Raul Gutierrez Segales [Mon, 29 Aug 2011 17:26:27 +0000 (18:26 +0100)]
e-d-s: use this.local_ids in linkable_property_to_links ()
Otherwise, this method can have a different behaviour depending
on the state of the current Persona depending on whether
this.local_ids was called before or not.
Raul Gutierrez Segales [Mon, 29 Aug 2011 17:25:00 +0000 (18:25 +0100)]
e-d-s: we should use the Persona's iid as local-ids
If we don't and use the contact_id as before, linking Personas
from different PersonaStores won't work.
Fixes: https://bugzilla.gnome.org/show_bug.cgi?id=657635
Raul Gutierrez Segales [Mon, 29 Aug 2011 14:47:37 +0000 (15:47 +0100)]
e-d-s: default trust level for PersonaStores is PARTIAL
Without this, we can't link personas from different
(e-d-s) PersonaStores.
Helps: https://bugzilla.gnome.org/show_bug.cgi?id=657635
Raul Gutierrez Segales [Wed, 31 Aug 2011 20:27:39 +0000 (21:27 +0100)]
e-d-s: make sure we remove GConf's configs between tests
When exercising GConf more than the usual, by creating multiple
address books or such, we might leave things in an inconsistent
state that might affect the tests that come afterwards.
Philip Withnall [Wed, 31 Aug 2011 18:30:10 +0000 (19:30 +0100)]
tracker: Fix the match-im-addresses test
Due to the Tracker tests now running with the tracker persona store as the
writeable store, Tracker personas in the tests are now linked by their
linkeable properties (as their persona store is now fully trusted due to
being the writeable store).
This caused the two personas in the match-im-addresses test to be linked
together implicitly by the aggregator, when the test case assumed they
weren't. This broke the test.
This commit fixes the test permanently by changing the way we consider
potential matches based on IM addresses. Previously, we only indicated a
potential match if the two individuals under consideration shared a common
IM address + protocol combination. This commit changes that so that two
individuals are now considered to potentially match if they share a common
IM address *only*. Protocols are no longer considered.
This allows us to change the way the personas are set up in the
match-im-addresses test so that they're no longer implicitly linked by the
aggregator.
Raul Gutierrez Segales [Wed, 31 Aug 2011 14:19:53 +0000 (15:19 +0100)]
e-d-s: fix indentation in tests backend
Raul Gutierrez Segales [Wed, 31 Aug 2011 11:19:00 +0000 (12:19 +0100)]
build: Bump the e-d-s dependency to 3.1.5
We totally missed this on previous releases and
it is now causing problems for packagers.
Philip Withnall [Tue, 9 Aug 2011 16:23:12 +0000 (18:23 +0200)]
Bug 653777 — Add a helper function to create a writable persona
This adds IndividualAggregator.ensure_individual_property_writeable(), which
returns a persona which has the specified property writeable in the specified
individual. If no such persona exists already, a new one is created and
linked to the individual. If that all fails, an error is returned.
This allows for clients to write to properties of personas with confidence
that the updates will be written out to the backend stores. (Achieved by
calling IndividualAggregator.ensure_individual_property_writeable() first
and only writing to the property if that call succeeds.)
Closes: bgo#653777
Philip Withnall [Thu, 11 Aug 2011 11:57:25 +0000 (13:57 +0200)]
key-file: Allow empty Kf.Personas to be added to the persona store
This is necessary for the tests for bgo#653777. It removes the error which
would previously be thrown when trying to add a persona to Kf.PersonaStore
with no properties.
Helps: bgo#653777
Philip Withnall [Thu, 11 Aug 2011 11:55:10 +0000 (13:55 +0200)]
core: Add PersonaStore:always-writeable-properties
This complements Persona:writeable-properties, listing the properties which
are guaranteed to always be writeable on the personas in a given persona
store. This is in contrast to Persona:writeable-properties, which may list
extra properties which are only writeable on that particular persona.
This could be the case with, for example, personas representing the user,
which might have extra writeable properties to allow the user to change their
alias or avatar.
Helps: bgo#653777
Travis Reitter [Mon, 29 Aug 2011 20:32:06 +0000 (13:32 -0700)]
Fix NEWS: there never was a 0.5.3 release.
