app.py

import os
import json
import time
os.environ['HF_HOME'] = '/tmp/cache'
os.environ['TORCH_HOME'] = '/tmp/cache'

from fastapi import FastAPI, File, UploadFile, HTTPException
from fastapi.middleware.cors import CORSMiddleware
from fastapi.responses import HTMLResponse
from PIL import Image
import torch
import io
import colorthief
import tempfile
import numpy as np

app = FastAPI(title="Fashion Classification API")

# Middleware CORS
app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
    expose_headers=["*"]
)

# --- ÉTAT DU MODÈLE ---
print("⚠️ Démarrage du chargement du modèle Marqo-FashionSigLIP...")
model = None
processor = None
model_loading = False
model_loaded = False
model_error = None

def load_fashion_model():
    global model, processor, model_loading, model_loaded, model_error
    
    model_loading = True
    try:
        from transformers import AutoModel, AutoProcessor
        
        model_name = "Marqo/Marqo-FashionSigLIP-Classification"
        
        print("📦 Téléchargement du modèle... (cela peut prendre 5-10 minutes)")
        
        # Charger le modèle SigLIP
        model = AutoModel.from_pretrained(
            model_name,
            cache_dir="/tmp/cache",
            torch_dtype=torch.float16,
            trust_remote_code=True
        )
        
        processor = AutoProcessor.from_pretrained(
            model_name,
            trust_remote_code=True
        )
        
        print("✅ Modèle Marqo-FashionSigLIP chargé avec succès !")
        model_loaded = True
        model_loading = False
        
    except Exception as e:
        print(f"❌ Erreur chargement modèle: {e}")
        model_error = str(e)
        model_loading = False
        import traceback
        traceback.print_exc()

# Démarrer le chargement IMMÉDIATEMENT
load_fashion_model()

# Catégories de mode
categories = [
    "t-shirt", "dress", "jeans", "shirt", "skirt", 
    "sneakers", "handbag", "jacket", "shorts", "sweater",
    "coat", "high heels", "blouse", "boots", "hat"
]

@app.get("/")
def read_root():
    return {
        "message": "Fashion Classification API is running!", 
        "status": "OK",
        "model_status": "loaded" if model_loaded else "loading" if model_loading else "error",
        "model_name": "Marqo-FashionSigLIP-Classification"
    }

@app.get("/health")
def health_check():
    return {
        "model_loaded": model_loaded,
        "model_loading": model_loading,
        "model_error": model_error,
        "status": "ready" if model_loaded else "loading" if model_loading else "error",
        "model_name": "Marqo-FashionSigLIP-Classification",
        "timestamp": time.time()
    }

@app.post("/analyze")
async def analyze_image(file: UploadFile = File(...)):
    # Vérifier si le modèle est chargé
    if not model_loaded:
        if model_loading:
            raise HTTPException(status_code=423, detail="Model still loading. Please wait 5-10 minutes and check /health")
        else:
            raise HTTPException(status_code=500, detail=f"Model failed to load: {model_error}")
    
    if model is None or processor is None:
        raise HTTPException(status_code=500, detail="Model not available")
    
    try:
        # Lire et préparer l'image
        contents = await file.read()
        image = Image.open(io.BytesIO(contents)).convert("RGB")
        image = image.resize((384, 384))
        
        # Traitement avec SigLIP
        inputs = processor(
            text=categories,
            images=image,
            return_tensors="pt",
            padding=True,
            truncation=True,
            max_length=64,
        )
        
        device = next(model.parameters()).device
        inputs = {k: v.to(device) for k, v in inputs.items()}
        
        with torch.no_grad():
            outputs = model(**inputs)
        
        logits_per_image = outputs.logits_per_image
        probs = torch.sigmoid(logits_per_image)
        probs = probs.cpu().numpy()[0]
        
        predicted_idx = np.argmax(probs)
        category_name = categories[predicted_idx]
        confidence_score = float(probs[predicted_idx])
        
        # Analyse couleur
        try:
            with tempfile.NamedTemporaryFile(suffix='.jpg', delete=False) as tmp:
                image.save(tmp, format='JPEG')
                tmp_path = tmp.name
            
            color_thief = colorthief.ColorThief(tmp_path)
            dominant_color = color_thief.get_color(quality=1)
            hex_color = '#%02x%02x%02x' % dominant_color
            os.unlink(tmp_path)
            
        except Exception:
            hex_color = "#000000"

        return {
            "category": category_name,
            "confidence": round(confidence_score, 4),
            "color_hex": hex_color,
            "model": "Marqo-FashionSigLIP-Classification"
        }

    except Exception as e:
        raise HTTPException(status_code=500, detail=f"Analysis error: {str(e)}")

# Interface de test avec statut de chargement
@app.get("/test-ui", response_class=HTMLResponse)
async def test_ui():
    return f"""
    <html>
        <head>
            <title>FashionSigLIP Detection</title>
            <style>
                body {{ font-family: Arial, sans-serif; margin: 40px; }}
                .container {{ max-width: 600px; margin: 0 auto; }}
                form {{ border: 2px dashed #ccc; padding: 30px; text-align: center; }}
                .status {{ padding: 15px; margin: 10px 0; border-radius: 5px; }}
                .loading {{ background: #fff3cd; color: #856404; }}
                .ready {{ background: #d4edda; color: #155724; }}
                .error {{ background: #f8d7da; color: #721c24; }}
            </style>
            <script>
                function checkStatus() {{
                    fetch('/health')
                        .then(response => response.json())
                        .then(data => {{
                            const statusDiv = document.getElementById('model-status');
                            const submitBtn = document.getElementById('submit-btn');
                            
                            if (data.model_loaded) {{
                                statusDiv.innerHTML = '✅ <b>Modèle chargé et prêt !</b>';
                                statusDiv.className = 'status ready';
                                submitBtn.disabled = false;
                            }} else if (data.model_loading) {{
                                statusDiv.innerHTML = '⏳ <b>Chargement du modèle en cours...</b><br>Cela peut prendre 5-10 minutes';
                                statusDiv.className = 'status loading';
                                submitBtn.disabled = true;
                                setTimeout(checkStatus, 5000); // Re-check dans 5 sec
                            }} else {{
                                statusDiv.innerHTML = '❌ <b>Erreur de chargement:</b><br>' + (data.model_error || 'Unknown error');
                                statusDiv.className = 'status error';
                                submitBtn.disabled = true;
                            }}
                        }});
                }}
                
                // Vérifier le statut au chargement de la page
                window.onload = checkStatus;
            </script>
        </head>
        <body>
            <div class="container">
                <h1>👗 FashionSigLIP Detector</h1>
                
                <div id="model-status" class="status loading">
                    Vérification du statut du modèle...
                </div>
                
                <form action="/analyze" method="post" enctype="multipart/form-data">
                    <h3>Uploader une image de vêtement :</h3>
                    <input type="file" name="file" accept="image/*" required>
                    <br><br>
                    <input type="submit" id="submit-btn" value="Analyser" disabled>
                </form>
                
                <div style="margin-top: 20px;">
                    <button onclick="checkStatus()">Actualiser le statut</button>
                    <button onclick="location.reload()">Rafraîchir la page</button>
                </div>
            </div>
        </body>
    </html>
    """