Git doesn't show this information in their "diff3" style either, and it
looks a bit strange having two different bracketed sections of text next
to each other, so I think it would be better to remove it.
When we have e.g. a bookmark name or a remote name and we want to
produce a revset from it, we may need to quote and escape it. This
patch implements a function for that.
Perhaps we'll want to more generally be able to format a whole
`RevsetExpression` later.
With this and "git --progress" PR #5519, remote git progress will be displayed.
The sideband callback is disabled in git2 code path if progress is not enabled.
If this were enabled, most "git fetch" tests would fail with git2 due to remote
progress messages. Since there's no git2 API to turn the remote progress off,
and the git2 code path is supposed to be phased out, I just inserted ad-hoc
workaround.
This patch adds thread-based implementation. On Unix, we can poll pipes without
using thread, but I don't think platform-dependent optimization is needed.
* Make the new `GitFetch` api public.
* Move `git::fetch` to `lib/tests/test_git.rs` as `git_fetch`, to minimize
churn in the tests. Update test call sites to use `git_fetch`
* Delete the `git::fetch` from `lib/src/git.rs`.
* Update `jj git clone` and `git_fetch` in `cli/src/git_utils.rs` to use
the new api directly. Removing one redundant layer of indirection.
* This fixes#4920 as it first fetches from all remotes before `import_refs()`
is called, so there is no race condition if the same commit is treated
differently in different remotes specified in the same command.
Original commit by @essiene
A missing newline would cause the following conflict markers to become
invalid when materializing the conflict, so we add a newline to prevent
this from happening. When parsing, if we parsed a conflict at the end of
the file which ended in a newline, we can check whether the file
actually did end in a newline, and then remove the newline if it didn't.
File conflicts have been simplified during materialization since 0.19.0,
so it should be safe to remove support for unsimplified conflicts now.
This will make implementing the next commit easier, since it won't have
to support unsimplified conflicts.
Currently, the Git subprocess tests only work on Linux due to a default
path being used for the `git` executable when `$PATH` is unset. This can
break if Git isn't installed at the expected path. Also, I believe it is
currently necessary to set the `$TEST_GIT_EXECUTABLE_PATH` environment
variable on Windows for tests to pass. Instead, we should use the user's
`$PATH` to locate the `git` executable, as well as any other executables
that are needed. This also makes `$TEST_GIT_EXECUTABLE_PATH` no longer
necessary, so it can be removed.
Closes#5217
Motivating use case:
[[--scope]]
--when.command = ["log"]
[--scope.ui]
pager = "less"
This adds a new (optional) field to `--when` config conditions, to
inspect the current command.
`--when.commands` is a list of space-separated values that matches a
space-separated command. To be specific:
--when.command = ["file"] # matches `jj file show`, `jj file list`, etc
--when.command = ["file show"] # matches `jj file show`, but *NOT* `jj file list`
--when.command = ["file", "log"] # matches `jj file` *OR* `jj log` (or subcommand of either)
When both `--when.commands` and `--when.repositories` are set, the
intersection is used.
In some cases, there are non trivial codepaths for fetching multiple
branches explicitly. In particular, it might be the case that fetching
works for n = 2 but not n = 3.
This commit changes a cli test to have 3 remotes and 3 branches, and a
lib test with 3 branches, only one of them succeds.
Fully qualified git refs (e.g., `+refs/head/main:refs/remotes/origin/main`)
are commonly used throughout the git glue code.
It is laborious and error prone to keep parsing refspecs when we need
only a segment of it.
This commits adds a type, `RefSpec`, which does exactly that
It's not uncommon to insert new table at a certain path. This could be modelled
as set_table(name, new_table), but the caller might want to merge items with
the existing table if any. This function can be used for both cases.
This was needed in order to migrate fix.tool-command to [fix.tools] table. We
don't have other use cases right now, but there will be something in future.
I'll make "jj abandon" delete bookmarks by default. This could be handled by
cmd_abandon(), but we'll need a repo API if we also want to change the behavior
of "jj rebase --skip-emptied".
I noticed that some config structures do not use their "Default" implementation, and have their default specified in a TOML file. I think specifying defaults in TOML is great, and it'd be good to delete those Default implementations, so that it does not diverge from the default in the TOML file.
There's only one non-test caller who doesn't need a complete map of rebased
commits.
FWIW, I tried to rewrite squash_commits() based on transform_descendants() API,
but it wasn't easy because of A-B+A=A rule. The merge order matters here.
Tag and bookmark names are usually ASCII, but they occasionally include Latin
or Han characters.
This doesn't fix the serialization problem, but should mitigate #5359.
The "identifier" rule will be changed to accept unicode characters. It's not
super important to reject non-ASCII string pattern prefixes or alias symbols,
but this is more consistent with templater and fileset. We can relax the rule
later if needed.
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
