ok/jj
1
0
Fork 0
forked from mirrors/jj
Code Pull requests Activity
jj/cli/tests/test_git_push.rs

826 lines
33 KiB
Rust
Raw Normal View History

copyright: change from "Google LLC" to "The Jujutsu Authors" Let's acknowledge everyone's contributions by replacing "Google LLC" in the copyright header by "The Jujutsu Authors". If I understand correctly, it won't have any legal effect, but maybe it still helps reduce concerns from contributors (though I haven't heard any concerns). Google employees can read about Google's policy at go/releasing/contributions#copyright.
2022-11-26 23:57:50 +00:00
// Copyright 2022 The Jujutsu Authors
cli: prevent pushing commits with conflicts It rarely makes sense to push commits with conflicts to a remote Git repo (which is the only kind of remote we support so far), so let's just error out instead of pushing a commit that others pulling from the remote probably can't make sense of. I've only added a simple test for the error case for now. `libgit2` doesn't support pushing to a local repo, so it's harder to test the success case. I suppose we'll have to have the regular `git` binary running local servers in test eventually. Closes #60.
2022-03-09 23:57:12 +00:00
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// https://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
use std::path::PathBuf;
cli: include hint when push is not fast-forward
2023-08-13 22:41:53 +00:00
use crate::common::{get_stderr_string, get_stdout_string, TestEnvironment};
tests: move `testutils` from `src/` to `tests/`
2022-03-30 17:47:11 +00:00
pub mod common;
cli: prevent pushing commits with conflicts It rarely makes sense to push commits with conflicts to a remote Git repo (which is the only kind of remote we support so far), so let's just error out instead of pushing a commit that others pulling from the remote probably can't make sense of. I've only added a simple test for the error case for now. `libgit2` doesn't support pushing to a local repo, so it's harder to test the success case. I suppose we'll have to have the regular `git` binary running local servers in test eventually. Closes #60.
2022-03-09 23:57:12 +00:00
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
fn set_up() -> (TestEnvironment, PathBuf) {
cli: prevent pushing commits with conflicts It rarely makes sense to push commits with conflicts to a remote Git repo (which is the only kind of remote we support so far), so let's just error out instead of pushing a commit that others pulling from the remote probably can't make sense of. I've only added a simple test for the error case for now. `libgit2` doesn't support pushing to a local repo, so it's harder to test the success case. I suppose we'll have to have the regular `git` binary running local servers in test eventually. Closes #60.
2022-03-09 23:57:12 +00:00
let test_env = TestEnvironment::default();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(test_env.env_root(), &["init", "--git", "origin"]);
test_git_push: create Git repo using `jj` commands Since we don't have the `git` binary available in the test environment, it's just easier to use `jj` commands to create Git repos.
2022-11-04 18:25:11 +00:00
let origin_path = test_env.env_root().join("origin");
let origin_git_repo_path = origin_path
.join(".jj")
.join("repo")
.join("store")
.join("git");
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&origin_path, &["describe", "-m=description 1"]);
test_env.jj_cmd_ok(&origin_path, &["branch", "create", "branch1"]);
test_env.jj_cmd_ok(&origin_path, &["new", "root()", "-m=description 2"]);
test_env.jj_cmd_ok(&origin_path, &["branch", "create", "branch2"]);
test_env.jj_cmd_ok(&origin_path, &["git", "export"]);
cli: prevent pushing commits with conflicts It rarely makes sense to push commits with conflicts to a remote Git repo (which is the only kind of remote we support so far), so let's just error out instead of pushing a commit that others pulling from the remote probably can't make sense of. I've only added a simple test for the error case for now. `libgit2` doesn't support pushing to a local repo, so it's harder to test the success case. I suppose we'll have to have the regular `git` binary running local servers in test eventually. Closes #60.
2022-03-09 23:57:12 +00:00
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(
tests: add a convenience function for running `jj` successfully We very often expect success, and we sometimes want to get the stdout, too. Let's add a convenience function for that. It saves a lot of lines of code.
2022-03-26 21:06:48 +00:00
test_env.env_root(),
tests: add test for successful push I think I had not added tests for successful push before because I thought there was some issue with testing it with libgit2. There *is* an issue, which is that libgit2 requires the remote to be bare when it's on local disk, but we can very easily make the git repo bare in this test. I also updated the error handling in the `git` module to not let follow-on errors hide the real error and to not fail if two branches moved to the same commit.
2022-07-02 02:50:51 +00:00
&[
"git",
"clone",
tests: set git_settings.auto_local_branch where it matters
2023-12-16 10:00:14 +00:00
"--config-toml=git.auto-local-branch=true",
test_git_push: create Git repo using `jj` commands Since we don't have the `git` binary available in the test environment, it's just easier to use `jj` commands to create Git repos.
2022-11-04 18:25:11 +00:00
origin_git_repo_path.to_str().unwrap(),
"local",
tests: add test for successful push I think I had not added tests for successful push before because I thought there was some issue with testing it with libgit2. There *is* an issue, which is that libgit2 requires the remote to be bare when it's on local disk, but we can very easily make the git repo bare in this test. I also updated the error handling in the `git` module to not let follow-on errors hide the real error and to not fail if two branches moved to the same commit.
2022-07-02 02:50:51 +00:00
],
tests: add a convenience function for running `jj` successfully We very often expect success, and we sometimes want to get the stdout, too. Let's add a convenience function for that. It saves a lot of lines of code.
2022-03-26 21:06:48 +00:00
);
test_git_push: create Git repo using `jj` commands Since we don't have the `git` binary available in the test environment, it's just easier to use `jj` commands to create Git repos.
2022-11-04 18:25:11 +00:00
let workspace_root = test_env.env_root().join("local");
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
(test_env, workspace_root)
}
cli: prevent pushing commits with conflicts It rarely makes sense to push commits with conflicts to a remote Git repo (which is the only kind of remote we support so far), so let's just error out instead of pushing a commit that others pulling from the remote probably can't make sense of. I've only added a simple test for the error case for now. `libgit2` doesn't support pushing to a local repo, so it's harder to test the success case. I suppose we'll have to have the regular `git` binary running local servers in test eventually. Closes #60.
2022-03-09 23:57:12 +00:00
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
#[test]
fn test_git_push_nothing() {
let (test_env, workspace_root) = set_up();
cli: prevent pushing commits with conflicts It rarely makes sense to push commits with conflicts to a remote Git repo (which is the only kind of remote we support so far), so let's just error out instead of pushing a commit that others pulling from the remote probably can't make sense of. I've only added a simple test for the error case for now. `libgit2` doesn't support pushing to a local repo, so it's harder to test the success case. I suppose we'll have to have the regular `git` binary running local servers in test eventually. Closes #60.
2022-03-09 23:57:12 +00:00
// No branches to push yet
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push", "--all"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
tests: re-run with `insta` crate version b9d99e87065b
2022-04-28 23:32:18 +00:00
Nothing changed.
"###);
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
}
cli: prevent pushing commits with conflicts It rarely makes sense to push commits with conflicts to a remote Git repo (which is the only kind of remote we support so far), so let's just error out instead of pushing a commit that others pulling from the remote probably can't make sense of. I've only added a simple test for the error case for now. `libgit2` doesn't support pushing to a local repo, so it's harder to test the success case. I suppose we'll have to have the regular `git` binary running local servers in test eventually. Closes #60.
2022-03-09 23:57:12 +00:00
tests: add test for successful push I think I had not added tests for successful push before because I thought there was some issue with testing it with libgit2. There *is* an issue, which is that libgit2 requires the remote to be bare when it's on local disk, but we can very easily make the git repo bare in this test. I also updated the error handling in the `git` module to not let follow-on errors hide the real error and to not fail if two branches moved to the same commit.
2022-07-02 02:50:51 +00:00
#[test]
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
fn test_git_push_current_branch() {
let (test_env, workspace_root) = set_up();
cli: make set of immutable commits configurable This adds a new `revset-aliases.immutable_heads()s` config for defining the set of immutable commits. The set is defined as the configured revset, as well as its ancestors, and the root commit commit (even if the configured set is empty). This patch also adds enforcement of the config where we already had checks preventing rewrite of the root commit. The working-copy commit is implicitly assumed to be writable in most cases. Specifically, we won't prevent amending the working copy even if the user includes it in the config but we do prevent `jj edit @` in that case. That seems good enough to me. Maybe we should emit a warning when the working copy is in the set of immutable commits. Maybe we should add support for something more like [Mercurial's phases](https://wiki.mercurial-scm.org/Phases), which is propagated on push and pull. There's already some affordance for that in the view object's `public_heads` field. However, this is simpler, especially since we can't propagate the phase to Git remotes, and seems like a good start. Also, it lets you say that commits authored by other users are immutable, for example. For now, the functionality is in the CLI library. I'm not sure if we want to move it into the library crate. I'm leaning towards letting library users do whatever they want without being restricted by immutable commits. I do think we should move the functionality into a future `ui-lib` or `ui-util` crate. That crate would have most of the functionality in the current `cli_util` module (but in a non-CLI-specific form).
2023-09-02 02:12:01 +00:00
test_env.add_config(r#"revset-aliases."immutable_heads()" = "none()""#);
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
// Update some branches. `branch1` is not a current branch, but `branch2` and
// `my-branch` are.
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
&workspace_root,
&["describe", "branch1", "-m", "modified branch1 commit"],
);
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["co", "branch2"]);
test_env.jj_cmd_ok(&workspace_root, &["branch", "set", "branch2"]);
test_env.jj_cmd_ok(&workspace_root, &["branch", "create", "my-branch"]);
test_env.jj_cmd_ok(&workspace_root, &["describe", "-m", "foo"]);
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
// Check the setup
cli: add "branch list --all" option to include all remote branches This will be the option to include non-tracking remote branches. We could add more fine-grained filtering flags, but I think --all is good enough and easier to remember. This patch also updates many of the test outputs to include synchronized remote branches. I think verbose outputs will help catch future bugs.
2023-10-18 06:12:17 +00:00
let stdout = test_env.jj_cmd_success(&workspace_root, &["branch", "list", "--all"]);
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
insta::assert_snapshot!(stdout, @r###"
cli templates: include change id and more in short commit description Include change id as well as (empty) and (conflict) markers in short commit description. Unlike `jj log`, * (conflict) is put in parentheses to distinguish it from the commit description when the output is not colored. * the change id is placed right before the commit id to put it into a predicteble place, even if the short commit description takes up more than one line. The commit id can be important for commit descriptions in commands like `abandon`.
