JCR-5152 Add method to check if a (local) name is valid according to JCR

spec

Author: kwin

jackrabbit-jcr-commons/src/main/java/org/apache/jackrabbit/util/Text.java

@@ -24,6 +24,7 @@
 import java.util.ArrayList;
 import java.util.BitSet;
 import java.util.Properties;
+import java.util.Set;
 
 /**
  * This Class provides some text related utilities
@@ -41,6 +42,8 @@ private Text() {
      */
     public static final char[] hexTable = "0123456789abcdef".toCharArray();
 
+    private static final Set<Character> INVALID_JCR_LOCAL_NAME_CHARS = Set.of( '/', ':', '[', ']', '*', '|');
+
     /**
      * Calculate an MD5 hash of the string given.
      *
@@ -466,6 +469,10 @@ public static String unescape(String string) {
      * char ::= nonspace | ' '
      * nonspace ::= (* Any Unicode character except: '/', ':', '[', ']', '*', '|' or any whitespace character *)
      * </pre>
+     * <p>
+     * Note that just using this method does not necessarily return a string which is a 
+     * <a href="https://s.apache.org/jcr-2.0-spec/3_Repository_Model.html#3.2.2%20Local%20Names">valid local name</a>.
+     * You still have to take care of invalid <a href="https://www.w3.org/TR/xml/#NT-Char">XML characters</a>.
      *
      * @param name the name to escape
      * @return the escaped name
@@ -567,6 +574,31 @@ public static String unescapeIllegalJcrChars(String name) {
         return buffer.toString();
     }
 
+    /**
+     * Checks if the given name is a valid JCR local name.
+     * <p>
+     * Note that the return value of {@link #escapeIllegalJcrChars(String)} is not necessarily a valid local name.
+     * You still have to take care of invalid <a href="https://www.w3.org/TR/xml/#NT-Char">XML characters</a>.
+     * 
+     * @param localName the string value to check
+     * @return <code>true</code> if the name is valid, <code>false</code> otherwise.
+     * @see <a href="https://s.apache.org/jcr-2.0-spec/3_Repository_Model.html#3.2.2%20Local%20Names">JCR 2.0 Spec, §3.2.2 Local Names</a>
+     * @see #escapeIllegalJcrChars(String)
+     * @since 2.6.0 (Apache Jackrabbit 2.24.0)
+     */
+    public static boolean isValidJcrLocalName(String localName) {
+        if (localName == null || localName.isEmpty()) {
+            return false;
+        }
+        // self or parent are invalid
+        if (localName.equals(".") || localName.equals("..")) {
+            return false;
+        }
+        return localName.chars().noneMatch(c -> 
+            INVALID_JCR_LOCAL_NAME_CHARS.contains((char) c) || !XMLChar.isValid(c)
+        );
+    }
+
     /**
      * Returns the name part of the path. If the given path is already a name
      * (i.e. contains no slashes) it is returned.

jackrabbit-jcr-commons/src/main/java/org/apache/jackrabbit/util/package-info.java

@@ -14,5 +14,5 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
-@org.osgi.annotation.versioning.Version("2.5.0")
+@org.osgi.annotation.versioning.Version("2.6.0")
 package org.apache.jackrabbit.util;

jackrabbit-jcr-commons/src/test/java/org/apache/jackrabbit/util/TextTest.java

@@ -202,4 +202,14 @@ public void testEscapeXML() {
     public void testEscapeHTML() {
         assertEquals("&amp;&lt;&gt;&#39;&quot;", Text.encodeIllegalHTMLCharacters("&<>'\""));
     }
+
+    public void testIsValidJcrLocalName() {
+        assertTrue(Text.isValidJcrLocalName("valid%Name..\n\r test.\"'"));
+        assertFalse(Text.isValidJcrLocalName("invalid|name"));
+        assertFalse(Text.isValidJcrLocalName("some:name"));
+        // containing non XML characters (unicode control character)
+        assertFalse(Text.isValidJcrLocalName("\u000FinvalidName"));
+        assertFalse(Text.isValidJcrLocalName(".."));
+        assertFalse(Text.isValidJcrLocalName("."));
+    }
 }