java: Improve text trimming performance (#85)

Precompiled the Pattern used in the utility methods.

Improves performance. `StringUtils.trim()` is more than 3 times faster.

Fixes: #84

CHANGELOG.md

@@ -9,6 +9,9 @@ This document is formatted according to the principles of [Keep A CHANGELOG](htt
 ## [Unreleased]
 - [Javascript, PHP, Ruby] Allow messages v19.1.4 to v20.0.1 to be used
 
+### Fixed
+- [Java] Improve text trim performance ([#84](https://github.com/cucumber/gherkin/issues/84))
+
 ## [26.0.2] - 2022-12-27
 ### Fixed
 - [Go] Fix module names (again)

java/src/main/java/io/cucumber/gherkin/StringUtils.java

@@ -1,28 +1,36 @@
 package io.cucumber.gherkin;
 
+import java.util.regex.Pattern;
+
 class StringUtils {
 
+    private static final Pattern LTRIM = Pattern.compile("^[ \\t\\n\\x0B\\f\\r\\x85\\xA0]+");
+    private static final Pattern LTRIM_KEEP_NEW_LINES = Pattern.compile("^[ \\t\\x0B\\f\\r\\x85\\xA0]+");
+    private static final Pattern RTRIM_KEEP_NEW_LINES = Pattern.compile("[ \\t\\x0B\\f\\r\\x85\\xA0]+$");
+    private static final Pattern RTRIM = Pattern.compile("[ \\t\\n\\x0B\\f\\r\\x85\\xA0]+$");
+    private static final Pattern TRIM = Pattern.compile("^[ \\t\\n\\x0B\\f\\r\\x85\\xA0]+|[ \\t\\n\\x0B\\f\\r\\x85\\xA0]+$");
+
     static String ltrim(String s) {
         // https://stackoverflow.com/questions/1060570/why-is-non-breaking-space-not-a-whitespace-character-in-java
-        return s.replaceAll("^[ \\t\\n\\x0B\\f\\r\\x85\\xA0]+", "");
+        return LTRIM.matcher(s).replaceAll("");
     }
 
     static String ltrimKeepNewLines(String s) {
         // https://stackoverflow.com/questions/1060570/why-is-non-breaking-space-not-a-whitespace-character-in-java
-        return s.replaceAll("^[ \\t\\x0B\\f\\r\\x85\\xA0]+", "");
+        return LTRIM_KEEP_NEW_LINES.matcher(s).replaceAll("");
     }
 
     static String rtrimKeepNewLines(String s) {
         // https://stackoverflow.com/questions/1060570/why-is-non-breaking-space-not-a-whitespace-character-in-java
-        return s.replaceAll("[ \\t\\x0B\\f\\r\\x85\\xA0]+$", "");
+        return RTRIM_KEEP_NEW_LINES.matcher(s).replaceAll("");
     }
 
     static String rtrim(String s) {
-        return s.replaceAll("[ \\t\\n\\x0B\\f\\r\\x85\\xA0]+$", "");
+        return RTRIM.matcher(s).replaceAll("");
     }
 
     static String trim(String s) {
-        return ltrim(rtrim(s));
+        return TRIM.matcher(s).replaceAll("");
     }
 
     static int symbolCount(String string) {

java/src/test/java/io/cucumber/gherkin/StringUtilsTest.java

@@ -0,0 +1,60 @@
+package io.cucumber.gherkin;
+
+import org.junit.jupiter.api.Disabled;
+import org.junit.jupiter.api.Test;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+
+public class StringUtilsTest {
+    private static final String WHITESPACE = "\u00A0 \t";
+    private static final String CUCUMBER = "🥒";
+
+    @Test
+    void testRtrim() {
+        assertEquals(WHITESPACE + CUCUMBER, StringUtils.rtrim(WHITESPACE + CUCUMBER + WHITESPACE));
+    }
+
+    @Test
+    void testRtrim_multiline() {
+        assertEquals("\n" + WHITESPACE + "\n" + WHITESPACE + CUCUMBER,
+                StringUtils.rtrim("\n" + WHITESPACE + "\n" + WHITESPACE + CUCUMBER + WHITESPACE + "\n" + WHITESPACE + "\n"));
+    }
+
+    @Test
+    void testRtrimKeepNewlines() {
+        assertEquals(WHITESPACE + CUCUMBER + "\n", StringUtils.rtrimKeepNewLines(WHITESPACE + CUCUMBER + "\n" + WHITESPACE));
+    }
+
+    @Test
+    void testLtrim() {
+        assertEquals(CUCUMBER + WHITESPACE, StringUtils.ltrim(WHITESPACE + CUCUMBER + WHITESPACE));
+    }
+
+    @Test
+    void testLtrim_multiline() {
+        assertEquals(CUCUMBER + WHITESPACE + "\n" + WHITESPACE + "\n",
+                StringUtils.ltrim("\n" + WHITESPACE + "\n" + WHITESPACE + CUCUMBER + WHITESPACE + "\n" + WHITESPACE + "\n"));
+    }
+
+    @Test
+    void testLtrimKeepNewlines() {
+        assertEquals("\n" + CUCUMBER + WHITESPACE, StringUtils.ltrimKeepNewLines(WHITESPACE + "\n" + CUCUMBER + WHITESPACE));
+    }
+
+    @Test
+    void testTrim() {
+        assertEquals(CUCUMBER, StringUtils.trim(WHITESPACE + CUCUMBER + WHITESPACE));
+    }
+
+    @Test
+    void testTrim_multiline() {
+        assertEquals(CUCUMBER,
+                StringUtils.trim("\n" + WHITESPACE + "\n" + WHITESPACE + CUCUMBER + WHITESPACE + "\n" + WHITESPACE + "\n"));
+    }
+
+    @Test
+    void testTrim_empty() {
+        assertEquals("",
+                StringUtils.trim(""));
+    }
+}