| YASH(1) | YASH(1) |
yash - a POSIX-compliant command line shell
yash [options...] [--] [operands...]
Yet another shell (yash) is a command line shell for UNIX-like operating systems. The shell conforms to the POSIX.1-2008 standard (for the most parts), and actually is more conforming than other POSIX-conforming shells. Moreover, it has many features that are used for interactive use, such as command history and command line editing.
This program can be freely modified and redistributed under the terms of GNU General Public License (Version 2). Use of this program is all at your own risk. There is no warranty and the author is not responsible for any consequences caused by use of this program.
This manual can be freely modified and redistributed under the terms of Creative Commons Attribution-ShareAlike 2.1 Japan.
Yash is developed and maintained by 渡邊裕貴 (WATANABE Yuki) aka Magicant. Yash development project and Yash’s homepage are hosted by OSDN.
When invoked as a program, yash performs the predefined initialization steps and repeatedly reads and executed commands. Command line arguments given in the invocation determines how the shell initializes itself and executes commands.
The syntax of command line arguments for yash conforms to POSIX. As defined in POSIX, arguments are separated into options and operands. For more detailed explanation about options and operands, see Command argument syntax. All options must come before operands. The interpretation of operands depends on options specified.
When you specify the -c (--cmdline) option, you must give at least one operand. The shell interprets and executes the first operand as a command string. The second operand, if any, is used to initialize the 0 special parameter. The other operands, if any, are used to initialize the positional parameters. When the -c (--cmdline) option is specified, the shell does not read any file or the standard input (unless the dot built-in is used).
If you specify the -s (--stdin) option, the shell reads the standard input, interprets the input as commands, and executes them. All the operands given are used to initialize the positional parameters. The 0 special parameter is initialized to the name the shell is invoked as.
If you specify neither the -c (--cmdline) nor -s (--stdin) option, the shell reads a file, interprets the file contents as commands, and executes them. The first operand specifies the pathname of the file. The remaining operands are used to initialize the positional parameters. If you do not give any operands, the shell reads the standard input as if the -s (--stdin) option is specified.
You cannot use both the -c (--cmdline) and -s (--stdin) options at a time.
If you specify either the --help or -V (--version) option, the shell never performs the usual initialization or command execution. Instead, it just prints brief usage (for --help) or version information (for -V and --version). If the -V (--version) option is accompanied by the -v (--verbose) option, the shell prints a list of the available optional features as well.
If you specify the -i (--interactive) option, the shell goes into the interactive mode. If you specify the +i (++interactive) option, conversely, the shell never goes into the interactive mode.
If you specify the -l (--login) option, the shell behaves as a login shell.
The --noprofile, --norcfile, --profile, and --rcfile options determine how the shell is initialized (see below for details).
In addition to the options described above, you can specify options that can be specified to the set built-in.
If the first operand is - and the options and the operands are not separated by --, the first operand is ignored.
Yash initializes itself as follows:
If the --profile=filename option is not specified, the shell reads ~/.yash_profile as a default.
If the --rcfile=filename option is not specified, the shell instead reads the following files:
Yash never automatically reads /etc/profile, /etc/yashrc, nor ~/.profile.
The shell reads, parses, and executes command line by line. If there is more than one command on a line, all the commands are parsed before executed. If a command is continued to next lines, the shell reads more enough lines to complete the command. On a syntax error, the shell neither reads nor executes any more commands.
A command is composed of one or more tokens. In the shell syntax, a token is a word that is part of a command. Normally, tokens are separated by whitespaces, that is, the space or tab character. Whitespaces inside a command substitution or a parameter expansion, however, do not separate tokens.
The following symbols have special meanings in the shell syntax and in most cases separate tokens:
; & | < > ( ) [newline]
The following symbols do not separate tokens, but have syntactic meanings:
$ ` \ " ' * ? [ # ~ = %
The following tokens are treated as keywords depending on the context in which they appear:
! { } [[ case do done elif else esac fi
for function if in then until while
A token is treated as a keyword when:
If a token begins with #, then the # and any following characters up to the end of the line are treated as a comment, which is completely ignored in syntax parsing.
If you want whitespaces, separator characters, or keywords described above to be treated as a normal characters, you must quote the characters using appropriate quotation marks. Quotation marks are not treated as normal characters unless they are themselves quoted. You can use the following three quotation marks:
The only exception about a backslash is the case where a backslash is followed by a newline. In this case, the two characters are treated as a line continuation rather than a newline being quoted. The two characters are removed from the input and the two lines surrounding the line continuation are concatenated into a single line.
Tokens that compose a command are subject to alias substitution. A token that matches the name of an alias that has already been defined is substituted with the value of the alias before the command is parsed.
Tokens that contain quotations are not alias-substituted since an alias name cannot contain quotation marks. Keywords and command separator characters are not alias-substituted either.
There are two kinds of aliases: normal aliases and global aliases. A normal alias can only substitute the first token of a command while a global alias can substitute any part of a command. Global aliases are yash extension that is not defined in POSIX.
If a token is alias-substituted with the value of a normal alias that ends with a whitespace, the next token is exceptionally subject to alias substitution for normal aliases.
The results of alias substitution are again subject to alias substitution for other aliases (but not for the aliases that have been already applied).
You can define aliases using the alias built-in and remove using the unalias built-in.
A command that does not start with a keyword token is a simple command. Simple commands are executed as defined in Execution of simple commands.
If the first and any number of following tokens of a simple command have the form name=value, they are interpreted as variable assignments. A variable name must consist of one or more alphabets, digits and/or underlines (_) and must not start with a digit. The first token that is not a variable assignment is considered as a command name and all the following tokens (whether or not they have the form name=value) as command arguments.
A variable assignment of the form var=(tokens) is interpreted as assignment to an array. You can write any number of tokens between a pair of parentheses. Tokens can be separated by not only spaces and tabs but also newlines.
A pipeline is a sequence of one or more simple commands, compound commands, and/or function definitions that are separated by |.
A pipeline that has more than one subcommand is executed by executing each subcommand of the pipeline in a subshell simultaneously. The standard output of each subcommand except the last one is redirected to the standard input of the next subcommand. The standard input of the first subcommand and the standard output of the last subcommand are not redirected.
The exit status of the pipeline is that of the last subcommand unless the pipe-fail option is enabled, in which case the exit status of the pipeline is that of the last subcommand that exits with a non-zero exit status. If all the subcommands exit with an exit status of zero, the exit status of the pipeline is also zero.
A pipeline can be prefixed by !, in which case the exit status of the pipeline is reversed: the exit status of the pipeline is 1 if that of the last subcommand is 0, and 0 otherwise.
Korn shell treats a word of the form !(...) as an extended pathname expansion pattern that is not defined in POSIX. In the POSIXly-correct mode, the tokens ! and ( must be separated by one or more white spaces.
When the execution of a pipeline finishes, at least the execution of the last subcommand has finished since the exit status of the last subcommand defines that of the whole pipeline. The execution of other subcommands, however, may not have finished then. On the other hand, the execution of the pipeline may not finish soon after that of the last subcommand finished because the shell may choose to wait for the execution of other subcommands to finish.
The POSIX standard allows executing any of subcommands in the current shell rather than subshells, though yash does not do so.
An and/or list is a sequence of one or more pipelines separated by && or ||.
An and/or list is executed by executing some of the pipelines conditionally. The first pipeline is always executed. The other pipelines are either executed or not executed according to the exit status of the previous pipelines.
The exit status of an and/or list is that of the last pipeline that was executed.
Normally, an and/or list must be terminated by a semicolon, ampersand, or newline. See Command separators and asynchronous commands.
The whole input to the shell must be composed of any number of and/or lists separated by a semicolon or ampersand. A terminating semicolon can be omitted if it is followed by ;;, ), or a newline. Otherwise, an and/or list must be terminated by a semicolon or ampersand.
If an and/or list is terminated by a semicolon, it is executed synchronously: the shell waits for the and/or list to finish before executing the next and/or list. If an and/or list is terminated by an ampersand, it is executed asynchronously: after the execution of the and/or list is started, the next and/or list is executed immediately. An asynchronous and/or list is always executed in a subshell and its exit status is zero.
If the shell is not doing job control, the standard input of an asynchronous and/or list is automatically redirected to /dev/null. Signal handlers of the and/or list for the SIGINT and SIGQUIT signals are set to “ignore” the signal so that the execution of the and/or list cannot be stopped by those signals.
When the execution of an asynchronous and/or list is started, the shell remembers its process ID. You can obtain the ID by referencing the ! special parameter. You can obtain the current and exit status of the asynchronous list as well by using the jobs and wait built-ins.
Compound commands provide you with programmatic control of shell command execution.
Grouping
A grouping is a list of commands that is treated as a simple command.
Normal grouping syntax
Subshell grouping syntax
The { and } tokens are keywords, which must be separated from other tokens. The ( and ) tokens, however, are special separators that need not to be separated.
In the normal grouping syntax, the commands in a grouping are executed in the current shell. In the subshell grouping syntax, the commands are executed in a new subshell.
In the POSIXly-correct mode, a grouping must contain at least one command. If the shell is not in the POSIXly-correct mode, a grouping may contain no commands.
The exit status of a grouping is that of the last command in the grouping. If the grouping contains no commands, its exit status is that of the last executed command before the grouping.
If command
The if command performs a conditional branch.
Basic if command syntax
Syntax with the else clause
Syntax with the elif clause
Syntax with the elif clause
For all the syntaxes, the execution of an if command starts with the execution of the condition commands that follows the if token. If the exit status of the condition commands is zero, the condition is considered as “true”. In this case, the body commands that follows the then token are executed and the execution of the if command finishes. If the exit status of the condition commands is non-zero, the condition is considered as “false”. In this case, the condition commands for the next elif clause are executed and the exit status is tested in the same manner as above. If there is no elif clause, the body commands that follow the else token are executed and the execution of the if command finishes. If there is no else clause either, the execution of the if command just ends.
An if command may have more than one elif-then clause.
The exit status of an if command is that of the body commands that were executed. The exit status is zero if no body commands were executed, that is, all the conditions were false and there was no else clause.
While and until loops
The while loop and until loop are simple loops with condition.
While loop syntax
Until loop syntax
If the shell is not in the POSIXly-correct mode, you can omit the condition and/or body commands of a while/until loop.
The execution of a while loop is started by executing the condition commands. If the exit status of the condition commands is zero, the shell executes the body commands and returns to the execution of the condition commands. The condition and body commands are repeatedly executed until the exit status of the condition commands is non-zero.
The body commands are not executed at all if the first execution of the condition commands yields a non-zero exit status.
An until loop is executed in the same manner as a while loop except that the condition to repeat the loop is reversed: the body commands are executed when the exit status of the condition commands is non-zero.
The exit status of a while/until loop is that of the last executed body command. The exit status is zero if the body commands are empty or were not executed at all.
For loop
The for loop repeats commands with a variable assigned one of given values in each round.
For loop syntax
for varname do command...; done
The word list after the in token may be empty, but the semicolon (or newline) before the do token is required even in that case. The words are not treated as keywords, but you need to quote separator characters (such as & and |) to include them as part of a word. The command list may be empty if not in the POSIXly-correct mode.
The varname must be a portable (ASCII-only) name in the POSIXly-correct mode.
The execution of a for loop is started by expanding the words in the same manner as in the execution of a simple command. If the in and word tokens are omitted, the shell assumes the word tokens to be "$@". Next, the following steps are taken for each word expanded (in the order the words were expanded):
By default, if a for loop is executed within a function, varname is created as a local variable, even if it already exists globally. Turning off the for-local shell option or enabling the POSIXly-correct mode mode will disable this behavior.
If the expansion of the words yields no words, no variable is created and the commands are not executed at all.
The exit status of a for loop is that of the last executed command. The exit status is zero if the commands are not empty and not executed at all. If the commands are empty, the exit status is that of the last executed command before the for loop.
If the variable is read-only, the execution of the for loop is interrupted and the exit status will be non-zero.
Case command
The case command performs a pattern matching to select commands to execute.
Case command syntax
Case item syntax
The word between the case and in tokens must be exactly one word. The word is not treated as a keyword, but you need to quote separator characters (such as & and |) to include them as part of the word. Between the in and esac tokens you can put any number of case items (may be none). You can omit the first ( token of a case item and the last ;; token before the esac token. If the last command of a case item is terminated by a semicolon, you can omit the semicolon as well. The commands in a case item may be empty.
The patterns in a case item are one or more tokens each separated by a | token.
The execution of a case command starts with subjecting the word to the four expansions. Next, the following steps are taken for each case item (in the order of appearance):
The exit status of a case command is that of the commands executed. The exit status is zero if no commands were executed, that is, there were no case items, no matching pattern was found, or no commands were associated with the matching pattern.
In the POSIXly-correct mode, the first pattern in a case item cannot be esac (even if you do not omit the ( token).
Double-bracket command
The double-bracket command is a syntactic construct that works similarly to the test built-in. It expands and evaluates the words between the brackets.
Double-bracket command syntax
The expression can be a single primary or combination of primaries and operators. The expression syntax is parsed when the command is parsed, not executed. Operators (either primary or non-primary) must not be quoted, or it will be parsed as a normal word.
When the command is executed, operand words are subjected to the four expansions, but not brace expansion, field splitting, or pathname expansion.
In the double-bracket command, the following primaries from the test built-in can be used:
Unary primaries
Binary primaries
Additionally, some binary primaries can be used to compare strings, which works slightly differently from those for the test built-in: The = primary treats the right-hand-side operand word as a pattern and tests if it matches the left-hand-side operand word. The == primary is the same as =. The != primary is negation of the = primary (reverse result).
The operand word of a primary must be quoted if it is ]] or can be confused with another primary operator.
More primaries may be added in future versions of the shell. You should quote any words that start with a hyphen.
The <= and >= binary primaries cannot be used in the double-bracket command because it cannot be parsed correctly in the shell grammar.
The following operands (listed in the descending order of precedence) can be used to combine primaries:
( expression )
! expression
expression && expression
expression || expression
Unlike the test built-in, neither -a nor -o can be used as a binary operator in the double-bracket command.
The exit status of the double-bracket command is 0 if expression is true, 1 if false, and 2 if it cannot be evaluated because of expansion error or any other reasons.
The double-bracket command is also supported in bash, ksh, mksh, and zsh, but not defined in the POSIX standard. The behavior slightly differs between the shells. The test built-in should be preferred over the double-bracket command for maximum portability.
The function definition command defines a function.
Function definition syntax
function funcname compound_command
function funcname ( ) compound_command
In the first syntax without the function keyword, funcname cannot contain any special characters such as semicolons and quotation marks. In the second and third syntax, which cannot be used in the POSIXly-correct mode, funcname is subjected to the four expansions when executed. In the POSIXly-correct mode, funcname is limited to a portable (ASCII-only) name.
When a function definition command is executed, a function whose name is funcname is defined with its body being compound_command.
