Codecov Report

Merging #1099 into master will increase coverage by 0.01%.
The diff coverage is 100%.

@@             Coverage Diff              @@
##             master    #1099      +/-   ##
============================================
+ Coverage     74.12%   74.13%   +0.01%     
- Complexity     3472     3476       +4     
============================================
  Files           385      385              
  Lines         10678    10683       +5     
  Branches       1309     1309              
============================================
+ Hits           7915     7920       +5     
  Misses         2302     2302              
  Partials        461      461

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 628cfb5...4e9b2bb. Read the comment docs.

I wouldn't enforce that, a user might like it that iterations are named the same. For example the navigation in IJ then works properly.

@marcphilipp that is another valid option, but I'd like this to be a conscious decision of the user, thus hidden behind a (possibly yet another) flag

For example the navigation in IJ then works properly.

Does it when using a pattern?

When using a pattern other than #featureName, the iterations are not named the same most probably. But if that is the pattern or there is actually no pattern used, the iterations are named all like the feature and navigation works. And with my fancy naming PR, you could rely on the the default naming to have the data variables in the name and if you want to have better IDE navigation temporarily, you can set a system variable to the pattern of #featureName and all iterations are named like the feature.

What are we trying to solve here?

1. You don't notice duplicated names, but you probably don't have a problem. Not ideal, but you probably won't go searching for the flag for a problem you don't even know about.
2. If you already noticed non-unique names, then you could just fix them.

So if we hide it behind a flag, it would probably only be used by you. However, if we don't do that, then we add the performance penalty to everyone. And I bet that at best only very few people are affected by this issue.

This could also be done with as a third-party extension, so instead of setting the flag you'd add the extension. I'm personally leaning to this solution.

At the end of the day, it is probably an issue with the reporting tool, at least for Spock 2.x, since it will report a UniqueId in addition to the displayed name.