from fastapi import FastAPI ,Request ,Form, UploadFile, File
from fastapi.responses import JSONResponse
from fastapi.responses import HTMLResponse, FileResponse
import os
import io
from PIL import ImageOps,Image ,ImageFilter
from transformers import pipeline
import matplotlib.pyplot as plt
import numpy as np


app = FastAPI()

# Root route
@app.get('/')
def hello_world():
    return "Hello World taha"

def get_segment_image(raw_image):
    pipe = pipeline("image-segmentation", model="Intel/dpt-large-ade")
    output = pipe(raw_image, points_per_batch=32)
    return output

def get_supported_segmentation(output):
    return [obj for obj in output if obj['label']=='person']


@app.post('/predict')
async  def predict(name: str = Form(),age: str = Form() , file: UploadFile = File(...)): 
    # Form(...) to accept input as web form ,may change when android /upload 
    '''
    contents = await file.read()
    
    image = Image.open(io.BytesIO(contents))
    

    return {
        "message": f"Your name is {name}, age is {age}",
        "filename": file.filename,
        "image:": str(np.array(image))  # Returns the original image size
    }
    '''
    contents = await file.read()
    image = Image.open(io.BytesIO(contents))
       
    # Process the image (example: convert to grayscale)
    processed_image = image.convert("L")

    # Save the processed image to a temporary file
    output_file_path = "tmp_processed_image.png"
    processed_image.save(output_file_path)

    # Return the processed image for download
    return FileResponse(output_file_path, media_type='image/png', filename="tmp_processed_image.png")