Implement incremental analysis and intelligent caching

[krystophny]

Summary

Implement incremental analysis to avoid re-analyzing unchanged code sections, significantly improving performance for large codebases and real-time IDE scenarios.

Current State

• Basic AST caching exists but was disabled due to memory safety issues
• test_incremental_analysis.f90 - 12 tests failing
• test_intelligent_caching.f90 - 11 tests failing
• test_cache_integration.f90 - Some basic caching works

Implementation Requirements

Phase 1: Safe AST Caching

• Implement deep copy mechanism for AST nodes
• Create cache invalidation strategy based on file modifications
• Add memory-safe cache storage with proper lifecycle management
• Implement cache hit/miss metrics

Phase 2: Incremental Analysis

• Track file dependencies and module usage
• Implement partial re-analysis for changed sections only
• Create dependency graph for cross-module analysis
• Handle cascading changes efficiently

Phase 3: Intelligent Caching Strategies

• LRU cache eviction for memory management
• Persistent cache for frequently accessed files
• Cache warming on startup
• Smart preloading based on usage patterns

Technical Design

Cache Key Generation

type :: cache_key_t
    character(len=:), allocatable :: file_path
    integer(int64) :: modification_time
    integer(int32) :: content_hash
end type

Incremental Analysis API

type :: incremental_analyzer_t
    type(ast_cache_t) :: cache
    type(dependency_graph_t) :: dependencies
contains
    procedure :: analyze_changes
    procedure :: get_affected_files
    procedure :: invalidate_cache_entry
end type

Performance Targets

• Cache hit rate > 80% for typical editing sessions
• <10ms response time for cached results
• <100ms for incremental re-analysis
• Memory usage < 100MB for typical projects

Integration Points

• LSP document synchronization events
• File watcher notifications
• CLI batch processing mode
• IDE save/format operations

Test Coverage

Expected to fix:

• test_incremental_analysis.f90 (12 tests)
• test_intelligent_caching.f90 (11 tests)
• test_cache_integration.f90 (remaining failures)
• test_diagnostic_performance.f90 (performance tests)

Benefits

• 10x performance improvement for large codebases
• Real-time analysis in IDEs becomes feasible
• Reduced memory footprint with smart eviction
• Better user experience with instant feedback

Dependencies

• Fixed AST deep copying mechanism
• File modification tracking
• Memory management utilities

Priority: HIGH

Critical for production use with large Fortran codebases and IDE integration scenarios where performance is crucial.

[krystophny]

Closing duplicate - intelligent caching implementation completed in Issue #14. All 41 intelligent caching tests now pass (100% success rate) and incremental analysis tests are at 68% completion.