mirror of
https://github.com/openai/codex.git
synced 2026-05-15 08:42:34 +00:00
cad4bbdd64
Build deterministic Python SDK integration coverage around the pinned app-server runtime and a local mock Responses server. Port behavioral coverage off direct SDK monkeypatches where the real app-server boundary is more useful. Co-authored-by: Codex <noreply@openai.com>
513 lines
17 KiB
Python
513 lines
17 KiB
Python
from __future__ import annotations
|
|
|
|
import asyncio
|
|
from collections import deque
|
|
from pathlib import Path
|
|
from types import SimpleNamespace
|
|
from typing import Any
|
|
|
|
import pytest
|
|
|
|
import openai_codex.api as public_api_module
|
|
from openai_codex.client import AppServerClient
|
|
from openai_codex.generated.v2_all import (
|
|
AgentMessageDeltaNotification,
|
|
ItemCompletedNotification,
|
|
MessagePhase,
|
|
TurnCompletedNotification,
|
|
TurnStartParams,
|
|
)
|
|
from openai_codex.models import InitializeResponse, Notification
|
|
from openai_codex.api import (
|
|
ApprovalMode,
|
|
AsyncCodex,
|
|
AsyncThread,
|
|
Codex,
|
|
Thread,
|
|
)
|
|
|
|
ROOT = Path(__file__).resolve().parents[1]
|
|
|
|
|
|
def _approval_settings(params: list[Any]) -> list[dict[str, object]]:
|
|
"""Return serialized approval settings from captured Pydantic params."""
|
|
return [
|
|
{
|
|
key: value
|
|
for key, value in param.model_dump(
|
|
by_alias=True,
|
|
exclude_none=True,
|
|
mode="json",
|
|
).items()
|
|
if key in {"approvalPolicy", "approvalsReviewer"}
|
|
}
|
|
for param in params
|
|
]
|
|
|
|
|
|
def _delta_notification(
|
|
*,
|
|
thread_id: str = "thread-1",
|
|
turn_id: str = "turn-1",
|
|
text: str = "delta-text",
|
|
) -> Notification:
|
|
return Notification(
|
|
method="item/agentMessage/delta",
|
|
payload=AgentMessageDeltaNotification.model_validate(
|
|
{
|
|
"delta": text,
|
|
"itemId": "item-1",
|
|
"threadId": thread_id,
|
|
"turnId": turn_id,
|
|
}
|
|
),
|
|
)
|
|
|
|
|
|
def _completed_notification(
|
|
*,
|
|
thread_id: str = "thread-1",
|
|
turn_id: str = "turn-1",
|
|
status: str = "completed",
|
|
error_message: str | None = None,
|
|
) -> Notification:
|
|
turn: dict[str, object] = {
|
|
"id": turn_id,
|
|
"items": [],
|
|
"status": status,
|
|
}
|
|
if error_message is not None:
|
|
turn["error"] = {"message": error_message}
|
|
return Notification(
|
|
method="turn/completed",
|
|
payload=TurnCompletedNotification.model_validate(
|
|
{
|
|
"threadId": thread_id,
|
|
"turn": turn,
|
|
}
|
|
),
|
|
)
|
|
|
|
|
|
def _item_completed_notification(
|
|
*,
|
|
thread_id: str = "thread-1",
|
|
turn_id: str = "turn-1",
|
|
text: str = "final text",
|
|
phase: MessagePhase | None = None,
|
|
) -> Notification:
|
|
"""Build a realistic completed-item notification accepted by generated models."""
