Anderson123 Dec 19, 2008   4:40 PM GMT

Sorry i didnt get you...can you please explain

0 pointsBadges:

philpl1jb Dec 19, 2008   5:10 PM GMT

That was an old RPG trick. This is a CL requirement -- check out the command CVTDAT If you have TATOOL look at ADDDAT or ADDDAT2

54,090 pointsBadges:

Anderson123 Dec 22, 2008   5:08 AM GMT

I have coverted to *cymd format.But sinde this is *char i will not be able to subtract from this.Can this be done only using julien...is there a way to subtract form the curren date

0 pointsBadges:

philpl1jb Dec 22, 2008   2:20 PM GMT

If you don't have TATOOLs with the ADDDAT command then yes convert to Julian, subtract 1 from the day part, if day part is 0 then subtract one from the year part and put the day part at 365 or 366 on leap year and convert to your desired date format. CL converts from *dec to *char and back in the CHGVAR command. So you will need to put the last three char of the julian into a *dec 3 variable using substring.

54,090 pointsBadges:

Anderson123 Dec 22, 2008   3:50 PM GMT

Thanks for the info...Is there a simpler process this one seems too elaborate...if not then i will proceed with ur suggestion.I dont have taatoll to use ADDDAT but if there are similar opcodes it could help

0 pointsBadges:

Anderson123 Dec 23, 2008   1:23 PM GMT

hI i used the below code to get a day before current date.But this will not work on jan 1st and i have not handled for a leap year. Can you please suggest.. PGM DCL VAR(&CURDT) TYPE(*CHAR) LEN(6) DCL VAR(&YYDDD) TYPE(*CHAR) LEN(5) DCL VAR(&JULIAN) TYPE(*DEC) LEN(5 0) RTVSYSVAL SYSVAL(QDATE) RTNVAR(&CURDT) CVTDAT DATE(&CURDT) TOVAR(&YYDDD) FROMFMT(*SYSVAL) + TOFMT(*JUL) TOSEP(*NONE) CHGVAR VAR(&JULIAN) VALUE(&YYDDD) CHGVAR VAR(&JULIAN) VALUE(&JULIAN-1) CHGVAR VAR(&YYDDD) VALUE(&JULIAN) CVTDAT DATE(&YYDDD) TOVAR(&DATE) FROMFMT(*JUL) + TOFMT(*SYSVAL) TOSEP(*NONE) SNDPGMMSG MSG(&CURDT) ENDPGM

0 pointsBadges:

philpl1jb Dec 23, 2008   3:00 PM GMT

PGM DCL VAR(&CURDT) TYPE(*CHAR) LEN(6) DCL VAR(&YYDDD) TYPE(*CHAR) LEN(5) DCL VAR(&JULIAN) TYPE(*DEC) LEN(5 0) RTVSYSVAL SYSVAL(QDATE) RTNVAR(&CURDT) CVTDAT DATE(&CURDT) TOVAR(&YYDDD) FROMFMT(*SYSVAL) + TOFMT(*JUL) TOSEP(*NONE) CHGVAR &DDDc substr(&Julian 3 5) CHGVAR &YYC substr(&Julian 1 2) CHGVar &DDDd &DDDc ChgVar &YYd &YYc CHGVar &DDDd (&DDDd-1) if (&DDDd = 0) then do chgvar &YYd (&YYd -1) chgvar &DDDd 365 <<-- right in here you would need to determine if the new &YYd was a factor or 4 EndDo CHGVAR VAR(&JULIAN) VALUE(&YYDDD) CHGVAR VAR(&JULIAN) VALUE(&JULIAN-1) CHGVAR VAR(&YYDDD) VALUE(&JULIAN) CVTDAT DATE(&YYDDD) TOVAR(&DATE) FROMFMT(*JUL) + TOFMT(*SYSVAL) TOSEP(*NONE) SNDPGMMSG MSG(&CURDT) ENDPGM

54,090 pointsBadges:

Anderson123 Dec 23, 2008   4:11 PM GMT

Hey Phil Thanks a lot...I was finally able to get it done with quite a bit of coding...one last question regarding this ...is there an opcode to check the factor of a number...can you explain how to build that condition...

0 pointsBadges:

philpl1jb Dec 23, 2008   4:31 PM GMT

No, of course not -- that would be too easy but I think this will work since &FCT has zero decimals dividing by 4 would loose the fractional part then multiply by 4 and compare with the original. If they are the same it's a leap year -- unless it isn't -- see below! dcl &FCT *dec 2 0 chgvar &FCT (&YYd/4) chgvar &FCT (&FCT * 4) if (&FCT = &YYd) The year 2000 is a leap year. Some programs know about this. They simply check whether a year is divisible by 4, and hence conclude that 2000 is a leap year. Or they are more rigorous and are aware of the exception: a year that is divisible by 100 as well is not a leap year, unless it is also divisible by 400. In other words, neither 1900 nor 2100 are leap years, but 2000 is. So this code is good till 2100 Phil

54,090 pointsBadges: