File size: 10,294 Bytes
f440f03 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 | """Tests for the shared public chat runtime."""
from __future__ import annotations
import pytest
from maris_core.space_chat import (
DEFAULT_CHAT_MODEL,
SpaceChatRequest,
build_space_chat_messages,
generate_space_chat_reply,
resolve_space_chat_models,
)
from maris_core.utils.hf_integration import HFIntegration
def test_space_chat_request_accepts_external_hf_model_ids() -> None:
request = SpaceChatRequest(message="Sveiki", model="Qwen/Qwen3-Coder-480B-A35B-Instruct")
assert request.model == "Qwen/Qwen3-Coder-480B-A35B-Instruct"
class _StopIterationChoices:
def __bool__(self) -> bool:
return True
def __getitem__(self, index: int) -> object:
raise StopIteration(index)
def __iter__(self):
return iter(())
class _MalformedChatClient:
def __init__(self, token: str | None = None) -> None:
self.token = token
self.generation_models: list[str] = []
def chat_completion(self, **_: object) -> dict[str, object]:
return {"choices": _StopIterationChoices()}
def text_generation(self, **kwargs: object) -> str:
self.generation_models.append(str(kwargs["model"]))
return "Fallback response from text_generation."
class _StopIterationChatClient:
def __init__(self, token: str | None = None) -> None:
self.token = token
self.generation_models: list[str] = []
def chat_completion(self, **_: object) -> dict[str, object]:
return next(iter(()))
def text_generation(self, **kwargs: object) -> str:
self.generation_models.append(str(kwargs["model"]))
return "Fallback response after StopIteration."
@pytest.mark.asyncio
async def test_generate_space_chat_reply_uses_hf_inference_space_runtime_config(
monkeypatch,
) -> None:
captured_kwargs: dict[str, object] = {}
async def fake_save_conversation(
self,
prompt: str,
response: str,
metadata: dict[str, object] | None = None,
) -> None:
del self, prompt, response, metadata
class _ConfiguredChatClient:
def __init__(self, **kwargs: object) -> None:
captured_kwargs.update(kwargs)
def chat_completion(self, **kwargs: object) -> dict[str, object]:
return {
"choices": [
{
"message": {
"content": f"Atbilde no {kwargs['model']}",
}
}
]
}
monkeypatch.setattr(HFIntegration, "save_conversation", fake_save_conversation)
monkeypatch.setenv("HF_INFERENCE_API_KEY", "hf_inference_secret")
response = await generate_space_chat_reply(
SpaceChatRequest(message="Palīdzi ar manu jautājumu"),
client_factory=_ConfiguredChatClient,
)
assert response.model == DEFAULT_CHAT_MODEL
assert captured_kwargs == {
"provider": "hf-inference",
"base_url": "https://api-inference.huggingface.co",
"token": "hf_inference_secret",
}
@pytest.mark.asyncio
async def test_generate_space_chat_reply_falls_back_on_malformed_chat_completion(
monkeypatch,
) -> None:
async def fake_save_conversation(
self,
prompt: str,
response: str,
metadata: dict[str, object] | None = None,
) -> None:
del self, prompt, response, metadata
monkeypatch.setattr(HFIntegration, "save_conversation", fake_save_conversation)
response = await generate_space_chat_reply(
SpaceChatRequest(message="Palīdzi ar manu jautājumu"),
client_factory=_MalformedChatClient,
)
assert response.response == "Fallback response from text_generation."
assert response.model == DEFAULT_CHAT_MODEL
@pytest.mark.asyncio
async def test_generate_space_chat_reply_falls_back_when_chat_completion_raises_stop_iteration(
monkeypatch,
) -> None:
async def fake_save_conversation(
self,
prompt: str,
response: str,
metadata: dict[str, object] | None = None,
) -> None:
del self, prompt, response, metadata
monkeypatch.setattr(HFIntegration, "save_conversation", fake_save_conversation)
response = await generate_space_chat_reply(
SpaceChatRequest(message="Palīdzi ar manu jautājumu"),
client_factory=_StopIterationChatClient,
)
assert response.response == "Fallback response after StopIteration."
