IF YOU WANT TO SKIP THE STORY AND GO TO THE QUESTION FAST LOOK BELOW:
Hi there. As most of you know, ext3 doesn't have a undelete, and if u rm -r at the root of a filesystem, well... everything is gone.
I did a rm -r /mnt instead of a rm -r ./mnt and lost all data, most of the backups too, (i backed up to other hard disks that are normally only mounted during backups are running, but I was testing backup scripts, so....)
ext3 deletes all the nodes for good, so no easy undelete for ext3, so I unmounted all partitions, and took the drive to professionals and see what can be done.
All the data is there, but no info about file names,dirs, etc... so they are going to do a complicated header scan through the whole disk (about 60 GB of data) and at the end I am going to have a long list of files named generically, 1.jpg, 2.jpg, 1.doc, 2.doc... the header anaylsis will hopefully find out what file type each file is, so at least i will know if its a doc or a jpg, and with some they will be able to put a guessed name on it.
I havent been able to assess the damage, but i think around 90% of all data is backed up. but those 10% are very important, so i am going to have to go through all the 10.000 jpg, 12000 tifs, and 6000 dwg files...
If i have to do in manually, i might as well jump of the bridge right away..
FOR THOSE WHO WANT TO GET TO THE QUESTION FAST:
So this is what i am looking for: a tool or script that will compare the files binary. I could imagine a script with cat, grep, diff, etc.. could probably do the trick, but i dont know how to script. I would have a long list of, for example, word documents, named somehow, and I would want the script to compare each one with every word document in a set of directories and its subdirectories. So I can forget about the ones that match, so I can concentrate on the ones that I know I don't have.
Does anybody have an idea if there are tools that already do this?
Or an easy script, or a pointer? All ideas and suggestions are welcome.
Thanks!