Added New York city's Taxi and Limousine Commission For Hire Vehicle trip support

[otegami]

Related Issue

• Add support for New York city Taxi and Limousine Commission trip record data #8

What I did

• added NYC TLC For Hire Vehicle trip support
• added NYC TLC High Volume FHV Trips support
• added simple unit test

Details

• FHV Trips Data Dictionary
• High Volume FHV Trips Data Dictionary

ref: https://www1.nyc.gov/site/tlc/about/tlc-trip-record-data.page

Notes for reviewers

Please give me some advice.

High Volume FHV Trips Data has over 300MB which means the number of data cases is 1,143,691.
It will take a lot of time to execute tests and it's so hard to create over 1 million instances in TLC::HighVolumeFHVTrip#each test.
So I decided to skip TLC::HighVolumeFHVTrip#each test because it has already asserted HighVolumeFHVTrip data in the #to_arrow test.

[kou]

So I decided to skip TLC::HighVolumeFHVTrip#each test because it has already asserted HighVolumeFHVTrip data in the #to_arrow test.

How about just checking the first record by each.first instead of skipping entirely.

If our #each uses return to_enum(__method__) {to_arrow.n_rows}, we can also check each.size.

[otegami]

Thank you for the comments.
I tried to check the only first record in the test. But the bottleneck is calling the raw_records method so it didn't improve.
I will fix this problem after implementing the Arrow::RecordBatch#each_raw_record.

• [Ruby] Add Arrow::RecordBatch#each_raw_record apache/arrow#33749

# /lib/high-volume-fhv-trip.rb

def each
  return to_enum(__method__) unless block_given?

  to_arrow.raw_records.each do |raw_record| # <- Here
    record = Record.new(*raw_record)
    yield(record)
  end
end