app.py

import os
import io
import sys
import re
import traceback
import subprocess
import warnings
import gradio as gr
import pandas as pd
from dotenv import load_dotenv
from crewai import Crew, Agent, Task, Process, LLM
from crewai_tools import FileReadTool
from pydantic import BaseModel, Field

# Filter out specific warnings
warnings.filterwarnings('ignore', category=FutureWarning, module='yfinance.*')

# Load environment variables
load_dotenv()

# Get API key from environment variables
OPENAI_API_KEY = os.getenv('OPENAI_API_KEY')
if not OPENAI_API_KEY:
    raise ValueError("OPENAI_API_KEY environment variable not set")

llm = LLM(
    model="openai/gpt-4o",
    api_key=OPENAI_API_KEY,
    temperature=0.7
)

# 1) Query parser agent
query_parser_agent = Agent(
    role="Stock Data Analyst",
    goal="Extract stock details and fetch required data from this user query: {query}.",
    backstory="You are a financial analyst specializing in stock market data retrieval.",
    llm=llm,
    verbose=True,
    memory=True,
)

# Need to define QueryAnalysisOutput class here as it's used by the task
class QueryAnalysisOutput(BaseModel):
    """Structured output for the query analysis task."""
    symbols: list[str] = Field(
        ...,
        json_schema_extra={"description": "List of stock ticker symbols (e.g., ['TSLA', 'AAPL'])."}
    )
    timeframe: str = Field(
        ...,
        json_schema_extra={"description": "Time period (e.g., '1d', '1mo', '1y')."}
    )
    action: str = Field(
        ...,
        json_schema_extra={"description": "Action to be performed (e.g., 'fetch', 'plot')."}
    )


query_parsing_task = Task(
    description="Analyze the user query and extract stock details.",
    expected_output="A dictionary with keys: 'symbol', 'timeframe', 'action'.",
    output_pydantic=QueryAnalysisOutput,
    agent=query_parser_agent,
)

# 2) Code writer agent
code_writer_agent = Agent(
    role="Senior Python Developer",
    goal="Write Python code to visualize stock data.",
    backstory="""You are a Senior Python developer specializing in stock market data visualization.
                 You are also a Pandas, Matplotlib and yfinance library expert.
                 You are skilled at writing production-ready Python code.
                 Ensure the code handles potential variations in the DataFrame structure returned by yfinance,
                 especially for different timeframes or delisted stocks.
                 Crucially, ensure the generated script saves any generated plot as 'plot.png' using `plt.savefig('plot.png')` before the script ends.""",
    llm=llm,
    verbose=True,
)

code_writer_task = Task(
    description="""Write Python code to visualize stock data based on the inputs from the stock analyst
                   where you would find stock symbol, timeframe and action.""",
    expected_output="A clean and executable Python script file (.py) for stock visualization.",
    agent=code_writer_agent,
)

# 3) Code output agent (instead of execution agent)
code_output_agent = Agent(
    role="Python Code Presenter",
    goal="Present the generated Python code for stock visualization.",
    backstory="You are an expert in presenting Python code in a clear and readable format.",
    allow_delegation=False,  # This agent just presents the code
    llm=llm,
    verbose=True,
)

code_output_task = Task(
    description="""Receive the Python code for stock visualization from the code writer agent and present it.""",
    expected_output="The complete Python script for stock visualization.",
    agent=code_output_agent,
)

crew = Crew(
    agents=[query_parser_agent, code_writer_agent, code_output_agent],  # Use code_output_agent
    tasks=[query_parsing_task, code_writer_task, code_output_task],  # Use code_output_task
    process=Process.sequential
)


def run_crewai_process(user_query, model, temperature):
    """
    Runs the CrewAI process, captures agent thoughts, gets generated code,
    executes the code, and returns results, including plot.

    Args:
        user_query (str): The user's query for the CrewAI process.
        model (str): The model to use for the LLM.
        temperature (float): The temperature to use for the LLM.

