| Re: array subscript conversion [message #55032 is a reply to message #55031] |
Wed, 25 July 2007 11:04   |
Foldy Lajos
Messages: 268
Registered: October 2001
|
Senior Member |
|
|
On Wed, 25 Jul 2007, mgalloy@gmail.com wrote:
> On Jul 25, 9:36 am, FÖLDY Lajos <fo...@rmki.kfki.hu> wrote:
>> Hi guys,
>>
>> according to the manual, array subscripts are converted to long (or long64
>> on 64 bit systems) before use if necessary, so an explicit conversion
>> should not affect the result.
>>
>> IDL> print, !version
>> { x86 linux unix linux 6.3 Mar 23 2006 32 64}
>> IDL>
>> IDL> a=lindgen(10)
>> IDL> print, a[[long(-1ull)]]
>> 0
>> IDL> print, a[[-1ull]]
>> 9
>>
>> Is it a bug or I am missing something?
>>
>> regards,
>> lajos
>
> All I see in the manual about converting subscripts to longs is:
>
> "Subscripts can be any type of vector or scalar expression. If a
> subscript expression is not integer, a longword integer copy is made
> and used to evaluate the subscript."
>
> around the middle of this page:
>
> http://idlastro.gsfc.nasa.gov/idl_html_help/Understanding_Ar ray_Subscripts.html
>
> By "integer", I think they mean the more general any integer type:
> byte, integer, long, etc. And in your case, the type is "integer", so
> no conversion is made.
>
> I am I missing another statement in the help that says something more
> explicit?
>
> Mike
> --
> www.michaelgalloy.com
>
You are right. I remembered "not long" instead of "not integer". And also
remembered IDL_MEMINT from idl_export.h, which represents memory offsets
and sizes, and assumed that IDL always uses IDL_MEMINT for subscripting
internally.
Second try:
IDL> a=lindgen(10)
IDL>
IDL> print, a[4294967296ll]
0
IDL> print, a[[4294967296ll]]
9
In the first case the scalar subscript is "integer", so no conversion is
needed. But it is converted to LONG (=0). Why?
regards,
lajos
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
