Staging a hunk from the focused main view advances the selection to the next
hunk, because staging removes the acted-on lines from the diff so the preserved
change-line ordinal lands on the next change. Toggling a hunk into a custom
patch left the selection sitting on the just-toggled hunk instead — the toggle
doesn't change the diff (only the inclusion set), so the same ordinal lands back
where it was, and you had to navigate by hand to build a patch hunk by hunk.
Give the toggle the same feel by advancing the reveal ordinal past the toggled
change lines (RevealSelectionAfterStaging gains an advanceBy arg, fed the toggled
change-line count). Staging and removal still pass 0 (their lines are consumed).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The secondary pane showed the custom patch with a bespoke in-memory render
(PatchBuilder.RenderAggregatedPatch / FormatView). That had two problems: it
could never be fed to a pager the way the main view's diff is (a stdin pager
might have worked, but an external diff tool like difftastic, which diffs two
files rather than a unified diff, could not), and its hand-rolled hunk/context
handling differed subtly from git's.
Materialize the patch instead as two real file trees under a temp dir — a/ holds
each patched file's "from"-side content, b/ that content with the patch applied —
and render it with `git diff --no-index`, reusing the exact pager wiring the main
view uses (stdin pager, external diff, or git's own colour as the raw fallback).
`--no-index` honors both GIT_PAGER and --ext-diff, so every pager type now renders
the custom patch like any other diff, and git computes the context, fixing the
quirks. Because the secondary is now an async diff task, the post-removal
selection reveal (which rides a task's restore) finally takes effect.
The trees are named a/b so that with --no-prefix the diff shows the real
repo-relative paths; added files are seeded empty in a/ so they pair up and show
their real paths rather than git's directory-comparison "added in b" form. The
patch builder owns the temp dir's lifetime (created on Start, removed on Reset)
and bumps a generation counter on every change, so the trees are rebuilt only
when the patch actually changes — covering the focused-main view and the old
explorer alike, without rebuilding on mere navigation.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The focused-main-view rework made the secondary pane actionable, but the old
patch-building explorer's secondary was inert, so its actions were never thought
through. Pressing space there routed through the same toggle handler as the main
pane, resolving the selection against the secondary's diff and mapping it to
patch-builder indices by line number. But the secondary shows the *aggregated*
custom patch, which renumbers included additions whenever an earlier addition in
the same hunk is excluded (Transform recomputes each hunk's +start). So the
shifted number resolved to the wrong line in the original diff — often adding an
unrelated line instead of removing the selected one.
Resolve the secondary selection by its *ordinal* among the change lines shown
instead: the custom-patch view renders exactly the included change lines in
order, so the k-th change line of a file is that file's k-th included change line
(PatchBuilder.IncludedChangeLineIndices), independent of the renumbering. Space
in the secondary now only ever removes, mirroring how space in the staging view's
staged pane unstages.
Discarding from the commit (the remove key) makes no sense in the custom-patch
preview — it would act on lines shown only as the patch, and space already
removes them — so it's disabled there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Dropping a hunk with `d` re-establishes the focused main view's selection
on the next surviving change, which feels great. But other operations that
rewrite the commit under the focused main view — moving a custom patch out
into the index, undoing right after a discard or a patch move — don't run
through the focused-main-view action handlers, so nothing was preserving
the selection. The stale gocui selection was left painted over the new
content, often as a large, now-meaningless range.
Rather than teach every such command to capture and restore the selection
(move-patch, undo, redo, and any future one), the focused main view now
preserves it itself, by its change-line ordinal, as the diff re-renders —
the command-agnostic counterpart of revealSelectionAfterPrimaryAction. The
diff side panels call it from their render-to-main before triggering the
render, so the restore rides the re-render.
It stands down unless the focused main view is current and shows a
selection, no precise restore is already pending (escape / post-stage
reveal / context-size place the selection more precisely), and the diff
command is actually changing. That last gate matters: a plain background
refresh re-renders the same commit's diff unchanged, and there the
selection — range and all — must be left alone; only a command change
(e.g. a rebase rewriting the commit's hash) means the content moved.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Wire the remaining diff panels — local commits, sub-commits, commit files,
stash, and reflog — into the focused main view's raw-diff fallback, the same way
the files panel already is: each computes DiffMainViewShouldRenderRaw and renders
its diff raw (no pager) when focused under a pager whose output we can't resolve,
so its selection stays toggleable into a custom patch (or, for reflog, navigable
and copyable).
The commit-diff panels build their diff through the diff helper, which can't
reach the staging helper, so the panel computes renderRaw and passes it in.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Wire the focused main view's `d` binding to a DiscardSelection action on the
side panel beneath, the discard counterpart of the primary action. As with
stage/toggle, what discard means is the panel's business:
- the files panel discards from the working tree, mirroring the staging
view's `d` — reverse-apply not cached on the unstaged side (destructive, so
confirmed), reverse-apply cached on the staged side (i.e. unstage);
- the commit panels (commit files, and the whole-commit diff of the local
commits / sub-commits / stash panels) remove the selected lines from the
commit via a rebase, mirroring the patch builder's "discard lines from
commit" — building a one-off patch from the selection and deleting it.
The commit-discard backend and its guard are shared free functions in
patch_building_from_main_view.go, paralleling the patch toggle. Discarding
from a commit is only possible on a local branch, so DiscardSelection grows a
disabled-reason companion on the FocusedMainViewActions interface: the `d`
binding is greyed with a reason on stash and other-branch sub-commits (never
rebaseable) and while a rebase is in progress, exactly as the patch builder
greys it. The files panel is always available (a zero-context diff is the one
error it reports inline, matching the staging view).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Until now space toggled lines into a custom patch only from the commit files
main view (the per-file diff). Register the same toggle handler on the commits,
sub-commits and stash panels, so a patch can be built straight from the
whole-commit (multi-file) diff their main view shows, without first diving into
the commit files panel.
The handler lives on SwitchToDiffFilesController, which is already bound to
exactly those three panels and already knows how to derive the patch target
(from/to/reverse/canRebase) for the selected ref — pulled out of enter() into a
shared canRebase helper and the reused FromAndToForDiff. The post-toggle refresh
is the cheap one: these panels have no per-file patch indicator to update, so we
just re-render their own main + secondary views rather than reloading the whole
commit list on every keystroke.
Sub-commits and stash didn't render the secondary patch view at all; give them
the same secondaryPatchPanelUpdateOpts the commits panel uses, and recompute the
inclusion gutter as their diff (re-)renders, so the cumulative patch and the
gutter track the toggle.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Pressing space in the commit-files focused main view now toggles the
selected line(s)/hunk into or out of the custom patch, instead of being a
no-op. It's the patch-building counterpart of working-tree staging from
the main view: the selection is resolved to change-line identities the
same way (ChangeLinesInViewRange), then mapped to the patch builder's
per-file line indices and added or removed.
The crux is that the commit's diff is unchanged by a toggle — only the
inclusion set changes — so unlike staging there's no async re-render to
ride. Membership is shown by an on-demand inclusion gutter: a reserved
left column painted with a marker on every change line currently in the
patch, recomputed synchronously after each toggle and on focusing the
main view, over the existing pager output. This is what lets patch
building work over a restructuring pager at all, where the old green
first-char overlay (which rewrites the rendered bytes) can't.
space routes to staging or patch building based on which handler the
panel beneath registers, via a new onTogglePatchFocusedMainView channel
(kept separate from staging because the post-action differs: sync gutter
repaint vs async reveal). Scoped to the commit-files panel for now;
commits/sub-commits/stash and the whole-commit multi-file diff follow.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Some tests assert that a specific commit subject does or doesn't occur
in the main view; interactive_rebase/outside_rebase_range_select.go is
an example for this, it asserts `t.Views().Main().Content(
DoesNotContain("commit 06"))`. The problem with this kind of assertion
and our test commit naming scheme is that the diff view begins with a
"commit <hash>" line, and when that hash happens to start with "06" the
assertion matched it and failed spuriously. This was usually masked by
our MaxAttempts=2 that we currently use for integration tests (it's
quite unlikely that the commit gets a hash beginning with "06" twice in
a row). However, we want to get to a state where we can set MaxAttempts
to 1, so make this more robust by changing our naming scheme.
When loading the files of a commit we passed --no-renames, so a rename
showed up as a separate delete and add rather than a single R entry.
