Quick hacky script to extract date and 2D length data from GPX files to CSV using gpxpy; at least compatible with Py2.7/3.4

extract_gpx_data.py

#!/usr/bin/env python
""" Extract date and 2D length data from GPX files in current directory. """
from __future__ import division, print_function
import decimal
import glob

import gpxpy


def main():
    gpx_filenames = glob.glob("*.gpx")
    print("Filename,Date (YYYY-MM-DD),Time,"
          "Length (km),ISO year,ISO week number")
    for gpx_filename in gpx_filenames:
        with open(gpx_filename, 'r') as f:
            gpx = gpxpy.parse(f.read())
            for track in gpx.tracks:
                extract_track_data(track, gpx_filename)


def extract_track_data(track, gpx_filename):
    """ Take GPX track and string filename; print comma separated data. """
    start_datetime, _ = track.get_time_bounds()
    iso_year, iso_week_number, _  = start_datetime.isocalendar()
    km_2d_length = decimal.Decimal(track.length_2d()/1000)
    output = ','.join([gpx_filename,
                       str(start_datetime.date()),
                       str(start_datetime.time()),
                       str(round(km_2d_length, 2)),
                       str(iso_year),
                       str(iso_week_number)])
    print(output)

if __name__ == '__main__':
    main()

TODO: maybe do the counting of data I want in this script, rather than using Pivot Tables in LibreOffice; or maybe using Jupyter is a better solution.

(My current use of this is finding out my total run distance per week — which is a Pivot Table of ISO week number and ISO year with sum of length as data — and the number of days I've run, which is a Pivot Table of ISO week number with unique count of date as data; this is generated separately using COUNTIF via https://stackoverflow.com/questions/11876238/simple-pivot-table-to-count-unique-values and ISO week number/count of unique count as data.) That works but is messing you have to do after extraction, which suggests doing it all in Python is nicer.