{ "content": [ { "type": "text", "text": "# zshbuiltins (info)\n\n## NAME\n\nzshbuiltins - zsh built-in commands\n\n## Sections\n\n- **NAME**\n- **SHELL BUILTIN COMMANDS**\n\nUse structuredContent.sections for detailed options, examples, and full documentation.\n" } ], "structuredContent": { "command": "zshbuiltins", "section": "", "mode": "info", "summary": "zshbuiltins - zsh built-in commands", "synopsis": null, "tldr_summary": null, "tldr_examples": [], "tldr_source": null, "flags": [], "examples": [], "see_also": [], "section_outline": [ { "name": "NAME", "lines": 2, "subsections": [] }, { "name": "SHELL BUILTIN COMMANDS", "lines": 2739, "subsections": [] } ], "sections": { "NAME": { "content": "zshbuiltins - zsh built-in commands\n", "subsections": [] }, "SHELL BUILTIN COMMANDS": { "content": "Some shell builtin commands take options as described in individual en-\ntries; these are often referred to in the list below as `flags' to\navoid confusion with shell options, which may also have an effect on\nthe behaviour of builtin commands. In this introductory section, `op-\ntion' always has the meaning of an option to a command that should be\nfamiliar to most command line users.\n\nTypically, options are single letters preceded by a hyphen (-). Op-\ntions that take an argument accept it either immediately following the\noption letter or after white space, for example `print -C3 {1..9}' or\n`print -C 3 {1..9}' are equivalent. Arguments to options are not the\nsame as arguments to the command; the documentation indicates which is\nwhich. Options that do not take an argument may be combined in a sin-\ngle word, for example `print -rca -- *' and `print -r -c -a -- *' are\nequivalent.\n\nSome shell builtin commands also take options that begin with `+' in-\nstead of `-'. The list below makes clear which commands these are.\n\nOptions (together with their individual arguments, if any) must appear\nin a group before any non-option arguments; once the first non-option\nargument has been found, option processing is terminated.\n\nAll builtin commands other than `echo' and precommand modifiers, even\nthose that have no options, can be given the argument `--' to terminate\noption processing. This indicates that the following words are non-op-\ntion arguments, but is otherwise ignored. This is useful in cases\nwhere arguments to the command may begin with `-'. For historical rea-\nsons, most builtin commands (including `echo') also recognize a single\n`-' in a separate word for this purpose; note that this is less stan-\ndard and use of `--' is recommended.\n\n- simple command\nSee the section `Precommand Modifiers' in zshmisc(1).\n\n. file [ arg ... ]\nRead commands from file and execute them in the current shell\nenvironment.\n\nIf file does not contain a slash, or if PATHDIRS is set, the\nshell looks in the components of $path to find the directory\ncontaining file. Files in the current directory are not read\nunless `.' appears somewhere in $path. If a file named\n`file.zwc' is found, is newer than file, and is the compiled\nform (created with the zcompile builtin) of file, then commands\nare read from that file instead of file.\n\nIf any arguments arg are given, they become the positional pa-\nrameters; the old positional parameters are restored when the\nfile is done executing. However, if no arguments are given, the\npositional parameters remain those of the calling context, and\nno restoring is done.\n\nIf file was not found the return status is 127; if file was\nfound but contained a syntax error the return status is 126;\nelse the return status is the exit status of the last command\nexecuted.\n\n: [ arg ... ]\nThis command does nothing, although normal argument expansions\nis performed which may have effects on shell parameters. A zero\nexit status is returned.\n\nalias [ {+|-}gmrsL ] [ name[=value] ... ]\nFor each name with a corresponding value, define an alias with\nthat value. A trailing space in value causes the next word to\nbe checked for alias expansion. If the -g flag is present, de-\nfine a global alias; global aliases are expanded even if they do\nnot occur in command position.\n\nIf the -s flag is present, define a suffix alias: if the command\nword on a command line is in the form `text.name', where text is\nany non-empty string, it is replaced by the text `value\ntext.name'. Note that name is treated as a literal string, not\na pattern. A trailing space in value is not special in this\ncase. For example,\n\nalias -s ps='gv --'\n\nwill cause the command `*.ps' to be expanded to `gv -- *.ps'.\nAs alias expansion is carried out earlier than globbing, the\n`*.ps' will then be expanded. Suffix aliases constitute a dif-\nferent name space from other aliases (so in the above example it\nis still possible to create an alias for the command ps) and the\ntwo sets are never listed together.\n\nFor each name with no value, print the value of name, if any.\nWith no arguments, print all currently defined aliases other\nthan suffix aliases. If the -m flag is given the arguments are\ntaken as patterns (they should be quoted to preserve them from\nbeing interpreted as glob patterns), and the aliases matching\nthese patterns are printed. When printing aliases and one of\nthe -g, -r or -s flags is present, restrict the printing to\nglobal, regular or suffix aliases, respectively; a regular alias\nis one which is neither a global nor a suffix alias. Using `+'\ninstead of `-', or ending the option list with a single `+',\nprevents the values of the aliases from being printed.\n\nIf the -L flag is present, then print each alias in a manner\nsuitable for putting in a startup script. The exit status is\nnonzero if a name (with no value) is given for which no alias\nhas been defined.\n\nFor more on aliases, include common problems, see the section\nALIASING in zshmisc(1).\n\nautoload [ {+|-}RTUXdkmrtWz ] [ -w ] [ name ... ]\nSee the section `Autoloading Functions' in zshmisc(1) for full\ndetails. The fpath parameter will be searched to find the func-\ntion definition when the function is first referenced.\n\nIf name consists of an absolute path, the function is defined to\nload from the file given (searching as usual for dump files in\nthe given location). The name of the function is the basename\n(non-directory part) of the file. It is normally an error if\nthe function is not found in the given location; however, if the\noption -d is given, searching for the function defaults to\n$fpath. If a function is loaded by absolute path, any functions\nloaded from it that are marked for autoload without an absolute\npath have the load path of the parent function temporarily\nprepended to $fpath.\n\nIf the option -r or -R is given, the function is searched for\nimmediately and the location is recorded internally for use when\nthe function is executed; a relative path is expanded using the\nvalue of $PWD. This protects against a change to $fpath after\nthe call to autoload. With -r, if the function is not found, it\nis silently left unresolved until execution; with -R, an error\nmessage is printed and command processing aborted immediately\nthe search fails, i.e. at the autoload command rather than at\nfunction execution..\n\nThe flag -X may be used only inside a shell function. It causes\nthe calling function to be marked for autoloading and then imme-\ndiately loaded and executed, with the current array of posi-\ntional parameters as arguments. This replaces the previous def-\ninition of the function. If no function definition is found, an\nerror is printed and the function remains undefined and marked\nfor autoloading. If an argument is given, it is used as a di-\nrectory (i.e. it does not include the name of the function) in\nwhich the function is to be found; this may be combined with the\n-d option to allow the function search to default to $fpath if\nit is not in the given location.\n\nThe flag +X attempts to load each name as an autoloaded func-\ntion, but does not execute it. The exit status is zero (suc-\ncess) if the function was not previously defined and a defini-\ntion for it was found. This does not replace any existing defi-\nnition of the function. The exit status is nonzero (failure) if\nthe function was already defined or when no definition was\nfound. In the latter case the function remains undefined and\nmarked for autoloading. If ksh-style autoloading is enabled,\nthe function created will contain the contents of the file plus\na call to the function itself appended to it, thus giving normal\nksh autoloading behaviour on the first call to the function. If\nthe -m flag is also given each name is treated as a pattern and\nall functions already marked for autoload that match the pattern\nare loaded.\n\nWith the -t flag, turn on execution tracing; with -T, turn on\nexecution tracing only for the current function, turning it off\non entry to any called functions that do not also have tracing\nenabled.\n\nWith the -U flag, alias expansion is suppressed when the func-\ntion is loaded.\n\nWith the -w flag, the names are taken as names of files compiled\nwith the zcompile builtin, and all functions defined in them are\nmarked for autoloading.\n\nThe flags -z and -k mark the function to be autoloaded using the\nzsh or ksh style, as if the option KSHAUTOLOAD were unset or\nwere set, respectively. The flags override the setting of the\noption at the time the function is loaded.\n\nNote that the autoload command makes no attempt to ensure the\nshell options set during the loading or execution of the file\nhave any particular value. For this, the emulate command can be\nused:\n\nemulate zsh -c 'autoload -Uz func'\n\narranges that when func is loaded the shell is in native zsh em-\nulation, and this emulation is also applied when func is run.\n\nSome of the functions of autoload are also provided by functions\n-u or functions -U, but autoload is a more comprehensive inter-\nface.\n\nbg [ job ... ]\njob ... &\nPut each specified job in the background, or the current job if\nnone is specified.\n\nbindkey\nSee the section `Zle Builtins' in zshzle(1).\n\nbreak [ n ]\nExit from an enclosing for, while, until, select or repeat loop.\nIf an arithmetic expression n is specified, then break n levels\ninstead of just one.\n\nbuiltin name [ args ... ]\nExecutes the builtin name, with the given args.\n\nbye Same as exit.\n\ncap See the section `The zsh/cap Module' in zshmodules(1).\n\ncd [ -qsLP ] [ arg ]\ncd [ -qsLP ] old new\ncd [ -qsLP ] {+|-}n\nChange the current directory. In the first form, change the\ncurrent directory to arg, or to the value of $HOME if arg is not\nspecified. If arg is `-', change to the previous directory.\n\nOtherwise, if arg begins with a slash, attempt to change to the\ndirectory given by arg.\n\nIf arg does not begin with a slash, the behaviour depends on\nwhether the current directory `.' occurs in the list of directo-\nries contained in the shell parameter cdpath. If it does not,\nfirst attempt to change to the directory arg under the current\ndirectory, and if that fails but cdpath is set and contains at\nleast one element attempt to change to the directory arg under\neach component of cdpath in turn until successful. If `.' oc-\ncurs in cdpath, then cdpath is searched strictly in order so\nthat `.' is only tried at the appropriate point.\n\nThe order of testing cdpath is modified if the option POSIXCD\nis set, as described in the documentation for the option.\n\nIf no directory is found, the option CDABLEVARS is set, and a\nparameter named arg exists whose value begins with a slash,\ntreat its value as the directory. In that case, the parameter\nis added to the named directory hash table.\n\nThe second form of cd substitutes the string new for the string\nold in the name of the current directory, and tries to change to\nthis new directory.\n\nThe third form of cd extracts an entry from the directory stack,\nand changes to that directory. An argument of the form `+n'\nidentifies a stack entry by counting from the left of the list\nshown by the dirs command, starting with zero. An argument of\nthe form `-n' counts from the right. If the PUSHDMINUS option\nis set, the meanings of `+' and `-' in this context are swapped.\nIf the POSIXCD option is set, this form of cd is not recognised\nand will be interpreted as the first form.\n\nIf the -q (quiet) option is specified, the hook function chpwd\nand the functions in the array chpwdfunctions are not called.\nThis is useful for calls to cd that do not change the environ-\nment seen by an interactive user.\n\nIf the -s option is specified, cd refuses to change the current\ndirectory if the given pathname contains symlinks. If the -P\noption is given or the CHASELINKS option is set, symbolic links\nare resolved to their true values. If the -L option is given\nsymbolic links are retained in the directory (and not resolved)\nregardless of the state of the CHASELINKS option.\n\nchdir Same as cd.\n\nclone See the section `The zsh/clone Module' in zshmodules(1).\n\ncommand [ -pvV ] simple command\nThe simple command argument is taken as an external command in-\nstead of a function or builtin and is executed. If the\nPOSIXBUILTINS option is set, builtins will also be executed but\ncertain special properties of them are suppressed. The -p flag\ncauses a default path to be searched instead of that in $path.\nWith the -v flag, command is similar to whence and with -V, it\nis equivalent to whence -v.\n\nSee also the section `Precommand Modifiers' in zshmisc(1).\n\ncomparguments\nSee the section `The zsh/computil Module' in zshmodules(1).\n\ncompcall\nSee the section `The zsh/compctl Module' in zshmodules(1).\n\ncompctl\nSee the section `The zsh/compctl Module' in zshmodules(1).\n\ncompdescribe\nSee the section `The zsh/computil Module' in zshmodules(1).\n\ncompfiles\nSee the section `The zsh/computil Module' in zshmodules(1).\n\ncompgroups\nSee the section `The zsh/computil Module' in zshmodules(1).\n\ncompquote\nSee the section `The zsh/computil Module' in zshmodules(1).\n\ncomptags\nSee the section `The zsh/computil Module' in zshmodules(1).\n\ncomptry\nSee the section `The zsh/computil Module' in zshmodules(1).\n\ncompvalues\nSee the section `The zsh/computil Module' in zshmodules(1).\n\ncontinue [ n ]\nResume the next iteration of the enclosing for, while, until,\nselect or repeat loop. If an arithmetic expression n is speci-\nfied, break out of n-1 loops and resume at the nth enclosing\nloop.\n\ndeclare\nSame as typeset.\n\ndirs [ -c ] [ arg ... ]\ndirs [ -lpv ]\nWith no arguments, print the contents of the directory stack.\nDirectories are added to this stack with the pushd command, and\nremoved with the cd or popd commands. If arguments are speci-\nfied, load them onto the directory stack, replacing anything\nthat was there, and push the current directory onto the stack.\n\n-c clear the directory stack.\n\n-l print directory names in full instead of using of using ~\nexpressions (see Dynamic and Static named directories in\nzshexpn(1)).