From bd7453adbac27db9d4b9501e5292832ce0473521 Mon Sep 17 00:00:00 2001
From: David Huynh <dfhuynh@gmail.com>
Date: Wed, 25 Aug 2010 17:35:16 +0000
Subject: [PATCH] Made sure to strip off charset from content-type when
 importing from URLs before looking up for the right importer.

git-svn-id: http://google-refine.googlecode.com/svn/trunk@1229 7d457c2a-affb-35e4-300a-418c747d4874
---
 .../gridworks/commands/project/CreateProjectCommand.java  | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/main/src/com/google/gridworks/commands/project/CreateProjectCommand.java b/main/src/com/google/gridworks/commands/project/CreateProjectCommand.java
index 6a90e58c6..07cf88545 100644
--- a/main/src/com/google/gridworks/commands/project/CreateProjectCommand.java
+++ b/main/src/com/google/gridworks/commands/project/CreateProjectCommand.java
@@ -372,7 +372,13 @@ public class CreateProjectCommand extends Command {
             }
 
             try {
-                importer = ImporterRegistry.guessImporter(connection.getContentType(), url.getPath());
+                String contentType = connection.getContentType();
+                int semicolon = contentType.indexOf(';');
+                if (semicolon >= 0) {
+                    contentType = contentType.substring(0, semicolon);
+                }
+                
+                importer = ImporterRegistry.guessImporter(contentType, url.getPath());
                 
                 internalInvokeImporter(project, importer, options, inputStream, connection.getContentEncoding());
             } finally {