Martin MarĂ¥k commented on ODFTOOLKIT-463:

I am having the same issue, writing a spreadsheet with 50 columns and 5000 
lines takes over 10 minutes, while the equivalent csv file is done in less than 
5 seconds. It seems that getCellByPosition(column,row) is the culprit.

> Performance issues
> ------------------
>                 Key: ODFTOOLKIT-463
>                 URL: https://issues.apache.org/jira/browse/ODFTOOLKIT-463
>             Project: ODF Toolkit
>          Issue Type: Bug
>            Reporter: Thomas Hackel
>            Priority: Minor
>         Attachments: odf-hotspots.png
> Version: 0.8.2-incubating  (not present in JIRA)
> I have different "output writers" for formats like ODF/CSV/POI etc.
> The ODF version is the slowest of all, it might be a usage problem or a 
> ODFToolkit issue.
> It seems that creating new Rows/Cells is consuming a lot of time.
> All other implementations have the QueryDSL.fetch method as the major 
> hotspot. With ODFToolkit this data gathering method is only at place 3.
> All what i am doing is creating a new row for each line of data:
> {code:java}
> row = table.getRowByIndex(rowNumber);
> {code}
> I am talking about only 6000 rows.

This message was sent by Atlassian JIRA

Reply via email to