PDFViewer instances must have canvas iterators

[maxpmaxp]

it would be more convenient to iterate pages like below:

for page_canvas in viewer:
    my_content = extract_my_content(page_canvas)

Iterator walks to the next page and renders it.