    Yields:
        tuple: A tuple containing the agent thoughts (str), the final answer (list of dicts),
               the generated code (str), the execution output (str), and plot file path (str or None).
    """
    # Create a string buffer to capture stdout
    output_buffer = io.StringIO()
    original_stdout = sys.stdout
    sys.stdout = output_buffer
    agent_thoughts = ""
    generated_code = ""
    execution_output = ""
    generated_plot_path = None
    final_answer_chat = [{"role": "user", "content": user_query}]

    try:
        # Initial status update with proper message format
        initial_message = {"role": "assistant", "content": "Starting CrewAI process..."}
        final_answer_chat = [{"role": "user", "content": str(user_query)}, initial_message]
        yield final_answer_chat, agent_thoughts, generated_code, execution_output, None, None

        # Run the crew process
        final_result = crew.kickoff(inputs={"query": user_query})

        # Get the captured CrewAI output (agent thoughts)
        agent_thoughts = output_buffer.getvalue()
        
        # Update with processing message
        processing_message = {"role": "assistant", "content": "Processing complete. Generating code..."}
        final_answer_chat = [{"role": "user", "content": str(user_query)}, processing_message]
        yield final_answer_chat, agent_thoughts, generated_code, execution_output, None, None

        # The final result is the generated code from the code_output_agent
        generated_code_raw = str(final_result).strip()

        # Use regex to extract the code block
        code_match = re.search(r"```python\n(.*?)\n```", generated_code_raw, re.DOTALL)
        if code_match:
            generated_code = code_match.group(1).strip()
        else:
            # If no code block is found, assume the entire output is code (or handle as error)
            generated_code = generated_code_raw
            if not generated_code.strip():  # Handle cases where output is empty or just whitespace
                execution_output = "CrewAI process completed, but no code was generated."
                final_answer_chat.append({"role": "assistant", "content": execution_output})
                yield agent_thoughts, final_answer_chat, generated_code, execution_output, generated_plot_path
                return  # Exit the generator

        # Format for Gradio Chatbot (list of dictionaries with 'role' and 'content' keys only)
        code_gen_message = {"role": "assistant", "content": "Code generation complete. See the 'Generated Code' box. Attempting to execute code..."}
        final_answer_chat = [{"role": "user", "content": str(user_query)}, code_gen_message]
        yield final_answer_chat, agent_thoughts, generated_code, execution_output, None, None

        # --- Execute the generated code ---
        # Check for common plot filename patterns
        # First check for symbol-specific plots (e.g., META_plot.png)
        symbol_plot_patterns = ['META_plot.png', 'AAPL_plot.png', 'MSFT_plot.png', 'GOOG_plot.png', 'TSLA_plot.png']
        
        # Also check for generic plot filenames
        generic_plot_patterns = ['plot.png', 'output.png', 'figure.png']
        
        # Combine all patterns to check
        plot_file_paths = symbol_plot_patterns + generic_plot_patterns

        if generated_code:
            try:
                # Write the generated code to a temporary file
                temp_script_path = "generated_script.py"
                with open(temp_script_path, "w") as f:
                    f.write(generated_code)

                # Read the generated script
                with open(temp_script_path, 'r') as f:
                    script_content = f.read()
                
                # Update yf.download() calls to include auto_adjust parameter
                def add_auto_adjust(match):
                    # Check if auto_adjust is already in the arguments
                    args = match.group(1).strip()
                    if 'auto_adjust' not in args:
                        # Add auto_adjust=True to the arguments
                        if args.endswith(','):
                            return f'yf.download({args} auto_adjust=True)'
                        elif args:  # If there are existing arguments
                            return f'yf.download({args}, auto_adjust=True)'
                        else:  # If no arguments
                            return 'yf.download(auto_adjust=True)'
                    return match.group(0)  # Return unchanged if auto_adjust is already present
                
                # This pattern matches yf.download() with any arguments
                script_content = re.sub(
                    r'yf\.download\(([^)]*)\)',
                    add_auto_adjust,
                    script_content
                )
                
                # Add helper functions at the beginning of the script
                helpers = """
# Standard plot filename to use
PLOT_FILENAME = 'generated_plot.png'

# Helper functions for data processing
def safe_get_column(df, column):
    # Handle case where column is a tuple (e.g., from multi-index)
    if isinstance(column, tuple):
        column = column[0]  # Take the first element of the tuple
    
