Capture values for all tests, make truly drop-in, refactor (#12)

This commit represents a significant evolution of the rspec-enriched_json gem, expanding its capabilities while simplifying its implementation.

The journey involved several pivots and discoveries about RSpec's internals.

Key Changes

1. Capture values for passing tests - Previously only failing tests had their expected/actual values captured. Now all tests benefit from structured data capture.
2. True drop-in replacement - Removed diff stripping functionality that was modifying RSpec's output, making this a pure enhancement rather than a modification.
3. Simplified serialization - Migrated from custom serialization to Oj (Optimized JSON) for better performance and circular reference handling.
4. Fixed predicate matcher support - BePredicate matchers now correctly capture their actual values instead of showing "null".
5. Added negation detection - Captures whether a matcher was used with not_to/to_not for better context.

The Journey

Challenges Faced

- Double-encoding issues - Initial implementation was double-encoding values, resulting in escaped quotes in the output. Fixed by ensuring single serialization pass.
- Memory management - Storing test data in a class variable risked memory leaks. Implemented automatic cleanup after test suite completion.
- Predicate matcher edge case - be_* matchers don't have an actual method, requiring special handling to extract their values.
- Test suite inception - Several specs were running RSpec within RSpec, causing infinite loops. Removed these in favor of focused unit tests.

Tracks Explored and Abandoned

- Safety limits - Started implementing string truncation and array limits, but realized Oj handles this better with its circular reference detection.
- Matcher-specific extraction - Began with custom logic for each matcher type, but discovered the universal wrapper approach was cleaner and more maintainable.
- Diff stripping - Initially modified RSpec's output to remove diffs from messages (since we capture them separately), but this made us not a true drop-in replacement.

Design Decisions

- Oj for serialization - Chose Oj over custom serialization for its robust handling of circular references, better performance, and maintenance-free approach.
- Universal wrapper - Instead of patching specific matchers, we intercept all expectations at the ExpectationHelper level for complete coverage.
- Respect RSpec's color mode - Now honors RSpec's color configuration for diff generation rather than forcing it off.
- Clean up test suite - Removed specs that were testing Oj's behavior rather than our library's functionality.

Technical Details

- Bumped version to 0.8.0 (breaking change due to diff behavior modification).
- Added GitHub Actions CI workflow for automated testing.
- Fixed StandardRB compliance throughout.
- Added comprehensive test coverage for new features.
- Updated documentation to reflect new capabilities.

Breaking Changes

- The formatter no longer strips diffs from exception messages. If you were relying on this behavior, the diffs will now appear in both the exception message and the structured diff field.

Author: raghubetina

.github/workflows/ci.yml

@@ -0,0 +1,38 @@
+name: CI
+
+on:
+  push:
+    branches: [ main ]
+  pull_request:
+    branches: [ main ]
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        ruby-version: ['3.2', '3.3', '3.4']
+    
+    steps:
+    - uses: actions/checkout@v4
+    
+    - name: Set up Ruby
+      uses: ruby/setup-ruby@v1
+      with:
+        ruby-version: ${{ matrix.ruby-version }}
+        bundler-cache: true
+    
+    - name: Run StandardRB
+      run: bundle exec standardrb
+    
+    - name: Run tests
+      run: bundle exec rspec
+      env:
+        COVERAGE: true
+    
+    - name: Upload coverage reports
+      uses: actions/upload-artifact@v4
+      if: always()
+      with:
+        name: coverage-report-${{ matrix.ruby-version }}
+        path: coverage/

.gitignore

@@ -21,13 +21,20 @@
 # Other gems for AI context
 /diffy/
 /super_diff/
+/fuzzy_match_poc/
+/oj/
+/parallel_tests/
+/amazing_print/
 
 # Bundler
 Gemfile.lock
 
-# Test output files (from old complex demo)
+# Test output files
+*.json
 enriched_output.json
 standard_output.json
+builtin_json_output.json
+vanilla_output.json
 
 # IDE
 .idea/
@@ -36,3 +43,4 @@ standard_output.json
 *.swo
 *~
 .DS_Store
+.claude/

.standard.yml

@@ -4,3 +4,7 @@ ignore:
   - 'spec/edge_cases_spec.rb:40'
   - 'spec/safety_limits_spec.rb:52'
   - 'rspec/**/*'
+  - 'fuzzy_match_poc/**/*'
+  - 'oj/**/*'
+  - 'diffy/**/*'
+  - 'super_diff/**/*'

CHANGELOG.md

@@ -5,6 +5,40 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.7.0] - 2025-07-18
+
+### Added
+- Add `negated` flag to detect when `not_to` or `to_not` is used
+- Add `passed` field to distinguish passing from failing tests in captured values
+
+### Changed
+- Major code simplification - removed unnecessary abstractions and comments
+- Updated documentation to reflect actual features (removed non-existent performance limits)
+- **BREAKING**: No longer removes diff from exception messages - now a true drop-in replacement
+
+### Fixed
+- Fixed spec files that were causing false CI failures
+- Updated integration tests to match current behavior
+
+### Removed
+- Removed automatic diff stripping from exception messages (introduced in 0.6.1)
+
+## [0.6.1] - 2025-07-18
+
+### Added
+- Capture expected/actual values for passing tests (not just failures)
+- Memory-safe implementation with cleanup after formatter completes
+- Special handling for Regexp serialization (human-readable format like `/pattern/flags`)
+- Comprehensive test coverage for new features
+
+### Fixed
+- Fixed key mismatch bug between storage and retrieval of test values
+- Fixed double-encoding issue in formatter that caused escaped strings in output
+
+### Changed
+- Upgraded to Oj for JSON serialization (better performance and object handling)
+- Improved error handling with detailed fallback information
+
 ## [0.5.0] - 2025-06-26
 