|
|
item: dict[str, object] = {
|
|
"id": "item-1",
|
|
"text": text,
|
|
"type": "agentMessage",
|
|
}
|
|
if phase is not None:
|
|
item["phase"] = phase.value
|
|
return Notification(
|
|
method="item/completed",
|
|
payload=ItemCompletedNotification.model_validate(
|
|
{
|
|
# The pinned runtime schema requires completion timestamps.
|
|
"completedAtMs": 1,
|
|
"item": item,
|
|
"threadId": thread_id,
|
|
"turnId": turn_id,
|
|
}
|
|
),
|
|
)
|
|
|
|
|
|
def test_codex_init_failure_closes_client(monkeypatch: pytest.MonkeyPatch) -> None:
|
|
closed: list[bool] = []
|
|
|
|
class FakeClient:
|
|
def __init__(self, config=None) -> None: # noqa: ANN001,ARG002
|
|
self._closed = False
|
|
|
|
def start(self) -> None:
|
|
return None
|
|
|
|
def initialize(self) -> InitializeResponse:
|
|
return InitializeResponse.model_validate({})
|
|
|
|
def close(self) -> None:
|
|
self._closed = True
|
|
closed.append(True)
|
|
|
|
monkeypatch.setattr(public_api_module, "AppServerClient", FakeClient)
|
|
|
|
with pytest.raises(RuntimeError, match="missing required metadata"):
|
|
Codex()
|
|
|
|
assert closed == [True]
|
|
|
|
|
|
def test_async_codex_init_failure_closes_client() -> None:
|
|
async def scenario() -> None:
|
|
codex = AsyncCodex()
|
|
close_calls = 0
|
|
|
|
async def fake_start() -> None:
|
|
return None
|
|
|
|
async def fake_initialize() -> InitializeResponse:
|
|
return InitializeResponse.model_validate({})
|
|
|
|
async def fake_close() -> None:
|
|
nonlocal close_calls
|
|
close_calls += 1
|
|
|
|
codex._client.start = fake_start # type: ignore[method-assign]
|
|
codex._client.initialize = fake_initialize # type: ignore[method-assign]
|
|
codex._client.close = fake_close # type: ignore[method-assign]
|
|
|
|
with pytest.raises(RuntimeError, match="missing required metadata"):
|
|
await codex.models()
|
|
|
|
assert close_calls == 1
|
|
assert codex._initialized is False
|
|
assert codex._init is None
|
|
|
|
asyncio.run(scenario())
|
|
|
|
|
|
def test_async_codex_initializes_only_once_under_concurrency() -> None:
|
|
async def scenario() -> None:
|
|
codex = AsyncCodex()
|
|
start_calls = 0
|
|
initialize_calls = 0
|
|
ready = asyncio.Event()
|
|
|
|
async def fake_start() -> None:
|
|
nonlocal start_calls
|
|
start_calls += 1
|
|
|
|
async def fake_initialize() -> InitializeResponse:
|
|
nonlocal initialize_calls
|
|
initialize_calls += 1
|
|
ready.set()
|
|
await asyncio.sleep(0.02)
|
|
return InitializeResponse.model_validate(
|
|
{
|
|
"userAgent": "codex-cli/1.2.3",
|
|
"serverInfo": {"name": "codex-cli", "version": "1.2.3"},
|
|
}
|
|
)
|
|
|
|
async def fake_model_list(include_hidden: bool = False): # noqa: ANN202,ARG001
|
|
await ready.wait()
|
|
return object()
|
|
|
|
codex._client.start = fake_start # type: ignore[method-assign]
|
|
codex._client.initialize = fake_initialize # type: ignore[method-assign]
|
|
codex._client.model_list = fake_model_list # type: ignore[method-assign]
|
|
|
|
await asyncio.gather(codex.models(), codex.models())
|
|
|
|
assert start_calls == 1
|
|
assert initialize_calls == 1
|
|
|
|
asyncio.run(scenario())
|
|
|
|
|
|
def _approval_mode_turn_params(approval_mode: ApprovalMode) -> TurnStartParams:
|
|
"""Build real generated turn params from one public approval mode."""
|
|
approval_policy, approvals_reviewer = public_api_module._approval_mode_settings(
|
|
approval_mode
|
|
)
|
|
return TurnStartParams(
|
|
thread_id="thread-1",
|
|
input=[],
|
|
approval_policy=approval_policy,
|
|
approvals_reviewer=approvals_reviewer,
|
|
)
|
|
|
|
|
|
def test_approval_modes_serialize_to_expected_start_params() -> None:
|
|
"""ApprovalMode should map to the app-server params sent for new work."""
