3394 commits

Author	SHA1	Message	Date
Scott Taylor	6e959fa12c	conflicts: allow stripped trailing whitespace in diffs ... Some editors strip trailing whitespace on save, which breaks any diffs which have context lines, since the parsing function expects them to start with a space. There's no visual difference between " \n" and "\n", so it seems reasonable to accept both.	2024-11-22 18:00:05 -06:00
Scott Taylor	efacbcbd45	conflicts: demo failed parse of diff with empty line	2024-11-22 18:00:05 -06:00
Scott Taylor	9674852dc7	conflicts: allow CRLF line endings on conflict markers ... Currently, conflict markers ending in CRLF line endings aren't allowed. I don't see any reason why we should reject them, since some editors/tools might produce CRLF automatically on Windows when saving files, which would break the conflicts otherwise.	2024-11-22 18:00:05 -06:00
Scott Taylor	ee7f829d4c	conflicts: demo failed parse of markers with CRLF	2024-11-22 18:00:05 -06:00
Yuya Nishihara	c6bb019d41	diff: optimize allocation of histogram entries for unique words ... ``` group new old ----- --- --- bench_diff_git_git_read_tree_c 1.00 34.5±0.26µs 1.32 45.7±0.11µs bench_diff_lines/modified/10k 1.00 28.2±0.10ms 1.19 33.5±0.69ms bench_diff_lines/modified/1k 1.00 2.6±0.01ms 1.15 3.0±0.01ms bench_diff_lines/reversed/10k 1.00 21.5±0.22ms 1.08 23.3±0.18ms bench_diff_lines/reversed/1k 1.00 364.8±11.96µs 1.22 445.1±8.99µs bench_diff_lines/unchanged/10k 1.00 1761.3±13.85µs 1.66 2.9±0.07ms bench_diff_lines/unchanged/1k 1.00 163.6±1.25µs 1.47 240.7±2.72µs ``` ``` % hyperfine --sort command --warmup 3 --runs 5 -L bin jj-0,jj-1 \ 'target/release-with-debug/{bin} --ignore-working-copy \ file annotate lib/src/revset.rs' Benchmark 1: target/release-with-debug/jj-0 .. Time (mean ± σ): 1.144 s ± 0.011 s [User: 1.088 s, System: 0.053 s] Range (min … max): 1.131 s … 1.159 s 5 runs Benchmark 2: target/release-with-debug/jj-1 .. Time (mean ± σ): 1.026 s ± 0.008 s [User: 0.975 s, System: 0.048 s] Range (min … max): 1.015 s … 1.035 s 5 runs ```	2024-11-22 08:20:55 +09:00
Yuya Nishihara	656a614f29	signing: pass &UserSettings in to backends ... If we add our own config storage implementation, .get_<type>() functions will be moved to UserSettings object.	2024-11-22 08:20:45 +09:00
Yuya Nishihara	7a7962bb9f	signing: propagate config errors	2024-11-22 08:20:45 +09:00
Yuya Nishihara	5e13a4a719	signing: remove unused SignInitError::Backend variant, simplify error handling ... Since signing backends shouldn't do non-trivial work at the initialization stage, the error type would never be categorized as an internal error.	2024-11-22 08:20:45 +09:00
Yuya Nishihara	7906b3f4a5	annotate: remove unneeded BString<->Vec round-trip	2024-11-21 10:50:37 +09:00
Yuya Nishihara	59a79fdcc0	conflicts: extract materialize_merge_result_to_bytes() helper ... We have many callers of materialize_merge_result() who just want in-memory buffer.	2024-11-21 10:50:37 +09:00
Yuya Nishihara	4cdead34b4	conflicts: make materialize_merge_result() accept reference type ... Because the output of diff.hunks() is a list of [&BStr]s, a Merge object reconstructed from a diff hunk will be Merge<&BStr>. I'm not pretty sure if we'll implement conflict diffs in that way, but this change should be harmless anyway.	2024-11-21 10:50:37 +09:00
Yuya Nishihara	2aa913b035	conflicts: extract inner block of materialize_merge_result() ... I'm going to make materialize_merge_result() accept reference type. This patch extracts large non-generic part to a separate function.	2024-11-21 10:50:37 +09:00
