Under a diff renderer that states which diff line each row shows, such
as delta, the queries about the lines of the main view read the rows by
those records alone. Still, they took the text of every row as well,
and for a diff of 300,000 lines this was most of the time each of them
took. Moving the selection by change block runs two of them for every
key press.
Take the records of the rows first, and their text only if no row
states a diff line.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Under a diff renderer that states which diff line each row shows, such
as delta, moving the selection by change block in a diff of 300,000
lines lags noticeably. One reason is that the queries about the lines
of the main view collect the records of every row from the row's
cells, going through every cell of every row. Under delta, a row has
about 130 cells.
Collect the records of a line once, when the write cursor moves on to
the next line, and take them from there. A line that is written to
again is read from its cells until it has been finished again.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
This is an improvement for users who use a side-by-side diff renderer on
a very large screen; in this scenario it was hard to see which hunk on
the right side was selected.
If a view doesn't wrap, a search match can be past its right edge, or
left of it once the view is scrolled sideways. Going to such a match
with /, n or N only scrolls the view vertically, so the match stays out
of sight. This matters most for a diff with gui.wrapLinesInDiffView set
to false, where a change is often at the end of a long line.
Scroll the view sideways if the match isn't on screen. Put the match a
third of the way across the view, so that some of the text before it
shows too. If the view can show the match without scrolling, go back to
the left edge instead. This applies to every view that doesn't wrap,
including the side panels.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The main view keeps its horizontal scroll position when a diff is
rendered again. If the new rendering is narrower, for example because
the only long line in it was just staged, the view can end up scrolled
past the end of every line and look empty. If one line still reaches
into the view, the view shows only a sliver of it, although the whole
diff would fit without scrolling.
Once a render has been read to the end, scroll the view back as far as
it takes to put the end of the widest line at the right edge. A long
diff that is read only in part is checked once the rest of it has been
read.
Nothing scrolls the main view horizontally yet, so this changes nothing
on its own. It prepares for scrolling the focused main view sideways
with H and L.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
If a view that shows the inclusion gutter is scrolled horizontally, its
content is drawn over the gutter. The marker of a marked line is
overwritten, and the content starts the gutter's width too far to the
left. The draw loop skips the cells scrolled past the left edge only
until x reaches 0, but the content area starts at the gutter's width.
Nothing scrolls such a view horizontally yet. A later commit lets the
focused main view scroll sideways, and the gutter shows there while
building a custom patch.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Nothing in lazygit calls View.Line or View.Word; they came with the
original gocui code and have never been used. They are also the only
callers of realPosition. That function ignores the inclusion gutter when
it maps a view position to the buffer. On top of that, Word uses the
returned column as an index into the line's cells, and this goes wrong
for wide characters.
Remove them, together with realPosition and the indexFunc helper. This
leaves contentX as the only code that maps a column of the view to a
column of the content.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Search matches beside the inclusion gutter are highlighted in the wrong
place, and hovering over or clicking a hyperlink there acts on the wrong
cell. The draw loop starts the content after the gutter, but the code
that maps a view column back to a column of the content only adds the
horizontal origin and ignores the gutter.
Add a helper that does both, and use it for search highlights, hover
and clicks.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
A double-width character that starts in the last column of a view is
drawn over the view's frame. One that is scrolled halfway out to the
left is skipped, and the column it reaches into isn't drawn at all. The
draw loop clips cells against the edges of the view as if every cell
were one column wide.
A terminal can't show half of a double-width character. So draw a blank
in the column that is on screen, in the character's colors, and let the
selection highlight cover that column like any other.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Actions that hand control to another application need visible
acknowledgement without moving a view's cursor or replacing renderer
colors. Reverse the narrow selection bar independently of selection
state, and clear transient flashes whenever the terminal UI suspends.
A view can need repainting even when its cached wrapping is still valid.
Track that state independently so content-only flushes do not overload
tainted, whose only job is to request a viewLines rebuild.
