url parser regex performance hack

git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@6524 6c8d7289-2bf4-0310-a012-ef5d649a1542

source/de/anomic/search/RankingProcess.java

@@ -641,7 +641,7 @@ public final class RankingProcess extends Thread {
         // take out relevant information for reference computation
         if ((resultEntry.url() == null) || (resultEntry.title() == null)) return;
         //final String[] urlcomps = htmlFilterContentScraper.urlComps(resultEntry.url().toNormalform(true, true)); // word components of the url
-        final String[] descrcomps = resultEntry.title().toLowerCase().split(DigestURI.splitrex); // words in the description
+        final String[] descrcomps = DigestURI.splitpattern.split(resultEntry.title().toLowerCase()); // words in the description
         
         // add references
         //addTopic(urlcomps);

source/de/anomic/search/ResultFetcher.java

@@ -373,7 +373,7 @@ public class ResultFetcher {
         // apply 'common-sense' heuristic using references
         final String urlstring = rentry.url().toNormalform(true, true);
         final String[] urlcomps = DigestURI.urlComps(urlstring);
-        final String[] descrcomps = rentry.title().toLowerCase().split(DigestURI.splitrex);
+        final String[] descrcomps = DigestURI.splitpattern.split(rentry.title().toLowerCase());
         Integer tc;
         for (int j = 0; j < urlcomps.length; j++) {
             tc = topwords.get(urlcomps[j]);

source/de/anomic/yacy/yacyClient.java

@@ -616,7 +616,7 @@ public final class yacyClient {
             final String references = result.get("references");
             yacyCore.log.logInfo("remote search (client): peer " + target.getName() + " sent references " + references);
             if (references != null) {
-                // add references twice, so they can be countet (must have at least 2 entries)
+                // add references twice, so they can be counted (must have at least 2 entries)
                 containerCache.addTopic(references.split(","));
                 containerCache.addTopic(references.split(","));
             }

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

@@ -402,7 +402,7 @@ public class ContentScraper extends AbstractScraper implements Scraper {
         if (s == null) s = metas.get("dc.description");
         if (s == null) s = "";
         if (s.length() == 0) {
-            return getTitle().toLowerCase().split(DigestURI.splitrex);
+            return DigestURI.splitpattern.split(getTitle().toLowerCase());
         }
         if (s.contains(",")) return s.split(" |,");
         if (s.contains(";")) return s.split(" |;");

source/net/yacy/kelondro/data/meta/DigestURI.java

@@ -938,11 +938,12 @@ public class DigestURI implements Serializable {
         return language;
     }
 
-    public static final String splitrex = " |/|\\(|\\)|-|\\:|_|\\.|,|\\?|!|'|" + '"';
+    private static final String splitrex = " |/|\\(|\\)|-|\\:|_|\\.|,|\\?|!|'|" + '"';
+    public static final Pattern splitpattern = Pattern.compile(splitrex);
     public static String[] urlComps(String normalizedURL) {
         final int p = normalizedURL.indexOf("//");
         if (p > 0) normalizedURL = normalizedURL.substring(p + 2);
-        return normalizedURL.toLowerCase().split(splitrex); // word components of the url
+        return splitpattern.split(normalizedURL.toLowerCase()); // word components of the url
     }
 
     public static void main(final String[] args) {