Status

gopy is in active development. The target is behavioural parity with CPython 3.14. The tables below describe the current state as of v0.12.x.

Subsystem readiness

Subsystem	State	Notes
Tokenizer	Complete	Indentation stack, f-strings, soft keywords, t-strings (PEP 750).
PEG parser	Complete	Full grammar including match.
AST	Complete	Generated from Python.asdl.
Symbol table	Complete	Two-pass resolver including __class__ cells, PEP 695.
Codegen	Complete	Every statement and expression form.
Flow graph	Partial	Jump threading, constant folding, dead-code removal, stack depth, exception tables.
Assembler	Complete	PEP 657 location tables, exception tables, const pools.
.pyc marshal	Partial	Wire format implemented; uses gopy magic number (not CPython-interoperable yet).
VM dispatch	Complete	Every Tier-1 opcode wired.
Specializer	Complete	All PEP 659 families.
Tier-2 trace	Partial	Projector and executor lifecycle in place; per-uop bodies being ported.
sys.monitoring	Complete	All PEP 669 events.
Legacy tracing	Complete	sys.settrace bridge.
GIL	Complete	Acquire / release, eval breaker.
Memory model	Complete	Refcount + generational cycle collector.
Generators / async	Complete	RESUME, YIELD_VALUE, RETURN_GENERATOR, SEND.
Import system	Complete	Bootstrap, path finder, inittab. Frozen modules via stdlibinit.
Exception machinery	Complete	PEP 654 groups, except*, PEP 657 tracebacks.
int / float	Complete	Big-int via objects/long.go.
str	Complete	UCS-1 / UCS-2 / UCS-4 kinded storage (PEP 393).
dict	Complete	Combined / split layout, version tag.
Other containers	Complete	list, tuple, set, frozenset, bytes, bytearray.

Modules

53 built-in module slots registered in stdlibinit. See Modules for the full catalogue with per-module state.

Standard-library coverage

Category	Available
Built-ins and core	builtins, sys, _io, io, errno, signal, gc
Numeric	math
Functional	_functools, functools, _operator
Containers	collections, _heapq (partial), _bisect (partial)
Text	re, _sre, _codecs, difflib, fnmatch, _string (partial)
Data	_json, _csv, _hashlib
OS / I/O	os, socket, _socket, _stat (partial)
Concurrency	_thread (partial), contextvars
Compression / encoding	zlib
Date / time	_time (partial), _datetime (partial)
Introspection	traceback, types, warnings, _warnings
Weakrefs	_weakref, weakref
Class machinery	_abc (partial), dataclasses, contextlib
Tooling	argparse, _tokenize, _opcode (partial)

What is not here yet (and commonly missed): asyncio, multiprocessing, ssl, subprocess, threading (the full module), urllib, http, email, pathlib, sqlite3. Open an issue if your use-case needs one of these and we will prioritise.

Version map

gopy	CPython baseline	Headline change
v0.10	3.14	First end-to-end Python program runs.
v0.11	3.14	PEP 659 adaptive specializer.
v0.12	3.14	Tier-2 trace projector and uop executor.
v0.12.1	3.14	CPython test vendor; spec 1700 gate.
v0.12.3	3.14	re, fnmatch ports.
v0.12.4	3.14	Lexer/tokenizer parity sweep.

Full release notes live in the changelog.

Active work

• Tier-2 uop coverage. ~14 of ~285 uops ported. The optimisation pipeline is in place; the bottleneck is the per-uop body translation.
• Tokenizer parity. v0.12.4 is closing the last gaps surfaced by the CPython test_tokenize suite.
• Standard-library port queue. re and fnmatch shipped in v0.12.3; next up are the parts of os and pathlib needed for stdlib tests to compile.
• .pyc interoperability. The marshal format works but uses gopy's own magic number. CPython-interoperable .pyc is on the roadmap.

Non-goals

• No alternative semantics. Anywhere gopy diverges from CPython, the divergence is a bug.
• No CPython C extension ABI. .so and .pyd modules are out of scope. Native modules are reimplemented in Go on demand.
• No Python 2.
• No bytecode stability across gopy versions until v1.0.

How to help

• File issues on the tracker for parity bugs. Minimal repro plus expected behaviour is the most useful shape.
• Pick a module from the partial list and translate one function at a time; see the contributor guide in CONTRIBUTING.md.
• Run the benchmark suite and report regressions.