416975 - Change memory and storage calendar to return parent items fitting the demanded range

Status: RESOLVED WORKSFORME

(Calendar :: Provider: Local Storage, defect)

Description

[Daniel Boelzle [:dbo]]

Attachment: fix

per bug 366177 comment #27 and following:

Memory calendar has a check if ITEM_FILTER_CLASS_OCCURRENCES has been set, but ITEM_FILTER_CLASS_OCCURRENCES only specifies whether the result is to be passed in its expanded form (or not).

Comment 1

[Sebastian Schwieger]

And where do you decide if to return expanded or unexpanded?
Also, it could be useful to special-case requests that have rangeStart = null and rangeEnd = null. We would not need to call getOccurrences but just return the unexpanded item. (Same idea applies to storageCalendar)

Comment 2

[Michiel van Leeuwen (email: mvl+moz@)]

from calICalendar.idl:
 * If set, return calIItemBase occurrences for all the appropriate instances,
 * as determined by an item's recurrenceInfo. 

and:
 * Note that the range is intended to act as a mask on the
 * occurrences, not just the initial recurring items.  So if a
 * getItems() call without ITEM_FILTER_CLASS_occurrenceS is made, all
 * events and todos which have occurrences inside the range should
 * be returned, even if some of those events or todos themselves
 * live outside the range.

I'm not sure what you want to change. I _think_ that the above comment agrees with the proposed change: always check the occurrences, no matter what the flag says.

I wonder if all the code that calls getItems really expect this, or relies on the broken current implementation.

Comment 3

[Daniel Boelzle [:dbo]]

Comment on attachment 302791 [details] [diff] [review]
fix

damn, I forgot to save the file before getting the diff...

Thinking about sebo's comment, I think we may run into problems for open ranges. I agree open ranges deserve special handling. Thus I better stick to the old (buggy behaviour) in the storage opt patch to not mix up that up too much.

Comment 4

[Daniel Boelzle [:dbo]]

(In reply to comment #2)
> I wonder if all the code that calls getItems really expect this, or relies on
> the broken current implementation.
AFAIK all code queries the expanded form.

Comment 5

[Philipp Preißer-Kewisch [:Fallen][☁️📆]]

Daniel, when you are back I'd appreciate a status on this bug. Can we close it?

Comment 6

[Wayne Mery (:wsmwk)]

(In reply to Philipp Kewisch [:Fallen] from comment #5)
> Daniel, when you are back I'd appreciate a status on this bug. Can we close
> it?

Daniel hasn't been active in recent years afaict

Comment 7

[Philipp Preißer-Kewisch [:Fallen][☁️📆]]

I think we can close this one until specific issues arise.