mirror of
https://github.com/headroomlabs-ai/headroom.git
synced 2026-11-22 13:29:19 +08:00
## Description Wrapped Grok Shell receives xAI model metadata with `context_length`, but Grok's custom-endpoint parser reads `context_window` and falls back to 256K when that alias is absent. Normalize the successful xAI model-list response at Headroom's provider metadata boundary, preserving the upstream value and the `grok-4.6` model id. Closes #3312. ## Type of Change - [x] Bug fix (non-breaking change that fixes an issue) - [ ] New feature (non-breaking change that adds functionality) - [ ] Breaking change (fix or feature that would cause existing functionality to change) - [ ] Documentation update - [ ] Performance improvement - [ ] Code refactoring (no functional changes) ## Changes Made - Translate a positive xAI `context_length` value to Grok's missing `context_window` field for the resolved `api.x.ai` model list. - Preserve model ids, original xAI fields, explicit context aliases, response status, and non-xAI model metadata. - Leave backend selection unchanged, so this fixes the 256K context fallback without rerouting inference from ChatCompletions to Responses. - Remove entity validators when rewriting model-list bodies, preventing cached metadata from being validated against the unmodified upstream body. ## Testing - [x] Unit tests pass (`uv run pytest tests/test_providers/test_grok_model_metadata.py tests/test_provider_model_metadata.py tests/test_provider_proxy_routes.py tests/test_provider_grok.py tests/test_cli/test_wrap_grok.py -q`) - [x] Linting passes (`uv run ruff check headroom/providers/grok/model_metadata.py headroom/providers/model_metadata.py headroom/proxy/handlers/openai.py tests/test_providers/test_grok_model_metadata.py tests/test_provider_model_metadata.py tests/test_provider_proxy_routes.py`) - [x] Type checking passes (`uv run mypy headroom`) - [x] New tests added for new functionality when applicable - [ ] Manual testing performed ### Test Output ```text uv run pytest tests/test_providers/test_grok_model_metadata.py tests/test_provider_model_metadata.py tests/test_provider_proxy_routes.py tests/test_provider_grok.py tests/test_cli/test_wrap_grok.py -q 46 passed, 1 warning in 8.41s uv run ruff check headroom/providers/grok/model_metadata.py headroom/providers/model_metadata.py headroom/proxy/handlers/openai.py tests/test_providers/test_grok_model_metadata.py tests/test_provider_model_metadata.py tests/test_provider_proxy_routes.py All checks passed! uv run ruff format headroom/providers/grok/model_metadata.py headroom/providers/model_metadata.py headroom/proxy/handlers/openai.py tests/test_providers/test_grok_model_metadata.py tests/test_provider_model_metadata.py tests/test_provider_proxy_routes.py --check 6 files already formatted uv run mypy headroom Success: no issues found in 529 source files ``` ## Real Behavior Proof - Environment: Windows, Headroom development build, focused route tests with a local OpenAI-compatible response fixture. - Exact command / steps: Run `tests/test_provider_proxy_routes.py::test_issue_3312_grok_model_metadata`; the fixture sends an xAI-shaped `/v1/models` entry containing `id: grok-4.6` and `context_length: 500000` through the real provider route. - Observed result: The patched route returns `id: grok-4.6`, preserves `context_length: 500000`, and adds `context_window: 500000`. - Not tested: live Grok Shell 1.0.5 session against `api.x.ai` ## Runtime Rollout Safety - Rollout-managed feature(s): None. - Minimum rollout channel: Stable/default. - Stable/default behavior changed: Yes, wrapped xAI model discovery now exposes the upstream context window to Grok. - Kill switch / disable path: Not applicable, revert the change to disable it. - Unsafe override required: No. - Qualification impact: None. - Rollback path: Revert the change; model-list responses return to the previous metadata shape. ## Review Readiness - [x] I have performed a self-review - [x] This PR is ready for human review ## Checklist - [x] My code follows the project's style guidelines - [x] I have performed a self-review of my code - [x] I have commented my code, particularly in hard-to-understand areas - [ ] I have made corresponding changes to the documentation - [ ] My changes generate no new warnings - [x] I have added tests that prove my fix is effective or that my feature works - [x] New and existing unit tests pass locally with my changes - [x] I did **not** edit `CHANGELOG.md`, it is generated by release-please from the Conventional Commit PR title. ## Screenshots (if applicable) Not applicable for this provider metadata change. ## Additional Notes The change is scoped to the xAI model-list schema boundary. There is no runtime rollout switch; the default behavior changes only for wrapped xAI model discovery, and reverting the change restores the previous metadata response. Live Grok Shell 1.0.5 verification against `api.x.ai` was not run; the available local binary is Grok 1.0.0.
