r3wp [groups: 83 posts: 189283]
  • Home
  • Script library
  • AltME Archive
  • Mailing list
  • Articles Index
  • Site search

World: r3wp

[Dialects] Questions about how to create dialects

Gregg wrote (in group Rebol vs Scheme):

I would *love* to see mini-primers on language design for Lisp, Forth, 
Logo, etc. in REBOL.

I've taken the first step for a BASIC dialect:

do http://www.fys.ku.dk/~niclasen/rebol/basic.r

It only knows a few commands so far: auto list new old
And these statements: end goto print rem run
And these functions: cos sin
Example of use:

   10 print "Hello World!"
   20 0
Hello World!
   10 print "Hello World!"
Very cool John. Now, let me throw another thought into the mix, just 
for fun. 

If you were to write a language interpreter long ago, you would do 
it in a low level language like ASM or, later, C. In those languages 
you didn't have high level constructs like we have in REBOL. Certain 
languages have very specific models; consider Lisp and Forth, each 
has a few core definitions and the rest of the language it built 
on those. Lisp has lists, Forth has blocks, etc. 

With REBOL, we can do things in many ways. 

1) Leverage all REBOL has to offer. For example, how hard would it 
be to write a simple Lisp system if you (basically) use blocks for 
lists and supply a few standard Lisp functions? Is eval'ing a Lisp 
paren/list different than DOing a REBOL block?

2) Write lower level code, simulating how you would have to write 
a language using something like C or ASM. You could go as far as 
writing a simple virtual machine with its own set of ops.

3) Write dialects that are designed for building specific kinds of 
languages, showing the core concepts of languages, where they're 
similar, and where they differ; tools for teaching language design. 

I think all of those approaches have something to offer.
[unknown: 9]
I think that making a Basic interpreter in Rebol is more useful, 
powerful, and educational than almost any other endeavour I have 
heard regarding Rebol (ever!).

-	People like the idea of Basic.  
-	Almost all programmers know some Basic
-	Although "basic" it tells programmers this is simple to do.
-	It is a great way to learn a dialect.

-	It has "news worthiness" would be good to write a compete dialect, 
and post on SlashDot.

-	If you can run QBasic - it would me instantly thousands if not 
millions of applications that could run instantly on any platform. 
Probably all would be faster even with graphics.
What may happen, is people (kids for example) would begin hacking 
old Basic applications rewritten in Rebol, to show off.

100 lines of Basic becoming 7 lines of Rebol for example.

There is a group of people hacking Nintendo emulators with a program 
that emulates the joystick, and attempt to play games in the shortest 
time possible.  It is very interesting, but why these types of things 
take off is that people can have fun, and without too much knowledge, 
show off their talents.
This Basic dialect parse in block mode, and this set some restrictions 
on the syntax, but it's probably faster and easier to program the 
parse rules. To make a QBasic would require, I used string parsing. 
Probably the same for most other languages. Unfortunately I don't 
have much time for this project atm., because I have 3 examins in 
the coming week, and after that I'm on vacation for 2 weeks. But 
I would like ot do more of this. Maybe we could make a real project 
with some goals!?
I added a few new things to the BASIC:

added DELETE command, added arguments to LIST, added STOP statement 
and some more (see source). Example of use:

>> do http://www.fys.ku.dk/~niclasen/rebol/basic.r
connecting to: www.fys.ku.dk
Script: "BASIC" (24-Jun-2007)

>auto 5 5
    5 print "Line 5"
   10 rem goto 20
   15 blab
   20 print "Line 20"
   25 stop
   30 0
Line 5

Mistake at line 15
>10 goto 20
Line 5
Line 20

STOP at line 25
[unknown: 9]
Yes, I was not suggesting YOU do this, but rather than it be a goal 
of the Rebol community...
I use this guide as a base for the BASIC interpreter: http://www.nvg.ntnu.no/bbc/doc/BBCUserGuide-1.00.pdf
I found it on this site: http://www.nvg.ntnu.no/bbc/docs.php3
I choosed that one, because I once owned a BBC Micro. I have no idea, 
how far that is from QBasic. But I guess Basic is Basic. They probably 
differ in stuff like graphics and sound.
To Gress's post:

1) Yes, it's interesting to find out, how much we get for free using 
e.g. blocks in REBOL to simulate lists in other languages. Maybe 
using hash! will benefit in some situations!?

2) Using rebcode is also a way to write lower level code. But it 
should also be possible to define REBOL functions, that work like 
(or simulate) the lower level commands in other languages. 

3) Yes, it would be interesting to find out, how languages differ 
in their cores.
LOL (I can't spell your name, sorry!)
*To Gregg's post*
Read before you post! Read before you post! Read before you post!

(Does that help? ;-) Well, we sometimes get a laugh, when people 
don't do that.)
Could this group we [web-public]?

It's an interesting discussion of techiqyes that deserve a wider 
Web public. Yes
Gress = Gregg. NP :-)

QBASIC - Ahhh, now we get to have some fun. QBASIC is not your old 
fashioned, line-numbered BASIC. I think it would be cool to do both, 
but the Q(Quick)BASIC language is much better for writing programs 
that actually do something. I've been tempted to do something like 
that myself. To actually run exisiting QB code, screen access would 
pretty much be a must-have feature. Nothing like those old character 
mode interfaces you know.
[unknown: 9]
Gregg, yes, you se my point....Sunanda, agreed.
If you want basic, why not take one and use r3-dll? :)
I also haven't got any problem with this group going web-public. 
I guess, it's just to change it.
Web public. Yes
A z-machine interpreter could be fun - plenty of infocom games to 
play. ... http://en.wikipedia.org/wiki/Z-machine
Thanks......I've changed the designation to [web-public]:


This guy has written a forth in javascript ... guess it's doable 
in Rebol
The start of a BBC BASIC interpreter using string parsing: http://www.fys.ku.dk/~niclasen/rebol/bbcbasic.r
How do we best get this rolling? I'm interested in making some language 
interpreters in REBOL, because I see future potential. Reichart and 
Gregg talked about QBASIC, which I don't know. Where are the specifications 
for that language? BASIC is a start, and along the line, I would 
like to dig into other languages as well. Is it best keeping it all 
public, or should we make new groups for teams interested in this?
Very nice...  As for the other questions...can't say...but well done 
sir.  And a nice reading code space to boot.
Nice John! I'm not sure how best to get rolling. I'm totally buried 
at the moment, but I can provide web space for the project if we 
need that. I also a number of grammars and things here, if we need 
Oddly, I don't seem to have a QBASIC grammar. :-\
I started on a VB grammar once, which is similar, and I still have 
most of my manuals here somewhere. :-)
Not sure if the various sites for lex/yacc tools and such might have 
Gregg, first I'm making a simple 'engine' or 'template' for BASIC. 
Today I implemented expressions, some simple string handling and 
a little more. I'll make conditions (IF) and loops (FOR), then that 
should be a good start to build on.
Uploaded new version of BBC BASIC intepreter. Added expressions and 
conditions. Added IF and INPUT. Example:

>> do http://www.fys.ku.dk/~niclasen/rebol/bbcbasic.r
connecting to: www.fys.ku.dk
Script: "BBC BASIC" (18-Jul-2007)
BASIC v. 0.0.2 

   10 input "Name",name$

   20 if name$="Carl" then print "Hi " name$ else print "Hello " name$
   30 0
Hello John
So far this is implemented:

Functions: COS, SIN
Expressions can beside unary +, - use: +, -, *, /, ^, (, )
Conditions can use: or, eor, and, =, <>, <=, >=, <, >
It's possible to add strings together and strings can also be tested 
in conditions.
3 datatypes is implemented. Examples:
a$ is a string
a% is an integer
a is a real.
*are implemented*
Use a zero (0) to leave AUTO. Pressing <Esc> will end the BASIC intepreter.
New version 0.1.0 of BBC BASIC. Added FOR ... NEXT loop. Example:

>> do http://www.fys.ku.dk/~niclasen/rebol/bbcbasic.r
connecting to: www.fys.ku.dk
Script: "BBC BASIC" (19-Jul-2007)
BASIC v. 0.1.0 

   10 for a=10 to pi step -2.3
   20 for n%=1 to 3 step 2
   30 print a n%
   40 next
   50 next
   60 0
                  10                   1
                  10                   3
                 7.7                   1
                 7.7                   3
                 5.4                   1
                 5.4                   3
[unknown: 9]
A modern BASIC should maybe be able to recognize both lower an UPPER 
case, like in my examples. The original BBC BASIC distinguish between 
upper and lower case. I should change my interpreter to work this 
way, so already written programs will work.
Or maybe the problem is only with variables!? Hmm
New version 0.1.1 of BBC BASIC. Added many keywords, mostly functions. 
To run:
>> do http://www.fys.ku.dk/~niclasen/rebol/bbcbasic.r

List of keywords: http://www.fys.ku.dk/~niclasen/rebol/bbcbasic.html
Version 0.2.0 of BBC BASIC uploaded. Added DATA, READ and RESTORE. 
Added GOSUB and RETURN. Added hex notation, made some changes and 
fixed bugs.
This is meant as an implementation of BBC BASIC originally found 
on the BBC Micro from british Acorn. It's not an emulator of that 
computer. For emulation of the BBC Micro, I recommend BeebEm.
One thing is to implement old languages, it could also be interesting 
to make a modern BASIC using block parsing in REBOL. That way, it 
would be possible to easily implement many of the datatypes found 
in REBOL. Also imagine to be able to have BASIC (or other language) 
code in the middle of a REBOL script, doing something like:
... (some REBOL code) ...
... (some BASIC code) ...
... (more REBOL code) ...
The BBC BASIC interpreter is now more than 800 lines of REBOL source 
(or 20278 bytes). It's 4568 bytes compressed.
In the early 80'ies, before the era of PCs, the Amiga, Mac and Linux, 
there were a whole range of home computers more or less expensive 
(the BBC Micro were in the high price range back then). They all 
came with some version of the BASIC language. It was also possible 
to program in assembler (the BBC had a 6502 CPU). Imagine a new homecomputer 
with REBOL and rebcode. And many other languages supported as REBOL 
dialects. It should just have a keyboard, network, some RAM and maybe 
a memory stick instead of a harddisk. And connection to a monitor. 
That shouldn't be too costly to produce. Would it sell if the price 
was right?
Oh, and it should have a low-energy CPU to run the scripts, something 
as the G3, I think IBM made to only use 1-2 Watt. Then it could be 
sold all over the World.
geomol, I think you'd have to price it like a midsized calculator
I think also we need a new age of exploratory computing (meaning, 
people exploring other than porn sites and youtube), and it requires 
its own hardware.