The "git" CLI chdir()s to the work tree root, so paths in config file are
usually resolved relative to the workspace root. OTOH, jj doesn't modify the
process environment, so libgit2 resolves remote paths relative to cwd, not to
the workspace root. To mitigate the problem, this patch makes "jj git remote"
sub commands to store resolved path in .git/config. It would be nice if we can
reconfigure in-memory git2 remote object to use absolute paths (or set up
in-memory named remote without writing a config file), but there's no usable
API afaik.
This behavior is different from "git remote add"/"set-url". I don't know the
rationale, but these commands don't resolve relative paths, whereas "git clone"
writes resolved path to .git/config. I think it's more consistent to make all
"jj git" sub commands resolve relative paths.
When running `cmd.spawn()` rust will by default inherit
the stderr of the parent, so `jj log test<TAB>`, would
print `There is no jj repo in "."` into the prompt.
UserSettings will be obtained from there. If operation template supported
"self.repo() -> Repo" method, RepoLoader would be needed. It's equivalent to
Repo in CommitTemplateLanguage.
This makes the graph compact. Short change ids are usually printed as a part
of commit summary. The --no-graph output is a bit harder to parse, but we can
still discriminate entries by change ids.
This is a middle ground. An inline table can still be overwritten or deleted
because it's syntactically a value. It would be annoying if "jj config set
colors.<name> '{ .. }'" couldn't be executed more than once because the existing
item was a table.
#5255
The list of migration rules is managed by CliRunner. I don't know if that's
needed, but in theory, an extension may insert migration rules as well as
default config layers.
Migration could be handled by ConfigEnv::resolve_config(), but it seemed rather
complicated because Vec<ConfigMigrationRule> cannot be cloned, and the scope of
these variables are a bit different.
If many files are conflicted, it would be nice to be able to resolve all
conflicts at once without having to run `jj resolve` multiple times.
This is especially nice for merge tools which try to automatically
resolve conflicts without user input, but it is also good for regular
merge editors like VS Code.
This change makes the behavior of `jj resolve` more consistent with
other commands which accept filesets since it will use the entire
fileset instead of picking an arbitrary file from the fileset.
Since we don't support passing directories to merge tools yet, the
current implementation just calls the merge tool repeatedly in a loop
until every file is resolved, or until an error occurs. If an error
occurs after successfully resolving at least one file, the transaction
is committed with all of the successful changes before returning the
error. This means the user can just close the editor at any point to
cancel resolution on all remaining files.
There are some experiments to try and compile `jj` to WebAssembly, so that we
might be able to do things like interactive web tutorials. One step for that
is making `git` support in `jj-cli` optional, because we can stub it out for
something more appropriate and it's otherwise a lot of porting annoyance for
both gitoxide and libgit2.
(On top of that, it might be a useful build configuration for other experiments
of mine where removing the need for the large libgit2 depchain is useful.)
As part of this, we need to mark `jj-lib` as having `default-features = false`
in the workspace dependency configuration; otherwise, the default behavior
for Cargo is to compile with all its default features, i.e. with git support
enabled, ignoring the `jj-cli` features clauses.
Other than that, it is fairly straightforward junk; it largely just sprinkles
some `#[cfg]` around liberally in order to make things work. It also adjusts the
CI pipeline so this is tested there, too, so we can progressively clean it up.
Signed-off-by: Austin Seipp <aseipp@pobox.com>
Certain tools (`diff`, `delta`) exit with code 1 to indicate there was
a difference. This allows selectively suppressing the "Tool exited with
... status" warning from jj when generating a diff.
example:
```toml
[merge.tools.delta]
diff-expected-exit-codes = [0, 1]
```
This backs out commit 0de36918e4. Documentation,
tests, and comments are updated accordingly. I also add ConfigTableLike type
alias as we decided to abstract table-like items away.
Closes#5255
If the parent tree contains conflicts, we want the index to also contain
a conflict to ensure that the use can't accidentally commit conflict
markers using `git commit`. Since we can't represent conflicts with more
than 2 sides in the Git index, we need to add a dummy conflict in this
case. We use ".jj-do-not-resolve-this-conflict" as the dummy conflict to
indicate to the user that they should not attempt to resolve this
conflict using Git.
Instead of setting the index to match the tree of HEAD, we now set the
index to the merged parent tree of the working copy commit. This means
that if you edit a merge commit, it will make the Git index look like it
would in the middle of a `git merge` operation (with all of the
successfully-merged files staged in the index).
If there are any 2-sided conflicts in the merged parent tree, then they
will be added to the index as conflicts. Since Git doesn't support
conflicts with more than 2 sides, many-sided conflicts are staged as the
first side of the conflict. The following commit will improve this.
This will give us more fine-grained control over what files we put in
the index, allowing us to create conflicted index states. We also still
need to use git2 to clean up the merge/rebase state, since gix doesn't
have any function for this currently.
These tests should still pass after we switch to gix for resetting the
index, so we need to make sure they don't rely on the cached index from
the `git2::Repository` instance.
