From 4149939c02ef9fc2e1b59ddb1a66f8b5843fd946 Mon Sep 17 00:00:00 2001
From: orbiter <orbiter@6c8d7289-2bf4-0310-a012-ef5d649a1542>
Date: Wed, 19 Jul 2006 23:18:06 +0000
Subject: [PATCH] better handling of whitespace for gettext quotation

git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@2310 6c8d7289-2bf4-0310-a012-ef5d649a1542
---
 .../htmlFilterContentTransformer.java         | 12 +++++----
 source/de/anomic/server/serverByteBuffer.java | 25 +++++++++++++++++++
 2 files changed, 32 insertions(+), 5 deletions(-)

diff --git a/source/de/anomic/htmlFilter/htmlFilterContentTransformer.java b/source/de/anomic/htmlFilter/htmlFilterContentTransformer.java
index d2429a2cc..a33eb7166 100644
--- a/source/de/anomic/htmlFilter/htmlFilterContentTransformer.java
+++ b/source/de/anomic/htmlFilter/htmlFilterContentTransformer.java
@@ -139,24 +139,26 @@ public class htmlFilterContentTransformer extends htmlFilterAbstractTransformer
     public byte[] transformText(byte[] text) {
         if (gettext) {
             serverByteBuffer sbb = new serverByteBuffer(text);
-            sbb.trim();
             //if (sbb.length() > 0) System.out.println("   TEXT: " + sbb.toString());
             serverByteBuffer[] sbbs = httpTemplate.splitQuotations(sbb);
             sbb = new serverByteBuffer();
             for (int i = 0; i < sbbs.length; i++) {
-                sbbs[i].trim();
-                if (sbbs[i].length() == 0) {
-                    sbb.append(' ');
+                if (sbbs[i].isWhitespace()) {
+                    sbb.append(sbbs[i]);
                 } else if ((sbbs[i].byteAt(0) == httpTemplate.hash) ||
                            (sbbs[i].startsWith(httpTemplate.dpdpa))) {
                     // this is a template or a part of a template
                     sbb.append(sbbs[i]);
                 } else {
                     // this is a text fragment, generate gettext quotation
+                    int ws = sbbs[i].whitespaceStart();
+                    int we = sbbs[i].whitespaceEnd();
+                    sbb.append(sbbs[i].getBytes(0, ws));
                     sbb.append('_');
                     sbb.append('(');
-                    sbb.append(sbbs[i]);
+                    sbb.append(sbbs[i].getBytes(ws, we));
                     sbb.append(')');
+                    sbb.append(sbbs[i].getBytes(we));
                 }
             }
             //if (sbb.length() > 0) System.out.println("GETTEXT: " + sbb.toString());
diff --git a/source/de/anomic/server/serverByteBuffer.java b/source/de/anomic/server/serverByteBuffer.java
index 829bfbdac..adc879364 100644
--- a/source/de/anomic/server/serverByteBuffer.java
+++ b/source/de/anomic/server/serverByteBuffer.java
@@ -290,6 +290,31 @@ public final class serverByteBuffer extends OutputStream {
         return trim(l, r);
     }
 
+    public boolean isWhitespace() {
+        // returns true, if trim() would result in an empty serverByteBuffer
+        for (int i = 0; i < length; i++) {
+            if (buffer[offset + i] > 32) return false;
+        }
+        return true;
+    }
+    
+    public int whitespaceStart() {
+        // returns number of whitespace bytes at the beginning of text
+        for (int i = 0; i < length; i++) {
+            if (buffer[offset + i] > 32) return i;
+        }
+        return length;
+    }
+    
+    public int whitespaceEnd() {
+        // returns position of whitespace at the end of text
+        for (int i = length - 1; i >= 0; i--) {
+            if (buffer[offset + i] > 32) return i + 1;
+        }
+        return 0;
+    }
+    
+    
     public String toString() {
         return new String(buffer, offset, length);
     }