Open kilianmh opened 2 months ago
It sounds reasonable. Although I like Lack to be minimal, there's not much sense to reimplement it in Lack.
Should response:set-cookies
be allowed to be of type cl-cookie:cookie-jar
and cl-cookie:cookie
, or only of the two?
Side note: Is there any specific reason, why to not allow (list of) cookie-strings as response:set-cookies
value? See fukamachi/ningle#43
It's not required, but it's better to accept Common Lisp objects other than cl-cookie's objects if it doesn't make much trouble.
Side note: Is there any specific reason, why to not allow (list of) cookie-strings as response:set-cookies value? See https://github.com/fukamachi/ningle/issues/43
No, I don't even remember the reason. I think an association list or hash table suits it.
Should we allow
response:set-cookies
slot in lack/response to be a instance of cl-cookie:cookie / cookie-jar? Then we could reuse cl-cookie:write-set-cookie-header for encoding the cookie-string.