A patch built from what is on screen has to show what is in it, over the
diff those lines were taken from — which may be a diff renderer's
rendering of it, whose bytes are none of ours to touch.
So the marks are drawn in a gutter over the content, from the lines the
patch holds rather than from where they were drawn last. They are worked
out again whenever a pane's content settles and whenever the patch
itself changes. Working them out as the content settles keeps them right
across a renderer switch, a change of context size, or a walk through
the commits.
They are shown whenever the main view shows the diff the patch is built
from, whether or not it has the focus. The pane beside the diff previews
the patch all the while, and marks that came and went with the focus
would look like a bug while browsing. A diff of any other commit gets
none.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The lines of a commit's diff that are in the custom patch being built have
to be shown as such over whatever a diff renderer made of that diff, whose
bytes we can't touch — and mustn't, since everything we know about a row is
keyed by where it sits in the content.
So the marks are a decoration drawn over the content instead: a column
reserved at the left of every line, blank but for the lines that are in,
with the content shifted along past it. Nothing about the content changes
but the width it has to wrap in.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The main view shows a diff renderer's picture of a diff, and that is not
what you want on your clipboard: a renderer may drop the +/- column, move
the line numbers into a gutter, group a hunk's deletions before its
additions, or leave lines out altogether. So copying takes the lines from
the diff itself, locating them by the identity of the selected rows.
Only the panel that produced the diff can produce it again, though: the
working tree's staged or unstaged side, a commit's, a stash entry's. The
new seam is there for that. It asks per file, so that copying three lines
of a commit's diff doesn't fetch the whole of it, and it will grow the
actions on a selection as staging and patch building arrive.
The clipboard gets the run of diff lines from the first selected line of
a file to the last, so that lines the renderer hid come along and the
result still reads as a diff. Headers count as selected lines too. A hunk
header names the first line of its hunk, so it is looked for the way a
line of the file is. A file header names no line at all, and a rendering
may draw it over any number of rows, so a selection touching one of them
takes the whole header.
As in the patch explorers, a selection that is all additions or all
deletions loses its +/- column, ready to be pasted into code. One that
reaches into a header keeps the columns, and what comes out is a patch
fragment rather than lines of code.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
Which pane a side of a file's diff appeared in depended on what else the file
had: the staged side had the lower pane while there were unstaged changes
above it, and took over the upper one when there weren't. So the same content
moved about, and which side a pane was showing was something the code had to
work out from the file's status rather than knowing from the pane.
Now each side has a pane of its own — unstaged above, staged below — shown
when there is something on it. A file with nothing unstaged shows its staged
changes in the lower pane alone, which then has the whole space, including
the label of the key that focuses it.
Nothing about this is visible to the user: the same diff appears in the same
place, with the same title and the same label, and the same keys focus and
scroll it. What it is for is the code, which no longer has to ask the file
what a pane is showing — and, once the diff can be staged from, no longer has
to make one key mean opposite things in the same pane.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
A range or hunk selection covers a stretch of the diff, and restoring only
the cursor left the other end pointing at whatever line the new rendering
happened to put there — with more context lines above, a selected hunk
would grow a tail of context it never covered.
So the other end is remembered by identity too, and put back before the
cursor. If it is a line the re-render dropped, the selection is left as the
single line we landed on: there is no telling which line inherits a
selection's edge, and a wrong guess acts on lines the user never chose.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
Pressing { or } re-rendered the diff with less or more context around each
change and dropped you back at the top of it, so finding your way back to
what you were reading was on you.
Remember where the view is before triggering the re-render, and have the
restore put it back there. The line to remember is the selected one when a
selection is showing, and the middle visible line otherwise — what you are
looking at, rather than the view's top edge. It is remembered by identity,
because the new rendering puts it on a different line of the view, and it
is found again by the records a diff renderer states for its rows, or by
parsing the rendering as a diff where it still looks like one.
The line may not be there at all afterwards: shrinking the context takes
context lines away. So the lines around it come along as fallbacks, nearest
first, and the view lands on the nearest one that survived, put back on the
screen row it was on — leaving what you were reading where it was, give or
take the line that went. The walk that collects them covers the whole diff
rather than stopping at the nearest change on either side: those always
survive a context-size change, but not everything a re-render can do to a
diff is that considerate.
Under a diff renderer that says nothing about its rows there is nothing to
look for at all, and the view is left at the offset it had instead: often not
the line it was on, but always closer to it than the top of the diff. That is
what the re-render asks for besides the restore, since it is a different
command from the one that produced what is on screen and would otherwise be
taken for a diff the user has never seen.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
A re-render builds into an off-screen buffer and swaps it in when it has
read enough to paint. Deciding where the new content should be shown means
reading it before that swap: afterwards it is on screen already, and
whatever we then scroll to has been seen at the wrong position first.
So expose the off-screen buffer's diff-line contents and line count, the
latter for telling when a line found there has a screenful below it. The
contents come in two forms, the whole buffer and everything from a given
line on, so that a reader following the render as it loads can look at each
line once instead of re-reading the buffer per line.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
Now that a diff view always carries a selection, the highlight fights the
diff itself for the line's colours. Painting the selection across the whole
line takes over the background, and a selected hunk becomes one solid block
with no boundary between what was removed and what replaced it — the more
lines you select, the less you can read.
This isn't specific to renderers like delta that say which side of the diff
a line is on by colouring its background, though they suffer most: git's own
output puts red and green text on that background, which reads badly too.
Since there is no rendering of a diff that a full-width highlight doesn't
degrade, there is nothing here worth configuring — the bar is simply what
diff views use.
Left edge only, and two columns wide: every convention for marking a row of
coloured content — change bars, diff gutters, selection gutters — puts the
marker on the left, and bracketing both edges reads as framing instead.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
Users who set hunk mode as their default get it when entering the staging
view, and the focused main view is on its way to replacing that view, so it
has to behave the same: focusing selects a whole change block, and clicking
a change line selects that line's block, ready to act on. A click on a
context line still selects only that line, since the click points at it
precisely — you may well want to edit it. A click inside the block that is
already selected does the same, giving hunk mode up and going back to line
by line. That keeps a single line reachable with the mouse in a block too
long to see the end of, and matches how a click inside a range selection
already collapses it.
The block offered up is the first one that begins on screen, so that its
whole extent can be seen before acting on it. Only when none does — a change
too long to fit on the screen — is the block that reaches into the view from
above taken instead, and the selection then extends to its first line off
screen: focusing a diff must not move it, so nothing here scrolls.
The staging view makes one exception, and so must we, or a file that is one
solid block of changes (a file you just created, or deleted) would come up
with all of it selected. The staging view asks the parsed patch; we ask the
rendered diff the same question, which needs no second git invocation and
works over a whole commit's diff, where the answer differs per file.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The focused main view could be scrolled but not pointed at: there was no
way to say "this line" or "these lines", which is what every line-level
interaction needs — editing a line, copying part of a diff, and, later,
staging. So a diff main view now always carries a selection while focused,
starting at the first change already on screen so that focusing doesn't
move the view.
The mode of the selection lives on the main context (a single line, a
range from a fixed anchor, or the change block around the cursor); the
selected line and the range anchor stay in the view itself, whose native
range select draws them, so there is no new highlight machinery. The modes
and their keys are the ones the staging view has, including that the arrow
keys step from block to block in hunk mode.
A pane holds something to select only when what it is showing is a diff
with changes in it, which the render works out once its content is final: a
placeholder message is not a diff, and neither is a diff with nothing in it
at all, such as a binary file's or an empty commit's. Whether a selection
is then drawn there follows from the context stack, as it does for every
other view.