|
|
assert {
|
|
mode.value: _approval_settings([_approval_mode_turn_params(mode)])[0]
|
|
for mode in ApprovalMode
|
|
} == {
|
|
"deny_all": {"approvalPolicy": "never"},
|
|
"auto_review": {
|
|
"approvalPolicy": "on-request",
|
|
"approvalsReviewer": "auto_review",
|
|
},
|
|
}
|
|
|
|
|
|
def test_unknown_approval_mode_is_rejected() -> None:
|
|
"""Invalid approval modes should fail before params are constructed."""
|
|
with pytest.raises(ValueError, match="deny_all, auto_review"):
|
|
public_api_module._approval_mode_settings("allow_all") # type: ignore[arg-type]
|
|
|
|
|
|
def test_thread_run_uses_last_completed_assistant_message_as_final_response() -> None:
|
|
client = AppServerClient()
|
|
first_item_notification = _item_completed_notification(text="First message")
|
|
second_item_notification = _item_completed_notification(text="Second message")
|
|
notifications: deque[Notification] = deque(
|
|
[
|
|
first_item_notification,
|
|
second_item_notification,
|
|
_completed_notification(),
|
|
]
|
|
)
|
|
client.next_turn_notification = lambda _turn_id: notifications.popleft() # type: ignore[method-assign]
|
|
client.turn_start = lambda thread_id, wire_input, *, params=None: SimpleNamespace( # noqa: ARG005,E731
|
|
turn=SimpleNamespace(id="turn-1")
|
|
)
|
|
|
|
result = Thread(client, "thread-1").run("hello")
|
|
|
|
assert result.final_response == "Second message"
|
|
assert result.items == [
|
|
first_item_notification.payload.item,
|
|
second_item_notification.payload.item,
|
|
]
|
|
|
|
|
|
def test_thread_run_preserves_empty_last_assistant_message() -> None:
|
|
client = AppServerClient()
|
|
first_item_notification = _item_completed_notification(text="First message")
|
|
second_item_notification = _item_completed_notification(text="")
|
|
notifications: deque[Notification] = deque(
|
|
[
|
|
first_item_notification,
|
|
second_item_notification,
|
|
_completed_notification(),
|
|
]
|
|
)
|
|
client.next_turn_notification = lambda _turn_id: notifications.popleft() # type: ignore[method-assign]
|
|
client.turn_start = lambda thread_id, wire_input, *, params=None: SimpleNamespace( # noqa: ARG005,E731
|
|
turn=SimpleNamespace(id="turn-1")
|
|
)
|
|
|
|
result = Thread(client, "thread-1").run("hello")
|
|
|
|
assert result.final_response == ""
|
|
assert result.items == [
|
|
first_item_notification.payload.item,
|
|
second_item_notification.payload.item,
|
|
]
|
|
|
|
|
|
def test_thread_run_prefers_explicit_final_answer_over_later_commentary() -> None:
|
|
client = AppServerClient()
|
|
final_answer_notification = _item_completed_notification(
|
|
text="Final answer",
|
|
phase=MessagePhase.final_answer,
|
|
)
|
|
commentary_notification = _item_completed_notification(
|
|
text="Commentary",
|
|
phase=MessagePhase.commentary,
|
|
)
|
|
notifications: deque[Notification] = deque(
|
|
[
|
|
final_answer_notification,
|
|
commentary_notification,
|
|
_completed_notification(),
|
|
]
|
|
)
|
|
client.next_turn_notification = lambda _turn_id: notifications.popleft() # type: ignore[method-assign]
|
|
client.turn_start = lambda thread_id, wire_input, *, params=None: SimpleNamespace( # noqa: ARG005,E731
|
|
turn=SimpleNamespace(id="turn-1")
|
|
)
|
|
|
|
result = Thread(client, "thread-1").run("hello")
|
|
|
|
assert result.final_response == "Final answer"
|
|
assert result.items == [
|
|
final_answer_notification.payload.item,
|
|
commentary_notification.payload.item,
|
|
]
|
|
|
|
|
|
def test_thread_run_returns_none_when_only_commentary_messages_complete() -> None:
|
|
client = AppServerClient()
|
|
commentary_notification = _item_completed_notification(
|
|
text="Commentary",
|
|
phase=MessagePhase.commentary,
|
|
)
|
|
notifications: deque[Notification] = deque(
|
|
[
|
|
commentary_notification,
|
|
_completed_notification(),
|
|
]
|
|
)
|
|
client.next_turn_notification = lambda _turn_id: notifications.popleft() # type: ignore[method-assign]
|
|
client.turn_start = lambda thread_id, wire_input, *, params=None: SimpleNamespace( # noqa: ARG005,E731
|
|
turn=SimpleNamespace(id="turn-1")
|
|
)
|
|
|
|
result = Thread(client, "thread-1").run("hello")
|
|
|
|
assert result.final_response is None
|
|
assert result.items == [commentary_notification.payload.item]
|
|
|
|
|
|
def test_thread_run_raises_on_failed_turn() -> None:
|
|
client = AppServerClient()
|
|
notifications: deque[Notification] = deque(
|
|
[
|
|
_completed_notification(status="failed", error_message="boom"),
|
|
]
|
|
)
|
|
client.next_turn_notification = lambda _turn_id: notifications.popleft() # type: ignore[method-assign]
|
|
client.turn_start = lambda thread_id, wire_input, *, params=None: SimpleNamespace( # noqa: ARG005,E731
|
|
turn=SimpleNamespace(id="turn-1")
|
|
)
|
|
|
|
with pytest.raises(RuntimeError, match="boom"):
|
|
Thread(client, "thread-1").run("hello")
|
|
|
|
|
|
def test_stream_text_registers_and_consumes_turn_notifications() -> None:
|
|
"""stream_text should register, consume, and unregister one turn queue."""
|
|
client = AppServerClient()
|
|
notifications: deque[Notification] = deque(
|
|
[
|
|
_delta_notification(text="first"),
|
|
_delta_notification(text="second"),
|
|
_completed_notification(),
|
|
]
|
|
)
|
|
calls: list[tuple[str, str]] = []
|
|
client.turn_start = lambda thread_id, input_items, *, params=None: SimpleNamespace( # noqa: ARG005,E731
|
|
turn=SimpleNamespace(id="turn-1")
|
|
)
|
|
|
|
def fake_register(turn_id: str) -> None:
|
|
"""Record registration for the turn created by stream_text."""
|
|
calls.append(("register", turn_id))
|
|
|
|
def fake_next(turn_id: str) -> Notification:
|
|
"""Return the next queued notification for stream_text."""
|
|
calls.append(("next", turn_id))
|
|
return notifications.popleft()
|
|
|
|
def fake_unregister(turn_id: str) -> None:
|
|
"""Record cleanup for the turn created by stream_text."""
|
|
calls.append(("unregister", turn_id))
|
|
|
|
client.register_turn_notifications = fake_register # type: ignore[method-assign]
|
|
client.next_turn_notification = fake_next # type: ignore[method-assign]
|
|
client.unregister_turn_notifications = fake_unregister # type: ignore[method-assign]
|
|
|
|
chunks = list(client.stream_text("thread-1", "hello"))
|
|
|
|
assert ([chunk.delta for chunk in chunks], calls) == (
|
|
["first", "second"],
|
|
[
|
|
("register", "turn-1"),
|
|
("next", "turn-1"),
|
|
("next", "turn-1"),
|
|
("next", "turn-1"),
|
|
("unregister", "turn-1"),
|
|
],
|
|
)
|
|
|
|
|
|
def test_async_thread_run_uses_last_completed_assistant_message_as_final_response() -> (
|
|
None
|
|
):
|
|
"""Async run should use the last final assistant message as the response text."""
|
|
|
|
async def scenario() -> None:
|
|
"""Feed two completed agent messages through the async per-turn stream."""
|
|
codex = AsyncCodex()
|
|
|
|
async def fake_ensure_initialized() -> None:
|
|
"""Avoid starting a real app-server process for this run test."""
|
|
return None
|
|
|
|
first_item_notification = _item_completed_notification(
|
|
text="First async message"
|
|
)
|
|
second_item_notification = _item_completed_notification(
|
|
text="Second async message"
|
|
)
|
|
notifications: deque[Notification] = deque(
|
|
[
|
|
first_item_notification,
|
|
second_item_notification,
|
|
_completed_notification(),
|
|
]
|
|
)
|
|
|
|
async def fake_turn_start(thread_id: str, wire_input: object, *, params=None): # noqa: ANN001,ANN202,ARG001
|
|
"""Return a synthetic turn id after AsyncThread.run builds input."""
|
|
return SimpleNamespace(turn=SimpleNamespace(id="turn-1"))
|
|
|
|
async def fake_next_notification(_turn_id: str) -> Notification:
|
|
"""Return the next queued notification for that synthetic turn."""
|
|
return notifications.popleft()
|
|
|
|
codex._ensure_initialized = fake_ensure_initialized # type: ignore[method-assign]
|
|
codex._client.turn_start = fake_turn_start # type: ignore[method-assign]
|
|
codex._client.next_turn_notification = fake_next_notification # type: ignore[method-assign]
|
|
|
|
result = await AsyncThread(codex, "thread-1").run("hello")
|
|
|
|
assert result.final_response == "Second async message"
|
|
assert result.items == [
|
|
first_item_notification.payload.item,
|
|
second_item_notification.payload.item,
|
|
]
|
|
|
|
asyncio.run(scenario())
|
|
|
|
|
|
def test_async_thread_run_returns_none_when_only_commentary_messages_complete() -> None:
|
|
"""Async Thread.run should ignore commentary-only messages for final text."""
|
|
|
|
async def scenario() -> None:
|
|
"""Feed a commentary item and completion through the async turn stream."""
|
|
codex = AsyncCodex()
|
|
|
|
async def fake_ensure_initialized() -> None:
|
|
"""Avoid starting a real app-server process for this run test."""
|
|
return None
|
|
|
|
commentary_notification = _item_completed_notification(
|
|
text="Commentary",
|
|
phase=MessagePhase.commentary,
|
|
)
|
|
notifications: deque[Notification] = deque(
|
|
[
|
|
commentary_notification,
|
|
_completed_notification(),
|
|
]
|
|
)
|
|
|
|
async def fake_turn_start(thread_id: str, wire_input: object, *, params=None): # noqa: ANN001,ANN202,ARG001
|
|
"""Return a synthetic turn id for commentary-only output."""
|
|
return SimpleNamespace(turn=SimpleNamespace(id="turn-1"))
|
|
|
|
async def fake_next_notification(_turn_id: str) -> Notification:
|
|
"""Return the next queued commentary/completion notification."""
|
|
return notifications.popleft()
|
|
|
|
codex._ensure_initialized = fake_ensure_initialized # type: ignore[method-assign]
|
|
codex._client.turn_start = fake_turn_start # type: ignore[method-assign]
|
|
codex._client.next_turn_notification = fake_next_notification # type: ignore[method-assign]
|
|
|
|
result = await AsyncThread(codex, "thread-1").run("hello")
|
|
|
|
assert result.final_response is None
|
|
assert result.items == [commentary_notification.payload.item]
|
|
|
|
asyncio.run(scenario())
|
|
|
|
|
|
def test_retry_examples_compare_status_with_enum() -> None:
|
|
for path in (
|
|
ROOT / "examples" / "10_error_handling_and_retry" / "sync.py",
|
|
ROOT / "examples" / "10_error_handling_and_retry" / "async.py",
|
|
):
|
|
source = path.read_text()
|
|
assert '== "failed"' not in source
|
|
assert "TurnStatus.failed" in source
|