Definitions.¶

A metacharacter is one of the following characters:

A blank is a tab or a space. An identifier is a sequence of letters, digits, or underscores starting with a letter or underscore. Identifiers are used as components of variable names. A vname is a sequence of one or more identifiers separated by a . and optionally preceded by a .. Vnames are used as function and variable names. A word is a sequence of characters from the character set defined by the current locale, excluding non-quoted metacharacters.

A command is a sequence of characters in the syntax of the shell language. The shell reads each command and carries out the desired action either directly or by invoking separate utilities. A built-in command is a command that is carried out by the shell itself without creating a separate process. Some commands are built in purely for convenience and are not documented here. Built-ins that cause side effects in the shell environment and built-ins that are found before performing a path search (see Execution below) are documented here. For historical reasons, some of these built-ins behave differently than other built-ins and are called special built-ins.

Commands.¶

A simple-command is a list of variable assignments (see Variable Assignments below) or a sequence of blank separated words which may be preceded by a list of variable assignments (see Environment below). The first word specifies the name of the command to be executed. Except as specified below, the remaining words are passed as arguments to the invoked command. The command name is passed as argument 0 (see exec(2)). The value of a simple-command is its exit status; 0-255 if it terminates normally; 256+signum if it terminates abnormally (the name of the signal corresponding to the exit status can be obtained via the -l option of the kill built-in utility).

A pipeline is a sequence of one or more commands separated by |. The standard output of each command but the last is connected by a socketpair(2) or (if the posix shell option is on) by a pipe(2) to the standard input of the next command. Each command except the last is run asynchronously in a subshell (see Subshells below). If the monitor or pipefail option is on, or the pipeline is preceded by the reserved word time, then the shell waits for all component commands in the pipeline to terminate; otherwise, the shell only waits for the last component command. The exit status of a pipeline is the exit status of its last component command, unless the pipefail option is enabled. Each pipeline can be preceded by the reserved word ! which causes the exit status of the pipeline to become 0 if the exit status of the last command is non-zero, and 1 if the exit status of the last command is 0.

A list is a sequence of one or more pipelines separated by ;, &, |&, &&, or ||, and optionally terminated by ;, &, or |&. Of these five symbols, ;, &, and |& have equal precedence, which is lower than that of && and ||. The symbols && and || also have equal precedence. A semicolon (;) causes sequential execution of the preceding pipeline; an ampersand (&) causes asynchronous execution of the preceding pipeline (i.e., the shell does not wait for that pipeline to finish). The symbol |& causes asynchronous execution of the preceding pipeline with a two-way pipe established to the parent shell; the standard input and output of the spawned pipeline can be written to and read from by the parent shell by applying the redirection operators <& and >& with arg p to commands and by using -p option of the built-in commands read and print described later. The symbol && (||) causes the list following it to be executed only if the preceding pipeline returns a zero (non-zero) value. One or more new-lines may appear in a list instead of a semicolon, to delimit a command. The first item of the first pipeline of a list that is a simple command not beginning with a redirection, and not occurring within a while, until, or if list, can be preceded by a semicolon. This semicolon is ignored unless the showme option is enabled as described with the set built-in below.

A command is either a simple-command or one of the following. Unless otherwise stated, the value returned by a command is that of the last simple-command executed in the command.

for vname [ in word ... ] ;do list ;done

Each time a for command is executed, vname is set to the next word taken from the in word list. If in word ... is omitted, then the for command executes the do list once for each positional parameter that is set starting from 1 (see Parameter Expansion below). Execution ends when there are no more words in the list.

for (( [expr1] ; [expr2] ; [expr3] )) ;do list ;done

The arithmetic expression expr1 is evaluated first (see Arithmetic Evaluation below). The arithmetic expression expr2 is repeatedly evaluated until it evaluates to zero and when non-zero, list is executed and the arithmetic expression expr3 evaluated. If any expression is omitted, then it behaves as if it evaluated to 1.

select vname [ in word ... ] ;do list ;done

A select command prints on standard error (file descriptor 2) the set of words, each preceded by a number. If in word ... is omitted, then the positional parameters starting from 1 are used instead (see Parameter Expansion below). The PS3 prompt is printed and a line is read from the standard input. If this line consists of the number of one of the listed words, then the value of the variable vname is set to the word corresponding to this number. If this line is empty, the selection list is printed again. Otherwise the value of the variable vname is set to the empty string. The contents of the line read from standard input is saved in the variable REPLY. The list is executed for each selection until a break or end-of-file is encountered. If the REPLY variable is set to the empty string by the execution of list, then the selection list is printed before displaying the PS3 prompt for the next selection.

case word in [ [(]pattern [ | pattern ] ... ) list ;; ] ... esac

A case command executes the list associated with the first pattern that matches word. The form of the patterns is the same as that used for pathname expansion (see Pathname Expansion below). The ;; operator causes execution of case to terminate. If ;& is used in place of ;; the next subsequent list, if any, is executed.

if list ;then list [ ;elif list ;then list ] ... [ ;else list ] ;fi

The list following if is executed and, if it returns a zero exit status, the list following the first then is executed. Otherwise, the list following elif is executed and, if its value is zero, the list following the next then is executed. Failing each successive elif list, the else list is executed. If the if list has non-zero exit status and there is no else list, then the if command returns a zero exit status.

while list ;do list ;done

until list ;do list ;done

A while command repeatedly executes the while list and, if the exit status of the last command in the list is zero, executes the do list; otherwise the loop terminates. If no commands in the do list are executed, then the while command returns a zero exit status; until may be used in place of while to negate the loop termination test.

while inputredirection ;do list ;done

Filescan loop. This is defined by a lone input redirection following while (see Input/Output below). It is faster than using the read built-in command in a regular while loop. The shell reads lines from the file or stream opened by inputredirection until the end is reached or the loop is broken. For each line read, the command list is executed with the line's contents assigned to the REPLY variable and the line's fields split into the positional parameters (see Field Splitting and Positional Parameters below). Within the list, standard input is redirected to /dev/null. If the posix compatibility shell option is on, this loop type is disabled and inputredirection is processed like a lone redirection in any other context.

((expression))

The expression is evaluated using the rules for arithmetic evaluation described below. If the value of the arithmetic expression is non-zero, the exit status is 0, otherwise the exit status is 1.

(list)

Execute list in a subshell (see Subshells below). Note, that if two adjacent open parentheses are needed for nesting, a space must be inserted to avoid evaluation as an arithmetic command as described above.

{ list;}

list is simply executed. Note that unlike the metacharacters ( and ), { and } are reserved words and must occur at the beginning of a line or after a ; in order to be recognized.

[[ expression ]]

Evaluates expression and returns a zero exit status when expression is true. See Conditional Expressions below, for a description of expression.

function varname { list ;}

varname () { list ;}

Define a function which is referenced by varname. A function whose varname contains a . is called a discipline function and the portion of the varname preceding the last . must refer to an existing variable. The body of the function is the list of commands between { and }. A function defined with the function varname syntax can also be used as an argument to the . special built-in command to get the equivalent behavior as if the varname() syntax were used to define it. (See Functions below.)

namespace identifier { list ;}

Defines or uses the name space identifier and runs the commands in list in this name space. (See Name Spaces below.)

& [ name [ arg... ] ]

Causes subsequent list commands terminated by & to be placed in the background job pool name. If name is omitted a default unnamed pool is used. Commands in a named background pool may be executed remotely.

time [ pipeline ]

If pipeline is omitted the user and system time for the current shell and completed child processes is printed on standard error. Otherwise, pipeline is executed and the elapsed time as well as the user and system time are printed on standard error. The TIMEFORMAT variable may be set to a format string that specifies how the timing information should be displayed. See Shell Variables below for a description of the TIMEFORMAT variable.

The following reserved words are recognized as reserved only when they are the first word of a command and are not quoted:

if then else elif fi case esac for while until do done { } function select time [[ ]] !

Variable Assignments.¶

One or more variable assignments can start a simple command or can be arguments to the typeset, enum, export, or readonly special built-in commands as well as to other declaration commands created as types. The syntax for an assignment is of the form:

varname=word

varname[word]=word

No space is permitted between varname and the = or between = and word.

varname=(assign_list)

No space is permitted between varname and the =. The variable varname is unset before the assignment. An assign_list can be one of the following:

In addition, a += can be used in place of the = to signify adding to or appending to the previous value. When += is applied to an arithmetic type, word is evaluated as an arithmetic expression and added to the current value. When applied to a string variable, the value defined by word is appended to the value. For compound assignments, the previous value is not unset and the new values are appended to the current ones provided that the types are compatible.

The right hand side of a variable assignment undergoes all the expansion listed below except word splitting, brace expansion, and pathname expansion. When the left hand side is an assignment is a compound variable and the right hand is the name of a compound variable, the compound variable on the right will be copied or appended to the compound variable on the left.

Comments.¶

A word beginning with # causes that word and all the following characters up to a new-line to be ignored.

Aliasing.¶

The first word of each command is replaced by the text of an alias if an alias for this word has been defined. An alias name consists of any number of characters excluding metacharacters, quoting characters, file expansion characters, parameter expansion and command substitution characters, the characters / and =. The replacement string can contain any valid shell script including the metacharacters listed above. The first word of each command in the replaced text, other than any that are in the process of being replaced, will be tested for aliases. If the last character of the alias value is a blank then the word following the alias will also be checked for alias substitution. Aliases can be used to redefine built-in commands but cannot be used to redefine the reserved words listed above. Aliases can be created and listed with the alias command and can be removed with the unalias command.

Aliasing is performed when scripts are read, not while they are executed. Therefore, for an alias to take effect, the alias definition command has to be executed before the command which references the alias is read.

The following aliases are automatically preset when the shell is invoked as an interactive shell. Preset aliases can be unset or redefined.

Tilde Expansion.¶

After alias substitution is performed, each word is checked to see if it begins with an unquoted ∼. For tilde expansion, word also refers to the word portion of parameter expansion (see Parameter Expansion below). If a word is preceded by a tilde, then it is checked up to a / to see if it matches a user name in the password database (see getpwname(3)). If a match is found, the ∼ and the matched login name are replaced by the login directory of the matched user. If no match is found, the original text is left unchanged. A ∼ by itself, or in front of a /, is replaced by $HOME, unless the HOME variable is unset, in which case the current user's home directory as configured in the operating system is used. A ∼ followed by a + or - is replaced by $PWD or $OLDPWD respectively.

In addition, when expanding a variable assignment (see Variable Assignments above), tilde expansion is attempted when the value of the assignment begins with a ∼, and when a ∼ appears after a :. A : also terminates a user name following a ∼.

The tilde expansion mechanism may be extended or modified by defining one of the discipline functions .sh.tilde.set or .sh.tilde.get (see Functions and Discipline Functions below). If either exists, then upon encountering a tilde word to expand, that function is called with the tilde word assigned to either .sh.value (for the .sh.tilde.set function) or .sh.tilde (for the .sh.tilde.get function). Performing tilde expansion within a discipline function will not recursively call that function, but default tilde expansion remains active, so literal tildes should still be quoted where required. Either function may assign a replacement string to .sh.value. If this value is non-empty and does not start with a ∼, it replaces the default tilde expansion when the function terminates. Otherwise, the tilde expansion is left unchanged.

Subshells.¶

A subshell is a separate execution environment that is a complete duplicate of the current shell environment, except for two things: all traps are reset to default except those for signals that are being ignored, and subshells cannot be interactive (i.e., they have no command prompt). Changes made within a subshell do not affect the parent environment and are lost when the subshell exits.

Particular care should be taken not to confuse a subshell with a newly invoked shell that is merely a child process of the current shell, and which (unlike a subshell) starts from scratch in terms of variables and functions and may be interactive. Beware of shell tutorials on the Internet that confuse these.

Subshells are not themselves invoked as commands. Instead, the following are automatically run in or from a subshell:

Creating processes is expensive, so as a performance optimization, a subshell of a non-interactive shell may share the process of its parent environment. Such a subshell is known as a virtual subshell. Subshells are virtual unless or until something (such as asynchronous execution, or an attempt to set a process limit using the ulimit built-in command, or other implementation- or system-defined requirements) makes it necessary to fork it into a separate process. Barring any bugs in the shell, virtual subshells should be indistinguishable from real subshells except by their execution speed and their process ID. See the description of the .sh.pid variable below for more information.

Command Substitution.¶

The standard output from a command list enclosed in parentheses preceded by a dollar sign ( $(list) ), or in a brace group preceded by a dollar sign ( ${ list;} ), or in a pair of grave accents (``) may be used as part or all of a word; trailing new-lines are removed. In the second case, the { and } are treated as a reserved words so that { must be followed by a blank and } must appear at the beginning of the line or follow a ;. In the third (obsolete) form, the string between the quotes is processed for special quoting characters before the command is executed (see Quoting below). The command substitution $(cat file) can be replaced by the equivalent but faster $(<file). The command substitution $(n<#) will expand to the current byte offset for file descriptor n. Except for the second form, the command list is run in a subshell so that no side effects are possible. For the second form, the final } will be recognized as a reserved word after any token.

Arithmetic Expansion.¶

An arithmetic expression enclosed in double parentheses preceded by a dollar sign ( $(()) ) is replaced by the value of the arithmetic expression within the double parentheses.

Process Substitution.¶

Each command argument of the form <(list) or >(list) will run process list asynchronously connected to some file in /dev/fd if this directory exists, or else a fifo in a temporary directory. The name of this file will become the argument to the command. If the form with > is selected then writing on this file will provide input for list. If < is used, then the file passed as an argument will contain the output of the list process. For example,

cuts fields 1 and 3 from the files file1 and file2 respectively, pastes the results together, and sends it to the processes process1 and process2, as well as putting it onto the standard output. Note that the file, which is passed as an argument to the command, is a UNIX pipe(2) so programs that expect to lseek(2) on the file will not work.

Process substitution of the form <(list) can also be used with the < redirection operator which causes the output of list to be standard input or the input for whatever file descriptor is specified.

Parameter Expansion.¶

A parameter is a variable, one or more digits, or any of the characters *, @, #, ?, -, $, and !. A variable is denoted by a vname. To create a variable whose vname contains a ., a variable whose vname consists of everything before the last . must already exist. A variable has a value and zero or more attributes. Variables can be assigned values and attributes by using the typeset special built-in command. The attributes supported by the shell are described later with the typeset special built-in command. Exported variables pass their attributes to the environment so that a newly invoked ksh that is a child or exec'd process of the current shell will automatically import them, unless the posix shell option is on.

The shell supports both indexed and associative arrays. An element of an array variable is referenced by a subscript. A subscript for an indexed array is denoted by an arithmetic expression (see Arithmetic Evaluation below) between a [ and a ]. To assign values to an indexed array, use vname=(value ...) or set -A vname value ... . The value of all non-negative subscripts must be in the range of 0 through 4,194,303. A negative subscript is treated as an offset from the maximum current index +1 so that -1 refers to the last element. Indexed arrays can be declared with the -a option to typeset. Indexed arrays need not be declared. Any reference to a variable with a valid subscript is legal and an array will be created if necessary.

An associative array is created with the -A option to typeset. A subscript for an associative array is denoted by a string enclosed between [ and ].

Referencing any array without a subscript is equivalent to referencing the array with subscript 0.

The value of a variable may be assigned by writing:

or

Note that no space is allowed before or after the =.

Attributes assigned by the typeset special built-in command apply to all elements of the array. An array element can be a simple variable, a compound variable or an array variable. An element of an indexed array can be either an indexed array or an associative array. An element of an associative array can also be either. To refer to an array element that is part of an array element, concatenate the subscript in brackets. For example, to refer to the foobar element of an associative array that is defined as the third element of the indexed array, use ${vname[3][foobar]}

A nameref is a variable that is a reference to another variable. A nameref is created with the -n attribute of typeset. The value of the variable at the time of the typeset command becomes the variable that will be referenced whenever the nameref variable is used. The name of a nameref cannot contain a .. When a variable or function name contains a ., and the portion of the name up to the first . matches the name of a nameref, the variable referred to is obtained by replacing the nameref portion with the name of the variable referenced by the nameref. If a nameref is used as the index of a for loop, a name reference is established for each item in the list. A nameref provides a convenient way to refer to the variable inside a function whose name is passed as an argument to a function. For example, if the name of a variable is passed as the first argument to a function, the command

inside the function causes references and assignments to var to be references and assignments to the variable whose name has been passed to the function.

If any of the floating point attributes, -E, -F, or -X, or the integer attribute, -i, is set for vname, then the value is subject to arithmetic evaluation as described below.

Positional parameters, parameters denoted by a number, may be assigned values with the set special built-in command. Parameter $0 is set from argument zero when the shell is invoked.

The character $ is used to introduce substitutable parameters.

${parameter}

The shell reads all the characters from ${ to the matching } as part of the same word even if it contains braces or metacharacters. The value, if any, of the parameter is substituted. The braces are required when parameter is followed by a letter, digit, or underscore that is not to be interpreted as part of its name, when the variable name contains a .. The braces are also required when a variable is subscripted unless it is part of an Arithmetic Expression or a Conditional Expression. If parameter is one or more digits then it is a positional parameter. A positional parameter of more than one digit must be enclosed in braces. If parameter is * or @, then all the positional parameters, starting with $1, are substituted (separated by a field separator character). If an array vname with last subscript * @, or for indexed arrays of the form sub1 .. sub2. is used, then the value for each of the elements between sub1 and sub2 inclusive (or all elements for * and @) is substituted, separated by the first character of the value of IFS.

${#parameter}

If parameter is * or @, the number of positional parameters is substituted. Otherwise, the length of the value of the parameter is substituted.

${#vname[*]}

${#vname[@]}

The number of elements in the array vname is substituted.

${@vname}

Expands to the type name (See Type Variables below) or attributes of the variable referred to by vname.

${!vname}

Expands to the name of the variable referred to by vname. This will be vname except when vname is a name reference.

${!vname[subscript]}

Expands to name of the subscript unless subscript is *, @. or of the form sub1 .. sub2. When subscript is *, the list of array subscripts for vname is generated. For a variable that is not an array, the value is 0 if the variable is set. Otherwise it is the empty string. When subscript is @, same as above, except that when used in double quotes, each array subscript yields a separate argument. When subscript is of the form sub1 .. sub2 it expands to the list of subscripts between sub1 and sub2 inclusive using the same quoting rules as @.

${!prefix@}

${!prefix*}

These both expand to the names of the variables whose names begin with prefix. The expansions otherwise work like $@ and $*, respectively (see under Quoting below).

${parameter:-word}

If parameter is set and has a non-empty value, then substitute its value; otherwise substitute word.

${parameter:=word}

If parameter is not set or has the empty string value, then set it to word; the value of the parameter is then substituted. Positional parameters may not be assigned to in this way.

${parameter:?word}

If parameter is set and has a non-empty value, then substitute its value; otherwise, print word and exit from the shell (if not interactive). If word is omitted then a standard message is printed.

${parameter:+word}

If parameter is set and has a non-empty value then substitute word; otherwise substitute the empty string.

In the above, word is not evaluated unless it is to be used as the substituted string, so that, in the following example, pwd is executed only if d is not set or has the empty string value:

If the colon (:) is omitted from the above expressions, then the shell only checks whether parameter is set or not.

${parameter:offset:length}

${parameter:offset}

Expands to the portion of the value of parameter starting at the character (counting from 0) determined by expanding offset as an arithmetic expression and consisting of the number of characters determined by the arithmetic expression defined by length. In the second form, the remainder of the value is used. If A negative offset counts backwards from the end of parameter. Note that one or more blanks is required in front of a minus sign to prevent the shell from interpreting the operator as :-. If parameter is * or @, or is an array name indexed by * or @, then offset and length refer to the array index and number of elements respectively. A negative offset is taken relative to one greater than the highest subscript for indexed arrays. The order for associative arrays is unspecified.

${parameter#pattern}

${parameter##pattern}

If the shell pattern matches the beginning of the value of parameter, then the value of this expansion is the value of the parameter with the matched portion deleted; otherwise the value of this parameter is substituted. In the first form the smallest matching pattern is deleted and in the second form the largest matching pattern is deleted. When parameter is @, *, or an array variable with subscript @ or *, the substring operation is applied to each element in turn.

${parameter%pattern}

${parameter%%pattern}

If the shell pattern matches the end of the value of parameter, then the value of this expansion is the value of the parameter with the matched part deleted; otherwise substitute the value of parameter. In the first form the smallest matching pattern is deleted and in the second form the largest matching pattern is deleted. When parameter is @, *, or an array variable with subscript @ or *, the substring operation is applied to each element in turn.

${parameter/pattern/string}

${parameter//pattern/string}

${parameter/#pattern/string}

${parameter/%pattern/string}

Expands parameter and replaces the longest match of pattern with the given string. Each occurrence of \n in string is replaced by the portion of parameter that matches the n-th subpattern. In the first form, only the first occurrence of pattern is replaced. In the second form, each match for pattern is replaced by the given string. The third form restricts the pattern match to the beginning of the string while the fourth form restricts the pattern match to the end of the string. When string is empty, the pattern will be deleted and the / in front of string may be omitted. When parameter is @, *, or an array variable with subscript @ or *, the substitution operation is applied to each element in turn. In this case, the string portion of word will be re-evaluated for each element.

Shell Variables.¶

The following parameters are automatically set by the shell:

The following variables are used by the shell:

The shell gives default values to PATH, PS1, PS2, PS3, PS4, MAILCHECK, FCEDIT, TMOUT and IFS, while HOME, SHELL, ENV, histchars, and MAIL are not set at all by the shell (although HOME is set by login(1)). On some systems MAIL and SHELL are also set by login(1).

Field Splitting.¶

After parameter expansion and command substitution, the results of substitutions are scanned for the field separator characters (those found in IFS) and split into distinct fields where such characters are found. Explicit empty fields ("" or ′′) are retained. Implicit empty fields (those resulting from parameters that are unset or have empty string values or from command substitutions yielding the empty string, and that are not quoted with "") are removed.

Brace Expansion.¶

If the braceexpand (-B) option is set then each of the fields resulting from IFS are checked to see if they contain one or more of the brace patterns {*,*}, {l1..l2} , {n1..n2} , {n1..n2% fmt} , {n1..n2 ..n3} , or {n1..n2 ..n3%fmt} , where * represents any character, l1,l2 are letters and n1,n2,n3 are signed numbers and fmt is a format specified as used by printf. In each case, fields are created by prepending the characters before the { and appending the characters after the } to each of the strings generated by the characters between the { and }. The resulting fields are checked to see if they have any brace patterns.

