Re: It is a bug?

[Daniel Kozak via Digitalmars-d-learn]

On Wednesday, 30 July 2014 at 16:14:56 UTC, Jesse Phillips wrote:

On Wednesday, 30 July 2014 at 07:08:17 UTC, Kozzi11 wrote:

#main.d:

import m.f;
class A {
   //class main.A member m is not accessible
   //mixin(t!(typeof(this), m));
   void m() {};
   //here is ok
   //mixin(t!(typeof(this), m));
}


The compiler is trying to construct type A. The first thing it 
sees is a need to get the members of the class. What members 
are in the class? I don't know I'm trying to mix one in right 
now!


Yes, this make sense. But why it is not a problem when 't' 
template is in the same file.

Re: pointer array?

[Daniel Kozak via Digitalmars-d-learn]

V Wed, 30 Jul 2014 14:33:51 +
seany via Digitalmars-d-learn digitalmars-d-learn@puremagic.com
napsáno:

 In Ali's excllent book, somehow one thing has escaped my 
 attention, and that it the mentioning of pointer arrays.
 
 Can pointers of any type of pointed variable be inserted in an 
 int array? Using to!(int) perhaps? If not directly, then what 
 else would achieve the same effect?

It depends on pointer size, for eg, on 64bit system with 64bit pointers
something like this should works:
long[] arr = (cast(long*)pointersArray.ptr)[0 .. pointersArray.length];
or
long[] arr = cast(long[])pointersArray; // but I am not sure if this is
OK

Re: Member access of __gshared global object

[Daniel Kozak via Digitalmars-d-learn]

V Thu, 31 Jul 2014 02:03:35 +
Puming via Digitalmars-d-learn digitalmars-d-learn@puremagic.com
napsáno:

 Hi,
 
 I'm writing this global Config class, with an AA member:
 
 ```d
 module my.config;
 
 class Config
 {
  Command[string] commands;
 }
 
 __gshared Config CONFIG;
 ```
 
 and initialize it in another module:
 
 ```d
 module my.app;
 
 import my.config;
 
 void main()
 {
CONFIG = new Config();
CONFIG.commands[bye] = new Command(...); // add commands
 }
 ```
 
 This is OK. But when I use a local variable to hold the commands 
 AA:
 
 ```
 auto cmds = CONFIG.commands;
 cmds[list] = new Command(...);
 ```
 
 The command list is not added.
 
 I guess what happened here was that `cmds` is a threadlocal 
 variable, so the compiler somehow copied the CONFIG.commands.
 
 My questions are:
 
 1. Are AAs reference type? if so, why does the compiler copy it?
 2. How do I reference a member of __gshared global objects?

can you post code somewhere? I try it and it works for me.

module main;

import std.stdio;
import config;

void main(string[] args)
{
CONFIG = new Config();
CONFIG.commands[bye] = yep;

auto cmds = CONFIG.commands;
cmds[list] = smt;

writeln(CONFIG.commands);
// Lets the user press Return before program returns
stdin.readln();
}

module config;

class Config
{
string[string] commands;
this()
{
// Constructor code
}
}

__gshared Config CONFIG;

Re: Split class declaration and definition

[Daniel Kozak via Digitalmars-d-learn]

V Thu, 31 Jul 2014 13:26:38 +
Kagamin via Digitalmars-d-learn digitalmars-d-learn@puremagic.com
napsáno:

 On Thursday, 31 July 2014 at 12:02:22 UTC, Kozzi11 wrote:
  module m;
  @someUda
  class C {
  void someFun();
  }
 
  @someUda
  class D {
  void anotherFun();
  }
 
  mixin(generateFunDefForClassesWithSomeUda!m);
 
 This is usually done by generating functions in the classes 
 directly.
 
 class C {
  mixin Generate!C;
 }

Yes this is how I do it now.

Re: A significant performance difference

[Daniel Kozak via Digitalmars-d-learn]

V Sun, 31 Aug 2014 10:55:31 +
bearophile via Digitalmars-d-learn digitalmars-d-learn@puremagic.com
napsáno:

 This is C++ code that solves one Euler problem:
 
 --
 #include stdio.h
 #include map
 
 const unsigned int H = 9, W = 12;
 
 const int g[6][3] = {{7, 0, H - 3},
   {1 + (1  H) + (1  (2 * H)), 0, H - 1},
   {3 + (1  H), 0, H - 2},
   {3 + (2  H), 0, H - 2},
   {1 + (1  H) + (2  H), 0, H - 2},
   {1 + (1  H) + (1  (H - 1)), 1, H - 1}};
 
 int main() {
  unsigned long long p, i, k;
  unsigned int j, l;
  std::mapunsigned int, unsigned long long x, y;
  x[0] = 1;
 
  for (i = 0; i  W; ++i) {
  y.clear();
  while (!x.empty()) {
  j = x.begin()-first;
  p = x.begin()-second;
  x.erase(x.begin());
 
  for (k = 0; k  H; ++k)
  if ((j  (1  k)) == 0)
  break;
 
  if (k == H)
  y[j  H] += p;
  else
  for (l = 0; l  6; ++l)
  if (k = g[l][1]  k = g[l][2])
  if ((j  (g[l][0]  k)) == 0)
  x[j + (g[l][0]  k)] += p;
  }
  x = y;
  }
 
  printf(%lld\n, y[0]);
  return 0;
 }
 --
 
 
 I have translated it to D like this (I know in D there are nicer 
 ways to write it, but I have tried to keep the look of the code 
 as much similar as possible to the C++ code):
 
 
 --
 import core.stdc.stdio;
 
 const uint H = 9, W = 12;
 
 const uint[3][6] g = [[7, 0, H - 3],
[1 + (1  H) + (1  (2 * H)), 0, H - 1],
[3 + (1  H), 0, H - 2],
[3 + (2  H), 0, H - 2],
[1 + (1  H) + (2  H), 0, H - 2],
[1 + (1  H) + (1  (H - 1)), 1, H - 1]];
 
 int main() {
  ulong p, i, k;
  uint j, l;
  ulong[uint] x, y;
  x[0] = 1;
 
  for (i = 0; i  W; ++i) {
  y = null;
  while (x.length) {
  j = x.byKey.front;
  p = x.byValue.front;
  x.remove(cast(int)j);
 
  for (k = 0; k  H; ++k)
  if ((j  (1  k)) == 0)
  break;
 
  if (k == H)
  y[j  H] += p;
  else
  for (l = 0; l  6; ++l)
  if (k = g[l][1]  k = g[l][2])
  if ((j  (g[l][0]  k)) == 0)
  x[j + (g[l][0]  k)] += p;
  }
  x = y;
  }
 
  printf(%lld\n, y[0]);
  return 0;
 }
 --
 
 
 The C++ code is much faster than the D code (I see the D code 30+ 
 times slower with dmd and about like 20 times with ldc2). One 
 difference between the C++ and D code is that the C++ map uses a 
 search tree (red-black probably), while the D code uses a hash.
 
 To test that algorithmic difference, if I replace the map in the 
 C++ code with a std::unordered_map (C++11):
 
 #include unordered_map
 ...
 std::unordered_mapunsigned int, unsigned long long x, y;
 
 
 then the run-time increases (more than two times) but it's still 
 much faster than the D code.
 
 Is it possible to fix the D code to increase its performance 
 (there are associative array libraries for D, but I have not 
 tried them in this program).
 
 Bye,
 bearophile

I think main problem is in calling delegates (x.byKey.front and
x.byValue.front;). If I change x.byValue.front with x[j], It makes
program a lot faster

Re: A significant performance difference

[Daniel Kozak via Digitalmars-d-learn]

V Mon, 1 Sep 2014 12:38:52 +0300
ketmar via Digitalmars-d-learn digitalmars-d-learn@puremagic.com
napsáno:

 On Mon, 01 Sep 2014 09:22:50 +
 bearophile via Digitalmars-d-learn digitalmars-d-learn@puremagic.com
 wrote:
 
  In theory the best solution is to improve the performance of the 
  byKey.front and byValue.front idioms.
 i found that slowdown is from _aaRange(), not from delegates.
 the following code is slow even w/o delegate creation:
 
 import core.stdc.stdio;
 
 ref K firstKey(T : V[K], V, K)(T aa) @property
 {
 return *cast(K*)_aaRangeFrontKey(_aaRange(cast(void*)aa));
 }
 
 ref V firstVal(T : V[K], V, K)(T aa) @property
 {
 return *cast(V*)_aaRangeFrontValue(_aaRange(cast(void*)aa));
 }
 
 
 int main() {
 long[int] aa;
 for (int i = 0; i  5; i++)
 aa[i] = i;
 
 long total = 0;
 
 while (aa.length) {
 //int k = aa.byKey.front;
 int k = aa.firstKey;
 //long v = aa.byValue.front;
 long v = aa.firstVal;
 aa.remove(k);
 total += k + v;
 }
 
 printf(%lld\n, total);
 return 0;
 }
 
 
 seems that we need two more hooks for AAs: `_aaFrontKey()` and
 `_aaFrontValue()`.
 
  x.pop() sounds nicer :-)
 ah, sure. i'm not very good in inventing names. ;-)
 

Yep, I found out something similar. with this ugly code

for (i = 0; i  W; ++i) {
y = null;
while (x.length) {

foreach(lj, lp ; x) {
j = lj;
p = lp;
x.remove(cast(int)j);
break;
}

...
}

The problem is with searching first element

Re: alias this cast

[Daniel Kozak via Digitalmars-d-learn]

V Thu, 11 Sep 2014 11:40:05 +
andre via Digitalmars-d-learn digitalmars-d-learn@puremagic.com
napsáno:

 Hi,
 
 I am 80% sure, the failing assertion is correct but please have a 
 look.

No it is not

assert(cast(A)cast(C)b); // this is OK

b is B so it does not know about having alias to A;

 Second assertion fails.
 
 Kind regards
 André
 
 class A{}
 
 class B{}
 
 class C : B
 {
   A a;
   alias a this;
   
   this()
   {
   a = new A();
   }
 }
 
 void main()
 {
   B b = new C();
 
   // OK
   assert(cast(C)b);
   
   // fails
   assert(cast(A)b);   
 }

Re: std.algorithm.reduce on an array of structs

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 11 September 2014 at 13:06:05 UTC, Colin wrote:

I have this test code:

struct Thing {
uint x;
}

void main(){
uint[] ar1 = [1, 2, 3, 4, 5];
auto min1 = ar1.reduce!((a,b) = a  b);
writefln(%s, min1);  // prints 1 as expected

Thing[] ar2 = [Thing(1), Thing(2), Thing(4)];
auto min2 = ar2.reduce!((a,b) = a.x  b.x);  //  - Wont 
Compile

writefln(%s, min2);
}

The line with Wont Compile on it has this error message:
/usr/include/dmd/phobos/std/algorithm.d(770): Error: cannot 
implicitly convert expression (__lambda2(result, 
front(_param_1))) of type bool to Thing
/usr/include/dmd/phobos/std/algorithm.d(791): Error: template 
instance t.main.reduce!((a, b) = a.x  b.x).reduce!(Thing, 
Thing[]) error instantiating

t.d(16):instantiated from here: reduce!(Thing[])


Any idea what I'm doing wrong here?
To me, the operation on ar2 should be pretty much identical to 
ar1, except for the use of the struct.



You are try to put uint to Thing. This is corect version:

import std.stdio;
import std.algorithm;

struct Thing {
uint x;
}

void main(){
uint[] ar1 = [1, 2, 3, 4, 5];
auto min1 = ar1.reduce!((a,b) = a  b);
writefln(%s, min1);  // prints 1 as expected

Thing[] ar2 = [Thing(1), Thing(2), Thing(4)];
auto min2 = ar2.reduce!((a,b) = a.x  b.x ? a : b);  //  -
Wont Compile
writefln(%s, min2);
}

Re: std.algorithm.reduce on an array of structs

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 11 September 2014 at 14:18:31 UTC, Colin wrote:


Ah ok. I get it.

Thanks daniel!


a quiet better version:

import std.stdio;
import std.algorithm;

struct Thing {
uint x;
alias x this;
}

void main(){
uint[] ar1 = [1, 2, 3, 4, 5];
auto min1 = ar1.reduce!((a,b) = min(a,b));
writefln(%s, min1);

Thing[] ar2 = [Thing(1), Thing(2), Thing(4)];
auto min2 = ar2.reduce!((a,b) = min(a,b));
writefln(%s, min2);
}

Re: std.algorithm.reduce on an array of structs

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 11 September 2014 at 14:39:53 UTC, Daniel Kozak 
wrote:

On Thursday, 11 September 2014 at 14:18:31 UTC, Colin wrote:


Ah ok. I get it.

Thanks daniel!


a quiet better version:

import std.stdio;
import std.algorithm;

struct Thing {
uint x;
alias x this;
}

void main(){
uint[] ar1 = [1, 2, 3, 4, 5];
auto min1 = ar1.reduce!((a,b) = min(a,b));
writefln(%s, min1);

Thing[] ar2 = [Thing(1), Thing(2), Thing(4)];
auto min2 = ar2.reduce!((a,b) = min(a,b));
writefln(%s, min2);
}


s/quiet/quite/

Re: std.algorithm.reduce on an array of structs

[Daniel Kozak via Digitalmars-d-learn]

V Thu, 11 Sep 2014 14:49:02 +
bearophile via Digitalmars-d-learn digitalmars-d-learn@puremagic.com
napsáno:

 Daniel Kozak:
 
 You can just use min:
 
 import std.stdio, std.algorithm;
 
 struct Thing {
  uint x;
  alias x this;
 }
 
 alias minimum = reduce!min;
 
 void main() {
   immutable ar1 = [10, 20, 30, 40, 50];
   ar1.minimum.writeln;
 
   immutable ar2 = [Thing(10), Thing(20), Thing(40)];
   ar2.minimum.writeln;
 }
 
 Bye,
 bearophile

Yep, this look the most idiomatic :).

Why there is no phobos function for minimum of array(range)?

Re: std.algorithm.reduce on an array of structs

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 11 September 2014 at 14:56:00 UTC, Daniel Kozak via 
Digitalmars-d-learn wrote:

V Thu, 11 Sep 2014 14:49:02 +
bearophile via Digitalmars-d-learn 
digitalmars-d-learn@puremagic.com

napsáno:


Daniel Kozak:

You can just use min:

import std.stdio, std.algorithm;

struct Thing {
 uint x;
 alias x this;
}

alias minimum = reduce!min;

void main() {
immutable ar1 = [10, 20, 30, 40, 50];
ar1.minimum.writeln;

immutable ar2 = [Thing(10), Thing(20), Thing(40)];
ar2.minimum.writeln;
}

Bye,
bearophile


Yep, this look the most idiomatic :).