Travis Reitter [Mon, 29 Aug 2011 17:58:24 +0000 (10:58 -0700)]
Use the key-file backend for writes in general tests.
See commit
ef36f347 for more details.
Raul Gutierrez Segales [Mon, 29 Aug 2011 11:17:15 +0000 (12:17 +0100)]
Post-release version bump
Philip Withnall [Mon, 29 Aug 2011 11:16:44 +0000 (12:16 +0100)]
build: Fix the NEWS file
More haste, less speed…
Philip Withnall [Sat, 27 Aug 2011 13:25:28 +0000 (14:25 +0100)]
Bug 645056 — TpLowlevel library should have only static public functions
Remove the TpLowlevel class from the TpLowlevel library and instead place
all the functions statically in the “FolksTpLowlevel” namespace.
Closes: bgo#645056
Raul Gutierrez Segales [Mon, 29 Aug 2011 09:18:50 +0000 (10:18 +0100)]
Release version 0.6.1
Raul Gutierrez Segales [Mon, 29 Aug 2011 09:49:56 +0000 (10:49 +0100)]
tracker: catch GLib.SpawnError when getting a connection
Raul Gutierrez Segales [Mon, 29 Aug 2011 09:02:05 +0000 (10:02 +0100)]
Add myself as maintainer so I can upload tarballs, etc
Philip Withnall [Sat, 27 Aug 2011 15:25:11 +0000 (16:25 +0100)]
tracker: Remove “alias” from the writeable properties list of Trf.Persona
Trf.Persona no longer implements AliasDetails.
Philip Withnall [Thu, 25 Aug 2011 18:17:27 +0000 (19:17 +0100)]
Bug 657332 — Add linking link-individuals
Add a `linking link-individuals` subcommand to folks-inspect. Note: this
appears to break readline command completion for the `linking` command in
certain situations (e.g. if you tab-complete the name of the command first).
I can't fix this for love or money. Pigs will fly before I manage to work
out why readline hates me so much.
Closes: bgo#657332
Travis Reitter [Fri, 26 Aug 2011 17:33:38 +0000 (10:33 -0700)]
Don't require that zero Individuals be removed after auto-linking.
Travis Reitter [Fri, 26 Aug 2011 17:31:36 +0000 (10:31 -0700)]
Make sure Tracker tests actually use the Tracker backend for writes.
Travis Reitter [Fri, 26 Aug 2011 17:16:17 +0000 (10:16 -0700)]
Ignore edsf-persona-store.c for translations.
Travis Reitter [Fri, 26 Aug 2011 16:41:05 +0000 (09:41 -0700)]
Properly handle 'url' TYPE for URLs in Tracker backend.
Travis Reitter [Fri, 26 Aug 2011 15:00:32 +0000 (08:00 -0700)]
Correctly spell ERROR_VALAFLAGS.
Travis Reitter [Fri, 26 Aug 2011 14:59:31 +0000 (07:59 -0700)]
Don't use ERROR_CFLAGS yet. Vala doesn't guarantee warning-free C.
Travis Reitter [Fri, 26 Aug 2011 14:47:44 +0000 (07:47 -0700)]
Use the key-file backend for writes in tests for non-writeable backends.
We could have used the EDS backend, but this helps keep the execution
of the tests as simple as possible.
This fixes a bug in the libsocialweb aggregation test where it was
using the EDS backend from the Aggregator default but its
FOLKS_BACKEND_STORE_KEY_FILE_PATH didn't include the EDS backend, so
it had nowhere to store its linking personas.
Raul Gutierrez Segales [Fri, 26 Aug 2011 12:19:59 +0000 (13:19 +0100)]
e-d-s: add urls to list of writeable properties
Raul Gutierrez Segales [Fri, 26 Aug 2011 10:55:18 +0000 (11:55 +0100)]
IndividualAggregator: initialize primary-store property to NULL
Raul Gutierrez Segales [Fri, 26 Aug 2011 10:32:05 +0000 (11:32 +0100)]
IndividualAggregator: emit notifications for primary-store property
Raul Gutierrez Segales [Fri, 26 Aug 2011 10:31:25 +0000 (11:31 +0100)]
IndividualAggregator: state that the primary-store can be null
Travis Reitter [Thu, 25 Aug 2011 22:11:59 +0000 (15:11 -0700)]
Hold a pointer to delegates; don't copy them.
