Thanks for your answer! It is pretty weird because all my tries are wrong.
If I try to index data with this field in the schema : <field name="content" type="text" indexed="true" required="true" stored="true" multiValued="true" /> It works perfectly (in the case of indexing extract from file (via solr cell/tika)), but it fails if I try to index plain text data! Is there something wrong ? And the field will never be able to index plain text data! (In any case, I haven't seen it). Another way to resolve the situation is to have 2 fields : Then, if I use these fields : <field name="content_docs" type="text" indexed="true" required="true" stored="true" multiValued="true" /> <field name="content_ptext" type="text" indexed="true" required="true" stored="true" multiValued="true" /> The first is to index data from extracted files, and the second is to index plain text data. With this kind of schema, I am able to index both values types, without any mistakes/problems. (but both fields have the same features!). Did you have any suggestions in order to help me ? (to summarize, my goal is to have only one field to share several data types) I have a (dummy?) question to ask : can we index data using the requestHandler "/update/extract/" without extracting any documents ? (I assume my evils are there ) Thank you (again!) -- View this message in context: http://lucene.472066.n3.nabble.com/How-to-set-a-common-field-to-several-values-types-tp2922192p2922573.html Sent from the Solr - User mailing list archive at Nabble.com.
