427 lines
12 KiB
Python
427 lines
12 KiB
Python
""" Unicode Mapping Parser and Codec Generator.
|
|
|
|
This script parses Unicode mapping files as available from the Unicode
|
|
site (ftp://ftp.unicode.org/Public/MAPPINGS/) and creates Python codec
|
|
modules from them. The codecs use the standard character mapping codec
|
|
to actually apply the mapping.
|
|
|
|
Synopsis: gencodec.py dir codec_prefix
|
|
|
|
All files in dir are scanned and those producing non-empty mappings
|
|
will be written to <codec_prefix><mapname>.py with <mapname> being the
|
|
first part of the map's filename ('a' in a.b.c.txt) converted to
|
|
lowercase with hyphens replaced by underscores.
|
|
|
|
The tool also writes marshalled versions of the mapping tables to the
|
|
same location (with .mapping extension).
|
|
|
|
Written by Marc-Andre Lemburg (mal@lemburg.com).
|
|
|
|
(c) Copyright CNRI, All Rights Reserved. NO WARRANTY.
|
|
(c) Copyright Guido van Rossum, 2000.
|
|
|
|
Table generation:
|
|
(c) Copyright Marc-Andre Lemburg, 2005.
|
|
Licensed to PSF under a Contributor Agreement.
|
|
|
|
"""#"
|
|
|
|
import re, os, time, marshal, codecs
|
|
|
|
# Maximum allowed size of charmap tables
|
|
MAX_TABLE_SIZE = 8192
|
|
|
|
# Standard undefined Unicode code point
|
|
UNI_UNDEFINED = unichr(0xFFFE)
|
|
|
|
mapRE = re.compile('((?:0x[0-9a-fA-F]+\+?)+)'
|
|
'\s+'
|
|
'((?:(?:0x[0-9a-fA-Z]+|<[A-Za-z]+>)\+?)*)'
|
|
'\s*'
|
|
'(#.+)?')
|
|
|
|
def parsecodes(codes,
|
|
len=len, filter=filter,range=range):
|
|
|
|
""" Converts code combinations to either a single code integer
|
|
or a tuple of integers.
|
|
|
|
meta-codes (in angular brackets, e.g. <LR> and <RL>) are
|
|
ignored.
|
|
|
|
Empty codes or illegal ones are returned as None.
|
|
|
|
"""
|
|
if not codes:
|
|
return None
|
|
l = codes.split('+')
|
|
if len(l) == 1:
|
|
return int(l[0],16)
|
|
for i in range(len(l)):
|
|
try:
|
|
l[i] = int(l[i],16)
|
|
except ValueError:
|
|
l[i] = None
|
|
l = filter(lambda x: x is not None, l)
|
|
if len(l) == 1:
|
|
return l[0]
|
|
else:
|
|
return tuple(l)
|
|
|
|
def readmap(filename):
|
|
|
|
f = open(filename,'r')
|
|
lines = f.readlines()
|
|
f.close()
|
|
enc2uni = {}
|
|
identity = []
|
|
unmapped = range(256)
|
|
|
|
# UTC mapping tables per convention don't include the identity
|
|
# mappings for code points 0x00 - 0x1F and 0x7F, unless these are
|
|
# explicitly mapped to different characters or undefined
|
|
for i in range(32) + [127]:
|
|
identity.append(i)
|
|
unmapped.remove(i)
|
|
enc2uni[i] = (i, 'CONTROL CHARACTER')
|
|
|
|
for line in lines:
|
|
line = line.strip()
|
|
if not line or line[0] == '#':
|
|
continue
|
|
m = mapRE.match(line)
|
|
if not m:
|
|
#print '* not matched: %s' % repr(line)
|
|
continue
|
|
enc,uni,comment = m.groups()
|
|
enc = parsecodes(enc)
|
|
uni = parsecodes(uni)
|
|
if comment is None:
|
|
comment = ''
|
|
else:
|
|
comment = comment[1:].strip()
|
|
if enc < 256:
|
|
if enc in unmapped:
|
|
unmapped.remove(enc)
|
|
if enc == uni:
|
|
identity.append(enc)
|
|
enc2uni[enc] = (uni,comment)
|
|
else:
|
|
enc2uni[enc] = (uni,comment)
|
|
|
|
# If there are more identity-mapped entries than unmapped entries,
|
|
# it pays to generate an identity dictionary first, and add explicit
|
|
# mappings to None for the rest
|
|
if len(identity) >= len(unmapped):
|
|
for enc in unmapped:
|
|
enc2uni[enc] = (None, "")
|
|
enc2uni['IDENTITY'] = 256
|
|
|
|
return enc2uni
|
|
|
|
def hexrepr(t, precision=4):
|
|
|
|
if t is None:
|
|
return 'None'
|
|
try:
|
|
len(t)
|
|
except:
|
|
return '0x%0*X' % (precision, t)
|
|
try:
|
|
return '(' + ', '.join(['0x%0*X' % (precision, item)
|
|
for item in t]) + ')'
|
|
except TypeError, why:
|
|
print '* failed to convert %r: %s' % (t, why)
|
|
raise
|
|
|
|
def python_mapdef_code(varname, map, comments=1, precisions=(2, 4)):
|
|
|
|
l = []
|
|
append = l.append
|
|
if map.has_key("IDENTITY"):
|
|
append("%s = codecs.make_identity_dict(range(%d))" %
|
|
(varname, map["IDENTITY"]))
|
|
append("%s.update({" % varname)
|
|
splits = 1
|
|
del map["IDENTITY"]
|
|
identity = 1
|
|
else:
|
|
append("%s = {" % varname)
|
|
splits = 0
|
|
identity = 0
|
|
|
|
mappings = map.items()
|
|
mappings.sort()
|
|
i = 0
|
|
key_precision, value_precision = precisions
|
|
for mapkey, mapvalue in mappings:
|
|
mapcomment = ''
|
|
if isinstance(mapkey, tuple):
|
|
(mapkey, mapcomment) = mapkey
|
|
if isinstance(mapvalue, tuple):
|
|
(mapvalue, mapcomment) = mapvalue
|
|
if mapkey is None:
|
|
continue
|
|
if (identity and
|
|
mapkey == mapvalue and
|
|
mapkey < 256):