Yuya Nishihara	5cc0bd0950	rewrite: fix duplicated commits to be rebased onto destination ... I believe this was an oversight. "jj duplicate" should duplicate commits (= patches), not trees. This patch adds a separate test file because test_rewrite.rs is pretty big, and we'll probably want to migrate CLI tests to jj-lib.	2024-11-21 10:49:51 +09:00
Yuya Nishihara	1973c712a3	log: emit working-copy branch first if included in the revset ... The working-copy revision is usually the latest commit, but it's not always true. This patch ensures that the wc branch is emitted first so the graph node order is less dependent on rewrites.	2024-11-20 10:50:16 +09:00
Yuya Nishihara	fb79f2024d	tests: do Result wrapping/unwrapping by graph helper	2024-11-20 10:50:16 +09:00
Yuya Nishihara	05c7da3db0	diff: reuse precomputed hash values ... This isn't always fast because it increases the chance of cache miss, but in practice, it makes "jj file annotate" faster. It's still slower than "git blame", though. Maybe we should also change the hash function. ``` group new old ----- --- --- bench_diff_git_git_read_tree_c 1.00 45.2±0.38µs 1.29 58.4±0.32µs bench_diff_lines/modified/10k 1.00 32.7±0.24ms 1.05 34.4±0.17ms bench_diff_lines/modified/1k 1.00 2.9±0.00ms 1.06 3.1±0.01ms bench_diff_lines/reversed/10k 1.00 22.7±0.18ms 1.02 23.2±0.29ms bench_diff_lines/reversed/1k 1.00 439.0±9.46µs 1.19 523.9±6.05µs bench_diff_lines/unchanged/10k 1.00 2.9±0.06ms 1.20 3.5±0.02ms bench_diff_lines/unchanged/1k 1.00 240.8±1.03µs 1.30 312.1±1.05µs ``` ``` % hyperfine --sort command --warmup 3 --runs 10 -L bin jj-0,jj-1 \ 'target/release-with-debug/{bin} --ignore-working-copy file annotate lib/src/revset.rs' Benchmark 1: target/release-with-debug/jj-0 .. Time (mean ± σ): 1.604 s ± 0.259 s [User: 1.543 s, System: 0.057 s] Range (min … max): 1.348 s … 1.917 s 10 runs Benchmark 2: target/release-with-debug/jj-1 .. Time (mean ± σ): 1.183 s ± 0.026 s [User: 1.118 s, System: 0.062 s] Range (min … max): 1.155 s … 1.237 s 10 runs ```	2024-11-20 10:36:08 +09:00
Yuya Nishihara	2c653564fd	diff: extract narrowed view type from DiffSource, add type-safe local index ... I'm going to add a Vec of precomputed hashes, and the Vec will be owned by DiffSource.	2024-11-20 10:36:08 +09:00
Yuya Nishihara	ed18b5bdc5	diff: cache hash values by Histogram ... Since patience diff is recursive, it makes some sense to reuse precomputed hash values. This patch migrates Histogram to remembering hashed values. The precomputed values will be cached globally by DiffSource. Technically, Histogram doesn't have to keep a separate copy of hash values, but this appears to give better perf than slicing text and hash value from two Vecs.	2024-11-20 10:36:08 +09:00
Luke Randall	068fa0f37e	revset: allow tags() to take a pattern for an argument ... This makes it more consistent with `bookmarks()`. Co-authored-by: Austin Seipp <aseipp@pobox.com>	2024-11-20 00:47:23 +00:00
Ilya Grigoriev	0c6c47101f	nightly clippy fixes	2024-11-18 18:49:05 -08:00
Benjamin Tan	4db4f413a7	revset: add fork_point function ... This can be used to find the fork point (best common ancestors) of a revset with an arbitrary number of commits, which cannot be expressed currently in the revset language.	2024-11-16 04:08:01 +08:00
Benjamin Tan	5138836cdc	rewrite: add duplicate_commits{,onto_parents} functions	2024-11-15 19:42:13 +08:00
