Improve the script itemization algorithm to match Unicode 8.0

Override preceding Common-s with a subsequent non-Inherited,
non-Common script.
This produces longer script runs, which automagically improves
the shaping quality (as we don't lose the context anymore),
the shaping performance (as we're typically shape a fewer runs),
and the fallback font selection (when the font supports more
than just a single language/script).

Task-number: QTBUG-29930
Change-Id: I1c55af30bd397871d7f1f6e062605517f5a7e5a1
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>

1 files changed, 7 insertions, 15 deletions

diff --git a/tests/auto/gui/text/qtextscriptengine/tst_qtextscriptengine.cpp b/tests/auto/gui/text/qtextscriptengine/tst_qtextscriptengine.cpp
index 18012fd347..e9e243f7ed 100644
--- a/tests/auto/gui/text/qtextscriptengine/tst_qtextscriptengine.cpp
+++ b/tests/auto/gui/text/qtextscriptengine/tst_qtextscriptengine.cpp
@@ -1225,29 +1225,21 @@ void tst_QTextScriptEngine::thaiWithZWJ()
     QTextLayout layout(s, font);
     QTextEngine *e = layout.engine();
     e->itemize();
-    QCOMPARE(e->layoutData->items.size(), 11);
+    QCOMPARE(e->layoutData->items.size(), 3);
 
     for (int item = 0; item < e->layoutData->items.size(); ++item)
         e->shape(item);
 
-    QCOMPARE(e->layoutData->items[0].num_glyphs, ushort(7));  // Thai: The ZWJ and ZWNJ characters are inherited, so should be part of the thai script
-    QCOMPARE(e->layoutData->items[1].num_glyphs, ushort(1));  // Common: The smart quotes cannot be handled by thai, so should be a separate item
-    QCOMPARE(e->layoutData->items[2].num_glyphs, ushort(1));  // Thai: Thai character
-    QCOMPARE(e->layoutData->items[3].num_glyphs, ushort(1));  // Common: Ellipsis
-    QCOMPARE(e->layoutData->items[4].num_glyphs, ushort(1));  // Thai: Thai character
-    QCOMPARE(e->layoutData->items[5].num_glyphs, ushort(1));  // Common: Smart quote
-    QCOMPARE(e->layoutData->items[6].num_glyphs, ushort(1));  // Thai: Thai character
-    QCOMPARE(e->layoutData->items[7].num_glyphs, ushort(1));  // Common: \xA0 = non-breaking space. Could be useful to have in thai, but not currently implemented
-    QCOMPARE(e->layoutData->items[8].num_glyphs, ushort(1));  // Thai: Thai character
-    QCOMPARE(e->layoutData->items[9].num_glyphs, ushort(1));  // Japanese: Kanji for tree
-    QCOMPARE(e->layoutData->items[10].num_glyphs, ushort(2)); // Thai: Thai character followed by superscript "a" which is of inherited type
+    QCOMPARE(e->layoutData->items[0].num_glyphs, ushort(15));  // Thai, Inherited and Common
+    QCOMPARE(e->layoutData->items[1].num_glyphs, ushort(1));  // Japanese: Kanji for tree
+    QCOMPARE(e->layoutData->items[2].num_glyphs, ushort(2)); // Thai: Thai character followed by superscript "a" which is of inherited type
 
     //A quick sanity check - check all the characters are individual clusters
     unsigned short *logClusters = e->layoutData->logClustersPtr;
-    for (int i = 0; i < 7; i++)
+    for (int i = 0; i < 15; i++)
         QCOMPARE(logClusters[i], ushort(i));
-    for (int i = 0; i < 10; i++)
-        QCOMPARE(logClusters[i+7], ushort(0));
+    for (int i = 0; i < 3; i++)
+        QCOMPARE(logClusters[i+15], ushort(0));
 
     // A thai implementation could either remove the ZWJ and ZWNJ characters, or hide them.
     // The current implementation hides them, so we test for that.