Re: BYDAY and YEARLY

[Bruce_Kahn@xxxxxxxxxxxxxxxx]

Dan wrote on 07/07/2003 03:16:10 PM:
> The issue is the use of BYDAY as follows to mean the last Sunday in
October.
> RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
> 
> The BYDAY on yearly refers to weeks of the year and not weeks of the
month
> so this really means the last Sunday of the year but only if it occurs
in
> October.  The last week of the year will always be in December
of course so
> the above rule never resolves to any valid dates.

I think a couple points from 2445 may
help here (all from Section 4.3.10 Recurrence Rule).  The first tidbit:

   Each BYDAY value can also be preceded
by a positive (+n) or negative
   (-n) integer. If present, this indicates the nth occurrence of
the
   specific day within the MONTHLY or YEARLY RRULE. For example, within
   a MONTHLY rule, +1MO (or simply 1MO) represents the first Monday
   within the month, whereas -1MO represents the last Monday of the
   month. If an integer modifier is not present, it means all days
of
   this type within the specified frequency. For example, within a
   MONTHLY rule, MO represents all Mondays within the month.

The second tidbit is:

   BYxxx rule parts modify the recurrence
in some manner. BYxxx rule
   parts for a period of time which is the same or greater than the
   frequency generally reduce or limit the number of occurrences of
the
   recurrence generated. For example, "FREQ=DAILY;BYMONTH=1"
reduces the
   number of recurrence instances from all days (if BYMONTH tag is
not
   present) to all days in January. BYxxx rule parts for a period
of
   time less than the frequency generally increase or expand the number
   of occurrences of the recurrence. For example,
   "FREQ=YEARLY;BYMONTH=1,2" increases the number of days
within the
   yearly recurrence set from 1 (if BYMONTH tag is not present) to
2.

   If multiple BYxxx rule parts are specified, then after evaluating
the
   specified FREQ and INTERVAL rule parts, the BYxxx rule parts are
   applied to the current set of evaluated occurrences in the following
   order: BYMONTH, BYWEEKNO, BYYEARDAY, BYMONTHDAY, BYDAY, BYHOUR,
   BYMINUTE, BYSECOND and BYSETPOS; then COUNT and UNTIL are evaluated.

This second tidbit describes the case
in Dans example rule; the BYDAY and BYMONTH modify the recurrence by expanding
the set.  However since the BYMONTH and BYDAY parts only have 1 value
each the set is not actually expanded but more modified.

By applying the first tidbit alone to
just RRULE:FREQ=YEARLY;BYDAY=-1SU alone I would get the last Sunday in
the year (the date is calculatable, the time is inherited from DTSTART).
 However as Ki noted (from the second tidbit) there are other modifiers
in play here and that is BYMONTH (or BYDAY if you follow the expressed
order in the last paragraph cited).

Thus evaluating the rule:

RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10

You start with DTSTART and apply the
FREQ and INTERVAL to get "Repeats yearly on the DTSTART date/time".

Then you apply the BYxxx rule parts
in the order proscribed.  The first would be BYMONTH resulting in
"Repeats yearly on the DTSTART date/time in the month of October"
(the day and time are inherited, the month is overridden by BYMONTH).  The
next step would be to apply the next BYxxx rule part we have, BYDAY.  This
would result in "Repeats yearly on the Last Sunday (at the DTSTART
time) in the month of October".

So the final set of instances would
be DTSTART and on the last Sunday in October that immediately follows the
DTSTART (it could be the next year or the same year depending on when DTSTART
was).  At least thats how I take the text from the RFCs on recurrence
rule evaluation.

In rechecking Dans text I see that he
said "The BYDAY on yearly refers to weeks
of the year and not weeks of the month"
but BYDAY is defined as:

   The BYDAY rule part specifies a COMMA
character (US-ASCII decimal 44)
   separated list of days of the week; MO indicates Monday; TU indicates
   Tuesday; WE indicates Wednesday; TH indicates Thursday; FR indicates
   Friday; SA indicates Saturday; SU indicates Sunday.

which is not quite the same.  BYWEEK
would be "weeks of the year".  BYDAY is defined as "days
of the week".

> If the opinion is that "RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10"
does mean
> the last Sunday in October, then that raises further questions.  For
> example, what dates would the following rule resolve to?
> RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10,11,12

With a BYMONTH=10,11,12 we would follow
the same evaluation logic above and get 3 instances (the last Sundays of
October, November and December) and that exactly jives w/the bit about
"expand the number of occurences of the recurrence" from 2445.

The same expansion logic can be confirmed
by using more than 1 value for the BYDAY as well.  

Bruce
===========================================================================
Bruce Kahn                  
             INet: Bruce_Kahn@xxxxxxxxxxxxxxxx
Messaging & Collaboration            
    Phone: 978.399.6496
IBM Software Group                
        FAX: and nothing but the FAX...
Standard disclaimers apply, even where prohibited by law...