Travis Reitter [Thu, 25 Aug 2011 22:06:19 +0000 (15:06 -0700)]
Use ERROR_VALAFLAGS as intended.
Travis Reitter [Thu, 25 Aug 2011 21:28:06 +0000 (14:28 -0700)]
Build tests with extra debug information.
Notably, this provides gdb with Vala source file line numbers for
backtraces (which should save a bit of digging through generated C
files).
Gdb still lists generated C line numbers for lines farther up the
stack when there's no obvious Vala line equivalent.
Travis Reitter [Wed, 24 Aug 2011 17:50:54 +0000 (10:50 -0700)]
Add test case for properties race condition.
Philip Withnall [Wed, 24 Aug 2011 20:23:30 +0000 (21:23 +0100)]
eds: Tell Vala to not notify of property changes in the setter functions
We only want to notify of property changes when we get notification back
from eds that the change was successful.
Philip Withnall [Mon, 22 Aug 2011 19:23:50 +0000 (20:23 +0100)]
build: Bump the Vala dependency to 0.13.3
This is due to commit
c0ecad63015078231934e63ec63f02b97defb730.
Raul Gutierrez Segales [Mon, 22 Aug 2011 12:51:28 +0000 (13:51 +0100)]
e-d-s: Vala dropped support for reordering GCancellable params
A couple of e-d-s methods (get_contact, modify_contact,
get_backend_property, etc) experienced an automatic API
change in it's Vala binding due to Vala dropping support
for reordering GCancellable params.
Fixes: https://bugzilla.gnome.org/show_bug.cgi?id=657067
Philip Withnall [Sun, 21 Aug 2011 11:10:22 +0000 (12:10 +0100)]
core: Fix some whitespace
Raul Gutierrez Segales [Wed, 17 Aug 2011 14:48:39 +0000 (15:48 +0100)]
e-d-s: ignore the type when comparing images
Raul Gutierrez Segales [Mon, 15 Aug 2011 12:30:42 +0000 (13:30 +0100)]
e-d-s: update the Avatar Details test
Now that we've reworked avatar handling and we've
introduced an Avatar cache, the avatars associated
to a newly created e-d-s contact might not be immediately
available after the corresponding Edsf.Persona is created.
Hence we need to listen to notifications for the avatar
property.
Raul Gutierrez Segales [Thu, 4 Aug 2011 15:09:12 +0000 (16:09 +0100)]
e-d-s: Only set avatar if store_avatar () was successful
Helps: https://bugzilla.gnome.org/show_bug.cgi?id=655374
Raul Gutierrez Segales [Thu, 4 Aug 2011 15:08:06 +0000 (16:08 +0100)]
e-d-s: When an avatar has been removed, set it to null
Though new_avatar was null, I prefer to even more
obvious about what is going on.
Helps: https://bugzilla.gnome.org/show_bug.cgi?id=655374
Raul Gutierrez Segales [Thu, 4 Aug 2011 14:55:09 +0000 (15:55 +0100)]
e-d-s: Don't infer we already have an avatar in our cache by its URI
From libfolks we should discourage using local URIs to
set an EContactPhoto so even if we get inlined EContactPhoto
just do a full content comparison.
Helps: https://bugzilla.gnome.org/show_bug.cgi?id=655374
Raul Gutierrez Segales [Thu, 4 Aug 2011 14:44:28 +0000 (15:44 +0100)]
Don't manipulate the Avatar Cache before e-d-s acks changes
Also, lets set EContactPhotos as INLINED because it doesn't
make sense to use URIs for local files. In the future, when
we support gravatars, we'll have to handle the case of
setting ContactPhotos as URIs for those.
Fixes: https://bugzilla.gnome.org/show_bug.cgi?id=655374
Travis Reitter [Thu, 28 Jul 2011 17:08:09 +0000 (10:08 -0700)]
Do Tracker tests last, since they're most likely to fail uncontrollably.
