From 1d50a3e24e254c965b07d8857d4edb162afac028 Mon Sep 17 00:00:00 2001
From: MM20 <15646950+MM2-0@users.noreply.github.com>
Date: Sat, 15 Oct 2022 15:58:22 +0200
Subject: [PATCH] Separate ligatures in search string matching

---
 ktx/src/main/java/de/mm20/launcher2/ktx/String.kt | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)

diff --git a/ktx/src/main/java/de/mm20/launcher2/ktx/String.kt b/ktx/src/main/java/de/mm20/launcher2/ktx/String.kt
index 62af5606..f003659a 100644
--- a/ktx/src/main/java/de/mm20/launcher2/ktx/String.kt
+++ b/ktx/src/main/java/de/mm20/launcher2/ktx/String.kt
@@ -10,16 +10,22 @@ fun String.decodeUrl(charset: String): String? {
 }
 
 /**
- * Normalize a string to lowercase ASCII
- * TODO: Only supports Chinese/Pinyin at the moment
+ * Normalize a string to lowercase string
+ * This is used for substring matching.
+ * Characters must be normalized independently so that
+ * A.contains(B) -> A.normalize().contains(B.normalize()) is true.
  */
 fun String.normalize(): String {
     return StringUtils.stripAccents(this.romanize().lowercase(Locale.getDefault()))
+        .replace("æ", "ae")
+        .replace("œ", "oe")
+        .replace("ß", "ss")
 }
 
 /**
  * Romanize a string, transliterate non-latin characters into latin
- * TODO: Only supports Chinese/Pinyin at the moment
+ * This is used for sorting. The resulting string represents the position where the original
+ * string should be sorted in the latin alphabet.
  */
 fun String.romanize(): String {
     return Pinyin.toPinyin(this, "")