*) Adding better https support for crawler

- solving problems with unkown certificates by implementing a dummy trust Manager
   - adding https support to robots-parser 
   - Seed File can now be downloaded from https resources
   - adapting plasmaHTCache.java to support https URLs properly

*) URL Normalization
   - sub URLs are now normalized properly during indexing
   - pointing urlNormalForm function of plasmaParser to htmlFilterContentScraper function
   - normalizing URLs which were received by a crawlOrder request

git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@1024 6c8d7289-2bf4-0310-a012-ef5d649a1542

htroot/yacy/crawlOrder.java

@@ -48,7 +48,9 @@ import java.util.Date;
 
 import de.anomic.http.httpHeader;
 import de.anomic.plasma.plasmaCrawlLURL;
+import de.anomic.plasma.plasmaParser;
 import de.anomic.plasma.plasmaSwitchboard;
+import de.anomic.plasma.plasmaURL;
 import de.anomic.server.serverObjects;
 import de.anomic.server.serverSwitch;
 import de.anomic.tools.crypt;
@@ -166,7 +168,19 @@ public final class crawlOrder {
                 int count = Math.min(urlv.size(), refv.size());
                 if (count == 1) {
                     // old method: only one url
-                    stackresult = stack(switchboard, (String) urlv.get(0), (String) refv.get(0), iam, youare);
+
+                    // normalizing URL
+                    String newURL = plasmaParser.urlNormalform((String)urlv.get(0));
+                    if (!newURL.equals(urlv.get(0))) {
+                        env.getLog().logWarning("crawlOrder: Received not normalized URL " + urlv.get(0));    
+                    }
+                    String refURL = plasmaParser.urlNormalform((String) refv.get(0));
+                    if ((refURL != null) && (!refURL.equals(refv.get(0)))) {
+                        env.getLog().logWarning("crawlOrder: Received not normalized Referer URL " + refv.get(0) + " of URL " + urlv.get(0));    
+                    }
+                    
+                    // adding URL to noticeURL Queue
+                    stackresult = stack(switchboard, newURL, refURL, iam, youare);
                     response = (String) stackresult[0];
                     reason = (String) stackresult[1];
                     lurl = (String) stackresult[2];

source/de/anomic/data/robotsParser.java

@@ -141,28 +141,21 @@ public final class robotsParser{
         return deny;
     }        
     
-    public static boolean containsRobotsData(URL nexturl) {
-        // generating the hostname:poart string needed to do a DB lookup
-        String urlHostPort = nexturl.getHost() + ":" + ((nexturl.getPort()==-1)?80:nexturl.getPort());
-        urlHostPort = urlHostPort.toLowerCase();
-        
-        // doing a DB lookup to determine if the robots data is already available
-        plasmaCrawlRobotsTxt.Entry robotsTxt4Host = plasmaSwitchboard.robots.getEntry(urlHostPort);
-        
-        // if we have not found any data or the data is older than 7 days, we need to load it from the remote server
-        if ((robotsTxt4Host == null) || (robotsTxt4Host.getLoadedDate() == null) ||
-                (System.currentTimeMillis() - robotsTxt4Host.getLoadedDate().getTime() > 7*24*60*60*1000)) {
-            return false;
-        }
-        return true;
-    }
-    
-    
     public static boolean isDisallowed(URL nexturl) {
         if (nexturl == null) throw new IllegalArgumentException();               
         
         // generating the hostname:poart string needed to do a DB lookup
-        String urlHostPort = nexturl.getHost() + ":" + ((nexturl.getPort()==-1)?80:nexturl.getPort());
+        String urlHostPort = null;
+        int port = nexturl.getPort();
+        if (port == -1) {
+            if (nexturl.getProtocol().equalsIgnoreCase("http")) {
+                port = 80;
+            } else if (nexturl.getProtocol().equalsIgnoreCase("https")) {
+                port = 443;
+            }
+            
+        }
+        urlHostPort = nexturl.getHost() + ":" + port;
         urlHostPort = urlHostPort.toLowerCase().intern();
         