Benjamin Tan	0a7f6bd8aa	rewrite: extract compute_commits_heads function ... This will be shared between `move_commits` and the new `duplicate_commits` function to be added.	2024-11-15 19:42:13 +08:00
dploch	0a5bc2bbed	workspace: move recovery commit logic into lib for sharing ... This is to facilitate automatic update-stale in extensions and in the CommandHelper layer.	2024-11-14 11:12:02 -05:00
dploch	afe25464fe	working_copy: move freshness calculation into lib for sharing ... This is to facilitate automatic update-stale in extensions and in the CommandHelper layer.	2024-11-14 11:12:02 -05:00
Yuya Nishihara	bfb7613d5d	cargo: bump gix to 0.67.0 (includes minor behavior change) ... - gix::object::tree::diff::change::Event::Rewrite is flattened - diff options are extracted to separate type 2b81e6c8bd - signature text now includes a trailing newline 4a6bbb1b79 The last change means that our SecureSig { sig } returned by GitBackend is now terminated by '\n'. I think this is harmless since textual signature is usually ends with '\n'.	2024-11-14 22:38:04 +09:00
Yuya Nishihara	eed32954b2	cargo: import fix_filter from gix::filter::plumbing ... I don't think we need to declare these dependencies separately because both are the optional dependencies enabled by the git feature. We'll also need the gix's "attributes" feature at some point. > attributes - Query attributes and excludes. Enables access to pathspecs, > worktree checkouts, filter-pipelines and submodules. https://docs.rs/gix/0.67.0/gix/index.html#feature-flags	2024-11-14 22:38:04 +09:00
Yuya Nishihara	168f07283a	cargo: whitespace cleanup, sort dependencies	2024-11-14 22:38:04 +09:00
Yuya Nishihara	48521d66b4	stacked_table: add context to error type, propagate lock error	2024-11-14 22:37:54 +09:00
Yuya Nishihara	aaf7f33804	lock: propagate error from lock(), remove panic from unlock path	2024-11-14 22:37:54 +09:00
Yuya Nishihara	6ffd4d4f63	lock: compile and test fallback module on all platforms ... This should help detect Windows issues early. Also fixed clippy warning.	2024-11-14 22:37:54 +09:00
Yuya Nishihara	1a2479cd2f	lock: reorganize platform-specific implementations as sub modules ... This will help rust-analyzer scan both fallback and unix impls.	2024-11-14 22:37:54 +09:00
Martin von Zweigbergk	dfc67e7051	simple_op_heads_store: propagate errors when trying to remove op head ... Any errors other than `NotFound` are unexpected.	2024-11-13 23:05:24 -08:00
Martin von Zweigbergk	de6da1a088	transaction: propagate errors from commit()	2024-11-13 23:05:24 -08:00
Martin von Zweigbergk	73186c1da4	op_heads_store: allow methods to return errors ... This was crashing our server at Google (when attempting to read from a repo that didn't exist).	2024-11-13 23:05:24 -08:00
Martin von Zweigbergk	5b844f630c	cargo: upgrade to sapling-renderdag ... There is now an updated version of `esl01-renderdag` published from the Sapling repo, so let's use that. That lets us remove the `bitflags` 1.x dependency and the `itertools` 0.10.x non-dev dependency.	2024-11-13 09:38:25 -08:00
Yuya Nishihara	7be4904982	tests: fix flakiness in shallow Git repo test ... This test reliably failed if I dropped tv_nsec part from statx(). Since we reload the repo now, several assertions get "fixed". I've added index().has_id() test to clarify that it's still broken.	2024-11-12 20:27:51 +09:00
