#! /usr/bin/env python

# Read and print an excel .xls spreadsheet.

# Thanks to http://scienceoss.com/read-excel-files-from-python/

import sys, xlrd

def print_sheet(sheetname, sh):

# Gnumeric makes null sheets, with 0 rows and 0 columns.

# Guard against division by zero in gnumeric spreadsheets:

if not sh or sh.ncols <= 0:

print("Null sheet", sheetname)

return

print("===== sheet", sheetname, ": rows", sh.nrows, "cols", sh.ncols)

# set a maximum on any one column.

# Ideally this would be smarter,

# and would let one long row expand if the rest were shorter.

maxlen = 80 / sh.ncols

# Get max len for each col

maxstr = [1] * sh.ncols

for rownum in range(sh.nrows):

for colnum in range(sh.ncols):

# Figure out how many characters this cell needs.

val = sh.cell(rownum, colnum).value

# This gets lots of null strings, so filter them out:

if not val:

continue

#print "type:", type(val), val

slen = len(str(val))

#if slen > maxlen: slen = maxlen

if slen > maxstr[colnum]:

maxstr[colnum] = slen

fmt = ""

for colnum in range(sh.ncols):

fmt += "%%-%ds" % (maxstr[colnum] + 1)

for rownum in range(sh.nrows):

# Should get the charset from the environment, but it's not clear

# even that is enough. In practice, this works to stdout without

# needing the encode, but to pipe through grep or redirect to a

# file it needs the encode.

print((fmt % tuple(map(str,

sh.row_values(rownum)))).encode(sys.getdefaultencoding(),

'backslashreplace'))

for filename in sys.argv[1:]:

wb = xlrd.open_workbook(filename)

#sh = wb.sheet_by_index(0)

for sheetname in wb.sheet_names():

print_sheet(sheetname, wb.sheet_by_name(sheetname))