app.py

from fastapi import FastAPI, HTTPException, Request
from pydantic import BaseModel
import requests
import subprocess
import time
import threading
import json
from typing import List, Optional, Dict, Any

app = FastAPI(title="Ollama OpenAI-Compatible API", description="OpenAI-compatible endpoint for Ollama")

# Add request logging middleware
@app.middleware("http")
async def log_requests(request: Request, call_next):
    print(f"🌐 Incoming request: {request.method} {request.url}")
    print(f"📋 Headers: {dict(request.headers)}")
    print(f"🔍 Path: {request.url.path}")
    print(f"❓ Query params: {request.url.query}")
    
    response = await call_next(request)
    
    print(f"📤 Response status: {response.status_code}")
    return response

# OpenAI-compatible models
class ChatMessage(BaseModel):
    role: str
    content: str

class ChatCompletionRequest(BaseModel):
    model: str
    messages: List[ChatMessage]
    temperature: Optional[float] = 0.7
    max_tokens: Optional[int] = None
    stream: Optional[bool] = False

class ChatCompletionResponse(BaseModel):
    id: str
    object: str = "chat.completion"
    created: int
    model: str
    choices: List[Dict[str, Any]]
    usage: Optional[Dict[str, int]] = None

class ModelInfo(BaseModel):
    id: str
    object: str = "model"
    created: int
    owned_by: str = "ollama"

# Global variable to track if Ollama is ready
ollama_ready = False

def wait_for_ollama():
    """Wait for Ollama server to be ready"""
    global ollama_ready
    max_retries = 30
    for i in range(max_retries):
        try:
            response = requests.get("http://localhost:11434/api/tags", timeout=5)
            if response.status_code == 200:
                ollama_ready = True
                print("Ollama server is ready!")
                return
        except:
            pass
        print(f"Waiting for Ollama server... ({i+1}/{max_retries})")
        time.sleep(10)
    print("Ollama server failed to start")

def pull_default_model():
    """Pull a small default model"""
    try:
        print("Checking for existing models...")
        # Model should already be pulled during build, but check anyway
        subprocess.run(["ollama", "list"], check=True, timeout=30)
        print("Models available")
    except Exception as e:
        print(f"Error checking models: {e}")

# Start background tasks
def initialize():
    wait_for_ollama()
    if ollama_ready:
        pull_default_model()

threading.Thread(target=initialize, daemon=True).start()

def convert_to_ollama_messages(messages: List[ChatMessage]) -> List[Dict[str, str]]:
    """Convert OpenAI format messages to Ollama format"""
    return [{"role": msg.role, "content": msg.content} for msg in messages]

@app.get("/")
async def root():
    print("Root endpoint accessed")
    return {
        "message": "Ollama OpenAI-Compatible API is running",
        "endpoints": {
            "chat_completions": "/v1/chat/completions",
            "models": "/v1/models"
        },
        "ready": ollama_ready
    }

@app.get("/health")
async def health():
    print("Health endpoint accessed")
    return {"status": "healthy", "ollama_ready": ollama_ready}

# Add the missing /v1 endpoint that nanocoder checks
@app.get("/v1")
async def v1_root():
    print("V1 root endpoint accessed")
    return {
        "message": "OpenAI-compatible API v1",
        "endpoints": {
            "chat_completions": "/v1/chat/completions",
            "models": "/v1/models"
        },
        "ready": ollama_ready
    }

