Hey there , thank you for the brief intro . Well , my goal is to read all the rows in a spreadsheet document with 50k rows. But i want to be able to do this with some paging , so that I won't need to hog the memory. We are running web servises on an apache tomcat and i cannot afford letting one thread use 600+MB of memory ,which is what happends when i read a spreadsheet with 50k rows.
Just 1-2 weeks ago , i sent a mail on the other mail group of odf -dev , saying ""I think odf toolkit needs to support an operation of reading the file in pages , like this project https://code.google.com/p/sqlsheet/"";. I think this kind of an approach would be a viable one. Actually , AFAIK, MSSQL databases can use an excel document as a table , just lets you mount and start working on reads. On 9 February 2014 14:36, Svante Schubert <[email protected]> wrote: > Am 22.10.2013 15:54, schrieb Homer Kommrad: > > Im reading a 50.000 rows ods file. I want to just read 10 lines , but > the > > whole file is loaded on the memory. Is there a way to read rows paged ? > > > All spreadsheets are tables and within a single XML file, the content.xml. > We are using the Xerces parser an XML push parser, I wonder if we need > to use a pull parser to archive this. > In addition is it really your goal to read the first 10 rows/columns or > do you want those with filled content? > > Thinking about very huge spreadsheets. How can we handle those > efficiently. Is it possible at all in an XML model? > Or is it not a problem of the model, but of the size of the spreadsheet? > Do we have to split a spreadsheet similar as huge databases have to be > split? > > Some late thoughts on that problem, I haven't had a answer earlier.. > Svante >
