Ddoc and struct members
Hi all! How do I generate documentation for struct members using ddoc? I don't understand if there is a way to generate documentation for all the members without putting a '///' over every member
Re: Ddoc and struct members
On 14/12/2017 3:16 AM, n00nb wrote: Hi all! How do I generate documentation for struct members using ddoc? I don't understand if there is a way to generate documentation for all the members without putting a '///' over every member There isn't.
Re: How to catch line number of exception without catching it ?
On Wednesday, 13 December 2017 at 18:24:09 UTC, Thomas wrote:
Or is there a better solution for tracing the error position
from root till the branch ?
Speaking of tracing exceptions, here's my favourite one .. so far
;-)
(I mean come on.. debugging is great fun!)
btw. If you compile/run this code on Windows, using LDC, you'll
get a nice little beep from the speaker (when it runs).
//
module test;
import std.stdio;
void main()
{
auto str = "hello";
string * ptr =
writeln(ptr[1]); //compile without using the -O option
// writeln(ptr.ptr[1]); // the above line was meant to be
this line.
}
//
Re: `Socket.receive` providing arbitrary packet sizes and hanging without sending EOF
On 12/13/2017 11:39 AM, Unazed Spectaculum wrote:
> ubyte[] receiveBytes(T)(T socket, size_t receiveCount)
> {
> ubyte[] buffer = new ubyte[receiveCount];
> size_t count = socket.receive(buffer);
Don't trust code you find on newsgroups. :o) You have to check the
returned value first. According to documentation, it can return
Socket.ERROR:
https://dlang.org/phobos/std_socket.html#.Socket.receive
> there is always a superfluous chunk
> which is awaiting data.
Can you show with complete code? Perhaps the stream is in blocking mode?
> No matter what way I try; my code doesn't seem to know when to quit
> regardless of the check. Also for the arbitrary packet sizes, I would've
> expected that if I received N bytes X times, the first X-1 times would
> be perfectly N not some unusual integer.
> Simply put, say I'm receiving 1024 bytes 5 times. The length of each
> item on the stack looks like:
>
> [720,
> 490,
> 1024,
> 103
> ]
Posix read(2) man page says
"It is not an error if this number is smaller than the number of bytes
requested; this may happen for example because fewer bytes are actually
available right now (maybe because we were close to end-of-file, or because
we are reading from a pipe, or from a terminal), or because read() was
interrupted by a signal."
Ali
Re: Global variable type does not match previous declaration
On Wednesday, 13 December 2017 at 21:38:49 UTC, Satoshi wrote: What means this error and how to solve it? object.d-mixin-1072(1112): Error: Global variable type does not match previous declaration with same mangled name: _D10TypeInfo_m6__initZ Actually, I'm working on OS with minimal D runtime and I'm unable to compile object.d source code: https://github.com/Rikarin/Trinix/blob/a42a6e1fb4b87374b3e5ad8b9be501b080655ccd/Kernel/object.d Thanks Compiling with ldc2 -debuglib= -defaultlib= -code-model=kernel -disable-red-zone -w -wi -de -O3 -mattr=-sse -I../ -of=obj-amd64/object.d.o -c -deps=obj-amd64/object.d.o.o.dep object.d
Global variable type does not match previous declaration
What means this error and how to solve it? object.d-mixin-1072(1112): Error: Global variable type does not match previous declaration with same mangled name: _D10TypeInfo_m6__initZ Actually, I'm working on OS with minimal D runtime and I'm unable to compile object.d source code: https://github.com/Rikarin/Trinix/blob/a42a6e1fb4b87374b3e5ad8b9be501b080655ccd/Kernel/object.d Thanks
How to statically link Derelict SDL2 satellite libs
I'm on Linux 64 bit, all SDL libraries (-dev versions) are
installed, dub is set up with:
"dependencies": {
"derelict-sdl2": "~>3.0.0-beta"
},
"subConfigurations": {
"derelict-sdl2": "derelict-sdl2-static"
},
"libs":
["sdl2"]
All the SDL core library functions work fine but when I try to
use SDL_image library's IMG_Load() or IMG_LoadTexture() I get a
linker error, for example:
undefined reference to `IMG_LoadTexture'
collect2: error: ld returned 1 exit status
Error: linker exited with status 1
dmd failed with exit code 1.
