Question: about originalPdfDocument.get_page_images(pageNumber, full=True), can't get full images list.

[AwesomeYuer]

Please provide all mandatory information!

Describe the bug (mandatory)

I want get full image list with parameter full=true , even though some images are same in pdf.
I can't get all image items from below code.
There are 5 images in the pdf, 4 of 5 are same,
So only top 2 unique image item are output.

Would you help to resolve it?

Thanks in advance.

To Reproduce (mandatory)

1. the Sample problem PDF document as below

https://github.com/AwesomeYuer/public-misc-share/blob/main/Simple.Rowspan.for.Pdf.Print.pdf

the pdf document is print by one html as below, all the images are inline base64 html format

https://github.com/AwesomeYuer/public-misc-share/blob/main/Simple.Rowspan.for.Pdf.Print.html

2.run the code in ipynb in vscode

pdfDocument = fitz.open(path)
page = pdfDocument[0] # Page 2
page.wrap_contents()
images = pdfDocument.get_page_images(0, full=True)
for i, item in enumerate(images):
    print(i, item)

Expected behavior (optional)

Describe what you expected to happen (if not obvious).

I want get full image list with parameter full=true , even though some images are same in pdf.

The full output list should be 5 items

Screenshots (optional)

If applicable, add screenshots to help explain your problem.

Your configuration (mandatory)

• Operating system, potentially version and bitness
• Python version, bitness
• PyMuPDF version, installation method (wheel or generated from source).

For example, the output of print(sys.version, "\n", sys.platform, "\n", fitz.__doc__) would be sufficient (for the first two bullets).

3.11.5 (main, Sep 11 2023, 13:54:46) [GCC 11.2.0] 
 linux 
 
PyMuPDF 1.23.5: Python bindings for the MuPDF 1.23.4 library.
Version date: 2023-10-11 00:00:01.
Built for Python 3.11 on linux (64-bit).

Additional context (optional)

Add any other context about the problem here.

[JorjMcKie]

This is a typical "Discussions" post - transferring to the respective tab.

[JorjMcKie]

Method page.get_images() only shows which image references are contained in the page's PDF object definition. This may or may not be the same set of images that the page actually displays. Whether you use the full parameter or not: this has a completely different meaning.
Please consult the documentation for more background.

To get a list of all image display commands that the page actually issues, use page.get_image_info() (please consult documentation for parameter details). This also contains duplicate images and images that may not have an xref (inline images).
Also, the sequence in that list is the sequence in which the display commands are being executed.

[AwesomeYuer]

@JorjMcKie thank you very much.

@AwesomeYuer

Would you help to modify the below code to get xref by using page.get_image_info?

pdfDocument = fitz.open(path)
page = pdfDocument[0]
page.wrap_contents()
images = pdfDocument.get_page_images(0, full=True)

for i, item in enumerate(images):
    print(i, item, type(item))
    old_xref = item[0]


print("===================================")


imagesInfo = page.get_image_info()

rects_list = []

for i, item in enumerate(imagesInfo):
    print(i, item, type(item))
    
    bb = item["bbox"]
    text_rect = fitz.Rect(bb[0], bb[1], bb[2], bb[3])
    rects_list.append(text_rect)
    
    # look here
    # Would you help to get the 'xref' of image from `image dict`?
    # ??????????
    old_xref = item ??????????????????
    # thanks in advance

[AwesomeYuer]

I know set parameter xrefs=True

[JorjMcKie]

Ok - so you are good?

Look, instead of doing this:

    bb = item["bbox"]
    text_rect = fitz.Rect(bb[0], bb[1], bb[2], bb[3])

... simply do this:

    text_rect = fitz.Rect(item["bbox"])

fitz.Rect understands any numeric sequence of length 4 and can directly make a rectangle from it.

[AwesomeYuer]

@JorjMcKie

I am good!

Big thanks for your answer and suggestion in time!