NCalc - Mathematical Expressions Evaluator for .NET

Closed Issue: Multiplication behavior [25202]

sebastienros — Tue, 03 Nov 2009 16:53:03 GMT

In version 1.3 of NCalc, the expression 2*2*2*2 produces the value 4. Is there a way to avoid this behavior other than 2*(2*(2*2))) ?
Comments:

THANK YOU !!!

Are you the only one to have tried this framework ;) I can't imagine such a bug, but actually it was there.

I have published the corrected version, i.e. NCalc 1.3.1

Updated Release: NCalc - 1.3.1 (nov. 03, 2009)

sebastienros — Tue, 03 Nov 2009 16:51:26 GMT

BUG: Major issue corrected with precedence. Could not imagine a so predictable issue to occur ! Shame on me. No test was checking it !!!

Released: NCalc - 1.3.1 (Nov 03, 2009)

Tue, 03 Nov 2009 16:51:26 GMT

BUG: Major issue corrected with precedence. Could not imagine a so predictable issue to occur ! Shame on me. No test was checking it !!!

Updated Release: NCalc - 1.3.1 (Jan 01, 2009)

sebastienros — Tue, 03 Nov 2009 16:50:38 GMT

BUG: Major issue corrected with precedence. Could not imagine a so predictable issue to occur ! Shame on me. No test was checking it !!!

Created Release: NCalc - 1.3.1 (Jan 01, 2009)

sebastienros — Tue, 03 Nov 2009 16:50:17 GMT

BUG: Major issue corrected with precedence. Could not imagine a so predictable issue to occur ! Shame on me. No test was checking it !!!

Created Issue: Multiplication behavior [25202]

wujonesj2 — Tue, 03 Nov 2009 15:36:01 GMT

In version 1.3 of NCalc, the expression 2*2*2*2 produces the value 4. Is there a way to avoid this behavior other than 2*(2*(2*2))) ?

Updated Wiki: operators

sebastienros — Tue, 27 Oct 2009 14:27:10 GMT

Operators

Expressions can be combined using operators. Each operator as a precedence priority. Here is the list of those expression's priority.

primary
unary
power
multiplicative
additive
relational
logical

Logical

These operators can do some logical comparison between other expressions:

or, ||
and, &&

  true or false and true

The and operator has more prioroty thand the or, thus in the example above, false and true is evaluated first.

Relational

=, ==, !=, <>
<, <=, >, >=

  1 < 2

Additive

+, -

  1 + 2 - 3

Multiplicative

*, /, %

 1 * 2 % 3

Bitwise

& (bitwise and), | (bitwise or), ^(bitwise xor), << (left shift), >>(right shift)

  2 >> 3

Unary

!, not, -, ~ (bitwise not)

  not true

Primary

(, )
values

  2 * ( 3 + 2 )

Updated Wiki: operators

sebastienros — Tue, 27 Oct 2009 14:26:52 GMT

Operators

Expressions can be combined using operators. Each operator as a precedence priority. Here is the list of those expression's priority.

primary
unary
power
multiplicative
additive
relational
logical

Logical

These operators can do some logical comparison between other expressions:

or, ||
and, &&

  true or false and true

The and operator has more prioroty thand the or, thus in the example above, false and true is evaluated first.

Relational

=, ==, !=, <>
<, <=, >, >=

  1 < 2

Additive

+, -

  1 + 2 - 3

Multiplicative

*, /, %

 1 * 2 % 3

Bitwise

& (bitwsie and), | (bitwise or), ^(bitwise xor), << (left shift), >>(right shift)

  2 >> 3

Unary

!, not, -, ~ (bitwise not)

  not true

Primary

(, )
values

  2 * ( 3 + 2 )

Released: NCalc - 1.3.0 (Oct 27, 2009)

Tue, 27 Oct 2009 13:08:12 GMT

NEW: Bitwise operators (&, |, ^, ~, <<, >>).
BREAK: Removed power operator, use Pow(x,y) instead.

Updated Release: NCalc - 1.3.0 (oct. 27, 2009)

sebastienros — Tue, 27 Oct 2009 13:08:12 GMT

NEW: Bitwise operators (&, |, ^, ~, <<, >>).
BREAK: Removed power operator, use Pow(x,y) instead.

