*) Bugfix for UTF-8: url names are now stored properly in stackcrawl, crawler, indexing queue and should be displayed correct on the gui

git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@2630 6c8d7289-2bf4-0310-a012-ef5d649a1542

source/de/anomic/plasma/plasmaCrawlNURL.java

@@ -500,7 +500,7 @@ public class plasmaCrawlNURL extends indexURL {
             this.initiator = entry.getColString(1, null);
             this.url = new URL(urlstring);
             this.referrer = (entry.empty(3)) ? dummyHash : entry.getColString(3, null);
-            this.name = (entry.empty(4)) ? "" : entry.getColString(4, null).trim();
+            this.name = (entry.empty(4)) ? "" : entry.getColString(4, "UTF-8").trim();
             this.loaddate = new Date(86400000 * entry.getColLong(5));
             this.profileHandle = (entry.empty(6)) ? null : entry.getColString(6, null).trim();
             this.depth = (int) entry.getColLong(7);
@@ -523,7 +523,7 @@ public class plasmaCrawlNURL extends indexURL {
                     (initiator == null) ? "".getBytes() : this.initiator.getBytes(),
                     this.url.toString().getBytes(),
                     this.referrer.getBytes(),
-                    this.name.getBytes(),
+                    this.name.getBytes("UTF-8"),
                     loaddatestr.getBytes(),
                     (this.profileHandle == null) ? null : this.profileHandle.getBytes(),
                     kelondroBase64Order.enhancedCoder.encodeLong(this.depth, urlCrawlDepthLength).getBytes(),

source/de/anomic/plasma/plasmaCrawlStacker.java

@@ -48,6 +48,7 @@ package de.anomic.plasma;
 
 import java.io.File;
 import java.io.IOException;
+import java.io.UnsupportedEncodingException;
 import java.net.InetAddress;
 import java.net.MalformedURLException;
 import java.util.Date;
@@ -560,12 +561,14 @@ public final class plasmaCrawlStacker {
             // store the hash in the hash cache
 
             // even if the entry exists, we simply overwrite it
-            byte[][] entry = new byte[][] { 
+            byte[][] entry = null;
+            try {
+                entry = new byte[][] { 
                     this.urlHash.getBytes(),
                     (this.initiator == null) ? "".getBytes() : this.initiator.getBytes(),
                     this.url.getBytes(),
                     this.referrerHash.getBytes(),
-                    this.name.getBytes(),
+                    this.name.getBytes("UTF-8"),
                     loaddatestr.getBytes(),
                     (this.profileHandle == null) ? null : this.profileHandle.getBytes(),
                     kelondroBase64Order.enhancedCoder.encodeLong(this.depth, indexURL.urlCrawlDepthLength).getBytes(),
@@ -574,6 +577,7 @@ public final class plasmaCrawlStacker {
                     this.flags.getBytes(),
                     normalizeHandle(this.handle).getBytes()
             };
+            } catch (UnsupportedEncodingException e) { /* ignore this */ }
             return entry;
         }        
         

source/de/anomic/plasma/plasmaSwitchboardQueue.java

@@ -110,7 +110,7 @@ public class plasmaSwitchboardQueue {
             (entry.initiator == null) ? indexURL.dummyHash.getBytes() : entry.initiator.getBytes(),
             kelondroBase64Order.enhancedCoder.encodeLong((long) entry.depth, indexURL.urlCrawlDepthLength).getBytes(),
             (entry.profileHandle == null) ? indexURL.dummyHash.getBytes() : entry.profileHandle.getBytes(),
-            (entry.anchorName == null) ? "-".getBytes() : entry.anchorName.getBytes()
+            (entry.anchorName == null) ? "-".getBytes("UTF-8") : entry.anchorName.getBytes("UTF-8")
         }));
     }