias expansion is always enabled, even in non-interactive shells.

@item
Reserved words appearing in a context where reserved words are recognized
do not undergo alias expansion.

@item
Alias expansion is performed when initially parsing a command substitution.
The default (non-posix)
mode generally defers it, when enabled, until the command
substitution is executed.
This means that command substitution will not
expand aliases that are defined after the command substitution is initially
parsed (e.g., as part of a function definition).

@item
The @code{time} reserved word may be used by itself as a simple command.
When used in this way, it displays timing statistics for the shell
and its completed children.
The @env{TIMEFORMAT} variable controls the format of the timing information.

@item
The parser does not recognize @code{time} as a reserved word if the next
token begins with a @samp{-}.

@item
When parsing and expanding a $@{@dots{}@} expansion that appears within
double quotes, single quotes are no longer special and cannot be used to
quote a closing brace or other special character, unless the operator is
one of those defined to perform pattern removal.
In this case, they do not have to appear as matched pairs.

@ignore
@item
When parsing @code{$()} command substitutions containing here-documents,
the parser does not allow a here-document to be delimited by the closing
right parenthesis.
The newline after the here-document delimiter is required.
@end ignore

@item
Redirection operators do not perform filename expansion on the word
in a redirection unless the shell is interactive.

@item
Redirection operators do not perform word splitting on the word in a
redirection.

@item
Function names may not be the same as one of the @sc{posix} special
builtins.

@item
Tilde expansion is only performed on assignments preceding a command
name, rather than on all assignment statements on the line, unless
the command is a declaration command.

@item
While variable indirection is available, it may not be applied to the
@samp{#} and @samp{?} special parameters.

@item
Expanding the @samp{*} special parameter in a pattern context where the
expansion is double-quoted does not treat the @code{$*} as if it were
double-quoted.

@item
A double quote character (@samp{"}) is treated specially when it appears
in a backquoted command substitution in the body of a here-document that
undergoes expansion.
That means, for example, that a backslash preceding a double quote
character will escape it and the backslash will be removed.

@item
Command substitutions don't set the @samp{?} special parameter.
The exit
status of a simple command without a command word is still the exit status
of the last command substitution that occurred while evaluating the variable
assignments and redirections in that command, but that does not happen until
after all of the assignments and redirections.

@item
Literal tildes that appear as the first character in elements of
the @env{PATH} variable are not expanded as described above
under @ref{Tilde Expansion}.

@item
Command lookup finds @sc{posix} special builtins before shell functions,
including output printed by the @code{type} and @code{command} builtins.

@item
Even if a shell function whose name contains a slash was defined before
entering @sc{posix} mode, the shell will not execute a function whose name
contains one or more slashes.

@item
When a command in the hash table no longer exists, Bash will re-search
@env{$PATH} to find the new location.
This is also available with @samp{shopt -s checkhash}.

@item
Bash will not insert a command without the execute bit set into the
command hash table, even if it returns it as a (last-ditch) result
from a @env{$PATH} search.

@item
Normally, when job control is not enabled,
the shell implicitly redirects the standard input of
asynchronous commands from @file{/dev/null}.
A redirection to the standard input in this command inhibits this
implicit redirection.
In @sc{posix} mode, a redirection that redirects file descriptor 0
to itself (e.g., @samp{<&0}) does not count as a redirection that
overrides the implicit redirection from @file{/dev/null}.

@item
The message printed by the job control code and builtins when a job
exits with a non-zero status is ``Done(status)''.

@item
The message printed by the job control code and builtins when a job
is stopped is ``Stopped(@var{signame})'', where @var{signame} is, for
example, @code{SIGTSTP}.

@item
If the shell is interactive, Bash does not perform job notifications
between executing commands in lists separated by @samp{;} or newline.
Non-interactive shells print status messages after a foreground job
in a list completes.

@item
If the shell is interactive, Bash waits until the next prompt before
printing the status of a background job that changes status or a foreground
job that terminates due to a signal.
Non-interactive shells print status messages after a foreground job
completes.

@item
Bash permanently removes jobs from the jobs table after notifying the
user of their termination via the @code{wait} or @code{jobs} builtins.
It removes the job from the jobs list after notifying the user of its
termination, but the status is still available via @code{wait}, as long
as @code{wait} is supplied a @sc{pid} argument.

@item
The @code{vi} editing mode will invoke the @code{vi} editor directly when
the @samp{v} command is run, instead of checking @code{$VISUAL} and
@code{$EDITOR}.

@item
Prompt expansion enables the
@sc{posix} @env{PS1} and @env{PS2} expansions of @samp{!} to
the history number and @samp{!!} to @samp{!},
and Bash performs parameter expansion on the values of @env{PS1} and
@env{PS2} regardless of the setting of the @code{promptvars} option.

@item
The default history file is @file{~/.sh_history} (this is the
default value the shell assigns to @env{$HISTFILE}).

@item
The @samp{!} character does not introduce history expansion within a
double-quoted string, even if the @code{histexpand} option is enabled.

@item
When printing shell function definitions (e.g., by @code{type}), Bash does
not print the @code{function} reserved word unless necessary.

@item
Non-interactive shells exit if a syntax error in an arithmetic expansion
results in an invalid expression.

@item
Non-interactive shells exit if a parameter expansion error occurs.

@item
If a @sc{posix} special builtin returns an error status, a
non-interactive shell exits.
The fatal errors are those listed in
the @sc{posix} standard, and include things like passing incorrect options,
redirection errors, variable assignment errors for assignments preceding
the command name, and so on.

@item
A non-interactive shell exits with an error status if a variable
assignment error occurs when no command name follows the assignment
statements.
A variable assignment error occurs, for example, when trying to assign
a value to a readonly variable.

@item
A non-interactive shell exits with an error status if a variable
assignment error occurs in an assignment statement preceding a special
builtin, but not with any other simple command.
For any other simple
command, the shell aborts execution of that command, and execution continues
at the top level ("the shell shall not perform any further processing of the
command in which the error occurred").

@item
A non-interactive shell exits with an error status if the iteration
variable in a @code{for} statement or the selection variable in a
@code{select} statement is a readonly variable or has an invalid name.

@item
Non-interactive shells exit if @var{filename} in @code{.} @var{filename}
is not found.

@item
Non-interactive shells exit if there is a syntax error in a script read
with the @code{.} or @code{source} builtins, or in a string processed by
the @code{eval} builtin.

@item
Non-interactive shells exit
if the @code{export}, @code{readonly} or @code{unset}
builtin commands get an argument
that is not a valid identifier, and they are not operating on shell
functions.
These errors force an exit because these are special builtins.

@item
Assignment statements preceding @sc{posix} special builtins
persist in the shell environment after the builtin completes.

@item
The @code{command} builtin does not prevent builtins that take assignment
statements as arguments from expanding them as assignment statements;
when not in @sc{posix} mode, declaration commands lose their assignment
statement expansion properties when preceded by @code{command}.

@item
Enabling @sc{posix} mode has the effect of setting the
@code{inherit_errexit} option, so
subshells spawned to execute command substitutions inherit the value of
the @option{-e} option from the parent shell.
When the @code{inherit_errexit} option is not enabled,
Bash clears the @option{-e} option in such subshells.

@item
Enabling @sc{posix} mode has the effect of setting the
@code{shift_verbose} option, so numeric arguments to @code{shift}
that exceed the number of positional parameters will result in an
error message.

@item
Enabling @sc{posix} mode has the effect of setting the
@code{interactive_comments} option (@pxref{Comments}).

@item
The @code{.} and @code{source} builtins do not search the current directory
for the filename argument if it is not found by searching @env{PATH}.

@item
When the @code{alias} builtin displays alias definitions, it does not
display them with a leading @samp{alias } unless the @option{-p} option
is supplied.

@item
The @code{bg} builtin uses the required format to describe each job placed
in the background, which does not include an indication of whether the job
is the current or previous job.

@item
When the @code{cd} builtin is invoked in logical mode, and the pathname
constructed from @code{$PWD} and the directory name supplied as an argument
does not refer to an existing directory, @code{cd} will fail instead of
falling back to physical mode.

@item
When the @code{cd} builtin cannot change a directory because the
length of the pathname 
constructed from @code{$PWD} and the directory name supplied as an argument
exceeds @code{PATH_MAX} when canonicalized, @code{cd} will
attempt to use the supplied directory name.

@item
When the @code{xpg_echo} option is enabled, Bash does not attempt to
interpret any arguments to @code{echo} as options.
@code{echo} displays each argument after converting escape sequences.

@item
The @code{export} and @code{readonly} builtin commands display their
output in the format required by @sc{posix}.

@item
When listing the history, the @code{fc} builtin does not include an
indication of whether or not a history entry has been modified.

@item
The default editor used by @code{fc} is @code{ed}.

@item
@code{fc} treats extra arguments as an error instead of ignoring them.

@item
If there are too many arguments supplied to @code{fc -s}, @code{fc} prints
an error message and returns failure.

@item
The output of @samp{kill -l} prints all the signal names on a single line,
separated by spaces, without the @samp{SIG} prefix.

@item
The @code{kill} builtin does not accept signal names with a @samp{SIG}
prefix.

@item
The @code{kill} builtin returns a failure status if any of the pid or job
arguments are invalid or if sending the specified signal to any of them
fails.
In default mode, @code{kill} returns success if the signal was
successfully sent to any of the specified processes.

@item
The @code{printf} builtin uses @code{double} (via @code{strtod}) to convert
arguments corresponding to floating point conversion specifiers, instead of
@code{long double} if it's available.
The @samp{L} length modifier forces
@code{printf} to use @code{long double} if it's available.

