[caldeveloper-l] SKIP in RRULEs

[Marten Gajda]

Hi Neil,

IMO duplicate instances generated from an RRULE should be handled just
like RRULE instances for which an RDATE exists. You merge them into one.
RFC 5545 speaks of a "recurrence set" and "a set is a collection of
distinct objects" (https://en.wikipedia.org/wiki/Set_(mathematics) ). So
by definition the recurrence set can not contain two equal instances.
This should answer 1 and 2.

Regarding 3, see the table on page 7
(https://tools.ietf.org/html/rfc7529#page-7)

   o  BYMONTH
   o  SKIP (for invalid month only)
   o  BYWEEKNO
   o  BYYEARDAY
   o  BYMONTHDAY
   o  SKIP (for invalid day)
   o  BYDAY
   o  BYHOUR
   o  BYMINUTE
   o  BYSECOND
   o  BYSETPOS
   o  COUNT
   o  UNTIL

It says that SKIP is applied before BYSETPOS. So In your example
2001-02-30 and 2001-02-31 would both wrap around to 2001-03-01 which
becomes the 2nd instance of this particular set (even though it's in the
next month), and also happens to be the first instance of the "March
Set". So in this case 3 instances are merged into one.

That's my interpretation and how I intended to implemented it. However,
I've just noticed that we return 2001-03-01 twice, which is not correct.
I'll fix that.

Cheers,

Marten

Am 19.10.18 um 06:23 schrieb Neil Jenkins:
> Hi everyone,
>
> I'm trying to extend the JSCalendar recurrence rule expansion
> algorithm
> <https://tools.ietf.org/html/draft-ietf-calext-jscalendar-08#section-4.3.1>
> to accommodate RFC7529
> <https://tools.ietf.org/html/rfc7529#section-4.1> and support RSCALE
> and SKIP. The former is easy, the latter seems to be a real pain. One
> big problem I'm running into is that the RFC seems to be quite
> under-specified when it comes to any kind of edge cases. So, can
> anyone help answer a few questions (presume |DTSTART=20010201| for all
> of these):
>
>  1. What happens if I generate duplicates due to SKIP; should they be
>     removed, or considered separate occurrences? e.g.
>     |RSCALE=GREGORIAN;FREQ=YEARLY;BYMONTH=2,3;BYMONTHDAY=1,31;SKIP=FORWARD|.
>     I presume that should not generate two instances of 2001-03-01
>     despite this being the result of
>     |BYMONTH=2;BYMONTHDAY=31;SKIP=FORWARD| as well as
>     |BYMONTH=3;BYMONTHDAY=1|.
>  2. What about if the duplicates occur in different periods, e.g.
>     |RSCALE=GREGORIAN;FREQ=MONTHLY;BYMONTHDAY=1,31;SKIP=FORWARD| –
>     should they still be removed?
>  3. How does this interact with |BYSETPOS|? Are the duplicates removed
>     before or after applying this? For
>     example, |RSCALE=GREGORIAN;FREQ=MONTHLY;BYMONTHDAY=1,30,31;BYSETPOS=1,2;SKIP=FORWARD|
>     – this might evaluate to 2001-02-01, 2001-02-30, 2001-02-31. We
>     apply bySetPos and dedup (in either order) and get 2001-02-01,
>     2001-03-01. OK, now we continue to March. is 2001-03-01 considered
>     for part of the |BYSETPOS| (so you get just 2001-03-30)? Or
>     because it's already been generated from the Feb candidates due to
>     the SKIP rule, is it eliminated before applying |BYSETPOS| (so you
>     get 2001-03-30 and 2001-03-31)?
>
>
> Looking forward to being enlightened on this,
>
> Neil.
>
> _______________________________________________
> caldeveloper-l mailing list
> caldeveloper-l at lists.calconnect.org
> http://lists.calconnect.org/listinfo.cgi/caldeveloper-l-calconnect.org

-- 
Marten Gajda
CEO

dmfs GmbH
Schandauer Straße 34
01309 Dresden
GERMANY

phone: +49 177 4427167
email: marten at dmfs.org

Managing Director: Marten Gajda
Registered address: Dresden
Registered No.: AG Dresden HRB 34881
VAT Reg. No.: DE303248743

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.calconnect.org/pipermail/caldeveloper-l-calconnect.org/attachments/20181019/78181d80/attachment.html>