fix composition (C/KC) of some sequences

an unassigned code point should be treated like a starter code point;
thus, if we have met an unassigned code point in the currently processed
sequence, we should start process the next one (but we know there are
no composed character for an unassigned starter -> simply skip it)

This patch doesn't apply to qtbase, commit 46b78113b22428e6f8540193fcf0e00591dbd724
solved the same issue in a different way.

Change-Id: I781a26024878d8b83a182b0fd4e681be2a6d9a27
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>

2 files changed, 15 insertions, 2 deletions

diff --git a/src/corelib/tools/qchar.cpp b/src/corelib/tools/qchar.cpp
index e5d7c82d29..017c0f232e 100644
--- a/src/corelib/tools/qchar.cpp
+++ b/src/corelib/tools/qchar.cpp
@@ -1546,7 +1546,7 @@ static void composeHelper(QString *str, QChar::UnicodeVersion version, int from)
 {
     QString &s = *str;
 
-    if (s.length() - from < 2)
+    if (from < 0 || s.length() - from < 2)
         return;
 
     // the loop can partly ignore high Unicode as all ligatures are in the BMP
@@ -1565,12 +1565,13 @@ static void composeHelper(QString *str, QChar::UnicodeVersion version, int from)
         const QUnicodeTables::Properties *p = qGetProp(uc);
         if (p->unicodeVersion == QChar::Unicode_Unassigned || p->unicodeVersion > version) {
             starter = -1; // to prevent starter == pos - 1
-            lastCombining = 0;
+            lastCombining = 255; // to prevent combining > lastCombining
             ++pos;
             continue;
         }
         int combining = p->combiningClass;
         if (starter == pos - 1 || combining > lastCombining) {
+            Q_ASSERT(starter >= from);
             // allowed to form ligature with S
             QChar ligature = ligatureHelper(s.at(starter).unicode(), uc);
             if (ligature.unicode()) {
diff --git a/tests/auto/qchar/tst_qchar.cpp b/tests/auto/qchar/tst_qchar.cpp
index 2f41d744a5..c6df716bdd 100644
--- a/tests/auto/qchar/tst_qchar.cpp
+++ b/tests/auto/qchar/tst_qchar.cpp
@@ -638,6 +638,18 @@ void tst_QChar::normalization()
 void tst_QChar::normalization_manual()
 {
     {
+        QString decomposed;
+        decomposed += QChar(0x41);
+        decomposed += QChar(0x0221); // assigned in 4.0
+        decomposed += QChar(0x300);
+
+        QVERIFY(decomposed.normalized(QString::NormalizationForm_C, QChar::Unicode_3_2) == decomposed);
+
+        decomposed[1] = QChar(0x037f); // unassigned in 6.1
+
+        QVERIFY(decomposed.normalized(QString::NormalizationForm_C) == decomposed);
+    }
+    {
         QString composed;
         composed += QChar(0xc0);
         QString decomposed;