warnings.warn("Discarded range with reserved name")
[u'Interim Testing', u'Year End Testing']
echo 'xlsx/SOX Controls Testing Template.xlsx' 0.00s user 0.00s system 41% cpu 0.002 total
python sheet_names_openpyxl.py 0.24s user 0.11s system 10% cpu 3.424 total
> \ls -1 **/*.xls? | python sheet_names_openpyxl.py --verbosity=info
> find . -print | python sheet_names_openpyxl.py --verbosity=info
> ls -d -1 **/*.*
"""
from sheet_names_api import main
def _openpyxl(filepath):
import openpyxl
# https://openpyxl.readthedocs.org/en/2.3.3/optimized.html
try:
wb = openpyxl.load_workbook(filepath, read_only=True)
# InvalidFileException: openpyxl does not support .txt file format, please check you can open it with Excel first. Supported formats are: .xlsx,.xlsm,.xltx,.xltm
# InvalidFileException: openpyxl does not support the old .xls file format, please use xlrd to read this file, or convert it to the more recent .xlsx file format.
# BadZipfile: File is not a zip file
return wb.get_sheet_names()
except Exception as e:
# Ignore unsupported files
if 'openpyxl does not support' in repr(e):
pass
# if 'openpyxl does not support the old .xls file format' in repr(e):
# pass
if __name__ == '__main__':
main(_openpyxl)
"1.2309041023254395": "xlsx/SOX Testing Status.xlsx",
"1.5334298610687256": "xlsm/Compare XML.xlsm"
}
"""
from sheet_names_api import main
def _xlrd(filepath):
import xlrd
# https://secure.simplistix.co.uk/svn/xlrd/trunk/xlrd/doc/xlrd.html?p=4966
# with xlrd.open_workbook(filepath, on_demand=True, ragged_rows=True) as wb:
# sheet_names = wb.sheet_names()
# return sheet_names
# How about with memory mapping? Nope, blows up on both xls and xlsx
import contextlib
import mmap
import os
length = 2**10 * 4
# length = 0 # whole file
with open(filepath, 'rb') as f:
# mmap throws if length is larger than file size
length = min(os.path.getsize(filepath), length)
with contextlib.closing(mmap.mmap(f.fileno(), length, access=mmap.ACCESS_READ)) as m,\
xlrd.open_workbook(on_demand=True, file_contents=m) as wb:
sheet_names = wb.sheet_names()
return sheet_names
if __name__ == '__main__':
main(_xlrd, ['xls/SMITH 2014 TRIP-Master List.xls'])
