convert.py

"""
Convert: filename of Gmail contacts -> XLSX file

Converts the contents of a filename where exported Gmail contacts can be found
into an XLSX spreadsheet where only the relevant information of the contacts is
found.
"""
import argparse
import codecs
import os
import os.path

import pandas as pd


OUTFILE_EXT = '.xlsx'
RELEVANT_FIELDS = [
    'First Name', 'Last Name', 'Job Title', 'Company', 'Categories',
    'Mobile Phone', 'Home Address', 'E-mail Address']


def parse_commandline():
    """Parses the command line"""
    parser = argparse.ArgumentParser()
    parser.add_argument('filename')
    return parser.parse_args()


def ensure_codecs(args, tmpfilename):
    """
    Ensures that the file we are reading the DataFrame from is encoded as
    utf-8
    """
    infile = codecs.open(args.filename, 'r', 'latin1')
    contents = infile.read()
    with codecs.open(tmpfilename, 'w', 'utf-8') as tmpfile:
        tmpfile.write(contents)


def filename_without_extension(filename):
    """
    Returns the base name of a filename, removing its extension, e.g.
    'input.csv' results in 'input'
    """
    return os.path.splitext(filename)[0]


def main():
    """Main processing"""
    tmpfilename = 'input.encoded.csv'
    args = parse_commandline()
    ensure_codecs(args, tmpfilename)
    dataframe = pd.read_csv(tmpfilename, index_col=False, header=0)
    outfile = filename_without_extension(args.filename) + OUTFILE_EXT
    out_enconding = 'utf8'
    # pylint: disable=no-member
    dataframe.to_excel(outfile, encoding=out_enconding, columns=RELEVANT_FIELDS)
    # pylint: enable=no-member
    os.remove(tmpfilename)

if __name__ == '__main__':
    main()