Skip to content

CHANGELOG.md

Latest commit

 

History

History
1101 lines (822 loc) · 47.1 KB

CHANGELOG.md

File metadata and controls

1101 lines (822 loc) · 47.1 KB

Changelog

All notable changes to debtmap will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[Unreleased]

[0.16.3] - 2026-04-14

Dependencies

  • Refreshed the Cargo lockfile for the 0.16.3 release
  • stillwater: 1.0.0 -> 1.0.1
  • openssl-src: 300.5.5+3.5.5 -> 300.6.0+3.6.2
  • tokio: 1.51.0 -> 1.52.0
  • openssl-sys: 0.9.112 -> 0.9.113

[0.16.2] - 2026-04-05

Dependencies

  • Refreshed the Cargo lockfile for the 0.16.2 release
  • lcov: 0.8.1 -> 0.8.2
  • tui-scrollview: 0.6.2 -> 0.6.4
  • env_logger: 0.11.9 -> 0.11.10
  • proptest: 1.10.0 -> 1.11.0

[0.16.1] - 2026-03-21

Fixed

  • Python Parameter Extraction - Analyzer no longer panics on typed default parameters
    • Hardened Python extractor parameter-name discovery for tree-sitter node variants
    • Added regression coverage for typed defaults plus *args and **kwargs

[0.16.0] - 2026-03-19

Added

  • AST-Based Python Analysis - Added first-class Python support powered by tree-sitter
    • New Python analyzer and extractor with function discovery and metric extraction
    • Complexity, entropy, and purity analysis for Python codebases
    • Added Python-focused regression coverage for extraction, complexity, and entropy

Fixed

  • Cross-Language Analysis Wiring - Restored and completed analyzer signal propagation

    • JavaScript and TypeScript analysis now fully wire extracted signals into downstream analysis
    • Python analysis now honors analyzer configuration and module-structure signals
    • Rust module-structure analysis now populates dependency graph relationships correctly

  • Discovery Safety - File discovery now ignores Git metadata paths

    • Prevents traversal into .git internals such as worktree metadata
    • Preserves normal source discovery while avoiding repository bookkeeping noise

  • TypeScript Determinism - Extraction and call graph construction now produce stable ordering

    • Functions and call edges are sorted deterministically during analysis
    • Added regression coverage for TypeScript determinism

  • TUI Scoring Progress - TUI now shows debt-scoring finalization progress more clearly

  • Benchmark Option Consistency - Bench analysis options now pass a fixed reference_time

Dependencies

  • Refreshed the Cargo lockfile and updated direct dependency minimums
  • clap: 4.5.60 -> 4.6.0
  • once_cell: 1.21.3 -> 1.21.4
  • assert_cmd: 2.1.2 -> 2.2.0
  • tempfile: 3.26.0 -> 3.27.0
  • openssl-sys: 0.9.111 -> 0.9.112
  • tracing-subscriber: 0.3.22 -> 0.3.23

[0.15.3] - 2026-03-13

Fixed

  • Analysis Determinism - Results are now bit-for-bit identical across multiple runs

    • LCOV record merging now correctly handles duplicate file entries instead of overwriting
    • Parallel call graph conversion now sorts nodes and edges for stable internal iteration order
    • Deterministic function updates in parallel coverage calculation, resolving overlap conflicts
    • Enforced sorted results for all call graph queries (callers, callees, search methods)
    • Stabilized floating-point summation order in risk context to prevent bit-level drift
    • Use of a fixed reference_time for all age-based risk calculations
    • Added secondary sort keys (file, line, function) to the prioritization pipeline
    • Sorted god object detection results and responsibility groups

  • TUI Robustness - Tracing output suppressed while TUI is active to prevent display corruption

  • ROI Test Compilation - Added missing chrono::Utc import to ROI unit tests

Added

  • Determinism Regression Tests - Permanent integration tests for stability
    • Targeted tests for LCOV parsing with adversarial shuffled input
    • Stability tests for call graph traversal and topological sorting
    • Floating-point summation order validation

[0.15.2] - 2026-03-12

Fixed

  • Profiling Report for Rust-Only Codebases - Language-specific timing no longer appears spuriously
    • TypeScript call graph timing now only recorded when TypeScript files exist
    • Fixed unconditional timing span creation in conditional block
    • Profiling reports now accurate for monolingual projects

Added

  • Profiling Regression Tests - Prevent unconditional timing pattern regressions
    • Test validates timing spans inside false conditionals aren't recorded
    • Test ensures language-specific timing doesn't appear for Rust-only projects

Documentation

  • Profiling Audit Report - Comprehensive audit of timing instrumentation
    • All 12 timing spans reviewed and validated
    • Pattern guidelines documented for future development
    • Related to commit 0a58e595

[0.15.1] - 2026-02-18

Fixed

  • Cyclomatic Complexity - else branches no longer incorrectly add to complexity

    • McCabe cyclomatic complexity counts decision points, not branches
    • if-else now correctly reports 2 (not 3)

  • Cognitive Complexity Unified - Consolidated to single spec-compliant implementation

    • Removed duplicate cognitive.rs module (-534 lines)
    • else blocks now correctly add +1 per SonarQube spec
    • Match arms no longer inflate complexity count

  • Nesting Depth for Else Blocks - Plain else blocks now at correct depth

    • Content inside else { if { } } correctly reports depth 2 (was 1)
    • else-if chains still correctly stay flat at depth 1

  • TypeScript Nesting Depth - Functions no longer count as nesting

    • data.map(x => x.filter(y => y)) now correctly reports 0 (was 2)
    • Nesting depth measures control flow only

  • TypeScript Cognitive Complexity - No longer double-counts callbacks

    • Parent function counts callback presence, not its internals
    • Callback complexity counted only when analyzing callback itself

[0.15.0] - 2026-02-16

Added

  • TypeScript/JavaScript Entropy Analysis - Entropy-based complexity dampening for JS/TS
    • Token extraction mapping tree-sitter nodes to weighted categories
    • Pattern detection for validation guards, method chains, switch cases, try-catch
    • Branch similarity analysis for if-else chains, switch cases, ternary expressions
    • Reduces false positives for repetitive validation code and similar constructs
    • Dampening applies when token entropy < 0.5 and pattern repetition detected

Fixed

  • TUI Detail View Footer - Corrected keybinding hints
    • Back navigation now correctly shows q/Esc instead of ←
    • Page navigation shows ←/→ for arrow keys

[0.14.0] - 2026-02-15

Added

  • TypeScript/JavaScript Analyzer - Full-featured analyzer with tree-sitter parsing
    • Cyclomatic and cognitive complexity calculation
    • Function extraction (declarations, arrows, methods, class constructors)
    • Async pattern detection (async/await, Promise chains, callbacks)
    • Functional chain detection (map/filter/reduce pipelines)
    • TypeScript-specific: any usage, type assertions, non-null assertions
    • JSX/TSX support with appropriate grammar selection
    • Call graph integration for cross-function analysis
    • Nesting depth and function length metrics

Fixed

  • Git History Context Fallback - Function-level git history now properly falls back to file-level

    • Previously returned 0 commits/authors when function wasn't found in git history
    • Now correctly falls back to file-level analysis with accurate commit data
    • Affects --context flag output showing git history metrics

  • TypeScript Cognitive Complexity - Fixed double-counting of nesting penalties

    • Nesting increment now happens once per construct, not twice

  • TypeScript Path Lookup - Fixed file path resolution in call graph building

    • Ensures TypeScript files are correctly linked in dependency analysis

  • Default Language Discovery - JavaScript and TypeScript now included in default languages

    • No longer requires explicit --languages javascript,typescript flag

Testing

  • Marked slow CLI integration tests as ignored (run with --ignored)
  • Added git history fallback integration test via ContextAggregator

[0.13.6] - 2026-02-14

Added

  • God Object Ignore Keyword - New god_object keyword for suppression annotations
    • // debtmap:ignore[god_object] - intentionally large coordinator
    • Specifically targets god object debt items

Changed

  • Function-Level Suppression Syntax - Changed from debtmap:allow to debtmap:ignore
    • More consistent with existing debtmap:ignore-start/end block syntax
    • Accepts both single dash (-) and double dash (--) as reason separator
    • Example: // debtmap:ignore[testing] - I/O dispatcher function

Fixed

  • Suppression Doc Comment Support - Recognize /// doc comments for annotations

    • Previously only // regular comments were recognized
    • Now /// debtmap:ignore[testing] - reason works correctly

  • Complexity Keyword Suppression - complexity keyword now suppresses both debt types

    • Matches both Complexity and ComplexityHotspot debt types
    • Added explicit hotspot keyword for hotspot-only suppression

  • Markdown Git Context Output - Aligned with TUI format for consistency

    • Added git context section to priority markdown formatter
    • Shows activity, stability, fix rate, age, contributors, and risk multiplier
    • Terminal formatter now shows commit counts instead of just changes/month

  • Allow Annotations in Coverage Mode - Function-level suppressions now work correctly

    • debtmap:ignore[testing] properly filters items in --lcov coverage mode

  • TUI Item Scores Header - Fixed inconsistent header styling in detail view

  • Context Test File References - Only include existing files in context suggestions

    • Filters out hypothetical *_test.rs paths that don't exist
    • Prevents misleading references when tests are inline modules

Security

  • RUSTSEC-2026-0002 - Updated ratatui 0.28 → 0.30 to fix vulnerability
  • RUSTSEC-2022-0081 - Removed unused cargo-modules dependency
  • Unmaintained bincode - Replaced bincode with postcard for serialization

Dependencies

  • Updated minimum dependency versions for better compatibility
  • cc: Updated to newer version
  • libc: Updated to newer version

Internal (Refactoring)

  • CLI Module Extraction - Extracted analyze and validate-improvement handlers
  • TUI Simplification - Simplified ResultsApp structure with better separation
  • Formatter Refactoring - Extracted tier header and blast radius helpers
  • Suppression System - Consolidated function-level ignore handling
  • Organization Module - Converted to iterator chains for functional style
  • Profiling Extraction - Separated profiling concern from main_inner

Testing

  • Added format_file_rationale tests
  • Added TUI navigation helper tests
  • Added missing dispatcher action tests
  • Added call graph tests for async and same-file call detection

Tech Debt Cleanup

  • Added debtmap:ignore annotations to intentionally complex functions:
    • with_retry - retry orchestration
    • build_adjustment_steps - multi-phase calculation
    • build_page_lines - rendering coordinator
    • execute_list_action - action dispatcher
    • execute_detail_action - detail view dispatcher
    • format_file_priority_item_with_verbosity - formatting coordinator
    • from (crate::core::errors::Error) - error conversion

[0.13.5] - 2026-02-11

Fixed

  • Role Multiplier Inversion - Corrected multipliers for accurate debt prioritization

