When we stage, discard or edit lines of the working tree's diff, we
read the file's diff again and look for the selected lines in it. If
none of them are there, nothing happens. The same goes for adding
lines of a commit's diff to a custom patch, or taking lines out of
it. This can happen if the file changed since its diff was shown, or
if we read the line numbers out of the diff on screen wrongly.
Show an error that names the file instead. As with a file that can't
be found, don't act on any file if the lines of one are missing.
Unlike the file check, this one happens after a new custom patch has
been started. If the error leaves that patch without any lines, reset
it, as we do when the last line is taken out of a patch.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
If the path in the diff header of a selected line doesn't name a file
of the files panel, staging or discarding the line does nothing. The
command log shows "Apply patch", but nothing else happens, so the user
has nothing to go on when reporting it. Adding a line of a commit's
diff to a custom patch fails in the same silent way if the path
doesn't name a file of that diff. A diff renderer that names files
differently can cause this, and so can a bug in how we read the path,
like the one we had with diff.mnemonicPrefix.
Show an error that names the path instead, and do nothing else. If a
selection spans several files, don't act on the files we did find,
because that would do only part of what the user asked for. Do the
same for opening a line of a diff in its pull request.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
A diff line carries the absolute path of its file, and both panels
acting on such a line turn it into the repo-relative one git speaks
themselves. Opening a line in a pull request needs that path too, to
name the file to GitHub by it.
Extract repoRelativePath, and have both panels use it. The files panel
gains the check for a path outside the repo that the other one had; a
path it used to pass on matches no file of the working tree either.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
A commit that moves an entire package elsewhere renames hundreds of
files, and passing every one of their old paths can push the command past
the length limit the OS imposes (~32k characters on Windows). Their
common parent directory does just as well whenever everything it holds
ends up in the diff anyway.
Deciding that needs to consider every file of the diff, not only those on
display, so the paths are now derived from the model rather than from the
tree; a status filter must not make a directory look emptier than it is.
Git limits its tree diff by the pathspec before it looks for renames, so
a directory only ever gets one end of a rename whose other end is outside
it. Nothing is left to pair up, and the file turns into an addition or a
deletion that the commit doesn't contain.
Pass the other end along with the directory. This is bounded by the
number of renames that cross the directory's boundary, so it costs
nothing at all for the vast majority of commits.
The files and commit files panels each had their own copy of this, one of
which used to be missing the previous path of a rename. Growing them
apart again is the last thing we want, since the next commit needs to
teach both of them about renames that cross a directory boundary.
The files panel version only returned paths for the filtered case, and
left it to WorktreeFileDiffCmdObj to derive the rest from the node; now
that all callers pass the paths in, that command doesn't need to know
about renames at all.