Shell Programming and Scripting

Hi All,

I am facing some problems with bad characters in my file.For example-

Code:

00000000509 TCI DEVOFFERS= 1

Now I want to remove all bad characters and replace with *.
Please suggest some solution.

Along with that How to mention a range of ASCII values in TR command to replace?

Say I want to replace ASCII value 32 -38(Dec) with Asterisk(*), how to perform with TR command in UNIX.


Please help.

Moderator's Comments:

Use code tags, thanks.

try sed

Code:

sed 's/[ #!"$%&]//g'  inputfile > newfile

The [ ] thing is called a "character class" in regular expressions

The behaviour of the "tr" command is Operating System dependent.

What Operating System and version you you have? What Shell are you using?



Btw. The equals sign in your example not in the ASCII range you define.
When working with "tr" to remove "bad" characters it is much easier if you can define the exact range of "good" characters.