The wcwidth() POSIX helper returns the number of cells that a character
occupies. This is normally 1, however, if it is 0 we simply discard this
input (this should never happen as all characters that wcwidth() returns 0
for should already be handled by the VTE layer).
But if it is >1, we write the character in the first cell, set the width
correctly and mark all following characters as width=0.
Everything else is left unchanged, that means, the running application has
to be aware of wcwidth() and correctly send two backspaces to clear a
width=2 character and so on.
This behavior almost exactly matches the xterm behavior. There are some
corner cases like erasing only one part of a multi-cell character etc.
that might differ. However, these are non-standard cases that no
application should never rely on.
Anyway, if we spot any of these differences between xterm and kmscon, we
should fix kmscon to behave like xterm does.
Thanks to Shixin for looking this up in other emulators.
Reported-by: Shixin Zeng
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Multi-cell glyphs are glyphs that span across multiple horizontal cells.
The font renderers already support this. This patch fixes the console
renderers to take advantage of this.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
The alternate screen is mainly used to avoid linking lines to the
scrollback buffer. This implements this feature so only main-screen lines
end up in the scrollback buffer.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We correctly free all new space and this is really needed due to rotations
in the active screen and scrollback buffer.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We didn't correctly initialize new lines on resize. In fact, we didn't
initialize new cells at all. This was all covered by a nasty-fix in the
screen-resize handler.
This fixes the line-resize handler to initialize all new cells correctly.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We have to free both, main and alternate buffers on shutdown. We only used
to free the currently active buffer.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
The VTE layer can now set the TSM_SCREEN_ALTERNATE flag to switch to the
alternate screen-buffer and back.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
This changes the line-allocation helpers of TSM-screen to allocate an
alternate buffer along with the main-buffer. We still keep the "lines"
pointer to the main-buffer all the time, but this will allow to implement
alternate-xterm-screens later.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
This adds a new helper that returns the UTF8 encoded data of the selected
parts in the tsm-screen object.
There is still much to do and it isn't a nice solution. However, it's a
proof-of-concept and works for now so we can just keep it.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We need to draw the final character of a selection with inversed
background, too. This is a bit tricky as the selection may be inversed
itself. Therefore, we just keep a flag that tells us whether the previous
character was selected and just draw the new character also selected.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
This implements UI selection for the console buffer. This selection is not
to be confused with PTY application selection provided by the
mouse-protocol via VT200 protocol.
Instead, this selection allows UIs to tell the TSM layer to select a
special part of the screen. Moreover, it allows selections to go into the
scrollback-buffer and to be bigger than a single screen. The UI can even
implement scrolling during selection to allow arbitrarily big selections.
There might still be some special cases where we need to fix selection.
However, it already works pretty well.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
If we have pointers somewhere into the scrollback-buffer, there is not
easy way to see which of two lines comes first. Therefore, we introduce
scrollback buffer IDs. These are unique and every line gets one assigned
when it is linked into the sb-buffer. These IDs guarantee that front lines
have lower IDs than bottom lines.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
This function is not used and I do not intend to use it anywhere soon.
Therefore, remove it so it doesn't generate sparse warnings.
The only reason why I kept is was that we might want to retrieve these
when increasing window size. However, this will not be implemented in the
near future as we need to keep a buffer-fill state for it.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
When drawing scroll-back buffer, we cannot guarantee that the lines are as
long as the current screen. Therefore, check that we aren't accessing out
of bounds and draw empty glyphs instead.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
If we scroll the screen during resize to keep the screen-content constant,
we should also move the cursor position. Otherwise, the application has an
inconsistent state.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We call screen_scroll_up() under special conditions when resizing. We
require it to allocate the lines with the new width even though we didn't
set it yet. Therefore, we need to set the width before calling it and then
adjust the height.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
When enabling the printf() logic for the logging subsystem, several
warnings were produced about missing or wrong specifiers. This fixes all
those occurrences.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
There is some nasty bug where we do not correctly clean buffers when
resizing. So we now clear the whole offscreen region on resize to go sure
everything is clean.
Note that cells can pretty much move everywhere during their
offscreen-life so this is a quite sophisticated task to keep track which
cells are clean and which not. Therefore, simply clean all of them when
taking them on-screen.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We already have an llog context so use it instead of relying on kmscon-log
subsystem. TSM-screen is now fully independent. Next step is TSM-vte.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Instead of accessing kmscon state from TSM, we now introduce options to
control the behavior of TSM-screens. Apart from logging, TSM is now
independent of any kmscon state/code.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Allow setting global options for screen objects. These are different from
flags as the latter affects the state-machine but options affect global
operation modes.
Options should be set by the application while flags are set by the VTE
handler. The latter might change often, while options should be set on
startup and then remain mainly constant.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We need to avoid logging to stderr directly in TSM so introduce the
tsm_log_t object similar to eloop.h.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
All TSM files use the "tsm_*" prefix and the object is now named "screen"
so rename the files to resemble this.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
