Files
OCRmyPDF/tests
James R. Barlow dfbb4c9275 Add regression test for blank page with non-zero MediaBox origin (#1709)
A page whose MediaBox has a non-zero origin (e.g. from PDF Arranger crops)
was rendered blank by --force-ocr in v16.12.0, because fix_pagepdf_boxes
offset the CropBox with the wrong sign, pushing it entirely outside the
image-page MediaBox. The behavior was fixed in v17.0.0 but had no
end-to-end test: the existing box tests only assert that ocrmypdf runs,
which the blank-page bug passed (it exited 0 with a valid, empty PDF/A).

Add a test that renders the output and asserts the visible content
survives, parametrized over both the ghostscript and pypdfium rasterizers
since the bug reproduced regardless of rasterizer.
2026-06-22 14:57:52 -07:00
..
2026-01-13 01:50:57 -08:00
2023-04-14 02:19:19 -07:00
2026-01-13 01:50:57 -08:00
2025-12-21 12:21:47 -08:00
2025-11-10 16:57:30 -08:00