         plasmaCrawlRobotsTxt.Entry robotsTxt4Host = null;
@@ -179,7 +172,7 @@ public final class robotsParser{
                 URL robotsURL = null;
                 // generating the proper url to download the robots txt
                 try {                 
-                    robotsURL = new URL(nexturl.getProtocol(),nexturl.getHost(),(nexturl.getPort()==-1)?80:nexturl.getPort(),"/robots.txt");
+                    robotsURL = new URL(nexturl.getProtocol(),nexturl.getHost(),port,"/robots.txt");
                 } catch (MalformedURLException e) {
                     serverLog.logSevere("ROBOTS","Unable to generate robots.txt URL for URL '" + nexturl.toString() + "'.");
                     return false;
@@ -249,9 +242,9 @@ public final class robotsParser{
             plasmaSwitchboard sb = plasmaSwitchboard.getSwitchboard();
             //TODO: adding Traffic statistic for robots download?
             if ((sb.remoteProxyConfig == null) || (!sb.remoteProxyConfig.useProxy())) {
-                con = httpc.getInstance(robotsURL.getHost(), robotsURL.getPort(), 10000, false);
+                con = httpc.getInstance(robotsURL.getHost(), robotsURL.getPort(), 10000, robotsURL.getProtocol().equalsIgnoreCase("https"));
             } else {
-                con = httpc.getInstance(robotsURL.getHost(), robotsURL.getPort(), 10000, false, sb.remoteProxyConfig);
+                con = httpc.getInstance(robotsURL.getHost(), robotsURL.getPort(), 10000, robotsURL.getProtocol().equalsIgnoreCase("https"), sb.remoteProxyConfig);
             }
             
             // if we previously have downloaded this robots.txt then we can set the if-modified-since header

source/de/anomic/htmlFilter/htmlFilterContentScraper.java

@@ -110,10 +110,25 @@ public class htmlFilterContentScraper extends htmlFilterAbstractScraper implemen
     public static String urlNormalform(String us) {
         if (us == null) return null;
         if (us.length() == 0) return null;
+        
+        /* TODO: what about 
+         * - case insensitive domain names
+         * - chars that should be escaped in URLs
+         */
         int p;
+        
+        // cutting of everything behind #
         if ((p = us.indexOf("#")) >= 0) us = us.substring(0, p);
-        if (us.endsWith(":80")) us = us.substring(0, us.length() - 3);
-        if ((p = us.indexOf(":80/")) >= 0) us = us.substring(0,p).concat(us.substring(p + 3));
+        
+        if (us.startsWith("https")) {
+            if (us.endsWith(":443")) us = us.substring(0, us.length() - 4);
+            p = us.indexOf(":443/");
+            if (p >= 0) us = us.substring(0,p).concat(us.substring(p + 4));            
+        } else if (us.startsWith("http")) {
+            if (us.endsWith(":80")) us = us.substring(0, us.length() - 3);
+            p = us.indexOf(":80/");
+            if (p >= 0) us = us.substring(0,p).concat(us.substring(p + 3));
+        } 
         if (((us.endsWith("/")) && (us.lastIndexOf('/', us.length() - 2) < 8))) us = us.substring(0, us.length() - 1);
         return us;
     }        

source/de/anomic/http/httpc.java

@@ -65,7 +65,12 @@ import java.util.TimeZone;
 import java.util.zip.GZIPInputStream;
 import java.util.zip.GZIPOutputStream;
 
+import javax.net.ssl.HostnameVerifier;
+import javax.net.ssl.HttpsURLConnection;
+import javax.net.ssl.SSLContext;
 import javax.net.ssl.SSLSocketFactory;
+import javax.net.ssl.TrustManager;
+import javax.net.ssl.X509TrustManager;
 
 import de.anomic.server.serverByteBuffer;
 import de.anomic.server.serverCodings;
@@ -176,6 +181,43 @@ public final class httpc {
         theHttpcPool = new httpcPool(new httpcFactory(),config);
     }
     
+    // initializing a dummy trustManager to enable https connections
+    static SSLSocketFactory theSSLSockFactory = null;
+    static {
+        // Create a trust manager that does not validate certificate chains
+        TrustManager[] trustAllCerts = new TrustManager[] { new X509TrustManager() {
+            public java.security.cert.X509Certificate[] getAcceptedIssuers() {
+                return null;
+            }
+ 
+            public void checkClientTrusted(
+                    java.security.cert.X509Certificate[] certs, String authType) {
+            }
+ 
+            public void checkServerTrusted(
+                    java.security.cert.X509Certificate[] certs, String authType) {
+            }
+        } };
+ 
+        // Install the all-trusting trust manager
+        try {
+            SSLContext sc = SSLContext.getInstance("SSL");
+            // Create empty HostnameVerifier
+            HostnameVerifier hv = new HostnameVerifier() {
+                public boolean verify(String urlHostName, javax.net.ssl.SSLSession session) {
+                    // logger.info("Warning: URL Host: "+urlHostName+"
+                    // vs."+session.getPeerHost());
+                    return true;
+                }
+            };
+ 
+            sc.init(null, trustAllCerts, new java.security.SecureRandom());
+            HttpsURLConnection.setDefaultSSLSocketFactory(theSSLSockFactory = sc.getSocketFactory());
+            HttpsURLConnection.setDefaultHostnameVerifier(hv);
+        } catch (Exception e) {
+        }        
+    }
+
     /**
      * A reusable readline buffer
      * @see serverByteBuffer
@@ -493,8 +535,9 @@ public final class httpc {
             }
 