A function definition command cannot be directly redirected. Any redirections that follow a function definition are associated with compound_command rather than the whole function definition command. In func() { cat; } >/dev/null, for example, it is not func() { cat; } but { cat; } that is redirected.
The exit status of a function definition is zero if the function was defined without errors, and non-zero otherwise.
Parameters are string values that are expanded in parameter expansion. There are three types of parameters: positional parameters, special parameters and variables.
Positional parameters are parameters that are identified by natural numbers. If there are three positional parameters, for example, they are identified as 1, 2, and 3. You can obtain the number of positional parameters by the # special parameter. The * and @ special parameters are expanded to all positional parameters.
Positional parameters are initialized from the shell’s command line arguments when the shell is started (see Command line arguments). In the initialization, the order of the operands are preserved as the order of the positional parameters.
When the shell executes a function call, positional parameters are changed to the arguments to the function call so that you can access the arguments while the function is being executed. Positional parameters are restored to the original values when the execution of the function is finished.
Positional parameters can be manipulated by built-in commands like set and shift.
Note that 0 is not a positional parameter but a special parameter.
Special parameters are parameters each identified by a single symbol. They cannot be directly assigned to by the user.
Yash provides the following special parameters:
0
#
$
-
?
!
*
If field-splitting is applied to an expansion result of this parameter, the value is first split into the original positional parameters and then further split depending on the current IFS variable. The first splitting is performed even if the IFS variable is empty.
@
Variables are parameters the user can assign values to. Each variable has a name that identifies it and a value that defines the results of expansion.
A variable name is composed of one or more alphanumeric characters and underscores (_). A name cannot start with a digit. Other characters may be used in a name depending on internationalization support of your environment.
Variables that are exported to external commands are called environment variables. They are passed to all external commands the shell invokes. Variables passed to the shell in invocation will be automatically exported.
You can assign to variables by a simple command as well as the typeset built-in. You can remove variables by using the unset built-in.
Variables used by the shell
The following variables are used by the shell for special purposes.
CDPATH
COLUMNS
COMMAND_NOT_FOUND_HANDLER
This feature is disabled in the POSIXly-correct mode.
DIRSTACK
ECHO_STYLE
ENV
FCEDIT
HANDLED
HISTFILE
HISTRMDUP
If the value of this variable is 1, for example, the most recent item is removed when a new item that have the same contents is added.
Items older than the nth recent item are not removed. No items are removed if the value of this variable is 0. All items are subject to removal if the variable value is greater than or equal to the value of the HISTSIZE variable.
HISTSIZE
HOME
IFS
LANG, LC_ALL, LC_COLLATE, LC_CTYPE, LC_MESSAGES, LC_MONETARY, LC_NUMERIC, LC_TIME
Unless the shell is interactive and not in the POSIXly-correct mode, the value of the LC_CTYPE variable is considered only when the shell is started. Once the shell has been initialized, changing the value of LC_CTYPE will have no effect on the shell’s behavior.
LINENO
In the interactive shell, the line number is reset to 1 each time the shell reads and executes a command.
If you assign to or remove this variable, it will no longer provide line numbers.
LINES
MAILCHECK
MAILPATH
NLSPATH
OLDPWD
OPTARG
OPTIND
PATH
PPID
PROMPT_COMMAND
PS1
PS1R
PS1S
PS2
PS2R
PS2S
PS4
PS4S
PWD
RANDOM
You can set the “seed” of random numbers by assigning a non-negative integer to the variable.
If you remove this variable, it will no longer work as a random number generator. If the shell was invoked in the POSIXly-correct mode, this variable does not work as a random number generator.
TERM
YASH_AFTER_CD
YASH_LOADPATH
YASH_LE_TIMEOUT
YASH_PS1, YASH_PS1R, YASH_PS1S, YASH_PS2, YASH_PS2R, YASH_PS2S, YASH_PS4, YASH_PS4S
YASH_VERSION
Arrays
An array is a variable that contains zero or more strings. The string values of an array are identified by natural numbers (like positional parameters).
You can assign values to an array by using a simple command as well as the array built-in. You can use the unset built-in to remove arrays.
Arrays cannot be exported as arrays. When an array is exported, it is treated as a normal variable whose value is a concatenation of all the array values, each separated by a colon.
Arrays are not supported in the POSIXly-correct mode.
Word expansion is substitution of part of a word with another particular string. There are seven types of word expansions:
These types of expansions are performed in the order specified above.
Tilde expansion, parameter expansion, command substitution, and arithmetic expansion are called the four expansions.
In tilde expansion, parts of words that start with a tilde (~) are substituted with particular pathnames. The part of each word that gets substituted is from the beginning of the word, which is a tilde, up to (but not including) the first slash (/) in the word. If the word does not contain a slash, the whole word is substituted. If any character in the substituted part is quoted, tilde expansion is not performed on the word.
The results of expansion are determined by the format of the substituted part:
~
~username
~+
~-
~+n, ~-n
When tilde expansion is performed on the value of a variable assignment that occurs during execution of a simple command, the value is considered as a colon-separated list of words and those words are each subject to tilde expansion. For example, the variable assignment
VAR=~/a:~/b:~/c
is equivalent to
VAR=/home/foo/a:/home/foo/b:/home/foo/c
if the value of HOME variable is /home/foo.
The POSIX standard does not prescribe how the shell should behave when it encounters an error during tilde expansion (e.g., when the HOME variable is not defined). Yash silently ignores any errors during tilde expansion; the part of the word that would be substituted is left intact.
In the POSIXly-correct mode, tilde expansion supports the formats of ~ and ~username only.
Parameter expansion expands to the value of a parameter.
The syntax of typical, simple parameter expansion is ${parameter}, which expands to the value of the parameter whose name is parameter. You can omit the braces (e.g., $parameter) if
If parameter is none of a special parameter, positional parameter, and variable, it is a syntax error. (Some shells other than yash may treat such a case as an expansion error.)
If the unset option is disabled and the parameter is an undefined variable, it is an expansion error. If the unset option is enabled, an undefined variable expands to the empty string.
More complex syntax of parameter expansion allows modifying the value of a parameter.
Parameter expansion
The spaces in the syntax definition above are for readability only and must be omitted. You can omit prefix, index, and/or modifier.
Prefix
The prefix, if any, must be a hash sign (#). If a parameter expansion has the prefix, the result of expansion is the number of characters in the value this expansion would be expanded to without the prefix.
Parameter name
The parameter name (parameter) must be either
The parameter expansion is expanded to the value of the parameter. If parameter is an array variable, the values of the array are field-split like the @ special parameter unless the index [*] is specified.
If parameter is another expansion, it is called a nested expansion. Nested expansion cannot be used in the POSIXly-correct mode. The braces ({ }) of a nested parameter expansion cannot be omitted.
Index
[word1,word2]
where word1 and word2 are parsed in the same manner as normal tokens except that they are always delimited by , or ] and can contain whitespace characters.
If there is an index in a parameter expansion, it is interpreted as follows:
If parameter is an array variable, the index specifies the part of the array. If parameter is either the * or @ special parameter, the index specifies the index range of positional parameters. In other cases, the index specifies the index range of a substring of the parameter value that is being expanded. In all cases, the specified range of the array values, positional parameters, or parameter value remains in the results of the expansion and other values are dropped.
If the interpretation of index is one or two integers, the following rules apply:
If the interpretation of index is one of *, @, and #, it is treated as follows:
*
@
#
If parameter is an array, the result of this parameter expansion will be the number of values in the array being expanded. If parameter is the * or @ special parameter, the result will be the number of current positional parameters. Otherwise, the result will be the number of characters in the value that is being expanded.
If a parameter expansion does not contain an index, it is assumed to be [@]. In the POSIXly-correct mode, index cannot be specified.
Example 1. Expansion of a normal variable
The following commands will print the string ABC:
var='123ABC789'
echo "${var[4,6]}"
Example 2. Expansion of positional parameters
The following commands will print the string 2 3 4:
set 1 2 3 4 5
echo "${*[2,-2]}"
Example 3. Expansion of an array
The following commands will print the string 2 3 4:
array=(1 2 3 4 5)
echo "${array[2,-2]}"
Modifier
You can modify the value to be expanded by using modifiers:
-word
+word
=word
?word
:-word, :+word, :=word, :?word
#word
##word
%word
%%word
/word1/word2
This modifier cannot be used in the POSIXly-correct mode.
/#word1/word2
/%word1/word2
//word1/word2
:/word1/word2
In all types of modifiers above, words are subjected to the four expansions when (and only when) they are used.
If parameter is an array variable or the @ or * special parameter, modifiers affect each value of the array or all positional parameters.
Command substitution expands to output of commands specified.
Command substitution
`commands`
When command substitution is evaluated, commands are executed by a subshell with output pipelined to the shell. When the commands finished, command substitution is substituted with the output of the commands. Any trailing newline characters in the output are ignored.
When command substitution of the form $(commands) is parsed, the commands are parsed carefully so that complex commands such as nested command substitution are parsed correctly. If commands start with (, you should put a space before commands so that the whole command substitution is not confused with arithmetic expansion. If the shell is in the POSIXly-correctly mode, the commands are parsed each time the command substitution is expanded; otherwise, commands are parsed only when the command substitution is parsed.
If command substitution is of the form `commands`, the commands are not parsed when the command substitution is parsed; the commands are parsed each time the command substitution is expanded. The end of commands is detected by the first backquote character (`) after the beginning of commands that is not quoted by a backslash. Backquotes that are part of commands (typically used for nested command substitution) must be quoted by backslashes. In commands, backslashes are treated as quotes only when preceding a dollar ($), backquote, newline, or another backslash. Additionally, if the command substitution occurs inside double quotes, double quotes in commands must be quoted with a backslash. Those backslashes are removed before commands are parsed.
Arithmetic expansion evaluates an arithmetic expression and expands to the value of the expression.
Arithmetic expansion
When arithmetic expansion is expanded, the expression is subject to parameter expansion, command substitution, and (nested) arithmetic expansion. The expression is parsed in (almost) same manner as an expression of the C programming language.
Yash allows an expression to be either an integer (of the long type in C) or a floating-point number (of the double type in C). An operation on integers yields an integer and an operation involving a floating-point number yields a floating-point number. In the POSIXly-correct mode, you can use integers only.
The following operators are available (in the order of precedence):
The ++ and -- operators cannot be used in the POSIXly-correct mode.
An atomic expression can be one of an integer literal, a floating-point number literal, and a variable. Literals are parsed in the same manner as in C. An octal integer literal starts with 0, and hexadecimal with 0x. A floating-point number literal may have an exponent (i.e. 1.23e+6). A variable with a non-numeric value will result in an error when parsed as a number. An unset variable is treated as a value of zero if the unset option is enabled.
In the POSIXly-correct mode, variables are always parsed as numbers. Otherwise, variables are parsed only when they are used as numbers in computation. Unparsed variables are left intact.
set +o posixly-correct foo=bar echo $((0 ? foo : foo)) # prints "bar" echo $((foo + 0)) # error
Brace expansion expands to several split words with preceding and succeeding portions duplicated to each split words. Brace expansion is expanded only when the brace-expand option is enabled.
Comma-separated brace expansion
Range brace expansion
{start..end..delta}
Comma-separated brace expansion is expanded to each comma-separated word. For example, a{1,2,3}b is expanded to the three words a1b, a2b, and a3b.
Range brace expansion is expanded to integers in the range defined by start and end. The difference between each integer can be defined by delta. If start is larger than end, the results will be in descending order. When ..delta is omitted, it defaults to 1 or -1. For example, a{1..3}b is expanded to the three words a1b, a2b, and a3b; and a{1..7..2}b to the four words a1b, a3b, a5b, and a7b.
Multiple brace expansions can be used in one word. Brace expansions can also be nested. You can quote braces and/or commas to prevent them from being treated as brace expansion.
Any errors in brace expansion are silently ignored.
In field splitting, words are split at predefined separators.
Field splitting can occur only within parts of words that resulted from parameter expansion, command substitution, and arithmetic expansion that are not between double-quotation marks. Expansion results of the @ special parameter are exceptionally split even between double-quotation marks.
Separators used in field splitting are defined by the value of the IFS variable. If the variable does not exist, the value is assumed to be the three characters of space, tab, and newline.
Characters included in the value of the IFS variable are called IFS characters. IFS characters that are any of space, tab, and newline are called IFS whitespace and other IFS characters are called IFS non-whitespace.
Field splitting is performed as follows:
Finally, the last word is removed from the results if:
Words are not split at all when the value of the IFS variable is empty.
Pathname expansion performs pattern matching and expands to pathnames matched by the pattern.
A word subjected to pathname expansion is treated as a pattern. If one or more pathnames are found that are matched by the pattern, the pathnames become the results of the pathname expansion.
Pathname expansion is not performed when the glob option is disabled.
The shell searches readable directories for matching pathnames. Unreadable directories are silently ignored.
The following options affect the behavior of pathname expansion:
null-glob
case-glob
dot-glob
mark-dirs
extended-glob
Any errors in pathname expansion are silently ignored. If the word is an invalid pattern, it just becomes the result. The results depend on the null-glob option when no matching pathnames are found.
Pattern matching is done for each filename (or pathname component) of pathnames. The shell skips matching for literal patterns that contain no wildcards or bracket expressions. As a result, the patterns /*/foo and /*/fo[o] may yield different expansion results when the case-glob option is disabled; for example, the pattern /*/fo[o] matches the pathname /bar/FOO but the pattern /*/foo does not because matching is skipped for foo.
Extension in pathname expansion
The following patterns can be used when the extended-glob option is enabled.
**
This pattern is not effective when appearing at the end of the whole pattern (i.e. foo/bar/**).
.**
***
.***
Pattern matching notation is a syntax of patterns that represent particular sets of strings. When a string is included in the set of strings a pattern represents, the pattern is said to match the string. Whether a pattern matches a string or not is defined as follows.
A character that is not quoted or any of special characters defined below is a normal character, which matches the character itself.
For example, the pattern abc matches the string abc, and not any other strings.
The character ? matches any single character.
For example, the pattern a?c matches any three-character strings that starts with a and ends with c, such as aac, abc, and a;c.
The character * matches any strings (of any length, including the empty string).
For example, the pattern a*c matches any string that starts with a and ends with c, such as ac, abc, and a;xyz;c.
A pattern that is enclosed by brackets ([ and ]) is a bracket expression. A bracket expression must have at least one character between the brackets. The characters between the brackets are interpreted as a bracket expression pattern, which is a below-defined special notation for bracket expression. A bracket expression pattern represents a set of characters. The bracket expression matches any one of the characters in the set the bracket expression pattern represents.
If the opening bracket ([) is followed by an exclamation mark (!), the exclamation is not treated as part of the bracket expression pattern and the whole bracket expression instead matches a character that is not included in the set the bracket expression pattern represents. If the opening bracket is followed by a caret (^), it is treated like an exclamation mark as above (but shells other than yash may treat the caret differently).
