|
History
| Log In
|
|
|
|
Found what I believe is a bug related to this one :
This works : [1,2,3][1..-1] This doesn't : LinkedList.new([1,2,3])[1..-1] Nick Sieger's comment : This seems like a Java integration issue, that the Java list isn't appropriately getting wrapped with Ruby array behavior. But, I'm not sure that it was ever meant to, so it could be a case of a missing feature. Could you add a comment to JRUBY-441? This patch causes the class on which constructors aren't found to be displayed in the exception message. Currently the exception merely reports that no constructors were found.
Another test case:
From the user list: passing a Ruby array as a java string array Bumping to 1.0 again...and this is in danger of falling off 1.0. If there's no additional patches for errors we'll just go with what we have and call this "done" enough for now. But I don't want to add this in RC2 right now.
Bumping to 1.x. The only patch provided for this doesn't apply cleanly anymore, and there's been a lot of rework for Java integration since this bug was reported. So I think it needs to be reexamined, and the final RC cycle doesn't seem like the right time to do it.
We should finally resolve this in 1.1, either with stock JI and better error handling or with Bill's new uberintegration.
Removing target release from issues that fit any of the following criteria:
|
|||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
Bug
Open
Major
Java integration code does not report errors well
Overriding RaiseException.toString() to print out the ruby backtrace is similar to what jython does, but it also goes against the description of Throwable.toString() as being a short description of the Exception.