Skip to content

The fastest, litest AI Gateway. Rust core with Python SDK. Call 100+ LLM APIs in OpenAI (or native) format with cost tracking, guardrails, load balancing, and logging [Bedrock, Azure, OpenAI, Anthropic, OpenAI, VertexAI, vLLM, Nvidia NIM]

14.9 KB
Raw Blame
from __future__ import annotations

import json
import os
import uuid
from typing import Final

import openai
import pytest
from integration._support.client import Gateway, Scenario
from integration._support.responses_vendor import same_response
from integration._support.wire import Reply, Request, Wire, wire_server
from integration._support.worker_readiness import wait_for_worker_readiness
from openai.types.responses import CompactedResponse
from pydantic import JsonValue, TypeAdapter

_MODEL: Final = "gpt-5"
_API_KEY: Final = "synthetic-responses-key"
_JSON_OBJECT: Final = TypeAdapter(dict[str, JsonValue])
_PROXY_WORKERS: Final = max(1, int(os.environ.get("INTEGRATION_PROXY_WORKERS", "1")))
_ALIASES: Final = ("/v1/responses/compact", "/responses/compact", "/openai/v1/responses/compact")
_FUNCTION_TOOL: Final = {
    "type": "function",
    "name": "lookup",
    "description": "Look up a record",
    "parameters": {"type": "object", "properties": {"query": {"type": "string"}}, "required": ["query"]},
    "strict": True,
}


def model_discovery_reply(model: str) -> Reply:
    return Reply(
        body=json.dumps(
            {"object": "list", "data": [{"id": model, "object": "model", "created": 1, "owned_by": "openai"}]}
        ).encode()
    )


def drain_contract_requests(wire: Wire) -> tuple[Request, ...]:
    requests: Final = wire.drain()
    discovery: Final = tuple(request for request in requests if request.target == "/v1/models")
    assert all(request.method == "GET" and request.body == b"" for request in discovery), requests
    return tuple(request for request in requests if request.target != "/v1/models")


def _deployment(scenario: Scenario, wire: Wire) -> str:
    model: Final = scenario.model(
        model=f"openai/{_MODEL}",
        api_key=_API_KEY,
        api_base=f"{wire.url}/v1",
    )
    wait_for_worker_readiness(scenario.gateway, model, 1, _PROXY_WORKERS)
    return model


def test_responses_compact_routes_forward_complete_request_and_typed_items(gateway: Gateway) -> None:
    reasoning_item: Final = {
        "id": f"rs_{uuid.uuid4().hex}",
        "type": "reasoning",
        "summary": [{"type": "summary_text", "text": "Prior reasoning"}],
        "encrypted_content": "synthetic-encrypted-content",
    }
    function_output_item: Final = {
        "type": "function_call_output",
        "call_id": "call_previous",
        "output": "The record is ready.",
    }
    requests: Final = (
        {
            "input": "Compact this conversation.",
        },
        {
            "input": [reasoning_item, function_output_item],
            "instructions": "Preserve the latest decision.",
            "tools": [_FUNCTION_TOOL],
        },
    )
    output_items: Final = [
        {
            "id": f"rs_compact_{uuid.uuid4().hex}",
            "type": "reasoning",
            "summary": [{"type": "summary_text", "text": "Compacted state"}],
            "encrypted_content": "synthetic-compact-state",
        },
        {
            "id": "fc_output_compacted",
            "type": "function_call_output",
            "call_id": "call_previous",
            "output": "The record is ready.",
            "status": "completed",
        },
    ]
    completed_id: Final = f"resp_compact_{uuid.uuid4().hex}"

    def respond(request: Request) -> Reply:
        if request.method == "GET" and request.target == "/v1/models":
            return model_discovery_reply(_MODEL)
        assert request.method == "POST"
        assert request.headers["authorization"] == f"Bearer {_API_KEY}", dict(request.headers)
        body: Final = _JSON_OBJECT.validate_json(request.body)
        compact_params: Final = (
            {
                "model": _MODEL,
                **requests[0],
            }
            if body.get("input") == requests[0]["input"]
            else {
                "model": _MODEL,
                **requests[1],
            }
        )
        assert body == compact_params, request.body
        return Reply(
            body=json.dumps(
                {
                    "id": completed_id,
                    "created_at": 1,
                    "object": "response.compaction",
                    "output": output_items,
                    "usage": {
                        "input_tokens": 7,
                        "input_tokens_details": {"cached_tokens": 0},
                        "output_tokens": 3,
                        "output_tokens_details": {"reasoning_tokens": 0},
                        "total_tokens": 10,
                    },
                }
            ).encode()
        )

