Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.soft-sys.math.mathematica > #13338 > unrolled thread

Dynamic scoping

Started byDaniel Huber <danielphuber@gmail.com>
First post2013-02-02 06:15 +0000
Last post2013-02-06 06:50 +0000
Articles 4 — 3 participants

Back to article view | Back to comp.soft-sys.math.mathematica


Contents

  Dynamic scoping Daniel Huber <danielphuber@gmail.com> - 2013-02-02 06:15 +0000
    Re: Dynamic scoping "Rolf.Mertig@gmail.com" <Rolf.Mertig@gmail.com> - 2013-02-03 07:49 +0000
      Re: Dynamic scoping Richard Fateman <fateman@cs.berkeley.edu> - 2013-02-05 08:06 +0000
    Re: Dynamic scoping Daniel Huber <danielphuber@gmail.com> - 2013-02-06 06:50 +0000

#13338 — Dynamic scoping

FromDaniel Huber <danielphuber@gmail.com>
Date2013-02-02 06:15 +0000
SubjectDynamic scoping
Message-ID<keiatl$qq3$1@smc.vnet.net>
Block is supposed to be a dynamic scoping construct. Therefore, in the following, "Print" should be an undefined function. But consider:

Block[{Plus},
 t = ToString[Plus[1, 2]];
 Print[t];
 ]

The output of this is: "3" and not "Plus[1,2]" as it would be if "Print" has no defined meaning.
cheers, Daniel

[toc] | [next] | [standalone]


#13342

From"Rolf.Mertig@gmail.com" <Rolf.Mertig@gmail.com>
Date2013-02-03 07:49 +0000
Message-ID<kel4qe$38d$1@smc.vnet.net>
In reply to#13338
Am Samstag, 2. Februar 2013 07:15:17 UTC+1 schrieb Daniel Huber:
> Block is supposed to be a dynamic scoping construct. Therefore, in the following, "Print" should be an undefined function. But consider:
> 
> 
> 
> Block[{Plus},
> 
>  t = ToString[Plus[1, 2]];
> 
>  Print[t];
> 
>  ]
> 
> 
> 
> The output of this is: "3" and not "Plus[1,2]" as it would be if "Print" has no defined meaning.
> 
> cheers, Daniel

Block[{Plus = "Plus"}, t = ToString[Plus[1, 2]];
 Print[t];]
will do what you want

[toc] | [prev] | [next] | [standalone]


#13392

FromRichard Fateman <fateman@cs.berkeley.edu>
Date2013-02-05 08:06 +0000
Message-ID<keqehp$ern$1@smc.vnet.net>
In reply to#13342
On 2/2/2013 11:49 PM, Rolf.Mertig@gmail.com wrote:
...

>
> Block[{Plus = "Plus"}, t = ToString[Plus[1, 2]];
>   Print[t];]
> will do what you want
>

No, I think what he wanted was to display a bug in Mathematica.
You just wrote another program.

To illuminate the bug I tried this:

Addition[x_,y]:=x+y;

Block[{Plus=Addition},Print[Plus[1,2]]]

which returns an infinite stream of errors that look like
Join::heads: Heads PacletManager`Package`LDCfindForDocResource and 
PacletManager`Package`MCfindForDocResource at positions 1 and 2 are 
expected to be the same.

I think the explanation needs some explanation, but setting that aside
I tried these:

Block[{P = Addition}, Print[P[1, 2]]]

prints 3

Block[{P = foo}, Print[P[1, 2]]]

prints foo[1,2]

An explanation may be simply that

Block[{Plus}} ....]

binds the name Plus not to some "undefined object" but to the value of 
Plus in the dynamically enclosing scope.  Indeed, this works as
expected:

Block[{Plus = hello}, Block[{Plus = Plus}, Plus[1, 2]]]

returns hello[1,2]

this explanation is not accurate because

Block[{Plus = hello}, Block[{Plus}, Plus[1, 2]]]

returns 3.


So Block[{Plus}, ...
binds the name Plus, or any other Global symbol? to the Global value.

I think that this is a bug.

RJF







[toc] | [prev] | [next] | [standalone]


#13398

FromDaniel Huber <danielphuber@gmail.com>
Date2013-02-06 06:50 +0000
Message-ID<kesuge$jut$1@smc.vnet.net>
In reply to#13338
On Saturday, February 2, 2013 7:15:17 AM UTC+1, Daniel Huber wrote:
> Block is supposed to be a dynamic scoping construct. Therefore, in the following, "Print" should be an undefined function. But consider:

OOOPS, there was a typo, it should read: "Plus" should be an undefined function." not Print of course.
One more thing, "dynamic scoping" is well defined, it means: the value of a variable is temporariliy stored away and set to undefined until the current scope is left. This is although true for all calls made from the current scope.
Therefore, the above behaviour is a BUG!
cheers, Daniel

[toc] | [prev] | [standalone]


Back to top | Article view | comp.soft-sys.math.mathematica


csiph-web