parser10.txt

ADaM is a data mining and image processing toolkit

  TParser parses and evaluates mathematical expressions 
specified at runtime. Its performance is remarkable 
- only 40-80% slower than similar compiled expression -
and it is by far the fastest parser on the freeware market.


  Quickstart:
  -----------
  [Some versions of TParser come with a comprehensive
   online documenation ]

  The programming interface is simple:
  - specify values for predefined variables in properties 
     A,B,C,D,E,X,Y or T;
  - optionally add new variables or removing existing ones;
  - specify expression to be evaluated in Expression property;
  - optionally change value of variables
  - retrieve computed value in Value property. 

  Example:
    Parser1.X := 100;
    Parser1.Y := 200;
    Parser1.Variable['z'] := 20;
    Parser1.Expression := 'sin(x)*cos(y)+z';
    Result := Parser1.Value;

  If you want to compute several values of the same 
expression with different sets of variables, specify the
expression once ONLY as this operation is rather time 
consuming, then assign new values for variables and 
retrieve the Value property as many times as you wish. 


  Example:

    for i := 1 to 100 do
    begin
      Parser1.X := i;
      Result := Parser1.Value;
    end;

  You also may add your own variables by specifying their 
names and values using the SetVariable method, accessing 
them via a property or through direct memory access.
(as usual variable names are not case-sensitive).

  Example (the following lines are all equivalent):

(1) Parser1.Variable['Test'] := 100;
    Test := Parser1.Variable['Test'];

(2) Parser1.SetVariable('Test', 100);
    Test := Parser1.GetVariable('Test');

(3) var
      PTest : PParserFloat; { pointer to memory }

      PTest := Parser1.SetVariable('Test', 100);
      PTest^ := 200; { set 'Test' = 200 }    
      Test := PTest^;

  There is no limit for expression length in the 32bit version, 
while the 16bit version has a restriction of 255 characters.


  Predefined variables:
    PI
  
  Accepted operators: + , - , * , / , ^ , MOD, DIV
  [ MOD and DIV implicitly perform a trunc() on their operands ]

  The following functions are supported; it doesn't matter
  if you use lower or upper case:

  [ NOTE: to activate some additional functions you need to
          remove the SpeedCompare conditional define
          in PARSER10.PAS !   ]

    COS, SIN, SINH, COSH, TAN, COTAN, ARCTAN, ARG,

    EXP, LN, LOG10, LOG2, LOGN,

    SQRT, SQR, POWER, INTPOWER,

    MIN, MAX, ABS, TRUNC, INT, CEIL, FLOOR, 

    HEAV (heav(x) is =1 for x=>0 and =0 for x<0),
    SIGN (sign(x) is 1 for x>1, 0 for x=0, -1 for x<0),
    ZERO (zero(x) is 0 for x=0, 1 for x<>0),
    PH (ph(x) = x - 2*pi*round(x/2/pi))
    RND (rnd(x) = int(x) * Random)
    RANDOM (random(X) = Random; the argument X is not used)
   
  Adding your own functions is easy, too.
  Either use ...

    AddFunctionOneParam   or
    AddFunctionTwoParam

  ... if you do not want to create a new class, or create a new
  class, inheriting from TParser and add your functions to
  the lists, as demonstrated in TParser.Create.


  IMPORTANT:
    DO NOT USE BLANKS (#32) IN THE EXPRESSION.
    THE PARSER IS UNABLE TO HANDLE THESE
    (and raises an exception in response).

  You can use bracketing (nestings) up to a level of 20.
  This can be increased at the expense of stack consumption
  by changing
    maxBracketLevels = 20;
  in P10BUILD.PAS. This should not be a problem.

  If you get a "Expression too complex" exception increase
    maxLevelWidth = 50;
  in P10BUILD.PAS. This should never happen.

  You can define your OWN variables at RUNTIME,
  in code you will use

     Parser1.Variable.Add('NAME', 123456)

  or simply (but slow)

     Parser1.Variable['ANOTHER'] := 1.23

  See the demonstration program for better techniques.


  IMPORTANT: 

  The used _mathematical_ routines behave exactly like Delphi
  runtime code in case of errors. Some people feel that this is
  a problem.

  This is not a parsing issue, but rather a lack of attention to
  the actual maths part (which is sloppy - deliberately)... Most 
  probably you will be using your own mathematical routines
  which are faster, more reliable, and provide more functionality.