column - Man Page

• Format the output of a command for a 30 characters wide display: printf "header1 header2\nbar foo\n" | column [-c|--output-width] 30
• Split columns automatically and auto-align them in a tabular format: printf "header1 header2\nbar foo\n" | column [-t|--table]
• Specify the column delimiter character for the --table option (e.g. "," for CSV) (defaults to whitespace): printf "header1,header2\nbar,foo\n" | column [-t|--table] [-s|--separator] ,
• Fill rows before filling columns: printf "header1\nbar\nfoobar\n" | column [-c|--output-width] 30 [-x|--fillrows]

tldr.sh

column [options] [file ...]

The column utility formats its input into multiple columns. It supports three modes:

fill columns before rows

fill rows before columns

create a table

Input is taken from file, or otherwise from standard input. Empty lines are ignored and all invalid multibyte sequences are encoded with the x<hex> convention.

The argument columns, for the --table-* options below, is a comma-separated list of user-supplied names defined with --table-columns name1,name2,..., or indices of columns as they appear in the input, beginning with 1, or names defined through a --table-column attribute. It’s possible to mix names and indices. The special placeholder '0' (e.g. -R0) may be used to specify all columns and '-1' (e.g. -R -1) to specify the last visible column. It’s possible to use ranges like '1-5' when addressing columns by indices.

-J,  --json

Use JSON output format to print the table. The option --table-columns is required and the option --table-name is recommended.

-c,  --output-width width

Output is formatted to a width specified as a number of characters. Note that input longer than width is not truncated by default. The default is the terminal width in interactive mode, and 80 columns in non-interactive mode. The column headers are never truncated.

The placeholder unlimited (or 0) can be used to prevent restricting the output width. This is recommended when redirecting output to a file.

(The original long name of this option was --columns; this name is deprecated since v2.30.)

-d,  --table-noheadings

Omit printing the header. This option allows having user-supplied column names on the command line, but keeps the header hidden when printing the table.

-o,  --output-separator string

Column delimiter for table output (default is two spaces).

-s,  --input-separator,  --separator separators

Possible input-item delimiters (default is whitespace).

-S,  --use-spaces number

When not in table mode, use spaces instead of tabulators to align the columns. This option specifies the minimum number of spaces that separate two columns.

-t,  --table

Determine the number of columns the input contains and create a table. Columns are by default delimited with whitespace, or with characters supplied using the --output-separator option. Table output is useful for pretty-printing.

--table-colorscheme name

Specifies color scheme. The default name is "column". For more details see the Colors section.

-C,  --table-column attributes

Define a column with a comma-separated list of attributes. This option can be used more than once, every use defines a single column. Attributes replace some of the --table-* options: for example, --table-column name=FOO,right defines a column where text is aligned to the right. The option is mutually exclusive with -N/--table-columns.

Supported attributes are:

name=string

Column name.

trunc

Truncate column text when necessary. The same as --table-truncate. Note that the column header is never truncated, and it is the minimal column width.

tree

Use this column for tree-like output. See --tree for more details.

right

Right align text. The same as --table-right.

width=number

Column width. It’s used only as a hint. To enforce it, specify the strictwidth attribute as well. The actual minimum width is the column header width.

strictwidth

Strictly follow the width= to set the minimum size of the column. The column can be larger if the output width provides enough space.

noextremes

Ignore unusually long content width. See --table-noextreme for more details.

hidden

Don’t print the column. See --table-hide for more details.

wrap

Allow the use of a multi-line cell for long text if necessary. Note that the column header is never wrapped and determines the minimum column width. See --table-wrap for more details.

wrapnl

Wrap at newline characters; newlines in cell data are treated as line breaks.

wrapzero

Wrap at zero (NUL) characters in cell data.

json=type

Define the column type for JSON output. Supported types are string, number, float, array-string, array-number and boolean.

color=name

Define the column color for output on the terminal. The name is a color name (black, blink, blue, bold, brown, cyan, darkgray, gray, green, halfbright, lightblue, lightcyan, lightgray, lightgreen, lightmagenta, lightred, magenta, red, reset, reverse, and yellow) or ANSI color sequence number(s) separated by a semicolon, but without the 'ESC[' prefix and 'm' suffix. For example, "37;41" defines sequences for a red background and white foreground. For more details see the Colors section.

colorkey=name

Addresses color as defined in the color scheme. See --table-colorscheme.

headercolor=name

Like color=, but define the color for column header only.

headercolorkey=name

Addresses color for header as defined in the color scheme. See --table-colorscheme.

-N,  --table-columns names

Specify column names with a comma-separated list. The names are used for the table header and for column addressing in option arguments. See also --table-column.

-l,  --table-columns-limit number

Specify the maximum number of input columns. The last column will contain all remaining line data if the limit is smaller than the number of columns in the input data.

-R,  --table-right columns

Right align text in the specified columns.

-T,  --table-truncate columns

Specify the columns where text can be truncated when necessary, otherwise very long table entries may be printed on multiple lines.

-E,  --table-noextreme columns

Specify the columns where longer-than-average content should be ignored when calculating the column width. The option influences width calculation and table formatting, but the printed text is not affected.

This option is active by default for the last visible column.

