Implement DoubleEndedIterator for ProcessResults

[shepmaster]

No description provided.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 94.43%. Comparing base (6814180) to head (0227457).
Report is 42 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #910      +/-   ##
==========================================
+ Coverage   94.38%   94.43%   +0.04%     
==========================================
  Files          48       48              
  Lines        6665     6829     +164     
==========================================
+ Hits         6291     6449     +158     
- Misses        374      380       +6     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[Philippe-Cholet]

This looks good to me, and I appreciate the update on specialization tests.

However, this leads to a change in meaning. That way, the first error encountered is not necessarily the first in the iterator order. I know that's the all point but I'm unsure about such change.
We can it.rev().process_results(..) if we only need it in reverse so this change only matters when we need both iteration sides. Do you have an example?

EDIT: I see you are the original author of process_results, thanks!

[phimuemue]

Thanks for this.

As for the DoubleEndedIterator: I think it's fine to stop with another error.

As for the use cases: try_rfold might be specialized for the underlying iterator.