dblock / distance_of_time_in_words / 34354810893

99%

push

github

Fix #160: don't fold unrelated UTC offset differences into the distance

offset_delta previously computed largest.utc_offset - smallest.utc_offset
for any two Time-like objects, assuming any offset difference represented
a real DST or tzdata transition of the same clock (as with #63 and #153).
That assumption breaks when the two times are simply expressed with
different, unrelated offsets (e.g. one in UTC, one with an explicit
"-08:00" offset) - there is no transition, so folding in the difference
double counts what Time#- already got right, producing wildly incorrect
results (e.g. reporting "9 hours" for an actual 1-hour difference).

Add a same_clock? guard: only fold the offset difference in when both
times are the same underlying clock, i.e. two ActiveSupport::TimeWithZone
values in the same zone, or two plain Time values in the process's system
zone (as opposed to one built with a fixed numeric offset, or in UTC mode).

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

5 of 5 new or added lines in 1 file covered. (100.0%)

278 of 278 relevant lines covered (100.0%)

624.13 hits per line