| Minor obscure error message [message #91815] |
Tue, 01 September 2015 12:12  |
wlandsman
Messages: 743
Registered: June 2000
|
Senior Member |
|
|
Suppose one wants to write !p.thick=2 but forgets the equals sign
IDL> !p.thick2
% Tag name THICK2 is undefined for structure !PLT.
% Execution halted at: $MAIN$
and one gets a very logical error message.
But if one precedes this with the syntactically correct statement !p.background=0
IDL> !p.background=0 & !p.thick2
% Object reference type required in this context: <STRUCT Array[1]>.
% Execution halted at: $MAIN$
one gets an obscure reference about requiring an object type. I'm not sure how IDL is getting confused here. It only happens when the two statements are on the same line.
IDL> print,!version
{ x86_64 linux unix linux 8.4 Sep 27 2014 64 64}
--Wayne
|
|
|
|
| Re: Minor obscure error message [message #91816 is a reply to message #91815] |
Tue, 01 September 2015 14:50  |
Dick Jackson
Messages: 347
Registered: August 1998
|
Senior Member |
|
|
Hi Wayne,
I think this is why: if there is an "&" indicating multiple statements, then "implied print" is not attempted at all on any of them. In your first example, IDL tried to interpret it as a method call on !p (and failed, with no message), then tried to interpret it as an implied print, and failed. Second time, IDL tried as a method call on !p, and that failed. Due to the "&", it did not try implied print, so it reported the first failure.
Here's a simpler example to show a similar effect:
IDL> !Pi
3.1415927
IDL> !Pi & !DPi
% Attempt to call undefined procedure: '!PI'.
Cheers,
-Dick
Dick Jackson Software Consulting Inc.
Victoria, BC, Canada --- http://www.d-jackson.com
On Tuesday, 1 September 2015 12:12:35 UTC-7, wlandsman wrote:
> Suppose one wants to write !p.thick=2 but forgets the equals sign
>
> IDL> !p.thick2
> % Tag name THICK2 is undefined for structure !PLT.
> % Execution halted at: $MAIN$
>
> and one gets a very logical error message.
>
> But if one precedes this with the syntactically correct statement !p.background=0
>
> IDL> !p.background=0 & !p.thick2
> % Object reference type required in this context: <STRUCT Array[1]>.
> % Execution halted at: $MAIN$
>
> one gets an obscure reference about requiring an object type. I'm not sure how IDL is getting confused here. It only happens when the two statements are on the same line.
>
> IDL> print,!version
> { x86_64 linux unix linux 8.4 Sep 27 2014 64 64}
>
>
> --Wayne
|
|
|
|
| Re: Minor obscure error message [message #91817 is a reply to message #91816] |
Tue, 01 September 2015 19:01 |
wlandsman
Messages: 743
Registered: June 2000
|
Senior Member |
|
|
Dick,
Thanks for the explanation! I was still confused for a bit about why IDL does not know that !P is a structure, and instead tries to interpret it as an object. The answer is likely that
IDL> structure.tag
can never be a valid syntax (if you don't allow implied print), but
IDL> object.method
can be a valid syntax. --Wayne
On Tuesday, September 1, 2015 at 5:50:09 PM UTC-4, Dick Jackson wrote:
> Hi Wayne,
>
> I think this is why: if there is an "&" indicating multiple statements, then "implied print" is not attempted at all on any of them. In your first example, IDL tried to interpret it as a method call on !p (and failed, with no message), then tried to interpret it as an implied print, and failed. Second time, IDL tried as a method call on !p, and that failed. Due to the "&", it did not try implied print, so it reported the first failure.
>
> Here's a simpler example to show a similar effect:
>
> IDL> !Pi
> 3.1415927
> IDL> !Pi & !DPi
> % Attempt to call undefined procedure: '!PI'.
>
> Cheers,
> -Dick
>
> Dick Jackson Software Consulting Inc.
> Victoria, BC, Canada --- http://www.d-jackson.com
>
> On Tuesday, 1 September 2015 12:12:35 UTC-7, wlandsman wrote:
>> Suppose one wants to write !p.thick=2 but forgets the equals sign
>>
>> IDL> !p.thick2
>> % Tag name THICK2 is undefined for structure !PLT.
>> % Execution halted at: $MAIN$
>>
>> and one gets a very logical error message.
>>
>> But if one precedes this with the syntactically correct statement !p.background=0
>>
>> IDL> !p.background=0 & !p.thick2
>> % Object reference type required in this context: <STRUCT Array[1]>.
>> % Execution halted at: $MAIN$
>>
>> one gets an obscure reference about requiring an object type. I'm not sure how IDL is getting confused here. It only happens when the two statements are on the same line.
>>
>> IDL> print,!version
>> { x86_64 linux unix linux 8.4 Sep 27 2014 64 64}
>>
>>
>> --Wayne
|
|
|
|
| Re: Minor obscure error message [message #91819 is a reply to message #91816] |
Wed, 02 September 2015 00:36 |
Yngvar Larsen
Messages: 134
Registered: January 2010
|
Senior Member |
|
|
On Tuesday, 1 September 2015 23:50:09 UTC+2, Dick Jackson wrote:
> Hi Wayne,
>
> I think this is why: if there is an "&" indicating multiple statements, then "implied print" is not attempted at all on any of them.
Apparently you are right. But then the question is: why not? Is there a compelling reason why expressions on a single line separated by "&" can't treated separately, with implied print where relevant? E.g. something like this:
IDL> !pi & a = 1 & !dpi & b = 0
3.1415927
3.1415926535897931
IDL> a, b
1
0
--
Yngvar
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
