## Changes
Add new option -tail to acceptance test runner. This logs output.txt in
real time with relative timestamp prefix.
## Why
When working on long running integration tests, it's useful to see what
stage they are at.
## Tests
Manually
```
~/work/cli/acceptance/selftest % testme -v -tail
+ go test .. -run ^TestAccept$/^selftest$ -v -tail
...
=== CONT TestAccept/selftest/diff
acceptance_test.go:774: 0.051 >>> diff.py out_dir_a out_dir_b
=== NAME TestAccept/selftest/basic
acceptance_test.go:774: 0.051 === Capturing STDERR
acceptance_test.go:774: 0.051 >>> python3 -c import sys; sys.stderr.write("STDERR\n")
=== NAME TestAccept/selftest/server
acceptance_test.go:774: 0.050 >>> curl -s http://127.0.0.1:55850/api/2.0/preview/scim/v2/Me
...
acceptance_test.go:774: 0.204 >>> /Users/denis.bilenko/work/cli/acceptance/build/darwin_arm64/databricks --version
=== NAME TestAccept/selftest/server
acceptance_test.go:774: 0.510 Error: Workspace path not found
...
```
Bumps [golang.org/x/text](https://github.com/golang/text) from 0.22.0 to
0.23.0.
<details>
<summary>Commits</summary>
<ul>
<li><a
href="566b44fc93"><code>566b44f</code></a>
go.mod: update golang.org/x dependencies</li>
<li><a
href="d5156da34e"><code>d5156da</code></a>
collate/build: do not use println in tests</li>
<li><a
href="221d88cef7"><code>221d88c</code></a>
x/text: fix scientific notation by removing extraneous spaces</li>
<li><a
href="b18c10747b"><code>b18c107</code></a>
internal/export/unicode: change C comment to mention unassigned code
points</li>
<li><a
href="835f8ac434"><code>835f8ac</code></a>
language: use a more straightforward return value</li>
<li><a
href="ae68efb59e"><code>ae68efb</code></a>
internal/export/unicode: add CategoryAliases, Cn, and LC</li>
<li><a
href="518d9c0174"><code>518d9c0</code></a>
all: upgrade go directive to at least 1.23.0 [generated]</li>
<li>See full diff in <a
href="https://github.com/golang/text/compare/v0.22.0...v0.23.0">compare
view</a></li>
</ul>
</details>
<br />
[](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)
Dependabot will resolve any conflicts with this PR as long as you don't
alter it yourself. You can also trigger a rebase manually by commenting
`@dependabot rebase`.
[//]: # (dependabot-automerge-start)
[//]: # (dependabot-automerge-end)
---
<details>
<summary>Dependabot commands and options</summary>
<br />
You can trigger Dependabot actions by commenting on this PR:
- `@dependabot rebase` will rebase this PR
- `@dependabot recreate` will recreate this PR, overwriting any edits
that have been made to it
- `@dependabot merge` will merge this PR after your CI passes on it
- `@dependabot squash and merge` will squash and merge this PR after
your CI passes on it
- `@dependabot cancel merge` will cancel a previously requested merge
and block automerging
- `@dependabot reopen` will reopen this PR if it is closed
- `@dependabot close` will close this PR and stop Dependabot recreating
it. You can achieve the same result by closing it manually
- `@dependabot show <dependency name> ignore conditions` will show all
of the ignore conditions of the specified dependency
- `@dependabot ignore this major version` will close this PR and stop
Dependabot creating any more for this major version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this minor version` will close this PR and stop
Dependabot creating any more for this minor version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this dependency` will close this PR and stop
Dependabot creating any more for this dependency (unless you reopen the
PR or upgrade to it yourself)
</details>
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
## Changes
<!-- Brief summary of your changes that is easy to understand -->
1. Change the cloud acceptance test for `bind/schema` to run locally
2. Add debug lines to the mock server
3. Change `fake_workspace` to create directories for imported files
## Why
<!-- Why are these changes needed? Provide the context that the reviewer
might be missing.
For example, were there any decisions behind the change that are not
reflected in the code itself? -->
1. Local version of the test run can indicate breaking changes faster
than the cloud version and it can be run locally without any predefined
environment variables
## Tests
<!-- How have you tested the changes? -->
1. Ran both acloud and local versions of test, both succeeded
<!-- If your PR needs to be included in the release notes for next
release,
add a separate entry in NEXT_CHANGELOG.md as part of your PR. -->
Bumps [golang.org/x/oauth2](https://github.com/golang/oauth2) from
0.26.0 to 0.28.0.
<details>
<summary>Commits</summary>
<ul>
<li><a
href="0042180b24"><code>0042180</code></a>
oauth2: Deep copy context client in NewClient</li>
<li><a
href="ce350bff61"><code>ce350bf</code></a>
oauth2: remove unneeded TokenSource implementation in transport
test</li>
<li><a
href="44967abe90"><code>44967ab</code></a>
google: fix typos</li>
<li><a
href="9c82a8cf7a"><code>9c82a8c</code></a>
oauth2.go: use a more straightforward return value</li>
<li><a
href="681b4d8edc"><code>681b4d8</code></a>
jws: split token into fixed number of parts</li>
<li><a
href="3f78298bee"><code>3f78298</code></a>
all: upgrade go directive to at least 1.23.0 [generated]</li>
<li><a
href="109dabf901"><code>109dabf</code></a>
endpoints: add links/provider for Discord</li>
<li><a
href="ac571fa341"><code>ac571fa</code></a>
oauth2: fix docs for Config.DeviceAuth</li>
<li><a
href="314ee5b92b"><code>314ee5b</code></a>
endpoints: add patreon endpoint</li>
<li>See full diff in <a
href="https://github.com/golang/oauth2/compare/v0.26.0...v0.28.0">compare
view</a></li>
</ul>
</details>
<br />
[](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)
Dependabot will resolve any conflicts with this PR as long as you don't
alter it yourself. You can also trigger a rebase manually by commenting
`@dependabot rebase`.
[//]: # (dependabot-automerge-start)
[//]: # (dependabot-automerge-end)
---
<details>
<summary>Dependabot commands and options</summary>
<br />
You can trigger Dependabot actions by commenting on this PR:
- `@dependabot rebase` will rebase this PR
- `@dependabot recreate` will recreate this PR, overwriting any edits
that have been made to it
- `@dependabot merge` will merge this PR after your CI passes on it
- `@dependabot squash and merge` will squash and merge this PR after
your CI passes on it
- `@dependabot cancel merge` will cancel a previously requested merge
and block automerging
- `@dependabot reopen` will reopen this PR if it is closed
- `@dependabot close` will close this PR and stop Dependabot recreating
it. You can achieve the same result by closing it manually
- `@dependabot show <dependency name> ignore conditions` will show all
of the ignore conditions of the specified dependency
- `@dependabot ignore this major version` will close this PR and stop
Dependabot creating any more for this major version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this minor version` will close this PR and stop
Dependabot creating any more for this minor version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this dependency` will close this PR and stop
Dependabot creating any more for this dependency (unless you reopen the
PR or upgrade to it yourself)
</details>
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Bumps [golang.org/x/term](https://github.com/golang/term) from 0.29.0 to
0.30.0.
<details>
<summary>Commits</summary>
<ul>
<li><a
href="04218fdaf7"><code>04218fd</code></a>
go.mod: update golang.org/x dependencies</li>
<li><a
href="208db03875"><code>208db03</code></a>
all: upgrade go directive to at least 1.23.0 [generated]</li>
<li>See full diff in <a
href="https://github.com/golang/term/compare/v0.29.0...v0.30.0">compare
view</a></li>
</ul>
</details>
<br />
[](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)
Dependabot will resolve any conflicts with this PR as long as you don't
alter it yourself. You can also trigger a rebase manually by commenting
`@dependabot rebase`.
