jasonf@cetemp.Eng.Sun.COM (Jason Freund) (08/08/90)
I need to be able to change dates in a database cell with commands like "+4w" == "Find the closest weekday to the day 4 weeks from a certain date", "-12d" == "Find the closest weekday 12 days earlier than this date.", and so on. I'll be writing in Pascal -- the code is no problem -- I'll give my function a "+/-", a number, and a "d/w/m" (day, week, month), and a date "MMDDYY". The function returns the closes weekday by adding/subtracting according to its parameters. Does anyone know of a general algorithm for dealing with calendars? It must that take into account leap years and be able to find what day of the week it is (MTWRF) so that it can make approximations to weekdays. I've seen some good ones before, (for computing how many days old you are and what day of the week you were born on) but I can't find any now. BTW, when is the next leap year? Thanks, Jason Freund, Sun Microsystems, jasonf@cetemp.Corp.sun.com <== summer address Deprtmnt of Computer Science, Univ California, Davis. freund@sakura.ucdavis.edu Quantum Link: JasonF5, Compu$erve: 72007,244, 690 Erie Dr, Sunnyvale, CA 94087 ------------------------------------------------------------------------------- STOLEN QUOTES -- Please give the authors credit if you know who they are! "To understand recursion, you need to understand recursion." "Wow! Virtual memory! Now I'm gonna build me a REALLY big ram disk!" "My other computer is a SUN3/50." "E. Pluribus UNIX" -- authors unkown
awd@dbase.A-T.COM (Alastair Dallas) (08/10/90)
In article <140352@sun.Eng.Sun.COM>, jasonf@cetemp.Eng.Sun.COM (Jason Freund) writes: > I need to be able to change dates in a database cell with commands like > "+4w" == "Find the closest weekday to the day 4 weeks from a certain date", > "-12d" == "Find the closest weekday 12 days earlier than this date.", and so > on. I'll be writing in Pascal -- the code is no problem -- I'll give my > function a "+/-", a number, and a "d/w/m" (day, week, month), and a date > "MMDDYY". The function returns the closes weekday by adding/subtracting > according to its parameters. Does anyone know of a general algorithm for > dealing with calendars? It must that take into account leap years and be able > to find what day of the week it is (MTWRF) so that it can make approximations > to weekdays. I've seen some good ones before, (for computing how many days old > you are and what day of the week you were born on) but I can't find any now. The collected algorithms of the ACM has an algorithm for converting m-d-y into an integer number of days from some start point a long time ago, and back again. That makes it easy to add days or weeks. For months, you're pretty much stuck adding to your m-d-y months and dealing with overflow into years. The ACM book describes the weekday values for the resulting integer modulus 7. (That is, if no_of_days % 7 equals 0, it's Sunday, or Saturday, or something like that.) > BTW, when is the next leap year? That's easy. Leap years occur every four years, except years evenly divisible by 100. The last exeception is that years evenly divisible by 400 _are_ leap years. Therefore, '92 and '96 are leap years and for the first time in four hundred years, the turn of the century will be a leap year. All because the solar year isn't exactly 365-1/4 days. /alastair/
