'''This python script is used to split the .arff file got from marsyas to .csv files, which is easier to be read into the matlab for further processs.

 

Create on 2012-7-13

 

@auther: mainred

'''

import xlrd, xlwt

import time

import re

import os

 

file_name = raw_input('please input name of the arff file:\n')

path = os.getcwd()

 

file_abs_path = path + '\\' + file_name + '.arff'

 

if not os.path.exists(file_abs_path):

    print "There is no %s.xls in such directory" % file_name

    time.sleep(2)

    exit()

 

data = xlrd.open_workbook(file_name+'.arff')

table = data.sheets()[0]

nrows = table.nrows

ncols = table.ncols

 

 

file = xlwt.Workbook()

 

pattern1 = r'[^\\]*$'

 

name = re.search(pattern1,table.row_values(68)[0]).group(0).split('.')[0]

 

tablew = file.add_sheet(name)

 

redundancy = ''

j = 0

pas = False

for i in range(70,nrows):

    if pas:

        pas = False

        continue

    elif table.row_values(i)[0].startswith('% filename'):

 

        name = name + '.csv'

        file.save(name)

        

        file = xlwt.Workbook()

        

        name = re.search(pattern1,table.row_values(i)[0]).group(0).split('.')[0]

        print name

        tablew = file.add_sheet(name)

        pas = True

        j = 0

        continue

    elif table.row_values(i)[0] == '':

        continue

    else:

        if j == 0:

            pattern2 = r'[^,]+?$'

            redundancy = re.search(pattern2, table.row_values(i)[0]).group(0)

            print redundancy

        tablew.write(j,0,table.row_values(i)[0].replace(redundancy,''))

        j = j + 1

    

name = name + '.csv'

file.save(name)