From jrheinlaender at gmx.de  Wed Jun  7 18:26:47 2017
From: jrheinlaender at gmx.de (=?UTF-8?Q?Jan_Rheinl=c3=a4nder?=)
Date: Wed, 7 Jun 2017 18:26:47 +0200
Subject: [GiNaC-devel] Substitution of wildcards
Message-ID: <f02c9b5b-74de-f9bd-ca20-d76e319c27b7@gmx.de>

Hi,

I'm wondering what is going on here (ginsh output):

> subs($1 * x  + 1/2 * x^2 * $2, match(a * x + b, $1 + x * $2));
b*a+1/2*a^3

What happened to "x" ? The match by itself correctly finds

> match(a * x + b, $1 + x * $2);
{$2==a,$1==b}

but the substitution result is weird.

Greetings,

    Jan




From kisilv at maths.leeds.ac.uk  Fri Jun  9 17:58:37 2017
From: kisilv at maths.leeds.ac.uk (Vladimir V. Kisil)
Date: Fri, 09 Jun 2017 16:58:37 +0100
Subject: [GiNaC-devel] Substitution of wildcards
In-Reply-To: <f02c9b5b-74de-f9bd-ca20-d76e319c27b7@gmx.de>
References: <f02c9b5b-74de-f9bd-ca20-d76e319c27b7@gmx.de>
Message-ID: <10599.1497023917@math-pc2069.leeds.ac.uk>

>>>>> On Wed, 7 Jun 2017 18:26:47 +0200, Jan Rheinl?nder <jrheinlaender at gmx.de> said:

    JR> Hi, I'm wondering what is going on here (ginsh output):

    >> subs($1 * x + 1/2 * x^2 * $2, match(a * x + b, $1 + x * $2));
    JR> b*a+1/2*a^3

    JR> What happened to "x" ? The match by itself correctly finds

    >> match(a * x + b, $1 + x * $2);
    JR> {$2==a,$1==b}

    JR> but the substitution result is weird.

    According to the GiNaC tutorial:

    "* A single wildcard matches any expression. "

    Thus, the wildcard $2 matches to x, then the expression

    $1 * $2 + 1/2 * $2^2 * $2

    is substituted using  {$2==a,$1==b}.

-- 
Vladimir V. Kisil                 http://www.maths.leeds.ac.uk/~kisilv/
  Book:     Geometry of Mobius Transformations     http://goo.gl/EaG2Vu
  Software: Geometry of cycles          http://moebinv.sourceforge.net/


From kisilv at maths.leeds.ac.uk  Mon Jun 12 16:42:21 2017
From: kisilv at maths.leeds.ac.uk (Vladimir V. Kisil)
Date: Mon, 12 Jun 2017 15:42:21 +0100
Subject: [GiNaC-devel] Substitution of wildcards
In-Reply-To: <10599.1497023917@math-pc2069.leeds.ac.uk>
References: <f02c9b5b-74de-f9bd-ca20-d76e319c27b7@gmx.de>
 <10599.1497023917@math-pc2069.leeds.ac.uk>
Message-ID: <25417.1497278541@math-pc2069.leeds.ac.uk>

>>>>> On Fri, 09 Jun 2017 16:58:37 +0100, "Vladimir V. Kisil" <kisilv at maths.leeds.ac.uk> said:

>>>>> On Wed, 7 Jun 2017 18:26:47 +0200, Jan Rheinl?nder <jrheinlaender at gmx.de> said:
    JR> Hi, I'm wondering what is going on here (ginsh output):

    >>> subs($1 * x + 1/2 * x^2 * $2, match(a * x + b, $1 + x * $2));
    JR> b*a+1/2*a^3

    JR> What happened to "x" ? The match by itself correctly finds

    The effect, which you are expecting, is achieved by by
  subs_options::no_pattern

#include <ginac/ginac.h>
using namespace std;
using namespace GiNaC;

int main(){
	realsymbol x("x"), a("a"), b("b");
	wildcard w0(0),  w1(1);
	exmap repl;
	(a * x + b).match(w0 + x * w1,repl);
	cout << (w0 * x + pow(x,2) * w1 / 2).subs(repl,subs_options::no_pattern) << endl;
	// -> b*x+1/2*a*x^2
}

-- 
Vladimir V. Kisil                 http://www.maths.leeds.ac.uk/~kisilv/
  Book:     Geometry of Mobius Transformations     http://goo.gl/EaG2Vu
  Software: Geometry of cycles          http://moebinv.sourceforge.net/


From jrheinlaender at gmx.de  Sat Jun 17 09:25:19 2017
From: jrheinlaender at gmx.de (=?UTF-8?Q?Jan_Rheinl=c3=a4nder?=)
Date: Sat, 17 Jun 2017 09:25:19 +0200
Subject: [GiNaC-devel] Compiler warnings in CLN
Message-ID: <8072a939-567e-f505-0962-8ba2880f0643@gmx.de>

Hi,

in cln/string.h it says:

--------------------

namespace cln {

struct cl_string;
^^^^

....

private:
// Friend declarations. They are for the compiler. Just ignore them.
        friend class cl_string;
                   ^^^

-----------------------

and the meticulous MSVC-compiler gives this a warning:

warning C4099: "cln::cl_string": Geben Sie den zuerst unter Verwendung
von "struct" und jetzt unter Verwendung von "class" gesehenen Namen ein

meaning that we should decide whether cl_string is a class or a struct.

Is there any interest in eliminating this kind of warning? Then I could
put together a list.

Greetings,

    Jan




From jrheinlaender at gmx.de  Mon Jun 19 18:24:21 2017
From: jrheinlaender at gmx.de (=?UTF-8?Q?Jan_Rheinl=c3=a4nder?=)
Date: Mon, 19 Jun 2017 18:24:21 +0200
Subject: [GiNaC-devel] is_exactly_a
Message-ID: <f2873445-27c3-b8ac-1116-5cc5ad0132b0@gmx.de>

Hi,

the GiNaC documentation says:

In contrast, 'is_exactly_a<T>(e)' allows you to check whether the
top-level object of an expression 'e' is an instance of the GiNaC class
'T', not including parent classes.

Shouldn't this read "not including subclasses"? Otherwise, I don't
understand what the documentation wants to say.

Greetings,

Jan




From kisilv at maths.leeds.ac.uk  Wed Jun 21 12:24:48 2017
From: kisilv at maths.leeds.ac.uk (Vladimir V. Kisil)
Date: Wed, 21 Jun 2017 11:24:48 +0100
Subject: [GiNaC-devel] is_exactly_a
In-Reply-To: <f2873445-27c3-b8ac-1116-5cc5ad0132b0@gmx.de>
References: <f2873445-27c3-b8ac-1116-5cc5ad0132b0@gmx.de>
Message-ID: <22506.1498040688@math-pc2069.leeds.ac.uk>

	Dear Jan,
>>>>> On Mon, 19 Jun 2017 18:24:21 +0200, Jan Rheinl?nder <jrheinlaender at gmx.de> said:
    JR> In contrast, 'is_exactly_a<T>(e)' allows you to check whether
    JR> the top-level object of an expression 'e' is an instance of the
    JR> GiNaC class 'T', not including parent classes.

    JR> Shouldn't this read "not including subclasses"? Otherwise, I
    JR> don't understand what the documentation wants to say.

    Consider:

	varidx k(symbol("k"),2);
	cout << is_a<idx>(k) << endl; // -> 1
	cout << is_exactly_a<idx>(k) << endl; // -> 0

  So, varidx k is an instance of idx, but not exactly idx---since the
  parent class idx of the class varidx is not included in the test
  is_exactly_a. The wording seems to be right to me.

  Best wishes,
  Vladimir
-- 
Vladimir V. Kisil                 http://www.maths.leeds.ac.uk/~kisilv/
  Book:     Geometry of Mobius Transformations     http://goo.gl/EaG2Vu
  Software: Geometry of cycles          http://moebinv.sourceforge.net/


From kreckel at in.terlu.de  Fri Jun 23 11:13:11 2017
From: kreckel at in.terlu.de (Richard B. Kreckel)
Date: Fri, 23 Jun 2017 11:13:11 +0200
Subject: [GiNaC-devel] Compiler warnings in CLN
In-Reply-To: <8072a939-567e-f505-0962-8ba2880f0643@gmx.de>
References: <8072a939-567e-f505-0962-8ba2880f0643@gmx.de>
Message-ID: <f3b98151-b120-6f17-7c9a-7fdda697c983@in.terlu.de>

On 06/17/2017 09:25 AM, Jan Rheinl?nder wrote:
> Is there any interest in eliminating this kind of warning? Then I could
> put together a list.

Sure! You may also consider sending a patch right away.

   -richy.
-- 
Richard B. Kreckel
<https://in.terlu.de/~kreckel/>