Re: [D1,static array] fill static multidimensional array.
On Fri, 22 Oct 2010 01:50:56 + (UTC)
%u wrote:
> What is the fastest way to fill a static multidimensional array?
>
> Looping over all dimension's elements sounds inefficient, especially as a
> static array is essentially continuous memory.
> What is the best practice?
>
> int[2][2][2] arr = 0;
> arr[] = 3; //Error: cannot implicitly convert expression (3) of type int to
> int[2u][2u][]
> arr[][][] = 3; // this would be fine too :)
>
> int[2][2][2] arr2 = [[[1,2],[3,4]],[[5,6],[7,8]]];
>
> Somehow I find it surprising that this copies the whole array.
> arr[] = arr2;
I would use the following feature (from ref):
Array Setting
If a slice operator appears as the lvalue of an assignment expression, and the
type of the rvalue is the same as the element type of the lvalue, then the
lvalue's array contents are set to the rvalue.
int[3] s;
int* p;
s[] = 3;// same as s[0] = 3, s[1] = 3, s[2] = 3
p[0..2] = 3;// same as p[0] = 3, p[1] = 3
Using this, one could fill a multidimensional array by setting, so to say, from
inside to outside. This is actually a variant of "(cast(T[n1*n2*n3])arr)[] =
e;" that bearophile proposed, but it does not need any low-level hack:
void main () {
int[2][2][2] arr3;
int[2][2] arr2;
int[2] arr1;
arr1[] = 1;
arr2[] = arr1;
arr3[] = arr2;
assert(arr3 == [[[1, 1], [1, 1]], [[1, 1], [1, 1]]]);
}
I find the syntax somewhat wrong, would prefere it explicit like:
arr.fill(element);
or
arrayFill(arr, element);
How does this method perform? would you be kind enough to include it in your
tests, %u? Even if it's slightly slower than "(cast(int[n1*n2*n3])arr)[] = e;",
I would use it because I find it cleaner.
Is D reflexive enough to know at runtime the number of dimensions of an array
and their sizes, and write a generic
arrayDeepFill(arr, element);
where element is the "terminal" value? Or maybe
arrayDeepFill(arr, element, [sizes]);
Denis
-- -- -- -- -- -- --
vit esse estrany ☣
spir.wikidot.com
custom exception type
Hello,
Where can one find descriptions of Throwable, Error, & Exception? (I mean, how
do you even know they exist?) I could finally guess the constructor must have a
string parameter used for error output.
Also, is it possible to implicitely reuse the superclass's constructor? I had
to write:
class E : Exception {
this (string msg) {
super(msg) ;
}
}
E.this performs nothing new. But without it, I get a compiler error:
trial.d(7): Error: constructor trial.E.this no match for implicit super()
call in constructor
Isn't the constructor inherited like other attributes?
Finally, is it possible to customize the error message construction, using eg
tostring? A big issue is that, currently, an exception's message is computed at
construction time, even if the exception will never be thrown, or more
ccommonly never be output -- because it is caught by a 'catch' clause. In some
cases, constructing the message can be costly; some programming schemes may
throw huge numbers of exceptions, all caught (or nearly all).
Example: in a parsing library, pattern match methods throw an instance of
MatchFailure when matching fails. When there is a pattern choice, there may be
numerous failures for each success. MatchFailure is just a clean way of
signaling this fact (*): each failure exception is caught at a higher level to
allow trying the alternative patterns. Since error messages can be rather
complicated, constructing them uselessly would multiply parsing time by a
rather big factor (in my case, ~ X 30!).
I guess tostring is the right feature for this: it would return the exception's
textual form, ie the message. (For information, this is how Python works.) I
tried to use it, but it seems to be simply ignored. What is the func/method
that constructs the text of an exception, eg what is implicitely called by
"writeln(e);"?
Denis
(*) This is exactly the programming pattern somewhere explained in D docs:
throw an exception instead of returning a fake value used as failure flag.
