Corning Credit Union parser

ccu.py

"""
PURPOSE: Parse a pile of 'spreadsheet' files downloaded from 
         Corning Credit Union eBranch.

FEAUTURES: - Parse and combine financial records from a list of raw
             files, converting the data into convenient types.
           - Plot stuff.

EXAMPLE USAGE:
import ccu
checking = ccu.read_checking(ccu.list_files('Downloads/checking/'))
ccu.running_balance(checking) # It's a plot!
"""

import os
import re
import decimal as dec
import pandas as pd
import matplotlib.pyplot as plt

dec.getcontext().prec = 2 # a datatype for money (2 decimal places)

def cast_money(s):
    pattern = '(\-?)\$?([1-9]+\.[0-9][0-9])'
    try:
        sign, stripped = re.search(pattern, s).group(1,2)
    except AttributeError:
        return s
    if sign:
        stripped = '-' + stripped
    return dec.Decimal(stripped)

def read_visa(filenames):
    """Read CSV files downloaded from CCU. Parse them, 
    concatenate all their data,
    and return a DataFrame array sorted by date."""
    if type(filenames) is str:
        filenames = list(filenames) # a single-element list
    df = pd.concat(
        [pd.read_csv(filename,
                     names=['date', 'description', 
                            'amount', 'fee',
                            'interest', 'principal', 
                            'balance','placeholder'],
                     parse_dates=[0],
                     converters = {'amount': cast_money,
                                   'fee': cast_money,
                                   'interest': cast_money,
                                   'principal': cast_money,
                                   'balance': cast_money})
         for filename in filenames])
    del df['placeholder'] # hanlded CCU's trailing comma
    return df.sort('date')

def read_checking(filenames):
    """Read CSV files downloaded from CCU. Parse them,
    concatenate all their data,
    and return a DataFrame array sorted by date."""
    if type(filenames) is str:
        filenames = list(filenames) # a single-element list
    df = pd.concat(
        [pd.read_csv(filename,
                     names=['date', 'description', 
                            'amount', 'balance',
                            'placeholder'], 
                     parse_dates=[0],
                     converters = {'amount': cast_money,
                                   'balance': cast_money})
         for filename in filenames])
    del df['placeholder'] # hanlded CCU's trailing comma
    return df.sort('date')

def running_balance(df):
    "Plot 'balance' versus 'date'."
    plt.plot(df['date'], df['balance'])
    plt.gcf().autofmt_xdate()
    plt.show()

def list_files(dir):
    "List the absolute paths of the files in a directory."
    return [os.path.join(dir, f) for f in os.listdir(dir)]