Skip to content

Commit 3706c1e

Browse files
restore safeguard
1 parent 42b83d2 commit 3706c1e

1 file changed

Lines changed: 9 additions & 2 deletions

File tree

mindee/input/local_input_source.py

Lines changed: 9 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -52,9 +52,16 @@ def __init__(self) -> None:
5252
self._check_mimetype()
5353
if self.is_pdf():
5454
self.file_object.seek(0)
55+
# Some broken (yet fixable) PDFs can cause pdfium to crash on open.
5556
if PYPDFIUM2_AVAILABLE:
56-
pdf = pdfium.PdfDocument(self.file_object)
57-
self.page_count = len(pdf)
57+
try:
58+
pdf = pdfium.PdfDocument(self.file_object)
59+
self.page_count = len(pdf)
60+
except pdfium.PdfiumError as e:
61+
logger.warning(
62+
"Could not open PDF file: %s due to %s", self.filename, e
63+
)
64+
self.page_count = 0
5865
self.file_object.seek(0)
5966
else:
6067
self.page_count = 0

0 commit comments

Comments
 (0)