day_of_week bug in core:time/datetime package

[adimemir-prog]

Context

• Operating System & Odin Version:
• Please paste odin report output:
• Odin: dev-2024-06-nightly:f745a1c47
• OS: Windows 10 Professional (version: 22H2), build 19045.4412
• CPU: AMD Ryzen 5 5600 6-Core Processor
• RAM: 16305 MiB
• Backend: LLVM 17.0.1

  Opinion

  I think the issue is that ordinals are reported as if in the Julian calendar but day_of_week function works as if it is working in the Gregorian calendar. Either incrementing the output of the unsafe_date_to_ordinal procedure by 1 (in fact this procedure seems to work correctly), or incrementing the output of the day_of_week procedure by 1 should fix this (a less invasive solution).

  Expected Behavior

  Entering main.
  -----------------------------------------
  1th of january 1AD has ordinal 1
  its day is <i am not sure of this one>
  -----------------------------------------
  4th july 2024 has ordinal 739071
  its day is Thursday

  Current Behavior

  Entering main.
  -----------------------------------------
  1th of january 1AD has ordinal 1
  its day is Sunday
  -----------------------------------------
  4th july 2024 has ordinal 739071
  its day is Wednesday

Steps to Reproduce

build this file below then run the output exe file

package main
import    "core:fmt"
import dt "core:time/datetime"
main :: proc() 
{
    fmt.println("Entering main.")
    fmt.println("-----------------------------------------")

    jan_1st_1ad := dt.Date{day=1,month=1,year=1}
    fmt.println("1th of january 1AD has ordinal", dt.unsafe_date_to_ordinal(jan_1st_1ad))
    fmt.println("its day is",dt.day_of_week(dt.unsafe_date_to_ordinal(jan_1st_1ad)))
    fmt.println("-----------------------------------------")
    jul_4th_2024ad := dt.Date {day=4, month=7, year=2024}
    fmt.println("4th july 2024 has ordinal", dt.unsafe_date_to_ordinal(jul_4th_2024ad))
    fmt.println("its day is", dt.day_of_week(dt.unsafe_date_to_ordinal(jul_4th_2024ad)))
}

Please provide detailed steps for reproducing the issue.

No additional compiler flags are used.

[Kelimion]

Good catch. Thanks.

[Kelimion]

or incrementing the output of the day_of_week procedure by 1 should fix this (a less invasive solution).

It's also the correct solution.