Initializing assoc array to non-null

[d coder]

Greetings

Is there a way to initialize an associative array to a non-null (but still
empty) state? The only way I know is by adding an element and then removing
it. Did I miss something obvious? Basically I want to write lines 7-8 in
the following code in a cleaner fashion. Any ideas?

Regards
- Puneet

class Foo { // 1
  string[int] _aa;  // 2
  public string[int] aa() { // 3
return _aa; // 4
  } // 5
  this() {  // 6
_aa[0] = "";// 7
_aa.remove(0);  // 8
  } // 9
}   // 10
void main() {   // 11
  Foo foo = new Foo;// 12
  string[int] bb = foo.aa;  // 13
  assert(bb !is null);  // 14
  bb[0] = "zero";   // 15
  assert(bb is foo.aa); // 16
}   // 17

Re: Initializing assoc array to non-null

[d coder]

Thanks Jonathan

Do you think this could make a good enhancement request?

Regards
- Puneet



On Sat, Jun 1, 2013 at 9:34 PM, Jonathan M Davis wrote:

> On Saturday, June 01, 2013 20:21:42 d coder wrote:
> > Greetings
> >
> > Is there a way to initialize an associative array to a non-null (but
> still
> > empty) state? The only way I know is by adding an element and then
> removing
> > it. Did I miss something obvious? Basically I want to write lines 7-8 in
> > the following code in a cleaner fashion. Any ideas?
>
> Unfortunately, at the moment, I believe that that's the best that you can
> do -
> though it's trivial enough to write a function that does that for you so
> that
> you can make it a one-liner:
>
> auto aa = initAA!(string[int])();
>
> or
>
> initAA(aa);
>
> - Jonathan M Davis
>

class destructors and sub-objects

[d coder]

Greetings

I have a question on class destructor method. D documentation for
destructors says:

"The garbage col­lec­tor is not guar­an­teed to run the de­struc­tor for
all un­ref­er­enced ob­jects. Fur­ther­more, the order in which the garbage
col­lec­tor calls de­struc­tors for un­ref­er­ence ob­jects is not
spec­i­fied. This means that when the garbage col­lec­tor calls a
de­struc­tor for an ob­ject of a class that has mem­bers that are
ref­er­ences to garbage col­lected ob­jects, those ref­er­ences may no
longer be valid. This means that de­struc­tors can­not ref­er­ence sub
ob­jects."


Now I have a class A and it refers to subobject B. But B may or may not
have other objects referring to it. As a result, an object of A is GCed,
but subobject B may or may not be getting garbage collected.

In this scenario, am I allowed to make a reference to subobject B from the
destructor of A? Is there a way to find out if B is garbage collected or
not?

Regards
- Puneet

Facing problems with Class Properties

[d coder]

Greetings All

I am trying to compile the following D2 code. The code giving compilation
issues is the "this()" function of the class Foo. The constructor basically
tries to initialize all the data members of the class, of type BaseClass and
of type BaseClass array.

I am using class property tupleof to iterate over members of the class. Then
I check the type of each member and if the member is a BaseClass array, I
new all the elements of the array. Otherwise if the member is of the type
BaseClass, I new it as it is.

The issue is that when I try to compile the program, I get the error
bug.d(10): Error: no property 'length' for type 'test.Bar'

I am explicitly checking the field type, and I am making sure that the field
is an array type, before looking for its length. So I am not sure why this
error appears. Please guide me.

Regards
Cherry

import std.stdio;
class BaseClass { }

class Bar: BaseClass { }

class Foo: BaseClass {
  this() {
foreach(i, f; this.tupleof) {
  if (is (typeof(f) : BaseClass[])) {
for (size_t j = 0; j < f.length; ++j) {
  f[j] = new typeof(f[j]) ();
}
  }
  if (is(typeof(f) : BaseClass)) {
f = new typeof(f) ();
  }
}
  }
  Bar instance1;
  Bar instance2;
  Bar [10] instances;
}

unittest {
  Foo foo;
  foo = new Foo;
}

Re: Facing problems with Class Properties

[d coder]

> if(false) {
>    for (size_t j = 0; j < f.length...)
>    ...
> }
>
> Semantically this code is wrong as you can't take the length of f which is 
> class Bar. The static if forces the compiler to not generate this code as it 
> is known to be false.
>

Thanks Jesse

What you are saying makes sense to me. The problem is that the
following code works perfectly. I have just commented out some part
and replaced it with some debug statements.

The output from this code makes me believe that "is" and "typeof" do
have some run-time semantics. Or is it a D2 bug. BTW I am using
Digital Mars D Compiler v2.050.

Regards
- Puneet

import std.stdio;
class BaseClass { }

class Bar: BaseClass { }

class Foo: BaseClass {
  this() {
foreach(i, f; this.tupleof) {
  if (is (typeof(this.tupleof[i]) : BaseClass[])) {
writeln("Creating new objects for all ARRAY types ", 
this.tupleof[i].stringof);
// for (size_t j = 0; j < this.tupleof[i].length; ++j) {
//   this.tupleof[i][j] = new typeof(this.tupleof[i][j]) ();
// }
  }
  if (is(typeof(this.tupleof[i]) : BaseClass)) {
writeln("Creating new objects for all NON-ARRAY types ",
this.tupleof[i].stringof);
// this.tupleof[i] = new typeof(this.tupleof[i]) ();
  }
}
  }
  Bar instance1;
  Bar instance2;
  Bar [10] instances;
}

unittest {
  Foo foo;
  foo = new Foo;
}

// I am getting the following output
// Creating new objects for all NON-ARRAY types this.instance1
// Creating new objects for all NON-ARRAY types this.instance2
// Creating new objects for all ARRAY types this.instances

Re: Facing problems with Class Properties

[d coder]

> Is there a reason you can't directly reference the members?  After all, you
> are writing the class.
>

Thanks Steve for your inputs.

Actually this is part of a bigger code that I am trying to create.
Though I am pretty new to D.

What I have posted is a reduced code that illustrated the issue I am
facing. In the actual code, I am using string mixins and there are too
many classes (all derived from BaseClass) and too many class
instances, and the constructor call for Bar (and other classes) is a
bit more involved. There is a need to make sure that all the
constructors are called in a seamless fashion.

In fact, I will not be the end-user. I would just be coding the
BaseClass and the mixin. The actual end-users are not supposed to be
quite D-literate and I would be asking them to code something to the
effect:

class Foo: BaseClass {
  mixin(CreateInstances());
  // Define your Bar and other sub-class instances here
  //...
  // More code here
}

Any ideas?

Regards
- Cherry

Re: Facing problems with Class Properties

[d coder]

> What you are saying makes sense to me. The problem is that the
> following code works perfectly. I have just commented out some part
> and replaced it with some debug statements.

Ah.. Now I think I understand.

This new code I have written will all be run at compile time. So in
this case, the foreach statement inside the constructor would be
reduced to a bunch of writeln statements at compile time and those
writeln would be executed at the run time. This will not happen with
the actual code since there are other typeof and is statements in
there that can not be run at runtime.

Did I get it right?

Re: Facing problems with Class Properties

[d coder]

> Another thing, is(T : U) simply means T is implicitly castable to U.  Due to
> a compiler bug, Bar[] is implicitly castable to BaseClass[].
>

Steve

I realize that I am using this compiler bug as a feature. It would be
kind of you to suggest me a code that would not exploit this bug. I
was thinking of using something to the effect:

if (__traits(isStaticArray, this.tupleof[i]) {
  if (is (typeof(this.tupleof[i][0]) : BaseModule)) {

Regards
- Cherry

Re: Facing problems with Class Properties

[d coder]

> Ah.. Now I think I understand.
>
> This new code I have written will all be run at compile time. So in
> this case, the foreach statement inside the constructor would be
> reduced to a bunch of writeln statements at compile time and those
> writeln would be executed at the run time. This will not happen with
> the actual code since there are other typeof and is statements in
> there that can not be run at runtime.
>
> Did I get it right?
>

If I got it right now, It will be possible for me to unroll the
foreach loop and the if statements in a mixin and that would work
well.
Or may be changing the "if" statement to "static if" would do the
trick. I will give it a try.

