basicparser.py

#! /usr/bin/python

# SPDX-License-Identifier: GPL-3.0-or-later
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program.  If not, see <https://www.gnu.org/licenses/>.

from basictoken import BASICToken as Token
from flowsignal import FlowSignal
import math
import random
try:
    from time import ticks_ms as monotonic
except:
    from time import monotonic


"""Implements a BASIC array, which may have up
to three dimensions of fixed size.

"""
class BASICArray:

    def __init__(self, dimensions, elem_type):
        """Initialises the object with the specified
        number of dimensions. Maximum number of
        dimensions is three

        :param dimensions: List of array dimensions and their
        corresponding sizes
        :param elem_type: Indicates whether the elements are strings ('str')
        or numbers ('num')

        """
        self.dims = min(3, len(dimensions))

        if self.dims == 0:
            raise SyntaxError("Zero dimensional array specified")

        # Check for invalid sizes and ensure int
        for i in range(self.dims):
            if dimensions[i] < 0:
                raise SyntaxError("Negative array size specified")
            # Allow sizes like 1.0f, but not 1.1f
            if int(dimensions[i]) != dimensions[i]:
                raise SyntaxError("Fractional array size specified")
            dimensions[i] = int(dimensions[i])

        # Store original dimensions for bounds checking
        self.original_dims = dimensions.copy()

        # MSBASIC: Initialize to Zero
        # MSBASIC: Overdim by one, as some dialects are 1 based and expect
        #          to use the last item at index = size
        if self.dims == 1:
            if elem_type == 'num':
                self.data = [0 for x in range(dimensions[0] + 1)]
            else:
                self.data = ['' for x in range(dimensions[0] + 1)]
        elif self.dims == 2:
            if elem_type == 'num':
                self.data = [
                     [0 for x in range(dimensions[1] + 1)] for x in range(dimensions[0] + 1)
                ]
            else:
                self.data = [
                    ['' for x in range(dimensions[1] + 1)] for x in range(dimensions[0] + 1)
                ]
        else:
            if elem_type == 'num':
                self.data = [
                    [
                        [0 for x in range(dimensions[2] + 1)]
                        for x in range(dimensions[1] + 1)
                    ]
                    for x in range(dimensions[0] + 1)
                ]
            else:
                self.data = [
                    [
                        ['' for x in range(dimensions[2] + 1)]
                        for x in range(dimensions[1] + 1)
                    ]
                    for x in range(dimensions[0] + 1)
                ]

    def pretty_print(self):
        print(str(self.data))

"""Implements a BASIC parser that parses a single
statement when supplied.

"""
class BASICParser:

    def __init__(self, basicdata):
        # Symbol table to hold variable names mapped
        # to values
        self.__symbol_table = {}

        # Stack on which to store operands
        # when evaluating expressions
        self.__operand_stack = []

        # BasicDATA structure containing program DATA Statements
        self.__data = basicdata
        # List to hold values read from DATA statements
        self.__data_values = []

        # These values will be
        # initialised on a per
        # statement basis
        self.__tokenlist = []
        self.__tokenindex = None

        # Previous flowsignal used to determine initializion of
        # loop variable
        self.last_flowsignal = None

        # Set to keep track of print column across multiple print statements
        self.__prnt_column = 0

        #file handle list
        self.__file_handles = {}

    def parse(self, tokenlist, line_number):
        """Must be initialised with the list of
        BTokens to be processed. These tokens
        represent a BASIC statement without
        its corresponding line number.

        :param tokenlist: The tokenized program statement
        :param line_number: The line number of the statement

        :return: The FlowSignal to indicate to the program
        how to branch if necessary, None otherwise

        """

        # Remember the line number to aid error reporting
        self.__line_number = line_number
        self.__tokenlist = []
        self.__tokenindex = 0
        linetokenindex = 0
        for token in tokenlist:
            # If statements will always be the last statement processed on a line so
            # any colons found after an IF are part of the condition execution statements
            # and will be processed in the recursive call to parse
            if token.category == token.IF:
                # process IF statement to move __tokenidex to the code block
                # of the THEN or ELSE and then call PARSE recursively to process that code block
                # this will terminate the token loop by RETURNing to the calling module
                #
                # **Warning** if an IF stmt is used in the THEN code block or multiple IF statement are used
                # in a THEN or ELSE block the block grouping is ambiguous and logical processing may not
                # function as expected. There is no ambiguity when single IF statements are placed within ELSE blocks
                linetokenindex += self.__tokenindex
                self.__tokenindex = 0
                self.__tokenlist = tokenlist[linetokenindex:]

                # Assign the first token
                self.__token = self.__tokenlist[0]
                flow = self.__stmt() # process IF statement
                if flow and (flow.ftype == FlowSignal.EXECUTE):
                    # recursive call to process THEN/ELSE block
                    try:
                        return self.parse(tokenlist[linetokenindex+self.__tokenindex:],line_number)
                    except RuntimeError as err:
                        raise RuntimeError(str(err)+' in line ' + str(self.__line_number))
                else:
                    # branch on original syntax 'IF cond THEN lineno [ELSE lineno]'
                    # in this syntax the then or else code block is not a legal basic statement
                    # so recursive processing can't be used
                    return flow
            elif token.category == token.COLON:
                # Found a COLON, process tokens found to this point
                linetokenindex += self.__tokenindex
                self.__tokenindex = 0

                # Assign the first token
                self.__token = self.__tokenlist[self.__tokenindex]

                flow = self.__stmt()
                if flow:
                    return flow

                linetokenindex += 1
                self.__tokenlist = []
            elif token.category == token.ELSE and self.__tokenlist[0].category != token.OPEN:
                # if we find an ELSE and we are not processing an OPEN statement, we must
                # be in a recursive call and be processing a THEN block
                # since we're processing the THEN block we are done if we hit an ELSE
                break
            else:
                self.__tokenlist.append(token)

        # reached end of statement, process tokens collected since last COLON (or from start if no COLONs)
        linetokenindex += self.__tokenindex
        self.__tokenindex = 0
        # Assign the first token
        self.__token = self.__tokenlist[self.__tokenindex]

        return self.__stmt()

    def __advance(self):
        """Advances to the next token

        """
        # Move to the next token
        self.__tokenindex += 1

        # Acquire the next token if there any left
        if not self.__tokenindex >= len(self.__tokenlist):
            self.__token = self.__tokenlist[self.__tokenindex]

    def __consume(self, expected_category):
        """Consumes a token from the list

        """
        if self.__token.category == expected_category:
            self.__advance()

        else:
            raise RuntimeError('Expecting ' + Token.catnames[expected_category] +
                               ' in line ' + str(self.__line_number))

    def __stmt(self):
        """Parses a program statement

        :return: The FlowSignal to indicate to the program
        how to branch if necessary, None otherwise

