Crate decy_ownership 

Ownership and lifetime inference for C-to-Rust conversion.

CRITICAL COMPONENT: This is the most important module for quality transpilation. Infers Rust ownership patterns and lifetime annotations from C pointer usage.

ML-Enhanced Features (DECY-ML-001, DECY-ML-003)

This crate includes ML-enhanced ownership inference features:

• OwnershipFeatures: 142-dimension feature vector for batch ML processing
• OwnershipDefect: 8-category defect taxonomy (DECY-O-001 through DECY-O-008)
• InferredOwnership: Predicted Rust ownership kinds
• OwnershipPrediction: Ownership with confidence score and fallback

Re-exports

pub use ml_features::default_pattern_library;

pub use ml_features::AllocationKind;

pub use ml_features::ErrorPattern;

pub use ml_features::ErrorSeverity;

pub use ml_features::FeatureExtractor;

pub use ml_features::InferredOwnership;

pub use ml_features::OwnershipDefect;

pub use ml_features::OwnershipErrorKind;

pub use ml_features::OwnershipFeatures;

pub use ml_features::OwnershipFeaturesBuilder;

pub use ml_features::OwnershipPrediction;

pub use ml_features::PatternLibrary;

pub use ml_features::Severity;

pub use ml_features::SuggestedFix;

pub use hybrid_classifier::ClassificationMethod;

pub use hybrid_classifier::HybridClassifier;

pub use hybrid_classifier::HybridMetrics;

pub use hybrid_classifier::HybridResult;

pub use hybrid_classifier::NullModel;

pub use hybrid_classifier::OwnershipModel;

pub use hybrid_classifier::DEFAULT_CONFIDENCE_THRESHOLD;

pub use ab_testing::ABExperiment;

pub use ab_testing::ABTestRunner;

pub use ab_testing::AssignmentStrategy;

pub use ab_testing::TestObservation;

pub use ab_testing::TestVariant;

pub use ab_testing::VariantMetrics;

pub use threshold_tuning::find_optimal_threshold;

pub use threshold_tuning::SelectionCriteria;

pub use threshold_tuning::ThresholdMetrics;

pub use threshold_tuning::ThresholdTuner;

pub use threshold_tuning::TuningResult;

pub use threshold_tuning::ValidationSample;

pub use model_versioning::ModelEntry;

pub use model_versioning::ModelQualityMetrics;

pub use model_versioning::ModelVersion;

pub use model_versioning::ModelVersionManager;

pub use model_versioning::QualityThresholds;

pub use model_versioning::RollbackResult;

pub use active_learning::ActiveLearner;

pub use active_learning::QueueStats;

pub use active_learning::SampleQueue;

pub use active_learning::SelectionStrategy;

pub use active_learning::UncertainSample;

pub use active_learning::UncertaintyCalculator;

pub use error_tracking::ErrorTracker;

pub use error_tracking::FeatureSuspiciousness;

pub use error_tracking::ImprovementSuggestion;

pub use error_tracking::InferenceError;

pub use error_tracking::PatternStats;

pub use error_tracking::SuggestionCategory;

pub use error_tracking::SuggestionPriority;

pub use retraining_pipeline::DataSplit;

pub use retraining_pipeline::ModelTrainer;

pub use retraining_pipeline::NullTrainer;

pub use retraining_pipeline::PipelineExecution;

pub use retraining_pipeline::RetrainingConfig;

pub use retraining_pipeline::RetrainingPipeline;

pub use retraining_pipeline::RetrainingResult;

pub use retraining_pipeline::RetrainingSchedule;

pub use retraining_pipeline::TrainingMetrics;

pub use retraining_pipeline::TrainingSample;

pub use training_data::CollectionResult;

pub use training_data::DataSource;

pub use training_data::DatasetStats;

pub use training_data::LabeledSample;

pub use training_data::SyntheticConfig;

pub use training_data::SyntheticDataGenerator;

pub use training_data::TrainingDataCollector;

pub use training_data::TrainingDataset;

pub use classifier::ClassifierEvaluator;

pub use classifier::ClassifierPrediction;

pub use classifier::ClassifierTrainer;

pub use classifier::EnsembleClassifier;

pub use classifier::EvaluationMetrics;

pub use classifier::OwnershipClassifier;

pub use classifier::RuleBasedClassifier;

pub use classifier::RuleWeights;

pub use classifier::TrainingConfig;

pub use classifier::TrainingResult;

pub use inference::OwnershipInference;

pub use inference::OwnershipKind;

pub use classifier_integration::classify_function_variables;

pub use classifier_integration::classify_with_rules;

pub use classifier_integration::extract_features_for_variable;

Modules

ab_testing

A/B testing framework for ownership inference (DECY-ML-013).

active_learning

Active learning for uncertain sample collection (DECY-ML-016).

array_slice

Array Parameter to Slice Transformation

borrow_gen

Borrow code generation from ownership inference.

classifier

ML classifier infrastructure for ownership inference (DECY-ML-011).

classifier_integration

Classifier integration layer for transpilation pipeline.

dataflow

Dataflow analysis for tracking pointer usage patterns.

error_tracking

CITL-based error tracking for ownership inference (DECY-ML-015).

hybrid_classifier

Hybrid ownership classification combining rules and ML (DECY-ML-012).

inference

Ownership inference from pointer usage patterns.

lifetime

Scope-based lifetime analysis for C-to-Rust conversion.

lifetime_gen

Lifetime annotation generation for function signatures.

ml_features

ML-enhanced ownership inference features.

model_versioning

Model versioning and rollback for ML-enhanced ownership inference (DECY-ML-017).

retraining_pipeline

Weekly model retraining pipeline (DECY-ML-018).

struct_lifetime

Struct field lifetime annotation generation.

threshold_tuning

Confidence threshold tuning for hybrid classification (DECY-ML-014).

training_data

Training data collection and management (DECY-ML-010).