    # Convert column to string in case it's not
    column = str(column)
    
    # Try exact match first
    if column in df.columns:
        return df[column]
        
    # Try case-insensitive match
    try:
        col_lower = column.lower()
        for col in df.columns:
            if str(col).lower() == col_lower:
                return df[col]
    except (AttributeError, TypeError):
        pass  # Skip case-insensitive matching if not applicable
    
    # If not found, try common variations
    variations = {
        'close': ['Close', 'Adj Close', 'close', 'adj close', 'CLOSE', 'Adj. Close'],
        'adj close': ['Adj Close', 'adj close', 'ADJ CLOSE', 'Close', 'close', 'CLOSE', 'Adj. Close']
    }
    
    for var_list in variations.values():
        for var in var_list:
            if var in df.columns:
                return df[var]
    
    # If still not found, try to find any column containing 'close'
    for col in df.columns:
        if 'close' in str(col).lower():
            return df[col]
            
    # If still not found, raise a helpful error
    raise KeyError(f"Column '{column}' not found in DataFrame. Available columns: {list(df.columns)}")

def show_plot(plt):
    try:
        # Use a non-interactive backend to avoid GUI issues
        import matplotlib
        matplotlib.use('Agg')
        
        # Create a temporary file to store the plot
        import io
        import base64
        
        # Save plot to a bytes buffer
        buf = io.BytesIO()
        plt.savefig(buf, format='png', bbox_inches='tight', dpi=100)
        plt.close()
        
        # Convert to base64 for display
        buf.seek(0)
        img_str = base64.b64encode(buf.read()).decode('utf-8')
        buf.close()
        
        # Return HTML to display the image
        return f'<img src="data:image/png;base64,{img_str}" />'
    except Exception as e:
        print(f"[ERROR] Failed to display plot: {str(e)}")
        return None

# Monkey patch DataFrame to add safe column access
import pandas as pd
pd.DataFrame.safe_get = safe_get_column
"""
                # Insert the helper functions after imports
                if 'import ' in script_content:
                    # Insert after the last import
                    last_import = script_content.rfind('import ')
                    insert_pos = script_content.find('\n', last_import) + 1
                    script_content = script_content[:insert_pos] + '\n' + helpers + script_content[insert_pos:]
                else:
                    # Insert at the beginning if no imports found
                    script_content = helpers + '\n' + script_content
                
                # Replace common column access patterns with our safe version
                script_content = script_content.replace("['Adj Close']", ".safe_get('close')")
                script_content = script_content.replace("['Close']", ".safe_get('close')")
                script_content = script_content.replace("['close']", ".safe_get('close')")
                
                # Replace plt.show() calls with our helper
                script_content = re.sub(
                    r'plt\\.show\\(\s*\\)',
                    r'print(show_plot(plt))',
                    script_content
                )
                
                # If no show() call is found, add one at the end of the script
                if 'plt.show()' not in script_content:
                    script_content += "\n# Display the plot if any figures exist\nif 'plt' in locals() and len(plt.get_fignums()) > 0:\n    print(show_plot(plt))\n"
                
                # Write the updated script back
                with open(temp_script_path, 'w') as f:
                    f.write(script_content)
                
                # Execute the temporary script using subprocess
                # Use python3 to ensure correct interpreter in Colab
                process = subprocess.run(
                    ["python3", temp_script_path],
                    capture_output=True,
                    text=True,  # Capture stdout and stderr as text
                    check=False  # Don't raise exception for non-zero exit codes
                )
                execution_output = process.stdout + process.stderr

                # Check for specific errors in execution output
                if "KeyError" in execution_output:
                    execution_output += "\n\nPotential Issue: The generated script encountered a KeyError. This might mean the script tried to access a column or data point that wasn't available for the specified stock or timeframe. Please try a different query or timeframe."
                elif "FileNotFoundError: [Errno 2] No such file or directory: 'plot.png'" in execution_output and "Figure(" in execution_output:
                    execution_output += "\n\nPlot Generation Issue: The script seems to have created a plot but did not save it to 'plot.png'. Please ensure the generated code includes `plt.savefig('plot.png')`."
                elif "FileNotFoundError: [Errno 2] No such file or directory: 'plot.png'" in execution_output:
                    execution_output += "\n\nPlot Generation Issue: The script ran, but the plot file was not created. Ensure the generated code includes commands to save the plot to 'plot.png'."