        """
        if self.__token.category in [Token.FOR, Token.IF, Token.NEXT,
                                     Token.ON, Token.WHILE, Token.WEND]:
            return self.__compoundstmt()

        else:
            return self.__simplestmt()

    def __simplestmt(self):
        """Parses a non-compound program statement

        :return: The FlowSignal to indicate to the program
        how to branch if necessary, None otherwise

        """
        if self.__token.category == Token.NAME:
            self.__assignmentstmt()
            return None

        elif self.__token.category == Token.PRINT:
            self.__printstmt()
            return None

        elif self.__token.category == Token.LET:
            self.__letstmt()
            return None

        elif self.__token.category == Token.GOTO:
            return self.__gotostmt()

        elif self.__token.category == Token.GOSUB:
            return self.__gosubstmt()

        elif self.__token.category == Token.RETURN:
            return self.__returnstmt()

        elif self.__token.category == Token.STOP:
            return self.__stopstmt()

        elif self.__token.category == Token.INPUT:
            self.__inputstmt()
            return None

        elif self.__token.category == Token.DIM:
            self.__dimstmt()
            return None

        elif self.__token.category == Token.RANDOMIZE:
            self.__randomizestmt()
            return None

        elif self.__token.category == Token.DATA:
            self.__datastmt()
            return None

        elif self.__token.category == Token.READ:
            self.__readstmt()
            return None

        elif self.__token.category == Token.RESTORE:
            self.__restorestmt()
            return None

        elif self.__token.category == Token.OPEN:
            return self.__openstmt()

        elif self.__token.category == Token.CLOSE:
            self.__closestmt()
            return None

        elif self.__token.category == Token.FSEEK:
            self.__fseekstmt()
            return None

        else:
            # Ignore comments, but raise an error
            # for anything else
            if self.__token.category != Token.REM:
                raise RuntimeError('Expecting program statement in line '
                                   + str(self.__line_number))

    def __printstmt(self):
        """Parses a PRINT statement, causing
        the value that is on top of the
        operand stack to be printed on
        the screen.

        """
        self.__advance()   # Advance past PRINT token

        fileIO = False
        if self.__token.category == Token.HASH:
            fileIO = True

            # Process the # keyword
            self.__consume(Token.HASH)

            # Acquire the file number
            self.__expr()
            filenum = self.__operand_stack.pop()

            if self.__file_handles.get(filenum) == None:
                raise RuntimeError("PRINT: file #"+str(filenum)+" not opened in line " + str(self.__line_number))

            # Process the comma
            if self.__tokenindex < len(self.__tokenlist) and self.__token.category != Token.COLON:
                self.__consume(Token.COMMA)

        # Check there are items to print
        if not self.__tokenindex >= len(self.__tokenlist):
            prntTab = (self.__token.category == Token.TAB)
            self.__logexpr()

            if prntTab:
                if self.__prnt_column >= len(self.__operand_stack[-1]):
                    if fileIO:
                        self.__file_handles[filenum].write("\n")
                    else:
                        print()
                    self.__prnt_column = 0

                current_pr_column = len(self.__operand_stack[-1]) - self.__prnt_column
                self.__prnt_column = len(self.__operand_stack.pop()) - 1
                if current_pr_column > 1:
                    if fileIO:
                        self.__file_handles[filenum].write(" "*(current_pr_column-1))
                    else:
                        print(" "*(current_pr_column-1), end="")
            else:
                self.__prnt_column += len(str(self.__operand_stack[-1]))
                if fileIO:
                    self.__file_handles[filenum].write('%s' %(self.__operand_stack.pop()))
                else:
                    print(self.__operand_stack.pop(), end='')

            while self.__token.category == Token.SEMICOLON:
                if self.__tokenindex == len(self.__tokenlist) - 1:
                    # If a semicolon ends this line, don't print
                    # a newline.. a-la ms-basic
                    self.__advance()
                    return
                self.__advance()
                prntTab = (self.__token.category == Token.TAB)
                self.__logexpr()

                if prntTab:
                    if self.__prnt_column >= len(self.__operand_stack[-1]):
                        if fileIO:
                            self.__file_handles[filenum].write("\n")
                        else:
                            print()
                        self.__prnt_column = 0
                    current_pr_column = len(self.__operand_stack[-1]) - self.__prnt_column
                    if fileIO:
                        self.__file_handles[filenum].write(" "*(current_pr_column-1))
                    else:
                        print(" "*(current_pr_column-1), end="")
                    self.__prnt_column = len(self.__operand_stack.pop()) - 1
                else:
                    self.__prnt_column += len(str(self.__operand_stack[-1]))
                    if fileIO:
                        self.__file_handles[filenum].write('%s' %(self.__operand_stack.pop()))
                    else:
                        print(self.__operand_stack.pop(), end='')

        # Final newline
        if fileIO:
            self.__file_handles[filenum].write("\n")
        else:
            print()
        self.__prnt_column = 0

    def __letstmt(self):
        """Parses a LET statement,
        consuming the LET keyword.
        """
        self.__advance()  # Advance past the LET token
        self.__assignmentstmt()

    def __gotostmt(self):
        """Parses a GOTO statement

        :return: A FlowSignal containing the target line number
        of the GOTO

        """
        self.__advance()  # Advance past GOTO token
        self.__expr()

        # Set up and return the flow signal
        return FlowSignal(ftarget=self.__operand_stack.pop())

    def __gosubstmt(self):
        """Parses a GOSUB statement

        :return: A FlowSignal containing the first line number
        of the subroutine

        """

        self.__advance()  # Advance past GOSUB token
        self.__expr()

        # Set up and return the flow signal
        return FlowSignal(ftarget=self.__operand_stack.pop(),
                          ftype=FlowSignal.GOSUB)

    def __returnstmt(self):
        """Parses a RETURN statement"""

        self.__advance()  # Advance past RETURN token

        # Set up and return the flow signal
        return FlowSignal(ftype=FlowSignal.RETURN)

    def __stopstmt(self):
        """Parses a STOP statement"""

        self.__advance()  # Advance past STOP token

        for handles in self.__file_handles:
            self.__file_handles[handles].close()
        self.__file_handles.clear()

        return FlowSignal(ftype=FlowSignal.STOP)

    def __assignmentstmt(self):
        """Parses an assignment statement,
        placing the corresponding
        variable and its value in the symbol
        table.

