The original intention (from memory) was that an IOException that occurred trying to produce the diff message should potentially be ignored. After several rounds of refactoring, I don't see that actually being the case anymore. I think you can treat the test as an over-optimistic coverage test, i.e. delete it but then add in whatever is required to reach 100% coverage at the end.
I played around with javaxdelta to do a binary diff. It seems to work well for large changes but seems to sometimes miss changes of less than 16 bytes in length due to the algorithm used. I am still sorting out whether that will work. This only affects how useful the reporting and log messages are rather than functionality. Cheers, Paul. Marc Guillemot wrote:
Hi, compareToExpected doesn't report any error when an IOException occurs during reference file retrieval and comparison. This seems to be work as designed because this is tested with a unit test. I don't understand what it brings and I'd like to throw a StepExecutionException if an IOException occurs. This would for instance avoid to have successfull tests, when the reference file is not found. Marc. -- View this message in context: http://www.nabble.com/-dev--compareToExpected%3A-why-are-IOException-ignored-t1810164.html#a4933126 Sent from the WebTest forum at Nabble.com. _______________________________________________ WebTest mailing list [email protected] http://lists.canoo.com/mailman/listinfo/webtest
_______________________________________________ WebTest mailing list [email protected] http://lists.canoo.com/mailman/listinfo/webtest