If the opening bracket (or the following exclamation or caret, if any) is followed by a closing bracket (]), it is treated as part of the bracket expression pattern rather than the end of the bracket expression. You cannot quote characters in the bracket expression pattern because quotation is treated before bracket expression.
An opening bracket in a pattern is treated as a normal character if it is not the beginning of a valid bracket expression.
A character that is not any of special characters defined below is a normal character, which represents the character itself.
For example, the bracket expression pattern abc represents the set of the three characters a, b, and c. The bracket expression [abc] therefore matches any of the three characters.
A hyphen preceded and followed by a character (or collating symbol) is a range expression, which represents the set of the two characters and all characters between the two in the collation order. A collation order is an order of characters that is defined in the locale data.
If a hyphen is followed by a closing bracket (]), the bracket is treated as the end of the bracket expression and the hyphen as a normal character.
For example, the range expression 3-5 represents the set of the three characters 3, 4, and 5. The bracket expression [3-5-] therefore matches one of the four characters 3, 4, 5, and -.
A collating symbol allows more than one character to be treated as a single character in matching. A collating symbol is made up of one or more characters enclosed by the special brackets [. and .].
One or more characters that are treated as a single character in matching are called a collating element. Precisely, a bracket expression pattern represents a set of collating elements and a bracket expression matches a collating element rather than a character, but we do not differentiate them for brevity here.
For example, the character combination “ch” was treated as a single character in the traditional Spanish language. If this character combination is registered as a collating element in the locale data, the bracket expression [[.ch.]df] matches one of ch, d, and f.
An equivalence class represents a set of characters that are considered equivalent. A equivalence class is made up of a character (or more precisely, a collating element) enclosed by the special brackets [= and =].
An equivalence class represents the set of characters that consists of the character enclosed by the brackets and the characters that are in the same primary equivalence class as the enclosed character. The shell consults the locale data for the definition of equivalence classes in the current locale.
For example, if the six characters a, à, á, â, ã, ä are defined to be in the same primary equivalence class, the bracket expressions [[=a=]], [[=à=]], and [[=á=]] match one of the six.
A character class represents a predefined set of characters. A character class is made up of a class name enclosed by the special brackets [: and :]. The shell consults the locale data for which class a character belongs to.
The following character classes can be used in all locales:
[:lower:]
[:upper:]
[:alpha:]
[:digit:]
[:xdigit:]
[:alnum:]
[:blank:]
[:space:]
[:punct:]
[:print:]
[:cntrl:]
For example, the bracket expression [[:lower:][:upper:]] matches a lower or upper case character. In addition to the classes listed above, other classes may be used depending on the definition of the current locale.
Redirection is a feature you can use to modify file descriptors of commands. By using redirection, you can execute commands with their standard input/output connected with files or devices other than the terminal.
You can do redirection by adding redirection operators to a command (simple command or compound command) In a simple command, redirection operators may appear anywhere in the command as long as operator tokens are separated from other tokens. In a compound command, redirection operators must appear at the end of the command.
Redirection operators are processed before the command body is executed. More than one redirection operator in a command are processed in the order of appearance. Redirection operators affect only the command in which they appear, except when they appear in an exec built-in without command operands. That is, file descriptors modified by redirection are restored after the command has finished.
A redirection operator starts with < or >. Redirection operators starting with < affects the standard input (file descriptor 0) by default. Redirection operators starting with > affects the standard output (file descriptor 1) by default. To affect another file descriptor, you can prefix a redirection operator with a non-negative integer; the operator will affect the file descriptor specified by the integer. The integer must immediately precede the < or > without any whitespaces in between. The integer must not be quoted, either.
The most common type of redirection is redirection to files.
Redirection of input
Redirection of output
>| token
>> token
Redirection of input and output
The token is subject to the four expansions. It is also subject to pathname expansion if the shell is interactive. The expansion result is treated as the pathname of the file to which redirection is performed. If the pathname expansion does not result in a single pathname, it is an error.
In redirection of input, the standard input is replaced with a file descriptor which is open for read-only access to the target file. If the target file cannot be opened for read-only access, it is an error.
In redirection of output, the standard output is replaced with a file descriptor which is open for write-only access to the target file. If the target file cannot be opened for write-only access, it is an error. If the target file does not exist, a new empty file is created and opened. If the target file already exists, the file is opened as follows:
In redirection of input and output, the standard input is replaced with a file descriptor which is open for read-and-write access to the target file. If the file does not exist, a new empty file is created and opened.
Socket redirection
If the pathname of the target file is of the form /dev/tcp/host/port or /dev/udp/host/port and the file cannot be opened in the usual manner, a new socket is opened for communication with the port of the host. The redirection replaces the standard input or output with the file descriptor to the socket.
A stream socket is opened for the form /dev/tcp/host/port and a datagram socket for the form /dev/udp/host/port. The protocol actually used for communication is determined by the socket library the shell uses. Typically, stream sockets use TCP and datagram sockets UDP.
In socket redirection, the file descriptor is both readable and writable regardless of the type of the redirection operator used.
Socket redirection is yash’s extension that is not defined in POSIX. Bash as well has socket redirection as extension.
Redirection allows duplicating or closing existing file descriptors.
Duplication of file descriptor
>& token
The token is subject to expansion as in redirection to files, but it is treated as a file descriptor rather than a pathname. Thus the expanded token must be a non-negative integer.
The <& and >& operators duplicate the file descriptor specified by token to the standard input and output, respectively. (The operators can be prefixed with a non-negative integer so that the file descriptor is duplicated to a file descriptor other than the standard input or output.)
If the expanded token is a single hyphen rather than a non-negative integer, the file descriptor is closed rather than duplicated. By default, the <& and >& operators close the standard input and output, respectively, but the operators can be prefixed with a non-negative integer so that another file descriptor is closed.
In the POSIXly-correct mode, a file descriptor must be readable when duplicated by the <& operator and writable when duplicated by the >& operator.
Here document and here string allow redirection to file descriptors that reads strings directly specified in shell commands.
Here document
<<- token
Here string
In a here document or here string, the standard input is replaced with a readable file descriptor. When the command reads from the file descriptor, it will read the contents of the here document/string, which is defined below.
When a here document operator (<< or <<-) appears in a command, the shell reads the contents of the here document starting from the next line. The contents of here documents are not parsed nor executed as commands. The token after the operand specifies a delimiter that indicates the end of the contents. (The token is not subject to any expansion, but quotation is processed.) The contents of the here document is terminated just before the first line containing the token only. When using the <<- operator, all tab characters at the beginning of each line in the here document contents are removed and the delimiter token may be preceded by tab characters.
If there are more than one here document operator on one line, the contents of the here documents are parsed in order: The contents of the first here document starts from the next line and ends before the first line containing the token that followed the first operator. Just after that line, the contents of the second here document starts, and so on.
The contents of here documents are treated literally: whitespaces, tabs, etc. remain as is. The exception is that, when the token is not quoted at all:
In here string, the token after the operator is subject to expansion as in redirection to files. The expansion result becomes the contents of the here string. A newline character is automatically appended to the end of here string contents.
Here string is yash’s extension that is not defined in POSIX. Other shells like bash, ksh, and zsh have the same feature.
Pipeline redirection allows opening pipelines that can be used for arbitrary purposes.
Pipeline redirection
The token is subject to expansion as in redirection to files, but it is treated as a file descriptor rather than a pathname. Thus the expanded token must be a non-negative integer.
Pipeline redirection opens a new pipeline. The standard output (or the file descriptor specified before the operator, if any) is replaced with the file descriptor open for writing to the pipeline. The file descriptor specified by token is replaced with the file descriptor open for reading from the pipeline.
Pipeline redirection is yash’s extension that is not defined in POSIX.
Process redirection creates a pipeline connected to another command.
Process redirection
>(command...)
In process redirection, the command specified is executed in a subshell. If the process redirection is of the form <(command...), the standard output of command is connected with a pipeline to the standard input of the command the redirection is associated with. If the process redirection is of the form >(command...), the standard input of command is connected with a pipeline to the standard output of the command the redirection is associated with.
Process redirection is yash’s extension that is not defined in POSIX. Bash and zsh have a feature called process substitution, which uses the same syntax as yash’s process redirection, but incompatibly differs in behavior.
This section describes how commands are executed.
A simple command is executed as follows:
In the following steps, the first word of the expansion results is referred to as command name, and the other words as command arguments. If there is only one word of the expansion results, there are no command argument words. If there are none of the expansion results, there is no command name either.
The assigned variables are automatically exported when the command name is specified or the all-export option is enabled.
If the command was executed, the exit status of this simple command is that of the executed command. If the algorithm failed to determine a command, no command is executed and the exit status is 127. If the shell failed to execute the determined command, the exit status is 126. If the executed command was killed by a signal, the exit status is the signal number plus 384.
Command search
A command that is executed in a simple command is determined by the command name using the following algorithm:
The value of the PATH variable is separated by colons. Each separated part is considered as a directory pathname (an empty pathname denotes the current working directory). The shell searches the directories (in the order of appearance) and checks if any directory directly contains an executable regular file whose name is equal to the command name. If such a file is found:
If no such file is found, no command is determined as the executed command.
When the shell finds a file that matches the command name during the search above, the shell remembers the pathname of the file if it is an absolute path. When the algorithm above is used for the same command name again, the shell skips searching and directly determines the command to be executed. If an executable regular file no longer exists at the remembered pathname, however, the shell searches again to update the remembered pathname. You can manage remembered pathnames using the hash built-in.
The shell exits when it reached the end of input and has parsed and executed all input commands or when the exit built-in is executed. The exit status of the shell is that of the last command the shell executed (or zero if no commands were executed). The exit status of the shell is always between 0 and 255 (inclusive). If the exit status of the last command is 256 or larger, the exit status of the shell will be the remainder of the exit status divided by 256.
If an exit handler has been registered by the trap built-in, the handler is executed just before the shell exits. The exit status of the commands executed in the handler does not affect the exit status of the shell.
If a non-interactive shell encountered one of the following errors, the shell immediately exits with a non-zero exit status:
Some shells other than yash exit when they fail to find a command to execute in command search.
Functions allow executing a compound command as a simple command. A function can be defined by the function definition command and executed by a simple command. You can use the unset built-in to remove function definitions.
There are no functions predefined when yash is started.
A function is executed by executing its body, which is a compound command. While the function is being executed, positional parameters are set to the arguments given to the function. The old positional parameters are restored when the function execution finishes.
Local variables
Local variables are temporary variables that are defined in a function and exist during the function execution only. They can be defined by the typeset built-in or implicitly created by a for loop. They are removed when the function execution finishes.
Local variables may hide variables that have already been defined before the function execution had started. An existing variable becomes inaccessible if a local variable of the same name is defined in a function. The old variable becomes accessible again when the function execution finishes.
You cannot create a local variable when not executing a function. A normal variable is created if you try to do so.
The shell holds following properties during execution.
Those properties are inherited from the invoker of the shell to the shell, and from the shell to each external command executed by the shell.
The properties can be changed during the execution of the shell by built-in commands, variable assignments, etc.
Subshells
A subshell is a copy of the shell process. Subshells are used in execution of groupings, pipelines, etc.
Subshells inherit functions, aliases, etc. defined in the shell as well as the properties above since subshells are copies of the shell process. Notable exceptions are:
Subshells are executed independently of the original shell, so changes of any properties above do not affect those of the original shell.
If the subshell contains a single trap built-in, some shells (but not yash) may not reset the traps on entry to the subshell.
The interactive mode is a mode of the shell intended for direct interaction with a user. If yash is in the interactive mode, it is called an interactive shell.
Whether a shell runs in the interactive mode or not is determined in the invocation of the shell. After the shell has started up, the interactive mode cannot be switched on or off.
When the shell is interactive:
The interactive shell prints a prompt just before it reads a command. The contents of the prompt is specified by the value of the PS1 and PS2 variables. The former is used for reading the first line of the command and the latter for other lines.
When the prompt is printed, the variable value is subjected to parameter expansion, command substitution, and arithmetic expansion (but note that the POSIX standard requires parameter expansion only). The result of the expansion is parsed by the rules below to make the actual prompt string, which is printed to the standard error.
In the POSIXly-correct mode, each exclamation mark (!) in the string is substituted with the command history number of the command that is being input. Two adjacent exclamation marks (!!) are printed as a single exclamation. Other characters are printed intact.
If the shell is not in the POSIXly-command mode, the following notations can be used to format the prompt string. Notations are replaced with the strings designated in the list below. Characters that are not interpreted as notations are printed intact.
\a
\e
\j
\n
\r
\!
\$
\\
\[, \]
\ffontspecs.
k
r
g
y
b
m
c
w
K
R
G
Y
B
M
C
W
t
d
s
u
v
b
i
o
x
D
The actual colors of font and background are defined by the terminal. Different terminals may use different colors.
In addition to the normal prompt, a prompt string can be displayed to the right of the cursor if line-editing is active. Those prompts are called right prompts. The contents of right prompts are defined by the value of the PS1R and PS2R variables, each corresponding to the PS1 and PS2 variables.
Using the above-said notations, the font style of command strings the user inputs can be changed as well as that of prompts. The font style of command strings is defined by the value of the PS1S and PS2S variables, each corresponding to the PS1 and PS2 variables. The value can contain the \ffontspecs. notation only.
When the shell is not in the POSIXly-correct mode, the prompt variables can be defined with a name prefixed with YASH_ (e.g. YASH_PS1). This allows using a different prompt string than that in the POSIXly-correct mode.
When the shell is not in the POSIXly-correct mode, the value of the PROMPT_COMMAND variable is executed before each prompt.
Command history is a feature of the shell that remembers executed commands to allow re-executing them later. Commands executed in the interactive mode are automatically saved in the command history. Saved commands can be edited and re-executed using line-editing and the fc and history built-ins.
Commands are saved line by line. Lines that do not contain any non-whitespace characters are not saved in the history. Lines that start with whitespaces are not saved when the hist-space option is on.
Command history is saved in a file. When history is first used after an interactive shell was started, the shell opens a file to save history in. The filename is specified by the value of the HISTFILE variable. If the file contains history data when opened, the data is restored to the shell’s history. The file contents are updated in real time as the user inputs commands into the shell. If the HISTFILE variable is not set or the file cannot be opened successfully, history is not saved in the file, but the history feature will be functional in all other respects.
The number of commands saved in history is specified by the value of the HISTSIZE variable. The shell automatically removes old history data so that the number of saved commands does not exceed the value. If the HISTSIZE variable is not set or its value is not a natural number, 500 items will be saved in history.
The shell looks at the value of the HISTFILE and HISTSIZE variables only when the history feature is first used after the shell was started. “The history feature is used” when:
Therefore, the variables should be set in initialization scripts.
When more than one instance of yash shares a single history file, all the shells use the same history data. As a result, commands that have been executed by a shell instance can be recalled on another shell instance. Shells sharing the same history should have the same HISTSIZE value so that they manipulate history data properly.
Yash’s history data file has its own format that is incompatible with other kinds of shells.
The HISTRMDUP variable can be set to remove duplicate history items.
