Extracting drawing to a new pdf

[faisalakhlaq]

Is your feature request related to a problem? Please describe.
I have a PDF document containing different paths (drawings). Can I extract each existing path in a new PDF document without using Shape?
The extracted drawings should be exactly the same as in the original PDF. With all the layers (OCG), colors, embedded / attached images e.t.c. Therefore, it will be wise not to create a new Shape object (I tried to do that but it loses the colors).

Describe the solution you'd like
Something along the lines:

        paths = page.get_drawings()  # extract existing drawings
        for index, path in enumerate(paths):
            rect = path['rect']
            module_pdf = fitz.open()
            outpage = module_pdf.new_page(width=rect.width, height=rect.height)        
            outpage.show_pdf_page(rect, path, 0)
            module_pdf.save(f"{index}.pdf")

Describe alternatives you've considered
I tried looping over the paths and creating shapes as demonstrated in many examples but I get only the outlines of the paths.

shape = outpage.new_shape()

for index, path in enumerate(paths):
    if item[0] == "l":  # line
            shape.draw_line(item[1], item[2])
           ....... 
    shape.finish(
        fill=path["fill"],
        ........
    shape.commit()
            outpdf.save(f"{index}.pdf")

**Additional contex
Below is the screenshot of the pdf and when extracted the expected output is 2 separate files each with one drawing.

t**

[JorjMcKie]

Having trouble to understand what of the old PDF you don't want in your new one. Just the text?
While your request as stated doesn't sound to be ever doable, you might be better off just removing the text using redactions.

[faisalakhlaq]

Many thanks for a very quick response. Was not expecting that. Sorry if my question was not well explained.

Having trouble to understand what of the old PDF you don't want in your new one.Just the text?
There is no text, only drawings. The above image (screenshot of the pdf) has 2 drawings (If I am using the terminology correctly). I tried to extract each drawing in a new pdf. So result is 2 pdfs each containing one drawing.

        paths = page.get_drawings()  # extract existing drawings
        for index, path in enumerate(paths):
            rect = path['rect']
            module_pdf = fitz.open()
            outpage = module_pdf.new_page(width=rect.width, height=rect.height)        
            outpage.show_pdf_page(rect, path, 0)
            module_pdf.save(f"{index}.pdf")

In the above code snippet if outpage.show_pdf_page(rect, path, 0) can work or something else like this exists then it solves the problem. show_pdf_page expects a pdf but I just want to put the path on it as I don't have any PDF.

While your request as stated doesn't sound to be ever doable
Good to know then I should stop exploring any further options.

[JorjMcKie]

What you can do is restricting the part of the source page rectangle that should be visible in the target page: tar_page.show_pdf_page(rect, source_pdf, source_pagr_no, clip=source_rect).

This reduces your problem to determnining the source_rect - presumably the rectangle that covers one of your drawings.
But in any case, this will not reflect any OCG conditions that may exist in the source document.

[JorjMcKie]

Let me convert this issue to a "Discussions" item. Will be easier to handle there.

[JorjMcKie]

To output each drawing (vector graphic) on the pages of a source PDF to a new page of some target PDF, you can thus do the following:

src = fitz.open("source.pdf")
out = fitz.open()  # output PDF having one drawing per page
for src_page in src:
    paths = src_page.get_drawings()  # vector graphics on source
    for path in paths:
        rect = path["rect"]  # rect covering the drawing
        tar_page = out.new_page(width=rect.width, height=rect.height)
        tar_page.show_pdf_page(tar_page.rect, src, src_page.number, clip=rect)
out.cose("paths.pdf")