I have:
import derelict.sdl2.image;
at the top of my program. I'm guessing that the "libs" section of
my dub subConfiguration needs to include the name of the image
satellite library?
Re: Understanding how dub works
On Wednesday, 13 December 2017 at 00:43:31 UTC, rikki cattermole
wrote:
1. Does dub compile dependencies as separate binaries? And if
yes how to specify where should be they placed?
Object/static files, but yes.
Where they go:
Windows: %APPDATA%/roaming/dub
Posix: ~/.dub
You don't need to change this and if you think you do, you're
wrong :)
2. It is possible to compile subpackage as a dynamic or static
library and link it to main binary file? Or just better create
separate dub package and use it as dependency?
Static yes, dynamic it won't link against (some bug last I
heard).
Have to do that manually.
"dependencies": { "mypackage:subpackage": "*" }
Change as required for SDL.
At first thanks for answers.
So going back to my first question. Code from dependencies is
compiled with mine project to one single executable?
Re: How to catch line number of exception without catching it ?
On Wednesday, 13 December 2017 at 18:24:09 UTC, Thomas wrote: So my question is: Is there a way to catch that line where the exception has happened without a catch ? Yes: use a debugger.
Re: How to catch line number of exception without catching it ?
On Wednesday, December 13, 2017 18:24:09 Thomas via Digitalmars-d-learn
wrote:
> Hi forks!
>
> I wanted to ask if there is a way to catch the line position on
> an exception without setting a try + catch block ?
> What I want is something like this:
>
>
> module main;
>
> import std.stdio;
> import std.conv;
>
> void foo()
> {
> scope(failure)
> {
> writeln("Got a failure in ", __FILE__, " ", __FUNCTION__, "!" );
> // what I miss is the line position with __LINE__
> }
>
> int x = to!int("1x"); // <-- here is the exception throwing
>
> }
>
> int main()
> {
> foo();
> return 0;
> }
>
>
> My goal is to trace the failure from the branch back to the root
> like this:
>
> (error) in "module2" at "foo2()" on line "..."
> (error) in "module1" at "foo1()" on line "..."
> (error) in "main" at "main()" on line "..."
>
> I don't want to set on every function a try + catch functionality
> because I think it is not a good coding style. Also I dont want
> to send on every possible position where an exception/error could
> happen (even when catched) the line position into an error log
> function. So I got the idea of writing out modulename and
> function with scope(failure) on the beginning of each function,
> which works great recursive, but only the line position is
> missing.
> So my question is: Is there a way to catch that line where the
> exception has happened without a catch ?
>
> Or is there a better solution for tracing the error position from
> root till the branch ?
If you want access to an exception in flight, you must catch it and rethrow
it. There is no other way.
But normally, if you want to see the stack when an exception was thrown,
you'd either use a debugger, or you'd look at the stack trace that gets
printed when the exception exits main (though unfortunately, as I understand
it, there's currently a regression, and the line numbers don't show up in
stack traces on Linux anymore, even when you compile with -g).
- Jonathan M Davis
`Socket.receive` providing arbitrary packet sizes and hanging without sending EOF
ubyte[] receiveBytes(T)(T socket, size_t receiveCount)
{
ubyte[] buffer = new ubyte[receiveCount];
size_t count = socket.receive(buffer);
return buffer[0 .. count];
}
string receiveAll(T)(T socket, size_t segmentSize = 1024)
{
ubyte[][] data;
size_t count = 0;
do
{
debug(1) writefln("Chunk %s", count);
data ~= receiveBytes(socket, segmentSize);
writeln(data[count]);
if (!data)
break;
} while(data[count++]);
char[] stringData;
foreach (elem; data)
stringData ~= elem;
debug(1) writeln(`Exiting "receiveAll"`);
return to!string(stringData);
}
I've tried many variations of the above code; both with the
retrieve integrated into the do-while loop in receiveAll; however
I was just seeing whether abstracting the code would let me spot
my issues a bit faster because it won't require me to make two
variables for the buffer and amount of bytes received, more like
the typical interface you get.
