agent.py

"""LangGraph Agent"""
import os
from dotenv import load_dotenv
from langgraph.graph import START, StateGraph, MessagesState
from langgraph.prebuilt import tools_condition
from langgraph.prebuilt import ToolNode
from langchain_google_genai import ChatGoogleGenerativeAI
from langchain_groq import ChatGroq
from langchain_huggingface import ChatHuggingFace, HuggingFaceEndpoint, HuggingFaceEmbeddings,  HuggingFacePipeline
from langchain_community.tools.tavily_search import TavilySearchResults
from langchain_community.document_loaders import WikipediaLoader
from langchain_community.document_loaders import ArxivLoader
from langchain_community.vectorstores import SupabaseVectorStore
from langchain_core.messages import SystemMessage, HumanMessage
from langchain_core.tools import tool
from langchain.tools.retriever import create_retriever_tool
from supabase.client import Client, create_client

load_dotenv()

@tool
def multiply(a: int, b: int) -> int:
    """Multiply two numbers.
    Args:
        a: first int
        b: second int
    """
    return a * b

@tool
def add(a: int, b: int) -> int:
    """Add two numbers.
    
    Args:
        a: first int
        b: second int
    """
    return a + b

@tool
def subtract(a: int, b: int) -> int:
    """Subtract two numbers.
    
    Args:
        a: first int
        b: second int
    """
    return a - b

@tool
def divide(a: int, b: int) -> int:
    """Divide two numbers.
    
    Args:
        a: first int
        b: second int
    """
    if b == 0:
        raise ValueError("Cannot divide by zero.")
    return a / b

@tool
def modulus(a: int, b: int) -> int:
    """Get the modulus of two numbers.
    
    Args:
        a: first int
        b: second int
    """
    return a % b

@tool
def wiki_search(query: str) -> str:
    """Search Wikipedia for a query and return maximum 2 results.
    
    Args:
        query: The search query."""
    search_docs = WikipediaLoader(query=query, load_max_docs=2).load()
    formatted_search_docs = "\n\n---\n\n".join(
        [
            f'<Document source="{doc.metadata["source"]}" page="{doc.metadata.get("page", "")}"/>\n{doc.page_content}\n</Document>'
            for doc in search_docs
        ])
    return {"wiki_results": formatted_search_docs}

# @tool
# def web_search(query: str) -> str:
#     """Search Tavily for a query and return maximum 3 results.
    
#     Args:
#         query: The search query."""
#     search_docs = TavilySearchResults(max_results=3).invoke(query=query)
#     formatted_search_docs = "\n\n---\n\n".join(
#         [
#             f'<Document source="{doc.metadata["source"]}" page="{doc.metadata.get("page", "")}"/>\n{doc.page_content}\n</Document>'
#             for doc in search_docs
#         ])
#     return {"web_results": formatted_search_docs}
@tool
def web_search(query: str) -> str:
    """Search Tavily for a query and return maximum 3 results.
    
    Args:
        query: The search query."""
    try:
        # Initialize the tool
        tavily_tool = TavilySearchResults(max_results=3)
        # Invoke it correctly
        search_results = tavily_tool.invoke(input=query) # <--- CORRECTED LINE

        # The result of TavilySearchResults.invoke is usually a list of strings or a single string.
        # Let's check its type and format accordingly.
        # Typically, TavilySearchResults directly returns a list of Document objects
        # or a list of dictionaries if you've configured it differently.
        # For the default, it's often a list of strings or a single concatenated string.
        # If it returns a list of Document objects (which is common for loaders/retrievers):
        if isinstance(search_results, list) and all(hasattr(doc, 'metadata') and hasattr(doc, 'page_content') for doc in search_results):
            formatted_search_docs = "\n\n---\n\n".join(
                [
                    f'<Document source="{doc.metadata.get("source", "N/A")}" page="{doc.metadata.get("page", "")}"/>\n{doc.page_content}\n</Document>'
                    for doc in search_results
                ])
        # If it returns a list of strings (less likely for Tavily in recent versions, but good to check)
        elif isinstance(search_results, list) and all(isinstance(item, str) for item in search_results):
            formatted_search_docs = "\n\n---\n\n".join(search_results)
        # If it returns a single string
        elif isinstance(search_results, str):
            formatted_search_docs = search_results
        else:
            # Fallback or handle unexpected format
            print(f"Unexpected Tavily search result format: {type(search_results)}")
            formatted_search_docs = str(search_results)

        return {"web_results": formatted_search_docs}
    except Exception as e:
        print(f"Error during Tavily search for query '{query}': {e}")
        return {"web_results": f"Error performing web search: {e}"}

@tool
def arvix_search(query: str) -> str:
    """Search Arxiv for a query and return maximum 3 result.
    
