How to mass check the correctness of files after restoring them using R-Studio?

Question

Data recovery

After another Win7 glitch, my friend restored the system, respectively, deleting all the data. R-Studio came to the rescue, but it could only find files of known types. There are many such files. The question is how to massively check file headers for compliance with the extension and delete incorrect files?

Answer 1

it could only find files of known types

- I'm not sure exactly, but this most likely means that she found these files by their headers (using the same GREP, for example). Another thing is that many files may not open for various reasons:
1) the file is fragmented, and the search by titles finds only continuous files;
2) false positive - "header" formally satisfies the rule, but what was actually found is not a header for this file type;
3) the body of the files is overwritten with new data;
4) other reasons.
I advise you to deal with each type of files separately. For example, you can use some kind of mass converter on pictures - compress all jpegs into jpegs with a minimum compression level. It will digest good files, and choke on bad ones. Maybe the converter can give them more meaningful exif names.

Answer 2

This is far from a free solution, but I used PowerGREP for similar tasks (recovering file types by signatures). This is a very flexible and versatile tool, but requires good knowledge of regular expressions.

Answer 3

Any program for finding duplicates by the contents of files will help in this difficult task.
For each type of files it can be different, there are no universal ones.
For graphic files, there are many of them. For audio, Similarity is not bad. Since the last use of similarity, in my opinion, it has already become a more functional combine (not only eats audio) + there is hardware acceleration
UPD: 150415
Throw out R-studio, rescan in Active Undelete (in v10.43, an almost professional option appeared that allows you to jump over bads in the area +/-512 sectors)

Answer 4

Write your own script or program, it seems not very difficult to check the known headers of the most popular files.