2023-07-11 01:46:38 +00:00
branch1: lzmmnrxq 19e00bf6 (empty) modified branch1 commit
cli templates: include hidden (AKA abandoned) status in short commit template This is mainly relevant for `jj abandon` and `jj branch list`. #2411
2023-10-24 01:43:12 +00:00
@origin (ahead by 1 commits, behind by 1 commits): lzmmnrxq hidden 45a3aa29 (empty) description 1
cli templates: include change id and more in short commit description Include change id as well as (empty) and (conflict) markers in short commit description. Unlike `jj log`, * (conflict) is put in parentheses to distinguish it from the commit description when the output is not colored. * the change id is placed right before the commit id to put it into a predicteble place, even if the short commit description takes up more than one line. The commit id can be important for commit descriptions in commands like `abandon`.
2023-07-11 01:46:38 +00:00
branch2: yostqsxw 10ee3363 (empty) foo
@origin (behind by 1 commits): rlzusymt 8476341e (empty) description 2
my-branch: yostqsxw 10ee3363 (empty) foo
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
"###);
// First dry-run. `branch1` should not get pushed.
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push", "--dry-run"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
Branch changes to push to origin:
cli: append "\n" to commit description specified by -m/--message Otherwise the description set by -m would differ from the one set by editor. This fixes test_describe() which says "make no changes", but previously "\n" would be added by the second "jj describe". As you can see, almost all hashes change in CLI tests. This means in-flight PRs will need to be rebased to update insta snapshots. Description text could be normalized by CommitBuilder, but the caller would have to normalize it beforehand to compare with the current description, so we would need an explicit function anyway. Another idea is to add a newtype that represents a normalized description, and make CommitBuilder require it. Commit::description() will return &Description in place of &str to ensure that commit.description() == raw_str wouldn't compile. Git CLI provides --cleanup=<mode> option to switch normalization rules, but I don't think we'll need such feature.
2022-12-21 04:47:10 +00:00
Move branch branch2 from 8476341eb395 to 10ee3363b259
Add branch my-branch to 10ee3363b259
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
Dry-run requested, not pushing.
"###);
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
Branch changes to push to origin:
cli: append "\n" to commit description specified by -m/--message Otherwise the description set by -m would differ from the one set by editor. This fixes test_describe() which says "make no changes", but previously "\n" would be added by the second "jj describe". As you can see, almost all hashes change in CLI tests. This means in-flight PRs will need to be rebased to update insta snapshots. Description text could be normalized by CommitBuilder, but the caller would have to normalize it beforehand to compare with the current description, so we would need an explicit function anyway. Another idea is to add a newtype that represents a normalized description, and make CommitBuilder require it. Commit::description() will return &Description in place of &str to ensure that commit.description() == raw_str wouldn't compile. Git CLI provides --cleanup=<mode> option to switch normalization rules, but I don't think we'll need such feature.
2022-12-21 04:47:10 +00:00
Move branch branch2 from 8476341eb395 to 10ee3363b259
Add branch my-branch to 10ee3363b259
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
"###);
cli: add "branch list --all" option to include all remote branches This will be the option to include non-tracking remote branches. We could add more fine-grained filtering flags, but I think --all is good enough and easier to remember. This patch also updates many of the test outputs to include synchronized remote branches. I think verbose outputs will help catch future bugs.
2023-10-18 06:12:17 +00:00
let stdout = test_env.jj_cmd_success(&workspace_root, &["branch", "list", "--all"]);
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
insta::assert_snapshot!(stdout, @r###"
cli templates: include change id and more in short commit description Include change id as well as (empty) and (conflict) markers in short commit description. Unlike `jj log`, * (conflict) is put in parentheses to distinguish it from the commit description when the output is not colored. * the change id is placed right before the commit id to put it into a predicteble place, even if the short commit description takes up more than one line. The commit id can be important for commit descriptions in commands like `abandon`.
2023-07-11 01:46:38 +00:00
branch1: lzmmnrxq 19e00bf6 (empty) modified branch1 commit
cli templates: include hidden (AKA abandoned) status in short commit template This is mainly relevant for `jj abandon` and `jj branch list`. #2411
2023-10-24 01:43:12 +00:00
@origin (ahead by 1 commits, behind by 1 commits): lzmmnrxq hidden 45a3aa29 (empty) description 1
cli templates: include change id and more in short commit description Include change id as well as (empty) and (conflict) markers in short commit description. Unlike `jj log`, * (conflict) is put in parentheses to distinguish it from the commit description when the output is not colored. * the change id is placed right before the commit id to put it into a predicteble place, even if the short commit description takes up more than one line. The commit id can be important for commit descriptions in commands like `abandon`.
2023-07-11 01:46:38 +00:00
branch2: yostqsxw 10ee3363 (empty) foo
cli: add "branch list --all" option to include all remote branches This will be the option to include non-tracking remote branches. We could add more fine-grained filtering flags, but I think --all is good enough and easier to remember. This patch also updates many of the test outputs to include synchronized remote branches. I think verbose outputs will help catch future bugs.
2023-10-18 06:12:17 +00:00
@origin: yostqsxw 10ee3363 (empty) foo
cli templates: include change id and more in short commit description Include change id as well as (empty) and (conflict) markers in short commit description. Unlike `jj log`, * (conflict) is put in parentheses to distinguish it from the commit description when the output is not colored. * the change id is placed right before the commit id to put it into a predicteble place, even if the short commit description takes up more than one line. The commit id can be important for commit descriptions in commands like `abandon`.
2023-07-11 01:46:38 +00:00
my-branch: yostqsxw 10ee3363 (empty) foo
cli: add "branch list --all" option to include all remote branches This will be the option to include non-tracking remote branches. We could add more fine-grained filtering flags, but I think --all is good enough and easier to remember. This patch also updates many of the test outputs to include synchronized remote branches. I think verbose outputs will help catch future bugs.
2023-10-18 06:12:17 +00:00
@origin: yostqsxw 10ee3363 (empty) foo
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
"###);
}
cli: search @- for branches to push when @ has none
2022-10-20 21:18:34 +00:00
#[test]
fn test_git_push_parent_branch() {
let (test_env, workspace_root) = set_up();
cli: make set of immutable commits configurable This adds a new `revset-aliases.immutable_heads()s` config for defining the set of immutable commits. The set is defined as the configured revset, as well as its ancestors, and the root commit commit (even if the configured set is empty). This patch also adds enforcement of the config where we already had checks preventing rewrite of the root commit. The working-copy commit is implicitly assumed to be writable in most cases. Specifically, we won't prevent amending the working copy even if the user includes it in the config but we do prevent `jj edit @` in that case. That seems good enough to me. Maybe we should emit a warning when the working copy is in the set of immutable commits. Maybe we should add support for something more like [Mercurial's phases](https://wiki.mercurial-scm.org/Phases), which is propagated on push and pull. There's already some affordance for that in the view object's `public_heads` field. However, this is simpler, especially since we can't propagate the phase to Git remotes, and seems like a good start. Also, it lets you say that commits authored by other users are immutable, for example. For now, the functionality is in the CLI library. I'm not sure if we want to move it into the library crate. I'm leaning towards letting library users do whatever they want without being restricted by immutable commits. I do think we should move the functionality into a future `ui-lib` or `ui-util` crate. That crate would have most of the functionality in the current `cli_util` module (but in a non-CLI-specific form).
2023-09-02 02:12:01 +00:00
test_env.add_config(r#"revset-aliases."immutable_heads()" = "none()""#);
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["edit", "branch1"]);
test_env.jj_cmd_ok(
cli: search @- for branches to push when @ has none
2022-10-20 21:18:34 +00:00
&workspace_root,
&["describe", "-m", "modified branch1 commit"],
);
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["new", "-m", "non-empty description"]);
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
std::fs::write(workspace_root.join("file"), "file").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
cli: search @- for branches to push when @ has none
2022-10-20 21:18:34 +00:00
Branch changes to push to origin:
cli: append "\n" to commit description specified by -m/--message Otherwise the description set by -m would differ from the one set by editor. This fixes test_describe() which says "make no changes", but previously "\n" would be added by the second "jj describe". As you can see, almost all hashes change in CLI tests. This means in-flight PRs will need to be rebased to update insta snapshots. Description text could be normalized by CommitBuilder, but the caller would have to normalize it beforehand to compare with the current description, so we would need an explicit function anyway. Another idea is to add a newtype that represents a normalized description, and make CommitBuilder require it. Commit::description() will return &Description in place of &str to ensure that commit.description() == raw_str wouldn't compile. Git CLI provides --cleanup=<mode> option to switch normalization rules, but I don't think we'll need such feature.
2022-12-21 04:47:10 +00:00
Force branch branch1 from 45a3aa29e907 to d47326d59ee1
cli: search @- for branches to push when @ has none
2022-10-20 21:18:34 +00:00
"###);
}
git push: do not consider @- if @ has non-empty content or description
2023-03-05 20:27:03 +00:00
#[test]
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
fn test_git_push_no_matching_branch() {
git push: do not consider @- if @ has non-empty content or description
2023-03-05 20:27:03 +00:00
let (test_env, workspace_root) = set_up();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["new"]);
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
git push: do not consider @- if @ has non-empty content or description
2023-03-05 20:27:03 +00:00
insta::assert_snapshot!(stderr, @r###"
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
No branches point to the specified revisions.
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
Nothing changed.
git push: do not consider @- if @ has non-empty content or description
2023-03-05 20:27:03 +00:00
"###);
}
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
#[test]
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
fn test_git_push_matching_branch_unchanged() {
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
let (test_env, workspace_root) = set_up();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["co", "branch1"]);
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
insta::assert_snapshot!(stderr, @r###"
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
No branches point to the specified revisions.
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
Nothing changed.
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
"###);
}
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
/// Test that `jj git push` without arguments pushes a branch to the specified
/// remote even if it's already up to date on another remote
/// (`remote_branches(remote=<remote>)..@` vs. `remote_branches()..@`).
cli: report no updates to current branch Fix a bug where `jj git push` would print "No current branch." when there is a current branch but it is unchanged. We were conflating the two because we print the message when no updates were performed, instead of only when no branches were found.
