app.py

import streamlit as st
import pandas as pd
import ftfy

def fix_complex_encoding(text):
    try:
        # First, try to fix text using ftfy
        fixed_text = ftfy.fix_text(text)
        # Additional encoding and decoding steps to handle more cases
        fixed_text = (fixed_text.encode('latin1', errors='ignore')
                                  .decode('utf-8', errors='ignore'))
        return fixed_text
    except Exception as e:
        return text

def process_dataframe(df):
    return df.applymap(fix_complex_encoding)

st.title("CSV Encoding Fixer")

uploaded_file = st.file_uploader("Upload a CSV file", type="csv")

if uploaded_file is not None:
    df = pd.read_csv(uploaded_file, dtype=str)
    st.write("Original Data:")
    st.dataframe(df)

    processed_df = process_dataframe(df)
    st.write("Processed Data:")
    st.dataframe(processed_df)

    output_csv = processed_df.to_csv(index=False).encode('utf-8')
    st.download_button(
        label="Download Processed CSV",
        data=output_csv,
        file_name='processed_file.csv',
        mime='text/csv'
    )