Skip to content

CLAUDE.md

Latest commit

 

History

History
95 lines (65 loc) · 2.76 KB

CLAUDE.md

File metadata and controls

95 lines (65 loc) · 2.76 KB
Raw

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Project Overview

mojibox is a CLI tool for flexible Unicode string manipulation and analysis. It supports processing text at three different levels:

  • Grapheme clusters (書記素クラスター) - default mode
  • Unicode code points (char mode)
  • Byte level (byte mode)

The tool is designed for handling diverse Japanese and multilingual text use cases.

Core Commands

Subcommands

  • iter: Expand strings one unit at a time by specified mode
  • len: Count string length by specified mode
  • take: Extract N units from the beginning
  • drop: Skip N units from the beginning and extract the rest

Options

  • --mode, -m: grapheme (default), char, byte
  • --engine, -e: unicode (unicode-segmentation), icu4x (icu_segmenter)

Development Commands

Since this is a Rust project, use standard Cargo commands:

# Build the project
cargo build

# Run tests
cargo test

# Run with arguments
cargo run -- <subcommand> <args>

# Build optimized release
cargo build --release

# Check code without building
cargo check

# Format code
cargo fmt

# Run linter
cargo clippy

Key Dependencies (as specified)

  • unicode-segmentation: For grapheme cluster, word, and sentence segmentation (UAX #29)
  • icu_segmenter (icu4x): For Unicode standard compliance and accurate multilingual segmentation
  • clap: For command-line argument parsing and subcommand management
  • anyhow: For simple error handling
  • thiserror: For custom error implementations
  • encoding_rs: For character encoding conversion when needed

Architecture Notes

The tool processes text input through different segmentation modes:

  1. Grapheme mode: Uses Unicode grapheme cluster boundaries (handles emoji, combining characters correctly)
  2. Char mode: Processes by Unicode code points
  3. Byte mode: Processes by individual bytes

Engine selection allows switching between unicode-segmentation and icu4x implementations for different accuracy/performance trade-offs.

Development Considerations

  • Handle UTF-8 incompatible systems/terminals gracefully
  • Be aware of segmentation differences between unicode-segmentation and icu4x
  • Consider performance and binary size implications, especially with ICU4X
  • Test with complex Unicode text including Japanese, emoji, and combining characters

Commit Guidelines

Follow Conventional Commits format:

<type>[optional scope]: <description>

[optional body]

[optional footer(s)]

Types: feat, fix, docs, style, refactor, test, chore

Examples:

  • feat: add iter subcommand for grapheme mode
  • fix: handle empty string input in len command
  • docs: update usage examples in spec.md