This commit is contained in:
@@ -0,0 +1,55 @@
|
||||
"""xAI model metadata normalization for Grok's custom endpoint."""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from collections.abc import Mapping
|
||||
from typing import Any
|
||||
from urllib.parse import urlsplit
|
||||
|
||||
from headroom.providers.grok.runtime import DEFAULT_API_URL
|
||||
|
||||
_XAI_API_HOST = urlsplit(DEFAULT_API_URL).hostname
|
||||
|
||||
|
||||
def is_xai_model_list_target(base_url: str) -> bool:
|
||||
"""Return whether ``base_url`` is the exact xAI model-list target."""
|
||||
parsed = urlsplit(base_url)
|
||||
return (
|
||||
parsed.hostname is not None
|
||||
and _XAI_API_HOST is not None
|
||||
and parsed.hostname.lower() == _XAI_API_HOST.lower()
|
||||
and parsed.path.rstrip("/") in {"", "/v1"}
|
||||
)
|
||||
|
||||
|
||||
def normalize_xai_model_metadata(payload: Any) -> Any:
|
||||
"""Add Grok's context alias to eligible xAI model-list entries."""
|
||||
if not isinstance(payload, dict) or not isinstance(payload.get("data"), list):
|
||||
return None
|
||||
|
||||
normalized_entries: list[Any] = []
|
||||
changed = False
|
||||
for entry in payload["data"]:
|
||||
if not isinstance(entry, Mapping):
|
||||
normalized_entries.append(entry)
|
||||
continue
|
||||
context_length = entry.get("context_length")
|
||||
if (
|
||||
isinstance(context_length, int)
|
||||
and not isinstance(context_length, bool)
|
||||
and context_length > 0
|
||||
and "context_window" not in entry
|
||||
and "contextWindow" not in entry
|
||||
):
|
||||
normalized_entry = dict(entry)
|
||||
normalized_entry["context_window"] = context_length
|
||||
normalized_entries.append(normalized_entry)
|
||||
changed = True
|
||||
else:
|
||||
normalized_entries.append(entry)
|
||||
|
||||
if not changed:
|
||||
return None
|
||||
normalized_payload = dict(payload)
|
||||
normalized_payload["data"] = normalized_entries
|
||||
return normalized_payload
|
||||
@@ -2,13 +2,19 @@
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import json
|
||||
from dataclasses import dataclass
|
||||
from typing import Any, cast
|
||||
from typing import Any, NoReturn, cast
|
||||
|
||||
from fastapi import Request
|
||||
from fastapi.responses import Response
|
||||
|
||||
from headroom.providers.codex.model_metadata import handle_chatgpt_model_metadata
|
||||
from headroom.providers.grok.model_metadata import (
|
||||
is_xai_model_list_target,
|
||||
normalize_xai_model_metadata,
|
||||
)
|
||||
from headroom.proxy.helpers import sanitize_forwarded_response_headers
|
||||
|
||||
|
||||
@dataclass(frozen=True, slots=True)
|
||||
@@ -23,6 +29,10 @@ class ModelMetadataEndpoint:
|
||||
MODEL_METADATA_LIST_ENDPOINT = ModelMetadataEndpoint("/v1/models", "/backend-api/models")
|
||||
|
||||
|
||||
def _reject_non_standard_json_constant(value: str) -> NoReturn:
|
||||
raise ValueError(f"non-standard JSON constant: {value}")
|
||||
|
||||
|
||||
def model_metadata_get_endpoint(model_id: str) -> ModelMetadataEndpoint:
|
||||
"""Return the single-model metadata endpoint for ``model_id``."""
