Commit graph

1733 commits

Author SHA1 Message Date
Yuya Nishihara
1f5835c267 cli: simplify help about default "log -r"
Since we've moved the default log revset to config/*.toml at 3dab92d2, we don't
have to repeat the default value. It can be queried by "jj config list". I also
split the help paragraphs.
2024-10-01 20:04:06 +09:00
Tyler Goffinet
9f8d936189 cli: color author and commiter names yellow
When `format_short_signature(signature)` is set to `signature.name()` the author names are not yellow like other signature types (eg email and username). When the commit signatures have no colors, they blend in making it hard to distinguish between signatures and commit messages.

If just `name` were set to `yellow`, just like email and username, it affects the colorization of branch names making them also yellow despite them being designated as magenta. Setting `author` and `committer` to `yellow` is specific enough to allow branches to keep their colors while still coloring signature names. This is known to affect signatures in both 'log' and 'show'.
2024-09-28 23:34:26 -07:00
Samuel Tardieu
f38c59f734 split: do not prevent all changes from going into the first commit
Let the user select all changes interactively and put them into
the first commit, and create a second commit with the possibility
of preserving the current commit message. This was previously only
possible in non-interactive mode by specifying matching paths, e.g.
".".  In both cases, a warning will be issued indicating that the second
commit is empty.
2024-09-27 13:33:05 +02:00
Samuel Tardieu
98c16c4e19 split: do not reference file paths in warning
jj split warning was potentially wrong in both interactive and
non-interactive modes when everything is put into the child commit:

- Non-interactive mode: "The given paths does not match any file:
  PATHS". The message is misleading, as the PATHS given on the command-line
  may match files but not match files containing changes.
- Interactive mode: "The given paths does not match any file: " while
  if possible that no paths were given on the command line.
2024-09-27 13:33:05 +02:00
Samuel Tardieu
baf25ba482 unsquash: deprecate in favor of squash and diffedit
Everything that could be done with `jj unsquash` can be done with
`jj squash` or `jj diffedit --preserve-descendants`.
2024-09-26 16:34:48 +02:00
Yuya Nishihara
9551794f19 cli: print operation summary on "op restore" as we do for "undo" 2024-09-25 16:30:22 +09:00
Yuya Nishihara
b30ce36c55 cli: print operation summary before committing transaction
It seemed a bit odd that working-copy update was printed before the "undo"
operation we made earlier.
2024-09-25 16:30:22 +09:00
Essien Ita Essien
895d53f395 Rename conflict and file revsets to conflicts and files.
See discussion thread in linked issue.

With this PR, all revset functions in [BUILTIN_FUNCTION_MAP](8d166c7642/lib/src/revset.rs (L570))
that return multiple values are either named in plural or the naming is hard to misunderstand (e.g. `reachable`)

Fixes: #4122
2024-09-24 20:02:49 +01:00
Samuel Tardieu
cf7847d784 restore: add --restore-descendants flag 2024-09-24 14:26:54 +02:00
Samuel Tardieu
e8deb08f78 abandon: add --restore-descendants flag 2024-09-24 10:52:06 +02:00
Samuel Tardieu
f76ee1872f diffedit: add --restore-descendants flag 2024-09-24 09:30:28 +02:00
Yuya Nishihara
05b90390ae cli: get author timestamp of description template in consistent way
There's a subtle difference. If the original commit was discardable, its author
timestamp is updated by CommitBuilder.
2024-09-24 07:17:33 +09:00
Yuya Nishihara
49e5585130 tests: snapshot some editor contents in test_describe_author() 2024-09-24 07:17:33 +09:00
Samuel Tardieu
69d1600d82 Inline variables in format string 2024-09-23 22:34:55 +02:00
Yuya Nishihara
49e45cc245 revset, templater: add deprecation warnings 2024-09-23 07:07:07 +09:00
Yuya Nishihara
11286b3072 templater: pass diagnostics receiver around 2024-09-23 07:07:07 +09:00
Yuya Nishihara
8b1760ca5d revset: pass diagnostics receiver around
Stacking at AliasExpanded node looks wonky. If we migrate error handling to
Diagnostics API, it might make sense to remove AliasExpanded node and add
node.aliases: vec![(id, span), ..] field instead.

Some closure arguments are inlined in order to help type inference.
2024-09-23 07:07:07 +09:00
Yuya Nishihara
4b477fa59e fileset: pass diagnostics receiver around, add printing function
CLI tests will be added later.
2024-09-23 07:07:07 +09:00
Yuya Nishihara
df8967970e revset, templater: make context message of nested errors less specific
So that these error variants can be reused as warning contexts.
2024-09-23 07:07:07 +09:00
Mateusz Mikuła
8dd3003bec refactor: mark Timestamp struct as Copy 2024-09-22 16:23:53 +02:00
Essien Ita Essien
76f40e5990 cli: reorder updating and reporting for consistency.
* See #4239 for details.
* For now, update working copy before reporting repo changes, so that
  potential errors in reporting changes don't leave the repo in a stale
  state.

