Adds the bar subcommand `gaps <amount>|<horizontal> <vertical>|<top>
<right> <bottom> <left>` to set gaps for swaybar. Due to restrictions on
margins for a layer_surface, only the sides that are anchored to an edge
of the screen can have gaps. Since there is support for per-side outer
gaps for workspaces, those should be able to be used instead for the
last side.
The positions "left" and "right" are not allowed by the man page, remove them
from the allowed positions. Also print an error to stderr if we default to the
bottom position.
Fixes#2878
* The loop functions are now prefixed with `loop_`.
* It is now easy to add timers to the loop.
* Timers are implemented using pollfd and timerfd, rather than manually
checking them when any other event happens to arrive.
Previously, when the bar was hidden, the height would be set to 0.
This meant that if the bar was empty upon reshow, it would not render
since the height was still 0, which made it seem there was a problem.
Now, the height is not reset, but the width is, to indicate upon reshow
that the layer surface needed reconfiguring.
Since wayland does not currently allow swaybar to create global
keybinds, this is handled within sway and sent to the bar using a custom
event, so as not to pollute existing events, called bar_state_update.
As well as adding the hidden_state property to the bar config struct,
this commit handles barconfig_update events when the mode or
hidden_state changes, and uses a new function determine_bar_visibility
to hide or show the bar as required, using, respectively,
destroy_layer_surface, which is also newly added, and add_layer_surface,
which has been changed to allow dynamically adding the surface.
The received json is handled outside of the case statement, which will
allow better extensibility.
This commit also introduces the variable bar_is_dirty, the return value
signifying whether the bar requires rendering.
This distinguishes the binding mode from the distinct config mode, as
well as removing mode_pango_markup from the config struct where it
should not be present.
When running swaymsg -v, the version returned is actually the version of
swaymsg itself, yet the message displayed was "sway version <version>".
This can create confusion if users update sway and swaymsg but don't
restart sway, then use swaymsg to check the version.
This patch changes the wording to be "swaymsg version <version>"
instead, and likewise for swaybar.
To get the version of a running sway instance, users should run swaymsg
-t get_version.
Sway sets a default status_command which runs date every second. This
patch removes this behaviour so the user can have a NULL status bar if
desired.
I had to swap swaybar's event_loop_poll and wl_display_flush so that it
would map the initial surface.
It's better to use DT_RPATH dynamic section of the elf binary to store
the paths of libraries to load instead of overwriting LD_LIBRARY_PATH
for the whole environment, causing surprises. This solution is much more
transparent and perfectly suitable for running contained installations
of wayland/wlroots/sway.
The code unsetting the LD_LIBRARY_PATH/LD_PRELOAD was also deleted as
it's a placebo security at best - we should trust the execution path
that leads us to running sway, and it's way too late to care about those
variables since we already started executing our compositor, thus we
would be compromised anyway.
This commit fixes a segfault in swaybar on FreeBSD that was caused by
using getdelim with EOF as delimiter on an infinite stream. The FreeBSD
implementation handles the "no more data, delimiter not found, and EOF
not reached" scenario as an error, so it can't be used to read the
output of status command. This commit replaces the getline/getdelim
calls with reading all available data from the stream in one go.
This prevents an signed-to-unsigned conversion error on buffer_index if
getdelim fails and returns -1, which caused swaybar to try to search the
header for the array and immediately failing
If the tokener parses incomplete json, it sets its error value to
json_tokener_continue. This means that extra json should be provided,
but the code was providing the entire object again. In the interest of
simplicity, the tokener is reset so that buffer_pos always points to the
start of the current object.
This was the source of numerous bugs, from hotplug events not being
received to segfaults because wl_display_roundtrip was making the
bar process unplug events while blocking in an iteration over all
outputs.
This now correctly handles an incoming json infinite array by shifting
most of the heavy listing to the json-c parser, as well as sending
multiple statuses at once. It also removes the struct
i3bar_protocol_state and moves its members into the status_line struct,
allowing the same buffer to be used for both protocols.
This now uses the getline function to receive the header, replacing
read_line_buffer, which has been deleted since it is otherwise unused.
Furthermore, once the protocol has been determined, the current status
is handled immediately to be shown (though this has not been added for
the i3bar protocol since it has not yet been rewritten to handle this).
This now uses getline to correctly handle multiple or long statuses. It
also removes the struct text_protocol_state and moves its members into
the status_line struct.
This prevents blocks from being destroyed before their hotspots are destroyed,
in case it is used for a pending click event that fires between the bar
receiving a new status, which destroys the block, and the bar rendering the new
status, which destroys the hotspot; this problem can be easily produced by
scrolling on a block that immediately causes a new status to be sent, with
multiple outputs
This does the following:
* Adds a baseline argument to get_text_size (the baseline is the
distance from the top of the texture to the baseline).
* Stores the baseline in the container when calculating the title
height.
* Takes the baseline into account when calculating the config's max font
height.
* When rendering, pads the textures according to the baseline so they
line up.
As well as ignoring scroll events on status elements when click_events
is enabled.
Previously, using the scroll wheel on a workspace button would switch to
that workspace instead of scrolling through them. Clicks and scrolling
on status elements would always be processed by swaybar, too. So in case
you were using scrolling as volume control on a status item, swaybar
would additionally scroll through your workspaces.
Uses 'visible' instead of 'focused', since we may scroll on a bar which
isn't the focused output.
We can't use "next_on_output" or "prev_on_output" to implement this,
because it only modify the focused output. So scrolling on an unfocused
output will affect the incorrect one.
We just use the "workspace name" command instead.
We defer the removal of entries until after the poll loop has finished.
Otherwise we may end up adjusting the poll array while we're still
reading from it, causing us to skip events.
