llama-metasearch

Llama Metasearch (llama-metasearch) is a powerful metasearch engine within the LlamaSearch AI ecosystem. It aggregates results from multiple underlying search sources, ranks them, and presents a unified set of results to the user.

Key Features

• Metasearch Engine: Core logic for querying multiple sources and combining results (metasearch.py).
• Source Aggregation: Fetches results from various configured search engines or APIs.
• Result Ranking: Implements algorithms to rank aggregated results effectively.
• Unified API: Provides a single point of access for querying diverse sources.
• Core Module: Manages the overall metasearch process (core.py).
• Configurable: Allows defining search sources, ranking parameters, and other settings (config.py).

Installation

pip install llama-metasearch
# Or install directly from GitHub for the latest version:
# pip install git+https://github.com/llamasearchai/llama-metasearch.git

Usage

(Usage examples demonstrating how to perform metasearch queries will be added here.)

# Placeholder for Python client usage
# from llama_metasearch import MetasearchClient, SearchConfig

# config = SearchConfig.load("config.yaml")
# client = MetasearchClient(config)

# # Perform a metasearch query
# results = client.search("artificial intelligence trends", sources=["web", "news", "academic"])
# for result in results:
#     print(f"[{result.source}] {result.title} - {result.url}")

Architecture Overview

graph TD
    A[User Query] --> B{Core Orchestrator (core.py)};
    B --> C{Metasearch Engine (metasearch.py)};
    C -- Queries --> D[Source 1 Interface];
    C -- Queries --> E[Source 2 Interface];
    C -- Queries --> F[...];
    D --> G((Source 1 API / DB));
    E --> H((Source 2 API / DB));
    F --> I((...));
    G -- Results --> C;
    H -- Results --> C;
    I -- Results --> C;
    C --> J{Result Aggregation & Ranking};
    J --> K[Unified Search Results];

    L[Configuration (config.py)] -- Configures --> B;
    L -- Configures --> C;
    L -- Configures --> D;
    L -- Configures --> E;
    L -- Configures --> F;

    style C fill:#f9f,stroke:#333,stroke-width:2px

1. Query Input: The user submits a search query.
2. Core Orchestrator: Manages the request flow.
3. Metasearch Engine: Receives the query and dispatches it to configured source interfaces.
4. Source Interfaces: Interact with the actual underlying search sources (APIs, databases, etc.).
5. Aggregation & Ranking: The engine gathers results from all sources, deduplicates, and ranks them.
6. Output: Presents a unified list of ranked results.
7. Configuration: Defines which sources to query, API keys, ranking strategies, etc.

Configuration

(Details on configuring search sources, API keys, ranking algorithms, result caching, etc., will be added here.)

Development

Setup

# Clone the repository
git clone https://github.com/llamasearchai/llama-metasearch.git
cd llama-metasearch

# Install in editable mode with development dependencies
pip install -e ".[dev]"

Testing

pytest tests/

Contributing

Contributions are welcome! Please refer to CONTRIBUTING.md and submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.