chemistry-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Florian Müller <f...@apache.org>
Subject Re: cancelCheckOut() not being called when running VersionSmokeTest from workbench
Date Fri, 17 Aug 2012 10:47:49 GMT
 Hi Sagara

 Check your server implementation. When the AtomPub binding is used, 
 cancelCheckOut() is not called directly. Instead 
 deleteObjectOrCancelCheckOut() is called. Your code has to test if the 
 provided object is a PWC. If so, you have to cancel the check out.

 The TCK tests fails if the cancellation of the check out was not 
 successful. Subsequent checks and calls fail because the test object is 
 in a wrong state.


 Florian




 Am Freitag, den 17.08.2012, 11:25 +0200 schrieb Sagara Paranagama 
 <sagara89dev@gmail.com>:
> Hi all,
>
> I'm running the VersionSmokeTest and it fails with the following 
> messages.
>
> FAILURE: Versioning Test Group (ATOMPUB)/Versioning Smoke Test: 
> Version
> series is not checked out but cmis:isVersionSeriesCheckedOut is not 
> FALSE!
> INFO: Versioning Test Group (ATOMPUB)/Versioning Smoke Test: True!
> FAILURE: Versioning Test Group (ATOMPUB)/Versioning Smoke Test: 
> Version
> series is not checked out but cmis:versionSeriesCheckedOutId has a 
> value!
> INFO: Versioning Test Group (ATOMPUB)/Versioning Smoke Test: Object 
> is not
> null!
> FAILURE: Versioning Test Group (ATOMPUB)/Versioning Smoke Test: 
> Version
> series is not checked out but cmis:versionSeriesCheckedOutBy has a 
> value!
> INFO: Versioning Test Group (ATOMPUB)/Versioning Smoke Test: Object 
> is not
> null!
> UNEXPECTED_EXCEPTION: Versioning Test Group (ATOMPUB)/Versioning 
> Smoke
> Test: Exception:
> 
> org.apache.chemistry.opencmis.commons.exceptions.CmisConstraintException:
> Document is already checked out
> /cmistck13451920595751879959029/versioningtest.txt
>
> The test creates a document (versioningtest.txt) and performs a 
> checkout.
> The checkout is working fine. (I get a hit in the breakpoints of my
> implementation of AbstractCmisService). However *cancelCheckOut()* 
> and *
> checkIn()* does not seem to be invoked in my my implementation. (No 
> hit in
> breakpoints)
>
> I checked out the source code for the test and the above two methods 
> should
> be invoked in my implementation, am I correct?
>
> And I'm guessing the test fails because cancelCheckOut is not getting
> invoked, and it reports the "version series is not checked out" 
> because it
> assumes cancelCheckout() worked correctly.
> Is there something wrong in my end? Any help on this would be 
> appreciated.
>
> Thanks.
> Sagara


Mime
View raw message