PDF tobytes Changing MD5 Hash

[TheCapybaraClub]

I would expect the hash of bytes resulting from tobytes would be static, but its not. Can you please explain why this happens and offer a solution for generating a static hash?

no_rot.pdf

import fitz
import hashlib

for i in range(5):
    pdfDoc = fitz.open('./no_rot.pdf')
    #file_bytes = pdfDoc.tobytes(garbage=4, deflate=True)
    file_bytes = pdfDoc.tobytes()
    ahash = hashlib.md5(file_bytes).hexdigest()
    print(f"Rnd: {i+1} Hash: {ahash}")

output:

Rnd: 1 Hash: 0604847b5bdcbf431de55ba8603115b9
Rnd: 2 Hash: ec03a4579676137c1b26b95babb2f105
Rnd: 3 Hash: dd243b2f5582bd8d89c08181e3d900a6
Rnd: 4 Hash: af71177f85bd20461aa77febee7f60c0
Rnd: 5 Hash: 43fb6b3dc2f2762983cb3830f0f13421

[JorjMcKie]

This is because the /ID key in the PDF trailer is recomputed on every save / tobytes based on a random generator.
Try to suppress this with the appropriate save / tobytes option.

[TheCapybaraClub]

Solution summary mentioned by @JorjMcKie
Set no_new_id=True in the tobytes method

import fitz
import hashlib

for i in range(5):
    pdfDoc = fitz.open('./no_rot.pdf')
    file_bytes = pdfDoc.tobytes(no_new_id=True)
    ahash = hashlib.md5(file_bytes).hexdigest()
    print(f"Rnd: {i+1} Hash: {ahash}")

output as expected:

Rnd: 1 Hash: 859aa17c704cdfa7d6e66f1f46e2be48
Rnd: 2 Hash: 859aa17c704cdfa7d6e66f1f46e2be48
Rnd: 3 Hash: 859aa17c704cdfa7d6e66f1f46e2be48
Rnd: 4 Hash: 859aa17c704cdfa7d6e66f1f46e2be48
Rnd: 5 Hash: 859aa17c704cdfa7d6e66f1f46e2be48