The idea is to yield the sharepoint data in batches. Currently, it loads the entire workbook into memory and then on top, reads the workbook and keeps another copy of it in memory as a list of dicts. This way, at least the second part consumes far less memory.
This code is untested right now.
The idea is to yield the sharepoint data in batches. Currently, it loads the entire workbook into memory and then on top, reads the workbook and keeps another copy of it in memory as a list of dicts. This way, at least the second part consumes far less memory.