fix escaped wildcard query on wildcard field

CHANGELOG.md

@@ -27,6 +27,7 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 ### Removed
 
 ### Fixed
+- Fix wildcard query containing escaped character ([#15737](https://github.com/opensearch-project/OpenSearch/pull/15737))
 
 ### Security
 

server/src/main/java/org/opensearch/index/mapper/WildcardFieldMapper.java

@@ -40,7 +40,6 @@
 import org.apache.lucene.util.automaton.RegExp;
 import org.opensearch.common.lucene.BytesRefs;
 import org.opensearch.common.lucene.Lucene;
-import org.opensearch.common.regex.Regex;
 import org.opensearch.common.unit.Fuzziness;
 import org.opensearch.core.xcontent.XContentParser;
 import org.opensearch.index.analysis.IndexAnalyzers;
@@ -430,22 +429,27 @@
                 finalValue = value;
             }
             Predicate<String> matchPredicate;
-            if (value.contains("?")) {
-                Automaton automaton = WildcardQuery.toAutomaton(new Term(name(), finalValue));
-                CompiledAutomaton compiledAutomaton = new CompiledAutomaton(automaton);
+            Automaton automaton = WildcardQuery.toAutomaton(new Term(name(), finalValue));
+            CompiledAutomaton compiledAutomaton = new CompiledAutomaton(automaton);
+            if (compiledAutomaton.type == CompiledAutomaton.AUTOMATON_TYPE.SINGLE) {
+                // when type equals SINGLE, #compiledAutomaton.runAutomaton is null
                 matchPredicate = s -> {
                     if (caseInsensitive) {
                         s = s.toLowerCase(Locale.ROOT);
                     }
-                    BytesRef valueBytes = BytesRefs.toBytesRef(s);
-                    return compiledAutomaton.runAutomaton.run(valueBytes.bytes, valueBytes.offset, valueBytes.length);
+                    return s.equals(finalValue);
                 };
+            } else if (compiledAutomaton.type == CompiledAutomaton.AUTOMATON_TYPE.ALL) {
+                return existsQuery(context);
+            } else if (compiledAutomaton.type == CompiledAutomaton.AUTOMATON_TYPE.NONE) {
+                return new MatchNoDocsQuery("Wildcard expression matches nothing");
             } else {
                 matchPredicate = s -> {
                     if (caseInsensitive) {
                         s = s.toLowerCase(Locale.ROOT);
                     }
-                    return Regex.simpleMatch(finalValue, s);
+                    BytesRef valueBytes = BytesRefs.toBytesRef(s);
+                    return compiledAutomaton.runAutomaton.run(valueBytes.bytes, valueBytes.offset, valueBytes.length);
                 };
             }
 
@@ -468,22 +472,30 @@
         // Package-private for testing
         static Set<String> getRequiredNGrams(String value) {
             Set<String> terms = new HashSet<>();
+
+            if (value.isEmpty()) {
+                return terms;
+            }
+
             int pos = 0;
+            String rawSequence = null;
             String currentSequence = null;
             if (!value.startsWith("?") && !value.startsWith("*")) {
                 // Can add prefix term
-                currentSequence = getNonWildcardSequence(value, 0);
+                rawSequence = getNonWildcardSequence(value, 0);
+                currentSequence = performEscape(rawSequence);
                 if (currentSequence.length() == 1) {
                     terms.add(new String(new char[] { 0, currentSequence.charAt(0) }));
                 } else {
                     terms.add(new String(new char[] { 0, currentSequence.charAt(0), currentSequence.charAt(1) }));
                 }
             } else {
                 pos = findNonWildcardSequence(value, pos);
-                currentSequence = getNonWildcardSequence(value, pos);
+                rawSequence = getNonWildcardSequence(value, pos);
             }
             while (pos < value.length()) {
-                boolean isEndOfValue = pos + currentSequence.length() == value.length();
+                boolean isEndOfValue = pos + rawSequence.length() == value.length();
+                currentSequence = performEscape(rawSequence);
                 if (!currentSequence.isEmpty() && currentSequence.length() < 3 && !isEndOfValue && pos > 0) {
                     // If this is a prefix or suffix of length < 3, then we already have a longer token including the anchor.
                     terms.add(currentSequence);
@@ -502,16 +514,16 @@
                         terms.add(new String(new char[] { a, b, 0 }));
                     }
                 }
-                pos = findNonWildcardSequence(value, pos + currentSequence.length());
-                currentSequence = getNonWildcardSequence(value, pos);
+                pos = findNonWildcardSequence(value, pos + rawSequence.length());
+                rawSequence = getNonWildcardSequence(value, pos);
             }
             return terms;
         }
 
         private static String getNonWildcardSequence(String value, int startFrom) {
             for (int i = startFrom; i < value.length(); i++) {
                 char c = value.charAt(i);
-                if (c == '?' || c == '*') {
+                if ((c == '?' || c == '*') && (i == 0 || value.charAt(i - 1) != '\\')) {
                     return value.substring(startFrom, i);
                 }
             }
@@ -529,6 +541,22 @@
             return value.length();
         }
 
+        private static String performEscape(String str) {
+            StringBuilder sb = new StringBuilder();
+            for (int i = 0; i < str.length(); i++) {
+                if (str.charAt(i) == '\\' && (i + 1) < str.length()) {
+                    char c = str.charAt(i + 1);
+                    if (c == '*' || c == '?') {
+                        i++;
+                    }
+                }
+                sb.append(str.charAt(i));
+            }
+            assert !sb.toString().contains("\\*");
+            assert !sb.toString().contains("\\?");
+            return sb.toString();
+        }
+
         @Override
         public Query regexpQuery(
             String value,

server/src/test/java/org/opensearch/index/mapper/WildcardFieldTypeTests.java

@@ -88,6 +88,32 @@ public void testWildcardQuery() {
         );
     }
 
+    public void testEscapedWildcardQuery() {
+        MappedFieldType ft = new WildcardFieldMapper.WildcardFieldType("field");
+        Set<String> expectedTerms = new HashSet<>();
+        expectedTerms.add(prefixAnchored("*"));
+        expectedTerms.add(suffixAnchored("*"));
+
+        BooleanQuery.Builder builder = new BooleanQuery.Builder();
+        for (String term : expectedTerms) {
+            builder.add(new TermQuery(new Term("field", term)), BooleanClause.Occur.FILTER);
+        }
+
+        assertEquals(
+            new WildcardFieldMapper.WildcardMatchingQuery("field", builder.build(), "\\**\\*"),
+            ft.wildcardQuery("\\**\\*", null, null)
+        );
+
+        assertEquals(new WildcardFieldMapper.WildcardMatchingQuery("field", builder.build(), "\\*"), ft.wildcardQuery("\\*", null, null));
+
+        expectedTerms.remove(suffixAnchored("*"));
+        builder = new BooleanQuery.Builder();
+        for (String term : expectedTerms) {
+            builder.add(new TermQuery(new Term("field", term)), BooleanClause.Occur.FILTER);
+        }
+        assertEquals(new WildcardFieldMapper.WildcardMatchingQuery("field", builder.build(), "\\**"), ft.wildcardQuery("\\**", null, null));
+    }
+
     public void testMultipleWildcardsInQuery() {
         final String pattern = "a?cd*efg?h";
         MappedFieldType ft = new WildcardFieldMapper.WildcardFieldType("field");