Create detection logic package for development dependencies

[Copilot]

Create Detection Logic Package

Implementing detection logic for development dependencies (cargo-machete, OpenTofu, Ansible, LXD) following the linting package pattern.

Implementation Plan

• Create package structure at packages/dependency-installer/
  • Create directory structure
  • Create Cargo.toml with dependencies (tracing, thiserror)
  • Add to workspace Cargo.toml
  • Create README.md
• Define core abstractions
  • ToolDetector trait in src/detector/mod.rs
  • DetectionError enum in src/errors.rs
  • Dependency enum for tool types
  • CheckResult struct for detection results
• Implement command utilities in src/command.rs
  • command_exists() function
  • execute_command() function
  • CommandError type
• Implement 4 detector structs
  • CargoMacheteDetector in src/detector/cargo_machete.rs
  • OpenTofuDetector in src/detector/opentofu.rs
  • AnsibleDetector in src/detector/ansible.rs
  • LxdDetector in src/detector/lxd.rs
• Create DependencyManager in src/manager.rs
  • new() to create with all detectors
  • check_all() to check all dependencies
  • get_detector() to get specific detector
• Add library exports in src/lib.rs
• Write comprehensive unit tests
  • Test each detector implementation
  • Test DependencyManager functionality
  • Test error handling
• Add working example demonstrating usage
• Run linting, build, and verify
  • cargo build ✓
  • cargo test ✓ (22 tests + 2 doctests passing)
  • cargo run --bin linter all ✓
• Address code review feedback
  • Document Unix-specific which command usage
  • Clarify detector creation pattern in DependencyManager
• Fix clippy linting issue
  • Use inline format args in example

Summary

Successfully created the packages/dependency-installer package with:

• ToolDetector trait: Abstract interface for tool detection with name() and is_installed() methods
• 4 Detector implementations: CargoMacheteDetector, OpenTofuDetector, AnsibleDetector, LxdDetector
• DependencyManager: Orchestrates all detectors with check_all() and get_detector() methods
• Command utilities: Helper functions for command execution (command_exists, execute_command)
• Comprehensive testing: 22 unit tests + 2 doc tests covering all functionality
• Error handling: Clear error types (DetectionError, CommandError) with context
• Logging: Structured logging with tracing for all operations
• Documentation: Platform support notes and implementation clarifications
• Example: Working example demonstrating dependency checking

All linters pass, all tests pass, workspace builds successfully, and the package follows the linting package pattern.

Security Summary

No security vulnerabilities identified. The package performs read-only operations:

• Executes which command to check for tool existence
• No file system modifications
• No network operations
• No user input processing
• No credential handling

Next Steps

This completes Phase 1 of Issue #114. The next phases will add:

• Phase 2 (Issue Create CLI Binary with Check Command #115): CLI binary with check command
• Phase 3 (Issue Create Docker Test Infrastructure #116): Docker test infrastructure
• Phase 4 (Issue Implement Installation Logic #117): Installation logic

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Create Detection Logic Package</issue_title>
<issue_description>Parent Issue: #113 - Create Dependency Installation Package for E2E Tests
Depends On: None (first phase)
Epic: #112 - Refactor and Improve E2E Test Execution

Overview

Create the package structure and implement detection logic to check if development dependencies (cargo-machete, OpenTofu, Ansible, LXD) are installed.

This is Phase 1 of 4 for building the dependency installation package.

Objectives

• Create packages/dependency-installer/ structure following packages/linting/ pattern
• Define ToolDetector trait for detection abstraction
• Implement 4 detector structs (CargoMachete, OpenTofu, Ansible, LXD)
• Create DependencyManager to coordinate detectors
• Add comprehensive unit tests with mocked commands
• Set up tracing for observability

Key Components

ToolDetector Trait:

trait ToolDetector {
    fn name(&self) -> &str;
    fn is_installed(&self) -> Result<bool, DetectionError>;
}

4 Detector Implementations:

• CargoMacheteDetector - checks cargo machete --version
• OpenTofuDetector - checks tofu --version
• AnsibleDetector - checks ansible --version
• LxdDetector - checks lxc --version

DependencyManager:

• Coordinates all detectors
• Provides check_all() method returning status for all tools

Acceptance Criteria

• Pre-commit checks pass
• Package structure follows linting package pattern
• ToolDetector trait is well-defined
• All 4 detectors are implemented with proper error handling
• DependencyManager works correctly
• Unit tests cover all detectors with mocked commands
• Logging provides clear visibility

Time Estimate

2-3 hours

Related Documentation

• Full specification: docs/issues/114-1-1-1-create-detection-logic-package.md
• Linting package reference: packages/linting/
• Error handling guide: docs/contributing/error-handling.md</issue_description>

Comments on the Issue (you are @copilot in this section)

• Fixes Create Detection Logic Package #114

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

[josecelano]

@copilot the linting workflow has failed:

2025-11-03T21:24:11.172995Z  INFO clippy: Running Rust Clippy linter...
error: variables can be used directly in the `format!` string

  --> packages/dependency-installer/examples/check_dependencies.rs:39:13
   |
39 |             eprintln!("Error checking dependencies: {}", e);
   |             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
   |
   = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#uninlined_format_args
   = note: `-D clippy::uninlined-format-args` implied by `-D clippy::pedantic`
   = help: to override `-D clippy::pedantic` add `#[allow(clippy::uninlined_format_args)]`
help: change this to
   |
39 -             eprintln!("Error checking dependencies: {}", e);
39 +             eprintln!("Error checking dependencies: {e}");
   |

error: could not compile `torrust-dependency-installer` (example "check_dependencies") due to 1 previous error

2025-11-03T21:24:41.967973Z ERROR clippy: Clippy linting failed. Please fix the issues above.

[Copilot]

@copilot the linting workflow has failed:

Fixed in commit 91d2eb3. Changed the format string to use inline format args: eprintln!("Error checking dependencies: {e}");

[josecelano]

Test Naming Convention Update

I've made a small refinement to align the tests with our project conventions:

Change: Renamed all 22 test functions from test_* pattern to it_should_* pattern as specified in docs/contributing/testing/unit-testing.md.

Examples:

• test_cargo_machete_detector_name → it_should_return_cargo_machete_detector_name
• test_dependency_manager_check_all_runs → it_should_check_all_dependencies_without_error
• test_command_exists_for_known_command → it_should_detect_existing_command

Verification:

• ✅ All 24 tests pass (22 unit tests + 2 doc tests)
• ✅ All linters pass (markdown, YAML, TOML, cspell, clippy, rustfmt, shellcheck)
• ✅ Full test suite passes (1,081 tests)

The implementation is ready to merge once CI workflows complete.

[josecelano]

ACK eaa0acc