[SOLR-443] POST queries don't declare its charset

• Type: Bug
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 1.2
• Fix Version/s: 1.3
• Component/s: clients - java

Environment

Tomcat 6.0.14

• Assignee: Unassigned
• Reporter: Andrew Schurman
• Created: 21/Dec/07 21:46
• Updated: 10/May/13 14:40
• Resolved: 03/Jul/08 02:24

Description

When sending a query via POST, the content-type is not set. The content charset for the POST parameters are set, but this only appears to be used for creating the Content-Length header in the commons library. Since a query is encoded in UTF-8, the http headers should also specify content type charset.

On Tomcat, this causes problems when the query string contains non-ascii characters (characters with accents and such) as it tries to parse the POST body in its default ISO-9886-1. There appears to be no way to set/change the default encoding for a message body on Tomcat.

Issue Links

• is depended upon by: SOLR-303 Distributed Search over HTTP
• relates to: SOLR-612 solrj should (optionally?) use POST for queries