Record sort order during the first sort by the value, and use it the 2nd time.

This avoids comparison of raw values (ScDPItemData) during the 2nd sort.
Comparison of raw values can be expensive especially when the item sets mostly
consist of string values.

sc/source/core/data/dpcache.cxx

@@ -247,8 +247,9 @@ struct Bucket
     ScDPItemData maValue;
     SCROW mnOrderIndex;
     SCROW mnDataIndex;
+    SCROW mnValueSortIndex;
     Bucket(const ScDPItemData& rValue, SCROW nOrder, SCROW nData) :
-        maValue(rValue), mnOrderIndex(nOrder), mnDataIndex(nData) {}
+        maValue(rValue), mnOrderIndex(nOrder), mnDataIndex(nData), mnValueSortIndex(0) {}
 };
 
 struct LessByValue : std::binary_function<Bucket, Bucket, bool>
@@ -259,6 +260,14 @@ struct LessByValue : std::binary_function<Bucket, Bucket, bool>
     }
 };
 
+struct LessByValueSortIndex : std::binary_function<Bucket, Bucket, bool>
+{
+    bool operator() (const Bucket& left, const Bucket& right) const
+    {
+        return left.mnValueSortIndex < right.mnValueSortIndex;
+    }
+};
+
 struct LessByDataIndex : std::binary_function<Bucket, Bucket, bool>
 {
     bool operator() (const Bucket& left, const Bucket& right) const
@@ -297,6 +306,17 @@ public:
     }
 };
 
+class TagValueSortOrder : std::unary_function<Bucket, void>
+{
+    SCROW mnCurIndex;
+public:
+    TagValueSortOrder() : mnCurIndex(0) {}
+    void operator() (Bucket& v)
+    {
+        v.mnValueSortIndex = mnCurIndex++;
+    }
+};
+
 void processBuckets(std::vector<Bucket>& aBuckets, ScDPCache::Field& rField)
 {
     if (aBuckets.empty())
@@ -305,6 +325,9 @@ void processBuckets(std::vector<Bucket>& aBuckets, ScDPCache::Field& rField)
     // Sort by the value.
     std::sort(aBuckets.begin(), aBuckets.end(), LessByValue());
 
+    // Remember this sort order.
+    std::for_each(aBuckets.begin(), aBuckets.end(), TagValueSortOrder());
+
     {
         // Set order index such that unique values have identical index value.
         SCROW nCurIndex = 0;
@@ -329,7 +352,7 @@ void processBuckets(std::vector<Bucket>& aBuckets, ScDPCache::Field& rField)
     std::for_each(aBuckets.begin(), aBuckets.end(), PushBackOrderIndex(rField.maData));
 
     // Sort by the value again.
-    std::sort(aBuckets.begin(), aBuckets.end(), LessByValue());
+    std::sort(aBuckets.begin(), aBuckets.end(), LessByValueSortIndex());
 
     // Unique by value.
     std::vector<Bucket>::iterator itUniqueEnd =