http://d.puremagic.com/issues/show_bug.cgi?id=5231

           Summary: BigInt lacks a normal toString()
           Product: D
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Phobos
        AssignedTo: nobody@puremagic.com
        ReportedBy: jmdavisProg@gmx.com


--- Comment #0 from Jonathan M Davis <jmdavisProg@gmx.com> 2010-11-17 21:02:36 PST ---
This program

import std.bigint;
import std.stdio;

void main()
{
    auto b = BigInt(42);
    writeln(b);
}


prints BigInt rather than 42. BigInt does not define a normal toString(). It
looks like it declares a version of toString() which takes a delegate and
format string in an attempt to have more control of what the string looks like.
However, this is useless for cases where you need an actual toString() -
particularly when functions which you have no control over call toString().
Normally, all types should define a toString() so that they can be printed, and
BigInt doesn't do that.

So, BigInt should declare a normal toString() - presumably one which prints out the BigInt in decimal form.

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

http://d.puremagic.com/issues/show_bug.cgi?id=5231


bearophile_hugs@eml.cc changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |bearophile_hugs@eml.cc


--- Comment #1 from bearophile_hugs@eml.cc 2010-11-17 23:10:31 PST ---
This is a dupe of my 4122

The lack of a normally usable toString is not acceptable.

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

http://d.puremagic.com/issues/show_bug.cgi?id=5231


Don <clugdbug@yahoo.com.au> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |clugdbug@yahoo.com.au


--- Comment #2 from Don <clugdbug@yahoo.com.au> 2010-11-18 00:03:42 PST ---
That's not really the correct solution.
BigInt should act like an int. Specifically,

BigInt b;
writefln(" b = %d, %x", b, b);
should just work.

This issue cannot be resolved until the definition of toString() is changed.
void toString() is a fundamentally broken design. It's wrong on many levels.

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

http://d.puremagic.com/issues/show_bug.cgi?id=5231



--- Comment #3 from Jonathan M Davis <jmdavisProg@gmx.com> 2010-11-18 00:24:35 PST ---
void toString()? Normally, it's string toString(). The only problem that I'm
aware of with regards to toString() design at the moment is the fact that it
must be _exactly_ string toString() and can't be const or pure or whatnot.

I don't see why BigInt can't just have a normal toString() which returns a
string representation of BigInt. Having a fancier toString() like it does now
may be useful, but I don't see how it precludes having a normal toString().

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

http://d.puremagic.com/issues/show_bug.cgi?id=5231


nfxjfg@gmail.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |nfxjfg@gmail.com


--- Comment #4 from nfxjfg@gmail.com 2010-11-18 00:56:03 PST ---
(In reply to comment #2)
> void toString() is a fundamentally broken design. It's wrong on many levels.

But it allows more control over formatting and potentially reduces memory allocation. string toString() seems more broken to me: no control, forces you to do memory allocation. (Another broken design issue is that _all_ objects in D have a toString() method, even if it doesn't make sense, but that is off-topic here.)

What is worrying is that void toString is nowhere documented. Does std.format use it or what?

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

http://d.puremagic.com/issues/show_bug.cgi?id=5231



--- Comment #5 from nfxjfg@gmail.com 2010-11-18 00:57:38 PST ---
(In reply to comment #4)
> What is worrying is that void toString is nowhere documented. Does std.format use it or what?

Should be: is it supposed to use it? Of course it doesn't right now. If it should, the bug report is about std.format/writefln, and not std.bigint.

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

http://d.puremagic.com/issues/show_bug.cgi?id=5231


Steven Schveighoffer <schveiguy@yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |schveiguy@yahoo.com


--- Comment #6 from Steven Schveighoffer <schveiguy@yahoo.com> 2010-11-18 05:10:23 PST ---
(In reply to comment #2)
> That's not really the correct solution.
> BigInt should act like an int. Specifically,
> 
> BigInt b;
> writefln(" b = %d, %x", b, b);
> should just work.
> 
> This issue cannot be resolved until the definition of toString() is changed.
> void toString() is a fundamentally broken design. It's wrong on many levels.

So BigInt's aren't printable via writeln in protest?  I guess I don't understand why you can't do this:

string toString()
{
  string retval;
  void sink(const(char)[] data) { retval ~= data; }
  toString(&sink, null);
  return retval;
}

It doesn't hurt/limit the current toString, does it?  And then it makes bigints printable via writeln.

BTW, toString's delegate is not scope, so you are going to allocate a closure for that...

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

http://d.puremagic.com/issues/show_bug.cgi?id=5231



--- Comment #7 from Don <clugdbug@yahoo.com.au> 2010-11-18 05:58:20 PST ---
(In reply to comment #6)
> (In reply to comment #2)
> So BigInt's aren't printable via writeln in protest?

Yes. I refuse to play any part in propagating the toString() abomination.
BigInt will NEVER have a string toString() function. No frigging way.

This needs to be fixed in writefln/format.

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

http://d.puremagic.com/issues/show_bug.cgi?id=5231



--- Comment #8 from Steven Schveighoffer <schveiguy@yahoo.com> 2010-11-18 06:30:23 PST ---
Pardon me for saying so, but that's too short-sighted.

first, it's not just a writeln/format change, it's a compiler change too.  The compiler is the one who decides whether to store a function pointer to toString in the TypeInfo_Struct.  Maybe you can help fix that...

Second, it's they way things currently work.  It's like saying you refuse to have const functions because they should be inout, but inout doesn't work. When toString is fixed, then you can remove the crufty function, and nobody cares whether it was ever there or not.  It looks to the outside like phobos is immature when it can't even print its own types, regardless of how inefficient it is.

Note that I 100% agree that the current system is crap, and needs to be completely redone similar to how you have implemented it, but it's not how it works now.  Can't BigInt just play along and we can push for changes to the system without making the library look like a stubborn child?

BTW, when the system is changed, I wouldn't want it to be called toString, since string may have nothing to do with it.  I'd call it something like format or output.

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

http://d.puremagic.com/issues/show_bug.cgi?id=5231



--- Comment #9 from Don <clugdbug@yahoo.com.au> 2010-11-18 08:42:45 PST ---
(In reply to comment #8)

> Second, it's they way things currently work.  It's like saying you refuse to have const functions because they should be inout, but inout doesn't work. When toString is fixed, then you can remove the crufty function, and nobody cares whether it was ever there or not.  It looks to the outside like phobos is immature when it can't even print its own types, regardless of how inefficient it is.

If it was just a question of inefficiency, I would have implemented it. The
issue is that it doesn't get the formatting string.
So
BigInt b;
writefln("%x %+d", b, b);

doesn't work, and cannot be made to work.

> Note that I 100% agree that the current system is crap, and needs to be completely redone similar to how you have implemented it, but it's not how it works now.  Can't BigInt just play along and we can push for changes to the system without making the library look like a stubborn child?

No. The format string is absolutely fundamental to the implementation of outputting BigInt and BigFloat. It's not just "a crap implementation".

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