Fixes: #4239
2024-09-22 15:02:21 +01:00
Mateusz Mikuła
2e9049b188 cli: add --author argument for commit and describe 2024-09-22 12:40:46 +02:00
Essien Ita Essien
0f5c5530df cli: Warn if specified remote branch not found for jj git fetch
* First fetch from remote.
* Then check tx.{base_repo(),repo}.view().remote_bookmarks_matching(<branch>, <remote>).
  This has to happen after the fetch has been done so the tx.repo() is updated.
* Warn if a branch is not found in any of the remotes used in the fetch. Note that the remotes
  used in the fetch can be a subset of the remotes configured for the repo, so the language
  of the warning tries to point that out.

Fixes: #4293
2024-09-21 14:09:24 +01:00
Yuya Nishihara
4533e81957 cli: pass &Ui to parse fileset/revset/template functions
Deprecation warnings will be printed there. auto_tracking_matcher(ui) could
be cached, but there aren't many callers right now, so it should be okay to
parse and emit warnings for each invocation. Other than that, the changes are
straightforward.
2024-09-20 23:01:12 +09:00
Yuya Nishihara
09cedd83a7 cli: add specialized function for pre-validated templates
I'll make parse_<language>_template() require &Ui, but these cached templates
should be re-constructible without access to a Ui.

Maybe we can split a parsed template object into RevsetExpression-like
evaluation tree and interpreter environment, but that'll be a big challenge.
2024-09-20 23:01:12 +09:00
Yuya Nishihara
3c04507097 cli: remove unneeded &mut from edit_multiple_descriptions() 2024-09-20 23:01:12 +09:00
Yuya Nishihara
dd9213a954 cli: leverage parse_operation_template() in cmd_op_show() 2024-09-20 23:01:12 +09:00
Yuya Nishihara
7e04af1729 cli: cache immutable heads revset expression
This will help simplify warning handling in future patches. I'm going to add
deprecation warnings to revset, so Ui will be required in order to parse a user
revset expression.

revset_util::parse_immutable_expression() is inlined as it's a thin wrapper
around parse_immutable_heads_expression().
2024-09-20 06:56:03 +09:00
Yuya Nishihara
1d86e325c6 cli: move check_repo_rewritable() to WorkspaceCommandEnvironment
This ensures that the repo argument is used instead of self.repo(). store() is
now obtained from the repo argument (which shouldn't matter, though.)
2024-09-20 06:56:03 +09:00
Yuya Nishihara
0c9bc37f10 diff: buffer color-words context lines
I'm testing simple conflicts diffs locally, and we'll probably need to handle
consecutive context hunks when we add some form of unmaterialized conflicts
diffs. Let's buffer context hunks (up to 1 right now.) The new code looks
simpler.
2024-09-19 20:01:51 +09:00
Yuya Nishihara
9f5ce934d2 diff: extract function that prints color-words diff lines
This block is relatively big, and will become even bigger if we add support
for unmaterialized conflicts diffs.
2024-09-19 20:01:51 +09:00
Yuya Nishihara
79e76c5dad diff: extract line number handling from show_color_words_diff_line()
This seems more consistent with !can_inline case, and we can remove extra
vec![] allocation from context line processing.
2024-09-19 20:01:51 +09:00
Samuel Tardieu
726cbbd66c diffedit: separate extra help information in its own paragraph 2024-09-18 09:11:07 +02:00
Vamsi Avula
a2e67bdd9b templates: add .normal_hex() for {Commit,Change}Id
This doesn't do much for CommitId but outputs the "forward" hex for
ChangeId, which can be useful to use as Gerrit Change-Id.
2024-09-17 21:11:00 -05:00
Samuel Tardieu
05c6d62c68 bookmark: add "b" alias
`jj bookmark` is a frequently used command. Its subcommands already have
one letter aliases. Defining `jj b` as an alias for `jj bookmarks` make
bookmarks really easy to use.
2024-09-17 18:24:26 +02:00
Essien Ita Essien
e25ec536ed next/prev: Fetch descendants with more correctness.
See context in [this discussion](https://github.com/martinvonz/jj/pull/3935#discussion_r1649520967)

Fixes #3947
2024-09-17 16:39:29 +01:00
Samuel Tardieu
6d0a092410 cli: remove unused dependency on hex 2024-09-17 12:34:31 +02:00
Samuel Tardieu
c1a032fad6 squash doc: --keep-emptied let the description unchanged 2024-09-17 11:49:48 +02:00
Kevin Liao
412ef36259 cli: Support renaming workspaces
fixes #4342
2024-09-16 19:35:36 -07:00
Samuel Tardieu
a7f32d3652 squash: accept -k as a shorthand for --keep-emptied
This eases the workflow in which a commit in the middle of the tree is
repeatedly squashed into its parent.
2024-09-15 18:31:47 +02:00
Lukas Wirth
7ff12ec8b0 op-diff: Use op_summary template 2024-09-15 14:11:33 +02:00
Lukas Wirth
8e727de2ab undo: Report what operation has been undone in jj op undo 2024-09-15 14:11:33 +02:00
Yuya Nishihara
cb07e5ab67 cli: borrow RepoLoader from Workspace
It should be slightly cheaper than constructing new RepoLoader.
2024-09-15 20:40:46 +09:00
Yuya Nishihara
9dc2cf802c cli: add helper to set up op template environment for the current repo state
We'll add a few more callers.
2024-09-15 20:40:46 +09:00
Yuya Nishihara
0c0e1b61e3 git-push: just say "Changes to push to <REMOTE>"
"Bookmark changes" sounds like changes will be bookmarked, and "Bookmark" here
is redundant. If we add support for pushing tags, this message will have to be
generalized anyway.
2024-09-15 07:06:56 +09:00
Yuya Nishihara
a48ecf3e62 git-push: remove redundant hash map lookup from print-changes loop 2024-09-15 07:06:56 +09:00
Yuya Nishihara
2f78c25d35 git-push: skip print-changes loop if --quiet 2024-09-15 07:06:56 +09:00
Yuya Nishihara
42f7908b8d git-push: extract function that prints bookmark changes to be pushed 2024-09-15 07:06:56 +09:00
Samuel Tardieu
6e72b1cfb0 git: add --remote option to clone command
This makes it easier to work with multiple remotes at once while
tracking the default branch of the remote used to create the local
repository:

```shell
$ jj git clone --remote upstream https://github.com/upstream-org/repo
$ cd repo
$ jj git remote add origin git@github.com:your-org/repo
$ jj config set --repo git.fetch upstream
```