In the first form, a field is created for each string between { and ,, between , and ,, and between , and }. The string represented by * can contain embedded matching { and } without quoting. Otherwise, each { and } with * must be quoted.

In the seconds form, l1 and l2 must both be either upper case or both be lower case characters in the C locale. In this case a field is created for each character from l1 through l2.

In the remaining forms, a field is created for each number starting at n1 and continuing until it reaches n2 incrementing n1 by n3. The cases where n3 is not specified behave as if n3 where 1 if n1<=n2 and -1 otherwise. If forms which specify %fmt any format flags, widths and precisions can be specified and fmt can end in any of the specifiers cdiouxX. For example, {a,z}{1..5..3%02d}{b..c}x expands to the 8 fields, a01bx, a01cx, a04bx, a04cx, z01bx, z01cx, z04bx and z04cx.

Pathname Expansion.¶

This is also known as globbing or sometimes filename generation. Following splitting, each field is scanned for the characters *, ?, (, and [ unless the -f option has been set. If one of these characters appears, then the word is regarded as a pattern. Each file name component that contains any pattern character is replaced with a lexicographically sorted set of names that matches the pattern from that directory. If no file name is found that matches the pattern, then that component of the filename is left unchanged unless the pattern is prefixed with ∼(N), in which case it is removed as described below. The special traversal names . and .. are never matched. If FIGNORE is set, then each file name component that matches the pattern defined by the value of FIGNORE is ignored when generating the matching filenames. If FIGNORE is not set, the character . at the start of each file name component will be ignored unless the first character of the pattern corresponding to this component is the character . itself. Note that, for uses of pattern matching other than pathname expansion, the / and . are not treated specially.

A pattern-list is a list of one or more patterns separated from each other with a & or |. A & signifies that all patterns must be matched whereas | requires that only one pattern be matched. Composite patterns can be formed with one or more of the following subpatterns:

By default, each pattern or subpattern will match the longest string possible consistent with generating the longest overall match. If more than one match is possible, the one starting closest to the beginning of the string will be chosen. However, for each of the above compound patterns, a - can be inserted in front of the ( to cause the shortest match to the specified pattern-list to be used.

When pattern-list is contained within parentheses, the backslash character \ is treated specially even when inside a character class. All ANSI C character escapes are recognized and match the specified character. In addition, the following escape sequences are recognized:

A pattern of the form %(pattern-pair(s)) is a subpattern that can be used to match nested character expressions. Each pattern-pair is a two-character sequence that cannot contain & or |. The first pattern-pair specifies the starting and ending characters for the match. Each subsequent pattern-pair represents the beginning and ending characters of a nested group that will be skipped over when counting starting and ending character matches. The behavior is unspecified when the first character of a pattern-pair is alphanumeric, except for the following:

Thus, %({}Q"E\), matches characters starting at { until the matching } is found, not counting any { or } that is inside a double-quoted string or preceded by the escape character \. Without the {}, this pattern matches any C language string.

Each subpattern in a composite pattern is numbered, starting at 1, by the location of the ( within the pattern. The sequence \n, where n is a single digit and \n comes after the nth subpattern, matches the same string as the subpattern itself.

Finally, a pattern can contain subpatterns of the form ∼(options:pattern-list), where either options or :pattern-list can be omitted. Unlike the other compound patterns, these subpatterns are not counted in the numbered subpatterns. :pattern-list must be omitted for options F, G, N, and V below. If options is present, it can consist of one or more of the following:

If both options and :pattern-list are specified, then the options apply only to pattern-list. Otherwise, these options remain in effect until they are disabled by a subsequent ∼(...) or at the end of the subpattern containing ∼(...).

Quoting.¶

Each of the metacharacters listed earlier (see Definitions above) has a special meaning to the shell and causes termination of a word unless quoted. A character may be quoted (i.e., made to stand for itself) by preceding it with a \. The pair \new-line is removed. All characters enclosed between a pair of single quote marks (′′) that is not preceded by a $ are quoted. A single quote cannot appear within the single quotes. A single quoted string preceded by an unquoted $ is processed as an ANSI C string except for the following:

Inside double quote marks (""), parameter and command substitution occur and \ quotes the characters \, `, ", and $. A $ in front of a double quoted string will be ignored in the "C" or "POSIX" locale, and may cause the string to be replaced by a locale specific string otherwise. The meaning of $* and $@ is identical when not quoted or when used as a variable assignment value or as a file name. However, when used as a command argument, "$*" is equivalent to "$1d$2d...", where d is the first character of the IFS variable, whereas "$@" is equivalent to "$1" "$2" .... Inside grave quote marks (``), \ quotes the characters \, `, and $. If the grave quotes occur within double quotes, then \ also quotes the character ".

The special meaning of reserved words or aliases can be removed by quoting any character of the reserved word. The recognition of function names or built-in command names listed below cannot be altered by quoting them.

Arithmetic Evaluation.¶

The shell performs arithmetic evaluation for arithmetic expansion, to evaluate an arithmetic command, to evaluate an indexed array subscript, and to evaluate arguments to the built-in commands shift and let as well as arguments to numeric format specifiers given to print -f and printf. Evaluations are performed using double precision floating point arithmetic or long double precision floating point for systems that provide this data type. Floating point constants follow the ANSI C programming language floating point conventions. The case-insensitive floating point constants NaN and Inf can be used to represent "not a number" and infinity respectively, unless the posix shell option is on. Integer constants follow the ANSI C programming language integer constant conventions although only single byte character constants are recognized and character casts are not recognized. In addition constants can be of the form [base#]n where base is a decimal number between two and sixty-four representing the arithmetic base and n is a number in that base. The digits above 9 are represented by the lower case letters, the upper case letters, @, and _ respectively. For bases less than or equal to 36, upper and lower case characters can be used interchangeably.

An arithmetic expression uses the same syntax, precedence, and associativity of expression as the C language. All the C language operators that apply to floating point quantities can be used. In addition, the operator ** can be used for exponentiation. It has higher precedence than multiplication and is left associative. In addition, when the value of an arithmetic variable or subexpression can be represented as a long integer, all C language integer arithmetic operations can be performed. Variables can be referenced by name within an arithmetic expression without using the parameter expansion syntax. When a variable is referenced, its value is evaluated as an arithmetic expression.

Any of the following math library functions that are in the C math library can be used within an arithmetic expression:

abs acos acosh asin asinh atan atan2 atanh cbrt ceil copysign cos cosh erf erfc exp exp10 exp2 expm1 fabs fdim finite float floor fma fmax fmin fmod fpclass fpclassify hypot ilogb int isfinite isgreater isgreaterequal isinf isinfinite isless islessequal islessgreater isnan isnormal issubnormal isunordered iszero j0 j1 jn ldexp lgamma log log10 log1p log2 logb nearbyint nextafter nexttoward pow remainder rint round scalb scalbn signbit sin sinh sqrt tan tanh tgamma trunc y0 y1 yn

In addition, arithmetic functions can be defined as shell functions with a variant of the function name syntax,

function .sh.math.name ident ... { list ;}

where name is the function name used in the arithmetic expression and each identifier, ident is a name reference to the long double precision floating point argument. The value of .sh.value when the function returns is the value of this function. User defined functions can take up to 3 arguments and override C math library functions.

An internal representation of a variable as a double precision floating point can be specified with the -E [n], -F [n], or -X [n] option of the typeset special built-in command. The -E option causes the expansion of the value to be represented using scientific notation when it is expanded. The optional option argument n defines the number of significant figures. The -F option causes the expansion to be represented as a floating decimal number when it is expanded. The -X option causes the expansion to be represented using the %a format defined by ISO C-99. The optional option argument n defines the number of places after the decimal (or radix) point in this case.

An internal integer representation of a variable can be specified with the -i [n] option of the typeset special built-in command. The optional option argument n specifies an arithmetic base to be used when expanding the variable. If you do not specify an arithmetic base, base 10 will be used.

Arithmetic evaluation is performed on the value of each assignment to a variable with the -E, -F, -X, or -i attribute. Assigning a floating point number to a variable whose type is an integer causes the fractional part to be truncated.

Prompting.¶

When used interactively, the shell prompts with the value of PS1 after expanding it for parameter expansion, command substitution, and arithmetic expansion, before reading a command. In addition, each single ! in the prompt is replaced by the command number. A !! is required to place ! in the prompt. If at any time a new-line is typed and further input is needed to complete a command, then the secondary prompt (i.e., the value of PS2) is issued.

Conditional Expressions.¶

A conditional expression is used with the [[ compound command to test attributes of files and to compare strings. Field splitting and pathname expansion are not performed on the words between [[ and ]]. Each expression can be constructed from one or more of the following unary or binary expressions:

The following obsolete arithmetic comparisons are also permitted:

In each of the above expressions, if file is of the form /dev/fd/n, where n is an integer, then the test is applied to the open file whose descriptor number is n.

A compound expression can be constructed from these primitives by using any of the following, listed in decreasing order of precedence:

Input/Output.¶

Before a command is executed, its input and output may be redirected using a special notation interpreted by the shell. The following may appear anywhere in a simple-command or may precede or follow a command and are not passed on to the invoked command. Command substitution, parameter expansion, and arithmetic expansion occur before word or digit is used except as noted below. Pathname expansion occurs only if the shell is interactive and the pattern matches a single file. Field splitting is not performed.

In each of the following redirections, if file is of the form /dev/sctp/host/port, /dev/tcp/host/port, or /dev/udp/host/port, where host is a hostname or host address, and port is a service given by name or an integer port number, then the redirection attempts to make a tcp, sctp or udp connection to the corresponding socket.

No intervening space is allowed between the characters of redirection operators.

<word

Use file word as standard input (file descriptor 0).

>word

Use file word as standard output (file descriptor 1). If the file does not exist then it is created. If the file exists, and the noclobber option is on, this causes an error; otherwise, it is truncated to zero length.

>|word

Same as >, except that it overrides the noclobber option.

>;word

Write output to a temporary file. If the command completes successfully rename it to word, otherwise, delete the temporary file. >;word cannot be used with the exec and redirect built-ins.

>>word

Use file word as standard output. If the file exists, then output is appended to it (by first seeking to the end-of-file); otherwise, the file is created.

<>word

Open file word for reading and writing as standard output. If the posix option is active, it defaults to standard input instead.

<>;word

The same as <>word except that if the command completes successfully, word is truncated to the offset at command completion. <>;word cannot be used with the exec and redirect built-ins.

<<[-]word

The shell input is read up to a line that is the same as word after any quoting has been removed, or to an end-of-file. No parameter expansion, command substitution, arithmetic expansion or pathname expansion is performed on word. The resulting document, called a here-document, becomes the standard input. If any character of word is quoted, then no interpretation is placed upon the characters of the document; otherwise, parameter expansion, command substitution, and arithmetic expansion occur, \new-line is ignored, and \ must be used to quote the characters \, $, `. If - is appended to <<, then all leading tabs are stripped from word and from the document. If # is appended to <<, then leading spaces and tabs will be stripped off the first line of the document and up to an equivalent indentation will be stripped from the remaining lines and from word. A tab stop is assumed to occur at every 8 columns for the purposes of determining the indentation.

<<<word

A short form of here document in which word becomes the contents of the here-document after any parameter expansion, command substitution, and arithmetic expansion occur.

<&digit

The standard input is duplicated from file descriptor digit (see dup(2)).

>&digit

The standard output is duplicated from file descriptor digit.