             // creating a socket
-            this.socket = (ssl) ? SSLSocketFactory.getDefault().createSocket()
-                                : new Socket();
+            this.socket = (ssl) 
+                        ? theSSLSockFactory.createSocket()
+                        : new Socket();
             
             // creating a socket address
             InetSocketAddress address = new InetSocketAddress(hostip, port);
@@ -700,7 +743,7 @@ public final class httpc {
 
         // send request
         if ((this.remoteProxyUse) && (!(method.equals(httpHeader.METHOD_CONNECT))))
-            path = "http://" + this.savedRemoteHost + path;
+            path = (this.savedRemoteHost.endsWith("443")?"https://":"http://") + this.savedRemoteHost + path;
         serverCore.send(this.clientOutput, method + " " + path + " HTTP/1.0"); // if set to HTTP/1.1, servers give time-outs?
 
         // send header

source/de/anomic/http/httpdFileHandler.java

@@ -296,7 +296,6 @@ public final class httpdFileHandler extends httpdAbstractHandler implements http
         String path       = conProp.getProperty(httpHeader.CONNECTION_PROP_PATH);
         String argsString = conProp.getProperty(httpHeader.CONNECTION_PROP_ARGS); // is null if no args were given
         String httpVersion= conProp.getProperty(httpHeader.CONNECTION_PROP_HTTP_VER);
-        String url = "http://" + requestHeader.get(httpHeader.HOST,"localhost") + path;
         
         // check hack attacks in path
         if (path.indexOf("..") >= 0) {

source/de/anomic/plasma/plasmaCrawlLURL.java

@@ -254,15 +254,17 @@ public final class plasmaCrawlLURL extends plasmaURL {
         return null;
     }
 
-    public void removeStack(int stack, int pos) {
+    public boolean removeStack(int stack, int pos) {
+        Object prevElement = null;
         switch (stack) {
-            case 1: externResultStack.remove(pos); break;
-            case 2: searchResultStack.remove(pos); break;
-            case 3: transfResultStack.remove(pos); break;
-            case 4: proxyResultStack.remove(pos); break;
-            case 5: lcrawlResultStack.remove(pos); break;
-            case 6: gcrawlResultStack.remove(pos); break;
+            case 1: prevElement = externResultStack.remove(pos); break;
+            case 2: prevElement = searchResultStack.remove(pos); break;
+            case 3: prevElement = transfResultStack.remove(pos); break;
+            case 4: prevElement = proxyResultStack.remove(pos); break;
+            case 5: prevElement = lcrawlResultStack.remove(pos); break;
+            case 6: prevElement = gcrawlResultStack.remove(pos); break;
         }
+        return prevElement != null;
     }
 
     public void clearStack(int stack) {
@@ -276,16 +278,18 @@ public final class plasmaCrawlLURL extends plasmaURL {
         }
     }
 
-    public void remove(String urlHash) {
-        super.remove(urlHash);
+    public boolean remove(String urlHash) {
+        boolean exists1 = super.remove(urlHash);
         for (int stack = 1; stack <= 6; stack++) {
             for (int i = getStackSize(stack) - 1; i >= 0; i--) {
                 if (getUrlHash(stack,i).equals(urlHash)) {
-                    removeStack(stack,i);
-                    return;
+                    boolean exits2 = removeStack(stack,i);
+                    exists1 = exists1 || exits2;
+                    return exists1;
                 }
             }
         }
+        return exists1;
     }
 
     private static SimpleDateFormat dayFormatter = new SimpleDateFormat("yyyy/MM/dd", Locale.US);

source/de/anomic/plasma/plasmaCrawlNURL.java

@@ -352,10 +352,12 @@ public class plasmaCrawlNURL extends plasmaURL {
         return new Entry(hash);
     }
 
