Ryan Barry has posted comments on this change.

Change subject: Use a better URL validator. RegexValidator adds flags
......................................................................


Patch Set 2:

Changing the __init__ of the validator isn't strictly necessary (we could use 
the same strategy as FQDN, which also used a flag). It simply seemed more 
obvious to do it explicitly than to expand a tuple, and provides a clear 
pattern for adding flags to other validators (if we need them)

There's not a practical solution for EL6. urlparse sees "http://foo::"; as part 
of the netloc, which is incorrect. We could explicitly cover the double colon 
case, but that leads us down a potential rabbit hole of someone finding another 
case where the URL validator breaks and us chasing it down. I can already tell 
you that urlparse doesn't have a problem with "--" in a domain, though it's 
invalid, and we'd need to catch that case as well.

urllib3 is a viable option once EL6 support is over. And I agree that this 
would be difficult to modify. But validating URLs correctly is very hard, and 
passing the existing doctests is a good indicator that it won't need to be 
modified anywhere. Additionally, since the regex came from the django project 
(I didn't write it), updating is easy.

I added a very complete doctest suite.

-- 
To view, visit https://gerrit.ovirt.org/42394
To unsubscribe, visit https://gerrit.ovirt.org/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I13429d0cc600446c8fd4c1be5e5a8752d6d992f7
Gerrit-PatchSet: 2
Gerrit-Project: ovirt-node
Gerrit-Branch: master
Gerrit-Owner: Ryan Barry <[email protected]>
Gerrit-Reviewer: Fabian Deutsch <[email protected]>
Gerrit-Reviewer: Jenkins CI
Gerrit-Reviewer: Ryan Barry <[email protected]>
Gerrit-Reviewer: [email protected]
Gerrit-HasComments: No
_______________________________________________
node-patches mailing list
[email protected]
http://lists.ovirt.org/mailman/listinfo/node-patches

Reply via email to