                # Check for the generated plot file in all possible locations
                generated_plot_path = None
                plot_found = False
                
                # First check the standard plot filename
                plot_file_paths = ['generated_plot.png', 'plot.png', 'META_plot.png', 'AAPL_plot.png', 'MSFT_plot.png', 'output.png']
                
                # Also check for any .png files in the current directory
                current_dir = os.path.abspath('.')
                png_files = [f for f in os.listdir(current_dir) 
                           if f.endswith('.png') and not f.startswith('gradio_')]
                
                # Add any found .png files to our search paths
                plot_file_paths.extend(png_files)
                
                # Make paths absolute and remove duplicates
                plot_file_paths = list(dict.fromkeys([os.path.abspath(f) for f in plot_file_paths]))
                
                print(f"[DEBUG] Looking for plot files in: {plot_file_paths}")
                
                for plot_file in plot_file_paths:
                    try:
                        if os.path.exists(plot_file) and os.path.getsize(plot_file) > 0:
                            print(f"[DEBUG] Found plot file: {plot_file}")
                            generated_plot_path = plot_file
                            plot_found = True
                            break
                    except Exception as e:
                        print(f"[DEBUG] Error checking plot file {plot_file}: {e}")
                        
                        # Add the plot to the execution output
                        try:
                            import base64
                            with open(plot_abs_path, 'rb') as img_file:
                                img_str = base64.b64encode(img_file.read()).decode('utf-8')
                                execution_output += f"\n\n![Generated Plot](data:image/png;base64,{img_str})"
                        except Exception as e:
                            execution_output += f"\n\nNote: Could not embed plot in output: {str(e)}"
                        
                        break
                
                if not plot_found:
                    # If no plot file was found, check the current directory for any .png files
                    current_dir = os.path.abspath('.')
                    png_files = [f for f in os.listdir(current_dir) if f.endswith('.png') and not f.startswith('gradio_')]
                    if png_files:
                        # Use the first .png file found
                        plot_abs_path = os.path.abspath(png_files[0])
                        generated_plot_path = plot_abs_path
                        print(f"Using plot file found at: {plot_abs_path}")
                        
                        # Add the plot to the execution output
                        try:
                            import base64
                            with open(plot_abs_path, 'rb') as img_file:
                                img_str = base64.b64encode(img_file.read()).decode('utf-8')
                                execution_output += f"\n\n![Generated Plot](data:image/png;base64,{img_str})"
                        except Exception as e:
                            execution_output += f"\n\nNote: Could not embed plot in output: {str(e)}"
                    else:
                        print(f"No plot file found in {current_dir}")
                        execution_output += "\nNo plot file was found after execution.\n\nMake sure the generated code includes:\n1. `plt.savefig('plot.png')` to save the plot\n2. `plt.close()` to close the figure after saving"

            except Exception as e:
                traceback_str = traceback.format_exc()
                execution_output = f"An error occurred during code execution: {e}\n{traceback_str}"

            finally:
                # Clean up the temporary script file
                if os.path.exists(temp_script_path):
                    os.remove(temp_script_path)

        else:
            execution_output = "No code was generated to execute."

        # Update final answer chat to reflect execution attempt
        execution_complete_msg = "Code execution finished. See 'Execution Output'."
        if generated_plot_path:
            plot_msg = f"Plot generated successfully at: {generated_plot_path}"
            final_answer_chat = [
                {"role": "user", "content": str(user_query)},
                {"role": "assistant", "content": execution_complete_msg},
                {"role": "assistant", "content": plot_msg}
            ]
            # Return the plot path to be displayed
            yield final_answer_chat, agent_thoughts, generated_code, execution_output, None, generated_plot_path
        else:
            no_plot_msg = "No plot was generated. Make sure your query includes a request for a visualization. Check the 'Execution Output' tab for any errors."
            final_answer_chat = [
                {"role": "user", "content": str(user_query)},
                {"role": "assistant", "content": execution_complete_msg},
                {"role": "assistant", "content": no_plot_msg}
            ]
            # Return None for plot path
            yield final_answer_chat, agent_thoughts, generated_code, execution_output, None, None

        yield agent_thoughts, final_answer_chat, generated_code, execution_output, generated_plot_path

