As the stored context from initialisation may have changed
by `setContext` by other operations when acting as a OpenChange
library. This would make having a login set to nil and forcing
a NSException when it attempts to set a nil key at
[SOGoAppointmentFolder:roleForComponentsWithAccessClass:forUser]
inside [SOGoAppointmentFolder:initializeQuickTablesAclsInContext].
When we have a contact with an email, avoid generating another one
with SOGoMailDomain value (normally we ended up with a contact with
two identical emails on 'emails' key and for multidomain source we
would had ended up with an email @localhost)
Happened in an imported vevent from Mozilla Thunderbird.
The crash was:
Sep 14 15:49:38 sogod [21063]: <0x6442DBF8[SOGoAppointmentFolder]:personal> missing 'c_startdate' in record?
Sep 14 15:49:38 sogod [21063]: <0x6442DBF8[SOGoAppointmentFolder]:personal> missing 'c_enddate' in record?
2015-09-14 15:49:38.927 sogod[21063] NGCalendarDateRange.m:37 Assertion failed in NGCalendarDateRange(instance), method initWithStartDate:endDate:. startDate MUST NOT be nil!
EXCEPTION: <NSException: 0x7fb96b3c0ed8> NAME:NSInternalInconsistencyException REASON:NGCalendarDateRange.m:37 Assertion failed in NGCalendarDateRange(instance), method initWithStartDate:endDate:. startDate MUST NOT be nil! INFO:(null)
The relevant ICS file lines are the following ones:
BEGIN:VEVENT
UID:040000008200E00074C5B7101A82E00800000000901646A7234BCE01000000000000000010000000E9152C8FF1C27D488C91967FAAFCC2B0
RECURRENCE-ID:20140513T100000Z
DURATION:PT1H
CLASS:PUBLIC
ATTENDEE;PARTSTAT=NEEDS-ACTION;ROLE=REQ-PARTICIPANT;RSVP=TRUE;CN=krsny >>
Ann Thierry K:mailto:krsny@example.com
END:VEVENT
When returning contacts we have to supply also the domain field.
Because in a multidomain environment UIDField is unique only in
the domain so an user must be identified as uid@domain.
So when creating http requests from client side, we have to use
uid@domain instead of only uid so the SOGoUser created on server
side when parsing the requests is created properly.
The folder names are encoded through the `asCSSIdentifier` and
`stringByEncodingImap4FolderName` functions when we store them as folder
keys. In addition, the prefix "folder" is added to the key.
The order in which these operations were done when storing the folder
keys (and reverted when retrieving them) wasn't consistent trough the
code. This led to problems such as creating twice a folder with a digit
at the beginning of its name.
The folder name goes now through the following operations when being
stored as a key (the retrieval reverts these in the reverse order):
* `stringByEncodingImap4FolderName`
* `asCSSIdentifier`
* Add "folder" prefix