2023-01-30 03:46:29 +00:00
#[test]
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
fn test_git_push_other_remote_has_branch() {
cli: report no updates to current branch Fix a bug where `jj git push` would print "No current branch." when there is a current branch but it is unchanged. We were conflating the two because we print the message when no updates were performed, instead of only when no branches were found.
2023-01-30 03:46:29 +00:00
let (test_env, workspace_root) = set_up();
cli: make set of immutable commits configurable This adds a new `revset-aliases.immutable_heads()s` config for defining the set of immutable commits. The set is defined as the configured revset, as well as its ancestors, and the root commit commit (even if the configured set is empty). This patch also adds enforcement of the config where we already had checks preventing rewrite of the root commit. The working-copy commit is implicitly assumed to be writable in most cases. Specifically, we won't prevent amending the working copy even if the user includes it in the config but we do prevent `jj edit @` in that case. That seems good enough to me. Maybe we should emit a warning when the working copy is in the set of immutable commits. Maybe we should add support for something more like [Mercurial's phases](https://wiki.mercurial-scm.org/Phases), which is propagated on push and pull. There's already some affordance for that in the view object's `public_heads` field. However, this is simpler, especially since we can't propagate the phase to Git remotes, and seems like a good start. Also, it lets you say that commits authored by other users are immutable, for example. For now, the functionality is in the CLI library. I'm not sure if we want to move it into the library crate. I'm leaning towards letting library users do whatever they want without being restricted by immutable commits. I do think we should move the functionality into a future `ui-lib` or `ui-util` crate. That crate would have most of the functionality in the current `cli_util` module (but in a non-CLI-specific form).
2023-09-02 02:12:01 +00:00
test_env.add_config(r#"revset-aliases."immutable_heads()" = "none()""#);
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
// Create another remote (but actually the same)
let other_remote_path = test_env
.env_root()
.join("origin")
.join(".jj")
.join("repo")
.join("store")
.join("git");
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
&workspace_root,
&[
"git",
"remote",
"add",
"other",
other_remote_path.to_str().unwrap(),
],
);
// Modify branch1 and push it to `origin`
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["edit", "branch1"]);
test_env.jj_cmd_ok(&workspace_root, &["describe", "-m=modified"]);
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
Branch changes to push to origin:
Force branch branch1 from 45a3aa29e907 to 50421a29358a
"###);
// Since it's already pushed to origin, nothing will happen if push again
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
insta::assert_snapshot!(stderr, @r###"
No branches point to the specified revisions.
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
Nothing changed.
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
"###);
// But it will still get pushed to another remote
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push", "--remote=other"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
Branch changes to push to other:
Add branch branch1 to 50421a29358a
"###);
cli: report no updates to current branch Fix a bug where `jj git push` would print "No current branch." when there is a current branch but it is unchanged. We were conflating the two because we print the message when no updates were performed, instead of only when no branches were found.
2023-01-30 03:46:29 +00:00
}
cli: include hint when push is not fast-forward
2023-08-13 22:41:53 +00:00
#[test]
fn test_git_push_not_fast_forward() {
let (test_env, workspace_root) = set_up();
// Move branch1 forward on the remote
let origin_path = test_env.env_root().join("origin");
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&origin_path, &["new", "branch1", "-m=remote"]);
cli: include hint when push is not fast-forward
2023-08-13 22:41:53 +00:00
std::fs::write(origin_path.join("remote"), "remote").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&origin_path, &["branch", "set", "branch1"]);
test_env.jj_cmd_ok(&origin_path, &["git", "export"]);
cli: include hint when push is not fast-forward
2023-08-13 22:41:53 +00:00
// Move branch1 forward to another commit locally
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["new", "branch1", "-m=local"]);
cli: include hint when push is not fast-forward
2023-08-13 22:41:53 +00:00
std::fs::write(workspace_root.join("local"), "local").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["branch", "set", "branch1"]);
cli: include hint when push is not fast-forward
2023-08-13 22:41:53 +00:00
// Pushing should fail
let assert = test_env
.jj_cmd(&workspace_root, &["git", "push"])
.assert()
.code(1);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(get_stdout_string(&assert), @"");
insta::assert_snapshot!(get_stderr_string(&assert), @r###"
cli: include hint when push is not fast-forward
2023-08-13 22:41:53 +00:00
Branch changes to push to origin:
Move branch branch1 from 45a3aa29e907 to c35839cb8e8c
Error: The push conflicts with changes made on the remote (it is not fast-forwardable).
Hint: Try fetching from the remote, then make the branch point to where you want it to be, and push again.
"###);
}
refs: classify push action based on tracking target Although this is logically correct, the error message is a bit cryptic. It's probably better to reject push if non-tracking remote branches exist. #1136
2023-10-14 14:54:35 +00:00
#[test]
fn test_git_push_locally_created_and_rewritten() {
let (test_env, workspace_root) = set_up();
// Ensure that remote branches aren't tracked automatically
test_env.add_config("git.auto-local-branch = false");
// Push locally-created branch
test_env.jj_cmd_ok(&workspace_root, &["new", "root()", "-mlocal 1"]);
tests: don't create new branch by "jj branch set"
2023-11-10 00:56:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["branch", "create", "my"]);
refs: classify push action based on tracking target Although this is logically correct, the error message is a bit cryptic. It's probably better to reject push if non-tracking remote branches exist. #1136
2023-10-14 14:54:35 +00:00
let (_stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push"]);
insta::assert_snapshot!(stderr, @r###"
Branch changes to push to origin:
Add branch my to fcc999921ce9
"###);
// Rewrite it and push again, which would fail if the pushed branch weren't
// set to "tracking"
test_env.jj_cmd_ok(&workspace_root, &["describe", "-mlocal 2"]);
cli: add "branch list --all" option to include all remote branches This will be the option to include non-tracking remote branches. We could add more fine-grained filtering flags, but I think --all is good enough and easier to remember. This patch also updates many of the test outputs to include synchronized remote branches. I think verbose outputs will help catch future bugs.
2023-10-18 06:12:17 +00:00
let stdout = test_env.jj_cmd_success(&workspace_root, &["branch", "list", "--all"]);
refs: classify push action based on tracking target Although this is logically correct, the error message is a bit cryptic. It's probably better to reject push if non-tracking remote branches exist. #1136
2023-10-14 14:54:35 +00:00
insta::assert_snapshot!(stdout, @r###"
branch1: lzmmnrxq 45a3aa29 (empty) description 1
cli: add "branch list --all" option to include all remote branches This will be the option to include non-tracking remote branches. We could add more fine-grained filtering flags, but I think --all is good enough and easier to remember. This patch also updates many of the test outputs to include synchronized remote branches. I think verbose outputs will help catch future bugs.
2023-10-18 06:12:17 +00:00
@origin: lzmmnrxq 45a3aa29 (empty) description 1
refs: classify push action based on tracking target Although this is logically correct, the error message is a bit cryptic. It's probably better to reject push if non-tracking remote branches exist. #1136
2023-10-14 14:54:35 +00:00
branch2: rlzusymt 8476341e (empty) description 2
cli: add "branch list --all" option to include all remote branches This will be the option to include non-tracking remote branches. We could add more fine-grained filtering flags, but I think --all is good enough and easier to remember. This patch also updates many of the test outputs to include synchronized remote branches. I think verbose outputs will help catch future bugs.
2023-10-18 06:12:17 +00:00
@origin: rlzusymt 8476341e (empty) description 2
refs: classify push action based on tracking target Although this is logically correct, the error message is a bit cryptic. It's probably better to reject push if non-tracking remote branches exist. #1136
2023-10-14 14:54:35 +00:00
my: vruxwmqv bde1d2e4 (empty) local 2
cli templates: include hidden (AKA abandoned) status in short commit template This is mainly relevant for `jj abandon` and `jj branch list`. #2411
2023-10-24 01:43:12 +00:00
@origin (ahead by 1 commits, behind by 1 commits): vruxwmqv hidden fcc99992 (empty) local 1
refs: classify push action based on tracking target Although this is logically correct, the error message is a bit cryptic. It's probably better to reject push if non-tracking remote branches exist. #1136
2023-10-14 14:54:35 +00:00
"###);
let (_stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push"]);
insta::assert_snapshot!(stderr, @r###"
Branch changes to push to origin:
Force branch my from fcc999921ce9 to bde1d2e44b2a
"###);
}
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
#[test]
Allow `jj git push` to push multiple branches/changes at once Also creates short arg `-b` for `--branch`.
2022-12-17 22:31:12 +00:00
fn test_git_push_multiple() {
tests: add test for successful push I think I had not added tests for successful push before because I thought there was some issue with testing it with libgit2. There *is* an issue, which is that libgit2 requires the remote to be bare when it's on local disk, but we can very easily make the git repo bare in this test. I also updated the error handling in the `git` module to not let follow-on errors hide the real error and to not fail if two branches moved to the same commit.
2022-07-02 02:50:51 +00:00
let (test_env, workspace_root) = set_up();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["branch", "delete", "branch1"]);
test_env.jj_cmd_ok(
tests: add test for successful push I think I had not added tests for successful push before because I thought there was some issue with testing it with libgit2. There *is* an issue, which is that libgit2 requires the remote to be bare when it's on local disk, but we can very easily make the git repo bare in this test. I also updated the error handling in the `git` module to not let follow-on errors hide the real error and to not fail if two branches moved to the same commit.
2022-07-02 02:50:51 +00:00
&workspace_root,
&["branch", "set", "--allow-backwards", "branch2"],
);
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["branch", "create", "my-branch"]);
test_env.jj_cmd_ok(&workspace_root, &["describe", "-m", "foo"]);
tests: add test for successful push I think I had not added tests for successful push before because I thought there was some issue with testing it with libgit2. There *is* an issue, which is that libgit2 requires the remote to be bare when it's on local disk, but we can very easily make the git repo bare in this test. I also updated the error handling in the `git` module to not let follow-on errors hide the real error and to not fail if two branches moved to the same commit.
2022-07-02 02:50:51 +00:00
// Check the setup
cli: add "branch list --all" option to include all remote branches This will be the option to include non-tracking remote branches. We could add more fine-grained filtering flags, but I think --all is good enough and easier to remember. This patch also updates many of the test outputs to include synchronized remote branches. I think verbose outputs will help catch future bugs.
