Unhandled null `max_row` from openpyxl

[jwilson232]

We are currently using the max_row field from openpyxl to guard against oversized spreadsheets. In some cases the field is unset and will raise a unhandled exception. Adding a default value for the max_row check would invalidate the checks, therefore we should handle the scenario where the field is unset by instead falling back to manually counting the rows.

[chrisarridge]

Thanks @jwilson232 ! This was introduced through PR #174, but I think the deployment might not have happened until recently, which is why we are only seeing this bug surface now. I found this guidance on read_only mode that might help us understand why we are seeing this (Google sheets might not be setting the sheet dimensions).

[jwilson232]

Hi @chrisarridge yes that does seem to be the case. I've tested Google sheets and they do seem to be missing the dimensions property, compared to excel sheets that do have it set. I've tested using sheet.calculate_dimension(force=True) and it does populate the max_row field for Google sheets, however it scans the entire sheet which seems a bit overkill for the large files we're trying to defend against. Since our max size is predefined, I'd suggest we iterate the rows and fail fast once we hit the limit to avoid reading a whole oversized file.