-    public synchronized void remove(String hash) {
+    public synchronized boolean remove(String hash) {
         try {
-            urlHashCache.remove(hash.getBytes());
-        } catch (IOException e) {}
+            return (urlHashCache.remove(hash.getBytes())!=null);
+        } catch (IOException e) {
+            return false;
+        }
     }
 
     public class Entry {

source/de/anomic/plasma/plasmaHTCache.java

@@ -436,8 +436,13 @@ public final class plasmaHTCache {
             remotePath = remotePath + "ndx";
         }
         remotePath = remotePath.replaceAll("[?&:]", "_"); // yes this is not reversible, but that is not needed
-        final int port = url.getPort();
-        if (port < 0 || port == 80) {
+        int port = url.getPort();
+        if (port < 0) {
+            if (url.getProtocol().equalsIgnoreCase("http"))       port = 80;
+            else if (url.getProtocol().equalsIgnoreCase("https")) port = 443;
+            else if (url.getProtocol().equalsIgnoreCase("ftp"))   port = 21;
+        }
+        if (port == 80) {
             return new File(this.cachePath, url.getHost() + remotePath);
         } else {
             return new File(this.cachePath, url.getHost() + "!" + port + remotePath);           
@@ -453,6 +458,8 @@ public final class plasmaHTCache {
 //      this.log.logFinest("plasmaHTCache: getURL:  IN: File=[" + f + "]");
         String s = f.toString().replace('\\', '/');
         final String c = cachePath.toString().replace('\\', '/');
+        
+        String protocol = "http";
         int pos = s.lastIndexOf(c);
         if (pos >= 0) {
             s = s.substring(pos + c.length());
@@ -466,12 +473,19 @@ public final class plasmaHTCache {
             
             pos = s.indexOf("!");
             if (pos >= 0) {
+                String temp = s.substring(pos + 1);
+                if (temp.startsWith("443/")) {
+                    protocol = "https";
+                } else if (temp.startsWith("21/")) {
+                    protocol = "ftp";
+                }
+                
                 s = s.substring(0, pos) + ":" + s.substring(pos + 1);
             }
             if (s.endsWith("ndx")) { s = s.substring(0, s.length() - 3); }
 //          this.log.logFinest("plasmaHTCache: getURL: OUT=" + s);    
             try {
-                return new URL("http://" + s);
+                return new URL(protocol + "://" + s);
             } catch (Exception e) {
                 return null;
             }

source/de/anomic/plasma/plasmaParser.java

@@ -649,15 +649,7 @@ public final class plasmaParser {
     }
     
     public static String urlNormalform(String us) {
-        if (us == null) return null;
-        if (us.length() == 0) return null;
-        int p;
-        if ((p = us.indexOf("#")) >= 0) us = us.substring(0, p);
-        if (us.endsWith(":80")) us = us.substring(0, us.length() - 3);
-        p = us.indexOf(":80/");
-        if (p >= 0) us = us.substring(0,p).concat(us.substring(p + 3)); 
-        if (((us.endsWith("/")) && (us.lastIndexOf('/', us.length() - 2) < 8))) us = us.substring(0, us.length() - 1);
-        return us;
+        return htmlFilterContentScraper.urlNormalform(us);
     }   
     
     static Map allReflinks(Map links) {

source/de/anomic/plasma/plasmaSwitchboard.java

@@ -1155,6 +1155,7 @@ public final class plasmaSwitchboard extends serverAbstractSwitch implements ser
                 while (i.hasNext()) {
                     e = (Map.Entry) i.next();
                     nexturlstring = (String) e.getKey();
+                    nexturlstring = plasmaParser.urlNormalform(nexturlstring);
   
                     sbStackCrawlThread.enqueue(nexturlstring, entry.url().toString(), initiatorHash, (String) e.getValue(), loadDate, entry.depth() + 1, entry.profile());
 
@@ -1217,11 +1218,8 @@ public final class plasmaSwitchboard extends serverAbstractSwitch implements ser
                                      );
                     
                     String urlHash = newEntry.hash();
-                    //log.logDebug("Remove NURL for '" + entry.normalizedURLString() + "'");
-                    urlPool.noticeURL.remove(urlHash); // worked-off
                     
