Rewrote split() and parseStringArray() not to use StringTokenizer because

it does undesirable things like not report multiple separators and skip
separators that show up at the beginning of the string. We prefer the more
rigorous approach of interpreting something like ",," as three empty
strings rather than nothing at all which is what StringTokenizer would
report.


git-svn-id: https://samskivert.googlecode.com/svn/trunk@567 6335cc39-0255-0410-8fd6-9bcaacd3b74c

projects/samskivert/src/java/com/samskivert/util/StringUtil.java

@@ -1,5 +1,5 @@
 //
-// $Id: StringUtil.java,v 1.27 2002/02/06 00:56:57 mdb Exp $
+// $Id: StringUtil.java,v 1.28 2002/02/07 08:41:59 mdb Exp $
 //
 // samskivert library - useful routines for java programs
 // Copyright (C) 2001 Michael Bayne
@@ -413,19 +413,32 @@ public class StringUtil
      */
     public static String[] parseStringArray (String source)
     {
+        int tcount = 0, tpos = -1, tstart = 0;
+
         // sort out escaped commas
         source = replace(source, ",,", "%COMMA%");
-        // now tokenize the string
-        StringTokenizer tok = new StringTokenizer(source, ",");
-        String[] vals = new String[tok.countTokens()];
-        for (int i = 0; tok.hasMoreTokens(); i++) {
-            // trim the whitespace from the token
-            String token = tok.nextToken().trim();
-            // undo the comma escaping
-            token = replace(token, "%COMMA%", ",");
-            vals[i] = token;
+
+        // count up the number of tokens
+        while ((tpos = source.indexOf(",", tpos+1)) != -1) {
+            tcount++;
         }
-        return vals;
+
+        String[] tokens = new String[tcount+1];
+        tpos = -1; tcount = 0;
+
+        // do the split
+        while ((tpos = source.indexOf(",", tpos+1)) != -1) {
+            tokens[tcount] = source.substring(tstart, tpos);
+            tokens[tcount] = replace(tokens[tcount].trim(), "%COMMA%", ",");
+            tstart = tpos+1;
+            tcount++;
+        }
+
+        // grab the last token
+        tokens[tcount] = source.substring(tstart);
+        tokens[tcount] = replace(tokens[tcount].trim(), "%COMMA%", ",");
+
+        return tokens;
     }
 
     /**
@@ -451,15 +464,31 @@ public class StringUtil
 
     /**
      * Splits the supplied string into components based on the specified
-     * separator character.
+     * separator string.
      */
     public static String[] split (String source, String sep)
     {
-        StringTokenizer tok = new StringTokenizer(source, sep);
-        String[] tokens = new String[tok.countTokens()];
-        for (int i = 0; tok.hasMoreTokens(); i++) {
-            tokens[i] = tok.nextToken();
+        int tcount = 0, tpos = -1, tstart = 0;
+
+        // count up the number of tokens
+        while ((tpos = source.indexOf(sep, tpos+1)) != -1) {
+            tcount++;
         }
+
+        String[] tokens = new String[tcount+1];
+        tpos = -1; tcount = 0;
+
+        // do the split
+        while ((tpos = source.indexOf(sep, tpos+1)) != -1) {
+            tokens[tcount] = source.substring(tstart, tpos);
+            tstart = tpos+1;
+            tcount++;
+        }
+
+        // grab the last token
+        tokens[tcount] = source.substring(tstart);
+        tokens[tcount] = replace(tokens[tcount].trim(), "%COMMA%", ",");
+
         return tokens;
     }