On Mon, 2013-05-13 at 20:37 +0200, Stefano Bagnara wrote: > 2013/5/13 Oleg Kalnichevski <[email protected]>: > > On Mon, 2013-05-13 at 15:34 +0200, Stefano Bagnara wrote: > > Prior to refactoring the test suites in question made no attempts to > > release file resources allocated by the tests. File backed resource > > streams were never closed. To me this is quite frivolous, even for test > > code. > > Make sense. I admit I've never cared about resource management in > tests (the JVM free them shutting down). >
Test cases can be run as a part of a larger test suite inside the same JVM and impact other tests. Sloppy resource management is simply a bad idea no matter the context. > > All other aspects of the test cases remained unchanged. I simply > > factored out bits of bootstrapping code that was common to all sample > > message based test cases. > > IIRC the testsuite declaration as a nested class and the code > duplication was needed in order to be able to run single tests in > Eclipse (and maybe in maven, but I don't remember the exact scope of > this statement). So I guess that extracting the common code could have > broken this feature (unfortunately I don't have eclipse on this > computer right now). > > I always found very useful to be able to test a single msg file from > the eclipse testrunner (and other junit testrunners) while I'm > debugging/fixing an issue in the parser or similar things, but I'm not > working anymore on mime4j so feel free to do what you prefer. > > At most I will fix it again if I will have to work again on mime4j. > > Stefano Still works for me. I can still run the test for individual messages in Eclipse after refactoring. Oleg