@item
The @code{pwd} builtin verifies that the value it prints is the same as the
current directory, even if it is not asked to check the file system with the
@option{-P} option.

@item
The @code{read} builtin may be interrupted by a signal for which a trap
has been set.
If Bash receives a trapped signal while executing @code{read}, the trap
handler executes and @code{read} returns an exit status greater than 128.

@item
When the @code{set} builtin is invoked without options, it does not display
shell function names and definitions.

@item
When the @code{set} builtin is invoked without options, it displays
variable values without quotes, unless they contain shell metacharacters,
even if the result contains nonprinting characters.

@item
The @code{test} builtin compares strings using the current locale when
evaluating the @samp{<} and @samp{>} binary operators.

@item
The @code{test} builtin's @option{-t} unary primary requires an argument.
Historical versions of @code{test} made the argument optional in certain
cases, and Bash attempts to accommodate those for backwards compatibility.

@item
The @code{trap} builtin displays signal names without the leading
@code{SIG}.

@item
The @code{trap} builtin doesn't check the first argument for a possible
signal specification and revert the signal handling to the original
disposition if it is, unless that argument consists solely of digits and
is a valid signal number.
If users want to reset the handler for a given
signal to the original disposition, they should use @samp{-} as the
first argument.

@item
@code{trap -p} without arguments displays signals whose dispositions are
set to SIG_DFL and those that were ignored when the shell started, not
just trapped signals.

@item
The @code{type} and @code{command} builtins will not report a non-executable
file as having been found, though the shell will attempt to execute such a
file if it is the only so-named file found in @code{$PATH}.

@item
The @code{ulimit} builtin uses a block size of 512 bytes for the @option{-c}
and @option{-f} options.

@item
The @code{unset} builtin with the @option{-v} option specified returns a
fatal error if it attempts to unset a @code{readonly} or @code{non-unsettable}
variable,
which causes a non-interactive shell to exit.

@item
When asked to unset a variable that appears in an assignment statement
preceding the command, the @code{unset} builtin attempts to unset a variable
of the same name in the current or previous scope as well.
This implements the required "if an assigned variable is further modified
by the utility, the modifications made by the utility shall persist" behavior.

@item
The arrival of @code{SIGCHLD}  when a trap is set on @code{SIGCHLD} does
not interrupt the @code{wait} builtin and cause it to return immediately.
The trap command is run once for each child that exits.

@item
Bash removes an exited background process's status from the list of such
statuses after the @code{wait} builtin returns it.

@end enumerate

There is additional @sc{posix} behavior that Bash does not implement by
default even when in @sc{posix} mode.
Specifically:

@enumerate

@item
@sc{posix} requires that word splitting be byte-oriented.
That is, each @emph{byte} in the value of @env{IFS} potentially splits a
word, even if that byte is part of a multibyte character in @env{IFS}
or part of multibyte character in the word.
Bash allows multibyte characters in the value of @env{IFS}, treating
a valid multibyte character as a single delimiter, and will not
split a valid multibyte character even if one of the bytes composing that
character appears in @env{IFS}.
This is @sc{posix} interpretation 1560, further modified by issue 1924.

@item
The @code{fc} builtin checks @code{$EDITOR} as a program to edit history
entries if @code{FCEDIT} is unset, rather than defaulting directly to
@code{ed}.
@code{fc} uses @code{ed} if @code{EDITOR} is unset.

@ignore
@item
A non-interactive shell does not exit if a variable assignment preceding
the @code{command} builtin or another non-special builtin fails.
There is ambiguity in @sc{posix} about this.
@end ignore

@item
Bash does not perform redirections before expanding variable assignments
preceding a simple command; it does not allow the redirections access to
the results of the variable assignments, but any side effects of expanding
the redirections take place first.
If compiled in @i{strict posix} mode, Bash performs redirections first.

@item
As noted above, Bash requires the @code{xpg_echo} option to be enabled for
the @code{echo} builtin to be fully conformant.

@end enumerate

Bash can be configured to be @sc{posix}-conformant by default, by specifying
the @option{--enable-strict-posix-default} to @code{configure} when building
(@pxref{Optional Features}).

@node Shell Compatibility Mode
@section Shell Compatibility Mode
@cindex Compatibility Level
@cindex Compatibility Mode

Bash-4.0 introduced the concept of a @dfn{shell compatibility level},
specified as a set of options to the shopt builtin
(@code{compat31},
@code{compat32},
@code{compat40},
@code{compat41},
and so on).
There is only one current
compatibility level -- each option is mutually exclusive.
The compatibility level is intended to allow users to select behavior
from previous versions that is incompatible with newer versions
while they migrate scripts to use current features and behavior.
It's intended to be a temporary solution.

This section does not mention behavior that is standard for a particular
version (e.g., setting @code{compat32} means that quoting the right hand
side of the regexp
matching operator quotes special regexp characters in the word, which is
default behavior in bash-3.2 and subsequent versions). 

If a user enables, say, @code{compat32}, it may affect the behavior of other
compatibility levels up to and including the current compatibility level.
The idea is that each compatibility level controls behavior that changed
in that version of Bash,
but that behavior may have been present in earlier versions.
For instance, the change to use locale-based comparisons with the @code{[[}
command came in bash-4.1, and earlier versions used ASCII-based comparisons,
so enabling @code{compat32} will enable ASCII-based comparisons as well.
That granularity may not be sufficient for
all uses, and as a result users should employ compatibility levels carefully.
Read the documentation for a particular feature to find out the
current behavior.

Bash-4.3 introduced a new shell variable: @env{BASH_COMPAT}.
The value assigned
to this variable (a decimal version number like 4.2, or an integer
corresponding to the @code{compat}@var{NN} option, like 42) determines the
compatibility level.

Starting with bash-4.4, Bash began deprecating older compatibility
levels.
Eventually, the options will be removed in favor of @env{BASH_COMPAT}.

Bash-5.0 was the final version for which there was an individual shopt
option for the previous version.
@env{BASH_COMPAT} is the only mechanism to control the compatibility level
in versions newer than bash-5.0.

The following table describes the behavior changes controlled by each
compatibility level setting.
The @code{compat}@var{NN} tag is used as shorthand for setting the
compatibility level
to @var{NN} using one of the following mechanisms.
For versions prior to bash-5.0, the compatibility level may be set using
the corresponding @code{compat}@var{NN} shopt option.
For bash-4.3 and later versions, the @env{BASH_COMPAT} variable is preferred,
and it is required for bash-5.1 and later versions.

@table @code
@item compat31
@itemize @bullet
@item
Quoting the rhs of the @code{[[} command's regexp matching operator (=~)
has no special effect
@end itemize

@item compat40
@itemize @bullet
@item
The @samp{<} and @samp{>} operators to the @code{[[} command do not
consider the current locale when comparing strings; they use ASCII
ordering.
Bash versions prior to bash-4.1 use ASCII collation and strcmp(3);
bash-4.1 and later use the current locale's collation sequence and
strcoll(3).
@end itemize

@item compat41
@itemize @bullet
@item
In @sc{posix} mode, @code{time} may be followed by options and still be
recognized as a reserved word (this is
@sc{posix}
interpretation 267).
@item
In @sc{posix} mode, the parser requires that an even number of single
quotes occur in the @var{word} portion of a double-quoted $@{@dots{}@}
parameter expansion and treats them specially, so that characters within
the single quotes are considered quoted
(this is
@sc{posix}
interpretation 221).
@end itemize

@item compat42
@itemize @bullet
@item
The replacement string in double-quoted pattern substitution does not
undergo quote removal, as it does in versions after bash-4.2.
@item
In @sc{posix} mode, single quotes are considered special when expanding
the @var{word} portion of a double-quoted $@{@dots{}@} parameter expansion
and can be used to quote a closing brace or other special character
(this is part of
@sc{posix}
interpretation 221);
in later versions, single quotes
are not special within double-quoted word expansions.
@end itemize

@item compat43
@itemize @bullet
@item
Word expansion errors are considered non-fatal errors that cause the
current command to fail, even in @sc{posix} mode
(the default behavior is to make them fatal errors that cause the shell
to exit).
@item
When executing a shell function, the loop state (while/until/etc.)
is not reset, so @code{break} or @code{continue} in that function will break
or continue loops in the calling context.
Bash-4.4 and later reset the loop state to prevent this.
@end itemize

@item compat44
@itemize @bullet
@item
The shell sets up the values used by @env{BASH_ARGV} and @env{BASH_ARGC}
so they can expand to the shell's positional parameters even if extended
debugging mode is not enabled.
@item
A subshell inherits loops from its parent context, so @code{break}
or @code{continue} will cause the subshell to exit.
Bash-5.0 and later reset the loop state to prevent the exit.
@item
Variable assignments preceding builtins like @code{export} and @code{readonly}
that set attributes continue to affect variables with the same
name in the calling environment even if the shell is not in
@sc{posix}
mode.
@end itemize

@item compat50 (set using BASH_COMPAT)
@itemize @bullet
@item
Bash-5.1 changed the way @code{$RANDOM} is generated to introduce slightly
more randomness.
If the shell compatibility level is set to 50 or lower, it reverts to
the method from bash-5.0 and previous versions,
so seeding the random number generator by assigning a value to
@env{RANDOM} will produce the same sequence as in bash-5.0.
@item
If the command hash table is empty, Bash versions prior to bash-5.1
printed an informational message to that effect, even when producing
output that can be reused as input.
Bash-5.1 suppresses that message when the @option{-l} option is supplied.
@end itemize

