Tailor grapheme boundaries so sequence emoji are one grapheme

Make it so it's not possible to position the cursor inside an emoji
formed by a sequence including zero-width joiners.

Bug: 25368653
Change-Id: I67ec0874cd1505f3c82ab91492ffc3d39a52fae6

libs/minikin/GraphemeBreak.cpp

@@ -64,6 +64,19 @@ bool isPureKiller(uint32_t c) {
             || c == 0xA953 || c == 0xABED || c == 0x11134 || c == 0x112EA || c == 0x1172B);
 }
 
+// Returns true if the character appears before or after zwj in a zwj emoji sequence. See
+// http://www.unicode.org/emoji/charts/emoji-zwj-sequences.html
+bool isZwjEmoji(uint32_t c) {
+    return (c == 0x2764       // HEAVY BLACK HEART
+            || c == 0x1F468   // MAN
+            || c == 0x1F469   // WOMAN
+            || c == 0x1F48B   // KISS MARK
+            || c == 0x1F466   // BOY
+            || c == 0x1F467   // GIRL
+            || c == 0x1F441   // EYE
+            || c == 0x1F5E8); // LEFT SPEECH BUBBLE
+}
+
 bool GraphemeBreak::isGraphemeBreak(const uint16_t* buf, size_t start, size_t count,
         size_t offset) {
     // This implementation closely follows Unicode Standard Annex #29 on
@@ -139,6 +152,19 @@ bool GraphemeBreak::isGraphemeBreak(const uint16_t* buf, size_t start, size_t co
             && u_getIntPropertyValue(c2, UCHAR_GENERAL_CATEGORY) == U_OTHER_LETTER) {
         return false;
     }
+    // Tailoring: make emoji sequences with ZWJ a single grapheme cluster
+    if (c1 == 0x200D && isZwjEmoji(c2) && offset_back > start) {
+        // look at character before ZWJ to see that both can participate in an emoji zwj sequence
+        uint32_t c0 = 0;
+        U16_PREV(buf, start, offset_back, c0);
+        if (c0 == 0xFE0F && offset_back > start) {
+            // skip over emoji variation selector
+            U16_PREV(buf, start, offset_back, c0);
+        }
+        if (isZwjEmoji(c0)) {
+            return false;
+        }
+    }
     // Rule GB10, Any ÷ Any
     return true;
 }

tests/GraphemeBreakTests.cpp

@@ -119,6 +119,21 @@ TEST(GraphemeBreak, tailoring) {
     EXPECT_FALSE(IsBreak("U+0915 U+094D | U+0915"));  // Devanagari ka+virama+ka
     EXPECT_FALSE(IsBreak("U+0E01 | U+0E3A U+0E01"));  // thai phinthu = pure killer
     EXPECT_TRUE(IsBreak("U+0E01 U+0E3A | U+0E01"));  // thai phinthu = pure killer
+
+    // suppress grapheme breaks in zwj emoji sequences, see
+    // http://www.unicode.org/emoji/charts/emoji-zwj-sequences.html
+    EXPECT_FALSE(IsBreak("U+1F469 U+200D | U+2764 U+FE0F U+200D U+1F48B U+200D U+1F468"));
+    EXPECT_FALSE(IsBreak("U+1F469 U+200D U+2764 U+FE0F U+200D | U+1F48B U+200D U+1F468"));
+    EXPECT_FALSE(IsBreak("U+1F469 U+200D U+2764 U+FE0F U+200D U+1F48B U+200D | U+1F468"));
+    EXPECT_FALSE(IsBreak("U+1F468 U+200D | U+1F469 U+200D U+1F466"));
+    EXPECT_FALSE(IsBreak("U+1F468 U+200D U+1F469 U+200D | U+1F466"));
+    EXPECT_FALSE(IsBreak("U+1F469 U+200D | U+1F469 U+200D U+1F467 U+200D U+1F466"));
+    EXPECT_FALSE(IsBreak("U+1F469 U+200D U+1F469 U+200D | U+1F467 U+200D U+1F466"));
+    EXPECT_FALSE(IsBreak("U+1F469 U+200D U+1F469 U+200D U+1F467 U+200D | U+1F466"));
+    EXPECT_FALSE(IsBreak("U+1F441 U+200D | U+1F5E8"));
+
+    // ARABIC LETTER BEH + ZWJ + heart, not a zwj emoji sequence, so we preserve the break
+    EXPECT_TRUE(IsBreak("U+0628 U+200D | U+2764"));
 }
 
 TEST(GraphemeBreak, offsets) {