\n\n-p print directory entries one per line.\n\n-v number the directories in the stack when printing.\n\ndisable [ -afmprs ] name ...\nTemporarily disable the named hash table elements or patterns.\nThe default is to disable builtin commands. This allows you to\nuse an external command with the same name as a builtin command.\nThe -a option causes disable to act on regular or global\naliases. The -s option causes disable to act on suffix aliases.\nThe -f option causes disable to act on shell functions. The -r\noptions causes disable to act on reserved words. Without argu-\nments all disabled hash table elements from the corresponding\nhash table are printed. With the -m flag the arguments are\ntaken as patterns (which should be quoted to prevent them from\nundergoing filename expansion), and all hash table elements from\nthe corresponding hash table matching these patterns are dis-\nabled. Disabled objects can be enabled with the enable command.\n\nWith the option -p, name ... refer to elements of the shell's\npattern syntax as described in the section `Filename Genera-\ntion'. Certain elements can be disabled separately, as given\nbelow.\n\nNote that patterns not allowed by the current settings for the\noptions EXTENDEDGLOB, KSHGLOB and SHGLOB are never enabled,\nregardless of the setting here. For example, if EXTENDEDGLOB\nis not active, the pattern ^ is ineffective even if `disable -p\n\"^\"' has not been issued. The list below indicates any option\nsettings that restrict the use of the pattern. It should be\nnoted that setting SHGLOB has a wider effect than merely dis-\nabling patterns as certain expressions, in particular those in-\nvolving parentheses, are parsed differently.\n\nThe following patterns may be disabled; all the strings need\nquoting on the command line to prevent them from being inter-\npreted immediately as patterns and the patterns are shown below\nin single quotes as a reminder.\n\n'?' The pattern character ? wherever it occurs, including\nwhen preceding a parenthesis with KSHGLOB.\n\n'*' The pattern character * wherever it occurs, including re-\ncursive globbing and when preceding a parenthesis with\nKSHGLOB.\n\n'[' Character classes.\n\n'<' (NOSHGLOB)\nNumeric ranges.\n\n'|' (NOSHGLOB)\nAlternation in grouped patterns, case statements, or\nKSHGLOB parenthesised expressions.\n\n'(' (NOSHGLOB)\nGrouping using single parentheses. Disabling this does\nnot disable the use of parentheses for KSHGLOB where\nthey are introduced by a special character, nor for glob\nqualifiers (use `setopt NOBAREGLOBQUAL' to disable\nglob qualifiers that use parentheses only).\n\n'~' (EXTENDEDGLOB)\nExclusion in the form A~B.\n\n'^' (EXTENDEDGLOB)\nExclusion in the form A^B.\n\n'#' (EXTENDEDGLOB)\nThe pattern character # wherever it occurs, both for rep-\netition of a previous pattern and for indicating globbing\nflags.\n\n'?(' (KSHGLOB)\nThe grouping form ?(...). Note this is also disabled if\n'?' is disabled.\n\n'*(' (KSHGLOB)\nThe grouping form *(...). Note this is also disabled if\n'*' is disabled.\n\n'+(' (KSHGLOB)\nThe grouping form +(...).\n\n'!(' (KSHGLOB)\nThe grouping form !(...).\n\n'@(' (KSHGLOB)\nThe grouping form @(...).\n\ndisown [ job ... ]\njob ... &|\njob ... &!\nRemove the specified jobs from the job table; the shell will no\nlonger report their status, and will not complain if you try to\nexit an interactive shell with them running or stopped. If no\njob is specified, disown the current job.\n\nIf the jobs are currently stopped and the AUTOCONTINUE option\nis not set, a warning is printed containing information about\nhow to make them running after they have been disowned. If one\nof the latter two forms is used, the jobs will automatically be\nmade running, independent of the setting of the AUTOCONTINUE\noption.\n\necho [ -neE ] [ arg ... ]\nWrite each arg on the standard output, with a space separating\neach one. If the -n flag is not present, print a newline at the\nend. echo recognizes the following escape sequences:\n\n\\a bell character\n\\b backspace\n\\c suppress subsequent characters and final newline\n\\e escape\n\\f form feed\n\\n linefeed (newline)\n\\r carriage return\n\\t horizontal tab\n\\v vertical tab\n\\\\ backslash\n\\0NNN character code in octal\n\\xNN character code in hexadecimal\n\\uNNNN unicode character code in hexadecimal\n\\UNNNNNNNN\nunicode character code in hexadecimal\n\nThe -E flag, or the BSDECHO option, can be used to disable\nthese escape sequences. In the latter case, -e flag can be used\nto enable them.\n\nNote that for standards compliance a double dash does not termi-\nnate option processing; instead, it is printed directly. How-\never, a single dash does terminate option processing, so the\nfirst dash, possibly following options, is not printed, but ev-\nerything following it is printed as an argument. The single\ndash behaviour is different from other shells. For a more por-\ntable way of printing text, see printf, and for a more control-\nlable way of printing text within zsh, see print.\n\nechotc See the section `The zsh/termcap Module' in zshmodules(1).\n\nechoti See the section `The zsh/terminfo Module' in zshmodules(1).\n\nemulate [ -lLR ] [ {zsh|sh|ksh|csh} [ flags ... ] ]\nWithout any argument print current emulation mode.\n\nWith single argument set up zsh options to emulate the specified\nshell as much as possible. csh will never be fully emulated.\nIf the argument is not one of the shells listed above, zsh will\nbe used as a default; more precisely, the tests performed on the\nargument are the same as those used to determine the emulation\nat startup based on the shell name, see the section COMPATIBIL-\nITY in zsh(1) . In addition to setting shell options, the com-\nmand also restores the pristine state of pattern enables, as if\nall patterns had been enabled using enable -p.\n\nIf the emulate command occurs inside a function that has been\nmarked for execution tracing with functions -t then the xtrace\noption will be turned on regardless of emulation mode or other\noptions. Note that code executed inside the function by the .,\nsource, or eval commands is not considered to be running di-\nrectly from the function, hence does not provoke this behaviour.\n\nIf the -R switch is given, all settable options are reset to\ntheir default value corresponding to the specified emulation\nmode, except for certain options describing the interactive en-\nvironment; otherwise, only those options likely to cause porta-\nbility problems in scripts and functions are altered. If the -L\nswitch is given, the options LOCALOPTIONS, LOCALPATTERNS and\nLOCALTRAPS will be set as well, causing the effects of the emu-\nlate command and any setopt, disable -p or enable -p, and trap\ncommands to be local to the immediately surrounding shell func-\ntion, if any; normally these options are turned off in all emu-\nlation modes except ksh. The -L switch is mutually exclusive\nwith the use of -c in flags.\n\nIf there is a single argument and the -l switch is given, the\noptions that would be set or unset (the latter indicated with\nthe prefix `no') are listed. -l can be combined with -L or -R\nand the list will be modified in the appropriate way. Note the\nlist does not depend on the current setting of options, i.e. it\nincludes all options that may in principle change, not just\nthose that would actually change.\n\nThe flags may be any of the invocation-time flags described in\nthe section INVOCATION in zsh(1), except that `-o EMACS' and `-o\nVI' may not be used. Flags such as `+r'/`+o RESTRICTED' may be\nprohibited in some circumstances.\n\nIf -c arg appears in flags, arg is evaluated while the requested\nemulation is temporarily in effect. In this case the emulation\nmode and all options are restored to their previous values be-\nfore emulate returns. The -R switch may precede the name of the\nshell to emulate; note this has a meaning distinct from includ-\ning -R in flags.\n\nUse of -c enables `sticky' emulation mode for functions defined\nwithin the evaluated expression: the emulation mode is associ-\nated thereafter with the function so that whenever the function\nis executed the emulation (respecting the -R switch, if present)\nand all options are set (and pattern disables cleared) before\nentry to the function, and the state is restored after exit. If\nthe function is called when the sticky emulation is already in\neffect, either within an `emulate shell -c' expression or within\nanother function with the same sticky emulation, entry and exit\nfrom the function do not cause options to be altered (except due\nto standard processing such as the LOCALOPTIONS option). This\nalso applies to functions marked for autoload within the sticky\nemulation; the appropriate set of options will be applied at the\npoint the function is loaded as well as when it is run.\n\nFor example:\n\nemulate sh -c 'fni() { setopt cshnullglob; }\nfno() { fni; }'\nfno\n\nThe two functions fni and fno are defined with sticky sh emula-\ntion. fno is then executed, causing options associated with em-\nulations to be set to their values in sh. fno then calls fni;\nbecause fni is also marked for sticky sh emulation, no option\nchanges take place on entry to or exit from it. Hence the op-\ntion cshnullglob, turned off by sh emulation, will be turned on\nwithin fni and remain on return to fno. On exit from fno, the\nemulation mode and all options will be restored to the state\nthey were in before entry to the temporary emulation.\n\nThe documentation above is typically sufficient for the intended\npurpose of executing code designed for other shells in a suit-\nable environment. More detailed rules follow.\n1. The sticky emulation environment provided by `emulate\nshell -c' is identical to that provided by entry to a\nfunction marked for sticky emulation as a consequence of\nbeing defined in such an environment. Hence, for exam-\nple, the sticky emulation is inherited by subfunctions\ndefined within functions with sticky emulation.\n2. No change of options takes place on entry to or exit from\nfunctions that are not marked for sticky emulation, other\nthan those that would normally take place, even if those\nfunctions are called within sticky emulation.\n3. No special handling is provided for functions marked for\nautoload nor for functions present in wordcode created by\nthe zcompile command.\n4. The presence or absence of the -R switch to emulate cor-\nresponds to different sticky emulation modes, so for ex-\nample `emulate sh -c', `emulate -R sh -c' and `emulate\ncsh -c' are treated as three distinct sticky emulations.\n5. Difference in shell options supplied in addition to the\nbasic emulation also mean the sticky emulations are dif-\nferent, so for example `emulate zsh -c' and `emulate zsh\n-o cbases -c' are treated as distinct sticky emulations.\n\nenable [ -afmprs ] name ...\nEnable the named hash table elements, presumably disabled ear-\nlier with disable. The default is to enable builtin commands.\nThe -a option causes enable to act on regular or global aliases.\nThe -s option causes enable to act on suffix aliases. The -f\noption causes enable to act on shell functions. The -r option\ncauses enable to act on reserved words. Without arguments all\nenabled hash table elements from the corresponding hash table\nare printed. With the -m flag the arguments are taken as pat-\nterns (should be quoted) and all hash table elements from the\ncorresponding hash table matching these patterns are enabled.\nEnabled objects can be disabled with the disable builtin com-\nmand.\n\nenable -p reenables patterns disabled with disable -p. Note\nthat it does not override globbing options; for example, `enable\n-p \"~\"' does not cause the pattern character ~ to be active un-\nless the EXTENDEDGLOB option is also set. To enable all possi-\nble patterns (so that they may be individually disabled with\ndisable -p), use `setopt EXTENDEDGLOB KSHGLOB NOSHGLOB'.\n\neval [ arg ... ]\nRead the arguments as input to the shell and execute the result-\ning command(s) in the current shell process. The return status\nis the same as if the commands had been executed directly by the\nshell; if there are no args or they contain no commands (i.e.\nare an empty string or whitespace) the return status is zero.\n\nexec [ -cl ] [ -a argv0 ] [ command [ arg ... ] ]\nReplace the current shell with command rather than forking. If\ncommand is a shell builtin command or a shell function, the\nshell executes it, and exits when the command is complete.\n\nWith -c clear the environment; with -l prepend - to the argv[0]\nstring of the command executed (to simulate a login shell); with\n-a argv0 set the argv[0] string of the command executed. See\nthe section `Precommand Modifiers' in zshmisc(1).\n\nIf the option POSIXBUILTINS is set, command is never inter-\npreted as a shell builtin command or shell function. This means\nfurther precommand modifiers such as builtin and noglob are also\nnot interpreted within the shell. Hence command is always found\nby searching the command path.\n\nIf command is omitted but any redirections are specified, then\nthe redirections will take effect in the current shell.\n\nexit [ n ]\nExit the shell with the exit status specified by an arithmetic\nexpression n; if none is specified, use the exit status from the\nlast command executed. An EOF condition will also cause the\nshell to exit, unless the IGNOREEOF option is set.\n\nSee notes at the end of the section JOBS in zshmisc(1) for some\npossibly unexpected interactions of the exit command with jobs.\n\nexport [ name[=value] ... ]\nThe specified names are marked for automatic export to the envi-\nronment of subsequently executed commands. Equivalent to type-\nset -gx. If a parameter specified does not already exist, it is\ncreated in the global scope.\n\nfalse [ arg ... ]\nDo nothing and return an exit status of 1.\n\nfc [ -e ename ] [ -LI ] [ -m match ] [ old=new ... ] [ first [ last ] ]\nfc -l [ -LI ] [ -nrdfEiD ] [ -t timefmt ] [ -m match ]\n[ old=new ... ] [ first [ last ] ]\nfc -p [ -a ] [ filename [ histsize [ savehistsize ] ] ]\nfc -P\nfc -ARWI [ filename ]\nThe fc command controls the interactive history mechanism. Note\nthat reading and writing of history options is only performed if\nthe shell is interactive. Usually this is detected automati-\ncally, but it can be forced by setting the interactive option\nwhen starting the shell.\n\nThe first two forms of this command select a range of events\nfrom first to last from the history list. The arguments first\nand last may be specified as a number or as a string. A nega-\ntive number is used as an offset to the current history event\nnumber. A string specifies the most recent event beginning with\nthe given string. All substitutions old=new, if any, are then\nperformed on the text of the events.