news@root.fiu.edu (08/11/90)
In article <140352@sun.Eng.Sun.COM> jasonf@cetemp.Eng.Sun.COM (Jason Freund) writes: > > I need to be able to change dates in a database cell with commands like >"+4w" == "Find the closest weekday to the day 4 weeks from a certain date", >"-12d" == "Find the closest weekday 12 days earlier than this date.", and so >on. I'll be writing in Pascal -- the code is no problem -- I'll give my >function a "+/-", a number, and a "d/w/m" (day, week, month), and a date >"MMDDYY". The function returns the closes weekday by adding/subtracting >according to its parameters. Does anyone know of a general algorithm for >dealing with calendars? It must that take into account leap years and be able >to find what day of the week it is (MTWRF) so that it can make approximations >to weekdays. I've seen some good ones before, (for computing how many days old >you are and what day of the week you were born on) but I can't find any now. > function canonical_day_number( date: YMD_date ) return integer is -- Expects a year, month, day combination and returns an integer -- which is the number of days since a very long time ago. To -- find the number of days between two dates, subtract these numbers. -- To find the day of the week, take the result mod 7. year2 : integer := date.year; month2 : integer := date.month; begin if month2 < 3 then month2 := month2 + 12; year2 := year2 - 1; end if; -- The following code avoids using floating point or really big numbers. return year2 * 365 + year2 / 4 - year2 / 100 + year2 / 400 +( month2 * 306 + 17 )/ 10 + date.day; end canonical_day_number;
Wright_RJ@cc.curtin.edu.au (Robert Wright) (08/13/90)
Excuse the ancient, shouted, UPPERCASE Fortran, but this still works to the best of my knowledge. This is really a very flexible algorithm. (cut after this signature block...) /-------------------------\ /-----------------------------------------------\ | Rob Wright |psi%050529452300070:Wright_RJ | | Curtin University |Wright_RJ@cc.curtin.edu.au | | Perth, Western Australia |Wright_RJ@cc.cut.oz.au | | Voice:+61 9 351 7385 |Wright_RJ%cc.curtin.edu.au@cunyvm.bitnet | | FAX: 09-351-2673 |uunet!munnari.oz!cc.curtin.edu.au!Wright_RJ | \-------------------------/ \-----------------------------------------------/ SUBROUTINE JULIAN(D1,M1,Y1,D2,M2,Y2,F,J2,W2,T2) IMPLICIT INTEGER*4 (A-Z) REAL FD1,FM1,FY1,FD2,FM2,FY2,FF,FJ2,FW2,FT2 C C ALGORITHM DERIVED FROM HART: SOFTWARE - PRACTICE AND EXPERIENCE C VOL 10, 405-417 (1980) C WHERE A MOST COMPLETE SYNTHESIS IS GIVEN. C C CONVERTED FROM HART'S BASIC TO FORTRAN BY R.J.WRIGHT C (WAIT COMPUTING CENTRE, SEPTEMBER 1980) C C LANGUAGE-CONVERSION PROBLEM: C BASIC STORES ALL NUMBERS IN FLOATING POINT AND C TRULY TRUNCATES WHEN CONVERTING TO INTEGER. C FORTRAN IN MANY IMPLEMENTATIONS ROUNDS WHEN C DOING THIS CONVERSION. THUS THIS ROUTINE FIRSTLY C CONVERTS ALL ARGUMENTS TO FLOATING POINT, WORKS C IN FLOATING POINT, AND DELIBERATELY TRUNCATES C USING THE AINT FUNCTION. THIS IS NOT QUITE AS EFFICIENT C AS MAY BE DESIRED, BUT IT WORKS! C C THE OTHER SIGNIFICANT CHANGE MADE WAS TO RE-ORDER THE M-D-Y C DATE CONVENTION TO THE MORE INTERNATIONALLY ACCEPTIBLE D-M-Y. C C ***** THE FOLLOWING COMMENTS ARE REPRODUCED ALMOST VERBATIM FROM HART ***** C C Julian conversion and inverse -- takes into account C leap years and the omission of February 29 in years evenly C divisible by 100 but not by 400 (the Gregorian convention). C C J2, the Julian date, is the exact number of days since C BC 4714 December 31. Noon on 1981 January 1 marks the C beginning of JD 2444606. C C D-M-Y, the Gregorian calendar, was adapted in the United C States on 1752 September 14 (JD 2361222). Prior to this C date the algorithm has no real meaning although it will C calculate imaginary dates back to AD 0 March 1. C C In the United States daylight saving time begins at 2:00 AM C the last Sunday of April and extends to 2:00 AM the last C Sunday of October. (These are respectively the first Sunday C beginning day 55, and the first Sunday beginning day 239). C This algorithm provides T2=0 during standard time and T2=1 C during daylight saving time. C Today's T2 minus yesterday's T2 is the daily adjustment to C the clock at 2:00 AM. C C Input arguments: C D1 day usually 0..31, can also be any integer C M1 month 0,1..12,13,14 C Y1 year 1753..future, must be 4 digits C (xx means 19xx) C C Output arguments: C D2 day 1..31 C M2 month 1..12 C Y2 year 1753..future C F flag 1=Jan,Feb; 0=Mar..Dec C J2 JD 2361222..future; 2444606=1-1-1981 C NOTE: J2 IS RETURNED AS 0 FOR INVALID DATE. C W2 weekday 0..6; Sunday..Saturday; (J2+1) MOD 7 C T2 time 0=standard; 1=daylight saving time C 2:00 AM adjustment = today's T2 minus yesterday's T2 C C C SOME POSSIBLE USES: C LAST MONTH C first day = JULIAN(1,M-1,Y,.......) C last day = JULIAN(0,M,Y,...) C C THIS MONTH C first day = JULIAN(1,M,Y,...) C last day = JULIAN(0,M+1,Y,...) C C NEXT MONTH C first day = JULIAN(1,M+1,Y,...) C last day = JULIAN(0,M+2,Y,...) C C X OR -X DAYS FROM D-M-Y C JULIAN(D+X,M,Y,...) C C DATE FOR THE XTH DAY OF THE YEAR C JULIAN(X,1,Y,...) C C DAYS BETWEEN TWO DATES C JULIAN(D1,M1,Y1,...,J1,...) C JULIAN(D2,M2,Y2,...,J2,...) C then take J2-J1 C C DAY OF THE YEAR C JULIAN(M,D,Y,...,J1,...) C JULIAN(0,1,Y,...,J2,...) C then take J2-J1 C C C COPY AND FLOAT THE INPUT INTEGER ARGUMENTS FD1=D1 FM1=M1 FY1=Y1 C FJ2=1. ! 1=no error IF(FY1.LT.0. .OR. FM1.LT.-1. .OR. FM1.GT.14.) FJ2=0. ! 0=error IF(FJ2.NE.1.)GOTO 1000 C FY2=FY1 ! year IF(FY2.LE.99.)FY2=FY2+1900. !assume 20th century FF=AINT((14.-FM1)/12.) ! 1=Jan,Feb; 0=Mar...Dec FJ2=AINT(30.61*(FM1+1.+FF*12.))+FD1+AINT(365.25*(FY2-FF)) FJ2=FJ2-AINT((((FY2-FF)/100.)+1.)*.75)+1720997. ! Julian day FW2=AINT(FJ2+1.-AINT((FJ2+1.)/7.)*7.) ! weekday: 0...6; Sun...Sat FY2=FJ2-1721119.1+AINT(.75*AINT((FJ2-1684594.75)/36524.25)) FD2=AINT(FY2-AINT(365.25*AINT(FY2/365.25))+122.2) ! 123...488 FY2=AINT(FY2/365.25)+AINT(FD2/429.) !year FM2=AINT(FD2/30.61)-1.-AINT(FD2/429.)*12. !month FD2=FD2-AINT(30.61*AINT(FD2/30.61)) !day FT2=INT(30.61*(FM2+1.+FF*12.))+FD2-FW2+1000. FT2=AINT(FT2/1177.)-AINT(FT2/1361.) ! 0=standard;1=daylight IF(FJ2.LT.2361222.)FJ2=0 !error if date < 14-Sep-1752 C FIX THE WORKING VALUES BACK TO INTEGERS AND RETURN C 1000 D1=FD1 M1=FM1 Y1=FY1 D2=FD2 M2=FM2 Y2=FY2 F=FF J2=FJ2 W2=FW2 T2=FT2 RETURN END
devine@shodha.dec.com (Bob Devine) (08/14/90)
In article <3091.26c69b52@cc.curtin.edu.au>, Wright_RJ@cc.curtin.edu.au (Robert Wright) writes: > Excuse the ancient, shouted, UPPERCASE Fortran, but this still works to the > best of my knowledge. This is really a very flexible algorithm. > > C In the United States daylight saving time begins at 2:00 AM > C the last Sunday of April and extends to 2:00 AM the last > C Sunday of October. Be very careful with hardcoding DST rules; the rules change unexpectantly The current start of DST is the _first_ Sunday in April. Bob Devine / DEC DB Engineering
paul@unhtel.uucp (Paul S. Sawyer) (08/14/90)
In article <1547@shodha.dec.com> devine@shodha.dec.com (Bob Devine) writes: > >Be very careful with hardcoding DST rules; the rules change unexpectantly >The current start of DST is the _first_ Sunday in April. > >Bob Devine / DEC DB Engineering As those of us well know who are stuck in UNIX SysV.2.... -- Paul S. Sawyer uunet!unh!unhtel!paul paul@unhtel.UUCP UNH Telecommunications attmail!psawyer p_sawyer@UNHH.BITNET Durham, NH 03824-3523 VOX: +1 603 862 3262 FAX: +1 603 862 2030
balden@van-bc.wimsey.bc.ca (Bruce Balden) (08/15/90)
Keywords: > SUBROUTINE JULIAN(D1,M1,Y1,D2,M2,Y2,F,J2,W2,T2) >C ALGORITHM DERIVED FROM HART: SOFTWARE - PRACTICE AND EXPERIENCE >C Julian conversion and inverse -- takes into account >C J2, the Julian date, is the exact number of days since >C D-M-Y, the Gregorian calendar, was adapted in the United Most elegant calendar algorithms depend on the fact that the calendar is much more regular if years are viewed as beginning in March, i.e. the zero'th day of the year is March 1 and the last day of the year is Feb 28 or 29. This allows the simple formula floor(30.6*MONTH+0.5) to calculate the day displacement. This "simple" formula can be complicated a little to avoid floating point arithmetic. On this simple observation and the idea of using Julian day numbers hang all elegant calendar algorithms. The usefulness of this approach is no accident: The year orginally really did begin on March 1. Look at the names of the months of the year: November "ninth month", September "seventh month", etc. Bruce Balden Grand Thaumaturge. Have a nice one.
koerberm@nixsin.UUCP (Mathias Koerber) (08/15/90)
Howdy, In article <669@dbase.A-T.COM> awd@dbase.A-T.COM (Alastair Dallas) writes: >In article <140352@sun.Eng.Sun.COM>, jasonf@cetemp.Eng.Sun.COM (Jason Freund) writes: >> I need to be able to change dates in a database cell with commands like >> "+4w" == "Find the closest weekday to the day 4 weeks from a certain date", [... stuff deleted ] >The collected algorithms of the ACM has an algorithm for converting [...] >dealing with overflow into years. The ACM book describes the weekday What is the ACM (guess: American Computing Monthly?,A magazine or so?). And what about the book?. Could anyone please e-mail me a short description of what it contains, and maybe an ISBN, price etc? Thx, Mathias -- Mathias Koerber |Tel: +65 / 7473828 ext 1852|Fax: +65 / 7474331 Nixdorf Computer Singapore|EUnet: koerber.sin@nixpbe |nerv: koerber.sin 2 Kallang Sector |uunet: uunet!linus!nixbur!koerber.sin Singapore 1334 |[Standard-disclaimer:All views personal... ]
mitchell@chance.uucp (George Mitchell) (08/27/90)
In article <1115@nixsin.UUCP> koerberm@nixsin.UUCP (Mathias Koerber) writes:
`What is the ACM (guess: American Computing Monthly?,A magazine or so?).