Issue is; when receiving any size buffer from the end-point
(tested with >1024 byte and <1024 byte buffers); there is always
a superfluous chunk which is awaiting data.
Listening: 0.0.0.0:6969
Client: somebody:58769
Chunk 0
[123, 34, 109, 101, 116, 104, 111, 100, 34, 58, 32, 34, 114, 101,
116, 114, 105, 101, 118, 101, 34, 44, 32, 34, 102, 105, 108, 101,
110, 97, 109, 101, 34, 58, 32, 34, 51, 48, 50, 54, 57, 50, 48,
49, 55, 50, 51, 54, 56, 54, 57, 49, 50, 49, 95, 97, 97, 97, 97,
97, 34, 44, 32, 34, 100, 97, 116, 97, 34, 58, 32, 34, 34, 125]
Chunk 1
[PAUSE]
Listening: 0.0.0.0:6969
Client: somebody:58767
Chunk 0
[123, 34, 109, 101, 116, 104, 111, 100, 34, 58, 32, 34, 114, 101,
116, 114, 105, 101, 118, 101, 34, 44, 32, 34, 102, 105, 108, 101,
110, 97, 109, 101, 34, 58, 32, 34, ... 97]
Chunk 1
[97, ... 125]
Chunk 2
[PAUSE]
No matter what way I try; my code doesn't seem to know when to
quit regardless of the check. Also for the arbitrary packet
sizes, I would've expected that if I received N bytes X times,
the first X-1 times would be perfectly N not some unusual integer.
Simply put, say I'm receiving 1024 bytes 5 times. The length of
each item on the stack looks like:
[720,
490,
1024,
103
]
Although I'd assume it'd be more like:
[1024,
1024,
289
]
What's up with this? It makes working with sockets so damn
tedious since there's no room for assumptions; can anybody
suggest an abstracted library for sockets in D or help with my
former issue?
How to catch line number of exception without catching it ?
Hi forks!
I wanted to ask if there is a way to catch the line position on
an exception without setting a try + catch block ?
What I want is something like this:
module main;
import std.stdio;
import std.conv;
void foo()
{
scope(failure)
{
writeln("Got a failure in ", __FILE__, " ", __FUNCTION__, "!" );
// what I miss is the line position with __LINE__
}
int x = to!int("1x"); // <-- here is the exception throwing
}
int main()
{
foo();
return 0;
}
My goal is to trace the failure from the branch back to the root
like this:
(error) in "module2" at "foo2()" on line "..."
(error) in "module1" at "foo1()" on line "..."
(error) in "main" at "main()" on line "..."
I don't want to set on every function a try + catch functionality
because I think it is not a good coding style. Also I dont want
to send on every possible position where an exception/error could
happen (even when catched) the line position into an error log
function. So I got the idea of writing out modulename and
function with scope(failure) on the beginning of each function,
which works great recursive, but only the line position is
missing.
So my question is: Is there a way to catch that line where the
exception has happened without a catch ?
Or is there a better solution for tracing the error position from
root till the branch ?
Thank you for your time!