\n\nIn addition to the number range,\n-I restricts to only internal events (not from $HISTFILE)\n-L restricts to only local events (not from other shells,\nsee SHAREHISTORY in zshoptions(1) -- note that $HISTFILE\nis considered local when read at startup)\n-m takes the first argument as a pattern (should be quoted)\nand only the history events matching this pattern are\nconsidered\n\nIf first is not specified, it will be set to -1 (the most recent\nevent), or to -16 if the -l flag is given. If last is not spec-\nified, it will be set to first, or to -1 if the -l flag is\ngiven. However, if the current event has added entries to the\nhistory with `print -s' or `fc -R', then the default last for -l\nincludes all new history entries since the current event began.\n\nWhen the -l flag is given, the resulting events are listed on\nstandard output. Otherwise the editor program specified by -e\nename is invoked on a file containing these history events. If\n-e is not given, the value of the parameter FCEDIT is used; if\nthat is not set the value of the parameter EDITOR is used; if\nthat is not set a builtin default, usually `vi' is used. If\nename is `-', no editor is invoked. When editing is complete,\nthe edited command is executed.\n\nThe flag -r reverses the order of the events and the flag -n\nsuppresses event numbers when listing.\n\nAlso when listing,\n-d prints timestamps for each event\n-f prints full time-date stamps in the US `MM/DD/YY hh:mm'\nformat\n-E prints full time-date stamps in the European `dd.mm.yyyy\nhh:mm' format\n-i prints full time-date stamps in ISO8601 `yyyy-mm-dd\nhh:mm' format\n-t fmt prints time and date stamps in the given format; fmt is\nformatted with the strftime function with the zsh exten-\nsions described for the %D{string} prompt format in the\nsection EXPANSION OF PROMPT SEQUENCES in zshmisc(1). The\nresulting formatted string must be no more than 256 char-\nacters or will not be printed\n-D prints elapsed times; may be combined with one of the op-\ntions above\n\n`fc -p' pushes the current history list onto a stack and\nswitches to a new history list. If the -a option is also speci-\nfied, this history list will be automatically popped when the\ncurrent function scope is exited, which is a much better solu-\ntion than creating a trap function to call `fc -P' manually. If\nno arguments are specified, the history list is left empty,\n$HISTFILE is unset, and $HISTSIZE & $SAVEHIST are set to their\ndefault values. If one argument is given, $HISTFILE is set to\nthat filename, $HISTSIZE & $SAVEHIST are left unchanged, and the\nhistory file is read in (if it exists) to initialize the new\nlist. If a second argument is specified, $HISTSIZE & $SAVEHIST\nare instead set to the single specified numeric value. Finally,\nif a third argument is specified, $SAVEHIST is set to a separate\nvalue from $HISTSIZE. You are free to change these environment\nvalues for the new history list however you desire in order to\nmanipulate the new history list.\n\n`fc -P' pops the history list back to an older list saved by `fc\n-p'. The current list is saved to its $HISTFILE before it is\ndestroyed (assuming that $HISTFILE and $SAVEHIST are set appro-\npriately, of course). The values of $HISTFILE, $HISTSIZE, and\n$SAVEHIST are restored to the values they had when `fc -p' was\ncalled. Note that this restoration can conflict with making\nthese variables \"local\", so your best bet is to avoid local dec-\nlarations for these variables in functions that use `fc -p'.\nThe one other guaranteed-safe combination is declaring these\nvariables to be local at the top of your function and using the\nautomatic option (-a) with `fc -p'. Finally, note that it is\nlegal to manually pop a push marked for automatic popping if you\nneed to do so before the function exits.\n\n`fc -R' reads the history from the given file, `fc -W' writes\nthe history out to the given file, and `fc -A' appends the his-\ntory out to the given file. If no filename is specified, the\n$HISTFILE is assumed. If the -I option is added to -R, only\nthose events that are not already contained within the internal\nhistory list are added. If the -I option is added to -A or -W,\nonly those events that are new since last incremental ap-\npend/write to the history file are appended/written. In any\ncase, the created file will have no more than $SAVEHIST entries.\n\nfg [ job ... ]\njob ...\nBring each specified job in turn to the foreground. If no job\nis specified, resume the current job.\n\nfloat [ {+|-}Hghlprtux ] [ {+|-}EFLRZ [ n ] ] [ name[=value] ... ]\nEquivalent to typeset -E, except that options irrelevant to\nfloating point numbers are not permitted.\n\nfunctions [ {+|-}UkmtTuWz ] [ -x num ] [ name ... ]\nfunctions -c oldfn newfn\nfunctions -M [-s] mathfn [ min [ max [ shellfn ] ] ]\nfunctions -M [ -m pattern ... ]\nfunctions +M [ -m ] mathfn ...\nEquivalent to typeset -f, with the exception of the -c, -x, -M\nand -W options. For functions -u and functions -U, see au-\ntoload, which provides additional options.\n\nThe -x option indicates that any functions output will have each\nleading tab for indentation, added by the shell to show syntac-\ntic structure, expanded to the given number num of spaces. num\ncan also be 0 to suppress all indentation.\n\nThe -W option turns on the option WARNNESTEDVAR for the named\nfunction or functions only. The option is turned off at the\nstart of nested functions (apart from anonoymous functions) un-\nless the called function also has the -W attribute.\n\nThe -c option causes oldfn to be copied to newfn. The copy is\nefficiently handled internally by reference counting. If oldfn\nwas marked for autoload it is first loaded and if this fails the\ncopy fails. Either function may subsequently be redefined with-\nout affecting the other. A typical idiom is that oldfn is the\nname of a library shell function which is then redefined to call\nnewfn, thereby installing a modified version of the function.\n\nUse of the -M option may not be combined with any of the options\nhandled by typeset -f.\n\nfunctions -M mathfn defines mathfn as the name of a mathematical\nfunction recognised in all forms of arithmetical expressions;\nsee the section `Arithmetic Evaluation' in zshmisc(1). By de-\nfault mathfn may take any number of comma-separated arguments.\nIf min is given, it must have exactly min args; if min and max\nare both given, it must have at least min and at most max args.\nmax may be -1 to indicate that there is no upper limit.\n\nBy default the function is implemented by a shell function of\nthe same name; if shellfn is specified it gives the name of the\ncorresponding shell function while mathfn remains the name used\nin arithmetical expressions. The name of the function in $0 is\nmathfn (not shellfn as would usually be the case), provided the\noption FUNCTIONARGZERO is in effect. The positional parameters\nin the shell function correspond to the arguments of the mathe-\nmatical function call. The result of the last arithmetical ex-\npression evaluated inside the shell function (even if it is a\nform that normally only returns a status) gives the result of\nthe mathematical function.\n\nIf the additional option -s is given to functions -M, the argu-\nment to the function is a single string: anything between the\nopening and matching closing parenthesis is passed to the func-\ntion as a single argument, even if it includes commas or white\nspace. The minimum and maximum argument specifiers must there-\nfore be 1 if given. An empty argument list is passed as a\nzero-length string.\n\nfunctions -M with no arguments lists all such user-defined func-\ntions in the same form as a definition. With the additional op-\ntion -m and a list of arguments, all functions whose mathfn\nmatches one of the pattern arguments are listed.\n\nfunction +M removes the list of mathematical functions; with the\nadditional option -m the arguments are treated as patterns and\nall functions whose mathfn matches the pattern are removed.\nNote that the shell function implementing the behaviour is not\nremoved (regardless of whether its name coincides with mathfn).\n\nFor example, the following prints the cube of 3:\n\nzmathcube() { (( $1 * $1 * $1 )) }\nfunctions -M cube 1 1 zmathcube\nprint $(( cube(3) ))\n\nThe following string function takes a single argument, including\nthe commas, so prints 11:\n\nstringfn() { (( $#1 )) }\nfunctions -Ms stringfn\nprint $(( stringfn(foo,bar,rod) ))\n\ngetcap See the section `The zsh/cap Module' in zshmodules(1).\n\ngetln [ -AclneE ] name ...\nRead the top value from the buffer stack and put it in the shell\nparameter name. Equivalent to read -zr.\n\ngetopts optstring name [ arg ... ]\nChecks the args for legal options. If the args are omitted, use\nthe positional parameters. A valid option argument begins with\na `+' or a `-'. An argument not beginning with a `+' or a `-',\nor the argument `--', ends the options. Note that a single `-'\nis not considered a valid option argument. optstring contains\nthe letters that getopts recognizes. If a letter is followed by\na `:', that option requires an argument. The options can be\nseparated from the argument by blanks.\n\nEach time it is invoked, getopts places the option letter it\nfinds in the shell parameter name, prepended with a `+' when arg\nbegins with a `+'. The index of the next arg is stored in\nOPTIND. The option argument, if any, is stored in OPTARG.\n\nThe first option to be examined may be changed by explicitly as-\nsigning to OPTIND. OPTIND has an initial value of 1, and is\nnormally set to 1 upon entry to a shell function and restored\nupon exit (this is disabled by the POSIXBUILTINS option). OP-\nTARG is not reset and retains its value from the most recent\ncall to getopts. If either of OPTIND or OPTARG is explicitly\nunset, it remains unset, and the index or option argument is not\nstored. The option itself is still stored in name in this case.\n\nA leading `:' in optstring causes getopts to store the letter of\nany invalid option in OPTARG, and to set name to `?' for an un-\nknown option and to `:' when a required argument is missing.\nOtherwise, getopts sets name to `?' and prints an error message\nwhen an option is invalid. The exit status is nonzero when\nthere are no more options.\n\nhash [ -Ldfmrv ] [ name[=value] ] ...\nhash can be used to directly modify the contents of the command\nhash table, and the named directory hash table. Normally one\nwould modify these tables by modifying one's PATH (for the com-\nmand hash table) or by creating appropriate shell parameters\n(for the named directory hash table). The choice of hash table\nto work on is determined by the -d option; without the option\nthe command hash table is used, and with the option the named\ndirectory hash table is used.\n\nA command name starting with a / is never hashed, whether by ex-\nplicit use of the hash command or otherwise. Such a command is\nalways found by direct look up in the file system.\n\nGiven no arguments, and neither the -r or -f options, the se-\nlected hash table will be listed in full.\n\nThe -r option causes the selected hash table to be emptied. It\nwill be subsequently rebuilt in the normal fashion. The -f op-\ntion causes the selected hash table to be fully rebuilt immedi-\nately. For the command hash table this hashes all the absolute\ndirectories in the PATH, and for the named directory hash table\nthis adds all users' home directories. These two options cannot\nbe used with any arguments.\n\nThe -m option causes the arguments to be taken as patterns\n(which should be quoted) and the elements of the hash table\nmatching those patterns are printed. This is the only way to\ndisplay a limited selection of hash table elements.\n\nFor each name with a corresponding value, put `name' in the se-\nlected hash table, associating it with the pathname `value'. In\nthe command hash table, this means that whenever `name' is used\nas a command argument, the shell will try to execute the file\ngiven by `value'. In the named directory hash table, this means\nthat `value' may be referred to as `~name'.\n\nFor each name with no corresponding value, attempt to add name\nto the hash table, checking what the appropriate value is in the\nnormal manner for that hash table. If an appropriate value\ncan't be found, then the hash table will be unchanged.\n\nThe -v option causes hash table entries to be listed as they are\nadded by explicit specification. If has no effect if used with\n-f.\n\nIf the -L flag is present, then each hash table entry is printed\nin the form of a call to hash.\n\nhistory\nSame as fc -l.\n\ninteger [ {+|-}Hghlprtux ] [ {+|-}LRZi [ n ] ] [ name[=value] ... ]\nEquivalent to typeset -i, except that options irrelevant to in-\ntegers are not permitted.\n\njobs [ -dlprs ] [ job ... ]\njobs -Z string\nLists information about each given job, or all jobs if job is\nomitted. The -l flag lists process IDs, and the -p flag lists\nprocess groups. If the -r flag is specified only running jobs\nwill be listed and if the -s flag is given only stopped jobs are\nshown. If the -d flag is given, the directory from which the\njob was started (which may not be the current directory of the\njob) will also be shown.\n\nThe -Z option replaces the shell's argument and environment\nspace with the given string, truncated if necessary to fit.\nThis will normally be visible in ps (ps(1)) listings. This fea-\nture is typically used by daemons, to indicate their state.\n\nkill [ -s signalname | -n signalnumber | -sig ] job ...\nkill -l [ sig ... ]\nSends either SIGTERM or the specified signal to the given jobs\nor processes. Signals are given by number or by names, with or\nwithout the `SIG' prefix. If the signal being sent is not\n`KILL' or `CONT', then the job will be sent a `CONT' signal if\nit is stopped. The argument job can be the process ID of a job\nnot in the job list. In the second form, kill -l, if sig is not\nspecified the signal names are listed. Otherwise, for each sig\nthat is a name, the corresponding signal number is listed. For\neach sig that is a signal number or a number representing the\nexit status of a process which was terminated or stopped by a\nsignal the name of the signal is printed.\n\nOn some systems, alternative signal names are allowed for a few\nsignals. Typical examples are SIGCHLD and SIGCLD or SIGPOLL and\nSIGIO, assuming they correspond to the same signal number. kill\n-l will only list the preferred form, however kill -l alt will\nshow if the alternative form corresponds to a signal number.\nFor example, under Linux kill -l IO and kill -l POLL both output\n29, hence kill -IO and kill -POLL have the same effect.\n\nMany systems will allow process IDs to be negative to kill a\nprocess group or zero to kill the current process group.\n\nlet arg ...\nEvaluate each arg as an arithmetic expression. See the section\n`Arithmetic Evaluation' in zshmisc(1) for a description of\narithmetic expressions. The exit status is 0 if the value of\nthe last expression is nonzero, 1 if it is zero, and 2 if an er-\nror occurred.\n\nlimit [ -hs ] [ resource [ limit ] ] ...