<&digit-

The file descriptor given by digit is moved to standard input.

>&digit-

The file descriptor given by digit is moved to standard output.

<&-

The standard input is closed.

>&-

The standard output is closed.

<&p

The input from the co-process is moved to standard input.

>&p

The output to the co-process is moved to standard output.

<#((expr))

Evaluate arithmetic expression expr and position file descriptor 0 to the resulting value bytes from the start of the file. The variables CUR and EOF evaluate to the current offset and end-of-file offset respectively when evaluating expr.

>#((offset))

The same as <# except applies to file descriptor 1.

<#pattern

Seeks forward to the beginning of the next line containing pattern.

<##pattern

The same as <# except that the portion of the file that is skipped is copied to standard output.

If one of the above is preceded by a digit, with no intervening space, then the file descriptor number referred to is that specified by the digit (instead of the default 0 or 1). If one of the above, other than >&- and the ># and <# forms, is preceded by {varname} with no intervening space, then a file descriptor number > 9 will be selected by the shell and stored in the variable varname, so it can be read from or written to with redirections like <& $varname or >& $varname. If >&- or the any of the ># and <# forms is preceded by {varname} the value of varname defines the file descriptor to close or position. For example:

means file descriptor 2 is to be opened for writing as a duplicate of file descriptor 1 and

means open file named file for reading and store the file descriptor number in variable n.

A special shorthand redirection operator &>word is available; it is equivalent to >word 2>&1. It cannot be preceded by any digit or variable name. This shorthand is disabled if the posix shell option is active.

The order in which redirections are specified is significant. The shell evaluates each redirection in terms of the (file descriptor, file) association at the time of evaluation. For example:

first associates file descriptor 1 with file fname. It then associates file descriptor 2 with the file associated with file descriptor 1 (i.e. fname). If the order of redirections were reversed, file descriptor 2 would be associated with the terminal (assuming file descriptor 1 had been) and then file descriptor 1 would be associated with file fname.

If a command is followed by & and job control is not active, then the default standard input for the command is the empty file /dev/null. Otherwise, the environment for the execution of a command contains the file descriptors of the invoking shell as modified by input/output specifications.

Environment.¶

The environment (see environ(7)) is a list of name-value pairs that is passed to an executed program in the same way as a normal argument list. The names must be identifiers and the values are character strings. The shell interacts with the environment in several ways. On invocation, the shell scans the environment and creates a variable for each name found, giving it the corresponding value and attributes and marking it export. Executed commands inherit the environment. If the user modifies the values of these variables or creates new ones, using the export or typeset -x commands, they become part of the environment. The environment seen by any executed command is thus composed of any name-value pairs originally inherited by the shell, whose values may be modified by the current shell, plus any additions which must be noted in export or typeset -x commands.

The environment for any simple-command or function may be augmented by prefixing it with one or more variable assignments. A variable assignment argument is a word of the form identifier=value. Thus:

are equivalent (as far as the above execution of cmd is concerned except for special built-in commands listed below - those that are marked with †).

If the obsolete -k option is set, all variable assignment arguments are placed in the environment, even if they occur after the command name. The following first prints a=b c and then c:

echo a=b c
set -k
echo a=b c

This feature is intended for use with scripts written for early versions of the shell and its use in new scripts is strongly discouraged. It is likely to disappear someday.

Functions.¶

For historical reasons, there are two ways to define functions, the name() syntax and the function name syntax, described in the Commands section above. Shell functions are read in and stored internally. Alias names are resolved when the function is read. Functions are executed like commands with the arguments passed as positional parameters. (See Execution below.)

Functions defined by the function name syntax and called by name execute in the same process as the caller and share all files and present working directory with the caller. Traps caught by the caller are reset to their default action inside the function. A trap condition that is not caught or ignored by the function causes the function to terminate and the condition to be passed on to the caller. A trap on EXIT set inside a function is executed in the environment of the caller after the function completes. Ordinarily, variables are shared between the calling program and the function. However, the typeset special built-in command used within a function defines local variables whose scope includes the current function. They can be passed to functions that they call in the variable assignment list that precedes the call or as arguments passed as name references. Errors within functions return control to the caller.

Functions defined with the name() syntax and functions defined with the function name syntax that are invoked with the . special built-in are executed in the caller's environment and share all variables and traps with the caller. Errors within these function executions cause the script that contains them to abort.

The special built-in command return is used to return from function calls.

Function names can be listed with the -f or +f option of the typeset special built-in command. The text of functions, when available, will also be listed with -f. Functions can be undefined with the -f option of the unset special built-in command.

Ordinarily, functions are unset when the shell executes a shell script. Functions that need to be defined across separate invocations of the shell should be placed in a directory and the FPATH variable should contain the name of this directory. They may also be specified in the ENV file.

Discipline Functions.¶

Each variable can have zero or more discipline functions associated with it. The shell initially understands the discipline names get, set, append, and unset but can be added when defining new types. On most systems others can be added at run time via the C programming interface extension provided by the builtin built-in utility. If the get discipline is defined for a variable, it is invoked whenever the given variable is referenced. If the variable .sh.value is assigned a value inside the discipline function, the referenced variable will evaluate to this value instead. If the set discipline is defined for a variable, it is invoked whenever the given variable is assigned a value. If the append discipline is defined for a variable, it is invoked whenever a value is appended to the given variable. The variable .sh.value is given the value of the variable before invoking the discipline, and the variable will be assigned the value of .sh.value after the discipline completes. If .sh.value is unset inside the discipline, then that value is unchanged. If the unset discipline is defined for a variable, it is invoked whenever the given variable is unset.

The variable .sh.name contains the name of the variable for which the discipline function is called, .sh.subscript is the subscript of the variable, and .sh.value will contain the value being assigned inside the set discipline function. The variable _ is a reference to the variable including the subscript if any. For the set discipline, changing .sh.value will change the value that gets assigned. Finally, the expansion ${var.name}, when name is the name of a discipline, and there is no variable of this name, is equivalent to the command substitution ${ var.name;}.

Name Spaces.¶

Commands and functions that are executed as part of the list of a namespace command that modify variables or create new ones, create a new variable whose name is the name of the name space as given by identifier preceded by .. When a variable whose name is name is referenced, it is first searched for using .identifier.name. Similarly, a function defined by a command in the namespace list is created using the name space name preceded by a ..

When the list of a namespace command contains a namespace command, the names of variables and functions that are created consist of the variable or function name preceded by the list of identifiers each preceded by ..

Outside of a name space, a variable or function created inside a name space can be referenced by preceding it with the name space name.

By default, variables starting with .sh are in the sh name space.

Type Variables.¶

Typed variables provide a way to create data structure and objects. A type can be defined either by a shared library, by the enum built-in command described below, or by using the new -T option of the typeset built-in command. With the -T option of typeset, the type name, specified as an option argument to -T, is set with a compound variable assignment that defines the type. Function definitions can appear inside the compound variable assignment and these become discipline functions for this type and can be invoked or redefined by each instance of the type. The function name create is treated specially. It is invoked for each instance of the type that is created but is not inherited and cannot be redefined for each instance.

When a type is defined a special built-in command of that name is added. These built-ins are declaration commands and follow the same expansion rules as the built-in commands described below that are marked with a ‡ symbol. These commands can subsequently be used inside further type definitions. The man page for these commands can be generated by using the --man option or any of the other -- options described with getopts. The -r, -a, -A, -h, and -S options of typeset are permitted with each of these new built-ins.

An instance of a type is created by invoking the type name followed by one or more instance names. Each instance of the type is initialized with a copy of the subvariables except for subvariables that are defined with the -S option. Variables defined with the -S are shared by all instances of the type. Each instance can change the value of any subvariable and can also define new discipline functions of the same names as those defined by the type definition as well as any standard discipline names. No additional subvariables can be defined for any instance.

When defining a type, if the value of a subvariable is not set and the -r attribute is specified, it causes the subvariable to be a required subvariable. Whenever an instance of a type is created, all required subvariables must be specified. These subvariables become read-only in each instance.

When unset is invoked on a subvariable within a type, and the -r attribute has not been specified for this field, the value is reset to the default value associative with the type. Invoking unset on a type instance not contained within another type deletes all subvariables and the variable itself.

A type definition can be derived from another type definition by defining the first subvariable name as _ and defining its type as the base type. Any remaining definitions will be additions and modifications that apply to the new type. If the new type name is the same as that of the base type, the type will be replaced and the original type will no longer be accessible.

The typeset command with the -T and no option argument or operands will write all the type definitions to standard output in a form that can be read in to create all they types.

Jobs.¶

If the monitor option of the set command is turned on, an interactive shell associates a job with each pipeline. It keeps a table of current jobs, printed by the jobs command, and assigns them small integer numbers. When a job is started asynchronously with &, the shell prints a line which looks like:

[1] 1234

indicating that the job which was started asynchronously was job number 1 and had one (top-level) process, whose process ID was 1234.

This paragraph and the next require features that are not in all versions of UNIX and may not apply. If you are running a job and wish to do something else you may hit the key ^Z (control-Z) which sends a STOP signal to the current job. The shell will then normally indicate that the job has been `Stopped', and print another prompt. You can then manipulate the state of this job, putting it in the background with the bg command, or run some other commands and then eventually bring the job back into the foreground with the foreground command fg. A ^Z takes effect immediately and is like an interrupt in that pending output and unread input are discarded when it is typed.

A job being run in the background will stop if it tries to read from the terminal. Background jobs are normally allowed to produce output, but this can be disabled by giving the command stty tostop. If you set this tty option, then background jobs will stop when they try to produce output like they do when they try to read input.

A job pool is a collection of jobs started with list & associated with a name.

There are several ways to refer to jobs in the shell. A job can be referred to by the process ID of any process of the job or by one of the following:

In addition, unless noted otherwise, wherever a job can be specified, the name of a background job pool can be used to represent all the jobs in that pool.

The shell learns immediately whenever a process changes state. It normally informs you whenever a job becomes blocked so that no further progress is possible, but only just before it prints a prompt. This is done so that it does not otherwise disturb your work. The notify option of the set command causes the shell to print these job change messages as soon as they occur.

When the monitor option is on, each background job that completes triggers any trap set for CHLD.

When you try to leave the shell while jobs are running or stopped, you will be warned that `You have stopped(running) jobs.' You may use the jobs command to see what they are. If you immediately try to exit again, the shell will not warn you a second time, and the stopped jobs will be terminated. When a login shell receives a HUP signal, it sends a HUP signal to each job that has not been disowned with the disown built-in command described below.

Signals.¶

The INT and QUIT signals for an invoked command are ignored if the command is followed by & and the monitor option is not active. Otherwise, signals have the values inherited by the shell from its parent (but see also the trap built-in command below).

Execution.¶

Each time a command is read, the above expansions and substitutions are carried out. If the command name matches one of the Special Built-in Commands listed below, it is executed within the current shell process. Next, the command name is checked to see if it matches a user defined function. If it does, the positional parameters are saved and then reset to the arguments of the function call. A function is also executed in the current shell process. When the function completes or issues a return, the positional parameter list is restored. For functions defined with the function name syntax, any trap set on EXIT within the function is executed. The exit value of a function is the value of the last command executed. If a command name is not a special built-in command or a user defined function, but it is one of the built-in commands listed below, it is executed in the current shell process.

The shell variables PATH followed by the variable FPATH defines the list of directories to search for the command name. Alternative directory names are separated by a colon (:). The default path is the value that was output by getconf PATH at the time ksh was compiled. The current directory can be specified by two or more adjacent colons, or by a colon at the beginning or end of the path list. If the command name contains a /, then the search path is not used. Otherwise, each directory in the list of directories defined by PATH and FPATH is checked in order. If the directory being searched is contained in FPATH and contains a file whose name matches the command being searched, then this file is loaded into the current shell environment as if it were the argument to the . command except that only preset aliases are expanded, and a function of the given name is executed as described above.

If this directory is not in FPATH the shell first determines whether there is a built-in version of a command corresponding to a given pathname and if so it is invoked in the current process. If no built-in is found, the shell checks for a file named .paths in this directory. If found and there is a line of the form FPATH=path where path names an existing directory then that directory is searched immediately after the current directory as if it were found in the FPATH variable. If path does not begin with /, it is checked for relative to the directory being searched.

The .paths file is then checked for a line of the form PLUGIN_LIB=libname [ : libname ] ... . Each library named by libname will be searched for as if it were an option argument to builtin -f, and if it contains a built-in of the specified name this will be executed instead of a command by this name. Any built-in loaded from a library found this way will be associated with the directory containing the .paths file so it will only execute if not found in an earlier directory.

Finally, the directory will be checked for a file of the given name. If the file has execute permission but is not an a.out file, it is assumed to be a file containing shell commands. A separate shell is spawned to read it. All non-exported variables are removed in this case. If the shell command file doesn't have read permission, or if the setuid and/or setgid bits are set on the file, then the shell executes an agent whose job it is to set up the permissions and execute the shell with the shell command file passed down as an open file. If the .paths contains a line of the form name=value in the first or second line, then the environment variable name is modified by prepending the directory specified by value to the directory list. If value is not an absolute directory, then it specifies a directory relative to the directory that the executable was found. If the environment variable name does not already exist it will be added to the environment list for the specified command. A parenthesized command is executed in a subshell without removing non-exported variables.

Command Re-entry.¶

The text of the last HISTSIZE (default 512) commands entered from a terminal device is saved in a history file. The file $HOME/.sh_history is used if the HISTFILE variable is not set or if the file it names is not writable. A shell can access the commands of all interactive shells which use the same named HISTFILE. The built-in command hist is used to list or edit a portion of this file. The portion of the file to be edited or listed can be selected by number or by giving the first character or characters of the command. A single command or range of commands can be specified. If you do not specify an editor program as an argument to hist then the value of the variable HISTEDIT is used. If HISTEDIT is unset, the obsolete variable FCEDIT is used. If FCEDIT is not defined, then /bin/ed is used. The edited command(s) is printed and re-executed upon leaving the editor unless you quit without writing. The -s option (and in obsolete versions, the editor name -) is used to skip the editing phase and to re-execute the command. In this case a substitution parameter of the form old=new can be used to modify the command before execution. For example, with the preset alias r, which is aliased to ′hist -s′, typing `r bad=good c' will re-execute the most recent command which starts with the letter c, replacing the first occurrence of the string bad with the string good.

History Expansion.¶

History expansions introduce words from the history list into the input stream, making it easy to repeat commands, repeat arguments of a previous command in the current command, or fix typos in the previous command. The history expansion facility is an alternative to history control via the fc or hist built-in command. To enable it, turn on the -H or histexpand option using the set command (see Built-in Commands below).

History expansions begin with the character !. They may begin anywhere in the input. The ! may be preceded by a \ or enclosed in single quotes to prevent its special meaning. A ! is also passed unchanged when it is followed by a space, tab, newline, = or (. History expansions do not nest. They are parsed separately before the shell parser is invoked, so they can override shell grammar rules.

By default, the expanded version of any line that contains a history expansion is printed, added to the history, and then immediately executed. History expansions are never added to the history themselves, regardless of whether they succeed or fail due to an error. Normally, this means that a command line with an erroneous history expansion is lost and needs to be retyped from scratch, but if the histreedit shell option is turned on and a line editor is active (see In-line Editing Options below), the erroneous line is pre-filled into the next prompt's input buffer for correcting. The histverify option causes the same to be done for successful history expansions, allowing verification and editing before execution.

A history expansion may have an event specification, which indicates the event from which words are to be taken, a word designator, which selects particular words from the chosen event, and/or a modifier, which manipulates the selected words.

An event specification can be:

For example, consider this bit of someone's history list as might be output by the hist -l command:

9 nroff -man wumpus.man
10 cp wumpus.man wumpus.man.old
11 vi wumpus.man
12 diff wumpus.man.old wumpus.man

The commands are shown with their event numbers. The current event, which we haven't typed in yet, is event 13. !11 and !-2 refer to event 11. !! refers to the previous event, 12. !! can be abbreviated ! if it is followed by : (see below). !n refers to event 9, which begins with n. !?old? also refers to event 12, which contains old. Without word designators or modifiers, history references simply expand to the entire event, so we might type !cp to redo the copy command or !!|more if the diff output scrolled off the top of the screen.

To select words from an event, the event specification can be followed by a : and a designator for the desired words. The words of an input line are numbered from 0, the first word (usually the command name) being 0, the second word (first argument) being 1, etc. The basic word designators are:

Selected words are inserted into the command line separated by single blanks. For example, the diff command in the previous example might have been typed as diff !!:1.old !!:1 (using :1 to select the first argument from the previous event) or diff !-2:2 !-2:1 to select and swap the arguments from the cp command. If we didn't care about the order of the diff, we might have said diff !-2:1-2 or simply diff !-2:*. The cp command might have been written cp wumpus.man !#:1.old, using # to refer to the current event. !n:- hurkle.man would reuse the first two words from the nroff command to say nroff -man hurkle.man.

The : separating the event specification from the word designator can be omitted if the argument selector begins with a ^, $, *, % or -. For example, our diff command might have been diff !!^.old !!^ or, equivalently, diff !!$.old !!$. However, if !! is abbreviated !, an argument selector beginning with - will be interpreted as an event specification.

The word(s) in a history reference can be edited by following them with one or more modifiers, each preceded by a colon (:):

Modifiers are applied to only the first modifiable word (unless g or a is used). It is an error for no word to be modifiable.

For example, the diff command might have been written as diff wumpus.man.old !#^:r, using :r to remove .old from the first argument on the same line (!#^). We might follow mail -s "I forgot my password" rot with !:s/rot/root to correct the spelling of root.

History expansions also occur when an input line begins with ^. When it is the first character on an input line, it is an abbreviation of !:s^. Thus we might have said ^rot^root to make the spelling correction in the previous example. This is the only history expansion that does not explicitly begin with !.

If a word on a command line begins with the history comment character #, history expansion is ignored for the rest of that line. This usually causes the shell parser (which uses the same character to signal a comment) to treat the rest of the line as a comment as well, but as history expansion is parsed separately from the shell grammar and with different rules, this cannot be guaranteed in all cases. If the history comment character is changed, the shell grammar comment character does not change along with it.

The three characters used to signal history expansion can be changed using the histchars shell variable; see Shell Variables above.

In-line Editing Options.¶

Normally, each command line entered from a terminal device is simply typed followed by a new-line (`RETURN' or `LINE FEED'). If either the emacs, gmacs, or vi option is active, the user can edit the command line. To be in either of these edit modes, set the corresponding option. An editing option is automatically selected each time the VISUAL or EDITOR variable is assigned a value matching any of these editor names; for details, see Shell Variables above under VISUAL.

The editing features require that the user's terminal accept `RETURN' as carriage return without line feed and that a space (` ') must overwrite the current character on the screen.

Unless the multiline option is on, the editing modes implement a concept where the user is looking through a window at the current line. The window width is the value of COLUMNS if it is defined, otherwise 80. If the window width is too small to display the prompt and leave at least 8 columns to enter input, the prompt is truncated from the left. If the line is longer than the window width minus two, a mark is displayed at the end of the window to notify the user. As the cursor moves and reaches the window boundaries, the window will be centered about the cursor. The mark is a > (<, *) if the line extends on the right (left, both) side(s) of the window.

The search commands in each edit mode provide access to the history file. Only strings are matched, not patterns, although a leading ^ in the string restricts the match to begin at the first character in the line.

Each of the edit modes has an operation to list the files or commands that match a partially entered word. When applied to the first word on the line, or the first word after a ;, |, &, or (, and the word does not begin with ∼ or contain a /, the list of aliases, functions, and executable commands defined by the PATH variable that could match the partial word is displayed. Otherwise, the list of files that match the given word is displayed. If the partially entered word does not contain any file expansion characters, a * is appended before generating these lists. After displaying the generated list, the input line is redrawn. These operations are called command name listing and file name listing, respectively. There are additional operations, referred to as command name completion and file name completion, which compute the list of matching commands or files, but instead of printing the list, replace the current word with a complete or partial match. For file name completion, if the match is unique, a / is appended if the file is a directory and a space is appended if the file is not a directory. Otherwise, the longest common prefix for all the matching files replaces the word. For command name completion, only the portion of the file names after the last / are used to find the longest command prefix. If only a single name matches this prefix, then the word is replaced with the command name followed by a space. When using a tab for completion that does not yield a unique match, a subsequent tab will provide a numbered list of matching alternatives. A specific selection can be made by entering the selection number followed by a tab. Neither completion nor listing operations are attempted before the first character in a line.

Key Bindings.¶

The KEYBD trap can be used to intercept keys as they are typed and change the characters that are actually seen by the shell. This trap is executed after each character (or sequence of characters when the first character is ESC) is entered while reading from a terminal. The variable .sh.edchar contains the character or character sequence which generated the trap. Changing the value of .sh.edchar in the trap action causes the shell to behave as if the new value were entered from the keyboard rather than the original value.

The variable .sh.edcol is set to the input column number of the cursor at the time of the input. The variable .sh.edmode is set to ESC when in vi insert mode (see below) and set to the empty string otherwise. Prepending ${.sh.editmode} to a value assigned to .sh.edchar will cause the shell to change to control mode if it is not already in this mode.

This trap is not invoked for characters entered as arguments to editing directives, or while reading input for a character search.

Emacs Editing Mode.¶

This mode is entered by enabling either the emacs or gmacs option. The only difference between these two modes is the way they handle ^T. To edit, the user moves the cursor to the point needing correction and then inserts or deletes characters or words as needed. All the editing commands are control characters or escape sequences. The notation for control characters is caret (^) followed by the character. For example, ^F is the notation for control F. This is entered by depressing `f' while holding down the `CTRL' (control) key. The `SHIFT' key is not depressed. (The notation ^? indicates the DEL (delete) key.)

The notation for escape sequences is M- followed by a character. For example, M-f (pronounced Meta f) is entered by depressing ESC (ASCII 033) followed by `f'. (M-F would be the notation for ESC followed by `SHIFT' (capital) `F'.)

All edit commands operate from any place on the line (not just at the beginning). Neither the `RETURN' nor the `LINE FEED' key is entered after edit commands except when noted.

The M-[ multi-character commands below are DEC VT220 escape sequences generated by special keys on standard PC keyboards, such as the arrow keys. You could type them directly but they are meant to recognize the keys in question, which are indicated in parentheses.

^F

Move cursor forward (right) one character.

M-[C

(Right arrow) Same as ^F.

M-f

Move cursor forward one word. (The emacs editor's idea of a word is a string of characters consisting of only letters, digits and underscores.)

M-[1;3C

(Alt-Right arrow) Same as M-f.

M-[1;5C

(Ctrl-Right arrow) Same as M-f.

M-[1;9C

(iTerm2 Alt-Right arrow) Same as M-f.

^B

Move cursor backward (left) one character.

M-[D

(Left arrow) Same as ^B.

M-b

Move cursor backward one word.

M-[1;3D

(Alt-Left arrow) Same as M-b.

M-[1;5D

(Ctrl-Left arrow) Same as M-b.

M-[1;9D

(iTerm2 Alt-Left arrow) Same as M-b.

^A

Move cursor to start of line.

M-[H

(Home) Same as ^A.

M-[1~

Same as ^A.

M-[7~

Same as ^A.

^E

Move cursor to end of line.

M-[F

(End) Same as ^E.

M-[4~

Same as ^E.

M-[8~

Same as ^E.

M-[Y

Same as ^E.

M-OA

(Up Arrow) Same as M-[A.

M-OB

(Down Arrow) Same as M-[B.

M-OC

(Right Arrow) Same as M-[C.

M-OD

(Left Arrow) Same as M-[D.

M-O5C

(Ctrl-Right Arrow) Same as M-f.

M-O5D

(Ctrl-Left Arrow) Same as M-b.

^]char

Move cursor forward to character char on current line.

M-^]char

Move cursor backward to character char on current line.

^X^X

Interchange the cursor and mark.

erase

(User defined erase character as defined by the stty(1) command, usually ^H .) Delete previous character.

lnext

(User defined literal next character as defined by the stty(1) command, or ^V if not defined.) Removes the next character's editing features (if any).

^D

Delete current character.

M-[3~

(Forward delete) Same as ^D.

M-d

Delete current word.

M-[3;5~

(Ctrl-Delete) Same as M-d.

M-^H

(Meta-backspace) Delete previous word.

M-h

Delete previous word.

M-^?

(Meta-DEL) Delete previous word (if your interrupt character is ^? (DEL, the default) then this command will not work).

^T

Transpose current character with previous character and advance the cursor in emacs mode. Transpose two previous characters in gmacs mode.

^C

Capitalize current character.

M-c

Capitalize current word.

M-l

Change the current word to lower case.

^K

Delete from the cursor to the end of the line. If preceded by a numerical parameter whose value is less than the current cursor position, then delete from given position up to the cursor. If preceded by a numerical parameter whose value is greater than the current cursor position, then delete from cursor up to given cursor position.

^W

Kill from the cursor to the mark.

M-p

Push the region from the cursor to the mark on the stack.

kill

(User defined kill character as defined by the stty command, usually ^U .) Kill the entire current line. If two kill characters are entered in succession, all kill characters from then on cause a line feed (useful when using paper terminals). A subsequent pair of kill characters undoes this change.

^Y

Restore last item removed from line. (Yank item back to the line.)

^L

Line feed and print current line.

M-^L

Clear the screen.

^@

(Null character) Set mark.

M-space

(Meta space) Set mark.

^J

(New line) Execute the current line.

^M

(Return) Execute the current line.

eof

End-of-file character, normally ^D, is processed as an End-of-file only if the current line is empty.

^P

Fetch previous command. Each time ^P is entered the previous command back in time is accessed. Moves back one line when not on the first line of a multi-line command.

M-[A

(Up arrow) If the cursor is at the end of the line, it is equivalent to ^R with string set to the contents of the current line. Otherwise, it is equivalent to ^P.

M-<

Fetch the least recent (oldest) history line.

M->

Fetch the most recent (youngest) history line.

^N

Fetch next command line. Each time ^N is entered the next command line forward in time is accessed.

M-[B

(Down arrow) Equivalent to ^N.

^Rstring

Reverse search history for a previous command line containing string. If a parameter of zero is given, the search is forward. String is terminated by a `RETURN' or `NEW LINE'. If string is preceded by a ^, the matched line must begin with string. If string is omitted, then the next command line containing the most recent string is accessed. In this case a parameter of zero reverses the direction of the search.

^G

Exit reverse search mode.

^O

Operate - Execute the current line and fetch the next line relative to current line from the history file.

M-digits

(Escape) Define numeric parameter, the digits are taken as a parameter to the next command. The commands that accept a parameter are ^F, ^B, erase, ^C, ^D, ^K, ^R, ^P, ^N, ^], M-., M-^], M-_, M-=, M-b, M-c, M-d, M-f, M-h, M-l, M-^H, and the arrow keys and forward-delete key.

M-letter

Soft-key - Your alias list is searched for an alias by the name _letter and if an alias of this name is defined, its value will be inserted on the input queue. The letter must not be one of the above meta-functions.

M-[letter

Soft-key - Your alias list is searched for an alias by the name __letter and if an alias of this name is defined, its value will be inserted on the input queue. This can be used to program function keys on many terminals.

M-.

The last word of the previous command is inserted on the line. If preceded by a numeric parameter, the value of this parameter determines which word to insert rather than the last word.

M-_

Same as M-..

M-*

Attempt pathname expansion on the current word. An asterisk is appended if the word doesn't match any file or contain any special pattern characters.

M-ESC

Command or file name completion as described above.

^I tab

Attempts command or file name completion as described above. If a partial completion occurs, repeating this will behave as if M-= were entered. If no match is found or entered after space, a tab is inserted.

M-=

If not preceded by a numeric parameter, it generates the list of matching commands or file names as described above. Otherwise, the word under the cursor is replaced by the item corresponding to the value of the numeric parameter from the most recently generated command or file list. If the cursor is not on a word, it is inserted instead.

^U

Multiply parameter of next command by 4.

\

If the backslashctrl shell option is on (which is the default setting), this escapes the next character. Editing characters, the user's erase, kill and interrupt (normally ^C) characters may be entered in a command line or in a search string if preceded by a \. The \ removes the next character's editing features (if any). See also lnext which is not subject to any shell option.

M-^V

Display version of the shell.

M-[2~

(Insert) Escape the next character.

M-#

If the line does not begin with a #, a # is inserted at the beginning of the line and after each new-line, and the line is entered. This causes a comment to be inserted in the history file. If the line begins with a #, the # is deleted and one # after each new-line is also deleted.

Vi Editing Mode.¶

There are two typing modes. Initially, when you enter a command you are in the input mode. To edit, the user enters control mode by typing ESC (033) and moves the cursor to the point needing correction and then inserts or deletes characters or words as needed. Most control commands accept an optional repeat count prior to the command.

The notation for control characters used below is ^ followed by a character. For instance, ^H is entered by holding down the Control key and pressing H. ^[ (Control+[) is equivalent to the ESC key. The notation for escape sequences is ^[ followed by one or more characters.

The ^[[ (ESC [) multi-character commands below are DEC VT220 escape sequences generated by special keys on standard PC keyboards, such as the arrow keys, which are indicated in parentheses. When in input mode, these keys will switch you to control mode before performing the associated action. These sequences can use preceding repeat count parameters, but only when the ^[ and the subsequent [ are entered into the input buffer at the same time, such as when pressing one of those keys.

     Input Edit Commands¶

     Motion Edit Commands¶

     Search Edit Commands¶

     Text Modification Edit Commands¶

     Other Edit Commands¶

Built-in Commands.¶

The simple-commands listed below are built in to the shell and are executed in the same process as the shell. The effects of any added Input/Output redirections are local to the command, except for the exec and redirect commands. Unless otherwise indicated, the output is written on standard output (file descriptor 1) and the exit status, when there is no syntax error, is zero. Except for :, true, false, and echo, all built-in commands accept -- to indicate end of options, and are self-documenting.

The self-documenting commands interpret the option --man as a request to display that command's own manual page, --help as a request to display the OPTIONS section from their manual page, and -? as a request to print a brief usage message. All these are processed as error messages, so they are written on standard error (file descriptor 2) and to pipe them into a pager such as more(1) you need to add a 2>&1 redirection before the |. The display of boldface text depends on whether standard error is on a terminal, so is disabled when using a pager. Exporting the ERROR_OPTIONS environment variable with a value containing emphasis will force this on; a value containing noemphasis forces it off. The test/[ command needs an additional -- argument to recognize self-documentation options, e.g. test --man --. The exec and redirect commands, as they make redirections permanent, should use self-documentation options in a subshell when redirecting, for example: (redirect --man) 2>&1. There are advanced output options as well; see getopts --man for more information.

Commands that are preceded by a † symbol below are special built-in commands and are treated specially in the following ways:

Commands that are preceded by a ‡ symbol below are declaration commands. Any following words that are in the format of a variable assignment are expanded with the same rules as a variable assignment. This means that tilde expansion is performed after the = sign, array assignments of the form varname=(assign_list) are supported, and field splitting and pathname expansion are not performed.

† : [ arg ... ]

The command only expands parameters.

† . name [ arg ... ]

If name is a function defined with the function name reserved word syntax, the function is executed in the current environment (as if it had been defined with the name() syntax). Otherwise if name refers to a file, the file is read in its entirety and the commands are executed in the current shell environment. The search path specified by PATH is used to find the directory containing the file. If any arguments arg are given, they become the positional parameters while processing the . command and the original positional parameters are restored upon completion. Otherwise the positional parameters are unchanged. The exit status is the exit status of the last command executed.

[ expression ]

The [ command is the same as test, with the exception that an additional closing ] argument is required. See test below.

alias [ -ptx ] [ name[ =value ] ] ...

alias with no arguments prints the list of aliases in the form name=value on standard output. The -p option causes the word alias to be inserted before each one. When one or more arguments are given, an alias is defined for each name whose value is given. A trailing space in value causes the next word to be checked for alias substitution. With the -t option, each name is looked up as a command in $PATH and its path is added to the hash table as a 'tracked alias'. If no name is given, this prints the hash table. See hash. Without the -t option, for each name in the argument list for which no value is given, the name and value of the alias is printed. The obsolete -x option has no effect in most contexts, although if it's used with -t it will suppress all output. The exit status is non-zero if a name is given, but no value, and no alias has been defined for the name.

autoload name ...

Marks each name undefined so that the FPATH variable will be searched to find the function definition when the function is referenced. The same as typeset -fu.

bg [ job... ]

This command is only on systems that support job control. Puts each specified job into the background. The current job is put in the background if job is not specified. See Jobs for a description of the format of job.

† break [ n ]

Exit from the enclosing for, while, until, or select loop, if any. If n is specified, then break n levels.

builtin [ -ds ] [ -f file ] [ name ... ]

If name is not specified, and no -f option is specified, the built-ins are printed on standard output. The -s option prints only the special built-ins. Otherwise, each name represents the pathname whose basename is the name of the built-in. The entry point function name is determined by prepending b_ to the built-in name. A built-in specified by a pathname will only be executed when that pathname would be found during the path search. Built-ins found in libraries loaded via the .paths file will associate with the pathname of the directory containing the .paths file.

The ISO C/C++ prototype is b_mycommand(int argc, char *argv[], void *context) for the builtin command mycommand where argv is array an of argc elements and context is an optional pointer to a Shell_t structure as described in <ast/shell.h>.

