Click here to Skip to main content
15,395,897 members
Articles / General Programming / Parser
Article
Posted 25 Feb 2022

Stats

7.3K views
191 downloads
6 bookmarked

Math Expression Evaluator for VBA

Rate me:
Please Sign up or sign in to vote.
5.00/5 (3 votes)
1 Mar 2022GPL36 min read
A class module developed to evaluate mathematical expressions given as VBA strings
The VBAexpressions class serves as an intermediary between user interfaces and the main VBA/custom functions exposed through it. The main objective of the development of the class is to demonstrate that all mathematical expressions can be evaluated computationally using an approach similar to the one we humans use: divide the function into sub-expressions, create a symbolic string (lambda) to draw the evaluation flow of the expression, split the sub-expressions into chunks of operations (tokens) by tokenization, evaluate all the tokens.

Introduction

The natural evolution of CSV Interface required the integration of a kind of mathematical expressions evaluator that would make it possible to filter CSV files with a considerable number of records. Nevertheless, this peculiar utility is little developed in VBA language, existing alternatives not very flexible and of difficult integration, as well as others unable to evaluate correctly some corner cases. This limitation put me on the way to develop an evaluator that meets the necessary requirements and at the same time can be integrated with a minimum amount of code in the target library.

Pursued Goals

The utility is required to meet the following requirements:

  1. The tool must be able to receive a text string as input and return the result obtained by executing the arithmetic, logical and binary relations operations contained in the expression.
  2. The user must be able to define the variables and decide the value used in each evaluation after parsing the expression.
  3. Built-in functions must be provided.
  4. The user must be able to register, at run time, and use his own functions without the need to append code to the developed tool.
  5. The variables found in each parsed expression, as well as the values relative to them, must be exposed to the user.
  6. The tool should avoid the use of hard-coded expressions, so variable values should be assigned as text strings.

The Road Way

The traditional way to evaluate mathematical expressions involves the conversion of infix expressions (operands separated by an operator) to postfix. In this way, the evaluation is performed using stacks, or parsed tree, which ensure the precedence of the operators defined in the mathematical expressions.

However, it has been decided to develop VBA Expressions as a tool that is able to evaluate infix expressions directly, although the difficulty is overwhelming, in order to demonstrate that the way humans evaluate an expression is adequate and relatively efficient. The analysis process, prior to evaluation, consists of the creation of an evaluation tree through the following steps:

  1. The expression is subdivided into related subexpressions using mathematical operators.
  2. Each subexpression is divided into tokens, each of which is defined by 2 arguments related by a mathematical operator.
  3. A token is created and stored for each subexpression.
  4. Each token is disaggregated into arguments, the lowest level of logic used, represented by a list, a variable or an operand.
  5. Once all tokens are parsed, the evaluation tree is complete.

The evaluation process follows the following path:

  1. The parser uses the first token in the evaluation tree as the entry point.
  2. All arguments of the token are then evaluated.
  3. The result of the token evaluation is saved.
  4. Evaluation continues until all tokens are evaluated.
  5. The evaluation result always corresponds to the last token evaluation.

The VBA Expressions methodology uses parentheses as entry points to tokenize expressions, so the tool allows some interesting hacks among which we can mention the support of arrays defined as text strings with a syntax similar to the one used by Java. Thus, parentheses are characters exclusively reserved for grouping subexpressions that must be evaluated as individual tokens, thus providing a way to evaluate highly complex expressions.

