Arabbit.428
net.lang.apl
utzoo!decvax!harpo!npois!alice!rabbit!ark
Fri Apr 9 09:00:19 1982
Re: More on ASCII APL
I do not see that the problem mentioned by ihuxi!otto is a serious
one, except for a special case that I will mention. First, each APL character
is a single entity, even if it is formed on a conventional terminal
by overstriking two characters. This is especially obvious to those
who have used APL 3270 terminals -- they have an extra case shift that
causes the "overstruck" characters to appear with single keystrokes.
Second, when you reperesent APL characters as lowercase words, you need
a rule to decide when a space adjacent to such a character is part of
the character and when it is a real space. It doesn't matter what the
exact details of the rule are, just that there be a rule and that it be
simple enough to know completely. One such rule might be that a keyword
is always preceded and followed by a "free" blank, unless the preceding
character is a left delimiter or the following character is a right
delimiter. A left delimiter is a quote, left bracket, or whatever,
and similarly for a right delimiter. Again, the content of the list
is not as important as its existence. Given this:
rho 'A rho B'
is unambiguously 3, as the spaces surrounding the second "rho" are part
of the symbol. Similarly,
'transpose X'[2]
is 'X' unambiguously.
The main problem with my proposal is that I have yet to figure out a way
of representing both the minus and negative signs that is not ugly.
-----------------------------------------------------------------
gopher://quux.org/ conversion by John Goerzen <
[email protected]>
of
http://communication.ucsd.edu/A-News/
This Usenet Oldnews Archive
article may be copied and distributed freely, provided:
1. There is no money collected for the text(s) of the articles.
2. The following notice remains appended to each copy:
The Usenet Oldnews Archive: Compilation Copyright (C) 1981, 1996
Bruce Jones, Henry Spencer, David Wiseman.