Iain Lane authored May 12, 2015 and Carlos Garnacho committed Jan 19, 2016

Somewhere between sqlite 3.8.7 and 3.8.10.2, sqlite seemed to fix
a bug on ORDER BY SOMESTRINGFUNCTION() clauses, where it would
invariably apply the BINARY collator, despite tracker specifying
its own one.

This is something that 2 libtracker-data tests were indirectly
testing, where queries are sorted by tracker:coalesce(), these now
happen to fail because the stored results inadvertently had the wrong
order.

https://bugzilla.gnome.org/show_bug.cgi?id=749262

Recent glib forbids duplicate test paths, as the test relies on that
in order to find the .rq/.out files, copy these files from graph-4*.

Iain Lane authored May 12, 2015 and Carlos Garnacho committed Jan 19, 2016

As of GLib 2.45.1, the inotify monitor backend is called this.

Duplicate test name

We shouldn't be passing a NULL string argument to g_regex_match.

https://bugzilla.gnome.org/show_bug.cgi?id=759280

Make it clear that tracker_miner_manager_index_file only tells the
miner to start indexing a file. It doesn't wait for the actual indexing
operation to complete.

https://bugzilla.gnome.org/show_bug.cgi?id=757366

As documented, tracker_miner_manager_index_file should return TRUE on
success, not FALSE.

https://bugzilla.gnome.org/show_bug.cgi?id=757366

Marius Gedminas authored Oct 23, 2015 and Carlos Garnacho committed Oct 23, 2015

libunistring uses UTF-8 strings without a trailing NUL byte.  We're
passing such strings to tracker_parser_unaccent_nfkd_string() from
function_sparql_unaccent() in the sqlite interface.  If the string has
no accented characters, writing a NUL byte at the end will step out of
bounds.  This causes memory corruption and crashes.

The other caller of tracker_parser_unaccent_nfkd_string() is
process_word_utf8(), and it looks like it wants a trailing NUL, so let's
add it there.

There are no more callers of the libunistring version of
tracker_parser_unaccent_nfkd_string().

(For extra confusion, the libicu version of
tracker_parser_unaccent_nfkd_string() deals with U+0000-terminated
UTF-16 strings.)

Should fix https://bugzilla.gnome.org/show_bug.cgi?id=746195

We might fail to get a TrackerSparqlConnection if the session was shut
down too early and the DBus connection closed. Since this is not a
programming error nor an example of a broken system, let's not abort
and use a WARNING instead of a CRITICAL.

https://bugzilla.gnome.org/show_bug.cgi?id=756869

https://bugzilla.gnome.org/show_bug.cgi?id=756877

https://bugzilla.gnome.org/show_bug.cgi?id=756877

These generated files shouldn't end up in POTFILES.skip, it's less
of a pain though than moving the generation of these files to
Makefiles so we can clean these up properly.

This property has maxCardinality=1, we are however possibly adding
multiple values there, either in both UUID/ISBN forms, or as multiple
UUIDs in faulty epubs.

ISBN should probably be its own rdf:Property, in the mean time, stick
to the first nie:identifier found, and ignore the rest.

https://bugzilla.gnome.org/show_bug.cgi?id=746401

Kunaal authored Mar 14, 2015 and Carlos Garnacho committed Apr 09, 2015

Glib now has a new class Glib.ListStore, which creates a conflict in Vala.

https://bugzilla.gnome.org/show_bug.cgi?id=746194

Kunaal authored Mar 23, 2015 and Carlos Garnacho committed Apr 09, 2015

Glib now has a new class Glib.ListStore, which creates a conflict in Vala.

https://bugzilla.gnome.org/show_bug.cgi?id=746194

SPARQL update strings can be very long if the entire text of a document
is being stored. Dumping these huge strings to the log eats up disk
space and makes them harder to follow.

However, often, the occurence of such an error indicates a broken
extractor. In those cases, knowing part of the SPARQL can help in
identifying the file that triggered the error.

Usually the "nie:plainTextContent" property is the last one in the
string, so we truncate the error messages at the first occurance of
this property to achieve the best of both worlds.

https://bugzilla.gnome.org/show_bug.cgi?id=735406

Firstly, allow specifying multiple --content options, and allow
specifying --content-non-recursive if you really want as well.

Also, don't generate an XDG user directories config. It seems a bit
weird to be doing this and setting each directory to the same location.

https://bugzilla.gnome.org/show_bug.cgi?id=742035

media_art_process is not set when running tracker-extract-info tests

https://bugzilla.gnome.org/show_bug.cgi?id=742035

Martin Kampas authored Oct 20, 2014 and Carlos Garnacho committed Dec 20, 2014

https://bugzilla.gnome.org/show_bug.cgi?id=678986

Martin Kampas authored Oct 07, 2014 and Carlos Garnacho committed Dec 20, 2014

Identified by MinerCrawlTest::test_07_move_from_monitored_to_monitored
(300-miner-basic-ops.py)

In item_move() it fails to get source_iri, does not check it's validity and
uses it in the DELETE expression of the SPARQL query constructed there.

Broken since d836f002 (libtracker-miner: Store iri transiently as GFile
qdata) - tracker_file_notifier_get_file_iri() is added 'force' argument
and (wrapped with lookup_file_urn) passed force=FALSE from item_move().
This call then fails for regular files because only directories are
cached once crawling has completed as stated in the comment in
finish_current_directory() in libtracker-miner/tracker-file-notifier.c.

https://bugzilla.gnome.org/show_bug.cgi?id=678986