unite_wos_files

#!/usr/bin/env bash 
# unite_wos_files: Unite files exported from Web of Science
# 
# Author: Helge Knüttel
# Date: 2019-2020

# Some sensible settings. Taken from https://kvz.io/bash-best-practices.html
set -o errexit  # exit when a command fails
set -o pipefail # catch failing commands in a pipe
set -o nounset  # exit when script tries to use undeclared variables
# set -o xtrace # Trace what gets executed. Useful for debugging.

# Help text
PrintHelp() {

  cat << "EOF"

USAGE

  unite_wos_files [FILE]...
  unite_wos_files -h|--help

DESCRIPTION

  Unite several export files from Web of Science into a single file

  Web of Science allows to export bibliographic records only in portions of up to 500 records. This script will unite such files into a single one.

  Output is written to STDOUT.

EXAMPLE

  unite_wos_files WoS_other_reference_software_r*.txt > WoS_other_reference_software_records_combined.txt
  
EOF
}

case "$1" in
	-h|--help )
		PrintHelp
		exit 0
		;;
	* )
		# Just go on
		;;
esac


HEADER="FN Clarivate Analytics Web of Science"
VERSION="VR 1.0"
ENDOFFILE="EF"

# Add byte order mark
# TODO: Endnote export files don't have a byte order mark. Will it hurt when importing into Endnote?
printf '\xEF\xBB\xBF%s\n' "$HEADER"
printf '%s\n'  "$VERSION"

for file in "$@"
do
	# There may be a byte-order mark before the header
	sed -e "/${HEADER}/d" -e "/^${VERSION}/d" -e "/^${ENDOFFILE}/d" "$file"
done

printf '%s' "$ENDOFFILE"