Step 1 is to dd the whole USB drive to a backup image. Then from a copy of that backup image you can start trying recovery from that.
On Sun, Jan 12, 2020 at 1:32 PM Keith Lofstrom <[email protected]> wrote: > The three book scanners near the checkout desk at the > PSU Millar library are somewhat difficult to use, but > better than my slow USB flatbed scanner at home. > > Yesterday I scanned three huge multipage files to a > Brand X "Cheap on Amazon" 2GB USB flash drive. I was > in a hurry, so I did not segment the files into smaller > chunks, or check the files with my laptop as I made them. > > Bad idea. > > I now have a flash drive which is 60% full, but no files > are listed in the directory. Either the files were too > large for the scanner, or the 40 character file names were. > The flash drive is formatted for VFAT16 DOS or somesuch. > > I hope to recover the files (if not the file names) and > avoid another 90 minute, 200+ page scanning marathon. > I can grep the drive image for strings; I don't see the > filenames, but grep shows about 100 strings like > /ProcSet [ /PDF /ImageB ] -or-f /ProcSet [ /PDF /ImageC ] > and some fragments. I tried using "testdisk" tools to > recover the three files; no joy. > > My best guess is that my overly-long scan files blew > the memory buffer on the PSU scanner, and it overwrote > garbage. There are signs above the scanners to "save > frequently" which I ignored (https://www.xkcd.com/293/) > > PERHAPS SOMEONE CAN SUGGEST CLEVER TOOLS to extract the > pdfs from the 1.2 gigabytes of "unlabeled something" on > the flash drive. Knowing how might help me help others > in the future. > > I expect I will only get my files by scanning them > again, properly, in small chunks PSU's feeble scanners > can handle. Meanwhile, "don't do that" is probably the > most help I can offer to others. > > Keith > > P.S. - the best book scanner I've used was at MIT Barker > Library; it images the book open 120 degrees, face up, > and accomodates the natural curve of the pages. The book > scanners in the Library of Congress are almost as good, > but you must stretch the pages flat to get focused images. > > -- > Keith Lofstrom [email protected] > _______________________________________________ > PLUG mailing list > [email protected] > http://lists.pdxlinux.org/mailman/listinfo/plug > _______________________________________________ PLUG mailing list [email protected] http://lists.pdxlinux.org/mailman/listinfo/plug