\nSet or display resource limits. Unless the -s flag is given,\nthe limit applies only the children of the shell. If -s is\ngiven without other arguments, the resource limits of the cur-\nrent shell is set to the previously set resource limits of the\nchildren.\n\nIf limit is not specified, print the current limit placed on re-\nsource, otherwise set the limit to the specified value. If the\n-h flag is given, use hard limits instead of soft limits. If no\nresource is given, print all limits.\n\nWhen looping over multiple resources, the shell will abort imme-\ndiately if it detects a badly formed argument. However, if it\nfails to set a limit for some other reason it will continue try-\ning to set the remaining limits.\n\nresource can be one of:\n\naddressspace\nMaximum amount of address space used.\naiomemorylocked\nMaximum amount of memory locked in RAM for AIO opera-\ntions.\naiooperations\nMaximum number of AIO operations.\ncachedthreads\nMaximum number of cached threads.\ncoredumpsize\nMaximum size of a core dump.\ncputime\nMaximum CPU seconds per process.\ndatasize\nMaximum data size (including stack) for each process.\ndescriptors\nMaximum value for a file descriptor.\nfilesize\nLargest single file allowed.\nkqueues\nMaximum number of kqueues allocated.\nmaxproc\nMaximum number of processes.\nmaxpthreads\nMaximum number of threads per process.\nmemorylocked\nMaximum amount of memory locked in RAM.\nmemoryuse\nMaximum resident set size.\nmsgqueue\nMaximum number of bytes in POSIX message queues.\nposixlocks\nMaximum number of POSIX locks per user.\npseudoterminals\nMaximum number of pseudo-terminals.\nresident\nMaximum resident set size.\nsigpending\nMaximum number of pending signals.\nsockbufsize\nMaximum size of all socket buffers.\nstacksize\nMaximum stack size for each process.\nswapsize\nMaximum amount of swap used.\nvmemorysize\nMaximum amount of virtual memory.\n\nWhich of these resource limits are available depends on the sys-\ntem. resource can be abbreviated to any unambiguous prefix. It\ncan also be an integer, which corresponds to the integer defined\nfor the resource by the operating system.\n\nIf argument corresponds to a number which is out of the range of\nthe resources configured into the shell, the shell will try to\nread or write the limit anyway, and will report an error if this\nfails. As the shell does not store such resources internally,\nan attempt to set the limit will fail unless the -s option is\npresent.\n\nlimit is a number, with an optional scaling factor, as follows:\n\nnh hours\nnk kilobytes (default)\nnm megabytes or minutes\nng gigabytes\n[mm:]ss\nminutes and seconds\n\nThe limit command is not made available by default when the\nshell starts in a mode emulating another shell. It can be made\navailable with the command `zmodload -F zsh/rlimits b:limit'.\n\nlocal [ {+|-}AHUahlprtux ] [ {+|-}EFLRZi [ n ] ] [ name[=value] ... ]\nSame as typeset, except that the options -g, and -f are not per-\nmitted. In this case the -x option does not force the use of\n-g, i.e. exported variables will be local to functions.\n\nlog List all users currently logged in who are affected by the cur-\nrent setting of the watch parameter.\n\nlogout [ n ]\nSame as exit, except that it only works in a login shell.\n\nnoglob simple command\nSee the section `Precommand Modifiers' in zshmisc(1).\n\npopd [ -q ] [ {+|-}n ]\nRemove an entry from the directory stack, and perform a cd to\nthe new top directory. With no argument, the current top entry\nis removed. An argument of the form `+n' identifies a stack en-\ntry by counting from the left of the list shown by the dirs com-\nmand, starting with zero. An argument of the form -n counts\nfrom the right. If the PUSHDMINUS option is set, the meanings\nof `+' and `-' in this context are swapped.\n\nIf the -q (quiet) option is specified, the hook function chpwd\nand the functions in the array $chpwdfunctions are not called,\nand the new directory stack is not printed. This is useful for\ncalls to popd that do not change the environment seen by an in-\nteractive user.\n\nprint [ -abcDilmnNoOpPrsSz ] [ -u n ] [ -f format ] [ -C cols ]\n[ -v name ] [ -xX tabstop ] [ -R [ -en ]] [ arg ... ]\nWith the `-f' option the arguments are printed as described by\nprintf. With no flags or with the flag `-', the arguments are\nprinted on the standard output as described by echo, with the\nfollowing differences: the escape sequence `\\M-x' (or `\\Mx')\nmetafies the character x (sets the highest bit), `\\C-x' (or\n`\\Cx') produces a control character (`\\C-@' and `\\C-?' give the\ncharacters NULL and delete), a character code in octal is repre-\nsented by `\\NNN' (instead of `\\0NNN'), and `\\E' is a synonym for\n`\\e'. Finally, if not in an escape sequence, `\\' escapes the\nfollowing character and is not printed.\n\n-a Print arguments with the column incrementing first. Only\nuseful with the -c and -C options.\n\n-b Recognize all the escape sequences defined for the bind-\nkey command, see the section `Zle Builtins' in zshzle(1).\n\n-c Print the arguments in columns. Unless -a is also given,\narguments are printed with the row incrementing first.\n\n-C cols\nPrint the arguments in cols columns. Unless -a is also\ngiven, arguments are printed with the row incrementing\nfirst.\n\n-D Treat the arguments as paths, replacing directory pre-\nfixes with ~ expressions corresponding to directory\nnames, as appropriate.\n\n-i If given together with -o or -O, sorting is performed\ncase-independently.\n\n-l Print the arguments separated by newlines instead of spa-\nces. Note: if the list of arguments is empty, print -l\nwill still output one empty line. To print a possi-\nbly-empty list of arguments one per line, use print -C1,\nas in `print -rC1 -- \"$list[@]\"'.\n\n-m Take the first argument as a pattern (should be quoted),\nand remove it from the argument list together with subse-\nquent arguments that do not match this pattern.\n\n-n Do not add a newline to the output.\n\n-N Print the arguments separated and terminated by nulls.\nAgain, print -rNC1 -- \"$list[@]\" is a canonical way to\nprint an arbitrary list as null-delimited records.\n\n-o Print the arguments sorted in ascending order.\n\n-O Print the arguments sorted in descending order.\n\n-p Print the arguments to the input of the coprocess.\n\n-P Perform prompt expansion (see EXPANSION OF PROMPT SE-\nQUENCES in zshmisc(1)). In combination with `-f', prompt\nescape sequences are parsed only within interpolated ar-\nguments, not within the format string.\n\n-r Ignore the escape conventions of echo.\n\n-R Emulate the BSD echo command, which does not process es-\ncape sequences unless the -e flag is given. The -n flag\nsuppresses the trailing newline. Only the -e and -n\nflags are recognized after -R; all other arguments and\noptions are printed.\n\n-s Place the results in the history list instead of on the\nstandard output. Each argument to the print command is\ntreated as a single word in the history, regardless of\nits content.\n\n-S Place the results in the history list instead of on the\nstandard output. In this case only a single argument is\nallowed; it will be split into words as if it were a full\nshell command line. The effect is similar to reading the\nline from a history file with the HISTLEXWORDS option\nactive.\n\n-u n Print the arguments to file descriptor n.\n\n-v name\nStore the printed arguments as the value of the parameter\nname.\n\n-x tab-stop\nExpand leading tabs on each line of output in the printed\nstring assuming a tab stop every tab-stop characters.\nThis is appropriate for formatting code that may be in-\ndented with tabs. Note that leading tabs of any argument\nto print, not just the first, are expanded, even if print\nis using spaces to separate arguments (the column count\nis maintained across arguments but may be incorrect on\noutput owing to previous unexpanded tabs).\n\nThe start of the output of each print command is assumed\nto be aligned with a tab stop. Widths of multibyte char-\nacters are handled if the option MULTIBYTE is in effect.\nThis option is ignored if other formatting options are in\neffect, namely column alignment or printf style, or if\noutput is to a special location such as shell history or\nthe command line editor.\n\n-X tab-stop\nThis is similar to -x, except that all tabs in the\nprinted string are expanded. This is appropriate if tabs\nin the arguments are being used to produce a table for-\nmat.\n\n-z Push the arguments onto the editing buffer stack, sepa-\nrated by spaces.\n\nIf any of `-m', `-o' or `-O' are used in combination with `-f'\nand there are no arguments (after the removal process in the\ncase of `-m') then nothing is printed.\n\nprintf [ -v name ] format [ arg ... ]\nPrint the arguments according to the format specification. For-\nmatting rules are the same as used in C. The same escape se-\nquences as for echo are recognised in the format. All C conver-\nsion specifications ending in one of csdiouxXeEfgGn are handled.\nIn addition to this, `%b' can be used instead of `%s' to cause\nescape sequences in the argument to be recognised and `%q' can\nbe used to quote the argument in such a way that allows it to be\nreused as shell input. With the numeric format specifiers, if\nthe corresponding argument starts with a quote character, the\nnumeric value of the following character is used as the number\nto print; otherwise the argument is evaluated as an arithmetic\nexpression. See the section `Arithmetic Evaluation' in zsh-\nmisc(1) for a description of arithmetic expressions. With `%n',\nthe corresponding argument is taken as an identifier which is\ncreated as an integer parameter.\n\nNormally, conversion specifications are applied to each argument\nin order but they can explicitly specify the nth argument is to\nbe used by replacing `%' by `%n$' and `*' by `*n$'. It is rec-\nommended that you do not mix references of this explicit style\nwith the normal style and the handling of such mixed styles may\nbe subject to future change.\n\nIf arguments remain unused after formatting, the format string\nis reused until all arguments have been consumed. With the print\nbuiltin, this can be suppressed by using the -r option. If more\narguments are required by the format than have been specified,\nthe behaviour is as if zero or an empty string had been speci-\nfied as the argument.\n\nThe -v option causes the output to be stored as the value of the\nparameter name, instead of printed. If name is an array and the\nformat string is reused when consuming arguments then one array\nelement will be used for each use of the format string.\n\npushd [ -qsLP ] [ arg ]\npushd [ -qsLP ] old new\npushd [ -qsLP ] {+|-}n\nChange the current directory, and push the old current directory\nonto the directory stack. In the first form, change the current\ndirectory to arg. If arg is not specified, change to the second\ndirectory on the stack (that is, exchange the top two entries),\nor change to $HOME if the PUSHDTOHOME option is set or if\nthere is only one entry on the stack. Otherwise, arg is inter-\npreted as it would be by cd. The meaning of old and new in the\nsecond form is also the same as for cd.\n\nThe third form of pushd changes directory by rotating the direc-\ntory list. An argument of the form `+n' identifies a stack en-\ntry by counting from the left of the list shown by the dirs com-\nmand, starting with zero. An argument of the form `-n' counts\nfrom the right. If the PUSHDMINUS option is set, the meanings\nof `+' and `-' in this context are swapped.\n\nIf the -q (quiet) option is specified, the hook function chpwd\nand the functions in the array $chpwdfunctions are not called,\nand the new directory stack is not printed. This is useful for\ncalls to pushd that do not change the environment seen by an in-\nteractive user.\n\nIf the option -q is not specified and the shell option\nPUSHDSILENT is not set, the directory stack will be printed af-\nter a pushd is performed.\n\nThe options -s, -L and -P have the same meanings as for the cd\nbuiltin.\n\npushln [ arg ... ]\nEquivalent to print -nz.\n\npwd [ -rLP ]\nPrint the absolute pathname of the current working directory.\nIf the -r or the -P flag is specified, or the CHASELINKS option\nis set and the -L flag is not given, the printed path will not\ncontain symbolic links.\n\nr Same as fc -e -.\n\nread [ -rszpqAclneE ] [ -t [ num ] ] [ -k [ num ] ] [ -d delim ]\n[ -u n ] [ name[?prompt] ] [ name ... ]\nRead one line and break it into fields using the characters in\n$IFS as separators, except as noted below. The first field is\nassigned to the first name, the second field to the second name,\netc., with leftover fields assigned to the last name. If name\nis omitted then REPLY is used for scalars and reply for arrays.\n\n-r Raw mode: a `\\' at the end of a line does not signify\nline continuation and backslashes in the line don't quote\nthe following character and are not removed.\n\n-s Don't echo back characters if reading from the terminal.\n\n-q Read only one character from the terminal and set name to\n`y' if this character was `y' or `Y' and to `n' other-\nwise. With this flag set the return status is zero only\nif the character was `y' or `Y'. This option may be used\nwith a timeout (see -t); if the read times out, or en-\ncounters end of file, status 2 is returned. Input is\nread from the terminal unless one of -u or -p is present.\nThis option may also be used within zle widgets.\n\n-k [ num ]\nRead only one (or num) characters. All are assigned to\nthe first name, without word splitting. This flag is ig-\nnored when -q is present. Input is read from the termi-\nnal unless one of -u or -p is present. This option may\nalso be used within zle widgets.\n\nNote that despite the mnemonic `key' this option does\nread full characters, which may consist of multiple bytes\nif the option MULTIBYTE is set.\n\n-z Read one entry from the editor buffer stack and assign it\nto the first name, without word splitting. Text is\npushed onto the stack with `print -z' or with push-line\nfrom the line editor (see zshzle(1)). This flag is ig-\nnored when the -k or -q flags are present.\n\n-e\n-E The input read is printed (echoed) to the standard out-\nput. If the -e flag is used, no input is assigned to the\nparameters.\n\n-A The first name is taken as the name of an array and all\nwords are assigned to it.\n\n-c\n-l These flags are allowed only if called inside a function\nused for completion (specified with the -K flag to com-\npctl). If the -c flag is given, the words of the current\ncommand are read. If the -l flag is given, the whole line\nis assigned as a scalar. If both flags are present, -l\nis used and -c is ignored.\n\n-n Together with -c, the number of the word the cursor is on\nis read. With -l, the index of the character the cursor\nis on is read. Note that the command name is word number\n1, not word 0, and that when the cursor is at the end of\nthe line, its character index is the length of the line\nplus one.