Hi, The XSSFWorkbook parameter to SXSSFWorkbook is intended as"template" with later writing large amounts of data in streaming mode.
If you need to read a large file and produce another large file, you will need to take a look at some of the event/streaming-based input functionality in Apache POI, see e.g. https://github.com/dearshor/poi-examples/blob/master/src/main/java/poi/xssf/eventusermodel/XLSX2CSV.java for an example of how to parse a sheet from an XLSX-file in a streaming mode. The classes OPCPackage and XSSFReader are used to extract the information from the file, then by implementing the SheetContentsHandler interface you have callbacks which allow you to process the data and populate a new SXSSFWorkbook with the values. Dominik. On Thu, Jan 16, 2020 at 4:57 PM Jerry Williamson < jerry.william...@e2open.com> wrote: > Hi, > > I have a system where POI was used to generated an XLSX file with > approximately 256 columns x 35,000 rows of data. > We used SXSSFWorkbook to create it in order to maintain a reasonably small > memory footprint. > The resulting workbook is 24-28MB depending on whether shared strings are > used. > > Sometime later, we want to add an additional sheet to the workbook and we, > again, want to create an SXSSFWorkbook to do the work. > > Unfortunately, the only mechanism available seems to be to create a an > XSSFWorkbook first and then pass that into the constructor of the > SXSSFWorkbook. The problem is that instantiating the XSSFWorkbook is > exhausting a 4GB heap if I bump the heap up to 8GB (in this case), I can > get it seems somewhat antithetical to the streaming approach in the first > place. > > Is there any way to create an SXSSFWorkbook for an existing file without > consuming all of that memory? > > Thanks > >
