Shell Programming and Scripting

Yes, sorry about the missing quotes around $line. I posted the suggestion "on the run" and didn't have time to try it out.

Here is a quick explanation of what is happening.

$0 = Current input line
substr(string, start, len) extract a substring starting at character position start for a length of len if len is not supplied it goes to end of string.

substr($0,23,4)+0==F
This get chars 23-26 of the input line
Adds zero (this converts the value to an integer which is needed as " 21" <> 21)
Compare this value with variable F if it's a match the code in {} is processed


{$0=substr($0,1,22) sprintf("%4d",T) substr($0,27)}
This rebuilds the input line with chars 1-22 + T (padded to 4 chars wide) + chars 27-end

1
This value is not zero in awk that is a true (not false) expression so the code in {} that follows is executed. In this case no code follows so the default action is performed which is "print the input line". Note that we may have changed the input line in the statement above (or we might not have all depending on the contents of T chars 23-26 of the input line)

This could also be expanded to:

Code:

awk -vF=$id -vT=$newid '
{ 
  id=substr($0,23,4)
  sub(/^ */,"",id)
  if(id == F) print substr($0,1,22) sprintf("%4d",T) substr($0,27)
  else print $0                                                   
}'

Because your using bash you could also use bash substrings (remember index starts at zero not 1), this solution will be much faster as it doesn't need to load and execute awk for each line processed:

Code:

function write_pdb() {
    ln=${line:22:4}
    ln=${ln##* }
    [ "$ln" = "$id" ] && line="${line:0:22}$(printf "%4d" $newid)${line:26}"
    echo "$line" >> $1
}

you are awesome man! thank you!