    • PureLogic: 1.2 → 0.7 (pure functions are easier to test, lower priority)
    • IOWrapper: 0.5 → 1.2 (I/O is harder to test, higher priority)
    • EntryPoint: 1.5 → 1.3 (entry points need integration tests)

  • I/O Pattern Detection - Added missing patterns for function classification

    • Process I/O: spawn, command, process, child, exec
    • Async I/O: timeout, async, await, tokio, future
    • Buffered I/O: bufreader, bufwriter

  • Purity-Based Role Classification - Impure functions no longer default to PureLogic

    • Functions detected as impure now classified as Unknown (neutral multiplier)
    • Prevents I/O orchestration functions from being misclassified

  • Context Test File References - Only include existing files in context suggestions

    • Filters out hypothetical *_test.rs paths that don't exist
    • Prevents misleading references when tests are inline modules

  • TUI Item Scores Header - Fixed inconsistent header styling in detail view

Added

  • Function-Level Debt Suppression - debtmap:ignore annotation support
    • Suppress specific functions from debt reports with inline comments
    • Useful for intentionally complex functions or false positives

Testing

  • Added call graph tests for async and same-file call detection
  • Validates tokio::spawn closure call attribution

[0.13.4] - 2026-02-08

Performance

  • Git Commit Processing - Parallelize commit processing with rayon for faster analysis
  • Git Diff Stats - Use diff.foreach() for O(N) stats calculation instead of O(N²)

Fixed

  • Windows Release Builds - Install OpenSSL via vcpkg for Windows release workflow

Documentation

  • LLMs.txt - Fixed inaccurate language and scoring claims

[0.13.3] - 2026-02-08

Added

  • TUI Content Scrolling - Detail view now supports scrolling through content with arrow keys
    • Enables viewing long debt item details without truncation

Fixed

  • TUI Navigation - Arrow keys now correctly navigate pages in detail view
  • Windows CI Builds - Install OpenSSL via vcpkg for Windows build compatibility
  • Dependency Compatibility - Added minimum time version constraint for CI stability

Changed

  • Dependencies - Updated Cargo.lock with latest dependency versions
    • memchr: 2.7.6 → 2.8.0
    • Updated various dependency version constraints

Internal (Refactoring)

  • God Object Analyzer - Split god object files into focused modules
  • Complexity Analysis - Extracted pure functions from detect_match_expression
  • General Refactoring - Extracted pure functions with improved test coverage

Testing

  • Added comprehensive tests for is_function_used_by_pattern
  • Added coverage for scoring, main CLI, and TUI navigation
  • Added edge case tests for ensure_balanced_distribution
  • Added coverage for TUI actions and navigation
  • Added tests for From<CoreError> implementation
  • Added coverage for git search functionality
  • Ignored slow trybuild UI test by default for faster CI

Infrastructure

  • Added --context flag to debtmap validate commands in CI
  • Applied cargo fmt formatting

[0.13.2] - 2026-02-06

Added

  • LLM Integration Resources
    • Added llms.txt following the emerging standard for LLM-friendly documentation
    • Added debtmap-analyst Claude skill for AI-assisted code analysis
    • Skill includes debtmap usage guidance and Rust/FP refactoring patterns

Changed

  • Dependencies - Updated Cargo.lock with latest dependency versions
    • trybuild: 1.0.114 → 1.0.115

Internal

  • Consolidated typos configuration into _typos.toml with cleaner formatting
  • Added file exclusions for *.lock and target/ directory in typos config

[0.13.1] - 2026-01-30

Fixed

  • Git Context - Fix zero git metrics when analyzing from subdirectories
    • GitHistoryProvider now uses git2's canonical workdir as repo_root
    • Path resolution handles symlinks (e.g., macOS /var → /private/var)
    • Fixes change frequency, bug density, age, and author metrics

Changed

  • Dependencies - Updated Cargo.lock with latest dependency versions
    • clap: 4.5.54 → 4.5.56
    • openssl-src: 300.5.4+3.5.4 → 300.5.5+3.5.5
    • zerocopy: 0.8.34 → 0.8.36

[0.13.0] - 2026-01-24

Added

  • Visual Dashboard - Interactive D3.js dashboard for exploring analysis results

    • Risk Quadrant visualization plotting functions by complexity vs coverage gap
    • Churn-based bubble sizing shows frequently changed code
    • Top Debt Items sortable table
    • Module Flow chord diagram
    • Risk Radar multi-dimensional comparison
    • Hosted at https://iepathos.github.io/debtmap/dashboard/
    • Also available locally via viz-dev/dashboard.html

  • File-Level debt_type - File items now include debt_type field in JSON output

    • Enables consistent filtering across function and file items

  • Smart Tooltip Positioning - Dashboard tooltips now stay within viewport bounds

Changed

  • Removed --format html output - Use the online dashboard with JSON output instead

Fixed

  • Dashboard Display Consistency - Dashboard now matches TUI behavior for debt item display
  • TUI Context Page Alignment - Fixed "context to read" label alignment to start at column 0 like other section headers
  • Test Race Condition - Fixed validation test race condition
  • Dashboard Auto-load - Dashboard no longer tries to auto-fetch on startup

Internal

  • Extracted pure functions from modules for improved testability and maintainability
  • Extracted tests and helpers in priority module
  • Deployed dashboard to GitHub Pages via docs workflow

[0.12.1] - 2026-01-15

Fixed

  • Duplicate Config Struct - Removed duplicate OutputConfig struct definition

Internal (Refactoring)

