A look inside "filter" function defintion

[Patrick Schluter]

On Tuesday, 2 August 2022 at 12:39:41 UTC, pascal111 wrote:
> On Tuesday, 2 August 2022 at 04:06:30 UTC, frame wrote:
>> On Monday, 1 August 2022 at 23:35:13 UTC, pascal111 wrote:
>>> This is the definition of "filter" function, and I think it 
>>> called itself within its definition. I'm guessing how it 
>>> works?
>>
>> It's a template that defines the function called "Eponymous 
>> Templates":
>> https://dlang.org/spec/template.html#implicit_template_properties
>>
>> A template generates code, it cannot be called, only 
>> instantiated.
>>
>> The common syntax is just a shortcut for using it. Otherwise 
>> you would need to write `filter!(a => a > 0).filter([1, -1, 2, 
>> 0, -3])`. Like UFCS, some magic the compiler does for you.
>
> Instantiation seems some complicated to me. I read "If a 
> template contains members whose name is the same as the 
> template identifier then these members are assumed to be 
> referred to in a template instantiation:" in the provided link, 
> but I'm still stuck. Do you have a down-to-earth example for 
> beginners to understand this concept?

A template is conceptually like a macro with parameters in C. An 
instantiation is like the using of the macro in your C program. 
The fundamental difference is, that the template is syntactically 
and semantically linked to the language. In C, the preprocessor 
was just a textual replacement done before the proper 
compilation. This meant that there are things that you couldn't 
do in the pre-processor (like `#if sizeof(int)==4`) and 
(horrible) things that never should have been possible (I used to 
use the C pre-processor with other languages like AutoLISP and 
dBase III).