Remove adjacent duplicates to optimize regex before processing

Signed-off-by: Craig Perkins <[email protected]>
cwperks · Nov 2, 2023 · c1ffbc4 · c1ffbc4
1 parent 8673fa9
commit c1ffbc4
Showing 1 changed file with 21 additions and 1 deletion.
diff --git a/server/src/main/java/org/opensearch/common/regex/Regex.java b/server/src/main/java/org/opensearch/common/regex/Regex.java
@@ -93,6 +93,25 @@ public static Automaton simpleMatchToAutomaton(String... patterns) {
         return Operations.union(automata);
     }
 
+    /**
+     *
+     * @param str - The input string to remove adjacent duplicate characters from
+     * @param target - The target character to remove duplicates of
+     * @return
+     */
+    public static String removeDuplicates(String str, char target) {
+        StringBuilder sb = new StringBuilder();
+        for (char c : str.toCharArray()) {
+            int size = sb.length();
+            if (size > 0 && c == target && sb.charAt(size - 1) == c) {
+                sb.deleteCharAt(size - 1);
+            } else {
+                sb.append(c);
+            }
+        }
+        return sb.toString();
+    }
+
     /**
      * Match a String against the given pattern, supporting the following simple
      * pattern styles: "xxx*", "*xxx", "*xxx*" and "xxx*yyy" matches (with an
@@ -104,7 +123,8 @@ public static Automaton simpleMatchToAutomaton(String... patterns) {
      * @return whether the String matches the given pattern
      */
     public static boolean simpleMatch(String pattern, String str) {
-        return simpleMatch(pattern, str, false);
+        String trimmedPattern = removeDuplicates(pattern, '*');
+        return simpleMatch(trimmedPattern, str, false);
     }
 
     /**