The idea is that InternalRevset will store a 'static boilerplate function that
borrows an 'index passed by function argument. This way, we can abstract the
index type over Arc<T> and &T without introducing too much generics.
I don't know why the dependabot didn't catch this, but there are things to
fix manually. EntryMode was changed to a u16 wrapper, and the enum was renamed
to EntryKind. Other than that, I don't find anything breaking our codebase.
Perhaps, this is the most controversial part. It could be moved to new
"segment" module (or something like "common"), but I think IndexSegment can be
considered a trait that enables the CompositeIndex abstraction.
Added pub(super) or pub where needed. I won't implement accessor methods on
IndexPositionByGeneration and IndexPosition as they are purely value types,
and protecting the inner values wouldn't make sense.
It seems better to have the caller pass the transaction description
when we finish the transaction than when we start it. That way we have
all the information we want to include more readily available.
Previously, the function relied on both the `self.parent_mapping` and
`self.rebased`. If `(A,B)` was in `parent_mapping` and `(B,C)` was in `rebased`,
`new_parents` would map `A` to `C`.
Now, `self.rebased` is ignored by `new_parents`. In the same situation,
DescendantRebaser is changed so that both `(A,B)` and `(B,C)` are in
`parent_mapping` before. `new_parents` now applies `parent_mapping` repeatedly,
and will map `A` to `C` in this situation.
## Cons
- The semantics are changed; `new_parents` now panics if `self.parent_mapping`
contain cycles. AFAICT, such cycles never happen in `jj` anyway, except for
one test that I had to fix. I think it's a sensible restriction to live with;
if you do want to swap children of two commits, you can call
`rebase_descendants` twice.
## Pros
- I find the new logic much easier to reason about. I plan to extract it into a
function, to be used in refactors for `jj rebase -r` and `jj new --after`. It
will make it much easier to have a correct implementation of `jj rebase -r
--after`, even when rebasing onto a descendant.
- The de-duplication is no longer O(n^2). I tried to keep the common case fast.
## Alternatives
- We could make `jj rebase` and `jj new` use a separate function with the
algorithm shown here, without changing DescendantRebaser. I believe that the new
algorithm makes DescendatRebaser easier to understand, though, and it feels more
elegant to reduce code duplication.
- The de-duplication optimization here is independent of other changes, and
could be used on its own.
into_segment() could be added instead of save_in(), but I decided to wrap
save_in(). save_in() may squash ancestor files, so it could be considered an
index-level operation.
default_index_store.rs is relatively big, and it contains types and impls in
arbitrary order. Let's split them into sub modules. After everything moved,
mod.rs will only contain tests.
I'm going to remove 'index lifetime from InternalRevset so Revset<'static>
can be easily constructed from DefaultReadonlyIndex. As the first step, this
series removes some lifetime complexity from EvaluationContext methods.
We don't need an descendant iterator API, but it helps to add separate function
to collect into HashSet<IndexPosition> instead of returning a pair of
ordered vec and set.
The wrapper type isn't needed for the mutable layer, but this mirrors the
readonly type structure. Test cases are also migrated to be using the index
wrapper so long as we don't have to care for the nesting of the segment files.
The idea is that the ReadonlyIndexSegment is a sub component of the index. The
Index trait could be implemented for any Segment type, but we don't need a
public interface to access sub segment as an index.
I'm going to split the internal Segment types and the public Index types
in order to clarify the layering concept. The public Index types will be
wrappers like DefaultReadonlyIndex.
Strictly speaking, ReadonlyIndexImpl is a segment + parent pointer pair,
but I think calling it a segment is pretty okay. It could be called a
ReadonlyIndexFile, but "File" can't apply to the mutable part.
Gitoxide errors don't implement PartialEq. We could instead stringify the
errors, but there aren't many callers who expect FailedRefExportReason to
be comparable.
Before, an absolute path would be saved in the git_target file if .git is a
symlink. That's not wrong, but seemed a bit weird. Let's consolidate the
behavior across .git file types.
Apparently, libgit2 doesn't deduce "core.bare" config from the directory name,
but gitoxide implements it correctly. So we shouldn't blindly canonicalize
the Git repository path. Fortunately, the saved git_target path isn't a fully-
canonicalized form (unless user explicitly sepcified "--git-repo ./.git"), so
we don't need a hack to remap git_target back to the symlink path.
is_colocated_git_workspace() is adjusted since the git_workdir is no longer
resolved from the fully-canonicalized repo path, at least in our code. Still we
have the ".git/.." fallback because test_init_git_colocated_symlink_gitlink()
would otherwise fail. I haven't figured out why, and the test might be actually
wrong compared to the git CLI behavior, but let's not change that for now.
Fixes#2668
This adds an initial `jj util gc` command, which simply calls `git gc`
when using the Git backend. That should already be useful in
non-colocated repos because it's not obvious how to GC (repack) such
repos. In my own jj repo, it shrunk `.jj/repo/store/` from 2.4 GiB to
780 MiB, and `jj log --ignore-working-copy` was sped up from 157 ms to
86 ms.
I haven't added any tests because the functionality depends on having
`git` binary on the PATH, which we don't yet depend on anywhere
else. I think we'll still be able to test much of the future parts of
garbage collection without a `git` binary because the interesting
parts are about manipulating the Git repo before calling `git gc` on
it.
I think this is a variant of the problem fixed by 7fda80fc22 "tree: simplify
conflict before resolving at hunk level." We need to simplify() the conflict
before and after extracting file ids because the source conflict values may
contain trees to be cancelled out, and the file values may differ only in exec
bits. Since the legacy tree passes a simplified conflict in to this function,
I made the merged tree do the same.
Fixes#2654
Otherwise an empty subtree would be added to the parent tree.
If the stored tree contained an empty subtree, simplify() wouldn't work
against new "absent" subtree representation. I don't know if there's a
such code path, but I believe it's very rare to encounter the problem.
#2654
