topic Shell script question using sed or awk in Operating System - Linux

Shell script question using sed or awk

Unix or Linux? — Mon, 17 Apr 2006 08:24:05 GMT

How do i convert individual characters to upper case and also remove redundant characters at the same time. My data is the following

john.doe

How do i convert it to:

John Doe

Re: Shell script question using sed or awk

Ivan Ferreira — Mon, 17 Apr 2006 08:35:11 GMT

Try this:

echo john.doe |tr [:lower:] [:upper:] |tr . " "

Re: Shell script question using sed or awk

Unix or Linux? — Mon, 17 Apr 2006 08:43:17 GMT

I think that will convert everything to upper case whereas I need specific characters converted. I need the first character and the character after the full stop.

Re: Shell script question using sed or awk

James R. Ferguson — Mon, 17 Apr 2006 08:59:34 GMT

Hi:

# echo "john doe"|perl -ne '@a=split;foreach (@a) {push @b,ucfirst};print "@b\n"'

Regards!

...JRF...

Re: Shell script question using sed or awk

Unix or Linux? — Mon, 17 Apr 2006 09:05:18 GMT

Great thank you, it worked. Is there any way that this can be done with awk or sed?

Re: Shell script question using sed or awk

Ivan Ferreira — Mon, 17 Apr 2006 09:11:58 GMT

Try with this:

#!/usr/bin/awk -f
# convert upper case letters to lower case
BEGIN {
SEPARATOR_CHAR=".";
LC="abcdefghijklmnopqrstuvwxyz";
UC="ABCDEFGHIJKLMNOPQRSTUVWXYZ";
}
{
out="";
separator=0;
# look at each character
for(i=1;i<=length($0);i++) {
# get the character to be checked
char=substr($0,i,1);
j=index(LC,char);
# is it the first letter?
if (( i == 1 ) || ( separator == 1 )) {
# print (char);
out = out substr(UC,j,1);
separator = 0;
}
else if (char == SEPARATOR_CHAR ) {
char=" ";
out = out char
separator=1
} else {
out = out char;
}

}
printf("%s\n", out);
}

Re: Shell script question using sed or awk

James R. Ferguson — Mon, 17 Apr 2006 09:16:35 GMT

Hi (again):

With 'awk' the problem becomes more tedious since you would have to isolate the first character of each "word" (say with 'substr'); convert it to uppercase with 'toupper'; and then convert the remaining characters of that word to lowercase with 'tolower'. Perl offers far, far many more features than does 'awk'.

Regards!

...JRF...

Re: Shell script question using sed or awk

Unix or Linux? — Mon, 17 Apr 2006 09:22:15 GMT

ok but the problem is I dont understand the statement

echo "john doe"|perl -ne '@a=split;foreach (@a) {push @b,ucfirst};print "@b\n"'

Could you give an explanation of the statement.
Thank you

Re: Shell script question using sed or awk

James R. Ferguson — Mon, 17 Apr 2006 09:42:42 GMT

Hi (again):

# echo "john doe"|perl -ne '@a=split;foreach (@a) {push @b,ucfirst};print "@b\n"'

The perl switch '-n' says create a read loop. For every line read, split on whitespace (blanks, or tabs) the line's elements into a list called "a". Now, walk each element of list "a", and uppercase the first character of each element. As you do this, "push" (store) the result into a new list called "b". When every element in a line has been handled, print the modified line.

You could use this with a file, like:

# perl -ne '@a=split;foreach (@a) {push @b,ucfirst};print "@b\n"' yourfile

Regards!

...JRF...

Re: Shell script question using sed or awk

Unix or Linux? — Mon, 17 Apr 2006 09:56:55 GMT

Great but it doesnt seem to append to a file. All the other statments do append to a file. My code is as follows:

> testFile
printf "\n" >> testFile
echo "john.doe"|perl -ne '@a=split;foreach (@a) {push @b,ucfirst};print "@b\n\n"' >> testFile
ps -U john.doe >> testFile

Its only the echo line that is not appending to testFile. Any ideas why?

Re: Shell script question using sed or awk

Unix or Linux? — Mon, 17 Apr 2006 10:11:00 GMT

Ignore last message, I spelled the filename incorrectly :(

Re: Shell script question using sed or awk

Unix or Linux? — Mon, 17 Apr 2006 10:11:54 GMT

Answer found. Thank you.

Re: Shell script question using sed or awk

Rodney Hills — Mon, 17 Apr 2006 10:13:01 GMT

Your example seems ok. I ran it and testFile has the results expected...

The first line-

>testFile

Which redirects all standard output for all commands that follow to testFile can be dangerous if you try to run additional commands (like cat testFile), without redirecting standard output back to your tty.

HTH

-- Rod Hills

Re: Shell script question using sed or awk

James R. Ferguson — Mon, 17 Apr 2006 13:34:36 GMT

Hi (again):

While my original solution met your requirements, there is a better approach.

If you want to preserve whitespace (i.e keep multiple, embedded spaces and tab characters), you can use:

# perl -ple 's,(\w+),\u\L$1,g'

If you want to eliminate redundanat, embedded whitespace, use:

# perl -ple 's,(\w+),\u\L$1,g;s,\s+, ,g'

As before, the first character of each "word" will be uppercased and the trailing characters lowercased.

Regards!

...JRF...