- Replace 'a' with 'before' and 'b' with 'after' in BuildPlans doc comment for clarity
- Fix backwards compatibility test case by swapping before/after files:
- before.yaml now correctly has fewer fields (older version)
- after.yaml now correctly has more fields (newer version with enhancements)
- Update comparison logic to filter fields from after that don't exist in before
- This correctly implements spec 4: after may have fields missing from before when backwards compatible
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Add test case BuildPlan_4 for BuildPlan spec 4: b may have fields missing from a if isBackwardsCompatible is true
- Implement backwards compatibility logic in comparison functions
- Add filterToCommonFields() to filter fields based on what exists in both structures
- Update compareDocumentLists() and compareStructures() to accept isBackwardsCompatible parameter
- When backwards compatible mode is enabled, only compare fields that exist in both structures
- Update test framework to support isBackwardsCompatible field in test cases
- Test verifies that missing labels, annotations, and nested fields don't cause failure in backwards compatible mode
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Reference BuildPlan spec items in test names (BuildPlan_3.1, BuildPlanFile_1, etc.)
- Add missing test cases for BuildPlan spec requirements:
- BuildPlanFile_2: each object matches exactly one unique object
- BuildPlan_5: fields in 'a' must exist in 'b'
- BuildPlan_6.1-6.3: null/empty/missing field equivalence
- BuildPlanValid_1-2: duplicate objects and same metadata.name
- Implement null/empty/missing field equivalence (spec 6)
- Add normalizeStructure() to handle spec 6 requirements
- Ensure null, [], and missing fields are treated as equivalent
- Update test messages to clarify spec coverage intent
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Add comprehensive test cases to cover all specification requirements:
- File spec requirement 1: equal number of BuildPlans
- BuildPlan spec 1&2: all fields must be equivalent
- BuildPlan spec 3: nested objects must match recursively
- BuildPlan spec 4: exact field value matching with exceptions
- BuildPlan spec 4.1: unordered spec.artifacts list
- BuildPlan spec 4.2: key order irrelevance
- Toleration 1: multiple identical BuildPlans treated as unique
- Toleration 2: multiple BuildPlans with same metadata.name
Update existing test names and messages to clearly reference the specific specification requirements they validate.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
This patch adds the output of the new stdio concurrenet compiler being
developed for v1alpha6 as after, which is semantically equivalent to the
v0.104.1 v1alpha5 holos show buildplans output.
Note the field orders and relative object order in the stream is quite
different.
- Refactor test infrastructure to support expectedErrors as a list instead of single expectedError
- Implement generic field difference extraction for any object additions
- Update extractTestErrors method to handle arbitrary field additions
- Migrate all test cases from expectedError to expectedErrors format
- Maintain backward compatibility with existing test expectations
- Add support for detecting and formatting any field additions in diff output
- Clean up go-cmp type annotations in error messages for better readability
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Implement the name field for all test cases to provide more descriptive
test names in the test output. This improves test readability and
maintenance.
- Add name field to testCase struct
- Update test logic to use name field when provided
- Add descriptive names to all test cases
- Maintain nested test structure with directory/name hierarchy
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Enhance test cases with descriptive messages that explain what each
test is verifying. These messages are displayed when tests fail,
improving debugging experience.
- Add msg field to testCase struct
- Update all testcase.json files with descriptive messages
- Use messages in test assertions for better failure output
- Ensure all tests continue to pass with new structure
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Simplify test case directory name from empty-files to empty.
- Rename testdata/empty-files/ to testdata/empty/
- Test continues to pass with cleaner naming
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Rename test configuration files from want.json to testcase.json for
better clarity of purpose. Updated test code to match the new naming.
- Rename want.json to testcase.json in testdata
- Update variable names from wantData to testcaseData
- Update error messages to reference testcase.json
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Move compare functionality from CLI package to dedicated internal/compare
package for better separation of concerns. CLI now acts as a thin wrapper
calling the compare package implementation.
- Create internal/compare package with Comparer type
- Move tests from CLI to compare package
- Update test fixtures path to testdata directory
- Replace CLI error with errors.NotImplemented()
- Clean up redundant test files from CLI package
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