@item compat51 (set using BASH_COMPAT)
@itemize @bullet
@item
The @code{unset} builtin will unset the array @code{a} given an argument like
@samp{a[@@]}.
Bash-5.2 will unset an element with key @samp{@@} (associative arrays)
or remove all the elements without unsetting the array (indexed arrays).
@item
Arithmetic commands ( ((@dots{})) ) and the expressions in an arithmetic for
statement can be expanded more than once.
@item
Expressions used as arguments to arithmetic operators in the @code{[[}
conditional command can be expanded more than once.
@item
The expressions in substring parameter brace expansion can be
expanded more than once.
@item
The expressions in the $(( @dots{} )) word expansion can be expanded
more than once.
@item
Arithmetic expressions used as indexed array subscripts can be
expanded more than once.
@item
@code{test -v}, when given an argument of @samp{A[@@]}, where @var{A} is
an existing associative array, will return true if the array has any set
elements.
Bash-5.2 will look for and report on a key named @samp{@@}.
@item
the $@{@var{parameter}[:]=@var{value}@} word expansion will return
@var{value}, before any variable-specific transformations have been
performed (e.g., converting to lowercase).
Bash-5.2 will return the final value assigned to the variable.
@item
Parsing command substitutions will behave as if extended globbing
(@pxref{The Shopt Builtin})
is enabled, so that parsing a command substitution containing an extglob
pattern (say, as part of a shell function) will not fail.
This assumes the intent is to enable extglob before the command is executed
and word expansions are performed.
It will fail at word expansion time if extglob hasn't been
enabled by the time the command is executed.
@end itemize

@item compat52 (set using BASH_COMPAT)
@itemize @bullet
@item
The @code{test} builtin uses its historical algorithm to parse parenthesized
subexpressions when given five or more arguments.
@item
If the @option{-p} or @option{-P} option is supplied to the @code{bind} builtin,
@code{bind} treats any arguments remaining after option processing
as bindable command names, and
displays any key sequences bound to those commands, instead of treating
the arguments as key sequences to bind.
@item
Interactive shells will notify the user of completed jobs while sourcing a
script.
Newer versions defer notification until script execution completes.
@ignore
@item
Bash will not try to execute a shell function whose name contains a slash.
Previous versions disallowed this in @sc{posix} mode but allowed it by
default.
@end ignore
@end itemize

@end table

@node Job Control
@chapter Job Control

This chapter discusses what job control is, how it works, and how
Bash allows you to access its facilities.

@menu
* Job Control Basics::		How job control works.
* Job Control Builtins::	Bash builtin commands used to interact
				with job control.
* Job Control Variables::	Variables Bash uses to customize job
				control.
@end menu

@node Job Control Basics
@section Job Control Basics
@cindex job control
@cindex foreground
@cindex background
@cindex suspending jobs

Job control
refers to the ability to selectively stop (suspend)
the execution of processes and continue (resume)
their execution at a later point.
A user typically employs
this facility via an interactive interface supplied jointly
by the operating system kernel's terminal driver and Bash.

The shell associates a @var{job} with each pipeline.
It keeps a
table of currently executing jobs, which the
@code{jobs} command will display.
Each job has a @dfn{job number}, which @code{jobs} displays between brackets.
Job numbers start at 1.
When Bash starts a job asynchronously, it prints a line that looks
like:
@example
[1] 25647
@end example
@noindent
indicating that this job is job number 1 and that the process @sc{id}
of the last process in the pipeline associated with this job is
25647.
All of the processes in a single pipeline are members of
the same job.
Bash uses the @var{job} abstraction as the basis for job control. 

To facilitate the implementation of the user interface to job control,
each process has a @dfn{process group @sc{id}}, and
the operating system maintains the notion of a current terminal
process group @sc{id}.
This terminal process group @sc{id} is associated with the
@dfn{controlling terminal}.

Processes that have the same process group ID are said to be part of
the same @dfn{process group}.
Members of the foreground process group (processes whose
process group @sc{id} is equal to the current terminal process group
@sc{id}) receive keyboard-generated signals such as @code{SIGINT}. 
Processes in the foreground process group are said to be
foreground processes.
Background processes
are those whose process group @sc{id} differs from the
controlling terminal's;
such processes are immune to keyboard-generated signals.
Only foreground processes are allowed to read from or,
if the user so specifies with
@code{stty tostop},
write to the controlling terminal.
The system sends a
@code{SIGTTIN} (@code{SIGTTOU})
signal to background processes which attempt to
read from (write to when @code{tostop} is in effect)
the terminal,
which, unless caught, suspends the process. 

If the operating system on which Bash is running supports
job control, Bash contains facilities to use it.
Typing the
@dfn{suspend} character (typically @samp{^Z}, Control-Z) while a
process is running stops that process
and returns control to Bash.
Typing the @dfn{delayed suspend} character
(typically @samp{^Y}, Control-Y) causes the process to stop
when it attempts to read input from the terminal,
and returns control to Bash.
The user then manipulates the state of
this job, using
the @code{bg} command to continue it in the background,
the @code{fg} command to continue it in the foreground, or
the @code{kill} command to kill it.
The suspend character
takes effect immediately, and has the additional side effect of
discarding any pending output and typeahead.
If you want to force a background process to stop, or stop a process
that's not associated with your terminal session,
send it the @code{SIGSTOP} signal using @code{kill}.

There are a number of ways to refer to a job in the shell.
The @samp{%} character introduces a @dfn{job specification} (jobspec).

Job number @code{n} may be referred to as @samp{%n}.
A job may also be referred to
using a prefix of the name used to start it,
or using a substring that appears in its command line.
For example, @samp{%ce} refers
to a job whose command name begins with @samp{ce}.
Using @samp{%?ce}, on the
other hand, refers to any job containing the string @samp{ce} in
its command line.
If the prefix or substring matches more than one job,
Bash reports an error.

The symbols @samp{%%} and  @samp{%+} refer to the shell's notion of the
@dfn{current job}.
A single @samp{%} (with no accompanying job specification) also refers
to the current job.
@samp{%-} refers to the @dfn{previous job}.
When a job starts in the background,
a job stops while in the foreground,
or a job is resumed in the background,
it becomes the current job.
The job that was the current job becomes the previous job.
When the current job terminates, the previous job becomes the
current job.
If there is only a single job, @samp{%+} and @samp{%-} can both be used
to refer to that job.
In output pertaining to jobs (e.g., the output of the @code{jobs}
command), the current job is always marked with a @samp{+}, and the
previous job with a @samp{-}. 

Simply naming a job can be used to bring it into the foreground:
@samp{%1} is a synonym for @samp{fg %1}, bringing job 1 from the
background into the foreground.
Similarly, @samp{%1 &} resumes
job 1 in the background, equivalent to @samp{bg %1}.

The shell learns immediately whenever a job changes state. 
Normally, Bash waits until it is about to print a prompt before
notifying the user about
changes in a job's status so as to not interrupt
any other output,
though it will notify of changes in a job's status after a
foreground command in
a list completes, before executing the next command in the list.
If the @option{-b} option to the @code{set} builtin is enabled,
Bash reports status changes immediately (@pxref{The Set Builtin}).
Bash executes any trap on @code{SIGCHLD}
for each child process that terminates.

When a job terminates and Bash notifies the user about it,
Bash removes the job from the jobs table.
It will not appear in @code{jobs} output, but @code{wait} will
report its exit status, as long as it's supplied the process ID
associated with the job as an argument.
When the table is empty, job numbers start over at 1.

If a user attempts to exit
Bash while jobs are stopped, (or running, if
the @code{checkjobs} option is enabled -- see @ref{The Shopt Builtin}), the
shell prints a warning message, and if the @code{checkjobs} option is
enabled, lists the jobs and their statuses.
The @code{jobs} command may then be used to inspect their status.
If the user immediately attempts to exit again,
without an intervening command,
Bash does not print another warning, and
terminates any stopped jobs.

When the shell is waiting for a job or process using the @code{wait}
builtin, and job control is enabled, @code{wait} will return when the
job changes state.
The @option{-f} option causes @code{wait} to wait
until the job or process terminates before returning.

@node Job Control Builtins
@section Job Control Builtins

@table @code

@item bg
@btindex bg
@example
bg [@var{jobspec} @dots{}]
@end example

Resume each suspended job @var{jobspec} in the background, as if it
had been started with @samp{&}.
If @var{jobspec} is not supplied, the shell uses its
notion of the current job.
@code{bg} returns  zero unless it is run when job control is not
enabled, or, when run with job control enabled, any
@var{jobspec} was not found or specifies a job
that was started without job control.

@item fg
@btindex fg
@example
fg [@var{jobspec}]
@end example

Resume the job @var{jobspec} in the foreground and make it the current job.
If @var{jobspec} is not supplied, @code{fg} resumes the current job.
The return status is that of the command placed into the foreground,
or non-zero if run when job control is disabled or, when run with
job control enabled, @var{jobspec} does not specify a valid job or
@var{jobspec} specifies a job that was started without job control.

@item jobs
@btindex jobs
@example
jobs [-lnprs] [@var{jobspec}]
jobs -x @var{command} [@var{arguments}]
@end example

The first form lists the active jobs.
The options have the following meanings:

@table @code
@item -l
List process @sc{id}s in addition to the normal information.

@item -n
Display information only about jobs that have changed status since
the user was last notified of their status.

@item -p
List only the process @sc{id} of the job's process group leader.

@item -r
Display only running jobs.

@item -s
Display only stopped jobs.
@end table

If @var{jobspec} is supplied,
@code{jobs} restricts output to information about that job. 
If @var{jobspec} is not supplied, @code{jobs} lists the status of all jobs.
The return status is zero unless an invalid option is encountered
or an invalid
@var{jobspec}
is supplied.

