mirror of
https://github.com/NousResearch/hermes-agent.git
synced 2026-11-17 21:32:52 +00:00
Neither kept test reaches a pyvenv.cfg read: the committed-generation path returns before the cron fall-through, and the gateway overlay never parses it. The `version=` arg, base/home dirs and the legacy cfg content were setup nothing consumed. The cron test's selected_venv patch looks inert on head but is what makes it red on base, so it now says so. Both tests still pass on head and fail with base prod files (fake-win32 harness).
65 lines
2.8 KiB
Python
65 lines
2.8 KiB
Python
"""Windows cron scripts overlay the committed generation, never the stale in-tree venv.
|
|
|
|
``cron/scheduler_script.py::_windows_cron_python_invocation`` answers "which interpreter and
|
|
which site-packages?" for a cron script child. On a PM-managed install it read the dependency
|
|
tree through ``pm.environments.selected_venv``, which falls back to ``base_venv`` and answers
|
|
the leftover pre-PM ``<root>/venv`` when no generation is recorded. That tree is built for
|
|
whichever interpreter created it, so overlaying it on the managed store Python loads a cp311
|
|
``pydantic_core`` on 3.14 and every script dies with
|
|
``ModuleNotFoundError: No module named 'pydantic_core._pydantic_core'`` — the cron sibling of
|
|
the gateway crash in #122183/#123650, which no PR covered.
|
|
|
|
Invariant: with a generation committed, that generation is the tree the child gets — never the
|
|
in-tree venv.
|
|
"""
|
|
|
|
import os
|
|
from pathlib import Path
|
|
|
|
import pytest
|
|
|
|
pytestmark = pytest.mark.platforms("windows")
|
|
|
|
|
|
def _write_venv(venv: Path) -> Path:
|
|
"""A fake Windows venv with the ``Scripts/python.exe`` shim a cron child is handed."""
|
|
(venv / "Lib" / "site-packages").mkdir(parents=True, exist_ok=True)
|
|
(venv / "Scripts").mkdir(parents=True, exist_ok=True)
|
|
(venv / "Scripts" / "python.exe").write_text("", encoding="utf-8")
|
|
return venv
|
|
|
|
|
|
def _site_packages_in(env_overlay: dict) -> Path | None:
|
|
for entry in env_overlay.get("PYTHONPATH", "").split(os.pathsep):
|
|
if Path(entry).name == "site-packages":
|
|
return Path(entry)
|
|
return None
|
|
|
|
|
|
def test_managed_install_uses_the_committed_generation_not_the_in_tree_venv(
|
|
tmp_path, monkeypatch,
|
|
):
|
|
"""The committed generation wins; the leftover pre-PM venv never reaches the child."""
|
|
from cron import scheduler_script as sched_script
|
|
|
|
stale = _write_venv(tmp_path / "repo" / "venv")
|
|
committed = _write_venv(tmp_path / "installs" / "gen" / "venv")
|
|
store = tmp_path / "store" / "python.exe"
|
|
store.parent.mkdir(parents=True)
|
|
store.write_text("", encoding="utf-8")
|
|
child = _write_venv(tmp_path / "child")
|
|
|
|
monkeypatch.setattr("hermes_cli._launchers.resolve_store_python", lambda _root: store)
|
|
monkeypatch.setattr("pm.environments.committed_venv", lambda _root: committed)
|
|
# Base read selected_venv; pointing it at the stale venv is what makes this red there.
|
|
monkeypatch.setattr("pm.environments.selected_venv", lambda _root: stale)
|
|
|
|
interpreter, env_overlay = sched_script._windows_cron_python_invocation(
|
|
str(child / "Scripts" / "python.exe")
|
|
)
|
|
|
|
assert interpreter == str(store)
|
|
overlay = _site_packages_in(env_overlay)
|
|
assert overlay is not None, "the committed generation must supply a site-packages entry"
|
|
assert overlay == committed / "Lib" / "site-packages"
|
|
assert not overlay.is_relative_to(stale)
|