chiark / gitweb /
gooswapper
5 years agoonly register gcal_link if necessary
Matthew Vernon [Fri, 19 Oct 2018 11:41:54 +0000 (12:41 +0100)]
only register gcal_link if necessary

5 years agomore verbose logging of recurrence
Matthew Vernon [Fri, 19 Oct 2018 10:21:52 +0000 (11:21 +0100)]
more verbose logging of recurrence

5 years agoadd utility function to pop an event from the cache
Matthew Vernon [Fri, 19 Oct 2018 10:21:36 +0000 (11:21 +0100)]
add utility function to pop an event from the cache

5 years agoStart on recurring event support
Matthew Vernon [Wed, 17 Oct 2018 16:04:30 +0000 (17:04 +0100)]
Start on recurring event support

...currently only daily events, but this is tested and works across a
DST change.

5 years agoMake start and end times be rendered in target timezone
Matthew Vernon [Wed, 17 Oct 2018 16:03:24 +0000 (17:03 +0100)]
Make start and end times be rendered in target timezone

This is consistent with how we were handling all-day events
previously, and means that recurring events work round DST changes
better.

5 years agorestrict fields passed to event.save()
Matthew Vernon [Wed, 10 Oct 2018 15:45:08 +0000 (16:45 +0100)]
restrict fields passed to event.save()

This avoids problems where the event has some fields we can't edit.

5 years agocheck for empty body
Matthew Vernon [Wed, 10 Oct 2018 15:44:36 +0000 (16:44 +0100)]
check for empty body

5 years agoRemove some no-longer needed lines
Matthew Vernon [Mon, 8 Oct 2018 10:35:26 +0000 (11:35 +0100)]
Remove some no-longer needed lines

These were part of getting to the bottom of why saving wasn't working,
which we have now fixed (see previous commit).

5 years agoAdd code to propagate changes, and match events
Matthew Vernon [Sat, 6 Oct 2018 21:54:10 +0000 (22:54 +0100)]
Add code to propagate changes, and match events

Updating is a fairly simple process of copying across the fields we
think we care about.

Matching looks for non-recurring events with the same title in the
same time range.

5 years agopass gcal_tz where needed
Matthew Vernon [Fri, 5 Oct 2018 14:13:09 +0000 (15:13 +0100)]
pass gcal_tz where needed

5 years agoSplit out building gcal event structure (no functional change)
Matthew Vernon [Fri, 5 Oct 2018 14:06:14 +0000 (15:06 +0100)]
Split out building gcal event structure (no functional change)

...this will enable us to use it for updates as well as creation

5 years agocorrect warning text
Matthew Vernon [Fri, 5 Oct 2018 14:00:11 +0000 (15:00 +0100)]
correct warning text

5 years agoCache the auto-discovery values from Exchange
Matthew Vernon [Fri, 5 Oct 2018 13:51:51 +0000 (14:51 +0100)]
Cache the auto-discovery values from Exchange

This saves us a bit of time.

5 years agoAlso support single non-all-day events
Matthew Vernon [Fri, 5 Oct 2018 11:17:05 +0000 (12:17 +0100)]
Also support single non-all-day events

5 years agostore gcal_link in cache immediately (rather than waiting for next cache update
Matthew Vernon [Fri, 5 Oct 2018 11:16:38 +0000 (12:16 +0100)]
store gcal_link in cache immediately (rather than waiting for next cache update

5 years agofix syntax of extendedProperties setting
Matthew Vernon [Fri, 5 Oct 2018 11:15:32 +0000 (12:15 +0100)]
fix syntax of extendedProperties setting

5 years agoRemove old commented-out deletion code
Matthew Vernon [Fri, 5 Oct 2018 10:23:25 +0000 (11:23 +0100)]
Remove old commented-out deletion code

During testing, it made life easier to automatically delete new events
sometimes. We don't want to do that any more :)

5 years agodelete linked events
Matthew Vernon [Fri, 5 Oct 2018 10:22:37 +0000 (11:22 +0100)]
delete linked events

5 years agoSwitch to using a named tuple for the cache
Matthew Vernon [Fri, 5 Oct 2018 10:06:36 +0000 (11:06 +0100)]
Switch to using a named tuple for the cache

This lets us store more than one field per event, fairly simply.

5 years agoRemove old part-written broken function
Matthew Vernon [Fri, 5 Oct 2018 09:06:47 +0000 (10:06 +0100)]
Remove old part-written broken function

This was a previous iteration that never actually worked (it led to
the GlobalObjectId work, which in turn led to just using item_id)

5 years agoRemove GlobalObjectId code
Matthew Vernon [Fri, 5 Oct 2018 09:05:51 +0000 (10:05 +0100)]
Remove GlobalObjectId code

Firstly, we can use item_id instead; secondly, it doesn't always
actually work :-/

5 years agoMove to keying by item_id
Matthew Vernon [Fri, 5 Oct 2018 09:04:15 +0000 (10:04 +0100)]
Move to keying by item_id

item_id is constant (changekey changes if an event is updated) and
searchable, so use it as primary key.

5 years agoRemove commented-out account line
Matthew Vernon [Fri, 5 Oct 2018 08:33:42 +0000 (09:33 +0100)]
Remove commented-out account line

(previous work thought that ServiceAccount might be misbehaving)

5 years agotidy up the exchange extended property
Matthew Vernon [Fri, 5 Oct 2018 08:33:03 +0000 (09:33 +0100)]
tidy up the exchange extended property

The two commented-out versions were wrong ("gooswapper") and unhelpful
(the number, stolen from the exchangelib docs); better is to use a
public string entry.

5 years agoInitial commit
Matthew Vernon [Fri, 5 Oct 2018 08:31:07 +0000 (09:31 +0100)]
Initial commit

This now can cope with new all-day events from exchange, and make the
corresponding event in gcal.

5 years agogit ignore file
Matthew Vernon [Fri, 5 Oct 2018 08:27:38 +0000 (09:27 +0100)]
git ignore file