app.py

import gradio as gr
import requests
import json

OLLAMA_URL = "http://localhost:11434"

MODELS = {
    "Qwen2.5-Coder 1.5B (Fastest)": "qwen2.5-coder:1.5b",
    "Qwen2.5-Coder 3B (Fast)": "qwen2.5-coder:3b",
    "Qwen2.5-Coder 7B (Quality)": "qwen2.5-coder:7b",
}

def check_ollama():
    try:
        r = requests.get(f"{OLLAMA_URL}/api/tags", timeout=5)
        return r.status_code == 200
    except:
        return False

def chat_stream(message, history, model_name, temperature, max_tokens):
    if not check_ollama():
        yield "⏳ Ollama starting... wait 30 seconds and try again."
        return
    
    model = MODELS.get(model_name, "qwen2.5-coder:3b")
    messages = [{"role": "system", "content": "You are an expert coding assistant. Always use markdown code blocks."}]
    
    for user_msg, assistant_msg in history:
        messages.append({"role": "user", "content": user_msg})
        if assistant_msg:
            messages.append({"role": "assistant", "content": assistant_msg})
    
    messages.append({"role": "user", "content": message})
    
    try:
        response = requests.post(
            f"{OLLAMA_URL}/api/chat",
            json={"model": model, "messages": messages, "stream": True, "options": {"temperature": temperature, "num_predict": max_tokens}},
            stream=True, timeout=300
        )
        
        full = ""
        for line in response.iter_lines():
            if line:
                try:
                    data = json.loads(line)
                    if "message" in data:
                        full += data["message"].get("content", "")
                        yield full
                except:
                    continue
    except Exception as e:
        yield f"Error: {e}"

def generate_code(prompt, language, model_name, max_tokens):
    if not prompt.strip():
        return "Please describe what you want."
    if not check_ollama():
        return "⏳ Ollama starting..."
    
    model = MODELS.get(model_name, "qwen2.5-coder:3b")
    full_prompt = f"Write {language} code for: {prompt}\n\nOutput ONLY code in a markdown block."
    
    try:
        r = requests.post(
            f"{OLLAMA_URL}/api/generate",
            json={"model": model, "prompt": full_prompt, "stream": False, "options": {"temperature": 0.3, "num_predict": max_tokens}},
            timeout=300
        )
        if r.status_code == 200:
            result = r.json().get("response", "")
            if "```" in result:
                parts = result.split("```")
                if len(parts) >= 2:
                    code = parts[1]
                    if "\n" in code:
                        code = code.split("\n", 1)[-1]
                    return code.strip()
            return result
        return f"Error: {r.text}"
    except Exception as e:
        return f"Error: {e}"

def explain_code(code, model_name, max_tokens):
    if not code.strip():
        return "Paste code to explain."
    if not check_ollama():
        return "⏳ Ollama starting..."
    
    model = MODELS.get(model_name, "qwen2.5-coder:3b")
    
    try:
        r = requests.post(
            f"{OLLAMA_URL}/api/generate",
            json={"model": model, "prompt": f"Explain this code:\n```\n{code}\n```", "stream": False, "options": {"num_predict": max_tokens}},
            timeout=300
        )
        return r.json().get("response", "") if r.status_code == 200 else f"Error: {r.text}"
    except Exception as e:
        return f"Error: {e}"

def fix_code(code, error, model_name, max_tokens):
    if not code.strip():
        return "Paste code to fix."
    if not check_ollama():
        return "⏳ Ollama starting..."
    
    model = MODELS.get(model_name, "qwen2.5-coder:3b")
    prompt = f"Fix this code:\n```\n{code}\n```\nError: {error or 'Not working'}"
    
    try:
        r = requests.post(
            f"{OLLAMA_URL}/api/generate",
            json={"model": model, "prompt": prompt, "stream": False, "options": {"temperature": 0.3, "num_predict": max_tokens}},
            timeout=300
        )
        return r.json().get("response", "") if r.status_code == 200 else f"Error: {r.text}"
    except Exception as e:
        return f"Error: {e}"

with gr.Blocks(title="GOD Coding Machine", theme=gr.themes.Soft(primary_hue="purple")) as demo:
    
    gr.Markdown("# 🔥 GOD Coding Machine\n**Docker Edition** • Qwen2.5-Coder running locally • No rate limits!")
    
    with gr.Row():
        model_dropdown = gr.Dropdown(choices=list(MODELS.keys()), value="Qwen2.5-Coder 3B (Fast)", label="🤖 Model")
        temperature = gr.Slider(0, 1, value=0.7, step=0.1, label="🌡️ Temperature")
        max_tokens = gr.Slider(256, 4096, value=2048, step=256, label="📏 Max Tokens")
    
    with gr.Tabs():
        with gr.TabItem("💬 Chat"):
            chatbot = gr.Chatbot(height=400)
            with gr.Row():
                msg = gr.Textbox(placeholder="Ask about coding...", show_label=False, scale=9)
                send = gr.Button("Send", variant="primary", scale=1)
            clear = gr.Button("Clear")
            gr.Examples(["Write a Python quicksort function", "Explain async/await in JavaScript"], inputs=msg)
        
        with gr.TabItem("⚡ Generate"):
            with gr.Row():
                with gr.Column():
                    gen_prompt = gr.Textbox(label="Describe what you want", lines=3)
                    gen_lang = gr.Dropdown(["Python", "JavaScript", "TypeScript", "Go", "Rust", "Java", "C++"], value="Python", label="Language")
                    gen_btn = gr.Button("Generate", variant="primary")
                gen_output = gr.Code(label="Code", language="python", lines=15)
        
        with gr.TabItem("🔍 Explain"):
            with gr.Row():
                explain_input = gr.Code(label="Paste code", lines=10)
                explain_output = gr.Markdown()
            explain_btn = gr.Button("Explain", variant="primary")
        
        with gr.TabItem("🔧 Fix"):
            with gr.Row():
                with gr.Column():
                    fix_input = gr.Code(label="Buggy code", lines=10)
                    fix_error = gr.Textbox(label="Error message", lines=2)
                    fix_btn = gr.Button("Fix", variant="primary")
                fix_output = gr.Markdown()
    
    def respond(message, history, model, temp, tokens):
        history = history or []
        for chunk in chat_stream(message, history, model, temp, tokens):
            yield history + [[message, chunk]], ""
    
    msg.submit(respond, [msg, chatbot, model_dropdown, temperature, max_tokens], [chatbot, msg])
    send.click(respond, [msg, chatbot, model_dropdown, temperature, max_tokens], [chatbot, msg])
    clear.click(lambda: [], None, chatbot)
    gen_btn.click(generate_code, [gen_prompt, gen_lang, model_dropdown, max_tokens], gen_output)
    explain_btn.click(explain_code, [explain_input, model_dropdown, max_tokens], explain_output)
    fix_btn.click(fix_code, [fix_input, fix_error, model_dropdown, max_tokens], fix_output)

demo.launch(server_name="0.0.0.0", server_port=7860)