\n\n-u n Input is read from file descriptor n.\n\n-p Input is read from the coprocess.\n\n-d delim\nInput is terminated by the first character of delim in-\nstead of by newline.\n\n-t [ num ]\nTest if input is available before attempting to read. If\nnum is present, it must begin with a digit and will be\nevaluated to give a number of seconds, which may be a\nfloating point number; in this case the read times out if\ninput is not available within this time. If num is not\npresent, it is taken to be zero, so that read returns im-\nmediately if no input is available. If no input is\navailable, return status 1 and do not set any variables.\n\nThis option is not available when reading from the editor\nbuffer with -z, when called from within completion with\n-c or -l, with -q which clears the input queue before\nreading, or within zle where other mechanisms should be\nused to test for input.\n\nNote that read does not attempt to alter the input pro-\ncessing mode. The default mode is canonical input, in\nwhich an entire line is read at a time, so usually `read\n-t' will not read anything until an entire line has been\ntyped. However, when reading from the terminal with -k\ninput is processed one key at a time; in this case, only\navailability of the first character is tested, so that\ne.g. `read -t -k 2' can still block on the second charac-\nter. Use two instances of `read -t -k' if this is not\nwhat is wanted.\n\nIf the first argument contains a `?', the remainder of this word\nis used as a prompt on standard error when the shell is interac-\ntive.\n\nThe value (exit status) of read is 1 when an end-of-file is en-\ncountered, or when -c or -l is present and the command is not\ncalled from a compctl function, or as described for -q. Other-\nwise the value is 0.\n\nThe behavior of some combinations of the -k, -p, -q, -u and -z\nflags is undefined. Presently -q cancels all the others, -p\ncancels -u, -k cancels -z, and otherwise -z cancels both -p and\n-u.\n\nThe -c or -l flags cancel any and all of -kpquz.\n\nreadonly\nSame as typeset -r. With the POSIXBUILTINS option set, same as\ntypeset -gr.\n\nrehash Same as hash -r.\n\nreturn [ n ]\nCauses a shell function or `.' script to return to the invoking\nscript with the return status specified by an arithmetic expres-\nsion n. If n is omitted, the return status is that of the last\ncommand executed.\n\nIf return was executed from a trap in a TRAPNAL function, the\neffect is different for zero and non-zero return status. With\nzero status (or after an implicit return at the end of the\ntrap), the shell will return to whatever it was previously pro-\ncessing; with a non-zero status, the shell will behave as inter-\nrupted except that the return status of the trap is retained.\nNote that the numeric value of the signal which caused the trap\nis passed as the first argument, so the statement `return\n$((128+$1))' will return the same status as if the signal had\nnot been trapped.\n\nsched See the section `The zsh/sched Module' in zshmodules(1).\n\nset [ {+|-}options | {+|-}o [ optionname ] ] ... [ {+|-}A [ name ] ]\n[ arg ... ]\nSet the options for the shell and/or set the positional parame-\nters, or declare and set an array. If the -s option is given,\nit causes the specified arguments to be sorted before assigning\nthem to the positional parameters (or to the array name if -A is\nused). With +s sort arguments in descending order. For the\nmeaning of the other flags, see zshoptions(1). Flags may be\nspecified by name using the -o option. If no option name is sup-\nplied with -o, the current option states are printed: see the\ndescription of setopt below for more information on the format.\nWith +o they are printed in a form that can be used as input to\nthe shell.\n\nIf the -A flag is specified, name is set to an array containing\nthe given args; if no name is specified, all arrays are printed\ntogether with their values.\n\nIf +A is used and name is an array, the given arguments will re-\nplace the initial elements of that array; if no name is speci-\nfied, all arrays are printed without their values.\n\nThe behaviour of arguments after -A name or +A name depends on\nwhether the option KSHARRAYS is set. If it is not set, all ar-\nguments following name are treated as values for the array, re-\ngardless of their form. If the option is set, normal option\nprocessing continues at that point; only regular arguments are\ntreated as values for the array. This means that\n\nset -A array -x -- foo\n\nsets array to `-x -- foo' if KSHARRAYS is not set, but sets the\narray to foo and turns on the option `-x' if it is set.\n\nIf the -A flag is not present, but there are arguments beyond\nthe options, the positional parameters are set. If the option\nlist (if any) is terminated by `--', and there are no further\narguments, the positional parameters will be unset.\n\nIf no arguments and no `--' are given, then the names and values\nof all parameters are printed on the standard output. If the\nonly argument is `+', the names of all parameters are printed.\n\nFor historical reasons, `set -' is treated as `set +xv' and `set\n- args' as `set +xv -- args' when in any other emulation mode\nthan zsh's native mode.\n\nsetcap See the section `The zsh/cap Module' in zshmodules(1).\n\nsetopt [ {+|-}options | {+|-}o optionname ] [ -m ] [ name ... ]\nSet the options for the shell. All options specified either\nwith flags or by name are set.\n\nIf no arguments are supplied, the names of all options currently\nset are printed. The form is chosen so as to minimize the dif-\nferences from the default options for the current emulation (the\ndefault emulation being native zsh, shown as in zshop-\ntions(1)). Options that are on by default for the emulation are\nshown with the prefix no only if they are off, while other op-\ntions are shown without the prefix no and only if they are on.\nIn addition to options changed from the default state by the\nuser, any options activated automatically by the shell (for ex-\nample, SHINSTDIN or INTERACTIVE) will be shown in the list.\nThe format is further modified by the option KSHOPTIONPRINT,\nhowever the rationale for choosing options with or without the\nno prefix remains the same in this case.\n\nIf the -m flag is given the arguments are taken as patterns\n(which should be quoted to protect them from filename expan-\nsion), and all options with names matching these patterns are\nset.\n\nNote that a bad option name does not cause execution of subse-\nquent shell code to be aborted; this is behaviour is different\nfrom that of `set -o'. This is because set is regarded as a\nspecial builtin by the POSIX standard, but setopt is not.\n\nshift [ -p ] [ n ] [ name ... ]\nThe positional parameters ${n+1} ... are renamed to $1 ...,\nwhere n is an arithmetic expression that defaults to 1. If any\nnames are given then the arrays with these names are shifted in-\nstead of the positional parameters.\n\nIf the option -p is given arguments are instead removed (popped)\nfrom the end rather than the start of the array.\n\nsource file [ arg ... ]\nSame as `.', except that the current directory is always\nsearched and is always searched first, before directories in\n$path.\n\nstat See the section `The zsh/stat Module' in zshmodules(1).\n\nsuspend [ -f ]\nSuspend the execution of the shell (send it a SIGTSTP) until it\nreceives a SIGCONT. Unless the -f option is given, this will\nrefuse to suspend a login shell.\n\ntest [ arg ... ]\n[ [ arg ... ] ]\nLike the system version of test. Added for compatibility; use\nconditional expressions instead (see the section `Conditional\nExpressions'). The main differences between the conditional ex-\npression syntax and the test and [ builtins are: these commands\nare not handled syntactically, so for example an empty variable\nexpansion may cause an argument to be omitted; syntax errors\ncause status 2 to be returned instead of a shell error; and\narithmetic operators expect integer arguments rather than arith-\nmetic expressions.\n\nThe command attempts to implement POSIX and its extensions where\nthese are specified. Unfortunately there are intrinsic ambigui-\nties in the syntax; in particular there is no distinction be-\ntween test operators and strings that resemble them. The stan-\ndard attempts to resolve these for small numbers of arguments\n(up to four); for five or more arguments compatibility cannot be\nrelied on. Users are urged wherever possible to use the `[['\ntest syntax which does not have these ambiguities.\n\ntimes Print the accumulated user and system times for the shell and\nfor processes run from the shell.\n\ntrap [ arg ] [ sig ... ]\narg is a series of commands (usually quoted to protect it from\nimmediate evaluation by the shell) to be read and executed when\nthe shell receives any of the signals specified by one or more\nsig args. Each sig can be given as a number, or as the name of\na signal either with or without the string SIG in front (e.g. 1,\nHUP, and SIGHUP are all the same signal).\n\nIf arg is `-', then the specified signals are reset to their de-\nfaults, or, if no sig args are present, all traps are reset.\n\nIf arg is an empty string, then the specified signals are ig-\nnored by the shell (and by the commands it invokes).\n\nIf arg is omitted but one or more sig args are provided (i.e.\nthe first argument is a valid signal number or name), the effect\nis the same as if arg had been specified as `-'.\n\nThe trap command with no arguments prints a list of commands as-\nsociated with each signal.\n\nIf sig is ZERR then arg will be executed after each command with\na nonzero exit status. ERR is an alias for ZERR on systems that\nhave no SIGERR signal (this is the usual case).\n\nIf sig is DEBUG then arg will be executed before each command if\nthe option DEBUGBEFORECMD is set (as it is by default), else\nafter each command. Here, a `command' is what is described as a\n`sublist' in the shell grammar, see the section SIMPLE COMMANDS\n& PIPELINES in zshmisc(1). If DEBUGBEFORECMD is set various\nadditional features are available. First, it is possible to\nskip the next command by setting the option ERREXIT; see the\ndescription of the ERREXIT option in zshoptions(1). Also, the\nshell parameter ZSHDEBUGCMD is set to the string corresponding\nto the command to be executed following the trap. Note that\nthis string is reconstructed from the internal format and may\nnot be formatted the same way as the original text. The parame-\nter is unset after the trap is executed.\n\nIf sig is 0 or EXIT and the trap statement is executed inside\nthe body of a function, then the command arg is executed after\nthe function completes. The value of $? at the start of execu-\ntion is the exit status of the shell or the return status of the\nfunction exiting. If sig is 0 or EXIT and the trap statement is\nnot executed inside the body of a function, then the command arg\nis executed when the shell terminates; the trap runs before any\nzshexit hook functions.\n\nZERR, DEBUG, and EXIT traps are not executed inside other traps.\nZERR and DEBUG traps are kept within subshells, while other\ntraps are reset.\n\nNote that traps defined with the trap builtin are slightly dif-\nferent from those defined as `TRAPNAL () { ... }', as the latter\nhave their own function environment (line numbers, local vari-\nables, etc.) while the former use the environment of the command\nin which they were called. For example,\n\ntrap 'print $LINENO' DEBUG\n\nwill print the line number of a command executed after it has\nrun, while\n\nTRAPDEBUG() { print $LINENO; }\n\nwill always print the number zero.\n\nAlternative signal names are allowed as described under kill\nabove. Defining a trap under either name causes any trap under\nan alternative name to be removed. However, it is recommended\nthat for consistency users stick exclusively to one name or an-\nother.\n\ntrue [ arg ... ]\nDo nothing and return an exit status of 0.\n\nttyctl [ -fu ]\nThe -f option freezes the tty (i.e. terminal or terminal emula-\ntor), and -u unfreezes it. When the tty is frozen, no changes\nmade to the tty settings by external programs will be honored by\nthe shell, except for changes in the size of the screen; the\nshell will simply reset the settings to their previous values as\nsoon as each command exits or is suspended. Thus, stty and sim-\nilar programs have no effect when the tty is frozen. Freezing\nthe tty does not cause the current state to be remembered: in-\nstead, it causes future changes to the state to be blocked.\n\nWithout options it reports whether the terminal is frozen or\nnot.\n\nNote that, regardless of whether the tty is frozen or not, the\nshell needs to change the settings when the line editor starts,\nso unfreezing the tty does not guarantee settings made on the\ncommand line are preserved. Strings of commands run between\nediting the command line will see a consistent tty state. See\nalso the shell variable STTY for a means of initialising the tty\nbefore running external commands.\n\ntype [ -wfpamsS ] name ...\nEquivalent to whence -v.\n\ntypeset [ {+|-}AHUaghlmrtux ] [ {+|-}EFLRZip [ n ] ]\n[ + ] [ name[=value] ... ]\ntypeset -T [ {+|-}Uglrux ] [ {+|-}LRZp [ n ] ]\n[ + | SCALAR[=value] array[=(value ...)] [ sep ] ]\ntypeset -f [ {+|-}TUkmtuz ] [ + ] [ name ... ]\nSet or display attributes and values for shell parameters.\n\nExcept as noted below for control flags that change the behav-\nior, a parameter is created for each name that does not already\nrefer to one. When inside a function, a new parameter is cre-\nated for every name (even those that already exist), and is un-\nset again when the function completes. See `Local Parameters'\nin zshparam(1). The same rules apply to special shell parame-\nters, which retain their special attributes when made local.\n\nFor each name=value assignment, the parameter name is set to\nvalue.\n\nIf the shell option TYPESETSILENT is not set, for each remain-\ning name that refers to a parameter that is already set, the\nname and value of the parameter are printed in the form of an\nassignment. Nothing is printed for newly-created parameters, or\nwhen any attribute flags listed below are given along with the\nname. Using `+' instead of minus to introduce an attribute\nturns it off.\n\nIf no name is present, the names and values of all parameters\nare printed. In this case the attribute flags restrict the dis-\nplay to only those parameters that have the specified at-\ntributes, and using `+' rather than `-' to introduce the flag\nsuppresses printing of the values of parameters when there is no\nparameter name.\n\nAll forms of the command handle scalar assignment. Array as-\nsignment is possible if any of the reserved words declare, ex-\nport, float, integer, local, readonly or typeset is matched when\nthe line is parsed (N.B. not when it is executed). In this case\nthe arguments are parsed as assignments, except that the `+='\nsyntax and the GLOBASSIGN option are not supported, and scalar\nvalues after = are not split further into words, even if ex-\npanded (regardless of the setting of the KSHTYPESET option;\nthis option is obsolete).