|
||||
return ModelMetadataEndpoint(
|
||||
@@ -49,7 +59,7 @@ async def handle_model_metadata_endpoint(
|
||||
if chatgpt_response is not None:
|
||||
return chatgpt_response
|
||||
|
||||
return cast(
|
||||
response = cast(
|
||||
Response,
|
||||
await proxy.handle_passthrough(
|
||||
request,
|
||||
@@ -58,3 +68,37 @@ async def handle_model_metadata_endpoint(
|
||||
provider_name,
|
||||
),
|
||||
)
|
||||
if (
|
||||
endpoint == MODEL_METADATA_LIST_ENDPOINT
|
||||
and 200 <= response.status_code < 300
|
||||
and is_xai_model_list_target(provider_api_base_url)
|
||||
):
|
||||
normalized_content: bytes | None = None
|
||||
try:
|
||||
payload = json.loads(
|
||||
response.body,
|
||||
parse_constant=_reject_non_standard_json_constant,
|
||||
)
|
||||
normalized_payload = normalize_xai_model_metadata(payload)
|
||||
if normalized_payload is not None:
|
||||
normalized_content = json.dumps(
|
||||
normalized_payload,
|
||||
separators=(",", ":"),
|
||||
ensure_ascii=False,
|
||||
).encode("utf-8")
|
||||
except (TypeError, UnicodeError, ValueError):
|
||||
normalized_payload = None
|
||||
if normalized_payload is not None and normalized_content is not None:
|
||||
headers = sanitize_forwarded_response_headers(
|
||||
response.headers,
|
||||
"etag",
|
||||
"last-modified",
|
||||
"cache-control",
|
||||
)
|
||||
headers["content-type"] = "application/json"
|
||||
return Response(
|
||||
content=normalized_content,
|
||||
status_code=response.status_code,
|
||||
headers=headers,
|
||||
)
|
||||
return response
|
||||
|
||||
@@ -10989,6 +10989,12 @@ class OpenAIHandlerMixin:
|
||||
separators=(",", ":"),
|
||||
ensure_ascii=False,
|
||||
).encode("utf-8")
|
||||
response_headers = _sanitize_forwarded_response_headers(
|
||||
response.headers,
|
||||
"etag",
|
||||
"last-modified",
|
||||
"cache-control",
|
||||
)
|
||||
response_headers["content-type"] = "application/json"
|
||||
|
||||
# Passthrough request: forwarded upstream with no transforms.
|
||||
|
||||
@@ -94,3 +94,253 @@ def test_handle_model_metadata_endpoint_falls_back_to_selected_provider(monkeypa
|
||||
|
||||
assert response.json() == {"provider": "anthropic", "sub_path": "models"}
|
||||
assert calls == [("https://api.anthropic.test", "models", "anthropic")]
|
||||
|
||||
|
||||
def test_grok_dispatch_adapts_xai_model_list(monkeypatch) -> None:
|
||||
async def no_chatgpt_metadata(http_client, request: Request, upstream_path: str) -> None:
|
||||
return None
|
||||
|
||||
monkeypatch.setattr(
|
||||
"headroom.providers.model_metadata.handle_chatgpt_model_metadata",
|
||||
no_chatgpt_metadata,
|
||||
)
|
||||
|
||||
class Proxy:
|
||||
http_client = "h2"
|
||||
|
||||
async def handle_passthrough(self, request, base_url, sub_path="", provider_name=""):
|
||||
return Response(
|
||||
content=b'{"object":"list","data":[{"id":"grok-4.6","context_length":500000}]}',
|
||||
status_code=200,
|
||||
headers={"content-type": "application/json"},
|
||||
)
|
||||
|
||||
app = FastAPI()
|
||||
|
||||
@app.get("/probe")
|
||||
async def probe(request: Request):
|
||||
return await handle_model_metadata_endpoint(
|
||||
Proxy(),
|
||||
request,
|
||||
endpoint=MODEL_METADATA_LIST_ENDPOINT,
|
||||
