UNIX for Dummies Questions & Answers

A very simple question but I have scoured the web and can't find an answer. How do I search for a character by ASCII code in a regular expression using grep?

For example, we use the End of Medium symbol as a delimiter in certain files. (this is ascii 031 in oct, displays as ^Y) I want to grep for CAT surrounded by this symbol. (displayed it looks like ^YCAT^Y) I've tried everything and can't find a way that works.

Thanks.

To enter a control character, type Ctrl-V followed by the control character you need (Ctrl-Y in this case):

Code:

$ grep '^YCAT^Y' testfile
CAT
$

You can't see the control characters in the output, but they are there (proving this by replacing them with a "X"):

Code:

$ grep '^YCAT^Y' testfile
CAT
$ grep '^YCAT^Y' testfile|sed 's/^Y/X/g'
XCATX
$

Or use bash/ksh/zsh string literals:

Code:

grep $'\C-YCAT\C-Y'

FYI
Wikipedia has a good ASCII chart that shows octal, decimal, hex and control character representations (scroll down a little):
https://secure.wikimedia.org/wikipedia/en/wiki/ASCII

---------- Post updated at 12:13 PM ---------- Previous update was at 12:11 PM ----------

I'm afraid that produced no output on our Solaris box using ksh. Am I missing something?

Try...

Code:

grep $(printf "\031CAT\031") file1

Code:

I'm afraid that produced no output on our Solaris box using ksh. Am I missing something?

Too many ksh-es. It works fine with ksh93t+ in my linux box.

Alternatively, you can "grep" with AWK, whose regular expression flavor supports octal escape sequences:

Code:

awk '/\031CAT\031/' file

Regards,
Alister