gotoextr

README

gotoextr

Small and durty program to extract GPX data from Google Location History.

Usage

1. Download your location history from Google Takeout as zip archive.
2. Run gotoextr with the downloaded archive as argument. For example to extract data for January 1, 2023 run:

gotoextr -s 2023-01-01 takeout-20230501T000000Z-001.zip

The output will be written to the file history_2023-01-01.gpx.

You can also manually extract Records.json and use it as parameter. Once extracted this file is quite big (several hundred MB).

Help message

$ gotoextr.exe -h
gotoextr [version: x.y.z] extract history data from Google Location History.

Usage:
  gotoextr [-h] -s <start> [options] <input>

Options:
  -h --help              Show this screen.
  -s <start>             Start date in YYYY-MM-DD format
  -e <end>               End date in YYYY-MM-DD format [default: <start>]
  -a <accuracy>          Keeps only locations with accuracy less than <accuracy> meters [default: 40]
  -t <tp>                New track if coordinates have less than <tp> digits in common [default: 1]  
  -g <sp>                New segment if coordinates have less than <sp> digits in common [default: 2]
  -f <format>            Output format (gpx|kml|tcx|csv|nmea) [default: gpx]
  -o <output>            Output file name [default: history_<start>_<end>.<format>]
  <input>                Input file name (zip or json)

Examples:
  gotoextr -s 2012-01-01 -e 2012-01-31 -a 40 takeout.zip

Installation

You can download the latest binary from the releases page.

Or you can install it from source:

go install github.com/kpym/gotoextr@latest

Why?

I use my travel history to geotag my photos. I often use a tracking application to record my positions, but occasionally (often?) I forget to launch it. In this case, Google Takout (Location history) helps me by extracting my tracks in GPX format.

Inspiration

This software is strongly inspired by location-history-json-converter 🙏. But since the pyton application is rather slow, I decided to make one in go that is about 10x faster. The original application is also more complete, I only implemented the features I needed.

License

MIT

Documentation

Overview

Read Records.json and extract history data. The input date is json with the following format { "locations": [

{
  "latitudeE7": 506553765,
  "longitudeE7": 30632229,
  "accuracy": 24,
  "timestamp": "2012-01-27T21:14:42.352Z"
  ...
},
  ...

] The file is very large, so we read it using json.Decoder. If the file is .zip (smaller) we try to read it in memory.