fix in html parser

source/net/yacy/cora/document/id/MultiProtocolURL.java

@@ -57,6 +57,7 @@ import net.yacy.cora.protocol.TimeoutRequest;
 import net.yacy.cora.protocol.ftp.FTPClient;
 import net.yacy.cora.protocol.http.HTTPClient;
 import net.yacy.cora.util.CommonPattern;
+import net.yacy.document.parser.html.CharacterCoding;
 
 /**
  * MultiProtocolURI provides a URL object for multiple protocols like http, https, ftp, smb and file
@@ -66,7 +67,6 @@ public class MultiProtocolURL implements Serializable, Comparable<MultiProtocolU
 
     public static final MultiProtocolURL POISON = new MultiProtocolURL(); // poison pill for concurrent link generators
 
-    private static final Pattern ampPattern = Pattern.compile(Pattern.quote("&amp;"));
     private static final long serialVersionUID = -1173233022912141884L;
     private static final long SMB_TIMEOUT = 5000;
 
@@ -636,7 +636,7 @@ public class MultiProtocolURL implements Serializable, Comparable<MultiProtocolU
         } else {
             this.searchpart = this.path.substring(r + 1);
             // strip &amp;
-            Matcher matcher = ampPattern.matcher(this.searchpart);
+            Matcher matcher = CharacterCoding.ampPattern.matcher(this.searchpart);
             while (matcher.find()) {
                 this.searchpart = matcher.replaceAll("&");
                 matcher.reset(this.searchpart);

source/net/yacy/document/parser/html/CharacterCoding.java

@@ -26,12 +26,15 @@ package net.yacy.document.parser.html;
 
 import java.util.HashMap;
 import java.util.Map;
+import java.util.regex.Pattern;
 
 /**
  * Contains methods to convert between Unicode and XML/HTML encoding.
  */
 public final class CharacterCoding {
 
+    /** Ampersand pattern */
+    public final static Pattern ampPattern = Pattern.compile(Pattern.quote("&"));
     /** Ampersand character in unicode encoding. */
     private static final char AMP_UNICODE = "\u0026".charAt(0);
     /** Ampersand character in HTML encoding. */
@@ -276,14 +279,15 @@ public final class CharacterCoding {
         }
         return sb.toString();
     }
-
+    
     /**
      * Replaces HTML-encoded characters with unicode representation.
      * @param text text with character to replace
      * @return text with replaced characters
      */
-    public static String html2unicode(final String text) {
+    public static String html2unicode(String text) {
         if (text == null) return null;
+        text = ampPattern.matcher(text).replaceAll("&"); // sometimes a double-replacement is necessary.
         int p = 0, p1, q;
         final StringBuilder sb = new StringBuilder(text.length());
         String s;

source/net/yacy/document/parser/html/ContentScraper.java

@@ -204,11 +204,12 @@ public class ContentScraper extends AbstractScraper implements Scraper {
     }
 
     @Override
-    public void scrapeText(final char[] newtext, final String insideTag) {
+    public void scrapeText(final char[] newtext0, final String insideTag) {
         // System.out.println("SCRAPE: " + UTF8.String(newtext));
         if (insideTag != null && ("script".equals(insideTag) || "style".equals(insideTag))) return;
         int p, pl, q, s = 0;
-
+        char[] newtext = CharacterCoding.html2unicode(new String(newtext0)).toCharArray();
+        
         // match evaluation pattern
         this.evaluationScores.match(Element.text, newtext);