NaN FloatDomainError after upgrading to 0.14.0

[stevehanson]

After upgrading to 0.14.0, I receive the following error after my tests run, presumably while the SimpleCov report is being generated:

vendor/bundle/gems/simplecov-0.14.0/lib/simplecov/source_file.rb:193:in `round': NaN (FloatDomainError)
  from vendor/bundle/gems/simplecov-0.14.0/lib/simplecov/source_file.rb:193:in `round_float'
  from vendor/bundle/gems/simplecov-0.14.0/lib/simplecov/source_file.rb:130:in `covered_strength'
  from vendor/bundle/gems/simplecov-0.14.0/lib/simplecov/file_list.rb:56:in `block in covered_strength'
  from vendor/bundle/gems/simplecov-0.14.0/lib/simplecov/file_list.rb:56:in `map'
  from vendor/bundle/gems/simplecov-0.14.0/lib/simplecov/file_list.rb:56:in `covered_strength'
  from (erb):7:in `formatted_file_list'
  from ~/.rbenv/versions/2.3.0/lib/ruby/2.3.0/erb.rb:864:in `eval'
  from ~/.rbenv/versions/2.3.0/lib/ruby/2.3.0/erb.rb:864:in `result'
  from vendor/bundle/gems/simplecov-html-0.10.0/lib/simplecov-html.rb:66:in `formatted_file_list'
  from (erb):21:in `block in format'

Downgrading to 0.13.0 resolved the issue for me. I'm on Ruby 2.3.0, Rails 4.2.5.

[PragTob]

Argh jeez, I knew this was gonna happen some day! Thanks for the report ❤️

So this error happens here - if you could give me the values of float and places in that part (by inserting a puts on your end there) that'd be amazing and helpful in tracking down the error source.

If you had something that reproduces, even better but that'll help get me started :)

I won't get to it today (am at work and have evening plans) but will try to make and publish a fix over the weekend.

Sorry for the inconvenience!

[chief]

Same here (got Infinity error)

ruby-2.4.0/gems/simplecov-0.14.0/lib/simplecov/source_file.rb:193:in `round': Infinity (FloatDomainError)

[stevehanson]

@PragTob No worries! It took me all of a couple minutes to downgrade the gem and move on, so no big deal.

Here is the puts output for float, places when I run a test. Let me know if I can do anything else to help, but no urgency on my end.

0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
0.0, 1
NaN, 1

[PragTob]

@stevehanson I'm on fixing this right now. No idea how NaN gets there... I identified the commit that is likely responsible (b35f569) and am cooking up a fix right now. It catches the 0.0 case, not the NaN one but I'll see... opening a PR soon

[stevehanson]

@PragTob cool. Feel free to ping me when you get the PR up, and I can test it out if you'd like.

[PragTob]

@stevehanson that's the idea, hope and intention. Thanks :)

[PragTob]

Fixed in 0.14.1 - thanks everyone