|
|
# No need to include identity mappings, since these
|
|
# are already set for the first 256 code points.
|
|
continue
|
|
key = hexrepr(mapkey, key_precision)
|
|
value = hexrepr(mapvalue, value_precision)
|
|
if mapcomment and comments:
|
|
append(' %s: %s,\t# %s' % (key, value, mapcomment))
|
|
else:
|
|
append(' %s: %s,' % (key, value))
|
|
i += 1
|
|
if i == 4096:
|
|
# Split the definition into parts to that the Python
|
|
# parser doesn't dump core
|
|
if splits == 0:
|
|
append('}')
|
|
else:
|
|
append('})')
|
|
append('%s.update({' % varname)
|
|
i = 0
|
|
splits = splits + 1
|
|
if splits == 0:
|
|
append('}')
|
|
else:
|
|
append('})')
|
|
|
|
return l
|
|
|
|
def python_tabledef_code(varname, map, comments=1, key_precision=2):
|
|
|
|
l = []
|
|
append = l.append
|
|
append('%s = (' % varname)
|
|
|
|
# Analyze map and create table dict
|
|
mappings = map.items()
|
|
mappings.sort()
|
|
table = {}
|
|
maxkey = 0
|
|
if map.has_key('IDENTITY'):
|
|
for key in range(256):
|
|
table[key] = (key, '')
|
|
maxkey = 255
|
|
del map['IDENTITY']
|
|
for mapkey, mapvalue in mappings:
|
|
mapcomment = ''
|
|
if isinstance(mapkey, tuple):
|
|
(mapkey, mapcomment) = mapkey
|
|
if isinstance(mapvalue, tuple):
|
|
(mapvalue, mapcomment) = mapvalue
|
|
if mapkey is None:
|
|
continue
|
|
table[mapkey] = (mapvalue, mapcomment)
|
|
if mapkey > maxkey:
|
|
maxkey = mapkey
|
|
if maxkey > MAX_TABLE_SIZE:
|
|
# Table too large
|
|
return None
|
|
|
|
# Create table code
|
|
for key in range(maxkey + 1):
|
|
if key not in table:
|
|
mapvalue = None
|
|
mapcomment = 'UNDEFINED'
|
|
else:
|
|
mapvalue, mapcomment = table[key]
|
|
if mapvalue is None:
|
|
mapchar = UNI_UNDEFINED
|
|
else:
|
|
if isinstance(mapvalue, tuple):
|
|
# 1-n mappings not supported
|
|
return None
|
|
else:
|
|
mapchar = unichr(mapvalue)
|
|
if mapcomment and comments:
|
|
append(' %r\t# %s -> %s' % (mapchar,
|
|
hexrepr(key, key_precision),
|
|
mapcomment))
|
|
else:
|
|
append(' %r' % mapchar)
|
|
|
|
append(')')
|
|
return l
|
|
|
|
def codegen(name, map, encodingname, comments=1):
|
|
|
|
""" Returns Python source for the given map.
|
|
|
|
Comments are included in the source, if comments is true (default).
|
|
|
|
"""
|
|
# Generate code
|
|
decoding_map_code = python_mapdef_code(
|
|
'decoding_map',
|
|
map,
|
|
comments=comments)
|
|
decoding_table_code = python_tabledef_code(
|
|
'decoding_table',
|
|
map,
|
|
comments=comments)
|
|
encoding_map_code = python_mapdef_code(
|
|
'encoding_map',
|
|
codecs.make_encoding_map(map),
|
|
comments=comments,
|
|
precisions=(4, 2))
|
|
|
|
if decoding_table_code:
|
|
suffix = 'table'
|
|
else:
|
|
suffix = 'map'
|
|
|
|
l = [
|
|
'''\
|
|
""" Python Character Mapping Codec %s generated from '%s' with gencodec.py.
