CEDET

In eieio.el, I see the following comment:

		  ;; TODO:
		  ;;  For other slots (protection, etc) we should get the
		  ;;  original value, and make sure each is equal to the
		  ;;  last value and throw an error, or accept it.

which seems as though I never got around to implementing this section.

If you'd like to take a stab at fixing the part for custom and
customgroup, that'd be ideal.  There is a good example just before
that comment.  Otherwise I'll look into sometime soon.

Thanks
Eric

>>> Pierre Lorenzon <pollock.nageoire@...> seems to think that:
>Hi,
>
>Sorry if this question has already been discussed here but I am
>a newbie in using eieio. Well I read in info documentation :
>"Every slot in parent
>     is replicated in the new class.  If two parents share the same
>     slot name, the parent which appears in the list first sets the
>     attributes for that slot.  If a slot in the new class' slot list
>     matches a parent, then the new specifications for the child class
>     override that of the parent."
>
>Well it does not seem to be the case for the custom attribute
>for instance. Indeed when a read teh code, I saw that 
>init, type (and maybe default) really override settings for
>parent class but not custom.
>
>Why ? Is it something that has to be done in the code or should
>it absolutely not be done if it would disturbe the widget
>definition process ?
>
>Bests
>
>Pierre
>
>
>-------------------------------------------------------------------------
>This SF.net email is sponsored by DB2 Express
>Download DB2 Express C - the FREE version of DB2 express and take
>control of your XML. No limits. Just data. Click to get it now.
>http://sourceforge.net/powerbar/db2/
>_______________________________________________
>cedet-eieio mailing list
>cedet-eieio@...
>https://lists.sourceforge.net/lists/listinfo/cedet-eieio
>

-- 
          Eric Ludlam:                 zappo@..., eric@...
   Home: http://www.ludlam.net            Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: http://www.gnu.org

Hi Eric,




From: "Eric M. Ludlam" <eric@...>
Subject: Re[1]: [cedet-eieio] Slot characteristic inheritance
Date: Wed, 6 Jun 2007 12:06:45 -0400

> In eieio.el, I see the following comment:
> 
> 		  ;; TODO:
> 		  ;;  For other slots (protection, etc) we should get the
> 		  ;;  original value, and make sure each is equal to the
> 		  ;;  last value and throw an error, or accept it.

		   Ok, I did not see thiscomment since I
		   investigated carefully version 0.17 of eieio
		   were this comment was not. Then I installed
		   vers 1.0 in the hope that something would
		   have change : but unfortunately ....  So
		   what is rellay the scope that must be
		   implemented ? If it is only a matter of
		   implementation shurely I can do it ! But so
		   waht ? Should the child custom match the
		   parent custom or is it allowed to completely
		   change custom between parent and children ?
		   I must confess that I would prefer the
		   second solution but I know that I have to
		   respect the package developers will ! 


> 
> which seems as though I never got around to implementing this section.
> 
> If you'd like to take a stab at fixing the part for custom and
> customgroup, that'd be ideal.  There is a good example just before
> that comment.  Otherwise I'll look into sometime soon.
> 
> Thanks
> Eric

  Regards

  Pierre

Hi Pierre,

  In the code section I mentioned below, each attribute of the slot
will need to be looked at to see if should be preserved from the
parent (and an error thrown if an overload attempted), or the new
attribute should apply.

  For example, the :allocation and :protection should not be allowed
to change, but I agree that the customization slot should be
overloadable.  The customize group, however, should probably remain
constant.

  If you would like to try your had at patching EIEIO, that would be
appreciated.  Starting with the version in CVS would likely be best.
Once you've made your change, running it through eieio-test.el is
important to make sure nothing is broken.  If nothing breaks, then it
is likely a new test will need to be added.

Thanks!
Eric 

>>> Pierre Lorenzon <pollock.nageoire@...> seems to think that:
>Hi Eric,
>
>
>
>
>From: "Eric M. Ludlam" <eric@...>
>Subject: Re[1]: [cedet-eieio] Slot characteristic inheritance
>Date: Wed, 6 Jun 2007 12:06:45 -0400
>
>> In eieio.el, I see the following comment:
>> 
>> 		  ;; TODO:
>> 		  ;;  For other slots (protection, etc) we should get the
>> 		  ;;  original value, and make sure each is equal to the
>> 		  ;;  last value and throw an error, or accept it.
>
>		   Ok, I did not see thiscomment since I
>		   investigated carefully version 0.17 of eieio
>		   were this comment was not. Then I installed
>		   vers 1.0 in the hope that something would
>		   have change : but unfortunately ....  So
>		   what is rellay the scope that must be
>		   implemented ? If it is only a matter of
>		   implementation shurely I can do it ! But so
>		   waht ? Should the child custom match the
>		   parent custom or is it allowed to completely
>		   change custom between parent and children ?
>		   I must confess that I would prefer the
>		   second solution but I know that I have to
>		   respect the package developers will ! 
>
>
>> 
>> which seems as though I never got around to implementing this section.
>> 
>> If you'd like to take a stab at fixing the part for custom and
>> customgroup, that'd be ideal.  There is a good example just before
>> that comment.  Otherwise I'll look into sometime soon.
>> 
>> Thanks
>> Eric
>
>  Regards
>
>  Pierre
>
>
>-------------------------------------------------------------------------
>This SF.net email is sponsored by DB2 Express
>Download DB2 Express C - the FREE version of DB2 express and take
>control of your XML. No limits. Just data. Click to get it now.
>http://sourceforge.net/powerbar/db2/
>_______________________________________________
>cedet-eieio mailing list
>cedet-eieio@...
>https://lists.sourceforge.net/lists/listinfo/cedet-eieio
>

-- 
          Eric Ludlam:                 zappo@..., eric@...
   Home: http://www.ludlam.net            Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: http://www.gnu.org

Hi Eric

I'll download the cvs version and try to make a patch.

From: "Eric M. Ludlam" <eric@...>
Subject: Re[2]: [cedet-eieio] Slot characteristic inheritance
Date: Wed, 6 Jun 2007 23:05:10 -0400

> Hi Pierre,
> 
>   In the code section I mentioned below, each attribute of the slot
> will need to be looked at to see if should be preserved from the
> parent (and an error thrown if an overload attempted), or the new
> attribute should apply.

  Ok it depends on the nature of attribute ! 


> 
>   For example, the :allocation and :protection should not be
>   allowed

    Sure ! It concerns the deep structure of the data
    structure. I do not remember if java for instance allows to
    change protection since I have not been devlopping in java
    since many years ! 


> to change, but I agree that the customization slot should be
> overloadable.  The customize group, however, should probably remain
> constant.

  OK for the customization. I think however that customization
  groupss could at least be incremented. When you implement
  child of a class it for more specialized purpose and the
  object will probably be involved in certain groups which are
  absolutely not concerned by the parent class. 

  Regards

  Pierre

>>> Pierre Lorenzon <pollock.nageoire@...> seems to think that:
>
>Hi Eric
>
>I'll download the cvs version and try to make a patch.

Great.  Thanks!

>From: "Eric M. Ludlam" <eric@...>
>Subject: Re[2]: [cedet-eieio] Slot characteristic inheritance
>Date: Wed, 6 Jun 2007 23:05:10 -0400
  [ ... ]
>> to change, but I agree that the customization slot should be
>> overloadable.  The customize group, however, should probably remain
>> constant.
>
>  OK for the customization. I think however that customization
>  groupss could at least be incremented. When you implement
>  child of a class it for more specialized purpose and the
>  object will probably be involved in certain groups which are
>  absolutely not concerned by the parent class. 
  [ ... ]

Good idea.  That would be a nice feature to have.

Eric

-- 
          Eric Ludlam:                 zappo@..., eric@...
   Home: http://www.ludlam.net            Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: http://www.gnu.org

Hi Eric,

Sorry but I did not found were to download the cvs version. I
browsed http://cedet.sourceforge.net/ and found cede-1.4 pre
but no link to the cvs repository. I am blind and at the moment
my braille display is not available so I browse the web only
with a speeh synthesis and it is possible that something escape
me ! 

Regards

Pierre

The CEDET project page is here:

http://sourceforge.net/projects/cedet

The description of the CVS access is here:

http://sourceforge.net/cvs/?group_id=17886

You can use the following unix commands to get a copy the CEDET trunk:

---------
cvs -d:pserver:anonymous@... login
 
cvs -z3 -d:pserver:anonymous@... co -P cedet
---------

Have fun
Eric


>>> Pierre Lorenzon <pollock.nageoire@...> seems to think that:
>Hi Eric,
>
>Sorry but I did not found were to download the cvs version. I
>browsed http://cedet.sourceforge.net/ and found cede-1.4 pre
>but no link to the cvs repository. I am blind and at the moment
>my braille display is not available so I browse the web only
>with a speeh synthesis and it is possible that something escape
>me ! 
>
>Regards
>
>Pierre
>

-- 
          Eric Ludlam:                 zappo@..., eric@...
   Home: http://www.ludlam.net            Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: http://www.gnu.org

Hi Eric,

Sorry but it took a while before I wrote the pieces of code
concerning slot inheritance. Inbetween I start to use ede and I
am very happy with it now ! 

First I think that I am not allowed to commit my changes to the
cvs repository for the eieio project. Hence I simply join the
modified eieio.el file. 

Modifications are between line 733 and 776. 

I ran the tests and they passed. Moreover I going to use
intensively the eieio package, hence if there is a bug
somewhere in the code I wrote, I'll certainly discover it soon
! 

Regards 

Pierre

Hi Again,

In my last modification of eieio.el, I did not implement any
protocol for the doc attribute. It seems clear for me that this
attribute should be overrriddable : Hence I implement it as
overriddable ! I send the modified eieio.el : modifications are
between line 733 and 785 in the join file.

Sure I ran the tests and they passed and used these new
features for my own classes and they produced the expected
result !

Regards

Pierre

Hi,

  I've investigated your changes, and they make sense.  I'm going to
write some tests for the changes and check them in sometime soon.

Thanks!
Eric


>>> From cedet-eieio-bounces@...  Tue Jun 26 08:10:05 2007 seems to think that:
>Hi Eric,
>
>Sorry but it took a while before I wrote the pieces of code
>concerning slot inheritance. Inbetween I start to use ede and I
>am very happy with it now ! 
>
>First I think that I am not allowed to commit my changes to the
>cvs repository for the eieio project. Hence I simply join the
>modified eieio.el file. 
>
>Modifications are between line 733 and 776. 
>
>I ran the tests and they passed. Moreover I going to use
>intensively the eieio package, hence if there is a bug
>somewhere in the code I wrote, I'll certainly discover it soon
>! 
>
>Regards 
>
>Pierre
  [ ... ]

-- 
          Eric Ludlam:                 zappo@..., eric@...
   Home: http://www.ludlam.net            Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: http://www.gnu.org

Hi Pierre,

  I wanted to let you know that in writing tests against your new
code, I've found some problems getting errors out when the protection
is overridden.  The current layout of the code makes the determination
a bit tricky.  The other stuff appears ok so far.

  I'll let you know when I resolve the issue.

Thanks
Eric

>>> Pierre Lorenzon <pollock.nageoire@...> seems to think that:
>----Next_Part(Sat_Jun_30_22_58_10_2007_759)--
>Content-Type: Text/Plain; charset=us-ascii
>Content-Transfer-Encoding: 7bit
>
>Hi Again,
>
>In my last modification of eieio.el, I did not implement any
>protocol for the doc attribute. It seems clear for me that this
>attribute should be overrriddable : Hence I implement it as
>overriddable ! I send the modified eieio.el : modifications are
>between line 733 and 785 in the join file.
>
>Sure I ran the tests and they passed and used these new
>features for my own classes and they produced the expected
>result !
>
>Regards
>
>Pierre
  [ ... ]

-- 
          Eric Ludlam:                 zappo@..., eric@...
   Home: http://www.ludlam.net            Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: http://www.gnu.org

Hi Pierre,

  I finally got part of your changes checked into CEDET/CVS for
eieio.el.  I had written some tests to try out your changes, and I
couldn't get them to pass.  It turns out that the protection (variable
prot) has a value of `nil' when it is public, so it skipped the test
for a protection change.

  It's testing well now, so those changes, plus a new set of tests are
in CVS.

  I still have to write some tests for the class allocated slots.
I'll get there eventually.

Thanks!
Eric


>>> Pierre Lorenzon <pollock.nageoire@...> seems to think that:
>----Next_Part(Tue_Jun_26_13_07_13_2007_720)--
>Content-Type: Text/Plain; charset=us-ascii
>Content-Transfer-Encoding: 7bit
>
>Hi Eric,
>
>Sorry but it took a while before I wrote the pieces of code
>concerning slot inheritance. Inbetween I start to use ede and I
>am very happy with it now ! 
>
>First I think that I am not allowed to commit my changes to the
>cvs repository for the eieio project. Hence I simply join the
>modified eieio.el file. 
>
>Modifications are between line 733 and 776. 
>
>I ran the tests and they passed. Moreover I going to use
>intensively the eieio package, hence if there is a bug
>somewhere in the code I wrote, I'll certainly discover it soon
>! 
>
>Regards 
>
>Pierre
  [ ... ]

-- 
          Eric Ludlam:                 zappo@..., eric@...
   Home: http://www.ludlam.net            Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: http://www.gnu.org