Hmmm - to me an assertion should not have a side effect, so Groovy 2.x
rejecting code like that, where the side effect is easily detected,
seems proper to me...
(The fact that Groovy has a way of interpreting the non-Boolean result
as Boolean through the Groovy-truth rules you mention is of course
necessary to allow this statement to compile; that does however not
necessarily make it sufficient)
Cheers,
mg
On 08/01/2020 22:06, Søren Berg Glasius wrote:
Hi Marcin,
I think it makes sense, that it does not fail, as a +=1 returns 2
which by Groovys asBoolean is true. I am more inclined to say that
Groovy 2.x had a bug, that was accepted as a compile error.
Just my two cents worth.
Best regards / Med venlig hilsen,
Søren Berg Glasius
Hedevej 1, Gl. Rye, 8680 Ry, Denmark
Mobile: +45 40 44 91 88, Skype: sbglasius
--- Press ESC once to quit - twice to save the changes.
Den ons. 8. jan. 2020 kl. 12.16 skrev Marcin Zajączkowski
<msz...@wp.pl <mailto:msz...@wp.pl>>:
Hi,
Playing with the failures which Groovy 3 brought to Spock tests
itself,
I noticed that Groovy 3 no longer rejects += and -= if used in an
assertion statement:
> def a = 1
> assert a += 1 //passes
> assert a == 2 //passes
While it behaves consistently, I wonder if that change it was
purposely
(it was failing in Groovy 2) or the check should be enhanced:
https://github.com/apache/groovy/commit/a17b9eb85835a5eaa306a893a0a8bcd51a49e299#diff-05fe8708ce3207739a2cb39a97426e8dR345
?
Marcin
--
https://blog.solidsoft.info/ - Working code is not enough
