[Java] Improve expression creation performance (#185, #187)

Author: jkronegg

CHANGELOG.md

@@ -6,6 +6,8 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/)
 and this project adheres to [Semantic Versioning](http://semver.org/).
 
 ## [Unreleased]
+### Fixed
+- [Java] Improve expression creation performance ([#187](https://github.com/cucumber/cucumber-expressions/pull/187), [#189](https://github.com/cucumber/cucumber-expressions/pull/189))
 
 ## [16.1.0] - 2022-11-28
 ### Added

java/src/main/java/io/cucumber/cucumberexpressions/CucumberExpression.java

@@ -24,7 +24,7 @@
 
 @API(status = API.Status.STABLE)
 public final class CucumberExpression implements Expression {
-    private static final Pattern ESCAPE_PATTERN = Pattern.compile("([\\\\^\\[({$.|?*+})\\]])");
+    private static final Pattern ESCAPE_PATTERN = Pattern.compile("[\\\\^\\[({$.|?*+})\\]]");
 
     private final List<ParameterType<?>> parameterTypes = new ArrayList<>();
     private final String source;
@@ -62,7 +62,7 @@ private String rewriteToRegex(Node node) {
     }
 
     private static String escapeRegex(String text) {
-        return ESCAPE_PATTERN.matcher(text).replaceAll("\\\\$1");
+        return ESCAPE_PATTERN.matcher(text).replaceAll("\\\\$0");
     }
 
     private String rewriteOptional(Node node) {

java/src/main/java/io/cucumber/cucumberexpressions/ExpressionFactory.java

@@ -11,15 +11,13 @@
  * using heuristics. This is particularly useful for languages that don't have a
  * literal syntax for regular expressions. In Java, a regular expression has to be represented as a String.
  *
- *  A string that starts with `^` and/or ends with `$` is considered a regular expression.
+ *  A string that starts with `^` and/or ends with `$` (or written in script style, i.e. starting with `/` 
+ *  and ending with `/`) is considered a regular expression.
  *  Everything else is considered a Cucumber expression.
  */
 @API(status = API.Status.STABLE)
 public final class ExpressionFactory {
 
-    private static final Pattern BEGIN_ANCHOR = Pattern.compile("^\\^.*");
-    private static final Pattern END_ANCHOR = Pattern.compile(".*\\$$");
-    private static final Pattern SCRIPT_STYLE_REGEXP = Pattern.compile("^/(.*)/$");
     private static final Pattern PARAMETER_PATTERN = Pattern.compile("((?:\\\\){0,2})\\{([^}]*)\\}");
 
     private final ParameterTypeRegistry parameterTypeRegistry;
@@ -29,14 +27,30 @@ public ExpressionFactory(ParameterTypeRegistry parameterTypeRegistry) {
     }
 
     public Expression createExpression(String expressionString) {
-        if (BEGIN_ANCHOR.matcher(expressionString).find() || END_ANCHOR.matcher(expressionString).find()) {
-            return createRegularExpressionWithAnchors(expressionString);
+        /* This method is called often (typically about number_of_steps x
+         * nbr_test_scenarios), thus performance is more important than
+         * readability here.
+         * Consequently, we check the first and last expressionString
+         * characters to determine whether we need to create a
+         * RegularExpression or a CucumberExpression (because character
+         * matching is faster than startsWith/endsWith and regexp matching).
+         */
+        int length = expressionString.length();
+        if (length == 0) {
+            return new CucumberExpression(expressionString, this.parameterTypeRegistry);
         }
-        Matcher m = SCRIPT_STYLE_REGEXP.matcher(expressionString);
-        if (m.find()) {
-            return new RegularExpression(Pattern.compile(m.group(1)), parameterTypeRegistry);
+
+        int lastCharIndex = length - 1;
+        char firstChar = expressionString.charAt(0);
+        char lastChar = expressionString.charAt(lastCharIndex);
+
+        if (firstChar == '^' || lastChar == '$') {
+            return this.createRegularExpressionWithAnchors(expressionString);
+        } else if (firstChar == '/' && lastChar == '/') {
+            return new RegularExpression(Pattern.compile(expressionString.substring(1, lastCharIndex)), this.parameterTypeRegistry);
         }
-        return new CucumberExpression(expressionString, parameterTypeRegistry);
+
+        return new CucumberExpression(expressionString, this.parameterTypeRegistry);
     }
 
     private RegularExpression createRegularExpressionWithAnchors(String expressionString) {

java/src/test/java/io/cucumber/cucumberexpressions/ExpressionFactoryTest.java

@@ -13,7 +13,12 @@
 import static org.junit.jupiter.api.Assertions.assertThrows;
 
 public class ExpressionFactoryTest {
-
+    
+    @Test
+    public void creates_cucumber_expression_for_empty() {
+        assertCucumberExpression("");
+    }
+    
     @Test
     public void creates_cucumber_expression_by_default() {
         assertCucumberExpression("strings are cukexp by default");