Open pc-coholic opened 5 months ago
This issue is now being hopefully fixed in https://github.com/psf/requests/pull/6731
A very quick fix would be to follow https://github.com/mikf/gallery-dl/commit/6cfbc1071fd88ed94776aa6f81f1582cda7fa5ab suit and place this within _create_sslcontext()
. But since an update for requests is probably imminent, we could also just close this issue and wait for the issue to disappear by itself :-)
Ok
@vog Sorry, but this problem is still current with requests 2.32.0
:
Could you reopen it?
Hi, the issue still persists with requests 2.32.3. I think this issue should be reopened
Ok, reopened. Pull requests are welcome.
@vog I'll be happy to provide a one-line pull request to fix the issue.
But I was wondering if you'd prefer putting in a workaround (which will be superfluous once 2.32.4 is released/the above linked PR is merged) or just wait for said PR to land in requests....
Never mind, we quickly added the fix of
on our own, and released requests_pkcs12 version 1.25:
Our local test suite runs fine with it.
Please check whether this solves the issue for you as well.
Works for me, was exactly what I was experimenting with, too :)
Thanks for your quick feedback. :+1:
Leaving this issue open as a reminder to revert that workaround as soon as
This might be a false positive - but from preliminary experimentation, it seems like a recent change in requests might cause issues with requests_pkcs12.
With requests 2.32.0, a breaking change was introduced that returned every call with that involved requests_pkcs12 with
SSLContext via a Transport Adapter
.requests 2.32.3 (still unreleased as of right now, apparently scheduled for Tuesday) aims to fix this issue (Allow for overriding of specific pool key params #6716), but now all calls involving requests_pkcs12 fail with
SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate
for me.I'll try to dig a little deeper and see if this is indeed something that requires changes here - I just wanted to provide a heads-up in case 2.32.3 is indeed shipping a breaking change.
Minimal example: