libBasic: implement extended grapheme cluster segmentation algorithm

This is only for the frontend, not for stdlib.  The implementation is very
slow, optimizing it is the next step.

rdar://16755123 rdar://16013860


Swift SVN r18928

include/swift/Basic/Unicode.h

@@ -28,6 +28,53 @@ static inline bool isSingleExtendedGraphemeCluster(StringRef S) {
   return First == S;
 }
 
+enum class GraphemeClusterBreakProperty : uint8_t {
+  Other,
+  CR,
+  LF,
+  Control,
+  Extend,
+  Regional_Indicator,
+  Prepend,
+  SpacingMark,
+  L,
+  V,
+  T,
+  LV,
+  LVT,
+};
+
+/// Extended grapheme cluster boundary rules, represented as a matrix.  Indexed
+/// by first code point, then by second code point in least-significant-bit
+/// order.  A set bit means that a boundary is prohibited between two code
+/// points.
+extern uint16_t ExtendedGraphemeClusterNoBoundaryRulesMatrix[];
+
+/// Returns the value of the Grapheme_Cluster_Break property for a given code
+/// point.
+GraphemeClusterBreakProperty getGraphemeClusterBreakProperty(uint32_t C);
+
+/// Returns true if there is always an extended grapheme cluster boundary
+/// after a code point with a given property value.  Use only for optimization,
+/// to skip calculating Grapheme_Cluster_Break property for the second code
+/// point.
+static inline bool
+isExtendedGraphemeClusterBoundaryAfter(GraphemeClusterBreakProperty GCB1) {
+  auto RuleRow =
+      ExtendedGraphemeClusterNoBoundaryRulesMatrix[static_cast<unsigned>(GCB1)];
+  return RuleRow == 0;
+}
+
+/// Determine if there is an extended grapheme cluster boundary between code
+/// points with given Grapheme_Cluster_Break property values.
+static inline bool
+isExtendedGraphemeClusterBoundary(GraphemeClusterBreakProperty GCB1,
+                                  GraphemeClusterBreakProperty GCB2) {
+  auto RuleRow =
+      ExtendedGraphemeClusterNoBoundaryRulesMatrix[static_cast<unsigned>(GCB1)];
+  return !(RuleRow & (1 << static_cast<unsigned>(GCB2)));
+}
+
 } // namespace unicode
 } // namespace swift