That made it impossible to work with a rename that also modifies the
file: the modifications were spread across a full deletion and a full
addition instead of appearing as the handful of lines that actually
changed. The staging view already shows renames and lets you stage
their hunks, so there was no good reason for the patch builder to
differ; the flag was only there because the commit-file parser couldn't
cope with the rename record format.
Switch the commit-file loader and the per-file diff to --find-renames,
teach the parser about the rename record (a status followed by two
paths), and carry the previous path through the patch builder so the
diff for a rename is loaded with both paths, which is what makes git
emit the rename in the first place.
A whole-file selection keeps the rename in the header, so the rename
moves or is discarded together with the file's contents. A partial
selection instead strips the rename metadata and points the header at
the new path, so applying the patch only changes the contents and
leaves the rename in place; the blob index line is kept so that a 3-way
apply can still fall back to a blob merge.
Discarding a renamed file from a commit now discards both the new and
the old path, so the new file is removed and the old one is restored.
Changing the rename similarity threshold refreshes the commit files
panel too, not just the files panel, so that a rename can turn into a
delete and add or back. It is disabled while building a patch, however,
because the patch builder caches each file's diff by path and would
desync if a rename changed into a delete and add underneath it.
Finally, copying a file's diff from the commit files panel now passes
both paths for a rename, so the copied diff shows the rename instead of
a new-file add.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Change working tree files and commit files panels to use filtering
(reducing the list) instead of search (highlighting matches). This
matches the behavior of other filterable views.
The text filter matches against the full file path, not just the
filename, which is more useful for navigating large directory trees.
When toggling a directory for a custom patch while a text filter is
active, only the visible filtered files in the directory are affected,
consistent with how staging a directory in the files panel works.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
I always press 'd' in the patch building view, expecting that I can do
exactly what I can do in the staging view, to find out I need to go
space -> ctrl+p -> d and I think it's time to honour the muscle memory
and support this convenience keybinding.
While it's true that the behavior is a little different from the staging panel,
where the staged lines are actually removed from the view and in many cases the
selection stays more or less in the same place, it is still very useful to move
to the next stageable thing in the custom patch building view too.
Unlike moving a patch to the index, applying or reverting a patch didn't
auto-stash, which means that applying a patch when there's a modified (but
unstaged) file in the working tree would error out with the message "error:
file1: does not match index", regardless of whether those modifications conflict
with the patch or not.
To fix this, we *could* add auto-stashing like we do for the "move patch to
index" command. However, in this case we rather simply stage the affected files
(after asking for confirmation). This has a few advantages:
- it only changes the staging state of those files that are contained in the
patch (whereas auto-stashing always changes all files to unstaged)
- it doesn't unnecessarily show a confirmation if none of the modified files are
affected by the patch
- if the patch conflicts with the modified files, the conflicts were "backwards"
("ours" was the patch, "theirs" the modified file); it is more logical if "ours"
is the current state of the file, and "theirs" is the patch.
It's a little unfortunate that the behavior isn't exactly the same as for "move
patch to index", but for that one we do need the auto-stash because of the
rebase that runs behind the scenes.
The tests show that this currently fails with the confusing error message "does
not match index", regardless of whether the patch conflicts with the
modifications or not. We'll improve this in the next commit.
I don't bother adding tests for reverting a patch, as the code is basically the
same as for apply.
This is functionality that works already, we only add the test for more complete
test coverage. However, there's a detail problem, and the test demonstrates
this: we keep the stash even if there was no conflict. We'll fix this next.
I almost broke this during the development of this branch, so add a test to
guard against that. The point here is that the stack remains intact, i.e. the
newly created commit is the last commit of the lower branch, and thus shows the
"*".
Previously we would render the diff for a directory to the main/secondary pair,
but a diff for a file to the staging/stagingSecondary pair. (And similar for
commit files: main/secondary for directories, but
patchBuilding/patchBuildingSecondary for files.)
I always found this confusing and couldn't really understand why we are doing
this; but now it gets in my way because I want to attach a controller to
main/secondary so that they can be focused. So change it to always use the main
context pair for everything we render from a side panel.
When you are in the middle of a rebase, and you cherry-pick a commit which
conflicts, it helps to be clear on whether you are prompted to continue the
cherry-pick or the rebase.
