for

shell loop construct for iteration

TLDR

Iterate over an explicit list

$ for item in [a b c]; do echo "$item"; done

Loop over matching files

$ for file in *.txt; do cat "$file"; done

Loop over a numeric range with brace expansion

$ for i in {1..10}; do echo "$i"; done

Loop with a C-style counter (bash, ksh, zsh)

$ for ((i=0; i<10; i++)); do echo "$i"; done

Iterate over the positional parameters

$ for arg do echo "$arg"; done

Iterate over array elements (bash, ksh, zsh)

$ for x in "${[array][@]}"; do echo "$x"; done

Loop over command output, one word per iteration

$ for user in $(cut -d: -f1 [/etc/passwd]); do echo "$user"; done

SYNOPSIS

for name [in words]; do commands; donefor (( init; condition; update )); do commands; done

for iterates over a list of words, assigning each to a variable and running the body once per word. It is a reserved word handled by the shell's parser, not an external command.The word list is produced by the shell's normal expansions, which is where its flexibility comes from: brace expansion ({1..10}) generates sequences, pathname expansion (*.txt) generates filenames, command substitution ($(...)) turns another program's output into items, and "${array[@]}" yields array elements. Crucially the list is fully expanded before the first iteration, so it is a fixed set of words rather than a stream.Omitting in entirely is a useful shorthand: for arg do ... done loops over "$@", the script's or function's arguments, correctly preserving arguments that contain spaces.The C-style form for ((init; condition; update)) is a separate construct that does arithmetic rather than word iteration. It is not POSIX, but is available in bash, ksh and zsh, and is the clearer choice when you need a counter with a computed bound.The loop's exit status is that of the last command executed, or 0 if the body never ran. break exits the loop and continue skips to the next item.

PARAMETERS

name

Variable assigned each item in turn. It stays set to the last value after the loop.

words

The list to iterate over, after all expansions. If in words is omitted, the loop iterates over the positional parameters, as if in "$@" had been written.

commands

The body, executed once per item.

Introduces the word list.

Begins the loop body.

done

Ends the loop.

CAVEATS

An unquoted $(command) in the word list is split on whitespace, not on lines, and each resulting word is then glob-expanded. This makes for line in $(cat file) wrong for any file whose lines contain spaces or wildcards: use while IFS= read -r line; do ...; done < file to read line by line.Always quote the variable inside the body ("$item"), or filenames with spaces will be split again when you use them.If a glob matches nothing, the shell leaves the pattern itself in the list by default, so **for f in *.txt runs once with the literal string *.txt in an empty directory. Guard with [ -e "$f" ], or set shopt -s nullglob** in bash.Because the list is expanded up front, a for loop over a very large glob or command output builds the entire list in memory first. For huge inputs prefer a while read loop or find -exec / xargs.The loop variable leaks: it remains set after done and is not scoped to the loop.

HISTORY

for dates to the Bourne shell (1977) and is standardized by POSIX. The arithmetic for ((...)) form came from the Korn shell and was adopted by bash in version 2.04, offering C programmers a familiar counting loop.

RESOURCES

Documentation

for