2023-10-18 06:12:17 +00:00
let stdout = test_env.jj_cmd_success(&workspace_root, &["branch", "list", "--all"]);
tests: add test for successful push I think I had not added tests for successful push before because I thought there was some issue with testing it with libgit2. There *is* an issue, which is that libgit2 requires the remote to be bare when it's on local disk, but we can very easily make the git repo bare in this test. I also updated the error handling in the `git` module to not let follow-on errors hide the real error and to not fail if two branches moved to the same commit.
2022-07-02 02:50:51 +00:00
insta::assert_snapshot!(stdout, @r###"
branch1 (deleted)
cli templates: include change id and more in short commit description Include change id as well as (empty) and (conflict) markers in short commit description. Unlike `jj log`, * (conflict) is put in parentheses to distinguish it from the commit description when the output is not colored. * the change id is placed right before the commit id to put it into a predicteble place, even if the short commit description takes up more than one line. The commit id can be important for commit descriptions in commands like `abandon`.
2023-07-11 01:46:38 +00:00
@origin: lzmmnrxq 45a3aa29 (empty) description 1
cli: remove weird line break from the "jj branch forget" hint The message seems a bit too long, but it doesn't make sense to insert \n in the middle of sentence.
2023-12-15 12:50:35 +00:00
(this branch will be *deleted permanently* on the remote on the next `jj git push`. Use `jj branch forget` to prevent this)
cli templates: include change id and more in short commit description Include change id as well as (empty) and (conflict) markers in short commit description. Unlike `jj log`, * (conflict) is put in parentheses to distinguish it from the commit description when the output is not colored. * the change id is placed right before the commit id to put it into a predicteble place, even if the short commit description takes up more than one line. The commit id can be important for commit descriptions in commands like `abandon`.
2023-07-11 01:46:38 +00:00
branch2: yqosqzyt 15dcdaa4 (empty) foo
@origin (ahead by 1 commits, behind by 1 commits): rlzusymt 8476341e (empty) description 2
my-branch: yqosqzyt 15dcdaa4 (empty) foo
tests: add test for successful push I think I had not added tests for successful push before because I thought there was some issue with testing it with libgit2. There *is* an issue, which is that libgit2 requires the remote to be bare when it's on local disk, but we can very easily make the git repo bare in this test. I also updated the error handling in the `git` module to not let follow-on errors hide the real error and to not fail if two branches moved to the same commit.
2022-07-02 02:50:51 +00:00
"###);
cli: add a `--dry-run` for `jj git push`
2022-06-29 02:10:45 +00:00
// First dry-run
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) =
test_env.jj_cmd_ok(&workspace_root, &["git", "push", "--all", "--dry-run"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
cli: add a `--dry-run` for `jj git push`
2022-06-29 02:10:45 +00:00
Branch changes to push to origin:
cli: append "\n" to commit description specified by -m/--message Otherwise the description set by -m would differ from the one set by editor. This fixes test_describe() which says "make no changes", but previously "\n" would be added by the second "jj describe". As you can see, almost all hashes change in CLI tests. This means in-flight PRs will need to be rebased to update insta snapshots. Description text could be normalized by CommitBuilder, but the caller would have to normalize it beforehand to compare with the current description, so we would need an explicit function anyway. Another idea is to add a newtype that represents a normalized description, and make CommitBuilder require it. Commit::description() will return &Description in place of &str to ensure that commit.description() == raw_str wouldn't compile. Git CLI provides --cleanup=<mode> option to switch normalization rules, but I don't think we'll need such feature.
2022-12-21 04:47:10 +00:00
Delete branch branch1 from 45a3aa29e907
Force branch branch2 from 8476341eb395 to 15dcdaa4f12f
Add branch my-branch to 15dcdaa4f12f
cli: add a `--dry-run` for `jj git push`
2022-06-29 02:10:45 +00:00
Dry-run requested, not pushing.
"###);
Allow `jj git push` to push multiple branches/changes at once Also creates short arg `-b` for `--branch`.
2022-12-17 22:31:12 +00:00
// Dry run requesting two specific branches
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(
Allow `jj git push` to push multiple branches/changes at once Also creates short arg `-b` for `--branch`.
2022-12-17 22:31:12 +00:00
&workspace_root,
&["git", "push", "-b=branch1", "-b=my-branch", "--dry-run"],
);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
Allow `jj git push` to push multiple branches/changes at once Also creates short arg `-b` for `--branch`.
2022-12-17 22:31:12 +00:00
Branch changes to push to origin:
cli: append "\n" to commit description specified by -m/--message Otherwise the description set by -m would differ from the one set by editor. This fixes test_describe() which says "make no changes", but previously "\n" would be added by the second "jj describe". As you can see, almost all hashes change in CLI tests. This means in-flight PRs will need to be rebased to update insta snapshots. Description text could be normalized by CommitBuilder, but the caller would have to normalize it beforehand to compare with the current description, so we would need an explicit function anyway. Another idea is to add a newtype that represents a normalized description, and make CommitBuilder require it. Commit::description() will return &Description in place of &str to ensure that commit.description() == raw_str wouldn't compile. Git CLI provides --cleanup=<mode> option to switch normalization rules, but I don't think we'll need such feature.
2022-12-21 04:47:10 +00:00
Delete branch branch1 from 45a3aa29e907
Add branch my-branch to 15dcdaa4f12f
Allow `jj git push` to push multiple branches/changes at once Also creates short arg `-b` for `--branch`.
2022-12-17 22:31:12 +00:00
Dry-run requested, not pushing.
"###);
jj git push: Make sure each branch is processed at most once
2022-12-19 02:25:04 +00:00
// Dry run requesting two specific branches twice
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(
jj git push: Make sure each branch is processed at most once
2022-12-19 02:25:04 +00:00
&workspace_root,
&[
"git",
"push",
"-b=branch1",
"-b=my-branch",
"-b=branch1",
cli: add string pattern support to "git push --branch" Since "jj git fetch --branch" supports glob patterns, users would expect that "jj git push --branch glob:.." also works. The error handling bits are copied from "branch" sub commands. We might want to extract it to a common helper function, but I haven't figured out a reasonable boundary point yet.
2023-10-24 06:31:26 +00:00
"-b=glob:my-*",
jj git push: Make sure each branch is processed at most once
2022-12-19 02:25:04 +00:00
"--dry-run",
],
);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
jj git push: Make sure each branch is processed at most once
2022-12-19 02:25:04 +00:00
Branch changes to push to origin:
cli: append "\n" to commit description specified by -m/--message Otherwise the description set by -m would differ from the one set by editor. This fixes test_describe() which says "make no changes", but previously "\n" would be added by the second "jj describe". As you can see, almost all hashes change in CLI tests. This means in-flight PRs will need to be rebased to update insta snapshots. Description text could be normalized by CommitBuilder, but the caller would have to normalize it beforehand to compare with the current description, so we would need an explicit function anyway. Another idea is to add a newtype that represents a normalized description, and make CommitBuilder require it. Commit::description() will return &Description in place of &str to ensure that commit.description() == raw_str wouldn't compile. Git CLI provides --cleanup=<mode> option to switch normalization rules, but I don't think we'll need such feature.
2022-12-21 04:47:10 +00:00
Delete branch branch1 from 45a3aa29e907
Add branch my-branch to 15dcdaa4f12f
jj git push: Make sure each branch is processed at most once
2022-12-19 02:25:04 +00:00
Dry-run requested, not pushing.
"###);
cli: add string pattern support to "git push --branch" Since "jj git fetch --branch" supports glob patterns, users would expect that "jj git push --branch glob:.." also works. The error handling bits are copied from "branch" sub commands. We might want to extract it to a common helper function, but I haven't figured out a reasonable boundary point yet.
2023-10-24 06:31:26 +00:00
// Dry run with glob pattern
let (stdout, stderr) = test_env.jj_cmd_ok(
&workspace_root,
&["git", "push", "-b=glob:branch?", "--dry-run"],
);
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
Branch changes to push to origin:
Delete branch branch1 from 45a3aa29e907
Force branch branch2 from 8476341eb395 to 15dcdaa4f12f
Dry-run requested, not pushing.
"###);
// Unmatched branch name is error
let stderr = test_env.jj_cmd_failure(&workspace_root, &["git", "push", "-b=foo"]);
insta::assert_snapshot!(stderr, @r###"
Error: No such branch: foo
"###);
let stderr = test_env.jj_cmd_failure(
&workspace_root,
&["git", "push", "-b=foo", "-b=glob:?branch"],
);
insta::assert_snapshot!(stderr, @r###"
Error: No matching branches for patterns: foo, ?branch
"###);
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push", "--all"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
cli: make `jj git push` print what it's going to push It's convenient to push all changed branches every time with `jj git push`, but sometimes I want to know which branches were actually pushed. This make the command print what it's going to do. I'll add a `--dry-run` mode and tests next.
2022-06-29 02:10:45 +00:00
Branch changes to push to origin:
cli: append "\n" to commit description specified by -m/--message Otherwise the description set by -m would differ from the one set by editor. This fixes test_describe() which says "make no changes", but previously "\n" would be added by the second "jj describe". As you can see, almost all hashes change in CLI tests. This means in-flight PRs will need to be rebased to update insta snapshots. Description text could be normalized by CommitBuilder, but the caller would have to normalize it beforehand to compare with the current description, so we would need an explicit function anyway. Another idea is to add a newtype that represents a normalized description, and make CommitBuilder require it. Commit::description() will return &Description in place of &str to ensure that commit.description() == raw_str wouldn't compile. Git CLI provides --cleanup=<mode> option to switch normalization rules, but I don't think we'll need such feature.
2022-12-21 04:47:10 +00:00
Delete branch branch1 from 45a3aa29e907
Force branch branch2 from 8476341eb395 to 15dcdaa4f12f
Add branch my-branch to 15dcdaa4f12f
cli: make `jj git push` print what it's going to push It's convenient to push all changed branches every time with `jj git push`, but sometimes I want to know which branches were actually pushed. This make the command print what it's going to do. I'll add a `--dry-run` mode and tests next.
