On Fri, Oct 30, 2009 at 2:04 AM, Avlesh Singh <avl...@gmail.com> wrote:
>>
>> One thing I thought about is if I can define my own
>> DocumentObjectBinder, so I can concatenate my entity names with the
>> IDs in the XML creation.
>>
>> Anyone knows if something like this can be done without modifying
>> Solrj sources? Is there any injection or plugin mecanism for this?
>>
> More details on the use-case please.
If I index a Book with ID=3, and then a Magazine with ID=3, I'll be
really removing my Book3 and indexing Magazine3. I want both entities
to be in the index.
The use case on DocumentObjectBinder is that I could override
toSolrInputDocument, and if field = ID, I could do: setField("id",
obj.getClass().getName() + obj.getId()) or something like that.
The goal is avoiding creating all the XMLs to be sent to Solr but
having the possibility of modifying them in some way.
Do you know how can I do that, or a better way of achieving the same results?
> Cheers
> Avlesh
>
> On Fri, Oct 30, 2009 at 2:16 AM, Christian López Espínola <
> penyask...@gmail.com> wrote:
>
>> Hi Israel,
>>
>> Thanks for your suggestion,
>>
>> On Thu, Oct 29, 2009 at 9:37 PM, Israel Ekpo <israele...@gmail.com> wrote:
>> > On Thu, Oct 29, 2009 at 3:31 PM, Christian López Espínola <
>> > penyask...@gmail.com> wrote:
>> >
>> >> Hi, my name is Christian and I'm a newbie introducing to solr (and
>> solrj).
>> >>
>> >> I'm working on a website where I want to index multiple entities, like
>> >> Book or Magazine.
>> >> The issue I'm facing is both of them have an attribute ID, which I
>> >> want to use as the uniqueKey on my schema, so I cannot identify
>> >> uniquely a document (because ID is saved in a database too, and it's
>> >> autonumeric).
>> >>
>> >> I'm sure that this is a common pattern, but I don't find the way of
>> solving
>> >> it.
>> >>
>> >> How do you usually solve this? Thanks in advance.
>> >>
>> >>
>> >> --
>> >> Cheers,
>> >>
>> >> Christian López Espínola <penyaskito>
>> >>
>> >
>> > Hi Christian,
>> >
>> > It looks like you are bringing in data to Solr from a database where
>> there
>> > are two separate tables.
>> >
>> > One for *Books* and another one for *Magazines*.
>> >
>> > If this is the case, you could define your uniqueKey element in Solr
>> schema
>> > to be a "string" instead of an integer then you can still load documents
>> > from both the books and magazines database tables but your could prefix
>> the
>> > uniqueKey field with "B" for books and "M" for magazines
>> >
>> > Like so :
>> >
>> > <field name="id" type="string" indexed="true" stored="true"
>> > required="true"/>
>> >
>> > <uniqueKey>id</uniqueKey>
>> >
>> > Then when loading the books or magazines into Solr you can create the
>> > documents with id fields like this
>> >
>> > <add>
>> > <doc>
>> > <field name="id">B14000</field>
>> > </doc>
>> > <doc>
>> > <field name="id">M14000</field>
>> > </doc>
>> > <doc>
>> > <field name="id">B14001</field>
>> > </doc>
>> > <doc>
>> > <field name="id">M14001</field>
>> > </doc>
>> > </add>
>> >
>> > I hope this helps
>>
>> This was my first thought, but in practice there isn't Book and
>> Magazine, but about 50 different entities, so I'm using the Field
>> annotation of solrj for simplifying my code (it manages for me the XML
>> creation, etc).
>> One thing I thought about is if I can define my own
>> DocumentObjectBinder, so I can concatenate my entity names with the
>> IDs in the XML creation.
>>
>> Anyone knows if something like this can be done without modifying
>> Solrj sources? Is there any injection or plugin mecanism for this?
>>
>> Thanks in advance.
>>
>>
>> > --
>> > "Good Enough" is not good enough.
>> > To give anything less than your best is to sacrifice the gift.
>> > Quality First. Measure Twice. Cut Once.
>> >
>>
>>
>>
>> --
>> Cheers,
>>
>> Christian López Espínola <penyaskito>
>>
>
--
Cheers,
Christian López Espínola <penyaskito>
