1. Re: subscripts (Long Post)
- Posted by Robert Craig <rds at ATTCANADA.NET>
May 12, 2000
-
Last edited May 13, 2000
Lewis Townsend writes:
>>sequence a:
>>
>>subscript value
>>
>>1 0
>>2 {0,0,{9},1}
>>3 {}
>>"Lewis" 99
>>"Townsend" {1, 2, 3}
> a ["First"] = 5
> ? a[1] -- is this equal to
> ? a["First"] -- this?
No, 1 and "First" are two different Euphoria objects,
so they would each have their own associated value.
A sequence could be implemented internally as
a list of object-value pairs. a[10] might have a value,
while a[4] has never been assigned a value,
so a[4] would not exist and would give you an error or something
if you tried to read it. You'd probably want some special "undefined"
value to be returned, rather than just getting a fatal subscript
error.
Subscripting a sequence would cause the interpreter
to search a list for an object. This could get very expensive,
so internally the interpreter would use a fast hash table lookup,
or some other trickery. If you never used anything but small
integers for subscripts, the implementation might
revert to the fast random access method that we have today.
I suppose length() would return the number of
object-value pairs.
Anyway this is all moot, as I have no plans
to go ahead with this.
Your idea could also be quite useful,
but I want to keep things simple.
Regards,
Rob Craig
Rapid Deployment Software
http://www.RapidEuphoria.com
