How to Install and Use choose in Linux for Fast and Human-Friendly Field and Column Cutting

Manipulating delimited tabular data, log files, and stream outputs on the Linux command line has long relied on utilities like cut and awk. However, both venerable tools exhibit noticeable ergonomic drawbacks. Standard cut cannot split on multiple whitespace characters without collapsing them first, lacks support for negative indexing from the end of a line, and cannot reorder fields. Meanwhile, writing multi-argument awk '{print $1, $3}' statements feels verbose for quick terminal one-liners. To solve these friction points, developers built choose. Written in Rust, choose is a fast, human-friendly command-line tool that brings Python-style slice syntax and regex-based field separation to shell data processing.

Why choose Outperforms Standard cut and awk

The choose command was engineered specifically to address common daily pain points encountered when parsing Unix command outputs:

• Python/Rust Slice Syntax: Slice fields using intuitive indices, ranges, and negative steps (e.g. choose 0 for the first column, choose -1 for the final column, or choose 2:5 for a range).
• Intelligent Whitespace Handling: By default, consecutive tabs and spaces are collapsed into single delimiters, meaning command outputs like ps aux or ls -l can be sliced immediately without piping through tr -s ' '.
• Field Reordering: Output columns in any custom sequence (such as choose 3 1 0), a capability impossible with standard cut.
• Arbitrary Character Slicing: Toggle character mode (-c) to extract specific letter ranges from lines.
• High Throughput: Benchmarks confirm choose matches or exceeds standard GNU cut throughput on multi-gigabyte dataset streams.

Installing choose on Linux

You can install choose using package managers, Cargo, or precompiled standalone binaries.

Method 1: Installing via Cargo (Universal)

If you have Rust and Cargo installed, compile the latest release directly from crates.io:

cargo install choose

Ensure ~/.cargo/bin is in your active shell $PATH:

choose --version

Method 2: Installing Precompiled Binaries from GitHub

For production servers without compiler tools, download the pre-built tarball from GitHub Releases:

VERSION=$(curl -s https://api.github.com/repos/theryangeary/choose/releases/latest | grep '"tag_name":' | sed -E 's/.*"([^"]+)".*/\\1/')
wget "https://github.com/theryangeary/choose/releases/download/${VERSION}/choose-x86_64-unknown-linux-musl.tar.gz"
tar -xzf choose-x86_64-unknown-linux-musl.tar.gz
sudo mv choose /usr/local/bin/
sudo chmod +x /usr/local/bin/choose

Method 3: Package Managers (Arch Linux & Homebrew)

On Arch Linux:

yay -S choose

On systems using Linuxbrew:

brew install choose-rust

Basic Field Slicing and Negative Indexing

Using choose is intuitive. It indexes fields starting from zero (0). Piping tabular data into choose extracts specific tokens immediately.

Extracting Single Columns and Last Columns

Consider piping the output of df -h. To isolate filesystem mount points (the final column) and device paths (the first column):

df -h | choose 0 -1

Notice how -1 dynamically references the very last item on every line, regardless of whether lines have varying column counts.

Range Slicing

Choose accepts familiar slice bounds start:end (where end is exclusive, matching Python conventions):

# Extract columns 1 through 3
cat access.log | choose 1:4

# Extract everything from column 2 onwards to the end of the line
cat access.log | choose 2:

# Extract everything up to column 3
cat access.log | choose :3

Specifying Custom Delimiters and Output Separators

While choose defaults to collapsing whitespace, you can define arbitrary input and output delimiters for structured formats like CSV or TSV.

Parsing CSV Data

Use -f (field separator) to set an explicit character or regex delimiter:

choose -f ',' 0 2 4 < users.csv

Customising Output Delimiters

By default, extracted fields are output separated by spaces. You can specify a custom output delimiter using -o:

# Read comma-separated values and output them tab-separated
choose -f ',' -o '	' 0 1 3 < data.csv

Character Slicing with Character Mode

To slice fixed-width strings by character positions rather than words or delimiters, pass the -c flag:

# Extract the first 10 characters of every line
cat hashes.txt | choose -c :10

# Extract the final 8 characters of every line
cat logs.txt | choose -c -8:

Real-World Pipeline Examples

Here is how choose simplifies everyday system administration pipelines:

1. Inspecting Top Memory Consuming Processes

ps aux --sort=-%mem | choose 1 2 3 -1 | head -n 10

Prints process IDs, CPU percentages, memory percentages, and executable command paths cleanly aligned.

2. Parsing IP Addresses from Network Interfaces

ip -o -4 addr show | choose 1 3

Instantly extracts network adapter names alongside their assigned CIDR network addresses without tedious regex formatting.