    Args:
        query: The search query."""
    search_docs = ArxivLoader(query=query, load_max_docs=3).load()
    formatted_search_docs = "\n\n---\n\n".join(
        [
            f'<Document source="{doc.metadata["source"]}" page="{doc.metadata.get("page", "")}"/>\n{doc.page_content[:1000]}\n</Document>'
            for doc in search_docs
        ])
    return {"arvix_results": formatted_search_docs}



# load the system prompt from the file
with open("system_prompt.txt", "r", encoding="utf-8") as f:
    system_prompt = f.read()

# System message
sys_msg = SystemMessage(content=system_prompt)

# build a retriever
embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-mpnet-base-v2") #  dim=768
supabase: Client = create_client(
    os.environ.get("SUPABASE_URL"), 
    os.environ.get("SUPABASE_SERVICE_KEY"))
vector_store = SupabaseVectorStore(
    client=supabase,
    embedding= embeddings,
    table_name="documents",
    query_name="match_documents_langchain",
)
create_retriever_tool = create_retriever_tool(
    retriever=vector_store.as_retriever(),
    name="Question Search",
    description="A tool to retrieve similar questions from a vector store.",
)



tools = [
    multiply,
    add,
    subtract,
    divide,
    modulus,
    wiki_search,
    web_search,
    arvix_search,
]

hf_token = os.environ.get('HF_TOKEN')
if not hf_token:
    raise ValueError("Hugging Face API token (HF_TOKEN) not found in environment variables.")

tavili_key = os.environ.get('TAVILY_API_KEY')
if not tavili_key:
    raise ValueError("Hugging Face API token (HF_TOKEN) not found in environment variables.")


# Build graph function
def build_graph(provider: str = "huggingface"):

    """Build the graph"""
    # Load environment variables from .env file
    if provider == "google":
        # Google Gemini
        llm = ChatGoogleGenerativeAI(model="gemini-2.0-flash", temperature=0)
    elif provider == "groq":
        # Groq https://console.groq.com/docs/models
        llm = ChatGroq(model="qwen-qwq-32b", temperature=0) # optional : qwen-qwq-32b gemma2-9b-it
    elif provider == "huggingface":
        repo_id = "togethercomputer/evo-1-131k-base"
        repo_id="HuggingFaceH4/zephyr-7b-beta",
        if not hf_token:
            raise ValueError("HF_TOKEN environment variable not set. It's required for Hugging Face provider.")
        llm = HuggingFaceEndpoint(
            repo_id="Qwen/Qwen2.5-Coder-32B-Instruct",
            provider="auto",
            task="text-generation",
            max_new_tokens=1000,
            do_sample=False,
            repetition_penalty=1.03,
            
        )
        llm = ChatHuggingFace(llm=llm)
    else:
        raise ValueError("Invalid provider. Choose 'google', 'groq' or 'huggingface'.")
    # Bind tools to LLM
    """Build the graph"""

  
    llm_with_tools = llm.bind_tools(tools)

    # Node
    def assistant(state: MessagesState):
        print("\n--- Assistant Node ---")
        print("Incoming messages to assistant:")
        for msg in state["messages"]:
            msg.pretty_print() #
        
        """Assistant node"""
        return {"messages": [llm_with_tools.invoke(state["messages"])]}
    
    def retriever(state: MessagesState):
        """Retriever node"""
        similar_question = vector_store.similarity_search(state["messages"][0].content)
        example_msg = HumanMessage(
            content=f"Here I provide a similar question and answer for reference: \n\n{similar_question[0].page_content}",
        )
        print("ex msgs"+[sys_msg] + state["messages"] + [example_msg])
        return {"messages": [sys_msg] + state["messages"] + [example_msg]}

    builder = StateGraph(MessagesState)
    builder.add_node("retriever", retriever)
    builder.add_node("assistant", assistant)
    builder.add_node("tools", ToolNode(tools))
    builder.add_edge(START, "assistant")
    builder.add_edge("retriever", "assistant")
    builder.add_conditional_edges(
        "assistant",
        tools_condition,
    )
    builder.add_edge("tools", "assistant")

    # Compile graph
    compiled_graph = builder.compile() # This line should already be there or be the next line

    # --- START: Add this visualization code ---
    try:
        print("Attempting to generate graph visualization...")
        image_filename = "langgraph_state_diagram.png"
        # Using draw_mermaid_png as it's often more robust
        image_bytes = compiled_graph.get_graph().draw_mermaid_png()
        with open(image_filename, "wb") as f:
            f.write(image_bytes)
        print(f"SUCCESS: Graph visualization saved to '{image_filename}'")

    except ImportError as e:
        print(f"WARNING: Could not generate graph image due to missing package: {e}. "
              "Ensure 'pygraphviz' and 'graphviz' (system) are installed, or Mermaid components are available.")
    except Exception as e:
        print(f"WARNING: An error occurred while generating the graph image: {e}")
        try:
            print("\nGraph (DOT format as fallback):\n", compiled_graph.get_graph().to_string())
        except Exception as dot_e:
            print(f"Could not even get DOT string: {dot_e}")
    # --- END: Visualization code ---

    return compiled_graph # This should be the last line of the function
    
# test
if __name__ == "__main__":
    question = "When was a picture of St. Thomas Aquinas first added to the Wikipedia page on the Principle of double effect?"
    # Build the graph
    graph = build_graph(provider="huggingface")
    # Run the graph
    messages = [HumanMessage(content=question)]

    print(messages)
    config = {"recursion_limit": 27}

    messages = graph.invoke({"messages": messages}, config=config)
    for m in messages["messages"]:
        m.pretty_print()