Thank you Steve and tank you Jesse for showing me the light.

Regards
- Cherry

List of derived types?

[d coder]

Greetings

I need a way to know (using traits or other compile time constructs)
all the types derived from a given type.
Is it possible in D?

Is it possible to get a list of all the user-defined classes? I could
use that to filter out the classes that I need.

Regards
Cherry

Re: List of derived types?

[d coder]

> I'm curious, why do you need that?
>

It is a long story.

But basically I am creating a platform wherein the users would be
deriving classes from some base classes that I write as part of the
platform. And the users would often be deriving many such classes.

The end-users would often not be programmers and would know little D
(as you can see I am also learning D :-). I want to automate some code
generation for them and I hope to do that by creating wrapper classes
that would shadow the classes that the end-users have written. Since
the and users would be instantiating these classes only inside a
particular class scope, I wanted to create some code inside that class
scope. Right now I am forcing the end-users to insert some mixin for
every class that they code. I wanted to automate that process.

Hope what I said makes sense to you.

Regards
Cherry

Hope what I said makes sense.

is expression for template structs/classes instances?

[d coder]

Greetings

I want to find if a given struct type is instantiated from a
particular template struct type. For example:

struct S (T)  {
  alias T Type;
  T t;
}

And later I want to find out if a given type is of type S(*)
(basically any type instantiated from template struct S). In fact I do
not know the type value T used at the time of instantiating S!(T).

I was looking at "is ( Type Identifier : TypeSpecialization ,
TemplateParameterList )" expression at
http://www.digitalmars.com/d/2.0/expression.html#IsExpression .
Thought there would be some way using that, but I could not find any.

Regards
Cherry

Re: is expression for template structs/classes instances?

[d coder]

> For instance, given your definiton of S, you could use
> _traits/std.traits to check that the type that you're testing has a member
> variable t. You could then check that S!(typeof(t)) was the same as the type
> that you were testing. So, if you get particularly cunning about it, I believe
> that it can be tested for in specific cases, but I don't believe that it can 
> be
> done in any general way.
>

Thanks Jonathan

That is exactly what I had thought of doing. I was conscious that it
may not be the cleanest way.
Now that you are saying a cleaner way may not exist, I will go ahead
and write the code.

Regards
- Cherry

Re: is expression for template structs/classes instances?

[d coder]

> S!int foo;
> static if ( is( typeof(foo) f == S!T, T ) ) {
>    // Here, T == int, f == typeof(foo)
> }
>
> Note that the syntax "is ( Type Identifier : TypeSpecialization ,
> TemplateParameterList )" is only usable inside static if.
>

Thanks Simen

I do know the template. I will try out your solution. Will let you
know if I face issues.

Regards
- Cherry

Re: is expression for template structs/classes instances?

[d coder]

> I do know the template. I will try out your solution. Will let you
> know if I face issues.
>

Simen

It works perfect, And this is exactly what I was looking for. If you
see my original post, I also thought this form of "is" expression
should work. Just could not get around to the right syntax.

With your help it is working now. I am using a slightly more elaborate
check which is obvious but I am writing it here to just let the list
know.

static if ( is( typeof(foo) f == S!T, T : int) ) {
  // foo is an object of type S!T
  // where T is convertible to int
}

Thanks once more
Warm Regards
- Cherry

Nested Structs

[d coder]

Greetings All

I have a situation where I have a struct nested inside a class. I
would like to make the enclosing class' members visible inside the
nested struct's constructor. I see that such preposition is feasible
for nested classes, but not for nested structs. Why so?

