* Start threading job output to RunStepRunner
* Strip ANSI
* Fix lint
* Use waitgroup to avoid test flakiness
* Move waitgroup higher
* Add ANSI test and use strings.Builder
* Fix lint
* Use errors.Wrap per style guide
* Create ShellCommandRunner to encapsulate streaming
* WIP: shell command runner
* Update signatures to propagate error finding version
* Fix log output
* Fix error checking
* Fix accidental whitespace stripping
* Remove unused struct field
* Fix error checking in terraform client
* Add unit tests to verify command output handler was called
* Remove err from async interface
* Remove duplicative log now that shell command runner does it
* Hide output in stream for env/multienv
* Add comment explaining goroutines
* Use printf for better macOS compatibility
* fix: add path to WorkingDir methods
`WorkingDirLocker` and `WorkingDir` are closely related -- the former acquires a lock, which ensures that the latter can safely operate on a given file path.
In #2131 we added `path` to the `WorkingDirLocker` lock key, but neglected to add the same to `WorkingDir`.
This commit adds `path` as an argument to certain `WorkingDir` methods, and includes `path` in the directory that we use to clone the repository for a given project.
Since `path` can include certain special characters such as `/`, we encode `path` as base32 when using it as part of a file path. This should ensure no special characters are used in the filesystem, and that the value can be decoded if desired (unlike hashes such as md5).
Additional changes:
- All calls to changed methods have been updated, including unit tests
- Mocks have been regenerated for `WorkingDir` and `WorkingDirLocker`
- `working_dir_test.go`
- Commands that operate on filesystem paths have been updated to include the base32 encoded `path` value
- When running `git init`, we include `-b master` as additional arguments, to ensure that `master` is our default branch (expected by the unit tests)
* Try fixing E2E tests
* Fix DefaultPendingPlanFinder
In addition to iterating over `workspaceDirs`, also iterate over `pathDirs`, and use both `workspace` and `path` to build `repoDir`.
* Fix DefaultPendingPlanFinder unit tests
* Fix DefaultProjectCommandBuilder unit tests
Co-authored-by: Kevin Snyder <kevinsnyder@KevinSnydersMBP.lan>
Co-authored-by: Kevin Snyder <kevinsnyder@ip-192-168-1-188.ec2.internal>
* fix: add path to DefaultWorkingDirLocker.TryLock()
In the release notes for v0.13.0: https://github.com/runatlantis/atlantis/blob/master/CHANGELOG.md#features-4
> Running in parallel is only supported if you're using workspaces to separate your projects. Projects in separate directories can not be run in parallel currently.
This commit adds `path` as an argument to `DefaultWorkingDirLocker.TryLock()` and includes `path` in the `workspaceKey` used to check if a project is locked. This should allow projects in separate directories to run in parallel.
To my knowledge, there is no functional reason that projects in separate directories cannot run in parallel.
All calls to `DefaultWorkingDirLocker.TryLock()` have been updated. A new unit test `TestTryLockWithDifferentPaths` has been added to test the behavior of locking two separate directories with the same workspace name.
* Add documntation for parallel_plan and parallel_apply options
Co-authored-by: Kevin Snyder <kevinsnyder@ip-192-168-4-61.ec2.internal>
Co-authored-by: Kevin Snyder <kevinsnyder@ip-10-60-10-94.ec2.internal>
* Moved CommandContext and CommandResult to models (#193)
* Moved CommandContext and CommandResult to models
* move from models to command
rename CommandContext -> Context
rename CommandResult -> Result
* moved command related helpers into command package
* move ProjectCommandContext and ProjectResult to command/project package
* move project command context and project result
* revert unrelated code
* move tests
* fix left over
* fix linting
* fix tests
* remove unused import
* fix project context dependencies
* fix depenedecies
* fix typo
* Add UUID for Log Streaming Job ID (#167)
* Update log handler to close buffered channels when an operation is complete (#170)
* Add preliminary check before registering new receivers in the log handler (#173)
* Using projectOutputBuffers to check for jobID instead of receiverBuffers (#181)
* Refactor log handler (#175)
* Reverting go.mod and go.sum
* Fix lint errors
* Fix linting
* ORCA-731 Force PR authors to rebase if master has diverged (#70)
* initial commit for undiverge apply req
* Update documentation
* Fix existing unit tests
* HasDiverged unit test
* project command runner unit test
* Global config tests cases
* update parser unit tests
* Update project_test unit tests
* Fix linting
* fix test case with logging output
* add space on readme markdown
* Remove unused variable
* remove deprecated flag option
* Use global args struct instead of passing in args
* Fix merge issue
* Update go-generate to run as a script
It was taking forever to run via the previous Makefile command. I don't
exactly know why. Perhaps because it was passing all the packages in as
multiple arguments?
Either way, this change allows it to actually run through in about 20
minutes.
* Regenerate all mocks using latest pegomock
Pegomock version v2.9.0
* Format mocks outside of regular mocks directory.
These mocks are outside of the regular directory due to an import cycle
causing us to have to have these not under a /mocks directory so they
need to be formatted since they're not excluded.
* Add back make test target
This is two changes:
Replacing all usages of SimpleLogger with it's interface SimpleLogging which makes it easier to swap out loggers going forward
Nukes SimpleLogger in favor of StructuredLogger which allows us to better analyze logs in our log management system we choose.
* Adding policy_check support into yaml config
* Added policy check model and runtime structs
* Adding BuildPolicyCheckCommand to ProjectCommandBuilder
* Return incorrectly deleted code
* Remove BuildAutoPolicyPlanCommand from ProjectCommandBuilder
* Split runAutoCommand into two functions
runAutoPlanCommand - does what originally RunAutoplancommand was doing
except now it returns CommandResult and []models.ProjectCommandContext
runAutoPolicyCheckCommand - accepts CommandContext, CommandResult, and
[]models.ProjectCommandContext as arguments and runs PolicyCheckStep runner
* Refactor RunCommentCommand
* Remove BuildPolicyCheckCommand and rename StepCmdExec back to TerraformExec
* Add policy step runner logic and conftest interfaces.
* Add show step runner to policy check stage.
* Adding models.PolicyCheckCommand to buildCtx
This also means buildPlanAllCommands call buildCtx twice once with
models.PlanCommand and once with models.PolicyCheckCommand
* Adding new project_command_builder that supports policy_check
* Refactoring PolicyCheck specific logic into a PolicyCheckProjectCommandBuilder
* Moving events.CommandContext to models.CommandContext this will allow me
to remove buildCtx method and move ProjectCommandContext creation into
models package
* Policy Owners might be different types, for that reason we are
refactoring Owners into its own struct with specific keys defining
different owner types
Co-authored-by: Nish Krishnan <nishk@lyft.com>
Co-authored-by: Nish Krishnan <nishkrishnan@users.noreply.github.com>
This version changes how we parse custom run steps.
Previously we used the go-shlex library to parse the steps but this
caused users issues because they'd have to use weird escaping.
Now we simply pass the step directy to sh -c without any parsing. The
parsing was unnecessary.
Split single top-level status check into one plan and one apply check.
These checks will also now track the whole pull request status rather
than just the last action.
Features:
- remote plans and applies will set a commit status with a link to the
Terraform Enterpise UI where the streaming output can be viewed.
- because remote plans don't support the -out flag, we diff the plan
output between plan and apply phases to ensure users are applying the
plan they expected
How:
We add a new capability to execute a terraform command while
concurrently reading its real-time output and performing actions
accordingly.
Automerging merges pull requests automatically if all plans have been
successfully applied.
* Save status of PR's to BoltDB so after each apply, we can check if
there are pending plans.
* Add new feature where we delete successful plans *unless* all plans
have succeeded *if* automerge is enabled. This was requested by users
because when automerge is enabled, they want to enforce that a pull
request's changes have been fully applied. They asked that plans not be
allowed to be applied "piecemeal" and instead, all plans must be
generated successfully prior to allowing any plans to be applied.
