Optimization API

The Optimization module provides compression and context optimization.

Context Optimization

class toonverter.optimization.ContextOptimizer(budget: int, policy: OptimizationPolicy | None = None, apply_lightweight_prepass: bool = False)

Optimizes data structure to fit within a token budget.

Parameters:

• budget – The target token budget (in tokens).

• policy – Optional OptimizationPolicy to guide degradation decisions.

• apply_lightweight_prepass – If True, runs a lightweight optimization pass (rounding, truncating) even if the data is within budget. Defaults to False.

optimize(data: Any) → Any

Main entry point. Returns a modified copy of data fitting the budget.

Smart Compression

class toonverter.optimization.SmartCompressor(min_length: int = 4, min_occurrences: int = 2, prefix: str = '@')

Optimizes data structure size using frequency-based dictionary compression.

Parameters:

• min_length – Minimum string length to consider for compression.

• min_occurrences – Minimum number of times a string must appear.

• prefix – Symbol prefix (default: “@”).

compress(data: Any) → dict[str, Any]

Compress data by extracting common strings into a symbol table.

decompress(compressed_data: dict[str, Any]) → Any

Decompress data using the embedded symbol table.