tools/search_tool.py

import os
import chromadb
from chromadb.utils import embedding_functions

# Configuration
CHROMA_PATH = "data/chroma_db"
COLLECTION_NAME = "nyaya_legal_docs"

def get_vector_collection():
    client = chromadb.PersistentClient(path=CHROMA_PATH)
    embedding_func = embedding_functions.DefaultEmbeddingFunction()
    return client.get_or_create_collection(
        name=COLLECTION_NAME,
        embedding_function=embedding_func
    )

def search_legal_documents(query: str):
    """
    Search across the indexed legal documents (Gazettes, Circulars, etc.) using semantic search.
    
    Args:
        query: The natural language query to search for in local documents.
    """
    try:
        collection = get_vector_collection()
        results = collection.query(
            query_texts=[query],
            n_results=3
        )
        if results['documents']:
            return "\n\n".join(results['documents'][0])
        return "No relevant local documents found."
    except Exception as e:
        return f"Error searching local documents: {e}"

import os
import requests

def search_indian_kanoon(query: str):
    """
    Search for legal cases, acts, and statutes on Indian Kanoon (Online).
    """
    token = os.getenv("INDIANKANOON_API_KEY")
    if not token:
        return f"Search result placeholder for: {query}. (API Token missing in .env. Using internal model knowledge.)"

    try:
        headers = {
            "Authorization": f"Token {token}",
            "Accept": "application/json"
        }
        params = {"formInput": query, "pagenum": 0}
        # Searching for the query - Indian Kanoon requires POST with params in the URL
        response = requests.post(
            f"https://api.indiankanoon.org/search/", 
            headers=headers, 
            params=params
        )
        data = response.json()
        
        if "docs" in data:
            results = data["docs"][:3]
            summary = "Found these cases on Indian Kanoon:\n"
            for res in results:
                summary += f"- {res['title']} (Source: {res.get('docsource', 'Unknown')}, ID: {res['tid']})\n"
            return summary
        return "No results found on Indian Kanoon."
    except Exception as e:
        return f"Error connecting to Indian Kanoon: {e}"

def get_act_details(act_name: str, section: str = None):
    """
    Retrieve details of a specific Indian Act or Section.
    
    Args:
        act_name: Name of the Act (e.g., 'IPC', 'DPDPA').
        section: Specific section number (optional).
    """
    return f"Details for {act_name} Section {section if section else 'All'}. (Data retrieval pending)"