ETag not supported to make conditional requests through API?

[instructure-analytics]

I notice that the GitHubObject has an eTag attribute, but I don't see the ability to pass this in to any of the different api calls. Am I missing something or is this a missing feature? I notice it's mentioned in the v2 API docs.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[raphaelyancey]

I also would very much like this to be implemented in PyGithub.

[jovezhong]

+1 If this lib can keep tracking the ETag and sending If-None-Match: last_etag in header, this will make us much easier to keep calling the GitHub REST API without being throttled or getting duplicated events

Example on https://docs.github.com/en/rest/reference/activity

$ curl -I https://api.github.com/users/tater/events

HTTP/2 200
X-Poll-Interval: 60
ETag: "a18c3bded88eb5dbb5c849a489412bf3"

The quotes around the ETag value are important
$ curl -I https://api.github.com/users/tater/events
$ -H 'If-None-Match: "a18c3bded88eb5dbb5c849a489412bf3"'

HTTP/2 304
X-Poll-Interval: 60

[michaelwooley]

I think that this can be achieved (to greater and lesser degrees) by making it easy for users to leverage requests-cache. (That is, I don't think that PyGithub needs to get into the business of cache management.)

In fact, requests-cache has a sample script that demonstrates use with the PyGithub specifically. I like that it works but I'm not crazy about the fact that it clobbers the entire requests package in order to do so. The docs also mention some other issues.

I have a draft branch implementing a second approach to this: master...michaelwooley:PyGithub:michaelwooley/add-custom-session . Basically, this allows someone to pass in a custom requests.Session similar to what happens in python-gitlab.

This colab notebook demos my progress so far.

I'm encountering issues with insuring that the headers are handled correctly, though. In particular:

• rate limit updates. (I actually am less concerned about this currently.) Don't want to pull stale rate limit info from a cached response.
• etag parsing This is something I need to look into more.

If there is still interest in this, I could add further tests and open up an MR.