A command line tool to select columns from a table-shaped input, where columns are separated by multiple whitespaces (or a delimiter of your choice). Supports filtering and formatting. Run table without any arguments to see available options.

Quick example:

• -h will treat the first line as the header and allow selecting the columns by name
• -s to choose which columns to output

docker images
#REPOSITORY          TAG       IMAGE ID       CREATED        SIZE
#bitnami/kafka       3.2       b7add9628c8e   4 weeks ago    657MB
#bitnami/zookeeper   3.8       000e247c0e4c   4 weeks ago    477MB
#openjdk             11.0      23d35e2be72f   7 weeks ago    650MB

docker images | table -h -s "TAG,IMAGE ID"

#output:
#3.2
#b7add9628c8e
#3.8
#000e247c0e4c
#11.0
#23d35e2be72f

1. regular select: -s/--select. select the columns by name (if --header flag is on) or by the index of the column.
2. excluding select: -S/--exclude-select selects all the columns except the requested.
3. projection: -p/--project allows to form an output string with placeholders for the requested columns. for example:

docker images | table -h -p "the image {REPOSITORY} is of size {SIZE}"

will output

#the image bitnami/kafka is of size 657MB
#the image bitnami/zookeeper is of size 477MB
#the image openjdk is of size 650MB"

You can filter only the rows you care about with -f/--filter. The filter works by searching if the cell includes the searched substring.

docker images:
#REPOSITORY          TAG       IMAGE ID       CREATED        SIZE
#bitnami/kafka       3.2       b7add9628c8e   4 weeks ago    657MB
#bitnami/zookeeper   3.8       000e247c0e4c   4 weeks ago    477MB
#openjdk             11.0      23d35e2be72f   7 weeks ago    650MB

docker images | table -h -s "TAG,IMAGE ID" -f "REPOSITORY=bitnami"

#output:
#3.2
#b7add9628c8e
#3.8
#000e247c0e4c

To split the columns by a different delimiter, such as dashes (-), you can use -d/--delimiter flag to choose a different character than whitespace. In the following example the whitespaces has been replaced by dashes we declare dash to be the delimiter:

docker images
#REPOSITORY----------TAG----IMAGE ID-------CREATED-------SIZE
#bitnami/kafka-------3.2----b7add9628c8e---4 weeks ago---657MB
#bitnami/zookeeper---3.8----000e247c0e4c---4 weeks ago---477MB
#openjdk-------------11.0---23d35e2be72f---7 weeks ago---650MB

docker images | table -h -s "TAG,IMAGE ID" -d -

#output:
#3.2
#b7add9628c8e
#3.8
#000e247c0e4c

To change the amount of times the delimiter has to show up consecutively to split the columns, you can use the -r/--delimiter-repeats flag. The following example shows a table input that is separated by a single tab (\t), so we will declare the delimiter to be the tab symbol, and the repetitions to be 1:

cat example.tsv
#CONST  123456  12.45

cat example.tsv | table -s 1 -d $'\t' -r 1

#output:
#123456

Note on special character escaping: to properly pass special characters like tab (\t) or new line (\n), it needs to be interpreted by the shell like so: $'\t'.

table doesn't handle properly a specific case: Right-justified column headers, for example notice the TIME column:

 ps -ef
 #  UID   PID  PPID   C STIME   TTY           TIME CMD
 #    0     1     0   0  9Mar23 ??       336:12.65 /sbin/launchd
 #    0   323     1   0  9Mar23 ??        43:20.14 /usr/libexec/logd

This case would not catch the width of the TIME column properly

While table can work with a JVM because it's written in Scala, it's intended to be a fast CLI tool, so it should be compiled to a native binary. The build requires Clang/LLVM installed for Scala Native compilation. You can use SDKMAN to install JVM-related dependencies

1. Scala-CLI or using SDKMAN: sdk install scalacli
2. Clang/LLVM - see Scala Native's instructions

From the root directory

scala-cli --power package --native .  -S 3.4.1 -o table

will create an executable named table

scala-cli test . --native