    except Exception as e:
        # If an error occurs during CrewAI process, return the error message
        traceback_str = traceback.format_exc()
        agent_thoughts += f"\nAn error occurred during CrewAI process: {e}\n{traceback_str}"
        error_message = f"An error occurred during CrewAI process: {e}"
        final_answer_chat = [
            {"role": "user", "content": str(user_query)},
            {"role": "assistant", "content": error_message}
        ]
        yield final_answer_chat, agent_thoughts, generated_code, execution_output, None, None

    finally:
        # Restore original stdout
        sys.stdout = original_stdout


def create_interface():
    """Create and return the Gradio interface."""
    with gr.Blocks(title="Financial Analytics Agent", theme=gr.themes.Soft()) as interface:
        gr.Markdown("# 📊 Financial Analytics Agent")
        gr.Markdown("Enter your financial query to analyze stock data and generate visualizations.")
        
        with gr.Row():
            with gr.Column(scale=2):
                user_query_input = gr.Textbox(
                    label="Enter your financial query",
                    placeholder="e.g., Show me the stock performance of AAPL and MSFT for the last year",
                    lines=3
                )
                submit_btn = gr.Button("Analyze", variant="primary")
                
                with gr.Accordion("Advanced Options", open=False):
                    gr.Markdown("### Model Settings")
                    model_dropdown = gr.Dropdown(
                        ["gpt-4o", "gpt-4-turbo", "gpt-3.5-turbo"],
                        value="gpt-4o",
                        label="Model"
                    )
                    temperature = gr.Slider(
                        minimum=0.1,
                        maximum=1.0,
                        value=0.7,
                        step=0.1,
                        label="Creativity (Temperature)"
                    )
            
            with gr.Column(scale=3):
                with gr.Tabs():
                    with gr.TabItem("Analysis"):
                        final_answer_chat = gr.Chatbot(
                            label="Analysis Results",
                            height=300,
                            show_copy_button=True,
                            type="messages"  # Explicitly set to use OpenAI-style message format
                        )
                    
                    with gr.TabItem("Agent Thoughts"):
                        agent_thoughts = gr.Textbox(
                            label="Agent Thinking Process",
                            interactive=False,
                            lines=15,
                            max_lines=30,
                            show_copy_button=True
                        )
                    
                    with gr.TabItem("Generated Code"):
                        generated_code = gr.Code(
                            label="Generated Python Code",
                            language="python",
                            interactive=False,
                            lines=15
                        )
                    
                    with gr.TabItem("Execution Output"):
                        execution_output = gr.Textbox(
                            label="Code Execution Output",
                            interactive=False,
                            lines=10,
                            show_copy_button=True
                        )
                
                with gr.Row():
                    with gr.Column():
                        plot_output = gr.Plot(
                            label="Generated Visualization",
                            visible=False
                        )
                        image_output = gr.Image(
                            label="Generated Plot",
                            type="filepath",
                            visible=False
                        )
        
        # Handle form submission
        inputs = [user_query_input, model_dropdown, temperature]
        outputs = [
            final_answer_chat,
            agent_thoughts,
            generated_code,
            execution_output,
            plot_output,  # Pass the actual component, not None
            image_output  # Pass the actual component, not None
        ]
        
        def process_results(chat, thoughts, code, output, plot_path):
            # This function will be called after run_crewai_process
            # Show the image in the image_output component
            return [
                chat, 
                thoughts, 
                code, 
                output, 
                gr.update(visible=plot_path is not None and os.path.exists(plot_path)), 
                gr.update(value=plot_path if (plot_path and os.path.exists(plot_path)) else None, 
                        visible=plot_path is not None and os.path.exists(plot_path))
            ]
        
        # First, run the crewAI process
        click_event = submit_btn.click(
            fn=run_crewai_process,
            inputs=inputs,
            outputs=outputs,
            api_name="analyze"
        )
        
        # Then update the UI with the results
        click_event.then(
            fn=process_results,
            inputs=[final_answer_chat, agent_thoughts, generated_code, execution_output, image_output],
            outputs=outputs
        )
    
    return interface


def main():
    """Run the Gradio interface."""
    interface = create_interface()
    interface.launch(share=False, server_name="0.0.0.0", server_port=7860)


if __name__ == "__main__":
    main()