from __future__ import annotations

from virtualenv.discovery.py_info import PythonInfo
from virtualenv.discovery.py_spec import PythonSpec

from .pep514 import discover_pythons

# Map of well-known organizations (as per PEP 514 Company Windows Registry key part) versus Python implementation
_IMPLEMENTATION_BY_ORG = {
    "ContinuumAnalytics": "CPython",
    "PythonCore": "CPython",
}


class Pep514PythonInfo(PythonInfo):
    """A Python information acquired from PEP-514."""


def propose_interpreters(spec, cache_dir, env):
    # see if PEP-514 entries are good

    # start with higher python versions in an effort to use the latest version available
    # and prefer PythonCore over conda pythons (as virtualenv is mostly used by non conda tools)
    existing = list(discover_pythons())
    existing.sort(
        key=lambda i: (*tuple(-1 if j is None else j for j in i[1:4]), 1 if i[0] == "PythonCore" else 0),
        reverse=True,
    )

    for name, major, minor, arch, threaded, exe, _ in existing:
        # Map well-known/most common organizations to a Python implementation, use the org name as a fallback for
        # backwards compatibility.
        implementation = _IMPLEMENTATION_BY_ORG.get(name, name)

        # Pre-filtering based on Windows Registry metadata, for CPython only
        skip_pre_filter = implementation.lower() != "cpython"
        registry_spec = PythonSpec(None, implementation, major, minor, None, arch, exe, free_threaded=threaded)
        if skip_pre_filter or registry_spec.satisfies(spec):
            interpreter = Pep514PythonInfo.from_exe(exe, cache_dir, env=env, raise_on_error=False)
            if interpreter is not None and interpreter.satisfies(spec, impl_must_match=True):
                yield interpreter  # Final filtering/matching using interpreter metadata


__all__ = [
    "Pep514PythonInfo",
    "propose_interpreters",
]