Apache Jackrabbit Oak 1.8索引 - Lucene不会在聚合节点中索引二进制属性

Question

我有以下索引：

oak:index
      jcr:primaryType = nt:unstructured
      dms-lucene-fulltext-index
         compatVersion = 2
         async = async
         jcr:primaryType = oak:QueryIndexDefinition
         evaluatePathRestrictions = true
         type = lucene
         tags = fulltext
         aggregates
            jcr:primaryType = nt:unstructured
            nt:file
               jcr:primaryType = nt:unstructured
               include0
                  path = jcr:content
                  jcr:primaryType = nt:unstructured

我在文件夹中有以下文件节点：

folder
   jcr:created = 2018-02-24T14:32:09.550+01:00
   jcr:createdBy = 
   jcr:primaryType = nt:folder
   jcr:uuid = 5c3e4689-84e9-4e34-8b14-029f62172812
   test.txt
      jcr:created = 2018-02-24T14:32:09.674+01:00
      jcr:createdBy = 14
      jcr:primaryType = nt:file
      jcr:content
         jcr:encoding = utf-8
         jcr:lastModifiedBy = 14
         jcr:mimeType = text/plain; charset=utf-8
         jcr:data = the quick brown fox
         jcr:lastModified = 2018-02-24T14:32:09.673+01:00
         jcr:primaryType = nt:resource
         jcr:uuid = 52f224e8-db57-4879-9d6a-94862f65fb8d

如果我执行以下查询，我会得到该文件：

SELECT * FROM [nt:file] WHERE ISDESCENDANTNODE('/folder') AND CONTAINS(*,'plain')

所以mimeType在索引中。但二进制不，导致后续查询没有结果：

SELECT * FROM [nt:file] WHERE ISDESCENDANTNODE('/folder') AND CONTAINS(*,'fox')

我希望有人能告诉我这里我做错了什么，谢谢！