        """
        left = self.__token.lexeme  # Save lexeme of
                                    # the current token
        self.__advance()

        if self.__token.category == Token.LEFTPAREN:
            # We are assigning to an array
            self.__arrayassignmentstmt(left)

        else:
            # We are assigning to a simple variable
            self.__consume(Token.ASSIGNOP)
            self.__logexpr()

            # Check that we are using the right variable name format
            right = self.__operand_stack.pop()

            if left.endswith('$') and not isinstance(right, str):
                raise SyntaxError('Syntax error: Attempt to assign non string to string variable' +
                                  ' in line ' + str(self.__line_number))

            elif not left.endswith('$') and isinstance(right, str):
                raise SyntaxError('Syntax error: Attempt to assign string to numeric variable' +
                                  ' in line ' + str(self.__line_number))

            self.__symbol_table[left] = right

    def __dimstmt(self):
        """Parses  DIM statement and creates a symbol
        table entry for an array of the specified
        dimensions.

        """
        self.__advance()  # Advance past DIM keyword

        # MSBASIC: allow dims of multiple arrays delimited by commas
        while True:
            # Extract the array name, append a suffix so
            # that we can distinguish from simple variables
            # in the symbol table
            name = self.__token.lexeme + '_array'
            self.__advance()  # Advance past array name

            self.__consume(Token.LEFTPAREN)

            # Extract the dimensions
            dimensions = []
            if not self.__tokenindex >= len(self.__tokenlist):
                self.__expr()
                dimensions.append(self.__operand_stack.pop())

                while self.__token.category == Token.COMMA:
                    self.__advance()  # Advance past comma
                    self.__expr()
                    dimensions.append(self.__operand_stack.pop())

            self.__consume(Token.RIGHTPAREN)

            if len(dimensions) > 3:
                raise SyntaxError(
                    'Maximum number of array dimensions is three '
                    + 'in line '
                    + str(self.__line_number)
                )

            # Ensure array is initialised with correct values
            # depending upon type
            if name.endswith('$_array'):
                self.__symbol_table[name] = BASICArray(dimensions, 'str')
            else:
                self.__symbol_table[name] = BASICArray(dimensions, 'num')

            if self.__tokenindex == len(self.__tokenlist):
                # We have parsed the last token here...
                return
            else:
                self.__consume(Token.COMMA)


    def __arrayassignmentstmt(self, name):
        """Parses an assignment to an array variable

        :param name: Array name

        """
        self.__consume(Token.LEFTPAREN)

        # Capture the index variables
        # Extract the dimensions
        indexvars = []
        if not self.__tokenindex >= len(self.__tokenlist):
            self.__expr()
            indexvars.append(self.__operand_stack.pop())

            while self.__token.category == Token.COMMA:
                self.__advance()  # Advance past comma
                self.__expr()
                indexvars.append(self.__operand_stack.pop())

        try:
            BASICarray = self.__symbol_table[name + '_array']

        except KeyError:
            raise KeyError('Array could not be found in line ' +
                           str(self.__line_number))

        if BASICarray.dims != len(indexvars):
            raise IndexError('Incorrect number of indices applied to array ' +
                             'in line ' + str(self.__line_number))

        self.__consume(Token.RIGHTPAREN)
        self.__consume(Token.ASSIGNOP)

        self.__logexpr()

        # Check that we are using the right variable name format
        right = self.__operand_stack.pop()

        if name.endswith('$') and not isinstance(right, str):
            raise SyntaxError('Attempt to assign non string to string array' +
                              ' in line ' + str(self.__line_number))

        elif not name.endswith('$') and isinstance(right, str):
            raise SyntaxError('Attempt to assign string to numeric array' +
                              ' in line ' + str(self.__line_number))

        # Assign to the specified array index
        self.__assign_array_val(BASICarray, indexvars, right)

    def __openstmt(self):
        """Parses an open statement, opens the indicated file and
        places the file handle into handle table
        """

        self.__advance() # Advance past OPEN token

        # Acquire the filename
        self.__logexpr()
        filename = self.__operand_stack.pop()

        # Process the FOR keyword
        self.__consume(Token.FOR)

        if self.__token.category == Token.INPUT:
            accessMode = "r"
        elif self.__token.category == Token.APPEND:
            accessMode = "r+"
        elif self.__token.category == Token.OUTPUT:
            accessMode = "w+"
        else:
            raise SyntaxError('Invalid Open access mode in line ' + str(self.__line_number))

        self.__advance() # Advance past access type

        if self.__token.lexeme != "AS":
            raise SyntaxError('Expecting AS in line ' + str(self.__line_number))

        self.__advance() # Advance past AS keyword

        # Process the # keyword
        self.__consume(Token.HASH)

        # Acquire the file number
        self.__expr()
        filenum = self.__operand_stack.pop()

        branchOnError = False
        if self.__token.category == Token.ELSE:
            branchOnError = True
            self.__advance() # Advance past ELSE

            if self.__token.category == Token.GOTO:
                self.__advance()    # Advance past optional GOTO

            self.__expr()

        if self.__file_handles.get(filenum) != None:
            if branchOnError:
                return FlowSignal(ftarget=self.__operand_stack.pop())
            else:
                raise RuntimeError("File #",filenum," already opened in line " + str(self.__line_number))

        try:
            self.__file_handles[filenum] = open(filename,accessMode)

        except:
            if branchOnError:
                return FlowSignal(ftarget=self.__operand_stack.pop())
            else:
                raise RuntimeError('File '+filename+' could not be opened in line ' + str(self.__line_number))

        if accessMode == "r+":
            # By checking the 'newlines' attribute the appropriate adjustment for
            # the file being opened can be determined.

            try:
                # newline attribute is only set after a line is read
                # manually identify newlines in case newlines attribute isn't supported
                fileline = self.__file_handles[filenum].readline()
                newlines = ""
                for ichar in range(len(fileline)-1,-1,-1):
                    if fileline[ichar] not in ['\n','\r']:
                        break
                    newlines = fileline[ichar:]
            except:
                newlines = None

            if hasattr(self.__file_handles[filenum],'newlines'):
                newlines = self.__file_handles[filenum].newlines

            self.__file_handles[filenum].seek(0)
            filelen = 0

            if newlines is not None:
                newlineAdj = len(newlines) - 1
            else:
                # If we wern't able to determine an appropriate value and
                # using version of Python that doesn't have newlines attribute
                # use adjustment appropriate for Windows formatted files
                newlineAdj = 1

            for lines in self.__file_handles[filenum]:
                filelen += len(lines)+newlineAdj

            self.__file_handles[filenum].seek(filelen)

        return None

    def __closestmt(self):
        """Parses a close, closes the file and removes
        the file handle from the handle table
        """

        self.__advance() # Advance past CLOSE token

        # Process the # keyword
        self.__consume(Token.HASH)

        # Acquire the file number
        self.__expr()
        filenum = self.__operand_stack.pop()

        if self.__file_handles.get(filenum) == None:
            raise RuntimeError("CLOSE: file #"+str(filenum)+" not opened in line " + str(self.__line_number))

        self.__file_handles[filenum].close()
        self.__file_handles.pop(filenum)

    def __fseekstmt(self):
        """Parses an fseek statement, seeks the indicated file position
        """

        self.__advance() # Advance past FSEEK token

        # Process the # keyword
        self.__consume(Token.HASH)

        # Acquire the file number
        self.__expr()
        filenum = self.__operand_stack.pop()

        if self.__file_handles.get(filenum) == None:
            raise RuntimeError("FSEEK: file #"+str(filenum)+" not opened in line " + str(self.__line_number))

        # Process the comma
        self.__consume(Token.COMMA)

        # Acquire the file position
        self.__expr()

        self.__file_handles[filenum].seek(self.__operand_stack.pop())

    def __inputstmt(self):
        """Parses an input statement, extracts the input
        from the user and places the values into the
        symbol table. Now supports array elements.