  • Complexity Reduction - Reduced complexity across multiple modules
    • Extracted pure helpers for complexity calculations
    • Reduced complexity in main_inner and find_cross_file_implementations
    • Fixed git context handling during refactoring
  • Progress System - Extracted progress helpers in unified_analysis module
  • Pure Function Extraction - Continued extracting pure helpers across modules for better testability

[0.12.0] - 2026-01-13

Fixed

  • Zero-Score Filtering - Items with zero score are now properly excluded from results

    • God objects with zero score no longer appear in output
    • All query types now correctly exclude zero-score items
    • Prevents noise from non-actionable debt items

  • Compare Command Input Types - Fixed input parsing to use output format types

    • Compare command now correctly parses JSON input files
    • Improved type consistency between output and input formats

  • Test Caller Filtering for Production Blast Radius (Spec 267)

    • Fixed path-based test attribute detection (#[tokio::test], #[actix_rt::test])
      • syn::Path::is_ident() only matches single identifiers, not paths
      • Now properly checks path segments for async test framework attributes
    • Fixed caller classification path mismatch
      • Caller strings have short paths (overflow.rs:func) but call graph has full paths
      • Added name-based fallback when path-based lookup fails
    • Fixed module-qualified name matching
      • Call graph stores test::func_name but caller strings only have func_name
      • Now matches if function name ends with ::name pattern
    • Fixed god object caller classification
      • File-scope debt items were bypassing caller classification entirely
      • Now properly separates production and test callers for god objects
    • Result: Production blast radius now correctly excludes test callers
      • Example: overflow.rs went from 129 (90+39) to 55 (16+39) after filtering 74 test callers

  • Score Dampening for Well-Tested Stable Cores (Spec 269)

    • Fixed coupling classification threshold from < 0.3 to <= 0.35
      • Instability of 0.3023 displays as "0.30" but failed strict < 0.3 check
      • Now correctly classifies stable modules with borderline instability
    • Applied score dampening for WellTestedCore classification in god objects
      • Well-tested stable cores (instability <= 0.35, 70%+ test callers) now get 80% score reduction
      • These foundations should not appear as priority debt items
    • Result: Well-tested stable files no longer flagged as high-priority debt
      • Example: overflow.rs (74 test / 16 production callers) now scores 10 (Low) instead of 50 (High)

  • Dependency Compatibility

    • Set minimum tokio version to 1.21 for JoinSet support
    • Ensures compatibility with async runtime features

Changed

  • MSRV Bump to 1.89 - Updated minimum supported Rust version to match toolchain
    • Aligns with current stable Rust features
    • Enables use of newer language features

Infrastructure

  • Updated actions/setup-python from v5 to v6
  • Updated github/codeql-action from v3 to v4

[0.11.0] - 2025-12-20

Added

  • LLM-Optimized Output Format (Spec 264)

    • New --llm output format designed for AI agent consumption
    • Structured context suggestions with specific, actionable guidance
    • Optimized for context window efficiency

  • TUI AI Sensor Model Redesign (Spec 265)

    • Redesigned TUI to emphasize AI sensor positioning
    • Clearer presentation of identification, severity, and structural context
    • Improved context suggestion display

  • Context Suggestions for God Objects (Spec 263)

    • God objects now generate targeted context suggestions
    • Provides specific guidance for AI agents analyzing large structs

Changed

  • Documentation Pivot to AI Sensor Model (Spec 266)
    • Updated documentation to reflect debtmap's role as an AI sensor
    • Emphasis on providing accurate identification and severity data to AI agents

Fixed

  • TUI Score Breakdown Transparency
    • Score breakdown now accurately reflects actual calculation
    • Coverage multiplier correctly displayed in calculation steps
    • Entropy dampening shown as multiplier only for god objects

Performance

  • Eliminated Redundant File I/O
    • Removed unnecessary file reads after debt scoring phase completes
    • Improved analysis speed for large codebases

[0.10.0] - 2025-12-19

Strategic Direction

  • AI Sensor Model Pivot (Specs 262-266)
    • Debtmap now positions itself as an "AI sensor" - providing accurate identification, quantified severity, and structural context to AI agents (Claude, Copilot, Cursor) that have semantic understanding
    • Planned removal of template-based recommendations in favor of context window suggestions
    • Key insight: Static analysis can accurately tell you WHERE problems are and HOW SEVERE they are; let AI determine HOW to fix them

Added

  • Score Calculation Transparency (Spec 260, 261)

    • New Score Breakdown detail page (Page 2) in TUI showing exact calculation steps
    • Explicit display of each scoring component: complexity, coverage, dependencies, contextual risk
    • Shows specific adjustments and their contributions to final score
    • Removed score clamping for more accurate representation
    • Removed Score0To100 and Score0To1 wrapper types in favor of direct values

  • Function-Level Git History Analysis (Spec 195)

    • Git blame integration for accurate contributor count per function
    • Function-level churn and authorship metrics
    • Continuous scoring for git history risk contribution

  • God Object Detection Improvements

    • Method Complexity Weighting (Spec 211) - Methods weighted by their individual complexity contribution
    • Accessor and Boilerplate Detection (Spec 209) - Identifies and discounts simple accessor methods
    • Pure Function Method Weighting (Spec 213) - Pure methods contribute less to god object score
    • Functional Decomposition Recognition (Spec 215) - Recognizes intentional functional decomposition patterns
    • Trait-Mandated Method Detection (Spec 217) - Methods required by trait implementations weighted differently
    • Test Code Exclusion (Spec 214) - Test code excluded from LOC metrics
    • Cohesion Gate (Spec 206) - Low cohesion now required before flagging as god object
    • Per-Struct LOC Fix (Spec 207) - Accurate LOC calculation for individual structs

