unicode-tool.py

#!/usr/bin/python3

import sys

if len(sys.argv) == 1:
	print('usage:  ./unicode-tool.py FILE', file=sys.stderr)
	print(file=sys.stderr)
	print('FILE format: [NAME::STRING|# comment...]\\n...', file=sys.stderr)
	sys.exit(1)

f = open(sys.argv[1], 'r')
data = f.read().split('\n')
f.close()

print('### Generated by unicode-tool.')
print()
print()

for line in data:
	if line == '' or line.startswith('#'):
		continue
	
	name = line.split('::')[0]
	value = line.split('::')[1].replace('\\:\\:', '::').replace('\\\\', '\\')
	
	print('# ``' + value + '``')
	
	G_value = bytes(value + '\0', encoding='utf-8')
	
	# (gnu-version of string)
	print('/alloc G_' + name + '[' + str(len(G_value)) + ']')
	
	for i in range(len(G_value)):
		print(str(G_value[i]) + ' ({G_' + name + '} ' + str(i) + ' +) =')
	
	# (xmtwolime-version of string)
	print('/alloc X_' + name + '[' + str(len(value) + 1) + ']')
	
	for i in range(len(value)):
		print(str(ord(value[i])) + ' ({X_' + name + '} ' + str(i) + ' +) =')
	print('0 ({X_' + name + '} ' + str((len(value) + 1) - 1) + ' +) =')
	
	print()