| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We display the a file's basename in the properties window title.
exmaple.txt Properties
If a file has a really long name, this makes the dialog way too wide.
example_of_a_very_very_very_very_very_very_very_very_very_very_very_very_very_very_long_name.txt Properties
Yes, that went past the 72 characters limit for a commit message lines,
which proves the point.
Instead, let's truncate the basename to a reasonable limit.
The full name is available in the Basic tab of properties dialog anyway.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=754388
|
|
|
|
|
|
|
|
|
|
| |
Files with very long names, or verbose translations, can make this
dialog window too large.
To ensure a sane width, set a maximum width for the labels.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=786701 as well as the
issue reported in https://gitlab.gnome.org/GNOME/nautilus/-/issues/1727#note_998181
|
|
|
|
|
|
|
|
| |
The NautilusFile backend now supports birth-time parameter.
It is now leveraged to display the date-created for a file
in properties dialogue.
Closes https://gitlab.gnome.org/GNOME/nautilus/-/issues/1566
|
|
|
|
|
|
|
| |
Now that we have backend support for showing and sorting by btime, add
an optional "Created" column to the list view.
This enables using the attribute as caption for icon view too.
|
|
|
|
| |
No user visible changes yet.
|
|
|
|
|
|
|
|
|
| |
GIO 2.66 introduces support for Linux statx stx_btime.[0]
First, let's add support for this attribute in our file models.
No user-visible changes yet.
[0] https://gitlab.gnome.org/GNOME/glib/-/commit/a396fa9027bf9aaab0697789fe38efec53b09431
|
|
|
|
|
|
|
|
| |
When converting the array into a list, the list takes the data. But the
original container is leaked.
Get the array inside the helper function and free it after its data is
taken. Also rename the helper funcion appropriately.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We extract an array constant strings from a GVariant dict using the
"^a&s" format string, which is equivalent to g_variant_get_strv() [0]
g_variant_get_srtv() is "(transfer container)", meaning the caller
owns the array.
However, we don't free the array, resulting in a memory leak. [2]
Use autocleanup to fix the leak.
[0] https://developer.gnome.org/glib/stable/gvariant-format-strings.html#gvariant-format-strings-convenience
[1] https://developer.gnome.org/glib/stable/glib-GVariant.html#g-variant-get-strv
|
| |
|
|
|
|
| |
Also respect code style (return type in different line).
|
|
|
|
|
|
|
|
|
|
|
|
| |
We are leaking references to the singleton object in multiple places.
It's not a problem during runtime, but this means it's not properly
finalized when the application quits, leaking the tracker connection,
which is detected by valgrind.
Add g_autoptr() and g_clear_object() where missing. Also, to prevent
future leaks, document nautilus_tag_manager_get() and uniformize its
use as either an autovariable initializer or a field initializer.
|
|
|
|
|
|
| |
Autocleanup works only at the end of the scope it's declared.
So, it must be declared inside the loop to work as intended here.
|
| |
|
| |
|
|
|
|
| |
g_object_get() returns a ref to the property object. We must unref it.
|
|
|
|
|
|
|
|
| |
We hold a ref since commit 6b16de613dc87b9f84d87a46ac5987b6d7087a5c
But we never release it on slot destruction, so it leaks.
Release it on destruction.
|
|
|
|
|
|
|
|
| |
g_value_set_object() increases the reference count of the passed object
But we already own a reference, so we end up leaking it.
Use g_value_take_object() to reuse the owned reference instead.
|
|
|
|
|
|
| |
It is visible in the about dialog and may be wrong.
Closes https://gitlab.gnome.org/GNOME/nautilus/-/issues/940
|
|
|
|
|
|
|
|
|
|
| |
When copying file(s), the Operations 'stop' button is elliptical, but it
should be circular.
So, replace our custom .nautilus-cirular-button class with .circular,
for which GTK employs the necessary stylesheet magic.
Fixes https://gitlab.gnome.org/GNOME/nautilus/-/issues/1483
|
|
|
|
|
|
|
|
|
|
|
|
| |
Label ids are employed to describe relations between widgets for
accessibility support.
However, we are using numbers for ids, which is harder for human
readers.
To make it easier, change the ids to meaningful names.
Closes https://gitlab.gnome.org/GNOME/nautilus/-/issues/1420
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Currently when closing the currently focused slot, the next slot to gain
focus is the earliest created. This is because the default behaviour,
which is to open the next tab at the same position or the new last tab
if the previous tab was closed, used in other gnome apps like epiphany
and terminal is overriden by manually setting active slot to earliest
inactive one.
To provide consistent behaviour, remove code that overwrites the default
gtk notebook tab closing behaviour.
Fixes https://gitlab.gnome.org/GNOME/nautilus/-/issues/1567
|
|
|
|
| |
They cut on boilerplate and improve readability.
|
|
|
|
|
|
|
|
|
| |
- Use GList API instead of reimplementing it:
- g_list_copy_deep()
- g_list_foreach()
- g_list_delete_link()
- NULL-initialize fully-owned lists.
- Don't reimplement nautilus_mime_get_default_application_for_file()
|
|
|
|
|
|
|
|
|
|
|
|
| |
When showing the properties for multiple files with the same MIME type,
The app_info variable is reassigned multiple times in a for loop to
the return value of nautilus_mime_get_default_application_for_file(),
which returns a caller-owned refference.
So, we leak a reference on each reassignment.
To fix this, declare the variable inside the loop block, to ensure
autocleanup after each loop iteration.
|
|
|
|
|
|
|
| |
Also fixes a few leaks:
*uri in setup_volume_usage_widget()
*error in set_as_default_clicked_cb()
*message in set_as_default_clicked_cb()
|
|
|
|
| |
Use modern GLib utilities for memory management.
|
|
|
|
|
|
|
|
|
|
| |
It's wrong to set a pointer to an object as data without incrementing
its refcount.
But there is actually no need to pass a real pointer as data here, as
we actually want a boolean.
Use a pointer conversion macros instead.
|
|
|
|
|
| |
Use "self" as symbol name for the NautilusPropertiesWindow* instance
in methods and signal handlers.
|
|
|
|
|
|
| |
ROW_PAD was a macro used by code the UI creation
code which has been removed in previous commits.
Now It's not useful.
|
|
|
|
| |
Unused since commit bd30a21a0ce1b40ca59814f731bd761670601aaa
|
|
|
|
|
|
|
| |
This reverts commit e3031953e40b4fe067d566ac6c403127d7b6c266
It worked around an issue in tracker, which is fixed now:
https://gitlab.gnome.org/GNOME/tracker/-/issues/252
|
|
|
|
|
|
| |
get_query_status() is a wrapper for tracker_sparql_cursor_next_finish()
and it's weird for _finish() not to be called directly by the
GAsyncReadyCallback function.
|
|
|
|
| |
See https://gitlab.gnome.org/GNOME/nautilus/-/merge_requests/553#note_989920
|
|
|
|
| |
Fixes: https://gitlab.gnome.org/GNOME/nautilus/issues/1319
|
|
|
|
|
|
|
|
|
|
|
| |
The default location (i.e. g_mount_get_default_location) should be
used when opening mounted, but we should not use it instead of the root
button in the path bar. Because it is currently impossible to navigate
to parent folders if default location is set. Let's change
nautilus_get_mounted_mount_for_root logic and return mount only for
root path so we can see the real root in the path bar.
https://gitlab.gnome.org/GNOME/nautilus/issues/1319
|
|
|
|
| |
(cherry picked from commit 8e67058fb368146183f0159292fa5ab483d9c4e9)
|
|
|
|
|
| |
Replace a guard check if/else statement with an early return and
minimize unnecessary function calls.
|
|
|
|
|
|
|
|
|
|
| |
The saved window state (whether the window is maximized and its initial
size) should be the state, that the user would most likely want the next
opened window to start with. As the tiled state doesn't make sense
without other windows and because it's not really possible to properly
restore it, it will not be saved anymore.
Fixes https://gitlab.gnome.org/GNOME/nautilus/-/issues/1685
|
|
|
|
|
|
|
|
|
| |
Since commit ece6b825, xdg-desktop-portal is used to set wallpapers. This
introduced the following build warning: "‘set_wallpaper_fallback’ defined
but not used", because the fallback is used only if it is built without
libportal support and it is not guraded by #ifdef. Let's use the fallback
also in the case of xdp_portal_set_wallpaper portal failure, which fixes
the warning as well.
|
|
|
|
|
|
|
| |
GNOME CI runners have been updated to the latest libseccomp so the
problem should no longer happen.
This reverts commit ab55380f200e5ea03116c5871607d125deff844c.
|
|
|
|
| |
(cherry picked from commit b62f9a7422edd181c92ef0d0ccce7b8d91860497)
|
|
|
|
|
|
|
|
| |
The pipeline currently fails with Fedora rawhide, because g-ir-scanner fails
with failures like: "ldd: error: you do not have read permission for
`/builds/GNOME/nautilus/_build/tmp-introspectgwhh729q/Nautilus-3.0'".
This obviously affects more projects:, e.g. GNOME/grilo!62. Let's use
Fedora latest for now as a workaround.
|
|
|
|
|
|
|
|
|
| |
The triage job is broken, which regularly causes CI failures. I've
made some attemts to fix it but I failed. I don't have capacity to
spent more time on it. Let's remove the job completely for now to
prevent the confusing CI failures.
Fixes: https://gitlab.gnome.org/GNOME/nautilus/-/issues/1625
|
|
|
|
|
| |
The Tracker 3 dependencies are installed currently from unofficial COPR
repository. Let's use the official Fedora packages instead.
|
|
|
|
|
|
| |
The GNOME runners are no more privileged and thus it is not possible to
use Docker from the pipeline. Let's use Buildah instead Docker to fix
the image generation.
|