Difference between revisions of "Calc/Performance/The Ou case"

Revision as of 14:24, 3 April 2009

Loading a large plain data file takes very long.

References:

• George Ou's blog entry
• The test case file (.sxc), convert to ODF .ods for profiling
• Same data, but zip'ed Excel-XML

Note that the numbers published in the article compare Excel .xls binary file format with Calc .ods, which is apples and oranges.

Findings:

• source/filter/xml/xmlsubti.cxx
  • 38% of time spent in ScMyTables::NewColumn() because of replicated use of aTableVec[nTableCount - 1] (vector::operator[])
    Note: percentage may be off due to compilation without optimization to obtain exact line numbers that may result in STLport's vector methods being differently compiled.
    • proposed fix: should obtain the pointer once instead.
  • Similar for other places where aTableVec[xxx] is used.

• TODO: Check all ScMyTables::.*() and ScMyTableData::.*()
  • Especially for 63342857 calls to AddColumn() and NewColumn() that result in 1168654944 calls to operator[] ...
  • 63081776 calls to AddColumn() originate from ScXMLTableRowCellContext::EndElement()
  • Those are highly suspicious and seem to indicate that too many temporary elements are created for empty columns/cells (needs verification).