Supported Workloads

Excellent Candidates

• Numeric computation (loops, arithmetic, aggregates)
• List/dict manipulation and transforms
• String processing and parsing
• Algorithmic problems (sorting, searching, graph traversal)
• Data validation and transformation pipelines

Good Candidates

• Classes and OOP (methods, inheritance)
• Exception handling with try/except
• Context managers (with statements)
• Pattern matching (match/case)
• Lambdas and higher-order functions

Falls Back to CPython

• Generators (yield, yield from)
• Async/await (async def, await, async for, async with)
• Dynamic code execution (eval, exec)

Tips for Maximum Speedup

• Use supported types: int, float, list, dict, str, tuple.
• Keep types stable across loop iterations.
• Avoid unsupported constructs in hot paths.
• Run pyvorin scan your_script.py to identify risks.