class Tokeniser(object):
modes = {
'expr': (tokens, token_indices),
'filename': (filename_tokens, {}),
}
def __init__(self, feeder):
self.pos = 0
self.feeder = feeder
self.prescanner = Prescanner(feeder)
self.code = self.prescanner.scan()
self.change_mode('expr')
def change_mode(self, mode):
self.mode = mode
self.tokens, self.token_indices = self.modes[mode]
def incomplete(self):
'get more code from the prescanner and continue'
self.prescanner.incomplete()
self.code += self.prescanner.scan()
def sntx_message(self, pos=None):
if pos is None:
pos = self.pos
pre, post = self.code[:pos], self.code[pos:].rstrip('\n')
if pos == 0:
self.feeder.message('Syntax', 'sntxb', post)
else:
self.feeder.message('Syntax', 'sntxf', pre, post)
def next(self):
'return next token'
self.skip_blank()
if self.pos >= len(self.code):
return Token('END', '', len(self.code))
# look for a matching pattern
indices = self.token_indices.get(self.code[self.pos], ())
if indices:
for index in indices:
tag, pattern = self.tokens[index]
match = pattern.match(self.code, self.pos)
if match is not None:
break
else:
for tag, pattern in self.tokens:
match = pattern.match(self.code, self.pos)
if match is not None:
break
# no matching pattern found
if match is None:
self.sntx_message()
raise ScanError()
# custom tokenisation rules defined with t_tag
override = getattr(self, 't_' + tag, None)
if override is not None:
return override(match)
else:
text = match.group(0)
self.pos = match.end(0)
return Token(tag, text, match.start(0))
def skip_blank(self):
'skip whitespace and comments'
comment = [] # start positions of comments
while True:
if self.pos >= len(self.code):
if comment:
self.incomplete()
else:
break
if comment:
if self.code.startswith('(*', self.pos):
comment.append(self.pos)
self.pos += 2
elif self.code.startswith('*)', self.pos):
comment.pop()
self.pos += 2
else:
self.pos += 1
elif self.code.startswith('(*', self.pos):
comment.append(self.pos)
self.pos += 2
elif self.code[self.pos] in ' \r\n\t':
self.pos += 1
else:
break
def t_String(self, match):
start, end = self.pos, None
self.pos += 1 # skip opening '"'
newlines = []
while True:
if self.pos >= len(self.code):
if end is None:
# reached end while still inside string
self.incomplete()
newlines.append(self.pos)
else:
break
c = self.code[self.pos]
if c == '"':
self.pos += 1
end = self.pos
break
elif c == '\\':
self.pos += 2
else:
self.pos += 1
indices = [start] + newlines + [end]
result = ''.join(self.code[indices[i]:indices[i + 1]]
for i in range(len(indices) - 1))
return Token('String', result, start)
def t_Number(self, match):
text = match.group(0)
pos = match.end(0)
if self.code[pos - 1:pos + 1] == '..':
# Trailing .. should be ignored. That is, `1..` is `Repeated[1]`.
text = text[:-1]
self.pos = pos - 1
else:
self.pos = pos
return Token('Number', text, match.start(0))
def token_mode(self, match, tag, mode):
'consume a token and switch mode'
text = match.group(0)
self.pos = match.end(0)
self.change_mode(mode)
return Token(tag, text, match.start(0))
def t_Get(self, match):
return self.token_mode(match, 'Get', 'filename')
def t_Put(self, match):
return self.token_mode(match, 'Put', 'filename')
def t_PutAppend(self, match):
return self.token_mode(match, 'PutAppend', 'filename')
def t_Filename(self, match):
return self.token_mode(match, 'Filename', 'expr')
class Tokeniser(object):
modes = {
'expr': (tokens, token_indices),
'filename': (filename_tokens, {}),
}
def __init__(self, feeder):
self.pos = 0
self.feeder = feeder
self.prescanner = Prescanner(feeder)
self.code = self.prescanner.scan()
self.change_mode('expr')
def change_mode(self, mode):
self.mode = mode
self.tokens, self.token_indices = self.modes[mode]
def incomplete(self):
'get more code from the prescanner and continue'
self.prescanner.incomplete()
self.code += self.prescanner.scan()
def sntx_message(self, pos=None):
if pos is None:
pos = self.pos
pre, post = self.code[:pos], self.code[pos:].rstrip('\n')
if pos == 0:
self.feeder.message('Syntax', 'sntxb', post)
else:
self.feeder.message('Syntax', 'sntxf', pre, post)
def next(self):
'return next token'
self.skip_blank()
if self.pos >= len(self.code):
return Token('END', '', len(self.code))
# look for a matching pattern
indices = self.token_indices.get(self.code[self.pos], ())
if indices:
for index in indices:
tag, pattern = self.tokens[index]
match = pattern.match(self.code, self.pos)
if match is not None:
break
else:
for tag, pattern in self.tokens:
match = pattern.match(self.code, self.pos)
if match is not None:
break
# no matching pattern found
if match is None:
self.sntx_message()
raise ScanError()
# custom tokenisation rules defined with t_tag
override = getattr(self, 't_' + tag, None)
if override is not None:
return override(match)
else:
text = match.group(0)
self.pos = match.end(0)
return Token(tag, text, match.start(0))
def skip_blank(self):
'skip whitespace and comments'
comment = [] # start positions of comments
while True:
if self.pos >= len(self.code):
if comment:
self.incomplete()
else:
break
if comment:
if self.code.startswith('(*', self.pos):
comment.append(self.pos)
self.pos += 2
elif self.code.startswith('*)', self.pos):
comment.pop()
self.pos += 2
else:
self.pos += 1
elif self.code.startswith('(*', self.pos):
comment.append(self.pos)
self.pos += 2
elif self.code[self.pos] in ' \r\n\t':
self.pos += 1
else:
break
def t_String(self, match):
start, end = self.pos, None
self.pos += 1 # skip opening '"'
newlines = []
while True:
if self.pos >= len(self.code):
if end is None:
# reached end while still inside string
self.incomplete()
newlines.append(self.pos)
else:
break
c = self.code[self.pos]
if c == '"':
self.pos += 1
end = self.pos
break
elif c == '\\':
self.pos += 2
else:
self.pos += 1
indices = [start] + newlines + [end]
result = ''.join(self.code[indices[i]:indices[i + 1]]
for i in range(len(indices) - 1))
return Token('String', result, start)
def t_Number(self, match):
text = match.group(0)
pos = match.end(0)
if self.code[pos - 1:pos + 1] == '..':
