fix: add mupdf exception for pdf extraction

engisalor · Jun 18, 2024 · 5f51358 · 5f51358
1 parent 2a5393f
commit 5f51358
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/corpusama/source/pdf.py b/corpusama/source/pdf.py
@@ -100,7 +100,7 @@ def _try_extract(file: str, clean: bool, n: int = 0) -> None:
         text = extract_text(file, clean)
         with open(file.with_suffix(".txt"), "w") as f:
             f.write(text)
-    except (fitz.fitz.FileDataError, RuntimeError) as e:
+    except (fitz.fitz.FileDataError, RuntimeError, fitz.mupdf.FzErrorFormat) as e:
         logger.warning(f"{n} - {file} - {e}")