Helps: bgo#655374 - Un-break Tracker avatar tests
Travis Reitter [Tue, 26 Jul 2011 22:53:02 +0000 (15:53 -0700)]
Add utility function to compare LoadableIcon content.
Unlike the various "equal()" functions for LoadableIcon and related classes,
this function actually compares image content (which can be necessary when
testing vs. images in the avatar cache, which may have a different path than an
otherwise-identical image).
Helps: bgo#655374 - Un-break Tracker avatar tests
Raul Gutierrez Segales [Wed, 17 Aug 2011 00:28:50 +0000 (01:28 +0100)]
Update NEWS file
Alexander Larsson [Tue, 16 Aug 2011 14:53:14 +0000 (16:53 +0200)]
eds: Support writing postal addresses as vcards
https://bugzilla.gnome.org/show_bug.cgi?id=656659
Alexander Larsson [Tue, 16 Aug 2011 12:06:33 +0000 (14:06 +0200)]
Generalize Edsf.PersonaStore._set_contact_attributes
We need to handle non-string field details values when we implement
postal addresses.
https://bugzilla.gnome.org/show_bug.cgi?id=656659
Alexander Larsson [Tue, 16 Aug 2011 12:06:10 +0000 (14:06 +0200)]
Read postal addresses as vcards
https://bugzilla.gnome.org/show_bug.cgi?id=656659
Alexander Larsson [Tue, 16 Aug 2011 12:04:44 +0000 (14:04 +0200)]
Break out common code in edsf backend
Most types use the same code to update the vcard params, so we
can share this code.
https://bugzilla.gnome.org/show_bug.cgi?id=656659
Philip Withnall [Tue, 16 Aug 2011 20:41:47 +0000 (21:41 +0100)]
inspect: Ensure to initialise out variables
This fixes a few compiler warnings.
Philip Withnall [Tue, 16 Aug 2011 20:41:32 +0000 (21:41 +0100)]
core: Fix use of deprecated Vala API
Philip Withnall [Tue, 16 Aug 2011 18:40:35 +0000 (19:40 +0100)]
eds: Don't copy EVCardAttributes at all
libebook does some icky things with EVCardAttributes it's passed, especially
when calling e_vcard_remove_attribute() (it removes by pointer comparison
and then unconditionally frees the passed-in attribute). Consequently, we
need to do everything we can to avoid copying EVCardAttributes unless we
want double-frees and memory leaks.
Raul Gutierrez Segales [Tue, 16 Aug 2011 19:03:03 +0000 (20:03 +0100)]
Don't use const string[] , Vala doesn't handle them correctly
Alexander Larsson [Tue, 16 Aug 2011 12:03:17 +0000 (14:03 +0200)]
Make sure we pass ENABLE_EDS to vala flags
Philip Withnall [Sun, 14 Aug 2011 17:48:37 +0000 (18:48 +0100)]
inspect: Remove workarounds for bgo#638363
See also: bgo#656467.
Travis Reitter [Sat, 13 Aug 2011 16:58:00 +0000 (18:58 +0200)]
Post-release version bump.
Travis Reitter [Sat, 13 Aug 2011 11:12:22 +0000 (13:12 +0200)]
Release version 0.6.0
Travis Reitter [Sat, 13 Aug 2011 16:27:13 +0000 (18:27 +0200)]
Update the checked files for valac requirement.
Travis Reitter [Sat, 13 Aug 2011 15:33:37 +0000 (17:33 +0200)]
Distribute EDS test tools.
Travis Reitter [Sat, 13 Aug 2011 11:00:32 +0000 (13:00 +0200)]
Un-break Docs.
Travis Reitter [Sat, 13 Aug 2011 10:33:39 +0000 (12:33 +0200)]
Write support for URLs in Edsf.Persona
Closes: bgo#655609 - Implement setter for URLs
Philip Withnall [Sat, 13 Aug 2011 17:56:30 +0000 (18:56 +0100)]
inspect: Fix listing GStrv properties
We have to work around bgo#656467 for the moment.
Philip Withnall [Sat, 13 Aug 2011 17:05:50 +0000 (18:05 +0100)]
inspect: Update folks-inspect for the FieldDetails changes
Travis Reitter [Sat, 13 Aug 2011 10:18:29 +0000 (12:18 +0200)]
Make the EDS backend's system address book the default writeable
Closes: bgo#648818 - Support the EDS backend being a primary
Travis Reitter [Sat, 13 Aug 2011 10:20:02 +0000 (12:20 +0200)]
Trivial NEWS clean-up.
Travis Reitter [Thu, 11 Aug 2011 14:44:53 +0000 (16:44 +0200)]
Remove FieldDetails (obsoleted by AbstractFieldDetails and implementors)
Travis Reitter [Thu, 11 Aug 2011 14:37:22 +0000 (16:37 +0200)]
Support vCard-like parameters for URLs
Closes: bgo#655922 - Rebase UrlDetails.urls upon an
AbstractFieldDetails-derived class
Travis Reitter [Thu, 11 Aug 2011 11:23:01 +0000 (13:23 +0200)]
Support vCard-like parameters for Roles
Closes: bug#655921 - Change RoleDetails.roles to support vCard-like arbitrary parameters
Travis Reitter [Tue, 9 Aug 2011 15:40:41 +0000 (17:40 +0200)]
Support vCard-like parameters for web service addresses
Closes: bgo#653682 - Change WebServiceDetails.web_service_addresses to support
vCard-like arbitrary parameters
Travis Reitter [Tue, 9 Aug 2011 13:59:16 +0000 (15:59 +0200)]
Rebase PhoneDetails.phone_numbers upon PhoneFieldDetails
Closes: bgo#655920 - Rebase PhoneDetails.phone_numbers upon an
AbstractFieldDetails-derived class
Travis Reitter [Mon, 8 Aug 2011 15:27:04 +0000 (17:27 +0200)]
Allow for Individual replacement in the EDS add-persona test.
Travis Reitter [Mon, 8 Aug 2011 14:08:36 +0000 (16:08 +0200)]
Rebase NoteDetails.notes upon NoteFieldDetails
Closes: bgo#655919 - Rebase NoteDetails.notes upon an
AbstractFieldDetails-derived class
Travis Reitter [Thu, 4 Aug 2011 01:43:17 +0000 (18:43 -0700)]
Set the eds backend as the primary for the eds tests.
This fixes the eds/link-personas test.
Travis Reitter [Wed, 3 Aug 2011 20:49:32 +0000 (13:49 -0700)]
Rebase EmailDetails.email_addresses upon EmailFieldDetails
Closes: bgo#655917 - Rebase EmailDetails.email_addresses upon an
AbstractFieldDetails-derived class
Travis Reitter [Wed, 3 Aug 2011 20:47:24 +0000 (13:47 -0700)]
Run key-file tests later, since it's a low-priority backend.
Travis Reitter [Wed, 3 Aug 2011 20:29:44 +0000 (13:29 -0700)]
Run Tracker tests last, since it's a low-priority backend.
They're also most likely to fail for reasons outside of our control.
Travis Reitter [Tue, 2 Aug 2011 22:12:24 +0000 (15:12 -0700)]
Add utility function to compare MultiMaps of AbstractFieldDetails.
Travis Reitter [Fri, 22 Jul 2011 23:49:40 +0000 (16:49 -0700)]
Clarify parameters for PostalAddressOwnerDetails.
Travis Reitter [Tue, 12 Jul 2011 18:20:56 +0000 (11:20 -0700)]
Add basic equality testing for ImFieldDetails.
Helps: bug#653680 - Change ImDetails.im_addresses to support vCard-like
arbitrary parameters
Travis Reitter [Thu, 14 Jul 2011 18:08:43 +0000 (11:08 -0700)]
Support vCard-like parameters for IM addresses
Helps: bug#653680 - Change ImDetails.im_addresses to support vCard-like
arbitrary parameters
Travis Reitter [Fri, 8 Jul 2011 23:35:15 +0000 (16:35 -0700)]
Clean up the way the Tracker backend tracks IM addresses.
Helps: bug#653680 - Change ImDetails.im_addresses to support vCard-like
arbitrary parameters
Travis Reitter [Fri, 8 Jul 2011 21:14:21 +0000 (14:14 -0700)]
Don't spill over 80 characters.
