clipka <ano### [at] anonymousorg> wrote:
> Am 22.05.2021 um 13:33 schrieb jr:
>
> > thinking that what your + CR's (ingenious) posts demonstrate is just how much an
> > 'istype()' like function is needed in SDL.  (I've a feeling it would not be too
> > difficult to implement, a developer's perspective would be good)
>
> "Pfffffffaaaaaaaaaaaaahahahahaha! BWAAAAHAHAHAHAHAHA!"
> (*manic laughter*)
> -- Anonymous developer.
>
> Keeping track of the type of things IN THE PARSER ITSELF is painful
> enough... add to that the trouble of coming up with a nice elegant
> syntax, questions such as whether this should work only on variables, or
> also literals or even complex expressions...
>
> ... oh, and then add to the mix that the parser doesn't go, "ah, let me
> evaluate and see what the result of this expression is", but rather "ah,
> I expect this particular type here; let me see how the following makes
> any sense under that presumption"...
> ...

and yet, WFP's current 'is_type()' is an excellent start on all that.  the thing
is that _any_ "introspection" support in SDL would be a step forward.


regards, jr.

Post a reply to this message

I had occasion to revisit this, as I wanted to determine the "size" of a vector,
because having separate macros to compare vectors is ridiculous.

The way that POV-Ray promotes vectors and does scalar+vector addition seemed to
frustrate things.  But since the parser "knows" things about the values that
we're using, I was wondering if I could use some of that occult knowledge and
somehow distinguish between scalar and vector values.

trying to write efficient code led me to discover that you can write rgb 3

but if I: store 3 in an array and then try to write rgb Array[0]
or try rgbf, rgbt, or rgbft 3
I get: "Parse Error: Attempted to redfine float identifier as colour
identifier."

But we CAN write rgb 3, which gets promoted to <3, 3, 3, 0, 0>
Whereas actual vectors just get padded with zeros, unless it's a 5D vector.

So we can check if rgb v0 has x=y=z, and also check if
the transmit value of rgb v0 = v0 + <0, 0, 0, 0, 0>

Because if v0 is non-zero, then the former will be 0, and the latter will be v0.

Perhaps there is more that we can do in the same vein.

Developmental scene is attached.


"Tor Olav Kristensen" <tor### [at] TOBEREMOVEDgmailcom> wrote:

> #macro VectorDimTest(v0)
>
>   #local vZ3D = <0, 0, 0>;
>   #local vZ4D = <0, 0, 0, 0>;
>   #local vZ5D = <0, 0, 0, 0, 0>;
>   #local v1 = v0 + 1;
>   #local v2 = v0 + 2;
>   #local D2 = ((vZ3D + v1).z = (vZ3D + v2).z);
>   #local D3 = ((vZ4D + v1).t = (vZ4D + v2).t);
>   #local D4 = ((vZ5D + v1).transmit = (vZ5D + v2).transmit);
>
>   #if (D2)
>       #debug "2D"
>   #else
>       #if (D3)
>           #debug "3D"
>       #else
>           #if (D4)
>               #debug "4D"
>           #else
>               #debug "5D or scalar"
>           #end // if
>       #end // if
>   #end // if
>
> #end // VectorDimTest

//#declare Vector = array mixed {pi, <1, 2>, <1, 2, 3>, <1, 2, 3, 4>, <1, 2, 3,
4, 5>}
//#declare Vector = array mixed {rgb pi, rgb <1, 2>, rgb <1, 2, 3>, rgb<1, 2, 3,
4>, rgb <1, 2, 3, 4, 5>}
#declare Vector = array mixed {
 rgb pi, rgb pi + 1, rgb pi + 2,
 rgb <1, 2>, rgb <1, 2>+1, rgb <1, 2>+2,
 rgb <1, 2, 3>, rgb <1, 2, 3>+1, rgb <1, 2, 3>+2,
 rgb<1, 2, 3, 4>, rgb<1, 2, 3, 4>+1, rgb<1, 2, 3, 4>+2,
 rgb <1, 2, 3, 4, 5>, rgb <1, 2, 3, 4, 5>+1, rgb <1, 2, 3, 4, 5>+2
}
//#declare Vector = array mixed {<0, 0>+1, <0, 0>+2, <0, 0, 0>+1, <0, 0, 0>+2,
<0, 0, 0, 0>+1, <0, 0, 0, 0>+2, <0, 0, 0, 0, 0>+1, <0, 0, 0, 0, 0>+2}

#declare VN = dimension_size (Vector, 1)-1;

#for (V, 0, VN)
 //#debug concat ("Vector = <", vstr(1+ceil((V+1)/2), Vector [V], ", ", 0, 0),
">           ")
 #debug concat ("Vector = <", vstr(5, Vector [V], ", ", 0, 0), ">           ")
 VectorDimTest(Vector [V])
 #debug "\n"
 #if (mod(V+1,3)=0)
  #debug "\n"
 #end
#end

#debug "\n"
#debug "\n"
#debug "\n"

Post a reply to this message

Attachments:
Download 'vectoroperations.pov.txt' (3 KB)

I do believe I have cracked the code!
I made a big spreadsheet to get a better overview of everything that was
happening with additions and automatic vector promotion.

It took me a while to get a good feel for what was possible, and what was just
going around in circles.
I was also concerned about all zero values being a special case.

Then I got some inspiration for leveraging the "asymmetry" afforded by the rgb
keyword.  Squaring v0 gives me all-positive values, so when I add 1 to the
result, I'm not losing the result I want when v0 is some form of -1.
Then I add the rgb conversion of that value, and that lets me differentiate
between scalars and all sizes of vectors.

I started making a huge list of test values, but I'm sure there's a better way
to automate the generation of 1D-5D values with all possible combinations of -,
0, +.

At the moment, I haven't run across any values that get improperly identified.
I separately list all vector components to show that accessing the individual
dot-notation components doesn't raise a "Bad operands for period operator"
error.

(I tried to use -GW or Warning_Console=Off, but the "Suspicious expression after
rgb" error doesn't get suppressed)

Check it out.
Hammer on it.
Try to break it.

- Bald "Nothing is Impossible!" Eagle

Post a reply to this message

Attachments:
Download 'vectoroperations.pov.txt' (4 KB)