Re: Bug: clear leaves global variables in symbol table

[Jaroslav Hajek]

On Mon, Jun 8, 2009 at 3:31 PM, John W. Eaton<address@hidden> wrote:
> On  8-Jun-2009, Jaroslav Hajek wrote:
>
> | On Sun, Jun 7, 2009 at 12:15 AM, Rik<address@hidden> wrote:
> | > 6/6/09
> | >
> | > In the Mercurial tip, 'clear' by itself should remove both local and
> | > global variables according to the documentation.  Instead it only
> | > appears to remove global variables but they still exist according to
> | > 'who global'.
> | >
> | > Code to reproduce:
> | > global abc = 1
> | > who
> | > =>
> | > Variables in the current scope:
> | >
> | > abc  ans
> | > ----------------
> | > clear
> | > who
> | > =>
> | > (blank)
> | > ---------------
> | > who global
> | > =>
> | > Global variables:
> | >
> | > abc
> | > ----------------
> | >
> | > The behavior is correct if 'clear -g' is used instead of just 'clear'.
> | >
> | > --Rik
> | >
> |
> | I checked in the following patch:
> | http://hg.savannah.gnu.org/hgweb/octave/rev/3a447ea9e77b
>
> I thnk we still have an incompatibility.  Given the function
>
>  function f ()
>  global x
>  x
>
> I think we should see the following behavior:
>
>  f
>    #=> x = []
>  global x
>  x = 1
>    #=>  x = 1
>  f
>    #=>  x = 1
>  clear x
>  x
>    #=>  x undefined
>  f
>    #=>  x = 1
>  clear functions
>  f
>    #=>  x = 1
>
> (this much is correct now)
>
>  clear global
>  x
>    #=>  x undefined
>  f
>    #=> x = []
>
> but I still see
>
>  x = 1
>
> for the value of x inside F, even after the "clear global" command.
>
> I haven't investigated to see why this is happening, or what the fix
> is, but wanted to make a note of it.
>
> jwe
>

Please try
http://hg.savannah.gnu.org/hgweb/octave/rev/73e36b147e45

cheers

-- 
RNDr. Jaroslav Hajek
computing expert & GNU Octave developer
Aeronautical Research and Test Institute (VZLU)
Prague, Czech Republic
url: www.highegg.matfyz.cz