no pages in doc after clean up

[hgdhot]

A pdf has 60+ pages, but after clean up with the code below, page_count is 0

def _clean_up(file_path):
doc = fitz.open(file_path)

try:
    tmp = BytesIO()
    tmp.write(doc.write(garbage=3, deflate=True))
    
    doc = fitz.Document('pdf', tmp.getvalue())
    tmp.close()
except Exception:
    ...

return doc

I guess there is something wrong with the pdf, how can I identify?

[JorjMcKie]

You actually executed some type of test in that you saved the PDF (to memory).
You could have looked into string pymupdf.TOOLS.mupdf_warnings() right after your first open - MuPDF will try automatic repairs if it encounters issues.
In your case, probably the PDF's internal structure is damaged that is called "page tree" - container of pointers to the single page objects.

I am slightly confused by the way you do that save to memory - I would have simply done pdfstream = doc.tobytes(...), and then tmp = pymupdf.open(stream=pdfstream).

You might as well loop through the single pages after original open to confirm that there is a page tree problem:

for page in doc:
    print(f"Successfully accessed {page.number=}")