#!/usr/bin/awk -f

function mystrtonum(str) {
	# adopted from http://www.gnu.org/software/gawk/manual/html_node/Strtonum-Function.html#Strtonum-Function
	if (str ~ /^0[xX][[:xdigit:]]+$/) {
		str = substr(str, 3) # lop off leading 0x
		n = length(str)
		ret = 0
		for (i = 1; i <= n; i++) {
			c = substr(str, i, 1)
			c = tolower(c)
			# index() returns 0 if c not in string,
			# includes c == "0"
			k = index("123456789abcdef", c)
			ret = ret * 16 + k
		}
	} else {
		ret = "NOT-A-HEX-NUMBER"
	}

	return ret
}

function add_hexrange (start, end) {
	from = mystrtonum("0x"start)
	to   = mystrtonum("0x"end)
	for ( i=from; i<=to; i++ )
		temp[i] = i
	result = temp[from]
	for ( i=from+1; i<=to; i++ )
		result = result " " temp[i]

	return result
}

BEGIN {
	FS = " "
	split("White_Space Dash Terminal_Punctuation STerm Pattern_White_Space", array_names)
}

{ 
	# skip comment and empty lines
	if ( $0 !~ /^#/ && NF != 0 ) {

		# cycle over array_names and do the math
		for (var in array_names) {
			if ( $0 ~ "; "array_names[var] ) {
				if ( $1 ~ /\.\./ ) {
					split($1, bounderies, "\.")
					array[var] = array[var] " " add_hexrange(bounderies[1], bounderies[3])
				} else {
					array[var] = array[var] " " mystrtonum("0x"$1)
				}
			}
		}
	}
}

END {
	print "/* This file is automatically generated by word-break-data.awk from PropList.txt */"
	for (var in array_names) {
		n = split(array[var], integers)
		print "static const uint32_t "array_names[var]"[]= {"
		for ( i=1; i<=n; i++) {
			if ( i == 1 ) {
				printf("\t0x%05X, ", integers[i])
			} else if ( (i-1)%8 == 0 ) {
				if ( i != n ) {
					printf("\n\t0x%05X, ", integers[i])
				} else {
					printf("\n\t0x%05X", integers[i])
				}
			} else if ( i != n ) {
				printf("0x%05X, ", integers[i])
			} else {
				printf("0x%05X", integers[i])
			}
		}
		print "\n};"
	}
}