[//]: # (dependabot-automerge-start)
[//]: # (dependabot-automerge-end)
---
<details>
<summary>Dependabot commands and options</summary>
<br />
You can trigger Dependabot actions by commenting on this PR:
- `@dependabot rebase` will rebase this PR
- `@dependabot recreate` will recreate this PR, overwriting any edits
that have been made to it
- `@dependabot merge` will merge this PR after your CI passes on it
- `@dependabot squash and merge` will squash and merge this PR after
your CI passes on it
- `@dependabot cancel merge` will cancel a previously requested merge
and block automerging
- `@dependabot reopen` will reopen this PR if it is closed
- `@dependabot close` will close this PR and stop Dependabot recreating
it. You can achieve the same result by closing it manually
- `@dependabot show <dependency name> ignore conditions` will show all
of the ignore conditions of the specified dependency
- `@dependabot ignore this major version` will close this PR and stop
Dependabot creating any more for this major version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this minor version` will close this PR and stop
Dependabot creating any more for this minor version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this dependency` will close this PR and stop
Dependabot creating any more for this dependency (unless you reopen the
PR or upgrade to it yourself)
</details>
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
## Changes
Set Go patch version in go.mod
## Why
Our dependabot PRs failing with the following diff
```
Run # Exit with status code 1 if there are differences (i.e. unformatted files)
diff --git a/go.mod b/go.mod
index 5c1cb1d..493152a 100644
--- a/go.mod
+++ b/go.mod
@@ -1,6 +1,[7](https://github.com/databricks/cli/actions/runs/13768953030/job/38502532441?pr=2459#step:5:8) @@
module github.com/databricks/cli
-go 1.23
+go 1.23.0
+
```
While I'm not sure why it started to happen, setting the patch version
at least should fix these PR.
This might be related to this change in Go
https://github.com/golang/go/issues/69095
## Changes
Add new CI jobs for testing Python code.
These jobs are separate from Golang code because they are executed
against a matrix of Python versions, and OS versions are not relevant.
We will get coverage with OS versions using acceptance tests. In
addition, we don't need to run these tests unless any of the Python code
is modified.
We use uv as a package manager to provision all CI tools:
- pyright to typecheck
- ruff to lint and format
experimental/python contains a Makefile with target similar to
repository root:
- make fmt
- make lint
- make build
## Tests
Using CI
<!-- How have you tested the changes? -->
---
NO_CHANGELOG=true
## Changes
<!-- Brief summary of your changes that is easy to understand -->
When a file is imported, FakeWorkspace now automatically identifies and
records all parent directories in the directories map, ensuring the
directory structure is properly maintained without requiring explicit
directory creation.
## Why
<!-- Why are these changes needed? Provide the context that the reviewer
might be missing.
For example, were there any decisions behind the change that are not
reflected in the code itself? -->
During an acceptance test, production code might want to check a root
folder for existence, this change makes sure that root folder is marked
as existing when at least one file is imported into the workspace
## Tests
<!-- How have you tested the changes? -->
Tests for another change using this improvement are green:
https://github.com/databricks/cli/pull/2463
<!-- If your PR needs to be included in the release notes for next
release,
add a separate entry in NEXT_CHANGELOG.md as part of your PR. -->
## Why
The command execution ID is a UUID that should be initialized once, and
then have a consistent read value for the duration of the command.
As we start emitting telemetry events, we also want to emit the command
execution ID in the telemetry payload. This library helps consolidate
all reads to the `cmd-exec-id` value, and ensures there's only one write
for the UUID.
relevant comment:
https://github.com/databricks/cli/pull/2432#discussion_r1982884235
## Tests
Unit tests.
## Changes
<!-- Summary of your changes that are easy to understand -->
This PR makes changes to the labs code base to allow for offline
installation of labs projects (like UCX). By passing a flag
--offline=true, the code will skip checking for project versions and
download code from GitHub and instead will look from the local
installation folder. This cmd is useful in systems where there is
internet restriction, the user should follow a set-up as follows:
- install a labs project on a machine which has internet
- zip and copy the file to the intended machine and
- run databricks labs install <project name>--offline=true
it will look for the code in the same install directory and if present
load from there.
Closes#1646
related to https://github.com/databrickslabs/ucx/issues/3418
## Tests
<!-- How is this tested? -->
Added unit test case and tested.
NO_CHANGELOG=true
---------
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: Pieter Noordhuis <pieter.noordhuis@databricks.com>
Co-authored-by: Lennart Kats (databricks) <lennart.kats@databricks.com>
Co-authored-by: Denis Bilenko <denis.bilenko@databricks.com>
Co-authored-by: Julia Crawford (Databricks) <julia.crawford@databricks.com>
Co-authored-by: Ilya Kuznetsov <ilya.kuznetsov@databricks.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Andrew Nester <andrew.nester@databricks.com>
Co-authored-by: Anton Nekipelov <226657+anton-107@users.noreply.github.com>
Co-authored-by: shreyas-goenka <88374338+shreyas-goenka@users.noreply.github.com>
## Changes
Removed Changelog check Github Action
## Why
Check for the Changelog and the necessity to add `NO_CHANGELOG=true`
appears to be quite noisy, so removing this
NO_CHANGELOG=true
## Changes
Fixes#1888
## Why
`databricks queries list` is using override to define a template to
display a list in a nice table form. This Go template was referencing to
non existing anymore fields and hence failing.