The grammar defining the expressions supported by the utility is shown below:

Expression    =     ([{"("}]  SubExpr [{Operator [{"("}] SubExpr 
                    [{")"}]}] [{")"}] | {["("] ["{"] List [{";" List}] ["}"] [")"]}
SubExpr       =     Token [{Operator Token}]
Token         =     [{Unary}] Argument [(Operator | Function) ["("] [{Unary}] [Argument] [")"]]
Argument      =     (List | Variable | Operand)
List          =     "{" ["{"] SubExpr [{";" SubExpr}] ["}"] "}"
Unary         =     "-" | "+" | ~
Operand       =     ({Digit} ["."] [{Digit}] ["E"("-" | "+"){Digit}] | (True | False))
Variable      =     Alphabet [{Decimal}] [{(Digit | Alphabet)}]
Alphabet      =     "A-Z" | "a-z"
Decimal       =     "."
Digit         =     "0-9"
Operator      =     "+" | "-" | "*" | "/" | "\" | "^" | "%" | "!" | "<" | "<=" | 
                    "<>" | ">" | ">=" | "=" | "&" | "|" | "||"
Function      =     "abs" | "sin" | "cos" | "min" |...|[UDF]

The precedence of the supported operators, as well as the operators themselves, are hard coded in the utility, so, so far, it is not possible to add new operators. The utility will prioritize the evaluation of operators according to the PEMDAS methodology:

1. ()               Grouping: evaluates functions arguments as well.
2. ! - +            Unary operators: exponentiation is the only operation that violates this. 
                    Ex.: `-2 ^ 2 = -4 | (-2)^ 2 = 4.
3. ^                Exponentiation: Although Excel and Matlab evaluate nested exponentiations 
                    from left to right, Google, mathematicians and several modern 
                    programming languages, such as Perl, Python and Ruby, evaluate this 
                    operation from right to left. VBA expressions also evals in Python way: 
                    a^b^c = a^(b^c).
4. \* / %           Multiplication, division, modulo: from left to right.
5. + -              Addition and subtraction: from left to right.
6. < <= <> >= = >   Binary relations.
7. ~                Logical negation.
8. &                Logical AND.
9. ||               Logical XOR.
10. |               Logical OR.

It should be noted that VBA does not allow the use of code reflection, so extensibility is limited to the CallByName function. In order to limit the use of hard-coded arguments, user-defined functions must accept a single variant type parameter, the function will receive a one-dimensional array of text strings with as many elements as arguments have been found when parsing the given expression.

The Code

The heart of the evaluations carried out by VBA Expressions lies in the Parse method, which is responsible for generating the evaluation tree. A fundamental aspect of this process is the differentiation between simple and compound expressions. A simple expression, in the eyes of the tool, is one that does not make use of parentheses and results in a single evaluation tree. For compound expressions, it is necessary to create as many evaluation trees as subexpressions are present in the expression. Here the code:

VB.NET
Private Sub Parse(ByRef Expression As String)
    Dim lambdaText As String
    Dim meLB As Long
    Dim meUB As Long
    Dim meCounter As Long
    Dim SimpleExpr As Boolean
    Dim TreeUB As Long
    Dim LbrCount As Long
    Dim RbrCount As Long
    
    On Error GoTo Parse_errHandler
    InitializeErrHandler
    LbrCount = CountParentheses(Expression, d_lParenthesis)
    RbrCount = CountParentheses(Expression, d_rParenthesis)
    If LbrCount <> RbrCount Then
        Exit Sub
    End If
    err.Clear
    SubTreeData() = GetSubTreeData(Expression)
    lambdaText = SerializeSubTree(Expression, SubTreeData)
    meLB = LBound(SubTreeData)
    meUB = UBound(SubTreeData)
    SimpleExpr = (meUB - meLB = 0 And lambdaText = SubTreeData(meUB))
    TreeUB = meUB + Abs(CLng(Not SimpleExpr)) 'Upper bound for ClusterTree array
    
