Otis Gospodnetic schrieb:
Hello Christoph,
I don't know enough to judge all changes you made with confidence, and don't have the code on this machine to apply the patch and look at it properly. However, it looks like seekInternal(long) should now be left with only the conditionals that can cause IOException, and nothing else. Why do we need to keep the body of the method then?
I discussed that already with Dmitry and he was not very passionate regarding these conditionals. Finally I kept them since they were required by Dmitry's test. But throwing an exception on seek is not a behavior implemented by the other subclasses of InputStream. I decided to remove the conditionals and slightly changed the unit test.
If more index checks are desirable, I would put them into InputStream. Despite that I added a "read past EOF" check to CSInputStream.readInternal to guarantee the same behavior as FSInputStream.
I also see getFilePointer() call, but can't tell where that method is. I'll assume it's from the parent InputStream.
yes.
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