In the example above, `upstream` is the repository containing the
reference source code that you might want to patch, while `origin` is
your fork where pull-request will be pushed. The branch `main@upstream`
will be tracked.
2024-09-13 18:14:57 +02:00
Yuya Nishihara
41ef9a76de cli: actually test if format_short_change_id() can be parsed
It's usually included in the commit_summary template, but not guaranteed.
2024-09-13 21:59:35 +09:00
Lukas Wirth
1410f2bee7 docs: Remove trailing whitespace in markdown files 2024-09-13 13:06:28 +02:00
Lukas Wirth
bccd8214df conflicts: Highlight change ids in rebase advice 2024-09-13 12:57:56 +02:00
Martin von Zweigbergk
63e616c801 git: restore support for git.push-branch-prefix config but deprecate it 2024-09-12 23:28:30 -07:00
Martin von Zweigbergk
c2220ca4d5 config-schema: update stale push-branch-prefix config
We missed this when we renamed `push-branch-prefix` to
`push-bookmark-prefix`. I changed the description slightly to try to
clarify that it's about the local bookmark that's created before
pushing it to the remote as a branch.
2024-09-12 23:00:20 -07:00
Martin von Zweigbergk
1aa2aec141 bookmarks: update some leftover uses of the word "branch" 2024-09-11 19:19:31 -07:00
Yuya Nishihara
d97a7ba86f cli: fix missed working-copy "bookmark" colors
Quoting style changed because of insta v1.40.0.
2024-09-12 09:02:16 +09:00
Yuya Nishihara
78edc6aba5 op log: add --op-diff option to embed operation diffs
This is basically "log -p" for "op log". The flag name has "op" because --diff
and --patch mean a similar thing in this context. Since -p implies --op-diff,
user can just do "op log -p" if he's okay with verbose op + content diffs.
Note that --no-graph affects both "op log" and "op diff" parts.

We might want to do some style changes later, such as inserting/deleting blank
lines, highlighting headers, etc.
2024-09-12 06:51:45 +09:00
Yuya Nishihara
050cde94ef op diff: apply word wrapping to non-graph template outputs
These contents may be embedded in "op log", and it's probably good to enable
word wrapping consistently.
2024-09-12 06:51:45 +09:00
Yuya Nishihara
babdf6b9c1 cli: relax error type of LogContentFormat::write() 2024-09-12 06:51:45 +09:00
Yuya Nishihara
6154827129 cli: make log word-wrapping helper calculate width eagerly
Multiple graphs will be nested in "op log" output, and things would be messy if
we had to calculate graph widths lazily. Let's simply make LogContentFormat
track the current available width no matter if ui.log-word-wrap is off.
2024-09-12 06:51:45 +09:00
Philip Metzger
d9c68e08b1 everything: Rename branches to bookmarks
Jujutsu's branches do not behave like Git branches, which is a major
hurdle for people adopting it from Git. They rather behave like
Mercurial's (hg) bookmarks. 

We've had multiple discussions about it in the last ~1.5 years about this rename in the Discord, 
where multiple people agreed that this _false_ familiarity does not help anyone. Initially we were 
reluctant to do it but overtime, more and more users agreed that `bookmark` was a better for name 
the current mechanism. This may be hard break for current `jj branch` users, but it will immensly 
help Jujutsu's future, by defining it as our first own term. The `[experimental-moving-branches]` 
config option is currently left alone, to force not another large config update for
users, since the last time this happened was when `jj log -T show` was removed, which immediately 
resulted in breaking users and introduced soft deprecations.

This name change will also make it easier to introduce Topics (#3402) as _topological branches_ 
with a easier model. 

This was mostly done via LSP, ripgrep and sed and a whole bunch of manual changes either from
me being lazy or thankfully pointed out by reviewers.
2024-09-11 18:54:45 +02:00
Yuya Nishihara
d04ff1213f op diff: set up diff/template context without using WorkspaceCommandHelper
I'll use a similar setup in "op log", but for each log entry. We might want to
extract some parts to helper function, but I don't have a good idea right now.