        """
        self.__advance()  # Advance past INPUT token

        fileIO = False
        if self.__token.category == Token.HASH:
            fileIO = True

            # Process the # keyword
            self.__consume(Token.HASH)

            # Acquire the file number
            self.__expr()
            filenum = self.__operand_stack.pop()

            if self.__file_handles.get(filenum) == None:
                raise RuntimeError("INPUT: file #"+str(filenum)+" not opened in line " + str(self.__line_number))

            # Process the comma
            self.__consume(Token.COMMA)

        prompt = '? '
        if self.__token.category == Token.STRING:
            if fileIO:
                raise SyntaxError('Input prompt specified for file I/O ' +
                                'in line ' + str(self.__line_number))

            # Acquire the input prompt
            self.__logexpr()
            prompt = self.__operand_stack.pop()
            self.__consume(Token.SEMICOLON)

        # Parse the target variables (simple variables or array elements)
        targets = []
        if not self.__tokenindex >= len(self.__tokenlist):
            # Parse first target
            targets.append(self.__parse_variable_target())

            while self.__token.category == Token.COMMA:
                self.__advance()  # Advance past comma
                targets.append(self.__parse_variable_target())

        valid_input = False
        while not valid_input:
            # Gather input from the user into the variables
            if fileIO:
                inputvals = ((self.__file_handles[filenum].readline().replace("\n","")).replace("\r","")).split(',', (len(targets)-1))
                valid_input = True
            else:
                inputvals = input(prompt).split(',', (len(targets)-1))

            try:
                for target in targets:
                    try:
                        right = inputvals.pop(0)

                        if target['is_string']:
                            value = str(right)
                        else:
                            # Try to convert to numeric
                            try:
                                if '.' in right:
                                    value = float(right)
                                else:
                                    value = int(right)
                            except ValueError:
                                if not fileIO:
                                    valid_input = False
                                    print('Non-numeric input provided to a numeric variable - redo from start')
                                    break
                                raise ValueError('Non-numeric input provided to a numeric variable ' +
                                               'in line ' + str(self.__line_number))

                        # Assign value to target (validates array indices if needed)
                        self.__assign_to_target(target, value)
                        valid_input = True

                    except IndexError:
                        # No more input to process
                        if not fileIO:
                            valid_input = False
                            print('Not enough values input - redo from start')
                            break
                        raise RuntimeError('Not enough input values in line ' + str(self.__line_number))

            except RuntimeError as e:
                if 'SUBSCRIPT ERROR' in str(e):
                    raise e  # Re-raise subscript errors immediately
                else:
                    raise e

    def __restorestmt(self):

        self.__advance() # Advance past RESTORE token

        # Acquire the line number
        self.__expr()

        self.__data_values.clear()
        self.__data.restore(self.__operand_stack.pop())

    def __datastmt(self):
        """Parses a DATA statement"""

    def __readstmt(self):
        """Parses a READ statement. Now supports array elements."""

        self.__advance()  # Advance past READ token

        # Parse the target variables (simple variables or array elements)
        targets = []
        if not self.__tokenindex >= len(self.__tokenlist):
            # Parse first target
            targets.append(self.__parse_variable_target())

            while self.__token.category == Token.COMMA:
                self.__advance()  # Advance past comma
                targets.append(self.__parse_variable_target())

        # Process each target
        for target in targets:
            # Validate array indices BEFORE reading data (per specification)
            if target['type'] == 'array':
                BASICarray = self.__symbol_table[target['array_name']]
                self.__validate_array_indices(BASICarray, target['indices'])

            # Get data value
            if len(self.__data_values) < 1:
                try:
                    self.__data_values = self.__data.readData(self.__line_number)
                except RuntimeError as e:
                    if "No DATA statements available" in str(e):
                        raise RuntimeError('? OUT OF DATA in line ' + str(self.__line_number))
                    else:
                        raise e

            if len(self.__data_values) < 1:
                raise RuntimeError('? OUT OF DATA in line ' + str(self.__line_number))

            right = self.__data_values.pop(0)

            if target['is_string']:
                # String target - accept any data type as string
                if isinstance(right, (int, float)):
                    # Convert numeric data to string
                    value = str(right)
                else:
                    # Already a string
                    value = right
            else:
                # Numeric target - must be numeric data
                try:
                    if isinstance(right, str):
                        # Try to parse string as number
                        numeric = float(right) if '.' in right else int(right)
                    else:
                        numeric = float(right)
                        if int(numeric) == numeric:
                            numeric = int(numeric)
                    value = numeric
                except ValueError:
                    raise ValueError('Non-numeric input provided to a numeric variable ' +
                                     'in line ' + str(self.__line_number))

            # Assign value to target
            self.__assign_to_target(target, value)

    def __expr(self):
        """Parses a numerical expression consisting
        of two terms being added or subtracted,
        leaving the result on the operand stack.

        """
        self.__term()  # Pushes value of left term
                       # onto top of stack

        while self.__token.category in [Token.PLUS, Token.MINUS]:
            savedcategory = self.__token.category
            self.__advance()
            self.__term()  # Pushes value of right term
                           # onto top of stack
            rightoperand = self.__operand_stack.pop()
            leftoperand = self.__operand_stack.pop()

            if savedcategory == Token.PLUS:
                self.__operand_stack.append(leftoperand + rightoperand)

            else:
                self.__operand_stack.append(leftoperand - rightoperand)

    def __term(self):
        """Parses a numerical expression consisting
        of two factors being multiplied together,
        leaving the result on the operand stack.

        """
        self.__sign = 1  # Initialise sign to keep track of unary
                         # minuses
        self.__factor()  # Leaves value of term on top of stack

        while self.__token.category in [Token.TIMES, Token.DIVIDE, Token.MODULO]:
            savedcategory = self.__token.category
            self.__advance()
            self.__sign = 1  # Initialise sign
            self.__factor()  # Leaves value of term on top of stack
            rightoperand = self.__operand_stack.pop()
            leftoperand = self.__operand_stack.pop()

            if savedcategory == Token.TIMES:
                self.__operand_stack.append(leftoperand * rightoperand)

            elif savedcategory == Token.DIVIDE:
                self.__operand_stack.append(leftoperand / rightoperand)

            else:
                self.__operand_stack.append(leftoperand % rightoperand)

    def __factor(self):
        """Evaluates a numerical expression
        and leaves its value on top of the
        operand stack.

