mirror of https://github.com/MISP/misp-modules
57 lines
1.7 KiB
Python
57 lines
1.7 KiB
Python
import json
|
|
import binascii
|
|
import np
|
|
import ezodf
|
|
import pandas_ods_reader
|
|
import io
|
|
import logging
|
|
|
|
misperrors = {'error': 'Error'}
|
|
mispattributes = {'input': ['attachment'],
|
|
'output': ['freetext', 'text']}
|
|
moduleinfo = {'version': '0.1', 'author': 'Sascha Rommelfangen',
|
|
'description': '.ods to freetext-import IOC extractor',
|
|
'module-type': ['expansion']}
|
|
|
|
moduleconfig = []
|
|
|
|
|
|
def handler(q=False):
|
|
if q is False:
|
|
return False
|
|
q = json.loads(q)
|
|
filename = q['attachment']
|
|
try:
|
|
ods_array = np.frombuffer(binascii.a2b_base64(q['data']), np.uint8)
|
|
except Exception as e:
|
|
print(e)
|
|
err = "Couldn't fetch attachment (JSON 'data' is empty). Are you using the 'Query enrichment' action?"
|
|
misperrors['error'] = err
|
|
print(err)
|
|
return misperrors
|
|
|
|
ods_content = ""
|
|
ods_file = io.BytesIO(ods_array)
|
|
doc = ezodf.opendoc(ods_file)
|
|
num_sheets = len(doc.sheets)
|
|
try:
|
|
for i in range(0, num_sheets):
|
|
ods = pandas_ods_reader.algo.read_data(pandas_ods_reader.parsers.ods, ods_file, i, headers=False)
|
|
ods_content = ods_content + "\n" + ods.to_string(max_rows=None)
|
|
return {'results': [{'types': ['freetext'], 'values': ods_content, 'comment': ".ods-to-text from file " + filename},
|
|
{'types': ['text'], 'values': ods_content, 'comment': ".ods-to-text from file " + filename}]}
|
|
except Exception as e:
|
|
logging.exception(e)
|
|
err = "Couldn't analyze file as .ods. Error was: " + str(e)
|
|
misperrors['error'] = err
|
|
return misperrors
|
|
|
|
|
|
def introspection():
|
|
return mispattributes
|
|
|
|
|
|
def version():
|
|
moduleinfo['config'] = moduleconfig
|
|
return moduleinfo
|