I would like to use the spec DSL inside my own parent class which may include helpers, etc. Here is a an example. https://gist.github.com/793330
This small change allows the describe_stack to set the first class to the parent class if it is a subclass of MiniTest::Spec before falling back to MiniTest::Spec itself. This way testers can subclass MiniTest::Spec with their own helpers and presumable their own framework on top of MiniTest::Spec without extending the base.
synchronizer apparently doesn't do deletions correctly
Well, I can confirm with this update, I no longer get any errors with 'rake test'.
Yeah... sorry about that. p4 -> git sync script apparently has a bug.
- Benchmarks in specs that didn't call bench_range would die. (zzak).
+ Completed doco on minitest/benchmark for specs.
[git-p4: depot-paths = "//src/minitest/dev/": change = 6086]
Add RDoc for MiniTest::Mock
[git-p4: depot-paths = "//src/minitest/dev/": change = 6089]
prepped for release
[git-p4: depot-paths = "//src/minitest/dev/": change = 6093]
- Fixed a new warning in 1.9.3.
[git-p4: depot-paths = "//src/minitest/dev/": change = 6095]
Allow class scopes to be used with spec dsl.
+1 I would have expected MiniTest::Spec to behave the same way that gist intended it to.
good luck but i doubt you're going to get much love from here. from the seattlerb profile page "HEY! We do NOT develop in git, so pull requests will be rejected. Submit patches via the bug reporter on rubyforge.org." personally i'll just continue keeping a fork up to date and manually applying your patch as long as possible...
I did create a ticket on rubyforge and linked the diff to this page.
And if that is not enough I just created a "patch" on the "tracker" of rubyforge. C'mon :) Aaron told me he would ask Ryan Davis if he could get on this since he finally started moving rails tests over to MiniTest::Spec as well and the first person that tried to use a describe blog would go WTF.
+1 especially after seeing this gist: https://gist.github.com/821558
@joslynesser: that gist works w/o this patch just fine, so I don't see what your point is.
@metaskills: I can't read/understand wtf is going on with this pull request. I took your tests and completely munged them into a proper reflective test (like test_structure) and made it pass.
The gist above posted by @joslynesser does work without the patch, but that is not the point. As I told @tenderlove, it will break the first time you try to use a helper in a describe block, for example "people(:one)" in a describe block will be a NoMethodError.
@zenspider Agreed, not sure WTF is happening with the pull request. This was my simple commit in my repo.
@zenspider: Sorry I didn't clarify. I meant to +1 this pull because of meta's comments in that gist, not that the gist itself failed.