 ### Changed

CLAUDE.md

@@ -22,7 +22,7 @@ This project creates a universal JSON output system for RSpec matchers:
 - `lib/rspec/enriched_json/formatters/enriched_json_formatter.rb` - JSON formatter that outputs enriched data
 - `rspec-enriched_json.gemspec` - Gem specification
 - `spec/` - Test suite with integration and unit tests
-- `demo_all_failures.rb` - Demo script showing various failure types
+- `demo.rb` - Demo script showing various failure types
 - `Gemfile` - Dependencies (rspec, standard)
 - `.standard.yml` - StandardRB configuration
 

README.md

@@ -1,5 +1,9 @@
 # RSpec::EnrichedJson
 
+[![CI](https://github.com/firstdraft/rspec-enriched_json/actions/workflows/ci.yml/badge.svg)](https://github.com/firstdraft/rspec-enriched_json/actions/workflows/ci.yml)
+[![Gem Version](https://badge.fury.io/rb/rspec-enriched_json.svg)](https://badge.fury.io/rb/rspec-enriched_json)
+[![Ruby Style Guide](https://img.shields.io/badge/code_style-standard-brightgreen.svg)](https://github.com/standardrb/standard)
+
 A drop-in replacement for RSpec's built-in JSON formatter that enriches the output with structured failure data. This makes it easy to programmatically analyze test results, extract expected/actual values, and build better CI/CD integrations.
 
 ## Quick Demo
@@ -30,7 +34,7 @@ This interactive demo script runs the same failing tests with both formatters an
 Add this line to your application's Gemfile:
 
 ```ruby
-gem 'rspec-enriched_json'
+gem "rspec-enriched_json"
 ```
 
 And then execute:
@@ -95,14 +99,14 @@ With this gem, you get structured data alongside the original message:
 
 ## Features
 
-- **Drop-in replacement**: Inherits from RSpec's JsonFormatter, maintaining 100% compatibility
-- **Structured data extraction**: Expected and actual values as proper JSON objects
-- **Rich object support**: Arrays, hashes, and custom objects are properly serialized
-- **Original message preservation**: When you override with a custom message, the original is preserved
-- **Graceful degradation**: Regular exceptions (non-expectation failures) work normally
-- **Enhanced metadata capture**: Test location, tags, hierarchy, and custom metadata
-- **Robust error recovery**: Handles objects that fail to serialize without crashing
-- **Diff information**: Includes `diffable` to help tools determine if values can be meaningfully diffed
+- **Drop-in replacement**: Inherits from RSpec's JsonFormatter, maintaining 100% compatibility.
+- **Structured data extraction**: Expected and actual values as proper JSON objects.
+- **Rich object support**: Arrays, hashes, and custom objects are properly serialized.
+- **Original message preservation**: When you override with a custom message, the original is preserved.
+- **Graceful degradation**: Regular exceptions (non-expectation failures) work normally.
+- **Enhanced metadata capture**: Test location, tags, hierarchy, and custom metadata.
+- **Robust error recovery**: Handles objects that fail to serialize without crashing.
+- **Diff information**: Includes `diffable` to help tools determine if values can be meaningfully diffed.
 
 ## Examples
 
@@ -147,36 +151,50 @@ end
 
 ## Use Cases
 
-- **CI/CD Integration**: Parse test results to create rich error reports
-- **Test Analytics**: Track which values commonly cause test failures  
-- **Debugging Tools**: Build tools that can display expected vs actual diffs
-- **Learning Platforms**: Provide detailed feedback on why tests failed
+- **CI/CD Integration**: Parse test results to create rich error reports.
+- **Test Analytics**: Track which values commonly cause test failures.
+- **Debugging Tools**: Build tools that can display expected vs actual diffs.
+- **Learning Platforms**: Provide detailed feedback on why tests failed.
 
 ## How It Works
 
 The gem works by:
 
-1. Patching RSpec's expectation system to capture structured data when expectations fail
-2. Extending the JsonFormatter to include this data in the JSON output
-3. Maintaining full backward compatibility with existing tools
+1. Patching RSpec's expectation system to capture structured data when expectations fail.
+2. Extending the JsonFormatter to include this data in the JSON output.
+3. Maintaining full backward compatibility with existing tools.
 
 ## Development
 
 After checking out the repo, run `bin/setup` to install dependencies. Then, run `rake spec` to run the tests.
 
-## Performance Considerations
+This project uses [StandardRB](https://github.com/standardrb/standard) for code formatting and style. Before committing:
+
+```bash
+# Check for style violations
+bundle exec standardrb
+
+# Auto-fix style violations
+bundle exec standardrb --fix
+```
+
+## Additional Features
+
+### Passing Test Value Capture
+The formatter also captures expected/actual values for passing tests, useful for:
+- Test analytics and insights.
+- Understanding test coverage patterns.
+- Debugging flaky tests.
 
-The enriched formatter adds minimal overhead:
-- Only processes failing tests (passing tests have no extra processing)
-- Limits serialization depth to prevent infinite recursion
-- Truncates large strings and collections to maintain reasonable output sizes
-- No impact on test execution time, only on failure reporting
+### Negation Detection
+Tests using `not_to` or `to_not` include a `negated: true` flag in the details.
 
-Default limits:
-- Max serialization depth: 5 levels
-- Max array size: 100 items
-- Max hash size: 100 keys
-- Max string length: 1000 characters
+### Serialization
+Values are serialized using [Oj](https://github.com/ohler55/oj) in object mode, providing:
+- Circular reference handling.
+- Proper Ruby object serialization.
+- Excellent performance.
+- Special handling for Regexp objects (serialized as inspect strings).
 
 ## Contributing