CommandHelper::operation_template_extensions() is removed because it's unlikely
to parse operation template without loading a workspace.
2024-09-10 07:06:43 +09:00
Yuya Nishihara
1e17078eb5 op log: leverage WorkspaceCommandEnvironment to parse template
I'll add an option to include diffs in "op log", and workspace_env will be used
in order to set up diff/template contexts per operation.

cmd_op_log() is split because of owned/borrowed type differences.
2024-09-10 07:06:43 +09:00
Yuya Nishihara
e0da3ec010 cli: move some template helpers to WorkspaceCommandEnvironment
These functions will be used by "op log"/"diff"/"show".

This patch also changed the error type as it's obvious that there are no other
errors to be returned.
2024-09-10 07:06:43 +09:00
Yuya Nishihara
1da9ebcec1 cli: cache parsed revsets.short-prefixes expression
It doesn't make sense to reparse revset expression. Let's reuse the parse
result. This also simplifies error handling bits.

OnceCell is switched to the std one as we no longer need get_or_try_init().
2024-09-10 07:06:43 +09:00
Yuya Nishihara
90791216c5 cli: move revset-related contexts to WorkspaceCommandEnvironment
This makes it clear that RevsetParseContext doesn't depend on repo view.
2024-09-10 07:06:43 +09:00
Yuya Nishihara
9ca6c99508 cli: extract WorkspaceCommandEnvironment that doesn't depend on a repo
Some "operation" commands need a workspace, plus multiple repo views. We
currently load WorkspaceCommandHelper twice for that reason. It works, but
would be messy if "op log" loaded WorkspaceCommandHelper for each log entry.

This patch starts splitting non-repo data from WorkspaceCommandHelper. The
workspace object isn't owned by the environment object so the object can be
freely discarded.
2024-09-10 07:06:43 +09:00
Yuya Nishihara
54443c9a8d cli: remove unneeded &mut Ui requirement from WorkspaceCommandHelper
These functions just print messages, and never change Ui state such as
pagination.
2024-09-10 07:06:43 +09:00
Martin von Zweigbergk
f36f4ad257 cli: make paths to auto-track configurable, add jj track
It's a pretty frequent request to have support for turning off
auto-tracking of new files and to have a command to manually track
them instead. This patch adds a `snapshot.auto-track` config to decide
which paths to auto-track (defaults to `all()`). It also adds a `jj
track` command to manually track the untracked paths.

This patch does not include displaying the untracked paths in `jj
status`, so for now this is probably only useful in colocated repos
where you can run `git status` to find the untracked files.

#323
2024-09-09 07:49:55 -07:00
Yuya Nishihara
653e8087da workspace: make workspace_root() and repo_path() return slice &Path
It's common to return &PathBuf as &Path.
2024-09-08 05:40:52 +09:00
Yuya Nishihara
47307556dd working_copy: pass SnapshotOptions by reference
Though SnapshotOptions can be cheaply cloned, it doesn't make much sense that
snapshot() consumes a settings-like object.
2024-09-08 04:51:21 +09:00
Martin von Zweigbergk
0b6714a33c workspace: take old repo path as separate argument
As I mentioned in recent patches, I'm about to remove
`ReadonlyRepo::repo_path()`.
2024-09-07 12:20:02 -07:00
Martin von Zweigbergk
bfb16a4c54 cli: get repo path from WorkspaceCommandHelper, not from ReadonlyRepo
I'd like to remove `ReadonlyRepo::repo_path()` since it doesn't make
sense when the repo is stored in a database.
2024-09-07 12:20:02 -07:00
Martin von Zweigbergk
8d090628c3 transaction: rename mut_repo() to idiomatic repo_mut()
We had both `repo()` and `mut_repo()` on `Transaction` and I think it
was easy to get confused and think that the former returned a
`&ReadonlyRepo` but both of them actually return a reference to
`MutableRepo` (the latter obviously returns a mutable reference). I
hope that renaming to the more idiomatic `repo_mut()` will help
clarify.

We could instead have renamed them to `mut_repo()` and
`mut_repo_mut()` but that seemed unnecessarily long. It would better
match the `mut_repo` variables we typically use, though.
2024-09-07 10:51:43 -07:00
Martin von Zweigbergk
e5e95225d4 cleanup: inline some short single-use mut_repo variables 2024-09-07 10:51:43 -07:00
Martin von Zweigbergk
5fbc819fc1 cleanup: avoid some unnecessary uses of mutable repo reference 2024-09-07 10:51:43 -07:00
Yuya Nishihara
5bae3eaa74 cli: remove unneeded Display for SparseInheritance
Since it's a ValueEnum, clap knows how to stringify the default value.
2024-09-07 10:17:37 +09:00
Yuya Nishihara
b263882e4b config-schema: add signing.backend=none to the list of enum variants 2024-09-07 10:17:08 +09:00
Yuya Nishihara
c3ed3e3781 evolog: use predecessor_ids() where commit object isn't needed 2024-09-07 09:09:27 +09:00
Yuya Nishihara
25819dd1c9 evolog: inline show_predecessor_patch() 2024-09-07 09:09:27 +09:00
Yuya Nishihara
03487af192 diff: extract inter-diff helper
This isn't as primitive as show_diff() or show_patch(), but there are three
callers. Let's add a helper method to remove repo argument.
2024-09-07 09:09:27 +09:00
Yuya Nishihara
f473d7bb7e op diff: pass merged repo in to DiffRenderer
Since DiffRenderer borrows a repo, it shouldn't be constructed from the base
WorkspaceCommandHelper.
2024-09-07 09:09:27 +09:00
Yuya Nishihara
b7caaf586b op diff: pass DiffRenderer in to show_op_diff() by reference
show_op_diff() doesn't need an ownership of renderer.
2024-09-07 09:09:17 +09:00
Yuya Nishihara
5bff5cdef0 op diff: create formatter by show_op_diff() callers
This will help embed operation diffs in "op log".
2024-09-07 09:09:17 +09:00
Yuya Nishihara
ea8f543d5a op show: accept root operation by mapping empty parents to root
I'm thinking of adding an option to embed operation diffs in "op log", and
"op log" shouldn't fail at the root operation. Let's make "op diff"/"show"
also work for consistency.
2024-09-07 09:09:17 +09:00
Austin Seipp
a31fe7f6d6 cli: implement workspace add --sparse-patterns
This flag implements three modes:

- `copy`: copy sparse patterns from parent
- `full`: do not copy sparse patterns from parent
- `empty`: clear all paths, equal to `set --clear`

This is useful for various tooling like tools that want to run a parallel
process that queries the build system (without running into locks/blocking.)

I think continuing to copy sparse patterns makes sense as the default behavior.

Signed-off-by: Austin Seipp <aseipp@pobox.com>
2024-09-06 16:22:47 -05:00
Danny Hooper
bf543402cc cli: fix: add --include-unchanged-files flag to allow fixing as yet unchanged files
This enables workflows like "insert a commit that reformats the code in one of
my project directories".

`jj fix --include-unchanged-files` is an easy way to fix everything in the repo.

`jj fix --include-unchanged-files <file...>` fixes all of the `<files>` even if they are
unchanged.

This is mostly orthogonal to other features, so not many tests are added.

This is a significant and simple enough improvement that I think it's
appropriate to make it here instead of waiting for a `jj run`-based solution.
2024-09-06 13:50:28 -05:00
Martin von Zweigbergk
fee706dfea signing: allow disabling by setting backend to "none"
We don't seem to have a way to override the config to disable
signing. This patch lets you do that by setting
`signing.backend="none"`.
2024-09-05 21:02:22 -07:00
Yuya Nishihara
6ea46b930b op diff: resolve GraphStyle by caller, remove &CommandHelper argument
show_op_diff() has many arguments. Let's make it a bit less.
2024-09-06 09:51:10 +09:00
Yuya Nishihara
dda0980c4d graphlog: inline node_template_for_key()
It's easier to follow than function call with 3 string parameters.
2024-09-06 09:51:10 +09:00
Yuya Nishihara
36ab165b57 cli: parse graph node settings strictly 2024-09-06 09:51:10 +09:00
Yuya Nishihara
406ead241b graphlog: load settings by caller
I'll make it error out if the style name is invalid.
2024-09-06 09:51:10 +09:00
Yuya Nishihara
bd7cbaaffb graphlog: add enum describing graph style 2024-09-06 09:51:10 +09:00
Yuya Nishihara
3ad54ad38d settings: move graph node helpers to CLI
I'm going to add enum GraphStyle, and it's specific to the CLI.
node_template_for_key() will be inlined later.
2024-09-06 09:51:10 +09:00
Yuya Nishihara
058249cc57 cli: update docs URLs to point to https://martinvonz.github.io/jj/latest/ 2024-09-06 09:22:04 +09:00
Martin von Zweigbergk
a0fae76622 cli: rename obslog to evolution-log/evolog
It seems everyone agrees that `obslog` is not an intuitive name. There
was some discussion about alternatives in #3592 and on #4146. The
alternatives included `evolution`, `evolutionlog`, `evolog`,
`rewritelog`, `revlog`, and `changelog`. It seemed like
`evolution-log`/`evolog` was the most popular option. That also
matches the command's current help text ("Show how a change has
evolved over time").
2024-09-05 13:45:17 -07:00
Martin von Zweigbergk
ce04e61dd4 tests: set $COLUMNS when running CLI tests
As reported in #4394, at least `test_show_command::test_show_basic`
can fail when run with a narrow terminal. This patch sets
`COLUMNS=100` in the environment when running tests so the CLI uses
that value instead of using the width of the user's terminal.
2024-09-05 08:56:04 -07:00
Ilya Grigoriev
6f44ed4f51 cli squash: explain --into in error message for merge commits 2024-09-04 21:38:55 -07:00
Martin von Zweigbergk
4a8d250f2c store: make write_file() async 2024-09-04 18:34:11 -07:00
Martin von Zweigbergk
8eb3d85b1c backend: make write methods async
This doesn't provide any benefit yet bit I think we've known for a
while that we want to make the backend write methods async. It's just
not been important to Google because we have the local daemon process
that makes our writes pretty fast. Regardless, this first commit just
changes the API and all callers immediately block for now, so it won't
help even on slow backends.
2024-09-04 18:34:11 -07:00
Kaleb Pace
dbb579adec cli/build-rs: remove metadata command dependency in favor of env vars
When building this project with [Nix/Crane](https://github.com/ipetkov/crane/discussions/693), if the `jj-cli` dependency is specified in `Cargo.toml` as a git-based crate, `cargo vendor` splits this workspace up into sub-crate directories, which causes `cargo metadata` to fail when searching for relative deps in the workspace root.

