Generate unicode ranges

unicode_range.sh

#!/bin/sh
#
# http://www.unicode.org/reports/tr44/#UnicodeData.txt
# http://www.unicode.org/Public/UNIDATA/UnicodeData.txt

total=""
prev_dec=
range_start_dec=

# Match general categories Lu, Ll, Lt, LC, Lm, Lo, L
for line_hex in $(awk -F ';' 'substr($3, 0, 1) == "L" {print $1}' UnicodeData.txt); do
	line_dec=$(( 0x$line_hex ))

	# Special case for first iteration.
	if [ "$range_start_dec" = "" ]; then
		range_start_dec=$line_dec
		prev_dec=$line_dec
		continue
	fi

	# There's a gap, append to total. Set new start.
	if [ $(( $line_dec - 1 )) -ne $prev_dec ]; then
		total="$total$(printf '\\u%0*x-\\u%0*x' 4 $range_start_dec 4 $prev_dec)"
		range_start_dec=$line_dec
	fi

	prev_dec=$line_dec
done
echo "$total"