Reach - Research Assistant CLI Tool

Reach is a command-line tool designed to enhance academic research by integrating search capabilities from Google, arXiv, and Gemini to provide comprehensive information gathering, summarization, and knowledge graph generation.

Features

• 🔍 Multi-source Search: Integrated search across Google, arXiv, and Gemini
• 📝 Web Content Extraction: Automatically extracts relevant content from web pages
• 📊 Summarization: Generates concise summaries of research content
• 🔗 Knowledge Graph Generation: Creates knowledge graphs to visualize relationships between concepts
• ⚙️ Configurable: Customize search parameters, keywords, and categories
• 🖥️ Terminal UI: Interactive terminal interface for query exploration and knowledge graph visualization
• 📚 Graph Database: High-performance graph database for storing relationship-based data

Prerequisites

• Rust (latest stable version)
• Python 3.12 or higher
• API keys:
  • Google Search API key
  • Google Custom Search Engine ID
  • Gemini API key

Installation

1. Clone the repository

git clone https://github.com/yourusername/reach.git
cd reach

2. Set up Python environment

# Create virtual environment
python -m venv .venv

# Activate virtual environment (Windows)
.venv\Scripts\activate

# Activate virtual environment (Linux/macOS)
source .venv/bin/activate

# Install dependencies
pip install --upgrade pip
pip install -U crawl4ai
crawl4ai-setup

3. Build the Rust project

cargo build --release

Configuration

1. Reach requires several API keys to function properly. You can configure these using the CLI:

# Launch the interactive configuration
cargo run -- config

# Or set each key individually
cargo run -- config --google-api-key YOUR_GOOGLE_API_KEY
cargo run -- config --search-engine-id YOUR_SEARCH_ENGINE_ID
cargo run -- config --gemini-api-key YOUR_GEMINI_API_KEY

2. You can customize your arXiv search preferences:

# Launch the interactive configuration
cargo run -- arxiv-config

# Show current configuration
cargo run -- arxiv-config --show

The arXiv configuration allows you to set:

• Keywords to include in searches
• Keywords to exclude from searches
• Specific authors to focus on
• arXiv categories to search within

Usage

Basic Search

# Search with default parameters
cargo run -- search "Diffusion Models"

# Specify result limit
cargo run -- search "Flow based Diffusion Models" --max-results 5

Generate Summaries

cargo run -- summarize "What are Diffusion Models?"

Generate Knowledge Graph

cargo run -- knowledge-graph "What are Flow based Diffusion Models?"

Interactive Terminal UI

# Launch the terminal user interface
cargo run -- tui

The terminal UI provides:

• Multiple session management
• Interactive conversation history
• Visual knowledge graph representation
• Two operational modes: Query and Search/Knowledge Graph Building

TUI Controls

• e - Enter edit mode for text input
• q - Quit the application
• t - Toggle between Query and Search modes
• n - Create a new session
• h - Show/hide help
• Left/Right arrows - Navigate between sessions
• Up/Down arrows - Scroll through message history
• Enter - Submit input (when in edit mode)
• Esc - Exit edit mode

Core Components

ReachDB

ReachDB is a high-performance graph database implementation in Rust, designed for efficient storage and traversal of relationship-based data. It uses memory-mapped files for fast access to node and relationship records.

Key Features

• Memory-mapped storage: Fast access to node and relationship data
• Bidirectional relationships: Each relationship connects source and target nodes
• User-defined relationship types: Custom relationship semantics through generics
• Efficient traversals: Iterators for relationship traversal
• Persistent storage: Data remains on disk between sessions

Example Usage

// Define relationship types
#[derive(Debug)]
enum RelationType {
    IsA(u8),
    HasA(u8),
    DependsOn(u8)
}

// Implement the UserDefinedRelationType trait
impl UserDefinedRelationType for RelationType {
    // ... implementation ...
}

// Open or create a new database
let mut db = Reachdb::<RelationType>::open("data", Some(10000), Some(10000))?;

// Add edges (automatically creates nodes if they don't exist)
db.add_edge("Person", "Human", "IS-A")?;
db.add_edge("Person", "Arms", "HAS-A")?;

ReachTUI

ReachTUI is built using Ratatui and Crossterm to create a responsive terminal UI that allows users to interact with the Reach knowledge graph system through multiple sessions.

Key Components

• Multiple session management: Work with different research topics in separate sessions
• Two operational modes: Query and Search/Knowledge Graph Building
• Interactive conversation history: View and navigate through past interactions
• Visual knowledge graph representation: See relationships between concepts
• Action tracking: Monitor actions performed during the research process

Layout

┌─────────────────────────────────────────────────────────────┐
│                         Sessions                            │
├────────────────────────────────────┬────────────────────────┤
│                                    │                        │
│    ┌──────────┐ ┌───────────────┐  │                        │
│    │          │ │               │  │                        │
│    │ Actions  │ │ Conversation  │  │   Knowledge Graph      │
│    │          │ │               │  │                        │
│    │          │ │               │  │                        │
│    └──────────┘ └───────────────┘  │                        │
│                                    │                        │
├────────────────────────────────────┴────────────────────────┤
│                       Input Field                           │
└─────────────────────────────────────────────────────────────┘

Metadata Module

The metadata module provides centralized access to important constants used throughout the Reach CLI application, including:

• Author information
• Version numbers
• Configuration file names

Project Structure

reach/
├── src/
│   ├── apis/                # API integrations (Google, Gemini, arXiv)
│   ├── config/              # Configuration handling
│   ├── metadata/            # Application metadata constants
│   │   └── README.md        # Metadata documentation
│   ├── reachdb/             # Graph database implementation
│   │   └── README.md        # ReachDB documentation
│   ├── rsearch/             # Research functionality
│   │   ├── knowledge_graph.rs  # Knowledge graph generation
│   │   └── utils.rs         # Utility functions
│   ├── reachtui/            # Terminal user interface components
│   │   └── README.md        # TUI documentation
│   ├── scripts/             # Python scripts for web scraping
│   ├── display/             # Output formatting
│   └── errors/              # Error handling
├── .venv/                   # Python virtual environment
└── data/                    # Output data storage
    ├── summaries.json       # Generated summaries
    └── knowledge_graph.json # Generated knowledge graphs

Development

Running Tests

# Run all tests
cargo test

# Run tests that require configuration
cargo test --features requires_config

GitHub Actions

The project includes GitHub Actions workflows that:

• Build the project
• Set up the Python environment
• Install dependencies
• Run tests

License

[Specify your license here]

Contributing

[Add contribution guidelines if applicable]

Author

Created by Me kshitiz4kaushik@gmail.com

Version

Current version: 1.0.0