2022-06-29 02:10:45 +00:00
"###);
cli: add "branch list --all" option to include all remote branches This will be the option to include non-tracking remote branches. We could add more fine-grained filtering flags, but I think --all is good enough and easier to remember. This patch also updates many of the test outputs to include synchronized remote branches. I think verbose outputs will help catch future bugs.
2023-10-18 06:12:17 +00:00
let stdout = test_env.jj_cmd_success(&workspace_root, &["branch", "list", "--all"]);
tests: add test for successful push I think I had not added tests for successful push before because I thought there was some issue with testing it with libgit2. There *is* an issue, which is that libgit2 requires the remote to be bare when it's on local disk, but we can very easily make the git repo bare in this test. I also updated the error handling in the `git` module to not let follow-on errors hide the real error and to not fail if two branches moved to the same commit.
2022-07-02 02:50:51 +00:00
insta::assert_snapshot!(stdout, @r###"
cli templates: include change id and more in short commit description Include change id as well as (empty) and (conflict) markers in short commit description. Unlike `jj log`, * (conflict) is put in parentheses to distinguish it from the commit description when the output is not colored. * the change id is placed right before the commit id to put it into a predicteble place, even if the short commit description takes up more than one line. The commit id can be important for commit descriptions in commands like `abandon`.
2023-07-11 01:46:38 +00:00
branch2: yqosqzyt 15dcdaa4 (empty) foo
cli: add "branch list --all" option to include all remote branches This will be the option to include non-tracking remote branches. We could add more fine-grained filtering flags, but I think --all is good enough and easier to remember. This patch also updates many of the test outputs to include synchronized remote branches. I think verbose outputs will help catch future bugs.
2023-10-18 06:12:17 +00:00
@origin: yqosqzyt 15dcdaa4 (empty) foo
cli templates: include change id and more in short commit description Include change id as well as (empty) and (conflict) markers in short commit description. Unlike `jj log`, * (conflict) is put in parentheses to distinguish it from the commit description when the output is not colored. * the change id is placed right before the commit id to put it into a predicteble place, even if the short commit description takes up more than one line. The commit id can be important for commit descriptions in commands like `abandon`.
2023-07-11 01:46:38 +00:00
my-branch: yqosqzyt 15dcdaa4 (empty) foo
cli: add "branch list --all" option to include all remote branches This will be the option to include non-tracking remote branches. We could add more fine-grained filtering flags, but I think --all is good enough and easier to remember. This patch also updates many of the test outputs to include synchronized remote branches. I think verbose outputs will help catch future bugs.
2023-10-18 06:12:17 +00:00
@origin: yqosqzyt 15dcdaa4 (empty) foo
tests: add test for successful push I think I had not added tests for successful push before because I thought there was some issue with testing it with libgit2. There *is* an issue, which is that libgit2 requires the remote to be bare when it's on local disk, but we can very easily make the git repo bare in this test. I also updated the error handling in the `git` module to not let follow-on errors hide the real error and to not fail if two branches moved to the same commit.
2022-07-02 02:50:51 +00:00
"###);
git: on push, update jj's view of remote branches without using import_refs() This means that the commits previously pinned by remote branches are no longer abandoned. I think that's more correct since "push" is the operation to propagate local view to remote, and uninteresting commits should have been locally abandoned.
2023-10-14 19:27:44 +00:00
let stdout = test_env.jj_cmd_success(&workspace_root, &["log", "-rall()"]);
insta::assert_snapshot!(stdout, @r###"
@ yqosqzyt test.user@example.com 2001-02-03 04:05:17.000 +07:00 branch2 my-branch 15dcdaa4
│ (empty) foo
│ ◉ rlzusymt test.user@example.com 2001-02-03 04:05:10.000 +07:00 8476341e
├─╯ (empty) description 2
│ ◉ lzmmnrxq test.user@example.com 2001-02-03 04:05:08.000 +07:00 45a3aa29
├─╯ (empty) description 1
◉ zzzzzzzz root() 00000000
"###);
tests: add test for successful push I think I had not added tests for successful push before because I thought there was some issue with testing it with libgit2. There *is* an issue, which is that libgit2 requires the remote to be bare when it's on local disk, but we can very easily make the git repo bare in this test. I also updated the error handling in the `git` module to not let follow-on errors hide the real error and to not fail if two branches moved to the same commit.
2022-07-02 02:50:51 +00:00
}
Add a test for `git push --change`
2022-12-17 22:31:12 +00:00
#[test]
fn test_git_push_changes() {
let (test_env, workspace_root) = set_up();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["describe", "-m", "foo"]);
Add a test for `git push --change`
2022-12-17 22:31:12 +00:00
std::fs::write(workspace_root.join("file"), "contents").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["new", "-m", "bar"]);
Add a test for `git push --change`
2022-12-17 22:31:12 +00:00
std::fs::write(workspace_root.join("file"), "modified").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push", "--change", "@"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
cli: use reverse-alphabet hex for branch name in `jj git push --change`
2023-02-12 19:49:05 +00:00
Creating branch push-yostqsxwqrlt for revision @
Add a test for `git push --change`
2022-12-17 22:31:12 +00:00
Branch changes to push to origin:
cli: use reverse-alphabet hex for branch name in `jj git push --change`
2023-02-12 19:49:05 +00:00
Add branch push-yostqsxwqrlt to 28d7620ea63a
Add a test for `git push --change`
2022-12-17 22:31:12 +00:00
"###);
Allow `jj git push` to push multiple branches/changes at once Also creates short arg `-b` for `--branch`.
2022-12-17 22:31:12 +00:00
// test pushing two changes at once
std::fs::write(workspace_root.join("file"), "modified2").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push", "-c=@", "-c=@-"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
cli: use reverse-alphabet hex for branch name in `jj git push --change`
2023-02-12 19:49:05 +00:00
Creating branch push-yqosqzytrlsw for revision @-
Allow `jj git push` to push multiple branches/changes at once Also creates short arg `-b` for `--branch`.
2022-12-17 22:31:12 +00:00
Branch changes to push to origin:
cli: use reverse-alphabet hex for branch name in `jj git push --change`
2023-02-12 19:49:05 +00:00
Force branch push-yostqsxwqrlt from 28d7620ea63a to 48d8c7948133
Add branch push-yqosqzytrlsw to fa16a14170fb
Allow `jj git push` to push multiple branches/changes at once Also creates short arg `-b` for `--branch`.
2022-12-17 22:31:12 +00:00
"###);
jj git push: Make sure each branch is processed at most once
2022-12-19 02:25:04 +00:00
// specifying the same change twice doesn't break things
std::fs::write(workspace_root.join("file"), "modified3").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push", "-c=@", "-c=@"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
jj git push: Make sure each branch is processed at most once
2022-12-19 02:25:04 +00:00
Branch changes to push to origin:
cli: use reverse-alphabet hex for branch name in `jj git push --change`
2023-02-12 19:49:05 +00:00
Force branch push-yostqsxwqrlt from 48d8c7948133 to b5f030322b1d
Shorten the git push branch when possible using the short change ID hash.
2023-01-11 18:30:31 +00:00
"###);
test_git_push: Test `git.push-branch-prefix` config
2023-08-07 22:41:42 +00:00
// Test changing `git.push-branch-prefix`. It causes us to push again.
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(
test_git_push: Test `git.push-branch-prefix` config
2023-08-07 22:41:42 +00:00
&workspace_root,
&[
"git",
"push",
"--config-toml",
r"git.push-branch-prefix='test-'",
"--change=@",
],
);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
test_git_push: Test `git.push-branch-prefix` config
2023-08-07 22:41:42 +00:00
Creating branch test-yostqsxwqrlt for revision @
Branch changes to push to origin:
Add branch test-yostqsxwqrlt to b5f030322b1d
"###);
Shorten the git push branch when possible using the short change ID hash.
2023-01-11 18:30:31 +00:00
}
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
#[test]
fn test_git_push_revisions() {
let (test_env, workspace_root) = set_up();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["describe", "-m", "foo"]);
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
std::fs::write(workspace_root.join("file"), "contents").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["new", "-m", "bar"]);
test_env.jj_cmd_ok(&workspace_root, &["branch", "create", "branch-1"]);
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
std::fs::write(workspace_root.join("file"), "modified").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["new", "-m", "baz"]);
test_env.jj_cmd_ok(&workspace_root, &["branch", "create", "branch-2a"]);
test_env.jj_cmd_ok(&workspace_root, &["branch", "create", "branch-2b"]);
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
std::fs::write(workspace_root.join("file"), "modified again").unwrap();
// Push an empty set
let stderr = test_env.jj_cmd_failure(&workspace_root, &["git", "push", "-r=none()"]);
insta::assert_snapshot!(stderr, @r###"
Error: Empty revision set
"###);
// Push a revision with no branches
cli: make `jj git push -r` just warn if no branches targeted If there are branches in the revset that don't need to be pushed because they already match the destination, we currently just print `Nothing changed.` It seems consistent with that to also treat it as success if there are no branches in the specified set to start with. This patch makes the command print a warning in that case instead.
2023-08-22 05:34:58 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push", "-r=@--"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
insta::assert_snapshot!(stderr, @r###"
cli: make `jj git push -r` just warn if no branches targeted If there are branches in the revset that don't need to be pushed because they already match the destination, we currently just print `Nothing changed.` It seems consistent with that to also treat it as success if there are no branches in the specified set to start with. This patch makes the command print a warning in that case instead.
2023-08-22 05:34:58 +00:00
No branches point to the specified revisions.
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
Nothing changed.
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
"###);
// Push a revision with a single branch
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) =
test_env.jj_cmd_ok(&workspace_root, &["git", "push", "-r=@-", "--dry-run"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
Branch changes to push to origin:
Add branch branch-1 to 7decc7932d9c
Dry-run requested, not pushing.
"###);
// Push multiple revisions of which some have branches
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
&workspace_root,
&["git", "push", "-r=@--", "-r=@-", "--dry-run"],
);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
Branch changes to push to origin:
Add branch branch-1 to 7decc7932d9c
Dry-run requested, not pushing.
"###);
// Push a revision with a multiple branches
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) =
test_env.jj_cmd_ok(&workspace_root, &["git", "push", "-r=@", "--dry-run"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
Branch changes to push to origin:
Add branch branch-2a to 1b45449e18d0
Add branch branch-2b to 1b45449e18d0
Dry-run requested, not pushing.
