[SOLR-2217] Odd response format when using extractOnly option with Solr Cell

• Type: Bug
• Status: Closed
• Priority: Minor
• Resolution: Cannot Reproduce
• Affects Version/s: 1.4.1
• Component/s: contrib - Solr Cell (Tika extraction)

Environment

Ubuntu 10.4 LTS (Lucid), Java version "1.6.0_18" OpenJDK Runtime Environment (IcedTea6 1.8.2) (6b18-1.8.2-4ubuntu2) OpenJDK 64-Bit Server VM (build 16.0-b13, mixed mode), Tomcat 6

• Assignee: Unassigned
• Reporter: Donovan Jimenez
• Created: 02/Nov/10 21:44
• Updated: 02/Oct/16 01:43
• Resolved: 02/Oct/16 01:43

Description

When using the extractOnly request parameter, the oas.handler.extraction.ExtractingDocumentLoader is using stream.getName() for parts of the response, but this name appears to be null because the serialized response will return an unnamed string and a list named "null_metadata". It seems more appropriate to use "content" (producing a named string "content" and list "content_metadata") or to use whatever oas.handler.extraction.SolrContentHandler is using for the content field name (coded to "content", but mappable by request parameters).

201 rsp.add(stream.getName(), writer.toString());
202 writer.close();
203 String[] names = metadata.names();
204 NamedList metadataNL = new NamedList();
205 for (int i = 0; i < names.length; i++)

{ 206 String[] vals = metadata.getValues(names[i]); 207 metadataNL.add(names[i], vals); 208 }

209 rsp.add(stream.getName() + "_metadata", metadataNL);

This is mostly to avoid having to use the odd empty string and null_metadata identifiers in unserialized data (like JSON, PHP, RUBY, etc)