  • Entropy Type Consolidation (Spec 218)

    • Unified entropy analysis types across the codebase
    • Improved entropy data propagation through god object analysis pipeline

  • Observability Improvements

    • Panic Hook with Crash Reports (Spec 207) - Structured crash reports for debugging
    • Structured Tracing with Spans (Spec 208) - Comprehensive tracing throughout analysis
    • Rayon Parallel Span Propagation (Spec 209) - Tracing spans properly propagate through parallel code
    • SourceMap Overflow Prevention (Spec 210) - Prevents crashes on large codebases

  • Output Quality Improvements

    • Output Invariant Testing (Spec 230) - Schema validation for all output formats
    • Debt Item Deduplication (Spec 231) - Eliminates duplicate items in output
    • Anti-Pattern Details (Spec 197) - Exposed in JSON and TUI
    • Cohesion Metrics (Spec 198) - Exposed in output for transparency

Changed

  • Unified Extraction Architecture (Specs 211-214)

    • Single-pass parsing for Rust files via UnifiedFileExtractor
    • Extraction adapters for consistent data transformation
    • Eliminated redundant AST traversals
    • Complete file parsing migration to unified extractor (Spec 204)
    • Consolidated god object detection to extraction adapter (Spec 212)

  • God Object Default Role

    • Changed default role classification from Orchestrator to PureLogic
    • More accurate default assumption for unclassified structs

  • Scoring Improvements

    • Isolated components now contribute zero dependency risk
    • Continuous (non-stepped) scoring for git history risk
    • Removed redundant god object warning from score breakdown
    • Fixed double-counted god_mult in calculation steps

Performance

  • Parallel Debt Item Scoring (Spec 196)

    • Rayon-based parallel scoring for debt items
    • Significant speedup on large codebases

  • File Line Count Caching (Spec 195)

    • Per-file line count caching eliminates redundant file reads

  • Shared Context Detection

    • ContextDetector shared across parallel metric processing
    • Eliminated redundant file I/O in Phase 3 analysis

Internal (Refactoring)

  • God Object Remediation

    • Split ResultsApp into focused modules (navigation, rendering, state)
    • Decomposed unified.rs god module into focused submodules
    • Split analyze.rs command into focused modules
    • Split validate.rs into focused modules
    • Split cfg_builder.rs into focused modules
    • Extracted clustering module from god object analysis
    • Decomposed evidence_calculator into focused modules
    • Split effects.rs into focused modules
    • Split trait_registry.rs into focused modules
    • Split module_structure.rs into focused modules
    • Unified TUI copy with rendering via shared builders

  • Nesting Calculation Consolidation (Specs 201-203)

    • Single source of truth for nesting depth calculation
    • Comprehensive test coverage for nesting scenarios
    • Fixed else-if chain nesting calculation (Spec 198)

  • God Object Extraction Adapter Rewrite (Spec 197)

    • Complete rewrite of extraction adapter for cleaner architecture

Fixed

  • TUI progress updates throughout all analysis stages
  • Correct calculation steps displayed for god objects vs functions
  • Detailed calculation steps shown for god objects in TUI
  • Dampened cyclomatic calculation using correct input

[0.9.2] - 2025-12-13

Added

  • Effects System Expansion (Spec 268)

    • Extended Stillwater effects integration across the codebase
    • Improved effect composition patterns for analysis pipelines

  • Effects-Based Progress System (Spec 262)

    • Replaced callback-based progress reporting with Stillwater effects
    • Cleaner separation of progress concerns from core analysis logic

  • Call Graph Purity Propagation (Spec 261)

    • Purity information now propagates through the call graph
    • Functions calling impure functions are correctly marked as impure
    • More accurate purity analysis across the entire codebase

  • Improved Purity Display Actionability (Spec 260)

    • Enhanced TUI display of purity information with actionable insights
    • Clearer guidance on why functions are marked pure or impure

  • Match Expression CFG Modeling (Spec 253)

    • Control flow graph now correctly models match expressions
    • Improved accuracy of data flow analysis through pattern matching

  • Pattern Destructuring Support for CFG (Spec 252)

    • CFG analysis now handles pattern destructuring correctly
    • Better tracking of variable bindings in complex patterns

Changed

  • Pure Core Extraction for Unified Analysis (Spec 265)

    • Extracted pure business logic from unified_analysis module
    • Improved testability and functional composition

  • Data Flow Module Decomposition (Spec 264)

    • Split monolithic data flow analysis into focused modules
    • Better separation of concerns in data flow analysis

  • Modular CLI Structure (Spec 267)

    • Decomposed main.rs into modular CLI components
    • Cleaner organization of command-line interface code

  • TUI Actions Decomposition (Spec 269)

    • Split TUI action handling into focused modules
    • Improved maintainability of interactive interface code

  • Simplified Mutation Analysis (Spec 257)

    • Reduced mutation analysis to binary signals (mutates/doesn't mutate)
    • Removed over-engineered complexity tracking
    • Clearer, more actionable mutation information

  • Removed CFG Liveness Complexity (Spec 258)

    • Simplified control flow graph analysis
    • Removed unnecessary liveness tracking overhead

  • Removed Dead Store Analysis (Spec 256)

    • Eliminated dead store detection which produced too many false positives
    • Focused analysis on higher-value signals

  • Purity Analysis Relocated

    • Moved purity analysis from patterns page to data flow page in TUI
    • More logical organization of analysis results

