fix(provider): normalize DeepSeek non-string message content

This commit is contained in:
hanyuanling
2026-04-27 15:43:41 +08:00
committed by Xubin Ren
parent 311a7fe36e
commit 8e0ce59c0e
2 changed files with 71 additions and 0 deletions
@@ -345,10 +345,25 @@ class OpenAICompatProvider(LLMProvider):
return json.dumps(arguments, ensure_ascii=False)
return "{}"
@staticmethod
def _coerce_content_to_string(content: Any) -> str | None:
"""Coerce block/list content into plain text for strict string-only APIs."""
if content is None or isinstance(content, str):
return content
text = OpenAICompatProvider._extract_text_content(content)
if isinstance(text, str) and text:
return text
try:
dumped = json.dumps(content, ensure_ascii=False)
except Exception:
dumped = str(content)
return dumped or "(empty)"
def _sanitize_messages(self, messages: list[dict[str, Any]]) -> list[dict[str, Any]]:
"""Strip non-standard keys, normalize tool_call IDs."""
sanitized = LLMProvider._sanitize_request_messages(messages, _ALLOWED_MSG_KEYS)
id_map: dict[str, str] = {}
force_string_content = bool(self._spec and self._spec.name == "deepseek")
def map_id(value: Any) -> Any:
if not isinstance(value, str):
@@ -382,6 +397,11 @@ class OpenAICompatProvider(LLMProvider):
clean["content"] = None
if "tool_call_id" in clean and clean["tool_call_id"]:
clean["tool_call_id"] = map_id(clean["tool_call_id"])
if (
force_string_content
and not (clean.get("role") == "assistant" and clean.get("tool_calls"))
):
clean["content"] = self._coerce_content_to_string(clean.get("content"))
return self._enforce_role_alternation(sanitized)
def _drop_deepseek_incomplete_reasoning_history(
+51
View File
@@ -929,6 +929,57 @@ def test_backfill_does_not_touch_messages_when_thinking_off() -> None:
assert "reasoning_content" not in msg
def test_deepseek_coerces_list_content_to_string() -> None:
"""DeepSeek chat endpoint expects message.content to be a string."""
spec = find_by_name("deepseek")
with patch("nanobot.providers.openai_compat_provider.AsyncOpenAI"):
p = OpenAICompatProvider(api_key="k", default_model="deepseek-chat", spec=spec)
kw = p._build_kwargs(
messages=[{
"role": "user",
"content": [
{"type": "text", "text": "hello "},
{"type": "text", "text": "world"},
],
}],
tools=None,
model="deepseek-chat",
max_tokens=1024,
temperature=0.7,
reasoning_effort=None,
tool_choice=None,
)
assert isinstance(kw["messages"][0]["content"], str)
assert "hello" in kw["messages"][0]["content"]
assert "world" in kw["messages"][0]["content"]
def test_non_deepseek_keeps_list_content() -> None:
"""Only DeepSeek should force string content; OpenAI-compatible providers keep blocks."""
spec = find_by_name("openai")
with patch("nanobot.providers.openai_compat_provider.AsyncOpenAI"):
p = OpenAICompatProvider(api_key="k", default_model="gpt-4o", spec=spec)
kw = p._build_kwargs(
messages=[{
"role": "user",
"content": [
{"type": "text", "text": "hello"},
],
}],
tools=None,
model="gpt-4o",
max_tokens=1024,
temperature=0.7,
reasoning_effort=None,
tool_choice=None,
)
assert isinstance(kw["messages"][0]["content"], list)
def test_openai_no_thinking_extra_body() -> None:
"""Non-thinking providers should never get extra_body for thinking."""
kw = _build_kwargs_for("openai", "gpt-4o", reasoning_effort="medium")