"###);
// Repeating a commit doesn't result in repeated messages about the branch
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
&workspace_root,
&["git", "push", "-r=@-", "-r=@-", "--dry-run"],
);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
Branch changes to push to origin:
Add branch branch-1 to 7decc7932d9c
Dry-run requested, not pushing.
"###);
}
cli: allow `jj git push --change xyz --branch foo` This was pretty simple. I simplified a bit by making the transaction description mention only branches, not changes. It still mentions the branches created for the changes, however. Also, since the operation "tags" contain the full command line, I think it'll still be relatively easy for the user to understand what the operation was about.
2023-06-03 06:24:28 +00:00
#[test]
fn test_git_push_mixed() {
let (test_env, workspace_root) = set_up();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["describe", "-m", "foo"]);
cli: allow `jj git push --change xyz --branch foo` This was pretty simple. I simplified a bit by making the transaction description mention only branches, not changes. It still mentions the branches created for the changes, however. Also, since the operation "tags" contain the full command line, I think it'll still be relatively easy for the user to understand what the operation was about.
2023-06-03 06:24:28 +00:00
std::fs::write(workspace_root.join("file"), "contents").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["new", "-m", "bar"]);
test_env.jj_cmd_ok(&workspace_root, &["branch", "create", "branch-1"]);
cli: allow `jj git push --change xyz --branch foo` This was pretty simple. I simplified a bit by making the transaction description mention only branches, not changes. It still mentions the branches created for the changes, however. Also, since the operation "tags" contain the full command line, I think it'll still be relatively easy for the user to understand what the operation was about.
2023-06-03 06:24:28 +00:00
std::fs::write(workspace_root.join("file"), "modified").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["new", "-m", "baz"]);
test_env.jj_cmd_ok(&workspace_root, &["branch", "create", "branch-2a"]);
test_env.jj_cmd_ok(&workspace_root, &["branch", "create", "branch-2b"]);
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
std::fs::write(workspace_root.join("file"), "modified again").unwrap();
cli: allow `jj git push --change xyz --branch foo` This was pretty simple. I simplified a bit by making the transaction description mention only branches, not changes. It still mentions the branches created for the changes, however. Also, since the operation "tags" contain the full command line, I think it'll still be relatively easy for the user to understand what the operation was about.
2023-06-03 06:24:28 +00:00
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(
cli: allow `jj git push --change xyz --branch foo` This was pretty simple. I simplified a bit by making the transaction description mention only branches, not changes. It still mentions the branches created for the changes, however. Also, since the operation "tags" contain the full command line, I think it'll still be relatively easy for the user to understand what the operation was about.
2023-06-03 06:24:28 +00:00
&workspace_root,
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
&["git", "push", "--change=@--", "--branch=branch-1", "-r=@"],
cli: allow `jj git push --change xyz --branch foo` This was pretty simple. I simplified a bit by making the transaction description mention only branches, not changes. It still mentions the branches created for the changes, however. Also, since the operation "tags" contain the full command line, I think it'll still be relatively easy for the user to understand what the operation was about.
2023-06-03 06:24:28 +00:00
);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
Creating branch push-yqosqzytrlsw for revision @--
cli: allow `jj git push --change xyz --branch foo` This was pretty simple. I simplified a bit by making the transaction description mention only branches, not changes. It still mentions the branches created for the changes, however. Also, since the operation "tags" contain the full command line, I think it'll still be relatively easy for the user to understand what the operation was about.
2023-06-03 06:24:28 +00:00
Branch changes to push to origin:
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
Add branch branch-1 to 7decc7932d9c
cli: allow `jj git push --change xyz --branch foo` This was pretty simple. I simplified a bit by making the transaction description mention only branches, not changes. It still mentions the branches created for the changes, however. Also, since the operation "tags" contain the full command line, I think it'll still be relatively easy for the user to understand what the operation was about.
2023-06-03 06:24:28 +00:00
Add branch push-yqosqzytrlsw to fa16a14170fb
cli: add `jj git push -r` for pushing branches pointing to revset I think I will find this useful in at least two cases: 1. When you already have a branch pointing to some commit, it's easier to do `jj git push -r xyz` than `jj git push --branch push-xyzxyzyxzxyz`. 2. When you have a stack of changes, it's useful to be able to push all of them at once. I think we should also update the default behavior of `jj git push` to be `jj git push -r 'remote_branches()..@'` or something like that. That removes the ugliness of having a default behavior that the user can't reproduce using flags. I'll leave that change for a separate PR.
2023-06-03 05:39:33 +00:00
Add branch branch-2a to 1b45449e18d0
Add branch branch-2b to 1b45449e18d0
cli: allow `jj git push --change xyz --branch foo` This was pretty simple. I simplified a bit by making the transaction description mention only branches, not changes. It still mentions the branches created for the changes, however. Also, since the operation "tags" contain the full command line, I think it'll still be relatively easy for the user to understand what the operation was about.
2023-06-03 06:24:28 +00:00
"###);
}
Shorten the git push branch when possible using the short change ID hash.
2023-01-11 18:30:31 +00:00
#[test]
fn test_git_push_existing_long_branch() {
let (test_env, workspace_root) = set_up();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["describe", "-m", "foo"]);
Shorten the git push branch when possible using the short change ID hash.
2023-01-11 18:30:31 +00:00
std::fs::write(workspace_root.join("file"), "contents").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(
Shorten the git push branch when possible using the short change ID hash.
2023-01-11 18:30:31 +00:00
&workspace_root,
backend: make random ChangeId fully random, remove UUID mask bits
2023-01-14 04:50:45 +00:00
&["branch", "create", "push-19b790168e73f7a73a98deae21e807c0"],
Shorten the git push branch when possible using the short change ID hash.
2023-01-11 18:30:31 +00:00
);
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push", "--change=@"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
Shorten the git push branch when possible using the short change ID hash.
2023-01-11 18:30:31 +00:00
Branch changes to push to origin:
backend: make random ChangeId fully random, remove UUID mask bits
2023-01-14 04:50:45 +00:00
Add branch push-19b790168e73f7a73a98deae21e807c0 to fa16a14170fb
jj git push: Make sure each branch is processed at most once
2022-12-19 02:25:04 +00:00
"###);
Add a test for `git push --change`
2022-12-17 22:31:12 +00:00
}
cli: don't use stale index to look up snapshotted working copy
2022-07-05 06:14:39 +00:00
#[test]
fn test_git_push_unsnapshotted_change() {
let (test_env, workspace_root) = set_up();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["describe", "-m", "foo"]);
cli: don't use stale index to look up snapshotted working copy
2022-07-05 06:14:39 +00:00
std::fs::write(workspace_root.join("file"), "contents").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["git", "push", "--change", "@"]);
cli: don't use stale index to look up snapshotted working copy
2022-07-05 06:14:39 +00:00
std::fs::write(workspace_root.join("file"), "modified").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["git", "push", "--change", "@"]);
cli: don't use stale index to look up snapshotted working copy
2022-07-05 06:14:39 +00:00
}
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
#[test]
fn test_git_push_conflict() {
let (test_env, workspace_root) = set_up();
cli: prevent pushing commits with conflicts It rarely makes sense to push commits with conflicts to a remote Git repo (which is the only kind of remote we support so far), so let's just error out instead of pushing a commit that others pulling from the remote probably can't make sense of. I've only added a simple test for the error case for now. `libgit2` doesn't support pushing to a local repo, so it's harder to test the success case. I suppose we'll have to have the regular `git` binary running local servers in test eventually. Closes #60.
2022-03-09 23:57:12 +00:00
std::fs::write(workspace_root.join("file"), "first").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["commit", "-m", "first"]);
cli: prevent pushing commits with conflicts It rarely makes sense to push commits with conflicts to a remote Git repo (which is the only kind of remote we support so far), so let's just error out instead of pushing a commit that others pulling from the remote probably can't make sense of. I've only added a simple test for the error case for now. `libgit2` doesn't support pushing to a local repo, so it's harder to test the success case. I suppose we'll have to have the regular `git` binary running local servers in test eventually. Closes #60.
2022-03-09 23:57:12 +00:00
std::fs::write(workspace_root.join("file"), "second").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["commit", "-m", "second"]);
cli: prevent pushing commits with conflicts It rarely makes sense to push commits with conflicts to a remote Git repo (which is the only kind of remote we support so far), so let's just error out instead of pushing a commit that others pulling from the remote probably can't make sense of. I've only added a simple test for the error case for now. `libgit2` doesn't support pushing to a local repo, so it's harder to test the success case. I suppose we'll have to have the regular `git` binary running local servers in test eventually. Closes #60.
2022-03-09 23:57:12 +00:00
std::fs::write(workspace_root.join("file"), "third").unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["rebase", "-r", "@", "-d", "@--"]);
tests: don't create new branch by "jj branch set"
2023-11-10 00:56:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["branch", "create", "my-branch"]);
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["describe", "-m", "third"]);
cli: push only branches pointing to `@` by default Since we now allow pushing open commits, we can implement support for pushing the "current" branch by defining a "current" branch as any branch pointing to `@`. That definition of a current/active seems to have been the consensus in discussion #411. Closes #246.
2022-07-13 04:14:02 +00:00
let stderr = test_env.jj_cmd_failure(&workspace_root, &["git", "push", "--all"]);
tests: re-run with `insta` crate version b9d99e87065b
2022-04-28 23:32:18 +00:00
insta::assert_snapshot!(stderr, @r###"
cli: enable tree-level conflicts by default I have used the tree-level conflict format for several weeks without problem (after the fix in 51b5d168aed0). Now - right after the 0.10.0 release - seems like a good time to enable the config by default. I enabled the config in our default configs in the CLI crate to reduce impact on tests (compared to changing the default in `settings.rs`).
2023-10-05 04:30:29 +00:00
Error: Won't push commit 1973d389875c since it has conflicts
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
"###);
}
#[test]
fn test_git_push_no_description() {
let (test_env, workspace_root) = set_up();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["branch", "create", "my-branch"]);
test_env.jj_cmd_ok(&workspace_root, &["describe", "-m="]);
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
let stderr =
test_env.jj_cmd_failure(&workspace_root, &["git", "push", "--branch", "my-branch"]);
insta::assert_snapshot!(stderr, @r###"
test_git_push: create Git repo using `jj` commands Since we don't have the `git` binary available in the test environment, it's just easier to use `jj` commands to create Git repos.
