Shell Programming and Scripting

I have a list of names and email addresses, like this. The <tab> markers are actually tabs.

joe.blow <tab> joe.blow@wherever.com
tom.t.hall <tab> tom.t.hall@wherever.com
john.r.smith <tab> john.r.smith@wherever.com
sally.jones <tab> sally.jones@state.or.us

I want to parse the data so that it will result in this:

joe <tab> blow <tab> joe.blow@wherever.com
tom <tab> hall<tab> tom.t.hall@wherever.com
john <tab> smith <tab> john.r.smith@wherever.com
sally <tab> jones <tab> sally.jones@state.or.us

So, all I need to do is remove the dot, if present, in the first part of each line. The second part (the email address) goes untouched.

So, I can easily use sed to go through and replace a 'character dot character', since sed looks for the first occurrence on each line, like this:

sed 's/\.[a-z]\./ /' emailfile.txt

That results in this:

joe.blow <tab> joe.blow@wherever.com
tom <tab> hall <tab> tom.t.hall@wherever.com
john <tab> smith <tab> john.r.smith@wherever.com
sally.jones <tab> sally.jones@state.or.us

If I run that sed command again, it will modify the email part (character.dot.character).

How can I get the result I need? Is sed the best tool for this?

awk?

Code:

awk -F'\t' '{ 
    split( $1,arr, ".")
    for(i=1; arr[i]>""; i++) {printf("%s\t",arr[i])}
    print $2
             } ' file

try this

Code:

sed 's/\.[[:alpha:]]\.\(.*\)$/ \1#/;/#/!s/\.\(.*\)$/ \1/;s/#$//' emailfile.txt

this is the right syntax for awk:

Code:

awk -F'\t' '{
   split( $1,arr, ".")
   if ( 3 in arr ) 
      { printf("%s %s\t",arr[1],arr[3]) } 
   else 
      { printf("%s %s\t",arr[1],arr[2]) }  
   print $2
            } ' emailfile.txt

since the given output is exactly as requested