-e,  --table-header-repeat

Print header line for each page.

-K,  --table-header-as-columns

Use the first input line as table header. This option is mutually exclusive with --table-columns and --table-column.

-W,  --table-wrap columns

Specify the columns where multi-line cells can be used for long text. By default, text is wrapped according to column width. Use --wrap-separator to wrap at custom separator characters instead. Note that the column header is never wrapped and determines the minimum column width.

--wrap-separator string

Use string as a separator for wrapping text within columns that have wrapping enabled. The separator is replaced with newlines when the text is displayed. This option requires table mode and columns with wrapping enabled (see --table-wrap). For example, use | to allow wrapping at pipe characters within column data.

-H,  --table-hide columns

Don’t print the specified columns. The special placeholder '-' may be used to hide all unnamed columns (see --table-columns).

-O,  --table-order columns

Specify the output column order.

-n,  --table-name name

Specify the table name used for JSON output. The default is "table".

-m,  --table-maxout

Fill all available space on output.

-L,  --keep-empty-lines

Preserve whitespace-only lines in the input. The default is to ignore all empty lines. This option’s original name was --table-empty-lines, but has since been deprecated because it gives the false impression that the option only applies to table mode.

-r,  --tree column

Specify the column to use for a tree-like output. Note that the circular dependencies and other anomalies in child and parent relation are silently ignored.

-i,  --tree-id column

Specify the column that contains each line’s unique child ID for a child-parent relation.

-p,  --tree-parent column

Specify the column that contains each line’s parent ID for a child-parent relation.

-x,  --fillrows

Fill rows before filling columns.

--color[=when]

Colorize the output. The optional argument when can be auto, never or always. If the when argument is omitted, it defaults to auto. The colors can be disabled; for the current built-in default see the --help output. See also the Colors section.

-h,  --help

Display help text and exit.

-V,  --version

Display version and exit.

The output colorization is implemented by terminal-colors.d(5) functionality. Implicit coloring can be disabled by an empty file

for the column command or for all tools by

Since version 2.41, the $NO_COLOR environment variable is also supported to disable output colorization unless explicitly enabled by a command-line option.

The user-specific $XDG_CONFIG_HOME/terminal-colors.d or $HOME/.config/terminal-colors.d overrides the global setting.

Note that the output colorization may be enabled by default, and in this case terminal-colors.d directories do not have to exist yet.

The default color scheme name is "column," and it can be overridden by

Then name[.disable|enable|scheme] file may be used in /etc/terminal-colors.d/, $XDG_CONFIG_HOME/terminal-colors.d, or $HOME/.config/terminal-colors.d/.

The logical color name (key) in the color scheme may be addressed by column properties colorkey= and headercolorkey=. For example

echo 'important red'           > ~/.config/terminal-colors.d/cooltable.scheme
echo 'important-header 37;41' >> ~/.config/terminal-colors.d/cooltable.scheme
echo -e "a b c\naa bb cc\naaa bbb ccc" | column -t \
        --table-colorscheme=cooltable \
        -C name=AAA,colorkey=important,headercolorkey=important-header \
        -C name=BBB,color=magenta,headercolor=cyan \
        -C name=CCC,color=green

This will create a color scheme with the name "cooltable" and colorize the first column header and data according to keys from the scheme. The other two columns will be colorized by direct color names.

COLUMNS

is used to determine the size of the screen if no other information is available.

LIBSMARTCOLS_DEBUG=all

enables libsmartcols debug output.

LIBSMARTCOLS_DEBUG_PADDING=on

use visible padding characters.

The column command appeared in 4.3BSD-Reno.

Version 2.23 changed the -s option to be non-greedy, for example:

printf "a:b:c\n1::3\n" | column -t -s ':'

Old output:

a  b  c
1  3

New output (since util-linux 2.23):

a  b  c
1     3

Historical versions of this tool indicated that "rows are filled before columns" by default, and that the -x option reverses this. This wording did not reflect the actual behavior, and it has since been corrected (see above). Other implementations of column may continue to use the older documentation, but the behavior should be identical in any case.

Print fstab with a header line and align numbers to the right:

sed 's/#.*//' /etc/fstab | column --table --table-columns SOURCE,TARGET,TYPE,OPTIONS,FREQ,PASS --table-right FREQ,PASS

Print fstab and hide unnamed columns:

sed 's/#.*//' /etc/fstab | column --table --table-columns SOURCE,TARGET,TYPE --table-hide -

Print a tree:

echo -e '1 0 A\n2 1 AA\n3 1 AB\n4 2 AAA\n5 2 AAB' | column --tree-id 1 --tree-parent 2 --tree 3
1  0  A
2  1  |-AA
4  2  | |-AAA
5  2  | `-AAB
3  1  `-AB

Print table with custom wrap separator:

echo -e 'Name:Description\nJohn:A|software|developer\nJane:A|data|scientist' | column --table --separator ':' --table-wrap 2 --wrap-separator '|'
Name  Description
John  A
      software
      developer
Jane  A
      data
      scientist

colrm(1), ls(1), paste(1), sort(1)

For bug reports, use the issue tracker.

The column command is part of the util-linux package which can be downloaded from Linux Kernel Archive.

colrm(1), nisplus_table(5).