Special built-ins cannot be bound to a pathname or deleted. The -d option deletes each of the given built-ins. On systems that support dynamic loading, the -f option names a shared library containing the code for built-ins. The shared library prefix and/or suffix, which depend on the system, can be omitted. Once a library is loaded, its symbols become available for subsequent invocations of builtin. Multiple libraries can be specified with separate invocations of the builtin command. Libraries are searched in the reverse order in which they are specified. When a library is loaded, it looks for a function in the library whose name is lib_init() and invokes this function with an argument of 0.

cd [ -L ] [ -eP ] [ arg ]

cd [ -L ] [ -eP ] old new

This command can be in either of two forms. In the first form it changes the current directory to arg. If arg is - the directory is changed to the previous directory. The shell variable HOME is the default arg. The variable PWD is set to the current directory. The shell variable CDPATH defines the search path for the directory containing arg. Alternative directory names are separated by a colon (:). The default path is the empty string (specifying the current directory). Note that the current directory may be specified by a dot (.) or by an empty path name, either of which can appear immediately after the equal sign or between the colon delimiters anywhere else in the path list. If arg begins with a / then the search path is not used. Otherwise, each directory in the path is searched for arg.

The second form of cd substitutes the string new for the string old in the current directory name, PWD, and tries to change to this new directory.

By default, symbolic link names are treated literally when finding the directory name. This is equivalent to the -L option. The -P option causes symbolic links to be resolved when determining the directory. The last instance of -L or -P on the command line determines which method is used.

If -e and -P are both in effect and the correct PWD could not be determined after successfully changing the directory, cd will return with exit status one and produce no output. If any other error occurs while both flags are active, the exit status is greater than one.

The cd command may not be executed by rksh.

command [ -pvxV ] name [ arg ... ]

With the -v option, command is equivalent to the built-in whence command described below. The -V option causes command to act like whence -v.

Without the -v or -V options, command executes name with the arguments given by arg. Functions and aliases will not be searched for when finding name. If name refers to a special built-in, as marked with † in this manual, command disables the special properties described above for that mark, executing the command as a regular built-in. (For example, using command set -o option-name prevents a script from terminating when an invalid option name is given.)

The -p option causes the operating system's standard utilities path (as output by getconf PATH) to be searched rather than the one defined by the value of PATH.

The -x option runs name as an external command, bypassing built-ins. If the arguments contain at least one word that expands to multiple arguments, such as "$@" or *.txt, then the -x option also allows executing external commands with argument lists that are longer than the operating system allows. This functionality is similar to xargs(1) but is easier to use. The shell does this by invoking the external command multiple times if needed, dividing the expanded argument list over the invocations. Any arguments that come before the first word that expands to multiple arguments, as well as any that follow the last such word, are considered static arguments and are repeated for each invocation. This allows each invocation to use the same command options, as well as the same trailing destination arguments for commands like cp(1) or mv(1). When all invocations are completed, command -x exits with the status of the invocation that had the highest exit status. (Note that command -x may still fail with an "argument list too long" error if a single argument exceeds the maximum length of the argument list, or if a long arguments list contains no word that expands to multiple arguments.)

‡ compound vname[=value] ...

Causes each vname to be a compound variable. The same as typeset -C.

† continue [ n ]

Resume the next iteration of the enclosing for, while, until, or select loop. If n is specified, then resume at the n-th enclosing loop.

disown [ job... ]

Causes the shell not to send a HUP signal to each given job, or all active jobs if job is omitted, when a login shell terminates.

echo [ arg ... ]

When the first arg does not begin with a -, and none of the arguments contain a \, then echo prints each of its arguments separated by a space and terminated by a new-line. Otherwise, the behavior of echo is system dependent and print or printf described below should be used. See echo(1) for usage and description.

‡ enum [ -i ] type[=(value ...) ] ...

Creates, for each type specified, an enumeration type declaration command named type. Variables of the created type can only store any one of the values given. For example, enum bool=(false true) creates a Boolean variable type of which variables may be declared like bool x=true y=false. If =(value ...) is omitted, then type must be an indexed array variable with at least two elements and the values are taken from this array variable. If -i is specified the values are case-insensitive. Declaration commands are created as special builtins that cannot be removed or overridden by shell functions. Each created declaration command has a --man option that shows documentation on its type's behavior and possible values.

† eval [ arg ... ]

The arguments are read as input to the shell and the resulting command(s) executed.

† exec [ -c ] [ -a name ] [ arg ... ]

If arg is given, the command specified by the arguments is executed in place of this shell without creating a new process. The value of the SHLVL environment variable is decreased by one, unless the shell replaced is a subshell. The -c option causes the environment to be cleared before applying variable assignments associated with the exec invocation. The -a option causes name rather than the first arg, to become argv[0] for the new process. If arg is not given and only I/O redirections are given, then this command persistently modifies file descriptors as in redirect.

† exit [ n ]

Causes the shell to exit with the exit status specified by n. The value will be the least significant 8 bits of n (if specified) or of the exit status of the last command executed. An end-of-file will also cause the shell to exit, except for an interactive shell that has the ignoreeof option turned on (see set below).

†‡ export [ -p ] [ name[=value] ] ...

If name is not given, the names and values of each variable with the export attribute are printed with the values quoted in a manner that allows them to be re-input. The export command is the same as typeset -x except that if you use export within a function, no local variable is created. The -p option causes the word export to be inserted before each one. Otherwise, the given names are marked for automatic export to the environment of subsequently-executed commands.

false

Does nothing, and exits 1. Used with until for infinite loops.

fc [ -e ename  ] [ -N num ] [ -nlr ] [ first [ last ] ]

fc -s [ old=new ] [ command ]

The same as hist.

fg [ job... ]

This command is only on systems that support job control. Each job specified is brought to the foreground and waited for in the specified order. Otherwise, the current job is brought into the foreground. See Jobs for a description of the format of job.

‡ float vname[=value] ...

Declares each vname to be a long floating point number. The same as typeset -lE.

functions [ -Stux ] [ name ... ]

Lists functions. The same as typeset -f.

getconf [ name [ pathname ] ]

Prints the current value of the configuration parameter given by name. The configuration parameters are defined by the IEEE POSIX 1003.1 and IEEE POSIX 1003.2 standards. (See pathconf(2) and sysconf(3).) The pathname argument is required for parameters whose value depends on the location in the file system. If no arguments are given, getconf prints the names and values of the current configuration parameters. The pathname / is used for each of the parameters that requires pathname.

getopts [ -a name ] optstring vname [ arg ... ]

Checks arg for legal options. If arg is omitted, the positional parameters are used. An option argument begins with a + or a -. An option not beginning with + or - or the argument -- ends the options. Options beginning with + are only recognized when optstring begins with a +. optstring contains the letters that getopts recognizes. If a letter is followed by a :, that option is expected to have an argument. The options can be separated from the argument by blanks. The option -? causes getopts to generate a usage message on standard error. The -a argument can be used to specify the name to use for the usage message, which defaults to $0.

getopts places the next option letter it finds inside variable vname each time it is invoked. The option letter will be prepended with a + when arg begins with a +. The index of the next arg is stored in OPTIND. The option argument, if any, gets stored in OPTARG.

A leading : in optstring causes getopts to store the letter of an invalid option in OPTARG, and to set vname to ? for an unknown option and to : when a required option argument is missing. Otherwise, getopts prints an error message. The exit status is non-zero when there are no more options.

There is no way to specify any of the options :, +, -, ?, [, and ]. The option # can only be specified as the first option.

hash [ -r ] [ utility ]

hash displays or modifies the hash table with the locations of recently used programs. If given no arguments, it lists all command/path associations (a.k.a. 'tracked aliases') in the hash table. Otherwise, hash performs a PATH search for each utility supplied and adds the result to the hash table. The -r option empties the hash table. This can also be achieved by resetting PATH.

hist [ -e ename  ] [ -N num ] [ -nlr ] [ first [ last ] ]

hist -s [ old=new ] [ command ]

In the first form, a range of commands from first to last is selected from the last HISTSIZE commands that were typed at the terminal. The arguments first and last may be specified as a number or as a string. A string is used to locate the most recent command starting with the given string. A negative number is used as an offset to the current command number. If the -l option is selected, the commands are listed on standard output. Otherwise, the editor program ename is invoked on a file containing these keyboard commands. If ename is not supplied, then the value of the variable HISTEDIT is used. If HISTEDIT is not set, then FCEDIT (default /bin/ed) is used as the editor. When editing is complete, the edited command(s) is executed if the changes have been saved. If last is not specified, then it will be set to first. If first is not specified, the default is the previous command for editing and -16 for listing. The option -r reverses the order of the commands and the option -n suppresses command numbers when listing. In the second form, command is interpreted as first described above and defaults to the last command executed. The resulting command is executed after the optional substitution old=new is performed. The option -N causes hist to start num commands back.

‡ integer vname[=value] ...

Declares each vname to be a long integer number. The same as typeset -li.

jobs [ -lnp ] [ job ... ]

Lists information about each given job; or all active jobs if job is omitted. The -l option lists process IDs in addition to the normal information. The -n option only displays jobs that have stopped or exited since last notified. The -p option causes only the process group to be listed. See Jobs for a description of the format of job.

kill [ -s signame ] job ...

kill [ -n signum ] job ...

kill -Ll [ sig ... ]

Sends either the TERM (terminate) signal or the specified signal to the specified jobs or processes. Signals are either given by number with the -n option or by name with the -s option (as given in <signal.h>, stripped of the prefix ``SIG'' with the exception that SIGCLD is named CHLD). For backward compatibility, the n and s can be omitted and the number or name placed immediately after the -. If the signal being sent is TERM (terminate) or HUP (hangup), then the job or process will be sent a CONT (continue) signal if it is stopped. The argument job can be the process ID of a process that is not a member of one of the active jobs. See Jobs for a description of the format of job. In the third form, kill -l or kill -L, if sig is not specified, the signal names are listed. The -l option lists only the signal names. The -L option lists each signal name and corresponding number. Otherwise, for each sig that is a name, the corresponding signal number is listed. For each sig that is a number, the signal name corresponding to the least significant 8 bits of sig is listed.

let arg ...

Each arg is a separate arithmetic expression to be evaluated. let only recognizes octal numbers starting with 0 when the set option letoctal is on. See Arithmetic Evaluation above for a description of arithmetic expression evaluation.

The exit status is 0 if the value of the last expression is non-zero, and 1 otherwise.

‡ nameref vname[=refname] ...

Declares each vname to be a variable name reference. The same as typeset -n.

print [ -CRenprsv ] [ -u unit ] [ -f format ] [ arg ... ]

With no options or with option - or --, each arg is printed on standard output. The -f option causes the arguments to be printed as described by printf. In this case, any e, n, r, R options are ignored. Otherwise, unless the -C, -R, -r, or -v are specified, the following escape conventions will be applied:

printf [ -v vname ] format [ arg ... ]

The arguments arg are printed on standard output in accordance with the ANSI C formatting rules associated with the format string format. If the number of arguments exceeds the number of format specifications, the format string is reused to format remaining arguments. The following extensions can also be used:

pwd [ -LP ]

Outputs the value of the current working directory. The -L option is the default; it prints the logical name of the current directory. If the -P option is given, all symbolic links are resolved from the name. The last instance of -L or -P on the command line determines which method is used.

read [ -ACSaprsv ] [ -d delim ] [ -n n ] [ -N n ] [ -t timeout ] [ -u unit ] [ vname?prompt ] [ vname ... ]

The shell input mechanism. One line is read and is broken up into fields using the characters in IFS as separators. The escape character, \, is used to remove any special meaning for the next character and for line continuation. The first field is assigned to the first vname, the second field to the second vname, etc., with leftover fields assigned to the last vname. If vname is omitted, then REPLY is used as the default vname. When vname has the binary attribute and -n or -N is specified, the bytes that are read are stored directly into the variable. If you append ?prompt to the first vname, then read will display prompt on standard error before reading if standard input is a terminal or pipe; the ? should be quoted to protect it from pathname expansion. The exit status is 0 unless an end-of-file is encountered or read has timed out. The options for the read command have meaning as follows:

†‡ readonly [ -p ] [ vname[=value] ] ...

If vname is not given, the names and values of each variable with the read-only attribute is printed with the values quoted in a manner that allows them to be re-input. The -p option causes the word readonly to be inserted before each one. Otherwise, the given vnames are marked read-only and these names cannot be changed by subsequent assignment. Unlike typeset -r , readonly does not create a function-local scope and the given vnames are marked globally read-only by default. When defining a type, if the value of a read-only subvariable is not defined, the value is required when creating each instance.

redirect

This command only accepts input/output redirections. It can open and close files and modify file descriptors from 0 to 9 as specified by the input/output redirection list (see the Input/Output section above), with the difference that the effect persists past the execution of the redirect command. When invoking another program, file descriptors greater than 2 that were opened with this mechanism are only passed on if they are explicitly redirected to themselves as part of the invocation (e.g. 4>&4) or if the posix option is set.

† return [ n ]

Causes a shell function, dot script (see . and source), or profile script to return to the invoking shell environment with the exit status specified by n. This status value can use the full signed integer range as shown by the commands getconf INT_MIN and getconf INT_MAX. A value outside that range will produce a warning and an exit status of 128. If n is omitted, then the value of $? is assumed, i.e., the exit status of the last command executed is passed on. If return is invoked while not in a function, dot script, or profile script, then it behaves the same as exit.

† set [ ±BCGHabefhkmnprstuvx ] [ ±o [ option ] ] ... [ ±A vname ] [ arg ... ]

The options for this command have meaning as follows:

† shift [ n ]

The positional parameters from $n+1 ... are renamed $1 ... , default n is 1. The parameter n can be any arithmetic expression that evaluates to a non-negative number less than or equal to $#.

sleep [ -s ] duration

Suspends execution for the number of decimal seconds or fractions of a second given by duration. duration can be an integer, floating point value or ISO 8601 duration specifying the length of time to sleep. The option -s causes the sleep builtin to terminate when it receives any signal. If duration is not specified in conjunction with -s, sleep will wait for a signal indefinitely.

source name [ arg ... ]

Same as ., except it is not treated as a special built-in command.

stop job ...

Sends a SIGSTOP signal to one or more processes specified by job, suspending them until they receive SIGCONT. The same as kill -s STOP.

suspend

Sends a SIGSTOP signal to the main shell process, suspending the script or child shell session until it receives SIGCONT (for instance, when typing fg in the parent shell). Equivalent to kill -s STOP "$$", except that it accepts no operands and refuses to suspend a login shell.

test expression

The test and [ commands execute conditional expressions similar to those specified for the [[ compound command under Conditional Expressions above, but with several important differences. The =, == and != operators test for string (in)equality without pattern matching; == is nonstandard and unportable. The && and || operators are not available. Instead, the -a and -o binary operators can be used, but they are fraught with pitfalls due to grammatical ambiguities and therefore deprecated in favor of invoking separate test commands. Most importantly, as test and [ are simple regular commands, field splitting and pathname expansion are performed on all their arguments and all aspects of regular shell grammar (such as redirection) remain active. This is usually harmful, so care must be taken to quote arguments and expansions to avoid this. To avoid the many pitfalls arising from these issues, the [[ compound command should be used instead. The primary purpose of the test and [ commands is compatibility with other shells that lack [[.

times

Displays the accumulated user and system CPU times, one line with the times used by the shell and another with those used by all of the shell's child processes. No options are supported. Seconds are zero-padded unless the posix shell option is on.

† trap [ -p ] [ action ] [ sig ] ...

The -p option causes the trap action associated with each trap as specified by the arguments to be printed with appropriate quoting. Otherwise, action will be processed as if it were an argument to eval when the shell receives signal(s) sig. Each sig can be given as a number or as the name of the signal. Trap commands are executed in order of signal number. Any attempt to set a trap on a signal that was ignored on entry to the current shell is ineffective. If action is omitted and the first sig is a number, or if action is -, then the trap(s) for each sig are reset to their original values. If action is the empty string, then this signal is ignored by the shell and by the commands it invokes. If sig is ERR then action will be executed whenever a command has a non-zero exit status. If sig is DEBUG then action will be executed before each command. The variable .sh.command will contain the current command line when action is running, in the same format as the output generated by the xtrace option (minus the preceding PS4 prompt). If the exit status of the trap is 2 the command will not be executed. If the exit status of the trap is 255 and inside a function or a dot script, the function or dot script will return. If sig is 0 or EXIT and the trap statement is executed inside the body of a function defined with the function name syntax, then the command action is executed after the function completes. If sig is 0 or EXIT for a trap set outside any function then the command action is executed on exit from the shell. If sig is KEYBD, then action will be executed whenever a key is read while in emacs, gmacs, or vi mode. The trap command with no arguments prints a list of commands associated with each signal number.

An exit or return without an argument in a trap action will preserve the exit status of the command that invoked the trap.

true

Does nothing, and exits 0. Used with while for infinite loops.

type [ -afpPqt ] name ...

The same as whence -v.

†‡ typeset [ ±ACHSbflmnprstux ] [ ±EFLRXZi[n] ] [ +-M [ mapname ] ] [ -T [ tname=(assign_list) ] ] [ -h str ] [ -a [type] ] [ vname[=value ] ] ...

Sets attributes and values for shell variables and functions. When invoked inside a function defined with the function name syntax, a new instance of the variable vname is created, and the variable's value and type are restored when the function completes. The following list of attributes may be specified:

ulimit [ -HSaMctdfkxlqenVuPpmrRbiswTv ] [ limit ]

Set or display a resource limit. The available resource limits are listed below. Many systems do not support one or more of these limits. The limit for a specified resource is set when limit is specified. The value of limit can be a number in the unit specified below with each resource, or the value unlimited. The -H and -S options specify whether the hard limit or the soft limit for the given resource is set. A hard limit cannot be increased once it is set. A soft limit can be increased up to the value of the hard limit. If neither the H nor S option is specified, the limit applies to both. The current resource limit is printed when limit is omitted. In this case, the soft limit is printed unless H is specified. When more than one resource is specified, then the limit name and unit is printed before the value.

umask [ -S ] [ mask ]

The user file-creation mask is set to mask (see umask(2)). mask can either be an octal number or a symbolic value as described in chmod(1). If a symbolic value is given, the new umask value is the complement of the result of applying mask to the complement of the previous umask value. If mask is omitted, the current value of the mask is printed. The -S option causes the mode to be printed as a symbolic value. Otherwise, the mask is printed in octal.

unalias [ -a ] name ...

The aliases given by the list of names are removed from the alias list. The -a option causes all the aliases to be unset.

† unset [ -fnv ] vname ...

The variables given by the list of vnames are unassigned, i.e., except for subvariables within a type, their values and attributes are erased. For subvariables of a type, the values are reset to the default value from the type definition. Readonly variables cannot be unset. If the -f option is set, then the names refer to function names. If the -v option is set, then the names refer to variable names. The -f option overrides -v. If -n is set and name is a name reference, then name will be unset rather than the variable that it references. The default is equivalent to -v. Unsetting LINENO, MAILCHECK, OPTARG, OPTIND, RANDOM, SECONDS, TMOUT, and _ removes their special meaning even if they are subsequently assigned to.

wait [ job ... ]

Wait for the specified job and report its termination status. If job is not given, then all currently active child processes are waited for. The exit status from this command is that of the last process waited for if job is specified; otherwise it is zero. See Jobs for a description of the format of job.

whence [ -afpPqtv ] name ...

For each name, indicate how it would be interpreted if used as a command name.

The -v option produces a more verbose report. The -f option skips the search for functions. The -p and -P options do a path search for name even if name is an alias, a function, or a reserved word. Both of these options turn off the -v option. The -q option causes whence to enter quiet mode. whence will return zero if all arguments are built-ins, functions, or are programs found on the path. The -t option only outputs the type of the given command. Like -p and -P, -t will turn off the -v option. The -a option is similar to the -v option but causes all interpretations of the given name to be reported.

Invocation.¶

If the shell is invoked by exec(2), initialization depends on argument zero ($0) as follows. If the first character of $0 is -, or the -l option is given on the invocation command line, then the shell is assumed to be a login shell. If the basename of the command path in $0 is rsh, rksh, or krsh, then the shell becomes restricted. If the basename is sh or rsh, or the -o posix option is given on the invocation command line, then the shell is initialized in full POSIX compliance mode (see the set builtin command above for more information). After this, if the shell was assumed to be a login shell, commands are read from /etc/profile and then from $HOME/.profile if it exists. Alternatively, the option -l causes the shell to be treated as a login shell. Next, for interactive shells, commands are read from the file named by ENV if the file exists, its name being determined by performing parameter expansion, command substitution, and arithmetic expansion on the value of that environment variable. If the -s option is not present and arg and a file by the name of arg exists, then it reads and executes this script. Otherwise, if the first arg does not contain a /, a path search is performed on the first arg to determine the name of the script to execute. The script arg must have execute permission and any setuid and setgid settings will be ignored. If the script is not found on the path, arg is processed as if it named a built-in command or function. Commands are then read as described below; the following options are interpreted by the shell when it is invoked:

-D

A list of all double quoted strings that are preceded by a $ will be printed on standard output and the shell will exit. This set of strings will be subject to language translation when the locale is not C or POSIX. No commands will be executed.

-E or -o rc or --rc

Read the file named by the ENV variable or by $HOME/.kshrc if not defined after the profiles. On by default for interactive shells. Use +E, +o rc or --norc to turn off.

-c

Read and execute a script from the first arg instead of a file. The second arg, if present, becomes that script's command name ($0). Any third and further args become positional parameters starting at $1.

-s

Read and execute a script from standard input instead of a file. The command name ($0) cannot be set. Any args become the positional parameters starting at $1. This option is forced on if no arg is given and is ignored if -c is also specified.

-i or -o interactive or --interactive

If the -i option is present or if the shell's standard input and standard error are attached to a terminal (as told by tcgetattr(3)), then this shell is interactive. In this case TERM is ignored (so that kill 0 does not kill an interactive shell) and INTR is caught and ignored (so that wait is interruptible). In all cases, QUIT is ignored by the shell.

-r or -o restricted or --restricted

If the -r option is present, the shell is a restricted shell.

The remaining options and arguments are described under the set command above. An optional - as the first argument is ignored.

Rksh Only.¶

Rksh is used to set up login names and execution environments whose capabilities are more controlled than those of the standard shell. The actions of rksh are identical to those of ksh, except that the following are disallowed:

The restrictions above are enforced after .profile and the ENV files are interpreted.

When a command to be executed is found to be a shell procedure, rksh invokes ksh to execute it. Thus, it is possible to provide to the end-user shell procedures that have access to the full power of the standard shell, while imposing a limited menu of commands; this scheme assumes that the end-user does not have write and execute permissions in the same directory.

The net effect of these rules is that the writer of the .profile has complete control over user actions, by performing guaranteed setup actions and leaving the user in an appropriate directory (probably not the login directory).

The system administrator often sets up a directory of commands (e.g., /usr/rbin) that can be safely invoked by rksh.