# Import packages
from datetime import datetime
from io import IOBase
from typing import List, Optional, Union
import pandas as pd
from pydantic import BaseModel, Field

from langchain.agents import AgentExecutor, create_tool_calling_agent
from langchain_core.language_models import BaseLanguageModel
from langchain_core.messages import (
    BaseMessage,
    HumanMessage,
    SystemMessage,
)
from langchain_core.prompts import (
    ChatPromptTemplate,
    MessagesPlaceholder,
)

from langchain_core.tools import Tool, BaseTool
from langchain_core.prompts.chat import (
    BaseMessagePromptTemplate,
    HumanMessagePromptTemplate,
)

from langchain_experimental.tools.python.tool import PythonAstREPLTool
from langchain_cohere.chat_models import ChatCohere

# Replace this cell with your actual cohere api key
os.env["COHERE_API_KEY"] = "cohere_api_key"

# Define prompts that we want to use in the csv agent
FUNCTIONS_WITH_DF = """
This is the result of `print(df.head())`:
{df_head}

Do note that the above df isn't the complete df. It is only the first {number_of_head_rows} rows of the df.
Use this as a sample to understand the structure of the df. However, donot use this to make any calculations directly!

The complete path for the csv files are for the corresponding dataframe is:
{csv_path}
"""  # noqa E501

FUNCTIONS_WITH_MULTI_DF = """
This is the result of `print(df.head())` for each dataframe:
{dfs_head}

Do note that the above dfs aren't the complete df. It is only the first {number_of_head_rows} rows of the df.
Use this as a sample to understand the structure of the df. However, donot use this to make any calculations directly!

The complete path for the csv files are for the corresponding dataframes are:
{csv_paths}
"""  # noqa E501

PREFIX_FUNCTIONS = """
You are working with a pandas dataframe in Python. The name of the dataframe is `df`."""  # noqa E501

MULTI_DF_PREFIX_FUNCTIONS = """
You are working with {num_dfs} pandas dataframes in Python named df1, df2, etc."""  # noqa E501

CSV_PREAMBLE = """## Task And Context
You use your advanced complex reasoning capabilities to help people by answering their questions and other requests interactively. You will be asked a very wide array of requests on all kinds of topics. You will be equipped with a wide range of search engines or similar tools to help you, which you use to research your answer. You may need to use multiple tools in parallel or sequentially to complete your task. You should focus on serving the user's needs as best you can, which will be wide-ranging. The current date is {current_date}

## Style Guide
Unless the user asks for a different style of answer, you should answer in full sentences, using proper grammar and spelling
"""  # noqa E501

# Define tools that we want the csv agent to have access to


def get_file_peek_tool() -> Tool:
    def file_peek(filename: str, num_rows: int = 5) -> str:
        """Returns the first textual contents of an uploaded file

        Args:
            table_path: the table path
            num_rows: the number of rows of the table to preview.
        """  # noqa E501
        if ".csv" in filename:
            return pd.read_csv(filename).head(num_rows).to_markdown()
        else:
            return "the table_path was not recognised"

    class file_peek_inputs(BaseModel):
        filename: str = Field(
            description="The name of the attached file to show a peek preview."
        )

    file_peek_tool = Tool(
        name="file_peek",
        description="The name of the attached file to show a peek preview.",  # noqa E501
        func=file_peek,
        args_schema=file_peek_inputs,
    )

    return file_peek_tool


def get_file_read_tool() -> Tool:
    def file_read(filename: str) -> str:
        """Returns the textual contents of an uploaded file, broken up in text chunks

        Args:
            filename (str): The name of the attached file to read.
        """  # noqa E501
        if ".csv" in filename:
            return pd.read_csv(filename).to_markdown()
        else:
            return "the table_path was not recognised"

    class file_read_inputs(BaseModel):
        filename: str = Field(
            description="The name of the attached file to read."
        )

    file_read_tool = Tool(
        name="file_read",
        description="Returns the textual contents of an uploaded file, broken up in text chunks",  # noqa E501
        func=file_read,
        args_schema=file_read_inputs,
    )

    return file_read_tool


def get_python_tool() -> Tool:
    """Returns a tool that will execute python code and return the output."""

    def python_interpreter(code: str) -> str:
        """A function that will return the output of the python code.

        Args:
            code: the python code to run.
        """
        return python_repl.run(code)

    python_repl = PythonAstREPLTool()
    python_tool = Tool(
        name="python_interpreter",
        description="Executes python code and returns the result. The code runs in a static sandbox without interactive mode, so print output or save output to a file.",  # noqa E501
        func=python_interpreter,
    )

    class PythonToolInput(BaseModel):
        code: str = Field(description="Python code to execute.")

    python_tool.args_schema = PythonToolInput
    return python_tool

def create_prompt(
    system_message: Optional[BaseMessage] = SystemMessage(
        content="You are a helpful AI assistant."
    ),
    extra_prompt_messages: Optional[
        List[BaseMessagePromptTemplate]
    ] = None,
) -> ChatPromptTemplate:
    """Create prompt for this agent.

    Args:
        system_message: Message to use as the system message that will be the
            first in the prompt.
        extra_prompt_messages: Prompt messages that will be placed between the
            system message and the new human input.