## Tests
Added acceptance test
## Changes
Update `docsgen` output to align with new documentation framework
1. New header style, previously we had brackets for `<name>`, now it's
in italic `_name_`
2. Updated broken markdown in OpenAPI descriptions
3. Table markdown has different structure in new framework
## Tests
Updated existing tests
NO_CHANGELOG=true
## Changes
Move the WorkspaceClient reader and setter from the root package to
`libs/command`.
## Why
This allows us to read the workspace client set in the context in all
CLI packages. This was not possible before because using
`root.WorkspaceClient` would often result in an import cycle.
This would also allow us to standardise reads of the workspace client to
be from the context in bundle commands. Today those are read from the
bundle object tree instead.
This is a natural followup to
https://github.com/databricks/cli/pull/2440.
## Tests
Existing tests and one new unit test.
NO_CHANGELOG=true
## Changes
<!-- Brief summary of your changes that is easy to understand -->
1. Created a new acceptance test for binding an existing schema to a
bundle
2. Added a feature flag support for acceptance test runner
## Why
<!-- Why are these changes needed? Provide the context that the reviewer
might be missing.
For example, were there any decisions behind the change that are not
reflected in the code itself? -->
1. The test is more simple than an existing integration test for bind
2. The test only can run in environments where Unity Catalog is enabled
## Tests
<!-- How have you tested the changes? -->
New test is passing in CI/CD
## Changelog
NO_CHANGELOG=true
## Changes
I forgot to add a unit test in the original PR, so adding it here:
https://github.com/databricks/cli/pull/2440
## Why
Unit testing code is a good practice. I believe we all should unit test
code.
## Tests
N/A
NO_CHANGELOG=true
## Changes
Make NO_CHANGELOG=true a default in PR template
## Why
Quite large amount of our changes are internal and don't need call outs
in CHANGELOG hence this is a good default for majority of our PRs.
Without NO_CHANGELOG=true CI check expects to have an entry in
NEXT_CHANGELOG file as part of same PR.
NO_CHANGELOG=true
## Changes
This PR moves `ConfigUsed` from the root package to `libs/command`.
## Why
Having the ConfigUsed function in the root package is a problem because
that means we cannot use that function from outside the `root` package
since doing so often leads to an import cycle (because `root` imports
everything implicitly).
Moving it to a separate package that consolidates the read interaction
and solves the import cycle issue. Example where this would have
simplified code:
https://github.com/databricks/cli/pull/2432#discussion_r1983368092
I'd like to send PRs to do the same for the workspace client and account
client as well. I'll wait however until this one is merged incase people
have concerns about the approach here.
## Tests
Existing tests.
## Changes
Enable Automated tagging workflow. The workflow uses the `tagging.py`
script to:
* Flush the NEXT_CHANGELOG.md` file into the `CHANGELOG.md`.
* Create a Release commit and push it to `main` (without a PR)
* Create a tag for the release.
Usage and Changelog management details can be found in
[go/deco/release](go/deco/release).
This PR does:
* Added `cli_legacy` mode. Each SDK has it's own "mode" used to
determine which templates are to be used.
* `api_changelog: true` enables API changelog to be appended to
NEXT_CHANGELOG when generating the SDK.
* Removed `.codegen/changelog.md.tmpl`. Changelog templates are now part
of universe.
* Workflow and `tagging.py` files are synced from universe as part of
SDK generation. This allows us to test them and keep them in sync with
other SDKs.
## Tests
Workflow enabled in databricks/terraform-provider-databricks
---------
Co-authored-by: Andrew Nester <andrew.nester@databricks.com>
## Changes
This PR also starts storing the DBR versions in the context.
Go patch file used:
```
@@
var x expression
@@
-dbr.MockRuntime(x, true)
+dbr.MockRuntime(x, dbr.Environment{IsDbr: true, Version: "15.4"})
@@
var x expression
@@
-dbr.MockRuntime(x, false)
+dbr.MockRuntime(x, dbr.Environment{})
```
ref: https://github.com/uber-go/gopatch
## Why
This localised all DBR version accesses to `libs/dbr`. Relevant comment:
https://github.com/databricks/cli/pull/2432#discussion_r1982878616
## Tests
Exiting tests are modified.
## Changes
- New library to patch existing whl with dynamic version. The version is
based on mtime of original wheel.
- New hidden command "selftest patchwhl" (for acceptance test).
## Why
Part of the project to deprecate dynamic version in setup.py template
(#1034). This would enable switching default-python to pyproject.toml
without dependency on setuptools-specific sections.
The dynamic version is needed to ensure that wheels are picked up by the
users code when doing development.
## Tests
- New unit tests that use uv to create envs on various versions and
install (and re-install) patched wheels there.
- New acceptance test that patches prebuilt whl and records the result
fully unzipped.
- New acceptance test helper setmtime.py to set mtime with nanosecond
precision cross-platform.
## Changes
The `schema` field implies the lifecycle of tables is no longer tied to
the lifecycle of the pipeline, as was the case with the `target` field.
More information about using the "catalog" and "schema" properties can
be found here:
https://docs.databricks.com/en/delta-live-tables/target-schema.html
## Tests
n/a
---------
Co-authored-by: Lennart Kats (databricks) <lennart.kats@databricks.com>
## Changes
- Do not modify or edit .gitignore in bundle root.
- Instead create .databricks/.gitignore with content set to "*"
## Why
Merging our changes into existing .gitignore is complicated and adding
.gitignore where it's not expected adds to the noise. Other tools also
use the approach in this PR (e.g. ruff creates .ruff_cache/.gitignore).
## Tests
- Modified templates/default-sql to capture this new file.
## Changes
- Added `service-principal-secrets` command
- Added `budget-policy-id` for apps
- `experiments.log-inputs` now requires `ID` parameter as an input
- Added `genie.get-space` command
- Added `providers.list-provider-share-assets` command
For the whole list of SDK changes see:
https://github.com/databricks/databricks-sdk-go/releases/tag/v0.59.0
## Why
VSCode for me resolves `script` as YAML files showing incorrect syntax
coloring. This setting forces VSCode to treat all files named `script`
as shell scripts leading to better syntax highlighting.
These files are common place in our acceptance tests.
## Tests
Manually.
There is no perf difference, it's fast in both cases:
```
~/work/cli % hyperfine 'ruff format -qn'
Benchmark 1: ruff format -qn
Time (mean ± σ): 22.5 ms ± 4.1 ms [User: 23.6 ms, System: 106.7 ms]
Range (min … max): 17.3 ms … 50.2 ms 52 runs
```
but it no longer produces .ruff_cache which can confuse acceptance test
runner
```
~/work/cli % find . -type d | grep ruff
./.ruff_cache
./.ruff_cache/0.8.3
./libs/notebook/testdata/.ruff_cache
./libs/notebook/testdata/.ruff_cache/0.8.3
./libs/sync/testdata/.ruff_cache
./libs/sync/testdata/.ruff_cache/0.8.3
./acceptance/bundle/templates/default-sql/.ruff_cache
./acceptance/bundle/templates/default-sql/.ruff_cache/0.8.3
```
## Changes
Switch to use GET workspaces-files/{name} instead of workspace/export
for state files.
## Why
`/api/2.0./workspaces-files/{name}` has a higher limit which allows to
export state files larger than 10 MBs (which is the current limit for
`workspace/export`). We don't use the same API for read in other places
and fully replacing existing Filer because it doesn't correct get the
file content for notebooks and returns "File Not Found" error instead.
## Tests
All existing tests pass
## Changes
This function provides all environment variables necessary to
authenticate the downstream applications to the same credentials as the
parent process.
It's used in https://github.com/databricks/cli/pull/2278 and will also
be useful for the `databricks bundle exec` command.
## Tests
Unit test.
## Changes
1. Changed `FindResourceByConfigKey` to return schema resources
2. Implemented the `Exists` method for schema resources
## Why
This PR adds support for schema resources in deployment operations,
enabling users to:
- Bind schemas using `databricks bundle deployment bind <myschema_key>
<schema_full_name>`
- Unbind schemas using `databricks bundle deployment unbind
<myschema_key>`
Where:
- `myschema_key` is a resource key defined in the bundle's .yml file
- `schema_full_name` references an existing schema in the Databricks
workspace
These capabilities allow for more flexible resource management of
schemas within bundles.
