main

import streamlit as st
import pandas as pd
from decimal import Decimal, ROUND_HALF_UP
import io

# Utility functions
def round_decimal(value):
    return Decimal(value).quantize(Decimal('0.01'), rounding=ROUND_HALF_UP)

def parse_amount(x):
    try:
        return Decimal(str(x).replace(',', ''))
    except:
        return Decimal('0.00')

def reconcile_ledger(statement_df, ledger_df):
    statement = statement_df.to_dict(orient='records')
    ledger = ledger_df.to_dict(orient='records')

    unmatched = []

    # Step 1: Cancel out contra entries
    for s in statement:
        s_amt = parse_amount(s['amount'])
        for l in ledger:
            l_amt = parse_amount(l['amount'])
            if s_amt == l_amt and s['type'] != l['type'] and not l.get('matched') and not s.get('matched'):
                s['matched'] = l['matched'] = True
                break
        if not s.get('matched'):
            unmatched.append(s)

    for l in ledger:
        if not l.get('matched'):
            unmatched.append(l)

    net_total = Decimal('0.00')
    for entry in unmatched:
        amt = parse_amount(entry['amount'])
        if entry['type'] == 'debit':
            net_total += amt
        else:
            net_total -= amt

    net_total = round_decimal(net_total)
    vat = round_decimal(net_total * Decimal('0.15'))
    total_with_vat = round_decimal(net_total + vat)
    discount = round_decimal(total_with_vat * Decimal('0.025'))
    final_payment = round_decimal(total_with_vat - discount)

    return {
        "net_total": net_total,
        "vat_15%": vat,
        "total_with_vat": total_with_vat,
        "discount_2.5%": discount,
        "total_payable": final_payment,
        "unmatched": pd.DataFrame(unmatched)
    }

# Streamlit App UI
st.set_page_config(page_title="Statement-Ledger Reconciliation Tool", layout="wide")
st.title("📊 Statement-Ledger Reconciliation Tool")

st.markdown("""
Upload your **Statement** and **Ledger** CSV files for automatic reconciliation.  
Ensure both files have columns: `id`, `amount`, `type` (`debit` or `credit`).
""")

statement_file = st.file_uploader("📄 Upload Statement CSV", type=["csv"])
ledger_file = st.file_uploader("📄 Upload Ledger CSV", type=["csv"])

if statement_file and ledger_file:
    try:
        statement_df = pd.read_csv(statement_file)
        ledger_df = pd.read_csv(ledger_file)

        required_cols = {'id', 'amount', 'type'}
        if required_cols.issubset(statement_df.columns) and required_cols.issubset(ledger_df.columns):
            result = reconcile_ledger(statement_df, ledger_df)
            unmatched_df = result['unmatched']

            st.markdown("---")
            st.subheader("🔍 Reconciliation Breakdown")

            col1, col2 = st.columns(2)
            with col1:
                st.metric("Net Total", f"R {result['net_total']}")
                st.metric("VAT (15%)", f"R {result['vat_15%']}")
            with col2:
                st.metric("Total Incl. VAT", f"R {result['total_with_vat']}")
                st.metric("Less Discount (2.5%)", f"- R {result['discount_2.5%']}")

            st.markdown("### ✅ **Total Amount Payable**")
            st.success(f"💰 R {result['total_payable']}")
            st.markdown("---")

            st.subheader("📋 Unmatched Transactions")
            st.dataframe(unmatched_df, use_container_width=True)

            # CSV Export
            csv_buffer = io.StringIO()
            unmatched_df.to_csv(csv_buffer, index=False)
            st.download_button(
                label="⬇️ Download Unmatched (CSV)",
                data=csv_buffer.getvalue(),
                file_name="unmatched_entries.csv",
                mime="text/csv"
            )

            # Excel Export
            excel_buffer = io.BytesIO()
            with pd.ExcelWriter(excel_buffer, engine='xlsxwriter') as writer:
                unmatched_df.to_excel(writer, index=False, sheet_name='Unmatched Entries')
                worksheet = writer.sheets['Unmatched Entries']
                for i, col in enumerate(unmatched_df.columns):
                    col_len = unmatched_df[col].astype(str).map(len).max()
                    worksheet.set_column(i, i, col_len + 5)
            st.download_button(
                label="⬇️ Download Unmatched (Excel)",
                data=excel_buffer.getvalue(),
                file_name="unmatched_entries.xlsx",
                mime="application/vnd.openxmlformats-officedocument.spreadsheetml.sheet"
            )

        else:
            st.error("❌ Missing required columns: 'id', 'amount', 'type' in one or both files.")
    except Exception as e:
        st.error(f"❗ An error occurred during processing:\n\n{e}")
else:
    st.info("📁 Please upload both Statement and Ledger CSV files to begin.")

Run NSIS > "Compile NSI script" > select your installer.nsi.

Output: StatementLedgerInstaller.exe
Double-click it — it installs your compiled .exe to Program Fil