feat: avoid looping through the tokens every single time

Signed-off-by: Andreas Reichel <andreas@manticore-projects.com>

Author: manticore-projects

src/main/jjtree/net/sf/jsqlparser/parser/JSqlParserCC.jjt

@@ -166,6 +166,46 @@ PARSER_END(CCJSqlParser)
 TOKEN_MGR_DECLS : {
     public FeatureConfiguration configuration = new FeatureConfiguration();
 
+    // Identify the index of the quoting/escaping tokens
+    public int charLiteralIndex = -1;
+    public int squaredBracketOpenIndex = -1;
+    {
+        for (int i=0;i<CCJSqlParserConstants.tokenImage.length;i++) {
+          if ( CCJSqlParserConstants.tokenImage[i].equals("<S_CHAR_LITERAL>") ) {
+              charLiteralIndex = i;
+              break;
+          }
+        }
+        for (int i=0;i<CCJSqlParserConstants.tokenImage.length;i++) {
+            if (CCJSqlParserConstants.tokenImage[i].equals("\"[\"")) {
+                squaredBracketOpenIndex = i;
+                break;
+            }
+        }
+    }
+
+    // Finds first occurrence of "\\'"
+    public static int indexOfSequence(String s, String target) {
+        int len = s.length();
+        for (int i = 0; i < len - 1; i++) {
+            if (s.charAt(i) == '\\' && s.charAt(i + 1) == '\'') {
+                return i;
+            }
+        }
+        return -1;
+    }
+
+    // Finds last occurrence of "\\''"
+    public static int lastIndexOfSequence(String s, String target) {
+        int len = s.length();
+        for (int i = len - 3; i >= 0; i--) {
+            if (s.charAt(i) == '\\' && s.charAt(i + 1) == '\'' && s.charAt(i + 2) == '\'') {
+                return i;
+            }
+        }
+        return -1;
+    }
+
     public void CommonTokenAction(Token t)
 	{
 		t.absoluteBegin = getCurrentTokenAbsolutePosition();
@@ -614,6 +654,7 @@ TOKEN : /* Statement Separators */
 TOKEN : /* Operators */
 {
     <OP_GREATERTHANEQUALS: ">" (<WHITESPACE>)* "=">
+|   <OP_COSINESIMILARITY: "<=>">
 |   <OP_MINORTHANEQUALS: "<" (<WHITESPACE>)* "=">
 |   <OP_NOTEQUALSSTANDARD: "<" (<WHITESPACE>)* ">">
 |   <OP_NOTEQUALSBANG: "!" (<WHITESPACE>)* "=">
@@ -738,33 +779,33 @@ TOKEN:
         // which contains the <SPECIAL_ESC>, then we will need to
         // 1) break the <S_CHAR_LITERAL> at <SPECIAL_ESC> close it with a "'"
         // 2) continue tokenizing after that <SPECIAL_ESC> with a new <S_CHAR_LITERAL> or any other Token
-        if ( !configuration.getAsBoolean(Feature.allowBackslashEscapeCharacter) && matchedToken.image.contains("\\'") ) {
-            matchedToken.image = image.substring( 0, image.indexOf("\\'") + 1 ) + "'";
-            for (int i=0;i<CCJSqlParserConstants.tokenImage.length;i++) {
-              if ( CCJSqlParserConstants.tokenImage[i].equals("<S_CHAR_LITERAL>") ) {
-                  matchedToken.kind = i;
-              }
+        boolean allowEscape = configuration.getAsBoolean(Feature.allowBackslashEscapeCharacter);
+        String img = matchedToken.image;
+        int pos;
+        if (!allowEscape) {
+            pos = indexOfSequence(img, "\\'");
+            if (pos > 0) {
+                matchedToken.image = image.substring(0, pos + 1) + "'";
+                matchedToken.kind = charLiteralIndex;
+                input_stream.backup(image.length() - matchedToken.image.length());
             }
-            input_stream.backup(image.length() - matchedToken.image.length() );
-        } else if ( configuration.getAsBoolean(Feature.allowBackslashEscapeCharacter) && matchedToken.image.contains("\\''") ) {
-            matchedToken.image = image.substring( 0, image.lastIndexOf("\\'") + 3);
-            for (int i=0;i<CCJSqlParserConstants.tokenImage.length;i++) {
-                if ( CCJSqlParserConstants.tokenImage[i].equals("<S_CHAR_LITERAL>") ) {
-                    matchedToken.kind = i;
-                }
+        } else {
+            pos = lastIndexOfSequence(img, "\\''");
+            if (pos > 0) {
+                matchedToken.image = image.substring(0, pos + 3);
+                matchedToken.kind = charLiteralIndex;
+                input_stream.backup(image.length() - matchedToken.image.length());
             }
-            input_stream.backup(image.length() - matchedToken.image.length() );
-      }
+        }
     }
 | < S_QUOTED_IDENTIFIER: "\"" ( "\"\"" | ~["\n","\r","\""])* "\"" | "$$" (~["$"])* "$$" | ("`" (~["\n","\r","`"])+ "`") | ( "[" (~["\n","\r","]"])* "]" ) >
     {
-        if ( !configuration.getAsBoolean(Feature.allowSquareBracketQuotation) && matchedToken.image.charAt(0) == '[' ) {
+        if ( !configuration.getAsBoolean(Feature.allowSquareBracketQuotation)
+            && matchedToken.image.charAt(0) == '[' ) {
+
             matchedToken.image = "[";
-            for (int i=0;i<CCJSqlParserConstants.tokenImage.length;i++) {
-                if (CCJSqlParserConstants.tokenImage[i].equals("\"[\"")) {
-                    matchedToken.kind = i;
-                }
-            }
+            // `squaredBracketOpenIndex` defined in TokenManagerDeclaration above
+            matchedToken.kind = squaredBracketOpenIndex;
             input_stream.backup(image.length() - 1);
         }
     }
@@ -4614,7 +4655,7 @@ Expression RegularCondition() #RegularCondition:
         | "-#" { result = new JsonOperator("-#"); }
         | "<->" { result = new GeometryDistance("<->"); }
         | "<#>" { result = new GeometryDistance("<#>"); }
-        | "<=>" { result = new CosineSimilarity(); }
+        | <OP_COSINESIMILARITY> { result = new CosineSimilarity(); }
     )
 