Fixed

  • Data Flow Page Visibility

    • Fixed Data Flow page not showing when purity info is available
    • Ensures users can access purity analysis results

  • Constants False Positive in Purity Analysis (Spec 259)

    • Fixed constants being incorrectly flagged in purity analysis
    • More accurate identification of truly impure code

  • False Positive Dead Stores for Pattern Bindings

    • Eliminated incorrect dead store warnings for pattern-bound variables

  • Option/Collection Unwrap Elimination (Spec 266)

    • Replaced unsafe unwrap calls on Options and collections with proper error handling
    • Improved robustness and safety

  • Critical Unwrap Elimination for Lock Safety (Spec 263)

    • Replaced unwrap on mutex locks with proper error handling
    • Prevents potential panics in concurrent code

Internal

  • Removed escape and taint analysis from data flow module (unused complexity)
  • Removed dead cfg_analysis checks from display logic
  • Applied clippy fixes across the codebase
  • Cleaned up implemented spec files

[0.9.1] - 2025-12-12

Added

  • Unified View Data Pipeline (Specs 250-253)

    • New PreparedDebtView unified data model for consistent output across all formats
    • View preparation pipeline separates data transformation from rendering
    • Output format unification ensures TUI, markdown, JSON, and terminal show identical data
    • Refactored format_split_recommendations_markdown for cleaner code structure

  • Accurate LOC Calculation (Spec 201)

    • Lines of code now calculated from actual source spans, not approximations
    • More precise function size metrics for complexity scoring

  • Clean Match Dispatcher Pattern Recognition (Spec 206)

    • Improved detection of clean match dispatcher patterns
    • Reduces false positives for well-structured match statements
    • Better identification of delegation patterns vs actual complexity

  • Data Flow Analysis for Impl Block Methods (Spec 202)

    • Enhanced data flow analysis now supports methods within impl blocks
    • Improved accuracy of mutation tracking in object-oriented Rust code

  • Multi-Debt Type Accumulation (Spec 228)

    • Functions now accumulate multiple independent debt classifications by default
    • Provides comprehensive technical debt assessment per function
    • Example: A function can be flagged as both a complexity hotspot AND have testing gaps
    • Supports three independent debt types: Testing Gaps, Complexity Hotspots, Dead Code
    • Each debt type appears as a separate entry in the output for the same function

Fixed

  • Severity Classification Consistency (Spec 251)

    • Fixed critical bug where severity labels were inconsistent between output formats
    • Unified severity classification to use 70/50/30 thresholds for 0-100 scale
    • TUI, markdown, JSON, and terminal output now show identical severity labels
    • Removed duplicated severity calculation logic from TUI modules
    • Improved clarity: scores 70+ are CRITICAL, 50+ are HIGH, 30+ are MEDIUM, <30 are LOW

  • Function Dependencies in TUI and JSON

    • Fixed function dependencies showing zeros in TUI and JSON output
    • Dependency counts now correctly propagate through all output formats

  • TUI Display Issues

    • Fixed popup areas to prevent background bleed-through
    • Improved clipboard copy to match rendered layout

  • Test Stability

    • Fixed doctest and eliminated test race condition
    • Improved test reliability across the suite

Changed

  • View Pipeline Refactoring
    • Split view_pipeline.rs into focused modules for better maintainability
    • Cleaner separation of concerns in the view layer

[0.8.0] - 2025-12-05

Breaking Changes

  • Legacy Code Cleanup for 1.0 Release (Spec 201)
    • Removed: src/organization/god_object/legacy_compat.rs module
      • This module provided backward compatibility shims that are no longer needed
      • All functionality is available through modern replacements
    • Removed: Deprecated formatter functions
      • format_priority_item_legacy() - Use format_priority_item() instead (uses pure + writer pattern)
      • apply_formatted_sections() - Functionality integrated into format_priority_item()
    • Renamed: Cognitive complexity function for clarity
      • calculate_cognitive_legacy() → calculate_cognitive_visitor_based()
      • Function behavior unchanged, name now accurately describes visitor-based implementation
    • Impact: External library users should migrate to modern equivalents
    • Migration: All modern replacements existed in v0.8.0+, no feature loss

Changed

  • God Object Detector Refactoring (Specs 181a-181i)

    • Completed migration of god object detection to pure functional architecture
    • Extracted pure functions for scoring, classification, and recommendation logic
    • Separated I/O operations (AST visiting) from pure computation
    • Organized into modular structure: types, thresholds, predicates, scoring, classifier, recommender, detector, ast_visitor
    • Removed legacy monolithic implementations (god_object_detector.rs, god_object_analysis.rs)
    • Improved testability and maintainability through functional composition
    • Updated ARCHITECTURE.md with new module structure documentation

  • Formatter Refactoring (Specs 203-207)

    • Separated pure formatting logic from I/O operations (Spec 203)
    • Consolidated pattern display system for consistent output (Spec 204)
    • Broke up monolithic formatter into focused modules (Spec 205)
    • Modularized formatter_markdown.rs for better maintainability (Spec 206)
    • Modularized formatter_verbosity.rs into separate components (Spec 207)
    • Improved code organization and testability of formatting layer

Added

  • Interactive TUI Results Explorer (Spec 211)

    • Specification added for keyboard-driven results exploration interface
    • Progressive disclosure design to handle large result sets (386+ items)
    • Planned features: search, filtering, sorting, and editor integration
    • Addresses scalability issues with flat text output for large codebases

  • Zen Minimalist TUI Progress Visualization (Spec 210)

