• "Bradeeoh" <bradeeoh crosswinds.net> Aug 28 2001
• Dan Hursh <hursh infonet.isl.net> Aug 28 2001
• nicO <nicolas.boulay ifrance.com> Sep 01 2001
• "kaffiene" <kaffiene xtra.co.nz> Sep 02 2001
• Dan Hursh <hursh infonet.isl.net> Sep 02 2001
• Axel Kittenberger <axel dtone.org> Sep 02 2001
• "Walter" <walter digitalmars.com> Sep 01 2001

"Bradeeoh" <bradeeoh crosswinds.net> writes:

If so, please tell me how unrealistic, just throwing out the idea.

Built-in array concatination.

byte[] b = new byte[3];
byte[] c = new byte[5];
byte[] d = b + c;
( d is now a new byte[], an empty byte[8] )

int[] e = new int[10];
d = c + e;    // error - can't mix byte[] and int[]

The idea is simple.  It just came up while working and I hit a particularly
nasty pocket of objects.

Of course, libraries could provide this sevice (not this nicely without
operator overloading...) and I've written my own libraries in both C++ and
Java to accomplish it.  But instead of calling functions, why not treat
arrays like Strings as far as simple '+' concatination goes?

I will admit that I didn't check the language spec before posting this
because I'm working and I can't take much time out.  I don't REMEMBER it
being there...  if it is already, I applaud Walter once again and return to
work gracefully.

If not, then any thoughts?

-Brady

Dan Hursh <hursh infonet.isl.net> writes:

Bradeeoh wrote:
 
 If so, please tell me how unrealistic, just throwing out the idea.
 
 Built-in array concatination.
 
 byte[] b = new byte[3];
 byte[] c = new byte[5];
 byte[] d = b + c;
 ( d is now a new byte[], an empty byte[8] )
 
 int[] e = new int[10];
 d = c + e;    // error - can't mix byte[] and int[]
 
 The idea is simple.  It just came up while working and I hit a particularly
 nasty pocket of objects.
 
 Of course, libraries could provide this sevice (not this nicely without
 operator overloading...) and I've written my own libraries in both C++ and
 Java to accomplish it.  But instead of calling functions, why not treat
 arrays like Strings as far as simple '+' concatination goes?
 
 I will admit that I didn't check the language spec before posting this
 because I'm working and I can't take much time out.  I don't REMEMBER it
 being there...  if it is already, I applaud Walter once again and return to
 work gracefully.
 
 If not, then any thoughts?
 
 -Brady



Actually, when I read the spec the first time I got the idea that:

	array0 = array1 + array2

for array's of the same size (that was required) meant:

	for(int i=0; i < array.length; i++)
		array0[i] = array1[i] + array2[i];

and I figured it was good that at least array addition would look like
addition.  When trying to verify my belief I reread the spec and
mournfully convinced myself that it was array concatenation.  I guess
I'd like to know which is it.

Dan

nicO <nicolas.boulay ifrance.com> writes:

Dan Hursh a écrit :
 
 Bradeeoh wrote:
 If so, please tell me how unrealistic, just throwing out the idea.

 Built-in array concatination.

 byte[] b = new byte[3];
 byte[] c = new byte[5];
 byte[] d = b + c;
 ( d is now a new byte[], an empty byte[8] )

 int[] e = new int[10];
 d = c + e;    // error - can't mix byte[] and int[]

 The idea is simple.  It just came up while working and I hit a particularly
 nasty pocket of objects.

 Of course, libraries could provide this sevice (not this nicely without
 operator overloading...) and I've written my own libraries in both C++ and
 Java to accomplish it.  But instead of calling functions, why not treat
 arrays like Strings as far as simple '+' concatination goes?

 I will admit that I didn't check the language spec before posting this
 because I'm working and I can't take much time out.  I don't REMEMBER it
 being there...  if it is already, I applaud Walter once again and return to
 work gracefully.

 If not, then any thoughts?

 -Brady


 Actually, when I read the spec the first time I got the idea that:
 
         array0 = array1 + array2
 
 for array's of the same size (that was required) meant:
 
         for(int i=0; i < array.length; i++)
                 array0[i] = array1[i] + array2[i];
 
 and I figured it was good that at least array addition would look like
 addition.  When trying to verify my belief I reread the spec and
 mournfully convinced myself that it was array concatenation.  I guess
 I'd like to know which is it.
 
 Dan


Maybe we could introduice new operator as '.' (dot) or 'concat'. So
introduice some other built-in operator. C introduice too few operator,
we should add '**' or '^' for power.


nicO
nicO

"kaffiene" <kaffiene xtra.co.nz> writes:

 Actually, when I read the spec the first time I got the idea that:

         array0 = array1 + array2

 for array's of the same size (that was required) meant:

         for(int i=0; i < array.length; i++)
                 array0[i] = array1[i] + array2[i];

 and I figured it was good that at least array addition would look like
 addition.  When trying to verify my belief I reread the spec and
 mournfully convinced myself that it was array concatenation.  I guess
 I'd like to know which is it.

 Dan


 Maybe we could introduice new operator as '.' (dot) or 'concat'. So
 introduice some other built-in operator. C introduice too few operator,
 we should add '**' or '^' for power.


^ is already used in C.

Dan Hursh <hursh infonet.isl.net> writes:

kaffiene wrote:
 
  >
 Actually, when I read the spec the first time I got the idea that:

         array0 = array1 + array2

 for array's of the same size (that was required) meant:

         for(int i=0; i < array.length; i++)
                 array0[i] = array1[i] + array2[i];

 and I figured it was good that at least array addition would look like
 addition.  When trying to verify my belief I reread the spec and
 mournfully convinced myself that it was array concatenation.  I guess
 I'd like to know which is it.

 Dan


 Maybe we could introduice new operator as '.' (dot) or 'concat'. So
 introduice some other built-in operator. C introduice too few operator,
 we should add '**' or '^' for power.


 ^ is already used in C.


For that matter, ** won't work because the is already an infix and
binary * operator.

	x ** y  =>  x * (*y)

I agree with the idea though.  I like perl because of it's liberal use
of infix and unary operations.  I like C++ because it let me do the
same, but it would be nice to have more operator symbol at my disposal,
and to be and to mess with precedence.  Since before first grade we are
taught infix operations.  They feel much more natural for anything
remotely numeric.

Dan

Axel Kittenberger <axel dtone.org> writes:

 For that matter, ** won't work because the is already an infix and
 binary * operator.
 
 x ** y  =>  x * (*y)


C is already broken in this regard:

     x /* y   =>   x / (*y)

Right? No, it's a comment :o)

- Axel

"Walter" <walter digitalmars.com> writes:

It's already there in the language <g>. -Walter

Bradeeoh wrote in message <9mfme5$v1g$1 digitaldaemon.com>...
If so, please tell me how unrealistic, just throwing out the idea.

Built-in array concatination.

byte[] b = new byte[3];
byte[] c = new byte[5];
byte[] d = b + c;
( d is now a new byte[], an empty byte[8] )

int[] e = new int[10];
d = c + e;    // error - can't mix byte[] and int[]

The idea is simple.  It just came up while working and I hit a particularly
nasty pocket of objects.

Of course, libraries could provide this sevice (not this nicely without
operator overloading...) and I've written my own libraries in both C++ and
Java to accomplish it.  But instead of calling functions, why not treat
arrays like Strings as far as simple '+' concatination goes?

I will admit that I didn't check the language spec before posting this
because I'm working and I can't take much time out.  I don't REMEMBER it
being there...  if it is already, I applaud Walter once again and return to
work gracefully.

If not, then any thoughts?

-Brady