AppPkg/Applications/Python: Add Python 2.7.2 sources since the release of Python...

[mirror_edk2.git] / AppPkg / Applications / Python / Python-2.7.2 / Tools / unicode / gencodec.py

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