1*ebfedea0SLionel Sambuc#!/usr/local/bin/python 2*ebfedea0SLionel Sambuc# -*- coding: iso-8859-1 -*- 3*ebfedea0SLionel Sambuc 4*ebfedea0SLionel Sambuc# Id 5*ebfedea0SLionel Sambuc 6*ebfedea0SLionel Sambuc# Copyright (c) 2004 Kungliga Tekniska Högskolan 7*ebfedea0SLionel Sambuc# (Royal Institute of Technology, Stockholm, Sweden). 8*ebfedea0SLionel Sambuc# All rights reserved. 9*ebfedea0SLionel Sambuc# 10*ebfedea0SLionel Sambuc# Redistribution and use in source and binary forms, with or without 11*ebfedea0SLionel Sambuc# modification, are permitted provided that the following conditions 12*ebfedea0SLionel Sambuc# are met: 13*ebfedea0SLionel Sambuc# 14*ebfedea0SLionel Sambuc# 1. Redistributions of source code must retain the above copyright 15*ebfedea0SLionel Sambuc# notice, this list of conditions and the following disclaimer. 16*ebfedea0SLionel Sambuc# 17*ebfedea0SLionel Sambuc# 2. Redistributions in binary form must reproduce the above copyright 18*ebfedea0SLionel Sambuc# notice, this list of conditions and the following disclaimer in the 19*ebfedea0SLionel Sambuc# documentation and/or other materials provided with the distribution. 20*ebfedea0SLionel Sambuc# 21*ebfedea0SLionel Sambuc# 3. Neither the name of the Institute nor the names of its contributors 22*ebfedea0SLionel Sambuc# may be used to endorse or promote products derived from this software 23*ebfedea0SLionel Sambuc# without specific prior written permission. 24*ebfedea0SLionel Sambuc# 25*ebfedea0SLionel Sambuc# THIS SOFTWARE IS PROVIDED BY THE INSTITUTE AND CONTRIBUTORS ``AS IS'' AND 26*ebfedea0SLionel Sambuc# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 27*ebfedea0SLionel Sambuc# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 28*ebfedea0SLionel Sambuc# ARE DISCLAIMED. IN NO EVENT SHALL THE INSTITUTE OR CONTRIBUTORS BE LIABLE 29*ebfedea0SLionel Sambuc# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 30*ebfedea0SLionel Sambuc# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 31*ebfedea0SLionel Sambuc# OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 32*ebfedea0SLionel Sambuc# HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 33*ebfedea0SLionel Sambuc# LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 34*ebfedea0SLionel Sambuc# OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 35*ebfedea0SLionel Sambuc# SUCH DAMAGE. 36*ebfedea0SLionel Sambuc 37*ebfedea0SLionel Sambucimport re 38*ebfedea0SLionel Sambucimport string 39*ebfedea0SLionel Sambuc 40*ebfedea0SLionel Sambucdef read(filename): 41*ebfedea0SLionel Sambuc """return a dict of unicode characters""" 42*ebfedea0SLionel Sambuc ud = open(filename, 'r') 43*ebfedea0SLionel Sambuc ret = {} 44*ebfedea0SLionel Sambuc while True: 45*ebfedea0SLionel Sambuc l = ud.readline() 46*ebfedea0SLionel Sambuc if not l: 47*ebfedea0SLionel Sambuc break 48*ebfedea0SLionel Sambuc l = re.sub('#.*$', '', l) 49*ebfedea0SLionel Sambuc if l == "\n": 50*ebfedea0SLionel Sambuc continue 51*ebfedea0SLionel Sambuc f = l.split(';') 52*ebfedea0SLionel Sambuc key = int(f[0], 0x10) 53*ebfedea0SLionel Sambuc if key in ret: 54*ebfedea0SLionel Sambuc raise Exception('Duplicate key in UnicodeData') 55*ebfedea0SLionel Sambuc ret[key] = f[1:] 56*ebfedea0SLionel Sambuc ud.close() 57*ebfedea0SLionel Sambuc return ret 58