-- -- -- -- -- -- --
vit esse estrany ☣
spir.wikidot.com
Re: custom exception type
E.this performs nothing new. But without it, I get a compiler error:
trial.d(7): Error: constructor trial.E.this no match for implicit
super() call in constructor
Isn't the constructor inherited like other attributes?
I think if you don't provide a constructor a default one is created:
this(){}
Then an implicit call to super() is inserted cause there's no explicit one.
But there is no this() without parameters in the Exception class if I'm
not mistaken. You can look it up in _object.d in the druntime source.
Re: custom exception type
On Fri, 22 Oct 2010 13:00:43 +0200, spir wrote:
> Hello,
>
> Where can one find descriptions of Throwable, Error, & Exception? (I
> mean, how do you even know they exist?) I could finally guess the
> constructor must have a string parameter used for error output.
Well, they should be in the documentation for the 'object' module, but I
see that they aren't. Until that is fixed, you can check out the
source. Throwable starts at line 1210 here:
http://www.dsource.org/projects/druntime/browser/trunk/src/object_.d
Exception and Error immediately follow it, but they don't really add
anything to Throwable.
> Also, is it possible to implicitely reuse the superclass's constructor?
> I had to write:
>
> class E : Exception {
> this (string msg) {
> super(msg) ;
> }
> }
>
> E.this performs nothing new. But without it, I get a compiler error:
> trial.d(7): Error: constructor trial.E.this no match for implicit
> super() call in constructor
> Isn't the constructor inherited like other attributes?
No. With the exception of a no-argument constructor, which Throwable
doesn't have, you have to call it explicitly with super(...).
> Finally, is it possible to customize the error message construction,
> using eg tostring? A big issue is that, currently, an exception's
> message is computed at construction time, even if the exception will
> never be thrown, or more ccommonly never be output -- because it is
> caught by a 'catch' clause. In some cases, constructing the message can
> be costly; some programming schemes may throw huge numbers of
> exceptions, all caught (or nearly all). Example: in a parsing library,
> pattern match methods throw an instance of MatchFailure when matching
> fails. When there is a pattern choice, there may be numerous failures
> for each success. MatchFailure is just a clean way of signaling this
> fact (*): each failure exception is caught at a higher level to allow
> trying the alternative patterns. Since error messages can be rather
> complicated, constructing them uselessly would multiply parsing time by
> a rather big factor (in my case, ~ X 30!). I guess tostring is the right
> feature for this: it would return the exception's textual form, ie the
> message. (For information, this is how Python works.) I tried to use it,
> but it seems to be simply ignored. What is the func/method that
> constructs the text of an exception, eg what is implicitely called by
> "writeln(e);"?
That would be toString(), not tostring().
-Lars
Re: [D1,static array] fill static multidimensional array.
Somehow I find it surprising that this copies the whole array. arr[] = arr2; Static arrays are value types.
Re: custom exception type
Although it is somewhat annoying that there is no default value for the
msg parameter in the first constructor, it is pretty easy to use the
mixin templates to overcome the issue:
public mixin template ExceptionCtorMixin() {
this(string msg = null, Throwable next = null) { super(msg, next); }
this(string msg, string file, size_t line, Throwable next = null) {
super(msg, file, line, next);
}
}
class MyException : Exception { mixin ExceptionCtorMixin; }
On 22.10.2010 13:00, spir wrote:
Hello,
Where can one find descriptions of Throwable, Error,& Exception? (I mean, how
do you even know they exist?) I could finally guess the constructor must have a
string parameter used for error output.
Also, is it possible to implicitely reuse the superclass's constructor? I had
to write:
class E : Exception {
this (string msg) {
super(msg) ;
}
}
E.this performs nothing new. But without it, I get a compiler error:
trial.d(7): Error: constructor trial.E.this no match for implicit super()
call in constructor
Isn't the constructor inherited like other attributes?
