[TYPO3-mvc] Conceptual Question with RELATION_HAS_ONE

[Jochen Rau]

Hi Jeff.

I am still working on that. But first I have to finish the CRUD 
operations of the new persistence layer.

Stay tuned ;-)

Jochen

Jeff Segars wrote:
> Jochen Rau wrote:
>> Hi Jeff.
>>
>>>>> Inside the extbase test cases,
>>>>> Tx_Extbase_Persistence_Mapper_DataMap_testcase contains a TCA
>>>>> definition for a blog author that references an author table rather
>>>>> than just being an input field as it is in blog_example. Since it's
>>>>> already in the test case, seems like a good example for my question :)
>>>>>
>>>>>> 'author' => array(
>>>>>> 'exclude' => 1,
>>>>>> 'label' =>
>>>>>> 'LLL:EXT:blog_example/Resources/Language/locallang_db.xml:tx_blogexample_domain_model_blog.author',
>>>>>>
>>>>>>
>>>>>> 'config' => array(
>>>>>> 'type' => 'select',
>>>>>> 'foreign_class' => 'Tx_BlogExample_Domain_Model_Author',
>>>>>> 'foreign_table' => 'tx_blogexample_domain_model_author',
>>>>>> 'maxitems' => 1,
>>>>>> )
>>>>>> ),
>>>>>
>>>>> From what I can see of the persistence framework (both in trunk and
>>>>> in persistence rewrite) it expects that this simple relation also
>>>>> provides a foreign_field and foreign_table field so that the relation
>>>>> is saved in the child table rather than the parent table.
>>>>>
>>>>> This of course doesn't match the blog model that would allow the same
>>>>> author to be assigned multiple blogs and doesn't work with the
>>>>> backend select type that saves a UID directly in the parent table.
>>
>> You are right. This is indeed a misconception inside Extbase. The
>> whole process from $TCA to the Data Map to the Domain Object is on my
>> list. By now I concentrated on the Data Map to Domain Object part.
>
> Thanks for the clarification. I was mainly concerned that I was missing
> something really obvious :)
>
>>>>> It seems like I must be missing something obvious here so if anyone
>>>>> can enlighten me, it would be greatly appreciated ... and if the
>>>>> answer is "I'm busy...wait for the persistence rewrite to get done
>>>>> and it will all work" that's perfectly fine too ;)
>>
>>> Hopefully, we'll get some word on the *right* way to do this soon.
>>
>> Yes. Definitely. I am going extend the Blog Example with an Author
>> Class this evening to test different concepts it in "real life". I
>> will give you feedback asap.
>
> Sound good. Once the code is in subversion I'll start testing
> immediately. If I can be of any help before then, just let me know.
>
> Thanks again for all your work here!
> Jeff

-- 
Every nit picked is a bug fixed