UNIX for Dummies Questions & Answers

I'm extremely confused by some of the comments I've seen in this thread.

First:

Code:

command1|command2

sets things up so the standard output of command 1 becomes the standard input for command2. This is exactly backwards from what you said above.

Then the question: What does the command:

Code:

cat f1 >> f2 | grep '^[A-Da-F]'

do?

First, the shell creates a pipe taking the output of cat f1 >> f2 and feeding it into the input of grep '^[A-Da-F]'. Then the shell sets up the two commands (cat and grep).

For the cat command, the shell sets the output from cat f1 to be appended to the end of the current contents of the file named f2. This causes the contents of the file f1 to be added to the end of the contents of the file f2. This output redirection causes the output side of the pipe from cat to grep set up above to be closed.

Assuming you're in the C or POSIX Locale, the grep command treats the argument '^[A-Da-F]' either as a request to have grep print any input line it reads if the first character on that line (indicated by the circumflex (^) at the start of the expression) is a collating symbol between A and D, inclusive, (i.e., A, B, C, or D) or to print a diagnostic message noting that the range expression a-F is invalid because F comes before a in the C and POSIX locales' collating sequences. In other locales the results are unspecified, but on many systems (if you don't get an invalid range expression error), you might also print lines where the first character on the line is one of the above with various diacritical markings (such as À, Á, or Ḉ).

Since the output from cat was redirected to f2 after the pipe was set up, the write end of the pipe will be closed and grep will hit end-of-file the first time it tries to read from the pipe. So, the grep command will never produce any output (other than the possible diagnostic message complaining about an invalid range expression).

I hope this helps.