An interactive shell can notify receipt of email. The shell periodically checks the modification date/time of a file specified by the user. If the file has been modified since the previous check, the shell prints a notification message (except when the shell is not in the POSIXly-correct mode and the file is empty). By specifying a mailbox file to be checked, the shell will print a message when the file has been modified, that is, some mail has been received.
Check is done just before the shell prints a command line prompt. The interval of checks can be specified by the MAILCHECK variable in seconds. If the variable value is 0, check is done before every prompt. If the variable value is not a non-negative integer, no checks are done.
The file whose modification time is checked is specified by the MAIL variable. The variable value should be set to the pathname of the file.
If you want to check more than one file or customize the notification message, you can set the MAILPATH variable instead of the MAIL variable. When the MAILPATH variable is set, the MAIL variable is ignored. The value of the MAILPATH variable should be set to one or more colon-separated pathnames of files to be checked. Each pathname can be followed by a percent sign (%) and a custom notification message, which is printed when the corresponding file has been modified. If the pathname contains a percent sign, it should be quoted by a backslash. The specified message is subject to parameter expansion. For example, if the value of the MAILPATH variable is /foo/mail%New mail!:/bar/mailbox%You've got mail:/baz/mail\%data, the shell will print
Job control is a function of the shell that executes multiple commands simultaneously and suspends/resumes the commands.
When job control is active:
When job control is inactive, processes executed by the shell have the same process group ID as the shell. The shell treats asynchronous commands as an uncontrolled job.
You can use the following built-ins to manipulate jobs:
jobs
fg and bg
wait
disown
kill
An interactive job-controlling shell reports jobs status before every prompt by default. You can set the following options to make the shell report status at other timings:
notify
notify-le
A job is removed from the shell’s job list when:
Jobs are not removed from the list when an interactive shell automatically reports the status of jobs.
The word “stop” is synonymous to “suspend” in the context of job control.
Some built-ins use the following notation, which is called job ID, to specify a job to operate on:
%, %%, %+
%-
%n
%string
%?string
The current job and previous job are jobs selected by the shell according to the following rules:
Yash has some options to modify the rules of the current/previous job selection. (The rules above have priority over the options below.)
cur-async
cur-bg
cur-stop
The current and previous jobs are not changed as long as the rules above are met.
The rules of the current/previous job selection defined in the POSIX standard are looser than yash’s rules above. Other POSIX-compliant shells may select the current and previous jobs differently.
Built-in commands are commands that are implemented in the shell and are executed by the shell without external programs.
There are three types of built-in commands in yash: special built-in commands, semi-special built-in commands and regular built-in commands.
Special built-in commands are much more important commands than others. They are executed regardless of whether the corresponding external commands exist or not. Results of variable assignments that occur in a simple command that invokes a special built-in last after the command has finished. Moreover, in the POSIXly-correct mode, a non-interactive shell immediately exits with a non-zero exit status when a redirect error, assignment error, or misuse of option or operand occurs in a special built-in command.
Semi special built-in commands are the second important built-in commands. They are executed regardless of whether the corresponding external commands exist or not. In other parts they are the same as regular built-in commands.
Regular built-in commands are less important built-in commands including commands that can be implemented as external commands or are not listed in POSIX. In the POSIXly-correct mode, a regular built-in is executed only when a corresponding external command is found in PATH.
In this section we explain common rules about command arguments. The built-in commands of yash follow the rules unless otherwise stated.
There are two types of command arguments. One is options and the other is operands. An option is an argument that starts with a hyphen (-) and changes the way the command behaves. Some options take arguments. An operand is an argument that is not an option and specifies objects the command operates on.
If you specify more than one option to a command, the order of the options are normally not significant. The order of operands, however, affects the command behavior.
An option is either a single-character option or a long option. A single-character option is identified by one alphabetic character. A long option is identified by multiple alphabetic characters. The POSIX standard only prescribes single-character options, so in the POSIXly-correct mode you cannot use long options.
A single-character option is composed of a hyphen followed by a letter. For example, -a is a single-character option. A single-character option that takes an argument requires the argument to be just after the option name.
Example 4. The set built-in and single-character options
For the set built-in, -m is a single-character option that does not take an argument and -o is one that takes an argument.
In these two command lines, errexit is the argument to the -o option.
In the second example above, the -o option and its argument are combined into a single command line argument. The POSIX standard deprecates that style and any POSIX-conforming applications must specify options and their arguments as separate command line arguments, although yash accepts both styles.
You can combine single-character options that do not take arguments into a single command line argument. For example, the three options -a, -b and -c can be combined into -abc.
A long option is composed of two hyphens followed by an option name. For example, --long-option is a long option. You can omit some last characters of a long option name as long as it is not ambiguous. For example, you can use --long instead of --long-option if there is no other options beginning with --long. Like a single-character option, a long option that takes an argument requires the argument to be a command line argument just after the option name or to be specified in the same command line argument as the option name, separated by an equal sign (=).
Example 5. The fc built-in and long options
For the fc built-in, --quiet is a long option that does not take an argument and --editor is one that takes an argument.
In these command lines, vi is the argument to the --editor option.
Arguments that are not options (nor arguments to them) are interpreted as operands. The POSIX standard requires all options should be specified before any operands. Therefore, in the POSIXly-correct mode, any arguments that come after the first operand are interpreted as operands (even if they look like options). If not in the POSIXly-correct mode, you can specify options after operand.
Regardless of whether the shell is in the POSIXly-correct mode or not, an argument that is just composed of two hyphens (--) can be used as a separator between options and operands. All command line arguments after the -- separator are interpreted as operands, so you can specify operands that start with a hyphen correctly using the separator.
Example 6. Options and operands to the set built-in
In this example, -a and -b are options and -c and -d are operands. The -- separator itself is neither an option nor an operand.
Regardless of whether the shell is in the POSIXly-correct mode or not, an argument that is just composed of a single hyphen (-) is interpreted as an operand.
With the line-editing feature, you can edit the command text when you input a command to an interactive shell. It not only works as a simple visual-interface editor, but also is integrated with the command history. You can recall, edit, and execute commands in the history with line-editing instead of using the fc built-in.
Line-editing has two editing modes, the vi and emacs modes, which each have their own key binding settings. By switching editing modes, you can change key bindings used in line-editing. Each mode has a corresponding shell option, which determines whether the mode is currently active or not. No more than one mode can be active at a time, so the options for the other modes are automatically turned off when you turn on the option for one mode. The whole line-editing feature is deactivated when those options are off.
When an interactive shell is started, the vi mode is automatically activated if the standard input and error are both connected to a terminal.
Line-editing can be used only when the standard input and error are both connected to a terminal. If not, the shell silently falls back to the normal input mechanism. While line-editing is being used, the shell uses the termios interface to change I/O settings of the terminal and the terminfo interface to parse input key sequences.
The following options can be set by the set built-in to enable line-editing and choose an editing mode to activate:
vi
emacs
The other line-editing-related options are:
le-always-rp
le-comp-debug
le-conv-meta
le-no-conv-meta
The le-conv-meta and le-no-conv-meta options cannot be both enabled at a time. When either is enabled, the other is automatically disabled. When neither is enabled, the 8th bit may be treated as a meta-key flag depending on terminfo data.
le-predict
le-predict-empty
le-prompt-sp
This option is enabled by default.
le-visible-bell
The vi mode is an editing mode that offers key bindings similar to that of the vi editor. The vi mode has two sub-modes that are switched during editing: the insert and command modes. The sub-mode is always reset to the insert mode when line-editing is started for a new command line. In the insert mode, most characters are inserted to the buffer as typed. In the command mode, input characters are treated as commands that move the cursor, insert/delete text, etc.
The emacs mode offers key bindings similar to the emacs editor. Most characters are inserted to the buffer as typed, but more characters are treated as commands than the vi insert mode.
Another sub-mode is used while you enter search keywords. The sub-mode is called the search mode, which offers slightly different key bindings depending on the active editing mode.
All characters the user enters while line-editing is active are treated as line-editing commands listed below. The bindkey built-in allows customizing the key bindings of each mode (except for the search mode).
The list below shows not only the functions of commands but also the default key bindings. The keywords “vi-insert”, “vi-command”, “vi-search”, “emacs”, “emacs-search” means the vi insert mode, the vi command mode, the search mode for the vi mode (the vi search mode), the emacs mode, and the search mode for the emacs mode (the emacs search mode), respectively.
Some commands take an argument that affects the function of the commands. For example, the forward-char command moves the cursor by as many characters as specified by the argument. To specify an argument, use the digit-argument command just before another command that takes an argument.
Basic editing commands
noop
vi-command
alert
self-insert
vi-insert, emacs
insert-tab
emacs
expect-verbatim
vi-insert, vi-search, emacs-search
emacs
digit-argument
This command can be bound to a digit or hyphen. To pass “12” as an argument to the forward-char command in the vi mode, for example, enter 12l.
vi-command
emacs
bol-or-digit
vi-command
accept-line
If command line prediction is active, the current prediction (if any) is ignored. See also the accept-prediction command.
vi-insert, vi-command, emacs, emacs-search
abort-line
vi-insert, vi-command, vi-search, emacs, emacs-search
eof
eof-if-empty
vi-insert, vi-command
eof-or-delete
emacs
accept-with-hash
vi-command
emacs
accept-prediction
setmode-viinsert
vi-command
setmode-vicommand
vi-insert
setmode-emacs
expect-char, abort-expect-char
redraw-all
vi-insert, vi-command, vi-search, emacs, emacs-search
clear-and-redraw-all
Motion commands
Motion commands move the cursor on the line. Most motion commands accept an argument. When passed an argument, they repeat the cursor motion as many times as specified by the argument. Passing “4” as an argument to the forward-char command, for example, advances the cursor by four characters.
The shell has several definitions of words as units of distance: A bigword is one or more adjacent non-whitespace characters. A semiword is one or more adjacent characters that contain no whitespaces or punctuations. An emacsword is one or more adjacent alphanumeric characters. A viword is either:
forward-char
vi-insert
vi-command
emacs
backward-char
vi-insert
vi-command
emacs
forward-bigword
vi-command
end-of-bigword
vi-command
backward-bigword
vi-command
forward-semiword
end-of-semiword
backward-semiword
forward-viword
vi-command
end-of-viword
vi-command
backward-viword
vi-command
forward-emacsword
emacs
backward-emacsword
emacs
beginning-of-line
vi-insert, vi-command
emacs
end-of-line
vi-insert
vi-command
emacs
go-to-column
vi-command
first-nonblank
vi-command
find-char
vi-command
emacs
find-char-rev
vi-command
emacs
till-char
vi-command
till-char-rev
vi-command
refind-char
vi-command
refind-char-rev
vi-command
Editing commands
Editing commands modify contents of the buffer. Most editing commands accept an argument. When passed an argument, they repeat the modification as many times as specified by the argument.
Texts deleted by commands whose name starts with “kill” are saved in kill ring, from which deleted contents can be restored to the buffer. The most recent 32 texts are kept in the kill ring.
delete-char
vi-insert, emacs
delete-bigword
delete-semiword
delete-viword
delete-emacsword
backward-delete-char
vi-insert, emacs
backward-delete-bigword
backward-delete-semiword
vi-insert
backward-delete-viword
backward-delete-emacsword
delete-line
forward-delete-line
backward-delete-line
vi-insert
kill-char
vi-command
kill-bigword
kill-semiword
kill-viword
kill-emacsword
emacs
backward-kill-char
vi-command
backward-kill-bigword
emacs
backward-kill-semiword
backward-kill-viword
backward-kill-emacsword
emacs
kill-line
forward-kill-line
emacs
backward-kill-line
emacs
put-before
vi-command
put
vi-command
put-left
emacs
put-pop
This command can be used only just after the put-before, put, put-left, or put-pop command.
emacs
undo
vi
emacs
undo-all
vi
emacs
cancel-undo
vi
cancel-undo-all
redo
vi-command
Completion commands
complete
complete-next-candidate
vi-insert, emacs
complete-prev-candidate
vi-insert, emacs
complete-next-column
complete-prev-column
complete-next-page
complete-prev-page
complete-list
If you pass no argument, a list of completion candidates is shown. Otherwise, the word is completed with the nth candidate where n is the argument.
emacs
complete-all
emacs
complete-max
complete-max-then-list
complete-max-then-next-candidate
complete-max-then-prev-candidate
clear-candidates
Vi-specific commands
vi-replace-char
vi-command
vi-insert-beginning
vi-command
vi-append
vi-command
vi-append-to-eol
vi-command
vi-replace
vi-command
vi-switch-case
vi-switch-case-char
vi-command
vi-yank
vi-command
vi-yank-to-eol
vi-command
vi-delete
vi-command
vi-delete-to-eol
vi-command
vi-change
vi-command
vi-change-to-eol
vi-command
vi-change-line
vi-command
vi-yank-and-change
vi-yank-and-change-to-eol
vi-yank-and-change-line
vi-substitute
vi-command
vi-append-last-bigword
vi-command
vi-exec-alias
vi-command
vi-edit-and-accept
vi-command
vi-complete-list
vi-command
vi-complete-all
vi-command
vi-complete-max
vi-command
vi-search-forward
vi-command
vi-search-backward
vi-command
Emacs-specific commands
emacs-transpose-chars
emacs
emacs-transpose-words
emacs
emacs-downcase-word
emacs
emacs-upcase-word
emacs
emacs-capitalize-word
emacs
emacs-delete-horizontal-space
emacs
emacs-just-one-space
emacs
emacs-search-forward
emacs
emacs-search-backward
emacs
History-related commands
oldest-history
newest-history
return-history
oldest-history-bol
vi-command
newest-history-bol
return-history-bol
vi-command
oldest-history-eol
emacs
newest-history-eol
return-history-eol
emacs
next-history
prev-history
next-history-bol
vi-command
prev-history-bol
vi-command
next-history-eol
vi-insert, emacs
prev-history-eol
vi-insert, emacs
search-again
vi-command
search-again-rev
vi-command
search-again-forward
search-again-backward
beginning-search-forward
beginning-search-backward
Search mode commands
srch-self-insert
vi-search, emacs-search
srch-backward-delete-char
vi-search, emacs-search
srch-backward-delete-line
vi-search, emacs-search
srch-continue-forward
emacs-search
srch-continue-backward
emacs-search
srch-accept-search
vi-search
emacs-search
srch-abort-search
vi-search
emacs-search
In the bindkey built-in, escape sequences are used to represent special keys such as function keys and arrow keys. Every escape sequence starts with a backslash (\) and thus there is also an escape sequence for a backslash itself.
Below are available escape sequences:
\\
\B
\D
\E
\H
\I
\L
\N
\P
\R
\U
\X
\!
\#
\$
\?
\^@
\^A, \^B, ..., \^Z
Note that Ctrl + I, Ctrl + J, and Ctrl + M are tab, newline, and carriage return, respectively.
\^[
\^\
\^]
\^^
\^_
\^?
\F00, \F01, ..., \F63
\a1
\a3
\b2
\bg
\bt
\c1
\c3
\ca
\cl
\cn
\co
\cp
\cr
\cs
\ct
\dl
\ei
\el
\es
\et
\ex
\fd
\hp
\il
\ll
\me
\mk
\ms
\mv
\nx
\on
\op
\pr
\pv
\rd
\re
\rf
\rh
\rp
\rs
\sf
\sl
\sr
\st
\su
\sv
\ud
\SE
\SH
\SI
\SL
\SR
\SX
\Sbg
\Scn
\Sco
\Scp
\Scr
\Sdl
\Sel
\Sex
\Sfd
\Shp
\Smg
\Smv
\Snx
\Sop
\Spr
\Spv
\Srd
\Sre
\Srp
\Ssu
\Ssv
\Sud
INTR, EOF, KILL, and ERASE are special characters configured by the stty command. In a typical configuration, they are sent by typing Ctrl+C, Ctrl+D, Ctrl+U, and Ctrl+H, respectively, but some configuration uses Ctrl+? instead of Ctrl+H for ERASE.
By using the complete and complete-next-candidate commands, etc., you can complete command names, options, and operands. By default, the complete-next-candidate command is bound with the Tab key in the vi insert and emacs modes.
Type a few first letters of a command name or pathname and hit the Tab key, and a list of matching names will be shown. You can choose a candidate from the list to complete the name by hitting the Tab key again. If there is only one matching name, no list will be shown and the name will directly be completed.
If the name to be completed contains characters like * and ?, it is treated as a pattern. The name on the command line will be directly substituted with all possible names matching the pattern (you cannot choose from a list).
Normally, command names are completed with command names and command arguments with pathnames. However, completion functions can be defined to refine completion results.
Completion details
When doing completion for the first time after the shell has been started, the INIT file is loaded as if the command string . -AL completion/INIT is executed. If the file is not found, it is silently ignored. This automatic loading is mainly intended for loading completion functions bundled with the shell, but you can let the shell load your own functions by putting a file in the load path.
When completing a command name, the shell executes the completion//command function and when completing a command argument, the completion//argument function. If those completion functions are not defined, the shell just completes with command names or pathnames. When completing other names, such as the user name in tilde expansion and the parameter name in parameter expansion, completion functions are never used: the shell just completes with user names, parameter names, or whatever applicable.
Completion functions are executed without any arguments. The following local variables are automatically defined while executing completion functions:
IFS
WORDS
TARGETWORD
Completion candidates are generated by executing the complete built-in during a completion function.
Completion functions must not perform I/O to the terminal, or displayed text will be corrupted. Completion functions should run as quickly as possible for better user experience.
While a completion function is being executed:
This is an experimental feature. When the le-predict option is enabled, the shell automatically tries to predict a command string fragment that follows the part of the command you have already typed in line-editing.
For example, assume you have once typed the command ls Documents. Next time you start typing ls Doc, the shell will show uments just after the cursor. If you are satisfied with this suggestion, you can move the cursor to the right by the forward-char or any other motion commands instead of typing the rest of the command. After moving the cursor after the last s, you can use the accept-line command to execute the command. You can also use the accept-prediction command to immediately execute the suggested command without moving the cursor.
To distinguish the typed part and the predicted part of a command string, you can change the font style of the typed part by setting the PS1S variable. Customizing the font style of the predicted part is not (yet) supported; it is always shown in the default style.
When you move the cursor to the right, the predicted part up to the cursor becomes the typed part as if you actually typed it. Moving the cursor to the left does not turn the typed part back to the prediction. Use deletion commands such as backward-delete-char to delete typed command fragment.
The predicted part of the command is shown only when the cursor is at the end of the typed part.
By default, the predicted part is shown after you start typing a command string. By enabling the le-predict-empty option, the predicted part is also shown before you type the first character of the command string.
The prediction algorithm suggests command fragments on the basis of the command history. The algorithm considers recent history entries more probable. It also takes command succession patterns into account. A predicted command fragment is not always a complete valid command because less probable part of the fragment is excluded from prediction.
Yash behaves as defined in POSIX.1-2008, Shell & Utilities for the most part, but some functionalities disobey POSIX for usability. When full POSIX-conformance is needed, you can enable the POSIXly-correct mode to make yash obey POSIX as much as possible.
If yash is started with the name “sh”, the POSIXly-correct mode is automatically enabled. The -o posixly-correct command-line option also enables the POSIXly-correct mode. After yash has been started, the POSIXly-correct mode can be enabled by executing the command string set -o posixly-correct.
When the POSIXly-correct mode is on, yash not only tries to obey the requirements by POSIX, but also treats as errors most conditions where the behavior is undefined or unspecified by POSIX. As a result, most yash-specific functionalities are disabled in the POSIXly-correct mode.
Below is the complete list of the behavioral differences between when yash is in the POSIXly-correct mode and when not. When the POSIXly-correct mode is enabled:
You need to set locale environment variables to enable Unicode character support.
If the variables have already been configured for your preference, you probably don’t have to do anything. To check the current locale configurations, you can use the locale command:
$ locale LANG= LC_CTYPE="en_US.utf8" LC_NUMERIC="en_US.utf8" LC_TIME="en_US.utf8" LC_COLLATE="en_US.utf8" LC_MONETARY="en_US.utf8" LC_MESSAGES="en_US.utf8" LC_PAPER="en_US.utf8" LC_NAME="en_US.utf8" LC_ADDRESS="en_US.utf8" LC_TELEPHONE="en_US.utf8" LC_MEASUREMENT="en_US.utf8" LC_IDENTIFICATION="en_US.utf8" LC_ALL=en_US.utf8
In this example, the locale command shows that all the locale setting categories are configured for the English language, the United States region, and the UTF-8 encoding.
If the current configuration does not seem to match your preference, set the LC_ALL variable like this:
export LC_ALL=en_US.utf8
If you want to use other languages, regions, or encodings, you have to set the variable to a different value. Please consult your OS’s documentation to learn how to configure these variables in detail.
If you want to apply the same configuration every time you start yash, write the command in ~/.yashrc or ~/.yash_profile.
If yash still rejects Unicode characters being entered, see the section below regarding line-editing.
First, type echo $TERM and see if it prints a sane value. xterm is the safest value that should work on any existing environment. Colored versions like xterm-16color and other terminal types like rxvt and vt100 may also work. All possible values can be listed by the toe command. It’s most desirable to choose a value that matches the actual terminal type you are using, but it might not work if the terminal type you chose is not supported on the system on which yash is running. If so, try changing the TERM value by export TERM=xterm, for example, to find a value that works.
If line-editing works but you have trouble entering Unicode (non-ASCII) characters, try enabling the le-no-conv-meta option by set -o le-no-conv-meta.
This chapter defines the syntax of the shell command language.
Some of the syntactic features described below are not supported in the POSIXly-correct mode.
The characters of the input source code are first delimited into tokens. Tokens are delimited so that the earlier token spans as long as possible. A sequence of one or more unquoted blank characters delimits a token.
The following tokens are the operator tokens:
& && ( ) ; ;; | || < << <& <( <<- <<< <> > >> >& >( >>| >| (newline)
Unlike other programming languages, the newline operator is a token rather than a white space.
Characters that are not blank nor part of an operator compose a word token. Words are parsed by the following parsing expression grammar:
Word
WordElement
' (!' .)* ' /
" QuoteElement* " /
Parameter /
Arithmetic /
CommandSubstitution
QuoteElement
Parameter /
Arithmetic /
CommandSubstitutionQuoted /
![`"\] .
Parameter
$ PortableName /
$ ParameterBody
PortableName
ParameterBody
ParameterNumber
ParameterName
[[:alnum:] _]+
ParameterIndex
ParameterIndexWord
ParameterMatch
(# / ## / % / %%) ParameterMatchWord /
(:/ / / [#%/]?) ParameterMatchWordNoSlash (/ ParameterMatchWord)?
ParameterMatchWord
ParameterMatchWordNoSlash
Arithmetic
ArithmeticBody
Parameter /
Arithmetic /
CommandSubstitution /
( ArithmeticBody ) /
![`()] .
CommandSubstitution
` CommandSubstitutionBody* `
CommandSubstitutionQuoted
` CommandSubstitutionBodyQuoted* `
CommandSubstitutionBody
!` .
CommandSubstitutionBodyQuoted
!` .
SpecialChar
The set of terminals of the grammar is the set of characters that can be handled on the environment in which the shell is run (a.k.a. execution character set), with the exception that the set does not contain the null character ('\0').
Strictly speaking, the definition above is not a complete parsing expression grammar because the rule for CommandSubstitution (Quoted) depends on CompleteProgram which is a non-terminal of the syntax.
Token classification
After a word token is delimited, the token may be further classified as an IO_NUMBER token, reserved word, name word, assignment word, or just normal word. Classification other than the normal word is applied only when applicable in the context in which the word appears. See Tokens and keywords for the list of the reserved words (keywords) and the context in which a word may be recognized as a reserved word.
A token is an IO_NUMBER token iff it is composed of digit characters only and immediately followed by < or >.
An assignment token is a token that starts with a name followed by =:
AssignmentWord
AssignmentPrefix
Name
Comments
A comment begins with # and continues up to (but not including) the next newline character. Comments are treated like a blank character and do not become part of a token. The initial # of a comment must appear as if it would otherwise be the first character of a word token; Other #s are just treated as part of a word token.
Comment
After tokens have been delimited, the sequence of the tokens is parsed according to the context-free grammar defined below, where *, +, and ? should be interpreted in the same manner as standard regular expression:
CompleteProgram
CompoundList
AndOrList
Pipeline
Command
FunctionDefinition |
SimpleCommand
CompoundCommand
Grouping |
IfCommand |
ForCommand |
WhileCommand |
CaseCommand |
DoubleBracketCommand |
FunctionCommand
Subshell
Grouping
IfCommand
ForCommand
WhileCommand
CaseCommand
CaseList
CaseItem
DoubleBracketCommand
Ors
Ands
Nots
Primary
Word (-ef | -eq | -ge | -gt | -le | -lt | -ne | -nt | -ot | -veq | -vge | -vgt | -vle | -vlt | -vne | = | == | === | =~ | != | !== | < | >) Word |
( Ors ) |
Word
FunctionCommand
FunctionDefinition
SimpleCommand
Word (Word | Redirection)*
Assignment
AssignmentPrefix( NL* (Word NL*)* )
Redirection
IO_NUMBER? <( CompleteProgram ) |
IO_NUMBER? >( CompleteProgram )
RedirectionOperator
NL
In the rule for Primary, Word tokens must not be ]]. Additionally, if a Primary starts with a Word, it must not be any of the possible unary operators allowed in the rule.
In the rule for SimpleCommand, a Word token is accepted only when the token cannot be parsed as the first token of an Assignment.
In the rule for Assignment, the ( token must immediately follow the AssignmentPrefix token, without any blank characters in between.
Here-document contents do not appear as part of the grammar above. They are parsed just after the newline (NL) token that follows the corresponding redirection operator.
Alias substitution
Word tokens are subject to alias substitution.
Tokens that are classified as reserved words are not subject to alias substitution.
The alias built-in defines and/or prints aliases.
The alias built-in defines and/or prints aliases as specified by operands. The printed aliases can be used as (part of) shell commands. The built-in prints all currently defined aliases when given no operands.
-g, --global
-p, --prefix
name
name=value
The exit status of the alias built-in is zero unless there is any error.
The characters that cannot be used in an alias name are the space, tab, newline, and any of =$<>\'"`;&|()#. You can use any characters in an alias value.
The alias built-in is a semi-special built-in.
The POSIX standard defines no options for the alias built-in, thus no options are available in the POSIXly correct mode.
The array built-in prints or modifies arrays.
When executed without any option or operands, the built-in prints all array definitions to the standard output in a form that can be parsed as commands.
When executed with name and values (but without an option), the built-in sets the values as the values of the array named name.
With the -d (--delete) option, the built-in removes the indexth values of the array named name. The number of values in the array will be decreased by the number of the indexes specified. If the indexth value does not exist, it is silently ignored.
With the -i (--insert) option, the built-in inserts values into the array named name. The number of values in the array will be increased by the number of the values specified. The values are inserted between the indexth and next values. If index is zero, the values are inserted before the first value. If index is larger than the number of values in the array, the values are appended after the last element.
With the -s (--set) option, the built-in sets value as the indexth value of the array named name. The array must have at least index values.
-d, --delete
-i, --insert
-s, --set
name
index
value
The exit status of the array built-in is zero unless there is any error.
The array built-in is not defined in the POSIX standard.
The command array name value... is equivalent to the assignment name=(value...).
The bg built-in resumes a job in the background.
The bg built-in sends the SIGCONT signal to the specified job. As a result, the job is resumed in the background (if it has been suspended).
The name of the job is printed when the job is resumed.
The built-in can be used only when job control is enabled.
job
More than one job can be specified at a time. The current job is resumed if none is specified.
The percent sign (%) at the beginning of a job ID can be omitted if the shell is not in the POSIXly-correct mode.
The exit status of the bg built-in is zero unless there is any error.
The bg built-in is a semi-special built-in.
The POSIX standard provides that the built-in shall have no effect when the job is already running. The bg built-in of yash, however, always sends the SIGCONT signal to the job.
The bindkey built-in prints or modifies key bindings used in line-editing.
When executed with the -l (--list) option, the built-in lists all available line-editing commands to the standard output.
When executed with one of the other options, the built-in prints or modifies key bindings for the editing mode specified by the option:
-a, --vi-command
-e, --emacs
-v, --vi-insert
key
command
The exit status of the bindkey built-in is zero unless there is any error.
The bindkey built-in is a semi-special built-in. In the POSIX standard, it is defined as a command with unspecified behavior.
The break built-in aborts a loop being executed.
When executed without the -i (--iteration) option, the built-in aborts a currently executed for, while, or until loop. When executed in nested loops, it aborts the nestth innermost loop. The default nest is one. If the number of currently executed nested loops is less than nest, the built-in aborts the outermost loop.
When executed with the -i (--iteration) option, the built-in aborts the currently executed (innermost) iterative execution.
-i, --iteration
nest
The exit status of the break built-in is:
The break built-in is a special built-in.
The POSIX standard defines no options for the break built-in; the built-in accepts no options in the POSIXly-correct mode.
Treatment of currently executed loops that are not lexically enclosing the break built-in is unspecified in POSIX. Examples of such loops include:
Yash does not allow breaking such loops.
The cd built-in changes the working directory.
The cd built-in changes the working directory to the directory specified by the operand.
The pathname of the new working directory is assigned to the PWD variable, whose previous value is again assigned to the OLDPWD variable. The new PWD value will not contain any . or .. components except when the shell is in the POSIXly-correct mode and the new pathname begins with /...
If directory is a relative path that does not start with ‘.’ or ‘..’, paths in the CDPATH variable are searched to find a new working directory. The search is done in a manner similar to the last step of command search, but a directory is sought instead of an executable regular file. If a new working directory was found from CDPATH, its pathname is printed to the standard output. If no applicable directory was found in the search, directory is simply treated as a pathname relative to the current working directory.
If the working directory was successfully changed, the value of the YASH_AFTER_CD variable is executed as a command unless the shell is in the POSIXly-correct mode. If the variable is an array, its values are executed iteratively (cf. eval built-in).
-L, --logical
-P, --physical
--default-directory=directory
The -L (--logical) and -P (--physical) options are mutually exclusive: only the last specified one is effective. If neither is specified, -L is assumed.
directory
If directory is a single hyphen (‘-’), the value of the OLDPWD variable is assumed for the new directory pathname, which is printed to the standard output.
If directory is omitted, the working directory is changed to the directory specified by the --default-directory=... option. If that option is not specified either, the default is the home directory.
The exit status of the cd built-in is zero if the working directory was successfully changed and non-zero if there was an error.
The cd built-in is a semi-special built-in.
The POSIX standard does not define the use of the YASH_AFTER_CD variable or the --default-directory=... option. The standard does not allow using an option with a single hyphen operand.
The exit status of the commands in the YASH_AFTER_CD variable does not affect that of the cd built-in.
The colon built-in does nothing.
The colon built-in does nothing. Any command line arguments are ignored.
The exit status of the colon built-in is zero.
The colon built-in is a special built-in.
Arguments are expanded and redirections are performed as usual. The colon and true built-ins have the same effect, but colon is a special built-in while true is a semi-special.
The command built-in executes or identifies a command.
Without the -v (--identify) or -V (--verbose-identify) option, the built-in executes command with arguments in the same manner as the last step of execution of simple commands. The command is treated as a built-in or external command or a function according to the options specified to the command built-in. The shell does not exit on argument syntax error etc. even if the command is a special built-in
With the -v (--identify) option, command is identified. If the command is found in $PATH, its full pathname is printed. If it is a keyword, function, or built-in that is not found in $PATH, the command name is simply printed. If it is an alias, it is printed in the form like alias ll='ls -l'. If the command is not found, nothing is printed and the exit status is non-zero.
The -V (--verbose-identify) option is similar to the -v (--identify) option, but the output format is more human-friendly.
-a, --alias
-b, --builtin-command
-e, --external-command
-f, --function
-k, --keyword
-p, --standard-path
-v, --identify
-V, --verbose-identify
If none of the -a (--alias), -b (--builtin-command), -e (--external-command), -f (--function), and -k (--keyword) options is specified, the following defaults are assumed:
Without the -v (--identify) or -V (--verbose-identify) option
With the -v (--identify) or -V (--verbose-identify) option
command
argument...
The exit status of the command built-in is:
Without the -v (--identify) or -V (--verbose-identify) option
With the -v (--identify) or -V (--verbose-identify) option
The command built-in is a semi-special built-in.
In the POSIXly-correct mode, options other than -p, -v, and -V cannot be used and at most one command can be specified. The POSIX standard does not allow specifying both -v and -V together, but yash does (only the last specified one is effective).
The complete built-in generates completion candidates. This built-in can only be executed from completion functions during command line completion.
The built-in generates completion candidates according to the specified arguments. No matter how candidates are generated, only candidates that match the word being completed are generated.
-A pattern, --accept=pattern
-D description, --description=description
-O, --option
-P prefix, --prefix=prefix
If the word being completed is file:///home/user/docume for example, the command line complete -P file:// -f will generate pathname candidates that complete /home/user/docume.
-R pattern, --reject=pattern
-S suffix, --suffix=suffix
-T, --no-termination
Options that select candidate types
-a, --alias
--array-variable
--bindkey
-b, --builtin-command
-c, --command
-d, --directory
--dirstack-index
--executable-file
--external-command
-f, --file
--finished-job
--function
--global-alias
-g, --group
-h, --hostname
-j, --job
-k, --keyword
--normal-alias
--regular-builtin
--running-job
--scalar-variable
--semi-special-builtin
--signal
--special-builtin
--stopped-job
-u, --username
-v, --variable
If the -d (--directory) option is specified without the -f (--file) option, the -S / -T options are assumed.
Generated candidates for job IDs do not have leading percent signs (%). If the word being completed starts with a percent sign, the -P % option should be specified.
Operands are treated as completion candidates.
The exit status of the built-in is zero if one or more candidates were generated, one if no candidates were generated, or larger than one if an error occurred.
The complete built-in is a semi-special built-in. In the POSIX standard, it is defined as a command with unspecified behavior.
The continue built-in skips an iteration of a loop being executed.
When executed without the -i (--iteration) option, the built-in aborts the current iteration of for, while, or until loop and starts the next iteration of the loop. When executed in nested loops, it affects the nestth innermost loop. The default nest is one. If the number of currently executed nested loops is less than nest, the built-in affects the outermost loop.
When executed with the -i (--iteration) option, the built-in aborts the current iteration of (innermost) iterative execution.
-i, --iteration
nest
The exit status of the continue built-in is:
The continue built-in is a special built-in.
The POSIX standard defines no options for the continue built-in; the built-in accepts no options in the POSIXly-correct mode.
Treatment of currently executed loops that are not lexically enclosing the continue built-in is unspecified in POSIX. Examples of such loops include:
Yash does not allow continuing such loops.
The dirs built-in prints the contents of the directory stack.
The directory stack is a feature that records history of working directories. You can use the pushd built-in to save a working directory in the directory stack, the popd built-in to recall the saved working directory, and the dirs built-in to see the stack contents. Those built-ins use the DIRSTACK array and the PWD variable to save the stack contents. Modifying the array means modifying the stack contents.
Directory stack entries are indexed by signed integers. The entry of index +0 is the current working directory, +1 is the last saved directory, +2 is the second last, and so on. Negative indices are in the reverse order: the entry of index -0 is the first saved directory, -1 is the second, and -n is the current working directory if the stack has n entries,
When executed without the -c (--clear) option, the dirs built-in prints the current contents of the directory stack to the standard output. With the -c (--clear) option, the built-in clears the directory stack.
-c, --clear
-v, --verbose
index
You can specify more than one index. If you do not specify any index, all the entries are printed.
The exit status of the dirs built-in is zero unless there is any error.
The dirs built-in is a semi-special built-in. In the POSIX standard, it is defined as a command with unspecified behavior.
The disown built-in removes jobs.
The disown built-in removes the specified jobs from the job list. The removed jobs will no longer be job-controlled, but the job processes continue execution (unless they have been suspended).
-a, --all
job
You can specify more than one job ID. If you do not specify any job ID, the current job is removed. If the shell is not in the POSIXly-correct mode, the %-prefix of the job ID can be omitted.
The exit status of the disown built-in is zero unless there is any error.
The disown built-in is a semi-special built-in. In the POSIX standard, it is defined as a command with unspecified behavior.
The dot built-in reads a file and executes commands in it.
The dot built-in reads the specified file, parses its contents as commands, and executes them in the current command execution environment.
If arguments are specified, positional parameters are temporarily set to them. The positional parameters will be restored when the dot built-in finishes. If no arguments are specified, the positional parameters are not changed.
If file does not contain any slashes, the shell searches $PATH for a readable (but not necessarily executable) shell script file whose name is file in the same manner as command search. If no such file was found, the shell searches the current working directory for a file unless in the POSIXly-correct mode. To ensure that the file in the current working directory is used, start file with ‘./’.
-A, --no-alias
-L, --autoload
The dot built-in treats as operands any command line arguments after the first operand.
file
arguments...
The exit status of the dot built-in is that of the last command executed. The exit status is zero if the file contains no commands to execute and non-zero if a file was not found or could not be opened.
The dot built-in is a special built-in.
A non-interactive shell immediately exits with a non-zero exit status if the dot built-in fails to find or open a file to execute.
The POSIX standard defines no options for the dot built-in; the built-in accepts no options in the POSIXly-correct mode.
The POSIX standard does not define the arguments... operands. It is an error to specify the arguments... operands in the POSIXly-correct mode.
The echo built-in prints its arguments.
The built-in treats all command line arguments as operands except for the options described below. Any word that cannot be parsed as an acceptable option is treated as an operand. Options must precede all operands. Syntax errors never happen in the echo built-in.
The echo built-in prints the operand strings followed by a newline to the standard output. The strings are each separated by a space.
Escape sequences
The ECHO_STYLE variable and the -e option enable escape sequences that are replaced with corresponding characters:
\a
\b
\c
\e
\f
\n
\r
\t
\v
\\
\0xxx
When escape sequences are not enabled, they are just printed intact.
ECHO_STYLE variable
The ECHO_STYLE variable defines which options are accepted and whether escape sequences are enabled by default. The variable value should be set to one of the following:
SYSV, XSI
BSD
GNU
ZSH
DASH
RAW
When the ECHO_STYLE variable is not set, it defaults to SYSV.
-n
-e
-E
The exit status of the echo built-in is zero unless there is any error.
The POSIX standard does not define the ECHO_STYLE variable nor any options for the built-in. According to POSIX, the behavior of the built-in is implementation-defined when the first argument is -n or when any argument contains a backslash. For maximum portability, the printf built-in should be preferred over the echo built-in.
Although many values for the ECHO_STYLE variable are defined on the basis of other existing implementations, yash is not intended to exactly imitate those originals. Zsh’s echo built-in interprets a single hyphen argument as a separator between options and operands. Yash does not support such use of hyphen.
The eval built-in evaluates operands as commands.
The eval built-in requires that all options precede operands. Any command line arguments after the first operand are all treated as operands.
The eval parses operands as commands and executes them in the current command execution environment.
When executed without the -i (--iteration) option, all the operands are concatenated into one string (with a space inserted between each operand) and parsed/executed at once.
With the -i (--iteration) option, the built-in performs iterative execution: operands are parsed/executed one by one. If the continue built-in is executed with the -i (--iteration) option during the iterative execution, the execution of the current operand is aborted and the next operand is parsed/executed immediately. The break built-in with the -i (--iteration) option is similar but the remaining operands are not parsed/executed. The value of the ? special parameter is saved before the iterative execution is started. The parameter value is restored to the saved one after each iteration.
-i, --iteration
command
The exit status is zero if no command was specified or command contained no actual command that can be executed. Otherwise, that is, if the eval built-in executed one or more commands, the exit status of the eval built-in is that of the last executed command.
The eval built-in is a special built-in.
The POSIX standard defines no options for the eval built-in; the built-in accepts no options in the POSIXly-correct mode.
The exec built-in replaces the shell process with another external command.
The exec built-in requires that all options precede operands. It is important so that options to the exec built-in are not confused with options to command. Any command line arguments after command are treated as arguments.
When the exec built-in is executed with command, the shell executes command with arguments in a manner similar to the last step of execution of a simple command. The differences are that command is always treated as an external command ignoring any existing functions and built-ins and that the exec system call that starts the external command is called in the current command execution environment instead of a subshell, replacing the shell process with the new command process.
If the shell is in the POSIXly-correct mode or not interactive, failure in execution of command causes the shell to exit immediately.
If an interactive shell that is not in the POSIXly-correct mode has a stopped job, the shell prints a warning message and refuses to execute command. Once the shell process is replaced with an external command, information about the shell’s jobs is lost, so you will have to resume or kill the stopped jobs by sending signals by hand. To force the shell to execute command regardless, specify the -f (--force) option.
When executed without command, the built-in does nothing. As a side effect, however, redirection applied to the built-in remains in the current command execution environment even after the built-in finished.
-a name, --as=name
-c, --clear
-f, --force
command
argument...
If the shell process was successfully replaced with the external command, there is no exit status since the shell process no longer exists.
The exit status is:
The exec built-in is a special built-in.
The POSIX standard defines no options for the exec built-in; the built-in accepts no options in the POSIXly-correct mode.
The exit built-in causes the shell process to exit.
The exit built-in causes the current shell (or subshell) process to exit.
If an interactive shell has a stopped job, the shell prints a warning message and refuses to exit. To force the shell to exit regardless, specify the -f (--force) option or execute the built-in twice in a row.
If an EXIT trap has been set, the shell executes the trap before exiting.
-f, --force
exit_status
If this operand is omitted, the exit status of the shell will be that of the last command executed before the exit built-in (but, if the built-in is executed during a trap, the exit status will be that of the last command before the trap is entered).
If exit_status is 256 or larger, the actual exit status will be the remainder of exit_status divided by 256.
Because the built-in causes the shell to exit, there is no exit status of the built-in.
As an exception, if the shell refused to exit, the exit status of the built-in is non-zero.
The exit built-in is a special built-in.
The POSIX standard defines no options for the exit built-in; the built-in accepts no options in the POSIXly-correct mode.
The POSIX standard provides that the exit_status operand should be between 0 and 255 (inclusive). Yash accepts integers larger than 255 as an extension.
If the built-in is executed during an EXIT trap, the shell just exits without executing the trap again. If exit_status was not specified, the exit status of the shell is what the exit status would be if the trap had not been set. See also Termination of the shell.
The export built-in marks variables for export to child processes.
The export built-in is equivalent to the typeset built-in with the -gx option.
The export built-in is a special built-in.
The POSIX standard defines the -p option only; other options cannot be used in the POSIXly-correct mode. The POSIX does not allow using the option together with operands.
The false built-in does nothing unsuccessfully.
The false built-in does nothing. Any command line arguments are ignored.
The exit status of the false built-in is non-zero.
The false built-in is a semi-special built-in.
The fc built-in re-executes or prints commands from command history.
When executed without the -l (--list) option, the built-in executes the commands in the command history range specified by the operands. If the -s (--silent) option is not specified, the shell invokes an editor which allows you to edit the commands before they are executed. The commands are executed when you quit the editor. If the -s (--silent) option is specified, the commands are immediately executed. In either case, the executed commands are printed to the standard output and added to the history.
When executed with the -l (--list) option, the built-in prints the commands in the command history range specified by the operands. By default, commands are printed with their history entry numbers, but output format can be changed using the -n (--no-numbers)) and -v (--verbose) options.
-e editor, --editor=editor
If this option is not specified, the value of the FCEDIT variable is used. If the variable is not set either, ed is used.
-l, --list
-n, --no-numbers
-q, --quiet
-r, --reverse
-s, --silent
-v, --verbose
start and end
If the first entry of the range that is specified by start is newer than the last entry of the range that is specified by end, the range is reversed as if the -r (--reverse) option was specified. (If the option is already specified, it is cancelled.)
The default values for start and end are:
| with -l | without -l | |
| start | -16 | -1 |
| end | -16 | same as start |
old=new
If commands was executed, the exit status of the fc built-in is that of the last executed command. Otherwise, the exit status is zero unless there is any error.
The fc built-in is a semi-special built-in.
The POSIX standard does not define the -q (--quiet) or -v (--verbose) options, so they cannot be used in the POSIXly-correct mode.
Command history cannot be modified during line-editing.
The fg built-in resumes a job in the foreground.
The fg built-in brings the specified job to the foreground and sends the SIGCONT signal to the job. As a result, the job is resumed in the foreground (if it has been suspended). The built-in then waits for the job to finish and returns the exit status of it.
