BAEL-6626 (#14345)

* BALE-6626
Check If a String Contains Non-Alphanumeric Characters in Java

* Move patterns to class level.

---------

Co-authored-by: parthiv39731 <parthiv39731@gmail.com>

core-java-modules/core-java-string-operations-6/src/main/java/com/baeldung/nonalphanumeric/NonAlphaNumRegexChecker.java

@@ -0,0 +1,50 @@
+package com.baeldung.nonalphanumeric;
+
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+
+public class NonAlphaNumRegexChecker {
+
+    private static final Pattern PATTERN_NON_ALPHNUM_ANYLANG = Pattern.compile("[^\\p{IsAlphabetic}\\p{IsDigit}]");
+    private static final Pattern PATTERN_NON_ALPHNUM_USASCII = Pattern.compile("[^a-zA-Z0-9]+");
+    /**
+     * checks if a non-alphanumeric character is present. this method would return true if
+     * it comes across a non english or non US-ASCII letter
+     *
+     * @param str - String to check for special character
+     * @return true if special character found else false
+     */
+    public static boolean isNonAlphanumeric(String str) {
+//      Pattern pattern = Pattern.compile("\\W");  //same as [^a-zA-Z0-9]+
+//      Pattern pattern = Pattern.compile("[^a-zA-Z0-9||\\s]+"); //ignores space
+        Matcher matcher = PATTERN_NON_ALPHNUM_USASCII.matcher(str);
+        return matcher.find();
+    }
+
+    /**
+     * Checks for non-alphanumeric characters from all language scripts
+     *
+     * @param input - String to check for special character
+     * @return true if special character found else false
+     */
+    public static boolean containsNonAlphanumeric(String input) {
+//      Pattern pattern = Pattern.compile("[^\\p{Alnum}]", Pattern.UNICODE_CHARACTER_CLASS); //Binary properties
+        Matcher matcher = PATTERN_NON_ALPHNUM_ANYLANG.matcher(input);
+        return matcher.find();
+    }
+
+    /**
+     * checks for non-alphanumeric character. it returns true if it detects any character other than the
+     * specified script argument. example of script - Character.UnicodeScript.GEORGIAN.name()
+     *
+     * @param input - String to check for special character
+     * @param script - language script
+     * @return true if special character found else false
+     */
+    public static boolean containsNonAlphanumeric(String input, String script) {
+        String regexScriptClass = "\\p{" + "Is" + script + "}";
+        Pattern pattern = Pattern.compile("[^" + regexScriptClass + "\\p{IsDigit}]"); //Binary properties
+        Matcher matcher = pattern.matcher(input);
+        return matcher.find();
+    }
+}

core-java-modules/core-java-string-operations-6/src/main/java/com/baeldung/nonalphanumeric/NonAlphaNumericChecker.java

@@ -0,0 +1,49 @@
+package com.baeldung.nonalphanumeric;
+
+import org.apache.commons.lang3.StringUtils;
+
+public class NonAlphaNumericChecker {
+    /**
+     * Checks for non-alphanumeric characters in any Unicode Script
+     * @param str - String to check for special characters
+     * @return true if special character found else false
+     */
+    public static boolean isNonAlphanumericAnyLangScript(String str) {
+        for (int i = 0; i < str.length(); i++) {
+            char c = str.charAt(i);
+            if (!Character.isLetterOrDigit(c)) {
+                return true;
+            }
+        }
+        return false;
+    }
+
+    /**
+     * checks for special characters,returns false if any character
+     * found other than the script argument
+     * @param str - String to check for special characters
+     * @param script - Language script
+     * @return true if special character found else false
+     */
+    public static boolean isNonAlphanumericInLangScript(String str, String script) {
+
+        for (int i = 0; i < str.length(); i++) {
+            char c = str.charAt(i);
+//            script e.g., Character.UnicodeScript.of(c).toString().equalsIgnoreCase(Character.UnicodeScript.LATIN.toString())
+            if (!Character.UnicodeScript.of(c).toString().equalsIgnoreCase(script)
+                    && !Character.isDigit(c)) {
+                return true;
+            }
+        }
+        return false;
+    }
+
+    /**
+     * checks for special characters in any lang
+     * @param str - String to check for special characters
+     * @return true if special character found else false
+     */
+    public static boolean isNonAlphanumericAnyLangScriptV2(String str) {
+        return !StringUtils.isAlphanumeric(str);
+    }
+}

core-java-modules/core-java-string-operations-6/src/test/java/com/baeldung/nonalphanumeric/NonAlphaNumRegexCheckerUnitTest.java