provider_api_base_url="https://api.x.ai/v1",
|
||||
provider_name="openai",
|
||||
)
|
||||
|
||||
with TestClient(app) as client:
|
||||
response = client.get("/probe")
|
||||
|
||||
assert response.json()["data"] == [
|
||||
{"id": "grok-4.6", "context_length": 500000, "context_window": 500000}
|
||||
]
|
||||
|
||||
|
||||
def test_grok_response_preserves_status_and_safe_headers_without_stale_framing(monkeypatch) -> None:
|
||||
async def no_chatgpt_metadata(http_client, request: Request, upstream_path: str) -> None:
|
||||
return None
|
||||
|
||||
monkeypatch.setattr(
|
||||
"headroom.providers.model_metadata.handle_chatgpt_model_metadata",
|
||||
no_chatgpt_metadata,
|
||||
)
|
||||
|
||||
class Proxy:
|
||||
http_client = "h2"
|
||||
|
||||
async def handle_passthrough(self, request, base_url, sub_path="", provider_name=""):
|
||||
return Response(
|
||||
content=b'{"data":[{"id":"grok","context_length":1}]}',
|
||||
status_code=206,
|
||||
headers={
|
||||
"content-type": "application/json",
|
||||
"x-upstream": "kept",
|
||||
"etag": '"stale"',
|
||||
"last-modified": "Thu, 01 Jan 1970 00:00:00 GMT",
|
||||
"cache-control": "max-age=60",
|
||||
"content-length": "44",
|
||||
"content-encoding": "gzip",
|
||||
"transfer-encoding": "chunked",
|
||||
},
|
||||
)
|
||||
|
||||
app = FastAPI()
|
||||
|
||||
@app.get("/probe")
|
||||
async def probe(request: Request):
|
||||
return await handle_model_metadata_endpoint(
|
||||
Proxy(),
|
||||
request,
|
||||
endpoint=MODEL_METADATA_LIST_ENDPOINT,
|
||||
provider_api_base_url="https://api.x.ai",
|
||||
provider_name="openai",
|
||||
)
|
||||
|
||||
with TestClient(app) as client:
|
||||
response = client.get("/probe")
|
||||
|
||||
assert response.status_code == 206
|
||||
assert response.headers["x-upstream"] == "kept"
|
||||
assert response.headers.get("etag") is None
|
||||
assert response.headers.get("last-modified") is None
|
||||
assert response.headers.get("cache-control") is None
|
||||
assert response.headers.get("content-encoding") is None
|
||||
assert response.headers.get("transfer-encoding") is None
|
||||
assert response.json()["data"][0]["context_window"] == 1
|
||||
|
||||
|
||||
def test_grok_non_success_and_non_json_responses_are_unchanged(monkeypatch) -> None:
|
||||
async def no_chatgpt_metadata(http_client, request: Request, upstream_path: str) -> None:
|
||||
return None
|
||||
|
||||
monkeypatch.setattr(
|
||||
"headroom.providers.model_metadata.handle_chatgpt_model_metadata",
|
||||
no_chatgpt_metadata,
|
||||
)
|
||||
responses = {
|
||||
"/error": Response(
|
||||
content=b'{"data":[{"context_length":500000}]}',
|
||||
status_code=500,
|
||||
headers={"x-upstream": "error"},
|
||||
),
|
||||
"/non-json": Response(
|
||||
content=b"upstream text",
|
||||
status_code=200,
|
||||
headers={"x-upstream": "text"},
|
||||
),
|
||||
"/surrogate": Response(
|
||||
content=b'{"data":[{"id":"grok","context_length":1,"label":"\\ud800"}]}',
|
||||
status_code=200,
|
||||
headers={"x-upstream": "surrogate"},
|
||||
),
|
||||
"/non-finite": Response(
|
||||
content=b'{"data":[{"id":"grok","context_length":1}],"extra":NaN}',
|
||||
status_code=200,
|
||||
headers={"x-upstream": "non-finite"},
|
||||
),