This commit simply changes how the crate version is determined, using Cargo's built-in environment variable [`CARGO_PKG_VERSION`](https://doc.rust-lang.org/cargo/reference/environment-variables.html)
2024-09-03 22:31:17 -07:00
Scott Taylor
52499e84cc config: add color for renamed and copied in summary
Right now, renamed and copied files don't have any color in the output
of `jj status`, and it makes them stand out. I think it's reasonable to
color renamed files the same as modified files, since renaming is like
modifying the path, and to color copied files the same as added files,
since they're basically just added files that happen to have similar
contents to an existing file.
2024-09-03 21:33:29 -05:00
Martin von Zweigbergk
6fdef8bf67 cli: make WorkspaceCommandHelper keep a CommandHelper
This avoids cloning `UserSettings` and some other data. I haven't
attempted to measure the performance impact (I expect it's tiny); this
is more about clarifying that there are not multiple different
versions of these fields.
2024-09-03 08:41:24 -07:00
Martin von Zweigbergk
34425a2501 cli: make CommandHelper cheaply cloneable
This wraps all the fields in `CommandHelper` in an `Rc` so
`CommandHelper` itself becomes cheap to clone (thanks to @yuja for the
idea). I'll use that next to avoid some cloning in
`WorkspaceCommandHelper`.
2024-09-03 08:41:24 -07:00
Martin von Zweigbergk
cc15ecf7c7 op log: change "resolve concurrent" to "reconcile divergent"
"Concurrent" operations are not necessarily actually concurrent, so
"divergent" seems like a better name. And "reconcile" seems like a
better term for merging them, though we also sometimes use "merge".
2024-08-30 21:56:11 -07:00
dploch
f963af3f16 workspace: turn WorkspaceLoader into a trait
Like https://github.com/martinvonz/jj/pull/4189, this allows extensions the ability to load the repo in an environment where the local filesystem is not accessible. This change allows such extensions to exist at the CLI layer where jj is invoked as a subprocess, rather than a library (common in testing).
2024-08-30 13:35:52 -04:00
Yuya Nishihara
d99a8508e4 rewrite: pass old parent ids in to mut_repo.new_parents() by slice 2024-08-29 08:59:22 +09:00
Yuya Nishihara
5af906d924 cli: change default inline threshold of color-words diffs
I played with max-inline-alternation = 3 for a couple of weeks, and it's pretty
good. I think somewhere between 2 and 4 is good default because one or two
remove + add sequences are easy to parse.
2024-08-28 10:33:33 +09:00
Yuya Nishihara
9c9e564dc4 conflicts: pass around conflict contents without materialization
We haven't decided how conflict diffs should be rendered, but whatever style
we'll choose, we'll need raw unmaterialized conflict contents.
2024-08-28 10:23:57 +09:00
Yuya Nishihara
73a8b13e4c conflicts: split materialized value to FileConflict and OtherConflict
FileConflict will be changed to not materialize Merge<BString>. I also updated
the revset engine to ignore non-file conflict. It doesn't make sense to grep
conflict description.
2024-08-28 10:23:57 +09:00
Yuya Nishihara
0c14a0a9ca conflicts: make describe() simply return string
I'll add more callers of id.describe(), and the output size wouldn't be large
enough to avoid allocation by using Write API.
2024-08-28 10:23:57 +09:00
Yuya Nishihara
1ba581b37c conflicts: replace ContentHunk with BString
ContentHunk is basically a nice wrapper around Vec<u8>. I think it would give
little benefit for type safety.
2024-08-28 10:23:57 +09:00
Austin Seipp
d928e43073 cli: add name() and about() methods to CliRunner
Useful for third party tools to override, along with the version.

Signed-off-by: Austin Seipp <aseipp@pobox.com>
2024-08-25 12:24:02 -05:00
Yuya Nishihara
6b65f8abec cli: move "untrack" to "file" subcommand
I don't think "jj untrack" is frequently used, and I think it is a "file"
command rather than "workspace".
2024-08-26 01:19:15 +09:00
Yuya Nishihara
3c15f8106a cli: add deprecation comment to cat, chmod, and files 2024-08-26 01:19:15 +09:00
Yuya Nishihara
3762f890f6 cli: extract helper that issues deprecation warning and invoke command fn 2024-08-26 01:19:15 +09:00
Yuya Nishihara
3b0b70a451 cli: fix default value of ui.allow-filesets in config schema
It's been enabled since 8fec7500c3 "cli: enable fileset by default."
2024-08-25 23:16:29 +09:00
Essien Ita Essien
f9dc0589c1 next/prev: update error message when no movement targets are found.
If movement commands don't find a target commit, they fail. However,
it's usually not intuitive why they fail because in non-edit mode the
start commit is the parent of the working commit.

Adding the start commit change hash to the error message makes it easier
for the user to figure out what is going on.

Also, specifying 'No **other** descendant...' helps make it clear what
`jj` is really looking for.