There should be no difference, but it's more consistent that all workspace/repo
commands use workspace_command.settings(), etc. than command.settings().
This ensures that WorkspaceCommandHelper created for initialized/cloned repo
refers to the settings object for that repo, not the settings loaded for the
cwd repo.
I also removed WorkspaceCommandEnvironment::settings() because it doesn't own
a workspace object. It's implementation detail that the env object holds a
copy of workspace.settings(), and the caller should be accessible to the
workspace object.
This patch doesn't fix DiffEditor::edit() API, which is fundamentally broken
if matcher argument is specified. I'm not sure if the builtin behavior is
correct or not. Suppose we add "jj diffedit FILESETS..", it would probably make
sense to leave unmatched paths unmodified because it is the command to edit the
destination (or right) tree. This is the edit_diff_external() behavior.
Fixes#5252
Disclaimer: this is the work of @necauqua and @julienvincent (see
#3141). I simply materialized the changes by rebasing them on latest
`main` and making the necessary adjustments to pass CI.
---
I had to fix an issue in `TestSignatureBackend::sign()`.
The following test was failing:
```
---- test_signature_templates::test_signature_templates stdout ----
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Snapshot Summary ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Snapshot: signature_templates
Source: cli/tests/test_signature_templates.rs:28
────────────────────────────────────────────────────────────────────────────────
Expression: stdout
────────────────────────────────────────────────────────────────────────────────
-old snapshot
+new results
────────────┬───────────────────────────────────────────────────────────────────
0 0 │ @ Commit ID: 05ac066d05701071af20e77506a0f2195194cbc9
1 1 │ │ Change ID: qpvuntsmwlqtpsluzzsnyyzlmlwvmlnu
2 2 │ │ Author: Test User <test.user@example.com> (2001-02-03 08:05:07)
3 3 │ │ Committer: Test User <test.user@example.com> (2001-02-03 08:05:07)
4 │-│ Signature: Good test signature
4 │+│ Signature: Bad test signature
5 5 │ │
6 6 │ │ (no description set)
7 7 │ │
8 8 │ ◆ Commit ID: 0000000000000000000000000000000000000000
────────────┴───────────────────────────────────────────────────────────────────
```
Print debugging revealed that the signature was bad, because of a
missing trailing `\n` in `TestSignatureBackend::sign()`.
```diff
diff --git a/lib/src/test_signing_backend.rs b/lib/src/test_signing_backend.rs
index d47fef1086..0ba249e358 100644
--- a/lib/src/test_signing_backend.rs
+++ b/lib/src/test_signing_backend.rs
@@ -59,6 +59,8 @@
let key = (!key.is_empty()).then_some(std::str::from_utf8(key).unwrap().to_owned());
let sig = self.sign(data, key.as_deref())?;
+ dbg!(&std::str::from_utf8(&signature).unwrap());
+ dbg!(&std::str::from_utf8(&sig).unwrap());
if sig == signature {
Ok(Verification::new(
SigStatus::Good,
```
```
[lib/src/test_signing_backend.rs:62:9] &std::str::from_utf8(&signature).unwrap() = \"--- JJ-TEST-SIGNATURE ---\\nKEY: \\n5300977ff3ecda4555bd86d383b070afac7b7459c07f762af918943975394a8261d244629e430c8554258904f16dd9c18d737f8969f2e7d849246db0d93cc004\\n\"
[lib/src/test_signing_backend.rs:63:9] &std::str::from_utf8(&sig).unwrap() = \"--- JJ-TEST-SIGNATURE ---\\nKEY: \\n5300977ff3ecda4555bd86d383b070afac7b7459c07f762af918943975394a8261d244629e430c8554258904f16dd9c18d737f8969f2e7d849246db0d93cc004\"
```
Thankfully, @yuja pointed out that libgit2 appends a trailing newline
(see bfb7613d5d).
Co-authored-by: necauqua <him@necauq.ua>
Co-authored-by: julienvincent <m@julienvincent.io>
If jj is compiled against musl libc, not all name services are available, and
getpwuid() can return null even if the system is configured properly. That's
the problem reported as #5231. Suppose operation.username exists mainly for
logging/tracing purposes, it should be better to include something less reliable
than leaving the field empty.
This patch also removes TODO comment about empty hostname/username. It's
unlikely that the hostname is invalid (as that would cause panic on older jj
versions), and $USER would probably be set on Unix.
The following deprecated commands have been removed for their `jj files`
alternatives:
- `jj cat` was deprecated for `jj file show` in 47bd6f4 and e6c2108.
- `jj chmod` was deprecated for `jj file chmod` in 47bd6f4.
- `jj files` was deprecated for `jj file list` in 5d307e6.
The idea is the same as diff_editor()/selector() API. This object will be passed
in to edit_*description() functions in place of (repo_path, settings) pair.
"ui.editor" isn't specific to editing commit descriptions, but it's mainly used
for that purpose. So I put the wrapper type in description_util.rs.