## Tests
Added a new integration test that tests bind and unbind methods together
with bundle deployment and destruction.
## Changes
1. remove t.Skip directive from TestAuthDescribeFailure integration test
2. remove checking the host in the test
## Why
1. This enables integration test that exercises `databricks auth
describe` command, which was previously throwing errors in the CI/CD
pipeline. For the nonexistent profile check to fail, we need to create a
new profile file for the CLI to check (otherwise the configuration skips
checking of the provided profile actually exists).
2. I removed the assertion for the host part because it is not relevant
for the test, and setting that up just required more code in the test
setup.
## Tests
The integration test is passing
## Changes
Added a warning when `config` section is used in apps
## Why
To avoid the confusion between using apps in DABs and outside of DABs,
we want to provide only one way of configuring apps runtime
configuration - by using `app.yml` file in the root of the app.
## Tests
Added acceptance tests
## Changes
Lock setuptools version to 75.8.2 (latest as of March 3, 2025)
## Why
As part of the tests `uv install` was installing latest version of
setuptools which led to all tests started to fail on Feb 25 when 75.8.1
setuptools version was released and which changed the naming of the
output built artifacts
https://setuptools.pypa.io/en/stable/history.html#v75-8-1
This change prevents us from breakages like this.
## Tests
Existing tests pass
## Changes
Same name libraries check only valid for local libraries. Local
libraries are only supported for Whl and Jar types. Hence we can
restrict matching pattern only to these libraries.
## Tests
Existing acceptance tests pass
## Changes
This PR adds the auth.EnvVars function, which is a list of all
environment variables that the SDK uses to read auth configuration.
This is useful for spawning subprocesses since you can unset all auth
environment variables to clean up the environment before configuring the
auth.
It's used in #2278 today and will also be useful for the `databricks
bundle exec` command.
## Tests
Unit test.
## Changes
Add mutex synchronization in cmdio logger Log() method.
## Why
Since we issue multiple calls to underlying writer, we should lock the
whole method, otherwise we can get broken messages. One that can be
easily reproduced today is
```
hyperfine -m 100 --show-output 'go test ./acceptance -run ^TestAccept$/^bundle$/^artifacts$/^whl_multiple$ -count=1'
...
-Uploading my_test_code-0.0.1-py3-none-any.whl...
-Uploading my_test_code_2-0.0.1-py3-none-any.whl...
+Uploading my_test_code-0.0.1-py3-none-any.whl...Uploading my_test_code_2-0.0.1-py3-none-any.whl...
Error: Command terminated with non-zero exit code 1 in benchmark iteration 54. Use the '-i'/'--ignore-failure' option if you want to ignore this. Alternatively, use the '--show-output' option to debug what went wrong.
```
An alternative could be to prepare a message fully in a local buffer and
write it in one call (I’m assuming underlying writer is still
synchronized). However, that’s more complicated and unclear if it’s
worth it, perf-wise.
## Tests
With this change I’m running the same hyperfine command with 1000
iterations with no failures.
## Changes
- Remove bundle.Parallel & bundle.ReadOnlyBundle.
- Add bundle.ApplyParallel, as a helper to migrate from bundle.Parallel.
- Keep ReadOnlyMutator as a separate type but it's now a subtype of
Mutator so it works on regular *Bundle. Having it as a separate type
prevents non-readonly mutators being passed to ApplyParallel
- validate.Validate becomes a function (was Mutator).
## Why
This a follow up to #2390 where we removed most of the tools to
construct chains of mutators. Same motivation applies here.
When it comes to read-only bundles, it's a leaky abstraction -- since
it's a shallow copy, it does not actually guarantee or enforce readonly
access to bundle. A better approach would be to run parallel operations
on independent narrowly-focused deep-copied structs, with just enough
information to carry out the task (this is not implemented here, but the
eventual goal). Now that we can just write regular code in phases and
not limited to mutator interface, we can switch to that approach.
## Tests
Existing tests.
---------
Co-authored-by: shreyas-goenka <88374338+shreyas-goenka@users.noreply.github.com>
Anton Nekipelov
Denis Bilenko
dependabot[bot]
Andrew Nester
Gleb Kanterov
shreyas-goenka
Hari Selvarajan
Ilya Kuznetsov
hectorcast-db
Pieter Noordhuis
Russell Clarey