Travis Reitter [Thu, 14 Jul 2011 18:38:31 +0000 (11:38 -0700)]
Support vCard-like parameters for addresses in PostalAddressDetails.
This also removes the older, less-functional, less-consistent
PostalAddress.types.
Helps: bgo#653679 - Change PostalAddressDetails.postal_addresses to
support vCard-like arbitrary parameters
Travis Reitter [Tue, 12 Jul 2011 18:48:43 +0000 (11:48 -0700)]
Add test for parameter equality in FieldDetails.
Helps: bgo#653679 - Change PostalAddressDetails.postal_addresses to
support vCard-like arbitrary parameters
Travis Reitter [Tue, 12 Jul 2011 18:17:12 +0000 (11:17 -0700)]
Add basic equality testing for FieldDetails.
Helps: bgo#653679 - Change PostalAddressDetails.postal_addresses to
support vCard-like arbitrary parameters
Travis Reitter [Fri, 15 Jul 2011 23:36:33 +0000 (16:36 -0700)]
Add parameters as an optional argument to the FieldDetails constructor.
Helps: bgo#653679 - Change PostalAddressDetails.postal_addresses to
support vCard-like arbitrary parameters
Travis Reitter [Wed, 13 Jul 2011 22:35:00 +0000 (15:35 -0700)]
Implement FieldDetails.equal()/hash() in terms of the default.
This allows explicit references to FieldDetails.equal()/hash() for
structures storing them (in case it changes from the
AbstractFieldDetails implementation in the future).
Helps: bgo#653679 - Change PostalAddressDetails.postal_addresses to
support vCard-like arbitrary parameters
Travis Reitter [Tue, 12 Jul 2011 00:20:54 +0000 (17:20 -0700)]
Add equal() and hash() functions to AbstractFieldDetails.
Helps: bgo#653679 - Change PostalAddressDetails.postal_addresses to
support vCard-like arbitrary parameters
Travis Reitter [Thu, 14 Jul 2011 23:57:01 +0000 (16:57 -0700)]
Centralize parameters to AbstractFieldDetails.
This cuts some lines of code from the derived classes.
Helps: bgo#653679 - Change PostalAddressDetails.postal_addresses to
support vCard-like arbitrary parameters
Travis Reitter [Thu, 7 Jul 2011 18:10:58 +0000 (11:10 -0700)]
Rebase FieldDetails upon an abstract and generic version of itself.
This should be useful for cases where the value has a very specific type
(eg, for PostalAddressDetails.postal_addresses). We'll create subclasses of
FieldDetails for those cases.
When we've created all the appropriate derived classes, we'll remove
FieldDetails entirely.
Helps: bgo#653679 - Change PostalAddressDetails.postal_addresses to
support vCard-like arbitrary parameters
Philip Withnall [Tue, 9 Aug 2011 10:49:31 +0000 (12:49 +0200)]
core: Expand the documentation of EmailDetails.email_addresses
Philip Withnall [Tue, 9 Aug 2011 10:23:27 +0000 (12:23 +0200)]
build: Add VAPI dependencies for eds backend
Philip Withnall [Tue, 9 Aug 2011 10:23:13 +0000 (12:23 +0200)]
core: Fix a typo in NoteDetails documentation
Philip Withnall [Tue, 9 Aug 2011 10:19:11 +0000 (12:19 +0200)]
Bug 653619 — Add company details
I dislike Tracker, ontologies, Sparql and drinking games.
Closes: bgo#653619
Philip Withnall [Sun, 7 Aug 2011 11:52:40 +0000 (13:52 +0200)]
eds: Fix a minor type problem in the add-contacts-stress-test
Philip Withnall [Sun, 7 Aug 2011 11:36:11 +0000 (13:36 +0200)]
build: Build eds backend by default
Signed-off-by: Raul Gutierrez Segales <rgs@collabora.co.uk>
Raul Gutierrez Segales [Fri, 5 Aug 2011 18:49:58 +0000 (19:49 +0100)]
e-d-s: fix link personas test
We've changed the way we set up the writeable store in
commit
5c062183b (bug #654907) and that broke this test.