    Returns:
        A prompt template to pass into this agent.
    """
    _prompts = extra_prompt_messages or []
    messages: List[Union[BaseMessagePromptTemplate, BaseMessage]]
    if system_message:
        messages = [system_message]
    else:
        messages = []

    messages.extend(
        [
            *_prompts,
            HumanMessagePromptTemplate.from_template("{input}"),
            MessagesPlaceholder(variable_name="agent_scratchpad"),
        ]
    )
    return ChatPromptTemplate(messages=messages)


def _get_csv_head_str(path: str, number_of_head_rows: int) -> str:
    with open(path, "r") as file:
        lines = []
        for _ in range(number_of_head_rows):
            lines.append(file.readline().strip("\n"))
        # validate that the head contents are well formatted csv

        return " ".join(lines)


def _get_prompt(
    path: Union[str, List[str]], number_of_head_rows: int
) -> ChatPromptTemplate:
    if isinstance(path, str):
        lines = _get_csv_head_str(path, number_of_head_rows)
        prompt_message = f"The user uploaded the following attachments:\nFilename: {path}\nWord Count: {count_words_in_file(path)}\nPreview: {lines}"  # noqa: E501

    elif isinstance(path, list):
        prompt_messages = []
        for file_path in path:
            lines = _get_csv_head_str(file_path, number_of_head_rows)
            prompt_messages.append(
                f"The user uploaded the following attachments:\nFilename: {file_path}\nWord Count: {count_words_in_file(file_path)}\nPreview: {lines}"  # noqa: E501
            )
        prompt_message = " ".join(prompt_messages)

    prompt = create_prompt(
        system_message=HumanMessage(prompt_message)
    )
    return prompt


def count_words_in_file(file_path: str) -> int:
    try:
        with open(file_path, "r") as file:
            content = file.readlines()
            words = [len(sentence.split()) for sentence in content]
            return sum(words)
    except FileNotFoundError:
        print("File not found.")
        return 0
    except Exception as e:
        print("An error occurred:", str(e))
        return 0

# Build the agent abstraction itself
def create_csv_agent(
    llm: BaseLanguageModel,
    path: Union[str, List[str]],
    extra_tools: List[BaseTool] = [],
    pandas_kwargs: Optional[dict] = None,
    prompt: Optional[ChatPromptTemplate] = None,
    number_of_head_rows: int = 5,
    verbose: bool = True,
    return_intermediate_steps: bool = True,
) -> AgentExecutor:
    """Create csv agent with the specified language model.

    Args:
        llm: Language model to use for the agent.
        path: A string path, or a list of string paths
            that can be read in as pandas DataFrames with pd.read_csv().
        number_of_head_rows: Number of rows to display in the prompt for sample data
        include_df_in_prompt: Display the DataFrame sample values in the prompt.
        pandas_kwargs: Named arguments to pass to pd.read_csv().
        prefix: Prompt prefix string.
        suffix: Prompt suffix string.
        prompt: Prompt to use for the agent. This takes precedence over the other prompt arguments, such as suffix and prefix.
        temp_path_dir: Temporary directory to store the csv files in for the python repl.
        delete_temp_path: Whether to delete the temporary directory after the agent is done. This only works if temp_path_dir is not provided.

    Returns:
        An AgentExecutor with the specified agent_type agent and access to
        a PythonREPL and any user-provided extra_tools.

    Example:
        .. code-block:: python

            from langchain_cohere import ChatCohere, create_csv_agent

            llm = ChatCohere(model="command-a-03-2025", temperature=0)
            agent_executor = create_csv_agent(
                llm,
                "titanic.csv"
            )
            resp = agent_executor.invoke({"input":"How many people were on the titanic?"})
            print(resp.get("output"))
    """  # noqa: E501
    try:
        import pandas as pd
    except ImportError:
        raise ImportError(
            "pandas package not found, please install with `pip install pandas`."
        )

    _kwargs = pandas_kwargs or {}
    if isinstance(path, (str)):
        df = pd.read_csv(path, **_kwargs)

    elif isinstance(path, list):
        df = []
        for item in path:
            if not isinstance(item, (str, IOBase)):
                raise ValueError(
                    f"Expected str or file-like object, got {type(path)}"
                )
            df.append(pd.read_csv(item, **_kwargs))
    else:
        raise ValueError(
            f"Expected str, list, or file-like object, got {type(path)}"
        )

    if not prompt:
        prompt = _get_prompt(path, number_of_head_rows)

    final_tools = [
        get_file_read_tool(),
        get_file_peek_tool(),
        get_python_tool(),
    ] + extra_tools
    if "preamble" in llm.__dict__ and not llm.__dict__.get(
        "preamble"
    ):
        llm = ChatCohere(**llm.__dict__)
        llm.preamble = CSV_PREAMBLE.format(
            current_date=datetime.now().strftime(
                "%A, %B %d, %Y %H:%M:%S"
            )
        )

    agent = create_tool_calling_agent(
        llm=llm, tools=final_tools, prompt=prompt
    )
    agent_executor = AgentExecutor(
        agent=agent,
        tools=final_tools,
        verbose=verbose,
        return_intermediate_steps=return_intermediate_steps,
    )
    return agent_executor

import csv

# Data to be written to the CSV file
data = [
    ["movie", "name", "num_tickets"],
    ["The Shawshank Redemption", "John", 2],
    ["The Shawshank Redemption", "Jerry", 2],
    ["The Shawshank Redemption", "Jack", 4],
    ["The Shawshank Redemption", "Jeremy", 2],
    ["Finding Nemo", "Darren", 3],
    ["Finding Nemo", "Jones", 2],
    ["Finding Nemo", "King", 1],
    ["Finding Nemo", "Penelope", 5],
]

file_path = "movies_tickets.csv"

with open(file_path, "w", newline="") as file:
    writer = csv.writer(file)
    writer.writerows(data)

print(f"CSV file created successfully at {file_path}.")

# Try out an example
llm = ChatCohere(model="command-a-03-2025", temperature=0)
agent_executor = create_csv_agent(llm, "movies_tickets.csv")
resp = agent_executor.invoke(
    {"input": "Who all watched Shawshank redemption?"}
)
print(resp.get("output"))