2016-08-23 Carlos Garnacho <carlosg@gnome.org>
Release 1.9.1
tests: Include new libtracker-data tests in dist
libtracker-miner: Move to next priority queue segment after deleting tail
If the tail element of a segment is being removed, we didn't move on to the
next one, so accounting of subsequent segments might get broken if further
elements are removed that are the head or the tail of a segment.
Also, just do the "jump to next segment" bit generic, so the affected paths
just turn on a fetch_segment boolean flag.
https://bugzilla.gnome.org/show_bug.cgi?id=769035
2016-08-23 Debarshi Ray <debarshir@gnome.org>
libtracker-control: Add convenience API for IndexFileForProcess
This mirrors the miner D-Bus API that is already available.
https://bugzilla.gnome.org/show_bug.cgi?id=757369
2016-08-22 Piotr Drąg <piotrdrag@gmail.com>
Updated Polish translation
2016-08-20 Balázs Úr <urbalazs@gmail.com>
Updated Hungarian translation
2016-08-19 Flo H <gnu.l10n.de@gmail.com>
Updated German translation
2016-08-19 Andika Triwidada <atriwidada@gnome.org>
Updated Indonesian translation
2016-08-17 Marek Černocký <marek@manet.cz>
Updated Czech translation
2016-08-10 Carlos Garnacho <carlosg@gnome.org>
configure: Do not hardcode dep on libpng
There is a --enable/disable-png check below.
2016-08-05 Dušan Kazik <prescott66@gmail.com>
Updated Slovak translation
2016-08-02 Daniel Mustieles <daniel.mustieles@gmail.com>
Updated Spanish translation
2016-07-31 Aurimas Černius <aurisc4@gmail.com>
Updated Lithuanian translation
2016-07-25 Piotr Drąg <piotrdrag@gmail.com>
Add Language headers to po files
Future versions of gettext will fail if this header is missing.
2016-07-18 Daniel Mustieles <daniel.mustieles@gmail.com>
Updated Spanish translation
2016-07-17 Carlos Garnacho <carlosg@gnome.org>
libtracker-sparql-backend: Turn failure to initialize direct backend to a warning
This usually indicates a wrong environment, we elegantly fallback to using the
bus backend so this is for the most part not noticeable by the user. Turn this
into a warning so the problem is more visible.
tracker: Fix possible warnings with the tracker search -[msivt] subcommands
We used to poke a third column when the "no search term given" queries just
return 2. This triggers warnings in the bus backend, but doesn't with the
direct backend, which just returns NULL.
Some people are seeing these warnings, which sounds bad enough because it
means tracker is forcing the "bus" backend for some reason. But of all ways
to notice this situation, this should not be one.
tests: Add tests around DELETE{}INSERT{}WHERE{} syntax
several (un)successful replacements of data are tested.
tests: Add tests around update syntax
So far, test for INSERT DATA, DELETE DATA and DELETE WHERE, which
became broken recently.
libtracker-data: Fix INSERT/DELETE DATA and DELETE WHERE syntax
This was broken with the introduction of DELETE{}INSERT{}WHERE{}.
Unbreak it, and check harder for invalid syntax situations, which
we should warn about.
libtracker-sparql: Make tracker URI GType private
We don't expose it anywhere in API, may be hidden as an implementation
detail.
libtracker-extract: Remove TrackerSparqlBuilder guarantee functions
Those are not used anymore, and libtracker-extract is private, so...
tests: Update tracker guarantee tests to check the TrackerResource functions
Those are the ones used now.
2016-07-16 Carlos Garnacho <carlosg@gnome.org>
libtracker-sparql: Add TrackerResource/NamespaceManager to docs
And document the missing functions.
libtracker-sparql: Fix small typo
Quite harmless, but triggers gtk-doc warnings.
libtracker-extract: Ensure UTF-8 content in contact/tag helpers
Those are created from untrusted data in some places, better to ensure
UTF-8 content in those.
tracker-extract-tiff: Ensure UTF-8 content in several TIFF fields
We can't rely on libtiff to do this.
tracker-extract-gif: Ensure UTF-8 content in nie:comment
We can't rely on libgif to do this.
tracker-extract-jpeg: Ensure UTF-8 content in nie:comment
We can't rely on libjpeg to do this.
libtracker-extract: Add tracker_guarantee_resource_utf8_string()
This is a wrapper to tracker_resource_set_string(), which additionally
validates the string to be UTF-8, and inserts only as much text as
could be validated.
Can be thought of as an internal replacement to
tracker_sparql_builder_object_unvalidated(), the fact that we deal with
non utf-8 data sources is an implementation detail to extractors, so
it doesn't make as much sense to make this a public function in the
shiny new TrackerResource API.
tracker-extract-mp3: Fix albumArtist coalescing
It seems the fix in commit 6dc2c85856 got oddly merged with the
TrackerResource work, this fixes coalescing so the performer is
used again as a last resort to determining the nmm:albumArtist.
2016-07-15 Sam Thursfield <sam@afuera.me.uk>
Merge branch 'sam/resource'
2016-07-14 Sam Thursfield <ssssam@gmail.com>
Use TrackerResource instead of TrackerSparqlBuilder in all extractors
For a long time, all the Tracker extractors have manually constructed a
SPARQL update command using TrackerSparqlBuilder to represent their
output. This commit changes all of them to use the TrackerResource
class instead, which makes the code a lot more concise and readable.
This introduces some API breaks in the internal libtracker-extract
library. This has been a private library since Tracker 0.16 or earlier,
so it's fine.
If the extractors only output SPARQL then they are only useful to
people who are using a SPARQL store. Now we can output a serialization
format like Turtle as well. This will hopefully make the extract modules
useful outside of Tracker itself.
I've tried to preserve the behaviour of the extractors as much as
possible, but there are two things that are now handled differently:
* nao:Tag resources are given a fixed URI based on the tag label, such
as <urn:tag:My_Tag>. Previously they were inserted as blank nodes,
so tracker-store would give them unique IDs like <urn:uuid:1234...>
* All extractors created nco:Contact resources for content publishers,
but previously some would assign fixed URIs based on the name
<urn:contact:James%20Joyce>, while others would insert them as blank
nodes so they would be assigned unique IDs like <urn:uuid:1234...>.
Now, all extractors create nco:Contact resources with fixed URIs
based on the content creator's name.
https://bugzilla.gnome.org/show_bug.cgi?id=767472
2016-07-14 Sam Thursfield <sam@afuera.me.uk>
libtracker-extract: Add resource-helpers module
This is factoring common code out of the individual extract modules.
https://bugzilla.gnome.org/show_bug.cgi?id=767472
2016-07-14 Sam Thursfield <ssssam@gmail.com>
libtracker-sparql: Add TrackerResource class
This provides a "resource-oriented" API for inserting and updating the
database. Rather than having to generate SPARQL queries, you can use the
TrackerResource abstraction to prepare information about a set of
resources, then generate a SPARQL query automatically.
TrackerResource can also serialize to Turtle directly.
https://bugzilla.gnome.org/show_bug.cgi?id=767472
2016-07-14 Sam Thursfield <sam@afuera.me.uk>
libtracker-sparql: Add TrackerNamespaceManager
This will keep track of a set of namespaces and their prefixes. Then,
when we are serializing a resource, we can use it.
This is separate from the Namespace and Ontologies classes in
libtracker-data. They may be able to make use of it though.
https://bugzilla.gnome.org/show_bug.cgi?id=767472
libtracker-sparql: Add TRACKER_TYPE_URI
This is the same as G_TYPE_STRING, but we need to treat strings and URIs
differently when generating SPARQL from the TrackerResource class, so we
need a separate GType to keep track of which is which when given a
GValue.
https://bugzilla.gnome.org/show_bug.cgi?id=767472
2016-07-14 Sam Thursfield <ssssam@gmail.com>
Bump GLib dependency to 2.44
So I can use G_DECLARE_DERIVABLE_TYPE() instead of writing out
everything manually.
https://bugzilla.gnome.org/show_bug.cgi?id=767472
2016-07-14 Sam Thursfield <sam@afuera.me.uk>
tracker-extract: Fix invalid SPARQL used when blacklisting files
This issue meant that files that triggered crashes never got
blacklisted, as the invalid SPARQL was ignored by the tracker-store.
So tracker-extract would spin on invalid files forever in some cases.
Clarify SPARQL example in manpage
https://bugzilla.gnome.org/show_bug.cgi?id=767246
Merge branch 'sam/extract-command'
2016-07-13 Sam Thursfield <sam@afuera.me.uk>
functional-tests: Add test case for re-extracting known files
This testcase should pass. I did think there was a bug here and
opened <https://bugzilla.gnome.org/show_bug.cgi?id=768706>, but
that turned out to be invalid.
functional-tests: Fix bug in await_resource_inserted()
As a performance optimisation, after the first GraphUpdated signal for a
resource, subsequent changes to that resource would be ignored. This is
actually dumb, because it means we ignore all the data added by the
extractor in some cases.
2016-07-12 Sam Thursfield <sam@afuera.me.uk>
tracker-extract-gstreamer: Only show GstDateTime message with verbosity=debug
I can't see any use for this message myself, maybe it should be
removed altogether:
GstDateTime was not complete, parts of the date/time were missing
(e.g. hours, minutes, seconds)
2016-07-04 Sam Thursfield <sam@afuera.me.uk>
Add example configuration 'uncrustify' formatting tool
This is useful for aligning variables in parameter names.
tracker-extract-gstreamer: Fix hash value format on 32-bit systems
The gsize type is 32 bits on a 32-bit system, but the hash value we
are printing is a guint64 so it's 64 bits on all systems. Note that
G_GINT64_MODIFIER is for both signed and unsigned int64 values.
2016-07-03 Adrien Bustany <adrien-xx-gnome@bustany.org>
Updated French translation
2016-06-30 Sam Thursfield <sam@afuera.me.uk>
sandbox: Support Python 3 as well as Python 2
sandbox: Fix MIME hack introduced in 8a758a03afada
It seems that new systems don't have this custom 'mime' dir (which
is good news), it's just some weird relic that exists on my laptop.
The workaround should now not break things in any case.
2016-06-28 Igor Gnatenko <ignatenko@src.gnome.org>
tracker-store: use @libexecdir@ in systemd service
References: https://bugzilla.gnome.org/show_bug.cgi?id=768127
2016-06-25 Sam Thursfield <sam@afuera.me.uk>
sandbox: Fix MIME-type detection inside the sandbox
It turns out that MIME-type detection depends on various files in
$XDG_DATA_HOME/mime, at least on my Fedora 24 system.
We need to link to these files in our overridden XDG_DATA_HOME dir if we
want extraction to work properly, otherwise everything is detected as
text/plain.
2016-06-24 Sam Thursfield <sam@afuera.me.uk>
libtracker-data: Fix pointer sign warnings
2016-06-22 Tiago Santos <tiagofsantos81@sapo.pt>
Updated Portuguese translation