We have a customer who is receiving a bunch of bad PDF files. When these PDF files are opened with Acrobat, they are displayed fine, but when closing the file, you are prompted to save the file ... which means that the error was recovered. These PDF files process OK, but the resulting output file has some content missing. When I process the same file again, but this time process it with the PDFTron Optimizer enabled with default settings, the resulting output PDF looks good.
We will take a look at these files, but in the meantime you can detect corrupt files that have been fixed using pdftron.PDF.PDFDoc.IsModified(). If IsModified() returns true, it means that there were some corruptions that caused changes in the document. You can then flag this type of files for review etc.