The IMAP service no longer blocks the event loop from progressing. When
syncing the event loop continues as usual, allow other parts of the
bridge to continue updating.
Once the sync is complete, it resets the event id back to the last
handled id before sync started and then we play catch up.
* Fix wrong context use for message downloads
* Fix delete of sync data failing due ErrNotFound
* Pre-allocate attachment data buffer before download
* Fix calculation of progress if message count is higher than total
Update to latest Gluon to allow access to the database for bridge. The
cache is placed in a `SharedCache` type to ensure that we respect calls
to `Connector.Close`.
In theory, this should never trigger an error, but this way we can catch
it if it happens.
https://github.com/ProtonMail/gluon/pull/391
Add two separate toggles to control event loop pausing. This is required
to prevent cases where the bridge requests the event loop to be paused
but a sync process completes and resumes the event loop.
For the loop to resume now both states need to be set to false. This
will be removed once GODT-2848 is implemented.
Enable the User Event, User Identity and User SMTP service. The services
don't persist any data at the moment to avoid conflict with the existing
event loop.
Rather than accessing the Bridge user list, each user register their
individual SMTP service with the server manager.
Note that some dependencies on the user data are hidden behind the
`UserInterface`. These will be removed in a future patch.
Refactor code to isolate the SMTP functionality in a dedicated SMTP
service for each user as discussed in the Bridge Service Architecture
RFC.
Some shared types have been moved from `user` to `usertypes` so that
they can be shared with Service and User Code.
Finally due to lack of recursive imports, the user data SMTP needs
access to is hidden behind an interface until the User Identity service
is implemented.
It should not be possible to reach this state on purpose. But due to
scheduling and synchronization variances, it is possible in theory that
a UserDeathEvent can occur at the same time as the bridge is closing,
causing a call to `User.Close()` to be executed 2 times.
To avoid this in the future we just clear the map after all the channels
have been closed.
This is not required in this call since address mode changes are always
proceeded by the removal of the IMAP user and then the creation of a new
IMAP user which will trigger the sync again.
It is possible, on slower machines, that the new event poll task is not
yet registered and attempts to cancel have nothing to cancel.
In this case, we need the refresh event to cancel the task, at that
point it is guaranteed that the task exists.
When fetching too many attachment bodies at once, the read can fail with
io.ErrUnexpectedEOF. In that case, we returun an error so the fetch is retried.
Report to sentry if we see some uncaught network err, but don't force
the user logout.
If we catch an uncaught json parser error we report the error to sentry
and let the user be logged out later.
Finally this patch also prints the error type in UserBadEvent sentry
report to further help diagnose issues.
When listing all a user's email addresses (e.g. for apple mail autoconf)
we need to exclude disabled addresses. Similarly, we need to remove
them from gluon if using split mode.
