www.digitalmars.com         C & C++   DMDScript  

digitalmars.D.bugs - [Issue 7176] New: Lambda => syntax for function and methods too

reply d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176

           Summary: Lambda => syntax for function and methods too
           Product: D
           Version: D2
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: DMD
        AssignedTo: nobody puremagic.com
        ReportedBy: bearophile_hugs eml.cc



(From an idea of Timon Gehr)

I think extending the applicability of the new lambda syntax to free
functions/methods is a nice idea, to shorten tiny functions/methods, that are
common enough:


struct C {
    int x;
    int getX() => x;
    int sqrX() => x ^^ 2;
}


Some examples from other languages:

A function to compute the arithmetic mean in Scala 2.7:

def mean(s: Seq[Int]) = s.foldLeft(0)(_+_) / s.size


From the Ada 2012 changes:
http://www.disca.upv.es/jorge/ae2010/slides/05-3_Language_Tech_Schonberg_Towards_Ada_2012.pdf


 To simplify the writing of pre/postconditions and predicates, allow
parametrized expressions (aka function bodies in package specs):


function Cube (X : integer) is (X ** 3);

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Dec 28 2011
next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176




There are 3 votes now. But this feature doesn't add a lot to D. This feature
looks nice, but I don't feel a need for it in my code.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jan 03 2012
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176


Jonathan M Davis <jmdavisProg gmx.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jmdavisProg gmx.com



PST ---
It would be too much of a departure from the normal syntax to enable the new
lambda syntax in general IMHO. It's useful for lambdas simply because without
it they risk being very verbose in what is already fairly dense code. Function
declarations don't really have that problem. Yes, the syntax is a bit verbose
if all you're doing is returning a value, but most functions do more than that,
and most functions are not declared in the midst of dense code like you
typically get with lambdas. This enhancement request is such a drastic
departure from the normal C-based syntax that I think that it would cause far
more harm than good.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jan 03 2012
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176


timon.gehr gmx.ch changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |timon.gehr gmx.ch



It is the same 'departure' as the one caused by the introduction of the new
lambda literals and therefore I cannot see how it can possibly cause any harm.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jan 03 2012
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176




PST ---
It's very different IMHO to introduce it in lambdas which are already part of
an expression and where the number of characters definitely matters than it is
to introduce it in normal function declarations. With declarations, they're on
their own instead of part of a larger expression. They just don't present the
same kind of gain and therefore don't merit the cost of the large departure in
syntax IMHO.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jan 03 2012
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176






Currently I am neutral toward this feature. I see it used in Scala and it looks
nice, but I don't think it will improve my D programs a lot.



 This enhancement request is such a drastic
 departure from the normal C-based syntax that I think that it would cause far
 more harm than good.


What kind of harm are you referring to?

I think it's not significantly bug-prone, and being already present in the
language (as lambda syntax) doesn't add a lot of complexity for the person that
has to learn D.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jan 03 2012
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176




PST ---

 What kind of harm are you referring to?


It doesn't fit with the rest of the language. The syntax is very different from
other declarations. This reduces readability and increases how much the
programmer has to deal with.

The verboseness of lambda expressions is a definite problem for readability, so
the syntax is arguably worth it for lambda expressions. But to then also use it
in declarations which don't have the same readibility problem is incurring that
cost where it's not worth it IMHO.

Obviously, this is perfectly valid as an enhancement request, but I hope that
the request is denied. I think that the lambda syntax is too different from
typical C-based syntax to be reasonable in normal function declarations.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jan 03 2012
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176


Jacob Carlborg <doob me.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |doob me.com



I could really have a use for this. I have a lot of methods that just returns a
single expression.

Another idea would be to allow optional braces for methods and functions, just
as for if-statements. This could be extended to all language features where
braces are used to make it more consistent.

In addition to the above we could make implicit returns possible to all
functions and methods.

I don't know which of these two ideas are farthest away from the normal C-based
syntax.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jan 03 2012
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176


David Piepgrass <qwertie256 gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |qwertie256 gmail.com



PDT ---
+1 from me. Limited C compatibility is one thing, but why should everything in
D look like C? Although I've written more code in C/C++ than any other
language, I haven't enjoyed it for many years now. I want a language that makes
me more productive, and I often use small functions (many of which return void
regardless of the expression type, btw, so that should be allowed.)

On the other hand, a lot of the small functions I write are boilerplate such as
property getters and forwarding functions in decorators, so maybe instead of a
special lambda syntax, what I really want is a few metaprograms to write those
functions for me.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jul 18 2012
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176







 On the other hand, a lot of the small functions I write are boilerplate such as
 property getters and forwarding functions in decorators, so maybe instead of a
 special lambda syntax, what I really want is a few metaprograms to write those
 functions for me.


I wouldn't mind some kind of property shortcut, like this:

class Foo
{
     property int bar;
}

Is lowered to this:

class Foo
{
    private int bar_;

     property int bar () { return bar_; }
     property int bar (int value) { return bar_ = value; }
}

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jul 18 2012
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176


Artem Borisovskiy <kolos80 bk.ru> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |kolos80 bk.ru



---

 class Foo
 {
      property int bar;
 }
 
 Is lowered to this:
 
 class Foo
 {
     private int bar_;
 
      property int bar () { return bar_; }
      property int bar (int value) { return bar_ = value; }
 }


Why not just make bar_ public? You do not add any code to the getter nor to the
setter anyway.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jul 20 2012
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176







 Why not just make bar_ public? You do not add any code to the getter nor to the
 setter anyway.


Perhaps I want it to be virtual, to be able to override it in a subclass.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jul 20 2012
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176




PDT ---




 Why not just make bar_ public? You do not add any code to the getter nor to the
 setter anyway.


 
 Perhaps I want it to be virtual, to be able to override it in a subclass.


Yes, or, quite often I want to write a trivial getter but a nontrivial setter.
So I'd like just the getter for free. Also, when the interface is going to be
exported, even a trivial property should often be a property instead of a
field, to avoid breaking binary compatibility if one changes one's mind and
wants to make it a property later (actually this even affects source
compatibility--a property can't be passed by reference).

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jul 20 2012
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176




PDT ---

 Yes, or, quite often I want to write a trivial getter but a nontrivial setter.
 So I'd like just the getter for free. Also, when the interface is going to be
 exported, even a trivial property should often be a property instead of a
 field, to avoid breaking binary compatibility if one changes one's mind and
 wants to make it a property later (actually this even affects source
 compatibility--a property can't be passed by reference).


That's why I've been tempted to suggest that  property on a variable made it so
that only operations which would still be legal if it were switched to being a
property function were allowed. I can't remember whether I ever actually opened
an enhancement request on that though. I'd have to go digging to find out.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jul 20 2012
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176







 Yes, or, quite often I want to write a trivial getter but a nontrivial setter.
 So I'd like just the getter for free. Also, when the interface is going to be
 exported, even a trivial property should often be a property instead of a
 field, to avoid breaking binary compatibility if one changes one's mind and
 wants to make it a property later (actually this even affects source
 compatibility--a property can't be passed by reference).


Other good points.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jul 20 2012
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176




After having used Scala a little, I now have changed my mind a little again.

In Scala you write:

def f3(x: Int, y: Int): Int = if (x == 0) x else x * y


This is current valid D code:

int f1(int x, int y) { return (x == 0) ? x : x ^^ 2; }

const f2 = (int x, int y) => (x == 0) ? x : x ^^ 2;



Allowing this in D is nice to reduce some syntax noise. So I now like this
idea:

int f4(int x, int y) => (x == 0) ? x : x ^^ 2;


In functional-style programming very short functions are common.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Mar 20 2013
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176


Nick Treleaven <ntrel-public yahoo.co.uk> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |ntrel-public yahoo.co.uk



07:15:30 PDT ---


 I could really have a use for this. I have a lot of methods that just returns a
 single expression.


I thought I'd add some hard data on this. There are quite a lot of these in
Phobos (edited results to only show larger count items):

$ git grep -Ec '\{\s*return\b' std/
std/algorithm.d:77
std/cpuid.d:27
std/format.d:35
std/functional.d:37
std/math.d:31
std/range.d:86
std/regex.d:44
std/traits.d:71
std/typecons.d:54
std/variant.d:23
std/xml.d:24

Admittedly, some of these may be false positives for e.g. lambdas, but a quick
scan through the results shows they are almost all one line function/method
definitions. I think this demonstrates a significant use case for the proposed
syntax.


 Another idea would be to allow optional braces for methods and functions, just
 as for if-statements.


That might not be ideal syntax with template constraints:
void foo(T)(T v) if (isFoo!T) writeln(v);
void foo(T)(T v) if (isFoo!T) => writeln(v);

The second syntax is clearer in distinguishing the constraint from if statement
syntax IMO.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Mar 20 2013
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176




---
I don't like this feature. Because:

1. it would reduce code readability.

   class LibClass {
     int foo() { return 1; }
     string bar() => "hi";
   }

   Mixing lambda syntax and normal function syntax looks messy.

2. Just only reducing 7 character is too small benefit.

   auto foo()=>expr;
   auto foo(){return expr;}

   With more complex function signature:

   ComplexReturnType!(..) foo(T, U, V)(T t, U u, V v) if (...)=>expr;
   ComplexReturnType!(..) foo(T, U, V)(T t, U u, V v) if (...){return expr;}

   Ratio will fall further.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Mar 20 2013
prev sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=7176






 I don't like this feature. Because:
 
 1. it would reduce code readability.
 


On the contrary! It also increases language consistency.


    class LibClass {
      int foo() { return 1; }
      string bar() => "hi";
    }
 
    Mixing lambda syntax and normal function syntax looks messy.
 


No. It is normal function syntax that looks messy in this case.

class LibClass {
    auto foo() => 1;
    auto bar() => "hi";
}



 2. Just only reducing 7 character is too small benefit.
 


7*_N_ characters. Also, it can get rid of additional indentation.


    auto foo()=>expr;
    auto foo(){return expr;}
 
    With more complex function signature:
 
    ComplexReturnType!(..) foo(T, U, V)(T t, U u, V v) if (...)=>expr;
    ComplexReturnType!(..) foo(T, U, V)(T t, U u, V v) if (...){return expr;}
 
    Ratio will fall further.


This is not a valid argument in any case.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Mar 20 2013