mirror of
https://invent.kde.org/system/dolphin
synced 2024-09-12 12:54:29 +00:00
KFileItemModel::insertItems(): guarantee O(N) run time complexity
This commit prevents repeated insertions of single items into the list m_itemData, which shift all following items by one position and result in O(N^2) worst case complexity for the entire function. Moreover, the hash m_items is updated only for the items starting from the first inserted/removed item to save some superfluous calculations of hash values. REVIEW: 110355
This commit is contained in:
parent
01761798a9
commit
bf85483c99
|
@ -956,9 +956,9 @@ void KFileItemModel::dispatchPendingItemsToInsert()
|
|||
}
|
||||
}
|
||||
|
||||
void KFileItemModel::insertItems(QList<ItemData*>& items)
|
||||
void KFileItemModel::insertItems(QList<ItemData*>& newItems)
|
||||
{
|
||||
if (items.isEmpty()) {
|
||||
if (newItems.isEmpty()) {
|
||||
return;
|
||||
}
|
||||
|
||||
|
@ -966,68 +966,81 @@ void KFileItemModel::insertItems(QList<ItemData*>& items)
|
|||
QElapsedTimer timer;
|
||||
timer.start();
|
||||
kDebug() << "===========================================================";
|
||||
kDebug() << "Inserting" << items.count() << "items";
|
||||
kDebug() << "Inserting" << newItems.count() << "items";
|
||||
#endif
|
||||
|
||||
m_groups.clear();
|
||||
|
||||
sort(items.begin(), items.end());
|
||||
sort(newItems.begin(), newItems.end());
|
||||
|
||||
#ifdef KFILEITEMMODEL_DEBUG
|
||||
kDebug() << "[TIME] Sorting:" << timer.elapsed();
|
||||
#endif
|
||||
|
||||
KItemRangeList itemRanges;
|
||||
int targetIndex = 0;
|
||||
int sourceIndex = 0;
|
||||
int insertedAtIndex = -1; // Index for the current item-range
|
||||
int insertedCount = 0; // Count for the current item-range
|
||||
int previouslyInsertedCount = 0; // Sum of previously inserted items for all ranges
|
||||
while (sourceIndex < items.count()) {
|
||||
// Find target index from m_items to insert the current item
|
||||
// in a sorted order
|
||||
const int previousTargetIndex = targetIndex;
|
||||
while (targetIndex < m_itemData.count()) {
|
||||
if (!lessThan(m_itemData.at(targetIndex), items.at(sourceIndex))) {
|
||||
break;
|
||||
const int existingItemCount = m_itemData.count();
|
||||
const int newItemCount = newItems.count();
|
||||
const int totalItemCount = existingItemCount + newItemCount;
|
||||
|
||||
if (existingItemCount == 0) {
|
||||
// Optimization for the common special case that there are no
|
||||
// items in the model yet. Happens, e.g., when entering a folder.
|
||||
m_itemData = newItems;
|
||||
itemRanges << KItemRange(0, newItemCount);
|
||||
} else {
|
||||
m_itemData.reserve(totalItemCount);
|
||||
for (int i = existingItemCount; i < totalItemCount; ++i) {
|
||||
m_itemData.append(0);
|
||||
}
|
||||
|
||||
// We build the new list m_items in reverse order to minimize
|
||||
// the number of moves and guarantee O(N) complexity.
|
||||
int targetIndex = totalItemCount - 1;
|
||||
int sourceIndexExistingItems = existingItemCount - 1;
|
||||
int sourceIndexNewItems = newItemCount - 1;
|
||||
|
||||
int rangeCount = 0;
|
||||
|
||||
while (sourceIndexNewItems >= 0) {
|
||||
ItemData* newItem = newItems.at(sourceIndexNewItems);
|
||||
if (sourceIndexExistingItems >= 0 && lessThan(newItem, m_itemData.at(sourceIndexExistingItems))) {
|
||||
// Move an existing item to its new position. If any new items
|
||||
// are behind it, push the item range to itemRanges.
