Add documentation for multi_cartesian_product when iterator is empty

[yizhepku]

Document the correct behavior of nullary cartesian product.

The bug was submitted in 2019 and #603 fixes it just fine. Please just merge both this PR and PR #603 and call it a day. I understand that review time is scarce, but having automated tests pass should be good enough for this kind of change.

[Philippe-Cholet]

I detect a bit of impatience and I get it, I do and I'm gonna look into it.

The "check format" fails, probably just the trailing spaces. If you could remove them, and possibly squash into one commit, that would be nice.

EDIT: The old 603 is now rebased.

[yizhepku]

The "check format" fails, probably just the trailing spaces. If you could remove them, and possibly squash into one commit, that would be nice.

Done.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (33541c4) 94.22% compared to head (7ab8480) 94.18%.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #834      +/-   ##
==========================================
- Coverage   94.22%   94.18%   -0.04%     
==========================================
  Files          48       48              
  Lines        6670     6676       +6     
==========================================
+ Hits         6285     6288       +3     
- Misses        385      388       +3     

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

[Philippe-Cholet]

#835 has been merged a few minutes ago. It's now time for your PR.
I rebased to see it work. Then I changed "self" for "the adapted iterator" in your commit.
Thanks for this.