Editorial: Avoid dividing by zero in ZonedDateTime#round.

[Ms2ger]

Ref. #1502.

[codecov]

Codecov Report

Merging #1582 (2de1f5d) into main (0e24fd3) will decrease coverage by 23.26%.
The diff coverage is 100.00%.

❗ Current head 2de1f5d differs from pull request most recent head be68077. Consider uploading reports for the commit be68077 to get more accurate results

@@             Coverage Diff             @@
##             main    #1582       +/-   ##
===========================================
- Coverage   95.58%   72.31%   -23.27%     
===========================================
  Files          19       18        -1     
  Lines       10933     4797     -6136     
  Branches     1715     1049      -666     
===========================================
- Hits        10450     3469     -6981     
- Misses        472     1034      +562     
- Partials       11      294      +283     

Flag	Coverage Δ
test262	72.31% <100.00%> (+0.01%)	⬆️
tests	?

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
polyfill/lib/zoneddatetime.mjs	89.41% <100.00%> (-9.24%)	⬇️
polyfill/lib/calendar.mjs	22.95% <0.00%> (-71.39%)	⬇️
polyfill/lib/intl.mjs	61.95% <0.00%> (-38.05%)	⬇️
polyfill/lib/instant.mjs	79.13% <0.00%> (-15.46%)	⬇️
polyfill/lib/plainmonthday.mjs	79.01% <0.00%> (-14.14%)	⬇️
polyfill/lib/plainyearmonth.mjs	83.42% <0.00%> (-12.99%)	⬇️
polyfill/lib/ecmascript.mjs	82.65% <0.00%> (-12.54%)	⬇️
polyfill/lib/plaindatetime.mjs	85.08% <0.00%> (-12.40%)	⬇️
polyfill/lib/timezone.mjs	81.66% <0.00%> (-11.80%)	⬇️
... and 8 more

---

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 0e24fd3...be68077. Read the comment docs.

[ptomato]

👍

[ljharb]

This kind of seems normative, not editorial? (altho presumably one that implementors are unlikely to get wrong)

[ptomato]

I considered that, but I thought it was an editorial error because the division operation on two mathematical values in RoundTime step 3.b would not be defined in ℝ if the divisor is 0.

However, I'm glad you prompted me to take another look, because it seems there is still an editorial error. The result of AddInstant and therefore AddZonedDateTime is a BigInt, meaning the division in RoundTime 3.b is on operands of mixed type, which is an editorial error according to Mathematical operations. @Ms2ger care to fix it up?

[Ms2ger]

Done in #1585.