{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import utils\n",
    "\n",
    "utils.load_env()\n",
    "os.environ['LANGCHAIN_TRACING_V2'] = \"false\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_core.messages import HumanMessage\n",
    "import operator\n",
    "import functools\n",
    "\n",
    "# for llm model\n",
    "from langchain_openai import ChatOpenAI\n",
    "from langchain.agents.format_scratchpad import format_to_openai_function_messages\n",
    "from tools import find_place_from_text, nearby_search\n",
    "from typing import Dict, List, Tuple, Annotated, Sequence, TypedDict\n",
    "from langchain.agents import (\n",
    "    AgentExecutor,\n",
    ")\n",
    "from langchain.agents.output_parsers import OpenAIFunctionsAgentOutputParser\n",
    "from langchain_community.chat_models import ChatOpenAI\n",
    "from langchain_community.tools.convert_to_openai import format_tool_to_openai_function\n",
    "from langchain_core.messages import (\n",
    "    AIMessage, \n",
    "    HumanMessage,\n",
    "    BaseMessage,\n",
    "    ToolMessage\n",
    ")\n",
    "from langchain_core.pydantic_v1 import BaseModel, Field\n",
    "from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder\n",
    "from langgraph.graph import END, StateGraph, START\n",
    "\n",
    "## Document vector store for context\n",
    "from langchain_core.runnables import RunnablePassthrough\n",
    "from langchain_chroma import Chroma\n",
    "from langchain_text_splitters import RecursiveCharacterTextSplitter\n",
    "from langchain_community.document_loaders import CSVLoader\n",
    "from langchain_openai import OpenAIEmbeddings\n",
    "import glob\n",
    "from langchain.tools import Tool\n",
    "\n",
    "def format_docs(docs):\n",
    "    return \"\\n\\n\".join(doc.page_content for doc in docs)\n",
    "\n",
    "# Specify the pattern\n",
    "file_pattern = \"document/*.csv\"\n",
    "file_paths = tuple(glob.glob(file_pattern))\n",
    "\n",
    "all_docs = []\n",
    "\n",
    "for file_path in file_paths:\n",
    "    loader = CSVLoader(file_path=file_path)\n",
    "    docs = loader.load()\n",
    "    all_docs.extend(docs)  # Add the documents to the list\n",
    "\n",
    "# Split text into chunks separated.\n",
    "text_splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=100)\n",
    "splits = text_splitter.split_documents(all_docs)\n",
    "\n",
    "# Text Vectorization.\n",
    "vectorstore = Chroma.from_documents(documents=splits, embedding=OpenAIEmbeddings())\n",
    "\n",
    "# Retrieve and generate using the relevant snippets of the blog.\n",
    "retriever = vectorstore.as_retriever()\n",
    "\n",
    "## tools and LLM\n",
    "\n",
    "retriever_tool = Tool(\n",
    "    name=\"Retriever\",\n",
    "    func=retriever.get_relevant_documents,\n",
    "    description=\"Use this tool to retrieve information about population, community and household expenditures.\"\n",
    ")\n",
    "\n",
    "# Bind the tools to the model\n",
    "tools = [retriever_tool, find_place_from_text, nearby_search]  # Include both tools if needed\n",
    "\n",
    "llm = ChatOpenAI(model=\"gpt-4o-mini\", temperature=0.0)\n",
    "\n",
    "## Create agents\n",
    "def create_agent(llm, tools, system_message: str):\n",
    "    \"\"\"Create an agent.\"\"\"\n",
    "    prompt = ChatPromptTemplate.from_messages(\n",
    "        [\n",
    "            (\n",
    "                \"system\",\n",
    "                \"You are a helpful AI assistant, collaborating with other assistants.\"\n",
    "                \" Use the provided tools to progress towards answering the question.\"\n",
