On Sun, Jun 6, 2010 at 8:42 AM, Vassil Dichev <[email protected]> wrote: > David has fixed the issue: > > http://www.assembla.com/spaces/liftweb/tickets/544 > > but broken other stuff in preparation for the new Lift 2.0 release. > OTOH, I can also revert the last fix to the tests (annotating the > get/post methods with type signatures), which is will make them more > readable.
How much work is the shift to the "return signature for the body of the response to Box[Array[Byte]]"? Might be easier to shift now and start getting ready for the 2.0 release. D. > > > On Thu, Jun 3, 2010 at 10:59 PM, Vassil Dichev <[email protected]> wrote: >> No sooner than you've written your mail, David has assigned the >> ticket. I don't expect it to take much considering the working version >> was there before the refactoring. >> >> On Thu, Jun 3, 2010 at 10:36 PM, Richard Hirsch <[email protected]> >> wrote: >>> Thanks for finding out what the problem is. >>> >>> Let's see how @dpp and the rest of the Lift team responds >>> >>> D. >>> >>> On Thu, Jun 3, 2010 at 9:00 PM, Vassil Dichev <[email protected]> wrote: >>>> There, I've debugged the issue to a refactoring in Lift and submitted a >>>> ticket: >>>> >>>> https://www.assembla.com/spaces/liftweb/tickets/544-testframework-throws-nullpointerexception-on-empty-body >>>> >>>> >>>> On Thu, Jun 3, 2010 at 8:18 AM, Vassil Dichev <[email protected]> wrote: >>>>> It seems that whenever empty content is expected, a >>>>> NullPointerException is thrown in the Lift helpers, because the whole >>>>> input stream is null. I'll try and see in the next days if this is >>>>> something that needs to be fixed in Lift or just the test mocks (which >>>>> were heavily refactored the last couple of days). >>>>> >>>>> >>>>> On Tue, Jun 1, 2010 at 4:01 PM, Vassil Dichev <[email protected]> wrote: >>>>>> Very possible, I noticed this too, but didn't have time to >>>>>> investigate. I'll find the last working build and check the changes. >>>>>> >>>>>> >>>>>> On Tue, Jun 1, 2010 at 3:29 PM, Richard Hirsch <[email protected]> >>>>>> wrote: >>>>>>> We now get some strange test failures - they are associated with GETs >>>>>>> instead of posts. >>>>>>> >>>>>>> Could there be a link with changes in the Lift Testkit? >>>>>>> >>>>>>> Tests in error: >>>>>>> API2 should /user/messages GET when no new messages exist, returns >>>>>>> 204 (no content) >>>>>>> API2 should /user/messages?timeout=2 GET when no new messages exist, >>>>>>> returns 204 (no content) >>>>>>> API2 should /pools/POOLID/messages GET with valid session and new >>>>>>> messages >>>>>>> API2 should /pools/POOLID/messages GET when no new messages exist, >>>>>>> returns 204 (no content) >>>>>>> API2 should /pools/POOLID/messages?timeout=2 GET with valid session >>>>>>> API2 should /pools/POOLID/messages?timeout=2 GET when no new >>>>>>> messages exist, returns 204 (no content) >>>>>>> >>>>>>> >>>>>>> On Tue, Jun 1, 2010 at 1:36 PM, Richard Hirsch <[email protected]> >>>>>>> wrote: >>>>>>>> Thanks - this fixed the problem >>>>>>>> >>>>>>>> On Tue, Jun 1, 2010 at 11:53 AM, Vassil Dichev <[email protected]> >>>>>>>> wrote: >>>>>>>>> I've committed a fix, try it now please. >>>>>>>>> >>>>>>>>> It should be easy to modify if there's a more elegant solution to >>>>>>>>> this problem. >>>>>>>>> >>>>>>>>> Vassil >>>>>>>>> >>>>>>>>> >>>>>>>>> On Tue, Jun 1, 2010 at 11:53 AM, Vassil Dichev <[email protected]> >>>>>>>>> wrote: >>>>>>>>>> Due to a refactoring it seems that now the type inferencer cannot put >>>>>>>>>> the right implicit and we must provide it a type hint. One possible >>>>>>>>>> fix is for all get and post methods to be replaced with: >>>>>>>>>> >>>>>>>>>> get[Response] >>>>>>>>>> post[Response] >>>>>>>>>> >>>>>>>>>> I'll try to submit a fix soon. >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> On Mon, May 31, 2010 at 2:22 PM, Richard Hirsch >>>>>>>>>> <[email protected]> wrote: >>>>>>>>>>> We are currently experiencing some errors during our compile. >>>>>>>>>>> >>>>>>>>>>> I think it is assocaited with changes in the lift Testkit. >>>>>>>>>>> >>>>>>>>>>> I've tried to find a solution but haven't been successful. Can >>>>>>>>>>> somebody else take a look and see if they spot the problem. >>>>>>>>>>> >>>>>>>>>>> D. >>>>>>>>>>> >>>>>>>>>>> [INFO] Compiling 7 source files to >>>>>>>>>>> /export/home/hudson/hudson-slave/workspace/ESME/server/target/test-classes >>>>>>>>>>> [WARNING] >>>>>>>>>>> /export/home/hudson/hudson-slave/workspace/ESME/server/src/test/scala/org/apache/esme/api/ApiTest.scala:84: >>>>>>>>>>> error: no implicit argument matching parameter type (String, >>>>>>>>>>> org.apache.commons.httpclient.HttpClient, >>>>>>>>>>> org.apache.commons.httpclient.HttpMethodBase) => T was found. >>>>>>>>>>> [WARNING] login <- post("/api/login", "token" -> token) !@ "Failed >>>>>>>>>>> to >>>>>>>>>>> log in" if (testSuccess(login)) >>>>>>>>>>> [WARNING] ^ [WARNING] >>>>>>>>>>> /export/home/hudson/hudson-slave/workspace/ESME/server/src/test/scala/org/apache/esme/api/API2Test.scala:51: >>>>>>>>>>> error: no implicit argument matching parameter type (String, >>>>>>>>>>> org.apache.commons.httpclient.HttpClient, >>>>>>>>>>> org.apache.commons.httpclient.HttpMethodBase) => Nothing was found. >>>>>>>>>>> [WARNING] val post_session = post("session", "token" -> token) >>>>>>>>>>> [WARNING] ^ [WARNING] >>>>>>>>>>> /export/home/hudson/hudson-slave/workspace/ESME/server/src/test/scala/org/apache/esme/api/API2Test.scala:94: >>>>>>>>>>> error: no implicit argument matching parameter type (String, >>>>>>>>>>> org.apache.commons.httpclient.HttpClient, >>>>>>>>>>> org.apache.commons.httpclient.HttpMethodBase) => T was found. >>>>>>>>>>> [WARNING] session <- post("session", "token" -> "000000") [WARNING] >>>>>>>>>>> ^ >>>>>>>>>>> [WARNING] >>>>>>>>>>> three errors found >>>>>>>>>>> >>>>>>>>>> >>>>>>>>> >>>>>>>> >>>>>>> >>>>>> >>>>> >>>> >>> >> >
