improve url hash computation for file path with mixed java & windows

file.separator to compute equal hashes (by normalizing path for computation)
+ expand test case for to check mixed java / windows file url notation
like e.g. file:///c:/test/file.html vs. file:///c:\test/file.html
- relates partially to http://mantis.tokeek.de/view.php?id=692

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

@@ -242,17 +242,24 @@ public class DigestURL extends MultiProtocolURL implements Serializable {
         }
 
         // find rootpath
+        final String normalizedPath;
+        if (this.isFile() && this.path.indexOf('\\') > -1) // for file protocol normalize path to java notation
+            normalizedPath = this.path.replace('\\','/'); // replace possible Windows pathseparator
+        else
+            normalizedPath = this.path;
+
         int rootpathStart = 0;
-        int rootpathEnd = this.path.length() - 1;
-        if (!this.path.isEmpty() && (this.path.charAt(0) == '/' || this.path.charAt(0) == '\\'))
+        int rootpathEnd = normalizedPath.length() - 1;
+        if (!normalizedPath.isEmpty() && (normalizedPath.charAt(0) == '/'))
             rootpathStart = 1;
-        if (this.path.endsWith("/"))
-            rootpathEnd = this.path.length() - 2;
-        p = this.path.indexOf('/', rootpathStart);
-        if (this.isFile() && p < 0) p = this.path.indexOf('\\', rootpathStart); // double-check for windows path (if it's a file url)
+        if (normalizedPath.endsWith("/"))
+            rootpathEnd = normalizedPath.length() - 2;
+        p = normalizedPath.indexOf('/', rootpathStart);
+        // following doesn't recognize mixed notation e.g. c:\\tmp/test.html correct -> solved by using normalized path
+        //if (this.isFile() && p < 0) p = this.path.indexOf('\\', rootpathStart); // double-check for windows path (if it's a file url)
         String rootpath = "";
         if (p > 0 && p < rootpathEnd) {
-            rootpath = this.path.substring(rootpathStart, p);
+            rootpath = normalizedPath.substring(rootpathStart, p);
         }
 
         // we collected enough information to compute the fragments that are

test/java/net/yacy/cora/document/id/DigestURLTest.java

@@ -1,7 +1,10 @@
 package net.yacy.cora.document.id;
 
 import java.net.MalformedURLException;
+import java.util.HashSet;
+import java.util.Set;
 import junit.framework.TestCase;
+import static junit.framework.TestCase.assertEquals;
 import net.yacy.cora.document.encoding.ASCII;
 import org.junit.Test;
 
@@ -37,16 +40,24 @@ public class DigestURLTest extends TestCase {
      */
     @Test
     public void testHash_ForFile() throws MalformedURLException {
-        String winUrlStr = "file:///C:\\tmp\\test.html"; // allowed Windows notation
         String javaUrlStr = "file:///C:/tmp/test.html"; // allowed Java notation for Windows file system
 
-        DigestURL winUrl = new DigestURL(winUrlStr);
-        DigestURL javaUrl = new DigestURL(javaUrlStr);
+        // allowed Windows notation
+        Set<String> testUrls = new HashSet();
+        testUrls.add("file:///C:\\tmp\\test.html");
+        testUrls.add("file:///C:/tmp\\test.html");
+        testUrls.add("file:///C:\\tmp/test.html");
+        testUrls.add("file:///C:/tmp/test.html");
 
-        String winHashResult = ASCII.String(winUrl.hash());
+        DigestURL javaUrl = new DigestURL(javaUrlStr);
         String javaHashResult = ASCII.String(javaUrl.hash());
-        
-        assertEquals("hash for same file url", javaHashResult, winHashResult);
+
+        // compare test url hash to default java Url notation
+        for (String str : testUrls) {
+            DigestURL winUrl = new DigestURL(str);
+            String winHashResult = ASCII.String(winUrl.hash());
+            assertEquals("hash for same file url "+str, javaHashResult, winHashResult);
+        }
 
     }