2

There is no file in my directory called a. I used the command "rm a 2>a" to redirect the standard output error to file a. But when I tried the "ls" command, there is no file called a. Can someone explain this?

Improve this question
3
  • 2
    voting to move to superuser
    – Ben Voigt
    Aug 2, 2010 at 17:07
  • Well ... what happens when you do not redirect? Is anything printed to stdout?
    – Fantomas
    Aug 2, 2010 at 17:11
  • 2
    Is this merely an academic question? What's your use case for using the same file name both for deletion and capturing the output?
    – Doug Harris
    Aug 2, 2010 at 18:09

2 Answers 2

Reset to default
10

2>a causes the shell to redirect the standard error stream to a file called a. This file is created before the command is started. rm is then executed. It finds a file called a (already created by the shell) and unlinks it.

Improve this answer
1
  • 4
    Right... the file still exists, so that anything rm writes to stderr after calling unlink can go into it, but it no longer has its name. When the process quits, the file finally goes away.
    – Ben Voigt
    Aug 2, 2010 at 17:17
3

a is created bbefore the rm command executes....its as simple as that!

Improve this answer
0

You must log in to answer this question.

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