\n\nExamples of the differences between command and reserved word\nparsing:\n\n# Reserved word parsing\ntypeset svar=$(echo one word) avar=(several words)\n\nThe above creates a scalar parameter svar and an array parameter\navar as if the assignments had been\n\nsvar=\"one word\"\navar=(several words)\n\nOn the other hand:\n\n# Normal builtin interface\nbuiltin typeset svar=$(echo two words)\n\nThe builtin keyword causes the above to use the standard builtin\ninterface to typeset in which argument parsing is performed in\nthe same way as for other commands. This example creates a\nscalar svar containing the value two and another scalar parame-\nter words with no value. An array value in this case would ei-\nther cause an error or be treated as an obscure set of glob\nqualifiers.\n\nArbitrary arguments are allowed if they take the form of assign-\nments after command line expansion; however, these only perform\nscalar assignment:\n\nvar='svar=val'\ntypeset $var\n\nThe above sets the scalar parameter svar to the value val.\nParentheses around the value within var would not cause array\nassignment as they will be treated as ordinary characters when\n$var is substituted. Any non-trivial expansion in the name part\nof the assignment causes the argument to be treated in this\nfashion:\n\ntypeset {var1,var2,var3}=name\n\nThe above syntax is valid, and has the expected effect of set-\nting the three parameters to the same value, but the command\nline is parsed as a set of three normal command line arguments\nto typeset after expansion. Hence it is not possible to assign\nto multiple arrays by this means.\n\nNote that each interface to any of the commands my be disabled\nseparately. For example, `disable -r typeset' disables the re-\nserved word interface to typeset, exposing the builtin inter-\nface, while `disable typeset' disables the builtin. Note that\ndisabling the reserved word interface for typeset may cause\nproblems with the output of `typeset -p', which assumes the re-\nserved word interface is available in order to restore array and\nassociative array values.\n\nUnlike parameter assignment statements, typeset's exit status on\nan assignment that involves a command substitution does not re-\nflect the exit status of the command substitution. Therefore,\nto test for an error in a command substitution, separate the\ndeclaration of the parameter from its initialization:\n\n# WRONG\ntypeset var1=$(exit 1) || echo \"Trouble with var1\"\n\n# RIGHT\ntypeset var1 && var1=$(exit 1) || echo \"Trouble with var1\"\n\nTo initialize a parameter param to a command output and mark it\nreadonly, use typeset -r param or readonly param after the pa-\nrameter assignment statement.\n\nIf no attribute flags are given, and either no name arguments\nare present or the flag +m is used, then each parameter name\nprinted is preceded by a list of the attributes of that parame-\nter (array, association, exported, float, integer, readonly, or\nundefined for autoloaded parameters not yet loaded). If +m is\nused with attribute flags, and all those flags are introduced\nwith +, the matching parameter names are printed but their val-\nues are not.\n\nThe following control flags change the behavior of typeset:\n\n+ If `+' appears by itself in a separate word as the last\noption, then the names of all parameters (functions with\n-f) are printed, but the values (function bodies) are\nnot. No name arguments may appear, and it is an error\nfor any other options to follow `+'. The effect of `+'\nis as if all attribute flags which precede it were given\nwith a `+' prefix. For example, `typeset -U +' is equiv-\nalent to `typeset +U' and displays the names of all ar-\nrays having the uniqueness attribute, whereas `typeset -f\n-U +' displays the names of all autoloadable functions.\nIf + is the only option, then type information (array,\nreadonly, etc.) is also printed for each parameter, in\nthe same manner as `typeset +m \"*\"'.\n\n-g The -g (global) means that any resulting parameter will\nnot be restricted to local scope. Note that this does\nnot necessarily mean that the parameter will be global,\nas the flag will apply to any existing parameter (even if\nunset) from an enclosing function. This flag does not\naffect the parameter after creation, hence it has no ef-\nfect when listing existing parameters, nor does the flag\n+g have any effect except in combination with -m (see be-\nlow).\n\n-m If the -m flag is given the name arguments are taken as\npatterns (use quoting to prevent these from being inter-\npreted as file patterns). With no attribute flags, all\nparameters (or functions with the -f flag) with matching\nnames are printed (the shell option TYPESETSILENT is not\nused in this case).\n\nIf the +g flag is combined with -m, a new local parameter\nis created for every matching parameter that is not al-\nready local. Otherwise -m applies all other flags or as-\nsignments to the existing parameters.\n\nExcept when assignments are made with name=value, using\n+m forces the matching parameters and their attributes to\nbe printed, even inside a function. Note that -m is ig-\nnored if no patterns are given, so `typeset -m' displays\nattributes but `typeset -a +m' does not.\n\n-p [ n ]\nIf the -p option is given, parameters and values are\nprinted in the form of a typeset command with an assign-\nment, regardless of other flags and options. Note that\nthe -H flag on parameters is respected; no value will be\nshown for these parameters.\n\n-p may be followed by an optional integer argument. Cur-\nrently only the value 1 is supported. In this case ar-\nrays and associative arrays are printed with newlines be-\ntween indented elements for readability.\n\n-T [ scalar[=value] array[=(value ...)] [ sep ] ]\nThis flag has a different meaning when used with -f; see\nbelow. Otherwise the -T option requires zero, two, or\nthree arguments to be present. With no arguments, the\nlist of parameters created in this fashion is shown.\nWith two or three arguments, the first two are the name\nof a scalar and of an array parameter (in that order)\nthat will be tied together in the manner of $PATH and\n$path. The optional third argument is a single-character\nseparator which will be used to join the elements of the\narray to form the scalar; if absent, a colon is used, as\nwith $PATH. Only the first character of the separator is\nsignificant; any remaining characters are ignored.\nMultibyte characters are not yet supported.\n\nOnly one of the scalar and array parameters may be as-\nsigned an initial value (the restrictions on assignment\nforms described above also apply).\n\nBoth the scalar and the array may be manipulated as nor-\nmal. If one is unset, the other will automatically be\nunset too. There is no way of untying the variables\nwithout unsetting them, nor of converting the type of one\nof them with another typeset command; +T does not work,\nassigning an array to scalar is an error, and assigning a\nscalar to array sets it to be a single-element array.\n\nNote that both `typeset -xT ...' and `export -T ...'\nwork, but only the scalar will be marked for export.\nSetting the value using the scalar version causes a split\non all separators (which cannot be quoted). It is possi-\nble to apply -T to two previously tied variables but with\na different separator character, in which case the vari-\nables remain joined as before but the separator is\nchanged.\n\nWhen an existing scalar is tied to a new array, the value\nof the scalar is preserved but no attribute other than\nexport will be preserved.\n\nAttribute flags that transform the final value (-L, -R, -Z, -l,\n-u) are only applied to the expanded value at the point of a pa-\nrameter expansion expression using `$'. They are not applied\nwhen a parameter is retrieved internally by the shell for any\npurpose.\n\nThe following attribute flags may be specified:\n\n-A The names refer to associative array parameters; see `Ar-\nray Parameters' in zshparam(1).\n\n-L [ n ]\nLeft justify and remove leading blanks from the value\nwhen the parameter is expanded. If n is nonzero, it de-\nfines the width of the field. If n is zero, the width is\ndetermined by the width of the value of the first assign-\nment. In the case of numeric parameters, the length of\nthe complete value assigned to the parameter is used to\ndetermine the width, not the value that would be output.\n\nThe width is the count of characters, which may be multi-\nbyte characters if the MULTIBYTE option is in effect.\nNote that the screen width of the character is not taken\ninto account; if this is required, use padding with pa-\nrameter expansion flags ${(ml...)...} as described in\n`Parameter Expansion Flags' in zshexpn(1).\n\nWhen the parameter is expanded, it is filled on the right\nwith blanks or truncated if necessary to fit the field.\nNote truncation can lead to unexpected results with nu-\nmeric parameters. Leading zeros are removed if the -Z\nflag is also set.\n\n-R [ n ]\nSimilar to -L, except that right justification is used;\nwhen the parameter is expanded, the field is left filled\nwith blanks or truncated from the end. May not be com-\nbined with the -Z flag.\n\n-U For arrays (but not for associative arrays), keep only\nthe first occurrence of each duplicated value. This may\nalso be set for tied parameters (see -T) or colon-sepa-\nrated special parameters like PATH or FIGNORE, etc. Note\nthe flag takes effect on assignment, and the type of the\nvariable being assigned to is determinative; for vari-\nables with shared values it is therefore recommended to\nset the flag for all interfaces, e.g. `typeset -U PATH\npath'.\n\nThis flag has a different meaning when used with -f; see\nbelow.\n\n-Z [ n ]\nSpecially handled if set along with the -L flag. Other-\nwise, similar to -R, except that leading zeros are used\nfor padding instead of blanks if the first non-blank\ncharacter is a digit. Numeric parameters are specially\nhandled: they are always eligible for padding with ze-\nroes, and the zeroes are inserted at an appropriate place\nin the output.\n\n-a The names refer to array parameters. An array parameter\nmay be created this way, but it may be assigned to in the\ntypeset statement only if the reserved word form of type-\nset is enabled (as it is by default). When displaying,\nboth normal and associative arrays are shown.\n\n-f The names refer to functions rather than parameters. No\nassignments can be made, and the only other valid flags\nare -t, -T, -k, -u, -U and -z. The flag -t turns on exe-\ncution tracing for this function; the flag -T does the\nsame, but turns off tracing for any named (not anonymous)\nfunction called from the present one, unless that func-\ntion also has the -t or -T flag. The -u and -U flags\ncause the function to be marked for autoloading; -U also\ncauses alias expansion to be suppressed when the function\nis loaded. See the description of the `autoload' builtin\nfor details.\n\nNote that the builtin functions provides the same basic\ncapabilities as typeset -f but gives access to a few ex-\ntra options; autoload gives further additional options\nfor the case typeset -fu and typeset -fU.\n\n-h Hide: only useful for special parameters (those marked\n`' in the table in zshparam(1)), and for local parame-\nters with the same name as a special parameter, though\nharmless for others. A special parameter with this at-\ntribute will not retain its special effect when made lo-\ncal. Thus after `typeset -h PATH', a function containing\n`typeset PATH' will create an ordinary local parameter\nwithout the usual behaviour of PATH. Alternatively, the\nlocal parameter may itself be given this attribute; hence\ninside a function `typeset -h PATH' creates an ordinary\nlocal parameter and the special PATH parameter is not al-\ntered in any way. It is also possible to create a local\nparameter using `typeset +h special', where the local\ncopy of special will retain its special properties re-\ngardless of having the -h attribute. Global special pa-\nrameters loaded from shell modules (currently those in\nzsh/mapfile and zsh/parameter) are automatically given\nthe -h attribute to avoid name clashes.\n\n-H Hide value: specifies that typeset will not display the\nvalue of the parameter when listing parameters; the dis-\nplay for such parameters is always as if the `+' flag had\nbeen given. Use of the parameter is in other respects\nnormal, and the option does not apply if the parameter is\nspecified by name, or by pattern with the -m option.\nThis is on by default for the parameters in the zsh/pa-\nrameter and zsh/mapfile modules. Note, however, that un-\nlike the -h flag this is also useful for non-special pa-\nrameters.\n\n-i [ n ]\nUse an internal integer representation. If n is nonzero\nit defines the output arithmetic base, otherwise it is\ndetermined by the first assignment. Bases from 2 to 36\ninclusive are allowed.\n\n-E [ n ]\nUse an internal double-precision floating point represen-\ntation. On output the variable will be converted to sci-\nentific notation. If n is nonzero it defines the number\nof significant figures to display; the default is ten.\n\n-F [ n ]\nUse an internal double-precision floating point represen-\ntation. On output the variable will be converted to\nfixed-point decimal notation. If n is nonzero it defines\nthe number of digits to display after the decimal point;\nthe default is ten.\n\n-l Convert the result to lower case whenever the parameter\nis expanded. The value is not converted when assigned.\n\n-r The given names are marked readonly. Note that if name\nis a special parameter, the readonly attribute can be\nturned on, but cannot then be turned off.\n\nIf the POSIXBUILTINS option is set, the readonly attri-\nbute is more restrictive: unset variables can be marked\nreadonly and cannot then be set; furthermore, the read-\nonly attribute cannot be removed from any variable.\n\nIt is still possible to change other attributes of the\nvariable though, some of which like -U or -Z would affect\nthe value. More generally, the readonly attribute should\nnot be relied on as a security mechanism.\n\nNote that in zsh (like in pdksh but unlike most other\nshells) it is still possible to create a local variable\nof the same name as this is considered a different vari-\nable (though this variable, too, can be marked readonly).\nSpecial variables that have been made readonly retain\ntheir value and readonly attribute when made local.\n\n-t Tags the named parameters. Tags have no special meaning\nto the shell. This flag has a different meaning when\nused with -f; see above.\n\n-u Convert the result to upper case whenever the parameter\nis expanded. The value is not converted when assigned.\nThis flag has a different meaning when used with -f; see\nabove.