[CI] Reorganize entrypoints tests (#27403)

Signed-off-by: chaunceyjiang <chaunceyjiang@gmail.com>

tests/entrypoints/openai/test_chat.py

@@ -599,145 +599,6 @@ async def test_structured_outputs_choice_chat_logprobs(
         assert item.logprob >= -9999.0, f"Failed (top_logprobs={top_logprobs})"
 
 
-@pytest.mark.asyncio
-async def test_named_tool_use(
-    client: openai.AsyncOpenAI,
-    sample_json_schema,
-):
-    messages = [
-        {"role": "system", "content": "you are a helpful assistant"},
-        {
-            "role": "user",
-            "content": (
-                "Give an example JSON for an employee profile using the specified tool."
-            ),
-        },
-    ]
-    tools = [
-        {
-            "type": "function",
-            "function": {
-                "name": "dummy_function_name",
-                "description": "This is a dummy function",
-                "parameters": sample_json_schema,
-            },
-        }
-    ]
-    tool_choice = {"type": "function", "function": {"name": "dummy_function_name"}}
-
-    # non-streaming
-
-    chat_completion = await client.chat.completions.create(
-        model=MODEL_NAME,
-        messages=messages,
-        max_completion_tokens=1000,
-        tools=tools,
-        tool_choice=tool_choice,
-    )
-    message = chat_completion.choices[0].message
-    assert len(message.content) == 0
-    json_string = message.tool_calls[0].function.arguments
-    json1 = json.loads(json_string)
-    jsonschema.validate(instance=json1, schema=sample_json_schema)
-
-    messages.append({"role": "assistant", "content": json_string})
-    messages.append(
-        {"role": "user", "content": "Give me another one with a different name and age"}
-    )
-
-    # streaming
-
-    stream = await client.chat.completions.create(
-        model=MODEL_NAME,
-        messages=messages,
-        max_completion_tokens=1000,
-        tools=tools,
-        tool_choice=tool_choice,
-        stream=True,
-    )
-
-    output = []
-    finish_reason_count = 0
-    async for chunk in stream:
-        delta = chunk.choices[0].delta
-        if delta.role:
-            assert delta.role == "assistant"
-        assert delta.content is None or len(delta.content) == 0
-        if delta.tool_calls:
-            output.append(delta.tool_calls[0].function.arguments)
-        if chunk.choices[0].finish_reason is not None:
-            finish_reason_count += 1
-    # finish reason should only return in last block
-    assert finish_reason_count == 1
-    json2 = json.loads("".join(output))
-    jsonschema.validate(instance=json2, schema=sample_json_schema)
-    assert json1["name"] != json2["name"]
-    assert json1["age"] != json2["age"]
-
-
-@pytest.mark.asyncio
-async def test_inconsistent_tool_choice_and_tools(
-    client: openai.AsyncOpenAI, sample_json_schema
-):
-    messages = [
-        {"role": "system", "content": "you are a helpful assistant"},
-        {
-            "role": "user",
-            "content": f"Give an example JSON for an employee profile that "
-            f"fits this schema: {sample_json_schema}",
-        },
-    ]
-
-    with pytest.raises(openai.BadRequestError):
-        await client.chat.completions.create(
-            model=MODEL_NAME,
-            messages=messages,
-            max_completion_tokens=1000,
-            tool_choice={
-                "type": "function",
-                "function": {"name": "dummy_function_name"},
-            },
-        )
-
-    with pytest.raises(openai.BadRequestError):
-        await client.chat.completions.create(
-            model=MODEL_NAME,
-            messages=messages,
-            max_completion_tokens=1000,
-            tools=[
-                {
-                    "type": "function",
-                    "function": {
-                        "name": "dummy_function_name",
-                        "description": "This is a dummy function",
-                        "parameters": sample_json_schema,
-                    },
-                }
-            ],
-            tool_choice={
-                "type": "function",
-                "function": {"name": "nondefined_function_name"},
-            },
-        )
-    with pytest.raises(openai.BadRequestError):
-        await client.chat.completions.create(
-            model=MODEL_NAME,
-            messages=messages,
-            max_completion_tokens=1000,
-            tools=[
-                {
-                    "type": "function",
-                    "function": {
-                        "name": "dummy_function_name",
-                        "description": "This is a dummy function",
-                        "parameters": sample_json_schema,
-                    },
-                }
-            ],
-            tool_choice={},
-        )
-
-
 @pytest.mark.asyncio
 async def test_response_format_json_object(client: openai.AsyncOpenAI):
     for _ in range(2):

tests/entrypoints/openai/test_completion_with_function_calling.py

@@ -2,7 +2,9 @@
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
 import datetime
+import json
 