Why there is no phobos function for minimum of array(range)?


or use alias minimum = reduce!a  b;
;)

Re: std.algorithm.reduce on an array of structs

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 11 September 2014 at 15:07:03 UTC, Daniel Kozak 
wrote:


or use alias minimum = reduce!a  b;
;)


ok this one does not work

Re: Is there a function that reads the entire contents of a std.stdio.File?

[Daniel Kozak via Digitalmars-d-learn]

V Tue, 16 Sep 2014 14:37:05 +
Jay via Digitalmars-d-learn digitalmars-d-learn@puremagic.com napsáno:

 all the functions/methods i've come across so far deal with 
 either streams or just file names (like std.file.read) and there 
 doesn't seem to be a way to wrap a std.stdio.File in a stream (or 
 is there?). i need a function that takes a std.stdio.File and 
 returns a string or byte array.


You can use rawRead:
http://dlang.org/phobos/std_stdio.html#.File.rawRead

parallel foreach hangs

[Daniel Kozak via Digitalmars-d-learn]

this code never end

import std.stdio;
import std.file;
import std.parallelism : parallel;
import std.algorithm : filter;

void main(string[] args)
{
foreach(d; parallel(args[1 .. $], 1))
{
auto phpFiles = 
filter!`endsWith(a.name,.php)`(dirEntries(d,SpanMode.depth));

writeln(phpFiles);
}
}

Re: Does D has C#'s string.Empty?

[Daniel Kozak via Digitalmars-d-learn]

V Thu, 25 Sep 2014 05:29:36 +
AsmMan via Digitalmars-d-learn digitalmars-d-learn@puremagic.com
napsáno:

 Does D has C#'s string.Empty?

string.init

Re: curl and proxy

[Daniel Kozak via Digitalmars-d-learn]

On Friday, 3 October 2014 at 11:13:11 UTC, Marc Schütz wrote:

On Friday, 3 October 2014 at 10:53:27 UTC, Marc Schütz wrote:

On Friday, 3 October 2014 at 04:57:28 UTC, AntonSotov wrote:

auto http = HTTP(dlang.org);
http.onReceive = (ubyte[] data)
{
  writeln(cast(string) (data));
  return data.length;
};
http.proxy = 192.168.111.111;
http.proxyPort = 1788;

WHAT HERE ?

http.perform();
//
how to make Сurl authorize on a proxy.
I specify proxyUser and proxyPassword?


I think there's currently no way. curl provides this as an 
option (CurlOption.proxyuserpwd):


   curl.set(CurlOption.proxyuserpwd, myuser:mypasswd);

But unfortunately the `curl` struct is a private member of 
`HTTP`...


https://github.com/D-Programming-Language/phobos/pull/2581

If you're able to build Phobos, could you give this PR a try? I 
cannot test it myself, because I don't know of a proxy 
requiring authentication.


Works OK for me

Thanks

Re: Best way to add slash to tail of the path

[Daniel Kozak via Digitalmars-d-learn]

Dne Thu, 27 Nov 2014 21:20:24 +0100 Suliman via Digitalmars-d-learn  
digitalmars-d-learn@puremagic.com napsal(a):



take a second look then. ;-) you'll find `buildPath()` here too.

Not better:

string foo = D:/code/txtDownloader;

writeln(foo);
foo = foo.buildPath;
foo ~= config.txt;
writeln(foo);


Running .\txtdownloader.exe
D:/code/txtDownloader
D:/code/txtDownloaderconfig.txt -- need:  
D:/code/txtDownloader/config.txt





what about:

string foo = D:/code/txtDownloader;
writeln(foo);
foo = buildPath(foo, config.txt);
writeln(foo);

Re: Best way to add slash to tail of the path

[Daniel Kozak via Digitalmars-d-learn]

try first few sentences and looked at the example ;)

Dne Thu, 27 Nov 2014 21:42:31 +0100 Suliman via Digitalmars-d-learn  
digitalmars-d-learn@puremagic.com napsal(a):


Could you quote for me part of docs where it's written? I really can't  
understand about what you are taking.



--
Vytvořeno poštovní aplikací Opery: http://www.opera.com/mail/

Re: function app.download (string[] links) is not callableusing argument types (string, string)

[Daniel Kozak via Digitalmars-d-learn]

Dne Thu, 27 Nov 2014 18:39:09 +0100 Suliman via Digitalmars-d-learn  
digitalmars-d-learn@puremagic.com napsal(a):



Is there any way to detect where collision was occurred?


Yes, read error description it has been app.download I guess
--
Vytvořeno poštovní aplikací Opery: http://www.opera.com/mail/

Re: Error: cannot return non-void from void function

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 27 November 2014 at 17:22:36 UTC, Suliman wrote:


ah, that's it! as spec says, D determines function return 
value from
the first 'return' statement it seen. in your case this is 
`return;`,

so function return type is determined to be `void`.

if you doing `auto` functions, try to arrange your code so the 
first

`return` returning the actual value.

besides, your code is wrong anyway, 'cause you can't have 
function that
returns both nothing and something. your first `return;` 
should
either return something, or must be changed to throwing some 
exception.


How I can terminate program? First return I used to terminate 
app if config file is exists.


You can use this:

auto parseConfig()
{
string txtlinks = buildPath(getcwd,notexist);
if(exists(txtlinks))
{
auto lines = File(txtlinks, r).byLine;
return lines;
}
writeln(Can't find input file with list of links.);
return typeof(return)();
}

it works somehow, but it is not good way how to do it.

Other way is use exit

if(!exists(txtlinks))
{
import core.runtime;
import std.c.process;
writeln(Can't find input file with list of links.);
Runtime.terminate();
exit(1);
}
But best way is use throw exception

if(!exists(txtlinks))
{
	throw new Exception(Can't find input file with list of 
links.);

}
and catch it somewhere from calling side

Re: Error: cannot return non-void from void function

[Daniel Kozak via Digitalmars-d-learn]

Dne Thu, 27 Nov 2014 22:21:52 +0100 ketmar via Digitalmars-d-learn  
digitalmars-d-learn@puremagic.com napsal(a):



On Thu, 27 Nov 2014 21:14:57 +
Daniel Kozak via Digitalmars-d-learn
digitalmars-d-learn@puremagic.com wrote:


import core.runtime;
import std.c.process;
writeln(Can't find input file with list of links.);
Runtime.terminate();
exit(1);

please-please-please don't teach people that! using such features
requiring deep understanding of how D runtime works, and how it
interacts with C runtime, with GC, with stack objects and so on.

it's better to now show people bad samples instead of telling them
don't do what i just wrote. ;-)


I know, I just can't help myself :).

Re: How to initialise array of ubytes?

[Daniel Kozak via Digitalmars-d-learn]

Dne Sat, 29 Nov 2014 21:10:41 +0100 Paul via Digitalmars-d-learn  
digitalmars-d-learn@puremagic.com napsal(a):



I'm trying to do this:

ubyte[MAPSIZE][MAPSIZE] map = 1;

but it doesn't work and I can't seem to cast the value to a ubyte (which  
looks rather ugly and out of place in D anyway). Is there a way to do  
this other than using a couple of loops?


Cheers

Paul


http://stackoverflow.com/questions/24600796/d-set-default-value-for-a-struct-member-which-is-a-multidimensional-static-arr/24754361#24754361
--
Vytvořeno poštovní aplikací Opery: http://www.opera.com/mail/

Re: How to initialise array of ubytes?

[Daniel Kozak via Digitalmars-d-learn]

On Saturday, 29 November 2014 at 20:45:34 UTC, bearophile wrote:

Daniel Kozak:


http://stackoverflow.com/questions/24600796/d-set-default-value-for-a-struct-member-which-is-a-multidimensional-static-arr/24754361#24754361


Do you also know why the simplest syntax doesn't work? Can't it 
be implemented and added to the D language?


Bye,
bearophile


I don't know. But this works too:

module main;
import std.stdio;

int[5][5] p;

static this() {
p = 1;
}

void main() {
writeln(p);
}

Re: math.log() benchmark of first 1 billion int using std.parallelism

[Daniel Kozak via Digitalmars-d-learn]

 I run Arch Linux on my PC. I compiled D programs using dmd-2.066 
 and used no compile arguments (dmd prog.d)

You should try use some arguments -O -release -inline -noboundscheck
and maybe try use gdc or ldc should help with performance

can you post your code in all languages somewhere? I like to try it on
my machine :)

Re: math.log() benchmark of first 1 billion int using std.parallelism

[Daniel Kozak via Digitalmars-d-learn]

On Monday, 22 December 2014 at 10:35:52 UTC, Daniel Kozak via 
Digitalmars-d-learn wrote:


I run Arch Linux on my PC. I compiled D programs using 
dmd-2.066 and used no compile arguments (dmd prog.d)


You should try use some arguments -O -release -inline 
-noboundscheck

and maybe try use gdc or ldc should help with performance

can you post your code in all languages somewhere? I like to 
try it on

my machine :)


Btw. try use C log function, maybe it would be faster:

import core.stdc.math;

Re: math.log() benchmark of first 1 billion int using std.parallelism

[Daniel Kozak via Digitalmars-d-learn]

That's very different to my results.

I see no important difference between ldc and dmd when using 
std.math, but when using core.stdc.math ldc halves its time 
where dmd only manages to get to ~80%


What CPU do you have? On my Intel Core i3 I have similar 
experience as Iov Gherman, but on my Amd FX4200 I have same 
results as you. Seems std.math.log is not good for my AMD CPU :)

Re: math.log() benchmark of first 1 billion int using std.parallelism

[Daniel Kozak via Digitalmars-d-learn]

On Tuesday, 23 December 2014 at 10:39:13 UTC, Iov Gherman wrote:


These multi-threaded benchmarks can be very sensitive to their 
environment, you should try running it with nice -20 and do 
multiple passes to get a vague idea of the variability in the 
result. Also, it's important to minimise the number of other 
running processes.


I did not use the nice parameter but I always ran them multiple 
times and choose the average time. My system has very few 
running processes, minimalist ArchLinux with Xfce4 so I don't 
think the running processes are affecting in any way my tests.


And what about single threaded version?

Btw. One reason why DMD is faster is because it use fyl2x X87 
instruction


here is version for others compilers:

import std.math, std.stdio, std.datetime;

enum SIZE = 100_000_000;

version(GNU)
{
real mylog(double x) pure nothrow
{
real result;
double y = LN2;
asm
{
fldl   %2\n
fldl   %1\n
fyl2x
: =t (result) : m (x), m (y);
}
return result;
}
}
else
{
real mylog(double x) pure nothrow
{
return yl2x(x, LN2);
}
}

void main() {

auto t1 = Clock.currTime();
auto logs = new double[SIZE];

foreach (i; 0 .. SIZE)
{
logs[i] = mylog(i + 1.0);
}

auto t2 = Clock.currTime();

writeln(time: , (t2 - t1));
}

But it is faster only on all Intel CPU, but on one of my AMD it 
is slower than core.stdc.log

Re: math.log() benchmark of first 1 billion int using std.parallelism

[Daniel Kozak via Digitalmars-d-learn]

On Tuesday, 23 December 2014 at 10:20:04 UTC, Iov Gherman wrote:

That's very different to my results.

I see no important difference between ldc and dmd when using 
std.math, but when using core.stdc.math ldc halves its time 
where dmd only manages to get to ~80%


I checked again today and the results are interesting, on my pc 
I don't see any difference between std.math and core.stdc.math 
with ldc. Here are the results with all compilers.


- with std.math:
dmd: 4 secs, 878 ms
ldc: 5 secs, 650 ms
gdc: 9 secs, 161 ms

- with core.stdc.math:
dmd: 5 secs, 991 ms
ldc: 5 secs, 572 ms
gdc: 7 secs, 957 ms


Btw. I just noticed small issue with D vs. java, you start 
messure in D before allocation, but in case of Java after 
allocation

Re: math.log() benchmark of first 1 billion int using std.parallelism

[Daniel Kozak via Digitalmars-d-learn]

On Tuesday, 23 December 2014 at 12:31:47 UTC, Iov Gherman wrote:


Btw. I just noticed small issue with D vs. java, you start 
messure in D before allocation, but in case of Java after 
allocation


Here is the java result for parallel processing after moving 
the start time as the first line in main. Still best result:


4 secs, 50 ms average


Java:

Exec time: 6 secs, 421 ms

LDC (-O3 -release -mcpu=native -singleobj -inline 
-boundscheck=off)


time: 5 secs, 321 ms, 877 μs, and 2 hnsecs

GDC(-O3 -frelease -march=native -finline -fno-bounds-check)

time: 5 secs, 237 ms, 453 μs, and 7 hnsecs

DMD(-O -release -inline -noboundscheck)
time: 5 secs, 107 ms, 931 μs, and 3 hnsecs

So all d compilers beat Java in my case:

but I have made some change in D version:

import std.parallelism, std.math, std.stdio, std.datetime;
import core.memory;

enum XMS = 3*1024*1024*1024; //3GB

version(GNU)
{
real mylog(double x) pure nothrow
{
double result;
double y = LN2;
asm
{
fldl   %2\n
fldl   %1\n
fyl2x\n
: =t (result) : m (x), m (y);
}

return result;
}
}
else
{
real mylog(double x) pure nothrow
{
return yl2x(x, LN2);
}
}

void main() {

GC.reserve(XMS);
auto t1 = Clock.currTime();


auto logs = new double[1_000_000_000];  
foreach(i, ref elem; taskPool.parallel(logs, 200)) {
elem = mylog(i + 1.0);
}


auto t2 = Clock.currTime();
writeln(time: , (t2 - t1)); 
}

Re: Is D's GC.calloc and C's memset played the same role?

[Daniel Kozak via Digitalmars-d-learn]

FrankLike via Digitalmars-d-learn píše v Út 23. 12. 2014 v 15:37 +:
 Today,I meet a question:get all processes names.
 
