scanning printed document

June 1, 2007 7:41 AM

Help! I printed out a long document last night on Microsoft Word (using ibook G4). When I looked at the file this morning, it only has two characters in it. I fear that I may have 'selected all' and then accidentally deleted it, then saved the document. But I have a clear double-spaced hard copy of the document. Is there any way to somehow scan the document and enter the text automatically, rather than retype the many, many words? (or some other way to retrieve the lost text?)
posted by dearleader to computers & internet (10 comments total)

With OCR scanning software like Omnipage you should be able to do it. But, OCR scanning is notoriously unreliable, or at least was the last time I used it.
posted by khaibit at 7:52 AM on June 1, 2007

If you have a scanner, text scanning with optical character recognition should be one of its basic functions, yes. If you don't have one, places like Kinko's should be able to help.

Check and see if there's any chance you have a previous .tmp version or revision tracking turned on in your word document, though. It might have a saved version of your text for you somewhere.
posted by jacquilynne at 7:53 AM on June 1, 2007

jacquilynne... thanks, great suggestion... but, uh, anyone know where to look for something like that?
posted by dearleader at 7:54 AM on June 1, 2007

If you remember what the file was named, you can search your hard drive for the file name. It will call up the version with the 2 characters in it, but also any temp versions. The temp version file name may be truncated at the beginning or end with a ~, so check for partial file names too.
posted by misanthropicsarah at 7:56 AM on June 1, 2007

One of the easiest ways to find temp files is to search by date - if you get it to list all files modified in the last 24 hours you can cut down your search space considerably.
posted by handee at 8:07 AM on June 1, 2007

Most scanners now come with OCR. Check your scanner.
posted by chuckdarwin at 8:34 AM on June 1, 2007

I should warn you that Kinko's will likely quote you an outrageous amount of money and give you a 24 hour turnaround. (Or at least I always used to)
posted by khaibit at 8:38 AM on June 1, 2007

FYI temp files may be invisible. So when you do a find select that option.
posted by Gungho at 10:07 AM on June 1, 2007

OCR scanning is pretty good these days. It's the sort of thing a lot of businesses have to do. I believe there are industrial copiers that can do it really quickly. You would still need to go through the text and check for errors, but it wouldn't be too bad.

Kinko's will be expensive. A lot of colleges and universities have copy centers on campus that could probably do the job for you and at a much lower price.
posted by PercussivePaul at 11:46 AM on June 1, 2007

Google has an open-source OCR program, which they "believe" would work with Apple's OSX. I've never used it, although ReadIris OCR worked reasonably well for me when I scanned typed, double-spaced pages into it.
posted by needs more cowbell at 3:57 PM on June 1, 2007