| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Found while building tracker 3.5.2 with clang 16, which enable
Wimplicit-function-declaration by default.
Without _GNU_SOURCE defined, tracker would fail in the configure phase
with Checking if "strftime 4-digit year modifier" runs: DID NOT
COMPILE error. And investigating the meson log we see
Compiler stderr:
/var/tmp/portage/app-misc/tracker-3.5.2/work/tracker-3.5.2-build/meson-private/tmplr2vm6b0/testfile.c:12:5: error: call to undeclared function 'gmtime_r'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
gmtime_r (×tamp, &tm);
And a little further up
Command line: clang /var/tmp/portage/app-misc/tracker-3.5.2/work/tracker-3.5.2-build/meson-private/tmplr2vm6b0/testfile.c -o /var/tmp/portage/app-misc/tracker-3.5.2/work/tracker-3.5.2-build/meson-private/tmplr2vm6b0/output.exe -O2 -pipe -march=native -DTRACKER_DEBUG -DG_DISABLE_CAST_CHECKS -D_FILE_OFFSET_BITS=64 -O0 -Werror=implicit-function-declaration -std=c99 -Wl,-O1 -Wl,--as-needed -fuse-ld=lld -rtlib=compiler-rt -unwindlib=libunwind -Wl,--as-needed
Seems like cc.run() doesn't pick up the general compiler args we set
earlier. We have already set it for the main build, so adding it just
for the configure test should be fine.
Signed-off-by: Brahmajit Das <brahmajit.xyz@gmail.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Different libc implementations (or different platforms) have different
bugs. Even though %Y is documented as a 4-digit number, and %C as
year/100 in a [00-99] range, they both crap out with years under 1000,
using respectively 3 and 1 digits.
This would be typically task for a width modifier (e.g. %4Y, or %2C),
but that was conversely found to break NixOS on Darwin.
Since the existing libc bugs paint us to a corner, detect an
appropriate modifier at build time to get a 4-digit year to get true
ISO 8601 in our supported [0001-9999] year range.
Closes: https://gitlab.gnome.org/GNOME/tracker/-/issues/402
|
| |
|
| |
|
| |
|
|\
| |
| |
| |
| | |
Migrate docs to gi-docgen (Take 2)
See merge request https://gitlab.gnome.org/GNOME/tracker/-/merge_requests/580
|
| |
| |
| |
| | |
We'll use some features from there.
|
|/
|
|
|
| |
The HAVE_LIBICU/LIBUNISTRING defines should not be needed for
anything now.
|
| |
|
|
|
|
|
| |
The autogenerated version has a more intuitive API on Vala and is more
up-to-date
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Since recently, GVDB repository includes a minimal meson.build
file to allow building as a subproject without additional
hassles (e.g. shipping supporting files at /subprojects/repofiles/)
Drop our internal copy of GVDB in favor of a subproject built
through Meson.
Since we're updating many years across, there has been GVDB API
updates that we need to adapt to: GvdbTable is no longer a refcounted
object, and gvdb_table_walk() is no longer offered to iterate across
values.
These largely affect our own set of GVDB tests though, the test
for gvdb_table_walk() was dropped, and so is the ref/unref one
(it basically does the same than gvdb/flat_strings, after dropping
the refcounting). These remaining tests stay useful, and should
ideally move into the GVDB repository, so it can run as a separate
suite here.
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
| |
For some reason this fails, but also is unused in this
project tree. We can stop asking for its existence.
|
|
|
|
|
| |
This generic method is available since meson 0.51 (which we already require),
while pkg.get_pkgconfig_variable is deprecated in 0.56.
|
|
|
|
|
| |
Tracker uses the fs module to build the documentation. It was
introduced in meson version 0.53.
|
| |
|
|
|
|
|
|
| |
Drop all the places left where we do specify things for Vala,
and rename the targets that had that name because of vala code
presence.
|
|\
| |
| |
| |
| | |
Build libtracker-sparql as a monolithic library
See merge request GNOME/tracker!504
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This was only a last resort for sqlite < 3.20, since module loading
API changed around that version, also it probably did us a disservice
since we barely updated the fts5.c file.
We no longer accept such old versions and we expect that sqlite has
a builtin FTS5 module, so we can drop this.
|
| |
| |
| |
| |
| |
| |
| |
| | |
Back in the day, 3.20.0 was stablished as the baseline to expect
a builtin fts5 module, due to module loading API changes in sqlite.
Fast forward to today, this version was released in Aug 2017 so it
is old enough to bump the dependency. This will allow us to drop our
in-tree fts5 module.
|
|/
|
|
|
| |
Also add the relevant versioning helper macros, so that we can
add new API right away.
|
| |
|
| |
|
|
|
|
|
| |
This allows distributions that ship both versions to choose which
version of libsoup will be linked.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Commit 8569c99d2926 ("Look for a Python installation with the tap
module") correctly added a dependency on the gi Python module for tests.
Commit 51ca035697a4 ("Make tracker:endpoint test work on macOS")
restructured Python probing by moving it to the top-level meson.build.
These two commits have the side-effect of exposing the fact that the
gi Python module is really only a test dependency, and should not be
required if building without tests.
As such, add a -Dtests=bool option to disable building tests. This
reduces the number of ninja targets from 232 to 188 on my non-test
build.
|
|
|
|
|
|
|
| |
We should use asciidoc's `a2x` driver instead of running xsltproc
ourselves. asciidoc 10 ships its stylesheets in a private directory.
Fixes: https://gitlab.gnome.org/GNOME/tracker/-/issues/347
|
| |
|
|
|
|
|
| |
Also add the relevant versioning helper macros, so that we can
add new API right away.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
This reverts commit 4446db1ba422b8f13cc96f75e49fbee7aa88d7d0.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
When building Tracker as a subproject, the CWD at build time is still
that of the root project. We want to check this, instead of the
subproject build root.
This makes the remote module correctly detected in the build dir
when building Tracker as a subproject (as it happens in tracker-miners
CI).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Tracker is an innocent bystander hindering libsoup3 port, since the
libsoup2/3->tracker->gtk3 dependency chain imposes a libsoup version,
graphical applications using libsoup cannot port at their own pace.
Make our remote code (connection & endpoint) be a private module
that is built against both versions of libsoup (if found), then we
pick one at runtime, with a preference on libsoup3 if libsoup2 .so
file is not seen in the already loaded libraries.
This patch should be reverted ASAP, once we can consider libsoup2
deprecated.
Fixes: https://gitlab.gnome.org/GNOME/tracker/-/issues/320
|
|
|
|
|
|
| |
This was used by tracker-extract which no longer lives in this repo.
https://gitlab.gnome.org/GNOME/tracker/-/issues/313
|
|
|
|
| |
currently it requires soup2 to be installed even if building against soup3
|
| |
|
|
|
|
|
|
| |
There was some experimentation on the version scheme to appease
https://gitlab.gnome.org/GNOME/tracker/-/issues/290, but it breaks
GNOME ftpadmin. Never again.
|
|
|
|
|
| |
This is redundant, and Fedora has made it a requirement to drop
unnecessary rpaths.
|
|
|
|
|
|
|
|
| |
This flag has detected some bugs and resulted in a number of good code
cleanups
However, some of these warnings cannot neatly be fixed, as GLib depends on
converting function pointers to (void *) and back again.
|
|
|
|
| |
Now that all occurrences have been fixed, we can enable this again.
|