app.py

import streamlit as st
import fitz  # PyMuPDF
import io
import zipfile
from PIL import Image

def convert_pdf_to_images(pdf_data, dpi_value):
    """
    Convert each page of a PDF into a list of PIL.Image objects using PyMuPDF.

    Args:
        pdf_data (bytes): PDF content in bytes
        dpi_value (int): DPI (resolution) for rasterization

    Returns:
        list: A list of PIL.Image objects representing each page
    """
    images_list = []
    # open PDF from memory
    with fitz.open(stream=pdf_data, filetype="pdf") as doc:
        for page_index in range(len(doc)):
            page = doc.load_page(page_index)
            # `matrix = fitz.Matrix(scale_x, scale_y)` で拡大率を指定
            # DPI は (72 * scale_x, 72 * scale_y) になるイメージ
            scale_val = dpi_value / 72.0  # PyMuPDF は 72dpi 基準
            matrix = fitz.Matrix(scale_val, scale_val)
            pix = page.get_pixmap(matrix=matrix)
            mode = "RGBA" if pix.alpha else "RGB"
            pil_image = Image.frombytes(mode, [pix.width, pix.height], pix.samples)
            images_list.append(pil_image)
    return images_list

def create_zip_from_images(images_list):
    """
    Create a ZIP file in memory from a list of PIL.Image objects.

    Args:
        images_list (list): A list of PIL.Image objects

    Returns:
        BytesIO: A BytesIO object representing the ZIP file
    """
    zip_buffer = io.BytesIO()
    with zipfile.ZipFile(zip_buffer, 'w', zipfile.ZIP_DEFLATED) as zf:
        for idx, img in enumerate(images_list):
            img_bytes = io.BytesIO()
            img.save(img_bytes, format="PNG")
            img_bytes.seek(0)
            zf.writestr(f"page_{idx+1}.png", img_bytes.read())
    zip_buffer.seek(0)
    return zip_buffer

def main():
    """
    Streamlit application entry point.
    Provides a PDF-to-image converter that outputs a ZIP file without using Poppler.
    """
    st.title("PDFを画像に変換 (PyMuPDF 版)")
    st.write("""
        Poppler を使わずに PyMuPDF を用いて PDF を各ページごとに PNG へ変換し、  
        ZIP にまとめてダウンロードできるサンプルアプリです。
    """)

    uploaded_file = st.file_uploader("PDFファイルをアップロードしてください", type=["pdf"])

    if uploaded_file is not None:
        dpi = st.slider("解像度（dpi）を選択してください", min_value=10, max_value=6000, value=300, step=10)

        if st.button("変換開始"):
            try:
                with st.spinner("変換中...しばらくお待ちください"):
                    images_converted = convert_pdf_to_images(uploaded_file.read(), dpi)
                    zip_file_bytes = create_zip_from_images(images_converted)
                
                st.success("変換が完了しました！")
                st.download_button(
                    label="ZIPファイルをダウンロード",
                    data=zip_file_bytes,
                    file_name="pdf_images.zip",
                    mime="application/zip"
                )

                # Preview images (optional)
                st.write("変換された画像のプレビュー:")
                for idx, image_item in enumerate(images_converted):
                    st.image(image_item, caption=f"ページ {idx+1}", use_column_width=True)

            except Exception as e:
                st.error(f"変換中にエラーが発生しました: {str(e)}")

if __name__ == "__main__":
    main()