2022-11-04 18:25:11 +00:00
Error: Won't push commit 5b36783cd11c since it has no description
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
"###);
}
#[test]
fn test_git_push_missing_author() {
let (test_env, workspace_root) = set_up();
let run_without_var = |var: &str, args: &[&str]| {
test_env
.jj_cmd(&workspace_root, args)
.env_remove(var)
.assert()
.success();
};
revset: convert root symbol to function The idea is that we can fully eliminate special symbols that would otherwise shadow user branches, tags, or change ID prefixes. Closes #2095
2023-09-03 03:47:23 +00:00
run_without_var("JJ_USER", &["checkout", "root()", "-m=initial"]);
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
run_without_var("JJ_USER", &["branch", "create", "missing-name"]);
let stderr = test_env.jj_cmd_failure(
&workspace_root,
&["git", "push", "--branch", "missing-name"],
);
insta::assert_snapshot!(stderr, @r###"
commits: use empty strings instead of placeholders for missing name or email This commit replaces the functions `UserSettings::user_name_placeholder()`` and `UserSettings::user_email_placeholder()` with `const` `&str`s to emphasize that the placeholder strings must not be changed to support commits without names or email addresses made before this change.
2023-08-16 21:10:54 +00:00
Error: Won't push commit 944313939bbd since it has no author and/or committer set
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
"###);
revset: convert root symbol to function The idea is that we can fully eliminate special symbols that would otherwise shadow user branches, tags, or change ID prefixes. Closes #2095
2023-09-03 03:47:23 +00:00
run_without_var("JJ_EMAIL", &["checkout", "root()", "-m=initial"]);
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
run_without_var("JJ_EMAIL", &["branch", "create", "missing-email"]);
tests: avoid about-to-be-deleted `close` command
2022-11-05 00:15:35 +00:00
let stderr =
test_env.jj_cmd_failure(&workspace_root, &["git", "push", "--branch=missing-email"]);
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
insta::assert_snapshot!(stderr, @r###"
commits: use empty strings instead of placeholders for missing name or email This commit replaces the functions `UserSettings::user_name_placeholder()`` and `UserSettings::user_email_placeholder()` with `const` `&str`s to emphasize that the placeholder strings must not be changed to support commits without names or email addresses made before this change.
2023-08-16 21:10:54 +00:00
Error: Won't push commit 59354714f789 since it has no author and/or committer set
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
"###);
}
#[test]
fn test_git_push_missing_committer() {
let (test_env, workspace_root) = set_up();
let run_without_var = |var: &str, args: &[&str]| {
test_env
.jj_cmd(&workspace_root, args)
.env_remove(var)
.assert()
.success();
};
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["branch", "create", "missing-name"]);
tests: avoid about-to-be-deleted `close` command
2022-11-05 00:15:35 +00:00
run_without_var("JJ_USER", &["describe", "-m=no committer name"]);
let stderr =
test_env.jj_cmd_failure(&workspace_root, &["git", "push", "--branch=missing-name"]);
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
insta::assert_snapshot!(stderr, @r###"
commits: use empty strings instead of placeholders for missing name or email This commit replaces the functions `UserSettings::user_name_placeholder()`` and `UserSettings::user_email_placeholder()` with `const` `&str`s to emphasize that the placeholder strings must not be changed to support commits without names or email addresses made before this change.
2023-08-16 21:10:54 +00:00
Error: Won't push commit 4fd190283d1a since it has no author and/or committer set
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
"###);
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["checkout", "root()"]);
test_env.jj_cmd_ok(&workspace_root, &["branch", "create", "missing-email"]);
tests: avoid about-to-be-deleted `close` command
2022-11-05 00:15:35 +00:00
run_without_var("JJ_EMAIL", &["describe", "-m=no committer email"]);
let stderr =
test_env.jj_cmd_failure(&workspace_root, &["git", "push", "--branch=missing-email"]);
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
insta::assert_snapshot!(stderr, @r###"
commits: use empty strings instead of placeholders for missing name or email This commit replaces the functions `UserSettings::user_name_placeholder()`` and `UserSettings::user_email_placeholder()` with `const` `&str`s to emphasize that the placeholder strings must not be changed to support commits without names or email addresses made before this change.
2023-08-16 21:10:54 +00:00
Error: Won't push commit eab97428a6ec since it has no author and/or committer set
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
"###);
// Test message when there are multiple reasons (missing committer and
// description)
tests: avoid about-to-be-deleted `close` command
2022-11-05 00:15:35 +00:00
run_without_var("JJ_EMAIL", &["describe", "-m=", "missing-email"]);
let stderr =
test_env.jj_cmd_failure(&workspace_root, &["git", "push", "--branch=missing-email"]);
cli: prevent pushing commits without description, author, or committer This patch prevents perhaps pushing commits with an empty description or the placeholder "(no user/email configured)" values for author/committer. Closes #322.
2022-05-27 15:54:47 +00:00
insta::assert_snapshot!(stderr, @r###"
commits: use empty strings instead of placeholders for missing name or email This commit replaces the functions `UserSettings::user_name_placeholder()`` and `UserSettings::user_email_placeholder()` with `const` `&str`s to emphasize that the placeholder strings must not be changed to support commits without names or email addresses made before this change.
2023-08-16 21:10:54 +00:00
Error: Won't push commit 1143ed607f54 since it has no description and it has no author and/or committer set
tests: re-run with `insta` crate version b9d99e87065b
2022-04-28 23:32:18 +00:00
"###);
cli: prevent pushing commits with conflicts It rarely makes sense to push commits with conflicts to a remote Git repo (which is the only kind of remote we support so far), so let's just error out instead of pushing a commit that others pulling from the remote probably can't make sense of. I've only added a simple test for the error case for now. `libgit2` doesn't support pushing to a local repo, so it's harder to test the success case. I suppose we'll have to have the regular `git` binary running local servers in test eventually. Closes #60.
2022-03-09 23:57:12 +00:00
}
git push: add --deleted option
2023-02-19 12:08:24 +00:00
#[test]
fn test_git_push_deleted() {
let (test_env, workspace_root) = set_up();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["branch", "delete", "branch1"]);
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push", "--deleted"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
git push: add --deleted option
2023-02-19 12:08:24 +00:00
Branch changes to push to origin:
Delete branch branch1 from 45a3aa29e907
git: on push, update jj's view of remote branches without using import_refs() This means that the commits previously pinned by remote branches are no longer abandoned. I think that's more correct since "push" is the operation to propagate local view to remote, and uninteresting commits should have been locally abandoned.
2023-10-14 19:27:44 +00:00
"###);
let stdout = test_env.jj_cmd_success(&workspace_root, &["log", "-rall()"]);
insta::assert_snapshot!(stdout, @r###"
◉ rlzusymt test.user@example.com 2001-02-03 04:05:10.000 +07:00 branch2 8476341e
│ (empty) description 2
│ ◉ lzmmnrxq test.user@example.com 2001-02-03 04:05:08.000 +07:00 45a3aa29
├─╯ (empty) description 1
│ @ yqosqzyt test.user@example.com 2001-02-03 04:05:13.000 +07:00 5b36783c
├─╯ (empty) (no description set)
◉ zzzzzzzz root() 00000000
git push: add --deleted option
2023-02-19 12:08:24 +00:00
"###);
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push", "--deleted"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
insta::assert_snapshot!(stderr, @r###"
git push: add --deleted option
2023-02-19 12:08:24 +00:00
Nothing changed.
"###);
}
tests: add "git push" tests involving conflicting branches
2023-07-03 09:03:07 +00:00
#[test]
fn test_git_push_conflicting_branches() {
let (test_env, workspace_root) = set_up();
tests: set git_settings.auto_local_branch where it matters
2023-12-16 10:00:14 +00:00
test_env.add_config("git.auto-local-branch = true");
tests: add "git push" tests involving conflicting branches
2023-07-03 09:03:07 +00:00
let git_repo = {
let mut git_repo_path = workspace_root.clone();
git_repo_path.extend([".jj", "repo", "store", "git"]);
git2::Repository::open(&git_repo_path).unwrap()
};
// Forget remote ref, move local ref, then fetch to create conflict.
git_repo
.find_reference("refs/remotes/origin/branch2")
.unwrap()
.delete()
.unwrap();
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["git", "import"]);
test_env.jj_cmd_ok(&workspace_root, &["new", "root()", "-m=description 3"]);
tests: don't create new branch by "jj branch set"
2023-11-10 00:56:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["branch", "create", "branch2"]);
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["git", "fetch"]);
cli: add "branch list --all" option to include all remote branches This will be the option to include non-tracking remote branches. We could add more fine-grained filtering flags, but I think --all is good enough and easier to remember. This patch also updates many of the test outputs to include synchronized remote branches. I think verbose outputs will help catch future bugs.
2023-10-18 06:12:17 +00:00
insta::assert_snapshot!(
test_env.jj_cmd_success(&workspace_root, &["branch", "list", "--all"]), @r###"
cli templates: include change id and more in short commit description Include change id as well as (empty) and (conflict) markers in short commit description. Unlike `jj log`, * (conflict) is put in parentheses to distinguish it from the commit description when the output is not colored. * the change id is placed right before the commit id to put it into a predicteble place, even if the short commit description takes up more than one line. The commit id can be important for commit descriptions in commands like `abandon`.
2023-07-11 01:46:38 +00:00
branch1: lzmmnrxq 45a3aa29 (empty) description 1
cli: add "branch list --all" option to include all remote branches This will be the option to include non-tracking remote branches. We could add more fine-grained filtering flags, but I think --all is good enough and easier to remember. This patch also updates many of the test outputs to include synchronized remote branches. I think verbose outputs will help catch future bugs.
2023-10-18 06:12:17 +00:00
@origin: lzmmnrxq 45a3aa29 (empty) description 1
tests: add "git push" tests involving conflicting branches
2023-07-03 09:03:07 +00:00
branch2 (conflicted):
cli templates: include change id and more in short commit description Include change id as well as (empty) and (conflict) markers in short commit description. Unlike `jj log`, * (conflict) is put in parentheses to distinguish it from the commit description when the output is not colored. * the change id is placed right before the commit id to put it into a predicteble place, even if the short commit description takes up more than one line. The commit id can be important for commit descriptions in commands like `abandon`.
