1ca1c9b0cSelric#!/usr/local/bin/python 2ca1c9b0cSelric# -*- coding: iso-8859-1 -*- 3ca1c9b0cSelric 4*b40995a4Selric# Id 5ca1c9b0cSelric 6ca1c9b0cSelric# Copyright (c) 2004 Kungliga Tekniska Högskolan 7ca1c9b0cSelric# (Royal Institute of Technology, Stockholm, Sweden). 8ca1c9b0cSelric# All rights reserved. 9ca1c9b0cSelric# 10ca1c9b0cSelric# Redistribution and use in source and binary forms, with or without 11ca1c9b0cSelric# modification, are permitted provided that the following conditions 12ca1c9b0cSelric# are met: 13ca1c9b0cSelric# 14ca1c9b0cSelric# 1. Redistributions of source code must retain the above copyright 15ca1c9b0cSelric# notice, this list of conditions and the following disclaimer. 16ca1c9b0cSelric# 17ca1c9b0cSelric# 2. Redistributions in binary form must reproduce the above copyright 18ca1c9b0cSelric# notice, this list of conditions and the following disclaimer in the 19ca1c9b0cSelric# documentation and/or other materials provided with the distribution. 20ca1c9b0cSelric# 21ca1c9b0cSelric# 3. Neither the name of the Institute nor the names of its contributors 22ca1c9b0cSelric# may be used to endorse or promote products derived from this software 23ca1c9b0cSelric# without specific prior written permission. 24ca1c9b0cSelric# 25ca1c9b0cSelric# THIS SOFTWARE IS PROVIDED BY THE INSTITUTE AND CONTRIBUTORS ``AS IS'' AND 26ca1c9b0cSelric# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 27ca1c9b0cSelric# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 28ca1c9b0cSelric# ARE DISCLAIMED. IN NO EVENT SHALL THE INSTITUTE OR CONTRIBUTORS BE LIABLE 29ca1c9b0cSelric# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 30ca1c9b0cSelric# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 31ca1c9b0cSelric# OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 32ca1c9b0cSelric# HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 33ca1c9b0cSelric# LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 34ca1c9b0cSelric# OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 35ca1c9b0cSelric# SUCH DAMAGE. 36ca1c9b0cSelric 37ca1c9b0cSelricimport re 38ca1c9b0cSelricimport string 39ca1c9b0cSelric 40ca1c9b0cSelricdef read(filename): 41ca1c9b0cSelric """return a dict of unicode characters""" 42ca1c9b0cSelric ud = open(filename, 'r') 43ca1c9b0cSelric ret = {} 44ca1c9b0cSelric while True: 45ca1c9b0cSelric l = ud.readline() 46ca1c9b0cSelric if not l: 47ca1c9b0cSelric break 48ca1c9b0cSelric l = re.sub('#.*$', '', l) 49ca1c9b0cSelric if l == "\n": 50ca1c9b0cSelric continue 51ca1c9b0cSelric f = l.split(';') 52ca1c9b0cSelric key = int(f[0], 0x10) 53ca1c9b0cSelric if key in ret: 54ca1c9b0cSelric raise Exception('Duplicate key in UnicodeData') 55ca1c9b0cSelric ret[key] = f[1:] 56ca1c9b0cSelric ud.close() 57ca1c9b0cSelric return ret 58