Updated Release: NCalc - 1.3.0 (Jan 01, 2009)

sebastienros — Tue, 27 Oct 2009 13:06:49 GMT

NEW: Bitwise operators (&, |, ^, ~, <<, >>).
BREAK: Removed power operator, use Pow(x,y) instead.

Created Release: NCalc - 1.3.0 (Jan 01, 2009)

sebastienros — Tue, 27 Oct 2009 13:06:18 GMT

NEW: Bitwise operators (&, |, ^, ~, <<, >>).
BREAK: Removed power operator, use Pow(x,y) instead.

New Post: Call for feedback !

leeatkinson — Mon, 21 Sep 2009 11:05:52 GMT

But lets suppose I'm using this is a long-running system. The system needs to evaluate many different expressions. Some may be identical, but yet some may only be used once. If the cache does not have the ability to remove objects that have not been used, and keep the most regularly-used ones, then eventutally, the cache will use too much memory.

ASP.NET cache has the ability to mange this much better. You could right your own one with the same sophistication, but wouldn't it be better to program to an interface, and let the users to select the implementation (or write their own).

Lee

New Post: Support for parameter initialization in code

leeatkinson — Mon, 21 Sep 2009 11:01:30 GMT

Thanks

The trouble is, I want the user to be able to set the expression, and that the expression is only evaluated when needed. It's really just to save multiple evaluation of the same expression during multiple tests against it.

Lee

New Post: Call for feedback !

sebastienros — Fri, 18 Sep 2009 15:52:03 GMT

The cache only contains the abstract tree of the expression. It means if you have "3 * x + y", the cache nows only that x and y are needed, but not the value. If you clear the cache, the next compilation will be exactly the same abstract tree result: + ( * ( 3, x ) , y ). When calling Evaluate(), this tree is taken from the cache and then only the parameters' value are resolved. If you create another Expression instance with the same text, the abstract tree will be taken from the tree, not doing a useless compilation. In three centuries, the same expression will be evaluated as the same abstract tree ;). So invalidating the cache is not needed.

The only reason to invalidate it is to gain memory, but it's already implemented internally using weak references.

New Post: Support for parameter initialization in code

sebastienros — Fri, 18 Sep 2009 15:45:52 GMT

You can evaluate it once and set the result as a parameter:

Expression e1 = new Expression("3.14 * x");
e1.SetParameter("x", 0.5);
// a is evaluated once
double a = e1.Evaluate();

Expression e = new Expression("cos(a) * b");
e.SetParameter("a", a);
e.SetParameter("b", 1);
e.Evaluate();
// define a new value for b
e.SetParameter("b", 2);
e.Evaluate();

And if you pass an Expression instance to a parameter, it will get evaluted each time it is needed, as the parameter might be dependent of another variable.

Updated Wiki: functions

sebastienros — Fri, 18 Sep 2009 14:30:02 GMT

The framework includes a set of already implemented functions.

Name	Description	Usage	Result
Abs	Returns the absolute value of a specified number.	Abs(-1)	1M
Acos	Returns the angle whose cosine is the specified number.	Acos(1)	0d
Asin	Returns the angle whose sine is the specified number.	Asin(0)	0d
Atan	Returns the angle whose tangent is the specified number.	Atan(0)	0d
Ceiling	Returns the smallest integer greater than or equal to the specified number.	Ceiling(1.5)	2d
Cos	Returns the cosine of the specified angle.	Cos(0)	1d
Exp	Returns e raised to the specified power.	Exp(0)	1d
Floor	Returns the largest integer less than or equal to the specified number.	Floor(1.5)	1d
IEEERemainder	Returns the remainder resulting from the division of a specified number by another specified number.	IEEERemainder(3, 2)	-1d
Log	Returns the logarithm of a specified number.	Log(1, 10)	0d
Log10	Returns the base 10 logarithm of a specified number.	Log10(1)	0d
Max	Returns the larger of two specified numbers.	Max(1, 2)	2
Min	Returns the smaller of two numbers.	Min(1, 2)	1
Pow	Returns a specified number raised to the specified power.	Pow(3, 2)	9d
Round	Rounds a value to the nearest integer or specified number of decimal places.	Round(3.222, 2)	3.22d
Sign	Returns a value indicating the sign of a number.	Sign(-10)	-1
Sin	Returns the sine of the specified angle.	Sin(0)	0d
Sqrt	Returns the square root of a specified number.	Sqrt(4)	2d
Tan	Returns the tangent of the specified angle.	Tan(0)	0d
Truncate	Calculates the integral part of a number.	Truncate(1.7)	1

It also includes other general purpose ones.

Name	Description	Usage	Result
in	Returns whether an element is in a set of values.	in(1 + 1, 1, 2, 3)	true
if	Returns a value based on a condition.	if(3 % 2 = 1, 'value is true', 'value is false')	'value is true'

Updated Wiki: values

sebastienros — Fri, 18 Sep 2009 14:29:48 GMT

Values

A value is a terminal token representing a concrete element. This can be:

an integer
a floating point number
a date time
a boolean
a string
a function
a parameter

Integers

They are represented using numbers.

They are evaluated as Int32.

Floating point numbers

Use the dot to define the decimal part.

123.456
.123

They are evaluated as Decimal.

Scientific notation

You can use the e to define power of ten (10^).

1.22e1
1e2
1e+2
1e-2
.1e-2
1e10

They are evaluated as Double

Dates and Times

Must be enclosed between sharps.

#2008/01/31# // for en-US culture

The are evaluated as DateTime. NCalc uses the current Culture to evaluate them.

Booleans

Booleans can be either true or false.

true

Strings

Any character between single quotes "'" are evaluated as String.

'hello'

You can escape special characters using \\, \', \n, \r, \t.

Function

A function is made of a name followed by braces, containing optionally any value as arguments.

  Abs(1), doSomehting(1, 'dummy')

Please read the functions page for details.

Parameters

A parameter as a name, and can be optionnaly contained inside brakets.

  2 + x, 2 + [x]

Please read the parameters page for details

Updated Wiki: operators

sebastienros — Fri, 18 Sep 2009 14:29:29 GMT

Operators

Expressions can be combined using operators. Each operator as a precedence priority. Here is the list of those expression's priority.

primary
unary
power
multiplicative
additive
relational
logical

Logical

These operators can do some logical comparison between other expressions:

or, ||
and, &&

  true or false and true

The and operator has more prioroty thand the or, thus in the example above, false and true is evaluated first.

Relational

=, ==, !=, <>
<, <=, >, >=

  1 < 2

Additive

+, -

  1 + 2 - 3

Multiplicative

*, /, %

 1 * 2 % 3

Power

  2 ^ 3

Unary

!, not, -

  not true

Primary

(, )
values

  2 * ( 3 + 2 )

Updated Wiki: description

sebastienros — Fri, 18 Sep 2009 14:28:53 GMT

Description

NCalc is a set of assemblies allowing expressions evaluation. The main class to use is Expression.
This class has a method Evaluate() returning the actual value of its String representation.

Example:

  Expression e = new Expression("2 * 3");
  object result = e.Evaluate();
  
  Console.WriteLine(result);

This example above first creates an instance of Expression using a valued constructor. This constructor takes a string as parameter.
Then the method Evaluate() is called to parse the string, and returns the actual value represented by the string.

To create expressions you can combine several operators and values.

Case sensitivity

By default, the evaluation process is case sensitive. This means every parameter and function evaluation will match using case. This behavior can be overriden using a specific evaluation option.

 Expression e = new Expression("aBs(-1)", EvaluateOptions.IgnoreCase)
 Debug.Assert(1M, e.Evaluate());

Handling errors

When the expression has a syntax error, the evaluation will throw an EvaluationException.

 try
 {
     new Expression("(3 + 2").Evaluate();
 }
 catch(EvaluationException e)
 {
     Console.WriteLine("Error catched: " + e.Message);
 }

Though, you can also detect syntax errors before the evaluation by using the HasErrors() method.

 Expression e = new Expression("a + b * (");
 if(e.HasErrors())
 {
     Console.WriteLine(e.Error);
 }

Expression caching

When Evaluate() is called on an expression, it is compiled once. If the same expression is reused the compilation is not executed again. Thus you can reuse Expression instances by changing the parameters, and you will gain in performance because only the trasversal of the expression tree will be done.

Moreover, each compiled expression is cached internaly, which means you don't even have to care about reusing an Expression instance, the framework will do it for you. The cache is automatically cleaned like the GC does when an Expression is no more used, or memory is needed (i.e. using WeakReference).

You can disable this behavior at the framework level by setting false to CacheEnabled.

 Expression.CacheEnabled = false;

You can also tell a specific Expression instance not to be taken from the cache.

 Expression e = new Expression("1 + 1", EvaluateOption.NoCache);