There should be a parameter in the Solr input like 'readAll="true"' to force the Tokenizer to read all attributes.
There should be a parameter in the Solr input like 'readAll="true"' to force the Tokenizer to read all attributes.