Hi Shawn, the DIH is doing the splitting:
... <field column="id" xpath="/documents/document/@id" /> <field column="dccreator_sort" xpath="/documents/document/element[@name='dccreator']/value" /> <field column="dccreator" splitBy="; " sourceColName="dccreator_sort" /> ... Bernd Am 18.02.2016 um 14:42 schrieb Shawn Heisey: > On 2/18/2016 3:45 AM, Bernd Fehling wrote: >> Now this is strange with solr 4.10.4, >> I have a multivalue string field for creator. >> <field name="creator" type="string" indexed="false" stored="true" >> multiValued="true" /> >> >> And a multivalue string field for f_person, prepared for facetting with >> docValues. >> <field name="f_person" type="string" indexed="true" stored="true" >> multiValued="true" docValues="true" /> >> >> To fill f_person I use copyField. >> <copyField source="creator" dest="f_person" /> >> >> The input to creator is 43470 bytes long with names, split at ";" for each >> subfield. >> Klionsky, Daniel J; JFA; CORA; Abdelmohsen, Kotb; Abe, Akihisa; ... > > How are you handling splitting that information into multiple pieces? > If it's done with analysis configuration in schema.xml, then the data > copied to f_person is *not* split into multiple values. The copyField > functionality always copies the the original input data -- *before* > analysis. > > If the information were split into multiple small values before it got > to Solr, then this error would not be happening. > > Thanks, > Shawn >