A click in the focused main view now has a meaning again: it selects the
line it points at.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
The count comes from the view's buffer, which a rendering task appends to on
its own goroutine, so reading it without the write mutex is a data race. Nobody
called it until now, which is why nothing has tripped over it; the next commit
does, from the UI thread while a render is still loading.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
Wherever two diff lines end up on one rendered line, a renderer emits
their records back to back: the deletion and the addition of a
modification collapsed into a single column, or a banner announcing a
file and its first hunk at once. A changed line that is empty is
rendered as its record alone. Attaching a record only to the cells it
precedes loses all of these — the last record of a run wins, and an
empty changed line becomes a line we can say nothing about at all.
Give such a record a cell of its own instead. It renders nothing, so the
diff looks the same, but the line keeps every record it was given.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
A diff renderer that restructures the diff — into columns, or with the
+/- markers replaced by colour — leaves us no way to tell which line of
which file a rendered row came from, which is what acting on the row
requires. The OSC 1717 protocol has the renderer say so directly: it
prefixes each line it renders with a record naming the file and the
line's position in the old and new versions of it.
Attach each record to the cells it precedes, so that a row's records
survive wrapping and the columns of a side-by-side rendering, and hand
them to readers together with the row's text: the two have to describe
the same buffer, and a re-render can rebuild it between two reads.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
A reader that parses a view's content rather than showing it wants the
text the writer wrote, and the cells don't always spell it. A tab is
expanded into the spaces it fills, so a line read back from the cells
ends in one to four spaces where the writer put a tab. A carriage
return moves the write cursor back to the start of the line, so the
text written after it overwrites what came before.
The parser of the main view's diff, which the next commit adds, meets
the first case in every header of a file whose path contains a space.
git terminates the path field of a "---" or "+++" line with a tab
then, and a parser reading the cells takes the spaces the tab became
for part of the path. That path names a file that doesn't exist, so
the file's lines can't be acted on.
Keep the text as written per line, from the first character on that
the cells spell differently, so that a line without a tab or a
carriage return costs nothing. LinesAsWritten hands it out the way
BufferLines hands out the cells' text.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
OverwriteLines is asked for a line and writes the one below it when the
write before it ended in a newline. The view holds such a newline back
until more content arrives, so that it doesn't end in an empty line,
and OverwriteLines moved the write cursor without letting go of it, so
the write that followed advanced to the next line first.
Move the cursor through SetWritePos, which drops the pending newline
along with it.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Everything about a diff's content — which file and line a row belongs
to, whether it's a change — is a property of the unwrapped buffer line,
while the cursor, clicks and the range selection all speak in view
lines, which count wrapped segments. Reading the content under the
cursor therefore needs the mapping in both directions, and doing it
outside gocui isn't possible: the wrapping is internal, and the caller
couldn't take the view's lock across the lookup and the read.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Wrapping the content at another width moves every line of it to a
different view line. The positions into the view are all view lines: the
scroll offset, the cursor, a range's anchor. Each of them is then left
pointing at a line it was never on. Committing the last of the staged
changes widens the main view by half a screen, and that moves a selected
hunk somewhere else entirely.
Carry the positions through the lines of content they were on. A position
always meant a line of content rather than a view line. The line the
cursor is on keeps the row it was drawn on, so it stays in front of the
user rather than the view scrolling under it; a view with no cursor on
screen keeps its own place instead. A range's ends go on the outermost
segments of their lines, since a range covers lines of content and not
the segments those lines are drawn as.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
A test asks which lines of a view are selected; a wrapping view's cursor and
range anchor answer in view lines, which count the segments each line is
drawn as. Going through the segment-to-line mapping keeps the answer in the
terms the question was asked in, and a line the selection covers several
segments of is reported once.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
The text of the selected line is always bold. Some users don't like
this (#2304), but there is no way to turn it off. Setting
selectedLineBgColor doesn't help. Its attributes apply to the text too,
so it can add bold, but it can't take it away.
Add gui.theme.selectedLineFgColor for the text of the selected line, in
focused and unfocused views alike. Its attributes are added to those of
the text, and a color replaces the colors of the text. It is [bold] by
default, so nothing changes unless you set it; 'default' leaves the
text as it is.
