This commit introduces a resilient, configurable retry mechanism for network requests.
Key changes include:
- A new `pkg/retry` package with a custom `http.Transport` that implements exponential backoff and jitter.
- Integration of the retry transport into the `website`, `pwa`, and `github` packages to handle transient network failures gracefully.
- New persistent CLI flags (`--retries`, `--retry-backoff`, `--retry-max`, `--retry-jitter`, `--no-retry`) to allow user configuration of the retry behavior.
- The flag-handling logic has been moved to a `PersistentPreRun` function to ensure user-provided values are parsed correctly.
- A basic retry mechanism has been added to the `vcs` package for git clone operations.
- Added unit tests for the retry transport.
This work is in progress, with the next steps being to implement support for the `Retry-After` header and unify the VCS retry logic with the global configuration.
Co-authored-by: Snider <631881+Snider@users.noreply.github.com>
Refactored the existing tests to use the `_Good`, `_Bad`, and `_Ugly`
testing convention. This provides a more structured approach to testing
and ensures that a wider range of scenarios are covered, including
valid inputs, invalid inputs, and edge cases.
In addition to refactoring the tests, this change also includes several
bug fixes that were uncovered by the new tests. These fixes improve the
robustness and reliability of the codebase.
The following packages and commands were affected:
- `pkg/datanode`
- `pkg/compress`
- `pkg/github`
- `pkg/matrix`
- `pkg/pwa`
- `pkg/vcs`
- `pkg/website`
- `cmd/all`
- `cmd/collect`
- `cmd/collect_github_repo`
- `cmd/collect_website`
- `cmd/compile`
- `cmd/root`
- `cmd/run`
- `cmd/serve`
This commit introduces a significant refactoring of the `cmd` package to improve testability and increases test coverage across the application.
Key changes include:
- Refactored Cobra commands to use `RunE` for better error handling and testing.
- Extracted business logic from command handlers into separate, testable functions.
- Added comprehensive unit tests for the `cmd`, `compress`, `github`, `logger`, and `pwa` packages.
- Added tests for missing command-line arguments, as requested.
- Implemented the `borg all` command to clone all public repositories for a GitHub user or organization.
- Restored and improved the `collect pwa` functionality.
- Removed duplicate code and fixed various bugs.
- Addressed a resource leak in the `all` command.
- Improved error handling in the `pwa` package.
- Refactored `main.go` to remove duplicated logic.
- Fixed several other minor bugs and inconsistencies.
This commit introduces a significant refactoring of the `cmd` package to improve testability and increases test coverage across the application.
Key changes include:
- Refactored Cobra commands to use `RunE` for better error handling and testing.
- Extracted business logic from command handlers into separate, testable functions.
- Added comprehensive unit tests for the `cmd`, `compress`, `github`, `logger`, and `pwa` packages.
- Added tests for missing command-line arguments, as requested.
- Implemented the `borg all` command to clone all public repositories for a GitHub user or organization.
- Restored and improved the `collect pwa` functionality.
- Removed duplicate code and fixed various bugs.
This commit refactors the codebase to use dependency injection for mocking external dependencies, removing the need for in-code mocking with the `BORG_PLEXSUS` environment variable.
- Interfaces have been created for external dependencies in the `pkg/vcs`, `pkg/github`, and `pkg/pwa` packages.
- The `cmd` package has been refactored to use these interfaces, with dependencies exposed as public variables for easy mocking in tests.
- The tests in `TDD/collect_commands_test.go` have been updated to inject mock implementations of these interfaces.
- The `BORG_PLEXSUS` environment variable has been removed from the codebase.
This commit addresses several issues in the testing framework:
- Corrected a malformed URL in the mock response data in `pkg/github/github.go`.
- Fixed a bug in the mock HTTP client in `pkg/mocks/http.go` where the response body was being exhausted after the first read. The client now returns a deep copy of the response for each call.
- Refactored the environment variable handling in `TDD/collect_commands_test.go` to use `t.Setenv`, ensuring proper cleanup after the test.
This commit introduces a TDD testing framework for the `collect` commands and resolves a build failure.
- A `TDD/` directory has been added to house tests for the `collect` commands.
- An environment variable `BORG_PLEXSUS=0` has been implemented to enable a mock mode, which prevents external network calls during testing.
- The `collect` commands have been updated to use the command's output streams, allowing for output capturing in tests.
