Python: Openpyxl outputs “None” for empty cells

Question

The below code should go through directories, open files, convert etc. The thing is when a cell is empty, the resulting CSV file outputs "None" in its place.

Any reason why and can this be remedied?

Thanks

import os
from openpyxl import load_workbook
import csv

for subdir, dirs, files in os.walk("C:\Users\Alan\Downloads\Knowledge\HOW DO I"):
    for file in files:
        filepath = subdir + os.sep + file

        wb = load_workbook(filename=filepath)
        sh = wb.active
        your_csv_file = open(filepath.replace(".xlsx","")+'_csv.csv','wb')
        wr = csv.writer(your_csv_file,quoting=csv.QUOTE_ALL)

        for rownum in sh.iter_rows():
            wr.writerow([unicode(val.value).encode('ascii','ignore') for val in rownum])

        your_csv_file.close()

Answer 1

OpenPyXl doesn't store empty cells (empty means without value, font, border, and so on). If you get a cell from a worksheet, it dynamically creates a new empty cell with a None value.

The current implementation (v2.4.0) of Worksheet.iter_rows() use Worksheet.cell() method which calls Cell() constructor with no value .

You need to change your code to handle "empty" cells:

for rownum in sh.iter_rows():
    values = [(u"" if cell.value is None else unicode(cell.value))
              for cell in rownum]
    wr.writerow([value.encode('ascii', 'ignore') for value in rownum])

Note: since you export your data to a CSV file for, presumably Windows users, you may choose a more useful encoding like: cp1252 .

Answer 2

Why? Because it seems right to equate "empty" with "None". I want 'NA' though, so I have something like:

def _transmap(dat):
    transmap = {
        # empty cells are going to be empty strings
        None: 'NA',
        # workaround for bug in openpyxl
        # https://bitbucket.org/openpyxl/openpyxl/issues/674/ 
        dt.datetime(1899, 12, 30, 0, 0): dt.time(0, 0),
        dt.datetime(1899, 12, 31, 0, 0): dt.datetime(1900, 1, 1, 0, 0),
    }
    return transmap[dat] if dat in transmap else dat

and then you would wite something like:

for rownum in sh.iter_rows():
            wr.writerow([unicode(_transmap(val.value)).encode('ascii','ignore') for val in rownum])