PC-Online 1998 February/PCOnline_02_1998.iso/filesbbs/os2/netrexx.arj/NETREXX.ZIP/NetRexx/nrover.doc

Text File | 1997-11-27 | 32.8 KB | 805 lines

                                                           NetRexx 1.122
NetRexx language quick start
============================

Copyright(c) IBM Corporation, 1996, 1997.  All rights reserved.

Introduction
""""""""""""
NetRexx is a new programming language derived from both Rexx and
Java(tm); it is a dialect of Rexx that retains the portability and
efficiency of Java, while being as easy to learn and to use as Rexx.

NetRexx is an effective alternative to the Java language.  With NetRexx,
you can create programs and applets for the Java environment more easily
than by programming in Java.  Using Java classes is especially easy in
NetRexx as you rarely have to worry about all the different types of
numbers and strings that Java requires.

This document summarises the main features of NetRexx, and is intended
to help you start using it quickly.  It's assumed that you have some
knowledge of programming in a language such as Rexx, C, BASIC, or Java,
but a knowledge of 'object-oriented' programming isn't needed.

This is not a complete tutorial, though -- think of it more as a
'taster'; it covers the main points of the language and shows some
examples you can try or modify.

For more samples (and examples of using Java classes from NetRexx), and
for more formal details of the language, please see the other NetRexx
documents at http://www2.hursley.ibm.com/netrexx/ -- you'll find the
NetRexx software to download there, too.

NetRexx programs
""""""""""""""""
The structure of a NetRexx program is extremely simple.  This sample
program, 'toast', is complete, documented, and executable as it stands:

  /* This wishes you the best of health. */
  say 'Cheers!'

This program consists of two lines: the first is an optional comment
that describes the purpose of the program, and the second is a SAY
statement.  SAY simply displays the result of the expression following
it -- in this case just a literal string (you can use either single or
double quotes around strings, as you prefer).

To run this program, edit a file called toast.nrx and copy or paste the
two lines above into it.  You can then use the NetRexxC Java program to
compile it, and the java command to run it:

  java COM.ibm.netrexx.process.NetRexxC toast
  java toast

You may also be able to use the NETREXXC command to compile and run the
program with a single command (details may vary -- see the installation
and user's guide document):

 netrexxc toast -run

Of course, NetRexx can do more than just display a character string.
Although the language has a simple syntax, and has a small number of
statement types, it is powerful; the language allows full access to the
rapidly growing collection of Java programs known as 'class libraries',
and allows new class libraries to be written in NetRexx.

The rest of this document introduces most of the features of NetRexx.
Since the economy, power, and clarity of expression in NetRexx is best
appreciated with use, you are urged to try using the language yourself.

Expressions and variables
"""""""""""""""""""""""""
Like SAY in the 'toast' example, many statements in NetRexx include
_expressions_ that will be evaluated.  NetRexx provides arithmetic
operators (including integer division, remainder, and power operators),
several concatenation operators, comparison operators, and logical
operators.  These can be used in any combination within a NetRexx
expression (provided, of course, that the data values are valid for
those operations).

All the operators act upon strings of characters (known as _Rexx
strings_), which may be of any length (typically limited only by the
amount of storage available).  Quotes (either single or double) are used
to indicate literal strings, and are optional if the literal string is
just a number.  For example, the expressions:

  '2' + '3'
  '2' + 3
    2 + 3

would all result in '5'.

The results of expressions are often assigned to _variables_, using a
conventional assignment syntax:

  var1=5            /* sets var1 to '5'    */
  var2=(var1+2)*10  /* sets var2 to '70' */

You can write the names of variables (and keywords) in whatever mixture
of uppercase and lowercase that you prefer; the language is not
case-sensitive.

This next sample program, 'greet', shows expressions used in various
ways:

  /* A short program to greet you.                      */
  /* First display a prompt:                            */
  say 'Please type your name and then press ENTER:'
  answer=ask               /* Get the reply into ANSWER */

  /* If no name was entered, then use a fixed greeting, */
  /* otherwise echo the name politely.                  */
  if answer='' then say 'Hello Stranger!'
               else say 'Hello' answer'!'

After displaying a prompt, the program reads a line of text from the
user ('ask' is a keyword provided by NetRexx) and assigns it to the
variable ANSWER.  This is then tested to see if any characters were
entered, and different actions are taken accordingly; if the user typed
'Fred' in response to the prompt, then the program would display:

  Hello Fred!

As you see, the expression on the last SAY (display) statement
concatenated the string 'Hello' to the value of variable ANSWER with a
blank in between them (the blank is here a valid operator, meaning
'concatenate with blank').  The string '!' is then directly concatenated
to the result built up so far.  These unobtrusive operators (the _blank
operator_ and abuttal) for concatenation are very natural and easy to
use, and make building text strings simple and clear.

The layout of statements is very flexible.  In the 'greet' example, for
instance, the IF statement could be laid out in a number of ways,
according to personal preference.  Line breaks can be added at either
side of the THEN (or following the ELSE).

In general, statements are ended by the end of a line.  To continue a
statement to a following line, you can use a hyphen (minus sign) just as
in English:

  say 'Here we have an expression that is quite long, so' -
      'it is split over two lines'

This acts as though the two lines were all on one line, with the hyphen
and any blanks around it being replaced by a single blank.  The net
result is two strings concatenated together (with a blank in between)
and then displayed.

When desired, multiple statements can be placed on one line with the
aid of the semicolon separator:

  if answer='Yes' then do; say 'OK!'; exit; end

(many people find multiple statements on one line hard to read, but
sometimes it is convenient).

Control statements
""""""""""""""""""
NetRexx provides a selection of _control_ statements, whose form was
chosen for readability and similarity to natural languages.  The control
statements include IF... THEN... ELSE (as in the 'greet' example) for
simple conditional processing:

  if ask='Yes' then say "You answered Yes"
               else say "You didn't answer Yes"

SELECT... WHEN... OTHERWISE... END for selecting from a number of
alternatives:

  select
    when a>0 then say 'greater than zero'
    when a<0 then say 'less than zero'
    otherwise say 'zero'
    end

DO... END for grouping:

  if a>3 then do
    say 'A is greater than 3; it will be set to zero'
    a=0
    end

and LOOP... END for repetition:

  loop i=1 to 10    /* repeat 10 times; I changes from 1 to 10 */
    say i
    end

The LOOP statement can be used to step a variable TO some limit, BY some
increment, FOR a specified number of iterations, and WHILE or UNTIL some
condition is satisfied.  LOOP FOREVER is also provided.  Loop execution
may be modified by LEAVE and ITERATE statements that significantly
reduce the complexity of many programs.

NetRexx arithmetic
""""""""""""""""""
Character strings in NetRexx are commonly used for arithmetic (assuming,
of course, that they represent numbers).  The string representation of
numbers can include integers, decimal notation, and exponential
notation; they are all treated the same way.  Here are a few:

  '1234'
  '12.03'
  '-12'
  '120e+7'

The arithmetic operations in NetRexx are designed for people rather than
machines, so are decimal rather than binary, do not overflow at certain
values, and follow the rules that people use for arithmetic. The
operations are completely defined by the ANSI standard for Rexx, so
correct implementations will always give the same results.

An unusual feature of NetRexx arithmetic is the NUMERIC statement: this
may be used to select the _arbitrary precision_ of calculations.  You
may calculate to whatever precision that you wish, for financial
calculations, perhaps, limited only by available memory.  For example:

  numeric digits 50
  say 1/7

which would display

  0.14285714285714285714285714285714285714285714285714

The numeric precision can be set for an entire program, or be adjusted
at will within the program.  The NUMERIC statement can also be used to
select the notation (_scientific_ or _engineering_) used for numbers in
exponential format.

NetRexx also provides simple access to the native binary arithmetic of
computers.  Using binary arithmetic offers many opportunities for
errors, but is useful when performance is paramount.  You select binary
arithmetic by adding the statement:

  options binary

at the top of a NetRexx program.  The language processor will then use
binary arithmetic instead of Rexx decimal arithmetic for calculations,
throughout the program.

Doing things with strings
"""""""""""""""""""""""""
Another thing Rexx is good for is manipulating strings in various ways.
NetRexx provides the same facilities as Rexx, but with a syntax that is
more like Java or other similar languages:

  phrase='Now is the time for a party'
  say phrase.word(7).pos('r')

The second line here can be read from left to right as "take the
variable 'phrase', find the seventh word, and then find the position
of the first 'r' in that word".  This would display '3' in this case,
because 'r' is the third character in 'party'.

(In Rexx, the second line above would have been written using nested
function calls:

  say pos('r', word(phrase, 7))

which is not as easy to read; you have to follow the nesting and then
backtrack from right to left to work out exactly what's going on.)

In the NetRexx syntax, at each point in the sequence of operations some
routine is acting on the result of what has gone before.  These routines
are called _methods_, to make the distinction from functions (which act
in isolation).  NetRexx provides (as methods) most of the functions that
were evolved for Rexx, for example:

  o changestr (change all occurrences of a substring to another)
  o copies (make multiple copies of a string)
  o lastpos (find rightmost occurrence)
  o left and right (return leftmost/rightmost character(s))
  o reverse (swap end-to-end)
  o space (pad between words with fixed spacing)
  o strip (remove leading and/or trailing white space)
  o pos and wordpos (find the position of string or a word in a string)
  o verify (check the contents of a string for selected characters)
  o word, wordindex, wordlength, and words (work with words)

These and the others like them, and the parsing described in the next
section, make it especially easy to process text with NetRexx.

Parsing strings
"""""""""""""""
The previous section described some of the string-handling facilities
available; NetRexx also provides Rexx string parsing, which is a fast
and simple way of breaking up strings of characters using simple pattern
matching.

A PARSE statement first specifies the string to be parsed, often taken
simply from a variable.  This is followed by a _template_ which
describes how the string is to be split up, and where the pieces are to
be put.

--- Parsing into words ---

The simplest form of parsing template consists of a list of variable
names.  The string being parsed is split up into words (sequences of
characters separated by blanks), and each word from the string is
assigned (copied) to the next variable in turn, from left to right.  The
final variable is treated specially in that it will be assigned a copy
of whatever is left of the original string and may therefore contain
several words. For example, in:

  parse 'This is a sentence.'   v1 v2 v3

the variable v1 would be assigned the value 'This', v2 would be assigned
the value 'is', and v3 would be assigned the value 'a sentence.'.

--- Literal patterns ---

A literal string may be used in a template as a pattern to split up the
string.  For example

  parse 'To be, or not to be?'   w1 ',' w2 w3 w4

would cause the string to be scanned for the comma, and then split at
that point; each section is then treated in just the same way as the
whole string was in the previous example.

Thus, w1 would be set to 'To be', w2 and w3 would be assigned the values
'or' and 'not', and w4 would be assigned the remainder: 'to be?'.  Note
that the pattern itself is not assigned to any variable.

The pattern may be specified as a variable, by putting the variable name
in parentheses.  The following statements:

  comma=','
  parse 'To be, or not to be?'   w1 (comma) w2 w3 w4

therefore have the same effect as the previous example.

--- Positional patterns ---

The third kind of parsing mechanism is the numeric positional pattern.
This works just like the string pattern except in syntax; it specifies a
column number (which may be absolute or relative, and derived from a
variable if necessary).

String patterns and positional patterns can be mixed.

Indexed variables
"""""""""""""""""
NetRexx provides an indexed variable mechanism, adapted from the
compound variables of Rexx.

NetRexx string variables can be referred to simply by name, or also by
their name qualified by another string (the _index_).  When an index is
used, a value associated with that index is either set or extracted; in
the latter case, the initial value of the variable is returned if the
index has not been used to set a value.  For example, the program:

  dognoise='bark'
  dognoise['pup']='yap'
  dognoise['bulldog']='grrrrr'
  say dognoise['pup'] dognoise['terrier'] dognoise['bulldog']

would display

  yap bark grrrrr

Any expression may be used inside the brackets; the resulting string is
used as the index.  Multiple dimensions may be used, if required:

  dognoise='bark'
  dognoise['spaniel', 'brown']='ruff'
  say dognoise['spaniel', 'brown'] dognoise['terrier']

which would display

  ruff bark

Here's a more complex example, a test program with a function (called
a _constant method_ in NetRexx) that removes all duplicate words from a
string of words:

  /* justonetest.nrx -- test the justone function.        */
  say justone('to be or not to be')    /* simple testcase */
  exit

  /* This removes duplicate words from a string, and      */
  /* shows the use of a variable (HADWORD) which is       */
  /* indexed by arbitrary data (words).                   */
  method justone(wordlist) constant
    hadword=0           /* show all possible words as new */
    outlist=''              /* initialize the output list */
    loop while wordlist\=''    /* loop while we have data */
      /* next, split WORDLIST into first word and residue */
      parse wordlist word wordlist
      if hadword[word] then iterate   /* loop if had word */
      hadword[word]=1   /* remember we have had this word */
      outlist=outlist word     /* add word to output list */
      end
    return outlist           /* finally return the result */

Running this program would display just the four words 'to', 'be', 'or',
and 'not'.

This example also uses the built-in _string parsing_ provided by the
PARSE statement.  In this instance, the value of WORDLIST is parsed,
with the first word of the value being assigned to the variable WORD and
the remainder being assigned back to WORDLIST (replacing the original
value).

[Author's note: since the notation for indexed variables looks just like
arrays (see the next section), but does not not suffer the restrictions
of arrays, I like to call them _disarrays_.]

Arrays
""""""
NetRexx also supports Java's fixed-size _arrays_.  These are an ordered
set of items, indexed by integers.  To use an array, you first have to
construct it; an individual item may then be selected by an index whose
value must be in the range 0 through N-1, where N is the number of items
in the array:

  array=String[3]            -- make an array of three Java Strings
  array[0]='String one'      -- set each array item
  array[1]='Another string'
  array[2]='foobar'
  loop i=0 to 2              -- display them
    say array[i]
    end

This example also shows NetRexx _line comments_; the sequence '--'
(outside of literal strings or '/*' comments) indicates that the
remainder of the line is not part of the program and is commentary.

Tracing
"""""""
Defined as part of the language, NetRexx tracing often provides useful
debugging information.  The flow of execution of programs may be traced,
and the execution trace can be viewed as it occurs or captured in a
file.  The trace can show each clause as it is executed, and optionally
show the results of expressions, _etc._  For example, the program:

  trace results
  number=1/7
  parse number before '.' after
  say after'.'before

would result in the trace:

  2 *=*   number=1/7
    >v> number "0.142857143"
  3 *=*   parse number before '.' after
    >v> before "0"
    >v> after "142857143"
  4 *=*   say after'.'before
    >>> "142857143.0"

where the lines marked with '*=*' are the statements in the program,
lines with '>v>' show results assigned to local variables, and
lines with '>>>' show results of un-named expressions.

Exception and error handling
""""""""""""""""""""""""""""
NetRexx doesn't have a GOTO statement, but a SIGNAL statement is
provided for abnormal transfer of control, such as when something
unusual occurs.  Using SIGNAL raises an _exception_; all control
statements are then 'unwound' until the exception is caught by a control
statement that specifies a suitable CATCH statement for handling the
exception.

Exceptions are also raised when various errors occur, such as attempting
to divide a number by zero.  For example:

  say 'Please enter a number:'
  number=ask
  do
    say 'The reciprocal of' number 'is:' 1/number
  catch Exception
    say 'Sorry, could not divide "'number'" into 1'
  end

Here, the CATCH statement will catch any exception that is raised when
the division is attempted (conversion error, divide by zero, etc.).

Any control statement that ends with END (DO, LOOP, or SELECT) may be
modified with one or more CATCH statements to handle exceptions.

Things that aren't strings
""""""""""""""""""""""""""
In all the examples so far, the data being manipulated (numbers, words,
and so on) are expressed as a string of characters.  Many things,
however, can be expressed more easily in some other way, so NetRexx
allows variables to refer to other collections of data, which are known
as _objects_.

Objects are defined by a name that lets NetRexx determine the data
and methods that are associated with the object.  This name identifies
the type of the object, and is usually called the _class_ of the object.

For example, an object of class Oblong might represent an oblong to be
manipulated and displayed.  The oblong could be defined by two values:
its width and its height.  These values are called the _properties_ of
the Oblong class.

Most methods associated with an object perform operations on the object;
for example a 'size' method might be provided to change the size of an
Oblong object.  Other methods are used to construct objects (just as
for NetRexx arrays, an object must be constructed before it can be
used).  In NetRexx and Java, these _constructor_ methods always have the
same name as the class of object that they build (Oblong, in this case).

Here's how an Oblong class might be written in NetRexx (by convention,
this would be written in a file called Oblong.nrx; Java expects the name
of the file to match the name of the class inside it):

  /* Oblong.nrx -- simple oblong class */
  class Oblong
    width       -- size (X dimension)
    height      -- size (Y dimension)

    /* Constructor method to make a new oblong */
    method Oblong(new_width, new_height)
      -- when we get here, a new (uninitialized) object has been
      -- created.  Copy the parameters we have been given to the
      -- properties of the object:
      width=new_width; height=new_height

    /* Change the size of an Oblong */
    method size(new_width, new_height) returns Oblong
      width=new_width; height=new_height
      return this   -- return the resized object

    /* Change the size of an Oblong, relative to its current size */
    method sizerelative(rel_width, rel_height) returns Oblong
      width=width+rel_width; height=height+rel_height
      return this

    /* 'Print' what we know about the oblong */
    method print
      say 'Oblong' width 'x' height

To summarize:

  1. A class is started by the 'class' statement, which names the
     class.

  2. The class statement is followed by a list of the properties of the
     object.  These can be assigned initial values, if required.

  3. The properties are followed by the methods of the object.  Each
     method is introduced by a method statement which names the method
     and describes the arguments that must be supplied to the method.
     The body of the method is ended by the next method statement (or
     by the end of the file).

The Oblong.nrx file is compiled just like any other NetRexx program, and
should create a _class file_ called Oblong.class.  Here's a program to
try out the Oblong class:

  /* tryOblong.nrx -- try the Oblong class */

  first=Oblong(5,3)              -- make an oblong
  first.print                    -- show it
  first.sizerelative(1,1).print  -- enlarge it and print it again

  second=Oblong(1,2)             -- make another oblong
  second.print                   -- and print it

when 'tryOblong.nrx' is compiled, you'll notice that the cross-reference
listing of variables shows that the variables 'first' and 'second' have
type 'Oblong'.  These variables refer to Oblongs, just as the variables
in earlier examples referred to Rexx strings.

Once a variable has been assigned a type, it can only refer to objects
of that type.  This helps avoid errors where a variable refers to an
object that it wasn't meant to.

--- Programs are classes, too ---

It's worth pointing out, here, that all the example programs in this
document are in fact classes (you may have noticed that compiling them
creates xxx.class files, where xxx is the name of the source file).  The
Java environment will allow a class to run as a stand-alone
_application_ if it has a constant method called 'main' which takes an
array of Java Strings as its argument.

If necessary (that is, if there is no class statement) NetRexx
automatically adds the necessary class and method statement, and also
a statement to convert the array of strings (each of which holds one
word from the command string) to a single Rexx string.  The 'toast'
example could therefore have been written:

  /* This wishes you the best of health. */
  class toast
    method main(argwords=String[]) constant; arg=Rexx(argwords)
      say 'Cheers!'


Extending classes
"""""""""""""""""
It's common, when dealing with objects, to take an existing class and
extend it.  One way to do this is to modify the source code of the
original class -- but this isn't always available, and with many
different people modifying a class, classes could rapidly get
over-complicated.

Languages that deal with objects, like NetRexx, therefore allow new
classes of objects to be set up which are derived from existing classes.
For example, if you wanted a different kind of Oblong in which the
Oblong had a new property that would be used when printing the Oblong as
a rectangle, you might define it thus:

  /* charOblong.nrx -- an oblong class with character */
  class charOblong extends Oblong
    printchar       -- the character for display

    /* Constructor method to make a new oblong with character */
    method charOblong(new_width, new_height, new_printchar)
      super(new_width, new_height)      -- make an oblong
      printchar=new_printchar           -- and set the print character

    /* 'Print' the oblong */
    method print
      loop for super.height
        say printchar.copies(super.width)
        end

There are several things worth noting about this example:

  1. The 'extends Oblong' on the class statement means that this class
     is an extension of the Oblong class.  The properties and methods of
     the Oblong class are _inherited_ by this class (that is, appear as
     though they were part of this class).

     Another common way of saying this is that 'charOblong' is a
     _subclass_ of 'Oblong' (and 'Oblong' is the _superclass_ of
     'charOblong').

  2. This class adds the 'printchar' property to the properties already
     defined for Oblong.

  3. The constructor for this class takes a width and height (just like
     Oblong) and adds a third argument to specify a print character.  It
     first invokes the constructor of its superclass (Oblong) to build
     an Oblong, and finally sets the printchar for the new object.

  4. The new charOblong object also prints differently, as a rectangle
     of characters, according to its dimension.  The 'print' method (as
     it has the same name and arguments -- none -- as that of the
     superclass) replaces (overrides) the 'print' method of Oblong.

  5. The other methods of Oblong are not overridden, and therefore can
     be used on charOblong objects.

The charOblong.nrx file is compiled just like Oblong.nrx was, and
should create a file called charOblong.class.  Here's a program to
try it out:

  /* trycharOblong.nrx -- try the charOblong class */

  first=charOblong(5,3,'#')      -- make an oblong
  first.print                    -- show it
  first.sizerelative(1,1).print  -- enlarge it and print it again

  second=charOblong(1,2,'*')     -- make another oblong
  second.print                   -- and print it

This should create the two charOblong objects, and print them out in a
simple 'character graphics' form.  Note the use of the method
'sizerelative' from Oblong to resize the charOblong object.

--- Optional arguments ---

All methods in NetRexx may have optional arguments (omitted from the
right) if desired.  For an argument to be optional, you must supply a
default value.  For example, if the charOblong constructor was to have a
default printchar value, its method statement could have been written:

  method charOblong(new_width, new_height, new_printchar='X')

which indicates that if no third argument is supplied then 'X' should be
used.  A program creating a charOblong could then simply write:

  first=charOblong(5,3)       -- make an oblong

which would have the same effect as if 'X' were specified as the third
argument.

Binary types and conversions
""""""""""""""""""""""""""""
The Java environment supports, and indeed requires, the notion of
fixed-precision 'primitive' binary types, which correspond closely to
the binary operations usually available at the hardware level in
computers.  In brief, these types are:

  o _byte, short, int,_ and _long_ -- signed integers that will fit in
    8, 16, 32, or 64 bits respectively

  o _float_ and _double_ -- signed floating point numbers that will fit
    in 32 or or 64 bits respectively.

  o _char_ -- an unsigned 16-bit quantity, holding a Unicode character

  o _boolean_ -- a 1-bit logical value, representing 'false' or 'true'.

Objects of these types are handled specially by the environment 'under
the covers' in order to achieve maximum efficiency; in particular, they
cannot be constructed like other objects -- their value is held
directly.  This distinction rarely matters to the NetRexx programmer: in
the case of string literals an object is constructed automatically; in
the case of an _int_ literal, an object is not constructed.

Further, NetRexx automatically allows the conversion between the various
forms of character strings in Java (String, char, char[], and Rexx) and
the primitive types listed above.  The 'golden rule' that is followed by
NetRexx is that any automatic conversion which is applied must not lose
information: either it can be determined at compile time that the
conversion is safe (as in int -> String) or it will be detected at run
time if the conversion fails (as in String -> int).

The automatic conversions greatly simplify the writing of programs for
the Java environment: the exact type of numeric and string-like method
arguments rarely needs to be a concern of the programmer.

For certain applications where early checking or performance override
other considerations, NetRexx provides options for different treatment
of the primitive types:

  1. options strictassign -- ensures exact type matching for all
     assignments.  No conversions (including those from shorter
     integers to longer ones) are applied.  This option provides
     stricter type-checking than Java, and ensures that all types are
     an exact match.

  2. options binary -- uses Java fixed precision arithmetic on binary
     types (also, literal numbers, for example, will be treated as
     binary, and local variables will be given 'native' Java types such
     as _int_ or _String_, where possible).

     Binary arithmetic currently gives better performance than Rexx
     decimal arithmetic, but places the burden of avoiding overflows and
     loss of information on the programmer.

The options statement (which may list more than one option) is placed
before the first class statement in a file.

You may also explicitly assign a type to an expression or variable:

   i=int 3000000    -- 'i' is an 'int' with initial value 3000000
   j=int 4000000    -- 'j' is an 'int' with initial value 4000000
   k=int            -- 'k' is assigned type 'int', with no initial value
   say i*j          -- carry out multiplication and display the result
   k=i*j            -- carry out multiplication and assign result to 'k'

This example also illustrates one difference between 'options nobinary'
and 'options binary'.  With the former (the NetRexx default) the SAY
would display '1.20000000E+13' and a Conversion overflow would be
reported when the same expression is assigned to the variable 'k'.

With 'options binary', binary arithmetic would be used for the
multiplications, and so no error would be detected; the SAY would
display '-138625024' and the variable 'k' takes the incorrect result.

--- Binary types in practice ---

In practice, explicit type assignment is only occasionally needed in
NetRexx.  Those conversions that are necessary for using existing
classes (or those that use 'options binary') are generally automatic.
For example, here is an "Applet" for use by Java-enabled browsers:

  /* A simple graphics Applet */
  class Rainbow extends Applet
    method paint(g=Graphics)                 -- called to repaint the window
      maxx=size.width-1
      maxy=size.height-1
      loop y=0 to maxy
        col=Color.getHSBColor(y/maxy, 1, 1)  -- select a colour
        g.setColor(col)                      -- set it
        g.drawLine(0, y, maxx, y)            -- and fill a slice
      end y

In this example, the variable 'col' will have type 'Color', and the
three arguments to the method 'getHSBColor' will all automatically be
converted to type 'float'.  As no overflows are possible in this
particular example, 'options binary' may be added to the top of the
program with no other changes being necessary.


Summary and Information Sources
"""""""""""""""""""""""""""""""
The NetRexx language, as you will have seen, allows the writing of
programs for the Java environment with a minimum of overhead and
'boilerplate syntax'; using NetRexx for writing Java classes could
increase your productivity by 30% or more.

Further, by simplifying the variety of numeric and string types of Java
down to a single class that follows the rules of Rexx strings,
programming is greatly simplified.  Where necessary, however, full
access to all Java types and classes is available.

Other examples are available, including both stand-alone applications
and samples of applets for Java-enabled browsers (for example, an applet
that plays an audio clip, and another that displays the time in
English).  You can find these from the NetRexx web pages, at

  http://www2.hursley.ibm.com/netrexx/

Also at that location, you'll find a more in-depth treatment of the
language, and downloadable packages containing the NetRexx software and
documentation.  The software should run on any platform that supports
the Java Development Kit.