From Wikipedia, the free encyclopedia
|Paradigm:||multi-paradigm: object-oriented, procedural, structured|
|Designed by:||Mike Cowlishaw|
|Developer:||Mike Cowlishaw & IBM|
|Typing discipline:||dynamic, everything is a string|
|Major implementations:||IBM NetREXX, Open Object Rexx, Regina, others|
|Dialects:||Open Object REXX, NetRexx|
|Influenced by:||PL/I, EXEC 2, BASIC|
REXX (REstructured eXtended eXecutor) is an interpreted programming language which was developed at IBM. It is a structured high-level programming language which was designed to be both easy to learn and easy to read. Both commercial and open source interpreters for REXX are available on a wide range of computing platforms, and compilers are available for IBM mainframes.
REXX has the following characteristics and features:
- character string basis
- dynamic data typing (no declarations)
- no reserved keywords (except in local context)
- arbitrary numerical precision
- decimal arithmetic (floating-point)
- a rich selection of built-in functions (especially string and word processing)
- automatic storage management
- crash protection
- content-addressable data structures
- straightforward access to system commands and facilities
- simple error-handling, and built-in tracing and debugger
- few artificial limitations
- simplified I/O facilities.
REXX has just twenty-three, largely self-evident, instructions (e.g., call, parse, and select) with minimal punctuation and formatting requirements. It is essentially an almost free-form language with only one data-type, the character string; this philosophy means that all data are visible (symbolic) and debugging and tracing are simplified.
REXX syntax looks similar to PL/I, but has fewer notations; this makes it harder to parse (by program) but easier to use.
REXX was designed and first implemented as an ‘own-time’ project between 20 March 1979 and mid-1982 by Mike Cowlishaw of IBM, originally as a scripting programming language to replace the languages EXEC and EXEC 2. It was designed to be a macro or scripting language for any system. As such, REXX is considered a precursor to Tcl and Python.
It was first described in public at the SHARE 56 conference in Houston, Texas, in 1981, where customer reaction, championed by Ted Johnston of SLAC, led to it being shipped as an IBM product in 1982.
Over the years IBM included REXX in almost all of its operating systems (VM/CMS, VM/GCS, MVS TSO/E, AS/400, VSE/ESA, AIX, CICS/ESA, PC-DOS, and OS/2), and has made versions available for Novell NetWare, Windows, Java, and Linux.
The first non-IBM version was written for PC-DOS by Charles Daney in 1984/5. Other versions have also been developed for Atari, Amiga, Unix (many variants), Solaris, DEC, Windows, Windows CE, PocketPC, MS-DOS, Palm OS, QNX, OS/2, Linux, BeOS, EPOC32, AtheOS, OpenVMS, OpenEdition, Macintosh, and Mac OS X. 
The Amiga version of Rexx, called ARexx was included with AmigaOS 2 onwards and was popular for scripting as well as application control. Many Amiga applications have "ARexx ports" built into them which allows control of the application via a user defined script.
Several freeware versions of Rexx are available. In 1992, the two most widely-used open-source ports appeared: Ian Collier's REXX/imc for Unix and Anders Christensen's Regina (later adopted by Mark Hessling) for Windows and Linux. BREXX is well-known for WinCE and PocketPC platforms.
Since the mid-1990s, two newer variants of REXX have appeared:
- NetRexx – which compiles to Java byte-code via Java source code; this has no reserved keywords at all, and uses the Java object model, and is therefore not upwards-compatible with ‘classic’ REXX.
- Object Rexx – which is an object-oriented upwards-compatible version of REXX.
In 1990, Cathy Dager of SLAC organized the first independent REXX symposium, which led to the forming of the REXX Language Association. Symposiums are held annually.
In plain text, Cowlishaw seems to prefer Rexx, whereas IBM documents and the majority of the web uses REXX. The ANSI standard uses the form preferred by the standardization committee, which has small capitals for the final three letters: REXX. This style is also used on the cover pages of "The Rexx Language" TRL and "The NetRexx Language" written by Michael Cowlishaw. Originally it was just called "Rex" because the author liked how it sounded, the extra "x" was added to avoid collisions with other products' names. The expansion of Rexx to the Restructured Extended Executor is a bacronym.
