passing static arrays to each! with a ref param [Re: Why can't static arrays be sorted?]
On Monday, 10 October 2016 at 16:46:55 UTC, Jonathan M Davis
wrote:
On Monday, October 10, 2016 16:29:41 TheGag96 via
Digitalmars-d-learn wrote:
On Saturday, 8 October 2016 at 21:14:43 UTC, Jon Degenhardt
wrote:
> This distinction is a bit on the nuanced side. Is it
> behaving as it should?
>
> --Jon
I think so? It's not being modified in the second case because
the array is being passed by value... "x" there is a reference
to an element of the copy created to be passed to each(). I
assume there's a good reason why ranges in general are passed
by value into these functions -- except in this one case, the
stuff inside range types copied when passed by value won't be
whole arrays, I'm guessing.
Whether it's by value depends entirely on the type of the
range. They're passed around, and copying them has whatever
semantics it has. In most cases, it copies the state of the
range but doesn't copy all of the elements (e.g. that's what
happens with a dynamic array, since it gets sliced). But if a
range is a class, then it's definitely a reference type. The
only way to properly save the state of a range is to call save.
But passing by ref would make no sense at all with input
ranges. It would completely kill chaining them. Almost all
range-based functions return rvalues.
- Jonathan M Davis
The example I gave uses ref parameters. On the surface it would
seem reasonable to that passing a static array by ref would allow
it to be modified, without having to slice it first. The
documentation says:
// If the range supports it, the value can be mutated in place
arr.each!((ref n) => n++);
assert(arr == [1, 2, 3, 4, 5]);
but, 'arr' is a dynamic array, so technically it's not describing
a static array (the opApply case).
Expanding the example, using foreach with ref parameters will
modify the static array in place, without slicing it. I would
have expected each! with a ref parameter to behave the same.
At a minimum this could be better documented, but it may also be
a bug.
Example:
T increment(T)(ref T x) { return x++; }
void main()
{
import std.algorithm : each;
int[] dynamicArray = [1, 2, 3, 4, 5];
int[5] staticArray = [1, 2, 3, 4, 5];
dynamicArray.each!(x => x++); // Dynamic array by
value
assert(dynamicArray == [1, 2, 3, 4, 5]); // ==> Not modified
dynamicArray.each!((ref x) => x++); // Dynamic array by
ref
assert(dynamicArray == [2, 3, 4, 5, 6]); // ==> Modified
staticArray[].each!((ref x) => x++); // Slice of static
array, by ref
assert(staticArray == [2, 3, 4, 5, 6]); // ==> Modified
staticArray.each!((ref x) => x++);// Static array by
ref
assert(staticArray == [2, 3, 4, 5, 6]); // ==> Not Modified
/* Similar to above, using foreach and ref params. */
foreach (ref x; dynamicArray) x.increment;
assert(dynamicArray == [3, 4, 5, 6, 7]); // Dynamic array =>
Modified
foreach (ref x; staticArray[]) x.increment;
assert(staticArray == [3, 4, 5, 6, 7]); // Static array
slice => Modified
foreach (ref x; staticArray) x.increment;
assert(staticArray == [4, 5, 6, 7, 8]); // Static array =>
Modified
}
Re: MemberDefaults trait
On 10/10/2016 04:50 PM, Ali Çehreli wrote:
> static if (!is(T == struct)) {
> static assert(T.stringof ~ " is not a struct type");
> }
Wow! That's a nice brain fart on my part. Ok, I can fix that one... :)
Ali
Re: Current State of the GC?
On 11/10/2016 10:12 AM, Martin Lundgren wrote: I've been reading up a bit on the D garbage collector. Seen mostly negative things about it. I've also seen a lot of proposals and what not, but not much about the current state of things. The latest page I can find about it is 2015H1. It mentions improving the GC and making libraries less reliant on it. However, I can't find *any* information about what GC improvements have been made. No up to date performance comparisons, etc. So what's been happening in memory management land lately? Bad GC seems like one of the Dlangs weak points, so showing improvements here could definitely bring more people in. Well I can't say what has happened since that half way document. Most of the work that goes on is minor tweaks and improvements that pretty much nobody outside of druntime knows about and that is quite all right. If you want to actually see all these things going on check out Github[0]. Anyway, most of the time the GC isn't a problem, contrary to popular belief. As long as you do tricks like reusing memory it will never fire and more importantly you won't be hit with memory allocation costs. So basically a double win for you in terms of speed. When dealing with multi threading you probably want to disable the GC collection and have predefined points so it can collect safely with no performance hits. In reality, except for some cycle counting points, you won't need a million dollar GC and even then they tend to fail at those sort of jobs. [0] https://github.com/dlang/druntime/tree/2db828bd4f21807254b770b3ec304f14596a9805/src/gc
MemberDefaults trait
Could you please review the following template to see whether it makes
sense. It produces an AliasSeq type consisting of the default values of
the members of a struct. It should and does support members that are
initialized with '= void'. I could not achieve this with std.traits or
__traits.
However, I'm not that happy with the use of __traits(compiles) below.
Can this be improved?
Thank you,
Ali
import std.meta;
/** Get as an expression tuple the default values of members of a struct. */
template MemberDefaults(T) {
static if (!is(T == struct)) {
static assert(T.stringof ~ " is not a struct type");
}
import std.traits : FieldNameTuple;
enum t = T.init;
alias fields = FieldNameTuple!T;
template get(size_t i) {
static if (__traits(compiles, { enum _ = t.tupleof[i]; })) {
enum get = t.tupleof[i];
}
else {
alias get = void;
}
}
template Impl(size_t i = 0) {
import std.meta : AliasSeq;
static if (i == fields.length) {
alias Impl = AliasSeq!();
} else {
alias Impl = AliasSeq!(get!i, Impl!(i+1));
}
}
alias MemberDefaults = Impl!();
}
unittest {
struct S {
int i = 42;
string s = "hello";
char c = 'c';
}
static assert(MemberDefaults!S == AliasSeq!(42, "hello", 'c'));
}
unittest {
struct S {
int i = 42;
int j = void;
}
static assert(MemberDefaults!S[0] == 42);
static assert(is(MemberDefaults!S[1] == void));
}
void main() {
}
Re: How to correctly display accented characters at the Windows prompt?
Windows. You can try command cp 65001 in the console window chcp 65001
Re: How to correctly display accented characters at the Windows prompt?
On Monday, 10 October 2016 at 19:31:14 UTC, Cleverson Casarin Uliana wrote: Hi Martin, indeed, here in my workplace Windows machine there is no "CP_UTF_8" codepage, nor there is 65001. I was waiting to codepage 65001 (UTF8) should be available on all modern Windows. You can try command cp 65001 in the console window (and change your font accordingly). Just for info, this is probably a d related bug, because the Racket language for example has a (reencode-output-port) function, and it Works. I don't know know about Racket, but D's problems are related to the C runtime library. Here is what I found out: -m32 (uses DM C library): output to stdout works, but stderr is broken (corrupt accented characters, or total output freeze). https://issues.dlang.org/show_bug.cgi?id=1448 -m32mscoff and -m64 (Microsoft C runtime): broken characters and some mysterious failures. https://issues.dlang.org/show_bug.cgi?id=15761 So this leaves transcoding as the only viable solution. Luckily, there is std.windows.charset.toMBSz function you can use.
Re: How to correctly display accented characters at the Windows prompt?
Hi Martin, indeed, here in my workplace Windows machine there is no "CP_UTF_8" codepage, nor there is 65001. I was waiting to try it later on my home machine, but since you say it's broken, then I'll need to look for a way to convert the actual string to the 850 codepage... Just for info, this is probably a d related bug, because the Racket language for example has a (reencode-output-port) function, and it Works. Greetings, Cleverson
Re: opIndexDispatch?
On Monday, October 10, 2016 19:01:19 Yuxuan Shui via Digitalmars-d-learn wrote: > Hi, > > Why is there no opIndexDispatch for overloading a[x].func() ? There's opIndex for overloading a[x], and then you can call a function on the return value. If you want some kind of opDispatch on the return value, then the return type will need to implement opDispatch. - Jonathan M Davis
opIndexDispatch?
Hi, Why is there no opIndexDispatch for overloading a[x].func() ?
Re: ptrdiff_t of class.tupleof entry
On Monday, October 10, 2016 17:57:15 Satoshi via Digitalmars-d-learn wrote:
> Hello,
> How can I take an ptrdiff_t of any tupleoff entry from class
> during compilation?