Re: Tuple Array Sorting
On Tuesday, 12 December 2017 at 19:00:01 UTC, Biotronic wrote:
On Tuesday, 12 December 2017 at 15:19:35 UTC, Vino wrote:
import std.algorithm: filter, map, sort;
import std.container.array;
import std.file: SpanMode, dirEntries, isDir ;
import std.stdio: writefln;
import std.typecons: Tuple, tuple;
import std.datetime.systime: SysTime;
void main () {
auto FFs = ["C:\\Temp\\sapnas2\\BACKUP",
"C:\\Temp\\sapnas2\\EXPORT", "C:\\Temp\\sapnas2\\PROD_TEAM"];
Array!(Tuple!(string, SysTime)) Result;
foreach(d; FFs[]) {
auto dFiles = Array!(Tuple!(string, SysTime))(dirEntries(d,
SpanMode.shallow).filter!(a => a.isDir).map!(a =>
tuple(a.name, a.timeCreated)));
foreach(e; dFiles) { Result ~= e; } }
writefln("%(%-(%-63s %.20s %)\n%)", Result[].sort!((a, b)
=> a[1] < b[1]));
}
Since there's little need to extract timeCreated and name
before sorting, here's a version that doesn't:
import std.algorithm : map, filter, sort;
import std.array : array;
import std.range : join;
import std.file : SpanMode, dirEntries, isDir;
import std.stdio : writefln;
import std.typecons : tuple;
void main() {
auto folders = [`C:\Windows`, `C:\Program Files`,
`C:\Users`];
auto sorted = folders
.map!(f => f.dirEntries(SpanMode.shallow))
.join
.filter!(e => e.isDir)
.array
.sort!((a,b) => a.timeCreated < b.timeCreated)
.map!(e => tuple(e.name, e.timeCreated.toSimpleString[0
.. 20]));
writefln("%(%-(%-63s %s %)\n%)", sorted);
}
And a version with normal loops, since the heavily range-based
version above can be a bit dense. These programs do essentially
the same thing:
import std.algorithm : sort;
import std.array : array;
import std.file : SpanMode, dirEntries, DirEntry, isDir;
import std.stdio : writefln;
import std.typecons : tuple, Tuple;
void main() {
auto folders = [`C:\Windows`, `C:\Program Files`,
`C:\Users`];
DirEntry[] subFolders;
foreach (folder; folders) {
auto children = dirEntries(folder, SpanMode.shallow);
foreach (child; children) {
if (child.isDir) subFolders ~= child;
}
}
subFolders.sort!((a,b) => a.timeCreated < b.timeCreated);
Tuple!(string, string)[] interestingParts;
foreach (subFolder; subFolders) {
interestingParts ~= tuple(subFolder.name,
subFolder.timeCreated.toSimpleString[0..20]);
}
writefln("%(%-(%-63s %s %)\n%)", interestingParts);
}
As you can see, I'm just chopping off the parts I don't like
from toSimpleString. It seems a good format function for dates
does not exist in Phobos.
--
Biotronic
Hi Biotronic,
I was able to find a solution using container array and also
date formatting, below is the code, please do let me know if you
find any issue, as i have tested the script and it is working as
expected.
Program:
import std.algorithm: filter, map, sort, each;
import std.container.array;
import std.file: SpanMode, dirEntries, isDir ;
import std.stdio: writeln,writefln;
import std.typecons: Tuple, tuple;
import std.datetime.systime: SysTime;
import std.conv;
void main () {
auto FFs = ["C:\\Temp\\sapnas2\\BACKUP",
"C:\\Temp\\sapnas2\\EXPORT"];
Array!(Tuple!(string, SysTime)) Sorted;
foreach(d; FFs[]) {
auto dFiles = Array!(Tuple!(string, SysTime))(dirEntries(d,
SpanMode.shallow).filter!(a => a.isDir).map!(a => tuple(a.name,
a.timeCreated)));
foreach(i; dFiles[]){ Sorted ~= i; }
Sorted[].sort!((a,b) => a[1] > b[1]).each!(e => writefln!"%-63s
%.20s"(e[0], e[1].to!string));
}
}
From,
Vino.B
Re: Container Array or tuples Sorting
On Wednesday, 13 December 2017 at 15:58:40 UTC, Vino wrote:
On Wednesday, 13 December 2017 at 15:16:50 UTC, Vino wrote:
Hi All,
Request your help, on how to sort a tuple container array, I
have raised the same topic in one of the other thread "Tuple
Array Sorting" and was addressed to use standard array rather
than container array, and i am not able to find any document
or example in the library for the same.
Eg: Program.
import std.algorithm: filter, map, sort;
import std.container.array;
import std.file: SpanMode, dirEntries, isDir ;
import std.stdio: writeln;
import std.typecons: Tuple, tuple;
import std.datetime.systime: SysTime;
void main () {
auto FFs = ["C:\\Temp\\BACKUP", "C:\\Temp\\EXPORT"];
Array!(Tuple!(string, SysTime)) Result;
foreach(d; FFs[]) {
auto dFiles = Array!(Tuple!(string, SysTime))(dirEntries(d,
SpanMode.shallow).filter!(a => a.isDir)
.sort!((a,b) => a.timeCreated > b.timeCreated)
.map!(a => tuple(a.name, a.timeCreated)));
writeln(dFiles[]);
} }
From,
Vino.B
HI All,
As per the message from the below forum I understand that
that we cannot perform a sorting on filtered result a container
array but the same can be performed form the standard array, so
i adjusted the above code as below and getting a different
error than what is discussed in the forum.
Forum:
"https://forum.dlang.org/post/mcteinnryudlqvbkq...@forum.dlang.org;
Program:
void main () {
auto FFs = ["C:\\Temp\\sapnas2\\BACKUP",
"C:\\Temp\\sapnas2\\EXPORT"];
Array!(Tuple!(string, SysTime)) Result;
foreach(d; FFs[]) {
auto dFiles = Array!(Tuple!(string, SysTime))(dirEntries(d,
SpanMode.shallow).filter!(a => a.isDir).map!(a => tuple(a.name,
a.timeCreated)))[]
.sort!((a,b) => a[1] > b[1]);
writeln(dFiles[]);
} }
Error:
Message.d(14): Error: function
Message.main.SortedRange!(RangeT!(Array!(Tuple!(string,
SysTime))), __lambda3).SortedRange.opSlice (uint a, uint b) is
not callab
le using argument types ()
Failed: ["dmd", "-v", "-o-", "Message.d", "-I."]
From,
Vino.B
Hi All,
Was able to find a solution and it is working as expected
import std.algorithm: filter, map, sort, each;
import std.container.array;
import std.file: SpanMode, dirEntries, isDir ;
import std.stdio: writeln,writefln;
import std.typecons: Tuple, tuple;
import std.datetime.systime: SysTime;
import std.conv;
void main () {
auto FFs = ["C:\\Temp\\BACKUP", "C:\\Temp\\EXPORT"];
Array!(Tuple!(string, SysTime)) Sorted;
foreach(d; FFs[]) {
auto dFiles = Array!(Tuple!(string, SysTime))(dirEntries(d,
SpanMode.shallow).filter!(a => a.isDir).map!(a => tuple(a.name,
a.timeCreated)));
foreach(i; dFiles[]){ Sorted ~= i; }
Sorted[].sort!((a,b) => a[1] > b[1]).each!(e => writefln!"%-63s
%.20s"(e[0], e[1].to!string));
}
}
From,
Vino.B
Re: Date Formating
On Wednesday, 13 December 2017 at 17:16:46 UTC, Vino wrote: On Wednesday, 13 December 2017 at 08:32:34 UTC, codephantom wrote: [...] Hi All, [...] Hi All, Thank you very much , was able to resolve the issue by changing the writefln line as below. Sorted[].sort!((a,b) => a[1] > b[1]).each!(e => writefln!"%-63s %.20s"(e[0], e[1].to!string)); From, Vino.B
Re: Date Formating
On Wednesday, 13 December 2017 at 08:32:34 UTC, codephantom wrote:
On Wednesday, 13 December 2017 at 07:35:40 UTC, Jonathan M
Davis wrote:
In general, you probably want to cast the SysTime to a
DateTime if you're going to do something like that.
yes, I would agree ;-)
Of course the intention was not really to just format it the
same way as Clock.currTime() does it, but rather to provide a
way to more easily customise the format, however one likes,
whenever one likes..
e.g.the following small change to the format string would make
it return: 20171213_1924_41
(that's more like something I'd use)
return
format("%04s%02s%02s_%02s%02s_%02s",
(d.year),
to!(int)(d.month),
(d.day),
(d.hour),
(d.minute),
(d.second)
);
Hi All,
Request your help on below program on how to format or cast
SysTime to DateTime
import std.algorithm: filter, map, sort;
import std.container.array;
import std.file: SpanMode, dirEntries, isDir ;
import std.stdio: writeln,writefln;
import std.typecons: Tuple, tuple;
import std.datetime.systime: SysTime;
void main () {
auto FFs = ["C:\\Temp\\BACKUP", "C:\\Temp\\\EXPORT"];
Array!(Tuple!(string, SysTime)) Sorted;
foreach(d; FFs[]) {
auto dFiles = Array!(Tuple!(string, SysTime))(dirEntries(d,
SpanMode.shallow).filter!(a => a.isDir).map!(a => tuple(a.name,
a.timeCreated)));
foreach(i; dFiles[]){ Sorted ~= i; }
writefln("%(%-(%-63s %s %)\n%)", Sorted[].sort!((a,b) => a[1] <
b[1]));
}
}
Output:
C:\Temp\BACKUP\DND3
2017-Sep-05 14:31:00.7037169
C:\Temp\BACKUP\DND5
2017-Sep-05 14:31:00.750517
C:\Temp\EXPORT\DND6
2017-Sep-05 14:31:00.8909172
C:\Temp\BACKUP\dir1
2017-Sep-06 16:06:42.7223837
C:\Temp\EXPORT\dir2
2017-Sep-06 16:06:43.1435864
C:\Temp\BACKUP\dir2
2017-Sep-09 22:44:11.7604069
C:\Temp\BACKUP\dir3
2017-Dec-10 06:56:07.5122231
C:\Temp\BACKUP\t1
2017-Dec-11 04:10:02.6413853
Required Output
C:\Temp\BACKUP\DND3
2017-Sep-05 14:31:00
C:\Temp\BACKUP\DND5
2017-Sep-05 14:31:00
C:\Temp\EXPORT\DND6
2017-Sep-05 14:31:00
C:\Temp\BACKUP\dir1
2017-Sep-06 16:06:42
C:\Temp\EXPORT\dir2
2017-Sep-06 16:06:43
C:\Temp\BACKUP\dir2
2017-Sep-09 22:44:11
C:\Temp\BACKUP\dir3
2017-Dec-10 06:56:07
C:\Temp\BACKUP\t1
2017-Dec-11 04:10:02
From,
Vino.B
Re: GUI program on Mac OS in D?
On Wednesday, 13 December 2017 at 15:17:59 UTC, Jacob Carlborg wrote: I forgot to mention that there have been several discussions around adding support for reference counted classes. Several of the mentioning interfacing with Objective-C is important/a requirement. Ok, good to know! I have another question about your Webkit test example... I see that you are doing some elaborate setup in order to bind the application delegate methods. Can you explain a bit about why you are doing it in this way instead of using the @selector attribute in a class? The thing is that I'm currently attempting to get it to work using the "easy" way using @selector but the methods in my AppDelegate class are not called. Not sure why that is, but I probably screwed something up :)
Re: Container Array or tuples Sorting
On Wednesday, 13 December 2017 at 15:16:50 UTC, Vino wrote:
Hi All,
Request your help, on how to sort a tuple container array, I
have raised the same topic in one of the other thread "Tuple
Array Sorting" and was addressed to use standard array rather
than container array, and i am not able to find any document or
example in the library for the same.
Eg: Program.
import std.algorithm: filter, map, sort;
import std.container.array;
import std.file: SpanMode, dirEntries, isDir ;
import std.stdio: writeln;
import std.typecons: Tuple, tuple;
import std.datetime.systime: SysTime;
void main () {
auto FFs = ["C:\\Temp\\BACKUP", "C:\\Temp\\EXPORT"];
Array!(Tuple!(string, SysTime)) Result;
foreach(d; FFs[]) {
auto dFiles = Array!(Tuple!(string, SysTime))(dirEntries(d,
SpanMode.shallow).filter!(a => a.isDir)
.sort!((a,b) => a.timeCreated > b.timeCreated)
.map!(a => tuple(a.name, a.timeCreated)));
writeln(dFiles[]);
} }
From,
Vino.B
HI All,
As per the message from the below forum I understand that that
we cannot perform a sorting on filtered result a container array
but the same can be performed form the standard array, so i
adjusted the above code as below and getting a different error
than what is discussed in the forum.
Forum:
"https://forum.dlang.org/post/mcteinnryudlqvbkq...@forum.dlang.org;
Program:
void main () {
auto FFs = ["C:\\Temp\\sapnas2\\BACKUP",
"C:\\Temp\\sapnas2\\EXPORT"];
Array!(Tuple!(string, SysTime)) Result;
foreach(d; FFs[]) {
auto dFiles = Array!(Tuple!(string, SysTime))(dirEntries(d,
SpanMode.shallow).filter!(a => a.isDir).map!(a => tuple(a.name,
a.timeCreated)))[]
.sort!((a,b) => a[1] > b[1]);
writeln(dFiles[]);
} }
Error:
Message.d(14): Error: function
Message.main.SortedRange!(RangeT!(Array!(Tuple!(string,
SysTime))), __lambda3).SortedRange.opSlice (uint a, uint b) is
not callab
le using argument types ()
Failed: ["dmd", "-v", "-o-", "Message.d", "-I."]
From,
Vino.B
Re: GUI program on Mac OS in D?
On 2017-12-13 16:07, Jacob Carlborg wrote: On 2017-12-13 13:18, mrphobby wrote: Would it be possible to somehow hook this up automatically to the D destructor perhaps? Interested in hearing your thoughts on this! As far as I know, the destructor is only called (automatically by the GC). Since the Objective-C objects are not created using the GC the destructor will never be called. I don't think it's possible to implement reference counting (that would call retain/release) for classes in D, since it's not possible to overload the assignment operator for classes [2] or implement a postblit [3]. The only option, as far as I know, would be to wrap the class in a struct that calls retain/release automatically. But then you need to box/unbox the wrapped class constantly and all the method signatures would probably need to use this wrapper as well. I forgot to mention that there have been several discussions around adding support for reference counted classes. Several of the mentioning interfacing with Objective-C is important/a requirement. https://wiki.dlang.org/Language_design_discussions#Automatic_Reference_Counting_.28ARC.29_as_an_alternative_to_D.27s_Garbage_Collector http://forum.dlang.org/post/n0nnu0$1tth$1...@digitalmars.com https://wiki.dlang.org/FatPointer#Interfacing_with_Objective-C -- /Jacob Carlborg
Container Array or tuples Sorting
Hi All,
Request your help, on how to sort a tuple container array, I
have raised the same topic in one of the other thread "Tuple
Array Sorting" and was addressed to use standard array rather
than container array, and i am not able to find any document or
example in the library for the same.
Eg: Program.
import std.algorithm: filter, map, sort;
import std.container.array;
import std.file: SpanMode, dirEntries, isDir ;
import std.stdio: writeln;
import std.typecons: Tuple, tuple;
import std.datetime.systime: SysTime;
void main () {
auto FFs = ["C:\\Temp\\BACKUP", "C:\\Temp\\EXPORT"];
Array!(Tuple!(string, SysTime)) Result;
foreach(d; FFs[]) {
auto dFiles = Array!(Tuple!(string, SysTime))(dirEntries(d,
SpanMode.shallow).filter!(a => a.isDir)
.sort!((a,b) => a.timeCreated > b.timeCreated)
.map!(a => tuple(a.name, a.timeCreated)));
writeln(dFiles[]);
} }
From,
Vino.B
Re: GUI program on Mac OS in D?
On 2017-12-13 13:18, mrphobby wrote: I have been taking a look at your example. Looks pretty neat! Some advanced mixin stuff there that looks pretty useful. They're pretty basic ;) However, as far as I can tell there is no handling of retain/release. No, that's correct. How would you incorporate this into your mixin system without having to put "release" in all the interface definitions? You can add methods for retain/release in the ClassTrait template [1]. Would it be possible to somehow hook this up automatically to the D destructor perhaps? Interested in hearing your thoughts on this! As far as I know, the destructor is only called (automatically by the GC). Since the Objective-C objects are not created using the GC the destructor will never be called. I don't think it's possible to implement reference counting (that would call retain/release) for classes in D, since it's not possible to overload the assignment operator for classes [2] or implement a postblit [3]. The only option, as far as I know, would be to wrap the class in a struct that calls retain/release automatically. But then you need to box/unbox the wrapped class constantly and all the method signatures would probably need to use this wrapper as well. [1] https://github.com/jacob-carlborg/d_webkit_test/blob/master/source/foundation/util.d#L28 [2] https://dlang.org/spec/operatoroverloading.html#assignment [3] https://dlang.org/spec/struct.html#struct-postblit -- /Jacob Carlborg
Re: GUI program on Mac OS in D?
On Thursday, 23 November 2017 at 17:28:43 UTC, Jacob Carlborg wrote: I have a simple example [2] of an application that shows a window with a WebKit view, i.e. and embedded browser. This works with the upstream DMD and LDC compilers. It basically only contains bindings for what I needed for that sample application. As you'll see there you need to use some parts of the Objective-C runtime to create class instances and subclasses. Also some gymnastics are required for class/static methods. I have been taking a look at your example. Looks pretty neat! Some advanced mixin stuff there that looks pretty useful. However, as far as I can tell there is no handling of retain/release. How would you incorporate this into your mixin system without having to put "release" in all the interface definitions? Would it be possible to somehow hook this up automatically to the D destructor perhaps? Interested in hearing your thoughts on this!
Re: AssocArray to string is ok,but how to get the AssocArray from string? Thanks
On Tuesday, 12 December 2017 at 17:32:15 UTC, Frank Like wrote:
Hi,everyone,
who can help me,about the "AssocArray to string is ok,but how
to get the AssocArray from string? ".
For example:
SysTime[][string] AATimes;
AATimes["a1"] =[SysTime(DateTime(2017, 1, 1, 12, 33,
33)),SysTime(DateTime(2017, 1, 2, 12, 33, 33))];
AATimes["a2"] =[SysTime(DateTime(2017, 1, 2, 12, 33,
33)),SysTime(DateTime(2017, 1, 3, 12, 33, 33))];
ubyte[] ua = cast(ubyte[])AATimes.to!string;
writeln("ua is ",ua);
string strTimes = cast(string)ua;
writeln("strTimes is ",strTimes);
But now,how to get the AATimes from string?
Thanks.
Frank.
Have a look at std.conv.parse
Re: What's the proper way to use std.getopt?
On Wednesday, 13 December 2017 at 07:37:17 UTC, Jonathan M Davis wrote: On Wednesday, December 13, 2017 06:55:46 bauss via Digitalmars-d-learn wrote: [...] If it works, it's a bug related to code lowering (since scope statements are always lowered to try-catch-finally blocks). You're not supposed to have access to the exception in a scope statement. - Jonathan M Davis Yeah that's what I thought. Just tested and it doesn't allow you to.
Re: Date Formating
On Wednesday, 13 December 2017 at 07:35:40 UTC, Jonathan M Davis
wrote:
In general, you probably want to cast the SysTime to a DateTime
if you're going to do something like that.
yes, I would agree ;-)
Of course the intention was not really to just format it the same
way as Clock.currTime() does it, but rather to provide a way to
more easily customise the format, however one likes, whenever one
likes..
e.g.the following small change to the format string would make it
return: 20171213_1924_41
(that's more like something I'd use)
return
format("%04s%02s%02s_%02s%02s_%02s",
(d.year),
to!(int)(d.month),
(d.day),
(d.hour),
(d.minute),
(d.second)
);