        """
        if self.__token.category == Token.PLUS:
            self.__advance()
            self.__factor()

        elif self.__token.category == Token.MINUS:
            self.__sign = -self.__sign
            self.__advance()
            self.__factor()

        elif self.__token.category == Token.UNSIGNEDINT:
            self.__operand_stack.append(self.__sign*int(self.__token.lexeme))
            self.__advance()

        elif self.__token.category == Token.UNSIGNEDFLOAT:
            self.__operand_stack.append(self.__sign*float(self.__token.lexeme))
            self.__advance()

        elif self.__token.category == Token.STRING:
            self.__operand_stack.append(self.__token.lexeme)
            self.__advance()

        elif (
            self.__token.category == Token.NAME
            and self.__token.category not in Token.functions
        ):
            # Check if this is a simple or array variable
            # MSBASIC Allows simple and complex variables to have the
            # same id.  This is probably a bad idea, but it's used in
            # some old example programs.  So check if next token is parens
            if (
                (self.__token.lexeme + "_array") in self.__symbol_table
                and self.__tokenindex < len(self.__tokenlist) - 1
                and self.__tokenlist[self.__tokenindex + 1].category == Token.LEFTPAREN
            ):
                # Capture the current lexeme
                arrayname = self.__token.lexeme + "_array"

                # Array must be processed
                # Capture the index variables
                self.__advance()  # Advance past the array name

                try:
                    self.__consume(Token.LEFTPAREN)
                    indexvars = []
                    if not self.__tokenindex >= len(self.__tokenlist):
                        self.__expr()
                        indexvars.append(self.__operand_stack.pop())

                        while self.__token.category == Token.COMMA:
                            self.__advance()  # Advance past comma
                            self.__expr()
                            indexvars.append(self.__operand_stack.pop())

                    BASICarray = self.__symbol_table[arrayname]
                    arrayval = self.__get_array_val(BASICarray, indexvars)

                    if arrayval != None:
                        self.__operand_stack.append(self.__sign * arrayval)

                    else:
                        raise IndexError(
                            "Empty array value returned in line "
                            + str(self.__line_number)
                        )
                except RuntimeError:
                    raise RuntimeError(
                        "Array used without index in line " + str(self.__line_number)
                    )

            elif self.__token.lexeme in self.__symbol_table:
                # Simple variable must be processed
                self.__operand_stack.append(self.__sign*self.__symbol_table[self.__token.lexeme])

            else:
                # default variables values for undefined variables.
                if self.__token.lexeme.endswith ("$"):
                    # default string
                    self.__operand_stack.append("")
                else:
                    #default int
                    self.__operand_stack.append(0)

            self.__advance()

        elif self.__token.category == Token.LEFTPAREN:
            self.__advance()

            # Save sign because expr() calls term() which resets
            # sign to 1
            savesign = self.__sign
            self.__logexpr()  # Value of expr is pushed onto stack

            if savesign == -1:
                # Change sign of expression
                self.__operand_stack[-1] = -self.__operand_stack[-1]

            self.__consume(Token.RIGHTPAREN)

        elif self.__token.category in Token.functions:
            self.__operand_stack.append(self.__evaluate_function(self.__token.category))

        else:
            raise RuntimeError('Expecting factor in numeric expression' +
                               ' in line ' + str(self.__line_number) +
                               self.__token.lexeme)

    def __get_array_val(self, BASICarray, indexvars):
        """Extracts the value from the given BASICArray at the specified indexes

        :param BASICarray: The BASICArray
        :param indexvars: The list of indexes, one for each dimension

        :return: The value at the indexed position in the array

        """
        self.__validate_array_indices(BASICarray, indexvars)

        # Fetch the value from the array
        try:
            if len(indexvars) == 1:
                arrayval = BASICarray.data[indexvars[0]]

            elif len(indexvars) == 2:
                arrayval = BASICarray.data[indexvars[0]][indexvars[1]]

            elif len(indexvars) == 3:
                arrayval = BASICarray.data[indexvars[0]][indexvars[1]][indexvars[2]]

        except IndexError:
            raise RuntimeError('SUBSCRIPT ERROR in line ' + str(self.__line_number))

        return arrayval

    def __validate_array_indices(self, BASICarray, indexvars):
        """Validates array indices and raises SUBSCRIPT ERROR if invalid

        :param BASICarray: The BASICArray to validate against
        :param indexvars: List of index values to validate
        :raises RuntimeError: If indices are out of bounds
        """
        if BASICarray.dims != len(indexvars):
            raise RuntimeError('SUBSCRIPT ERROR in line ' + str(self.__line_number))

        for i, index in enumerate(indexvars):
            # Convert float to int (truncate toward zero)
            if isinstance(index, float):
                index = int(index)
                indexvars[i] = index

            # Check bounds - negative or greater than declared dimension
            if index < 0 or index > BASICarray.original_dims[i]:
                raise RuntimeError('SUBSCRIPT ERROR in line ' + str(self.__line_number))

    def __assign_array_val(self, BASICarray, indexvars, value):
        """Assigns a value to an array element after validating indices

        :param BASICarray: The BASICArray
        :param indexvars: List of indices
        :param value: Value to assign
        """
        self.__validate_array_indices(BASICarray, indexvars)

        try:
            if len(indexvars) == 1:
                BASICarray.data[indexvars[0]] = value
            elif len(indexvars) == 2:
                BASICarray.data[indexvars[0]][indexvars[1]] = value
            elif len(indexvars) == 3:
                BASICarray.data[indexvars[0]][indexvars[1]][indexvars[2]] = value
        except IndexError:
            raise RuntimeError('SUBSCRIPT ERROR in line ' + str(self.__line_number))

    def __parse_variable_target(self):
        """Parses a variable target which can be a simple variable or array element

        :return: Dictionary with 'type' ('simple' or 'array'), 'name' (variable name),
                 'indices' (list of index values for arrays), 'is_string' (boolean)
        """
        if self.__token.category != Token.NAME:
            raise SyntaxError('Expecting variable name in line ' + str(self.__line_number))

        var_name = self.__token.lexeme
        is_string = var_name.endswith('$')
        self.__advance()

        # Check if this is an array element
        if self.__token.category == Token.LEFTPAREN:
            # Array element target
            array_name = var_name + '_array'

