Mercurial > hgrepos > Python2 > PyMuPDF

comparison tests/test_2548.py @ 1:1d09e1dec1d9 upstream

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
ADD: PyMuPDF v1.26.4: the original sdist. It does not yet contain MuPDF. This normally will be downloaded when building PyMuPDF.
author Franz Glasner <fzglas.hg@dom66.de>
date Mon, 15 Sep 2025 11:37:51 +0200
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 1:1d09e1dec1d9
1 import os
2
3 import pymupdf
4
5 root = os.path.abspath(f'{__file__}/../..')
6
7 def test_2548():
8 """Text extraction should fail because of PDF structure cycle.
9
10 Old MuPDF version did not detect the loop.
11 """
12 print(f'test_2548(): {pymupdf.mupdf_version_tuple=}')
13 pymupdf.TOOLS.mupdf_warnings(reset=True)
14 doc = pymupdf.open(f'{root}/tests/resources/test_2548.pdf')
15 e = False
16 for page in doc:
17 try:
18 _ = page.get_text()
19 except Exception as ee:
20 print(f'test_2548: {ee=}')
21 if hasattr(pymupdf, 'mupdf'):
22 # Rebased.
23 expected = "RuntimeError('code=2: cycle in structure tree')"
24 else:
25 # Classic.
26 expected = "RuntimeError('cycle in structure tree')"
27 assert repr(ee) == expected, f'Expected {expected=} but got {repr(ee)=}.'
28 e = True
29 wt = pymupdf.TOOLS.mupdf_warnings()
30 print(f'test_2548(): {wt=}')
31
32 # This checks that PyMuPDF 1.23.7 fixes this bug, and also that earlier
33 # versions with updated MuPDF also fix the bug.
34 rebased = hasattr(pymupdf, 'mupdf')
35 if pymupdf.mupdf_version_tuple >= (1, 27):
36 expected = 'format error: No common ancestor in structure tree\nstructure tree broken, assume tree is missing'
37 expected = '\n'.join([expected] * 5)
38 else:
39 expected = 'format error: cycle in structure tree\nstructure tree broken, assume tree is missing'
40 if rebased:
41 assert wt == expected, f'expected:\n {expected!r}\nwt:\n {wt!r}\n'
42 assert not e