Prerequisites

Download an export of all your data from Endomondo

• The script will work to upload any directory of tcx files, but the main use case for this repo is migrating workouts from Endomondo to Strava
• From within your Endomondo account settings, you can request a full archive download of all of your data. Once requested, it currently seems to take a few days for this to be prepared, but it will eventually be ready to download.
• The script will expect to find a directory called 'data' in the application root. Create this directory 'data', Unzip the downloaded archive, and move the contents of the directory 'endomondo_$date' into it. You should now have a directory 'Workouts' containing the tcx files (as well as json files which contain some additional data so leave these here) in the data directory.

Install node etc and package dependencies

• Tested on node version (14.15.1)

• Recommended to install latest lts version via nvm (https://github.com/nvm-sh/nvm/blob/master/README.md)

  nvm install --lts

  nvm install-latest-npm

• Install dependencies

  npm install

Get a Strava Client ID and Secret

1. Follow instructions on "How to create and account" (https://developers.strava.com/docs/getting-started/#account). You just need to fill:
   1. Application name: StravaImporter
   2. Website: https://github.com/fishion/StravaImporter/
   3. Authorization Callback Domain: localhost
2. Make a note of your Client ID and Client Secret - you'll be prompted for them later. If you prefer, you can export them as environment variables STRAVA_CLIENT_ID and STRAVA_CLIENT_SECRET

run the script

1. run "node index.js" to start the upload script
2. Follow the instructions of the script to give the script appropriate permissions
3. Strava API is rate limited (100 API calls in any 15 min period, 1000 in a day) so depending on how many workouts you have to upload, it may take some time to complete. Script gives plenty of output to confirm when it is rate limited.
4. As script runs, it moves successfully processed files to 'Workouts_Success'. Any failures will be moved to 'Workout_Failures'. This way, the script can be safely stopped part way through and restarted if necessary.

NOTE : Activity mapping

• TCX format files only seem to support ‘biking’, ‘running’, ‘hiking’, ‘walking’ and ‘swimming’ activities with everything else marked as 'other'. These are only activity values recognised by Strava tcx imports and the only activities identified in the Endomondo tcx exports.
• The correct activity names are in fact exported alongside the tcx files in json files. Strava doesn't support importing these JSON files directly.
• The script will first attempt to upload all the tcx files, and then it will read through all the json files, extract the correct activities, map them to activity types supported in Strava, and then update the activities in Strava to the correct type. Strava will create default names for the activities such as 'Morning Run', and the script attempts to update these to the correct activity type too.

Run with docker

Build the image locally:

$ docker build . -t strava-importer

Run the image with your data and the Strava client credentials:

$ docker run \
  -it --rm \
  -v "$PWD/data":/usr/src/app/data \
  -e STRAVA_CLIENT_ID=$YOUR_APP_CLIENT_ID \
  -e STRAVA_CLIENT_SECRET=$YOUR_APP_CLIENT_SECRET \
  -w /usr/src/app \
  strava-importer \
  node index.js