I have a file (call it A, for reference) that may be a fragment of some other extant file on my system. I can't use cmp because I don't know how many bytes may be missing from the start of A (or, at least, I can't use it without brute forcing through the -i flag). Is there a way for me to discover whether A is already existent on my system (using GNU tools, or any other linux program)? Or will I have to botch together a c++ program to do the job? Note: efficiency is desirable since the files that A has to be compared with may be numerous.

1 Answer 1


Not knowing how big the chunk (A) is ... have you considered grep'ing for its content using -a?

  • echo -e "\x00\x11\x22\x33" > file; echo -e "\x00\x11" > snip; grep -a snip file doesn't seem to work
    – daisy
    Oct 15, 2012 at 23:59
  • No, I hadn't considered grep! That'll do.
    – Zorawar
    Oct 16, 2012 at 0:09
  • @warl0ck: grep needs a pattern, so try grep -a "$(cat snip)" file. But, shouldn't there be a problem with echo's newlines? Your example should fail regardless of this correction, shouldn't it? On my system it succeeds...
    – Zorawar
    Oct 16, 2012 at 0:11
  • In fact, why use grep -a at all? Can't I just use grep? Taking warl0ck's example, echo -e "\x00\x11\x22\x33" > file; echo -e "\x00\x11Surplus Gibberish" > snip; grep -a "$(cat snip)" file succeeds. Is that expected behaviour? If I remove the -a flag, then the command fails, as I'd want it to.
    – Zorawar
    Oct 16, 2012 at 0:25
  • I'd be surprised if that worked. Neither of snip or file is wholly included in the other in your example; I wouldn't expect a match, no matter what method of invocation.
    – tink
    Oct 16, 2012 at 0:38

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .