Text insertion fails due to missing /Resources object

[C-Saunders]

Description of the bug

Per the information here and related issue here, inserting text should work without first calling clean_contents() on the page. I've recently run into a document where it did not work.

The error raised is pymupdf.mupdf.FzErrorArgument: code=4: not a dict (null) for both TextWriter.write_text and Page.insert_text

We've had some issues with clean_contents() in the past, and it seems best to avoid it if we can, but we can add it back into our process if needed (potentially we'd do it conditionally if we hit this error).

How to reproduce the bug

error_example.pdf

import pymupdf

doc = pymupdf.open("./error_example.pdf")
page = doc[0]

try:
  page.insert_text((100, 100), "hey there", fontsize=20)
  print("no clean - no error")
except pymupdf.mupdf.FzErrorArgument:
  print("no clean - caught pymupdf.mupdf.FzErrorArgument")

page.clean_contents()

try:
  page.insert_text((100, 100), "hey there", fontsize=20)
  print("cleaned - no error")
except pymupdf.mupdf.FzErrorArgument:
  print("cleaned - caught pymupdf.mupdf.FzErrorArgument")

PyMuPDF version

1.25.3

Operating system

MacOS

Python version

3.11

[JorjMcKie]

This was an easy fix of 2 lines. We just need to check whether the page has a Resources object and give it a new empty one if not.

[JorjMcKie]

Here is your test script, modified in the same way as that fix also works.
Try it out if you want!

import pymupdf

doc = pymupdf.open("error_example.pdf")
page = doc[0]

# following lines represent the fix underway
pdfpage = pymupdf._as_pdf_page(page)
resources = pymupdf.mupdf.pdf_dict_get(pdfpage.obj(), pymupdf.PDF_NAME("Resources"))
if not resources.pdf_is_dict():
    pymupdf.mupdf.pdf_dict_put_dict(pdfpage.obj(), pymupdf.PDF_NAME("Resources"), 5)
# end of fix -------------------------------------
try:
    page.insert_text((100, 100), "hey there", fontsize=20)
    print("no clean - no error")
except pymupdf.mupdf.FzErrorArgument:
    print("no clean - caught pymupdf.mupdf.FzErrorArgument")

page.clean_contents()

try:
    page.insert_text((100, 100), "hey there", fontsize=20)
    print("cleaned - no error")
except pymupdf.mupdf.FzErrorArgument:
    print("cleaned - caught pymupdf.mupdf.FzErrorArgument")

[julian-smith-artifex-com]

Fixed in PyMuPDF-1.25.5.