Closed zarino closed 10 years ago
An error is now presented when the limit is hit, but this is a long way off actually fixing the problem:
The more foresightful solution would be to replace xlwt with something like PyExcelerate or OpenPyXL.
See https://github.com/scraperwiki/spreadsheet-download-tool/issues/41 for more discussion of alternative libraries.
IIRC OpenPyXL was used before and moved away from due to poor performance. xlsxwriter
looks like it might be plausible.
I've had a support email about this. It's a rather serious regression for our paying customers.
It was a deliberate feature that the CSVs would always be generated regardless of whether the data was too big for the XLS file format.
@paulfurley apologies, I did quite a bit of refactoring. I thought I had preserved the old behaviour, but evidently not. Is this a regression? If so, it wasn't clear that this was a requirement.
Extracting a dataset with more than 65536 rows results in this exception:
Not only is the current Excel worksheet discarded, so is the entire Excel file, and all CSV files too.
There are a few options here:
Incidentally, the exception is not communicated to the user, who just sees an infinite spinner. That's covered by this issue: https://github.com/scraperwiki/spreadsheet-download-tool/issues/48