    ReDim EvalTree(meLB To TreeUB)
    For meCounter = meLB To TreeUB
        InitBuffer EvalTree(meCounter)        'Reserve storage for tree branches/Expressions
        If meCounter < TreeUB Then            'Tokenize branches
            TokenizeSubExpr SubTreeData(meCounter), SubTreeData, EvalTree(meCounter)
        Else
            If Not SimpleExpr Then            'Tokenize main tree/lambda expression
                TokenizeSubExpr lambdaText, SubTreeData, EvalTree(meCounter)
            Else
                TokenizeSubExpr SubTreeData(meCounter), SubTreeData, EvalTree(meCounter)
            End If
        End If
        ShrinkBuffer EvalTree(meCounter)
    Next meCounter
    If ValidTree Then    'Call the validate function
        GeneratedTree = True
    Else
        GeneratedTree = False
    End If
End Sub

Each subexpression is tokenized or subdivided into its fundamental components using the TokenizeSubExpr method. Here, the differentiation is made between regular subexpressions, argument lists and arrays.

VB.NET
Private Sub TokenizeSubExpr(ByRef Expression As String, _
        ByRef SubExpressionsData() As String, ByRef outBuffer As ClusterTree)
    Dim tmpReplacement As String
    Dim ExpCopy As String
    Dim tmpArgs() As String
    Dim taIcounter As Long
    Dim OperationIndex As Long
    Dim tmpIndex As Long
    
    tmpIndex = UBound(SubExpressionsData) + 1
    OperationIndex = tmpIndex
    ExpCopy = Expression
    tmpReplacement = GetSubstStr(OperationIndex)
    If Not ExpCopy Like "*{{*}}*" Then
        Select Case InStrB(1, ExpCopy, P_SEPARATORCHAR)
            Case 0 'Regular sub-expression
                GetRootedTree ExpCopy, tmpReplacement, OperationIndex, outBuffer
                outBuffer.CompCluster = False
            Case Else 'Composite function argument
                tmpArgs() = Split(ExpCopy, P_SEPARATORCHAR)
                For taIcounter = LBound(tmpArgs) To UBound(tmpArgs)
                    GetRootedTree tmpArgs(taIcounter), tmpReplacement, _
                                                       OperationIndex, outBuffer
                Next taIcounter
                outBuffer.CompCluster = True
                outBuffer.CompArrCluster = False
        End Select
    Else    'Composite array function argument
        outBuffer.ClusterArrBounds = SplitArrBranch(ExpCopy, tmpArgs)
        If outBuffer.ClusterArrBounds(0) <> -1 Then    'Splitting argument success
            For taIcounter = LBound(tmpArgs) To UBound(tmpArgs)
                GetRootedTree tmpArgs(taIcounter), tmpReplacement, OperationIndex, outBuffer
            Next taIcounter
            outBuffer.CompCluster = True
            outBuffer.CompArrCluster = True
        Else
            'Todo: Code here for trap error of missing () in a composite array 
            'and standard input
        End If
    End If
End Sub

The GetRootedTree method extracts all binary operations on the tokens until they are reduced to a single argument.

VB.NET
Private Sub GetRootedTree(ByRef SubExpression As String, ByRef tmpReplacement As String, _
                          ByRef OperationIndex As Long, ByRef outBuffer As ClusterTree)
    Dim vToken As Token
    Dim switch As Boolean
    Dim tmpPos As Long
    Dim OperandInBundle As Boolean
    Dim PrevChar As String
    
    Do
        SubExpression = ApplyLawOfSigns(SubExpression)
        vToken = GetEvalToken(SubExpression)
        '@--------------------------------------------------------------------
        ' Mask worked token
        tmpPos = InStrB(1, SubExpression, vToken.DefString)
        If tmpPos > 2 Then
            PrevChar = MidB$(SubExpression, tmpPos - 2, 2)
            OperandInBundle = (InStrB(1, op_AllItems, PrevChar))
            Do While Not OperandInBundle And tmpPos > 2    'Tokens starts with a operator 
                                                           'or with a null string
                tmpPos = InStrB(tmpPos + 2, SubExpression, vToken.DefString)
                PrevChar = MidB$(SubExpression, tmpPos - 2, 2)
                OperandInBundle = (InStrB(1, op_AllItems, PrevChar))
            Loop
        End If
        SubExpression = MidB$(SubExpression, 1, tmpPos - 1) & _
           tmpReplacement & MidB$(SubExpression, tmpPos + LenB(vToken.DefString))
        AppendToBuffer outBuffer, vToken 'Save to target token ClusterTree
        switch = (SubExpression <> tmpReplacement)
        If switch Then
            OperationIndex = OperationIndex + 1
            tmpReplacement = GetSubstStr(OperationIndex)
        End If
    Loop While switch
End Sub

When these processes finish without any alert, the result obtained is an evaluation tree ready to be processed. It is at this point where the Compute method comes into play, which is responsible for processing the evaluation tree and returning a result. The process to evaluate is relatively simple code, the difficulty lies in the relationship between results of those evaluation trees related to each other, in which case it is necessary to link previous results with other operations to compute new results.

To make the linking of results possible and feasible, the variable BaseIndex was devised, which indicates whether an argument is related to a branch of the same evaluation tree and/or to results obtained when evaluating another tree. The Compute method runs through the evaluation tree of each token, extracting its operands and saving the result after evaluating each one. The result obtained after evaluating arrays (with m rows of n columns) and lists are stored in the format in which they were entered.

VB.NET
Private Function Compute() As String
    Dim B As Long
    Dim t As Long
    Dim i As Long
    Dim OperationIndex As Long
    Dim BaseIndex As Long
    Dim PrevOP1 As String
    Dim PrevOP2 As String
    
    BaseIndex = UBound(SubTreeData) + 1
    For B = LBound(EvalTree) To UBound(EvalTree)              'Loop all subexpressions
        OperationIndex = BaseIndex
        For t = 0 To EvalTree(B).Index                        'Loop all tokens
            OperationIndex = OperationIndex + 1
            If Not EvalTree(B).Storage(t).ConstantToken Then  'Gallop
                If P_GALLOPING_MODE Then
                    PrevOP1 = EvalTree(B).Storage(t).Arg1.Operand
                    PrevOP2 = EvalTree(B).Storage(t).Arg2.Operand
                End If
                GetOperands EvalTree(B).Storage(t), EvalTree(B), BaseIndex
                BottomLevelEval EvalTree(B).Storage(t)
                If P_GALLOPING_MODE Then
                    EvalTree(B).Storage(t).ConstantToken = _
                                (PrevOP1 = EvalTree(B).Storage(t).Arg1.Operand And _
                                 PrevOP2 = EvalTree(B).Storage(t).Arg2.Operand)
                End If
            End If
        Next t
        If Not EvalTree(B).CompCluster Then
            EvalTree(B).EvalResult = EvalTree(B).Storage(t - 1).EvalResult
        Else            'The ClusterTree contains a composite function args as expression
            Dim tmpResult() As String
            ReDim tmpResult(0 To EvalTree(B).Index)
            '@--------------------------------------------------------------------
            ' Loop all sub-expression tokens results
            For i = 0 To EvalTree(B).Index
                tmpResult(i) = EvalTree(B).Storage(i).EvalResult
            Next i
            If Not EvalTree(B).CompArrCluster Then   'Function Argument
                EvalTree(B).EvalResult = Join$(tmpResult, P_SEPARATORCHAR)
            Else        'Array function Argument
                EvalTree(B).EvalResult = JoinArrFunctArg_
                (tmpResult, EvalTree(B).ClusterArrBounds(0), EvalTree(B).ClusterArrBounds(1))
            End If
        End If
    Next B
    Compute = EvalTree(B - 1).EvalResult
    ComputedTree = True
End Function

The operands are obtained through the GetOperands and GetOperand methods.

VB.NET
Private Sub GetOperands(ByRef CurToken As Token, ByRef CurTree As ClusterTree, _
                                                 ByRef BaseIndex As Long)
                                    
    '@--------------------------------------------------------------------
    ' Get the first operand
    GetOperand CurToken, CurToken.Arg1, CurTree, BaseIndex
    '@--------------------------------------------------------------------
    ' Get the second operand if required
    If Not CurToken.OperationToken = otNull Then
        GetOperand CurToken, CurToken.Arg2, CurTree, BaseIndex
    End If
End Sub

Private Sub GetOperand(ByRef CurToken As Token, ByRef CurArg As Argument, _
                        ByRef CurTree As ClusterTree, ByRef BaseIndex As Long)
                                    
    If CurArg.Implicit Then
        If CurArg.FunctionIn Then 'Implicit function
            If CurArg.FactorialIn Then
                If CurArg.LinkedIndex >= BaseIndex Then 'Data on current tree
                    CurArg.Operand = Factorial(EvalFunction_
                    (CurTree.Storage(CurArg.LinkedIndex - BaseIndex).EvalResult, _
                                        CurArg.FuncName, CurArg.UDFFunctionIn))
                Else 'Data is main tree
                    CurArg.Operand = Factorial(EvalFunction(EvalTree_
                                     (CurArg.LinkedIndex).EvalResult, _
                                      CurArg.FuncName, CurArg.UDFFunctionIn))
                End If
            Else 'Eval
                If CurArg.LinkedIndex >= BaseIndex Then
                    CurArg.Operand = EvalFunction(CurTree.Storage_
                                     (CurArg.LinkedIndex - BaseIndex).EvalResult, _
                                      CurArg.FuncName, CurArg.UDFFunctionIn)
                Else
                    CurArg.Operand = EvalFunction(EvalTree(CurArg.LinkedIndex).EvalResult, _
                                        CurArg.FuncName, CurArg.UDFFunctionIn)
                End If
            End If
        Else    'Return data
            If CurArg.LinkedIndex >= BaseIndex Then
                If CurArg.FactorialIn Then
                    CurArg.Operand = Factorial(CurTree.Storage_
                                     (CurArg.LinkedIndex - BaseIndex).EvalResult)
                Else
                    CurArg.Operand = CurTree.Storage(CurArg.LinkedIndex - BaseIndex).EvalResult
                End If
            Else
                If CurArg.FactorialIn Then
                    CurArg.Operand = Factorial(EvalTree(CurArg.LinkedIndex).EvalResult)
                Else
                    CurArg.Operand = EvalTree(CurArg.LinkedIndex).EvalResult
                End If
            End If
        End If
        If AscW(CurArg.DefString) = 45 Then
            CurArg.Operand = ApplyLawOfSigns(op_minus + CurArg.Operand)
        End If
    Else     'Explicit function or data
        If CurArg.LinkedVar > -1 Then             'Variable substitution
            If CurArg.FactorialIn Then            'Operate factorials
                CurArg.Operand = Factorial(ExprVariables.Storage(CurArg.LinkedVar).value)
            Else
                CurArg.Operand = ExprVariables.Storage(CurArg.LinkedVar).value
            End If
            If AscW(CurArg.DefString) = 45 Then
                CurArg.Operand = ApplyLawOfSigns(op_minus + CurArg.Operand)
            End If
        Else
            If CurArg.FactorialIn Then
                CurArg.Operand = Factorial(MidB$(CurArg.DefString, 1, _
                                           LenB(CurArg.DefString) - 2))
            Else
                CurArg.Operand = CurArg.DefString
            End If
        End If
    End If
End Sub

Finally, each token is evaluated by the BottomLevelEval method:

VB.NET
Private Sub BottomLevelEval(ByRef aToken As Token)
    If aToken.OperationToken < 8 Then 'Arithmetic operators
        Select Case aToken.OperationToken
            Case OperatorToken.otSum
                aToken.EvalResult = CastOPtype(aToken.Arg1.Operand, _
                                               aToken.Arg1.NegationFlagOn) _
                                               + CastOPtype(aToken.Arg2.Operand, _
                                               aToken.Arg2.NegationFlagOn)
            Case OperatorToken.otDiff
                aToken.EvalResult = CastOPtype(aToken.Arg1.Operand, _
                                               aToken.Arg1.NegationFlagOn) - _
                                    CastOPtype(aToken.Arg2.Operand, _
                                               aToken.Arg2.NegationFlagOn)
            Case OperatorToken.otMultiplication
                aToken.EvalResult = CastOPtype(aToken.Arg1.Operand, _
                                               aToken.Arg1.NegationFlagOn) * _
                                    CastOPtype(aToken.Arg2.Operand, _
                                               aToken.Arg2.NegationFlagOn)
            Case OperatorToken.otDivision
                aToken.EvalResult = CastOPtype(aToken.Arg1.Operand, _
                                               aToken.Arg1.NegationFlagOn) / _
                                    CastOPtype(aToken.Arg2.Operand, _
                                               aToken.Arg2.NegationFlagOn)
            Case OperatorToken.otPower
                Dim kFctr As Double
                If AscW(aToken.Arg1.DefString) = 45 Then
                    kFctr = -1
                Else
                    kFctr = 1
                End If
                aToken.EvalResult = kFctr * CastOPtype(aToken.Arg1.Operand, _
                                            aToken.Arg1.NegationFlagOn) ^ _
                                            CastOPtype(aToken.Arg2.Operand, _
                                            aToken.Arg2.NegationFlagOn)
            Case OperatorToken.otMod
                aToken.EvalResult = CastOPtype(aToken.Arg1.Operand, _
                                               aToken.Arg1.NegationFlagOn) Mod _
                                    CastOPtype(aToken.Arg2.Operand, _
                                               aToken.Arg2.NegationFlagOn)
            Case OperatorToken.otIntDiv
                aToken.EvalResult = Floor(CastOPtype(aToken.Arg1.Operand, _
                                                     aToken.Arg1.NegationFlagOn) / _
                                    CastOPtype(aToken.Arg2.Operand, _
                                               aToken.Arg2.NegationFlagOn))
            Case Else
                If aToken.Logical Then
                    If aToken.Arg1.NegationFlagOn Then
                        If AscW(aToken.Arg1.Operand) <> 126 Then '"~"
                            aToken.EvalResult = Not CBool(aToken.Arg1.Operand)
                        Else
                            aToken.EvalResult = _
                                   Not CBool(MidB$(aToken.Arg1.Operand, 3))
                        End If
                    Else
                        aToken.EvalResult = CBool(aToken.Arg1.Operand)
                    End If
                Else
                    aToken.EvalResult = aToken.Arg1.Operand
                End If
        End Select
    Else
        If aToken.OperationToken < 14 Then 'Comparison operators
            Select Case aToken.OperationToken
                Case OperatorToken.otEqual
                    aToken.EvalResult = (CastOPtype(aToken.Arg1.Operand, _
                                                    aToken.Arg1.NegationFlagOn) = _
                                        CastOPtype(aToken.Arg2.Operand, _
                                                   aToken.Arg2.NegationFlagOn))
                Case OperatorToken.otNotEqual
                    aToken.EvalResult = (CastOPtype(aToken.Arg1.Operand, _
                                                    aToken.Arg1.NegationFlagOn) <> _
                                        CastOPtype(aToken.Arg2.Operand, _
                                                   aToken.Arg2.NegationFlagOn))
                Case OperatorToken.otGreaterThan
                    aToken.EvalResult = CastOPtype(aToken.Arg1.Operand, _
                                                   aToken.Arg1.NegationFlagOn) > _
                                        CastOPtype(aToken.Arg2.Operand, _
                                                   aToken.Arg2.NegationFlagOn)
                Case OperatorToken.otLessThan
                    aToken.EvalResult = CastOPtype(aToken.Arg1.Operand, _
                                                   aToken.Arg1.NegationFlagOn) < _
                                        CastOPtype(aToken.Arg2.Operand, _
                                                   aToken.Arg2.NegationFlagOn)
                Case OperatorToken.otGreaterThanOrEqual
                    aToken.EvalResult = CastOPtype(aToken.Arg1.Operand, _
                                                   aToken.Arg1.NegationFlagOn) >= _
                                        CastOPtype(aToken.Arg2.Operand, _
                                                   aToken.Arg2.NegationFlagOn)
                Case Else
                    aToken.EvalResult = CastOPtype(aToken.Arg1.Operand, _
                                                   aToken.Arg1.NegationFlagOn) <= _
                                        CastOPtype(aToken.Arg2.Operand, _
                                                   aToken.Arg2.NegationFlagOn)
            End Select
        Else 'Logical operators
            Dim tmpBooleans() As Boolean
            Select Case aToken.OperationToken
                Case OperatorToken.otLogicalAND
                    tmpBooleans() = GetLogicalNeg(aToken)
                    aToken.EvalResult = tmpBooleans(0) And tmpBooleans(1)
                Case OperatorToken.otLogicalOR
                    tmpBooleans() = GetLogicalNeg(aToken)
                    aToken.EvalResult = tmpBooleans(0) Or tmpBooleans(1)
                Case Else
                    tmpBooleans() = GetLogicalNeg(aToken)
                    aToken.EvalResult = tmpBooleans(0) Xor tmpBooleans(1)
            End Select
        End If
    End If
End Sub

The EvalFunction method is in charge of evaluating the functions included in the parsed expression, as mentioned before, the user-defined functions (UDF) will receive an array of text strings, but given the requirements of the CallByName function the only argument that the UDFs must exhibit must be of variant type.

VB.NET
Private Function EvalFunction(ByRef Argument As String, _
        ByRef FunctionName As String, Optional IsUDF As Boolean = False) As String
    If Not IsUDF Then
        Select Case FunctionName
            Case "Absolute"
                EvalFunction = Absolute(Argument)
            Case "ArcSin"
                EvalFunction = ArcSin(Argument)
            Case "ArcCos"
                EvalFunction = ArcCos(Argument)
            Case "ArcTan"
                EvalFunction = ArcTan(Argument)
            Case "Average"
                EvalFunction = average(Argument)
            Case "Cosin"
                EvalFunction = Cosin(Argument)
            Case "aCeiling"
                EvalFunction = aCeiling(Argument)
            Case "ExpEuler"
                EvalFunction = ExpEuler(Argument)
            Case "aFloor"
                EvalFunction = aFloor(Argument)
            Case "Gamma"
                EvalFunction = tGamma(Argument)
            Case "Logarithm"
                EvalFunction = Logarithm(Argument)
            Case "LgN"
                EvalFunction = LgN(Argument)
            Case "LN"
                EvalFunction = LN(Argument)
            Case "Max"
                EvalFunction = Max(Argument)
            Case "Min"
                EvalFunction = Min(Argument)
            Case "Percent"
                EvalFunction = Percent(Argument)
            Case "Power"
                EvalFunction = Power(Argument)
            Case "Sign"
                EvalFunction = Sign(Argument)
            Case "Sine"
                EvalFunction = Sine(Argument)
            Case "SquareRoot"
                EvalFunction = SquareRoot(Argument)
            Case "Tangent"
                EvalFunction = Tangent(Argument)
            Case Else
                'Rise an error for not found function
        End Select
    Else
        EvalFunction = EvalUDF(FunctionName, Argument)
    End If
End Function

Private Function EvalUDF(ByRef UDFname As String, ByRef Expression As String) As String
    Dim args As Variant
    Dim tmpEval As String
    Dim UDFidx As Long
    
    UDFidx = GetCBItemIdx(UserDefFunctions, UDFname)
    If UDFidx > -1 Then              'Only declared functions are called
        args = SplitArgs(Expression) 'Pass a string array to UDF functions
        tmpEval = CallByName(callback(UserDefFunctions.Storage(UDFidx).value), _
                             UDFname, VbMethod, args)
        EvalUDF = tmpEval
    End If
End Function

Using the Code

Let's start with the evaluation of an expression with multiple parentheses:

VB.NET
Sub SimpleMathEval()
    Dim Evaluator As VBAexpressions
    Set Evaluator = New VBAexpressions
    With Evaluator
        .Create "(((((((((((-123.456-654.321)*1.1)*2.2)*3.3)+4.4)+5.5)+_
                 6.6)*7.7)*8.8)+9.9)+10.10)"
        If .ReadyToEval Then    'Evaluates only if the expression was successfully parsed.
            .Eval
        End If
    End With
End Sub

Now let's see how to evaluate an expression that contains variables whose values are defined at the moment the evaluation of the expression is requested.

VB.NET
Sub LateVariableAssignment()
    Dim Evaluator As VBAexpressions
    Set Evaluator = New VBAexpressions
    With Evaluator
        .Create "Pi.e * 5.2Pie.1 + 3.1Pie"
        If .ReadyToEval Then
            Debug.Print "Variables: "; .CurrentVariables    'Print the list of parsed variables
            .Eval ("Pi.e=1; Pie.1=2; Pie=3")                'Late variable assignment
            Debug.Print .Expression; " = "; .Result; _
                        "; for: "; .CurrentVarValues        'Print stored result, expression 
                                                            'and values used in evaluation
        End If
    End With
End Sub

The values of the variables can also be defined prior to calling the Eval method.

VB.NET
Sub EarlyVariableAssignment()
    Dim Evaluator As VBAexpressions
    Set Evaluator = New VBAexpressions
    With Evaluator
        .Create "Pi.e * 5.2Pie.1 + 3.1Pie"
        If .ReadyToEval Then
            Debug.Print "Variables: "; .CurrentVariables
            .VarValue("Pi.e") = 1
            .VarValue("Pie.1") = 2
            .VarValue("Pie") = 3
            .Eval
            Debug.Print .Expression; " = "; .Result; _
                        "; for: "; .CurrentVarValues
        End If
    End With
End Sub

The trigonometric functions included in the library can be evaluated using radians or degrees.

VB.NET
Sub TrigFunctions()
    Dim Evaluator As VBAexpressions
    Set Evaluator = New VBAexpressions
    With Evaluator
        .Create "asin(sin(30))"
        If .ReadyToEval Then
            .Degrees = True               'Eval in degrees
            .Eval
        End If
    End With
End Sub

An important feature of VBA Expressions is that the user can use the GallopingMode property to avoid reevaluating those tokens whose value remains constant between calls to the Eval method.

Points of Interest

This method has proven to be quite flexible, since it follows the same mental process we use when evaluating expressions on paper, allowing, with the use of parentheses, some very powerful hacks. This virtue was exploited when it came to admitting arrays and lists.

Although corner cases have been selected that have caused other solutions to fail, much more testing is needed to ensure that the tool contains a negligible amount of programming errors. Feel free to suggest expressions that can be used in the tests.

History

  • 25th February, 2022: First release

License

This article, along with any associated source code and files, is licensed under The GNU General Public License (GPLv3)

Share

About the Author

W. García
Dominican Republic Dominican Republic
No Biography provided

Comments and Discussions

 
QuestionUnits Pin
Ned Ganchovski28-Feb-22 1:22
MemberNed Ganchovski28-Feb-22 1:22 

General General    News News    Suggestion Suggestion    Question Question    Bug Bug    Answer Answer    Joke Joke    Praise Praise    Rant Rant    Admin Admin   

Use Ctrl+Left/Right to switch messages, Ctrl+Up/Down to switch threads, Ctrl+Shift+Left/Right to switch pages.