The name of the job is printed when the job is resumed. When not in the POSIXly-correct mode, the job number is also printed.
The built-in can be used only when job control is enabled.
job
If more than one job is specified, they are resumed in order, one at a time. The current job is resumed if none is specified.
The percent sign (%) at the beginning of a job ID can be omitted if the shell is not in the POSIXly-correct mode.
The exit status of the fg built-in is that of the (last) job resumed. The exit status is non-zero when there was some error.
The fg built-in is a semi-special built-in.
You cannot specify more than one job in the POSIXly-correct mode.
The getopts built-in parses command options.
The getopts built-in parses single-character options that appear in arguments. Each time the built-in is invoked, it parses one option and assigns the option character to variable.
The optionlist operand is a list of option characters that should be accepted by the parser. In optionlist, an option that takes an argument should be specified as the option character followed by a colon. For example, if you want the -a, -b and -c options to be parsed and the -b option to take an argument, then optionlist should be ab:c.
When an option that takes an argument is parsed, the argument is assigned to the OPTARG variable.
When an option that is not specified in optionlist is found or when an option argument is missing, the result depends on the first character of optionlist:
The built-in parses one option for each execution. For all options in a set of command line arguments to be parsed, the built-in has to be executed repeatedly with the same arguments. The built-in uses the OPTIND variable to remember which argument should be parsed next. When the built-in is invoked for the first time, the variable value must be 1, which is the default value. You must not modify the variable until all the options have been parsed, when the built-in sets the variable to the index of the first operand in arguments. (If there are no operands, it will be set to the number of arguments plus one.)
When you want to start parsing a new set of arguments, you have to reset the OPTIND variable to 1 beforehand.
optionlist
variable
arguments
When no arguments are given, the positional parameters are parsed.
If an option is found, whether or not it is specified in optionlist, the exit status is zero. If there is no more option to be parsed, the exit status is non-zero.
aopt=false bopt= copt=false while getopts ab:c opt do
case $opt in
a) aopt=true ;;
b) bopt=$OPTARG ;;
c) copt=true ;;
\?) return 2 ;;
esac done if $aopt; then echo Option -a specified; fi if [ -n "$bopt" ]; then echo Option -b $bopt specified; fi if $copt; then echo Option -c specified; fi shift $((OPTIND - 1)) echo Operands are: $*
In arguments that are parsed, options must precede operands. The built-in ends parsing when it encounters the first operand.
The getopts built-in is a semi-special built-in.
The POSIX standard does not specify what will happen when the OPTIND variable is assigned a value other than 1.
In the POSIXly-correct mode, option characters in optionlist must be alphanumeric.
The hash built-in remembers, forgets, or reports command locations.
When executed with commands but without options, the built-in immediately performs command path search and caches commands' full paths.
When executed with the -r (--remove) option, it removes the paths of commands (or all cached paths if none specified) from the cache.
When executed without options or commands, it prints the currently cached paths to the standard output.
With the -d (--directory) option, the built-in does the same things to the home directory cache, rather than the command path cache. Cached home directory paths are used in tilde expansion.
-a, --all
Without this option, paths for built-ins are not printed.
-d, --directory
-r, --remove
command
user
The exit status of the hash built-in is zero unless there is any error.
The shell automatically caches command and directory paths when executing a command or performing tilde expansion, so normally there is no need to use this built-in explicitly to cache paths.
Assigning a value to the PATH variable removes all command paths from the cache as if hash -r was executed.
The POSIX standard defines the -r option only: other options cannot be used in the POSIXly-correct mode.
The hash built-in is a semi-special built-in.
The help built-in prints usage of built-ins.
The help built-in prints a description of built-ins.
built-ins
The exit status of the help built-in is zero unless there is any error.
The help built-in is a semi-special built-in. In the POSIX standard, it is defined as a command with unspecified behavior.
Many built-ins of yash accept the --help option that prints the same description.
The history built-in prints or edits command history.
The history built-in prints or edits command history.
When executed with an option, the built-in edits history according to the option. If more than one option is specified, each option is processed in order.
When executed with the count operand, the built-in prints the most recent count history entries to the standard output in the same manner as the fc built-in.
When executed with neither options nor operands, the built-in prints the whole history.
-c, --clear
-d entry, --delete=entry
-F, --flush-file
-r file, --read=file
-s command, --set=command
-w file, --write=file
count
The exit status of the history built-in is zero unless there is any error.
The history built-in is a semi-special built-in. In the POSIX standard, it is defined as a command with unspecified behavior.
Command history cannot be modified during line-editing.
The jobs built-in reports job status.
The jobs built-in prints information of jobs the shell is currently controlling.
By default, the following information is printed for each job, line by line:
-l, --verbose
-n, --new
-p, --pgid-only
-r, --running-only
-s, --stopped-only
jobs
The percent sign (%) at the beginning of a job ID can be omitted if the shell is not in the POSIXly-correct mode.
The exit status of the jobs built-in is zero unless there is any error.
The jobs built-in is a semi-special built-in.
The POSIX standard defines the -l and -p options only: other options cannot be used in the POSIXly-correct mode. In the POSIXly-correct mode, the effect of the -l option is different in that status is reported for each job rather than for each process.
The process group ID of a job executed by yash is equal to the process ID of the first command of the pipeline that forms the job.
The kill built-in sends a signal to processes.
The kill built-in requires that all options precede operands. Any command line arguments after the first operand are all treated as operands.
When executed without the -l option, the built-in sends a signal to processes. The signal sent can be specified by option. The SIGTERM signal is sent if no signal is specified.
When executed with the -l option, the built-in prints information of signals to the standard output. If no signal is specified, information of all signals is printed.
Signal-specifying options
-signal, -s signal, -n signal
You can specify at most one signal-specifying option at a time.
Other options
-l
-v
Without this option, the built-in prints the signal name only. This option adds the signal number and a short description.
When the -v option is specified, the -l option can be omitted.
processes
Processes can be specified by the process ID, the process group ID, or the job ID. The process group ID must be prefixed with a hyphen (-) so that it is not treated as a process ID.
When 0 is specified as process, the signal is sent to the process group to which the shell process belongs. When -1 is specified, the signal is sent to all processes on the system.
signal
The signal can be specified by the name, the number, or the exit status of a command that was killed by the signal.
The exit status of the kill built-in is zero unless there is any error. If the signal was sent to at least one process, the exit status is zero even if the signal was not sent to all of the specified processes.
The kill built-in is a semi-special built-in.
Command arguments that start with a hyphen should be used with care. The command kill -1 -2, for example, sends signal 1 to process group 2 since -1 is treated as a signal-specifying option and -2 as an operand that specifies a process group. The commands kill -- -1 -2 and kill -TERM -1 -2, on the other hand, treats both -1 and -2 as operands.
The POSIX standard does not define the -n or -v options, so they cannot be used in the POSIXly-correct mode. The standard does not allow specifying a signal number as the argument of the -s option or a signal name as the signal operand.
The standard requires signal names to be specified without the SIG prefix, like INT and QUIT. If the shell is not in the POSIXly-correct mode, the built-in accepts SIG-prefixed signal names as well.
The local built-in prints or sets local variables.
The local built-in is equivalent to the typeset built-in except that the -f (--functions) and -g (--global) options cannot be used.
The local built-in is a semi-special built-in. In the POSIX standard, it is defined as a command with unspecified behavior.
The popd built-in pops a directory from the directory stack.
The popd built-in removes the last entry from the directory stack, returning to the previous working directory. If index is given, the entry specified by index is removed instead of the last one.
index
If omitted, +0 (the last entry) is assumed.
The exit status of the popd built-in is zero unless there is any error.
It is an error to use this built-in when there is only one directory stack entry.
The popd built-in is a semi-special built-in. In the POSIX standard, it is defined as a command with unspecified behavior.
The printf built-in prints formatted values.
The printf built-in formats values according to format and prints them to the standard output. Unlike the echo built-in, the printf built-in does not print a newline automatically.
The formatting process is very similar to that of the printf function in the C programming language. You can use conversion specifications (which start with %) and escape sequences (which start with \) in format. Any other characters that are not part of a conversion specification or escape sequence are printed literally.
Conversion specifications
A conversion specification starts with a percent sign (%).
A conversion specification except %% consumes a value, which is formatted according to the specification and printed. Each conversion specification consumes one value in the order of appearance. If there are more values than conversion specifications, the entire format is re-processed until all the values are consumed. If a value to be consumed is missing, it is assumed to be an empty string (if the specification requires a string) or zero (if a number). If no values are given, format is processed just once.
Available conversion specifications are:
%d, %i
%u
%o
%x
%X
%f
%F
%e
%E
%g
%G
%c
%s
%b
%%
For %g and %G, the specification that is actually used is %f or %F if the exponent part is between -5 and the precision (exclusive); %e or %E otherwise.
In a conversion specification except %%, the leading percent sign may be followed by flags, field width, and/or precision in this order.
The flags are a sequence of any number of the following characters:
Minus sign (-)
Plus sign (+)
Space ( )
Hash sign (#)
Zero (0)
A field width is specified as a decimal integer that has no leading zeros.
A field width defines a minimum byte count of a formatted value. If the formatted value does not reach the minimum byte count, so many spaces are prepended that the printed value has the specified byte count.
A precision is specified as a period (.) followed by a decimal integer. If the integer is omitted after the period, the precision is assumed to be zero.
For conversion specifications %d, %i, %u, %o, %x, and %X, a precision defines a minimum digit count. If the formatted integer does not reach the minimum digit count, so many zeros are prepended that the printed integer has the specified number of digits. The default precision is one for these conversion specifications.
For conversion specifications %e, %E, %f, and %F, a precision defines the number of digits after the decimal mark. The default precision is six for these conversion specifications.
For conversion specifications %g, and %G, a precision defines a maximum number of significant digits in the printed value. The default precision is six for these conversion specifications.
For conversion specifications %s, and %b, a precision defines a maximum byte count of the printed string. The default precision is infinity for these conversion specifications.
In the conversion specification %08.3f, the zero flag is specified, the field width is 8, and the precision is 3. If this specification is applied to value 12.34, the output will be 0012.340.
Escape sequences
The following escape sequences are recognized in format:
\a
\b
\f
\n
\r
\t
\v
\\
\"
\'
\xxx
format
values
A value is either a number or a string.
When a numeric value is required, value can be a single or double quotation followed by a character, instead of a normal number. For example, the command printf '%d' '"3' will print 51 on a typical environment where character 3 has character code 51.
The exit status of the printf built-in is zero unless there is any error.
The POSIX standard does not precisely define how multibyte characters should be handled by the built-in. When you use the %s conversion specification with precision or the %c conversion specification, you may obtain unexpected results if the formatted value contains a character that is represented by more than one byte. Yash never prints only part of the bytes that represent a single multibyte character because all multibyte characters are converted to wide characters when processed in the shell.
If the shell is not in the POSIXly-correct mode and the “long double” floating-point arithmetic is supported on the running system, then “long double” is used for floating-point conversion specifications. Otherwise, “double” is used.
The pushd built-in pushes a directory into the directory stack.
The pushd built-in changes the working directory to directory in the same manner as the cd built-in and adds it to the directory stack. If the working directory could not be changed successfully, the stack is not modified.
The pushd built-in accepts the following option as well as the options that can be used for the cd built-in:
--remove-duplicates
directory
If directory is a single hyphen (‘-’), the value of the OLDPWD variable is assumed for the new directory pathname, which is printed to the standard output.
If directory is an integer with a plus or minus sign, it is considered as an entry index of the directory stack. The entry is removed from the stack and then pushed to the stack again.
If directory is omitted, the working directory is changed to the directory specified by the --default-directory=... option. If that option is not specified either, the default is index +1.
The exit status of the pushd built-in is zero unless there is any error.
The pushd built-in is a semi-special built-in. In the POSIX standard, it is defined as a command with unspecified behavior.
The pwd built-in prints the current working directory.
The pwd built-in prints an absolute path to the shell’s current working directory to the standard output.
-L, --logical
-P, --physical
The -L (--logical) and -P (--physical) options are mutually exclusive: only the last specified one is effective. If neither is specified, -L is assumed.
The exit status of the pwd built-in is zero unless there is any error.
The pwd built-in is a semi-special built-in.
The read built-in reads a line from the standard input.
The read built-in reads a line of string from the standard input and assigns it to the specified variables.
If the -r (--raw-mode) option is specified, all characters in the line are treated literally.
If the -r (--raw-mode) option is not specified, backslashes in the line are treated as quotations. If a backslash is at the end of the line, it is treated as a line continuation. When the built-in reads the next line, the PS2 variable is used as a prompt if the shell is interactive and the standard input is a terminal.
The input line is subject to field splitting. The resulting words are assigned to variables in order. If there are more words than variables, the last variable is assigned all the remaining words (as if the words were not split). If the words are fewer than variables, the remaining variables are assigned empty strings.
-A, --array
-e, --line-editing
To use line-editing, all of the following conditions must also be met:
-P, --ps1
-p prompt, --prompt=prompt
-r, --raw-mode
variables
The exit status of the read built-in is zero unless there is any error.
Note that the exit status is non-zero if an end of input is encountered before reading the entire line.
The read built-in is a semi-special built-in.
The POSIX standard defines the -r option only: other options cannot be used in the POSIXly-correct mode.
The PS1R and PS1S variables affect the behavior of line-editing if the PS1 prompt is used. The same for PS2.
The readonly built-in makes variables and functions read-only.
The readonly built-in is equivalent to the typeset built-in with the -gr option.
The readonly built-in is a special built-in.
The POSIX standard defines the -p option only; other options cannot be used in the POSIXly-correct mode. The POSIX does not allow using the option together with operands.
The return built-in returns from a function or script.
When executed without the -n (--no-return) option, one of the following happens:
When executed with the -n (--no-return) option, the built-in does nothing but return the specified exit_status.
-n, --no-return
exit_status
The value must be a non-negative integer.
If omitted, the exit status of the last executed command is used. (But when the shell is executing a trap, the exit status of the last command before the trap is used.)
The exit status of the return built-in is defined by the exit_status operand. The exit status is used also as the exit status of the terminated function, script, or the shell.
The return built-in is a special built-in.
The POSIX standard provides that the exit_status operand should be between 0 and 255 (inclusive). Yash accepts integers larger than 255 as an extension.
In the POSIX standard, the behavior of the return built-in is defined only when the shell is executing a function or script.
The POSIX standard defines no options for the return built-in; the built-in accepts no options in the POSIXly-correct mode.
The set built-in sets shell options and positional parameters.
The set built-in requires that all options precede operands. Any command line arguments after the first operand are all treated as operands.
When executed without any command arguments, the built-in prints a list of all existing variables to the standard input in a form that can be reused as commands that will restore the variable definitions.
When -o is the only command argument, the built-in prints a list of shell options with their current settings. When +o is the only command argument, the built-in prints commands that can be reused to restore the current shell option settings.