Finally, is it possible to customize the error message construction, using eg
tostring? A big issue is that, currently, an exception's message is computed at
construction time, even if the exception will never be thrown, or more
ccommonly never be output -- because it is caught by a 'catch' clause. In some
cases, constructing the message can be costly; some programming schemes may
throw huge numbers of exceptions, all caught (or nearly all).
Example: in a parsing library, pattern match methods throw an instance of
MatchFailure when matching fails. When there is a pattern choice, there may be
numerous failures for each success. MatchFailure is just a clean way of
signaling this fact (*): each failure exception is caught at a higher level to
allow trying the alternative patterns. Since error messages can be rather
complicated, constructing them uselessly would multiply parsing time by a
rather big factor (in my case, ~ X 30!).
I guess tostring is the right feature for this: it would return the exception's textual
form, ie the message. (For information, this is how Python works.) I tried to use it, but
it seems to be simply ignored. What is the func/method that constructs the text of an
exception, eg what is implicitely called by "writeln(e);"?
Denis
(*) This is exactly the programming pattern somewhere explained in D docs:
throw an exception instead of returning a fake value used as failure flag.
-- -- -- -- -- -- --
vit esse estrany ☣
spir.wikidot.com
Re: [D1,static array] fill static multidimensional array.
== Quote from spir (denis.s...@gmail.com)'s article
> How does this method perform? would you be kind enough to include it in you
> r tests, %u? Even if it's slightly slower than "(cast(int[n1*n2*n3])arr)[]
> = e;", I would use it because I find it cleaner.
But it does take up more memory (assuming you don't generate the arrays every
fill).
And, it performs worse than the casts (with this array length).
If you really want to investigate, I suggest disassembling ;)
Here is the added code:
int[50][50] arr2;
int[50] arr1;
timer.start();
for(int i = 0; i < 10_000; i++) {
arr1[] = i;
arr2[] = arr1;
arr3[] = arr2;
}
timer.stop();
writefln("123 : ",timer.periodCount());
And killing almost every process on my computer the new timings are:
[x][y][z] : 37..
[x][y][]: 38..
cast(int[*]): 28..
cast(int[1]): 28..
123 : 36..
fill: 28..
> Is D reflexive enough to know at runtime the number of dimensions of an arr
> ay and their sizes, and write a generic
> arrayDeepFill(arr, element);
> where element is the "terminal" value? Or maybe
> arrayDeepFill(arr, element, [sizes]);
Yeah, wouldn't that be nice ?!
arr.Fill(elem);
:D
template BaseType(T: T[]) { alias BaseType!(T) BaseType; }
template BaseType(T) { alias T BaseType; }
void Fill(T,E)(T arr, E elem) {
static assert( isStaticArray!(T) );
static assert( is( BaseType!(T) == E ) );
(cast(E[arr.sizeof / elem.sizeof])arr)[] = elem;
}
[D1,__traits] D1 has __traits?
http://www.digitalmars.com/d/1.0/template.html .. assert(__traits(isRef, x[i])); .. static assert(!__traits(compiles, min(3, y) = 10)); These won't compile with my D1 :(
Re: [D1,__traits] D1 has __traits?
On 2010-10-22 20:57, %u wrote: http://www.digitalmars.com/d/1.0/template.html .. assert(__traits(isRef, x[i])); .. static assert(!__traits(compiles, min(3, y) = 10)); These won't compile with my D1 :( D1 neither has __traits or auto ref. I guess that the section was accidentally put in the D1 specification. -- /Jacob Carlborg
Re: [D1,__traits] D1 has __traits?
== Quote from Jacob Carlborg (d...@me.com)'s article > On 2010-10-22 20:57, %u wrote: > > http://www.digitalmars.com/d/1.0/template.html > > .. > > assert(__traits(isRef, x[i])); > > .. > > static assert(!__traits(compiles, min(3, y) = 10)); > > > > These won't compile with my D1 :( > D1 neither has __traits or auto ref. I guess that the section was > accidentally put in the D1 specification. filed