\n\n-x Mark for automatic export to the environment of subse-\nquently executed commands. If the option GLOBALEXPORT\nis set, this implies the option -g, unless +g is also ex-\nplicitly given; in other words the parameter is not made\nlocal to the enclosing function. This is for compatibil-\nity with previous versions of zsh.\n\nulimit [ -HSa ] [ { -bcdfiklmnpqrsTtvwx | -N resource } [ limit ] ... ]\nSet or display resource limits of the shell and the processes\nstarted by the shell. The value of limit can be a number in the\nunit specified below or one of the values `unlimited', which re-\nmoves the limit on the resource, or `hard', which uses the cur-\nrent value of the hard limit on the resource.\n\nBy default, only soft limits are manipulated. If the -H flag is\ngiven use hard limits instead of soft limits. If the -S flag is\ngiven together with the -H flag set both hard and soft limits.\n\nIf no options are used, the file size limit (-f) is assumed.\n\nIf limit is omitted the current value of the specified resources\nare printed. When more than one resource value is printed, the\nlimit name and unit is printed before each value.\n\nWhen looping over multiple resources, the shell will abort imme-\ndiately if it detects a badly formed argument. However, if it\nfails to set a limit for some other reason it will continue try-\ning to set the remaining limits.\n\nNot all the following resources are supported on all systems.\nRunning ulimit -a will show which are supported.\n\n-a Lists all of the current resource limits.\n-b Socket buffer size in bytes (N.B. not kilobytes)\n-c 512-byte blocks on the size of core dumps.\n-d Kilobytes on the size of the data segment.\n-f 512-byte blocks on the size of files written.\n-i The number of pending signals.\n-k The number of kqueues allocated.\n-l Kilobytes on the size of locked-in memory.\n-m Kilobytes on the size of physical memory.\n-n open file descriptors.\n-p The number of pseudo-terminals.\n-q Bytes in POSIX message queues.\n-r Maximum real time priority. On some systems where this\nis not available, such as NetBSD, this has the same ef-\nfect as -T for compatibility with sh.\n-s Kilobytes on the size of the stack.\n-T The number of simultaneous threads available to the user.\n-t CPU seconds to be used.\n-u The number of processes available to the user.\n-v Kilobytes on the size of virtual memory. On some systems\nthis refers to the limit called `address space'.\n-w Kilobytes on the size of swapped out memory.\n-x The number of locks on files.\n\nA resource may also be specified by integer in the form `-N re-\nsource', where resource corresponds to the integer defined for\nthe resource by the operating system. This may be used to set\nthe limits for resources known to the shell which do not corre-\nspond to option letters. Such limits will be shown by number in\nthe output of `ulimit -a'.\n\nThe number may alternatively be out of the range of limits com-\npiled into the shell. The shell will try to read or write the\nlimit anyway, and will report an error if this fails.\n\numask [ -S ] [ mask ]\nThe umask is set to mask. mask can be either an octal number or\na symbolic value as described in chmod(1). If mask is omitted,\nthe current value is printed. The -S option causes the mask to\nbe printed as a symbolic value. Otherwise, the mask is printed\nas an octal number. Note that in the symbolic form the permis-\nsions you specify are those which are to be allowed (not denied)\nto the users specified.\n\nunalias [ -ams ] name ...\nRemoves aliases. This command works the same as unhash -a, ex-\ncept that the -a option removes all regular or global aliases,\nor with -s all suffix aliases: in this case no name arguments\nmay appear. The options -m (remove by pattern) and -s without\n-a (remove listed suffix aliases) behave as for unhash -a. Note\nthat the meaning of -a is different between unalias and unhash.\n\nunfunction\nSame as unhash -f.\n\nunhash [ -adfms ] name ...\nRemove the element named name from an internal hash table. The\ndefault is remove elements from the command hash table. The -a\noption causes unhash to remove regular or global aliases; note\nwhen removing a global aliases that the argument must be quoted\nto prevent it from being expanded before being passed to the\ncommand. The -s option causes unhash to remove suffix aliases.\nThe -f option causes unhash to remove shell functions. The -d\noptions causes unhash to remove named directories. If the -m\nflag is given the arguments are taken as patterns (should be\nquoted) and all elements of the corresponding hash table with\nmatching names will be removed.\n\nunlimit [ -hs ] resource ...\nThe resource limit for each resource is set to the hard limit.\nIf the -h flag is given and the shell has appropriate privi-\nleges, the hard resource limit for each resource is removed.\nThe resources of the shell process are only changed if the -s\nflag is given.\n\nThe unlimit command is not made available by default when the\nshell starts in a mode emulating another shell. It can be made\navailable with the command `zmodload -F zsh/rlimits b:unlimit'.\n\nunset [ -fmv ] name ...\nEach named parameter is unset. Local parameters remain local\neven if unset; they appear unset within scope, but the previous\nvalue will still reappear when the scope ends.\n\nIndividual elements of associative array parameters may be unset\nby using subscript syntax on name, which should be quoted (or\nthe entire command prefixed with noglob) to protect the sub-\nscript from filename generation.\n\nIf the -m flag is specified the arguments are taken as patterns\n(should be quoted) and all parameters with matching names are\nunset. Note that this cannot be used when unsetting associative\narray elements, as the subscript will be treated as part of the\npattern.\n\nThe -v flag specifies that name refers to parameters. This is\nthe default behaviour.\n\nunset -f is equivalent to unfunction.\n\nunsetopt [ {+|-}options | {+|-}o optionname ] [ name ... ]\nUnset the options for the shell. All options specified either\nwith flags or by name are unset. If no arguments are supplied,\nthe names of all options currently unset are printed. If the -m\nflag is given the arguments are taken as patterns (which should\nbe quoted to preserve them from being interpreted as glob pat-\nterns), and all options with names matching these patterns are\nunset.\n\nvared See the section `Zle Builtins' in zshzle(1).\n\nwait [ job ... ]\nWait for the specified jobs or processes. If job is not given\nthen all currently active child processes are waited for. Each\njob can be either a job specification or the process ID of a job\nin the job table. The exit status from this command is that of\nthe job waited for. If job represents an unknown job or process\nID, a warning is printed (unless the POSIXBUILTINS option is\nset) and the exit status is 127.\n\nIt is possible to wait for recent processes (specified by\nprocess ID, not by job) that were running in the background even\nif the process has exited. Typically the process ID will be\nrecorded by capturing the value of the variable $! immediately\nafter the process has been started. There is a limit on the\nnumber of process IDs remembered by the shell; this is given by\nthe value of the system configuration parameter CHILDMAX. When\nthis limit is reached, older process IDs are discarded, least\nrecently started processes first.\n\nNote there is no protection against the process ID wrapping,\ni.e. if the wait is not executed soon enough there is a chance\nthe process waited for is the wrong one. A conflict implies\nboth process IDs have been generated by the shell, as other pro-\ncesses are not recorded, and that the user is potentially inter-\nested in both, so this problem is intrinsic to process IDs.\n\nwhence [ -vcwfpamsS ] [ -x num ] name ...\nFor each name, indicate how it would be interpreted if used as a\ncommand name.\n\nIf name is not an alias, built-in command, external command,\nshell function, hashed command, or a reserved word, the exit\nstatus shall be non-zero, and -- if -v, -c, or -w was passed --\na message will be written to standard output. (This is differ-\nent from other shells that write that message to standard er-\nror.)\n\nwhence is most useful when name is only the last path component\nof a command, i.e. does not include a `/'; in particular, pat-\ntern matching only succeeds if just the non-directory component\nof the command is passed.\n\n-v Produce a more verbose report.\n\n-c Print the results in a csh-like format. This takes\nprecedence over -v.\n\n-w For each name, print `name: word' where word is one of\nalias, builtin, command, function, hashed, reserved or\nnone, according as name corresponds to an alias, a\nbuilt-in command, an external command, a shell function,\na command defined with the hash builtin, a reserved word,\nor is not recognised. This takes precedence over -v and\n-c.\n\n-f Causes the contents of a shell function to be displayed,\nwhich would otherwise not happen unless the -c flag were\nused.\n\n-p Do a path search for name even if it is an alias, re-\nserved word, shell function or builtin.\n\n-a Do a search for all occurrences of name throughout the\ncommand path. Normally only the first occurrence is\nprinted.\n\n-m The arguments are taken as patterns (pattern characters\nshould be quoted), and the information is displayed for\neach command matching one of these patterns.\n\n-s If a pathname contains symlinks, print the symlink-free\npathname as well.\n\n-S As -s, but if the pathname had to be resolved by follow-\ning multiple symlinks, the intermediate steps are\nprinted, too. The symlink resolved at each step might be\nanywhere in the path.\n\n-x num Expand tabs when outputting shell functions using the -c\noption. This has the same effect as the -x option to the\nfunctions builtin.\n\nwhere [ -wpmsS ] [ -x num ] name ...\nEquivalent to whence -ca.\n\nwhich [ -wpamsS ] [ -x num ] name ...\nEquivalent to whence -c.\n\nzcompile [ -U ] [ -z | -k ] [ -R | -M ] file [ name ... ]\nzcompile -ca [ -m ] [ -R | -M ] file [ name ... ]\nzcompile -t file [ name ... ]\nThis builtin command can be used to compile functions or\nscripts, storing the compiled form in a file, and to examine\nfiles containing the compiled form. This allows faster au-\ntoloading of functions and sourcing of scripts by avoiding pars-\ning of the text when the files are read.\n\nThe first form (without the -c, -a or -t options) creates a com-\npiled file. If only the file argument is given, the output file\nhas the name `file.zwc' and will be placed in the same directory\nas the file. The shell will load the compiled file instead of\nthe normal function file when the function is autoloaded; see\nthe section `Autoloading Functions' in zshmisc(1) for a descrip-\ntion of how autoloaded functions are searched. The extension\n.zwc stands for `zsh word code'.\n\nIf there is at least one name argument, all the named files are\ncompiled into the output file given as the first argument. If\nfile does not end in .zwc, this extension is automatically ap-\npended. Files containing multiple compiled functions are called\n`digest' files, and are intended to be used as elements of the\nFPATH/fpath special array.\n\nThe second form, with the -c or -a options, writes the compiled\ndefinitions for all the named functions into file. For -c, the\nnames must be functions currently defined in the shell, not\nthose marked for autoloading. Undefined functions that are\nmarked for autoloading may be written by using the -a option, in\nwhich case the fpath is searched and the contents of the defini-\ntion files for those functions, if found, are compiled into\nfile. If both -c and -a are given, names of both defined func-\ntions and functions marked for autoloading may be given. In ei-\nther case, the functions in files written with the -c or -a op-\ntion will be autoloaded as if the KSHAUTOLOAD option were un-\nset.\n\nThe reason for handling loaded and not-yet-loaded functions with\ndifferent options is that some definition files for autoloading\ndefine multiple functions, including the function with the same\nname as the file, and, at the end, call that function. In such\ncases the output of `zcompile -c' does not include the addi-\ntional functions defined in the file, and any other initializa-\ntion code in the file is lost. Using `zcompile -a' captures all\nthis extra information.\n\nIf the -m option is combined with -c or -a, the names are used\nas patterns and all functions whose names match one of these\npatterns will be written. If no name is given, the definitions\nof all functions currently defined or marked as autoloaded will\nbe written.\n\nNote the second form cannot be used for compiling functions that\ninclude redirections as part of the definition rather than\nwithin the body of the function; for example\n\nfn1() { { ... } >~/logfile }\n\ncan be compiled but\n\nfn1() { ... } >~/logfile\n\ncannot. It is possible to use the first form of zcompile to\ncompile autoloadable functions that include the full function\ndefinition instead of just the body of the function.\n\nThe third form, with the -t option, examines an existing com-\npiled file. Without further arguments, the names of the origi-\nnal files compiled into it are listed. The first line of output\nshows the version of the shell which compiled the file and how\nthe file will be used (i.e. by reading it directly or by mapping\nit into memory). With arguments, nothing is output and the re-\nturn status is set to zero if definitions for all names were\nfound in the compiled file, and non-zero if the definition for\nat least one name was not found.\n\nOther options:\n\n-U Aliases are not expanded when compiling the named files.\n\n-R When the compiled file is read, its contents are copied\ninto the shell's memory, rather than memory-mapped (see\n-M). This happens automatically on systems that do not\nsupport memory mapping.\n\nWhen compiling scripts instead of autoloadable functions,\nit is often desirable to use this option; otherwise the\nwhole file, including the code to define functions which\nhave already been defined, will remain mapped, conse-\nquently wasting memory.\n\n-M The compiled file is mapped into the shell's memory when\nread. This is done in such a way that multiple instances\nof the shell running on the same host will share this\nmapped file. If neither -R nor -M is given, the zcompile\nbuiltin decides what to do based on the size of the com-\npiled file.\n\n-k\n-z These options are used when the compiled file contains\nfunctions which are to be autoloaded. If -z is given, the\nfunction will be autoloaded as if the KSHAUTOLOAD option\nis not set, even if it is set at the time the compiled\nfile is read, while if the -k is given, the function will\nbe loaded as if KSHAUTOLOAD is set. These options also\ntake precedence over any -k or -z options specified to\nthe autoload builtin. If neither of these options is\ngiven, the function will be loaded as determined by the\nsetting of the KSHAUTOLOAD option at the time the com-\npiled file is read.