Part of #3947
2024-08-25 09:58:04 +01:00
Yuya Nishihara
8d166c7642 templater: integrate copy tracking in commit.diff() template 2024-08-24 15:47:08 +09:00
Martin von Zweigbergk
b78c83e9fe status: report copies and renames 2024-08-23 18:51:02 -07:00
Yuya Nishihara
f1be5c59f3 cleanup: consistently import std::fmt 2024-08-23 13:05:27 +09:00
Yuya Nishihara
87fb169266 cleanup: remove redundant ::{self} from use declarations 2024-08-23 13:05:27 +09:00
Yuya Nishihara
4ffd3a3055 diff: pass tree diff stream in to inner show_*() functions as before 2024-08-23 10:29:12 +09:00
Yuya Nishihara
f5187fa063 copies: determine copy/rename operation by CopiesTreeDiffStream
Not all callers need this information, but I assumed it's relatively cheap to
look up the source path in the target tree compared to diffing.

This could be represented as Regular(_)|Copied(_, _)|Renamed(_, _), but it's
a bit weird if Copied and Renamed were separate variants. Instead, I decided
to wrap copy metadata in Option.
2024-08-23 10:29:12 +09:00
Yuya Nishihara
08262eb152 copies: extract (source, target) path pair to separate type
This patch adds accessor methods as I'm going to change the underlying data
types. Since entry values are consumed separately, these methods are implemented
on CopiesTreeDiffEntryPath, not on *TreeDiffEntry.
2024-08-23 10:29:12 +09:00
Yuya Nishihara
43bf195314 merged_tree: rename diff entry field from "value" to "values"
It seems a slightly better, and aligns with the local variable name in
materialized_diff_stream().
2024-08-23 10:29:12 +09:00
Matt Kulukundis
ac7fa1abb9 docs: improve md format for jj fix 2024-08-22 16:03:44 -04:00
Matt Kulukundis
8ead72e99f formatting only: switch to Item level import ganularity 2024-08-22 14:52:54 -04:00
Yuya Nishihara
352a4a0eea copies: filter rename source entries by CopiesTreeDiffStream 2024-08-22 20:17:19 +09:00
Yuya Nishihara
7684ab5994 diff: omit rename source entries from color-words diffs
Rename source entries will be handled by CopiesTreeDiffStream.
2024-08-22 20:17:19 +09:00
Yuya Nishihara
2cffcc9323 copies: provide source path mapping by CopyRecords
All for/has_source/target() combinations are added for API consistency.
2024-08-22 20:17:19 +09:00
Yuya Nishihara
5e356ffd24 diff: filter out uninteresting copy records by matcher
Git reports a rename source as deleted if the rename target is excluded. I
think that's because Git restricts the search space to the specified paths. For
example, Git doesn't also recognize a rename if the source path is excluded
whereas jj does.

I don't think we need to copy the exact behavior of Git, so this patch just
moves matcher application to earlier stage. This change will help remove
collect_copied_sources().

The added get_copy_records() helper could be moved to jj_lib, but we'll probably
want a stream version of this function in library, and writing a stream adapter
isn't as simple as iterator.
2024-08-22 20:17:19 +09:00
Yuya Nishihara
d85e66bbb4 copies: turn add_records() into non-stream API, block_on_stream() by caller
This is simpler, and I think it's generally better to not spawn executor in
library code.
2024-08-22 20:17:19 +09:00
Yuya Nishihara
a83dadd5a9 diff: add option to display complex color-words diffs without inlining
In this patch, I use the number of adds<->removes alternation as a threshold,
which approximates the visual complexity of diff hunks. I don't think user can
choose the threshold intuitively, but we need a config knob to try out some.
I set `max-inline-alternation = 3` locally. 0 and 1 mean "disable inlining"
and "inline adds-only/removes-only lines" respectively.

I've added "diff.<format>" config namespace assuming "ui.diff" will be
reorganized as "ui.diff-formatter" or something. #3327

