Store the garth session information and upload the fit file via garth

[jrast]

To reduce the requests during garmin login, the garth session information is stored similar to the withing session data. Also garth is used to upload the fit-file and cloudscrapper is no longer required. I don't know if we can get rid of cloudscrapper or if there is still an advantage if the upload is done with cloudscrapper.

As with my other PR: If you like these changes let me know and I can cleanup. For example the directory in which the garth session is stored should be defined at a central place.

[stynoo]

Imho, it looks like this is the way to go. Garth does all the heavy lifting and should be able to handle MFA as well (not sure if extra code is needed to save the oauth1/2 tokens somewhere)
The directory should ideally be the same place where .withings_user.json is stored. Iirc it's /root in the containers.

As for cloudscraper: this was a drop in replacement that fixed the last auth issues. If Garth works without it then I don't see a reason to keep it.

[longstone]

This question should be in the MR before. However, is this method static anymore? We reference garth.* stuff.
If yes, my bad im still kind of beginner in python/pep's.

[longstone]

As with my other PR: If you like these changes let me know and I can cleanup. For example the directory in which the garth session is stored should be defined at a central place.

Yes! I very appreciate every PR making this simple and maintanable.
As @stynoo already said, a file should be where the other files are. Currently /root

[jrast]

With matin/garth#9 merged in garth, the garmin upload can be further simplified. Currently the implementation is more of a proof-of-concept, there are some points I personally would change:

• Create a custom garth client instead of using the global one (might be overkill)
• I don't get why the methods of the GarminConnect class are static, this might be something which was required some time ago. But with the current implementation there are better ways.
• If garth is providing the upload method, the GarminConnect class can be remove entierly and everything can be done in the sync_garmin method directly.

[jrast]

Imho, it looks like this is the way to go. Garth does all the heavy lifting and should be able to handle MFA as well (not sure if extra code is needed to save the oauth1/2 tokens somewhere)

Storing the tokens reduces the network request considerably, making the upload process more robust as less things can fail.

The directory should ideally be the same place where .withings_user.json is stored. Iirc it's /root in the containers.

I will look into this.

[matin]

Garth now natively supports upload as of 0.4.32.

with open("12129115726_ACTIVITY.fit", "rb") as f:
    uploaded = garth.client.upload(f)

Let me know if anyone runs into issues, and I can fix them.

[matin]

@jrast if you decide to do this ...

Create a custom garth client instead of using the global one (might be overkill)

take a look at how garminconnect creates a custom Garth Client vs using the global one.