|
||||
}
|
||||
|
||||
class Proxy:
|
||||
http_client = "h2"
|
||||
|
||||
async def handle_passthrough(self, request, base_url, sub_path="", provider_name=""):
|
||||
return responses[request.url.path]
|
||||
|
||||
app = FastAPI()
|
||||
|
||||
@app.get("/{kind}")
|
||||
async def probe(request: Request, kind: str):
|
||||
return await handle_model_metadata_endpoint(
|
||||
Proxy(),
|
||||
request,
|
||||
endpoint=MODEL_METADATA_LIST_ENDPOINT,
|
||||
provider_api_base_url="https://api.x.ai",
|
||||
provider_name="openai",
|
||||
)
|
||||
|
||||
with TestClient(app) as client:
|
||||
error = client.get("/error")
|
||||
non_json = client.get("/non-json")
|
||||
surrogate = client.get("/surrogate")
|
||||
non_finite = client.get("/non-finite")
|
||||
|
||||
assert error.status_code == 500
|
||||
assert error.content == b'{"data":[{"context_length":500000}]}'
|
||||
assert error.headers["x-upstream"] == "error"
|
||||
assert non_json.status_code == 200
|
||||
assert non_json.content == b"upstream text"
|
||||
assert non_json.headers["x-upstream"] == "text"
|
||||
assert surrogate.content == b'{"data":[{"id":"grok","context_length":1,"label":"\\ud800"}]}'
|
||||
assert surrogate.headers["x-upstream"] == "surrogate"
|
||||
assert non_finite.content == b'{"data":[{"id":"grok","context_length":1}],"extra":NaN}'
|
||||
assert non_finite.headers["x-upstream"] == "non-finite"
|
||||
|
||||
|
||||
def test_grok_response_unchanged_preserves_entity_validators(monkeypatch) -> None:
|
||||
async def no_chatgpt_metadata(http_client, request: Request, upstream_path: str) -> None:
|
||||
return None
|
||||
|
||||
monkeypatch.setattr(
|
||||
"headroom.providers.model_metadata.handle_chatgpt_model_metadata",
|
||||
no_chatgpt_metadata,
|
||||
)
|
||||
|
||||
class Proxy:
|
||||
http_client = "h2"
|
||||
|
||||
async def handle_passthrough(self, request, base_url, sub_path="", provider_name=""):
|
||||
return Response(
|
||||
content=b'{"data":[{"id":"grok","context_length":1,"context_window":1}]}',
|
||||
status_code=200,
|
||||
headers={
|
||||
"content-type": "application/json",
|
||||
"etag": '"current"',
|
||||
"last-modified": "Thu, 01 Jan 1970 00:00:00 GMT",
|
||||
"cache-control": "max-age=60",
|
||||
},
|
||||
)
|
||||
|
||||
app = FastAPI()
|
||||
|
||||
@app.get("/probe")
|
||||
async def probe(request: Request):
|
||||
return await handle_model_metadata_endpoint(
|
||||
Proxy(),
|
||||
request,
|
||||
endpoint=MODEL_METADATA_LIST_ENDPOINT,
|
||||
provider_api_base_url="https://api.x.ai",
|
||||
provider_name="openai",
|
||||
)
|
||||
|
||||
with TestClient(app) as client:
|
||||
response = client.get("/probe")
|
||||
|
||||
assert response.headers["etag"] == '"current"'
|
||||
assert response.headers["last-modified"] == "Thu, 01 Jan 1970 00:00:00 GMT"
|
||||
assert response.headers["cache-control"] == "max-age=60"
|
||||
|
||||
|
||||
def test_grok_negative_space_bypasses_non_xai_detail_and_aliases(monkeypatch) -> None:
|
||||
async def no_chatgpt_metadata(http_client, request: Request, upstream_path: str) -> None:
|
||||
return None
|
||||
|
||||
monkeypatch.setattr(
|
||||
"headroom.providers.model_metadata.handle_chatgpt_model_metadata",
|
||||
no_chatgpt_metadata,