+import jsonschema
 import openai  # use the official client for correctness check
 import pytest
 import pytest_asyncio
@@ -341,3 +343,142 @@ async def test_no_args_tool_call(
     else:
         # No tool called — just print model's direct reply
         assert message.content is not None
+
+
+@pytest.mark.asyncio
+async def test_named_tool_use(
+    client: openai.AsyncOpenAI,
+    sample_json_schema,
+):
+    messages = [
+        {"role": "system", "content": "you are a helpful assistant"},
+        {
+            "role": "user",
+            "content": (
+                "Give an example JSON for an employee profile using the specified tool."
+            ),
+        },
+    ]
+    tools = [
+        {
+            "type": "function",
+            "function": {
+                "name": "dummy_function_name",
+                "description": "This is a dummy function",
+                "parameters": sample_json_schema,
+            },
+        }
+    ]
+    tool_choice = {"type": "function", "function": {"name": "dummy_function_name"}}
+
+    # non-streaming
+
+    chat_completion = await client.chat.completions.create(
+        model=MODEL_NAME,
+        messages=messages,
+        max_completion_tokens=1000,
+        tools=tools,
+        tool_choice=tool_choice,
+    )
+    message = chat_completion.choices[0].message
+    assert len(message.content) == 0
+    json_string = message.tool_calls[0].function.arguments
+    json1 = json.loads(json_string)
+    jsonschema.validate(instance=json1, schema=sample_json_schema)
+
+    messages.append({"role": "assistant", "content": json_string})
+    messages.append(
+        {"role": "user", "content": "Give me another one with a different name and age"}
+    )
+
+    # streaming
+
+    stream = await client.chat.completions.create(
+        model=MODEL_NAME,
+        messages=messages,
+        max_completion_tokens=1000,
+        tools=tools,
+        tool_choice=tool_choice,
+        stream=True,
+    )
+
+    output = []
+    finish_reason_count = 0
+    async for chunk in stream:
+        delta = chunk.choices[0].delta
+        if delta.role:
+            assert delta.role == "assistant"
+        assert delta.content is None or len(delta.content) == 0
+        if delta.tool_calls:
+            output.append(delta.tool_calls[0].function.arguments)
+        if chunk.choices[0].finish_reason is not None:
+            finish_reason_count += 1
+    # finish reason should only return in last block
+    assert finish_reason_count == 1
+    json2 = json.loads("".join(output))
+    jsonschema.validate(instance=json2, schema=sample_json_schema)
+    assert json1["name"] != json2["name"]
+    assert json1["age"] != json2["age"]
+
+
+@pytest.mark.asyncio
+async def test_inconsistent_tool_choice_and_tools(
+    client: openai.AsyncOpenAI, sample_json_schema
+):
+    messages = [
+        {"role": "system", "content": "you are a helpful assistant"},
+        {
+            "role": "user",
+            "content": f"Give an example JSON for an employee profile that "
+            f"fits this schema: {sample_json_schema}",
+        },
+    ]
+
+    with pytest.raises(openai.BadRequestError):
+        await client.chat.completions.create(
+            model=MODEL_NAME,
+            messages=messages,
+            max_completion_tokens=1000,
+            tool_choice={
+                "type": "function",
+                "function": {"name": "dummy_function_name"},
+            },
+        )
+
+    with pytest.raises(openai.BadRequestError):
+        await client.chat.completions.create(
+            model=MODEL_NAME,
+            messages=messages,
+            max_completion_tokens=1000,
+            tools=[
+                {
+                    "type": "function",
+                    "function": {
+                        "name": "dummy_function_name",
+                        "description": "This is a dummy function",
+                        "parameters": sample_json_schema,
+                    },
+                }
+            ],
+            tool_choice={
+                "type": "function",
+                "function": {"name": "nondefined_function_name"},
+            },
+        )
+    with pytest.raises(openai.BadRequestError):
+        await client.chat.completions.create(
+            model=MODEL_NAME,
+            messages=messages,
+            max_completion_tokens=1000,
+            tools=[
+                {
+                    "type": "function",
+                    "function": {
+                        "name": "dummy_function_name",
+                        "description": "This is a dummy function",
+                        "parameters": sample_json_schema,
+                    },
+                }
+            ],
+            tool_choice={},
+        )