>
>
> I need something like this:
>
> template Test(T, int i) {
> enum Test = [i] - [0];
> }
>
>
> and then use it like:
>
> class AnyClass {
> int a;
> int b;
> int c;
> string d;
>
> }
>
> enum Addr = Test!(AnyClass, 3);
>
> void assignInt!(T)(AnyClass cls, T value) {
> *(cast(T *)cast(void *)cls + Addr) = value;
> }
>
>
>
> This is just an example how I need to use it.
> Is something like this possible to do?
You can use the offsetof property of a member variable to find out the
offset between its address and the address of the beginning of the class or
struct that it's a member of.
- Jonathan M Davis
ptrdiff_t of class.tupleof entry
Hello,
How can I take an ptrdiff_t of any tupleoff entry from class
during compilation?
I need something like this:
template Test(T, int i) {
enum Test = [i] - [0];
}
and then use it like:
class AnyClass {
int a;
int b;
int c;
string d;
}
enum Addr = Test!(AnyClass, 3);
void assignInt!(T)(AnyClass cls, T value) {
*(cast(T *)cast(void *)cls + Addr) = value;
}
This is just an example how I need to use it.
Is something like this possible to do?
Re: How to correctly display accented characters at the Windows prompt?
Thanks, Cleverson Call SetConsoleOutputCP(CP_UTF8). No, this may appear to to work, but in reality, it's broken. The only reliable way is to convert to the native windows codepage.
Re: isRvalue trait
On Monday, 10 October 2016 at 12:22:54 UTC, Marc Schütz wrote: I would like to overload to an identity op. If the compiler is smart enough to understand what `moveEmplace()` does, it could already do this automatically. Doh! My mistake. I'll use `moveEmplace`, then. Thx!
Re: Easy sockets - don't exist yet?
On Monday, 10 October 2016 at 07:37:48 UTC, Bauss wrote: Wrote some pretty simple sockets that you could use (Based on vibe.d though.) https://github.com/bausshf/cheetah Hi, Yes I saw it, but not sure. Does it make sense to use vibe.d only for sockets. I mean, it like a web framework with many dependencies etc...
Re: Easy sockets - don't exist yet?
On Monday, 10 October 2016 at 02:54:09 UTC, Jonathan M Davis wrote: On Monday, October 10, 2016 01:43:54 Konstantin Kutsevalov via So, it's simply gone. But if someone wants to propose a replacement, they're certainly still free to do so. - Jonathan M Davis I see, thank you for answer
Re: Why can't static arrays be sorted?
On Monday, October 10, 2016 16:29:41 TheGag96 via Digitalmars-d-learn wrote: > On Saturday, 8 October 2016 at 21:14:43 UTC, Jon Degenhardt wrote: > > This distinction is a bit on the nuanced side. Is it behaving > > as it should? > > > > --Jon > > I think so? It's not being modified in the second case because > the array is being passed by value... "x" there is a reference to > an element of the copy created to be passed to each(). I assume > there's a good reason why ranges in general are passed by value > into these functions -- except in this one case, the stuff inside > range types copied when passed by value won't be whole arrays, > I'm guessing. Whether it's by value depends entirely on the type of the range. They're passed around, and copying them has whatever semantics it has. In most cases, it copies the state of the range but doesn't copy all of the elements (e.g. that's what happens with a dynamic array, since it gets sliced). But if a range is a class, then it's definitely a reference type. The only way to properly save the state of a range is to call save. But passing by ref would make no sense at all with input ranges. It would completely kill chaining them. Almost all range-based functions return rvalues. - Jonathan M Davis
Re: Why can't static arrays be sorted?
On Saturday, 8 October 2016 at 21:14:43 UTC, Jon Degenhardt wrote: This distinction is a bit on the nuanced side. Is it behaving as it should? --Jon I think so? It's not being modified in the second case because the array is being passed by value... "x" there is a reference to an element of the copy created to be passed to each(). I assume there's a good reason why ranges in general are passed by value into these functions -- except in this one case, the stuff inside range types copied when passed by value won't be whole arrays, I'm guessing.
Re: How to correctly display accented characters at the Windows prompt?
On Monday, October 10, 2016 13:24:09 Cleverson Casarin Uliana via Digitalmars- d-learn wrote: > Hello John and all, how do you search for a given function to see > where it is defined/declared? I tried to find SetConsoleOutputCP by > myself, but the search embeded in the documentation cannot find > anything in the phobos library refference, and searching the entire > site returns forum posts only. It looks like it's in core.sys.windows.wincon; - Jonathan M Davis
Re: How to correctly display accented characters at the Windows prompt?
Hello John and all, how do you search for a given function to see where it is defined/declared? I tried to find SetConsoleOutputCP by myself, but the search embeded in the documentation cannot find anything in the phobos library refference, and searching the entire site returns forum posts only. Thanks, Cleverson
Re: How to correctly display accented characters at the Windows prompt?
On Monday, 10 October 2016 at 14:14:08 UTC, Cleverson Casarin Uliana wrote: Hello, when I run a compiled Windows executable at the console, letters outside the ascii range such as ç and ã do not display propperly. Is there any d function to change the console code page on the fly? My Windows console code page, which is for Brazilian Portuguese, is at 850. Perhaps an alternative would be to convert d strings from Unicode to the 850 code page on the fly. Thanks, Cleverson Call SetConsoleOutputCP(CP_UTF8).
How to correctly display accented characters at the Windows prompt?
Hello, when I run a compiled Windows executable at the console, letters outside the ascii range such as ç and ã do not display propperly. Is there any d function to change the console code page on the fly? My Windows console code page, which is for Brazilian Portuguese, is at 850. Perhaps an alternative would be to convert d strings from Unicode to the 850 code page on the fly. Thanks, Cleverson
Re: isRvalue trait
On Monday, 10 October 2016 at 11:46:01 UTC, Nordlöw wrote:
At
https://github.com/nordlow/phobos-next/blob/master/src/moval.d
I've implemented a helper function for creating r-value out of
l-values defined as
E movedToRvalue(E)(ref E e)
{
import std.algorithm.mutation : move;
E value;
move(e, value); // this can be optimized
return value;
}
Doesn't the second overload of `move()` already do what you want?
https://dlang.org/phobos/std_algorithm_mutation.html#.move
Note that in your implementation, you needlessly initialize
`value`, which then needs to be properly destroyed by `move()`,
which the Phobos implementation avoids:
https://github.com/dlang/phobos/blob/master/std/algorithm/mutation.d#L1083-L1088
For the case when movedToRvalue is called with an r-value, such
as in,
static assert(__traits(compiles, { consume(S(14)); }));
I would like to overload to an identity op.
If the compiler is smart enough to understand what
`moveEmplace()` does, it could already do this automatically.
Re: isRvalue trait
On Monday, 10 October 2016 at 11:51:09 UTC, Nordlöw wrote: Found it: http://forum.dlang.org/post/n8m8bh$2vgc$1...@digitalmars.com https://issues.dlang.org/show_bug.cgi?id=15634 Ok, so I added `isLvalue` and `isRvalue` to https://github.com/nordlow/phobos-next/blob/master/src/moval.d Now the next follow-up question becomes how to use it to restrict/overload my definition of `movedToRvalue()` since `isLvalue` takes an alias as argument and `movedToRvalue` takes a reference to an instance. How to solve this?
Re: isRvalue trait
On Monday, 10 October 2016 at 11:46:01 UTC, Nordlöw wrote: Is there one? Found it: http://forum.dlang.org/post/n8m8bh$2vgc$1...@digitalmars.com https://issues.dlang.org/show_bug.cgi?id=15634
isRvalue trait
At
https://github.com/nordlow/phobos-next/blob/master/src/moval.d
I've implemented a helper function for creating r-value out of
l-values defined as
E movedToRvalue(E)(ref E e)
{
import std.algorithm.mutation : move;
E value;
move(e, value); // this can be optimized
return value;
}
For the case when movedToRvalue is called with an r-value, such
as in,
static assert(__traits(compiles, { consume(S(14)); }));
I would like to overload to an identity op.
Is this currently possible somehow? I cannot find any trait
`isRvalue` that fulfills
static assert(isRvalue!(S(14)));
Is there one?
Re: When to call GC.{add,remove}Range in containers?
On Monday, 10 October 2016 at 08:26:45 UTC, Kagamin wrote: On Monday, 10 October 2016 at 07:12:10 UTC, Nordlöw wrote: should not be scanned by the GC. Shouldn't be a problem. What do you mean? I'm talking about an optimization; don't call addRange when we don't need to, because the internal stored `E* _ptr` is maintained manually by calls to `*alloc` and `free`.
Re: How do I load a shared library?
Fixed this issue, stackoverflow helped http://stackoverflow.com/questions/39929495 just compiled the library with -fPIC -m32 -shared
Re: How do I load a shared library?
anyone? so there is no way to get my 50,000 Line of code to work again? All that code to waste? P.S: I've tried doing these: Tried to use GDC, no luck used -defaultlib=libphobos2.so, no luck removed all functions from library, compiled it empty, yet, dlopen gives segFault. AND: This same code used to work on ubuntu 14.04 (i386), now I'm on xubuntu 16.04 (amd64).
Re: weighted round robin
On Saturday, 8 October 2016 at 22:48:53 UTC, vino wrote:
Hi,
Can some one guide me on how to implement the weighted round
robin, below is what i tried or any other better ways to do it
Main Requirement : Incoming socket connection has to be sent to
3 servers in the weighted round robin fashion.
Prog:1
import std.stdio;
import std.range;
import std.range.primitives;
void main()
{
auto a = [1,2,3]; // E.g :Server Array
auto b = [1,2,3,4,5]; // E.g: Socket Array
auto r = roundRobin(a, b);
writeln(r);
}
OUTPUT : [1, 1, 2, 2, 3, 3, 4, 5]
Requirement : [1, 1, 2, 2, 3, 3,1,4,2,5]
auto r = roundRobin(a.cycle, b.cycle);
Beware though that this yields an infinite range. If you just
need one round, you can use:
import std.algorithm.comparison : max;
writeln(r.take(max(a.length, b.length)));
Re: When to call GC.{add,remove}Range in containers?
On Monday, 10 October 2016 at 07:12:10 UTC, Nordlöw wrote: should not be scanned by the GC. Shouldn't be a problem.
Re: dub command line in config?
On Sunday, 9 October 2016 at 20:03:58 UTC, WhatMeWorry wrote: On Sunday, 9 October 2016 at 19:11:52 UTC, Jinx wrote: On Sunday, 9 October 2016 at 08:52:55 UTC, rikki cattermole wrote: On 09/10/2016 9:17 PM, Jinx wrote: [...] That is enough. Mike Parker has presented a workaround that you can implement. He has also shown how you can contact those that have the power to make this happen as you desire. But as shown by other related issues[0], you may not get what you want. [0] https://github.com/dlang/dub/issues/940 Yessim boss! plez dunt hurt me! he shoed ma how too puck da cotton n giv me dis hur bag. I alredy no hew to puck da cotton n hud me own bag. Hey Jinx. Why don't you rename yourself to Jerk. This is not a company - this is open source. I understand that one can be frustrated sometimes, but it should be kept in mind that all contributions happen voluntarily in people's free time. If you want something to be changed (and can't wait that someone does it in their free time), please submit a PR.
Re: Easy sockets - don't exist yet?
Wrote some pretty simple sockets that you could use (Based on vibe.d though.) https://github.com/bausshf/cheetah
When to call GC.{add,remove}Range in containers?
Which std.trait should be used to statically check whether I
should call GC.{add,remove}Range on the elements of D containers?
`std.container.array.Array` currently uses `hasIndirections` but
a comment on the same line says it should use `hasPointers`
instead.
containers-em uses
template shouldAddGCRange(T)
{
import std.traits;
enum shouldAddGCRange = isPointer!T || hasIndirections!T || is
(T == class);
}
Quite some inconsistency here.
And what the case `Array!(Array!int))`? The wrapper Array!int
contains contains a non-GC allocate pointer to ints and should
not be scanned by the GC. Do we need another Container-trait for
this? Or do we need to tag pointers with a special attribute that
tells whether it has been allocated by GC or not.
Re: Problems with "shared"
On Sunday, 9 October 2016 at 21:32:23 UTC, ag0aep6g wrote:
On 10/09/2016 10:57 PM, jython234 wrote:
1. Where do I use the "shared" keyword?
What doesn't work is creating an unshared object of a class
that only has a shared constructor. To create both shared and
unshared objects, you need either two constructors or a `pure`
one.
class C
{
this() pure {}
}
void main()
{
auto u = new C; /* works */
auto s = new shared C; /* too */
}
Why this ins't in doc? :O I didn't know about using pure in this
way.