    • Beautiful ratatui-based TUI for real-time analysis progress
    • 9-stage pipeline visualization with smooth 60 FPS animations
    • Responsive layout adapting to terminal size (minimal/compact/standard/full)
    • Hierarchical progress with active stage expansion and rich statistics
    • Pure functional rendering logic separated from state management
    • Added --no-tui and --quiet flags for flexibility

  • Composable Pipeline Architecture (Spec 209)

    • Type-safe composable pipeline system for analysis stages
    • Stage trait with PureStage and FallibleStage implementations
    • PipelineBuilder with fluent API for pipeline construction
    • Conditional stage support via .when() method
    • Automatic progress reporting and per-stage timing
    • Foundation for replacing monolithic unified_analysis_computation

  • Pure Function Extraction (Spec 208)

    • Extracted pure business logic into focused pipeline stage modules
    • Created modules: call_graph, purity, debt, scoring, filtering, aggregation
    • All functions pure, small (<20 lines), and well-documented
    • Improved testability and composability of analysis logic

  • Stillwater Effects Integration (Spec 207)

    • Comprehensive documentation for Stillwater effects system
    • Effect system architecture added to ARCHITECTURE.md
    • Examples demonstrating effect composition patterns
    • Reader, Retry, and Validation pattern documentation
    • Migration strategy for imperative to effect-based code

  • Batched Git History Optimization (Spec 206)

    • Single comprehensive git log query replaces multiple per-file queries
    • Pure parsing functions for git log output transformation
    • HashMap-based lookups for O(1) access to git history
    • 25x+ performance improvement: 260s → <10s for --context analysis
    • Integrated with GitHistoryProvider as optimized fast path

  • Behavioral Decomposition Refactoring

    • Extracted types module for core data structures
    • Extracted behavioral categorization into separate module
    • Extracted clustering algorithms module
    • Extracted field analysis and recommendations module
    • Improved code organization and maintainability

  • Lock-Free Context Sharing for Parallel Risk Analysis (Spec 204)

    • Implemented lock-free context sharing using Arc-wrapped immutable structures
    • Enables efficient parallel risk analysis without synchronization overhead
    • Improved performance for multi-threaded analysis workflows
    • Thread-safe access to shared context during risk assessment

  • State Machine Arm-Level Analysis (Spec 203)

    • Deep analysis of state machine implementation patterns
    • Detects complex state transitions and arm-level complexity
    • Identifies potential issues in state machine design
    • Enhanced detection of state transition anti-patterns

  • Contextual Risk Integration for Priority Scoring (Spec 202)

    • Integrated contextual risk factors into priority scoring algorithm
    • Risk scores now consider surrounding code context and dependencies
    • More accurate prioritization of technical debt items
    • Added contextual risk display to priority output for better insights

  • Enhanced Progress Feedback (Spec 201)

    • Complete progress feedback for all analysis phases
    • Live updates during file discovery, parsing, and analysis
    • Unified progress display across all operations
    • Clear indication of call graph preparation and multi-pass analysis phases

  • Output Quality Improvements (Spec 201)

    • Filter "no action needed" items from analysis output
    • Only show actionable technical debt items
    • Cleaner, more focused recommendations

  • Unified Progress Flow Display (Spec 195)

    • Consistent progress indicators across all analysis stages
    • Real-time feedback on long-running operations
    • Improved user experience for large codebase analysis

Improved

  • Analysis Architecture (Specs 202, 208)

    • Inverted multi-pass analysis default for better performance (Spec 202)
    • Multi-pass analysis now explicitly enabled when needed
    • Merged dual responsibility classification systems (Spec 208)
    • Extracted severity and coverage classification for clarity (Spec 202)
    • Simplified classification logic and reduced code duplication

  • Data Flow Analysis (Spec 201)

    • Enhanced mutation tracking capabilities
    • Better identification of state transition patterns
    • Improved detection of mutable state usage

Fixed

  • Fixed TUI display corruption and output loss issues
  • Fixed duplicate progress output in TUI statistics
  • Fixed clustering integration tests after behavioral_decomposition refactor
  • Fixed spec 210 implementation gaps for TUI system
  • Fixed spec 209 implementation gaps for pipeline architecture
  • Fixed spec 206 batched git history implementation
  • Fixed parameters with leading underscore to follow Rust conventions
  • Resolved state machine pattern detector test failures
  • Fixed CI pipeline issues in contextual risk tests
  • Resolved test failures in formatter and coverage tests
  • Updated domain pattern tests to use new API
  • Completed implementation gaps in specs 181i, 201, 202, 203, and 204
  • Removed useless test with assert!(true) from concise_recommendation
  • Fixed file-level debt item formatting and display
  • Updated color validation tests after formatter refactoring
  • Fixed god_object_config_rs_test to reference existing files after refactoring

Internal

  • Added spec 211 for Interactive TUI Results Explorer
  • Added specs 207-209 for Stillwater pipeline refactoring
  • Added comprehensive module documentation for behavioral_decomposition
  • Added composable pipeline architecture documentation to ARCHITECTURE.md
  • Created example demonstrating effect composition patterns
  • Applied automated code formatting across the codebase
  • Applied clippy fixes for better code quality
  • Removed completed spec files (181b, 181e, 181h, 206, 210)
  • Removed misplaced example files from src/
  • Updated comprehensive book documentation to fix drift across all chapters
  • Restructured analysis guide into multi-subsection format for better organization
  • Added missing chapters for call graph and boilerplate detection