If the @option{-x} option is supplied, @code{jobs} replaces any
@var{jobspec} found in @var{command} or @var{arguments} with the
corresponding process group @sc{id}, and executes @var{command},
passing it @var{argument}s, returning its exit status. 

@item kill
@btindex kill
@example
kill [-s @var{sigspec}] [-n @var{signum}] [-@var{sigspec}] @var{id} [@dots{}]
kill -l|-L [@var{exit_status}]
@end example

Send a signal specified by @var{sigspec} or @var{signum} to the processes
named by each @var{id}.
Each @var{id} may be a
job specification @var{jobspec} or process @sc{id} @var{pid}.
@var{sigspec} is either a case-insensitive signal name such as
@code{SIGINT} (with or without the @code{SIG} prefix)
or a signal number; @var{signum} is a signal number.
If @var{sigspec} and @var{signum} are not present, @code{kill}
sends @code{SIGTERM}.

The @option{-l} option lists the signal names.
If any arguments are supplied when @option{-l} is supplied,
@code{kill} lists the names of the signals corresponding to the arguments,
and the return status is zero.
@var{exit_status} is a number specifying a signal number or the exit
status of a process terminated by a signal;
if it is supplied, @code{kill} prints the name of the signal that caused
the process to terminate.
@code{kill} assumes that process exit statuses are greater than 128;
anything less than that is a signal number.
The @option{-L} option is equivalent to @option{-l}.

The return status is zero if at least one signal was successfully sent,
or non-zero if an error occurs or an invalid option is encountered.

@item wait
@btindex wait
@example
wait [-fn] [-p @var{varname}] [@var{id} @dots{}]
@end example

Wait until the child process specified by each @var{id} exits and
return the exit status of the last @var{id}.
Each @var{id} may be a process @sc{id} @var{pid}
or a job specification @var{jobspec};
if a jobspec is supplied, @code{wait} waits for all processes in the job.

If no options or @var{id}s are supplied,
@code{wait} waits for all running background jobs and
the last-executed process substitution,
if its process id is the same as @var{$!},
and the return status is zero.

If the @option{-n} option is supplied, @code{wait} waits for any one of
the @var{id}s or,
if no @var{id}s are supplied, any job or process substitution,
to complete and returns its exit status.
If none of the supplied @var{id}s is a child of the shell,
or if no arguments are supplied and the shell has no unwaited-for children,
the exit status is 127.

If the @option{-p} option is supplied, @code{wait} assigns
the process or job identifier of the job
for which the exit status is returned to the
variable @var{varname} named by the option argument.
The variable,
which cannot be readonly,
will be unset initially, before any assignment.
This is useful only when used with the @option{-n} option.

Supplying the @option{-f} option, when job control is enabled,
forces @code{wait} to wait for each @var{id} to terminate before
returning its status, instead of returning when it changes status.
If there are no @var{id} arguments,
@code{wait} waits until all background processes have terminated.

If none of the @var{id}s specify one of the shell's an active child
processes, the return status is 127.
If @code{wait} is interrupted by a signal,
any @var{varname} will remain unset,
and the return status will be greater
than 128, as described above (@pxref{Signals}).
Otherwise, the return status is the exit  status of the last @var{id}.

@item disown
@btindex disown
@example
disown [-ar] [-h] [@var{id} @dots{}]
@end example

Without options, remove each @var{id} from the table of
active jobs.
Each @var{id} may be a job specification @var{jobspec}
or a process @sc{id} @var{pid};
if @var{id} is a @var{pid},
@code{disown} uses the job containing @var{pid} as @var{jobspec}.

If the @option{-h} option is supplied,
@code{disown} does not remove the jobs corresponding to each @code{id}
from the jobs table,
but rather marks them so the shell does not send
@code{SIGHUP}
to the job if the shell receives a
@code{SIGHUP}.

If no @var{id} is supplied, the @option{-a} option means to remove or
mark all jobs; the @option{-r} option without an @var{id}
argument removes or marks running jobs.
If no @var{id} is supplied,
and neither the @option{-a} nor the @option{-r} option is supplied,
@code{disown} removes or marks the current job.

The return value is 0 unless an @var{id} does not specify a valid job.

@item suspend
@btindex suspend
@example
suspend [-f]
@end example

Suspend the execution of this shell until it receives a
@code{SIGCONT} signal.
A login shell, or a shell without job control enabled,
cannot be suspended; the @option{-f}
option will override this and force the suspension.
The return status is 0 unless the shell is a login shell
or job control is not enabled
and
@option{-f}
is not supplied.

@end table

When job control is not active, the @code{kill} and @code{wait}
builtins do not accept @var{jobspec} arguments.
They must be supplied process @sc{id}s.

@node Job Control Variables
@section Job Control Variables

@vtable @code

@item auto_resume
This variable controls how the shell interacts with the user and
job control.
If this variable exists then simple commands
consisting of only a single word,
without redirections, are treated as candidates for resumption
of an existing job.
There is no ambiguity allowed; if there is more than one job
beginning with or containing the word, then
this selects the most recently accessed job.
The name of a stopped job, in this context, is the command line
used to start it, as displayed by @code{jobs}.
If this variable is set to the value @samp{exact},
the word must match the name of a stopped job exactly;
if set to @samp{substring},
the word needs to match a substring of the name of a stopped job.
The @samp{substring} value provides functionality
analogous to the @samp{%?string} job @sc{id} (@pxref{Job Control Basics}).
If set to any other value (e.g., @samp{prefix}),
the word must be a prefix of a stopped job's name;
this provides functionality analogous to the @samp{%string} job @sc{id}.

@end vtable

@set readline-appendix
@set history-appendix
@cindex Readline, how to use
@include rluser.texi
@cindex History, how to use
@include hsuser.texi
@clear readline-appendix
@clear history-appendix

@node Installing Bash
@chapter Installing Bash

This chapter provides basic instructions for installing Bash on
the various supported platforms.
The distribution supports the
@sc{gnu} operating systems, nearly every version of Unix, and several
non-Unix systems such as BeOS and Interix.
Other independent ports exist for Windows platforms.

@menu
* Basic Installation::	Installation instructions.
* Compilers and Options::	How to set special options for various
				systems.
* Compiling For Multiple Architectures::	How to compile Bash for more
						than one kind of system from
						the same source tree.
* Installation Names::	How to set the various paths used by the installation.
* Specifying the System Type::	How to configure Bash for a particular system.
* Sharing Defaults::	How to share default configuration values among GNU
			programs.
* Operation Controls::	Options recognized by the configuration program.
* Optional Features::	How to enable and disable optional features when
			building Bash.
@end menu

@node Basic Installation
@section Basic Installation
@cindex installation
@cindex configuration
@cindex Bash installation
@cindex Bash configuration

These are installation instructions for Bash.

The simplest way to compile Bash is:

@enumerate
@item
@code{cd} to the directory containing the source code and type
@samp{./configure} to configure Bash for your system.
If you're using @code{csh} on an old version of System V, you might
need to type @samp{sh ./configure} instead to prevent @code{csh}
from trying to execute @code{configure} itself.

Running @code{configure} takes some time.
While running, it prints messages telling which features it is
checking for.

@item
Type @samp{make} to compile Bash and build the @code{bashbug} bug
reporting script.

@item
Optionally, type @samp{make tests} to run the Bash test suite.

@item
Type @samp{make install} to install @code{bash} and @code{bashbug}.
This will also install the manual pages and Info file, message translation
files, some supplemental documentation, a number of example loadable
builtin commands, and a set of header files for developing loadable
builtins.
You may need additional privileges to install @code{bash} to your
desired destination, which may require @samp{sudo make install}.
More information about controlling the locations where @code{bash} and
other files are installed is below (@pxref{Installation Names}).

@end enumerate

The @code{configure} shell script attempts to guess correct
values for various system-dependent variables used during compilation.
It uses those values to create a @file{Makefile} in
each directory of the package (the top directory, the
@file{builtins}, @file{doc}, @file{po}, and @file{support} directories,
each directory under @file{lib}, and several others).
It also creates a
@file{config.h} file containing system-dependent definitions. 
Finally, it creates a shell script named @code{config.status} that you
can run in the future to recreate the current configuration, a
file @file{config.cache} that saves the results of its tests to
speed up reconfiguring, and a file @file{config.log} containing
compiler output (useful mainly for debugging @code{configure}). 
If at some point
@file{config.cache} contains results you don't want to keep, you
may remove or edit it. 

To find out more about the options and arguments that the
@code{configure} script understands, type 

@example
bash-4.2$ ./configure --help
@end example

@noindent
at the Bash prompt in your Bash source directory.

If you want to build Bash in a directory separate from the source
directory -- to build for multiple architectures, for example --
just use the full path to the configure script.
The following commands
will build Bash in a directory under @file{/usr/local/build} from
the source code in @file{/usr/local/src/bash-4.4}:

@example
mkdir /usr/local/build/bash-4.4
cd /usr/local/build/bash-4.4
bash /usr/local/src/bash-4.4/configure
make
@end example

See @ref{Compiling For Multiple Architectures} for more information
about building in a directory separate from the source.

If you need to do unusual things to compile Bash, please
try to figure out how @code{configure} could check whether or not
to do them, and mail diffs or instructions to
@email{bash-maintainers@@gnu.org} so they can be
considered for the next release.

The file @file{configure.ac} is used to create @code{configure}
by a program called Autoconf.
You only need @file{configure.ac} if you want to change it or regenerate
@code{configure} using a newer version of Autoconf.
If you do this, make sure you are using Autoconf version 2.69 or
newer.

