COmanage Developers List

[Marie Huynh <>]

In following cake conventions, I added the lft and rght columns that
are required for the built in tree functions.

https://spaces.internet2.edu/display/COmanage/cm_cous

I moved some code to the model and replaced some loops with some
cakeisms.  Please take a look at the diff file that I've attached.

Thanks,

Marie

On Thu, Sep 29, 2011 at 12:38 PM, Marie Huynh 
<>
 wrote:
> I think these are all the relevant changes I made for co172.
> changes.tar includes the files.  co172 is the diff.
>
> I kept getting missing file errors when doing this:
>
> svn diff --changelist co172
>
> Creating dummy files and adding it to the ignore list didn't help.  So
> I did this instead:
>
> svn st --changelist co172 > test
> tail test -n +3  | grep -v '?' | awk '{ print $2 }' | xargs svn diff
> --diff-cmd diff -x -uw > co172diff
>
> There's gotta be a better way... right?
>
> Marie
>
> On Tue, Sep 13, 2011 at 4:52 AM, Benn Oshrin 
> <>
>  wrote:
>> (bring back to dev list)
>>
>> On 9/13/11 4:36 AM, Marie Huynh wrote:
>>>
>>> I ended up adding in these two:
>>>
>>>     'er.cou.haschild' =>  'COUs with children can not be deleted',
>>>     'er.cou.wrongco' =>  'COUs must be in the same CO',
>>
>> I try to keep these as short as possible, so I'd probably do something like
>>
>>  er.cou.child
>>  er.cou.co or er.cou.sameco
>>
>> and fd.cou.parent, fd.cou.parent.desc, and er.cou.cycle for the other ones
>> you might need...
>>
>>> I made the parent COU field on the edit/add pages a dropdown populated
>>> only with COUs in the same CO (other than itself).  I think it needs a
>>> check for cycles.  What error message would you prefer?  403 COU
>>> Family Tree cycle?  Alternatively, I could also remove COUs from the
>>> dropdown that would cause such a cycle so the user shouldn't be able
>>> to cause the error.
>>
>> Good catch re cycles. The REST error is needed regardless of the dropdown
>> being properly populated, but it seems right to not put it in the dropdown
>> (perhaps also with a text note that you can't select a parent that is a or
>> has any child of the current COU).
>>
>> How about
>>
>>  403 Parent Would Create Cycle
>>
>> ?
>>
>>> I'm still trying to figure out what all needs to be done for item #6
>>> and what all is connected to it.  Whee!
>>
>> Yeah, that one's going to be fun to figure out :) I think it's going to be 
>> a
>> relatively small code diff, though.
>>
>> -Benn-
>>
>

Attachment: co172diff2
Description: Binary data