            # Check if array exists
            if array_name not in self.__symbol_table:
                raise RuntimeError('Array not dimensioned in line ' + str(self.__line_number))

            self.__advance()  # Past LEFTPAREN

            # Parse index expressions
            indexvars = []
            self.__expr()
            indexvars.append(self.__operand_stack.pop())

            while self.__token.category == Token.COMMA:
                self.__advance()  # Past comma
                self.__expr()
                indexvars.append(self.__operand_stack.pop())

            self.__consume(Token.RIGHTPAREN)

            return {
                'type': 'array',
                'name': var_name,
                'array_name': array_name,
                'indices': indexvars,
                'is_string': is_string
            }
        else:
            # Simple variable target
            return {
                'type': 'simple',
                'name': var_name,
                'is_string': is_string
            }

    def __assign_to_target(self, target, value):
        """Assigns a value to a target (simple variable or array element)

        :param target: Target dictionary from __parse_variable_target
        :param value: Value to assign
        """
        # Type checking
        if target['is_string'] and not isinstance(value, str):
            raise ValueError('Non-string input provided to a string variable in line ' +
                           str(self.__line_number))
        elif not target['is_string'] and isinstance(value, str):
            raise ValueError('Non-numeric input provided to a numeric variable in line ' +
                           str(self.__line_number))

        if target['type'] == 'simple':
            self.__symbol_table[target['name']] = value
        else:  # array
            BASICarray = self.__symbol_table[target['array_name']]
            self.__assign_array_val(BASICarray, target['indices'], value)

    def __compoundstmt(self):
        """Parses compound statements,
        specifically if-then-else and
        loops

        :return: The FlowSignal to indicate to the program
        how to branch if necessary, None otherwise

        """
        if self.__token.category == Token.FOR:
            return self.__forstmt()

        elif self.__token.category == Token.NEXT:
            return self.__nextstmt()

        elif self.__token.category == Token.IF:
            return self.__ifstmt()

        elif self.__token.category == Token.ON:
            return self.__ongosubstmt()

        elif self.__token.category == Token.WHILE:
            return self.__whilestmt()

        elif self.__token.category == Token.WEND:
            return self.__wendstmt()

    def __ifstmt(self):
        """Parses if-then-else
        statements

        :return: The FlowSignal to indicate to the program
        how to branch if necessary, None otherwise

        """

        self.__advance()  # Advance past IF token
        self.__logexpr()

        # Save result of expression
        saveval = self.__operand_stack.pop()

        # Process the THEN part and save the jump value
        self.__consume(Token.THEN)

        if self.__token.category != Token.UNSIGNEDINT:
            if saveval:
                return FlowSignal(ftype=FlowSignal.EXECUTE)
        else:
            self.__expr()
            target = self.__operand_stack.pop()

            # Jump if the expression evaluated to True
            if saveval:
                # Set up and return the flow signal
                return FlowSignal(ftarget=target)

        # advance to ELSE
        while self.__tokenindex < len(self.__tokenlist) and self.__token.category != Token.ELSE:
            self.__advance()

        # See if there is an ELSE part
        if self.__token.category == Token.ELSE:
            self.__advance()

            if self.__token.category != Token.UNSIGNEDINT:
                return FlowSignal(ftype=FlowSignal.EXECUTE)
            else:

                self.__expr()

                # Set up and return the flow signal
                return FlowSignal(ftarget=self.__operand_stack.pop())

        else:
            # No ELSE action
            return None

    def __forstmt(self):
        """Parses for loops

        :return: The FlowSignal to indicate that
        a loop start has been processed

        """

        # Set up default loop increment value
        step = 1

        self.__advance()  # Advance past FOR token

        # Process the loop variable initialisation
        loop_variable = self.__token.lexeme  # Save lexeme of
                                             # the current token

        if loop_variable.endswith('$'):
            raise SyntaxError('Syntax error: Loop variable is not numeric' +
                              ' in line ' + str(self.__line_number))

        self.__advance()  # Advance past loop variable
        self.__consume(Token.ASSIGNOP)
        self.__expr()

        # Check that we are using the right variable name format
        # for numeric variables
        start_val = self.__operand_stack.pop()

        # Advance past the 'TO' keyword
        self.__consume(Token.TO)

        # Process the terminating value
        self.__expr()
        end_val = self.__operand_stack.pop()

        # Check if there is a STEP value
        increment = True
        if not self.__tokenindex >= len(self.__tokenlist):
            self.__consume(Token.STEP)

            # Acquire the step value
            self.__expr()
            step = self.__operand_stack.pop()

            # Check whether we are decrementing or
            # incrementing
            if step == 0:
                raise IndexError('Zero step value supplied for loop' +
                                 ' in line ' + str(self.__line_number))

            elif step < 0:
                increment = False

        # Now determine the status of the loop

        # Note that we cannot use the presence of the loop variable in
        # the symbol table for this test, as the same variable may already
        # have been instantiated elsewhere in the program
        #
        # Need to initialize the loop variable anytime the for
        # statement is reached from a statement other than an active NEXT.

        from_next = False
        if self.last_flowsignal:
            if self.last_flowsignal.ftype == FlowSignal.LOOP_REPEAT:
                from_next = True

        if not from_next:
            self.__symbol_table[loop_variable] = start_val

        else:
            # We need to modify the loop variable
            # according to the STEP value
            self.__symbol_table[loop_variable] += step

        # If the loop variable has reached the end value,
        # remove it from the set of extant loop variables to signal that
        # this is the last loop iteration
        stop = False
        if increment and self.__symbol_table[loop_variable] > end_val:
            stop = True

        elif not increment and self.__symbol_table[loop_variable] < end_val:
            stop = True

        if stop:
            # Loop must terminate
            return FlowSignal(ftype=FlowSignal.LOOP_SKIP,
                              ftarget=loop_variable)
        else:
            # Set up and return the flow signal
            return FlowSignal(ftype=FlowSignal.LOOP_BEGIN,floop_var=loop_variable)

    def __nextstmt(self):
        """Processes a NEXT statement that terminates
        a loop

        :return: A FlowSignal indicating that a loop
        has been processed

        """

        self.__advance()  # Advance past NEXT token

        # Process the loop variable initialisation
        loop_variable = self.__token.lexeme  # Save lexeme of
                                             # the current token

        if loop_variable.endswith('$'):
            raise SyntaxError('Syntax error: Loop variable is not numeric' +
                              ' in line ' + str(self.__line_number))

        return FlowSignal(ftype=FlowSignal.LOOP_REPEAT,floop_var=loop_variable)

    def __whilestmt(self):
        """Parses WHILE loops

        :return: The FlowSignal to indicate that
        a WHILE loop start has been processed

        """

        self.__advance()  # Advance past WHILE token
        self.__logexpr()  # Evaluate the condition

