def test_charmapencode(self):
if test_support.due_to_ironpython_bug("http://tkbgitvstfat01:8080/WorkItemTracking/WorkItem.aspx?artifactMoniker=303938"):
return
# For charmap encodings the replacement string will be
# mapped through the encoding again. This means, that
# to be able to use e.g. the "replace" handler, the
# charmap has to have a mapping for "?".
charmap = dict([ (ord(c), 2*c.upper()) for c in "abcdefgh"])
sin = u"abc"
sout = "AABBCC"
self.assertEquals(codecs.charmap_encode(sin, "strict", charmap)[0], sout)
sin = u"abcA"
self.assertRaises(UnicodeError, codecs.charmap_encode, sin, "strict", charmap)
charmap[ord("?")] = "XYZ"
sin = u"abcDEF"
sout = "AABBCCXYZXYZXYZ"
self.assertEquals(codecs.charmap_encode(sin, "replace", charmap)[0], sout)
charmap[ord("?")] = u"XYZ"
self.assertRaises(TypeError, codecs.charmap_encode, sin, "replace", charmap)
charmap[ord("?")] = u"XYZ"
self.assertRaises(TypeError, codecs.charmap_encode, sin, "replace", charmap)
def test_charmap_decode_1(self):
import codecs
assert codecs.charmap_encode(u'xxx') == ('xxx', 3)
assert codecs.charmap_encode(u'xxx', 'strict', {ord('x'): 'XX'}) == ('XXXXXX', 3)
res = codecs.charmap_decode("\x00\x01\x02", "replace", u"ab")
assert res == (u"ab\ufffd", 3)
res = codecs.charmap_decode("\x00\x01\x02", "replace", u"ab\ufffe")
assert res == (u'ab\ufffd', 3)
def test_charmap_decode_1(self):
import codecs
assert codecs.charmap_encode(u"xxx") == ("xxx", 3)
assert codecs.charmap_encode(u"xxx", "strict", {ord("x"): "XX"}) == ("XXXXXX", 3)
res = codecs.charmap_decode("\x00\x01\x02", "replace", u"ab")
assert res == (u"ab\ufffd", 3)
res = codecs.charmap_decode("\x00\x01\x02", "replace", u"ab\ufffe")
assert res == (u"ab\ufffd", 3)
def getPayload( self ):
print self.challenge
hash = password( self.passwd, self.challenge )
packet = "0" + DELIM1 + self.nick + DELIM1 + "6" + DELIM1 + codecs.charmap_encode(hash[ 0 ])[0] + DELIM1 + "96" + \
DELIM1 + codecs.charmap_encode(hash[ 1 ])[0] + DELIM1 + "2" + DELIM1 + "1" + DELIM1 + "1" + DELIM1 + \
self.nick + DELIM1
printPacket( packet )
return packet
def test_charmap_encode_replace(self):
charmap = dict([(c, bytes([c, c]).upper()) for c in b"abcdefgh"])
charmap[ord("?")] = b"XYZ"
import codecs
sin = "abcDEF"
sout = codecs.charmap_encode(sin, "replace", charmap)[0]
assert sout == b"AABBCCXYZXYZXYZ"
def doExecute( self, sock, words ):
# where is the song file?
server = self.config.getString( "playing", "server" )
filename = self.config.getString( "playing", "file" )
# get the file from the server
conn = httplib.HTTPConnection( server )
conn.request( "GET", filename )
response = conn.getresponse()
if response.status != 200:
raise Command.CommandError( "Could not access song file" )
file = response.read()
# parse the XML
dom = xml.dom.minidom.parseString(file)
node = dom.getElementsByTagName( "media" )[0]
artist = node.getElementsByTagName("artist")[0].firstChild.nodeValue.strip()
song = node.getElementsByTagName( "title" )[0].firstChild.nodeValue.strip()
album = node.getElementsByTagName( "album" )[0].firstChild.nodeValue.strip()
year = node.getElementsByTagName( "year" )[0].firstChild.nodeValue.strip()
# show it
msg = "%(artist)s: %(song)s [%(album)s, %(year)s]" % locals()
msg = codecs.charmap_encode(msg)[0]
self.sendMessage( sock, msg )
def test_charmap_encode_replace(self):
charmap = dict([ (ord(c), 2*c.upper()) for c in "abcdefgh"])
charmap[ord("?")] = "XYZ"
import codecs
sin = u"abcDEF"
sout = codecs.charmap_encode(sin, "replace", charmap)[0]
assert sout == "AABBCCXYZXYZXYZ"
def test_charmap_encode(self):
#Sanity
self.assertEqual(codecs.charmap_encode("abc"),
('abc', 3))
self.assertEqual(codecs.charmap_encode("abc", "strict"),
('abc', 3))
self.assertEqual(codecs.charmap_encode("", "strict", {}),
('', 0))
charmap = dict([ (ord(c), c.upper()) for c in "abcdefgh"])
self.assertEqual(codecs.charmap_encode(u"abc", "strict", charmap),
('ABC', 3))
#Sanity Negative
self.assertRaises(UnicodeEncodeError, codecs.charmap_encode, "abc", "strict", {})
def test_charmapencode(self):
