I'm using hashdeep to calculate checksums, recursively for files. However, I need to exclude certain directories and so the solution would seem to be to use find to feed a file list to hashdeep. This is mostly working well.
hashdeep -c md5 -f <(find . -type f -and -not -path '\./excludeme*')
For testing, this can be simplified to:
hashdeep -c md5 -f <(find . -type f)
The trouble is that I've started seeing files that have been named in an unfortunate way. e.g. they have a carriage return in the file name. I didn't put it there but it's not my place to go and fix it.
We can manufacture a file like that with:
touch file$'\r'
hashdeep on its own can cope with this:
$ hashdeep -c md5 -r .
%%%% HASHDEEP-1.0
%%%% size,md5,filename
## Invoked from: /home/myusername/hashdeeptest
## $ hashdeep -c md5 -r .
##
0,d41d8cd98f00b204e9800998ecf8427e,/home/myusername/hashdeeptest/file
If I use the invocation with find from above though, I get:
./file: No such file or directory
I suspect there is some way to get find to feed the file name to hashdeep in a way that it finds palatable, but at the moment that eludes me.
Many thanks for a solution!