Some other metrics I've tried:
```
// Per-line alternation. This also works well, but can't measure complexity of
// changes across lines.
fn count_max_diff_alternation_per_line(diff_lines: &[DiffLine]) -> usize {
    diff_lines
        .iter()
        .map(|line| {
            let sides = line.hunks.iter().map(|&(side, _)| side);
            sides
                .filter(|&side| side != DiffLineHunkSide::Both)
                .dedup() // omit e.g. left->both->left
                .count()
        })
        .max()
        .unwrap_or(0)
}

// Per-line occupancy of changes. Large diffs don't always look complex.
fn max_diff_token_ratio_per_line(diff_lines: &[DiffLine]) -> f32 {
    diff_lines
        .iter()
        .filter_map(|line| {
            let [both_len, left_len, right_len] =
                line.hunks.iter().fold([0, 0, 0], |mut acc, (side, data)| {
                    let index = match side {
                        DiffLineHunkSide::Both => 0,
                        DiffLineHunkSide::Left => 1,
                        DiffLineHunkSide::Right => 2,
                    };
                    acc[index] += data.len();
                    acc
                });
            // left/right-only change is readable
            (left_len != 0 && right_len != 0).then(|| {
                let diff_len = left_len + right_len;
                let total_len = both_len + left_len + right_len;
                (diff_len as f32) / (total_len as f32)
            })
        })
        .reduce(f32::max)
        .unwrap_or(0.0)
}

// Total occupancy of changes. Large diffs don't always look complex.
fn total_change_ratio(diff_lines: &[DiffLine]) -> f32 {
    let (diff_len, total_len) = diff_lines
        .iter()
        .flat_map(|line| &line.hunks)
        .fold((0, 0), |(diff_len, total_len), (side, data)| {
            let l = data.len();
            match side {
                DiffLineHunkSide::Both => (diff_len, total_len + l),
                DiffLineHunkSide::Left => (diff_len + l, total_len + l),
                DiffLineHunkSide::Right => (diff_len + l, total_len + l),
            }
        });
    (diff_len as f32) / (total_len as f32)
}
```
2024-08-21 17:48:52 +09:00
Yuya Nishihara
be9b7ed88f diff: pad last newline by show_color_words_diff_line()
Though this is needed only for the last line, checking it for each line is
cheap. As I'm going to add another rendering style, the condition to pad "\n"
would become more complicated.
2024-08-21 17:48:52 +09:00
Yuya Nishihara
bb686c1459 diff: extract helper that prints color-words line numbers 2024-08-21 17:48:52 +09:00
Yuya Nishihara
6b2e191b2b diff: extract helpers that process word-level diffs without interleaving
These functions will be reused by non-inline version of color-words diffs.
2024-08-21 17:48:52 +09:00
Yuya Nishihara
decd913cea cli: introduce options struct specific to color-words diffs
I'll add a threshold whether to inline word changes.
2024-08-21 17:48:52 +09:00
Yuya Nishihara
5ff4e2d0a2 cli: pass &DiffFormatArgs in to default_diff_format()
Suppose we add -w/--ignore-all-space flag for example, it's probably better to
pass optional parameters by struct, not by separate arguments.
2024-08-21 17:48:52 +09:00
Yuya Nishihara
f76db7fa4b cli: flatten diff-args-to-formats translation
I'll add more parameters to color-words diffs, and the format variant will no
longer be trivially constructible.
2024-08-21 17:48:52 +09:00
Essien Ita Essien
bb018a54c3 next/prev: Add config flag to control prev/next edit behaviour.
* We started with a tristate flag where:
    - Auto - Maintain current behaviour. This edits if
      the wc parent is not a head commit. Else, it will
      create a new commit on the parent of the wc in
      the direction of movement.
    - Always - Always edit
    - Never - Never edit, prefer the new+squash workflow.
  However, consensus the review thread is that `auto` mode where we try to infer when to
  switch to `edit mode`, should be removed. So `ui.movement.edit` is a boolean flag now.
    - true: edit mode
    - false: new+squash mode
* Also add a `--no-edit` flag as the explicit inverse of `--edit` and
  ensure both flags take precedence over the config.
* Update tests that assumed edit mode inference, to specify `--edit` explicitly.

NOTE: #4302 was squashed into this commit, so see that closed PR for review history.

Part of #3947
2024-08-20 15:46:00 +01:00
Marie Ramlow
e3fcae2152 merge_tools.toml: add VSCodium as a merge tool
[VSCodium](https://vscodium.com/) is a free/libre distribution of
Microsoft's Visual Studio Code editor, it's functionally more or less
the same, but distributed under a FOSS license, unlike VS Code.

This adds VSCodium as a merge tool.
2024-08-19 22:37:00 -07:00
Marijan Smetko
0852724c76 Warn user about the working copy when configuring the author 2024-08-19 17:09:30 +02:00
Martin von Zweigbergk
3acb89e7cc merged_tree: remove TreeDiffEntry::source 2024-08-18 22:16:41 -07:00
Martin von Zweigbergk
721aa1238c copies: add a separate diff stream item type with copy info
The goal is to have the new item type know if it represent a copy, a
rename, a deleted rename source, or a regular copy-unrelated item.
2024-08-18 22:16:41 -07:00
Martin von Zweigbergk
70598498b0 merged_tree: provide separate version of diff_stream() with copy info
I plan to provide a richer version of `TreeDiffEntry` with copy info
(and to make `TreeDiffEntry` itself "poorer"). Most callers want to
know about copies/renames, but at least working copy implementations
probably don't. This patch adds separate `diff_stream()` and
`diff_stream_with_copies()` so we can provide the simpler interface
for callers that don't need copy info.
2024-08-18 22:16:41 -07:00
Martin von Zweigbergk
fd9a236be5 copies: move CopyRecords to new copies module
Copy/rename handling is complicated. It seems worth having a module
for it. I'm going to add more content to it next.
2024-08-18 22:16:41 -07:00
Martin von Zweigbergk
aa0fbd9f3f drop "support" for legacy tree config
The tree-level conflicts have worked well in practice and we don't
want to allow users to use legacy trees for new commits. We don't
really support legacy trees very well since 0590f8bece anyway.
2024-08-18 07:19:50 -07:00
Essien Ita Essien
e2a5c83e5c Follow ups on post-submission comments for #4282.
* Derive a bunch of standard and useful traits for `movement_util::Direction`
  as it is a simple type. Importantly `Copy`.
* Return `&'static str` from Direction.cmd()
* Refactor out `MovementArgs` to reduce the number of arguments
  to `movement_util::move_to_commit`.
* Implement `From<&NextArgs/&PrevArgs>` for MovementArgs

Part of #3947
2024-08-18 10:43:08 +01:00
Yuya Nishihara
1be955ea4e diff: simplify conditions whether to emit color-words context lines
This appears to fix redundant "    ..." line for empty diffs.
2024-08-18 12:40:07 +09:00