def next_tamil_letter(self):
self.handle = codecs.open(self.filename,'r','utf-8')
for letter in utf8.get_letters_iterable(self.handle.read()):
if ( utf8.istamil( letter ) ):
yield letter
return
def next_tamil_letter(self):
self.handle = codecs.open(self.filename, 'r', 'utf-8')
for letter in utf8.get_letters(self.handle.read()):
if (utf8.istamil(letter)):
yield letter
raise StopIteration
def tokenize(self,data=None):
""" do hard-work of tokenizing and
put EzhilLexemes into the tokens[] Q """
if ( self.debug ): print(u"Start of Ezhil lexer - begin tokenize")
if ( self.stdin_mode ):
if ( self.debug ): print(self.tokens)
## cleanup the Q for stdin_mode of any EOF that can remain.
if ( len(self.tokens) != 0 ):
self.match( EzhilToken.EOF )
if( len(self.tokens) != 0 ):
raise ScannerException("Lexer: token Q has previous session tokens ")
self.tokens = list()
else:
if hasattr(self.File,'data'):
data = self.File.data
elif self.encoding == "utf-8":
data = self.File.read()
elif self.encoding == "tscii":
if self.debug: print("Loading TSCII converted data -> ")
data = self.converted_data
else:
assert False
if ( self.debug ): print(data)
idx = 0
tok_start_idx = 0
while ( idx < len( data ) ):
c = data[idx]
if ( self.debug ): print(idx,c)
if ( istamil( c ) or c.isalpha( ) or c == u'_' ):
tok_start_idx = idx
s = c; idx = idx + 1
while ( ( idx < len( data ) )
and self.is_allowed_for_identifier( data[idx] ) ):
s = s + data[idx]
idx = idx + 1
if idx < len(data) and not data[idx].isspace():
if data[idx] in ['#','$','@','\'','"']:
raise ScannerException("Lexer: token %s is not valid for identifier, with prefix %s"%(data[idx],s))
self.get_lexeme( s , tok_start_idx )
elif ( c.isspace() ): # or c in u' 'or c == u'\t' or c == u'\n'
if ( c == u'\n' ):
##actual col = idx - col_idx
self.update_line_col(idx)
idx = idx + 1
elif ( c == u'\r' ):
idx = idx + 1
continue
elif ( c == u'#' ):
## single line skip comments like Python/Octave
start = idx;
while ( idx < len( data ) and not (data[idx] in [u'\r',u'\n']) ):
idx = idx + 1
if ( idx < len(data) and data[idx] == u'\r' ):
idx = idx + 1
end = idx
self.comments[self.line]= data[start:end]
elif ( c.isdigit() ): #or c == '+' or c == '-' ):
num = c
tok_start_idx = idx
idx = idx + 1
## FIXME: this prevents you from +.xyz, or -.xyz use 0.xyz
## instead. also may throw an error if we exceed
## buffer-length.
if ( c in [u'+',u'-'] and ( idx < len( data ) )
and not data[idx].isdigit() ):
self.get_lexeme( c , idx )
continue
in_sci_notation = False
while ( ( idx < len( data) )
and ( data[idx].isdigit() or data[idx] in [u'+',u'-',u'e',u'E',u'.']) ):
if ( data[idx] in [u'+',u'-'] and not in_sci_notation ):
break;
elif( data[idx] in [u'e',u'E'] ):
in_sci_notation = True
num = num + data[idx]
idx = idx + 1
self.get_lexeme( num , tok_start_idx )
elif ( c == u"\"" ):
tok_start_idx = idx
s = c; idx = idx + 1
while ( idx < len( data ) and
( data[idx] != u'\"' ) ):
if ( data[idx] == u'\\' ):
idx = idx + 1
if ( data[idx] == u'n' ):
s = s + u'\n'
elif ( data[idx] == u't' ):
s = s +u'\t'
else:
s = s + data[idx]
else:
s = s + data[idx]
idx = idx + 1
s = s+data[idx]
idx = idx + 1
self.get_lexeme( s , tok_start_idx )
elif ( c in self.unary_binary_ops ):
tok_start_idx = idx
if ( len(data) > ( 1 + idx )
and data[idx+1] in [u'=',u'|',u'&',u'>',u'<'] ):
c = c +data[idx+1]
idx = idx + 1
self.get_lexeme( c , tok_start_idx )
idx = idx + 1
elif c == u";":
# treat as newline
idx = idx + 1
continue
else:
tok_start_idx = idx
idx = idx + 1
self.get_lexeme( c , tok_start_idx )
tok_start_idx = idx
## close the file if not stdin_mode
if ( not self.stdin_mode ): self.File.close()