In other cases, the built-in changes shell option settings and/or positional parameters.
When one or more options are specified, the built-in enables or disables the shell options. A normal hyphen-prefixed option enables a shell option. An option that is prefixed with a plus (+) instead of a hyphen disables a shell option. For example, options -m, -o monitor, and --monitor enable the monitor option and options +m, +o monitor, ++monitor disable it.
The name of a long option is case-insensitive and may include irrelevant non-alphanumeric characters, which are ignored. For example, options --le-comp-debug and --LeCompDebug are equivalent. If no is prepended to the name of a long option, the meaning is reversed. For example, --noallexport is equivalent to ++allexport and ++nonotify to --notify.
An option can be specified in one of the following forms:
Not all options can be specified as single-character options.
The available options are:
all-export (-a)
brace-expand
case-glob
clobber (+C)
cur-async, cur-bg, cur-stop
dot-glob
emacs
empty-last-field
err-exit (-e)
err-return
exec (+n)
extended-glob
for-local
glob (+f)
hash-on-def (-h)
hist-space
ignore-eof
le-always-rp, le-comp-debug, le-conv-meta, le-no-conv-meta, le-predict, le-predict-empty, le-prompt-sp, le-visible-bell
mark-dirs
monitor (-m)
notify (-b)
notify-le
null-glob
pipe-fail
posixly-correct
trace-all
unset (+u)
verbose (-v)
vi
x-trace (-x)
If one or more operands are passed to the set built-in, current positional parameters are all removed and the operands are set as new positional parameters. If the -- separator (cf. syntax of command arguments) is passed, the positional parameters are set even when there are no operands, in which case new positional parameters will be nothing.
The exit status of the set built-in is zero unless there is any error.
The set built-in is a special built-in.
In the POSIX standard, available shell options are much limited. The standard does not define:
The options defined in the standard are:
Yash does not support the nolog option, which prevents function definitions from being added to command history.
The shift built-in removes some positional parameters or array values.
The shift built-in removes the first count positional parameters or array values, where count is specified by the operand.
-A array, --array=array
count
It is an error if the actual number of positional parameters or array values is less than count. If omitted, the default value is one. If negative, the last -count positional parameters or array values are removed instead of the first ones.
The exit status of the shift built-in is zero unless there is any error.
The shift built-in is a special built-in.
The number of positional parameters can be obtained with the # special parameter. The number of array values can be obtained with ${array[#]}.
The POSIX standard defines no options for the shift built-in; the built-in accepts no options in the POSIXly-correct mode.
Negative operands are not allowed in the POSIXly-correct mode.
The suspend built-in suspends the shell.
The suspend built-in sends a SIGSTOP signal to all processes in the process group to which the shell process belongs. The signal suspends the processes (including the shell). The suspended processes resume when they receive a SIGCONT signal.
If the shell is interactive and its process group ID is equal to the process ID of the session leader, the shell prints a warning message and refuses to send a signal unless the -f (--force) option is specified. (In such a case, there is no other job-controlling shell that can send a SIGCONT signal to resume the suspended shell, so the shell could never be resumed.)
-f, --force
The exit status is zero if the signal was successfully sent and non-zero otherwise.
The suspend built-in is a semi-special built-in. In the POSIX standard, it is defined as a command with unspecified behavior.
The test built-in evaluates an expression.
The test built-in does not distinguish options and operands; all command line arguments are interpreted as expression. If the built-in is executed with the name [, expression must be followed by ].
The test built-in evaluates expression as a conditional expression that is made up of operators and operands described below. The exit status is 0 if the condition is true and 1 otherwise.
The unary operators below test a file. If the operand file is a symbolic link, the file referred to by the link is tested (except for the -h and -L operators).
-b file
-c file
-d file
-e file
-f file
-G file
-g file
-h file
-k file
-L file
-N file
-O file
-p file
-r file
-S file
-s file
-u file
-w file
-x file
The unary operator below tests a file descriptor:
-t fd
The unary operators below test a string:
-n string
-z string
The unary operator below tests a shell option:
-o ?option
-o option
The binary operators below compare files. Non-existing files are considered older than any existing files.
file1 -nt file2
file1 -ot file2
file1 -ef file2
The binary operators below compare strings:
string1 = string2, string1 == string2
string1 != string2
The binary operators below compare strings according to the alphabetic order in the current locale:
string1 === string2
string1 !== string2
string1 < string2
string1 <= string2
string1 > string2
string1 >= string2
The binary operator below performs pattern matching:
string =~ pattern
The binary operators below compare integers:
v1 -eq v2
v1 -ne v2
v1 -gt v2
v1 -ge v2
v1 -lt v2
v1 -le v2
The binary operators below compare version numbers:
v1 -veq v2
v1 -vne v2
v1 -vgt v2
v1 -vge v2
v1 -vlt v2
v1 -vle v2
The operators below can be used to make complex expressions:
! expression
( expression )
expression1 -a expression2
expression1 -o expression2
If the expression is a single word without operators, the -n operator is assumed. An empty expression evaluates to false.
Comparison of version numbers
Comparison of version numbers is similar to comparison of strings in alphabetic order. The differences are:
For example, version numbers 0.1.2-3 and 00.001.02-3 are equal and 0.2.1 is smaller than 0.10.0.
The exit status of the test built-in is 0 if expression is true and 1 otherwise. The exit status is 2 if expression cannot be evaluated because of a syntax error or any other reasons.
Complex expressions may cause confusion and should be avoided. Use the shell’s compound commands. For example, [ 1 -eq 1 ] && [ -t = 1 ] && ! [ foo ] is preferred over [ 1 -eq 1 -a -t = 1 -a ! foo ].
The POSIX standard provides that the exit status should be larger than 1 on error. The POSIX standard does not define the following operators: -G, -k, -N, -O, -nt, -ot, -ef, ==, ===, !==, <, <=, >, >=, =~, -veq, -vne, -vgt, -vge, -vlt, and -vle. POSIX neither specifies -o as a unary operator.
The times built-in prints CPU time usage.
The times built-in prints the CPU times consumed by the shell process and its child processes to the standard output.
The built-in prints two lines: the first line shows the CPU time of the shell process and the second one that of its child processes (not including those which have not terminated). Each line shows the CPU times consumed in the user and system mode.
The exit status of the times built-in is zero unless there is any error.
The times built-in is a special built-in.
The trap built-in sets or prints signal handlers.
The trap built-in sets or prints actions that are taken when the shell receives signals. (Those actions are called traps.)
When executed with action and one or more signals, the built-in sets the traps for signals to action. If the shell receives one of the signals, the action will be taken.
If the first operand is signal_number instead of action, the built-in resets the traps for signal_number and signals as if action was -.
When executed with the -p (--print) option or with no operands, the built-in prints currently set traps to the standard output in a format that can be executed as commands that restore the current traps. If one or more signals are specified, only those signals are printed. Otherwise, all signals with non-default actions are printed. (In some situations, however, the built-in may print previous trap settings instead of the current. See notes below.)
-p, --print
action
If action is a single hyphen (-), the action is reset to the default action that is defined by the operating system. If action is an empty string, the signal is ignored on receipt. Otherwise, action is treated as a command string: the string is parsed and executed as commands when the signal is received. (If a signal is received while a command is being executed, the action is taken just after the command finishes.)
signal
If signal is number 0 or name EXIT, it is treated as a special imaginary signal that is always received when the shell exits. The action set for this signal is taken when the shell exits normally.
signal_number
The exit status of the trap built-in is zero unless there is any error.
The trap built-in is a special built-in.
The POSIX standard defines no options for the trap built-in; the built-in accepts no options in the POSIXly-correct mode.
The POSIX standard requires that signal names must be specified without the SIG-prefix, like INT and QUIT. As an extension, yash accepts SIG-prefixed names like SIGINT and SIGQUIT and treats signal names case-insensitively.
Reusing output of the built-in
Output of the trap built-in can be saved in a variable, which can be later executed by the eval built-in to restore the traps.
saved_traps=$(trap) trap '...' INT eval "$saved_traps"
There are some tricks behind the scenes to allow this idiom. You use a command substitution to save the output of the trap built-in in the variable. The command substitution is executed in a subshell. The subshell resets all traps (except ignored ones) at the beginning of itself. This seemingly would result in (almost) empty output from the built-in that would fail to restore the traps as expected.
To avoid that pitfall, POSIX requires the shell to follow one of the two options below:
Yash obeys the second.
The true built-in does nothing successfully.
The true built-in does nothing. Any command line arguments are ignored.
The exit status of the true built-in is zero.
The true built-in is a semi-special built-in.
The true and colon built-ins have the same effect, but true is a semi-special built-in while colon is a special.
The type built-in identifies a command.
The type built-in is equivalent to the command built-in with the -V option.
The POSIX standard does not define the relation between the type and command built-ins. The standard does not define options for the type built-in.
At least one command operand must be specified in the POSIXly-correct mode.
The type built-in is a semi-special built-in.
The typeset built-in prints or sets variables or functions.
If executed without the -f (--functions) option, the typeset built-in prints or sets variables to the standard output. Otherwise, it prints or sets functions.
If executed with the -p (--print) option, the built-in prints the variables or functions specified by operands. Without the option, it sets variables or functions. If no operands are specified, it prints all existing variables or functions, regardless of whether the -p (--print) option is specified.
-f, --functions
-g, --global
When printing variables, all existing variables including global variables are printed if this option is specified. Without this option, only local variables are printed.
-p, --print
-r, --readonly
When printing variables or functions, print read-only variables or functions only.
-x, --export
When printing variables, print exported variables only.
-X, --unexport
variable (without value)
Without the -p (--print) option, the variable is defined (if not yet defined) but its value is not set nor changed. Variables that are defined without values are treated as unset in parameter expansion.
variable=value
The value is assigned to the variable (regardless of the -p (--print) option).
function
The exit status of the typeset built-in is zero unless there is any error.
A global variable cannot be newly defined if a local variable has already been defined with the same name. The local variable will be set regardless of the -g (--global) option.
The typeset built-in is a semi-special built-in. In the POSIX standard, it is defined as a command with unspecified behavior.
The export and readonly built-ins are equivalent to the typeset built-in with the -gx and -gr options, respectively. The local built-in is equivalent to the typeset built-in except that the -f (--functions) and -g (--global) options cannot be used.
The ulimit built-in sets or prints a resource limit.
The ulimit built-in sets or prints a resource limit.
If executed with the -a (--all) option, the built-in prints the current limits for all resource types. Otherwise, it sets or prints the limit of a single resource type. The resource type can be specified by the options listed below. The resource limits will affect the current shell process and all commands invoked from the shell.
Each resource type has two limit values: the hard and soft limit. You can change a soft limit freely as long as it does not exceed the hard limit. You can decrease a hard limit but cannot increase it without a proper permission.
When the -H (--hard) or -S (--soft) option is specified, the built-in sets or prints the hard or soft limit, respectively. If neither of the options is specified, the built-in sets both the hard and soft limit or prints the soft limit.
-H, --hard
-S, --soft
-a, --all
The following options specify the type of resources. If none of them is specified, -f is the default. The types of resources that can be set depend on the operating system.
-c, --core
-d, --data
-e, --nice
-f, --fsize
-i, --sigpending
-l, --memlock
-m, --rss
-n, --nofile
-q, --msgqueue
-r, --rtprio
-s, --stack
-t, --cpu
-u, --nproc
-v, --as
-x, --locks
limit
The value must be a non-negative integer or one of hard, soft, and unlimited. If value is hard or soft, the new limit is set to the current hard or soft limit. If limit is not specified, the current limit is printed.
The exit status of the ulimit built-in is zero unless there is any error.
The POSIX standard defines no options other than -f. It neither defines hard, soft, or unlimited for limit values.
The ulimit built-in is a semi-special built-in.
The umask built-in sets or prints the file mode creation mask.
If executed without the mask operand, the built-in prints the current file mode creation mask of the shell to the standard output in a form that can later be used as mask to restore the current mask.
Otherwise, the built-in sets the file mode creation mask to mask.
-S, --symbolic
mask
Octal integer form
In the octal integer form, the mask is specified as a non-negative octal integer that is the sum of the following permissions:
0400
0200
0100
0040
0020
0010
0004
0002
0001
Symbolic form
In the symbolic form, the mask is specified as a symbolic expression that denotes permissions that are not included in the mask.
The entire expression is one or more clauses separated by comma. A clause is a sequence of whos followed by one or more actions.
A who is one of:
u
g
o
a
An empty sequence of whos is equivalent to who a.
An action is an operator followed by permission. An operator is one of:
=
+
-
and permission is one of:
r
w
x
X
s
u
g
o
but more than one of r, w, x, X, and s can be specified after a single operand.
For example, the command umask u=rwx,go+r-w
The exit status of the umask built-in is zero unless there is any error.
The umask built-in is a semi-special built-in.
The POSIX standard does not require the default output format (used when the -S option is not specified) to be the octal integer form.
The unalias built-in undefines aliases.
The unalias built-in removes the definition of the aliases specified by operands.
-a, --all
name
The exit status of the unalias built-in is zero unless there is any error. It is an error to specify the name of a non-existing alias as name.
The unalias built-in is a semi-special built-in.
The unset built-in undefines variables or functions.
The unset built-in removes the definition of the variables or functions specified by operands.
It is not an error if any of the specified variables or functions do not exist; they are silently ignored.
-f, --functions
-v, --variables
These options are mutually exclusive: only the last specified one is effective. If neither is specified, -v is assumed.
name
The exit status of the unset built-in is zero unless there is any error.
The unset built-in is a special built-in.
Although yash does not do so, the POSIX standard allows removing a function if neither of the -f and -v options is specified and the specified variable does not exist.
At least one name operand must be specified in the POSIXly-correct mode.
The wait built-in waits for jobs to terminate.
The wait built-in waits for background jobs to terminate. If job control is enabled, stopped jobs are considered as terminated.
The built-in can be used to wait for asynchronous commands if job control is disabled.
If the shell receives a signal while the built-in is waiting and if a trap has been set for the signal, then the trap is executed and the built-in immediately finishes (without waiting for the jobs). If the shell receives a SIGINT signal when job control is enabled, the built-in aborts waiting.
If the shell is interactive, job-controlling, and not in the POSIXly-correct mode, the job status is printed when the job is terminated or stopped.
job
If no jobs are specified, the built-in waits for all existing jobs.
If the specified job does not exist, the job is considered to have terminated with the exit status of 127.
If no jobs were specified and the built-in successfully waited for all the jobs, the exit status is zero. If one or more jobs were specified, the exit status is that of the last job.
If the built-in was aborted by a signal, the exit status is an integer (> 128) that denotes the signal. If there was any other error, the exit status is between 1 and 126 (inclusive).
The wait built-in is a semi-special built-in.
The process ID of the last process of a job can be obtained by the ! special parameter. You can use the jobs built-in as well to obtain process IDs of job processes.
Yuki Watanabe <magicant@users.osdn.me>
| 2022-07-10 | 2.52 |