This is probably a duplicate as I see other date/time parsing related issues, but filing it anyways just in case

Environment Information

• jruby 10.0.0.1 (3.4.2) 2025-05-07 79cf1e4 Java HotSpot(TM) 64-Bit Server VM 24.0.1+9-30 on 24.0.1+9-30 +indy +jit [x86_64-linux]
• No command line flags passed in or relevant env vars set
• Linux6.12.27-amd64 break script engine #1 SMP PREEMPT_DYNAMIC Debian 6.12.27-1 (2025-05-06) x86_64 GNU/Linux

Actual Behavior

The following specs:

        let(:value) { { year: "2020", month: 1, day: 2, hours: 3, minutes: 4, seconds: 5, zone: "-0700" } }

        it { is_expected.to eq(Time.new(2020, 1, 2, 3, 4, 5, in: "-0700")) }

and

      let(:value) { "2020-01-02 03:04:05.500 -1200" }

      it { is_expected.to eq(Time.new(2020, 1, 2, 3, 4, 5.5, in: "-12:00")) } 

result in:

     Failure/Error: it { is_expected.to eq(Time.new(2020, 1, 2, 3, 4, 5, in: "-0700")) }
     
       expected: 2020-01-02 03:04:00.000000000 -0700
            got: 2020-01-02 03:04:05.000000000 -0700
     # ./spec/builtin_types/datetime_spec.rb:33:in 'block in <main>'

and

     Failure/Error: it { is_expected.to eq(Time.new(2020, 1, 2, 3, 4, 5.5, in: "-12:00")) }
     
       expected: 2020-01-02 03:04:00.000000000 -1200
            got: 2020-01-02 03:04:05.500000000 -1200

Expected Behavior

In MRI, these specs pass.

To reproduce

I can actually reproduce this locally in IRB with the following:

irb(main):001> Time.new(2020, 1, 2, 3, 4, 5, in: "-0700")
=> 2020-01-02 03:04:00 -0700
irb(main):002> 

Looks like the seconds are coming out as 0 even though a 5 was passed in.

I could give a more complicated way of reproducing if needed involving cloning/checking out a branch and running rspec. Just let me know!