From 605a88b4ff7c4682c02aff65faab48776aa91cf2 Mon Sep 17 00:00:00 2001
From: Graham Triggs <gtriggs@duraspace.org>
Date: Mon, 4 Jul 2016 19:13:20 +0100
Subject: [PATCH] [VIVO-1237] Add sitemap for profiles

---
 .../webapp/sitemap/SiteMapServlet.java        | 115 ++++++++++++++++++
 webapp/src/main/webapp/WEB-INF/web.xml        |  14 +++
 2 files changed, 129 insertions(+)
 create mode 100644 api/src/main/java/org/vivoweb/webapp/sitemap/SiteMapServlet.java

diff --git a/api/src/main/java/org/vivoweb/webapp/sitemap/SiteMapServlet.java b/api/src/main/java/org/vivoweb/webapp/sitemap/SiteMapServlet.java
new file mode 100644
index 00000000..ac1cb4f4
--- /dev/null
+++ b/api/src/main/java/org/vivoweb/webapp/sitemap/SiteMapServlet.java
@@ -0,0 +1,115 @@
+/* $This file is distributed under the terms of the license in /doc/license.txt$ */
+
+package org.vivoweb.webapp.sitemap;
+
+import com.hp.hpl.jena.query.QuerySolution;
+import edu.cornell.mannlib.vitro.webapp.controller.VitroHttpServlet;
+import edu.cornell.mannlib.vitro.webapp.controller.VitroRequest;
+import edu.cornell.mannlib.vitro.webapp.controller.freemarker.UrlBuilder;
+import edu.cornell.mannlib.vitro.webapp.rdfservice.RDFServiceException;
+import edu.cornell.mannlib.vitro.webapp.rdfservice.ResultSetConsumer;
+import edu.cornell.mannlib.vitro.webapp.visualization.constants.QueryConstants;
+import org.apache.axis.utils.StringUtils;
+import org.apache.commons.io.IOUtils;
+
+import javax.servlet.ServletException;
+import javax.servlet.ServletOutputStream;
+import javax.servlet.http.HttpServletRequest;
+import javax.servlet.http.HttpServletResponse;
+import java.io.FileInputStream;
+import java.io.IOException;
+import java.io.InputStream;
+import java.io.OutputStream;
+
+public class SiteMapServlet extends VitroHttpServlet {
+    @Override
+    protected void doGet(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
+        String requestURI  = request.getRequestURI();
+
+        if (requestURI != null) {
+            if (requestURI.contains("robots.txt")) {
+                String robotsPath = getServletContext().getRealPath("/robots.txt");
+                String contextPath = request.getContextPath();
+
+                StringBuilder builder = new StringBuilder("Sitemap: ");
+                builder.append(getSchemeAndServer(request));
+
+                if (!"/".equals(contextPath)) {
+                    builder.append(contextPath);
+                }
+                builder.append("/sitemap.xml");
+
+                response.getOutputStream().println(builder.toString());
+                response.getOutputStream().println();
+
+                InputStream is = null;
+                try {
+                    is = new FileInputStream(robotsPath);
+                    IOUtils.copy(is, response.getOutputStream());
+                } finally {
+                    if (is != null) {
+                        is.close();
+                    }
+                }
+            } else {
+                final String schemeAndServer = getSchemeAndServer(request);
+
+                String query = QueryConstants.getSparqlPrefixQuery() +
+                        "SELECT ?person\n" +
+                        "WHERE\n" +
+                        "{\n" +
+                        "  ?person a foaf:Person .\n" +
+                        "}\n";
+
+                final VitroRequest vreq = new VitroRequest(request);
+                final ServletOutputStream out = response.getOutputStream();
+
+                out.println("<?xml version=\"1.0\" encoding=\"UTF-8\"?>");
+                out.println("<urlset xmlns=\"http://www.sitemaps.org/schemas/sitemap/0.9\">");
+
+                try {
+                    vreq.getRDFService().sparqlSelectQuery(query, new ResultSetConsumer() {
+                        @Override
+                        protected void processQuerySolution(QuerySolution qs) {
+                            String person = qs.getResource("person").getURI();
+
+                            String profileUrl = UrlBuilder.getIndividualProfileUrl(person, vreq);
+
+                            if (!StringUtils.isEmpty(profileUrl)) {
+                                try {
+                                    out.println("  <url>");
+                                    out.println("    <loc>" + schemeAndServer + profileUrl + "</loc>");
+                                    out.println("    <changefreq>weekly</changefreq>");
+                                    out.println("  </url>");
+                                } catch (Exception e) {
+                                }
+                            }
+                        }
+                    });
+                } catch (RDFServiceException rse) {
+                }
+
+                out.println("</urlset>");
+            }
+        }
+        super.doGet(request, response);
+    }
+
+    private String getSchemeAndServer(HttpServletRequest request) {
+        String serverName  = request.getServerName();
+        int serverPort     = request.getServerPort();
+        String scheme      = request.getScheme();
+
+        if ("http".equalsIgnoreCase(scheme) || "https".equalsIgnoreCase(scheme)) {
+            StringBuilder builder = new StringBuilder();
+            builder.append(scheme).append("://").append(serverName);
+            if (("http".equalsIgnoreCase(scheme) && serverPort != 80) || ("https".equalsIgnoreCase(scheme) && serverPort != 443) ) {
+                builder.append(":").append(serverPort);
+            }
+
+            return builder.toString();
+        }
+
+        return "";
+    }
+}
diff --git a/webapp/src/main/webapp/WEB-INF/web.xml b/webapp/src/main/webapp/WEB-INF/web.xml
index 2f7b50ce..525428a7 100644
--- a/webapp/src/main/webapp/WEB-INF/web.xml
+++ b/webapp/src/main/webapp/WEB-INF/web.xml
@@ -1324,6 +1324,20 @@
     <url-pattern>/selectLocale</url-pattern>
   </servlet-mapping>
 
+  <servlet>
+    <description>SiteMap support</description>
+    <servlet-name>SiteMapServlet</servlet-name>
+    <servlet-class>org.vivoweb.webapp.sitemap.SiteMapServlet</servlet-class>
+  </servlet>
+  <servlet-mapping>
+    <servlet-name>SiteMapServlet</servlet-name>
+    <url-pattern>/robots.txt</url-pattern>
+  </servlet-mapping>
+  <servlet-mapping>
+    <servlet-name>SiteMapServlet</servlet-name>
+    <url-pattern>/sitemap.xml</url-pattern>
+  </servlet-mapping>
+
     
   <!-- ============================== servlet-mappings ======================== -->
    <servlet-mapping>