CVS: bzflag/src/common WordFilter.cxx,1.23,1.24

[Frank T. <ch...@us...>]

Update of /cvsroot/bzflag/bzflag/src/common
In directory sc8-pr-cvs1:/tmp/cvs-serv17716/src/common

Modified Files:
	WordFilter.cxx 
Log Message:
fix some warnings because a char was used as array index; can cuase problem when char is signed and contains values<0 (>=128); does filter work at all for 8 bit characters? can that occur?

Index: WordFilter.cxx
===================================================================
RCS file: /cvsroot/bzflag/bzflag/src/common/WordFilter.cxx,v
retrieving revision 1.23
retrieving revision 1.24
diff -w -u -r1.23 -r1.24
--- WordFilter.cxx	27 Aug 2003 01:00:53 -0000	1.23
+++ WordFilter.cxx	29 Aug 2003 21:42:41 -0000	1.24
@@ -42,8 +42,11 @@
 
     word = line.substr(startPosition, endPosition-startPosition);
     findWord.word = word;
-    if (filters[word[0]].find(findWord) != \
-	filters[word[0]].end()) {
+    int firstchar = word[0];
+    if (firstchar < 0)
+      firstchar+=256; // FIXME, not very clean?
+    if (filters[firstchar].find(findWord) != \
+	filters[firstchar].end()) {
 
       /* fill with asterisks */
       //      memset(input+startPosition,'*', endPosition-startPosition);
@@ -86,14 +89,14 @@
   startPosition = 0;
 
   int endPosition;
-  char wordIndices[MAX_WORDS];
+  unsigned char wordIndices[MAX_WORDS];
   unsigned int wordIndexLength=0;
 
   char characterIndex;
 
   /* clear memory for arrays */
-  memset(matchPair, 0, MAX_WORDS * 2 * sizeof(int));
-  memset(wordIndices, 0, MAX_WORDS * sizeof(char));
+  memset(matchPair, 0, MAX_WORDS * 2 * sizeof matchPair[0]);
+  memset(wordIndices, 0, MAX_WORDS * sizeof wordIndices[0]);
 
   /* iterate over all the words start position in the input and keep track
    * of the starting character of each word (we only need to check those)
@@ -637,12 +640,15 @@
     newFilter.expression = expression;
     newFilter.compiled = getCompiledExpression(expression);
 
+    int firstchar = tolower(word[0]);
+    if (firstchar < 0)
+      firstchar += 256; // FIXME, not too clean?
     /* check if the word is already added */
-    if (filters[tolower(word[0])].find(newFilter) != \
-	filters[word[0]].end()) {
+    if (filters[firstchar].find(newFilter) != \
+	filters[firstchar].end()) {
       return false;
     } else {
-      filters[tolower(word[0])].insert(newFilter);
+      filters[firstchar].insert(newFilter);
     }
     return true;
   }