@app.get("/v1/models")
async def list_models():
    """List available models in OpenAI format"""
    print("V1 models endpoint accessed")
    if not ollama_ready:
        print("Ollama not ready, returning 503")
        raise HTTPException(status_code=503, detail="Ollama server not ready")
    
    try:
        response = requests.get("http://localhost:11434/api/tags", timeout=10)
        if response.status_code == 200:
            ollama_models = response.json()
            models = []
            for model in ollama_models.get("models", []):
                models.append({
                    "id": model["name"],
                    "object": "model",
                    "created": int(time.time()),
                    "owned_by": "ollama"
                })
            print(f"Returning {len(models)} models")
            return {"object": "list", "data": models}
        else:
            print(f"Ollama tags returned {response.status_code}")
            raise HTTPException(status_code=500, detail="Failed to fetch models")
    except requests.exceptions.RequestException as e:
        print(f"Error fetching models: {e}")
        raise HTTPException(status_code=500, detail=str(e))

@app.post("/v1/chat/completions")
async def create_chat_completion(request: ChatCompletionRequest):
    """OpenAI-compatible chat completions endpoint - no authentication required"""
    print(f"Chat completions endpoint accessed with model: {request.model}")
    
    if not ollama_ready:
        print("Ollama not ready, returning 503")
        raise HTTPException(status_code=503, detail="Ollama server not ready")
    
    try:
        # Convert to Ollama format
        ollama_messages = convert_to_ollama_messages(request.messages)
        print(f"Converted {len(ollama_messages)} messages for Ollama")
        
        ollama_request = {
            "model": request.model,
            "messages": ollama_messages,
            "stream": False,
            "options": {
                "num_ctx": 32768  # Maximum context length
            }
        }
        
        if request.temperature is not None:
            ollama_request["options"]["temperature"] = request.temperature
        if request.max_tokens is not None:
            ollama_request["options"]["num_predict"] = request.max_tokens

        print(f"Making request to Ollama with: {ollama_request}")
        
        # Make request to Ollama
        response = requests.post(
            "http://localhost:11434/api/chat",
            json=ollama_request,
            timeout=120
        )
        
        print(f"Ollama response status: {response.status_code}")
        
        if response.status_code == 200:
            ollama_response = response.json()
            print("Successfully got response from Ollama")
            
            # Convert to OpenAI format
            openai_response = {
                "id": f"chatcmpl-{int(time.time())}",
                "object": "chat.completion",
                "created": int(time.time()),
                "model": request.model,
                "choices": [{
                    "index": 0,
                    "message": {
                        "role": "assistant",
                        "content": ollama_response.get("message", {}).get("content", "")
                    },
                    "finish_reason": "stop"
                }],
                "usage": {
                    "prompt_tokens": ollama_response.get("prompt_eval_count", 0),
                    "completion_tokens": ollama_response.get("eval_count", 0),
                    "total_tokens": ollama_response.get("prompt_eval_count", 0) + ollama_response.get("eval_count", 0)
                }
            }
            
            return openai_response
        else:
            print(f"Ollama returned error: {response.status_code} - {response.text}")
            raise HTTPException(status_code=500, detail=f"Ollama error: {response.status_code}")
            
    except requests.exceptions.RequestException as e:
        print(f"Request exception: {e}")
        raise HTTPException(status_code=500, detail=str(e))

# Legacy endpoints for direct Ollama access
@app.post("/api/generate")
async def generate(request: dict):
    print("Generate endpoint accessed")
    if not ollama_ready:
        raise HTTPException(status_code=503, detail="Ollama server not ready")
    
    try:
        response = requests.post(
            "http://localhost:11434/api/generate",
            json=request,
            timeout=120
        )
        return response.json()
    except requests.exceptions.RequestException as e:
        raise HTTPException(status_code=500, detail=str(e))

@app.post("/api/chat")
async def chat(request: dict):
    print("Chat endpoint accessed")
    if not ollama_ready:
        raise HTTPException(status_code=503, detail="Ollama server not ready")
    
    try:
        response = requests.post(
            "http://localhost:11434/api/chat",
            json=request,
            timeout=120
        )
        return response.json()
    except requests.exceptions.RequestException as e:
        raise HTTPException(status_code=500, detail=str(e))

if __name__ == "__main__":
    import uvicorn
    uvicorn.run(app, host="0.0.0.0", port=7860)