app.py

import gradio as gr
import tensorflow as tf
import numpy as np
from PIL import Image
import json
from gtts import gTTS
import os
import tempfile
from health_logic import generate_advice
from deep_translator import GoogleTranslator
from datetime import datetime
from fpdf import FPDF

# Load model
model = tf.keras.models.load_model("food_vision_model.keras")

# Load metadata
with open("food_info.json", "r") as f:
    food_info = json.load(f)

class_names = list(food_info.keys())

# Language code mapping for gTTS
LANG_CODE = {
    "English": "en",
    "Hausa": "ha",
    "Yoruba": "yo",
    "Igbo": "ig"
}

# Risk color level
def get_risk_level(score):
    if score <= 30:
        return "🟢 Low Risk"
    elif score <= 70:
        return "🟡 Medium Risk"
    else:
        return "🔴 High Risk"

# PDF Generator using Unicode-safe font
class UnicodePDF(FPDF):
    def __init__(self):
        super().__init__()
        self.add_font("DejaVu", "", "DejaVuSans.ttf", uni=True)
        self.add_font("DejaVu", "B", "DejaVuSans.ttf", uni=True)
        self.set_font("DejaVu", size=12)

def generate_pdf(info, advice, risk_score, risk_level, flags, conditions, lang, display_name, image_path):
    pdf = UnicodePDF()
    pdf.add_page()
    pdf.set_font("DejaVu", "B", 16)
    pdf.cell(0, 10, "HoodHealth Pro+ Report", ln=True)

    pdf.set_font("DejaVu", size=12)
    pdf.cell(0, 10, f"Date: {datetime.now().strftime('%Y-%m-%d %H:%M')}", ln=True)
    pdf.ln(10)

    if os.path.exists(image_path):
        pdf.image(image_path, x=10, y=40, w=60)

    pdf.set_xy(75, 40)
    pdf.multi_cell(0, 10,
        f"🍽️ Food: {display_name}\n"
        f"🌍 Ethnicity: {info['ethnicity']}\n"
        f"🥦 Ingredients: {info['ingredients']}"
    )

    pdf.ln(35)
    pdf.multi_cell(0, 10,
        f"🔥 Calories: {info['calories']} kcal\n"
        f"🍞 Carbs: {info['carbs']}g\n"
        f"🥩 Protein: {info['protein']}g\n"
        f"🧈 Fat: {info['fat']}g"
    )

    pdf.ln(5)
    pdf.multi_cell(0, 10,
        f"🌱 Diet Type: {info['diet_type']}\n"
        f"🔁 Substitute: {info.get('substitute', 'None')}"
    )

    pdf.ln(10)
    pdf.set_font("DejaVu", "B", 14)
    pdf.cell(0, 10, "Health Analysis", ln=True)
    pdf.set_font("DejaVu", size=12)

    pdf.multi_cell(0, 10,
        f"📋 Selected Conditions: {', '.join(conditions)}\n"
        f"📊 Risk Score: {risk_score}% ({risk_level})\n"
        f"⚠️ Risk Factors: {', '.join(flags) if flags else 'None'}\n"
        f"✅ Advice: {advice}"
    )

    pdf.ln(5)
    pdf.set_font("DejaVu", "I", 10)
    pdf.multi_cell(0, 10, "*Note: This advice is generated using simplified nutritional rules and is not a medical diagnosis.*")

    temp_pdf = os.path.join(tempfile.gettempdir(), "report.pdf")
    pdf.output(temp_pdf)

    return temp_pdf

# Main logic
def classify_food(image, conditions, language):
    if not conditions:
        return "⚠️ Please select at least one health condition.", None, None, None

    image = image.convert("RGB").resize((224, 224))
    img_array = tf.keras.preprocessing.image.img_to_array(image) / 255.0
    img_array = np.expand_dims(img_array, axis=0)

    predictions = model.predict(img_array)[0]
    predicted_index = np.argmax(predictions)
    predicted_class = class_names[predicted_index]
    confidence = round(float(predictions[predicted_index]) * 100, 2)

    info = food_info[predicted_class]
    display_name = info.get("display_name", predicted_class.title())

    advice, risk_score, flags = generate_advice(info, conditions)
    risk_level = get_risk_level(risk_score)

    # Translate
    if language != "English":
        try:
            translated = GoogleTranslator(source="auto", target=LANG_CODE[language]).translate(advice)
            advice = translated
        except:
            advice += " (⚠️ Translation failed)"

    # TTS with fallback
    tts_audio = None
    try:
        tts_lang = LANG_CODE.get(language, "en")
        tts = gTTS(text=advice, lang=tts_lang)
        tts_audio = os.path.join(tempfile.gettempdir(), "tts.mp3")
        tts.save(tts_audio)
    except Exception as e:
        print(f"TTS error: {e}")
        advice += " (⚠️ Voice advice unavailable)"

    # Save image and PDF
    tmp_img = os.path.join(tempfile.gettempdir(), "upload.jpg")
    image.save(tmp_img)
    pdf_path = generate_pdf(info, advice, risk_score, risk_level, flags, conditions, language, display_name, tmp_img)

    # Output
    result = (
        f"🧠 *Predictions*:\n"
        f"🍲 {display_name} – {confidence}%\n\n"
        f"📉 *Top Confidence*: {confidence}%\n"
        f"📊 *Risk Score*: {risk_score}% ({risk_level})\n"
        f"⚠️ *Risk Factors*: {', '.join(flags) if flags else 'None'}\n"
        f"✅ *Advice*: {advice}\n"
        f"🔁 *Substitute*: {info.get('substitute', 'None')}\n\n"
        f"📝 This advice is generated using simple nutrition rules and is not a medical diagnosis."
    )

    return result, tts_audio, f"🗣️ Language: {language}", pdf_path

# Interface
interface = gr.Interface(
    fn=classify_food,
    inputs=[
        gr.Image(type="pil", label="Upload Food Image"),
        gr.CheckboxGroup(
            label="Select Health Conditions",
            choices=["Normal", "Diabetic", "Hypertensive", "Weight Loss", "Malnourished", "Pregnant/Nursing", "Cholesterol Watch"]
        ),
        gr.Dropdown(label="Language for TTS", choices=list(LANG_CODE.keys()), value="English")
    ],
    outputs=[
        gr.Textbox(label="Result", lines=10),
        gr.Audio(label="Hear Advice"),
        gr.Textbox(label="TTS Language"),
        gr.File(label="Download Health Report (PDF)")
    ],
    title="🥗 FoodHealth Pro+",
    description="Upload a food image. Get full nutrition info, risk score, health advice with voice support in your language and download a detailed PDF."
)

if __name__ == "__main__":
    interface.launch()