support \p{L} Unicode property escapes (issue #583)

rbri · rbri · commit 776d1a4607f5 · 2023-04-26T10:01:32.000+02:00
diff --git a/src/main/java/org/htmlunit/javascript/regexp/RegExpJsToJavaConverter.java b/src/main/java/org/htmlunit/javascript/regexp/RegExpJsToJavaConverter.java
@@ -17,6 +17,7 @@
 import java.util.ArrayDeque;
 import java.util.ArrayList;
 import java.util.Deque;
+import java.util.HashMap;
 import java.util.List;
 
 /**
@@ -37,13 +38,20 @@
 public class RegExpJsToJavaConverter {
 
     private static final String DIGITS = "0123456789";
+    private static final HashMap<String, String> UNICODE_ESCAPES;
 
     private Tape tape_;
     private boolean insideCharClass_;
     private boolean insideRepetition_;
     private Deque<Subexpresion> parsingSubexpressions_;
     private List<Subexpresion> subexpressions_;
 
+    static {
+        UNICODE_ESCAPES = new HashMap<>();
+        UNICODE_ESCAPES.put("L", "L");
+        UNICODE_ESCAPES.put("Letter", "L");
+    }
+
     /**
      * Helper to encapsulate the transformations.
      */
@@ -373,6 +381,34 @@ private void processEscapeSequence() {
             return;
         }
 
+        // Unicode property escape
+        if ('p' == escapeSequence) {
+            int next = tape_.read();
+            if (next > -1) {
+                if (next == '{') {
+                    final int uPos = tape_.currentPos_;
+                    do {
+                        next = tape_.read();
+                    }
+                    while (next > -1 && next != '}');
+                    if (next == '}') {
+                        final String escape = tape_.tape_.substring(uPos, tape_.currentPos_ - 1);
+                        final String replace = UNICODE_ESCAPES.get(escape);
+                        if (replace != null) {
+                            tape_.tape_.replace(uPos, uPos + escape.length(), replace);
+                            return;
+                        }
+                    }
+
+                    // back to the old behavior
+                    tape_.move(uPos - tape_.currentPos_ - 3);
+                    tape_.remove(1);
+                }
+            }
+
+            return;
+        }
+
         if ("ACEFGHIJKLMNOPQRTUVXYZaeghijklmpqyz".indexOf(escapeSequence) > -1) {
             // no need to escape this chars
             tape_.move(-2);
diff --git a/src/test/java/org/htmlunit/javascript/regexp/RegExpJsToJavaConverter2Test.java b/src/test/java/org/htmlunit/javascript/regexp/RegExpJsToJavaConverter2Test.java
@@ -86,6 +86,16 @@ public void validationPatternUnicodeCodePointEscapesFails() throws Exception {
         validation("123\\u{1D306}", "123&#x1D307;");
     }
 
+    /**
+     * @throws Exception if an error occurs
+     */
+    @Test
+    @Alerts("true")
+    public void validationPatternUnicodePropertyEscapeL() throws Exception {
+        validation("\\p{L}*", "Html");
+        validation("\\p{L}*", "&#x043C&#x0439&#x0440");
+    }
+
     private void validation(final String pattern, final String value) throws Exception {
         final String html =
                 "<html><head>\n"
diff --git a/src/test/java/org/htmlunit/javascript/regexp/RegExpJsToJavaConverterTest.java b/src/test/java/org/htmlunit/javascript/regexp/RegExpJsToJavaConverterTest.java
@@ -383,4 +383,17 @@ public void unicode() {
         assertEquals("\\u{FFFFD", regExpJsToJavaConverter.convert("\\u{FFFFD"));
         assertEquals("\\x{FFFFD}\\}", regExpJsToJavaConverter.convert("\\u{FFFFD}}"));
     }
+
+    /**
+     * Unicode property escapes.
+     */
+    @Test
+    public void unicodePropertyEscapes() {
+        final RegExpJsToJavaConverter regExpJsToJavaConverter = new RegExpJsToJavaConverter();
+
+        assertEquals("\\p{L}0-9", regExpJsToJavaConverter.convert("\\p{L}0-9"));
+        assertEquals("\\p{L}0-9", regExpJsToJavaConverter.convert("\\p{Letter}0-9"));
+
+        assertEquals("p\\{html\\}0-9", regExpJsToJavaConverter.convert("\\p{html}0-9"));
+    }
 }
