import gradio as gr
import pandas as pd


def process_csv(df):

    # Check for commas in any of the columns and filter out those rows
    df = df[~df.apply(lambda row: row.astype(str).str.contains(',').any(), axis=1)]
    # Split the 'title' column on ' -' and keep only the first part
    df['title'] = df['title'].str.split(' -').str[0]
    df = df.rename(columns={'request_text': 'Veta', 'intent_1': 'Intent', 'title': 'ID Odpovedi'})
    # Apply a filter to eliminate rows with the 'Veta' column that have less than or equal to 3 words
    df = df[df['Veta'].str.split().str.len() > 3]
    # Delete all rows where 'Intent' contains 'Cognitive' or 'SmallTalk'
    df = df.dropna(subset=['Intent'])
    df = df[~df['Intent'].str.contains('Cognitive|SmallTalk')]
    # Using the pandas dataframe sample function to get 200 random rows
    df = df.sample(200, random_state=42, replace=True)
    return df


def process_and_save(file):
    
    # Load the CSV file into a DataFrame
    df = pd.read_csv(file.name, index_col=False)
    df.reset_index(drop=True, inplace=True)
    # Process the DataFrame using your script
    processed_df = process_csv(df)
    # Save the processed DataFrame to an Excel file
    output_file = 'processed_output.csv'
    processed_df.to_csv(output_file, index=False)

    return output_file


# Define the Gradio interface
interface = gr.Interface(
    fn=process_and_save,
    inputs=gr.File(label="Upload CSV File"),
    outputs=gr.File(label="Download Excel File"),
    title="Zprocesuj denní testy pro Testing Tool",
    description="Načti export denních testů z PowerBI v csv a klikni na 'Submit' Pak stáhni zprocesovaný csv soubor."
)

if __name__ == "__main__":
    interface.launch(share=True)