     ( LOOKAHEAD(2) <K_PRIOR> rightExpression=ComparisonItem() { oraclePrior = EqualsTo.ORACLE_PRIOR_END; }

src/site/sphinx/_static/jmh_results.txt

@@ -17,4 +17,17 @@ JSQLParserBenchmark.parseSQLStatements        5.2  avgt   15  388.453 ± 13.149
 Benchmark                               (version)  Mode  Cnt    Score   Error  Units
 JSQLParserBenchmark.parseSQLStatements     latest  avgt   30   83.504 ± 1.557  ms/op
 JSQLParserBenchmark.parseSQLStatements        5.2  avgt   30  400.876 ± 8.291  ms/op
-JSQLParserBenchmark.parseSQLStatements        5.1  avgt   30   85.731 ± 1.288  ms/op
+JSQLParserBenchmark.parseSQLStatements        5.1  avgt   30   85.731 ± 1.288  ms/op
+
+
+-- Token Manipulation
+Before Optimization                  (version)  Mode  Cnt  Score   Error  Units
+JSQLParserBenchmark.parseQuotedText     latest  avgt   30  0.421 ± 0.008  ms/op
+
+After Optimization                   (version)  Mode  Cnt  Score   Error  Units
+JSQLParserBenchmark.parseQuotedText     latest  avgt   30  0.366 ± 0.009  ms/op
+
+Benchmark                            (version)  Mode  Cnt  Score   Error  Units
+JSQLParserBenchmark.parseQuotedText     latest  avgt   30  0.414 ± 0.003  ms/op
+Benchmark                            (version)  Mode  Cnt  Score   Error  Units
+JSQLParserBenchmark.parseQuotedText     latest  avgt   30  0.418 ± 0.003  ms/op

src/test/java/net/sf/jsqlparser/benchmark/JSQLParserBenchmark.java

@@ -12,6 +12,7 @@
 import net.sf.jsqlparser.parser.CCJSqlParser;
 import net.sf.jsqlparser.statement.Statements;
 import org.openjdk.jmh.annotations.*;
+import org.openjdk.jmh.infra.Blackhole;
 
 import java.io.IOException;
 import java.io.InputStream;
@@ -72,14 +73,27 @@ private Path downloadJsqlparserJar(String version) throws IOException {
     }
 
     @Benchmark
-    public void parseSQLStatements() throws Exception {
+    public void parseSQLStatements(Blackhole blackhole) throws Exception {
         final Statements statements = runner.parseStatements(
                 sqlContent,
                 executorService,
                 (Consumer<CCJSqlParser>) parser -> {
                     // No-op consumer (or you can log/validate each parser if desired)
                 });
-        assert statements.size() == 4;
+        blackhole.consume(statements);
+    }
+
+    @Benchmark
+    public void parseQuotedText(Blackhole blackhole) throws Exception {
+        String sqlStr = "SELECT ('\\'', 'a');\n"
+                + "INSERT INTO recycle_record (a,f) VALUES ('\\'anything', 'abc');\n"
+                + "INSERT INTO recycle_record (a,f) VALUES ('\\'','83653692186728700711687663398101');\n";
+
+        final Statements statements = runner.parseStatements(
+                sqlStr,
+                executorService,
+                (Consumer<CCJSqlParser>) parser -> parser.withBackslashEscapeCharacter(true));
+        blackhole.consume(statements);
     }
 
     @TearDown(Level.Trial)