Skip to content
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.

Commit 778531f

Browse files
stainless-app[bot]stainless-app[bot]
committedJun 18, 2025·
fix(tests): fix: tests which call HTTP endpoints directly with the example parameters
1 parent d08d4f4 commit 778531f

File tree

1 file changed

+40
-105
lines changed

1 file changed

+40
-105
lines changed
 

‎tests/test_client.py

Lines changed: 40 additions & 105 deletions
Original file line numberDiff line numberDiff line change
@@ -23,9 +23,7 @@
2323

2424
from groq import Groq, AsyncGroq, APIResponseValidationError
2525
from groq._types import Omit
26-
from groq._utils import maybe_transform
2726
from groq._models import BaseModel, FinalRequestOptions
28-
from groq._constants import RAW_RESPONSE_HEADER
2927
from groq._exceptions import GroqError, APIStatusError, APITimeoutError, APIResponseValidationError
3028
from groq._base_client import (
3129
DEFAULT_TIMEOUT,
@@ -35,7 +33,6 @@
3533
DefaultAsyncHttpxClient,
3634
make_request_options,
3735
)
38-
from groq.types.chat.completion_create_params import CompletionCreateParams
3936

4037
from .utils import update_env
4138

@@ -707,68 +704,37 @@ def test_parse_retry_after_header(self, remaining_retries: int, retry_after: str
707704

708705
@mock.patch("groq._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout)
709706
@pytest.mark.respx(base_url=base_url)
710-
def test_retrying_timeout_errors_doesnt_leak(self, respx_mock: MockRouter) -> None:
707+
def test_retrying_timeout_errors_doesnt_leak(self, respx_mock: MockRouter, client: Groq) -> None:
711708
respx_mock.post("/openai/v1/chat/completions").mock(side_effect=httpx.TimeoutException("Test timeout error"))
712709

713710
with pytest.raises(APITimeoutError):
714-
self.client.post(
715-
"/openai/v1/chat/completions",
716-
body=cast(
717-
object,
718-
maybe_transform(
719-
dict(
720-
messages=[
721-
{
722-
"role": "system",
723-
"content": "You are a helpful assistant.",
724-
},
725-
{
726-
"role": "user",
727-
"content": "Explain the importance of low latency LLMs",
728-
},
729-
],
730-
model="llama3-8b-8192",
731-
),
732-
CompletionCreateParams,
733-
),
734-
),
735-
cast_to=httpx.Response,
736-
options={"headers": {RAW_RESPONSE_HEADER: "stream"}},
737-
)
711+
client.chat.completions.with_streaming_response.create(
712+
messages=[
713+
{
714+
"content": "content",
715+
"role": "system",
716+
}
717+
],
718+
model="meta-llama/llama-4-scout-17b-16e-instruct",
719+
).__enter__()
738720

739721
assert _get_open_connections(self.client) == 0
740722

741723
@mock.patch("groq._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout)
742724
@pytest.mark.respx(base_url=base_url)
743-
def test_retrying_status_errors_doesnt_leak(self, respx_mock: MockRouter) -> None:
725+
def test_retrying_status_errors_doesnt_leak(self, respx_mock: MockRouter, client: Groq) -> None:
744726
respx_mock.post("/openai/v1/chat/completions").mock(return_value=httpx.Response(500))
745727

746728
with pytest.raises(APIStatusError):
747-
self.client.post(
748-
"/openai/v1/chat/completions",
749-
body=cast(
750-
object,
751-
maybe_transform(
752-
dict(
753-
messages=[
754-
{
755-
"role": "system",
756-
"content": "You are a helpful assistant.",
757-
},
758-
{
759-
"role": "user",
760-
"content": "Explain the importance of low latency LLMs",
761-
},
762-
],
763-
model="llama3-8b-8192",
764-
),
765-
CompletionCreateParams,
766-
),
767-
),
768-
cast_to=httpx.Response,
769-
options={"headers": {RAW_RESPONSE_HEADER: "stream"}},
770-
)
771-
729+
client.chat.completions.with_streaming_response.create(
730+
messages=[
731+
{
732+
"content": "content",
733+
"role": "system",
734+
}
735+
],
736+
model="meta-llama/llama-4-scout-17b-16e-instruct",
737+
).__enter__()
772738
assert _get_open_connections(self.client) == 0
773739

774740
@pytest.mark.parametrize("failures_before_success", [0, 2, 4])
@@ -1582,68 +1548,37 @@ async def test_parse_retry_after_header(self, remaining_retries: int, retry_afte
15821548

15831549
@mock.patch("groq._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout)
15841550
@pytest.mark.respx(base_url=base_url)
1585-
async def test_retrying_timeout_errors_doesnt_leak(self, respx_mock: MockRouter) -> None:
1551+
async def test_retrying_timeout_errors_doesnt_leak(self, respx_mock: MockRouter, async_client: AsyncGroq) -> None:
15861552
respx_mock.post("/openai/v1/chat/completions").mock(side_effect=httpx.TimeoutException("Test timeout error"))
15871553

15881554
with pytest.raises(APITimeoutError):
1589-
await self.client.post(
1590-
"/openai/v1/chat/completions",
1591-
body=cast(
1592-
object,
1593-
maybe_transform(
1594-
dict(
1595-
messages=[
1596-
{
1597-
"role": "system",
1598-
"content": "You are a helpful assistant.",
1599-
},
1600-
{
1601-
"role": "user",
1602-
"content": "Explain the importance of low latency LLMs",
1603-
},
1604-
],
1605-
model="llama3-8b-8192",
1606-
),
1607-
CompletionCreateParams,
1608-
),
1609-
),
1610-
cast_to=httpx.Response,
1611-
options={"headers": {RAW_RESPONSE_HEADER: "stream"}},
1612-
)
1555+
await async_client.chat.completions.with_streaming_response.create(
1556+
messages=[
1557+
{
1558+
"content": "content",
1559+
"role": "system",
1560+
}
1561+
],
1562+
model="meta-llama/llama-4-scout-17b-16e-instruct",
1563+
).__aenter__()
16131564

16141565
assert _get_open_connections(self.client) == 0
16151566

16161567
@mock.patch("groq._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout)
16171568
@pytest.mark.respx(base_url=base_url)
1618-
async def test_retrying_status_errors_doesnt_leak(self, respx_mock: MockRouter) -> None:
1569+
async def test_retrying_status_errors_doesnt_leak(self, respx_mock: MockRouter, async_client: AsyncGroq) -> None:
16191570
respx_mock.post("/openai/v1/chat/completions").mock(return_value=httpx.Response(500))
16201571

16211572
with pytest.raises(APIStatusError):
1622-
await self.client.post(
1623-
"/openai/v1/chat/completions",
1624-
body=cast(
1625-
object,
1626-
maybe_transform(
1627-
dict(
1628-
messages=[
1629-
{
1630-
"role": "system",
1631-
"content": "You are a helpful assistant.",
1632-
},
1633-
{
1634-
"role": "user",
1635-
"content": "Explain the importance of low latency LLMs",
1636-
},
1637-
],
1638-
model="llama3-8b-8192",
1639-
),
1640-
CompletionCreateParams,
1641-
),
1642-
),
1643-
cast_to=httpx.Response,
1644-
options={"headers": {RAW_RESPONSE_HEADER: "stream"}},
1645-
)
1646-
1573+
await async_client.chat.completions.with_streaming_response.create(
1574+
messages=[
1575+
{
1576+
"content": "content",
1577+
"role": "system",
1578+
}
1579+
],
1580+
model="meta-llama/llama-4-scout-17b-16e-instruct",
1581+
).__aenter__()
16471582
assert _get_open_connections(self.client) == 0
16481583

16491584
@pytest.mark.parametrize("failures_before_success", [0, 2, 4])

0 commit comments

Comments
 (0)
Please sign in to comment.