Implemented cookies - closes issue 82: https://github.com/mikeal/request/issues/82 #102
Conversation
| @@ -406,6 +421,15 @@ Request.prototype.request = function () { | |||
| response.body = JSON.parse(response.body) | |||
| } catch (e) {} | |||
| } | |||
| if (response.statusCode == 200) { | |||
| // only add cookies if no custom jar is defined (by the user) | |||
There was a problem hiding this comment.
No need to 'remember' the cookie if the user wants to go 'stealth' - by setting the jar option
|
Some things I forgot to mention:
|
| @@ -137,6 +140,18 @@ Request.prototype.request = function () { | |||
| setHost = true | |||
| } | |||
|
|
|||
| // if user tries to define a bad cookie jar | |||
There was a problem hiding this comment.
instanceof checks are harmful and unnecessarily restrictive.
assume people are satisfying enough of the contract with the object they pass in, failure will be obvious if they don't.
There was a problem hiding this comment.
that's right, thing will fail anyway so no need for that
|
i am slightly confused as to the usage of the jar option. in my original api i had assume that the jar would hold the state of any cookies set. if no jar was passed (which is default) then cookie support would be off. i see a few checks where jar not being set actually turns on cookie handling which is confusing so i would like a little clarification. i'm also wondering if this jar implementation handles cookies set for different domains and is domain aware? if not, that's not the end of the world so long as someone can add support later. |
|
i see the support for domains in the cookie implementation, nice. i also like the request.cookie() and jar.add(cookie) stuff, very nice. still confused though why setting the jar stops cookies from being set in it. |
|
Well I thought cookies should be enabled by default (I thought people expect request to behave like the browser here) and if they really don't want cookies (or they want specific cookies) the default option should help with that. If you want I can change this to be the other way around, I just thought it would be more helpful like this. |
|
mainly, i think that if self.jar is set that we should still set new cookiies on it. you might want to keep two jars around for different requests, so you'll want them to be set by the setcookie header calls. if someone wants to only use cookies they set explicitly they could just override the add() function in their jar object. i'm conflicted about using a global jar by default. i guess it's fine, but setting jar:false should work to disable cookie handling. |
|
Should be as you want now. |
|
i'm sick right now so i don't have a lot of time to review and test this. i'm going to merge it but hold off on pushing a new release to npm for a few days until i feel better and can sit down with it. thanks for the patch, i've been wanting this for a while. |
Implemented cookies - closes issue 82: #82
ghost
mentioned this pull request
Implemented cookies - closes issue 82: #82
I've also added tests and updated the README file.