Putting bold into the default of selectedLineBgColor instead wouldn't
work well. To turn it off, you would have to replace the whole list, and
lose the color that is computed from the terminal's background.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
gocui draws the palette colors 0 to 7 on the selected line in their
bright variants. This was meant for the blue highlight on a dark
background, but in most dark palettes the bright variants are so close
to the normal ones that it's barely visible. Elsewhere it makes the text
harder to read. In many palettes, the bright variants are lighter, and
they don't work on the light highlight of a light background. Solarized
maps most of its bright colors to its grays, so colored text on the
selected line turns gray.
Draw the text of the selected line in the same colors as on the other
lines.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Rendering a view's content again while a search is on leaves the "x of y"
describing the content that has just been replaced. The status is worked
out when the search is typed and again when a key steps through the
matches, and a render is neither. Change the diff context size while
searching the focused main view, and the count stays as it was, however
many matches the wider context brought in or took away.
Run the search again over the new content once the render has finished
putting it there.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
A view's search positions were worked out again from every write, and
each of those walks the whole view. Content arrives a line at a time, so
rendering into a searched view costs a walk per line. Streaming 2000
lines takes 565ms, where the same render into an unsearched view takes
about 10ms.
Mark the positions stale on a write instead, and work them out where they
are read: when the view is drawn, when a key steps through the matches,
when the status is asked for. That is at most once a frame, and the same
2000 lines now take 8ms.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
A view can only be drawn with the focused frame and title colors while it
is the current view, but a panel made of an outer view and an editable
field embedded in it has to look focused as a whole, whichever of the two
the keyboard is pointed at.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
It existed for the incremental re-render: a shorter render left the previous
one's view lines in the tail (deliberately, to avoid a blank frame), and this
cleared them once the new content was fully read. Async renders now build
off-screen and swap in whole, so refreshViewLinesIfNeeded truncates the view
lines to the buffer and no tail can form. All the call at end-of-input still
did was discard every wrapped line and force the whole buffer to be re-wrapped
on the next draw, which is pure work on a large diff.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
A cmd/pty re-render used to overwrite the displayed buffer from the top
down as lines arrived, relying on keeping the previous render's view-line
tail to avoid a blank frame. That left the view showing a mixture of old
and new content while loading, and any reader (draw, clicks, the
view-line mapping) could observe a half-written buffer at the wrong
scroll.
Instead, build the new content in a second, off-screen viewBuffer: until
the task has read enough to paint, writes go there and the displayed
buffer — and so everything every reader sees — is left untouched. Once the
task reaches its first-paint point (InitialRefreshAfter, or EOF for short
content) it swaps the off-screen buffer in atomically, so the view jumps
straight from the previous render to the new one with no intermediate
frame. Subsequent lines append to the now-displayed buffer.
Swapping at the first-paint point means the displayed buffer is only a
viewport tall when it appears and then grows as the rest streams in toward
the count needed for an accurate scrollbar. The scrollbar is sized from the
displayed buffer's height, so left to itself the thumb would shrink and
snap back during that growth (most visibly: the files panel's periodic
refresh making the thumb jump while scrolled down). The total height the
scrollbar needs is a strictly later quantity than the viewport-fill paint,
so no single early swap can have both right. FreezeScrollbarHeight therefore
records the view's height when a load begins and the scrollbar is held there
— growing only if the new content turns out taller — until the load ends; a
synchronous render superseding the load releases it. This mirrors the layout
clamp, which already ignores the partial content height while a view loads.
With the swap doing a wholesale replace, refreshViewLinesIfNeeded can
truncate the view lines to the current buffer: there is no longer a
half-loaded shorter buffer whose tail we must keep showing, so a stale
tail never forms. clear()/Reset() abandon any in-progress off-screen
render so a synchronous SetContent after a stopped task writes to the
display.
The swap holds writeMutex for now; it could later move to the main thread.
Flicker behaviour still needs interactive verification (LAZYGIT_SLOW_RENDER
+ a real diff renderer).
