
Python3.src.antlr4.Token.py Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of antlr4-runtime-testsuite Show documentation
Show all versions of antlr4-runtime-testsuite Show documentation
A collection of tests for ANTLR 4 Runtime libraries.
#[The "BSD license"]
# Copyright (c) 2012 Terence Parr
# Copyright (c) 2012 Sam Harwell
# Copyright (c) 2014 Eric Vergnaud
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions
# are met:
#
# 1. Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# 2. Redistributions in binary form must reproduce the above copyright
# notice, this list of conditions and the following disclaimer in the
# documentation and/or other materials provided with the distribution.
# 3. The name of the author may not be used to endorse or promote products
# derived from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
# IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
# OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
# IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
# INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
# NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
# THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#
# A token has properties: text, type, line, character position in the line
# (so we can ignore tabs), token channel, index, and source from which
# we obtained this token.
from io import StringIO
class Token (object):
INVALID_TYPE = 0
# During lookahead operations, this "token" signifies we hit rule end ATN state
# and did not follow it despite needing to.
EPSILON = -2
MIN_USER_TOKEN_TYPE = 1
EOF = -1
# All tokens go to the parser (unless skip() is called in that rule)
# on a particular "channel". The parser tunes to a particular channel
# so that whitespace etc... can go to the parser on a "hidden" channel.
DEFAULT_CHANNEL = 0
# Anything on different channel than DEFAULT_CHANNEL is not parsed
# by parser.
HIDDEN_CHANNEL = 1
def __init__(self):
self.source = None
self.type = None # token type of the token
self.channel = None # The parser ignores everything not on DEFAULT_CHANNEL
self.start = None # optional; return -1 if not implemented.
self.stop = None # optional; return -1 if not implemented.
self.tokenIndex = None # from 0..n-1 of the token object in the input stream
self.line = None # line=1..n of the 1st character
self.column = None # beginning of the line at which it occurs, 0..n-1
self._text = None # text of the token.
@property
def text(self):
return self._text
# Explicitly set the text for this token. If {code text} is not
# {@code null}, then {@link #getText} will return this value rather than
# extracting the text from the input.
#
# @param text The explicit text of the token, or {@code null} if the text
# should be obtained from the input along with the start and stop indexes
# of the token.
@text.setter
def text(self, text:str):
self._text = text
def getTokenSource(self):
return self.source[0]
def getInputStream(self):
return self.source[1]
class CommonToken(Token):
# An empty {@link Pair} which is used as the default value of
# {@link #source} for tokens that do not have a source.
EMPTY_SOURCE = (None, None)
def __init__(self, source:tuple = EMPTY_SOURCE, type:int = None, channel:int=Token.DEFAULT_CHANNEL, start:int=-1, stop:int=-1):
super().__init__()
self.source = source
self.type = type
self.channel = channel
self.start = start
self.stop = stop
self.tokenIndex = -1
if source[0] is not None:
self.line = source[0].line
self.column = source[0].column
else:
self.column = -1
# Constructs a new {@link CommonToken} as a copy of another {@link Token}.
#
#
# If {@code oldToken} is also a {@link CommonToken} instance, the newly
# constructed token will share a reference to the {@link #text} field and
# the {@link Pair} stored in {@link #source}. Otherwise, {@link #text} will
# be assigned the result of calling {@link #getText}, and {@link #source}
# will be constructed from the result of {@link Token#getTokenSource} and
# {@link Token#getInputStream}.
#
# @param oldToken The token to copy.
#
def clone(self):
t = CommonToken(self.source, self.type, self.channel, self.start, self.stop)
t.tokenIndex = self.tokenIndex
t.line = self.line
t.column = self.column
t.text = self.text
return t
@property
def text(self):
if self._text is not None:
return self._text
input = self.getInputStream()
if input is None:
return None
n = input.size
if self.start < n and self.stop < n:
return input.getText(self.start, self.stop)
else:
return ""
@text.setter
def text(self, text:str):
self._text = text
def __str__(self):
with StringIO() as buf:
buf.write("[@")
buf.write(str(self.tokenIndex))
buf.write(",")
buf.write(str(self.start))
buf.write(":")
buf.write(str(self.stop))
buf.write("='")
txt = self.text
if txt is not None:
txt = txt.replace("\n","\\n")
txt = txt.replace("\r","\\r")
txt = txt.replace("\t","\\t")
else:
txt = ""
buf.write(txt)
buf.write("',<")
buf.write(str(self.type))
buf.write(">")
if self.channel > 0:
buf.write(",channel=")
buf.write(str(self.channel))
buf.write(",")
buf.write(str(self.line))
buf.write(":")
buf.write(str(self.column))
buf.write("]")
return buf.getvalue()
© 2015 - 2025 Weber Informatics LLC | Privacy Policy