-                    if (((processCase == 4) || (processCase == 5) || (processCase == 6)) &&
-			(entry.profile().localIndexing())) {
+                    if (((processCase == 4) || (processCase == 5) || (processCase == 6)) && (entry.profile().localIndexing())) {
                         // remove stopwords
                         log.logInfo("Excluded " + condenser.excludeWords(stopwords) + " words in URL " + entry.url());
                         indexingEndTime = System.currentTimeMillis();
@@ -1287,6 +1285,7 @@ public final class plasmaSwitchboard extends serverAbstractSwitch implements ser
                         
                         if (log.isLoggable(Level.INFO)) {
                             log.logInfo("*Indexed " + words + " words in URL " + entry.url() + 
+                                        " [" + entry.urlHash() + "]" +
                                         "\n\tDescription:  " + descr + 
                                         "\n\tMimeType: "  + document.getMimeType() + " | " + 
                                         "Size: " + document.text.length + " bytes | " +
@@ -1328,10 +1327,17 @@ public final class plasmaSwitchboard extends serverAbstractSwitch implements ser
         } catch (IOException e) {
             log.logSevere("ERROR in plasmaSwitchboard.process(): " + e.toString());
         } finally {
+            // removing current entry from in process list
             synchronized (this.indexingTasksInProcess) {
                 this.indexingTasksInProcess.remove(entry.urlHash());
             }            
             
+            // removing current entry from notice URL queue
+            boolean removed = urlPool.noticeURL.remove(entry.urlHash()); // worked-off
+            if (!removed) {
+                log.logFinest("Unable to remove indexed URL " + entry.url() + " from Crawler Queue. This could be because of an URL redirect.");
+            }
+            
             // explicit delete/free resources
             if ((entry != null) && (entry.profile() != null) && (!(entry.profile().storeHTCache()))) {
                 plasmaHTCache.filesInUse.remove(entry.cacheFile());

source/de/anomic/plasma/plasmaURL.java

@@ -452,11 +452,14 @@ public class plasmaURL {
 	}
     }
 
-    public void remove(String urlHash) {
-	try {
-            existsIndex.remove(urlHash);
-	    urlHashCache.remove(urlHash.getBytes());
-	} catch (IOException e) {}
+    public boolean remove(String urlHash) {
+        try {
+            boolean existsInIndex = this.existsIndex.remove(urlHash); 
+            boolean existsInCache = (this.urlHashCache.remove(urlHash.getBytes())!= null);
+            return existsInIndex || existsInCache;
+        } catch (IOException e) {
+            return false;
+        }
     }
 
     public static final int flagTypeID(String hash) {
@@ -495,7 +498,15 @@ public class plasmaURL {
             dom = dom.substring(p + 1);
         }
         int port = url.getPort();
-        if (port <= 0) port = (isHTTP) ? 80 : 21;
+        if (port <= 0) {
+            if (isHTTP) {
+                port = 80;
+            } else if (url.getProtocol().equalsIgnoreCase("https")) {
+                port = 443;
+            } else {
+                port = 21;
+            }
+        }
         String path = url.getPath();
         if (path.startsWith("/")) path = path.substring(1);
         if (path.endsWith("/")) path = path.substring(0, path.length() - 1);

source/de/anomic/yacy/yacyCore.java

@@ -676,7 +676,12 @@ public class yacyCore {
             try{
                 final String seedURLStr = sb.getConfig("seedURL", "");
                 if (seedURLStr.length() == 0) { throw new MalformedURLException("The seed-file url must not be empty."); }
-                if (!seedURLStr.toLowerCase().startsWith("http://")) { throw new MalformedURLException("Unsupported protocol."); }
+                if (!(
+                        seedURLStr.toLowerCase().startsWith("http://") ||
+                        seedURLStr.toLowerCase().startsWith("https://")
+                )){ 
+                    throw new MalformedURLException("Unsupported protocol."); 
+                }
                 seedURL = new URL(seedURLStr);
             } catch(MalformedURLException e) {
                 final String errorMsg = "Malformed seed file URL '" + sb.getConfig("seedURL", "") + "'. " + e.getMessage();

source/de/anomic/yacy/yacyPeerActions.java

@@ -157,7 +157,10 @@ public class yacyPeerActions {
         for (int i = 0; i < superseed.size(); i++) {
             if (Thread.currentThread().isInterrupted()) break;
             seedListFileURL = (String) superseed.any();
-            if (seedListFileURL.startsWith("http://")) {
+            if (
+                    seedListFileURL.startsWith("http://") || 
+                    seedListFileURL.startsWith("https://")
+            ) {
                 // load the seed list
                 try {
                     httpHeader reqHeader = new httpHeader();