[SOLR-9205] Parse schema in LukeResponse

• Type: Improvement
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Fix Version/s: 6.3
• Component/s: SolrJ
• Assignee: Alan Woodward
• Reporter: Fengtan
• Created: 11/Jun/16 02:37
• Updated: 09/Nov/16 08:39
• Resolved: 30/Sep/16 09:59

Description

LukeRequestHandler (/admin/luke) lists schema flags using two fields named "schema" and "flags".

For instance on my local machine http://localhost:8983/solr/collection1/admin/luke returns something like this:

<lst name="id">
  <str name="type">string</str>
  <str name="schema">I-S-----OF-----l</str>
</lst>

And http://localhost:8983/solr/collection1/admin/luke?show=schema returns something like this:

<lst name="id">
  <str name="type">string</str>
  <str name="flags">I-S-----OF-----l</str>
</lst>

However, when processing a LukeRequest in SolrJ, only the "flags" field is parsed into a Set of FieldFlag objects. The "schema" field is left as a String, and as a result is hard to process by client applications who do not know how to parse "I-S----OF----l".

Here is an example that illustrates the problem:

public class MyClass {
  public static void main(String[] args) throws Exception {
    SolrClient client = new HttpSolrClient("http://localhost:8983/solr/collection1");
    LukeRequest request = new LukeRequest();
    LukeResponse response = request.process(client);
    for (LukeResponse.FieldInfo field:response.getFieldInfo().values()) {
      System.out.println(field.getSchema());
      // field.getSchema() returns "I-S-----OF------" (i.e. a String) which is not much meaningful for SolrJ applications
      // Ideally field.getSchema() should return something like "[INDEXED, STORED, OMIT_NORMS, OMIT_TF]" (i.e. a EnumSet<FieldFlag>) which is meaningful for SolrJ applications
    }
  }
}

It is probably fine to parse both fields the same way in SolrJ since LukeRequestHandler populates them the same way.