Lucene matching non-empty fields.

example.scala

// Skip attachments (images, documents, etc.) - all those files without "_id".
booleanQuery.add (new NotEmptyQuery (new Term ("alt__id")), BooleanClause.Occur.MUST)
// Skip removed objects (_removed = time of removal in milliseconds).
booleanQuery.add (new NotEmptyQuery (new Term ("alt__removed")), BooleanClause.Occur.MUST_NOT)

NotEmptyQuery.scala

import org.apache.lucene.search._, org.apache.lucene.document._, org.apache.lucene.index._

class NotEmptyQuery (emptyTerm: Term) extends MultiTermQuery {
  override def toString (field: String) = "NotEmptyQuery(" + (if ((field ne null) && field.length != 0) field else emptyTerm.field) + ")"
  override def getEnum (reader: IndexReader) = new FilteredTermEnum {
    setEnum (reader.terms (emptyTerm)) // Position at the start of the field terms.
    var theEnd = false
    override def termCompare (term: Term): Boolean =
      if (term.field == emptyTerm.field && term.text != emptyTerm.text) true else {theEnd = true; false}
    override def difference (): Float = 1
    override def endEnum (): Boolean = theEnd
  }
}