[rspec-devel] RSpec inside JRuby
ola.bini at gmail.com
Wed May 23 12:29:22 EDT 2007
Nick Sieger wrote:
> On 5/23/07, aslak hellesoy <aslak.hellesoy at gmail.com> wrote:
>> Hi all,
>> The fine JRuby team declared at RailsConf that they will bundle RSpec
>> with JRuby when it ships. That's awesome, thanks again guys.
>> I hope this can be the RSpec 1.0.1 version (not yet released) - it has
>> some minor tweaks that are needed to get all of RSpec's specs to run
>> on JRuby. (This is something Ola Bini and I paired on at RailsConf).
>> We'll release 1.0.1 if the JRuby team ask us for it.
>> I am currently able to run *all* of RSpec core's specs on JRuby (HEAD).
>> LET'S KEEP IT THAT WAY.
>> I'm asking that all RSpec developers make sure RSpec always runs
>> against JRuby until they ship 1.0. Here is how:
>> svn co http://svn.codehaus.org/jruby jruby
>> cd jruby/jruby
>> ant clean
>> # put JRuby's bin dir on your PATH
>> cd ../testsuites/rspec
>> mkdir target
>> jruby -S rake checkout_code
>> cd target/rspec
>> jruby bin/spec spec -c
>> I'm getting a bunch of hpricot warnings but otherwise it's 100%
>> gems/hpricot-0.5.140-jruby/lib/hpricot/parse.rb:15 warning: multiple
>> values for a block parameter (4 for 1)
> This is also in CI: http://jruby.thresher.com/browse/RSPEC-TRUNK
> Although since RSpec checkout is part of the build, it won't get run
> when RSpec code changes at the moment.
> Right now there are 5 failures which come and go, it seems like we
> (JRuby) have some randomly recurring issue with Object#inspect or
> something like that, because the actual output only differs by #<...>.
> One of us will have to look at it.
I actually believe this has something to do with the "hex" string output
from Object#inspect not always being the same length. Sometimes it's one
char shorter or longer. If there's any regex or similar in RSpec that
looks at this, that may be the cause of these failures.
Ola Bini (http://ola-bini.blogspot.com)
JvYAML, RbYAML, JRuby and Jatha contributor
System Developer, Karolinska Institutet (http://www.ki.se)
OLogix Consulting (http://www.ologix.com)
"Yields falsehood when quined" yields falsehood when quined.
More information about the rspec-devel