@@ -0,0 +1,117 @@
+package com.baeldung.nonalphanumeric;
+
+import org.junit.Test;
+
+import static org.junit.Assert.*;
+
+
+public class NonAlphaNumRegexCheckerUnitTest {
+    @Test
+    public void whenStrLatinOrHasNonAlphaNum_ThenRetTrue() {
+        //alphabets with special character
+        String str1 = "W$nder^ful";
+        //digits with special character
+        String str2 = "123$%45";
+        //alphabets and digits with special characters
+        String str3 = "W@nd$r123$%45";
+        //Error message
+        String ERROR_MSG = "Test failed, no alphanumeric char found in ";
+
+        assertTrue(ERROR_MSG + str1, NonAlphaNumRegexChecker.isNonAlphanumeric(str1));
+        assertTrue(ERROR_MSG + str2, NonAlphaNumRegexChecker.isNonAlphanumeric(str2));
+        assertTrue(ERROR_MSG + str3, NonAlphaNumRegexChecker.isNonAlphanumeric(str3));
+    }
+
+    @Test
+    public void whenStrLatinOrHasNoNonAlphaNum_ThenRetFalse() {
+        //only alphabets
+        String str1 = "Wonderful";
+        //only digits
+        String str2 = "12345";
+        //mix of alphabet and digit
+        String str3 = "5Won6der1234";
+        //Error message
+        String ERROR_MSG = "Test failed, non alphanumeric char found in ";
+
+        assertFalse(ERROR_MSG + str1, NonAlphaNumRegexChecker.isNonAlphanumeric(str1));
+        assertFalse(ERROR_MSG + str2, NonAlphaNumRegexChecker.isNonAlphanumeric(str2));
+        assertFalse(ERROR_MSG + str3, NonAlphaNumRegexChecker.isNonAlphanumeric(str3));
+    }
+
+    @Test
+    public void whenStrNonLatinOrHasNonAlphaNum_ThenRetTrue() {
+        //special character in Georgian text
+        String str1 = "##მშვენიერი@";
+        //special character with Turkish  text
+        String str2 = "müthiş#$";
+        //No special character in Georgian text
+        String str3 = "მშვენიერი";
+        //Error message
+        String ERROR_MSG = "Test failed, no alphanumeric char found in ";
+
+        assertTrue(ERROR_MSG + str1, NonAlphaNumRegexChecker.isNonAlphanumeric(str1));
+        assertTrue(ERROR_MSG + str2, NonAlphaNumRegexChecker.isNonAlphanumeric(str2));
+        assertTrue(ERROR_MSG + str3, NonAlphaNumRegexChecker.isNonAlphanumeric(str3));
+    }
+
+    @Test
+    public void whenStrAnyLangOrHasNonAlphaNum_ThenRetTrue() {
+        //special character in Georgian text
+        String str1 = "##მშვენიერი@";
+        //special character with Turkish  text
+        String str2 = "müthiş#$";
+        //Error message
+        String ERROR_MSG = "Test failed, no alphanumeric char found in ";
+
+        assertTrue(ERROR_MSG + str1, NonAlphaNumRegexChecker.containsNonAlphanumeric(str1));
+        assertTrue(ERROR_MSG + str2, NonAlphaNumRegexChecker.containsNonAlphanumeric(str2));
+    }
+
+    @Test
+    public void whenStrAnyLangOrHasNoNonAlphaNum_ThenRetFalse() {
+        //Georgian text with no special char
+        String str1 = "მშვენიერი";
+        //Turkish  text with no special char
+        String str2 = "müthiş";
+        //Latin text with no special char
+        String str3 = "Wonderful";
+        //Error message
+        String ERROR_MSG = "Test failed, no alphanumeric char found in ";
+
+        assertFalse(ERROR_MSG + str1, NonAlphaNumRegexChecker.containsNonAlphanumeric(str1));
+        assertFalse(ERROR_MSG + str2, NonAlphaNumRegexChecker.containsNonAlphanumeric(str2));
+        assertFalse(ERROR_MSG + str3, NonAlphaNumRegexChecker.containsNonAlphanumeric(str3));
+    }
+
+    @Test
+    public void givenLang_whenStrHasDiffLangOrHasNonAlphaNum_ThenRetTrue() {
+        String script1 = Character.UnicodeScript.MYANMAR.name(); //script used for Burmese Lang
+         //text in Burmese with special char
+        String str1 = "အံ့ဩ%စ##ရာ";
+        //special character and english char in Burmese
+        String str2 = "အံ့ADFဩစ%ရာ*^";
+        //English character in Burmese
+        String str3 = "အံ့ဩစTရာWon";
+        //Error message
+        String ERROR_MSG = "Test failed, no alphanumeric char found in ";
+
+        assertTrue(ERROR_MSG + str1, NonAlphaNumRegexChecker.containsNonAlphanumeric(str1, script1));
+        assertTrue(ERROR_MSG + str2, NonAlphaNumRegexChecker.containsNonAlphanumeric(str2, script1));
+        assertTrue(ERROR_MSG + str3, NonAlphaNumRegexChecker.containsNonAlphanumeric(str3, script1));
+    }
+
+    @Test
+    public void givenLang_whenStrHasSameLangOrHasNoNonAlphaNum_ThenRetFalse() {
+        String script1 = Character.UnicodeScript.MYANMAR.name(); //script used for Burmese Lang
+        String script2 = Character.UnicodeScript.GREEK.name();
+        //text in Burmese
+        String str1 = "အံ့ဩစရာ";
+        //text in Greek
+        String str2 = "Εκπληκτικός";
+        //Error message
+        String ERROR_MSG = "Test failed, alphanumeric char found in ";
+
+        assertFalse(ERROR_MSG + str1, NonAlphaNumRegexChecker.containsNonAlphanumeric(str1, script1));
+        assertFalse(ERROR_MSG + str2, NonAlphaNumRegexChecker.containsNonAlphanumeric(str2, script2));
+    }
+}

core-java-modules/core-java-string-operations-6/src/test/java/com/baeldung/nonalphanumeric/NonAlphaNumericCheckerUnitTest.java

@@ -0,0 +1,159 @@
+package com.baeldung.nonalphanumeric;
+
+import org.junit.Test;
+
+import static org.junit.Assert.assertFalse;
+import static org.junit.Assert.assertTrue;
+
+public class NonAlphaNumericCheckerUnitTest {
+    @Test
+    public void whenStrLatinOrHasNonAlphaNum_ThenRetTrue() {
+        String latin = Character.UnicodeScript.LATIN.name();
+        //alphabets with special character
+        String str1 = "W$nder^ful";
+        //digits with special character
+        String str2 = "123$%45";
+        //alphabets and digits with special characters
+        String str3 = "W@nd$r123$%45";
+        //Error message
+        String ERROR_MSG = "Test failed, no alphanumeric char found in ";
+
+        assertTrue(ERROR_MSG + str1, NonAlphaNumericChecker.isNonAlphanumericInLangScript(str1, latin));
+        assertTrue(ERROR_MSG + str2, NonAlphaNumericChecker.isNonAlphanumericInLangScript(str2, latin));
+        assertTrue(ERROR_MSG + str3, NonAlphaNumericChecker.isNonAlphanumericInLangScript(str3, latin));
+    }
+
+    @Test
+    public void whenStrLatinOrHasNoNonAlphaNum_ThenRetFalse() {
+        String latin = Character.UnicodeScript.LATIN.name();
+        //only alphabets
+        String str1 = "Wonderful";
+        //only digits
+        String str2 = "12345";
+        //mix of alphabet and digit
+        String str3 = "5Won6der1234";
+        //Error message
+        String ERROR_MSG = "Test failed, non alphanumeric char found in ";
+
+        assertFalse(ERROR_MSG + str1, NonAlphaNumericChecker.isNonAlphanumericInLangScript(str1, latin));
+        assertFalse(ERROR_MSG + str2, NonAlphaNumericChecker.isNonAlphanumericInLangScript(str2, latin));
+        assertFalse(ERROR_MSG + str3, NonAlphaNumericChecker.isNonAlphanumericInLangScript(str3, latin));
+    }
+
+    @Test
+    public void whenStrNonLatinOrHasNonAlphaNum_ThenRetTrue() {
+        String latin = Character.UnicodeScript.LATIN.name();
+        //special character in Georgian text
+        String str1 = "##მშვენიერი@";
+        //special character with Turkish  text
+        String str2 = "müthiş#$";
+        //No special character in Georgian text
+        String str3 = "მშვენიერი";
+        //Error message
+        String ERROR_MSG = "Test failed, no alphanumeric char found in ";
+
+        assertTrue(ERROR_MSG + str1, NonAlphaNumericChecker.isNonAlphanumericInLangScript(str1, latin));
+        assertTrue(ERROR_MSG + str2, NonAlphaNumericChecker.isNonAlphanumericInLangScript(str2, latin));
+        assertTrue(ERROR_MSG + str3, NonAlphaNumericChecker.isNonAlphanumericInLangScript(str3, latin));
+    }
+
+    @Test
+    public void whenStrAnyLangOrHasNonAlphaNum_ThenRetTrue() {
+        //special character in Georgian text
+        String str1 = "##მშვენიერი@";
+        //special character with Turkish  text
+        String str2 = "müthiş#$";
+        //Error message
+        String ERROR_MSG = "Test failed, no alphanumeric char found in ";
+
+        assertTrue(ERROR_MSG + str1, NonAlphaNumericChecker.isNonAlphanumericAnyLangScript(str1));
+        assertTrue(ERROR_MSG + str2, NonAlphaNumericChecker.isNonAlphanumericAnyLangScript(str2));
+    }
+
+    @Test
+    public void whenStrAnyLangOrHasNoNonAlphaNum_ThenRetFalse() {
+        //Georgian text with no special char
+        String str1 = "მშვენიერი";
+        //Turkish  text with no special char
+        String str2 = "müthiş";
+        //Latin text with no special char
+        String str3 = "Wonderful";
+        //Error message
+        String ERROR_MSG = "Test failed, no alphanumeric char found in ";
+
+        assertFalse(ERROR_MSG + str1, NonAlphaNumericChecker.isNonAlphanumericAnyLangScript(str1));
+        assertFalse(ERROR_MSG + str2, NonAlphaNumericChecker.isNonAlphanumericAnyLangScript(str2));
+        assertFalse(ERROR_MSG + str3, NonAlphaNumericChecker.isNonAlphanumericAnyLangScript(str3));
+    }
+
+    @Test
+    public void givenLang_whenStrHasDiffLangOrHasNonAlphaNum_ThenRetTrue() {
+        String script1 = Character.UnicodeScript.MYANMAR.name(); //script used for Burmese Lang
+        //text in Burmese with special char
+        String str1 = "အံ့ဩ%စ##ရာ";
+        //special character and english char in Burmese
+        String str2 = "အံ့ADFဩစ%ရာ*^";
+        //English character in Burmese
+        String str3 = "အံ့ဩစTရာWon";
+        //Error message
+        String ERROR_MSG = "Test failed, no alphanumeric char found in ";
+
+        assertTrue(ERROR_MSG + str1, NonAlphaNumericChecker.isNonAlphanumericInLangScript(str1, script1));
+        assertTrue(ERROR_MSG + str2, NonAlphaNumericChecker.isNonAlphanumericInLangScript(str2, script1));
+        assertTrue(ERROR_MSG + str3, NonAlphaNumericChecker.isNonAlphanumericInLangScript(str3, script1));
+    }
+
+    @Test
+    public void givenLang_whenStrHasSameLangOrHasNoNonAlphaNum_ThenRetFalse() {
+        String script1 = Character.UnicodeScript.MYANMAR.name(); //script used for Burmese Lang
+        String script2 = Character.UnicodeScript.GREEK.name();
+        //text in Burmese
+        String str1 = "အံ့ဩစရာ";
+        //text in Greek
+        String str2 = "Εκπληκτικός";
+        //Error message
+        String ERROR_MSG = "Test failed, alphanumeric char found in ";
+
+        assertFalse(ERROR_MSG + str1, NonAlphaNumericChecker.isNonAlphanumericInLangScript(str1, script1));
+        assertFalse(ERROR_MSG + str2, NonAlphaNumericChecker.isNonAlphanumericInLangScript(str2, script2));
+    }
+
+    @Test
+    public void givenComLangImpl_whenStrAnyLangOrHasNonAlphaNum_ThenRetTrue() {
+        //special character in Georgian text
+        String str1 = "##მშვენიერი@";
+        //special character with Turkish  text
+        String str2 = "müthiş#$";
+        //Error message
+        String ERROR_MSG = "Test failed, no alphanumeric char found in ";
+
+        assertTrue(ERROR_MSG + str1, NonAlphaNumericChecker.isNonAlphanumericAnyLangScriptV2(str1));
+        assertTrue(ERROR_MSG + str2, NonAlphaNumericChecker.isNonAlphanumericAnyLangScriptV2(str2));
+    }
+
+    @Test
+    public void givenComLangImpl_whenStrAnyLangOrHasNoNonAlphaNum_ThenRetFalse() {
+        //Georgian text with no special char
+        String str1 = "მშვენიერი";
+        //Turkish  text with no special char
+        String str2 = "müthiş";
+        //Latin text with no special char
+        String str3 = "Wonderful";
+        //Burmese in Myanmar script
+//        String str4 = "အံ့ဩစရာ";
+        //only digits
+        String str5 = "3465";
+        //Error message
+        String ERROR_MSG = "Test failed, alphanumeric char found in ";
+
+        assertFalse(ERROR_MSG + str1, NonAlphaNumericChecker.isNonAlphanumericAnyLangScriptV2(str1));
+        assertFalse(ERROR_MSG + str2, NonAlphaNumericChecker.isNonAlphanumericAnyLangScriptV2(str2));
+        assertFalse(ERROR_MSG + str3, NonAlphaNumericChecker.isNonAlphanumericAnyLangScriptV2(str3));
+        //StringUtils.isAlphanumeric is unable to support supplementary unicode character and hence fails
+        //assertFalse(ERROR_MSG + str4, NonAlphaNumericChecker.isNonAlphanumericAnyLangScriptV2(str4));
+        assertFalse(ERROR_MSG + str5, NonAlphaNumericChecker.isNonAlphanumericAnyLangScriptV2(str5));
+    }
+}
+
+
+