    with wire_server(respond) as wire, gateway.scenario() as scenario:
        model: Final = _deployment(scenario, wire)
        for alias in _ALIASES:
            for request_body in requests:
                response: Final = gateway.request(
                    "POST",
                    alias,
                    {"model": model, **request_body},
                )
                assert response.status_code == 200, response.text
                compacted: Final = CompactedResponse.model_validate_json(response.content)
                assert compacted.id != completed_id, compacted.model_dump_json()
                assert same_response(compacted.id, completed_id), compacted.model_dump_json()
                assert compacted.model_dump(mode="json") == {
                    "id": compacted.id,
                    "created_at": 1,
                    "error": None,
                    "incomplete_details": None,
                    "instructions": None,
                    "metadata": None,
                    "model": model,
                    "object": "response.compaction",
                    "output": [
                        {
                            "id": output_items[0]["id"],
                            "summary": [{"text": "Compacted state", "type": "summary_text"}],
                            "type": "reasoning",
                            "content": None,
                            "encrypted_content": "synthetic-compact-state",
                            "status": None,
                        },
                        {
                            "id": "fc_output_compacted",
                            "call_id": "call_previous",
                            "output": "The record is ready.",
                            "status": "completed",
                            "type": "function_call_output",
                            "created_by": None,
                        },
                    ],
                    "parallel_tool_calls": None,
                    "temperature": None,
                    "tool_choice": None,
                    "tools": None,
                    "top_p": None,
                    "max_output_tokens": None,
                    "previous_response_id": None,
                    "reasoning": None,
                    "status": None,
                    "text": None,
                    "truncation": None,
                    "usage": {
                        "input_tokens": 7,
                        "input_tokens_details": {
                            "audio_tokens": None,
                            "cached_tokens": 0,
                            "cached_tokens_details": None,
                            "image_tokens": None,
                            "text_tokens": None,
                            "video_tokens": None,
                        },
                        "output_tokens": 3,
                        "output_tokens_details": {
                            "audio_tokens": None,
                            "reasoning_tokens": 0,
                            "text_tokens": None,
                        },
                        "total_tokens": 10,
                        "cost": None,
                    },
                    "user": None,
                    "store": None,
                }, response.text
        assert [request.target for request in drain_contract_requests(wire)] == ["/v1/responses/compact"] * 6


@pytest.mark.parametrize("prefix", ("/v1", "", "/openai/v1"))
def test_sdk_compact_reaches_the_compact_endpoint_with_its_full_body(gateway: Gateway, prefix: str) -> None:
    completed_id: Final = f"resp_compact_sdk_{uuid.uuid4().hex}"
    input_items: Final = [
        {"role": "user", "content": [{"type": "input_text", "text": "Summarize the patch so far."}]},
        {
            "id": f"rs_{uuid.uuid4().hex}",
            "type": "reasoning",
            "summary": [{"type": "summary_text", "text": "Prior reasoning"}],
            "encrypted_content": "synthetic-encrypted-content",
        },
        {"type": "function_call_output", "call_id": "call_previous", "output": "The record is ready."},
    ]
    output_item: Final = {
        "id": f"rs_compact_{uuid.uuid4().hex}",
        "type": "reasoning",
        "summary": [{"type": "summary_text", "text": "Compacted state"}],
        "encrypted_content": "synthetic-compact-state",
    }

    def respond(request: Request) -> Reply:
        if request.method == "GET" and request.target == "/v1/models":
            return model_discovery_reply(_MODEL)
        assert (request.method, request.target) == ("POST", "/v1/responses/compact"), request.target
        assert request.headers["authorization"] == f"Bearer {_API_KEY}", dict(request.headers)
        assert _JSON_OBJECT.validate_json(request.body) == {
            "model": _MODEL,
            "input": input_items,
            "instructions": "Preserve the latest decision.",
        }, request.body
        return Reply(
            body=json.dumps(
                {
                    "id": completed_id,
                    "created_at": 1,
                    "object": "response.compaction",
                    "output": [output_item],
                    "usage": {
                        "input_tokens": 7,
                        "input_tokens_details": {"cached_tokens": 0},
                        "output_tokens": 3,
                        "output_tokens_details": {"reasoning_tokens": 0},
                        "total_tokens": 10,
                    },
                }
            ).encode()
        )