\n\nThese options may also appear as many times as necessary\nbetween the listed names to specify the loading style of\nall following functions, up to the next -k or -z.\n\nThe created file always contains two versions of the com-\npiled format, one for big-endian machines and one for\nsmall-endian machines. The upshot of this is that the\ncompiled file is machine independent and if it is read or\nmapped, only one half of the file is actually used (and\nmapped).\n\nzformat\nSee the section `The zsh/zutil Module' in zshmodules(1).\n\nzftp See the section `The zsh/zftp Module' in zshmodules(1).\n\nzle See the section `Zle Builtins' in zshzle(1).\n\nzmodload [ -dL ] [ -s ] [ ... ]\nzmodload -F [ -alLme -P param ] module [ [+-]feature ... ]\nzmodload -e [ -A ] [ ... ]\nzmodload [ -a [ -bcpf [ -I ] ] ] [ -iL ] ...\nzmodload -u [ -abcdpf [ -I ] ] [ -iL ] ...\nzmodload -A [ -L ] [ modalias[=module] ... ]\nzmodload -R modalias ...\nPerforms operations relating to zsh's loadable modules. Loading\nof modules while the shell is running (`dynamical loading') is\nnot available on all operating systems, or on all installations\non a particular operating system, although the zmodload command\nitself is always available and can be used to manipulate modules\nbuilt into versions of the shell executable without dynamical\nloading.\n\nWithout arguments the names of all currently loaded binary mod-\nules are printed. The -L option causes this list to be in the\nform of a series of zmodload commands. Forms with arguments\nare:\n\nzmodload [ -is ] name ...\nzmodload -u [ -i ] name ...\nIn the simplest case, zmodload loads a binary module.\nThe module must be in a file with a name consisting of\nthe specified name followed by a standard suffix, usually\n`.so' (`.sl' on HPUX). If the module to be loaded is al-\nready loaded the duplicate module is ignored. If zmod-\nload detects an inconsistency, such as an invalid module\nname or circular dependency list, the current code block\nis aborted. If it is available, the module is loaded if\nnecessary, while if it is not available, non-zero status\nis silently returned. The option -i is accepted for com-\npatibility but has no effect.\n\nThe named module is searched for in the same way a com-\nmand is, using $modulepath instead of $path. However,\nthe path search is performed even when the module name\ncontains a `/', which it usually does. There is no way\nto prevent the path search.\n\nIf the module supports features (see below), zmodload\ntries to enable all features when loading a module. If\nthe module was successfully loaded but not all features\ncould be enabled, zmodload returns status 2.\n\nIf the option -s is given, no error is printed if the\nmodule was not available (though other errors indicating\na problem with the module are printed). The return sta-\ntus indicates if the module was loaded. This is appro-\npriate if the caller considers the module optional.\n\nWith -u, zmodload unloads modules. The same name must be\ngiven that was given when the module was loaded, but it\nis not necessary for the module to exist in the file sys-\ntem. The -i option suppresses the error if the module is\nalready unloaded (or was never loaded).\n\nEach module has a boot and a cleanup function. The mod-\nule will not be loaded if its boot function fails. Simi-\nlarly a module can only be unloaded if its cleanup func-\ntion runs successfully.\n\nzmodload -F [ -almLe -P param ] module [ [+-]feature ... ]\nzmodload -F allows more selective control over the fea-\ntures provided by modules. With no options apart from\n-F, the module named module is loaded, if it was not al-\nready loaded, and the list of features is set to the re-\nquired state. If no features are specified, the module\nis loaded, if it was not already loaded, but the state of\nfeatures is unchanged. Each feature may be preceded by a\n+ to turn the feature on, or - to turn it off; the + is\nassumed if neither character is present. Any feature not\nexplicitly mentioned is left in its current state; if the\nmodule was not previously loaded this means any such fea-\ntures will remain disabled. The return status is zero if\nall features were set, 1 if the module failed to load,\nand 2 if some features could not be set (for example, a\nparameter couldn't be added because there was a different\nparameter of the same name) but the module was loaded.\n\nThe standard features are builtins, conditions, parame-\nters and math functions; these are indicated by the pre-\nfix `b:', `c:' (`C:' for an infix condition), `p:' and\n`f:', respectively, followed by the name that the corre-\nsponding feature would have in the shell. For example,\n`b:strftime' indicates a builtin named strftime and\np:EPOCHSECONDS indicates a parameter named EPOCHSECONDS.\nThe module may provide other (`abstract') features of its\nown as indicated by its documentation; these have no pre-\nfix.\n\nWith -l or -L, features provided by the module are\nlisted. With -l alone, a list of features together with\ntheir states is shown, one feature per line. With -L\nalone, a zmodload -F command that would cause enabled\nfeatures of the module to be turned on is shown. With\n-lL, a zmodload -F command that would cause all the fea-\ntures to be set to their current state is shown. If one\nof these combinations is given with the option -P param\nthen the parameter param is set to an array of features,\neither features together with their state or (if -L alone\nis given) enabled features.\n\nWith the option -L the module name may be omitted; then a\nlist of all enabled features for all modules providing\nfeatures is printed in the form of zmodload -F commands.\nIf -l is also given, the state of both enabled and dis-\nabled features is output in that form.\n\nA set of features may be provided together with -l or -L\nand a module name; in that case only the state of those\nfeatures is considered. Each feature may be preceded by\n+ or - but the character has no effect. If no set of\nfeatures is provided, all features are considered.\n\nWith -e, the command first tests that the module is\nloaded; if it is not, status 1 is returned. If the mod-\nule is loaded, the list of features given as an argument\nis examined. Any feature given with no prefix is simply\ntested to see if the module provides it; any feature\ngiven with a prefix + or - is tested to see if is pro-\nvided and in the given state. If the tests on all fea-\ntures in the list succeed, status 0 is returned, else\nstatus 1.\n\nWith -m, each entry in the given list of features is\ntaken as a pattern to be matched against the list of fea-\ntures provided by the module. An initial + or - must be\ngiven explicitly. This may not be combined with the -a\noption as autoloads must be specified explicitly.\n\nWith -a, the given list of features is marked for au-\ntoload from the specified module, which may not yet be\nloaded. An optional + may appear before the feature\nname. If the feature is prefixed with -, any existing\nautoload is removed. The options -l and -L may be used\nto list autoloads. Autoloading is specific to individual\nfeatures; when the module is loaded only the requested\nfeature is enabled. Autoload requests are preserved if\nthe module is subsequently unloaded until an explicit\n`zmodload -Fa module -feature' is issued. It is not an\nerror to request an autoload for a feature of a module\nthat is already loaded.\n\nWhen the module is loaded each autoload is checked\nagainst the features actually provided by the module; if\nthe feature is not provided the autoload request is\ndeleted. A warning message is output; if the module is\nbeing loaded to provide a different feature, and that au-\ntoload is successful, there is no effect on the status of\nthe current command. If the module is already loaded at\nthe time when zmodload -Fa is run, an error message is\nprinted and status 1 returned.\n\nzmodload -Fa can be used with the -l, -L, -e and -P op-\ntions for listing and testing the existence of autoload-\nable features. In this case -l is ignored if -L is spec-\nified. zmodload -FaL with no module name lists autoloads\nfor all modules.\n\nNote that only standard features as described above can\nbe autoloaded; other features require the module to be\nloaded before enabling.\n\nzmodload -d [ -L ] [ name ]\nzmodload -d name dep ...\nzmodload -ud name [ dep ... ]\nThe -d option can be used to specify module dependencies.\nThe modules named in the second and subsequent arguments\nwill be loaded before the module named in the first argu-\nment.\n\nWith -d and one argument, all dependencies for that mod-\nule are listed. With -d and no arguments, all module de-\npendencies are listed. This listing is by default in a\nMakefile-like format. The -L option changes this format\nto a list of zmodload -d commands.\n\nIf -d and -u are both used, dependencies are removed. If\nonly one argument is given, all dependencies for that\nmodule are removed.\n\nzmodload -ab [ -L ]\nzmodload -ab [ -i ] name [ builtin ... ]\nzmodload -ub [ -i ] builtin ...\nThe -ab option defines autoloaded builtins. It defines\nthe specified builtins. When any of those builtins is\ncalled, the module specified in the first argument is\nloaded and all its features are enabled (for selective\ncontrol of features use `zmodload -F -a' as described\nabove). If only the name is given, one builtin is de-\nfined, with the same name as the module. -i suppresses\nthe error if the builtin is already defined or au-\ntoloaded, but not if another builtin of the same name is\nalready defined.\n\nWith -ab and no arguments, all autoloaded builtins are\nlisted, with the module name (if different) shown in\nparentheses after the builtin name. The -L option\nchanges this format to a list of zmodload -a commands.\n\nIf -b is used together with the -u option, it removes\nbuiltins previously defined with -ab. This is only pos-\nsible if the builtin is not yet loaded. -i suppresses\nthe error if the builtin is already removed (or never ex-\nisted).\n\nAutoload requests are retained if the module is subse-\nquently unloaded until an explicit `zmodload -ub builtin'\nis issued.\n\nzmodload -ac [ -IL ]\nzmodload -ac [ -iI ] name [ cond ... ]\nzmodload -uc [ -iI ] cond ...\nThe -ac option is used to define autoloaded condition\ncodes. The cond strings give the names of the conditions\ndefined by the module. The optional -I option is used to\ndefine infix condition names. Without this option prefix\ncondition names are defined.\n\nIf given no condition names, all defined names are listed\n(as a series of zmodload commands if the -L option is\ngiven).\n\nThe -uc option removes definitions for autoloaded condi-\ntions.\n\nzmodload -ap [ -L ]\nzmodload -ap [ -i ] name [ parameter ... ]\nzmodload -up [ -i ] parameter ...\nThe -p option is like the -b and -c options, but makes\nzmodload work on autoloaded parameters instead.\n\nzmodload -af [ -L ]\nzmodload -af [ -i ] name [ function ... ]\nzmodload -uf [ -i ] function ...\nThe -f option is like the -b, -p, and -c options, but\nmakes zmodload work on autoloaded math functions instead.\n\nzmodload -a [ -L ]\nzmodload -a [ -i ] name [ builtin ... ]\nzmodload -ua [ -i ] builtin ...\nEquivalent to -ab and -ub.\n\nzmodload -e [ -A ] [ string ... ]\nThe -e option without arguments lists all loaded modules;\nif the -A option is also given, module aliases corre-\nsponding to loaded modules are also shown. If arguments\nare provided, nothing is printed; the return status is\nset to zero if all strings given as arguments are names\nof loaded modules and to one if at least on string is not\nthe name of a loaded module. This can be used to test\nfor the availability of things implemented by modules.\nIn this case, any aliases are automatically resolved and\nthe -A flag is not used.\n\nzmodload -A [ -L ] [ modalias[=module] ... ]\nFor each argument, if both modalias and module are given,\ndefine modalias to be an alias for the module module. If\nthe module modalias is ever subsequently requested, ei-\nther via a call to zmodload or implicitly, the shell will\nattempt to load module instead. If module is not given,\nshow the definition of modalias. If no arguments are\ngiven, list all defined module aliases. When listing, if\nthe -L flag was also given, list the definition as a\nzmodload command to recreate the alias.\n\nThe existence of aliases for modules is completely inde-\npendent of whether the name resolved is actually loaded\nas a module: while the alias exists, loading and unload-\ning the module under any alias has exactly the same ef-\nfect as using the resolved name, and does not affect the\nconnection between the alias and the resolved name which\ncan be removed either by zmodload -R or by redefining the\nalias. Chains of aliases (i.e. where the first resolved\nname is itself an alias) are valid so long as these are\nnot circular. As the aliases take the same format as\nmodule names, they may include path separators: in this\ncase, there is no requirement for any part of the path\nnamed to exist as the alias will be resolved first. For\nexample, `any/old/alias' is always a valid alias.\n\nDependencies added to aliased modules are actually added\nto the resolved module; these remain if the alias is re-\nmoved. It is valid to create an alias whose name is one\nof the standard shell modules and which resolves to a\ndifferent module. However, if a module has dependencies,\nit will not be possible to use the module name as an\nalias as the module will already be marked as a loadable\nmodule in its own right.\n\nApart from the above, aliases can be used in the zmodload\ncommand anywhere module names are required. However,\naliases will not be shown in lists of loaded modules with\na bare `zmodload'.\n\nzmodload -R modalias ...\nFor each modalias argument that was previously defined as\na module alias via zmodload -A, delete the alias. If any\nwas not defined, an error is caused and the remainder of\nthe line is ignored.\n\nNote that zsh makes no distinction between modules that were\nlinked into the shell and modules that are loaded dynamically.\nIn both cases this builtin command has to be used to make avail-\nable the builtins and other things defined by modules (unless\nthe module is autoloaded on these definitions). This is true\neven for systems that don't support dynamic loading of modules.\n\nzparseopts\nSee the section `The zsh/zutil Module' in zshmodules(1).\n\nzprof See the section `The zsh/zprof Module' in zshmodules(1).\n\nzpty See the section `The zsh/zpty Module' in zshmodules(1).\n\nzregexparse\nSee the section `The zsh/zutil Module' in zshmodules(1).\n\nzsocket\nSee the section `The zsh/net/socket Module' in zshmodules(1).\n\nzstyle See the section `The zsh/zutil Module' in zshmodules(1).\n\nztcp See the section `The zsh/net/tcp Module' in zshmodules(1).\n\nzsh 5.8.1 February 12, 2022 ZSHBUILTINS(1)", "subsections": [] } } } }