Rut

rut is a tool for selecting bytes, characters, or fields from files. It is heavily based on cut, but with additional support for delimiting fields with a regular expression. Note that there are multiple implementations of cut which differ in various ways. rut is primarily based on the GNU Coreutils implementation.

Below is a table briefly describing rut's features and how it compares to cut.

Option	Description	POSIX cut 1	GNU Coreutils cut	rut
-b	Select bytes.	✔	✔ (also supports --bytes)	✔ (also supports --bytes)
-c	Select characters.	✔	⚠ (also supports --characters; behaves the same as -b)	✔ (also supports --characters; requires UTF-8 input)
-f	Select fields (strings separated by a delimiter).	✔	✔ (also supports --fields; treats each byte as a character, without regard for encoding)	✔ (also supports --fields; requires UTF-8 input)
-d	Specify a single character delimiter when used with -f.	✔	⚠ (also supports --delimiter; requires single byte character)	✔ (also supports --delimiter; must be a UTF-8 character)
-s	Do not print lines without a delimiter. Normal behavior is to print the full line.	✔	✔ (also supports --only-delimited)	✔ (also supports --only-delimited)
-n	Do not split multi-byte characters when used with -b.	✔	⚠ (no-op)	⚠ (no-op)
--output-delimiter	Specify a string to delimit selected fields. Normal behavior is to use delimiter.	❌	✔	✔ (also supports -o)
--complement	Select the complement of selected bytes/characters/fields.	❌	✔	✔
-z / --zero-terminated	Delimit "lines" with a zero byte rather than a newline	❌	✔	✔
-r / --regex-delimiter	Specific a regular expression as a delimiter when used with -f.	❌	❌	✔

1. This column describes the POSIX definition of cut and not any particular implementation.

Use

rut is intended to be a drop-in replacement for the GNU Coreutils implementation of cut in many cases. In particular, you should be able to replace cut with rut for any valid cut command with ASCII input. It is considered a bug if it does not.

rut also adds support for an additional options and UTF-8 encoded input. This has the following consequences.

• Some cut commands which would fail due to invalid or unrecognized options will pass with rut.
• The output for commands using -c or -f will be different for non-ASCII input.

Examples

Select bytes from a file:

$ rut -b1-5 tests/files/ascii.txt
abcde
a b c
a_b_c
a:b:c

Select from stdin:

$ cat tests/files/ascii.txt | rut -b1-5
abcde
a b c
a_b_c
a:b:c

Comparison of bytes versus characters:

$ rut -b1-4 tests/files/utf8.txt
abcd
αβ
abα
😀
$ rut -c1-4 tests/files/utf8.txt
abcd
αβγδ
abαβ
😀😁😂😃

Select fields:

$ rut -f1 -dd tests/files/ascii.txt
abc
a b c 
a_b_c_
$ rut -f2,4,6 -d'_' -s tests/files/ascii.txt
b_d_f

Select fields with regex delimiter:

$ rut -f 2-4,6-8 -r '[ _:]' -s -o# tests/files/ascii.txt
b#c#d#f#g#h
b#c#d#f#g#h
b#c#d#f#g#h

Test and Build

rut is written in Rust. It has been tested with Rust 1.45.2 but may work with earlier or later versions. The following instructions assume you have Rust installed.

To run the full test suite, run:

$ cargo test

To build rut in debug mode (faster compile time, slower executable), use:

$ cargo build

For release mode (slower compile times, faster executable), use:

$ cargo build --release

Package for Release

Build scripts are provided to create a release package. Follow the instructions below to build an archive (.zip or .tar.gz) in the target directory.

Windows

$ .\Create-Release.ps1

Linux

$ ./create-release.sh