assert response.model == DEFAULT_CHAT_MODEL
def test_resolve_space_chat_models_appends_hidden_fallbacks(monkeypatch) -> None:
monkeypatch.setenv("MARIS_CHAT_MODEL", "MarisUK/maris-ai-text")
monkeypatch.setenv(
"MARIS_CHAT_FALLBACK_MODELS",
"MarisUK/maris-assistant-runtime-fallback,Qwen/Qwen3-Coder-480B-A35B-Instruct",
)
models = resolve_space_chat_models("MarisUK/offline-model")
assert models == (
"MarisUK/offline-model",
"MarisUK/maris-ai-text",
"MarisUK/maris-assistant-runtime-fallback",
"Qwen/Qwen3-Coder-480B-A35B-Instruct",
)
def test_build_space_chat_messages_drops_pending_duplicate_user_turn() -> None:
messages = build_space_chat_messages(
SpaceChatRequest(
message="Palīdzi ar plānu",
history=[
{"role": "user", "content": "Iepriekšējais jautājums"},
{"role": "assistant", "content": "Iepriekšējā atbilde"},
{"role": "user", "content": "Palīdzi ar plānu"},
],
)
)
assert messages[-3:] == [
{"role": "user", "content": "Iepriekšējais jautājums"},
{"role": "assistant", "content": "Iepriekšējā atbilde"},
{"role": "user", "content": "Palīdzi ar plānu"},
]
class _RecoveringChatClient:
def __init__(self, token: str | None = None) -> None:
self.token = token
self.chat_models: list[str] = []
def chat_completion(self, **kwargs: object) -> dict[str, object]:
model = str(kwargs["model"])
self.chat_models.append(model)
if model == "MarisUK/offline-model":
raise OSError("primary model unavailable")
return {
"choices": [
{
"message": {
"content": f"Atbilde no {model}",
}
}
]
}
@pytest.mark.asyncio
async def test_generate_space_chat_reply_tries_fallback_models_when_primary_fails(
monkeypatch,
) -> None:
async def fake_save_conversation(
self,
prompt: str,
response: str,
metadata: dict[str, object] | None = None,
) -> None:
del self, prompt, response, metadata
monkeypatch.setattr(HFIntegration, "save_conversation", fake_save_conversation)
monkeypatch.setenv("MARIS_CHAT_MODEL", "MarisUK/offline-model")
monkeypatch.setenv("MARIS_CHAT_MODELS", "")
monkeypatch.setenv("MARIS_CHAT_FALLBACK_MODELS", "Qwen/Qwen3-Coder-480B-A35B-Instruct")
response = await generate_space_chat_reply(
SpaceChatRequest(message="Sveiki"),
client_factory=_RecoveringChatClient,
)
assert response.response == "Atbilde no Qwen/Qwen3-Coder-480B-A35B-Instruct"
assert response.model == "Qwen/Qwen3-Coder-480B-A35B-Instruct"
class _UnavailableChatClient:
def __init__(self, token: str | None = None) -> None:
self.token = token
def chat_completion(self, **kwargs: object) -> dict[str, object]:
del kwargs
raise RuntimeError("provider rejected model")
def text_generation(self, **kwargs: object) -> str:
del kwargs
raise RuntimeError("generation unavailable")
@pytest.mark.asyncio
async def test_generate_space_chat_reply_returns_emergency_fallback_when_all_models_fail(
monkeypatch,
) -> None:
async def fake_save_conversation(
self,
prompt: str,
response: str,
metadata: dict[str, object] | None = None,
) -> None:
del self, prompt, metadata
assert "drošo fallback režīmu" in response
monkeypatch.setattr(HFIntegration, "save_conversation", fake_save_conversation)
response = await generate_space_chat_reply(
SpaceChatRequest(message="Sveiki", model="deepseek-ai/DeepSeek-V3.2"),
client_factory=_UnavailableChatClient,
)
assert "drošo fallback režīmu" in response.response
assert response.model == DEFAULT_CHAT_MODEL
@pytest.mark.asyncio
async def test_generate_space_chat_reply_saves_rich_metadata(monkeypatch) -> None:
captured_metadata: dict[str, object] = {}
async def fake_save_conversation(
self,
prompt: str,
response: str,
metadata: dict[str, object] | None = None,
) -> None:
del self
assert prompt == "Sveiki"
assert response == "Atbilde no MarisUK/maris-ai-text"
captured_metadata.update(metadata or {})
class _HealthyChatClient:
def __init__(self, token: str | None = None) -> None:
self.token = token
def chat_completion(self, **kwargs: object) -> dict[str, object]:
return {
"choices": [
{
"message": {
"content": f"Atbilde no {kwargs['model']}",
}
}
]
}
monkeypatch.setattr(HFIntegration, "save_conversation", fake_save_conversation)
response = await generate_space_chat_reply(
SpaceChatRequest(
message="Sveiki",
history=[{"role": "user", "content": "Sveiki"}],
persona_id="strategist",
session_id="space-session-1",
),
client_factory=_HealthyChatClient,
)
assert response.model == DEFAULT_CHAT_MODEL
assert captured_metadata == {
"session_id": "space-session-1",
"persona_id": "strategist",
"requested_model": DEFAULT_CHAT_MODEL,
"resolved_model": DEFAULT_CHAT_MODEL,
"history_messages": 1,
"detected_emotion": "neutral",
"emotion_confidence": 0.45,
"response_style": "clear_grounded",
"space_repo": "MarisUK/maris.ai.chat",
"public_space_chat": True,
}
|