|
|
|
|
"""#"
|
|
|
|
import codecs
|
|
|
|
### Codec APIs
|
|
|
|
class Codec(codecs.Codec):
|
|
|
|
def encode(self,input,errors='strict'):
|
|
return codecs.charmap_encode(input,errors,encoding_%s)
|
|
|
|
def decode(self,input,errors='strict'):
|
|
return codecs.charmap_decode(input,errors,decoding_%s)
|
|
''' % (encodingname, name, suffix, suffix)]
|
|
l.append('''\
|
|
class IncrementalEncoder(codecs.IncrementalEncoder):
|
|
def encode(self, input, final=False):
|
|
return codecs.charmap_encode(input,self.errors,encoding_%s)[0]
|
|
|
|
class IncrementalDecoder(codecs.IncrementalDecoder):
|
|
def decode(self, input, final=False):
|
|
return codecs.charmap_decode(input,self.errors,decoding_%s)[0]''' %
|
|
(suffix, suffix))
|
|
|
|
l.append('''
|
|
class StreamWriter(Codec,codecs.StreamWriter):
|
|
pass
|
|
|
|
class StreamReader(Codec,codecs.StreamReader):
|
|
pass
|
|
|
|
### encodings module API
|
|
|
|
def getregentry():
|
|
return codecs.CodecInfo(
|
|
name=%r,
|
|
encode=Codec().encode,
|
|
decode=Codec().decode,
|
|
incrementalencoder=IncrementalEncoder,
|
|
incrementaldecoder=IncrementalDecoder,
|
|
streamreader=StreamReader,
|
|
streamwriter=StreamWriter,
|
|
)
|
|
''' % encodingname.replace('_', '-'))
|
|
|
|
# Add decoding table or map (with preference to the table)
|
|
if not decoding_table_code:
|
|
l.append('''
|
|
### Decoding Map
|
|
''')
|
|
l.extend(decoding_map_code)
|
|
else:
|
|
l.append('''
|
|
### Decoding Table
|
|
''')
|
|
l.extend(decoding_table_code)
|
|
|
|
# Add encoding map
|
|
if decoding_table_code:
|
|
l.append('''
|
|
### Encoding table
|
|
encoding_table=codecs.charmap_build(decoding_table)
|
|
''')
|
|
else:
|
|
l.append('''
|
|
### Encoding Map
|
|
''')
|
|
l.extend(encoding_map_code)
|
|
|
|
# Final new-line
|
|
l.append('')
|
|
|
|
return '\n'.join(l).expandtabs()
|
|
|
|
def pymap(name,map,pyfile,encodingname,comments=1):
|
|
|
|
code = codegen(name,map,encodingname,comments)
|
|
f = open(pyfile,'w')
|
|
f.write(code)
|
|
f.close()
|
|
|
|
def marshalmap(name,map,marshalfile):
|
|
|
|
d = {}
|
|
for e,(u,c) in map.items():
|
|
d[e] = (u,c)
|
|
f = open(marshalfile,'wb')
|
|
marshal.dump(d,f)
|
|
f.close()
|
|
|
|
def convertdir(dir, dirprefix='', nameprefix='', comments=1):
|
|
|
|
mapnames = os.listdir(dir)
|
|
for mapname in mapnames:
|
|
mappathname = os.path.join(dir, mapname)
|
|
if not os.path.isfile(mappathname):
|
|
continue
|
|
name = os.path.split(mapname)[1]
|
|
name = name.replace('-','_')
|
|
name = name.split('.')[0]
|
|
name = name.lower()
|
|
name = nameprefix + name
|
|
codefile = name + '.py'
|
|
marshalfile = name + '.mapping'
|
|
print 'converting %s to %s and %s' % (mapname,
|
|
dirprefix + codefile,
|
|
dirprefix + marshalfile)
|
|
try:
|
|
map = readmap(os.path.join(dir,mapname))
|
|
if not map:
|
|
print '* map is empty; skipping'
|
|
else:
|
|
pymap(mappathname, map, dirprefix + codefile,name,comments)
|
|
marshalmap(mappathname, map, dirprefix + marshalfile)
|
|
except ValueError, why:
|
|
print '* conversion failed: %s' % why
|
|
raise
|
|
|
|
def rewritepythondir(dir, dirprefix='', comments=1):
|
|
|
|
mapnames = os.listdir(dir)
|
|
for mapname in mapnames:
|
|
if not mapname.endswith('.mapping'):
|
|
continue
|
|
name = mapname[:-len('.mapping')]
|
|
codefile = name + '.py'
|
|
print 'converting %s to %s' % (mapname,
|
|
dirprefix + codefile)
|
|
try:
|
|
map = marshal.load(open(os.path.join(dir,mapname),
|
|
'rb'))
|
|
if not map:
|
|
print '* map is empty; skipping'
|
|
else:
|
|
pymap(mapname, map, dirprefix + codefile,name,comments)
|
|
except ValueError, why:
|
|
print '* conversion failed: %s' % why
|
|
|
|
if __name__ == '__main__':
|
|
|
|
import sys
|
|
if 1:
|
|
convertdir(*sys.argv[1:])
|
|
else:
|
|
rewritepythondir(*sys.argv[1:])
|