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
write, writeCells, makeWriteable, parseInput and
autoRenderHyperlinksInCurrentLine produced cells into v.buf; move them onto
viewBuffer so they can write into any buffer, not just the displayed one.
The display-side effects that don't belong to content production —
tainting, clearing hover, updating search positions — stay behind in the
View.write wrapper, which delegates the actual writing to v.buf.write(v).
Render config the writer needs (Editable, colors, width, tab width,
hyperlink auto-render) is read from the passed View. Behaviour-preserving:
the wrapper still always targets v.buf.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The fields that make up a view's content and the act of writing to it —
the cell buffer (lines), the write cursor (wx/wy), the escape-sequence
decoder (ei) and the held-newline flag (pendingNewline) — were loose
fields on View. Bundle them into a viewBuffer struct that View holds by
pointer. This is a behaviour-preserving prep refactor: every access just
goes through v.buf now. It sets up rendering into a second, off-screen
viewBuffer that can be swapped in atomically, so an async re-render never
exposes a half-written buffer to readers.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
hyperlinkAt (the click path) and onMouseMove/findHyperlinkAt (hover) read
v.viewLines without holding writeMutex, unlike every other reader. They run
on the event-handling goroutine, so a re-render on the task goroutine can
shrink or rebuild viewLines between the bounds check and the indexing,
causing an out-of-range panic (observed: "index out of range [60] with
length 0" while hovering during a diff re-render).
Take writeMutex for the duration, like the other viewLines readers do, so
the check and the access see the same slice.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Reading a view's internal buffer belongs on the view itself, next to
findHyperlinkAt, rather than in the event loop; and the view is where the
lock that guards that buffer can be taken.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Several methods assigned v.ox and v.oy directly: SetOrigin, CopyContent,
the wrap/autoscroll branches in draw, FocusPoint, and
Scroll{Up,Down,Left,Right}. Funnelling them all through SetOriginX and
SetOriginY gives a single place to observe (or set a breakpoint on)
every change to a view's scroll position, which makes debugging scroll
behaviour much easier.
This means those call sites now also get the setters' `< 0` clamps, but
that is behaviour-preserving in every case: each assigned value is
already >= 0. calculateNewOrigin never returns a negative number;
CopyContent copies origins that are themselves always >= 0; and the draw
and scroll writes are all guarded (or fed only non-negative amounts) so
the result can't go below zero.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
When using lazygit in a multi-tab terminal it is useful to see if the
lazygit tab is currently active; ghostty does a very good job at dimming
down the inactive tabs, but VS Code's builtin terminal does not, so
indicate this on our side by removing the green highlight from panel
frames and tab titles, and showing the selection as inactive like we do
for a side panel when the main view is focused.
moveMainContextToTop copies the current top view's content into the view
it's promoting, to avoid a flicker. The source can be a main view with a
live streaming task (e.g. resolving a conflict promotes the merge-conflicts
view over a main view that's mid-diff), and CopyContent both read and
published that source's buffer unsafely:
- it read the source's lines/viewLines while locking only the
destination, racing the task's concurrent Write; and
- it aliased the source's row slices into the destination, so the
source's ongoing appends (growslice reading the shared array) and
refreshViewLinesIfNeeded's in-place wrapping-cache writes (&lines[i])
kept racing this view's rendering after the copy.
Lock the source for the read, and shallow-clone the row slices so the
destination gets its own arrays. The per-row cell data is immutable once
written, so it stays shared -- the clone cost is proportional to the
number of rows, not their contents.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A view's line buffer, its viewLines/tainted flags, and its hover state
are all written from the command-task goroutine (under writeMutex) as it
renders. But three accessors reached that same state from the UI thread
without the lock: SetView and the GUI-resize path cleared a view's lines
directly, viewsToRedrawContentOnly read the tainted flag, and Buffer read
the line buffer. Each raced a rendering task.
Guard them with writeMutex, matching the view's other buffer accessors.