# Trailing .. should be ignored. That is, `1..` is `Repeated[1]`.
text = text[:-1]
self.pos = pos - 1
else:
self.pos = pos
return Token('Number', text, match.start(0))
def token_mode(self, match, tag, mode):
'consume a token and switch mode'
text = match.group(0)
self.pos = match.end(0)
self.change_mode(mode)
return Token(tag, text, match.start(0))
def t_Get(self, match):
return self.token_mode(match, 'Get', 'filename')
def t_Put(self, match):
return self.token_mode(match, 'Put', 'filename')
def t_PutAppend(self, match):
return self.token_mode(match, 'PutAppend', 'filename')
def t_Filename(self, match):
return self.token_mode(match, 'Filename', 'expr')
class Tokeniser(object):
modes = {
"expr": (tokens, token_indices),
"filename": (filename_tokens, {}),
}
def __init__(self, feeder):
self.pos = 0
self.feeder = feeder
self.prescanner = Prescanner(feeder)
self.code = self.prescanner.scan()
self.change_mode("expr")
def change_mode(self, mode):
self.mode = mode
self.tokens, self.token_indices = self.modes[mode]
def incomplete(self):
"get more code from the prescanner and continue"
self.prescanner.incomplete()
self.code += self.prescanner.scan()
def sntx_message(self, pos=None):
if pos is None:
pos = self.pos
pre, post = self.code[:pos], self.code[pos:].rstrip("\n")
if pos == 0:
self.feeder.message("Syntax", "sntxb", post)
else:
self.feeder.message("Syntax", "sntxf", pre, post)
def next(self):
"return next token"
self.skip_blank()
if self.pos >= len(self.code):
return Token("END", "", len(self.code))
# look for a matching pattern
indices = self.token_indices.get(self.code[self.pos], ())
if indices:
for index in indices:
tag, pattern = self.tokens[index]
match = pattern.match(self.code, self.pos)
if match is not None:
break
else:
for tag, pattern in self.tokens:
match = pattern.match(self.code, self.pos)
if match is not None:
break
# no matching pattern found
if match is None:
self.sntx_message()
raise ScanError()
# custom tokenisation rules defined with t_tag
override = getattr(self, "t_" + tag, None)
if override is not None:
return override(match)
else:
text = match.group(0)
self.pos = match.end(0)
return Token(tag, text, match.start(0))
def skip_blank(self):
"skip whitespace and comments"
comment = [] # start positions of comments
while True:
if self.pos >= len(self.code):
if comment:
self.incomplete()
else:
break
if comment:
if self.code.startswith("(*", self.pos):
comment.append(self.pos)
self.pos += 2
elif self.code.startswith("*)", self.pos):
comment.pop()
self.pos += 2
else:
self.pos += 1
elif self.code.startswith("(*", self.pos):
comment.append(self.pos)
self.pos += 2
elif self.code[self.pos] in " \r\n\t":
self.pos += 1
else:
break
def t_String(self, match):
start, end = self.pos, None
self.pos += 1 # skip opening '"'
newlines = []
while True:
if self.pos >= len(self.code):
if end is None:
# reached end while still inside string
self.incomplete()
newlines.append(self.pos)
else:
break
c = self.code[self.pos]
if c == '"':
self.pos += 1
end = self.pos
break
elif c == "\\":
self.pos += 2
else:
self.pos += 1
indices = [start] + newlines + [end]
result = "".join(self.code[indices[i]:indices[i + 1]]
for i in range(len(indices) - 1))
return Token("String", result, start)
def t_Number(self, match):
text = match.group(0)
pos = match.end(0)
if self.code[pos - 1:pos + 1] == "..":
# Trailing .. should be ignored. That is, `1..` is `Repeated[1]`.
text = text[:-1]
self.pos = pos - 1
else:
self.pos = pos
return Token("Number", text, match.start(0))
def token_mode(self, match, tag, mode):
"consume a token and switch mode"
text = match.group(0)
self.pos = match.end(0)
self.change_mode(mode)
return Token(tag, text, match.start(0))
def t_Get(self, match):
return self.token_mode(match, "Get", "filename")
def t_Put(self, match):
return self.token_mode(match, "Put", "filename")
def t_PutAppend(self, match):
return self.token_mode(match, "PutAppend", "filename")
def t_Filename(self, match):
return self.token_mode(match, "Filename", "expr")