The loop control structure in Rexx always begins with a DO and ends with an END but comes in several varieties.
do until [condition] [instructions] end
do while [condition] [instructions] end
With an index variable:
do i = x [to y [[by z]] [instructions] end
The step increment (z above) may be omitted and defaults to 1. The upper limit (y above) can also be omitted, which makes the loop continue forever. You can also loop forever without an index variable with this:
do forever [instructions] end
A program can break out of the current loop with the leave instruction (which is the normal way to exit a "forever" loop), or can short-circuit it with the iterate instruction. The do while and do until forms are equivalent to:
do forever if [condition] then leave [instructions] end
do forever [instructions] if [condition] then leave end
Testing conditions with IF
if [condition] then do [instructions] end else do [instructions] end
For single instructions, DO and END can also be omitted:
if [condition] then [instruction] else [instruction]
Testing for multiple conditions
SELECT is REXX's CASE structure, like many other constructs derived from PL/I:
select when [condition] then [instruction] when [condition] then do [instructions] end otherwise [instructions] or NOP end
NOP indicates no instruction is to be executed.
Variables in REXX are typeless, and initially are evaluated as their names, in upper case. Thus a variable's type can vary with its use in the program:
say hello /* => HELLO */ hello = 25 say hello /* => 25 */ hello = "say 5 + 3" say hello /* => say 5 + 3 */ interpret hello /* => 8 */ drop hello say hello /* => HELLO */
Unlike many other programming languages, classic REXX has no direct support for arrays of variables addressed by a numerical index. Instead it provides compound variables. A compound variable consists of a stem followed by a tail. A . (dot) is used to join the stem to the tail. If the tails used are numeric, it is easy to produce the same effect as an array.
do i = 1 to 10 stem.i = 10 - i end
Afterwards the following variables with the following values exist: stem.1 = 9, stem.2 = 8, stem.3 = 7...
Unlike arrays, the index for a stem variable is not required to have an integer value. For example, the following code is valid:
i = 'Monday' stem.i = 2
In Rexx it is also possible to set a default value for a stem.
stem. = 'Unknown' stem.1 = 'USA' stem.44 = 'UK' stem.33 = 'France'
After these assignments the term stem.3 would produce 'Unknown'.
The whole stem can also be erased with the DROP statement.
This also has the effect of removing any default value set previously.
By convention (and not as part of the language) the compound stem.0 is often used to keep track of how many items are in a stem, for example a procedure to add a word to a list might be coded like this:
add_word: procedure expose dictionary. parse arg w n = dictionary.0 + 1 dictionary.n = w dictionary.0 = n return
It is also possible to have multiple elements in the tail of a compound variable. For example:
m = 'July' d = 15 y = 2005 day.y.m.d = 'Friday'
Multiple numerical tail elements can be used to provide the effect of a multi-dimensional array.
Features similar to Rexx compound variables are found in many other languages (associative arrays in AWK, hashes in Perl, Vectors in Java, etc). Most of these languages provide an instruction to iterate over all the keys (or tails in Rexx terms) of such a construct, but this is lacking in classic Rexx. Instead it is necessary to keep auxiliary lists of tail values as appropriate. For example in a program to count words the following procedure might be used to record each occurrence of a word.
add_word: procedure expose count. word_list parse arg w . count.w = count.w + 1 /* assume count. has been set to 0 */ if count.w = 1 then word_list = word_list w return
and then later
do i = 1 to words(word_list) w = word(word_list,i) say w count.w end
At the cost of some opacity it is possible to combine these techniques into a single stem.
add_word: procedure expose dictionary. parse arg w . dictionary.w = dictionary.w + 1 if dictionary.w = 1 /* assume dictionary. = 0 */ then do n = dictionary.0+1 dictionary.n = w dictionary.0 = n end return
do i = 1 to dictionary.0 w = dictionary.i say i w dictionary.w end
However Rexx provides no safety net here, so if one of your words happens to be a whole number less than dictionary.0 the above technique will fail mysteriously.
Recent implementations of Rexx, include IBM's Object Rexx and the open source implementations like Regina and OORexx include a new language construct to simplify iteration over the value of a stem.
do i over stem. say i '-->' stem.i end
The PARSE instruction is particularly powerful; it combines some useful string-handling functions. Its syntax is:
parse [upper] origin template
where origin specifies the source:
- arg (arguments, at top level tail of command line)
- linein (standard input, e.g. keyboard)
- pull (REXX data queue or standard input)
- source (info on how program was executed)
- value (an expression) with
- the keyword with is required to indicate where the expression ends
- var (a variable)
- version (version/release number)
and template can be:
- list of variables
- column number delimiters
- literal delimiters
upper is optional; it you specify it, data will be converted to upper case.
Using a list of variables as template
myVar = "John Smith" parse var MyVar firstName lastName say "First name is:" firstName say "Last name is:" lastName
displays the following
First name is: John Last name is: Smith
Using a delimiter as template:
myVar = "Smith, John" parse var MyVar LastName "," FirstName say "First name is:" firstName say "Last name is:" lastName
also displays the following
First name is: John Last name is: Smith
Using column number delimiters:
myVar = "(202) 123-1234" parse var MyVar 2 AreaCode 5 7 SubNumber say "Area code is:" AreaCode say "Subscriber number is:" SubNumber
displays the following
Area code is: 202 Subscriber number is: 123-1234
A template can use a combination of variables, literal delimiters, and column number delimiters.
The INTERPRET instruction is very powerful and one of the two reasons why writing REXX compilers isn't trivial (the other reason is REXX's decimal arbitrary precision arithmetic):
/* a touch of LISP */ X = 'square' interpret 'say' X || '(4) ; exit' SQUARE: return arg(1) * arg(1)
This displays 16 and exits. Because anything in REXX are strings, even rational numbers with exponents, and last but not least complete programs, REXX offers to interpret strings as programs.
This feature was used to implement something like function parameters, e.g. passing SIN, COS, etc. to a procedure SIMPSON to calculate integrals.
Note that REXX offers only basic math. functions like ABS, DIGITS, MAX, MIN, SIGN, RANDOM, and a complete set of hex. plus binary conversions with bit operations, anything else like SIN has to be implemented from scratch or using external libraries. The latter typically don't support arbitrary precision.
Later versions (non-classic) support CALL variable constructs, together with the built-in function VALUE this allows to avoid many cases of INTERPRET. This is a classic program:
/* terminated by input "exit" or similar */ do forever ; interpret linein() ; end
A slightly more sophisticated REXX calculator:
X = 'input BYE to quit' do until X = 'BYE' ; interpret 'say' X ; pull X ; end
say digits() fuzz() form() /* => 9 0 SCIENTIFIC */ say 999999999 + 1 /* => 1.000000000E+9 */ numeric digits 10 /* only limited by available memory */ say 999999999 + 1 /* => 1000000000 */
say 0.9999999999 = 1 /* => 0 (false) */ numeric fuzz 3 say 0.99999999 = 1 /* => 1 (true) */ say 0.99999999 == 1 /* => 0 (false) */
say 100 * 123456789 /* => 1.23456789E+10 */ numeric form engineering say 100 * 123456789 /* => 12.34567890E+9 */
The REXX SIGNAL instruction is intended for abnormal changes in the flow of control (see the next section), however, it can be misused and treated like the GOTO in other languages (although it is not strictly equivalent, because it terminates loops and other constructs). This can produce difficult to read code.
Error handling and exceptions
It is possible in REXX to intercept and deal with errors and other exceptions, using the SIGNAL instruction. There are seven system conditions: ERROR, FAILURE, HALT, NOVALUE, NOTREADY, LOSTDIGITS and SYNTAX. Handling of each can be switched on and off in the source code as desired.
This example will run until stopped by the user:
signal on halt; do a = 1 say a do 100000 /* a delay */ end end halt: say "The program was stopped by the user" exit
Virtually all serious REXX programs contain signal on novalue or a similar statement. This disables the "feature", where undefined variables get their own (upper case) name as value. The status of a variable can be checked with the built-in function SYMBOL returning VAR for defined variables.
|ERROR||Positive RC from a system command|
|FAILURE||Negative RC for a system command (e.g. command doesn't exist)|
|HALT||Abnormal termination (see above)|
|NOVALUE||An unset variable was referenced (see above)|
|NOTREADY||Input or output error (e.g. read attempts beyond end of file)|
|SYNTAX||Invalid program syntax, or some error condition not covered above|
|LOSTDIGITS||Significant digits are lost (ANSI REXX, not in TRL second edition)|
When a condition is handled by SIGNAL ON, the SIGL and RC system variables can be analyzed to understand the situation. RC contains the REXX error code and SIGL contains the line number where the error arose.
Beginning with REXX version 4 conditions can get names, and there's also a CALL ON construct. That's handy if external functions do not necessarily exist:
CHCP: procedure /* protect SIGNAL settings */ signal on syntax name CHCP.TRAP return SysQueryProcessCodePage() CHCP.TRAP: return 1004 /* windows-1252 on OS/2 */
- XEDIT - text editor(s) with native REXX support
- ^ Cowlishaw, Mike. IBM REXX Brief History. IBM. Retrieved on 2006-08-15.
- ^ Cowlishaw, Mike (1981-02-18). REX -- A Command Programming Language. SHARE. Retrieved on 2006-08-15.
- ^ Rexx Implementations. RexxLA. Retrieved on 2006-08-15.
- The Rexx Language: A Practical Approach to Programming ( Prentice Hall, 1990), by Michael Cowlishaw, ISBN 0-13-780651-5
- The NetRexx Language ( Prentice Hall, 1997), by Michael Cowlishaw, ISBN 0-13-806332-X
- Programming in REXX (McGraw-Hill, 1990), by Charles Daney, ISBN 0-07-015305-1
- REXX with OS/2, TSO, & CMS Features (M V S Training, 1999), by Gabriel Gargiulo, ISBN 1-892559-03-X
- Down to Earth Rexx (Perfect Niche Software, 2000), by William Schindler, ISBN 0-9677590-0-5
- Rexx Programmer's Reference (Wiley/Wrox, 2005), by Howard Fosdick, ISBN 0-7645-7996-7
- [http://www.amazon.de REXX Grundlagen für die z/OS Praxis by Johannes Deuring 2005, Germany
- Complete list of 49 books through 2000
- Mike Cowlishaw at IBM
- REXX language page at IBM
- REXX Language Association
- REXX Information-- Downloads, tools, tutorials, reference materials, etc.
- Regina: open-source (LGPL) interpreter for Linux, BSD, Windows, etc.
- REXX/imc: open-source (nonstandard license) interpreter for Unix and Linux systems.
- BREXX: open-source (GPL) interpreter for DOS, Linux, Windows CE, etc.
- Reginald: free interpreter for Windows.
- roo!: free (Kilowatt Software) interpreter for Windows with object-oriented extensions.
- r4: free (Kilowatt Software) interpreter for Windows.
- REXX for Palm OS: shareware (Jaxo Inc.) interpreter for Palm OS.
- Personal REXX: commercial (Quercus Systems) interpreter for Windows, OS/2 and DOS.
- S/REXX: commercial (Benaroya) interpreter for UNIX and Windows.
- uni-REXX: commercial (The Workstation Group Ltd.) interpreter for UNIX.
- Rexx for everyone: An introduction by David Mertz for IBM developerWorks.
- Vladimir Zabrodsky's Album of Algorithms and Techniques for Standard Rexx
- Vladimir Zabrodsky's An Introduction to the Rexx Programming Language
- PLEAC-REXX: Programming Language Examples Alike Cookbook for REXX
- Tips & tricks 3.60 by Bernd Schemmer (OS/2 INF format, 755 KB ZIP, 2004)
- Rexx Frequently Asked Questions (FAQ)
- Introductory Rexx Tutorial - SHARE, Spring 1997