modules/parser.py

"""Module for parsing Zerodha CSV exports."""

import pandas as pd
import io

def clean_numeric_series(series):
    """Removes commas and % signs from a pandas Series, then converts to float."""
    return series.astype(str).str.replace(',', '', regex=False).str.replace('%', '', regex=False).astype(float)

def load_portfolio(file_path_or_string):
    """
    Reads a Zerodha portfolio CSV and returns a cleaned pandas DataFrame.
    
    Accepts either a file path ending in .csv or a raw CSV text string.
    """
    try:
        if file_path_or_string.strip().endswith('.csv') and '\n' not in file_path_or_string:
            df = pd.read_csv(file_path_or_string.strip())
        else:
            df = pd.read_csv(io.StringIO(file_path_or_string.strip()))
            
        # Strip whitespace from column names
        df.columns = df.columns.str.strip()
        
        # Rename columns. Using a flexible mapping in case of slight variations (like "Qty." vs "Qty")
        col_mapping = {
            "Instrument": "symbol",
            "Qty.": "qty",
            "Qty": "qty",
            "Avg. cost": "avg_cost",
            "LTP": "ltp",
            "Invested": "invested",
            "Cur. val": "cur_val",
            "P&L": "pnl",
            "Net chg.": "net_chg",
            "Day chg.": "day_chg"
        }
        
        df = df.rename(columns=col_mapping)
                
        # Drop rows where symbol is empty or NaN
        if 'symbol' in df.columns:
            df = df.dropna(subset=['symbol'])
            df = df[df['symbol'].astype(str).str.strip() != '']
        
        # Convert numeric columns
        numeric_cols = ['qty', 'avg_cost', 'ltp', 'invested', 'cur_val', 'pnl', 'net_chg', 'day_chg']
        for col in numeric_cols:
            if col in df.columns:
                df[col] = clean_numeric_series(df[col])
                
        return df
    
    except Exception as e:
        raise ValueError(f"Failed to load or parse portfolio: {e}")

def validate_portfolio(df):
    """
    Validates the parsed portfolio DataFrame.
    
    Returns (True, None) if valid.
    Returns (False, error_message) if invalid.
    """
    try:
        if df.empty:
            return False, "Portfolio DataFrame is empty (no rows)."
            
        required_cols = ['symbol', 'qty', 'avg_cost', 'ltp', 'invested', 'cur_val', 'pnl', 'net_chg', 'day_chg']
        missing_cols = [col for col in required_cols if col not in df.columns]
        
        if missing_cols:
            found_cols = list(df.columns)
            return False, f"Missing required columns: {', '.join(missing_cols)}\nFound columns: {', '.join(found_cols)}"
            
        if df['symbol'].isnull().any() or (df['symbol'].astype(str).str.strip() == '').any():
            return False, "Found rows with null or empty symbols."
            
        return True, None
        
    except Exception as e:
        return False, f"Validation failed with error: {e}"

def portfolio_to_text(df):
    """
    Converts the cleaned DataFrame into a formatted text summary for AI prompts.
    """
    lines = []
    
    for _, row in df.iterrows():
        # SYMBOL | Qty: X | Avg: ₹Y | LTP: ₹Z | P&L: ₹A (B%) | Day: C%
        sym = row['symbol']
        qty = row['qty']
        avg = row['avg_cost']
        ltp = row['ltp']
        pnl = row['pnl']
        net = row['net_chg']
        day = row['day_chg']
        
        line = f"{sym} | Qty: {qty:g} | Avg: ₹{avg:.2f} | LTP: ₹{ltp:.2f} | P&L: ₹{pnl:.2f} ({net:+.2f}%) | Day: {day:+.2f}%"
        lines.append(line)
        
    lines.append("-" * 40)
    
    total_invested = df['invested'].sum() if 'invested' in df.columns else 0.0
    total_cur_val = df['cur_val'].sum() if 'cur_val' in df.columns else 0.0
    
    lines.append(f"Total Invested: ₹{total_invested:.2f} | Total Current Value: ₹{total_cur_val:.2f}")
    
    return "\n".join(lines)


if __name__ == "__main__":
    import sys
    if sys.platform == 'win32':
        sys.stdout.reconfigure(encoding='utf-8')
        
    # Simple test with sample data matching Zerodha format
    sample_csv = '''Instrument,Qty.,Avg. cost,LTP,Invested,Cur. val,P&L,Net chg.,Day chg.
    RELIANCE,10,"2,450.50","2,500.00","24,505.00","25,000.00","495.00",2.02%,0.50%
    TCS,5,"3,200.00","3,150.00","16,000.00","15,750.00","-250.00",-1.56%,-0.20%
    ,10,100,100,1000,1000,0,0%,0%
    '''
    
    print("Loading sample portfolio...")
    df = load_portfolio(sample_csv)
    
    print("\nValidating DataFrame...")
    is_valid, err = validate_portfolio(df)
    print(f"Is valid? {is_valid}")
    if not is_valid:
        print(f"Error: {err}")
        
    print("\nDataFrame Output:")
    print(df.to_string())
    
    print("\nFormatted Text:")
    print(portfolio_to_text(df))