kombu v0.2.0

Kombu

Kombu simplifies interacting with tabulated CLI data. Pipe the output of a shell command to kombu, and apply javascript functions to it. It's like using awk, without having to learn awk.

Example

ls | kb "data.map((n) => n.concat(' ✨'))"
README.md ✨
node_modules ✨
package.json ✨
src ✨
yarn.lock ✨

Many shell commands output tabulated data:

$ df
Filesystem    512-blocks      Used Available Capacity iused      ifree %iused  Mounted on
/dev/disk1     487653376 110844640 376296736    23%  876686 4294090593    0%   /
devfs                370       370         0   100%     640          0  100%   /dev

Tabulated data piped to kombu (which I've aliased to kb) is read into a 2D array named data. The user supplies javascript statements to transform the data. In this example, the full data set is returned by supplying the statement data, and kb logs it automatically.

$ df | kb "data"
Filesystem 512-blocks Used Available Capacity iused ifree %iused Mounted on
/dev/disk1 487653376 110841568 376299808 23% 876685 4294090594 0% /
devfs 370 370 0 100% 640 0 100% /dev

Say we want to get a list of File System names. First, we slice data to remove the columb titles (note kombu can remove titles automatically with the -t, --title flag):

$ df | kb "data.slice(1)"
/dev/disk1 487653376 110841896 376299480 23% 876666 4294090613 0% /
devfs 370 370 0 100% 640 0 100% /dev

Then, we apply a map to the array to select the first element of each row:

$ df | kb "data.slice(1)" | kb "data.map((d) => d[0])"
/dev/disk1
devfs

IO

Kombu aims to provide an intuitive experience. The type of data should be what you intuitively expect it to be:

• single strings -> strings
• 1D lists -> array
• 2D tables -> nested array

Likewise, output is printed in a sensible format.

API

Usage: <command> | kombu [options] statement

Kombu is a command-line tool for manipulating tabulated data.

Tabulated data should be piped to kombu, and a statement which transforms the
data should be supplied. The data is parsed, and made available to the statment
via the variable 'data'.

Options:
    -h, --help          Show this message and exit
    -t, --title         Remove titles, assumed to be at row 0

Alternatives

• awk
• FuncShell, which kombu was inspired by