Shell Programming and Scripting

I have a textfile which I am parsing with awk. The lines do not have the same number of fields, so sometimes $3 is the last field, sometimes not.

When I do a 'printf("%s, %s\n", $2, $3)', if $3 is the last field in the line, when I cat the file the output looks something like this:
,someemail@someprovider.com

If I 'vi' the file it looks like this:
NAME^M^M,someemail@someprovider.com

Oh, and the textfile is a DOS file ported over to Unix. How can I get rid of those two newlines where $3 is the last field?

^M is not a newline, it's a carriage return, also known as \r in some places. DOS text files end lines in \r\n, UNIX ones only end in \n, which causes occasional confusion since the \r may be taken as part of the literal text instead of being part of the line separator.

By itself, the terminal takes \r to mean 'send the cursor to the beginning of the same line'. So when you cat it, it prints NAME to the terminal, goes back to the beginning, and overwrites it with ,someemail@someprovider.com, then goes to the next line.

Try deleting all carriage returns for each line by doing gsub(/\r/, ""); before you do any checks or processing.

Or run dos2unix - DOS/MAC to UNIX text file format converter on your file to get rid of CR (carriage return):

Code:

dos2unix file

Most systems don't have dos2unix.

Any system will have tr.

Code:

tr -d '\r' < dostext > unixtext