You are viewing a plain text version of this content. The canonical link for it is here.
Posted to solr-user@lucene.apache.org by Nawab Zada Asad Iqbal <kh...@gmail.com> on 2017/06/05 07:23:35 UTC
Field x is not multivalued and destination for multiple copyFields
*Hi, I have a field 'name_token' which gets value via copyFields on
different language-specific fields (e.g. name_en , name_it, name_es, etc.)
If I can ensure that, *
*only one of these language-specific fields will have a value for a given
document, is it ok to ignore this warning:"IndexSchema Field name_token is
not multivalued and destination for multiple copyFields"*
*Also, what will happen if my a record happens to have values in two
language specific name fields (e.g. if a name or word exists in two
languages: name_zh and name_ja ). *
*My understanding is that the value is same anyways, so there is no
drawback, but can it result in an exception?*
*Regards*
*Nawab*
Re: Field x is not multivalued and destination for multiple
copyFields
Posted by Rick Leir <rl...@leirtech.com>.
Nawab,
If you have multivalued=true and you index a document with two or more
name_* fields, then the name_token will have two or more values, all of
which will be searchable. I think this is what you want, because names
can be quite different in different languages. For example the country names
de Deutschland
en Germany
fr Allemagne
es Alemania
or names for 'shoe'
schuh
zapato
cipela
If you have multivalued=false and try to index two or more languages for
the name, it would cause an exception to be thrown in the indexing
thread and the document will not be indexed (schema validation will fail).
If you use multivalued=false on a field which is a copyfield
destination, this is not a common usage pattern and in my mind (using a
northern idiom) you are "walking on thin ice". Tell us how it goes.
Cheers -- Rick
On 2017-06-05 03:23 AM, Nawab Zada Asad Iqbal wrote:
> *Hi, I have a field 'name_token' which gets value via copyFields on
> different language-specific fields (e.g. name_en , name_it, name_es, etc.)
> If I can ensure that, *
>
>
>
> *only one of these language-specific fields will have a value for a given
> document, is it ok to ignore this warning:"IndexSchema Field name_token is
> not multivalued and destination for multiple copyFields"*
>
>
> *Also, what will happen if my a record happens to have values in two
> language specific name fields (e.g. if a name or word exists in two
> languages: name_zh and name_ja ). *
>
>
>
> *My understanding is that the value is same anyways, so there is no
> drawback, but can it result in an exception?*
>
> *Regards*
>
> *Nawab*
>