    "                \" If you are unable to fully answer, that's OK, another assistant with different tools \"\n",
    "                \" will help where you left off. Execute what you can to make progress.\"\n",
    "                \" If you or any of the other assistants have the final answer or deliverable,\"\n",
    "                \" prefix your response with FINAL ANSWER so the team knows to stop.\"\n",
    "                \" You have access to the following tools: {tool_names}.\\n{system_message}\",\n",
    "            ),\n",
    "            MessagesPlaceholder(variable_name=\"messages\"),\n",
    "        ]\n",
    "    )\n",
    "    prompt = prompt.partial(system_message=system_message)\n",
    "    prompt = prompt.partial(tool_names=\", \".join([tool.name for tool in tools]))\n",
    "    llm_with_tools = llm.bind(functions=[format_tool_to_openai_function(t) for t in tools])\n",
    "    # return prompt | llm.bind_tools(tools)\n",
    "    agent = prompt | llm\n",
    "    return agent\n",
    "\n",
    "\n",
    "## Define state\n",
    "# This defines the object that is passed between each node\n",
    "# in the graph. We will create different nodes for each agent and tool\n",
    "class AgentState(TypedDict):\n",
    "    messages: Annotated[Sequence[BaseMessage], operator.add]\n",
    "    sender: str\n",
    "\n",
    "\n",
    "# Helper function to create a node for a given agent\n",
    "def agent_node(state, agent, name):\n",
    "    result = agent.invoke(state)\n",
    "    # We convert the agent output into a format that is suitable to append to the global state\n",
    "    if isinstance(result, ToolMessage):\n",
    "        pass\n",
    "    else:\n",
    "        result = AIMessage(**result.dict(exclude={\"type\", \"name\"}), name=name)\n",
    "    return {\n",
    "        \"messages\": [result],\n",
    "        # Since we have a strict workflow, we can\n",
    "        # track the sender so we know who to pass to next.\n",
    "        \"sender\": name,\n",
    "    }\n",
    "\n",
    "\n",
    "## Define Agents Node\n",
    "# Research agent and node\n",
    "agent_meta = utils.load_agent_meta()\n",
    "agent_name = [meta['name'] for meta in agent_meta]\n",
    "\n",
    "agents={}\n",
    "agent_nodes={}\n",
    "\n",
    "for meta in agent_meta:\n",
    "    name = meta['name']\n",
    "    prompt = meta['prompt']\n",
    "    \n",
    "    agents[name] = create_agent(\n",
    "            llm,\n",
    "            tools,\n",
    "            system_message=prompt,\n",
    "        )\n",
    "    \n",
    "    agent_nodes[name] = functools.partial(agent_node, agent=agents[name], name=name)\n",
    "\n",
    "\n",
    "## Define Tool Node\n",
    "from langgraph.prebuilt import ToolNode\n",
    "from typing import Literal\n",
    "\n",
    "tool_node = ToolNode(tools)\n",
    "\n",
    "def router(state) -> Literal[\"call_tool\", \"__end__\", \"continue\"]:\n",
    "    # This is the router\n",
    "    messages = state[\"messages\"]\n",
    "    last_message = messages[-1]\n",
    "    if last_message.tool_calls:\n",
    "        # The previous agent is invoking a tool\n",
    "        return \"call_tool\"\n",
    "    if \"FINAL ANSWER\" in last_message.content:\n",
    "        # Any agent decided the work is done\n",
    "        return \"__end__\"\n",
    "    return \"continue\"\n",
    "\n",
    "\n",
    "## Workflow Graph\n",
    "workflow = StateGraph(AgentState)\n",
    "\n",
    "# add agent nodes\n",
    "for name, node in agent_nodes.items():\n",
    "    workflow.add_node(name, node)\n",
    "    \n",
    "workflow.add_node(\"call_tool\", tool_node)\n",
    "\n",
    "\n",
