Andy, I am supportive of this because we clearly have at least a couple folks in the community who know and are interested in Groovy. I just want to make sure we aren't increasing the burden on people to setup their environments. As long as this is seamless from a Maven user perspective (and be mindful of minimum maven versions) then I'm good with it. I am slightly concerned about maintainability of these unit tests relative to Java knowledge we appear to have in the community.
With the above in mind i am a +1. Thanks Joe On Mon, Jan 4, 2016 at 6:42 PM, Matt Burgess <[email protected]> wrote: > +1, Groovy is a great way to rapidly generate tests, remove boilerplate, and > enable powerful test frameworks like Spock. > > Sent from my iPhone > >> On Jan 4, 2016, at 7:29 PM, Andy LoPresto <[email protected]> wrote: >> >> I am considering writing unit tests in for new development/regression >> testing in Groovy. There are numerous advantages to this [1][2] (such as map >> coercion, relaxed permissions on dependency injection, etc.). Mocking large >> and complex objects, such as NiFiProperties, when only one feature is under >> test is especially easy. I plan to write “Java-style” unit tests, but this >> would also make TDD/BDD frameworks like Spock or Cucumber much easier to use. >> >> I figured before doing this I would poll the community and see if anyone >> strongly objects? In previous situations, I have created a custom Maven >> profile which only runs when triggered (by an environment variable, current >> username, etc.) to avoid polluting the environment of anyone who doesn’t >> want the Groovy test dependencies installed. >> >> Does anyone have thoughts on this? >> >> >> [1] http://www.ibm.com/developerworks/java/library/j-pg11094/index.html >> [2] >> https://keyholesoftware.com/2015/04/13/short-on-time-switch-to-groovy-for-unit-testing/ >> >> >> Andy LoPresto >> [email protected] >> PGP Fingerprint: 70EC B3E5 98A6 5A3F D3C4 BACE 3C6E F65B 2F7D EF69 >>
