use clj-http's wrap-cookies instead of ring's because formats have diverged#15
use clj-http's wrap-cookies instead of ring's because formats have diverged#15bmaddy wants to merge 1 commit intotailrecursion:masterfrom
Conversation
|
👍 my understanding of all of the cookie fields is a little rusty, but I do think for a proxy server, being a bit "looser" with spec adherence makes sense. Alternatively, one could delete the keys in the |
|
Just curious @webmonarch is there some way to override this implementation easily as you mentioned above? I want to make sure I'm not missing a neat piece of clojure. The only thing I can think of is a fork in my own repo to implement the desired cookie behavior. |
|
hey @aft-luke, yeah, nothing neat that I know of. I forked for our internal projects. The code is pretty small so a good ol' copy + paste would probably work too. Good luck! |
|
Thanks @webmonarch. I ended up writing my own handler that's 80% copy/paste as well. |
This switches to use clj-http's
wrap-cookiesinstead of ring's because their formats have diverged. clj-http returns cookies something like this:but ring no longer accepts
:discardand:versionas shown here: ring-clojure/ring@c05004f#diff-c055e0aeef031084c68bf6da304ed4f7L25This leads to the following error when there is cookie info in the response: