Path: news.daimi.aau.dk!news.daimi.aau.dk!eernst
From: eernst@quercus.daimi.aau.dk (Erik Ernst)
Newsgroups: comp.lang.beta
Subject: Re: a few basic questions
Date: 04 May 1995 00:17:01 GMT
Organization: DAIMI, Computer Science Dept. of Aarhus Univ.
Lines: 107
Distribution: world
Message-ID: <EERNST.95May4021701@quercus.daimi.aau.dk>
References: <bakulD7rut5.FLw@netcom.com>
NNTP-Posting-Host: quercus.daimi.aau.dk
In-reply-to: bakul@netcom.com's message of Sat, 29 Apr 1995 01:00:41 GMT

In article <bakulD7rut5.FLw@netcom.com> bakul@netcom.com (Bakul Shah) writes:

   I have been reading the Beta book (Object-Oriented Programming in
   the BETA Programming Language) and have some questions.  I looked
   though the Beta FAQ and various publicly available on-line
   sources but did not find answers to these quesions.

   1. Are Beta identifiers case sensitive?

No. (Exception: names of externals are case sensitive in cases like 

        someFunction: external(# in,out: @integer enter in exit out #);

where there is no "do 'someFunction'->callC" or similar to specify the 
name of the function to call. In this case the name of the external 
is used as the default name of the called function).

   2. Are Beta keywords case sensitive?  In various sources
      I have seen `INNER' as well as `inner' which leads me
      to believe may be they are not.

No, they are not. 

   3. What characters are allowed in identifiers?  More generally,
      grammar described in the appendix A of the Beta book does
      not give any syntax rules for terminals symbols
      such as identifiers and numbers.

Identifier      = {'a'-'z','A'-'Z','_'}{'a'-'z','A'-'Z','0'-'9','_'}*
LiteralInteger  = {'0'-'9'}+
                | '0' ('x'|'X') {'0'-'9','a'-'f','A'-'F'}+
LiteralReal     = {'0'-'9'}+ '.' {'0'-'9'}+
                | {'0'-'9'}+ ('.' {'0'-'9'}+)? ('e'|'E') {'0'-'9'}+

I hope this notation is readable even though it is not quite
standard---unquoted parentheses are used for grouping, and braces are
used to construct sets; '+', '*', and '?' are the usual regular
language constructors; '-' constructs ranges of characters; so only
the quoted characters appear in the resulting strings.  Of course,
'Identifier' is really the set of strings specified above _except_ the
set of reserved words in all possible mixtures of lowercase and
uppercase letters.

   4. Syntax for the If-imperative does not specify (if E then I if)
      as valid but I have seen this syntax used in a number of
      examples elsewhere (i.e. not in the Beta book).  Which is
      right?

The "simple if-statement" has been introduced after the BETA book went
to press.  It also allows an else-part, 

<SimpleIfImp> ::= '(' 'if' <Evaluation> 'then' <Imperatives> 
	          ( 'else' <Imperatives> )? 'if' ')'

   5. Page 62 shows two forms for labeled imperatives.  The second
      form ``(L: Imp1; ... :L)'' does not match what is in appendix
      A.  I suspect appendix A is wrong. Also, what is the `scope'
      of a label? For example, is ``(L: (L: 1 :L) :L)'' valid?

``(L: Imp1; ... :L)'' has been removed from the language.  One can
always use ``L: (# do Imp1; ... #)'' to the same effect.  This even 
generates quite efficient code as long as the inserted object 
'(# .. #)' has no prefix and declares no attributes of its own. 

   6. I am curious about why Beta only allows one dimensional arrays.
      Why is
	   col:(# cell: [columns] @real #);
	   m1: [rows]@col
      preferable to
	   m1:[rows][columns]@real
      ?

I don't think the first one is preferable, just easier to implement
and specify (just think about something like

  ``234554->m1<?what-should-I-write-here-to-mean-all-of-them?>.extend''

or would we allow ``234554->m1[17].extend''?).  It's tedious to have
to write it as it is now and to use it like ``m1[2].cell[3]'', so
perhaps some nice, clean extension to arbitrary-dimensional
repetitions will be proposed (and hopefully implemented right after
that :-).

   I realize that the Beta book is not a formal language reference
   so please don't think of the above questions as a criticism of it.
   I am simply curious.

Me too!

So what kind of information have I actually supplied here?  Just the
kind of information available to any person with access to the Mjølner
implementation of the language.  It might be wrong, or it might be
correct, when talking about _the language_ BETA.  It does seem to be
in line with the behaviour of the Mjølner BETA compiler, though, and
it somehow seems "reasonable" to me.

On the other hand, I think these questions are very good
demonstrations of the fact that we need a concise language reference
manual, so that questions like these will have a definite answer.
One never knows, such one might just turn up ;-)


  cheers, 

--
Erik Ernst                            eernst@daimi.aau.dk
Computer Science Department of Aarhus University, Denmark