strepsirrhini/office-gobmx
9
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

fdo#35943: Better performance with pivot table refresh.

Browse source 
When the source data range contains trailing rows, simply skip them, and
don't even bother iterating them.  Apparently sometimes users specify a
data range with a huge amount of trailing empty rows, which would slow
down the pivot table refresh for no good reason.

But we do still need to keep the original end row position, in case the
pivot table needs to include empty cells in the output.

Change-Id: I2c73c368837b8e322e12b25ddf31429488961f06
This commit is contained in:
Kohei Yoshida 2012-12-15 00:46:12 -05:00
parent 33c5cd7857
commit 696644dbfb
3 changed files with 43 additions and 6 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

1
sc/inc/dpcache.hxx
View file

@ -121,6 +121,7 @@ private:
LabelsType maLabelNames; // Stores dimension names.
mdds::flat_segment_tree<SCROW, bool> maEmptyRows;
SCROW mnDataSize;
SCROW mnRowCount;
bool mbDisposing;

45
sc/source/core/data/dpcache.cxx
View file

@ -56,6 +56,7 @@ ScDPCache::ScDPCache(ScDocument* pDoc) :
mnColumnCount ( 0 ),
maEmptyRows(0, MAXROW, true),
mnDataSize(-1),
mnRowCount(0),
mbDisposing(false)
{
}
@ -311,6 +312,16 @@ bool ScDPCache::InitFromDoc(ScDocument* pDoc, const ScRange& rRange)
mnColumnCount = nEndCol - nStartCol + 1;
// this row count must include the trailing empty rows.
mnRowCount = nEndRow - nStartRow; // skip the topmost label row.
// Skip trailing empty rows if exists.
SCCOL nCol1 = nStartCol, nCol2 = nEndCol;
SCROW nRow1 = nStartRow, nRow2 = nEndRow;
pDoc->ShrinkToDataArea(nDocTab, nCol1, nRow1, nCol2, nRow2);
bool bTailEmptyRows = nEndRow > nRow2; // Trailing empty rows exist.
nEndRow = nRow2;
maFields.reserve(mnColumnCount);
for (size_t i = 0; i < static_cast<size_t>(mnColumnCount); ++i)
maFields.push_back(new Field);
@ -342,6 +353,17 @@ bool ScDPCache::InitFromDoc(ScDocument* pDoc, const ScRange& rRange)
}
processBuckets(aBuckets, rField);
if (bTailEmptyRows)
{
// If the last item is not empty, append one. Note that the items
// are sorted, and empty item should come last when sorted.
if (rField.maItems.empty() || !rField.maItems.back().IsEmpty())
{
aData.SetEmpty();
rField.maItems.push_back(aData);
}
}
}
PostInit();
@ -404,6 +426,9 @@ bool ScDPCache::InitFromDataBase(DBConnector& rDB)
rDB.finish();
if (!maFields.empty())
mnRowCount = maFields[0].maData.size();
PostInit();
return true;
}
@ -684,6 +709,8 @@ void ScDPCache::PostInit()
void ScDPCache::Clear()
{
mnColumnCount = 0;
mnRowCount = 0;
maFields.clear();
maLabelNames.clear();
maGroupFields.clear();
@ -723,7 +750,18 @@ SCROW ScDPCache::GetItemDataId(sal_uInt16 nDim, SCROW nRow, bool bRepeatIfEmpty)
OSL_ENSURE(nDim < mnColumnCount, "ScDPTableDataCache::GetItemDataId ");
const Field& rField = maFields[nDim];
if (bRepeatIfEmpty)
if (static_cast<size_t>(nRow) >= rField.maData.size())
{
// nRow is in the trailing empty rows area.
if (bRepeatIfEmpty)
nRow = rField.maData.size()-1; // Move to the last non-empty row.
else
// Return the last item, which should always be empty if the
// initialization has skipped trailing empty rows.
return rField.maItems.size()-1;
}
else if (bRepeatIfEmpty)
{
while (nRow > 0 && rField.maItems[rField.maData[nRow]].IsEmpty())
--nRow;
@ -772,10 +810,7 @@ const ScDPItemData* ScDPCache::GetItemDataById(long nDim, SCROW nId) const
SCROW ScDPCache::GetRowCount() const
{
if (maFields.empty() || maFields[0].maData.empty())
return 0;
return maFields[0].maData.size();
return mnRowCount;
}
SCROW ScDPCache::GetDataSize() const

3
sc/source/core/data/dpitemdata.cxx
View file

@ -33,7 +33,8 @@ sal_Int32 ScDPItemData::Compare(const ScDPItemData& rA, const ScDPItemData& rB)
{
if (rA.meType != rB.meType)
{
// group value, value and string in this order.
// group value, value and string in this order. Ensure that the empty
// type comes last.
return rA.meType < rB.meType ? -1 : 1;
}