prep for 2.5 release
reduce dependency on output in cuke scenario
dot placement matters
.not_to in addition to .to_not
- Closes #50.
update Guardfile - use actual regexps, not strings that look like them
Additional improvements to raise_error matcher docs.
Improved docs for raise_error
- Closes #58.
Fix exist matcher so that it uses either #exist? or #exists?
Fix specs that are failing for me on MRI 1.8.7 p330.
We can't count on the order Hash#inspect prints the elements on ruby 1.8.
prep for 2.4.0 release
update expected output in cukes to align with changes in rspec-core
reorg cuke features
remove feature duplicated in rspec-core
more more consistency across failure messages for operator matchers
More consistency across failure messages for operator matchers.
Also cleaned up the related cukes. There was some duplication between
the operator and "be" matcher cukes, and the latter had some examples
that I prefer to discourage.
dev support: add Guardfile
more improvments to RDoc for change matcher
improved documentation for change method
- Closes #51.
dev-support: update relish task
bump version to 2.4.0.pre
prep for 2.3.0 release
align Gemfiles across rspec dev repos
fix warnings on include spec
Define include matcher as diffable
Modify the specs so they match output rather than specify it exactly.
- Closes #49.
use ZenTest 4.4.2 instead of autotest
Changed expectation to use regex.
- Closes #48.
lock autotest at 4.4.5
unless it should be if ...
clean up spec_helper
expect the right message regardless of hash ordering
No need to have separate 1.8 and 1.9 logic for checking existence of …
Fixed spec so it runs on ruby 1.9.1.
Also fixed indentation.