get_highest_row/column are unreliable
Created originally on Bitbucket by private (Alex Stapleton)
These functions advertise that they return the highest cells with data in the sheet. This turns out to not always be true.
The most obvious example is that an empty workbook saved using LibreOffice has a highest column index of 1025. (An empty Microsoft Excel workbook has a highest column of 1.)
I also have workbooks (originally created using an as yet unidentified version of Microsoft Excel) that report a highest row of 65536 despite only containing a couple of hundred populated rows. Resaving these workbooks with LibreOffice results in different XML but the same highest and lowest rows.
I think is is probably something to do with how the dimensions are determined and an interaction with the cell and column styling information in the underlying XML files.
My hacky solution to this is to walk sheet._cells
and work out the true values for these but relying on undocumented private members is less than ideal.
I think changing the behaviour of get_highest_row/column might not be desireable but it does seem like the documentatio needs updating and perhaps some methods that do what they currently advertise themselves as doing should be added?