comp.lang.idl-pvwave archive
Messages from Usenet group comp.lang.idl-pvwave, compiled by Paulo Penteado
Home » Public Forums » archive » Still missing features in IDL 8
Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend  		Return to the default flat view Create a new topic Submit Reply
Re: Still missing features in IDL 8 [message #73448 is a reply to message #73381] Sat, 06 November 2010 12:03 Go to previous messageGo to previous message
JDS is currently offline  JDS
Messages: 94
Registered: March 2009
Member
On Nov 4, 5:40 pm, Paulo Penteado <pp.pente...@gmail.com> wrote:
> On Nov 4, 9:29 pm, JD Smith <jdtsmith.nos...@yahoo.com> wrote:
>
>> Lists and hashes, while very welcome (!), are somewhat cumbersome to
>> use with out some syntactic sugar along the lines of what has been
>> discussed.  In addition to:
>
>>  list[1,2,3]
>
>> for a list/array containing another list/array, containing another
>> list/array, we need, e.g.:
>
>>  list[3,'foo',2:5]
>
>> for a list containing a hash or structure, containing an array or
>> list.
>
> That is just what I was doing.
>
>>   Currently you must use
>
>>  (list[3])['foo']
>
>> A better syntax, if it could be arranged, would be:
>
>>  list[1][2]
>
>> or
>
>>  list[3]['foo']
>
>> Even better if these could function correctly on the LHS of an
>> assignment.
>
> I agree that way would be better, not just because it would not mess
> with the 8D limits. But that would require changes to the language,
> for the new syntax, and a very different way for the overloadbrackets
> method to work.

I don't see how 8D limits come in. The limits are the number of
possible arguments to the _OVERLOADBRACKETSLEFTSIDE method, which is
much larger than 8. Unfortunately, the proposed method is deeply
flawed, as there is a unavoidable ambiguity between array and list
indices. Consider:


IDL> la=objarr(2,2)
IDL> for i=1b,4 do la[i-1]=list([i,2*i^2],100b+[i,2*i^2])
IDL> for i=0,3 do print,la[i]
1 2
101 102
2 8
102 108
3 18
103 118
4 32
104 132

Now consider the proposed syntax: la[0,1,1]

Which indices apply to which entity? I.e., is this:

IDL> print,(la[0,1])[1]
103 118

or

IDL> print,((la[0])[1])[1]
102

One option would be to say that arrays of lists gobble up as many
indices as they have dimensions, but that is unnecessarily limiting
and unintuitive, since you might like to index them using a single
scalar as I have. You might also say each list gobbles only only a
single index.

This also point out how the other syntax is superior:

la[0,1][1] vs. la[0][1][1]

would be different things. But I agree, this is not easy to
accommodate at this point.

A few more comments on LIST and HASH:

We need LISTARR and HASHARR to go with OBJARR.

You should be able to assign a single scalar variable to "hash slices"
to set each of the specified keys to that same value, just as you can
assign a single value to a subset of lists:

IDL> a=hash({one:1,two:2})
IDL> a[ ['one','two'] ] = 4 ;; this should work!
% Key and Value must have the same number of elements.
% Error occurred at: HASH::_OVERLOADBRACKETSLEFTSIDE
% $MAIN$
% Execution halted at: $MAIN$


JD

Report message to a moderator

[Message index]
 
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic: modify the attribute value within a tag container
Next Topic: Re: how to calculate latitude and longitude from range and direction(azimuth and elevation)

-=] Back to Top [=-
[ Syndicate this forum (XML) ] [ RSS ] [ PDF ]

Current Time: Fri Oct 10 02:51:40 PDT 2025

Total time taken to generate the page: 0.16080 seconds
.:: Contact :: Home ::.

Powered by: FUDforum 3.0.7.
Copyright ©2001-2013 FUDforum Bulletin Board Software