#!/usr/bin/env python
# -*- coding: utf-8 -*-

# Name: create-unicodedata-dicts-prop-value-alias.py
# Goal: Create dictionaries for unicodedata package contining property value aliases in terms of abbreviated names and long names.
# Fixes: http://bugs.python.org/issue16684
# Date: 2012-12-23
# Author: Pander <pander@users.sourceforge.net>
# License: Python license

import urllib2

inbc = False
ingc = False
ineaw = False

dictbc = {}
dictgc = {}
dicteaw = {}

aliasfile = urllib2.urlopen('http://www.unicode.org/Public/UNIDATA/PropertyValueAliases.txt')

for line in aliasfile.readlines():
    line = line.strip()

    if line == '# Bidi_Class (bc)':
        inbc = True
        continue
    elif line == '# General_Category (gc)':
        ingc = True
        continue
    elif line == '# East_Asian_Width (ea)':
        ineaw = True
        continue

    if inbc:
        if line == '':
            if len(dictbc) != 0:
                # passed all bc lines
                inbc = False
            continue
        line = line.split(';')
        abbrname = line[1].strip()
        longname = line[2].strip()
        if '#' in longname:
            # strip comments
            longname = longname.split('#')[0].strip()
        dictbc[abbrname] = longname
    elif ingc:
        if line == '':
            if len(dictgc) != 0:
                # passed all gc lines
                ingc = False
            continue
        line = line.split(';')
        abbrname = line[1].strip()
        longname = line[2].strip()
        if '#' in longname:
            # strip comments
            longname = longname.split('#')[0].strip()
        dictgc[abbrname] = longname
    elif ineaw:
        if line == '':
            if len(dicteaw) != 0:
                # passed all eaw lines
                ineaw = False
            continue
        line = line.split(';')
        abbrname = line[1].strip()
        longname = line[2].strip()
        if '#' in longname:
            # strip comments
            longname = longname.split('#')[0].strip()
        dicteaw[abbrname] = longname

print 'bcdict = {'
for bc in dictbc.iterkeys():
    if len(bc) == 1:
        print '    \'%s\'  : \'%s\',' %(bc, dictbc[bc])
    elif len(bc) == 2:
        print '    \'%s\' : \'%s\',' %(bc, dictbc[bc])
    else:
        print '    \'%s\': \'%s\',' %(bc, dictbc[bc])
print '}'
print

print 'gcdict = {'
for gc in dictgc.iterkeys():
    if len(gc) == 1:
        print '    \'%s\' : \'%s\',' %(gc, dictgc[gc])
    else:
        print '    \'%s\': \'%s\',' %(gc, dictgc[gc])
print '}'
print

print 'eawdict = {'
for eaw in dicteaw.iterkeys():
    if len(eaw) == 1:
        print '    \'%s\' : \'%s\',' %(eaw, dicteaw[eaw])
    else:
        print '    \'%s\': \'%s\',' %(eaw, dicteaw[eaw])
print '}'
print