# For charmap encodings the replacement string will be
# mapped through the encoding again. This means, that
# to be able to use e.g. the "replace" handler, the
# charmap has to have a mapping for "?".
charmap = dict((ord(c), bytes(2*c.upper(), 'ascii')) for c in "abcdefgh")
sin = "abc"
sout = b"AABBCC"
self.assertEquals(codecs.charmap_encode(sin, "strict", charmap)[0], sout)
sin = "abcA"
self.assertRaises(UnicodeError, codecs.charmap_encode, sin, "strict", charmap)
charmap[ord("?")] = b"XYZ"
sin = "abcDEF"
sout = b"AABBCCXYZXYZXYZ"
self.assertEquals(codecs.charmap_encode(sin, "replace", charmap)[0], sout)
charmap[ord("?")] = "XYZ" # wrong type in mapping
self.assertRaises(TypeError, codecs.charmap_encode, sin, "replace", charmap)
def _encode(input,errors='strict'):
# split to see if we have any 'extended' characters
runs=unicode_splitter.split(input)
# now iterate through handling any 'multibyte' ourselves
out_str=list()
consumed=0
extended=extended_encode_map.keys()
for run in runs:
if len(run)==1 and run[0] in extended:
out_str.append(extended_indicator+extended_encode_map[run])
consumed+=1
else:
# pass it to the standard encoder
out,cons=codecs.charmap_encode(run,errors,encoding_table)
out_str.append(out)
consumed+=cons
return (''.join(out_str),consumed)
def encode(self, input, errors="strict"):
return codecs.charmap_encode(input, errors, encoding_table)
def encode(self, char, final=False):
return codecs.charmap_encode(char, self.errors, ENCODING_MAP)[0]
""" Python Character Mapping Codec generated from 'CP852.TXT' with gencodec.py.
def encode(self, input, final=False):
data, nbytes = codecs.charmap_encode(input, self.errors,
encoding_map)
return data
def encode(self, input, final = False):
return codecs.charmap_encode(input, self.errors, self.mapping)[0]
def encode(self, input, final=False):
return codecs.charmap_encode(input, self.errors, self.mapping)[0]
""" Python Character Mapping Codec generated from 'CP860.TXT' with gencodec.py.
def encode(self, input, final = False):
return codecs.charmap_encode(input, self.errors, encoding_table)[0]
#!/usr/bin/env python3
# encoding: utf-8
#
# Copyright (c) 2010 Doug Hellmann. All rights reserved.
#
"""Error handling with character map encodings.
"""
#end_pymotw_header
import codecs
from codecs_invertcaps_charmap import encoding_map
text = 'pi: \u03c0'
for error in ['ignore', 'replace', 'strict']:
try:
encoded = codecs.charmap_encode(text, error, encoding_map)
except UnicodeEncodeError as err:
encoded = str(err)
print('{:7}: {}'.format(error, encoded))
""" Python Character Mapping Codec generated from 'CP857.TXT' with gencodec.py.
def encode(self,input,errors='strict'):
return codecs.charmap_encode(input,errors,encoding_table)
def encode(self, input, final=False): return codecs.charmap_encode(input,self.errors,encmap_hd44780_a02)[0]
def encode_a02(self,input,errors='strict'): return codecs.charmap_encode(input,errors,encmap_hd44780_a02)
def test_charmap_build(self):
decodemap = ''.join([unichr(i).upper() if chr(i).islower() else unichr(i).lower() for i in xrange(256)])
encodemap = codecs.charmap_build(decodemap)
self.assertEqual(codecs.charmap_decode(u'Hello World', 'strict', decodemap), ('hELLO wORLD', 11))
self.assertEqual(codecs.charmap_encode(u'Hello World', 'strict', encodemap), ('hELLO wORLD', 11))
""" Python Character Mapping Codec generated from 'CP500.TXT' with gencodec.py.
""" Python Character Mapping Codec generated from 'CP1257.TXT' with gencodec.py.
def encode(self, _input, errors='strict'):
return codecs.charmap_encode(_input, errors, encoding_map)
""" Python Character Mapping Codec generated from 'KOI8-R.TXT' with gencodec.py.
""" Python Character Mapping Codec generated from 'CP1006.TXT' with gencodec.py.
#
# Copyright (c) 2010 Doug Hellmann. All rights reserved.
#
"""Character mapping encoder
"""
#end_pymotw_header
import codecs
import string
# Map every character to itself
decoding_map = codecs.make_identity_dict(range(256))
# Make a list of pairs of ordinal values for the lower and upper case
# letters
pairs = zip([ ord(c) for c in string.ascii_lowercase],
[ ord(c) for c in string.ascii_uppercase])
# Modify the mapping to convert upper to lower and lower to upper.
decoding_map.update( dict( (upper, lower) for (lower, upper) in pairs) )
decoding_map.update( dict( (lower, upper) for (lower, upper) in pairs) )
# Create a separate encoding map.
encoding_map = codecs.make_encoding_map(decoding_map)
if __name__ == '__main__':
print codecs.charmap_encode('abc.DEF', 'strict', encoding_map)
print codecs.charmap_decode('abc.DEF', 'strict', decoding_map)
print encoding_map == decoding_map
""" Python Character Mapping Codec generated from 'CP424.TXT' with gencodec.py.
def encode(self, input, errors="strict"):
return codecs.charmap_encode(input, errors, encoding_map)
""" Python Character Mapping Codec generated from 'CP1256.TXT' with gencodec.py.
#!/usr/bin/python
# -*- coding: utf-8 -*-
#
# Copyright (c) 2008 Doug Hellmann All rights reserved.
#
import codecs
import string
# Map every character to itself
decoding_map = codecs.make_identity_dict(range(256))
# Make a list of pairs of ordinal values for the lower and upper case
# letters
pairs = zip([ord(c) for c in string.ascii_lowercase],
[ord(c) for c in string.ascii_uppercase])
# Modify the mapping to convert upper to lower and lower to upper.
decoding_map.update(dict((upper, lower) for (lower, upper) in pairs))
decoding_map.update(dict((lower, upper) for (lower, upper) in pairs))
# Create a separate encoding map.
encoding_map = codecs.make_encoding_map(decoding_map)
if __name__ == '__main__':
print codecs.charmap_encode('abc.DEF', 'strict', encoding_map)
print codecs.charmap_decode('abc.DEF', 'strict', decoding_map)
print encoding_map == decoding_map
""" Python Character Mapping Codec generated from '8859-9.TXT' with gencodec.py.
def encode(self, input, final=False): return codecs.charmap_encode(input,self.errors,encmap_hd44780_a02)[0]
""" Python Character Mapping Codec generated from '8859-10.TXT' with gencodec.py.
def encode(self, _input, final=False):
return codecs.charmap_encode(_input, self.errors, encoding_map)[0]
def encode(self, char, errors='strict'):
return codecs.charmap_encode(char, errors, ENCODING_MAP)
def encode(self, input, final = False):
return codecs.charmap_encode(input, self.errors, encoding_table)[0]
def encode(self, input, errors=error_handling):
return codecs.charmap_encode(input, errors, encoding_map)
# Map every character to itself
decoding_map = codecs.make_identity_dict( xrange(256) )
# Make a list of pairs of ordinal values for all lcase and ucase letters
pairs = zip( [ord(c) for c in string.ascii_lowercase],
[ord(c) for c in string.ascii_uppercase] )
# modify the mapping to convert upper to lower and lower to upper
decoding_map.update( dict( (upper, lower) for (lower, upper) in pairs ) )
decoding_map.update( dict( (lower, upper) for (lower, upper) in pairs ) )
# Create a separate encoding map
encoding_map = codecs.make_encoding_map( decoding_map )
print (codecs.charmap_encode('abc.DEF', error_handling, encoding_map))
print (codecs.charmap_decode('abc.DEF', error_handling, decoding_map))
print encoding_map == decoding_map
print
# by default, char map encoders and decoders support the standard error methods
# since this charmap only includes [a-zA-Z], the u"pi: π" from earlier fails
for error in ['ignore', 'replace', 'strict']:
try:
encoded = codecs.charmap_encode(data, error, encoding_map)
except UnicodeEncodeError, err:
encoded = str(err)
print '{:7} {}'.format(error, encoded)
print
# After defining a en/decoding maps, a few additonal classes have to be set up
""" Python Character Mapping Codec generated from 'ICELAND.TXT' with gencodec.py.
#!/usr/bin/env python
# -*- coding: utf-8 -*-
import codecs
from codecs_invertcaps_charmap import encoding_map
text = u'pi: π'
for error in [ 'ignore', 'replace', 'strict' ]:
try:
encoded = codecs.charmap_encode(text, error, encoding_map)
except UnicodeEncodeError, err:
encoded = str(err)
print '{:7}: {}'.format(error, encoded)
""" Python Character Mapping Codec generated from 'CP875.TXT' with gencodec.py.
""" Python Character Mapping Codec generated from '8859-8.TXT' with gencodec.py.
def encode(self,input,errors='strict'):
return codecs.charmap_encode(input,errors,encoding_map)
""" Python Character Mapping Codec generated from 'LATIN2.TXT' with gencodec.py.