VIVO-1008 Refine the mime type: if no extension, then HTML

When behind Apache, some HTML pages were coming out as text/plain

webapp/src/edu/cornell/mannlib/vitro/webapp/filters/CharsetEncodingFilter.java

@@ -3,11 +3,11 @@
 package edu.cornell.mannlib.vitro.webapp.filters;
 
 import java.io.IOException;
+import java.io.UnsupportedEncodingException;
 
 import javax.servlet.Filter;
 import javax.servlet.FilterChain;
 import javax.servlet.FilterConfig;
-import javax.servlet.ServletContext;
 import javax.servlet.ServletException;
 import javax.servlet.ServletRequest;
 import javax.servlet.ServletResponse;
@@ -16,6 +16,7 @@ import javax.servlet.http.HttpServletRequest;
 import org.apache.commons.fileupload.servlet.ServletFileUpload;
 
 public class CharsetEncodingFilter implements Filter {
+
 	@Override
 	public void init(FilterConfig filterConfig) throws ServletException {
 		// Nothing to set up
@@ -26,25 +27,51 @@ public class CharsetEncodingFilter implements Filter {
 			ServletResponse servletResponse, FilterChain filterChain)
 			throws IOException, ServletException {
 		HttpServletRequest req = (HttpServletRequest) servletRequest;
-		ServletContext ctx = req.getSession().getServletContext();
-		String url = req.getRequestURL().toString();
 
-		if (!ServletFileUpload.isMultipartContent(req)) {
-			servletRequest.setCharacterEncoding("UTF-8");
-		}
-
-		if (req.getContentType() == null) {
-			String typeFromContext = ctx.getMimeType(url);
-			if (typeFromContext == null) {
-				servletResponse.setContentType("text/html;charset=UTF-8");
-			} else {
-				servletResponse.setContentType(typeFromContext);
-			}
-		}
+		setCharacterEncodingOnRequest(req);
+		setContentTypeOnResponse(req, servletResponse);
 
 		filterChain.doFilter(servletRequest, servletResponse);
 	}
 
+	private void setCharacterEncodingOnRequest(HttpServletRequest req)
+			throws UnsupportedEncodingException {
+		if (!ServletFileUpload.isMultipartContent(req)) {
+			req.setCharacterEncoding("UTF-8");
+		}
+	}
+
+	private void setContentTypeOnResponse(HttpServletRequest req,
+			ServletResponse servletResponse) {
+		if (servletResponse.getContentType() != null) {
+			return;
+		}
+		
+		String uri = req.getRequestURI();
+		if (!hasExtension(uri)) {
+			servletResponse.setContentType("text/html;charset=UTF-8");
+			return;
+		}
+
+		String typeFromContext = req.getSession().getServletContext()
+				.getMimeType(uri);
+		if (typeFromContext == null) {
+			servletResponse.setContentType("text/html;charset=UTF-8");
+			return;
+		}
+
+		servletResponse.setContentType(typeFromContext);
+		servletResponse.setCharacterEncoding("UTF-8");
+	}
+
+	private boolean hasExtension(String uri) {
+		return uri.matches(".+" // some stuff
+				+ "\\." // a literal period
+				+ "[^/]+" // some stuff that's not a slash
+				+ "$" // the end of the URI
+		);
+	}
+
 	@Override
 	public void destroy() {
 		// Nothing to tear down