        # Save result of expression
        condition_result = self.__operand_stack.pop()

        # Check if the condition is true
        if condition_result:
            # Condition is true, execute the loop body
            return FlowSignal(ftype=FlowSignal.WHILE_BEGIN)
        else:
            # Condition is false, skip the loop
            return FlowSignal(ftype=FlowSignal.WHILE_SKIP)

    def __wendstmt(self):
        """Processes a WEND statement that terminates
        a WHILE loop

        :return: A FlowSignal indicating that the loop
        should repeat (return to WHILE)

        """

        self.__advance()  # Advance past WEND token

        # Return signal to repeat the WHILE loop
        return FlowSignal(ftype=FlowSignal.WHILE_REPEAT)

    def __ongosubstmt(self):
        """Process the ON-GOSUB statement

        :return: A FlowSignal indicating the subroutine line number
        if the condition is true, None otherwise

        """

        self.__advance()  # Advance past ON token
        self.__expr()

        # Save result of expression
        saveval = self.__operand_stack.pop()

        if self.__token.category == Token.GOTO:
            self.__consume(Token.GOTO)
            branchtype = 1
        else:
            self.__consume(Token.GOSUB)
            branchtype = 2

        branch_values = []
        # Acquire the comma separated values
        if not self.__tokenindex >= len(self.__tokenlist):
            self.__expr()
            branch_values.append(self.__operand_stack.pop())

            while self.__token.category == Token.COMMA:
                self.__advance()  # Advance past comma
                self.__expr()
                branch_values.append(self.__operand_stack.pop())

        if saveval < 1 or saveval > len(branch_values) or len(branch_values) == 0:
            return None
        elif branchtype == 1:
            return FlowSignal(ftarget=branch_values[saveval-1])
        else:
            return FlowSignal(ftarget=branch_values[saveval-1],
                              ftype=FlowSignal.GOSUB)

    def __relexpr(self):
        """Parses a relational expression
        """
        self.__expr()

        # Since BASIC uses same operator for both
        # assignment and equality, we need to check for this
        if self.__token.category == Token.ASSIGNOP:
            self.__token.category = Token.EQUAL

        if self.__token.category in [Token.LESSER, Token.LESSEQUAL,
                              Token.GREATER, Token.GREATEQUAL,
                              Token.EQUAL, Token.NOTEQUAL]:
            savecat = self.__token.category
            self.__advance()
            self.__expr()

            right = self.__operand_stack.pop()
            left = self.__operand_stack.pop()

            if savecat == Token.EQUAL:
                self.__operand_stack.append(left == right)  # Push True or False

            elif savecat == Token.NOTEQUAL:
                self.__operand_stack.append(left != right)  # Push True or False

            elif savecat == Token.LESSER:
                self.__operand_stack.append(left < right)  # Push True or False

            elif savecat == Token.GREATER:
                self.__operand_stack.append(left > right)  # Push True or False

            elif savecat == Token.LESSEQUAL:
                self.__operand_stack.append(left <= right)  # Push True or False

            elif savecat == Token.GREATEQUAL:
                self.__operand_stack.append(left >= right)  # Push True or False

    def __logexpr(self):
        """Parses a logical expression
        """
        self.__notexpr()

        while self.__token.category in [Token.OR, Token.AND]:
            savecat = self.__token.category
            self.__advance()
            self.__notexpr()

            right = self.__operand_stack.pop()
            left = self.__operand_stack.pop()

            if savecat == Token.OR:
                self.__operand_stack.append(left or right)  # Push True or False

            elif savecat == Token.AND:
                self.__operand_stack.append(left and right)  # Push True or False

    def __notexpr(self):
        """Parses a logical not expression
        """
        if self.__token.category == Token.NOT:
            self.__advance()
            self.__relexpr()
            right = self.__operand_stack.pop()
            self.__operand_stack.append(not right)
        else:
            self.__relexpr()

    def __evaluate_function(self, category):
        """Evaluate the function in the statement
        and return the result.

        :return: The result of the function

        """

        self.__advance()  # Advance past function name

        # Process arguments according to function
        if category == Token.RND:
            self.__consume(Token.LEFTPAREN)

            self.__expr()
            arg = self.__operand_stack.pop()

            self.__consume(Token.RIGHTPAREN)
            # MSBASIC basic reseeds with negative values
            # as arg to RND... not sure if it returned anything
            # Zero returns the last value again (not implemented)
            # Any positive value returns random fload btw 0 and 1
            if arg < 0:
                random.seed(arg)

            return random.random()

        if category == Token.PI:
            return math.pi

        if category == Token.RNDINT:
            self.__consume(Token.LEFTPAREN)

            self.__expr()
            lo = self.__operand_stack.pop()

            self.__consume(Token.COMMA)

            self.__expr()
            hi = self.__operand_stack.pop()

            self.__consume(Token.RIGHTPAREN)

            try:
                return random.randint(lo, hi)

            except ValueError:
                raise ValueError("Invalid value supplied to RNDINT in line " +
                                 str(self.__line_number))

        if category == Token.MAX:
            self.__consume(Token.LEFTPAREN)

            self.__expr()
            value_list = [self.__operand_stack.pop()]

            while self.__token.category == Token.COMMA:
                self.__advance() # Advance past comma
                self.__expr()
                value_list.append(self.__operand_stack.pop())

            self.__consume(Token.RIGHTPAREN)

            try:
                return max(*value_list)

            except TypeError:
                raise TypeError("Invalid type supplied to MAX in line " +
                                 str(self.__line_number))

        if category == Token.MIN:
            self.__consume(Token.LEFTPAREN)

            self.__expr()
            value_list = [self.__operand_stack.pop()]

            while self.__token.category == Token.COMMA:
                self.__advance() # Advance past comma
                self.__expr()
                value_list.append(self.__operand_stack.pop())

            self.__consume(Token.RIGHTPAREN)

            try:
                return min(*value_list)

            except TypeError:
                raise TypeError("Invalid type supplied to MIN in line " +
                                 str(self.__line_number))

        if category == Token.POW:
            self.__consume(Token.LEFTPAREN)

            self.__expr()
            base = self.__operand_stack.pop()

            self.__consume(Token.COMMA)

            self.__expr()
            exponent = self.__operand_stack.pop()

            self.__consume(Token.RIGHTPAREN)

            try:
                return math.pow(base, exponent)

            except ValueError:
                raise ValueError("Invalid value supplied to POW in line " +
                                 str(self.__line_number))

        if category == Token.TERNARY:
            self.__consume(Token.LEFTPAREN)

            self.__logexpr()
            condition = self.__operand_stack.pop()

            self.__consume(Token.COMMA)

            self.__expr()
            whentrue = self.__operand_stack.pop()

            self.__consume(Token.COMMA)

            self.__expr()
            whenfalse = self.__operand_stack.pop()

            self.__consume(Token.RIGHTPAREN)

            return whentrue if condition else whenfalse

        if category == Token.LEFT:
            self.__consume(Token.LEFTPAREN)

            self.__expr()
            instring = self.__operand_stack.pop()

            self.__consume(Token.COMMA)

            self.__expr()
            chars = self.__operand_stack.pop()

            self.__consume(Token.RIGHTPAREN)

            try:
                return instring[:chars]

            except TypeError:
                raise TypeError("Invalid type supplied to LEFT$ in line " +
                                 str(self.__line_number))

        if category == Token.RIGHT:
            self.__consume(Token.LEFTPAREN)

            self.__expr()
            instring = self.__operand_stack.pop()

            self.__consume(Token.COMMA)

            self.__expr()
            chars = self.__operand_stack.pop()

            self.__consume(Token.RIGHTPAREN)

            try:
                return instring[-chars:]

            except TypeError:
                raise TypeError("Invalid type supplied to RIGHT$ in line " +
                                 str(self.__line_number))

        if category == Token.MID:
            self.__consume(Token.LEFTPAREN)

            self.__expr()
            instring = self.__operand_stack.pop()

            self.__consume(Token.COMMA)

            self.__expr()
            # Older basic dialets were always 1 based
            start = self.__operand_stack.pop() - 1

            if self.__token.category == Token.COMMA:
                self.__advance() # Advance past comma
                self.__expr()
                chars = self.__operand_stack.pop()
            else:
                chars = None

            self.__consume(Token.RIGHTPAREN)

            try:
                if chars:
                    return instring[start:start+chars]
                else:
                    return instring[start:]

            except TypeError:
                raise TypeError("Invalid type supplied to MID$ in line " +
                                 str(self.__line_number))

        if category == Token.INSTR:
            self.__consume(Token.LEFTPAREN)

            self.__expr()
            hackstackstring = self.__operand_stack.pop()
            if not isinstance(hackstackstring, str):
                raise TypeError("Invalid type supplied to INSTR in line " +
                                 str(self.__line_number))

            self.__consume(Token.COMMA)

            self.__expr()
            needlestring = self.__operand_stack.pop()

            start = end = None
            if self.__token.category == Token.COMMA:
                self.__advance() # Advance past comma
                self.__expr()
                # Older basic dialets were always 1 based
                start = self.__operand_stack.pop() -1

                if self.__token.category == Token.COMMA:
                    self.__advance() # Advance past comma
                    self.__expr()
                    end = self.__operand_stack.pop() -1

            self.__consume(Token.RIGHTPAREN)

            try:
                # Older basis dialets are 1 based, so the return value
                # here needs to be incremented by one.  ALSO
                # this moves the -1 not found value to 0
                # which indicated not found in most dialects
                return hackstackstring.find(needlestring, start, end) + 1

            except TypeError:
                raise TypeError("Invalid type supplied to INSTR in line " +
                                 str(self.__line_number))

        self.__consume(Token.LEFTPAREN)

        self.__expr()
        value = self.__operand_stack.pop()

        self.__consume(Token.RIGHTPAREN)

        if category == Token.SQR:
            try:
                return math.sqrt(value)

            except ValueError:
                raise ValueError("Invalid value supplied to SQR in line " +
                                 str(self.__line_number))

        elif category == Token.ABS:
            try:
                return abs(value)

            except ValueError:
                raise ValueError("Invalid value supplied to ABS in line " +
                                 str(self.__line_number))

        elif category == Token.ATN:
            try:
                return math.atan(value)

            except ValueError:
                raise ValueError("Invalid value supplied to ATN in line " +
                                 str(self.__line_number))

        elif category == Token.COS:
            try:
                return math.cos(value)

            except ValueError:
                raise ValueError("Invalid value supplied to COS in line " +
                                 str(self.__line_number))

        elif category == Token.EXP:
            try:
                return math.exp(value)

            except ValueError:
                raise ValueError("Invalid value supplied to EXP in line " +
                                 str(self.__line_number))

        elif category == Token.INT:
            try:
                return math.floor(value)

            except ValueError:
                raise ValueError("Invalid value supplied to INT in line " +
                                 str(self.__line_number))

        elif category == Token.ROUND:
            try:
                return round(value)

            except TypeError:
                raise TypeError("Invalid type supplied to LEN in line " +
                                 str(self.__line_number))

        elif category == Token.LOG:
            try:
                return math.log(value)

            except ValueError:
                raise ValueError("Invalid value supplied to LOG in line " +
                                 str(self.__line_number))

        elif category == Token.SIN:
            try:
                return math.sin(value)

            except ValueError:
                raise ValueError("Invalid value supplied to SIN in line " +
                                 str(self.__line_number))

        elif category == Token.TAN:
            try:
                return math.tan(value)

            except ValueError:
                raise ValueError("Invalid value supplied to TAN in line " +
                                 str(self.__line_number))

        elif category == Token.CHR:
            try:
                return chr(value)

            except TypeError:
                raise TypeError("Invalid type supplied to CHR$ in line " +
                                 str(self.__line_number))

            except ValueError:
                raise ValueError("Invalid value supplied to CHR$ in line " +
                                 str(self.__line_number))

        elif category == Token.ASC:
            try:
                return ord(value)

            except TypeError:
                raise TypeError("Invalid type supplied to ASC in line " +
                                 str(self.__line_number))

            except ValueError:
                raise ValueError("Invalid value supplied to ASC in line " +
                                 str(self.__line_number))

        elif category == Token.STR:
            return str(value)

        elif category == Token.VAL:
            try:
                numeric = float(value)
                if int(numeric) == numeric:
                    return int(numeric)
                return numeric

            # Like other BASIC variants, non-numeric strings return 0
            except ValueError:
                return 0

        elif category == Token.LEN:
            try:
                return len(value)

            except TypeError:
                raise TypeError("Invalid type supplied to LEN in line " +
                                 str(self.__line_number))

        elif category == Token.UPPER:
            if not isinstance(value, str):
                raise TypeError("Invalid type supplied to UPPER$ in line " +
                                 str(self.__line_number))

            return value.upper()

        elif category == Token.LOWER:
            if not isinstance(value, str):
                raise TypeError("Invalid type supplied to LOWER$ in line " +
                                 str(self.__line_number))

            return value.lower()

        elif category == Token.TAB:
            if isinstance(value, int):
                return " "*value

            else:
                raise TypeError("Invalid type supplied to TAB in line " +
                                 str(self.__line_number))

        else:
            raise SyntaxError("Unrecognised function in line " +
                              str(self.__line_number))

    def __randomizestmt(self):
        """Implements a function to seed the random
        number generator

        """
        self.__advance()  # Advance past RANDOMIZE token

        if not self.__tokenindex >= len(self.__tokenlist):
            self.__expr()  # Process the seed
            seed = self.__operand_stack.pop()

            random.seed(seed)

        else:
            random.seed(int(monotonic()))