1. Suggestion for Euphoria 2.3, 2.2a, 2.2.1, whatever
I noticed several things that C has that Euphoria doesn't. Among these are
structs and switch statements. It should be easy to implement a struct like:
struct thisStruct do
integer a, b
sequence st
end struct
And as for the switch statements:
switch a do
case 1 do
puts(1,"a is 1\n")
elsif 2 do
puts(1,"a is 2\n")
else
puts(1,"a is some other value\n")
end case
end switch
And one feature that even secondbestlanguageintheworld Java doesn't have!
Using plain old math operators on sequences. Maybe you could say, 'st =
"hi"' instead of the current 'equals(st,"hi")' or 'compare(st,"hi")=0'. This
would be much clearer.
- Thx for listening
2. Re: Suggestion for Euphoria 2.3, 2.2a, 2.2.1, whatever
On Thu, 07 Sep 2000, you wrote:
> I noticed several things that C has that Euphoria doesn't. Among these are
> structs and switch statements. It should be easy to implement a struct like:
>
> struct thisStruct do
> integer a, b
> sequence st
> end struct
Since a structure is very likely to be constant during its' existence,
why not declare it in much the same way:
structure x =
integer a,b,
sequence st
end structure
> And as for the switch statements:
>
> switch a do
> case 1 do
> puts(1,"a is 1\n")
> elsif 2 do
> puts(1,"a is 2\n")
> else
> puts(1,"a is some other value\n")
> end case
> end switch
Useful, but ugly.
Why not simplify it somewhat:
case a
= 1 then puts(1,"a is 1"),
= 2 then puts(1,"a is 2"),
< 5 then puts(1,"a is between 3 and 5")
else
puts(1,"ERROR")
end case
--
Regards,
Irv
3. Re: Suggestion for Euphoria 2.3, 2.2a, 2.2.1, whatever
----- Original Message -----
From: irv <irv at ELLIJAY.COM>
To: <EUPHORIA at LISTSERV.MUOHIO.EDU>
Sent: Friday, September 08, 2000 10:26 AM
Subject: Re: Suggestion for Euphoria 2.3, 2.2a, 2.2.1, whatever
> On Thu, 07 Sep 2000, you wrote:
> >
> > switch a do
> > case 1 do
> > puts(1,"a is 1\n")
> > elsif 2 do
> > puts(1,"a is 2\n")
> > else
> > puts(1,"a is some other value\n")
> > end case
> > end switch
>
> Useful, but ugly.
> Why not simplify it somewhat:
> case a
> = 1 then puts(1,"a is 1"),
> = 2 then puts(1,"a is 2"),
> < 5 then puts(1,"a is between 3 and 5")
> else
> puts(1,"ERROR")
> end case
>
> --
> Regards,
> Irv
Or perhaps
case a
1: puts(1,"a is 1"),
2: puts(1,"a is 2"),
<5: puts(1,"a is between 3 and 5")
otherwise: puts(1,"ERROR")
end case
which I find more readable. The "otherwise" statement makes it easy to
distinguish case blocks from if blocks at a glance, even if internally
they'd work the same way. The same philosophy evidently inspired the
"unless" statement (can't remember where), actually an "if not" in disguise,
as in
unless a = b
puts(1,"Alert! A no longer equals B")
end unless
While I'm at it, my own wish list might include something equivalent to the
COBOL level 88 definitions, which let you assign a name to a value, series
of values, or range. In Eu this could be accomplished thru the CONSTANT
statement, as in
constant CHILD = {0..12}, TEEN = {13..19}, OLDGEEZER {20..999}
constant PRIME = {1,2,3,5,7,11}
constant USER_SAYS_OK = {"y", "yes", "ok", "si", "oui"}
This would also help towards shorter, more readable code. As it would
obviously apply sequence processing, subscripting might be allowed, also for
recursive redefinition:
constant BABY = CHILD[0..3], USER_SOUNDS_FRENCH = USER_SAYS_OK[5]
Thus you'd write
case users_age
baby: puts(1,"Please call Mom.\n")
child: puts(1,"Sorry, kid. This is a boring adult math puzzle.\n")
teen: puts(1,"Welcome to Goreville!\n")
oldgeezer: puts(1,"Go get a life!\n")
otherwise: puts(1,"Hi, Methuselah!\n")
end case
and nicer-looking ifs, which could get very complicated in just a few words:
if teen_user and after_midnight and long_program_run and weekday
puts(1,"Go get some sleep.")
end if
Wish I could...
Gerardo
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com
