choose

choose is a simple, expressive alternative to cut and basic awk field-selection one-liners. It uses an intuitive syntax for selecting fields and ranges from lines of text, and is fast enough to handle large streams without breaking a sweat.

Features

• Intuitive field selection — select fields by index using familiar Python-like range syntax
• Negative indexing — count fields from the end with negative indices
• Custom delimiters — split on any regex pattern, not just whitespace or a single character
• Output delimiter control — join selected fields with any separator you choose
• Rust-speed — processes large inputs significantly faster than equivalent awk scripts

Installation

cargo install choose

Or via your package manager:

# macOS
brew install choose-rust

# Arch Linux
yay -S choose

# Debian / Fedora
# Pre-built Linux binaries are available on the
# [releases page](https://github.com/theryangeary/choose/releases).

Usage

# Select the first field (0-indexed) from each line
echo "one two three" | choose 0
# → one

# Select the third field
echo "one two three" | choose 2
# → three

# Select a range of fields (inclusive)
echo "one two three four" | choose 1:3
# → two three four

# Select from field 1 to the end
echo "one two three four" | choose 1:
# → two three four

# Select the last field using negative indexing
echo "one two three four" | choose -1
# → four

# Use a custom delimiter (split on commas)
echo "a,b,c,d" | choose -f ',' 2
# → c

# Use a custom output delimiter
echo "one two three" | choose -o ',' 0:2
# → one,two,three

# Practical example: get the second column of ps output
ps aux | choose 1

Comparison with cut

# cut requires 1-based indexing and a specific delimiter flag
cut -d' ' -f2 file.txt

# choose is more natural
choose 1 < file.txt

choose really shines in interactive shell sessions where you just want to grab a column quickly without remembering awk's arcane field syntax or cut's 1-based indexing.