These are reads/clears of state writeMutex already protects, not new
callers of it -- the view's geometry stays outside the mutex.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A command task streams its output into a view from its own goroutine. To
track soft-wraps (so cursor-positioning escapes from a pager land on the
right line) the write path read the view's live InnerWidth, and the pty
setup read its InnerSize -- both off the UI thread, racing the UI thread
mutating the view's dimensions during layout.
Capture the width on the UI thread instead and hand it to the task: the
escape interpreter keeps a screenColMax it reads from, seeded in NewView
and refreshed per render via View.SetContentWidth (called from
newCmdTask/newPtyTask before the task's goroutine starts), and the pty
size is computed in the after-layout callback rather than in the task's
start func. The view's dimensions stay UI-thread-only; the task uses the
snapshot rather than reading them live.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
refreshViewLinesIfNeeded re-wrapped every line of the buffer whenever
the view was tainted. That's cheap for short content, but scrolling a
long diff calls it constantly: adjustDownwardScrollAmount queries
ViewLinesHeight on every scroll event, and each newly-read line taints
the view, so every notch re-wrapped the entire buffer. Wrapping measures
each cell's width (uniseg) and allocates per line, so once you'd scrolled
far enough down the diff, scrolling turned sluggish - the cost grew with
how much had been read. (A CPU profile of scrolling deep in a long diff
put 77% of the time in lineWrap, reached almost entirely via
ViewLinesHeight rather than draw.)
Cache each line's wrapped result on the lineType, keyed by the width it
was wrapped at, and only re-wrap lines that have actually changed since
the last refresh. A firstDirtyLine index, updated in the same three
places that set `tainted` (write, clearViewLines' callers, SetHighlight),
marks the lowest line that might have changed; lines below it with a
matching cached width reuse their cached wrapping. The cache lives on the
line, so it's freed with the line when the view's content is replaced
(e.g. selecting a different commit) - it doesn't accumulate across a
session.
The wrapping cost per scroll now scales with the number of lines just
read, not with the total size of the buffer, so scrolling stays smooth
no matter how far down you are.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
ConPTY compresses runs of default-colored spaces into ECH + CUF
(\x1b[NX\x1b[NC) instead of emitting them literally. ECH is still a
no-op for us — our buffer is built sequentially and has nothing to
erase — but CUF has to materialize as N visible space cells so the
gap actually appears, otherwise content the child wrote with leading
indentation slides left against the preceding cell.
The view's cursorForward branch reuses the same machinery as tab
expansion: substitute the trigger byte for a space and let the
repeatCount path emit the cells under the parser-tracked SGR. The
existing notifyCellsWritten plumbing then advances screenCol over
the gap, keeping subsequent CUP targets aligned.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
ConPTY presents its child's output as a screen buffer and uses CUP /
CUD / CNL / VPA to skip over blank rows rather than emitting LFs. The
previous behaviour swallowed all of those and the visible content
collapsed together. Now the escape parser tracks the screen-relative
cursor row, and any CSI that moves the cursor past the current row
emits a cursorDown instruction that the view turns into the matching
number of empty lines.
Column tracking is deliberately omitted: doing it correctly would mean
duplicating the view's grapheme-cluster width math in the parser, and
ConPTY in practice positions to column 1 after a CR-equivalent, which
the existing wx-reset path already handles. ConPTY-internal scrolling
needs no special handling either: it only emits cursor-positioning
escapes within the first, un-scrolled screenful — once its screen
scrolls it switches to plain linefeeds, which the view advances on
directly regardless of the tracked cursor.
Backward cursor moves are silently dropped — the view's buffer is
append-style and can't undo earlier writes. The exception is cursor-home
(CUP to row 1): ConPTY emits it at the start of every screen, so rather
than drop it we re-anchor the row tracking to the current write position.
Without that, a view not rewound in lockstep with ConPTY's screen (the
command log, which streams pty output without a rewind) accumulates
drift, and every later absolute CUP becomes a dropped backward move that
collapses the rows ConPTY positioned with.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
