AIL/ail-framework
7
0
Fork 0
mirror of https://github.com/ail-project/ail-framework.git synced 2024-11-10 08:38:28 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix: [import_dir] remove special characters

Browse source
This commit is contained in:
Terrtia 2020-02-25 14:41:45 +01:00
parent 72fe8a2461
commit bdf2fce332
No known key found for this signature in database
GPG key ID: 1E1B1F50D84613D0
1 changed files with 14 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

18
bin/import_dir.py
View file

@ -10,6 +10,7 @@ import argparse
import os import os
import time, datetime import time, datetime
import magic import magic
import re
''' '''
' '
@ -50,6 +51,9 @@ def is_hierachy_valid(path):
correctDate = False correctDate = False
return correctDate return correctDate
def sanitize_str(str_var, invalid_char_regex):
res = re.sub(invalid_char_regex, "-", str_var)
return res.replace(' ', '_')
if __name__ == "__main__": if __name__ == "__main__":
parser = argparse.ArgumentParser(description='Take files from a directory and push them into a 0MQ feed.') parser = argparse.ArgumentParser(description='Take files from a directory and push them into a 0MQ feed.')
@ -67,6 +71,9 @@ if __name__ == "__main__":
socket.bind("tcp://*:{}".format(args.port)) socket.bind("tcp://*:{}".format(args.port))
time.sleep(1) #Important, avoid loosing the 1 message time.sleep(1) #Important, avoid loosing the 1 message
invalid_char = r'[\\/*?&%=:"<>|#\\\']'
invalid_char_dir = r'[\\*?&%=:"<>|#\\\']'
for dirname, dirnames, filenames in os.walk(args.directory): for dirname, dirnames, filenames in os.walk(args.directory):
for filename in filenames: for filename in filenames:
complete_path = os.path.join(dirname, filename) complete_path = os.path.join(dirname, filename)
@ -85,17 +92,20 @@ if __name__ == "__main__":
if not is_hierachy_valid(complete_path): if not is_hierachy_valid(complete_path):
now = datetime.datetime.now() now = datetime.datetime.now()
paste_name = complete_path.split('/')[-1] paste_name = complete_path.split('/')[-1]
paste_name = sanitize_str(paste_name, invalid_char)
directory = complete_path.split('/')[-2] directory = complete_path.split('/')[-2]
directory = sanitize_str(directory, invalid_char_dir)
wanted_path = os.path.join(directory, now.strftime("%Y"), now.strftime("%m"), now.strftime("%d"), paste_name) wanted_path = os.path.join(directory, now.strftime("%Y"), now.strftime("%m"), now.strftime("%d"), paste_name)
wanted_path = os.path.relpath(wanted_path)
else: else:
#take wanted path of the file #take wanted path of the file
wanted_path = os.path.realpath(complete_path) wanted_path = os.path.relpath(complete_path)
wanted_path = wanted_path.split('/') wanted_path = wanted_path.split('/')
wanted_path = '/'.join(wanted_path[-(4+args.hierarchy):]) wanted_path = '/'.join(wanted_path[-(4+args.hierarchy):])
wanted_path = sanitize_str(wanted_path, invalid_char_dir)
# remove whitespace # sanitize feeder_name
wanted_path = wanted_path.replace(' ', '') feeder_name = os.path.relpath(sanitize_str(args.name, invalid_char))
feeder_name = args.name.replace(' ', '')
path_to_send = 'import_dir/' + feeder_name + '>>' + wanted_path path_to_send = 'import_dir/' + feeder_name + '>>' + wanted_path
s = b' '.join( [ args.channel.encode(), path_to_send.encode(), base64.b64encode(messagedata) ] ) s = b' '.join( [ args.channel.encode(), path_to_send.encode(), base64.b64encode(messagedata) ] )