 --C++ CODE-
 #include stdafx.h
 #include windows.h
 #include stdio.h//C standard I/O
 #include tlhelp32.h
 
 int _tmain(int argc, _TCHAR* argv[])
 {
  HANDLE 
 hProcessSnap=CreateToolhelp32Snapshot(TH32CS_SNAPPROCESS,0);
 
  if(hProcessSnap==INVALID_HANDLE_VALUE)
  {
  _tprintf(_T(CreateToolhelp32Snapshot error!\n));
  return -1;
  }
 
  PROCESSENTRY32 pe32;
  pe32.dwSize = sizeof(PROCESSENTRY32);
 
  BOOL bMore=Process32First(hProcessSnap,pe32);
  int i=0;
 
  _tprintf(_T(PID\t thread nums \t name \n));
 
  while(bMore)
  {
  bMore=Process32Next(hProcessSnap,pe32);
  _tprintf(_T(%u\t),pe32.th32ProcessID);
  _tprintf(_T(%u\t),pe32.cntThreads);
  _tprintf(_T(%s\n),pe32.szExeFile);
 
  i++;
  }
 
  CloseHandle(hProcessSnap);
  _tprintf(_T(Count:%d\n),i);
 
  return 0;
 }
 D code--
 import std.stdio;
 import std.string;
 import core.sys.windows.windows;
 import core.memory;
 import win32.tlhelp32;
 
 void main()
 {
   
  HANDLE 
 hProcessSnap=CreateToolhelp32Snapshot(TH32CS_SNAPPROCESS,0);
 
  if(hProcessSnap is null)
  {
  writeln(CreateToolhelp32Snapshot error!\n);
  return ;
  }
 
  PROCESSENTRY32* pe32 = 
 cast(PROCESSENTRY32*)GC.calloc(PROCESSENTRY32.sizeof);
 
   pe32.dwSize = PROCESSENTRY32.sizeof;
 
  bool bMore=cast(bool)Process32First(hProcessSnap,pe32);
  int i=0;
 
  writeln(PID\t thread nums\t name \n);
 
  while(bMore)
  {
  bMore=cast(bool)Process32Next(hProcessSnap,pe32);
 string s = cast(string)pe32.szExeFile;
 auto a = s.indexOf('\0');
 if(a =0)
  
 writeln(\t,pe32.th32ProcessID,\t,pe32.cntThreads,\t,s[0..a]);
  i++;
  }
 
  CloseHandle(hProcessSnap);
  writeln(format(count:%d,i));
 
  return ;
 }
 ---end--
 you will find the different:
   D: PROCESSENTRY32* pe32 = 
 cast(PROCESSENTRY32*)GC.calloc(PROCESSENTRY32.sizeof);
 
 C++:PROCESSENTRY32 pe32;
 
 GC.calloc means: memset ?!
calloc means alloc cleared memory same as malloc but clear all bits to
zero

Re: Call of rmdir in destructor causes InvalidMemoryOperationError

[Daniel Kozak via Digitalmars-d-learn]

Timo Gransch via Digitalmars-d-learn píše v Čt 01. 01. 2015 v 16:14
+0100:
 Hi,
 
 I have a class which unzips an archive into a temporary directory below the 
 system temp folder. I want to delete this temporary directory in the class's 
 destructor, but when I call rmdir there, I get an
 
 core.exception.InvalidMemoryOperationError@(0)
 
 The effect is not limited to this special case. Whenever I call rmdir in the 
 destructor, no matter for which directory, I get the same error.
 
 Environment: DMD v2.066.1 (from D-Apt) on Ubuntu 14.10 (32 and 64 bit).
 
 Sample code:
 
 // ---
 module main;
 
 import std.stdio;
 import std.file;
 import std.path;
 
 class RmdirTest
 {
   private string sTempDir;
 
   this(string sInstanceName)
   {
   writeln(Constructor called);
   sTempDir=tempDir() ~ dirSeparator ~ sInstanceName;
   mkdir (sTempDir);
   }
 
   ~this()
   {
   writeln(Destructor called);
 
   if (sTempDir !is null)
   {
   rmdir(sTempDir);
   }
   }
 }
 
 void main(string[] args)
 {
   RmdirTest rmDirTest=new RmdirTest(123);
 }
 // ---
 
 Console output is:
 
 Constructor called
 Destructor called
 core.exception.InvalidMemoryOperationError@(0)
 
 The directory /tmp/123 is created, but not deleted.
 
 When I change the line (sTempDir !is null) to ((sTempDir !is null)  
 (exists(sTempDir))  (isDir(sTempDir))), the exception is thrown already on 
 this line, so obviously  the problem also applies to other file functions 
 like exists and isDir.
 
 Is there any solution for this?
 
 Thanks and best regards,
 Timo
You shoud not use destructor for this operation. In D there is no
guarantee that class destructor will be called. So you can probablly use
struct instead of class or add some method and called it explicitly when
needed

Re: Scoped external function declaration

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 1 January 2015 at 17:51:46 UTC, novice2 wrote:

I want to use external or C function.
It used only one time from one D function.
I want do declare C function inside D function.
I don't want to declare C function in global scope.

Is my wish correct?
Reduced code:



extern (C) int getch();
void main() {
  getch();
}
//compiled OK



void main() {
  extern (C) int getch();
  getch();
}
//Error 42: Symbol Undefined __D4test4mainFZ5getchUZi


you can use local import


cfun.d:
module cfun;
extern (C) int getc();

main.d:
module main;
void main() {
import cfun;
getc();
}

What exactly shared means?

[Daniel Kozak via Digitalmars-d-learn]

I always think that shared should be use to make variable global 
across threads (similar to __gshared) with some synchronize 
protection. But this code doesn't work (app is stuck on _aaGetX 
or _aaRehash ):


shared double[size_t] logsA;

void main() {

auto logs = new double[1_000_000];

foreach(i, ref elem; parallel(logs, 4)) {
elem = log(i + 1.0);
logsA[i]= elem;
}
}


But when I add synchronized block it is OK:

shared double[size_t] logsA;

void main() {

auto logs = new double[1_000_000];

foreach(i, ref elem; parallel(logs, 4)) {
elem = log(i + 1.0);
synchronized {
logsA[i]= elem;
}
}
}

Re: static class vs. static struct

[Daniel Kozak via Digitalmars-d-learn]

On Tuesday, 27 January 2015 at 09:01:39 UTC, ref2401 wrote:
For several times I've met struct(or static struct) usage in 
Phobos for singleton pattern implementation. Unfortunately now 
i can remember only core.runtime.Runtime.
So I've got a question. Why do Phobos guys use struct or static 
struct for or singleton pattern implementation? Why don't use 
static final class for this purpose?


I do not think this is a singleton pattern (no instance). I see 
it much more like namespace in case of core.runtime.Runtime. And 
yes static final class could do that too but struct looks better 
than final class and you can disable this on structs

Re: static class vs. static struct

[Daniel Kozak via Digitalmars-d-learn]

On Tuesday, 27 January 2015 at 09:36:49 UTC, Daniel Kozak wrote:

On Tuesday, 27 January 2015 at 09:01:39 UTC, ref2401 wrote:
For several times I've met struct(or static struct) usage in 
Phobos for singleton pattern implementation. Unfortunately now 
i can remember only core.runtime.Runtime.
So I've got a question. Why do Phobos guys use struct or 
static struct for or singleton pattern implementation? Why 
don't use static final class for this purpose?


I do not think this is a singleton pattern (no instance). I see 
it much more like namespace in case of core.runtime.Runtime. 
And yes static final class could do that too but struct looks 
better than final class and you can disable this on structs


import std.stdio;
import std.conv;

struct S
{
@disable this();
}

final class C
{
}

void main() {
writeln(C.sizeof);
writeln(S.sizeof);
}

Re: Why can't functions and struct types have the same name?

[Daniel Kozak via Digitalmars-d-learn]

On Monday, 26 January 2015 at 11:15:26 UTC, Joakim wrote:
Right now, any attempt to have symbols with the same name 
errors out, regardless of how they're used.  This caused a 
problem for me because I'm trying to use a third-party C 
library that defines a struct type called socket and my code 
calls that library and some networking modules from druntime.  
Since core.sys.posix.sys.socket happens to contain a function 
called socket also, dmd gets confused when I try to create a 
socket* and use it in my code.


I would think it'd know I can't do that with a function name.  
Is such symbol disambiguation a convenience that just isn't 
implemented yet or something that can't/won't be done?


Right now, I had to go through and selectively import all 14 
symbols I needed from the 3 druntime modules that publicly 
import core.sys.posix.sys.socket, so that the function socket 
isn't imported.  Seems like a pain that can be mitigated by a 
smarter compiler.


You can use static import, or alias to solve this issue

Re: For those ready to take the challenge

[Daniel Kozak via Digitalmars-d-learn]

Vladimir Panteleev via Digitalmars-d-learn píše v So 10. 01. 2015 v
07:42 +:
 On Saturday, 10 January 2015 at 02:10:04 UTC, Jesse Phillips 
 wrote:
  On Friday, 9 January 2015 at 13:50:29 UTC, eles wrote:
  https://codegolf.stackexchange.com/questions/44278/debunking-stroustrups-debunking-of-the-myth-c-is-for-large-complicated-pro
 
  Link to answer in D:
  http://codegolf.stackexchange.com/a/44417/13362
 
 I think byLine is not necessary. By default . will not match line 
 breaks.
 
 One statement solution:
 
 import std.net.curl, std.stdio;
 import std.algorithm, std.regex;
 
 void main() {
   get(http://www.stroustrup.com/C++.html;)
   .matchAll(`a.*?href=(.*)`)
   .map!(m = m[1])
   .each!writeln();
 }
 
 Requires Phobos PR#2024 ;)
Oh here is it, I was looking for each. I think it is allready in a
phobos but I can not find. Now I know why :D

Re: Parameterized enum does not work

[Daniel Kozak via Digitalmars-d-learn]

On Friday, 9 January 2015 at 06:17:53 UTC, Andre wrote:

Hi,

Should following coding work?

string lpad(ubyte length, long n)
{
import std.string: rightJustify;
import std.conv: to;
return rightJustify(to!string(n), length, '0');
}

enum lpad14(long n) = lpad(14, n);

void main()
{
lpad14(123);
}

There is following error from dmd:

source\app.d(12): Error: template app.lpad14 cannot deduce 
function from argumen

t types !()(int), candidates are:
source\app.d(8):app.lpad14(long n)

Kind regards
André


What are you trying to do?

Re: Parameterized enum does not work

[Daniel Kozak via Digitalmars-d-learn]

On Friday, 9 January 2015 at 07:50:53 UTC, Daniel Kozak wrote:

On Friday, 9 January 2015 at 06:17:53 UTC, Andre wrote:

Hi,

Should following coding work?

string lpad(ubyte length, long n)
{
import std.string: rightJustify;
import std.conv: to;
return rightJustify(to!string(n), length, '0');
}

enum lpad14(long n) = lpad(14, n);

void main()
{
lpad14(123);
}

There is following error from dmd:

source\app.d(12): Error: template app.lpad14 cannot deduce 
function from argumen

t types !()(int), candidates are:
source\app.d(8):app.lpad14(long n)

Kind regards
André


What are you trying to do?


OK I probably see it now :):

import std.stdio;

string lpad(ubyte length, long n)
{
import std.string: rightJustify;
import std.conv: to;
return rightJustify(to!string(n), length, '0');
}

enum lpad14(long n) = lpad(14, n);

void main()
{
writeln(lpad14!(123));
}

Re: Parameterized enum does not work

[Daniel Kozak via Digitalmars-d-learn]

On Friday, 9 January 2015 at 07:52:50 UTC, Daniel Kozak wrote:

On Friday, 9 January 2015 at 07:50:53 UTC, Daniel Kozak wrote:

On Friday, 9 January 2015 at 06:17:53 UTC, Andre wrote:

Hi,

Should following coding work?

string lpad(ubyte length, long n)
{
import std.string: rightJustify;
import std.conv: to;
return rightJustify(to!string(n), length, '0');
}

enum lpad14(long n) = lpad(14, n);

void main()
{
lpad14(123);
}

There is following error from dmd:

source\app.d(12): Error: template app.lpad14 cannot deduce 
function from argumen

t types !()(int), candidates are:
source\app.d(8):app.lpad14(long n)

Kind regards
André


What are you trying to do?


OK I probably see it now :):

import std.stdio;

string lpad(ubyte length, long n)
{
import std.string: rightJustify;
import std.conv: to;
return rightJustify(to!string(n), length, '0');
}

enum lpad14(long n) = lpad(14, n);

void main()
{
writeln(lpad14!(123));
}


However this is only for compile time, if you want runtime 
version you need something like this:


import std.stdio;

string lpad(ubyte length)(long n)
{
import std.string: rightJustify;
import std.conv: to;
return rightJustify(to!string(n), length, '0');
}

alias lpad14 = lpad!14;

void main()
{
writeln(lpad14(123));
}

Re: Endless static this call when used a thread in it

[Daniel Kozak via Digitalmars-d-learn]

On Tuesday, 13 January 2015 at 14:02:45 UTC, Daniel Kozák via 
Digitalmars-d-learn wrote:

V Tue, 13 Jan 2015 13:56:05 +
tcak via Digitalmars-d-learn digitalmars-d-learn@puremagic.com
napsáno:


On Tuesday, 13 January 2015 at 13:53:11 UTC, tcak wrote:
 I have written the following code:

 test.d
 ==
 import core.thread;
 import std.stdio;

 void threadFunc(){
writeln(Thread func);
 }

 public static this(){
auto t = new Thread( threadFunc );
t.start();

writeln(Static init);
 }

 void main(){
  writeln(End of main);
 }


 run
 ==
 rdmd test.d


 result
 ==
 Static init
 Thread func
 Static init
 Thread func
 Static init
 Thread func
 Static init
 Thread func
 Static init
 Thread func
 Static init
 Thread func
 Static init
 Thread func
 Static init
 Thread func
 Sta...

 Is this normal, what's happening?

When I defined static init with shared

public shared static this()

it works normal now. But it doesn't explain above issue. 
What's the relation between a new thread and a module's 
initialiser?


I am not sure but my guess is

static this needs to be called before anything else in module 
so when
you try call threadFunc it looks if static this has been called 
and

finished which is not true so it call it again

And here is better explanation

http://dlang.org/module.html#staticorder

Static constructors are code that gets executed to initialize a 
module or a class before the main() function gets called.

...

Static constructors and static destructors run on thread local 
data, and are run whenever threads are created or destroyed.

