Reject or accept matching a set of IP addresses.

2026-09-26 07:35:39 +00:00 · 2012-04-02 18:16:23 -07:00
parent e90fc2965b
commit f9f9e62f4c
2 changed files with 94 additions and 0 deletions
@@ -0,0 +1,7 @@
+*.class
+bin
+commons/target
+dist/target
+engine/target
+modules/target
+hq/target
@@ -0,0 +1,87 @@
+package org.archive.modules.deciderules;
+
+import static org.archive.modules.CoreAttributeConstants.A_DNS_SERVER_IP_LABEL;
+
+import java.net.InetAddress;
+import java.util.Collections;
+import java.util.Set;
+
+import org.archive.modules.CrawlURI;
+import org.archive.modules.net.CrawlHost;
+import org.archive.modules.net.ServerCache;
+import org.springframework.beans.factory.annotation.Autowired;
+
+/**
+ * <pre>
+ * &lt;bean class="org.archive.modules.deciderules.IpAddressSetDecideRule"&gt;
+ *  &lt;property name="ipAddresses"&gt;
+ *   &lt;set&gt;
+ *    &lt;value&gt;127.0.0.1&lt;/value&gt;
+ *    &lt;value&gt;69.89.27.209&lt;/value&gt;
+ *   &lt;/set&gt;
+ *  &lt;/property&gt;
+ *  &lt;property name='decision' value='REJECT' /&gt;
+ * &lt;/bean&gt;
+ * </pre>
+ * 
+ * @contributor Travis Wellman <travis@archive.org>
+ */
+
+class IpAddressSetDecideRule extends PredicatedDecideRule {
+
+//    private static final Logger LOGGER = Logger.getLogger(IpAddressSetDecideRule.class.getCanonicalName());
+    private static final long serialVersionUID = -3670434739183271441L;
+    private Set<String> ipAddresses;
+    
+    /**
+     * @return the addresses being matched
+     */
+    public Set<String> getIpAddresses() {
+        return Collections.unmodifiableSet(ipAddresses);
+    }
+
+    /**
+     * @param ipAddresses the addresses to match
+     */
+    public void setIpAddresses(Set<String> ipAddresses) {
+        this.ipAddresses = ipAddresses;
+    }
+
+    @Override
+    protected boolean evaluate(CrawlURI curi) {
+        String hostAddress = getHostAddress(curi);
+        System.out.println("hostAddress for "+ curi +" "+ hostAddress);
+        return hostAddress != null &&
+                ipAddresses.contains(hostAddress.intern());
+    }
+
+    transient protected ServerCache serverCache;
+    public ServerCache getServerCache() {
+        return this.serverCache;
+    }
+    @Autowired
+    public void setServerCache(ServerCache serverCache) {
+        this.serverCache = serverCache;
+    }
+    
+    /**
+     * from WriterPoolProcessor
+     * 
+     * @param curi CrawlURI
+     * @return String of IP address
+     */
+    protected String getHostAddress(CrawlURI curi) {
+        // special handling for DNS URIs: want address of DNS server
+        if (curi.getUURI().getScheme().toLowerCase().equals("dns")) {
+            return (String)curi.getData().get(A_DNS_SERVER_IP_LABEL);
+        }
+        // otherwise, host referenced in URI
+        // TODO:FIXME: have fetcher insert exact IP contacted into curi,
+        // use that rather than inferred by CrawlHost lookup 
+        CrawlHost crlh = getServerCache().getHostFor(curi.getUURI());
+        if (crlh == null) { return null; }
+        InetAddress inetadd = crlh.getIP();
+        if (inetadd == null) { return null; }
+        return inetadd.getHostAddress();
+    }
+}