Shell Programming and Scripting

I been trying to get this right. I have trying to get rid of spaces in between the character < and the character >. Everytime I try, sed gets too greedy and do the whole line.

Ex.

< T AG 1> Hello, how are you doing?
<Tag 2> I am doing fine.

I want this:

<TAG1> Hello, how are you doing?
<Tag2> I am doing fine.


What I keep on getting is this:

<TAG1>Hello,howareyoudoing?
<Tag2>Iamdoingfine.


Does anyone has a way to only do part of a line when I find that pattern using sed?

This does not seem to work. Keeps giving me illegal statement error.

Use nawk or /usr/xpg4/bin/awk on Solaris.

Regards

This seems to partially work. It works fine for lines that have a tag <Some tag> but it causes unintended side effect. It also does lines that don't have tag in it.
Ex.
I am fine. This is line 200
It becomes:
Iamfine.Thisisline200

Is there any way to get around this? Let me expand the example a little further.
Ex.
< T AG 1> Hello, how are you doing?
<Tag 2> I am doing fine.
This is a test < T A G 3> I
am doing this to test how to
get rid of spaces on part of a
line.


I just want to find a way to only get rid of spaces between the < and the > and nothing else on that line and nothing else on lines that don't have < and >.

awk -F'<|>' '{gsub(/ /,"",$2);print $1 "<" $2 ">" $3}' file | sed 's/<>//'

Using bash or ksh93:

Code:

while IFS= read -r line
do
  while :
  do
    case $line in
      *"<"*">"*)
           left=${line%%<*}
           line=${line#"$left"}
           tag=${line%%>*}
           line=${line##*"$tag"}
           printf "%s" "$left" "${tag// /}"
           ;;
      *) break;;
    esac
  done
  printf "%s\n" "$line"
done