You can remove the program binaries and object files from the
source code directory by typing @samp{make clean}.
To also remove the
files that @code{configure} created (so you can compile Bash for
a different kind of computer), type @samp{make distclean}.

@node Compilers and Options
@section Compilers and Options

Some systems require unusual options for compilation or linking
that the @code{configure} script does not know about.
You can give @code{configure} initial values for variables by setting
them in the environment.
Using a Bourne-compatible shell, you can do that on the command line
like this:

@example
CC=c89 CFLAGS=-O2 LIBS=-lposix ./configure
@end example

On systems that have the @code{env} program, you can do it like this:

@example
env CPPFLAGS=-I/usr/local/include LDFLAGS=-s ./configure
@end example

The configuration process uses GCC to build Bash if it
is available.

@node Compiling For Multiple Architectures
@section Compiling For Multiple Architectures

You can compile Bash for more than one kind of computer at the
same time, by placing the object files for each architecture in their
own directory.
To do this, you must use a version of @code{make} that
supports the @code{VPATH} variable, such as GNU @code{make}.
@code{cd} to the
directory where you want the object files and executables to go and run
the @code{configure} script from the source directory
(@pxref{Basic Installation}).
You may need to
supply the @option{--srcdir=PATH} argument to tell @code{configure}
where the source files are.
@code{configure} automatically checks for the
source code in the directory that @code{configure} is in and in @file{..}.

If you have to use a @code{make} that does not support the @code{VPATH}
variable, you can compile Bash for one architecture at a
time in the source code directory.
After you have installed
Bash for one architecture, use @samp{make distclean} before
reconfiguring for another architecture.

Alternatively, if your system supports symbolic links, you can use the
@file{support/mkclone} script to create a build tree which has
symbolic links back to each file in the source directory.
Here's an
example that creates a build directory in the current directory from a
source directory @file{/usr/gnu/src/bash-2.0}:

@example
bash /usr/gnu/src/bash-2.0/support/mkclone -s /usr/gnu/src/bash-2.0 .
@end example

@noindent
The @code{mkclone} script requires Bash, so you must have already built
Bash for at least one architecture before you can create build
directories for other architectures.

@node Installation Names
@section Installation Names

By default, @samp{make install} will install into
@file{/usr/local/bin}, @file{/usr/local/man}, etc.;
that is, the @dfn{installation prefix} defaults to @file{/usr/local}.
You can specify an installation prefix other than @file{/usr/local} by
giving @code{configure} the option @option{--prefix=@var{PATH}},
or by specifying a value for the @env{prefix} @samp{make}
variable when running @samp{make install}
(e.g., @samp{make install prefix=@var{PATH}}).
The @env{prefix} variable provides a default for @env{exec_prefix} and
other variables used when installing Bash.

You can specify separate installation prefixes for
architecture-specific files and architecture-independent files. 
If you give @code{configure} the option
@option{--exec-prefix=@var{PATH}}, @samp{make install} will use
@var{PATH} as the prefix for installing programs and libraries.
Documentation and other data files will still use the regular prefix.

If you would like to change the installation locations for a single run,
you can specify these variables as arguments to @code{make}:
@samp{make install exec_prefix=/} will install @code{bash} and
@code{bashbug} into @file{/bin} instead of the default @file{/usr/local/bin}.

If you want to see the files Bash will install and where it will install
them without changing anything on your system, specify the variable
@env{DESTDIR} as an argument to @code{make}.
Its value should be the absolute directory path you'd like to use as the
root of your sample installation tree.
For example,

@example
mkdir /fs1/bash-install
make install DESTDIR=/fs1/bash-install
@end example

@noindent
will install @code{bash} into @file{/fs1/bash-install/usr/local/bin/bash},
the documentation into directories within
@file{/fs1/bash-install/usr/local/share}, the example loadable builtins into
@file{/fs1/bash-install/usr/local/lib/bash}, and so on.
You can use the usual @env{exec_prefix} and @env{prefix} variables to alter
the directory paths beneath the value of @env{DESTDIR}.

The GNU Makefile standards provide a more complete description of these
variables and their effects.

@node Specifying the System Type
@section Specifying the System Type

There may be some features @code{configure} can not figure out
automatically, but needs to determine by the type of host Bash
will run on.
Usually @code{configure} can figure that
out, but if it prints a message saying it can not guess the host
type, give it the @option{--host=TYPE} option.
@samp{TYPE} can
either be a short name for the system type, such as @samp{sun4},
or a canonical name with three fields: @samp{CPU-COMPANY-SYSTEM}
(e.g., @samp{i386-unknown-freebsd4.2}).

See the file @file{support/config.sub} for the possible
values of each field. 

@node Sharing Defaults
@section Sharing Defaults

If you want to set default values for @code{configure} scripts to
share, you can create a site shell script called
@code{config.site} that gives default values for variables like
@code{CC}, @code{cache_file}, and @code{prefix}.
@code{configure}
looks for @file{PREFIX/share/config.site} if it exists, then
@file{PREFIX/etc/config.site} if it exists.
Or, you can set the
@code{CONFIG_SITE} environment variable to the location of the site
script.
A warning: the Bash @code{configure} looks for a site script,
but not all @code{configure} scripts do.

@node Operation Controls
@section Operation Controls

@code{configure} recognizes the following options to control how it
operates.

@table @code

@item --cache-file=@var{file}
Use and save the results of the tests in
@var{file} instead of @file{./config.cache}.
Set @var{file} to @file{/dev/null} to disable caching,
for debugging @code{configure}. 

@item --help
Print a summary of the options to @code{configure}, and exit.

@item --quiet
@itemx --silent
@itemx -q
Do not print messages saying which checks are being made.

@item --srcdir=@var{dir}
Look for the Bash source code in directory @var{dir}.
Usually @code{configure} can determine that directory automatically.

@item --version
Print the version of Autoconf used to generate the @code{configure}
script, and exit.
@end table

@code{configure} also accepts some other, not widely used, boilerplate
options.
@samp{configure --help} prints the complete list.

@node Optional Features
@section Optional Features

The Bash @code{configure} has a number of @option{--enable-@var{feature}}
options, where @var{feature} indicates an optional part of Bash.
There are also several @option{--with-@var{package}} options,
where @var{package} is something like @samp{bash-malloc} or @samp{afs}.
To turn off the default use of a package, use
@option{--without-@var{package}}.
To configure Bash without a feature
that is enabled by default, use @option{--disable-@var{feature}}.

Here is a complete list of the @option{--enable-} and @option{--with-}
options that the Bash @code{configure} recognizes. 

@table @code
@item --with-afs
Define if you are using the Andrew File System from Transarc.

@item --with-bash-malloc
Use the Bash version of
@code{malloc} in the directory @file{lib/malloc}.
This is not the same
@code{malloc} that appears in @sc{gnu} libc, but a custom version
originally derived from the 4.2 @sc{bsd} @code{malloc}.
This @code{malloc} is very fast, but wastes some space on each allocation,
though it uses several techniques to minimize the waste.
This option is enabled by default.
The @file{NOTES} file contains a list of systems for
which this should be turned off, and @code{configure} disables this
option automatically for a number of systems.

@item --with-curses[=@var{LIBNAME}]
Use the curses library instead of the termcap library as the library
where the linker can find the termcap functions.
@code{configure} usually chooses this automatically, since most systems
include the termcap functions in the curses library.
If @var{LIBNAME} is supplied, @code{configure} does not search for an
appropriate library and uses @var{LIBNAME} instead.
@var{LIBNAME} should be either an argument for the linker
(e.g., @option{-l@var{libname}})
or a filename
(e.g., @file{/opt/local/lib/libncursesw.so}).

@item --with-gnu-malloc
A synonym for @code{--with-bash-malloc}.

@item --with-installed-readline[=@var{PREFIX}]
Define this to make Bash link with a locally-installed version of Readline
rather than the version in @file{lib/readline}.
This works only with Readline 5.0 and later versions.
If @var{PREFIX} is @code{yes} or not
supplied, @code{configure} uses the values of the make variables
@code{includedir} and @code{libdir}, which are subdirectories of @code{prefix}
by default, to find the installed version of Readline if it is not in
the standard system include and library directories.
If @var{PREFIX} is @code{no}, Bash links with the version in
@file{lib/readline}.
If @var{PREFIX} is set to any other value, @code{configure} treats it as
a directory pathname and looks for
the installed version of Readline in subdirectories of that directory
(include files in @var{PREFIX}/@code{include} and the library in
@var{PREFIX}/@code{lib}).
The Bash default is to link with a static library built in the
@file{lib/readline} subdirectory of the build directory.

@item --with-libintl-prefix[=@var{PREFIX}]
Define this to make Bash link with a locally-installed version of the
libintl library instead of the version in @file{lib/intl}.

@item --with-libiconv-prefix[=@var{PREFIX}]
Define this to make Bash look for libiconv in @var{PREFIX} instead of the
standard system locations.
The Bash distribution does not include this library.

@item --enable-minimal-config
This produces a shell with minimal features, closer to the historical
Bourne shell.
@end table

There are several @option{--enable-} options that alter how Bash is
compiled, linked, and installed, rather than changing run-time features.

@table @code
@item --enable-largefile
Enable support for @url{http://www.unix.org/version2/whatsnew/lfs20mar.html,
large files} if the operating system requires special compiler options
to build programs which can access large files.
This is enabled by
default, if the operating system provides large file support.

@item --enable-profiling
This builds a Bash binary that produces profiling information to be
processed by @code{gprof} each time it is executed.

@item --enable-separate-helpfiles
Use external files for the documentation displayed by the @code{help} builtin
instead of storing the text internally.

@item --enable-static-link
This causes Bash to be linked statically, if @code{gcc} is being used.
This could be used to build a version to use as root's shell.

@end table

The @samp{minimal-config} option can be used to disable all of
the following options, but it is processed first, so individual
options may be enabled using @samp{enable-@var{feature}}. 

All of the following options except for
@samp{alt-array-implementation},
@samp{disabled-builtins},
@samp{direxpand-default},
@samp{strict-posix-default},
and
@samp{xpg-echo-default} are
enabled by default, unless the operating system does not provide the
necessary support.

@table @code
@item --enable-alias
Allow alias expansion and include the @code{alias} and @code{unalias}
builtins (@pxref{Aliases}).

@item --enable-alt-array-implementation
This builds Bash using an alternate implementation of arrays
(@pxref{Arrays}) that provides faster access at the expense of using
more memory (sometimes many times more, depending on how sparse an array is).

@item --enable-arith-for-command
Include support for the alternate form of the @code{for} command
that behaves like the C language @code{for} statement
(@pxref{Looping Constructs}).

@item --enable-array-variables
Include support for one-dimensional array shell variables
(@pxref{Arrays}).

@item --enable-bang-history
Include support for @code{csh}-like history substitution
(@pxref{History Interaction}).

@item --enable-bash-source-fullpath-default
Set the default value of the @code{bash_source_fullpath} shell option
described above under @ref{The Shopt Builtin} to be enabled.
This controls how filenames are assigned to the @code{BASH_SOURCE}
array variable.

@item --enable-brace-expansion
Include @code{csh}-like brace expansion
( @code{b@{a,b@}c} @expansion{} @code{bac bbc} ).
See @ref{Brace Expansion}, for a complete description.

@item --enable-casemod-attributes
Include support for case-modifying attributes in the @code{declare} builtin
and assignment statements.
Variables with the @code{uppercase} attribute,
for example, will have their values converted to uppercase upon assignment.

@item --enable-casemod-expansion
Include support for case-modifying word expansions.

@item --enable-command-timing
Include support for recognizing @code{time} as a reserved word and for
displaying timing statistics for the pipeline following @code{time}
(@pxref{Pipelines}).
This allows timing pipelines, shell compound commands, shell builtins,
and shell functions, which an external command cannot do easily.

@item --enable-cond-command
Include support for the @code{[[} conditional command.
(@pxref{Conditional Constructs}).

@item --enable-cond-regexp
Include support for matching @sc{posix} regular expressions using the
@samp{=~} binary operator in the @code{[[} conditional command.
(@pxref{Conditional Constructs}).

@item --enable-coprocesses
Include support for coprocesses and the @code{coproc} reserved word
(@pxref{Pipelines}).

@item --enable-debugger
Include support for the Bash debugger (distributed separately).

@item --enable-dev-fd-stat-broken
If calling @code{stat} on /dev/fd/@var{N} returns different results than
calling @code{fstat} on file descriptor @var{N}, supply this option to
enable a workaround.
This has implications for conditional commands that test file attributes.

@item --enable-direxpand-default
Cause the @code{direxpand} shell option (@pxref{The Shopt Builtin})
to be enabled by default when the shell starts.
It is normally disabled by default.

@item --enable-directory-stack
Include support for a @code{csh}-like directory stack and the
@code{pushd}, @code{popd}, and @code{dirs} builtins
(@pxref{The Directory Stack}).

@item --enable-disabled-builtins
Allow builtin commands to be invoked via @samp{builtin xxx}
even after @code{xxx} has been disabled using @samp{enable -n xxx}.
See @ref{Bash Builtins}, for details of the @code{builtin} and
@code{enable} builtin commands.

@item --enable-dparen-arithmetic
Include support for the @code{((@dots{}))} command
(@pxref{Conditional Constructs}).

@item --enable-extended-glob
Include support for the extended pattern matching features described
above under @ref{Pattern Matching}.

@item --enable-extended-glob-default
Set the default value of the @code{extglob} shell option described
above under @ref{The Shopt Builtin} to be enabled.

@item --enable-function-import
Include support for importing function definitions exported by another
instance of the shell from the environment.
This option is enabled by default.

@item --enable-glob-asciiranges-default
Set the default value of the @code{globasciiranges} shell option described
above under @ref{The Shopt Builtin} to be enabled.
This controls the behavior of character ranges when used in pattern matching
bracket expressions.

@item --enable-help-builtin
Include the @code{help} builtin, which displays help on shell builtins and
variables (@pxref{Bash Builtins}).

@item --enable-history
Include command history and the @code{fc} and @code{history}
builtin commands (@pxref{Bash History Facilities}).

@item --enable-job-control
This enables the job control features (@pxref{Job Control}),
if the operating system supports them.

@item --enable-multibyte
This enables support for multibyte characters if the operating
system provides the necessary support.

@item --enable-net-redirections
This enables the special handling of filenames of the form
@code{/dev/tcp/@var{host}/@var{port}} and
@code{/dev/udp/@var{host}/@var{port}}
when used in redirections (@pxref{Redirections}).

@item --enable-process-substitution
This enables process substitution (@pxref{Process Substitution}) if
the operating system provides the necessary support.

@item --enable-progcomp
Enable the programmable completion facilities
(@pxref{Programmable Completion}).
If Readline is not enabled, this option has no effect.

@item --enable-prompt-string-decoding
Turn on the interpretation of a number of backslash-escaped characters
in the @env{$PS0}, @env{$PS1}, @env{$PS2}, and @env{$PS4} prompt
strings.
See @ref{Controlling the Prompt}, for a complete list of prompt
string escape sequences.

@item --enable-readline
Include support for command-line editing and history with the Bash
version of the Readline library (@pxref{Command Line Editing}).

@item --enable-restricted
Include support for a @dfn{restricted shell}.
If this is enabled,
Bash enters a restricted mode when called as @code{rbash}.
See @ref{The Restricted Shell}, for a description of restricted mode.

@item --enable-select
Include the @code{select} compound command, which allows generation of
simple menus (@pxref{Conditional Constructs}).

@item --enable-single-help-strings
Store the text displayed by the @code{help} builtin as a single string for
each help topic.
This aids in translating the text to different languages.
You may need to disable this if your compiler cannot handle very long string
literals.

@item --enable-strict-posix-default
Make Bash @sc{posix}-conformant by default (@pxref{Bash POSIX Mode}).

@item --enable-translatable-strings
Enable support for @code{$"@var{string}"} translatable strings
(@pxref{Locale Translation}).

@item --enable-usg-echo-default
A synonym for @code{--enable-xpg-echo-default}.

@item --enable-xpg-echo-default
Make the @code{echo} builtin expand backslash-escaped characters by default,
without requiring the @option{-e} option.
This sets the default value of the @code{xpg_echo} shell option to @code{on},
which makes the Bash @code{echo} behave more like the version specified in
the Single Unix Specification, version 3.
@xref{Bash Builtins}, for a description of the escape sequences that
@code{echo} recognizes.
@end table

The file @file{config-top.h} contains C Preprocessor
@samp{#define} statements for options which are not settable from
@code{configure}.
Some of these are not meant to be changed; beware of the consequences if
you do.
Read the comments associated with each definition for more
information about its effect.

@node Reporting Bugs
@appendix Reporting Bugs

Please report all bugs you find in Bash.
But first, you should
make sure that it really is a bug, and that it appears in the latest
version of Bash.
The latest released version of Bash is always available for FTP from
@uref{ftp://ftp.gnu.org/pub/gnu/bash/} and from
@uref{http://git.savannah.gnu.org/cgit/bash.git/snapshot/bash-master.tar.gz}.

Once you have determined that a bug actually exists, use the
@code{bashbug} command to submit a bug report or use the form at the
@uref{https://savannah.gnu.org/projects/bash/,Bash project page}.
If you have a fix, you are encouraged to submit that as well!
Suggestions and ``philosophical'' bug reports may be mailed
to @email{bug-bash@@gnu.org} or @email{help-bash@@gnu.org}.

All bug reports should include:
@itemize @bullet
@item
The version number of Bash.
@item
The hardware and operating system.
@item
The compiler used to compile Bash.
@item
A description of the bug behavior.
@item
A short script or ``recipe'' which exercises the bug and may be used
to reproduce it.
@end itemize

@noindent
@code{bashbug} inserts the first three items automatically into
the template it provides for filing a bug report.

Please send all reports concerning this manual to
@email{bug-bash@@gnu.org}.

@node Major Differences From The Bourne Shell
@appendix Major Differences From The Bourne Shell

Bash implements essentially the same grammar, parameter and
variable expansion, redirection, and quoting as the Bourne Shell. 
Bash uses the @sc{posix} standard as the specification of
how these features are to be implemented and how they should behave.
There are some
differences between the traditional Bourne shell and Bash; this
section quickly details the differences of significance.
A number of these differences are explained in greater depth in
previous sections.
This section uses the version of @code{sh} included in SVR4.2 (the
last version of the historical Bourne shell) as the baseline reference.

@itemize @bullet

@item
Bash is @sc{posix}-conformant, even where the @sc{posix} specification
differs from traditional @code{sh} behavior (@pxref{Bash POSIX Mode}).

@item
Bash has multi-character invocation options (@pxref{Invoking Bash}).

@item
The Bash restricted mode is more useful (@pxref{The Restricted Shell});
the SVR4.2 shell restricted mode is too limited.

@item
Bash has command-line editing (@pxref{Command Line Editing}) and
the @code{bind} builtin.

@item
Bash provides a programmable word completion mechanism
(@pxref{Programmable Completion}), and builtin commands
@code{complete}, @code{compgen}, and @code{compopt}, to
manipulate it.

@item
Bash decodes a number of backslash-escape sequences in the prompt string
variables (@code{PS0}, @code{PS1}, @code{PS2}, and @code{PS4})
(@pxref{Controlling the Prompt}).

@item
Bash expands and displays the @code{PS0} prompt string variable.

@item
Bash runs commands from the @code{PROMPT_COMMAND} array variable before
issuing each primary prompt.

@item
Bash has command history (@pxref{Bash History Facilities}) and the
@code{history} and @code{fc} builtins to manipulate it.
The Bash history list maintains timestamp information and uses the
value of the @code{HISTTIMEFORMAT} variable to display it.

@item
Bash implements @code{csh}-like history expansion
(@pxref{History Interaction}).

@item
Bash supports the @code{$'@dots{}'} quoting syntax, which expands ANSI-C
backslash-escaped characters in the text between the single quotes
(@pxref{ANSI-C Quoting}).

@item
Bash supports the @code{$"@dots{}"} quoting syntax and performs
locale-specific translation of the characters between the double
quotes.
The @option{-D}, @option{--dump-strings}, and @option{--dump-po-strings}
invocation options list the translatable strings found in a script
(@pxref{Locale Translation}).

@item
Bash includes brace expansion (@pxref{Brace Expansion}) and tilde
expansion (@pxref{Tilde Expansion}).

@item
Bash implements command aliases and the @code{alias} and @code{unalias}
builtins (@pxref{Aliases}).

@item
Bash implements the @code{!} reserved word to negate the return value of
a pipeline (@pxref{Pipelines}).
This is very useful when an @code{if} statement needs to act only if a
test fails.
The Bash @samp{-o pipefail} option to @code{set} will cause a pipeline to
return a failure status if any command fails
(@pxref{The Set Builtin}).

@item
Bash has the @code{time} reserved word and command timing (@pxref{Pipelines}).
The display of the timing statistics may be controlled with the
@env{TIMEFORMAT} variable.

@item
Bash provides coprocesses and the @code{coproc} reserved word
(@pxref{Coprocesses}).

@item
Bash implements the @code{for (( @var{expr1} ; @var{expr2} ; @var{expr3} ))}
arithmetic for command, similar to the C language (@pxref{Looping Constructs}).

@item
Bash includes the @code{select} compound command, which allows the
generation of simple menus (@pxref{Conditional Constructs}).

@item
Bash includes the @code{[[} compound command, which makes conditional
testing part of the shell grammar (@pxref{Conditional Constructs}), including
optional regular expression matching.

@item
Bash provides optional case-insensitive matching for the @code{case} and
@code{[[} constructs  (@pxref{Conditional Constructs}).

@item
Bash provides additional @code{case} statement action list terminators:
@samp{;&} and @samp{;;&} (@pxref{Conditional Constructs}).

@item
Bash provides shell arithmetic,
the @code{((} compound command (@pxref{Conditional Constructs}),
the @code{let} builtin,
and arithmetic expansion (@pxref{Shell Arithmetic}).

@item
Bash has one-dimensional array variables (@pxref{Arrays}), and the
appropriate variable expansions and assignment syntax to use them.
Several of the Bash builtins take options to act on arrays.
Bash provides a number of built-in array variables.

@item
Variables present in the shell's initial environment are automatically
exported to child processes (@pxref{Command Execution Environment}).
The Bourne shell does not normally do
this unless the variables are explicitly marked using the @code{export}
command.

@item
Bash can expand positional parameters beyond @code{$9} using
@code{$@{@var{num}@}} (@pxref{Shell Parameter Expansion}).

@item
Bash supports the @samp{+=} assignment operator, which appends to the value
of the variable named on the left hand side
(@pxref{Shell Parameters}).

@item
Bash includes the @sc{posix} pattern removal @samp{%}, @samp{#}, @samp{%%}
and @samp{##} expansions to remove leading or trailing substrings from
variable values (@pxref{Shell Parameter Expansion}).

@item
The expansion @code{$@{#xx@}}, which returns the length of @code{$@{xx@}},
is supported (@pxref{Shell Parameter Expansion}).

@item
The expansion @code{$@{var:}@var{offset}@code{[:}@var{length}@code{]@}},
which expands to the substring of @code{var}'s value of length
@var{length}, beginning at @var{offset}, is present
(@pxref{Shell Parameter Expansion}).

@item
The expansion
@code{$@{@var{var}/[/]}@var{pattern}@code{[/}@var{replacement}@code{]@}},
which matches @var{pattern} and replaces it with @var{replacement} in
the value of @var{var}, is available (@pxref{Shell Parameter Expansion}),
with a mechanism to use the matched text in @var{replacement}.

@item
The expansion @code{$@{!@var{prefix}*@}} expansion, which expands to
the names of all shell variables whose names begin with @var{prefix},
is available (@pxref{Shell Parameter Expansion}).

@item
Bash has indirect variable expansion using @code{$@{!word@}}
(@pxref{Shell Parameter Expansion}) and implements the @code{nameref}
variable attribute for automatic indirect variable expansion.

@item
Bash includes a set of parameter transformation word expansions of the
form @code{$@{var@@X@}}, where @samp{X} specifies the transformation
(@pxref{Shell Parameter Expansion}).

@item
The @sc{posix} @code{$()} form of command substitution
is implemented (@pxref{Command Substitution}),
and preferred to the Bourne shell's @code{``} (which
is also implemented for backwards compatibility).

@item
Bash implements a variant of command substitution that runs the enclosed
command in the current shell execution environment:
@code{$@{ @var{command};@}} or @code{$@{|@var{command};@}}
(@pxref{Command Substitution}).

@item
Bash has process substitution (@pxref{Process Substitution}).

@item
Bash automatically assigns variables that provide information about the
current user (@env{UID}, @env{EUID}, and @env{GROUPS}), the current host
(@env{HOSTTYPE}, @env{OSTYPE}, @env{MACHTYPE}, and @env{HOSTNAME}),
and the instance of Bash that is running (@env{BASH},
@env{BASH_VERSION}, and @env{BASH_VERSINFO}).
@xref{Bash Variables}, for details.

@item
Bash uses many variables to provide functionality and customize shell
behavior that the Bourne shell does not.
Examples include @env{RANDOM}, @env{SRANDOM}, @env{EPOCHSECONDS},
@env{EPOCHREALTIME},
@env{TIMEFORMAT}, @env{BASHPID}, @env{BASH_XTRACEFD},
@env{GLOBIGNORE}, @env{HISTIGNORE}, and @env{BASH_VERSION}.
@xref{Bash Variables}, for a complete list.

@item
Bash uses the @code{GLOBSORT} shell variable to control how to sort
the results of filename expansion (@pxref{Filename Expansion}).

@item
Bash uses the @env{IFS} variable to split only the results of expansion,
not all words (@pxref{Word Splitting}).
This closes a longstanding shell security hole.

@item
The filename expansion bracket expression code uses @samp{!} and @samp{^}
to negate the set of characters between the brackets
(@pxref{Filename Expansion}).
The Bourne shell uses only @samp{!}.

@item
Bash implements the full set of @sc{posix} filename expansion operators,
including character classes, equivalence classes, and
collating symbols (@pxref{Filename Expansion}).

@item
Bash implements extended pattern matching features when the @code{extglob}
shell option is enabled (@pxref{Pattern Matching}).

@item
The @code{globstar} shell option extends filename expansion to recursively
scan directories and subdirectories for matching filenames
(@pxref{Pattern Matching}).

@item
It is possible to have a variable and a function with the same name;
@code{sh} does not separate the two name spaces.

@item
Bash functions are permitted to have local variables using the
@code{local} builtin, and thus users can write useful recursive functions
(@pxref{Bash Builtins}).

@item
Bash performs filename expansion on filenames specified as operands
to input and output redirection operators (@pxref{Redirections}).

@item
Bash contains the @samp{<>} redirection operator, allowing a file to be
opened for both reading and writing, and the @samp{&>} redirection
operator, for directing standard output and standard error to the same
file (@pxref{Redirections}).

@item
Bash includes the @samp{<<<} redirection operator, allowing a string to
be used as the standard input to a command (@pxref{Redirections}).

@item
Bash implements the @samp{[n]<&@var{word}} and @samp{[n]>&@var{word}}
redirection operators, which move one file descriptor to another.

@item
Bash treats a number of filenames specially when they are
used in redirection operators (@pxref{Redirections}).

@item
Bash provides the @{@var{var}@}<@var{word} capability to have the
shell allocate file descriptors for redirections and assign them
to @var{var} (@pxref{Redirections}).
This works with multiple redirection operators.

@item
Bash can open network connections to arbitrary machines and services
with the redirection operators (@pxref{Redirections}).

@item
The @code{noclobber} option is available to avoid overwriting existing
files with output redirection (@pxref{The Set Builtin}).
The @samp{>|} redirection operator may be used to override @code{noclobber}.

@item
Variable assignments preceding commands affect only that command, even
builtins and functions (@pxref{Environment}).
In @code{sh}, all variable assignments 
preceding commands are global unless the command is executed from the
file system.

@item
Bash includes a number of features to support a separate debugger for
shell scripts:
variables (@code{BASH_ARGC}, @code{BASH_ARGV}, @code{BASH_LINENO},
@code{BASH_SOURCE}),
the @code{DEBUG}, @code{RETURN}, and @code{ERR} traps,
@samp{declare -F},
and
the @code{caller} builtin.

@item
Bash implements a @code{csh}-like directory stack, and provides the
@code{pushd}, @code{popd}, and @code{dirs} builtins to manipulate it
(@pxref{The Directory Stack}).
Bash also makes the directory stack visible as the value of the
@env{DIRSTACK} shell variable.

