Commit 287ce125 authored by Mattia Bondanza's avatar Mattia Bondanza

Inserted a basic cli to specify the directory to analyze.

parent 1dd00bea
......@@ -5,9 +5,20 @@ from fnmatch import fnmatch
from PyPDF2 import PdfFileReader
from searchISBN import *
import re
import argparse
root = '/home/mattia/Nextcloud/Università/Libreria/'
pattern = "*"
#-----Argument parsing-------#
parser = argparse.ArgumentParser( description="BAO takes a directory that contains many pdf files and "
"analyzes them, to tag the files with the correct metadata." )
parser.add_argument( 'directory', nargs = 1, type=str, help='The directory that contains the'
' pdf file to analyze', metavar='/directory/' )
parser.add_argument( '-p', '--pattern', nargs = 1,
dest='pattern', type=str, required=False, default='*', help='The pattern the file have to match to be'
' analyzed', metavar='path' )
args = parser.parse_args()
type_stat = {}
total_files = 0.0
......@@ -15,9 +26,9 @@ metadata = {}
books = []
for path, subdirs, files in os.walk(root):
for path, subdirs, files in os.walk([0]):
for name in files:
if fnmatch(name, pattern):
if fnmatch(name, args.pattern[0]):
#print os.path.join(path, name)
total_files += 1
#file type analysis, the try / except is to manage different version
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment