Think that "sort=False" does the trick!!

Thanks for help, I think you can close the issue... Do you want I leave the
solution with the issue?

Richard

On Thu, May 3, 2012 at 10:02 AM, Richard Vézina <[email protected]
> wrote:

> Ok, I remove the IS_IN_DB validator and now I get unordered list... So,
> look like it the problem...
>
> I will try to change it...
>
> Richard
>
>
> On Thu, May 3, 2012 at 9:53 AM, Richard Vézina <
> [email protected]> wrote:
>
>> I use IS_IN_DB with a set not the table directly...
>>
>> analyst_set = db((db.auth_group.role=='analyst') & 
>> (db.auth_user.id==db.auth_membership.user_id)
>> & (db.auth_membership.group_id==db.auth_group.id))
>>
>> requires=IS_IN_DB(analyst_set,'auth_user.id','%(initials)s',\
>>             multiple=True,cache=(cache.ram, 3600)),
>>
>> So, maybe it is the IS_IN_DB validator that order the list at the end of
>> the process before returning the list to the DAL.
>>
>> Richard
>>
>> On Wed, May 2, 2012 at 11:12 PM, Massimo Di Pierro <
>> [email protected]> wrote:
>>
>>> Are you talking about the IS_IN_SET() validator? I need to see some
>>> code. There is nothing in DAL that cases this. It must be a validator. So
>>> it depends on what you are using.
>>>
>>>
>>> On Wednesday, 2 May 2012 17:29:35 UTC-5, Richard wrote:
>>>>
>>>> Ok DAL do will return [1,2,3] in case of the user has select the first
>>>> 3 element of the referenced table.
>>>>
>>>> What Choosen plugin return (and what I would to be stored in database
>>>> field) could be [1,3,2] or [3,1,2] or what ever other order in which the
>>>> user had pick those referenced elements. Choosen allow that when use
>>>> multiple option.
>>>>
>>>> So I would like that the DAL return at the database level the order in
>>>> which the element have been entered in the field in case of use of plugin
>>>> like choosen. Off course it is not applying if someone is not using a
>>>> plugin because you can't know in which order the element have been 
>>>> selected.
>>>>
>>>> Now, I think that what prevent the DAL to return a unordered list to
>>>> the database is the use of dict...
>>>>
>>>> I think I found a work around... But I didn't have time to implement
>>>> it... It consist of creating an other field to store the content of the
>>>> form.vars.field (list:reference field). So having 2 fields instead of 1...
>>>>
>>>> Hope it more clear.
>>>>
>>>> Richard
>>>>
>>>> On Wed, May 2, 2012 at 4:52 PM, Massimo Di Pierro <
>>>> [email protected]> wrote:
>>>>
>>>>> Richard, I am at a loss about this. the dal does not sort, nor does
>>>>> the validator. Can you show me and example to reproduce this sorting?
>>>>>
>>>>>
>>>>> On Thursday, 19 April 2012 16:14:50 UTC-5, Richard wrote:
>>>>>>
>>>>>>   Issue 768 <http://code.google.com/p/web2py/issues/detail?id=768>: Allow
>>>>>> insert of unsorted or unordered list:reference type field values
>>>>>> On Thu, Apr 19, 2012 at 5:08 PM, Richard Vézina <
>>>>>> [email protected]> wrote:
>>>>>>
>>>>>>> Found this http://stackoverflow.com/****questions/1867861/python-**
>>>>>>> dicti**onary-keep-keys-values-**in-**same-order-as-declared<http://stackoverflow.com/questions/1867861/python-dictionary-keep-keys-values-in-same-order-as-declared>
>>>>>>>  on
>>>>>>> stackoverflow where the most rated post redirect here :
>>>>>>> http://dev.pocoo.org/hg/**sandbo**x/raw-file/tip/odict.py<http://dev.pocoo.org/hg/sandbox/raw-file/tip/odict.py>
>>>>>>>
>>>>>>> Notice, it will download odict.py into your computer that contain a
>>>>>>> class supposing to allow preservation of initial order in a python 
>>>>>>> dict...
>>>>>>>
>>>>>>> I think, it could be nice to have this class added to web2py in
>>>>>>> order to allow insert into database of unordered list:reference field
>>>>>>> input...
>>>>>>>
>>>>>>> Since this could be slow (not in case of few value I think), it
>>>>>>> could be a choice offer to the user like : list:reference sorted or
>>>>>>> list:reference unsorted...
>>>>>>>
>>>>>>> What do you think??
>>>>>>>
>>>>>>> Richard
>>>>>>>
>>>>>>>
>>>>>>> On Thu, Apr 19, 2012 at 4:59 PM, Richard Vézina <
>>>>>>> [email protected]> wrote:
>>>>>>>
>>>>>>>> Some progress...
>>>>>>>>
>>>>>>>> I just make a check and chosen plugin multiple select option return
>>>>>>>> a unordered list of the picked element by the user...
>>>>>>>>
>>>>>>>> So, now, I think that it could be nice if web2py allow some way to
>>>>>>>> determine in case of use of chosen kind of plugin to store a unordered 
>>>>>>>> list
>>>>>>>> in list:reference field...
>>>>>>>>
>>>>>>>> I think that it is near line 207 (web2py 1.99.4) gluon/sqlhtml that
>>>>>>>> we have to make change :
>>>>>>>>
>>>>>>>> class OptionsWidget(FormWidget):
>>>>>>>>
>>>>>>>>     @staticmethod
>>>>>>>>     def has_options(field):
>>>>>>>>         """
>>>>>>>>         checks if the field has selectable options
>>>>>>>>
>>>>>>>>         :param field: the field needing checking
>>>>>>>>         :returns: True if the field has options
>>>>>>>>         """
>>>>>>>>
>>>>>>>>         return hasattr(field.requires, 'options')
>>>>>>>>
>>>>>>>>     @classmethod
>>>>>>>>     def widget(cls, field, value, **attributes):
>>>>>>>>         """
>>>>>>>>         generates a SELECT tag, including OPTIONs (only 1 option
>>>>>>>> allowed)
>>>>>>>>
>>>>>>>>         see also: :meth:`FormWidget.widget`
>>>>>>>>         """
>>>>>>>>         default = dict(value=value) # HERE!!!!!!!!
>>>>>>>>
>>>>>>>> I am not pretty sure actually... And I don't really know, what
>>>>>>>> could be done to keep submit multiple list order in case of plugin 
>>>>>>>> use...
>>>>>>>>
>>>>>>>> I try naively to replace
>>>>>>>>
>>>>>>>> default = dict(value=value)
>>>>>>>>
>>>>>>>> By
>>>>>>>>
>>>>>>>> defautl = value
>>>>>>>>
>>>>>>>> But it failed.
>>>>>>>>
>>>>>>>> Richard
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Thu, Apr 19, 2012 at 3:43 PM, Richard Vézina <
>>>>>>>> [email protected]> wrote:
>>>>>>>>
>>>>>>>>> This is a part of the solution I think :
>>>>>>>>>
>>>>>>>>> http://www.ryancramer.com/**jour**nal/entries/select_**multiple/<http://www.ryancramer.com/journal/entries/select_multiple/>
>>>>>>>>>
>>>>>>>>> Richard
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On Thu, Apr 19, 2012 at 3:08 PM, Richard <
>>>>>>>>> [email protected]> wrote:
>>>>>>>>>
>>>>>>>>>> Hello,
>>>>>>>>>>
>>>>>>>>>> I would know if there is a way to tell to list:reference to not
>>>>>>>>>> order the id of the reference table to keep the order in wich the 
>>>>>>>>>> user has
>>>>>>>>>> pick those element... At the same time I just thought that this 
>>>>>>>>>> concept is
>>>>>>>>>> useless except in case you use some javascript plugin like Chosen 
>>>>>>>>>> Multiple
>>>>>>>>>> Select 
>>>>>>>>>> (http://harvesthq.github.com/**c**hosen/<http://harvesthq.github.com/chosen/>)
>>>>>>>>>> because, you pick the element in a list and there is no way to know 
>>>>>>>>>> which
>>>>>>>>>> element in the list the user has selected first.
>>>>>>>>>>
>>>>>>>>>> My need is to allow users to select element from an other table
>>>>>>>>>> and keep a more informative those selected thing. Like the 
>>>>>>>>>> ingredient on
>>>>>>>>>> the back panel of food box, the first ingredient more abondant than 
>>>>>>>>>> the
>>>>>>>>>> second an so on.
>>>>>>>>>>
>>>>>>>>>> So, for now I think I am stock with a text field. I would avoid a
>>>>>>>>>> free text field because I would prefer user to always use the same 
>>>>>>>>>> name for
>>>>>>>>>> the same thing, but I can't do it with text type field. What I 
>>>>>>>>>> thought is a
>>>>>>>>>> list:reference field and a button beside the field to allow allow 
>>>>>>>>>> user to
>>>>>>>>>> add missing element name in the foreign table if need without leave 
>>>>>>>>>> the
>>>>>>>>>> original form. This can be done easily with the select or add web2py 
>>>>>>>>>> slice
>>>>>>>>>> : http://www.web2pyslices.com/****slice/show/1446/widget-select-*
>>>>>>>>>> ***or-add-option<http://www.web2pyslices.com/slice/show/1446/widget-select-or-add-option>.
>>>>>>>>>> Thanks to the author.
>>>>>>>>>>
>>>>>>>>>> But, I think that keeping the order of the selected element can't
>>>>>>>>>> be done??
>>>>>>>>>>
>>>>>>>>>> Thanks for help.
>>>>>>>>>>
>>>>>>>>>> Richard
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>
>>>>
>>
>

Reply via email to