Fixing pymupdf.mupdf.FzErrorFormat crash by recasting as an `Imposs…

…ibleParsingError` (#474)

paperqa/readers.py

@@ -21,7 +21,14 @@ def parse_pdf_to_pages(path: Path) -> ParsedText:
         total_length = 0
 
         for i in range(file.page_count):
-            page = file.load_page(i)
+            try:
+                page = file.load_page(i)
+            except pymupdf.mupdf.FzErrorFormat as exc:
+                raise ImpossibleParsingError(
+                    f"Page loading via {pymupdf.__name__} failed on page {i} of"
+                    f" {file.page_count} for the PDF at path {path}, likely this PDF"
+                    " file is corrupt"
+                ) from exc
             pages[str(i + 1)] = page.get_text("text", sort=True)
             total_length += len(pages[str(i + 1)])