app.py

import gradio as gr
import os
import tempfile
import numpy as np
import cv2
import img2pdf
import shutil
import uuid
from pathlib import Path
from pdf2image import convert_from_path  # pip install pdf2image opencv-python numpy img2pdf

# Bounding box: (x1, y1, x2, y2) in fractions of image size
default_bbox = (0.75, 0.9, 0.98, 0.98)

def expand_bbox(bbox, expand_factor=0.3):
    x_min, y_min, x_max, y_max = bbox
    width = x_max - x_min
    height = y_max - y_min

    new_x_min = max(0, x_min - width * expand_factor)
    new_y_min = max(0, y_min - height * expand_factor)
    new_x_max = min(1, x_max + width * expand_factor)
    new_y_max = min(1, y_max + height * expand_factor)
    return (new_x_min, new_y_min, new_x_max, new_y_max)

def inpaint_image(image_np, bbox, color=(255, 255, 255)):
    h, w, _ = image_np.shape
    x1 = int(bbox[0] * w)
    y1 = int(bbox[1] * h)
    x2 = int(bbox[2] * w)
    y2 = int(bbox[3] * h)
    inpainted_image = image_np.copy()
    cv2.rectangle(inpainted_image, (x1, y1), (x2, y2), color, -1)
    return inpainted_image

def process_pdf(pdf_file):
    try:
        pdf_path = pdf_file.name if pdf_file else "wifi_basics_teaching_genspark.pdf"
        assert os.path.exists(pdf_path), f"File not found: {pdf_path}"

        page_inpaint_colors = {
            0: (255, 255, 255),
            1: (0, 0, 0)
        }

        with tempfile.TemporaryDirectory() as tmpdir:
            images = convert_from_path(pdf_path, dpi=300, output_folder=tmpdir, fmt="png")
            output_images = []

            for i, img in enumerate(images):
                np_img = np.array(img)[:, :, ::-1]
                bbox = expand_bbox(default_bbox, expand_factor=0.3)
                color = page_inpaint_colors.get(i, (255, 255, 255))
                inpainted = inpaint_image(np_img, bbox, color=color)

                img_path = os.path.join(tmpdir, f"page_{i}.png")
                cv2.imwrite(img_path, inpainted)
                output_images.append(img_path)

            temp_pdf_path = os.path.join(tmpdir, "inpainted_output.pdf")
            with open(temp_pdf_path, "wb") as f:
                f.write(img2pdf.convert([Path(p) for p in sorted(output_images)]))

        # Save to persistent path
        final_pdf_path = f"inpainted_output_{uuid.uuid4().hex[:8]}.pdf"
        shutil.copy(temp_pdf_path, final_pdf_path)
        return final_pdf_path

    except Exception as e:
        print(f"❌ Error during PDF processing: {e}")
        return None

# Gradio UI
with gr.Blocks() as demo:
    gr.Markdown("### 🧽 PDF Logo Inpainting (Solid Fill, Expand + Color Config)")
    with gr.Row():
        pdf_input = gr.File(
            label="Upload PDF (or leave empty to use test.pdf)",
            file_types=[".pdf"],
            interactive=True
        )
        pdf_output = gr.File(label="Download Inpainted PDF")

    run_button = gr.Button("Process and Inpaint")
    run_button.click(fn=process_pdf, inputs=pdf_input, outputs=pdf_output)

demo.launch()