    "workflow.add_conditional_edges(\n",
    "    \"analyst\",\n",
    "    router,\n",
    "    {\"continue\": \"data collector\", \"call_tool\": \"call_tool\", \"__end__\": END}\n",
    ")\n",
    "\n",
    "workflow.add_conditional_edges(\n",
    "    \"data collector\",\n",
    "    router,\n",
    "    {\"continue\": \"reporter\", \"call_tool\": \"call_tool\", \"__end__\": END}\n",
    ")\n",
    "\n",
    "workflow.add_conditional_edges(\n",
    "    \"reporter\",\n",
    "    router,\n",
    "    {\"continue\": \"data collector\", \"__end__\": END}\n",
    ")\n",
    "\n",
    "workflow.add_conditional_edges(\n",
    "    \"call_tool\",\n",
    "    # Each agent node updates the 'sender' field\n",
    "    # the tool calling node does not, meaning\n",
    "    # this edge will route back to the original agent\n",
    "    # who invoked the tool\n",
    "    lambda x: x[\"sender\"],\n",
    "    {name: name for name in agent_name},\n",
    ")\n",
    "workflow.add_edge(START, \"analyst\")\n",
    "graph = workflow.compile()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from IPython.display import Image, display\n",
    "\n",
    "# try:\n",
    "#     display(Image(graph.get_graph(xray=True).draw_mermaid_png()))\n",
    "# except Exception:\n",
    "#     # This requires some extra dependencies and is optional\n",
    "#     pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# graph = workflow.compile()\n",
    "\n",
    "# events = graph.stream(\n",
    "#     {\n",
    "#         \"messages\": [\n",
    "#             HumanMessage(\n",
    "#                 content=\"ค้นหาร้านกาแฟใกล้มาบุญครอง พร้อมวิเคราะห์จำนวนประชากร\"\n",
    "#             )\n",
    "#         ],\n",
    "#     },\n",
    "#     # Maximum number of steps to take in the graph\n",
    "#     {\"recursion_limit\": 10},\n",
    "# )\n",
    "# for s in events:\n",
    "#     print(s)\n",
    "#     print(\"----\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'FINAL ANSWER\\n\\nรายงานการวิเคราะห์ร้านกาแฟใกล้มาบุญครอง:\\n\\n### รายชื่อร้านกาแฟ\\n1. **ร้านกาแฟ A** - เมนูหลากหลาย บรรยากาศน่านั่ง\\n2. **ร้านกาแฟ B** - กาแฟสด ขนมเค้กอร่อย\\n3. **ร้านกาแฟ C** - Wi-Fi ฟรี พื้นที่นั่งทำงาน\\n4. **ร้านกาแฟ D** - กิจกรรมดนตรีสด\\n5. **ร้านกาแฟ E** - เมนูเอกลักษณ์ การตกแต่งสวยงาม\\n\\n### การวิเคราะห์\\n- **ประเภทของร้านกาแฟ**: มีความหลากหลาย เช่น บรรยากาศน่านั่ง, เน้นกาแฟสด, และการจัดกิจกรรม\\n- **กลุ่มเป้าหมาย**: นักศึกษา, คนทำงาน, นักท่องเที่ยว\\n- **โอกาสทางการตลาด**: สร้างความแตกต่าง, ใช้โซเชียลมีเดีย, จัดกิจกรรมพิเศษ\\n\\n### ข้อเสนอแนะ\\n- สร้างเอกลักษณ์เฉพาะตัว\\n- ใช้กลยุทธ์การตลาดที่เหมาะสม\\n- จัดกิจกรรมเพื่อดึงดูดลูกค้า\\n\\nหากต้องการข้อมูลเพิ่มเติมหรือรายละเอียดเฉพาะเจาะจงเกี่ยวกับร้านกาแฟใด ๆ โปรดแจ้งให้ฉันทราบ!'"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def submitUserMessage(user_input: str) -> str:\n",
    "    graph = workflow.compile()\n",
    "\n",
    "    events = graph.stream(\n",
    "        {\n",
    "            \"messages\": [\n",
    "                HumanMessage(\n",
    "                    content=user_input\n",
    "                )\n",
    "            ],\n",
    "        },\n",
    "        # Maximum number of steps to take in the graph\n",
    "        {\"recursion_limit\": 15},\n",
    "    )\n",
    "    \n",
    "    events = [e for e in events]\n",
    "    \n",
    "    response = list(events[-1].values())[0][\"messages\"][0]\n",
    "    response = response.content\n",
    "    response = response.replace(\"FINAL ANSWER: \", \"\")\n",
    "    \n",
    "    return response\n",
    "\n",
    "#submitUserMessage(\"วิเคราะห์การเปิดร้านกาแฟใกล้มาบุญครอง\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}