- A `pkg/mocks` package has been added to provide mock implementations for testing.
- The `.gitignore` file has been updated to exclude generated `.datanode` files.
- The "flag redefined" build error has been fixed by refactoring the root command initialization in `cmd/root.go` to prevent duplicate flag definitions.
This commit introduces a TDD testing framework for the `collect` commands.
- A `TDD/` directory has been added to house the tests.
- An environment variable `BORG_PLEXSUS=0` has been implemented to enable a mock mode, which prevents external network calls during testing.
- The `collect` commands have been updated to use the command's output streams, allowing for output capturing in tests.
- A `pkg/mocks` package has been added to provide mock implementations for testing.
- The `.gitignore` file has been updated to exclude generated `.datanode` files.
This commit addresses several issues identified in a code review to improve the overall quality and robustness of the application.
Key changes include:
- Added safe type assertions with `nil` checks when retrieving the logger from the context to prevent panics.
- Moved the `bar.Finish()` call to be inside the loop in the `all` command, so each progress bar finishes after its corresponding repository is cloned.
- Added a check for context cancellation at the start of the pagination loop in the GitHub client to prevent unnecessary API calls.
- Ensured the authenticated client is used consistently, even when falling back to the organization endpoint.
- Added `nil` checks for the progress bar parameter in the `website` and `pwa` packages to prevent panics.
- Updated the `golang.org/x/oauth2` dependency to a patched release to address a reported vulnerability.
This commit introduces two key improvements to the application:
1. **Authenticated GitHub API Access:** The GitHub client now uses a personal access token (PAT) from the `GITHUB_TOKEN` environment variable if it is available. This increases the rate limit for GitHub API requests, making the tool more robust for users who need to collect a large number of repositories.
2. **Structured Logging:** The application now uses the standard library's `slog` package for structured logging. A `--verbose` flag has been added to the root command to control the log level, allowing for more detailed output when needed. This makes the application's output more consistent and easier to parse.
This commit addresses several issues identified in a code review to improve the overall quality and robustness of the application.
Key changes include:
- Refactored `cmd.Execute()` to return an error instead of calling `os.Exit`, making the application more testable.
- Fixed critical issues in `cmd/main_test.go`, including renaming `TestMain` to avoid conflicts and removing the brittle E2E test.
- Improved the GitHub API client in `pkg/github/github.go` by:
- Fixing a resource leak where an HTTP response body was not being closed.
- Restoring a parameterized function to improve testability.
- Adding support for `context.Context` and API pagination for robustness.
- Updated the `.github/workflows/go.yml` CI workflow to use the `Taskfile.yml` for building and testing, ensuring consistency.
- Added a `test` task to `Taskfile.yml`.
- Ran `go mod tidy` and fixed several unused import errors.
This commit fixes a critical build issue where the application was being compiled as an archive instead of an executable. This was caused by the absence of a `main` package.
The following changes have been made to resolve this and improve the development process:
- A `main.go` file has been added to the root of the project to serve as the application's entry point.
- A `Taskfile.yml` has been introduced to standardize the build, run, and testing processes.
- The build process has been corrected to produce a runnable binary.
- An end-to-end test (`TestE2E`) has been added to the test suite. This test builds the application and runs it with the `--help` flag to ensure the binary is always executable, preventing similar build regressions in the future.
This commit introduces several maintenance improvements to the repository.
- A `go.work` file has been added to define the workspace and make the project easier to work with.
- The module path in `go.mod` has been updated to use a GitHub URL, and all import paths have been updated accordingly.
- `examples` and `docs` directories have been created.
- The `examples` directory contains scripts that demonstrate the tool's functionality.
- The `docs` directory contains documentation for the project.
- Tests have been added to the `pkg/github` package following the `_Good`, `_Bad`, `_Ugly` convention.
- The missing `pkg/borg` package has been added to resolve a build error.
This commit introduces the core functionality of the Borg Data Collector.
- Adds the `collect` command to clone a single Git repository and store it in a Trix cube.
- Adds the `collect all` command to clone all public repositories from a GitHub user or organization.
- Implements the Trix cube as a `tar` archive.
- Adds the `ingest` command to add files to a Trix cube.
- Adds the `cat` command to extract files from a Trix cube.
- Integrates Borg-themed status messages for a more engaging user experience.
