Update 2026-05-13 16:43:53

dataclaw-api/app/core/llm_provider.py

@@ -0,0 +1,87 @@
+import os
+from typing import Optional, Dict
+
+from nanobot.providers.azure_openai_provider import AzureOpenAIProvider
+from nanobot.providers.openai_codex_provider import OpenAICodexProvider
+from nanobot.providers.registry import find_by_name
+
+from app.core.patched_openai_compat_provider import PatchedOpenAICompatProvider
+
+
+def normalize_provider_name(provider: Optional[str]) -> Optional[str]:
+    if not provider:
+        return None
+    normalized = provider.strip().lower()
+    alias_map = {
+        "azure": "azure_openai",
+        "local": "vllm",
+    }
+    return alias_map.get(normalized, normalized)
+
+
+def _running_in_docker() -> bool:
+    # Best-effort, cross-platform detection.
+    if os.environ.get("DATACLAW_RUNNING_IN_DOCKER", "").strip().lower() in ("1", "true", "yes", "y"):
+        return True
+    return os.path.exists("/.dockerenv")
+
+
+def _rewrite_localhost_api_base(api_base: Optional[str]) -> Optional[str]:
+    """
+    When running inside Docker, `localhost` points to the container itself.
+    For host-local LLMs (Ollama/vLLM), users often configure `http://localhost:...`,
+    which breaks in containers. We rewrite it to `host.docker.internal`.
+    """
+    if not api_base:
+        return api_base
+    base = api_base.strip()
+    if base.startswith("http://localhost") or base.startswith("https://localhost"):
+        return base.replace("://localhost", "://host.docker.internal", 1)
+    if base.startswith("http://127.0.0.1") or base.startswith("https://127.0.0.1"):
+        return base.replace("://127.0.0.1", "://host.docker.internal", 1)
+    return api_base
+
+
+def build_llm_provider(
+    *,
+    model: str,
+    provider: Optional[str] = None,
+    api_key: Optional[str] = None,
+    api_base: Optional[str] = None,
+    extra_headers: Optional[Dict[str, str]] = None,
+):
+    provider_name = normalize_provider_name(provider)
+    spec = find_by_name(provider_name) if provider_name else None
+    backend = spec.backend if spec else "openai_compat"
+    if _running_in_docker():
+        api_base = _rewrite_localhost_api_base(api_base)
+
+    if backend == "openai_codex" or model.startswith("openai-codex/"):
+        return OpenAICodexProvider(default_model=model)
+
+    if backend == "azure_openai":
+        if not api_key or not api_base:
+            raise ValueError("Azure OpenAI requires api_key and api_base.")
+        return AzureOpenAIProvider(
+            api_key=api_key,
+            api_base=api_base,
+            default_model=model,
+        )
+
+    if backend == "anthropic":
+        from nanobot.providers.anthropic_provider import AnthropicProvider
+
+        return AnthropicProvider(
+            api_key=api_key,
+            api_base=api_base,
+            default_model=model,
+            extra_headers=extra_headers,
+        )
+
+    return PatchedOpenAICompatProvider(
+        api_key=api_key,
+        api_base=api_base,
+        default_model=model,
+        extra_headers=extra_headers,
+        spec=spec,
+    )