Switch from lubridate to hms

[polettif]

Sorry for the kinda erratic development but I had to revisit #26. Apparently tidyverse functions don't work with lubridate (see github issue). That's why I changed the implementation of set_hms_times to work with hms::hms. That way, joining and otherwise filtering tables seems to work. The time functions of lubridate would've been a bit overkill anyways, since we don't have to work with timezones and so on.

Maybe more important: There's now tests covering the two methods added in #26. I also fixed a bug with date ranges.

[codecov-io]

Codecov Report

Merging #28 into master will increase coverage by 5.49%.
The diff coverage is 100%.

@@            Coverage Diff            @@
##           master     #28      +/-   ##
=========================================
+ Coverage   50.61%   56.1%   +5.49%     
=========================================
  Files          11      11              
  Lines         735     745      +10     
=========================================
+ Hits          372     418      +46     
+ Misses        363     327      -36

Impacted Files	Coverage Δ
R/time.R	96.55% <100%> (+75.71%)	⬆️

---

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 d6ea344...cc3bb7f. Read the comment docs.

[polettif]

Fixed the failing tests.

@tbuckl I saw you made set_hms_times internal. What's your proposed workflow to use it? Because if it's not exported we should include in read_gtfs?

[tbuckl]

first of all, thanks so much for the tests. everyone can appreciate tests.

regarding making set_hms_times internal, the improvements you had made this month to read_gtfs seemed so useful to the functioning and clarity of the entire package that i wanted to make sure to get it into CRAN before i had my head wrapped around set_hms_times. it looks like a useful function i just want to be able to write a small example before including it. as you say, i think its ideal usage is in read_gtfs, perhaps as an optional argument (e.g. parse_time=TRUE)

[tbuckl]

also, good catch with lubridate/hms!

[polettif]

regarding making set_hms_times internal, the improvements you had made this month to read_gtfs seemed so useful to the functioning and clarity of the entire package that i wanted to make sure to get it into CRAN before i had my head wrapped around set_hms_times. it looks like a useful function i just want to be able to write a small example before including it. as you say, i think its ideal usage is in read_gtfs, perhaps as an optional argument (e.g. parse_time=TRUE)

No worries, there's no need to rush things