Re: filtering of text files
When I tried this, it mapped all punctuation to 'Z'. It seems
like the second argument needs something like [*] tacked
on (enough nothings to make arg2 as big as arg1). This
didn't work, and I couldn't find a "null" special character in
the tr man pages.
Might I suggest:
cat filename | tr '[:lower:][:punct:]' '[:upper:][a*]' | sed 's/a//g' > filename2
Since the lower-to-upper translation has been performed, 'a' (or any other
lower-case letter) will safely match all punctuation (i.e., sed won't be
deleting any a's that are really a's and not punctuation marks).
Marc Mongeon
>>> Ian Peters <itp@gnu.org> 05/06 7:22 AM >>>
On Thu, May 06, 1999 at 12:12:37PM +0000, Richard Harran wrote:
cat filename | tr '[:lower:][:punct:]' '[:upper:]' > filename2
Reply to: