Add arabic normalizer for search

OsmAnd-java/src/main/java/net/osmand/CollatorStringMatcher.java

@@ -1,5 +1,7 @@
 package net.osmand;
 
+import net.osmand.util.ArabicNormalizer;
+
 import java.util.Locale;
 
 
@@ -55,9 +57,17 @@ public Collator getCollator() {
 	public boolean matches(String name) {
 		return cmatches(collator, name, part, mode);
 	}
-
-
+
 	public static boolean cmatches(Collator collator, String fullName, String part, StringMatcherMode mode){
+		String withoutDiacritic = ArabicNormalizer.normalize(fullName);
+		boolean matchDiacritic = false;
+		if (!fullName.equals(withoutDiacritic)) {
+			matchDiacritic = cmatchInternal(collator, withoutDiacritic, part, mode);
+		}
+		return matchDiacritic || cmatchInternal(collator, fullName, part, mode);
+	}
+
+	private static boolean cmatchInternal(Collator collator, String fullName, String part, StringMatcherMode mode){
 		switch (mode) {
 		case CHECK_CONTAINS:
 			return ccontains(collator, fullName, part); 

OsmAnd-java/src/main/java/net/osmand/search/core/SearchPhrase.java

@@ -16,6 +16,7 @@
 import net.osmand.data.QuadRect;
 import net.osmand.osm.AbstractPoiType;
 import net.osmand.util.Algorithms;
+import net.osmand.util.ArabicNormalizer;
 import net.osmand.util.LocationParser;
 import net.osmand.util.MapUtils;
 
@@ -802,7 +803,12 @@ public boolean matches(Collection<String> map) {
 
 		@Override
 		public boolean matches(String name) {
-			return sm.matches(name);
+			String normalized = ArabicNormalizer.normalize(name);
+			boolean matchDiacritic = false;
+			if (!name.equals(normalized)) {
+				matchDiacritic = sm.matches(normalized);
+			}
+			return matchDiacritic || sm.matches(name);
 		}
 
 	}

OsmAnd-java/src/main/java/net/osmand/util/ArabicNormalizer.java

@@ -0,0 +1,35 @@
+package net.osmand.util;
+
+import java.text.Normalizer;
+import java.util.regex.Pattern;
+
+public class ArabicNormalizer {
+
+    private static final Pattern DIACRITICS_PATTERN = Pattern.compile("\\p{Mn}");
+
+    public static String normalize(String text) {
+        if (text == null) {
+            return null;  // Handle null input
+        }
+
+        String normalized = Normalizer.normalize(text, Normalizer.Form.NFD);
+        normalized = DIACRITICS_PATTERN.matcher(normalized).replaceAll(""); // Remove diacritics efficiently
+
+        // Hamza variations
+        normalized = normalized.replace("إ", "ا"); // Initial hamza on alif
+        normalized = normalized.replace("أ", "ا"); // Initial hamza on waw
+        normalized = normalized.replace("ئ", "ي"); // Hamza on ya' (This should be 'ي' not 'ا' for better accuracy)
+        normalized = normalized.replace("ؤ", "و"); // Hamza on waw
+
+        // Other normalizations
+        normalized = normalized.replace("آ", "ا");  // Alif madda
+        normalized = normalized.replace("ى", "ي"); // Final form of ya'
+        normalized = normalized.replace("ة", "ه"); // Teh marbuta to ha'
+
+        // Kashida
+        normalized = normalized.trim().replaceAll("\u0640", "");// Kashida
+
+        return normalized;
+    }
+
+}