my_io.py

doc = None 

def render_page(page):
    # Render the page as a PNG image with a resolution of 150 DPI
    pix = page.get_pixmap(dpi=150)
    image = Image.frombytes("RGB", [pix.width, pix.height], pix.samples)
    return image

def ready(file, page_num:int):

    global doc    

    # Use PyMuPDF to render the pages of the uploaded document
    doc = pymupdf.open(file)
    
    if page_num > doc.page_count:
        page_num = doc.page_count

    page_range = range(0, page_num)
    
    images = []

    for item in page_range:
        images.append(render_page(doc[item]))

    return images

def convertToMD(page_num:int, checkboxes:str = None, radios:str = None):

    version = f"Processed using: {pymupdf.version=}, {pymupdf4llm.version=}"
    choice_table_strategy = radios
    choice_page_separators = False
    choice_embed_images = False

    if checkboxes is not None:
        for n in checkboxes:
            if n == "Separate pages":
                choice_page_separators = True
            if n == "Embed images":
                choice_embed_images = True

    if doc == None:
        raise gr.Error(message="Please upload a PDF")
    
    if page_num > doc.page_count:
        page_num = doc.page_count

    print(f"page num={page_num}")
    page_range = range(0, page_num)
    md = pymupdf4llm.to_markdown(doc, 
                                 pages = page_range)
    return md, md, version

def convertComplete():
    print("conversion complete")