Re: final methods by default

[Daniel Kozak via Digitalmars-d-learn]

On Fri, 20 Mar 2015 22:11:51 +
weaselcat via Digitalmars-d-learn digitalmars-d-learn@puremagic.com wrote:

 On Friday, 20 March 2015 at 14:25:22 UTC, ref2401 wrote:
  Why aren't methods of class final by default?
 
 history
 
 use final class, it should devirtualize all methods.
 see: https://github.com/D-Programming-Language/dmd/pull/4427

Yes, but you can not extend final class. Ok you can still use UFCS but it
is not elegand solution.

Re: final methods by default

[Daniel Kozak via Digitalmars-d-learn]

On Fri, 20 Mar 2015 16:27:04 -0700
Jonathan M Davis via Digitalmars-d-learn digitalmars-d-learn@puremagic.com
wrote:

 On Friday, March 20, 2015 23:53:14 Daniel Kozak via Digitalmars-d-learn wrote:
 
  On Fri, 20 Mar 2015 22:11:51 +
  weaselcat via Digitalmars-d-learn digitalmars-d-learn@puremagic.com wrote:
 
   On Friday, 20 March 2015 at 14:25:22 UTC, ref2401 wrote:
Why aren't methods of class final by default?
  
   history
  
   use final class, it should devirtualize all methods.
   see: https://github.com/D-Programming-Language/dmd/pull/4427
 
  Yes, but you can not extend final class. Ok you can still use UFCS but it
  is not elegand solution.
 
 Then you can just do
 
 class Foo
 {
 final:
 // methods...
 }
 
 or
 
 class Foo
 {
 final
 {
 // methods...
 }
 }
 
 And even if you couldn't do that, you could always mark each function with
 final individually.
 

Yes I know that and use it. Not often because I use struct and templates so I
need to marks methods as final occasionally ;-)

Re: Unicode exception raise when replacing underscore with space

[Daniel Kozak via Digitalmars-d-learn]

On Tuesday, 13 January 2015 at 20:30:16 UTC, Nordlöw wrote:
On Tuesday, 13 January 2015 at 13:01:56 UTC, Daniel Kozák via 
Digitalmars-d-learn wrote:

What do I need to do/add to avoid auto-decoding here?


std.array.replace(x, `_`, ` `);


Thanks! What about adding See alsos in the docs that relate 
these two with respect to auto-decoding?


I am not sure, it doesn`t exactly do the same. And to be fair 
std.array.replace use internaly std.algorithm.find which use in 
some scenario auto-decoding. So to be sure no autodecoding 
occured you must used something like that:


string x = some_text;
auto res = std.array.replace(cast(byte[])x, [byte('_')], 
[byte(' ')]);

writeln(cast(string)res);

Re: Static function template

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 7 May 2015 at 10:19:44 UTC, Lemonfiend wrote:
Is it not possible to have a static function template with the 
same name as the non-static version?


struct S
{
int i;

auto foo(T)(int j) {
i=j;
}

static auto foo(T)(int j) {
S s;
s.foo!T(j);
return s;
}
}

void main()
{
auto s = S.foo!bool(1);
}

Error: need 'this' for 'foo' of type '(int j)'


Another thinks which is wierd is than on 2.066 it prints:

test.d(17): Error: need 'this' for 'foo' of type 'pure nothrow 
@nogc @safe void(int j)'


So it seems auto deduction for attributes does not work anymore 
:( (in this case I guess)

Re: Static function template

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 7 May 2015 at 11:18:17 UTC, Daniel Kozak wrote:

On Thursday, 7 May 2015 at 11:15:02 UTC, Daniel Kozak wrote:

On Thursday, 7 May 2015 at 11:08:50 UTC, Daniel Kozák wrote:


On Thu, 07 May 2015 10:46:19 +
Lemonfiend via Digitalmars-d-learn 
digitalmars-d-learn@puremagic.com

wrote:


On Thursday, 7 May 2015 at 10:43:28 UTC, Daniel Kozak wrote:
 On Thursday, 7 May 2015 at 10:39:09 UTC, Daniel Kozák 
 wrote:


 On Thu, 07 May 2015 10:33:44 +
 Vadim Lopatin via Digitalmars-d-learn
 digitalmars-d-learn@puremagic.com wrote:

 struct S
 {
   int i;
 
   auto foo2(T)(int j) {

   i=j;
   }
 
   static S foo(T)(int j) {

   S s;
   s.foo2!T(j);
   return s;
   }
 }
 
 void main()

 {
   auto s = S.foo!bool(1);
 }

 As I said, it is not bug. It is OK. There is no way how 
 you can
 distinguish between static and non static methods or even 
 field in some

 cases.

 e.g.:

 import std.stdio;

 struct S
 {
string foo = Please select me?;
string foo() { return (No, select me?); };
 	static string foo() { return (I am better than the 
 otters :D?); };

 }

 void main()
 {
auto s = S();
writeln(s.foo);
 }

Well it's clear to me now why it shouldn't work.

However, the error msg is not clear on the problem. Imo it 
should give a conflict error like in your previous example. 
That would make it clear what's happened/allowed.




Yep, I think you are right even this example make useless and
wrong error message:

struct S
{
  static S foo(T)(int j) {
  S s;
  return s;
  }
  static S foo(T)(int j) {
  S s;
  return s;
  }
}

void main()
{
  auto s = S.foo!bool(1);
}


test.d(15): Error: need 'this' for 'foo' of type '(int j)' // 
WTF?


btw. it is a regresion from 2.067 on 2.066 and before it makes 
much
better error. Even for OP code when is modified (static must 
be declared before non static one) it have a better error msg.


test.d(6): Error: test.S.foo called with argument types (int) 
matches both:

test.d(4): test.S.foo!bool.foo(int j)
and:
test.d(10): test.S.foo!bool.foo(int j)
test.d(18): Error: template instance test.S.foo!bool error 
instantiating


But only when static one is declared before non static one, so 
even on 2.066 it was not ideal. So I think we should open two 
issues, probably one for regression and one for enhancment


OK both are regressions, cause on dmd 2.063 it is much much 
better


test.d(19): Error: template test.S.foo matches more than one 
template declaration, test.d(3):foo(T)(int j) and 
test.d(8):foo(T)(int j)
test.d(19): Error: need 'this' for 'foo' of type 'pure nothrow 
@safe void(int j)'


https://issues.dlang.org/show_bug.cgi?id=14554

Re: Static function template

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 7 May 2015 at 11:15:02 UTC, Daniel Kozak wrote:

On Thursday, 7 May 2015 at 11:08:50 UTC, Daniel Kozák wrote:


On Thu, 07 May 2015 10:46:19 +
Lemonfiend via Digitalmars-d-learn 
digitalmars-d-learn@puremagic.com

wrote:


On Thursday, 7 May 2015 at 10:43:28 UTC, Daniel Kozak wrote:
 On Thursday, 7 May 2015 at 10:39:09 UTC, Daniel Kozák wrote:

 On Thu, 07 May 2015 10:33:44 +
 Vadim Lopatin via Digitalmars-d-learn
 digitalmars-d-learn@puremagic.com wrote:

 struct S
 {
int i;
 
auto foo2(T)(int j) {

i=j;
}
 
static S foo(T)(int j) {

S s;
s.foo2!T(j);
return s;
}
 }
 
 void main()

 {
auto s = S.foo!bool(1);
 }

 As I said, it is not bug. It is OK. There is no way how 
 you can
 distinguish between static and non static methods or even 
 field in some

 cases.

 e.g.:

 import std.stdio;

 struct S
 {
string foo = Please select me?;
string foo() { return (No, select me?); };
 	static string foo() { return (I am better than the otters 
 :D?); };

 }

 void main()
 {
auto s = S();
writeln(s.foo);
 }

Well it's clear to me now why it shouldn't work.

However, the error msg is not clear on the problem. Imo it 
should give a conflict error like in your previous example. 
That would make it clear what's happened/allowed.




Yep, I think you are right even this example make useless and
wrong error message:

struct S
{
   static S foo(T)(int j) {
   S s;
   return s;
   }
   static S foo(T)(int j) {
   S s;
   return s;
   }
}

void main()
{
   auto s = S.foo!bool(1);
}


test.d(15): Error: need 'this' for 'foo' of type '(int j)' // 
WTF?


btw. it is a regresion from 2.067 on 2.066 and before it makes 
much
better error. Even for OP code when is modified (static must be 
declared before non static one) it have a better error msg.


test.d(6): Error: test.S.foo called with argument types (int) 
matches both:

test.d(4): test.S.foo!bool.foo(int j)
and:
test.d(10): test.S.foo!bool.foo(int j)
test.d(18): Error: template instance test.S.foo!bool error 
instantiating


But only when static one is declared before non static one, so 
even on 2.066 it was not ideal. So I think we should open two 
issues, probably one for regression and one for enhancment


OK both are regressions, cause on dmd 2.063 it is much much better

test.d(19): Error: template test.S.foo matches more than one 
template declaration, test.d(3):foo(T)(int j) and 
test.d(8):foo(T)(int j)
test.d(19): Error: need 'this' for 'foo' of type 'pure nothrow 
@safe void(int j)'

Re: Static function template

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 7 May 2015 at 11:08:50 UTC, Daniel Kozák wrote:


On Thu, 07 May 2015 10:46:19 +
Lemonfiend via Digitalmars-d-learn 
digitalmars-d-learn@puremagic.com

wrote:


On Thursday, 7 May 2015 at 10:43:28 UTC, Daniel Kozak wrote:
 On Thursday, 7 May 2015 at 10:39:09 UTC, Daniel Kozák wrote:

 On Thu, 07 May 2015 10:33:44 +
 Vadim Lopatin via Digitalmars-d-learn
 digitalmars-d-learn@puremagic.com wrote:

 struct S
 {
 int i;
 
 auto foo2(T)(int j) {

 i=j;
 }
 
 static S foo(T)(int j) {

 S s;
 s.foo2!T(j);
 return s;
 }
 }
 
 void main()

 {
 auto s = S.foo!bool(1);
 }

 As I said, it is not bug. It is OK. There is no way how you 
 can
 distinguish between static and non static methods or even 
 field in some

 cases.

 e.g.:

 import std.stdio;

 struct S
 {
string foo = Please select me?;
string foo() { return (No, select me?); };
 	static string foo() { return (I am better than the otters 
 :D?); };

 }

 void main()
 {
auto s = S();
writeln(s.foo);
 }

Well it's clear to me now why it shouldn't work.

However, the error msg is not clear on the problem. Imo it 
should give a conflict error like in your previous example. 
That would make it clear what's happened/allowed.




Yep, I think you are right even this example make useless and
wrong error message:

struct S
{
static S foo(T)(int j) {
S s;
return s;
}
static S foo(T)(int j) {
S s;
return s;
}
}

void main()
{
auto s = S.foo!bool(1);
}


test.d(15): Error: need 'this' for 'foo' of type '(int j)' // 
WTF?


btw. it is a regresion from 2.067 on 2.066 and before it makes 
much
better error. Even for OP code when is modified (static must be 
declared before non static one) it have a better error msg.


test.d(6): Error: test.S.foo called with argument types (int) 
matches both:

test.d(4): test.S.foo!bool.foo(int j)
and:
test.d(10): test.S.foo!bool.foo(int j)
test.d(18): Error: template instance test.S.foo!bool error 
instantiating


But only when static one is declared before non static one, so 
even on 2.066 it was not ideal. So I think we should open two 
issues, probably one for regression and one for enhancment

Re: Static function template

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 7 May 2015 at 10:39:09 UTC, Daniel Kozák wrote:


On Thu, 07 May 2015 10:33:44 +
Vadim Lopatin via Digitalmars-d-learn
digitalmars-d-learn@puremagic.com wrote:


struct S
{
 int i;

 auto foo2(T)(int j) {
 i=j;
 }

 static S foo(T)(int j) {
 S s;
 s.foo2!T(j);
 return s;
 }
}

void main()
{
 auto s = S.foo!bool(1);
}


As I said, it is not bug. It is OK. There is no way how you can
distinguish between static and non static methods or even field 
in some

cases.


e.g.:

import std.stdio;

struct S
{
string foo = Please select me?;
string foo() { return (No, select me?); };
	static string foo() { return (I am better than the otters 
:D?); };

}

void main()
{
auto s = S();
writeln(s.foo);
}

Re: How to create a mutable array of strings?

[Daniel Kozak via Digitalmars-d-learn]

On Sun, 17 May 2015 09:06:38 +
Dennis Ritchie via Digitalmars-d-learn digitalmars-d-learn@puremagic.com
wrote:

 Hi,
 It seems to me, or D do not create mutable array of strings?
 
 How to create a mutable equivalent of a string array?
 
 -
 string[] s = [foo, bar];
 // s[1][1] = 't'; // immutable expression s[1][1]

auto s = [foo.dup, bar.dup];

Re: How to create a mutable array of strings?

[Daniel Kozak via Digitalmars-d-learn]

On Sun, 17 May 2015 09:06:38 +
Dennis Ritchie via Digitalmars-d-learn digitalmars-d-learn@puremagic.com
wrote:

 Hi,
 It seems to me, or D do not create mutable array of strings?
 
 How to create a mutable equivalent of a string array?
 
 -
 string[] s = [foo, bar];
 // s[1][1] = 't'; // immutable expression s[1][1]

or you can use cast if you are sure thats what you really need:
auto s = [cast(char[])foo, cast(char[])bar];

or

auto s = cast(char[][])[foo, bar];

Re: How to create a mutable array of strings?

[Daniel Kozak via Digitalmars-d-learn]

On Sunday, 17 May 2015 at 09:18:15 UTC, Daniel Kozak wrote:


On Sun, 17 May 2015 09:06:38 +
Dennis Ritchie via Digitalmars-d-learn 
digitalmars-d-learn@puremagic.com

wrote:


Hi,
It seems to me, or D do not create mutable array of strings?

How to create a mutable equivalent of a string array?

-
string[] s = [foo, bar];
// s[1][1] = 't'; // immutable expression s[1][1]


or you can use cast if you are sure thats what you really need:
auto s = [cast(char[])foo, cast(char[])bar];

or

auto s = cast(char[][])[foo, bar];


But I am unsure if this will work in case where immutable data 
will occupied read only memory.

Re: How to create a mutable array of strings?

[Daniel Kozak via Digitalmars-d-learn]

On Sunday, 17 May 2015 at 09:21:58 UTC, Marc Schütz wrote:

On Sunday, 17 May 2015 at 09:18:15 UTC, Daniel Kozak wrote:


On Sun, 17 May 2015 09:06:38 +
Dennis Ritchie via Digitalmars-d-learn 
digitalmars-d-learn@puremagic.com

wrote:


Hi,
It seems to me, or D do not create mutable array of strings?

How to create a mutable equivalent of a string array?

-
string[] s = [foo, bar];
// s[1][1] = 't'; // immutable expression s[1][1]


or you can use cast if you are sure thats what you really need:
auto s = [cast(char[])foo, cast(char[])bar];

or

auto s = cast(char[][])[foo, bar];


That's not a good idea. I haven't checked, but this will likely 
segfault on mutation


Yep, you are right, shame on me :)

Re: What wrong?

[Daniel Kozak via Digitalmars-d-learn]

On Friday, 15 May 2015 at 09:20:32 UTC, Gary Willoughby wrote:

On Friday, 15 May 2015 at 07:51:29 UTC, thedeemon wrote:

On Saturday, 2 May 2015 at 02:51:52 UTC, Fyodor Ustinov wrote:

Simple code:

http://pastebin.com/raw.php?i=7jVeMFXQ

What I'm doing wrong?


Try using class instead of struct.
Last time I played with std.concurrency it used Variants to 
store the messages, so when something bigger than a little 
basic value or a reference, like a class object, is sent it 
behaves unpredictably: can crash or shit garbage. Trying to 
send structs larger than ~20 bytes usually caused problems.


Please raise a bugzilla issue for this.


this commit cause the issue:
https://github.com/D-Programming-Language/phobos/commit/45fda72192ff5b878ebe915db0ffb9f6504cca8f

but it is probably just a trigger not a real cause.

Re: ICE?

[Daniel Kozak via Digitalmars-d-learn]

On Sunday, 17 May 2015 at 09:25:33 UTC, Namespace wrote:
Is this error an ICE? I think so, because I see the internal 
filename, but I'm not sure.


Error: e2ir: cannot cast malloc(length * 8u) of type void* to 
type char[]


https://github.com/D-Programming-Language/dmd/pull/4667

Re: How to create a mutable array of strings?

[Daniel Kozak via Digitalmars-d-learn]

On Sunday, 17 May 2015 at 09:20:17 UTC, Dennis Ritchie wrote:

On Sunday, 17 May 2015 at 09:18:15 UTC, Daniel Kozak wrote:

auto s = cast(char[][])[foo, bar];


Thanks. This version I was completely satisfied.


So maybe this one would be ok with you too :)

auto s = to!(char[][])([foo, bar]);

Re: How to create a mutable array of strings?

[Daniel Kozak via Digitalmars-d-learn]

On Sun, 17 May 2015 09:39:21 +
Dennis Ritchie via Digitalmars-d-learn digitalmars-d-learn@puremagic.com
wrote:

 I remembered code Ali Çereli. It really helped:
 http://forum.dlang.org/thread/ulhtlyxxclihaseef...@forum.dlang.org#post-mihl6m:241che:241:40digitalmars.com
 
 -
 import std.stdio, std.traits, std.range, std.algorithm;
 
 auto deepDup(A)(A arr)
   if (isArray!A)
 {
   static if (isArray!(ElementType!A)) {
   return arr.map!(a = a.deepDup).array;
   
   } else {
   return arr.dup;
   }
 }
 
 void main() {
 
   auto s = [foo, bar].deepDup;
 
   s[1][1] = 't';
 
   writeln(s);
 }
 -
 http://rextester.com/QBFH12695
 
 P.S. Need to enable deepDup in Phobos.

allready there:
auto s = [foo, bar].map!a.dup.array;

:)

Re: ICE?

[Daniel Kozak via Digitalmars-d-learn]

On Sunday, 17 May 2015 at 09:59:41 UTC, Daniel Kozak wrote:


On Sun, 17 May 2015 09:33:27 +
Namespace via Digitalmars-d-learn 
digitalmars-d-learn@puremagic.com wrote:



On Sunday, 17 May 2015 at 09:30:16 UTC, Gary Willoughby wrote:
 On Sunday, 17 May 2015 at 09:25:33 UTC, Namespace wrote:
 Is this error an ICE? I think so, because I see the 
 internal filename, but I'm not sure.


 Error: e2ir: cannot cast malloc(length * 8u) of type void* 
 to type char[]


 Have you got a code sample to reproduce this?

Of course:


void main() {
import core.stdc.stdlib : malloc, free;

auto ptr = cast(char[]) malloc(42);
}



I guess it should be:

auto ptr = cast(char*)malloc(42)[0 .. 42];


or this if you want ptr to be char[] and not a pointer to char:

auto ptr = (cast(char*)malloc(42))[0 .. 42];
or
auto ptr = cast(char[])malloc(42)[0 .. 42];

Re: ICE?

[Daniel Kozak via Digitalmars-d-learn]

On Sun, 17 May 2015 10:17:42 +
anonymous via Digitalmars-d-learn digitalmars-d-learn@puremagic.com wrote:

 On Sunday, 17 May 2015 at 10:09:11 UTC, Daniel Kozak wrote:
  On Sunday, 17 May 2015 at 09:25:33 UTC, Namespace wrote:
 [...]
  Error: e2ir: cannot cast malloc(length * 8u) of type void* to 
  type char[]
 
  I would say this is not an ICE just normal error message.
 
 e2ir:  shouldn't be there, though.

Yep

Re: ICE?

[Daniel Kozak via Digitalmars-d-learn]

On Sun, 17 May 2015 09:33:27 +
Namespace via Digitalmars-d-learn digitalmars-d-learn@puremagic.com wrote:

 On Sunday, 17 May 2015 at 09:30:16 UTC, Gary Willoughby wrote:
  On Sunday, 17 May 2015 at 09:25:33 UTC, Namespace wrote:
  Is this error an ICE? I think so, because I see the internal 
  filename, but I'm not sure.
 
  Error: e2ir: cannot cast malloc(length * 8u) of type void* to 
  type char[]
 
  Have you got a code sample to reproduce this?
 
 Of course:
 
 
 void main() {
   import core.stdc.stdlib : malloc, free;
 
   auto ptr = cast(char[]) malloc(42);
 }
 

I guess it should be:

auto ptr = cast(char*)malloc(42)[0 .. 42];

Re: How to create a mutable array of strings?

[Daniel Kozak via Digitalmars-d-learn]

On Sunday, 17 May 2015 at 09:57:05 UTC, Daniel Kozak wrote:


On Sun, 17 May 2015 09:39:21 +
Dennis Ritchie via Digitalmars-d-learn 
digitalmars-d-learn@puremagic.com

wrote:


I remembered code Ali Çereli. It really helped:
http://forum.dlang.org/thread/ulhtlyxxclihaseef...@forum.dlang.org#post-mihl6m:241che:241:40digitalmars.com

-
import std.stdio, std.traits, std.range, std.algorithm;

auto deepDup(A)(A arr)
if (isArray!A)
{
static if (isArray!(ElementType!A)) {
return arr.map!(a = a.deepDup).array;

} else {
return arr.dup;
}
}




void main() {

auto s = [foo, bar].deepDup;

s[1][1] = 't';

writeln(s);
}
-
http://rextester.com/QBFH12695

P.S. Need to enable deepDup in Phobos.


allready there:
auto s = [foo, bar].map!a.dup.array;

:)

Ouch ignore this one :D

Re: ICE?

[Daniel Kozak via Digitalmars-d-learn]

On Sunday, 17 May 2015 at 09:25:33 UTC, Namespace wrote:
Is this error an ICE? I think so, because I see the internal 
filename, but I'm not sure.


Error: e2ir: cannot cast malloc(length * 8u) of type void* to 
type char[]


I would say this is not an ICE just normal error message.

Re: ICE?

[Daniel Kozak via Digitalmars-d-learn]

On Sun, 17 May 2015 10:36:33 +
Namespace via Digitalmars-d-learn digitalmars-d-learn@puremagic.com wrote:

 On Sunday, 17 May 2015 at 09:59:41 UTC, Daniel Kozak wrote:
 
  On Sun, 17 May 2015 09:33:27 +
  Namespace via Digitalmars-d-learn 
  digitalmars-d-learn@puremagic.com wrote:
 
  On Sunday, 17 May 2015 at 09:30:16 UTC, Gary Willoughby wrote:
   On Sunday, 17 May 2015 at 09:25:33 UTC, Namespace wrote:
   Is this error an ICE? I think so, because I see the 
   internal filename, but I'm not sure.
  
   Error: e2ir: cannot cast malloc(length * 8u) of type void* 
   to type char[]
  
   Have you got a code sample to reproduce this?
  
  Of course:
  
  
  void main() {
 import core.stdc.stdlib : malloc, free;
  
 auto ptr = cast(char[]) malloc(42);
  }
  
 
  I guess it should be:
 
  auto ptr = cast(char*)malloc(42)[0 .. 42];
 
 That would work, but I did it on purpose. I wanted to test 
 whether such dangerous code compiles or whether the compiler is 
 smart enough and hits the alarm.

I see.

for eg.:

void main() {
import core.stdc.stdlib : malloc, free;
static struct arr
{
size_t length;
void* ptr;
}
auto ptr = cast(arr) malloc(42);
}

prints:

test2.d(9): Error: cannot cast expression malloc(42LU) of type void* to arr

which makes sense, even e2ir text is not here.

Re: Printing an std.container.Array

[Daniel Kozak via Digitalmars-d-learn]

On Thu, 16 Apr 2015 19:55:52 +
Bayan Rafeh via Digitalmars-d-learn digitalmars-d-learn@puremagic.com wrote:

 Executing this code:
 
 import std.container.array;
 import std.stdio;
 
 
 int main() {
   writeln(Array!int([1, 2]));
   return 0;
 }
 
 outputs the following:
 
 Array!int(RefCounted!(Payload,
 cast(RefCountedAutoInitialize)0)(RefCountedStore(B694B0)))
 
 
 The strange thing is that this works fine:
 
 import std.container.array;
 import std.stdio;
 
 int main() {
   writeln(Array!int([1, 2])[0..$]);
   return 0;
 }
 
 [1, 2]
 
 How am I supposed to interpret this?

https://github.com/D-Programming-Language/phobos/pull/2875

Re: Printing an std.container.Array

[Daniel Kozak via Digitalmars-d-learn]

On Thu, 16 Apr 2015 13:05:48 -0700
H. S. Teoh via Digitalmars-d-learn digitalmars-d-learn@puremagic.com wrote:

 On Thu, Apr 16, 2015 at 07:55:52PM +, Bayan Rafeh via Digitalmars-d-learn
 wrote:
  Executing this code:
  
  import std.container.array;
  import std.stdio;
  
  
  int main() {
  writeln(Array!int([1, 2]));
  return 0;
  }
  
  outputs the following:
  
  Array!int(RefCounted!(Payload,
  cast(RefCountedAutoInitialize)0)(RefCountedStore(B694B0)))
  
  
  The strange thing is that this works fine:
  
  import std.container.array;
  import std.stdio;
  
  int main() {
  writeln(Array!int([1, 2])[0..$]);
  return 0;
  }
  
  [1, 2]
  
  How am I supposed to interpret this?
 
 Try slicing the Array before passing it to writeln?
 
   writeln(Array!int([1, 2])[]);
 
 Basically, there is a distinction between a container and a range that
 spans the items in a container. The conventional syntax for getting a
 range over a container's contents is the slicing operator [].
 
 
 T
 

Yep, but problem is almost no one expect this, or know this. We definitely
should do better.

Re: Converting void* to D array

[Daniel Kozak via Digitalmars-d-learn]

On Wednesday, 15 April 2015 at 04:43:39 UTC, Daniel Kozák wrote:


On Wed, 15 Apr 2015 04:24:20 +
Craig Dillabaugh via Digitalmars-d-learn
digitalmars-d-learn@puremagic.com wrote:


Hi.
I want to call a C library function that returns a data buffer 
as a void*.  How do I convert the resulting void* into 
something I can process in D?


//I have the following function from the GDAL C library.
extern(C) CPLErr GDALReadBlock( GDALRasterBandH, int, int, 
void* );



So I have (GByte is defined in the GDAL library):

void* buffer = malloc( GByte.sizeof * x_block_size * 
y_block_size );


I fill the buffer (and ignore any errors :o)

GDALReadBlock( AGDALRasterBandHInstance, xblock, yblock, 
buffer );



Now, how can I access the data in buffer?

Or you probably can do it like this:

auto buffer = new GByte[xblock*yblock];

GDALReadBlock( AGDALRasterBandHInstance, xblock, yblock,
(cast void*)buffer.ptr );
But in this case memory will be scan by GC. Which probably is not 
something what you want.

Re: alias this of non-public member

[Daniel Kozak via Digitalmars-d-learn]

On Tue, 07 Apr 2015 16:40:29 +
via Digitalmars-d-learn digitalmars-d-learn@puremagic.com wrote:

 Hi!
 
 Excuse me if this is obvious, but I can't recall coming across 
 anything similar and a quick search returns nothing relevant:
 
 struct Foo {
 }
 
 struct FooWrapper {
alias x_ this;
private Foo* x_; // doesn't work, as x_ is private
 }
 
 Basically, I want x_ to never be visible, except through the 
 alias this mechanism, at which point it should instead be seen 
 as public.
 
 Assuming something like this is not already possible in a clean 
 way, I would like to suggest a tiny(I think) addition to the 
 language:
 
 struct FooWrapper {
public alias x_ this; // overrides the visibility through the 
 alias;
private Foo* x_;
 }
 
 
 While I think this would be useful for the language, the reason I 
 want such a wrapper, is because I want to give opIndex, toString, 
 to a pointer, or, in fact just value semantics, while keeping the 
 rest of the interface through the pointer.
 
 I thought about using a class instead of a struct pointer, but I 
 am not sure about the memory layout for classes, nor about the 
 efficiency of overriding Object's methods, so I didn't want to 
 risk making it any less efficient. If someone could shed some 
 light about D's class memory layout and general performance 
 differences to a simple struct (or a C++ class for that matter), 
 that would also be great. In general, more information about 
 these sort of things would be great for us also-C++ programmers. 
 :)

Works for me:

struct M
{
void callMe() {
writeln(Ring...);
}
}

struct S
{
alias m this;
private M m;
}

void main(string[] args)
{
S s;
s.callMe();
}

Re: alias this of non-public member

[Daniel Kozak via Digitalmars-d-learn]

On Tuesday, 7 April 2015 at 17:21:09 UTC, Daniel Kozak wrote:


On Tue, 07 Apr 2015 16:40:29 +
via Digitalmars-d-learn digitalmars-d-learn@puremagic.com 
wrote:



Hi!

Excuse me if this is obvious, but I can't recall coming across 
anything similar and a quick search returns nothing relevant:


struct Foo {
}

struct FooWrapper {
   alias x_ this;
   private Foo* x_; // doesn't work, as x_ is private
}

Basically, I want x_ to never be visible, except through the 
alias this mechanism, at which point it should instead be 
seen as public.


Assuming something like this is not already possible in a 
clean way, I would like to suggest a tiny(I think) addition to 
the language:


struct FooWrapper {
   public alias x_ this; // overrides the visibility through 
the alias;

   private Foo* x_;
}


While I think this would be useful for the language, the 
reason I want such a wrapper, is because I want to give 
opIndex, toString, to a pointer, or, in fact just value 
semantics, while keeping the rest of the interface through the 
pointer.


I thought about using a class instead of a struct pointer, but 
I am not sure about the memory layout for classes, nor about 
the efficiency of overriding Object's methods, so I didn't 
want to risk making it any less efficient. If someone could 
shed some light about D's class memory layout and general 
performance differences to a simple struct (or a C++ class for 
that matter), that would also be great. In general, more 
information about these sort of things would be great for us 
also-C++ programmers. :)


Works for me:

struct M
{
void callMe() {
writeln(Ring...);
}
}

struct S
{
alias m this;
private M m;
}

void main(string[] args)
{
S s;
s.callMe();
}


module some;
import std.stdio;

Another way is use template mixin:

private mixin template M()
{
int someVar = 7;
public void callMe() {
writeln(Call);
}

public void callMe2() {
writeln(Call2);
}
}

struct S
{
mixin M;
}




module main;
import some;
void main(string[] args)
{
S s;
s.callMe();
s.callMe2();
}

Re: alias this of non-public member

[Daniel Kozak via Digitalmars-d-learn]

On Tuesday, 7 April 2015 at 17:43:08 UTC, Daniel Kozak wrote:

On Tuesday, 7 April 2015 at 17:21:09 UTC, Daniel Kozak wrote:


On Tue, 07 Apr 2015 16:40:29 +
via Digitalmars-d-learn digitalmars-d-learn@puremagic.com 
wrote:



Hi!

Excuse me if this is obvious, but I can't recall coming 
across anything similar and a quick search returns nothing 
relevant:


struct Foo {
}

struct FooWrapper {
  alias x_ this;
  private Foo* x_; // doesn't work, as x_ is private
}

Basically, I want x_ to never be visible, except through the 
alias this mechanism, at which point it should instead be 
seen as public.


Assuming something like this is not already possible in a 
clean way, I would like to suggest a tiny(I think) addition 
to the language:


struct FooWrapper {
  public alias x_ this; // overrides the visibility through 
the alias;

  private Foo* x_;
}


While I think this would be useful for the language, the 
reason I want such a wrapper, is because I want to give 
opIndex, toString, to a pointer, or, in fact just value 
semantics, while keeping the rest of the interface through 
the pointer.


I thought about using a class instead of a struct pointer, 
but I am not sure about the memory layout for classes, nor 
about the efficiency of overriding Object's methods, so I 
didn't want to risk making it any less efficient. If someone 
could shed some light about D's class memory layout and 
general performance differences to a simple struct (or a C++ 
class for that matter), that would also be great. In general, 
more information about these sort of things would be great 
for us also-C++ programmers. :)


Works for me:

struct M
{
void callMe() {
writeln(Ring...);
}
}

struct S
{
alias m this;
private M m;
}

void main(string[] args)
{
S s;
s.callMe();
}


module some;
import std.stdio;

Another way is use template mixin:

private mixin template M()
{
int someVar = 7;
public void callMe() {
writeln(Call);
}

public void callMe2() {
writeln(Call2);
}
}

struct S
{
mixin M;
}




module main;
import some;
void main(string[] args)
{
S s;
s.callMe();
s.callMe2();
}


And maybe Proxy can be use for your use case:
http://dlang.org/phobos/std_typecons.html#.Proxy

Re: Parameter is null by default. No value is given. Code says it is not null.

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 9 April 2015 at 14:42:33 UTC, Daniel Kozak wrote:

On Thursday, 9 April 2015 at 14:30:07 UTC, Daniel Kozak wrote:

On Thursday, 9 April 2015 at 14:25:56 UTC, Daniel Kozak wrote:

On Thursday, 9 April 2015 at 14:16:00 UTC, tcak wrote:
By the way, I am using DMD64 D Compiler v2.067.0 on Ubuntu 
14.04.


I have Archlinux DMD64 D Compiler v2.067.0 and it works OK 
for me.


WOW

rdmd app.d(without params):
Ok rdmd and dub works because they are use ldc, but do not me 
ask how. I always think that dub and rdmd should use dmd 
compiler until I tell them otherwise


I try it with DMD64 D Compiler v2.066 and same problem occured. 
So probably some backend problem. You should create an issue on 
https://issues.dlang.org

Re: Parameter is null by default. No value is given. Code says it is not null.

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 9 April 2015 at 14:30:07 UTC, Daniel Kozak wrote:

On Thursday, 9 April 2015 at 14:25:56 UTC, Daniel Kozak wrote:

On Thursday, 9 April 2015 at 14:16:00 UTC, tcak wrote:
By the way, I am using DMD64 D Compiler v2.067.0 on Ubuntu 
14.04.


I have Archlinux DMD64 D Compiler v2.067.0 and it works OK for 
me.


WOW

rdmd app.d(without params):
Ok rdmd and dub works because they are use ldc, but do not me ask 
how. I always think that dub and rdmd should use dmd compiler 
until I tell them otherwise

Re: Parameter is null by default. No value is given. Code says it is not null.

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 9 April 2015 at 14:25:56 UTC, Daniel Kozak wrote:

On Thursday, 9 April 2015 at 14:16:00 UTC, tcak wrote:
By the way, I am using DMD64 D Compiler v2.067.0 on Ubuntu 
14.04.


I have Archlinux DMD64 D Compiler v2.067.0 and it works OK for 
me.


WOW

rdmd app.d(without params):
Name: A 1  Expires null: true  Path equals null: true  Domain 
null: true
Name: A 2  Expires null: true  Path equals null: true  Domain 
null: true
Name: A 3  Expires null: true  Path equals null: true  Domain 
null: true
Name: A 4  Expires null: true  Path equals null: true  Domain 
null: true
Name: A 5  Expires null: true  Path equals null: true  Domain 
null: true

~
Name: A 6  Expires null: true  Path equals null: true  Domain 
null: true
Name: A 7  Expires null: true  Path equals null: true  Domain 
null: true
Name: A 8  Expires null: true  Path equals null: true  Domain 
null: true
Name: A 9  Expires null: true  Path equals null: true  Domain 
null: true

~
Name: A10  Expires null: true  Path equals null: true  Domain 
null: true
Name: A11  Expires null: true  Path equals null: true  Domain 
null: true
Name: A12  Expires null: true  Path equals null: true  Domain 
null: true
Name: A13  Expires null: true  Path equals null: true  Domain 
null: true




dmd -O:
Name: A 1  Expires null: true  Path equals null: false  Domain 
null: false
Name: A 2  Expires null: true  Path equals null: false  Domain 
null: false
Name: A 3  Expires null: true  Path equals null: false  Domain 
null: false
Name: A 4  Expires null: true  Path equals null: false  Domain 
null: false
Name: A 5  Expires null: true  Path equals null: false  Domain 
null: false

~
Name: A 6  Expires null: true  Path equals null: false  Domain 
null: false
Name: A 7  Expires null: true  Path equals null: false  Domain 
null: false
Name: A 8  Expires null: true  Path equals null: false  Domain 
null: false
Name: A 9  Expires null: true  Path equals null: false  Domain 
null: false

~
Name: A10  Expires null: true  Path equals null: false  Domain 
null: false
Name: A11  Expires null: true  Path equals null: false  Domain 
null: false
Name: A12  Expires null: true  Path equals null: false  Domain 
null: false
Name: A13  Expires null: true  Path equals null: false  Domain 
null: false


dmd -release:
Name: A 1  Expires null: true  Path equals null: true  Domain 
null: true
Name: A 2  Expires null: true  Path equals null: false  Domain 
null: false
Name: A 3  Expires null: true  Path equals null: false  Domain 
null: false
Name: A 4  Expires null: true  Path equals null: false  Domain 
null: false
Name: A 5  Expires null: true  Path equals null: false  Domain 
null: false

~
Name: A 6  Expires null: true  Path equals null: false  Domain 
null: false
Name: A 7  Expires null: true  Path equals null: false  Domain 
null: false
Name: A 8  Expires null: true  Path equals null: false  Domain 
null: false
Name: A 9  Expires null: true  Path equals null: false  Domain 
null: false

~
Name: A10  Expires null: true  Path equals null: false  Domain 
null: false
Name: A11  Expires null: true  Path equals null: false  Domain 
null: false
Name: A12  Expires null: true  Path equals null: false  Domain 
null: false
Name: A13  Expires null: true  Path equals null: false  Domain 
null: false


with ldc everything is ok

Re: Parameter is null by default. No value is given. Code says it is not null.

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 9 April 2015 at 14:16:00 UTC, tcak wrote:
By the way, I am using DMD64 D Compiler v2.067.0 on Ubuntu 
14.04.


I have Archlinux DMD64 D Compiler v2.067.0 and it works OK for me.

Re: Template type deduction and specialization

[Daniel Kozak via Digitalmars-d-learn]

On Wednesday, 20 May 2015 at 09:24:28 UTC, Daniel Kozák wrote:


On Wed, 20 May 2015 06:31:11 +
Mike Parker via Digitalmars-d-learn 
digitalmars-d-learn@puremagic.com

wrote:

I don't understand why this behaves as it does. Given the 
following two templates:


```
void printVal(T)(T t) {
writeln(t);
}
void printVal(T : T*)(T* t) {
writeln(*t);
}
```

I find that I actually have to explicitly instantiate the 
template with a pointer type to get the specialization.


```
void main() {
int x = 100;
printVal(x);
int* px = x;
printVal(px);// prints the address
 printVal!(int*)(px)  // prints 100
}
```

Intuitively, I would expect the specialization to be deduced 
without explicit instantiation. Assuming this isn't a bug 
(I've been unable to turn up anything in Bugzilla), could 
someone in the know explain the rationale behind this?


Because it cannot deduce type T:

try this:

void printVal(T : T*)(T* t) {
writeln(*t);
}

void main() {
int x = 100;
int* px = x;
printVal(px);
}

It will print error.

My advise is not to use T:T* or T:T[] it works only when 
explicitly

instantiate. Is better use T:M*,M or T:M[], M because it works
automaticly and you have both types available.

import std.stdio;

void printVal(T)(T t) {
writeln(t);
}

void printVal(T:M*,M)(T t) {
writeln(*t);
}

void main() {
int x = 100;
printVal(x);
int* px = x;
printVal(px);// prints the 100
}


DOCS: http://dlang.org/template.html#function-templates
says: Function template type parameters that are to be implicitly 
deduced may not have specializations:

Re: Template type deduction and specialization

[Daniel Kozak via Digitalmars-d-learn]

On Wednesday, 20 May 2015 at 07:27:53 UTC, jklp wrote:


---
import std.stdio;

void printVal(T)(T t) {
writeln(t);
}

void printVal(T: T)(T* t) {
writeln(*t);
}

void main() {
int x = 100;
printVal(x);
int* px = x;
printVal(px);
}
---

here it's selected correctly without explicit instantiation. 
But honestly i don't know why since the  asterisk is removed 
from the T it looks quite incorrect.


No it is correct it is same as:
void printVal(T: int)(T* t) {
writeln(*t);
}

Re: Template type deduction and specialization

[Daniel Kozak via Digitalmars-d-learn]

On Wednesday, 20 May 2015 at 09:35:48 UTC, Jonathan M Davis wrote:


Well, if

printVal!(int*)(px);

prints 100, then that's a bug. It should print the address. In 
fact, it
should be _impossible_ for the second overload of printVal to 
ever be

instantiated


IMHO thats not true, it should print 100. This is what spec say.

void printVal(T : T*)(T* t) {
writeln(*t);
}

T is deduce to be int so we have

void printVal(int* t) {
writeln(*t);
}

which will print value not address

Re: Utf8 to Utf32 cast cost

[Daniel Kozak via Digitalmars-d-learn]

On Mon, 08 Jun 2015 18:16:57 +
Anonymouse via Digitalmars-d-learn digitalmars-d-learn@puremagic.com wrote:

 On Monday, 8 June 2015 at 11:44:47 UTC, Daniel Kozák wrote:
  No difference even with GC.disable() results are same.
 
 Profile! Callgrind is your friend~
Yep, but I dont care, I am the one who makes transcode faster, so I am happy
with results :P. 

P.S. I care and probably when I have some spare time I will
improve to!dstring too

Re: Abstract sockets (linux)

[Daniel Kozak via Digitalmars-d-learn]

On Thu, 25 Jun 2015 15:56:04 +
freeman via Digitalmars-d-learn digitalmars-d-learn@puremagic.com wrote:

 I am having trouble using abstract sockets on Linux.
 
 Here is sample python code that works, which works:
  ptm_sockname = \0/var/run/ptmd.socket
  sock = socket.socket(socket.AF_UNIX, socket.SOCK_STREAM)
  sock.connect(ptm_sockname)
  sock.setblocking(1)
  sock.sendall('get-status detail')
 
 Similar code in D, which does not work:
  string socket_name = \0/var/run/ptmd.socket;
  auto address = new UnixAddress(socket_name);
  auto sock = new Socket(AddressFamily.UNIX, SocketType.STREAM);
  scope(exit) sock.close();
  sock.blocking = true;
  sock.connect(address);
  sock.send(get-status detail);
 
 This is the equivalent with socat, which works:
  $ echo get-status detail | socat - 
 ABSTRACT-CLIENT:/var/run/ptmd.socket
 
 My test D program exits on connect:
 std.socket.SocketOSException@runtime/phobos/std/socket.d(2674): 
 Unable to connect socket: Connection refused
 
 Any pointers?

instead of:
string socket_name = \0/var/run/ptmd.socket;
try:
string socket_name = /var/run/ptmd.socket;
works for me

Re: Qualified destructors / immutable objects

[Daniel Kozak via Digitalmars-d-learn]

On Friday, 12 June 2015 at 15:36:22 UTC, anonymous wrote:

no need for ~this() to modify immutable data:

class C {
int a;

this(int a) {
this.a = a;
}
}

struct S {
C elem = new C(42);
}

void main() {
import std.stdio;
immutable(S) s1;

//  Error: cannot modify immutable expression s1.elem.a
// s1.elem.a = 43;

writeln(s1.elem.a);

S s2;
s2.elem.a = 123;
writeln(s1.elem.a);
}

Prints:
42
123


Is there an existing issue on issue.dlang.org? If not can you 
report it

Re: Weird result of getsockopt

[Daniel Kozak via Digitalmars-d-learn]

On Sunday, 24 May 2015 at 16:51:44 UTC, CodeSun wrote:

Hello guys,
Today, I found a weird problem when I was learning to enable 
SO_KEEPALIVE for a specific socket. I use setsockopt to enable 
keepalive firstly, and then use getsockopt to show if it is 
enabled correctly.


My code snippet is listed below:

Dlang version:

import core.sys.posix.sys.socket;
import core.sys.posix.netinet.in_;
import std.c.stdio;

void main() {
int sfd = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP);
int flag = 1;
size_t size;
	setsockopt(sfd, SOL_SOCKET, SO_KEEPALIVE, flag, 
cast(socklen_t)flag.sizeof);

flag = 0;
	getsockopt(sfd, SOL_SOCKET, SO_KEEPALIVE, flag, 
cast(socklen_t*)size);

printf(%d\n, flag);
}

C version:

#include sys/socket.h
#include arpa/inet.h
#include stdio.h

int main() {
int sfd = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP);
int flag = 1;
size_t size;
	setsockopt(sfd, SOL_SOCKET, SO_KEEPALIVE, flag, 
(socklen_t)sizeof(flag));

flag = 0;
	getsockopt(sfd, SOL_SOCKET, SO_KEEPALIVE, flag, 
(socklen_t*)size);

printf(%d\n, flag);
return 0;
}

Dlang version always prints 0, which means keepalive is not 
enabled, while C version can almost display 1  all the time as 
expected.


So is there anything wrong inside the code? If not, whose 
behavior is correct?


Cause your code is wrong:
If the size of the option value is greater than option_len, the 
value stored in the object pointed to by the option_value
   argument shall be silently truncated. Otherwise, the 
object pointed to by the option_len argument shall be modified to 
indicate the  actual

   length of the value.

So because you have size set to 0 it will not work, you mast call 
it again and than it will probably work.


In C this work because size is not initialize which mean it could 
be anything

Re: Weird result of getsockopt

[Daniel Kozak via Digitalmars-d-learn]

On Sunday, 24 May 2015 at 21:13:02 UTC, Daniel Kozak wrote:

On Sunday, 24 May 2015 at 21:11:34 UTC, Daniel Kozak wrote:

On Sunday, 24 May 2015 at 16:51:44 UTC, CodeSun wrote:

Hello guys,
Today, I found a weird problem when I was learning to enable 
SO_KEEPALIVE for a specific socket. I use setsockopt to 
enable keepalive firstly, and then use getsockopt to show if 
it is enabled correctly.


My code snippet is listed below:

Dlang version:

import core.sys.posix.sys.socket;
import core.sys.posix.netinet.in_;
import std.c.stdio;

void main() {
int sfd = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP);
int flag = 1;
size_t size;
	setsockopt(sfd, SOL_SOCKET, SO_KEEPALIVE, flag, 
cast(socklen_t)flag.sizeof);

flag = 0;
	getsockopt(sfd, SOL_SOCKET, SO_KEEPALIVE, flag, 
cast(socklen_t*)size);

printf(%d\n, flag);
}

C version:

#include sys/socket.h
#include arpa/inet.h
#include stdio.h

int main() {
int sfd = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP);
int flag = 1;
size_t size;
	setsockopt(sfd, SOL_SOCKET, SO_KEEPALIVE, flag, 
(socklen_t)sizeof(flag));

flag = 0;
	getsockopt(sfd, SOL_SOCKET, SO_KEEPALIVE, flag, 
(socklen_t*)size);

printf(%d\n, flag);
return 0;
}

Dlang version always prints 0, which means keepalive is not 
enabled, while C version can almost display 1  all the time 
as expected.


So is there anything wrong inside the code? If not, whose 
behavior is correct?


Cause your code is wrong:
If the size of the option value is greater than option_len, 
the value stored in the object pointed to by the option_value
  argument shall be silently truncated. Otherwise, the 
object pointed to by the option_len argument shall be modified 
to indicate the  actual

  length of the value.

So because you have size set to 0 it will not work, you mast 
call it again and than it will probably work.


In C this work because size is not initialize which mean it 
could be anything

*call it again with right size and than it will probably work


I mean call it with something big enought, size than should be 
modified to real size

Re: Weird result of getsockopt

[Daniel Kozak via Digitalmars-d-learn]

On Sunday, 24 May 2015 at 21:11:34 UTC, Daniel Kozak wrote:

On Sunday, 24 May 2015 at 16:51:44 UTC, CodeSun wrote:

Hello guys,
Today, I found a weird problem when I was learning to enable 
SO_KEEPALIVE for a specific socket. I use setsockopt to enable 
keepalive firstly, and then use getsockopt to show if it is 
enabled correctly.


My code snippet is listed below:

Dlang version:

import core.sys.posix.sys.socket;
import core.sys.posix.netinet.in_;
import std.c.stdio;

void main() {
int sfd = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP);
int flag = 1;
size_t size;
	setsockopt(sfd, SOL_SOCKET, SO_KEEPALIVE, flag, 
cast(socklen_t)flag.sizeof);

flag = 0;
	getsockopt(sfd, SOL_SOCKET, SO_KEEPALIVE, flag, 
cast(socklen_t*)size);

printf(%d\n, flag);
}

C version:

#include sys/socket.h
#include arpa/inet.h
#include stdio.h

int main() {
int sfd = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP);
int flag = 1;
size_t size;
	setsockopt(sfd, SOL_SOCKET, SO_KEEPALIVE, flag, 
(socklen_t)sizeof(flag));

flag = 0;
	getsockopt(sfd, SOL_SOCKET, SO_KEEPALIVE, flag, 
(socklen_t*)size);

printf(%d\n, flag);
return 0;
}

Dlang version always prints 0, which means keepalive is not 
enabled, while C version can almost display 1  all the time as 
expected.


So is there anything wrong inside the code? If not, whose 
behavior is correct?


Cause your code is wrong:
If the size of the option value is greater than option_len, 
the value stored in the object pointed to by the option_value
   argument shall be silently truncated. Otherwise, the 
object pointed to by the option_len argument shall be modified 
to indicate the  actual

   length of the value.

So because you have size set to 0 it will not work, you mast 
call it again and than it will probably work.


In C this work because size is not initialize which mean it 
could be anything

*call it again with right size and than it will probably work

Re: Template type deduction and specialization

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 21 May 2015 at 13:12:36 UTC, Daniel Kozák wrote:


On Thu, 21 May 2015 08:54:54 -0400
Steven Schveighoffer via Digitalmars-d-learn
digitalmars-d-learn@puremagic.com wrote:


On 5/21/15 2:35 AM, Daniel Kozák via Digitalmars-d-learn wrote:

 On Wed, 20 May 2015 17:23:05 -0700
 Ali Çehreli via Digitalmars-d-learn
 digitalmars-d-learn@puremagic.com wrote:

 On 05/20/2015 04:10 PM, Mike Parker wrote:
 On Wednesday, 20 May 2015 at 13:46:22 UTC, Daniel Kozák 
 wrote:

 DOC say  `may not have` not `must not have` ;-)


 OK, if that's the intent, it needs to be reworded. As it 
 stands,
 it looks more like it's saying specialization is not 
 permissible,

 rather than what might be possible.

 That's the only meaning that I get: The doc means must 
 not. Yet,

 as you've shown, the behavior does not match the doc.

 Ali

 1.) we could fix just doc - easiest, but inconsistent

Before doing this, we have to understand what works and what 
doesn't. It's not clear to me.


 2.) remove implicit deduction even for fun(T:char)(T c) and 
 all

 other specialization - code breakage so imho not good

I don't think this is possible, this would break lots of 
existing

code.

 3.) fix doc and allow even fun(T:T*)(T* p) - same as 2

I agree with this fix. I don't understand why specialization 
should disqualify IFTI. Can someone explain this rationale 
besides because

the docs say so?



But this will break more code than 2. So it is impossible to 
fix it.


Not more, but it will be worst, because it could change behaviour 
of program without error message

Re: Static arrays inside struct and class - bug?

[Daniel Kozak via Digitalmars-d-learn]

V Sat, 01 Aug 2015 19:21:36 +
NX via Digitalmars-d-learn digitalmars-d-learn@puremagic.com napsáno:

 On Saturday, 1 August 2015 at 18:50:09 UTC, Daniel Kozak wrote:
  No you don't. You still use static allocation for array
 
 Can clarify why does that happen and I still suspect it's a 
 static allocation it would increase output exe if it was really 
 that static..?

No it would not increase output exe. Problem is with definition:
type[size] val; // static declaration so compilere check max 16M.

But you are right, in your case it could be improved and such
declaration could work.


because:

S {
  byte[16*1024*1024*1024] arr;
}

void main() {
auto s = new S();
if (s is null) {
  // error cannont allocate enought memory
}
   
}

but:

void main() {
  byte[16*1024*1024*1024] arr; // impossible to check if is allocated
}

Maybe you can open an enhancment on issues.dlang.org

Re: Static arrays inside struct and class - bug?

[Daniel Kozak via Digitalmars-d-learn]

On Saturday, 1 August 2015 at 18:07:51 UTC, NX wrote:

On Saturday, 1 August 2015 at 17:29:54 UTC, Adam D. Ruppe wrote:

Sorry, I can't see _the_ point in that. I understand that could 
be a problem if it was a global array but this scenery is 
completely wrong in my view. I'm already going to dynamically 
allocate it


No you don't. You still use static allocation for array

Re: Static arrays inside struct and class - bug?

[Daniel Kozak via Digitalmars-d-learn]

V Sat, 01 Aug 2015 19:16:16 +
NX via Digitalmars-d-learn digitalmars-d-learn@puremagic.com napsáno:

 On Saturday, 1 August 2015 at 18:47:00 UTC, Daniel Kozak wrote:
  Still same problem, You can`t allocate more then 16M on stack. 
  Use dynamic allocation
 
 I don't think new MyStruct allocates on stack, actually 
 allocating ~16MB on stack will immediatelly crash the program 
 which is not the case with NewExpression.

My fault It is not on stack, but still it is a static allocation

Re: Static arrays inside struct and class - bug?

[Daniel Kozak via Digitalmars-d-learn]

V Sat, 01 Aug 2015 18:07:50 +
NX via Digitalmars-d-learn digitalmars-d-learn@puremagic.com napsáno:

 On Saturday, 1 August 2015 at 17:29:54 UTC, Adam D. Ruppe wrote:
  On Saturday, 1 August 2015 at 17:22:40 UTC, NX wrote:
  I wonder if the followings are compiler bugs:
 
  No, it is by design, the idea is to keep static arrays smallish 
  so null references will be caught by the processor. (An overly 
  large static array could allow indexing it through a null 
  pointer to potentially reach another object.)
 
  The easiest workaround is to just dynamically allocate such 
  huge arrays:
 
  byte[] arr = new byte[](1024*1024*16);
  ReadProcessMemory(Proc, 0xdeadbeef, arr.ptr, arr.length, null);
 
  The arr.ptr and arr.length are the key arguments there.
 
 Sorry, I can't see _the_ point in that. I understand that could 
 be a problem if it was a global array but this scenery is 
 completely wrong in my view. I'm already going to dynamically 
 allocate it and my problem is actually a lot complex than what I 
 showed there, I not even allowed to do this:
 
 struct stuff
 {
 byte[1024*1024*16] arr; // Error: index 16777216 overflow for 
 static array
 }
 //...
 stuff* data = new stuff;
 ReadProcessMemory(Proc, (void*)0xA970F4, data, stuff.sizeof, 
 null);
 
 Here 
 (https://gist.github.com/NightmareX1337/6408287d7823c8a4ba20) is 
 the real issue if anyone want to see the real-world problem with 
 long lines of code

Still same problem, You can`t allocate more then 16M on stack. Use
dynamic allocation

Re: [dmd2.068] Bug or future?

[Daniel Kozak via Digitalmars-d-learn]

On Friday, 7 August 2015 at 06:26:21 UTC, VlasovRoman wrote:

I have some code:

import std.stdio;

auto dot(T, R)(T x, R y) {
return x * y;
}

struct Vector(T)
{
alias selftype = Vector!T;
int len = 5;
pure:
const @property{
static if( is( typeof( dot( selftype.init, 
selftype.init ) ) ) ){

auto len2() {return len * len;}
}

static if(is(typeof(T.init * T.init) == T)) {
auto e() {return len;}
}
}
}


void main() {
Vector!(float) vec;
float x = vec.len2();
writeln(x);
x = vec.e();
writeln(x);
}

I get error by compiler when i build this:
main.d(30): Error: no property 'len2' for type 'Vector!float', 
did you mean 'len'?


In dmd 2.067 is normaly.
is it Bug or enhancements?


Does not work in 2.067 for me.

Btw. you do not need to do this:

alias selftype = Vector!T;

You can just use Vector, or:

alias selftype = Vector;

if you prefer selftype as a name.

Re: zlib performance

[Daniel Kozak via Digitalmars-d-learn]

On Friday, 7 August 2015 at 09:12:32 UTC, yawniek wrote:

On Friday, 7 August 2015 at 08:50:11 UTC, Daniel Kozák wrote:

 ldc[2] -O -release -boundscheck=off -singleobj  app.d


ldc 0.15.2 beta2
2.86s user 0.55s system 77% cpu 4.392 total

v2.068-devel-8f81ffc
2.86s user 0.67s system 78% cpu 4.476 total

v2.067
2.88s user 0.67s system 78% cpu 4.529 total




i can now reproduce the results and indeed, its faster than 
zcat:

on a c4.xlarge aws instance running archlinux and dmd v2.067
same file as above on my macbook.

best run: 2.72s user 0.39s system 99% cpu 3.134 total
worst run: 3.47s user 0.46s system 99% cpu 3.970 total

zcat:
best: 4.45s user 0.28s system 99% cpu 4.764 total
worst: 4.99s user 0.57s system 99% cpu 5.568 total


so i guess on os x there is still something to be optimized


Can you try it without write operation (comment out all write)? 
And than try it without uncompression?



// without compression:

void main(string[] args)
{
  auto f = File(args[1], r);
  foreach (buffer; f.byChunk(4096))
  {
  write(cast(char[])buffer);
  }
}

// without write:

void main(string[] args)
{
  auto f = File(args[1], r);
  auto uncompressor = new UnCompress(HeaderFormat.gzip);

  foreach (buffer; f.byChunk(4096))
  {
  auto uncompressed = 
cast(char[])(uncompressor.uncompress(buffer));

  }
  uncompressor.flush;
}

Re: What D really needs :D

[Daniel Kozak via Digitalmars-d-learn]

On Saturday, 25 July 2015 at 18:30:16 UTC, Daniel Kozak wrote:

https://github.com/avinassh/rockstar?utm_content=buffer64b3cutm_medium=socialutm_source=twitter.comutm_campaign=buffer


Wrong forum, someone put learn on top.

http://forum.dlang.org/post/ooybhdoonxoxkzdsz...@forum.dlang.org

What D really needs :D

[Daniel Kozak via Digitalmars-d-learn]

https://github.com/avinassh/rockstar?utm_content=buffer64b3cutm_medium=socialutm_source=twitter.comutm_campaign=buffer

Re: static linking

[Daniel Kozak via Digitalmars-d-learn]

On Sat, 25 Jul 2015 18:02:46 +
Laeeth Isharc via Digitalmars-d-learn digitalmars-d-learn@puremagic.com wrote:

 Hi.
 
 Is static linking with dmd still broken on linux?  If so, can I 
 link statically with gdc or ldc, and if so how?
 
 https://issues.dlang.org/show_bug.cgi?id=12268
 
 I am trying to compile a D binary to run on AWS lambda.  If I 
 cannot link statically, which files should I include in the zip 
 upload - libphobos2.so, libdruntime-linux64so.o ?
 
 Thanks.
 
 
 Laeeth.

Do you mean dynamic linking?

I do not have any problems with dmd.
Can you please post more details what you are trying?

*.so are dynamic, *.[l]a are static

Re: OSX Foundation framework D binding

[Daniel Kozak via Digitalmars-d-learn]

V Wed, 11 Nov 2015 06:17:00 +
Vadim Lopatin via Digitalmars-d-learn
 napsáno:

> Hello,
> 
> I'm working on native Cocoa backend for DlangUI GUI library under 
> OSX.
> Is there any ready to use bindings for easy accessing Cocoa API?
> Probably, there is some HelloWorld program which creates window 
> and draws something?
> 
> 
> Best regards,
>   Vadim

I find only this one: http://code.dlang.org/packages/derelict-cocoa

Re: my first D program (and benchmark against perl)

[Daniel Kozak via Digitalmars-d-learn]

On Wednesday, 11 November 2015 at 13:32:00 UTC, perlancar wrote:
Here's my first non-hello-world D program, which is a direct 
translation from the Perl version. I was trying to get a feel 
about D's performance:


...

While I am quite impressed with how easy I was able to write D, 
I am not so impressed with the performance. Using rdmd (build 
20151103), the D program runs in 17.127s while the Perl version 
runs in 11.391s (so the D version is quite a bit *slower* than 
Perl's). While using gdc (Debian 4.9.2-10), I am able to run it 
in 3.988s (only about 3x faster than Perl's version).


I understand that string processing (concatenation, allocation) 
is quite optimized in Perl, I was wondering if the D version 
could still be sped up significantly?


Main problem is with allocations and with stripLeft, here is my 
version which is 10x faster than perls even with DMD. With LDC is 
12x faster


import std.stdio;
import std.array : appender;
import std.range;


auto fmttable(T)(T table) {
 auto res = appender!(string)();
 res.reserve(64);

 if (table.length == 0) return "";

 // column widths
 auto widths = new int[](table[0].length);

 foreach (rownum, row; table) { 
 foreach (colnum, cell; row) {  
 if (cell.length > widths[colnum])
widths[colnum] = cast(int)cell.length;
 }
 }

 foreach (row; table) {
 res.put("|");
 foreach (colnum, cell; row) {
 int l = widths[colnum] - cast(int)cell.length;
 res.put(cell);
 if (l)
 res.put(' '.repeat().take(l)); 
 res.put("|");
 }
 res.put("\n");
 }

 return res.data;
}

void main() {

auto table = [
["row1.1", "row1.2  ", "row1.3"],
["row2.1", "row2.2", "row2.3"],
["row3.1", "row3.2", "row3.3  "],
["row4.1", "row4.2", "row4.3"],
["row5.1", "row5.2", "row5.3"],
];

write(fmttable(table));
for (int i=0; i < 100; ++i) {
fmttable(table);
}
}

Re: my first D program (and benchmark against perl)

[Daniel Kozak via Digitalmars-d-learn]

V Thu, 12 Nov 2015 09:12:32 +
Daniel Kozak via Digitalmars-d-learn
<digitalmars-d-learn@puremagic.com> napsáno:

> On Wednesday, 11 November 2015 at 13:32:00 UTC, perlancar wrote:
> > Here's my first non-hello-world D program, which is a direct 
> > translation from the Perl version. I was trying to get a feel 
> > about D's performance:
> >
> > ...
> >
> > While I am quite impressed with how easy I was able to write D, 
> > I am not so impressed with the performance. Using rdmd (build 
> > 20151103), the D program runs in 17.127s while the Perl version 
> > runs in 11.391s (so the D version is quite a bit *slower* than 
> > Perl's). While using gdc (Debian 4.9.2-10), I am able to run it 
> > in 3.988s (only about 3x faster than Perl's version).
> >
> > I understand that string processing (concatenation, allocation) 
> > is quite optimized in Perl, I was wondering if the D version 
> > could still be sped up significantly?  
> 
> Main problem is with allocations and with stripLeft, here is my 
> version which is 10x faster than perls even with DMD. With LDC is 
> 12x faster
> 
> import std.stdio;
> import std.array : appender;
> import std.range;
> 
> 
> auto fmttable(T)(T table) {
>   auto res = appender!(string)();
>   res.reserve(64);
> 
>   if (table.length == 0) return "";
> 
>   // column widths
>   auto widths = new int[](table[0].length);
> 
>   foreach (rownum, row; table) {
> foreach (colnum, cell; row) { 
>   if (cell.length > widths[colnum])
>   widths[colnum] = cast(int)cell.length;
>   }
>   }
> 
>   foreach (row; table) {
>   res.put("|");
>   foreach (colnum, cell; row) {
>   int l = widths[colnum] - cast(int)cell.length;
>   res.put(cell);
>   if (l)
>res.put('
> '.repeat().take(l)); res.put("|");
>   }
>   res.put("\n");
>   }
> 
>   return res.data;
> }
> 
> void main() {
>   
>   auto table = [
>   ["row1.1", "row1.2  ", "row1.3"],
>   ["row2.1", "row2.2", "row2.3"],
>   ["row3.1", "row3.2", "row3.3  "],
>   ["row4.1", "row4.2", "row4.3"],
>   ["row5.1", "row5.2", "row5.3"],
>   ];
> 
>   write(fmttable(table));
>   for (int i=0; i < 100; ++i) {
>   fmttable(table);
>   }
> }
> 
> 

or with ~ operator:

import std.stdio;

auto fmttable(string[][] table) {

import std.array : appender, uninitializedArray;
import std.range : take, repeat;
import std.exception : assumeUnique;

auto res = appender(uninitializedArray!(char[])(128));
res.clear();
 
if (table.length == 0) return "";
// column widths
auto widths = new int[](table[0].length);

foreach (rownum, row; table) {   
foreach (colnum, cell; row) { 
if (cell.length > widths[colnum])
widths[colnum] = cast(int)cell.length;
}  
} 

foreach (row; table) {
res ~= "|";
foreach (colnum, cell; row) {
int l = widths[colnum] - cast(int)cell.length;
res ~= cell;
if (l) 
res ~= ' '.repeat().take(l);
res ~= "|";
}
res.put("\n");
}

 return res.data.assumeUnique();
}

void main() {

auto table = [
["row1.1", "row1.2  ", "row1.3"],
["row2.1", "row2.2", "row2.3"],
["row3.1", "row3.2", "row3.3  "],
["row4.1", "row4.2", "row4.3"],
["row5.1", "row5.2", "row5.3"],
];

write(fmttable(table));
for (int i=0; i < 100; ++i) {
fmttable(table);
}
}

Re: my first D program (and benchmark against perl)

[Daniel Kozak via Digitalmars-d-learn]

V Thu, 12 Nov 2015 11:03:38 +
Tobias Pankrath via Digitalmars-d-learn
 napsáno:

> > or with ~ operator:
> >
> > import std.stdio;
> >
> > [...]  
> 
> Did anyone check that the last loop isn't optimized out?

Yes, it is not optimized out

> Could also be improved further if you make the function take an
> output range and reuse one appender for every call, but that might be
> to far off the original perl solution.

I agree, that would be to far off the original solution.

Re: my first D program (and benchmark against perl)

[Daniel Kozak via Digitalmars-d-learn]

V Thu, 12 Nov 2015 12:13:10 +
perlancar via Digitalmars-d-learn 
napsáno:

> On Wednesday, 11 November 2015 at 14:20:51 UTC, Rikki Cattermole 
> wrote:
> > I turned it into mostly using large allocations, instead of 
> > small ones.
> > Although I'd recommend using Appender instead of my custom 
> > functions for this.
> >
> > Oh and for me, I got it at 2 secs, 513 ms, 397 μs, and 5 
> > hnsecs. Unoptimized, using dmd.
> > When release mode is enabled on dmd: 1 sec, 550 ms, 838 μs, and 
> > 9 hnsecs. So significant improvement even with dmds awful 
> > optimizer.  
> 
> Hi Rikki,
> 
> Thanks. With your version, I've managed to be ~4x faster:
> 
> dmd  : 0m1.588s
> dmd (release): 0m1.010s
> gdc  : 0m2.093s
> ldc  : 0m1.594s
> 
> Perl version : 0m11.391s
> 
> So, I'm satisfied enough with the speed for now. Turns out dmd is 
> not always slower.

It depends which flags do you use on ldc and gdc


ldc (-singleobj -release -O3 -boundscheck=off)
gdc (-O3 -finline -frelease -fno-bounds-check)

Re: my first D program (and benchmark against perl)

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 12 November 2015 at 12:25:08 UTC, Daniel Kozak wrote:

V Thu, 12 Nov 2015 12:13:10 +
perlancar via Digitalmars-d-learn 


napsáno:

On Wednesday, 11 November 2015 at 14:20:51 UTC, Rikki 
Cattermole wrote:

> I turned it into mostly using large allocations, instead of
> small ones.
> Although I'd recommend using Appender instead of my custom
> functions for this.
>
> Oh and for me, I got it at 2 secs, 513 ms, 397 μs, and 5
> hnsecs. Unoptimized, using dmd.
> When release mode is enabled on dmd: 1 sec, 550 ms, 838 μs, 
> and

> 9 hnsecs. So significant improvement even with dmds awful
> optimizer.

Hi Rikki,

Thanks. With your version, I've managed to be ~4x faster:

dmd  : 0m1.588s
dmd (release): 0m1.010s
gdc  : 0m2.093s
ldc  : 0m1.594s

Perl version : 0m11.391s

So, I'm satisfied enough with the speed for now. Turns out dmd 
is not always slower.


It depends which flags do you use on ldc and gdc


ldc (-singleobj -release -O3 -boundscheck=off)
gdc (-O3 -finline -frelease -fno-bounds-check)


import std.stdio;

auto fmttable(string[][] table) {

import std.array : appender, uninitializedArray;
import std.range : take, repeat;
import std.exception : assumeUnique;


if (table.length == 0) return "";
// column widths
auto widths = new int[](table[0].length);
	size_t total = (table[0].length + 1) * table.length + 
table.length;	


foreach (rownum, row; table) {
foreach (colnum, cell; row) {
if (cell.length > widths[colnum])
widths[colnum] = cast(int)cell.length;
}
}

foreach (colWidth; widths)
{
total += colWidth * table.length;
}   

auto res = appender(uninitializedArray!(char[])(total));
res.clear();

foreach (row; table) {
res ~= "|";
foreach (colnum, cell; row) {
int l = widths[colnum] - cast(int)cell.length;
res ~= cell;
if (l)
res ~= ' '.repeat().take(l);
res ~= "|";
}
res.put("\n");
}

 return res.data.assumeUnique();
}

void main() {

auto table = [
["row1.1", "row1.2  ", "row1.3"],
["row2.1", "row2.2", "row2.3"],
["row3.1", "row3.2", "row3.3  "],
["row4.1", "row4.2", "row4.3"],
["row5.1", "row5.2", "row5.3"],
];

writeln(fmttable(table));
for (int i=0; i < 100; ++i) {
fmttable(table);
}
}

dmd -O -release -inline -boundscheck=off  asciitable.d

real0m1.463s
user0m1.453s
sys 0m0.003s


ldc2 -singleobj -release -O3 -boundscheck=off asciitable.d

real0m0.945s
user0m0.940s
sys 0m0.000s

gdc -O3 -finline -frelease -fno-bounds-check -o asciitable 
asciitable.d


real0m0.618s
user0m0.613s
sys 0m0.000s


perl:

real0m14.198s
user0m14.170s
sys 0m0.000s

Re: my first D program (and benchmark against perl)

[Daniel Kozak via Digitalmars-d-learn]

On Thursday, 12 November 2015 at 12:49:55 UTC, Daniel Kozak wrote:
On Thursday, 12 November 2015 at 12:25:08 UTC, Daniel Kozak 
wrote:

... 
auto res = appender(uninitializedArray!(char[])(total));
res.clear();
...


this is faster for DMD and ldc:

auto res = appender!(string)();
res.reserve(total);

but for gdc(fronend version 2.066) it makes it two times slower 
(same for dmd, ldc 2.066 and older)