## and manually add an EOF statement.
eof_tok = EzhilLexeme("",EzhilToken.EOF )
eof_tok.set_line_col( self.get_line_col( tok_start_idx ) )
self.tokens.append( eof_tok )
if ( self.debug ): print(u"before reverse"); self.dump_tokens()
self.tokens.reverse()
if ( self.debug ): print(u"after reverse"); self.dump_tokens()
return
def tokenize(self, data=None):
""" do hard-work of tokenizing and
put EzhilLexemes into the tokens[] Q """
if (self.debug): print(u"Start of Ezhil lexer - begin tokenize")
if (self.stdin_mode):
if (self.debug): print(self.tokens)
## cleanup the Q for stdin_mode of any EOF that can remain.
if (len(self.tokens) != 0):
self.match(EzhilToken.EOF)
if (len(self.tokens) != 0):
raise ScannerException(
"Lexer: token Q has previous session tokens ")
self.tokens = list()
else:
if hasattr(self.File, 'data'):
if (self.debug): print("data attribute")
data = self.File.data
elif self.encoding == "utf-8":
data = self.File.read()
elif self.encoding == "tscii":
if self.debug: print("Loading TSCII converted data -> ")
data = self.converted_data
else:
assert False
if (self.debug): print(data)
idx = 0
tok_start_idx = 0
while (idx < len(data)):
c = data[idx]
if (self.debug): print(idx, c)
if (istamil(c) or c.isalpha() or c == u'_'):
tok_start_idx = idx
s = c
idx = idx + 1
while ((idx < len(data))
and self.is_allowed_for_identifier(data[idx])):
s = s + data[idx]
idx = idx + 1
if idx < len(data) and not data[idx].isspace():
if data[idx] in ['#', '$', '@', '\'', '"']:
raise ScannerException(
"Lexer: token %s is not valid for identifier, with prefix %s"
% (data[idx], s))
self.get_lexeme(s, tok_start_idx)
elif (c.isspace()): # or c in u' 'or c == u'\t' or c == u'\n'
if (c == u'\n'):
##actual col = idx - col_idx
self.update_line_col(idx)
idx = idx + 1
elif (c == u'\r'):
idx = idx + 1
continue
elif (c == u'#'):
## single line skip comments like Python/Octave
start = idx
while (idx < len(data) and not (data[idx] in [u'\r', u'\n'])):
idx = idx + 1
if (idx < len(data) and data[idx] == u'\r'):
idx = idx + 1
end = idx
self.comments[self.line] = data[start:end]
elif (c.isdigit()): #or c == '+' or c == '-' ):
num = c
tok_start_idx = idx
idx = idx + 1
## FIXME: this prevents you from +.xyz, or -.xyz use 0.xyz
## instead. also may throw an error if we exceed
## buffer-length.
if (c in [u'+', u'-'] and (idx < len(data))
and not data[idx].isdigit()):
self.get_lexeme(c, idx)
continue
in_sci_notation = False
while ((idx < len(data))
and (data[idx].isdigit()
or data[idx] in [u'+', u'-', u'e', u'E', u'.'])):
if (data[idx] in [u'+', u'-'] and not in_sci_notation):
break
elif (data[idx] in [u'e', u'E']):
in_sci_notation = True
num = num + data[idx]
idx = idx + 1
self.get_lexeme(num, tok_start_idx)
elif (c == u"\""):
tok_start_idx = idx
s = c
idx = idx + 1
while (idx < len(data) and (data[idx] != u'\"')):
if (data[idx] == u'\\'):
idx = idx + 1
if (data[idx] == u'n'):
s = s + u'\n'
elif (data[idx] == u't'):
s = s + u'\t'
else:
s = s + data[idx]
else:
s = s + data[idx]
idx = idx + 1
s = s + data[idx]
idx = idx + 1
self.get_lexeme(s, tok_start_idx)
elif (c in self.unary_binary_ops):
tok_start_idx = idx
if (len(data) > (1 + idx)
and data[idx + 1] in [u'=', u'|', u'&', u'>', u'<']):
c = c + data[idx + 1]
idx = idx + 1
self.get_lexeme(c, tok_start_idx)
idx = idx + 1
elif c == u";":
# treat as newline
idx = idx + 1
continue
else:
tok_start_idx = idx
idx = idx + 1
self.get_lexeme(c, tok_start_idx)
tok_start_idx = idx
## close the file if not stdin_mode
if (not self.stdin_mode): self.File.close()
## and manually add an EOF statement.
eof_tok = EzhilLexeme("", EzhilToken.EOF)
eof_tok.set_line_col(self.get_line_col(tok_start_idx))
self.tokens.append(eof_tok)
if (self.debug):
print(u"before reverse")
self.dump_tokens()
self.tokens.reverse()
if (self.debug):
print(u"after reverse")
self.dump_tokens()
return
def next_tamil_letter(self):
self.handle = codecs.open(self.filename,'r','utf-8')
for letter in utf8.get_letters(self.handle.read()):
if ( utf8.istamil( letter ) ):
yield letter
raise StopIteration
def next_tamil_letter(self):
self.handle = codecs.open(self.filename,'r','utf-8')
for letter in utf8.get_letters_iterable(self.handle.read()):
if ( utf8.istamil( letter ) ):
yield letter
return
