This sub-directory contains utility progams used for developping the Date::Calendar::Persian module, especially its test files.

build-04-conv.el

This script builds test data to include in t/04-conversion.rakutest. These test data are generated from Emacs' calendar/cal-persia.el.

How to use

First, ensure the path to calendar/cal-persia.el and to calendar/calendar.el is correct. Your Emacs version may be a version other than 25.1.

Second, run the script from the command line, redirecting stdout to a temporary file. Do not bother with stderr. This generates conversion data for each New Year's day for Persian years 2 to 2000, for each New Year's eve for Persian years 1 to 1999 and for each month begin and end in Persian year 1400.

Third, choose a significant sample of test dates. Checking more than 4000 dates is overkill, so delete some of them to have a much smaller sample, still checking enough special cases such as leap years.

You may wish to move the year 1400 dates to their chronological place instead of leaving them at the end of the test data. But this is not required.

Insert the test data in t/04-conversion.rakutest and add manually a line for the Persian epoch, 1 Farvardin 1 (22 March 622 in the proleptic Gregorian calendar, that is 19 March in the Julian calendar).

build-08-week.raku

This script helps building data to check the week-related attributes in Persian dates objects. It uses a partially implemented Date::Calendar::Persian module.

First Step

Comment out then list(... True) line and uncomment the list(... False) line.

Run the script.

Check that the displayed days of week correspond to the real days: either check with what build-04-conv.el has generated or use the calendar functions in Emacs.

• Shanbe → Saturday

• Yek shanbe → Sunday

• Do shanbe → Monday

• Se shanbe → Tuesday

• Chahar shanbe → Wednesday

• Panj shanbe → Thursday

• Jumee → Friday

Find an interval of years with a significant sample of years changes. It should contain at least the three cases where the week-year is 53-week long:

• Normal year beginning on Se shanbe (e.g. year 1447),

• Leap year beginning on Do shanbe (e.g. year 1441),

• Leap year beginning on Se shanbe (e.g. year 1453).

Please note that you have to read two successive lines, the first one for the beginning of the year, the second one for the year type: normal / leap.

Second Step

Comment out the list(... False) line and uncomment the list(... True) line. Update the span of years in this line to include the sample chosen in the first step.

Run the script, redirecting stdout to a text file.

Edit this file. On each line, update the nnnn-Wnn-n value. From right to left:

• Check that the day-of-week is the proper one.

• Change the week number from "00" to the proper value, "01", "02", "51, "52" or even "53".

• Change the week-year value by adding or subtracting 1 to the current value (which has been initialised with the normal year value).

For number litterals (month numbers, day numbers and day-of-year numbers), remove the leading zeros, e.g. replacing "001" by "1" preceded by two spaces and replacing "02" by "2" preceded by one space.

Insert the resulting text into t/08-week.rakutest.

gener-test-0.1.0.raku

This script use 0.0.x modules to generate test data for 0.1.y modules. Read the instructions to use within the script's POD.