[SOLVED] Unexpected double printing of awk when printing unmatching pattern

Issue

I have data organized as lines (no columns). Lines altern between ">name" and "data", such that:

>name1
textA
>name2
textB
>name3
textC

I want to remove lines with a given name and the associated data - e.g., remove the data for >name3, meaning that both the line >name3 and the textC line should be removed.

I am using:

awk 'BEGIN {RS = ">"; ORS = ""} !/name3/ {print">"; print $0}' FILE

However, the output is as following:

>>name1
textA
>name2
textB

I have tried several alternatives but I did not manage to get the first line right (e.g., either the ">" is doubled or completely missing).

Solution

The record separator is >.
The first character of the file is >.
Thus, the first record of the file is the empty string before the first record separator.

You want:

awk '
  BEGIN {RS = ">"; ORS = ""}
  length && $1 != "name3" {print RS $0}
' file

>name1
textA
>name2
textB

---

An alternate way to solve the problem:

paste - - < file | grep -v '>name3[[:blank:]]' | tr '\t' '\n'

Answered By - glenn jackman

Answer Checked By - Senaida (WPSolving Volunteer)