Are there some alternative constructs which could give me this
behavior for nested constructs?

Regards
- Cherry

Re: Nested Structs

[d coder]

> A struct nested in a class does not have a hidden "outer" pointer as a
> nested class does.  It's because a struct is generally more bare-bones than
> a class (which has loads of hidden pieces: vtable, interfaces, classinfo,
> etc.).  Also, instantiating such a struct does not tie it to a class
> instance.
>

Thanks Steve

As far as I am concerned, this is a very limiting behavior wrt structs
nested inside classes. I have also observed that if you define a
constant (using enum for example) inside the enclosing class, the
constant remains visible in the nested struct too. So I believe the
language is making the hidden "outer" scope visible to the nested
struct (or maybe there is some other mechanism for enums that I am not
aware of).

> You need to implement this behavior on your own:
>

Actually I am trying to define a DSEL and the end users are not
typically programmers. I am in a situation where the end-user decides
the unit in an enclosing class (unit of length as an inch for example)
and later whenever he instantiates a line or a shape, the particular
unit is automatically considered at the time of instantiation. For my
application, I need value semantics and therefor using nested classes
in place of structs can not be considered.

Making the end-users pass the unit or its context (in form of this
pointer) everytime they instantiate a shape would be atrocious. It
would defeat the very purpose of letting the end-user define a unit.

I know I am not on the right mailing group, but I want to ask for this
particular language enhancement. I believe this would make the
behavior of the language more straight wrt nested structs inside
classes (are not these beasts expected to serve like nested classes or
even structs nested inside function scopes?). I hope my requirement is
generic enough and fits the bill for an enhancement request.

Regards
- Cherry

Maximum Number of Threads?

[d coder]

Greetings

  Is there a limit on the maximum number of threads that can be
spawned? Or does it just depend on the value in
/proc/sys/kernel/threads-max on a linux system?

Regards
- Cherry

Dynamic and Static Casting

[d coder]

Greetings All

I have learnt that D has only one casting operator and that is 'cast'.
The same operator assumes different functionality depending on the
context in which it he being used.

Now I have a situation where I have to downcast an object and I am
sure of the objects type and thereby I am sure that the downcast would
only be successful. To make the operation faster, in C++ I could have
used static_cast operator, thus giving the RTTI a skip. Would this be
possible in D? Can I force a static_cast which downcasting?

Regards
- Cherry

Re: Dynamic and Static Casting

[d coder]

Thanks Lars and Bearophile, I will give it a try.

I understand that static downcasting is dangerous. But there are
places where efficiency is paramount and you are sure that the casting
is safe. So I wholeheartedly second your proposal to have the stuff in
phobos.

Regards
- Cherry

Number of references to a Class Object

[d coder]

Greetings

I am in a situation where I need to know the number of references to a
class' object. To explain, I have an array of class objects and I
occasionally process this array. But if a particular object in this
array is not being garbage collected just because it is part of this
array, I would like to loose that object by making it null.

So either I need to find out the number of references to that object
and in case there is only one reference, I force the object null. If
this is not possible, I am sure there would be alternative solutions
since this should be a common situation faced by programmers.

Please help.

Regards
- Cherry

Re: Number of references to a Class Object

[d coder]

> I believe what you're referring to is generally called a Weak
> Reference, which is a reference that the GC doesn't consider when
> deciding to keep an object alive, but that the GC will update if an
> object dies.
> There's a feature request at 
> http://d.puremagic.com/issues/show_bug.cgi?id=4151
>

Thanks Andrew

Is there a way fro the users like myself to vote up an issue on DMD Bugzilla.

Regards
- Cherry

Re: Number of references to a Class Object

[d coder]

> Also tango (for D 1.0) implements it.
> Link:
> http://www.dsource.org/projects/tango/docs/current/tango.core.WeakRef.html
>
> Might be worth a look if you are going to implement it for D 2.0.
>

I looked at the D1 implementation. It depends on GC methods
weakPointerCreate and weakPointerDestroy. These functions are
implemented in D1 GC as extern C functions.

It seems most of this code should be directly portable to D2, it would
certainly require changes in the DMD2 source code. I have never worked
at that level and have used D2 only for past couple of months. While I
can give that a blind try,  it would be useful only if it gets
excepted in DMD2.

What do you guys suggest?

Regards
- Cherry

Re: Number of references to a Class Object

[d coder]

> If you know roughly what to do and want to take a stab at producing a viable
> patch to fix the problem, then feel free. If it's solid, it may get in. I 
> don't
> know. It doesn't hurt to try though (as long as you're prepared for the fact
> that it may not be accepted).

Thanks for letting me know of the licensing issues. I do not think I
have the expertize required to fix this myself. So I hang my shoes
here and wait for somebody else to get stuck in a similar situation.

Regards
- Cherry

Re: Number of references to a Class Object

[d coder]

> However, it's not generally an issue, because you shouldn't normally be 
> keeping
> references around for stuff that isn't used anymore. The garbage collector is
> smart enough to deal with circular references and the like, so the biggest 
> cases
> where you'd normally need weak references aren't actually a problem. You're
> trying to do something fairly abnormal here.
>

Thanks Jonathon, with some effort I hope to wriggle out of the situation.

Regards
- Cherry

Error: this for ~this needs to be type foo not type foo[1u][1u]

[d coder]

Greetings

I am getting this error when I am instantiating a struct array with a
single element inside another class and only if there is the
destructor for the struct is explicitly defined. Is it a known error?

Here is a minimized snippet that gives error:

$ rdmd --main -unittest test.d
Error: this for ~this needs to be type foo not type foo[1u][1u]
Error: this for ~this needs to be type foo not type foo[1u]


// test.d

struct foo {
  int foofoo;
  ~this() { // no error if explicit destructor not
// defined
  }
}

class bar {
  foo fred;
  foo[2][2] foofoo;
  foo[1] frop;  // this gives error
  foo[1][1] fropfrop;   // this too
}

unittest {
  foo frop;
  foo[1][1] fropfrop;   // this works
  bar burp;
}

Are function pointers compile time constants?

[d coder]

Greetings

I tried to initialize a struct member with a function pointer, and
found that DMD2 did not like it. Are not function pointers compile
time constants? And why they should not be?

Regards
- Cherry

Re: Are function pointers compile time constants?

[d coder]

Thanks Simon.

Doubt about Synchronized Code Clocks

[d coder]

Greetings

I have a doubt about synchronized code blocks.

I learnt that in Java the synchronized keyword has two fold effect.
Firstly it locks the code to make sure that only a single thread gets
access to the code block at a given time. Secondly, it makes sure that
the data elements accessed inside the code block are not "stale". As a
result Java programming practice is to synchronize even access
functions that access shared data elements.

How about D? Does D synchronized keyword just result in mutex locking
of the code block? Or does it also ensure that the accessed shared
data elements are not stale?

Thanks and Regards
- Puneet

Re: Doubt about Synchronized Code Clocks

[d coder]

> I'm afraid that I have no idea what would be "stale" about a shared
variable.
> sychronized uses a mutex, and if you want to avoid race conditions, you
need to
> use mutexes or something similar when dealing with shared variables. But I
don't
> know what would be "stale" about a variable.
>

One thread modifies a shared variable and the other thread still gets an old
value. I do not know if this is applicable to D at all. Just wanted to get a
clarification after I read an article in "Java Concurrency in Practice"
book. I quote a relevant paragraph:

Locking is not just about mutual exclusion; it is also about memory
> visibility. To ensure that all threads see the most up-to-date values of
> shared mutable variables, the reading and writing must synchronize on a
> common lock.


Regards

Why is it necessary to use synchronized functions when passing shared variables?

[d coder]

Greetings

Why is it necessary to use synchronized functions when passing shared
variables? I get error even when I am not modifying the shared variable in
the function.
Kindly look at the following code. I get a compile error unless I declare
the functions parent and root synchronized.

The compile error says:
test.d(13): Error: function test.HierObject.root () const is not callable
using argument types () shared const

Thanks and Regards
- Puneet

// Reduced Code
import std.exception;

// synchronized // Compiles without error when uncommented
class HierObject {
  private shared HierObject _root;
  private shared HierObject _parent;

  shared(const(HierObject)) root() const {
if(_root) return _root;
else {
  enforce(_parent,
  "HierObject Instance does not have a parent!");
  return this.parent().root();
}
  }

  shared(const(HierObject)) parent() const {
enforce(_parent);
return _parent;
  }
}

Re: Why is it necessary to use synchronized functions when passing shared variables?

[d coder]

>
>
> It's probably complaining because using shared without synchronizing is
> generally very foolish. Now, I would have _thought_ that it would still
> work
> without, but I apparently not. Regardless, I'm not sure why you'd want to
> use
> shared anything without synchronizing your access of it.



Thanks Jonathan

Actually in my actual use case, I am using synchronized at code block level
-- to limit the scope of locking. I am doing this to mitigate possible
inefficiency due to indiscriminate use of mutex locked code.

But D is forcing me to synchronize at function level, thus making most of my
code go under mutex locks.

Regards
- Puneet

Do I need to declare instances of Mutexes, Semaphores, and Barriers shared?

[d coder]

Greetings

Do I need to declare instances of Semaphores, Mutexes and Barriers shared?
In most situations these objects would be naturally accessed by many
threads.

Regards
- Puneet

Why is the struct instance being copied here?

[d coder]

Greetings

Please look at the code down here. When compiled and run, I get the message
"Call to postblit" printed. I think it is because of the foreach block,
because the variable "i" is not declared as ref there. Is there a way to
make it a ref?

Regards
- Puneet

import std.stdio;

struct Foo {
  this(this) {
writeln("Call to postblit");
  }
}

class Bar {
  Foo foo;
  this() {
foreach(i, f; this.tupleof) {
  // do nothing
}
  }
}

void main()
{
  Bar bar = new Bar();
}

Next Release

[d coder]

Greetings All

It has been 2 months since we had release 2.052. Just wondering when is the
2.053 release planned?

Also, is there an automated way to create a snapshot release from the git
repositories?

Regards
- Puneet

Unable to pass shared class method as delegate

[d coder]

Greetings

I am facing problem passing a shared class method as an argument to a
function. Look at the following minimized code snippet.

class Foo {
  shared // compiles when commented out
void bar() {}
}

void frop(void delegate() dg) {
}

void main() {
  shared Foo foo = new shared(Foo);
  frop(&foo.bar);
}

I get the following errors (using dmd 2.052)
dg.d(11): Error: function dg.frop (void delegate() dg) is not callable using
argument types (void delegate() shared)
dg.d(11): Error: cannot implicitly convert expression (&foo.bar) of type
void delegate() shared to void delegate()

Please suggest a valid signature for the function frop, so that it can take
shared delegates.

Regards
- Puneet

Re: Unable to pass shared class method as delegate

[d coder]

>
>
> I've copy and pasted my reply to Benjamin below, you should be able to
> adapt it to your needs:
>

Stupid me. Should have checked the list before posting.


>
> The only way I've found to do this that works is using an alias - this is
> probably worth a bug report if there isn't one already.


I checked and found that this is already reported
http://d.puremagic.com/issues/buglist.cgi?quicksearch=shared+delegate

And since Walter says that const issues are high on his priority list, I
hope this would get resolved soon.

Regards
- Puneet