[0.7.0] - 2025-12-02

Added

  • State Transition Metrics Display (Spec 190)

    • Display state transition metrics in all output formats (JSON, table, detailed text)
    • Show state transition patterns in priority recommendations
    • Added metrics for mutation frequency, transition complexity, and state field counts

  • Enhanced State Field Detection (Spec 202)

    • Pattern-based heuristics for identifying state fields in Rust code
    • Detection of common naming patterns (state, cache, buffer, queue, pool, registry, etc.)
    • Improved accuracy in identifying mutable state and tracking state transitions

Changed - Strategic Direction

  • Rust-Only Focus: Debtmap now exclusively supports Rust code analysis
    • Removed Python, JavaScript, and TypeScript language support
    • Enables deeper Rust language integration and more accurate analysis
    • Allows focus on Rust-specific patterns, idioms, and best practices
    • Significantly reduced dependency footprint and complexity

Removed

  • Python Language Support (Spec 191)

    • Removed Python AST extraction and analysis
    • Removed Python-specific analyzers: asyncio patterns, dead code detection, exception flow, purity analysis
    • Removed Python complexity calculations and pattern detection
    • Removed Python resource tracking: async resources, circular references, context managers, unbounded collections
    • Removed Python testing analyzers: test detection, assertion patterns, flaky test detection, framework detection
    • Removed Python-specific benchmarks and integration tests
    • Removed 17,000+ lines of Python-specific code

  • JavaScript and TypeScript Support (Spec 192)

    • Removed JavaScript complexity analysis
    • Removed JavaScript dead code detection
    • Removed JavaScript resource management tracking
    • Removed JavaScript testing analyzers
    • Removed 3,000+ lines of JavaScript/TypeScript code
    • Removed framework pattern detection for web frameworks

  • Multi-Language Infrastructure

    • Removed cross-language analysis utilities
    • Removed language-agnostic pattern extraction
    • Simplified file walker to focus on .rs files
    • Removed language-specific configuration options

Changed

  • Dependency Updates
    • Updated stillwater from 0.11.0 to 0.13.0 for improved bracket analysis
    • Updated criterion from 0.7.0 to 0.8.0 for enhanced benchmarking capabilities

Improved

  • Architecture Refinement (Spec 181)

    • Implemented "Pure Core, Imperative Shell" pattern for analyzers
    • Separated I/O operations from pure analysis logic
    • Improved testability and composability of analysis functions
    • Better adherence to functional programming principles

  • Code Quality

    • Applied Stillwater evaluation recommendations
    • Improved code organization and modularity
    • Enhanced type safety and error handling
    • Reduced cyclomatic complexity in key modules

Fixed

  • Stillwater API Compatibility

    • Updated bracket API usage for stillwater 0.13.0 compatibility
    • Ensured correct bracket analysis integration with latest stillwater version

  • Tier Classification Accuracy (Commit 2b037ce0)

    • Fixed tier classification to use sophisticated scoring metrics instead of raw cyclomatic complexity
    • Tier classification now respects weighted complexity (30% cyclomatic + 70% cognitive), entropy dampening, nesting depth, and other advanced metrics
    • Functions with high cognitive complexity (e.g., cyclo=9, cognitive=16, nesting=4) now correctly appear as T2 (Complex) instead of being hidden as T4 (Maintenance)
    • Resolves false negatives where high-cognitive-load functions were incorrectly filtered from analysis results
    • Example: reconcile_state() in examples/rethinking-code-quality-analysis now correctly ranks as #1 priority
    • Impact: Tier classification now aligns with debtmap's core philosophy that cognitive load and nesting matter more than raw cyclomatic complexity

Benefits of Rust-Only Approach

  • Accuracy: Native syn-based AST parsing provides 100% Rust language coverage
  • Performance: Optimized for Rust analysis without multi-language overhead
  • Codebase Size: Reduced from ~51,000 lines to ~38,000 lines (-25%)
  • Dependencies: Streamlined dependency tree by removing multi-language parsers
  • Features: Deep integration with Rust-specific constructs:
    • Trait implementations
    • Macro expansion tracking
    • Lifetime and ownership analysis
    • Async/await pattern detection
    • Property-based test detection
  • Maintenance: Simpler codebase focused on a single language ecosystem

Migration Guide

For projects using debtmap on Python, JavaScript, or TypeScript codebases:

  • Use debtmap v0.6.0 or earlier for multi-language support
  • Consider language-specific alternatives:
    • Python: pylint, radon, vulture
    • JavaScript/TypeScript: ESLint, SonarJS, CodeClimate

For Rust projects:

  • No changes required - continue using debtmap as normal
  • Expect improved accuracy and new Rust-specific features in future releases

Documentation

  • Updated ARCHITECTURE.md to reflect Rust-only implementation (Spec 193)
  • Updated analysis guide to focus on Rust metrics
  • Updated getting started guide to remove Python/JS/TS references
  • Updated README with Rust-only focus and capabilities
  • Updated CONTRIBUTING.md with Rust-specific development guidelines
  • All documentation now consistently describes Rust-only capabilities

Internal

  • Added STILLWATER_EVALUATION.md documenting architectural assessment
  • Created specs 181-200 for future improvements:
    • Terminal output UX enhancements (specs 194-200)
    • State transition metrics display (spec 190)
    • Additional architectural refinements (specs 182-189)
  • Added spec 201 for data flow analysis of state transitions and mutation tracking
  • Added spec 202 for enhanced state field detection with pattern-based heuristics
  • Applied automated code formatting across the codebase