Read file - Posted on July Mon 19th 3:21 AM (Never Expires) - Format: python
  1. #!/usr/bin/python
  2. from csv import reader
  3. import csv,re, sys, getopt
  4.  
  5. def main(argv):
  6.     inputfile = ''
  7.     outputfile = ''
  8.     try:
  9.         opts, args = getopt.getopt(argv,"hi:o:",["ifile=","ofile="])
  10.     except getopt.GetoptError:
  11.         print 'convert.py -i <inputfile> -o <outputfile>'
  12.         sys.exit(2)
  13.     for opt, arg in opts:
  14.         if opt == '-h':
  15.             print 'convert.py -i <inputfile> -o <outputfile>'
  16.             sys.exit()
  17.         elif opt in ("-i", "--ifile"):
  18.             inputfile = arg
  19.         elif opt in ("-o", "--ofile"):
  20.             outputfile = arg
  21.     print 'Input file is ', inputfile
  22.     print 'Output file is ', outputfile
  23.     regex = re.compile(r"\s*\b(?!href|src)[^\s]+=\"[^\"]*\"(?=[^<>]*>)", re.IGNORECASE)
  24.     with open(inputfile, 'r') as read_obj, open(outputfile, 'wb') as write_obj:
  25.         csv_reader = reader(read_obj,  quoting=csv.QUOTE_MINIMAL,  delimiter=',', quotechar='"', skipinitialspace=True, lineterminator = '\r\n')
  26.         writer = csv.writer(write_obj, quoting=csv.QUOTE_NONNUMERIC,  delimiter=',', quotechar='"', skipinitialspace=True)
  27.         for row in csv_reader:
  28.             row.append(regex.sub(r"", row[2]))
  29.             writer.writerow(row)
  30.  
  31. if __name__ == "__main__":
  32.    main(sys.argv[1:])

New Paste

Paste Options

Recent Pastes

99 days ago

helloupdate

342 days ago

FACEBOOK LINKEDI

347 days ago

BUSINESS DATABAS

439 days ago

Untitled

447 days ago

Untitled

447 days ago

Untitled

447 days ago

Untitled

447 days ago

Untitled

447 days ago

Untitled

447 days ago

Untitled