Tsuchinoko

Unsupported Features

This document lists Python features NOT currently supported by Tsuchinoko transpiler.

Language Constructs

Statements

• del statement (deleting variables or elements)
• match statement (Python 3.10+ pattern matching)
• type statement (Python 3.12+ type alias syntax)
• global statement (declaring global variables)
• nonlocal statement (nested function variable binding)
• Walrus operator (:= assignment expression)

Async/Await

• async def (coroutine definitions)
• await expressions
• async for (asynchronous iteration)
• async with (asynchronous context managers)

Generators

• yield statement (generator functions)
• yield from (generator delegation)
• Generator expressions ((x for x in items))

[!NOTE] List comprehensions, dict comprehensions, and set comprehensions are supported.

Context Managers

• Custom context managers (__enter__ / __exit__ protocol)

[!NOTE] Basic with open(...) as f: and similar patterns are supported (V1.6.0).

Arguments

[!NOTE] *args (positional var-args) and **kwargs (keyword var-args) are supported (V1.6.0).

Exception Handling

• Custom exception classes (defining your own exception types)

Class Features

Inheritance & OOP

• Multiple inheritance
• Abstract base classes (abc module)
• Metaclasses

Decorators

• Custom decorators (except @staticmethod, @dataclass, @property)
• Class methods (@classmethod)

[!NOTE] Single inheritance, super(), and @property are supported (V1.6.0).

Magic Methods

• __repr__, __str__ (string representation)
• __call__ (callable objects)
• __slots__ (memory optimization)
• __getitem__, __setitem__, __delitem__ (container protocol)
• __iter__, __next__ (iterator protocol)
• __len__, __contains__ (container protocol)
• __hash__, __eq__ (hashing and equality)
• Operator overloading (__add__, __sub__, __mul__, etc.)

Built-in Types

• Complex numbers (complex, j suffix)
• Bytes/Bytearray (b"...", bytearray)
• Frozenset (frozenset())
• Decimal (decimal.Decimal)
• Fraction (fractions.Fraction)
• Memoryview (memoryview)
• Slice objects (slice())
• Ellipsis (...)
• Range as type (range objects used outside for loops)

Built-in Functions (Native Transpilation)

• Reflection: getattr(), setattr(), hasattr(), delattr()
• Type checking: issubclass(), type() (isinstance is partially supported)
• Introspection: dir(), vars(), locals(), globals()
• Object identity: id(), hash()
• Iteration: iter(), next()
• Dynamic execution: exec(), eval(), compile()
• Formatting: format(), repr()

[!NOTE] isinstance() and super() are supported (V1.6.0).

• Memory: memoryview(), bytearray()

[!NOTE] Many of these can be used via Resident Worker.

Operators & Expressions

[!NOTE] Chained comparisons (a < b < c) are supported (V1.6.0).

Standard Library (Native Transpilation)

These modules cannot be transpiled to pure Rust but work via Resident Worker:

• File I/O (open(), file operations)
• Regular expressions (re module)
• Date/Time (datetime module)
• Collections (collections module: deque, Counter, OrderedDict)
• Itertools (itertools module)
• Functools (functools module: partial, reduce)
• Module system (complex multi-file projects with relative imports)
• Pickle (pickle module)
• JSON (json module) - use Rust’s serde_json instead
• OS/Sys (os, sys modules)
• Threading/Multiprocessing (threading, multiprocessing)
• Networking (socket, http, urllib)
• Subprocess (subprocess module)

Resident Worker Support ✅

Libraries that work via IPC (not native Rust):

• numpy, pandas, scipy, opencv (cv2)
• Any library in your Python environment

Partial / Not Supported Syntax in Resident Libraries

Even with Resident Worker:

• Type aliases with external types: NDInt = npt.NDArray[np.int64]
• Advanced operator overloading: df[df["col"] > 5] (Pandas filtering)
• Object-specific methods: Some methods may lose type information

Notes

Features listed here may be added in future versions. For feature requests, please open an issue on the GitHub repository.

This site is open source. Improve this page.