@item
Bash allows a function to override a builtin with the same name, and provides
access to that builtin's functionality within the function via the
@code{builtin} and @code{command} builtins (@pxref{Bash Builtins}).

@item
Bash includes the @code{caller} builtin
(@pxref{Bash Builtins}), which displays the context of
any active subroutine call (a shell function or a script executed with
the @code{.} or @code{source} builtins).
This supports the Bash debugger.

@item
The Bash @code{cd} and @code{pwd} builtins (@pxref{Bourne Shell Builtins})
each take @option{-L} and @option{-P} options to switch between logical and
physical modes.

@item
The @code{command} builtin allows selectively skipping shell functions
when performing command lookup (@pxref{Bash Builtins}).

@item
Bash uses the @code{declare} builtin to modify the full set of variable
and function attributes, and to assign values to variables.

@item
The @code{disown} builtin can remove a job from the internal shell
job table (@pxref{Job Control Builtins}) or suppress sending
@code{SIGHUP} to a job when the shell exits as the result of a
@code{SIGHUP}.

@item
The @code{enable} builtin (@pxref{Bash Builtins}) can enable or disable
individual builtins
and implements support for dynamically loading
builtin commands from shared objects.

@item
The Bash @code{exec} builtin takes additional options that allow users
to control the contents of the environment passed to the executed
command, and what the zeroth argument to the command is to be
(@pxref{Bourne Shell Builtins}).

@item
Shell functions may be exported to children via the environment
using @code{export -f} (@pxref{Shell Functions}).

@item
The Bash @code{export} and @code{readonly} builtins
(@pxref{Bourne Shell Builtins}) can
take a @option{-f} option to act on shell functions, a @option{-p} option to
display variables with various attributes set in a format that can be
used as shell input, a @option{-n} option to remove various variable
attributes, and @samp{name=value} arguments to set variable attributes
and values simultaneously.

@item
The Bash @code{hash} builtin allows a name to be associated with
an arbitrary filename, even when that filename cannot be found by
searching the @env{$PATH}, using @samp{hash -p}
(@pxref{Bourne Shell Builtins}).

@item
Bash includes a @code{help} builtin for quick reference to shell
facilities (@pxref{Bash Builtins}).

@item
Bash includes the @code{mapfile} builtin to quickly read the contents
of a file into an indexed array variable (@pxref{Bash Builtins}).

@item
The @code{printf} builtin is available to display formatted output
(@pxref{Bash Builtins}),
and has additional custom format specifiers and an option to assign
the formatted output directly to a shell variable.

@item
The Bash @code{read} builtin (@pxref{Bash Builtins})
will read a line ending in @samp{\} with
the @option{-r} option, and will use the @env{REPLY} variable as a
default if no non-option arguments are supplied.

@item
The @code{read} builtin (@pxref{Bash Builtins})
accepts a prompt string with the @option{-p} option and will use
Readline to obtain the line when given the @option{-e} or @option{-E}
options,
with the ability to insert text into the line using the @option{-i}
option.
The @code{read} builtin also has additional options to control input:
the @option{-s} option will turn off echoing of input characters as
they are read, the @option{-t} option will allow @code{read} to time out
if input does not arrive within a specified number of seconds, the
@option{-n} option will allow reading only a specified number of
characters rather than a full line, and the @option{-d} option will read
until a particular character rather than newline.

@item
The @code{return} builtin may be used to abort execution of scripts
executed with the @code{.} or @code{source} builtins
(@pxref{Bourne Shell Builtins}).

@item
Bash has much more optional behavior controllable with the @code{set}
builtin (@pxref{The Set Builtin}).

@item
The @option{-x} (@option{xtrace}) option displays commands other than
simple commands when performing an execution trace
(@pxref{The Set Builtin}).

@item
Bash includes the @code{shopt} builtin, for finer control of shell
optional capabilities (@pxref{The Shopt Builtin}), and allows these options
to be set and unset at shell invocation (@pxref{Invoking Bash}).

@item
The @code{test} builtin (@pxref{Bourne Shell Builtins})
is slightly different, as it implements the @sc{posix} algorithm,
which specifies the behavior based on the number of arguments.

@item
The @code{trap} builtin (@pxref{Bourne Shell Builtins}) allows a
@code{DEBUG} pseudo-signal specification, similar to @code{EXIT}.
Commands specified with a @code{DEBUG} trap are executed before every
simple command, @code{for} command, @code{case} command,
@code{select} command, every arithmetic @code{for} command, and before
the first command executes in a shell function.
The @code{DEBUG} trap is not inherited by shell functions unless the
function has been given the @code{trace} attribute or the
@code{functrace} option has been enabled using the @code{shopt} builtin.
The @code{extdebug} shell option has additional effects on the
@code{DEBUG} trap.

The @code{trap} builtin (@pxref{Bourne Shell Builtins}) allows an
@code{ERR} pseudo-signal specification, similar to @code{EXIT} and @code{DEBUG}.
Commands specified with an @code{ERR} trap are executed after a simple
command fails, with a few exceptions.
The @code{ERR} trap is not inherited by shell functions unless the
@code{-o errtrace} option to the @code{set} builtin is enabled.

The @code{trap} builtin (@pxref{Bourne Shell Builtins}) allows a
@code{RETURN} pseudo-signal specification, similar to
@code{EXIT} and @code{DEBUG}.
Commands specified with a @code{RETURN} trap are executed before
execution resumes after a shell function or a shell script executed with
@code{.} or @code{source} returns.
The @code{RETURN} trap is not inherited by shell functions unless the
function has been given the @code{trace} attribute or the
@code{functrace} option has been enabled using the @code{shopt} builtin.

@item
The Bash @code{type} builtin is more extensive and gives more information
about the names it finds (@pxref{Bash Builtins}).

@item
The @code{ulimit} builtin provides control over many more per-process
resources (@pxref{Bash Builtins}).

@item
The Bash @code{umask} builtin uses the @option{-p} option to display
the output in the form of a @code{umask} command
that may be reused as input (@pxref{Bourne Shell Builtins}).

@item
The Bash @code{wait} builtin has a @option{-n} option to wait for the
next child to exit, possibly selecting from a list of supplied jobs,
and the @option{-p} option to store information about a terminated
child process in a shell variable.

@item 
The SVR4.2 shell behaves differently when invoked as @code{jsh}
(it turns on job control).

@item
The SVR4.2 shell has two privilege-related builtins
(@code{mldmode} and @code{priv}) not present in Bash.

@item
Bash does not have the @code{stop} or @code{newgrp} builtins.

@item
Bash does not use the @env{SHACCT} variable or perform shell accounting.

@item
The SVR4.2 @code{sh} uses a @env{TIMEOUT} variable like Bash uses
@env{TMOUT}.

@end itemize

@noindent
More features unique to Bash may be found in @ref{Bash Features}.

@appendixsec Implementation Differences From The SVR4.2 Shell

Since Bash is a completely new implementation, it does not suffer from
many of the limitations of the SVR4.2 shell.
For instance:

@itemize @bullet

@item
Bash does not fork a subshell when redirecting into or out of
a shell control structure such as  an @code{if} or @code{while}
statement.

@item
Bash does not allow unbalanced quotes.
The SVR4.2 shell will silently
insert a needed closing quote at @code{EOF} under certain circumstances.
This can be the cause of some hard-to-find errors.

@item
The SVR4.2 shell uses a baroque memory management scheme based on
trapping @code{SIGSEGV}.
If the shell is started from a process with
@code{SIGSEGV} blocked (e.g., by using the @code{system()} C library
function call), it misbehaves badly.

@item
In a questionable attempt at security, the SVR4.2 shell,
when invoked without the @option{-p} option, will alter its real
and effective @sc{uid} and @sc{gid} if they are less than some
magic threshold value, commonly 100.
This can lead to unexpected results.

@item
The SVR4.2 shell does not allow users to trap @code{SIGSEGV},
@code{SIGALRM}, or @code{SIGCHLD}.

@item
The SVR4.2 shell does not allow the @env{IFS}, @env{MAILCHECK},
@env{PATH}, @env{PS1}, or @env{PS2} variables to be unset.

@item
The SVR4.2 shell treats @samp{^} as the undocumented equivalent of
@samp{|}.

@item
Bash allows multiple option arguments when it is invoked (@code{-x -v});
the SVR4.2 shell allows only one option argument (@code{-xv}).
In fact, some versions of the shell dump core if the second argument begins
with a @samp{-}.

@item
The SVR4.2 shell exits a script if any builtin fails; Bash exits
a script only if one of the @sc{posix} special builtins fails, and
only for certain failures, as enumerated in the @sc{posix} standard.

@item
If the @code{lastpipe} option is enabled, and job control is not active,
Bash runs the last element of a pipeline in the current shell execution
environment.

@end itemize

@node GNU Free Documentation License
@appendix GNU Free Documentation License

@include fdl.texi

@node Indexes
@appendix Indexes

@menu
* Builtin Index::		Index of Bash builtin commands.
* Reserved Word Index::		Index of Bash reserved words.
* Variable Index::		Quick reference helps you find the
				variable you want.
* Function Index::		Index of bindable Readline functions.
* Concept Index::		General index for concepts described in
				this manual.
@end menu

@node Builtin Index
@appendixsec Index of Shell Builtin Commands
@printindex bt

@node Reserved Word Index
@appendixsec Index of Shell Reserved Words
@printindex rw

@node Variable Index
@appendixsec Parameter and Variable Index
@printindex vr

@node Function Index
@appendixsec Function Index
@printindex fn

@node Concept Index
@appendixsec Concept Index
@printindex cp

@bye