Yuya Nishihara	062a1bceb9	local_working_copy: on check out, skip entries conflicting with untracked dirs ... This seems more consistent because file->directory conflicts are skipped.	2024-11-12 16:12:12 +09:00
Yuya Nishihara	f3a75c5c46	local_working_copy: on check out, ignore diff of Git submodule ids ... This is different from skipped paths because the file state has to remain as FileType::GitSubmodule in order to ignore the submodule directory when snapshotting. Fixes #4825.	2024-11-12 16:12:12 +09:00
Yuya Nishihara	4983db563f	local_working_copy: migrate Git submodule test to MergedTreeBuilder ... I also removed tx.commit() because the test doesn't rely on the committed operation.	2024-11-12 16:12:12 +09:00
Benjamin Tan	1aad724798	repo: remove MutableRepo::rebase_descendants_return_map ... This function is merely a simple wrapper around `MutableRepo::rebase_descendants_with_options_return_map`.	2024-11-12 14:00:00 +08:00
Benjamin Tan	44ec7d0ee9	repo: remove MutableRepo::rebase_descendants_with_options ... This function had no callers within the codebase.	2024-11-12 14:00:00 +08:00
Benjamin Tan	cb1e0fbafc	rewrite: rebase_commits_with_options: avoid cloning of new parents	2024-11-12 14:00:00 +08:00
Yuya Nishihara	3144a8cb9e	annotate: add line_ranges() and compact_line_ranges() iterator ... They allow callers to test range overlaps with e.g. diff hunks. "jj absorb" will leverage compact_line_ranges().	2024-11-12 08:26:42 +09:00
Yuya Nishihara	077bac8be1	annotate: add low-level function to specify starting file content ... In "jj absorb", we'll need to calculate annotation from the parent tree. It's usually identical to the tree of the parent commit, but this is not true for a merge commit. Since I'm not sure how we'll process conflict trees in general, this patch adds a minimal API to specify a single file content, not a MergedTree.	2024-11-12 08:26:42 +09:00
Yuya Nishihara	85e0a8b068	annotate: add option to not search all ancestors of starting commit ... The primary use case is to exclude immutable commits when calculating line ranges to absorb. For example, "jj absorb" will build annotation of @ revision with domain = mutable().	2024-11-12 08:26:42 +09:00
Benjamin Tan	0e67ef9184	rewrite: avoid abandoned commit parent lookup in rebase_commit_with_options ... This is an optimization to avoid fetching the parent commit of an abandoned commit after rebasing, given that it might not even be used.	2024-11-12 01:33:12 +08:00
Benjamin Tan	9bd7e7707f	repo: add docs for MutableRepo::rebase_descendants_* functions	2024-11-12 01:33:12 +08:00
Benjamin Tan	d7f929fefb	repo: group MutableRepo::rebase_descendants_* functions together	2024-11-12 01:33:12 +08:00
Benjamin Tan	18faaf72a3	rewrite: remove DescendantRebaser ... Due to the gradual rewrite to use the `MutableRepo::transform_descendants` API, `DescendantRebaser` is no longer used in `MutableRepo::rebase_descendants`, and is only used (indirectly through `MutableRepo::rebase_descendants_return_map`) in `rewrite::squash_commits`. `DescendantRebaser` is removed since it contains a lot of logic similar to `MutableRepo::transform_descendants`. Instead, `MutableRepo::rebase_descendants_with_options_return_map` is rewritten to use `MutableRepo::transform_descendants` directly, and the other `MutableRepo::rebase_descendants_{return_map,with_options}` functions call `MutableRepo::rebase_descendants_with_options_return_map` directly. `MutableRepo::rebase_descendants_return_rebaser` is also removed.	2024-11-12 01:33:12 +08:00