Add provider implementations (OpenAI, Anthropic, Ollama)

src/promptforge/providers/ollama.py

@@ -0,0 +1,106 @@
+import asyncio
+import time
+from typing import Any, AsyncIterator, Dict, Optional
+import httpx
+
+from .base import ProviderBase, ProviderResponse
+from ..core.exceptions import ProviderError
+
+
+class OllamaProvider(ProviderBase):
+    def __init__(
+        self,
+        model: str = "llama2",
+        temperature: float = 0.7,
+        base_url: str = "http://localhost:11434",
+        **kwargs,
+    ):
+        super().__init__(None, model, temperature, **kwargs)
+        self.base_url = base_url.rstrip('/')
+
+    @property
+    def name(self) -> str:
+        return "ollama"
+
+    async def complete(
+        self,
+        prompt: str,
+        system_prompt: Optional[str] = None,
+        max_tokens: Optional[int] = None,
+        **kwargs,
+    ) -> ProviderResponse:
+        start_time = time.time()
+
+        try:
+            async with httpx.AsyncClient() as client:
+                payload = {
+                    "model": self.model,
+                    "prompt": prompt,
+                    "stream": False,
+                    "options": {
+                        "temperature": self.temperature,
+                    }
+                }
+                if max_tokens:
+                    payload["options"]["num_predict"] = max_tokens
+
+                response = await client.post(
+                    f"{self.base_url}/api/generate",
+                    json=payload,
+                    timeout=120.0
+                )
+
+                response.raise_for_status()
+                data = response.json()
+
+                latency_ms = (time.time() - start_time) * 1000
+
+                return ProviderResponse(
+                    content=data.get("response", ""),
+                    model=self.model,
+                    provider=self.name,
+                    latency_ms=latency_ms,
+                )
+        except httpx.HTTPStatusError as e:
+            raise ProviderError(f"Ollama HTTP error: {e}")
+        except httpx.RequestError as e:
+            raise ProviderError(f"Ollama connection error: {e}")
+
+    async def stream_complete(
+        self,
+        prompt: str,
+        system_prompt: Optional[str] = None,
+        max_tokens: Optional[int] = None,
+        **kwargs,
+    ) -> AsyncIterator[str]:
+        try:
+            async with httpx.AsyncClient() as client:
+                payload = {
+                    "model": self.model,
+                    "prompt": prompt,
+                    "stream": True,
+                    "options": {
+                        "temperature": self.temperature,
+                    }
+                }
+                if max_tokens:
+                    payload["options"]["num_predict"] = max_tokens
+
+                async with client.stream(
+                    "POST",
+                    f"{self.base_url}/api/generate",
+                    json=payload,
+                    timeout=120.0
+                ) as response:
+                    async for line in response.aiter_lines():
+                        import json
+                        data = json.loads(line)
+                        if "response" in data:
+                            yield data["response"]
+        except httpx.HTTPStatusError as e:
+            raise ProviderError(f"Ollama HTTP error: {e}")
+        except httpx.RequestError as e:
+            raise ProviderError(f"Ollama connection error: {e}")
+
+    def validate_api_key(self) -> bool:
+        return True