    with wire_server(respond) as wire, gateway.scenario() as scenario:
        model: Final = _deployment(scenario, wire)
        client: Final = openai.OpenAI(base_url=f"{gateway.client.base_url}{prefix}", api_key=gateway.key, max_retries=0)
        compacted: Final = client.responses.compact(
            model=model,
            input=input_items,
            instructions="Preserve the latest decision.",
        )
        assert compacted.id != completed_id and same_response(compacted.id, completed_id), compacted.model_dump_json()
        assert compacted.model_dump(mode="json", exclude_none=True) == {
            "id": compacted.id,
            "created_at": 1,
            "model": model,
            "object": "response.compaction",
            "output": [output_item],
            "usage": {
                "input_tokens": 7,
                "input_tokens_details": {"cached_tokens": 0},
                "output_tokens": 3,
                "output_tokens_details": {"reasoning_tokens": 0},
                "total_tokens": 10,
            },
        }, compacted.model_dump_json()
        assert [(request.method, request.target) for request in drain_contract_requests(wire)] == [
            ("POST", "/v1/responses/compact")
        ]


@pytest.mark.parametrize("prefix", ("/v1", "", "/openai/v1"))
def test_compact_previous_response_id_reaches_the_provider_as_the_raw_id(
    gateway: Gateway,
    prefix: str,
) -> None:
    raw_id: Final = f"resp_compact_prev_{uuid.uuid4().hex}"
    completed_id: Final = f"resp_compact_result_{uuid.uuid4().hex}"
    input_items: Final = [{"role": "user", "content": [{"type": "input_text", "text": "Summarize the patch so far."}]}]
    output_item: Final = {
        "id": f"rs_compact_{uuid.uuid4().hex}",
        "type": "reasoning",
        "summary": [{"type": "summary_text", "text": "Compacted patch"}],
        "encrypted_content": "synthetic-compact-state",
    }

    def respond(request: Request) -> Reply:
        if request.method == "GET" and request.target == "/v1/models":
            return model_discovery_reply(_MODEL)
        if request.method == "POST" and request.target == "/v1/responses":
            return Reply(
                body=json.dumps(
                    {
                        "id": raw_id,
                        "object": "response",
                        "created_at": 1,
                        "status": "completed",
                        "model": _MODEL,
                        "output": [],
                    }
                ).encode()
            )
        return Reply(
            body=json.dumps(
                {
                    "id": completed_id,
                    "created_at": 1,
                    "object": "response.compaction",
                    "output": [output_item],
                    "usage": {
                        "input_tokens": 7,
                        "input_tokens_details": {"cached_tokens": 0},
                        "output_tokens": 3,
                        "output_tokens_details": {"reasoning_tokens": 0},
                        "total_tokens": 10,
                    },
                }
            ).encode()
        )

    with wire_server(respond) as wire, gateway.scenario() as scenario:
        model: Final = _deployment(scenario, wire)
        client: Final = openai.OpenAI(
            base_url=f"{gateway.client.base_url}{prefix}",
            api_key=gateway.key,
            max_retries=0,
        )
        created: Final = client.responses.create(model=model, input="create before compact")
        assert created.id != raw_id, created.model_dump_json()
        compact_response: Final = client.responses.with_raw_response.compact(
            model=model,
            previous_response_id=created.id,
            input=input_items,
        )
        compacted: Final = compact_response.parse()
        assert compacted.object == "response.compaction", compact_response.http_response.text
        assert compacted.created_at == 1, compact_response.http_response.text
        assert same_response(compacted.id, completed_id), compact_response.http_response.text
        assert len(compacted.output) == 1, compact_response.http_response.text
        assert compacted.output[0].type == "reasoning", compact_response.http_response.text
        assert compacted.output[0].summary[0].text == "Compacted patch", compact_response.http_response.text
        requests: Final = drain_contract_requests(wire)
        assert [(request.method, request.target) for request in requests] == [
            ("POST", "/v1/responses"),
            ("POST", "/v1/responses/compact"),
        ], requests
        assert _JSON_OBJECT.validate_json(requests[0].body) == {
            "model": _MODEL,
            "input": "create before compact",
        }, requests[0].body
        assert _JSON_OBJECT.validate_json(requests[1].body) == {
            "model": _MODEL,
            "previous_response_id": raw_id,
            "input": input_items,
        }, compact_response.http_response.text