Sitemap extension support

- add extension support to SiteMapTester
- list extension attributes in SiteMapURL.toString()
- update change log

CHANGES.txt

@@ -1,6 +1,7 @@
 Crawler-Commons Change Log
 
 Current Development 0.11-SNAPSHOT (yyyy-mm-dd)
+- [Sitemaps] Add support for sitemap extensions (tuxnco, sebastian-nagel) #35, #36, #149, #162
 - [Sitemaps] Use the Java 8 date and time API (java.time.*) to parse dates in sitemaps (sebastian-nagel) #217
 - [Robots] Fix for handling URLs with query parameters but no path (kkrugler) #215
 

src/main/java/crawlercommons/sitemaps/SiteMapTester.java

@@ -44,6 +44,8 @@ public class SiteMapTester {
             LOG.error("Java properties:");
             LOG.error("  sitemap.strictNamespace");
             LOG.error("                  if true sitemaps are required to use the standard namespace URI");
+            LOG.error("  sitemap.extensions");
+            LOG.error("                  if true enable sitemap extension parsing");
         } else {
             URL url = new URL(args[0]);
             String mt = (args.length > 1) ? args[1] : null;
@@ -64,6 +66,11 @@ public class SiteMapTester {
         boolean strictNamespace = new Boolean(System.getProperty("sitemap.strictNamespace"));
         saxParser.setStrictNamespace(strictNamespace);
 
+        boolean enableExtensions = new Boolean(System.getProperty("sitemap.extensions"));
+        if (enableExtensions) {
+            saxParser.enableExtensions();
+        }
+
         AbstractSiteMap sm = null;
         // guesses the mimetype
         if (mt == null || mt.equals("")) {
@@ -80,7 +87,11 @@ public class SiteMapTester {
         } else {
             Collection<SiteMapURL> links = ((SiteMap) sm).getSiteMapUrls();
             for (SiteMapURL smu : links) {
-                LOG.info(smu.getUrl().toString());
+                if (enableExtensions) {
+                    LOG.info(smu.toString());
+                } else {
+                    LOG.info(smu.getUrl().toString());
+                }
             }
         }
     }

src/main/java/crawlercommons/sitemaps/SiteMapURL.java

@@ -25,6 +25,7 @@ import java.time.ZonedDateTime;
 import java.util.Date;
 import java.util.Locale;
 import java.util.Map;
+import java.util.Map.Entry;
 import java.util.TreeMap;
 
 import crawlercommons.sitemaps.extension.Extension;
@@ -371,6 +372,13 @@ public class SiteMapURL {
         sb.append(", lastMod = ").append((lastModified == null) ? "null" : SiteMap.W3C_FULLDATE_FORMATTER_UTC.format(lastModified.toInstant()));
         sb.append(", changeFreq = ").append(changeFreq);
         sb.append(", priority = ").append(priority);
+        if (attributes != null) {
+            for (Entry<Extension, ExtensionMetadata[]> e : attributes.entrySet()) {
+                for (ExtensionMetadata m : e.getValue()) {
+                    sb.append(", ").append(m.toString());
+                }
+            }
+        }
 
         return sb.toString();
     }