API creation of tasks via XML

[Kagilum]

Hi,

I’m unable to create tasks using the API.

I’ve been able to create stories, update stories, accept stories, plan stories to a sprint, get all features, get all sprints, and get all tasks using the API fairly well.

Here are 3 cURL requests that I made in series (with some information replaced):

curl -H «Accept: application/xml» -u user:password -k –request GET http://xxxxxxxx/icescrum/ws/p/PKEY/story/83

curl -H «Accept: application/xml» -u user:password -k –request GET http://xxxxxxxx/icescrum/ws/p/PKEY/sprint/3

curl -H «Accept: application/xml» -u user:password -k –request POST –verbose http://xxxxxxxx/icescrum/ws/p/PKEY/task -d «833CLAIM TASK<
/name>«

The first 2 were successful; however, the 3rd request which is were I attempted to create the task had failed, and I believe it printed a stack trace. Here’s part of the response:

Grails Runtime Exception


Error Details






Error 500: Executing action [save] of controlle

r [org.icescrum.web.presentation.app.TaskController]  caused exception: Cannot i

nvoke method remove() on null object


Servlet: grails


URI: /icescrum/grails/task/save.dispatch




Exception Message: Cannot invoke method

remove() on null object 


Caused by: Cannot invoke method remove(

) on null object 


Class: CorsFilter 


At Line: [66] 


Code Snippet:














Stack Trace




java.lang.NullPointerException: Cannot invoke method remove()


   at org.icescrum.web.presentation.app.TaskController$_closure4.doCall(Tas


   at org.icescrum.web.presentation.app.TaskController$_closure4.doCall(Tas


   at grails.plugin.springcache.web.GrailsFragmentCachingFilter.doFilter(Gr


   at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExec


   at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor



java.lang.Thread.run(Thread.java:662)




Any idea what I could have been doing wrong? Is there any more information that would be helpful for me to provide?
Thanks,

Michael

		
	

[Nicolas Noullet]

Hello,

There was two problems:
– The Accept header was set but the Content-Type one was required. This one is very strange because the library we use should treat the Accept and the Content-Type headers in the same way (which is not a good thing, but as far as I know it was enforced). I must give another look at this problem.
– The id of sub-resources must be included as attributes
The documentation was wrong about that so we have fixed it.

Here is an example of a working curl command that creates a task :

curl -H "Content-Type: application/xml" -u user:password -k --request POST --verbose http://xxxxxxxx/icescrum/ws/p/PKEY/task -d "CLAIM TASK"

[Kagilum]

Thank you! I’m currently circumventing this issue by just using JSON, but it seemed like it was a problem with data returned as XML attributes. Hopefully, this will fix the problem I’ve been having.

I was already setting both accept and content-type as «application/xml» in my code for all requests, but I had only been specifying one in my curl requests.
I also had issues with feature.id and sprint.id in creating stories and updating stories, which I have been circumventing by using the separate 2 requests for specifically associating a feature or planning in a sprint, but this probably resolves those issues as well.

[Autor]

Entradas