|
||||
)
|
||||
|
||||
class Proxy:
|
||||
http_client = "h2"
|
||||
|
||||
async def handle_passthrough(self, request, base_url, sub_path="", provider_name=""):
|
||||
if request.url.path == "/non-xai":
|
||||
content = b'{"data":[{"context_length":500000}]}'
|
||||
elif request.url.path == "/detail":
|
||||
content = b'{"id":"grok-4.6","context_length":500000}'
|
||||
else:
|
||||
content = b'{"data":[{"context_length":500000,"contextWindow":123}]}'
|
||||
return Response(content=content, status_code=200, media_type="application/json")
|
||||
|
||||
app = FastAPI()
|
||||
|
||||
@app.get("/{kind}")
|
||||
async def probe(request: Request, kind: str):
|
||||
return await handle_model_metadata_endpoint(
|
||||
Proxy(),
|
||||
request,
|
||||
endpoint=MODEL_METADATA_LIST_ENDPOINT
|
||||
if kind == "list"
|
||||
else model_metadata_get_endpoint("grok"),
|
||||
provider_api_base_url="https://api.x.ai"
|
||||
if kind != "non-xai"
|
||||
else "https://api.openai.com",
|
||||
provider_name="openai",
|
||||
)
|
||||
|
||||
with TestClient(app) as client:
|
||||
non_xai = client.get("/non-xai")
|
||||
detail = client.get("/detail")
|
||||
alias = client.get("/list")
|
||||
|
||||
assert non_xai.content == b'{"data":[{"context_length":500000}]}'
|
||||
assert detail.content == b'{"id":"grok-4.6","context_length":500000}'
|
||||
assert alias.json()["data"][0]["contextWindow"] == 123
|
||||
|
||||
@@ -1261,6 +1261,11 @@ def test_anthropic_model_metadata_strips_ansi_model_ids() -> None:
|
||||
{"id": "claude-sonnet-4-5[1m]", "object": "model"},
|
||||
],
|
||||
},
|
||||
headers={
|
||||
"etag": '"stale"',
|
||||
"last-modified": "Thu, 01 Jan 1970 00:00:00 GMT",
|
||||
"cache-control": "max-age=60",
|
||||
},
|
||||
)
|
||||
|
||||
async def aclose(self) -> None:
|
||||
@@ -1272,6 +1277,9 @@ def test_anthropic_model_metadata_strips_ansi_model_ids() -> None:
|
||||
response = client.get("/v1/models", headers={"x-api-key": "sk-ant-test"})
|
||||
|
||||
assert response.status_code == 200
|
||||
assert response.headers.get("etag") is None
|
||||
assert response.headers.get("last-modified") is None
|
||||
assert response.headers.get("cache-control") is None
|
||||
assert response.json()["data"] == [
|
||||
{"id": "claude-opus-4-8", "object": "model"},
|
||||
{"id": "claude-sonnet-4-5", "object": "model"},
|
||||
@@ -1309,6 +1317,57 @@ def test_anthropic_model_detail_path_strips_ansi_model_id() -> None:
|
||||
]
|
||||
|
||||
|
||||
def test_issue_3312_grok_model_metadata() -> None:
|
||||
fixture = {
|
||||
"model": "grok-4.6",
|
||||
"context_length": 500000,
|
||||
}
|
||||
|
||||
class FakeAsyncClient:
|
||||
async def request(self, method, url, **kwargs): # type: ignore[no-untyped-def]
|
||||
return httpx.Response(
|
||||
200,
|
||||
json={
|
||||
"object": "list",
|
||||
"data": [
|
||||
{
|
||||
"id": fixture["model"],
|
||||
"object": "model",
|
||||
"context_length": fixture["context_length"],
|
||||
}
|
||||
],
|
||||
},
|
||||
)
|
||||
|
||||
async def aclose(self) -> None:
|
||||
return None
|
||||
|
||||
app = create_app(
|
||||
ProxyConfig(
|
||||
optimize=False,
|
||||
cache_enabled=False,
|
||||
rate_limit_enabled=False,
|
||||
anthropic_api_url="https://api.anthropic.test",
|
||||
openai_api_url="https://api.x.ai",
|
||||
gemini_api_url="https://api.gemini.test",
|
||||
cloudcode_api_url="https://cloudcode.test",
|
||||
vertex_api_url="https://vertex.test",
|
||||
)
|
||||
)
|
||||
with TestClient(app) as client:
|
||||
client.app.state.proxy.http_client = FakeAsyncClient()
|
||||
response = client.get("/v1/models", headers={"authorization": "Bearer sk-xai-test"})
|
||||
|
||||
entry = response.json()["data"][0]
|
||||
assert response.status_code == 200
|
||||
assert entry == {
|
||||
"id": "grok-4.6",
|
||||
"object": "model",
|
||||
"context_length": 500000,
|
||||
"context_window": 500000,
|
||||
}
|
||||
|
||||
|
||||
def test_anthropic_messages_strips_ansi_model_id_before_upstream() -> None:
|
||||
class FakeAsyncClient:
|
||||
def __init__(self) -> None:
|
||||
|
||||
@@ -0,0 +1,64 @@
|
||||
from __future__ import annotations
|
||||
|
||||
import pytest
|
||||
|
||||
from headroom.providers.grok.model_metadata import (
|
||||
is_xai_model_list_target,
|
||||
normalize_xai_model_metadata,
|
||||
)
|
||||
from headroom.providers.grok.runtime import DEFAULT_API_URL
|
||||
|
||||
|
||||
@pytest.mark.parametrize(
|
||||
("url", "expected"),
|
||||
[
|
||||
(DEFAULT_API_URL, True),
|
||||
(f"{DEFAULT_API_URL}/v1/", True),
|
||||
("https://api.x.ai.evil.test", False),
|
||||
("https://api.x.ai/v2", False),
|
||||
("https://api.openai.com", False),
|
||||
],
|
||||
)
|
||||
def test_xai_model_list_target_uses_exact_host_and_optional_v1_path(
|
||||
url: str, expected: bool
|
||||
) -> None:
|
||||
assert is_xai_model_list_target(url) is expected
|
||||
|
||||
|
||||
def test_normalize_xai_model_metadata_copies_only_positive_integer_context_length() -> None:
|
||||
payload = {
|
||||
"object": "list",
|
||||
"data": [
|
||||
{"id": "grok-4.6", "context_length": 500000, "nested": {"keep": True}},
|
||||
{"id": "zero", "context_length": 0},
|
||||
{"id": "negative", "context_length": -1},
|
||||
{"id": "bool", "context_length": True},
|
||||
{"id": "string", "context_length": "500000"},
|
||||
{"id": "snake", "context_length": 10, "context_window": 20},
|
||||
{"id": "camel", "context_length": 10, "contextWindow": 30},
|
||||
"not-an-entry",
|
||||
],
|
||||
}
|
||||
|
||||
normalized = normalize_xai_model_metadata(payload)
|
||||
|
||||
assert normalized["object"] == "list"
|
||||
assert normalized["data"][0] == {
|
||||
"id": "grok-4.6",
|
||||
"context_length": 500000,
|
||||
"context_window": 500000,
|
||||
"nested": {"keep": True},
|
||||
}
|
||||
assert normalized["data"][1:] == payload["data"][1:]
|
||||
assert payload["data"][0] == {
|
||||
"id": "grok-4.6",
|
||||
"context_length": 500000,
|
||||
"nested": {"keep": True},
|
||||
}
|
||||
|
||||
|
||||
def test_normalize_xai_model_metadata_returns_none_when_no_entry_changes() -> None:
|
||||
assert normalize_xai_model_metadata({"data": [{"id": "grok", "context_length": 0}]}) is None
|
||||
assert normalize_xai_model_metadata({"data": [{"id": "grok", "contextWindow": 10}]}) is None
|
||||
assert normalize_xai_model_metadata({"data": []}) is None
|
||||
assert normalize_xai_model_metadata({"data": {}}) is None
|
||||
Reference in New Issue
Block a user