Remove tab, new line, carriage return until encountered char

Question

I have a CSV file that has record separator of ,\n. One of the columns, has the possibility of containing tabs and new lines, which I wish to remove. How do I remove these chars (\t\n\r) using sed and/or awk, until encountering the record separator?

Answer 1

Here you go:

cat input_file | tr -d '\r\t' | sed ': loop_start; /,$/b; N; s/\n//; b loop_start'

Explanation:

1. tr is used to delete (-d) all \r and \t chars.
2. Then sed loops (: loop_start; ...; b loop_start) and
3. joins all lines (N) and
4. removes the newline (s/\n//)
5. until it finds the record separator (/,$/) and
6. branches to the end of the script (b) where it will start the script over with the next line of input.