2023-07-11 01:46:38 +00:00
+ yostqsxw 8e670e2d (empty) description 3
+ rlzusymt 8476341e (empty) description 2
@origin (behind by 1 commits): rlzusymt 8476341e (empty) description 2
tests: add "git push" tests involving conflicting branches
2023-07-03 09:03:07 +00:00
"###);
let bump_branch1 = || {
tests: replace jj_cmd_success() involving mutation to allow stderr output
2023-10-10 11:59:18 +00:00
test_env.jj_cmd_ok(&workspace_root, &["new", "branch1", "-m=bump"]);
test_env.jj_cmd_ok(&workspace_root, &["branch", "set", "branch1"]);
tests: add "git push" tests involving conflicting branches
2023-07-03 09:03:07 +00:00
};
// Conflicting branch at @
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
tests: add "git push" tests involving conflicting branches
2023-07-03 09:03:07 +00:00
insta::assert_snapshot!(stderr, @r###"
cli: make `jj git push` default to `-r 'remote_branches()..@'` The way `jj git push` without arguments chooses branches pointing to either `@` or `@-` is unusual and difficult to explain. Now that we have `-r`, we could instead default it to `-r '@-::@'`. However, I think it seems likely that users will want to push all local branches leading up to `@` from the closest remote branch. That's typically what I want. This patch changes the default to do that.
2023-08-21 20:11:40 +00:00
Branch branch2 is conflicted
cli: show hints for branches rejected to push
2023-10-16 19:25:52 +00:00
Hint: Run `jj branch list` to inspect, and use `jj branch set` to fix it up.
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
Nothing changed.
tests: add "git push" tests involving conflicting branches
2023-07-03 09:03:07 +00:00
"###);
// --branch should be blocked by conflicting branch
let stderr = test_env.jj_cmd_failure(&workspace_root, &["git", "push", "--branch", "branch2"]);
insta::assert_snapshot!(stderr, @r###"
Error: Branch branch2 is conflicted
cli: show hints for branches rejected to push
2023-10-16 19:25:52 +00:00
Hint: Run `jj branch list` to inspect, and use `jj branch set` to fix it up.
tests: add "git push" tests involving conflicting branches
2023-07-03 09:03:07 +00:00
"###);
// --all shouldn't be blocked by conflicting branch
bump_branch1();
cli: warn if loosely selected push targets include conflicted branches
2023-07-03 08:52:31 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push", "--all"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
cli: warn if loosely selected push targets include conflicted branches
2023-07-03 08:52:31 +00:00
insta::assert_snapshot!(stderr, @r###"
Branch branch2 is conflicted
cli: show hints for branches rejected to push
2023-10-16 19:25:52 +00:00
Hint: Run `jj branch list` to inspect, and use `jj branch set` to fix it up.
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
Branch changes to push to origin:
Move branch branch1 from 45a3aa29e907 to fd1d63e031ea
cli: warn if loosely selected push targets include conflicted branches
2023-07-03 08:52:31 +00:00
"###);
tests: add "git push" tests involving conflicting branches
2023-07-03 09:03:07 +00:00
// --revisions shouldn't be blocked by conflicting branch
bump_branch1();
cli: warn if loosely selected push targets include conflicted branches
2023-07-03 08:52:31 +00:00
let (stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push", "-rall()"]);
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
insta::assert_snapshot!(stdout, @"");
cli: warn if loosely selected push targets include conflicted branches
2023-07-03 08:52:31 +00:00
insta::assert_snapshot!(stderr, @r###"
Branch branch2 is conflicted
cli: show hints for branches rejected to push
2023-10-16 19:25:52 +00:00
Hint: Run `jj branch list` to inspect, and use `jj branch set` to fix it up.
cli: send status messages to stderr, specify stdout/stderr explicitly Many of &mut UI can be changed to immutable borrows, but I'm not gonna update them in this patch.
2023-10-10 11:07:06 +00:00
Branch changes to push to origin:
Move branch branch1 from fd1d63e031ea to 8263cf992d33
cli: warn if loosely selected push targets include conflicted branches
2023-07-03 08:52:31 +00:00
"###);
tests: add "git push" tests involving conflicting branches
2023-07-03 09:03:07 +00:00
}
git: prohibit push to remote named "git" Since I'm going to make git::push_branches() update the repo view internally, it should fail fast if the remote name is reserved. Before, the problem was detected on git::import_refs().
2023-10-14 20:11:19 +00:00
refs: classify push action based on tracking target Although this is logically correct, the error message is a bit cryptic. It's probably better to reject push if non-tracking remote branches exist. #1136
2023-10-14 14:54:35 +00:00
#[test]
fn test_git_push_deleted_untracked() {
let (test_env, workspace_root) = set_up();
// Absent local branch shouldn't be considered "deleted" compared to
// non-tracking remote branch.
test_env.jj_cmd_ok(&workspace_root, &["branch", "delete", "branch1"]);
test_env.jj_cmd_ok(&workspace_root, &["branch", "untrack", "branch1@origin"]);
let (_stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push", "--deleted"]);
insta::assert_snapshot!(stderr, @r###"
Nothing changed.
"###);
let stderr = test_env.jj_cmd_failure(&workspace_root, &["git", "push", "--branch=branch1"]);
insta::assert_snapshot!(stderr, @r###"
cli: add string pattern support to "git push --branch" Since "jj git fetch --branch" supports glob patterns, users would expect that "jj git push --branch glob:.." also works. The error handling bits are copied from "branch" sub commands. We might want to extract it to a common helper function, but I haven't figured out a reasonable boundary point yet.
2023-10-24 06:31:26 +00:00
Error: No such branch: branch1
refs: classify push action based on tracking target Although this is logically correct, the error message is a bit cryptic. It's probably better to reject push if non-tracking remote branches exist. #1136
2023-10-14 14:54:35 +00:00
"###);
}
#[test]
fn test_git_push_moved_forward_untracked() {
let (test_env, workspace_root) = set_up();
test_env.jj_cmd_ok(&workspace_root, &["new", "branch1", "-mmoved branch1"]);
test_env.jj_cmd_ok(&workspace_root, &["branch", "set", "branch1"]);
test_env.jj_cmd_ok(&workspace_root, &["branch", "untrack", "branch1@origin"]);
let (_stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push"]);
insta::assert_snapshot!(stderr, @r###"
refs: error out push if non-tracking remote branches exist We can provide more actionable error message than "not fast-forwardable". If the push was fast-forwardable, "jj branch track" should be able to merge the remote branch without conflicts, so the added step would be minimal.
2023-10-15 17:43:45 +00:00
Non-tracking remote branch branch1@origin exists
cli: show hints for branches rejected to push
2023-10-16 19:25:52 +00:00
Hint: Run `jj branch track branch1@origin` to import the remote branch.
refs: error out push if non-tracking remote branches exist We can provide more actionable error message than "not fast-forwardable". If the push was fast-forwardable, "jj branch track" should be able to merge the remote branch without conflicts, so the added step would be minimal.
2023-10-15 17:43:45 +00:00
Nothing changed.
refs: classify push action based on tracking target Although this is logically correct, the error message is a bit cryptic. It's probably better to reject push if non-tracking remote branches exist. #1136
2023-10-14 14:54:35 +00:00
"###);
}
#[test]
fn test_git_push_moved_sideways_untracked() {
let (test_env, workspace_root) = set_up();
test_env.jj_cmd_ok(&workspace_root, &["new", "root()", "-mmoved branch1"]);
test_env.jj_cmd_ok(
&workspace_root,
&["branch", "set", "--allow-backwards", "branch1"],
);
test_env.jj_cmd_ok(&workspace_root, &["branch", "untrack", "branch1@origin"]);
refs: error out push if non-tracking remote branches exist We can provide more actionable error message than "not fast-forwardable". If the push was fast-forwardable, "jj branch track" should be able to merge the remote branch without conflicts, so the added step would be minimal.
2023-10-15 17:43:45 +00:00
let (_stdout, stderr) = test_env.jj_cmd_ok(&workspace_root, &["git", "push"]);
refs: classify push action based on tracking target Although this is logically correct, the error message is a bit cryptic. It's probably better to reject push if non-tracking remote branches exist. #1136
2023-10-14 14:54:35 +00:00
insta::assert_snapshot!(stderr, @r###"
refs: error out push if non-tracking remote branches exist We can provide more actionable error message than "not fast-forwardable". If the push was fast-forwardable, "jj branch track" should be able to merge the remote branch without conflicts, so the added step would be minimal.
2023-10-15 17:43:45 +00:00
Non-tracking remote branch branch1@origin exists
cli: show hints for branches rejected to push
2023-10-16 19:25:52 +00:00
Hint: Run `jj branch track branch1@origin` to import the remote branch.
refs: error out push if non-tracking remote branches exist We can provide more actionable error message than "not fast-forwardable". If the push was fast-forwardable, "jj branch track" should be able to merge the remote branch without conflicts, so the added step would be minimal.
2023-10-15 17:43:45 +00:00
Nothing changed.
refs: classify push action based on tracking target Although this is logically correct, the error message is a bit cryptic. It's probably better to reject push if non-tracking remote branches exist. #1136
2023-10-14 14:54:35 +00:00
"###);
}
git: prohibit push to remote named "git" Since I'm going to make git::push_branches() update the repo view internally, it should fail fast if the remote name is reserved. Before, the problem was detected on git::import_refs().
2023-10-14 20:11:19 +00:00
#[test]
fn test_git_push_to_remote_named_git() {
let (test_env, workspace_root) = set_up();
let git_repo = {
let mut git_repo_path = workspace_root.clone();
git_repo_path.extend([".jj", "repo", "store", "git"]);
git2::Repository::open(&git_repo_path).unwrap()
};
git_repo.remote_rename("origin", "git").unwrap();
let stderr =
test_env.jj_cmd_failure(&workspace_root, &["git", "push", "--all", "--remote=git"]);
insta::assert_snapshot!(stderr, @r###"
Branch changes to push to git:
Add branch branch1 to 45a3aa29e907
Add branch branch2 to 8476341eb395
Error: Git remote named 'git' is reserved for local Git repository
"###);
}
Copy permalink