PyMISP/pymisp/tools/csvloader.py

49 lines
1.9 KiB
Python

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
from pathlib import Path
import csv
from pymisp import MISPObject
class CSVLoader():
def __init__(self, template_name: str, csv_path: Path, fieldnames: list=[], has_fieldnames=False):
self.template_name = template_name
self.csv_path = csv_path
self.fieldnames = [f.strip().lower() for f in fieldnames]
if not self.fieldnames:
# If the user doesn't pass fieldnames, we assume the CSV has them.
self.has_fieldnames = True
else:
self.has_fieldnames = has_fieldnames
def load(self):
objects = []
with open(self.csv_path, newline='') as csvfile:
reader = csv.reader(csvfile)
if self.has_fieldnames:
# The file has fieldnames, we either ignore it, or validate its validity
fieldnames = [f.strip().lower() for f in reader.__next__()]
if not self.fieldnames:
self.fieldnames = fieldnames
if not self.fieldnames:
raise Exception(f'No fieldnames, impossible to create objects.')
else:
# Check if the CSV file has a header, and if it matches with the object template
tmp_object = MISPObject(self.template_name)
allowed_fieldnames = list(tmp_object._definition['attributes'].keys())
for fieldname in self.fieldnames:
if fieldname not in allowed_fieldnames:
raise Exception(f'{fieldname} is not a valid object relation for {self.template_name}: {allowed_fieldnames}')
for row in reader:
tmp_object = MISPObject(self.template_name)
for object_relation, value in zip(self.fieldnames, row):
tmp_object.add_attribute(object_relation, value=value)
objects.append(tmp_object)
return objects