|
||||
if (rangeCount > 0) {
|
||||
itemRanges << KItemRange(sourceIndexExistingItems + 1, rangeCount);
|
||||
rangeCount = 0;
|
||||
}
|
||||
|
||||
m_itemData[targetIndex] = m_itemData.at(sourceIndexExistingItems);
|
||||
--sourceIndexExistingItems;
|
||||
} else {
|
||||
// Insert a new item into the list.
|
||||
++rangeCount;
|
||||
m_itemData[targetIndex] = newItem;
|
||||
--sourceIndexNewItems;
|
||||
}
|
||||
++targetIndex;
|
||||
--targetIndex;
|
||||
}
|
||||
|
||||
if (targetIndex - previousTargetIndex > 0 && insertedAtIndex >= 0) {
|
||||
itemRanges << KItemRange(insertedAtIndex, insertedCount);
|
||||
previouslyInsertedCount += insertedCount;
|
||||
insertedAtIndex = targetIndex - previouslyInsertedCount;
|
||||
insertedCount = 0;
|
||||
// Push the final item range to itemRanges.
|
||||
if (rangeCount > 0) {
|
||||
itemRanges << KItemRange(sourceIndexExistingItems + 1, rangeCount);
|
||||
}
|
||||
|
||||
// Insert item at the position targetIndex by transferring
|
||||
// the ownership of the item-data from 'items' to m_itemData.
|
||||
// m_items will be inserted after the loop (see comment below)
|
||||
m_itemData.insert(targetIndex, items.at(sourceIndex));
|
||||
++insertedCount;
|
||||
|
||||
if (insertedAtIndex < 0) {
|
||||
insertedAtIndex = targetIndex;
|
||||
Q_ASSERT(previouslyInsertedCount == 0);
|
||||
}
|
||||
++targetIndex;
|
||||
++sourceIndex;
|
||||
// Note that itemRanges is still sorted in reverse order.
|
||||
std::reverse(itemRanges.begin(), itemRanges.end());
|
||||
}
|
||||
|
||||
// The indexes of all m_items must be adjusted, not only the index
|
||||
// of the new items
|
||||
const int itemDataCount = m_itemData.count();
|
||||
m_items.reserve(itemDataCount);
|
||||
for (int i = 0; i < itemDataCount; ++i) {
|
||||
// The indexes starting from the first inserted item must be adjusted.
|
||||
m_items.reserve(totalItemCount);
|
||||
for (int i = itemRanges.front().index; i < totalItemCount; ++i) {
|
||||
m_items.insert(m_itemData.at(i)->item.url(), i);
|
||||
}
|
||||
|
||||
itemRanges << KItemRange(insertedAtIndex, insertedCount);
|
||||
emit itemsInserted(itemRanges);
|
||||
|
||||
#ifdef KFILEITEMMODEL_DEBUG
|
||||
kDebug() << "[TIME] Inserting of" << items.count() << "items:" << timer.elapsed();
|
||||
kDebug() << "[TIME] Inserting of" << newItems.count() << "items:" << timer.elapsed();
|
||||
#endif
|
||||
}
|
||||
|
||||
|
@ -1119,10 +1132,10 @@ void KFileItemModel::removeItems(const KFileItemList& items, RemoveItemsBehavior
|
|||
|
||||
m_itemData.erase(m_itemData.end() - indexesToRemove.count(), m_itemData.end());
|
||||
|
||||
// Step 3: Adjust indexes in the hash m_items. Note that all indexes
|
||||
// might have been changed by the removal of the items.
|
||||
// Step 3: Adjust indexes in the hash m_items, starting from the
|
||||
// index of the first removed item.
|
||||
const int newItemDataCount = m_itemData.count();
|
||||
for (int i = 0; i < newItemDataCount; ++i) {
|
||||
for (int i = itemRanges.front().index; i < newItemDataCount; ++i) {
|
||||
m_items.insert(m_itemData.at(i)->item.url(), i);
|
||||
}
|
||||
|
||||
|
|
Loading…
Reference in a new issue