Assert the entire lines using Equals instead of Contains. This makes the tests a
bit easier to read, and it makes it much easier to decide how they need to be
changed when we change the layout (like we do in the last commit of this
branch).
It is true that this requires changing all these tests for any future UI
changes, but I think this is a good price to pay; those adaptions are trivial
and can be done without thinking.
There are two ways to jump to the editor on a specific line: pressing `e` in the
staging or patch building panels, or clicking on a hyperlink in a delta diff. In
both cases, this works perfectly in the unstaged changes view, but in other
views (either staged changes, or an older commit) it can often jump to the wrong
line; this happens when there are further changes to the file being viewed in
later commits or in unstaged changes.
This commit fixes this so that you end up on the right line in these cases.
When pressing `e` on line 5 in a diff of an older commit, we expect it to take
us to line 5 in that file. But we end up on line 2, because the file had further
changes both in newer commits, and in the unstaged changes of the working copy.
This updates gocui to include https://github.com/jesseduffield/gocui/pull/68 and
https://github.com/jesseduffield/gocui/pull/69, which changes views to not have
an extra blank line at the end when content ending in a newline character is
written to them. This makes text views more consistent with list views, which
don't have a blank line after the last list entry either.
Several custom patch commands on parts of an added file would fail with the
confusing error message "error: new file XXX depends on old contents". These
were dropping the custom patch from the original commit, moving the patch to a
new commit, moving it to a later commit, or moving it to the index.
We fix this by converting the patch header from an added file to a diff against
an empty file. We do this not just for the purpose of applying the patch, but
also for rendering it and copying it to the clip board. I'm not sure it matters
much in these cases, but it does feel more correct for a filtered patch to be
presented this way.
This currently works (albeit with a bit of manual work, as the user needs to
resolve conflicts), and we add this test just to make sure that we don't break
it with the following change.
This currently works, we add it as a regression test to make sure we don't break
it. It is an interesting test because it turns the deletion of the file in the
moved-from commit into a modification.
As part of this, you must now press enter on a merge conflict file
to focus the merge view; you can no longer press space and if you do
it will raise an error.
This is the highest priority of the escape actions because it's the thing you're
most likely to want to do upon hitting escape if you have a range selected.
Applying this to the staging/patch-building views is tricky: if we want this logic
for when a range of lines is selected, we'll also need to apply it when a hunk
is selected too. I still think it's worth it though: I've often accidentally
escaped from the staging view when trying to cancel a range selection.
This adds range select ability in two ways:
1) Sticky: like what we already have with the staging view i.e. press v then use arrow keys
2) Non-sticky: where you just use shift+up/down to expand the range
The state machine works like this:
(no range, press 'v') -> sticky range
(no range, press arrow) -> no range
(no range, press shift+arrow) -> nonsticky range
(sticky range, press 'v') -> no range
(sticky range, press arrow) -> sticky range
(sticky range, press shift+arrow) -> nonsticky range
(nonsticky range, press 'v') -> no range
(nonsticky range, press arrow) -> no range
(nonsticky range, press shift+arrow) -> nonsticky range
The test apply_in_reverse_with_conflict.go fails in git versions 2.30.8 and
earlier. Apparently the output "Applied patch to 'file2' cleanly" was only added
more recently. It's not essential that we check this output.
We have not been good at consistent casing so far. Now we use 'Sentence case' everywhere. EVERYWHERE.
Also Removing 'Lc' prefix from i18n field names: the 'Lc' stood for lowercase but now that everything
is in 'Sentence case' there's no need for the distinction.
I've got a couple lower case things I've kept: namely, things that show up in parentheses.
By constructing an arg vector manually, we no longer need to quote arguments
Mandate that args must be passed when building a command
Now you need to provide an args array when building a command.
There are a handful of places where we need to deal with a string,
such as with user-defined custom commands, and for those we now require
that at the callsite they use str.ToArgv to do that. I don't want
to provide a method out of the box for it because I want to discourage its
use.
For some reason we were invoking a command through a shell when amending a
commit, and I don't believe we needed to do that as there was nothing user-
supplied about the command. So I've switched to using a regular command out-
side the shell there