`And what about the book?. Could anyone please e-mail me a short description
`of what it contains, and maybe an ISBN, price etc?
Would it be appropriate to post a periodic (monthly?) welcome
message for comp.software.eng that contains some pointers to
news.announce.newusers and provides some of the information we
expect most readers (and more posters) to possess?
--
George Mitchell, MITRE, MS Z676, 7525 Colshire Dr, McLean, VA 22102
email: gmitchel@mitre.org [alt: mitchell@community-chest.mitre.org]
vmail: 703/883-6029 FAX: 703/883-5519mcgregor@hemlock.Atherton.COM (Scott McGregor) (08/28/90)
> What is the ACM (guess: American Computing Monthly?,A magazine or so?). > And what about the book?. Could anyone please e-mail me a short description > of what it contains, and maybe an ISBN, price etc? ACM is the Association for Computer Machinery. It is one of the two major professional societies for people working in the computer field. (The IEEE (Institute of Electrical and Electronics Engineers) Computer Society is the other major professional society). Both societies publish many important books (such as Collected Algorithms mentioned above) and magazines (such as Communications of the ACM, IEEE Computer, IEEE Software...). They sponsor many conferences and special interest groups (SIGs). Both are based in the US, and sponsor most activities there, but have international arms. Much time and effort can be saved be reusing work already done by others and published by these societies. Good technical libraries well have many of their publications available, so they are accessible if you want them, probably even in Singapore. Note, I posted this rather than replying privately, because I have found a number of US educated engineers don't know about or refer to the Collected Algorithms of the ACM and other useful ACM and IEEE books and journals. I believe that this may be a defect in our current computer science engineering programs, where we often encourage reinvention rather than library research and and reuse. Later when these engineers move from the halls of academia into industry becomes the dreaded Not Invented Here syndrome that undermines reuse strategies). The number of people who are unfamiliar with Ivan Sutherland's Sketchpad, Doug Engelbart's NLS Augment system, and other contributions of the 60s is staggering, yet many people are trying to reinvent this same technology today. A historical perspective can often help one to develop "new" technology more quickly. We should at least be thankful that the many people are once again "rediscovering" virtual memory--now for personal computers. Scott McGregor
norm@oglvee.UUCP (Norman Joseph) (08/29/90)
In <1115@nixsin.UUCP> koerberm@nixsin.UUCP (Mathias Koerber) writes: >>The collected algorithms of the ACM has an algorithm for converting [...] >What is the ACM (guess: American Computing Monthly?,A magazine or so?). ACM is The Association for Computing Machinery, a professional society of the computing community founded in 1947. For membership information, write to their headquarters: ACM Membership and Marketing Services 11 West 42nd Street New York, NY 10036 Tel: 212 869-7440 Telex: 421686 >And what about the book?. The Collected Algorithms of the ACM are in four volumes, with updates quarterly, and costs US $150 for ACM members. -- Norm Joseph - (amanue!oglvee!norm) cgh!amanue!oglvee!norm@dsi.com, or Oglevee Computer Systems, Inc. ditka!oglvee!norm@daver.bungi.com ---<*>--- "Shucking Usenet oysters in pursuit of a pearl." -- Bill Kennedy
torkil@Pacesetter.COM (Torkil Hammer) (08/31/90)
# # The Collected Algorithms of the ACM are in four volumes, with updates # quarterly, and costs US $150 for ACM members. I really hope that the ACM algorithm is identical to the Julian Day Number used by astronomers. We don't need another day by day calendar. Mon, Jan 1, 1990 in the gregorian calendar we use is JD 2447893, though the definition really is that JD 0 is Mon, Jan 1, 4713 BC in the (proleptic) julian calendar, that was used before the gregorian. Specifically JD 0.0 is defined to be noon GMT on that day. Can anyone with knowledge of the ACM algorithm confirm? Torkil Hammer