[BioRuby] Travis-CI & bundler problem

[Iain Barnett]

On 7 Nov 2013, at 17:13, Joachim Baran wrote:

> Hi!
> 
>   Thanks for putting work into this!

No bother.

> 
>   So, you are getting rspec to fail on your local machine? That is interesting, because I might have some file lingering around on my machine that makes it work for me then!
> 
> Joachim

Try sandboxing the bundle:

    bundle install --binstubs --path vendor

will install all executables in a bin dir inside the project root directory, and all the libraries into project-root/vendor/ruby/VERSION/... If you then use

    bundle exec rspec

or

    bin/rspec

it should use only the local version and not touch any other installed files.

Regards,
Iain


> 
> 
> 
> On 7 November 2013 11:48, Iain Barnett <iainspeed at gmail.com> wrote:
> 
> On 7 Nov 2013, at 16:03, Joachim Baran wrote:
> 
>> Hi!
>> 
>> On 7 November 2013 10:20, Iain Barnett <iainspeed at gmail.com> wrote:
>> How long do the specs normally take to run for you? I started them a while ago (using Ruby v2) and they seem to be stuck, which would indicate to me that I may end up with a blown stack.
>>   Normally the tests run for a little bit over a minute. The current tests that have the stack problem (not all of them have it) fail in under a minute.
>>  
>> Can you post a link to one of these so I can see what you mean about it being a TravisCI problem?
>>   I have not done too much research into this, but I got the impression that it is a set-up problem on TravisCI's side by reading through the comments of this report: http://jira.codehaus.org/browse/JRUBY-6203
>> 
>> Joachim
>> 
> 
> I've changed some of the specs and code around and now I can get the specs to run, but I get a failure and also a stack level error, using v1.9.3. I've pushed the changes to https://github.com/yb66/BioInterchange/tree/rspec-changes.
> 
> It's quite possible I've broken something with my changes - especially for 1.8, which I've not run this against - but there's nothing that should be fundamental though as I've not touched the workings of the library, just the way it loads. Let me know if you see anything wrong since you know the code a lot better, and I'll change it and try again.
> 
> 
> I get the following output on running rspec:
> 
> <snip>
> 
> BioInterchange::Phylogenetics::CDAORDFWriter
>   serialization of tree models
>     Given an empty document
>       should be zero
>     Given a single Newick tree
>       should eq 103 (FAILED - 1)
> 
> <snip>
> 
> BioInterchange::TextMining::RDFWriter
>   serialization of text-mining documents
>     empty document
>     document with two entities (FAILED - 2)
>     full advanced json document (FAILED - 3)
> 
> Failures:
> 
>   1) BioInterchange::Phylogenetics::CDAORDFWriter serialization of tree models Given a single Newick tree 
>      Failure/Error: it { should eq 103 }
>        
>        expected: 103
>             got: 0
>        
>        (compared using ==)
>      # ./spec/phylogenetics_spec.rb:79:in `block (4 levels) in <top (required)>'
> 
>   2) BioInterchange::TextMining::RDFWriter serialization of text-mining documents document with two entities
>      Failure/Error: Unable to find matching line from backtrace
>      SystemStackError:
>        stack level too deep
>      # /Users/iainuser/RubyProjects/BioInterchange/vendor/ruby/2.0.0/gems/rdf-1.0.10.1/lib/rdf/query/pattern.rb:36
> 
>   3) BioInterchange::TextMining::RDFWriter serialization of text-mining documents full advanced json document
>      Failure/Error: Unable to find matching line from backtrace
>      SystemStackError:
>        stack level too deep
>      # /Users/iainuser/RubyProjects/BioInterchange/vendor/ruby/2.0.0/gems/rdf-1.0.10.1/lib/rdf/query/pattern.rb:36
> 
> 
> 
> Regards,
> Iain
> 
>