From d7f545721bec93da1cdfb6abe096c83189dd8802 Mon Sep 17 00:00:00 2001
From: Max Kellermann <max.kellermann@gmail.com>
Date: Tue, 15 Nov 2022 17:44:54 +0100
Subject: [PATCH] lib/icu/Canonicalize: use unorm2_normalize() instead of
 u_strFoldCase()

unorm2_normalize() can case-fold as well, plus it applies Unicode
normalization which MPD should do for proper string comparisons.
---
 NEWS                         |  1 +
 doc/protocol.rst             |  4 +++-
 src/lib/icu/Canonicalize.cxx | 11 ++++++-----
 test/TestStringFilter.cxx    | 38 ++++++++++++++++++++++++++++++++++++
 4 files changed, 48 insertions(+), 6 deletions(-)

diff --git a/NEWS b/NEWS
index 6ea828dfc..b7b96af42 100644
--- a/NEWS
+++ b/NEWS
@@ -6,6 +6,7 @@ ver 0.24 (not yet released)
   - limit "player" idle events to the current partition
   - operator "starts_with"
   - show PCRE support in "config" response
+  - apply Unicode normalization to case-insensitive filter expressions
 * archive
   - add option to disable archive plugins in mpd.conf
 * input
diff --git a/doc/protocol.rst b/doc/protocol.rst
index 9d8125698..8fde63888 100644
--- a/doc/protocol.rst
+++ b/doc/protocol.rst
@@ -235,7 +235,9 @@ of:
   (album == 'BAR'))`
 
 The :command:`find` commands are case sensitive, while
-:command:`search` and related commands ignore case.
+:command:`search` and related commands ignore case.  The latter also
+applies `Unicode normalization <https://unicode.org/reports/tr15/>`__
+if MPD was compiled with `ICU <https://icu.unicode.org/>`__ support.
 
 Prior to MPD 0.21, the syntax looked like this::
 
diff --git a/src/lib/icu/Canonicalize.cxx b/src/lib/icu/Canonicalize.cxx
index 8f97cb9f5..08528c592 100644
--- a/src/lib/icu/Canonicalize.cxx
+++ b/src/lib/icu/Canonicalize.cxx
@@ -25,7 +25,7 @@
 #include "util/AllocatedString.hxx"
 
 #ifdef HAVE_ICU
-#include "FoldCase.hxx"
+#include "Normalize.hxx"
 #include "Util.hxx"
 #include "util/AllocatedArray.hxx"
 #include "util/SpanCast.hxx"
@@ -39,10 +39,11 @@ try {
 	if (u.data() == nullptr)
 		return {src};
 
-	if (fold_case)
-		if (auto folded = IcuFoldCase(ToStringView(std::span{u}));
-		    folded != nullptr)
-			u = std::move(folded);
+	if (auto n = fold_case
+	    ? IcuNormalizeCaseFold(ToStringView(std::span{u}))
+	    : IcuNormalize(ToStringView(std::span{u}));
+	    n != nullptr)
+		u = std::move(n);
 
 	return UCharToUTF8(ToStringView(std::span{u}));
 #else
diff --git a/test/TestStringFilter.cxx b/test/TestStringFilter.cxx
index 80ac38122..4b526759b 100644
--- a/test/TestStringFilter.cxx
+++ b/test/TestStringFilter.cxx
@@ -97,6 +97,10 @@ TEST_F(StringFilterTest, Latin)
 	const StringFilter f{"nëedlé", false, false, false, false};
 
 	EXPECT_TRUE(f.Match("nëedlé"));
+#if defined(HAVE_ICU) || defined(_WIN32)
+	EXPECT_TRUE(f.Match("nëedl\u00e9"));
+	// TODO EXPECT_TRUE(f.Match("nëedl\u0065\u0301"));
+#endif
 	EXPECT_FALSE(f.Match("NËEDLÉ"));
 	EXPECT_FALSE(f.Match("needlé"));
 	EXPECT_FALSE(f.Match("néedlé"));
@@ -109,13 +113,47 @@ TEST_F(StringFilterTest, Latin)
 	EXPECT_FALSE(f.Match("FOOnëedleBAR"));
 }
 
+#if defined(HAVE_ICU) || defined(_WIN32)
+
+TEST_F(StringFilterTest, Normalize)
+{
+	const StringFilter f{"1①H", true, false, false, false};
+
+	EXPECT_TRUE(f.Match("1①H"));
+	EXPECT_TRUE(f.Match("¹₁H"));
+	EXPECT_TRUE(f.Match("①1ℌ"));
+	EXPECT_TRUE(f.Match("①1ℍ"));
+	EXPECT_FALSE(f.Match("21H"));
+
+#ifndef _WIN32
+	// fails with Windows CompareStringEx()
+	EXPECT_TRUE(StringFilter("dž", true, false, false, false).Match("dž"));
+#endif
+
+	EXPECT_TRUE(StringFilter("\u212b", true, false, false, false).Match("\u0041\u030a"));
+	EXPECT_TRUE(StringFilter("\u212b", true, false, false, false).Match("\u00c5"));
+
+	EXPECT_TRUE(StringFilter("\u1e69", true, false, false, false).Match("\u0073\u0323\u0307"));
+
+#ifndef _WIN32
+	// fails with Windows CompareStringEx()
+	EXPECT_TRUE(StringFilter("\u1e69", true, false, false, false).Match("\u0073\u0307\u0323"));
+#endif
+}
+
+#endif
+
 TEST_F(StringFilterTest, FoldCase)
 {
 	const StringFilter f{"nëedlé", true, false, false, false};
 
 	EXPECT_TRUE(f.Match("nëedlé"));
 #if defined(HAVE_ICU) || defined(_WIN32)
+	EXPECT_TRUE(f.Match("nëedl\u00e9"));
+	EXPECT_TRUE(f.Match("nëedl\u0065\u0301"));
 	EXPECT_TRUE(f.Match("NËEDLÉ"));
+	EXPECT_TRUE(f.Match("NËEDL\u00c9"));
+	EXPECT_TRUE(f.Match("NËEDL\u0045\u0301"));
 #endif
 	EXPECT_FALSE(f.Match("needlé"));
 	EXPECT_FALSE(f.Match("néedlé"));