问题
I am trying to test my Python program, which takes in either .zip or .Z files and decompresses them using Python's zipfile module or Unix's gzip, respectively. It makes sure that the filetype is either .zip or .Z (in the latter case, using Unix's file command) before trying to do anything. I wanted to test my error handling in the very rare case in which a verified archive file errors out while being decompressed. So basically, I want to feed it a corrupt .Z file.
Someone suggested that I could use Unix's dd command to just mess up a good .Z file and use that as my bad input. I couldn't find any examples for using dd for this use case and was hoping someone could provide a simple example. I know I shouldn't mess with the header, since that's where the metadata is that tells us it is a .Z file. So I know I need to mess up some of the middle and some of the end... Thanks for any help.
回答1:
You could just have used a hex editor like hexedit.
Since you ask though,
dd if=/dev/urandom of=yourfile.z bs=1024 seek=$((RANDOM%10)) count=1 conv=notrunc
will rewrite garbage to a random 1024b block out of the first ten in a file.
来源：https://stackoverflow.com/questions/24130538/testing-intentionally-corrupt-a-z-file-using-dd