def coref_the_following_colon(stri):
sentence2 = ' '
final_txt = ''
fl = len(final_txt)
# list1 = the_following_colon_lst()
list1 = load_lists(fpath)['TFCL']
list1 = list1.replace("'", "").strip('][').split(', ')
sentences = sent_tokenize(stri)
l = len(sentences)
c = 0
for sentence in sentences:
c += 1
for value in list1:
if value in sentence:
sentence.strip(
) # to get ride of possible space at the end of sentence
if sentence[-1] == ".":
sentence = sentence[:-1] # removes the dot from the end
if ":" in sentence:
one = sentence.split(value)[0]
two = sentence.split(value)[1]
# sentence2 = sentence.split(":")[0].replace(value[:-1],sentence.split(":",1)[1]) + ". " # replace the token with value
# sentence2 = sentence.replace(value, sentence.split(value)[1]) + ". " # replace the token with value
sentence2 = sentence.replace(value, " ") + ". "
final_txt += " " + sentence2
p = final_txt
fl += 1
break
if c > fl:
final_txt += " " + sentence
fl += 1
return final_txt
def colon_seprator_multiplication(stri):
coref_reference_list = load_lists(fpath)['TFL']
coref_reference_list = coref_reference_list.replace(
"'", "").strip('][').split(', ')
stri = stri.rstrip()
stri = stri.rstrip('.')
result = ""
for item in sent_tokenize(stri):
flag = False
for refrence in coref_reference_list:
if refrence in item and ":" in item:
sentence_splits = item.split(":", 1)
y = iocs.list_of_iocs(item.split(":", 1)[1])
if y:
sentence_replicas = [sentence_splits[0].rstrip(":")
] * len(y)
for i in range(len(sentence_replicas)):
result += sentence_replicas[i].replace(refrence,
y[i]) + " . "
else:
item = sentence_splits[0].replace(refrence,
sentence_splits[1])
result += item
flag = True
break
if flag == False:
result += item
result += " "
if result.rstrip("")[-1] != ".":
result += "."
return result
def astriks(lis):
apps_process = load_lists(fpath)['APPs-PROCESS']
apps_process = apps_process.replace("'", "").strip('][').split(' , ')
updated_list = [[] for x in range(len(lis))]
for jj, lst in enumerate(lis):
for i in range(len(lst)):
if ":" not in lst[i]:
lst[i] = "TMP: " + lst[i]
leftnode = lst[i].split(":", 1)[0]
rightnode = lst[i].split(":", 1)[1]
lOFioc = iocs.list_of_iocs(rightnode)
found_app = [
app for app in apps_process if app in rightnode.lower()
]
if len(found_app) > 1:
found_app = list(group_partials(found_app))
found_app = process_convert(found_app)
else:
found_app = process_convert(found_app)
if not lOFioc:
if leftnode.lower() != "v" and not found_app:
updated_list[jj].append(leftnode + ": *")
elif leftnode.lower() == "v":
updated_list[jj].append(lst[i])
elif leftnode.lower() != "v" and found_app:
for process in found_app:
updated_list[jj].append(leftnode + ": " + process)
elif len(lOFioc) == 1:
updated_list[jj].append(leftnode + ": " + lOFioc[0])
elif len(lOFioc) >= 2:
updated_list[jj].append(leftnode + ": " + lOFioc[0])
for index in range(1, len(lOFioc)):
updated_list[jj].append("ARG-NEW: " + lOFioc[index])
return updated_list
def on_the_windows_x_only():
# on_the_windows_x_list = load_lists_microsoft.on_the_windows_x_lst()
on_the_windows_x_list = load_lists(fpath)['MS_OTW']
on_the_windows_x_list = on_the_windows_x_list.replace(
"'", "").strip('][').split(', ')
lst = perform_following_action()
for i in lst:
for j in on_the_windows_x_list:
if j == i:
lst.remove(i)
# break
return lst
def perform_following_action(
): # When Virus:Win32/Funlove.4099 runs, it performs the following actions:
perform_following_action_list = load_lists(fpath)['MS_PFA']
perform_following_action_list = perform_following_action_list.replace(
"'", "").strip('][').split(', ')
lst = remove_analysis_by()
for i in lst:
for j in perform_following_action_list:
if j in i:
lst.remove(i)
break
return lst
def zero_word_verb(stri):
doc = nlp(stri.strip())
main_verbs = load_lists(fpath)['verbs']
main_verbs = main_verbs.replace("'", "").strip('][').split(', ')
if not (doc[0].tag_ == "MD") and\
not (doc[0].tag_ == "VB" and
str(doc[0]).lower() in main_verbs) and\
not (doc[0].tag_ == "VB" and
str(doc[0]).lower() not in main_verbs) and\
not(str(doc[0]).lower() in main_verbs):
return False
else:
return True
def verb_and_verb(txt):
verbs_list = load_lists(fpath)['verbs']
doc = nlp(txt)
result = ""
for i in range(len(doc) + 2):
if doc[i].pos_ == "VERB" and doc[i + 1].pos_ == "CCONJ" and doc[
i + 2].pos_ == "VERB":
if doc[i].text in verbs_list and doc[i + 2].text in verbs_list:
candidate = doc[i].text + " " + doc[i +
1].text + " " + doc[i +
2].text
result += txt.replace(candidate, doc[i].text) + " "
result += txt.replace(candidate, doc[i + 2].text)
break
return result
def removable_token(
): # When Virus:Win32/Funlove.4099 runs, it performs the following actions:
removable_token_list = load_lists(fpath)['RTL']
removable_token_list = removable_token_list.replace(
"'", "").strip('][').split(', ')
lst = on_the_windows_x_only()
for id, value in enumerate(lst):
for j in removable_token_list:
if value.strip().startswith(
j
): #### definetly remember we should use only startswith()for proper matching
# lst.remove(value)
lst[id] = value.replace(j, " ")
# break
return lst
def following_subject(txt):
following_subject_list = load_lists(fpath)['TFSL']
txt = txt.rstrip()
txt = txt.rstrip('.')
result = ""
for sent in sent_tokenize(txt):
for item in following_subject_list:
if item in sent and ":" in sent:
old_subj = item
new_sub = sent.split(":", 1)[1]
y = iocs.list_of_iocs(sent)
if y:
sentence_replicas = [new_sub[0]] * len(y)
for i in range(len(sentence_replicas)):
k = sent.split(":", 1)[0]
l = k.replace(old_subj, " ")
result += y[i] + l + " . "
break
return result
def findSVOs(tokens):
svos = []
is_pas = _is_passive(tokens)
verbs = [tok for tok in tokens if _is_non_aux_verb(tok)]
# @kia
if verbs == []:
main_verbs = load_lists(fpath)['verbs']
main_verbs = main_verbs.replace("'", "").strip('][').split(', ')
verbs = [tok for tok in tokens if str(tok) in main_verbs]
visited = set() # recursion detection
for v in verbs:
subs, verbNegated = _get_all_subs(v)
# hopefully there are subs, if not, don't examine this verb any longer
if len(subs) > 0:
isConjVerb, conjV = _right_of_verb_is_conj_verb(v)
if isConjVerb:
v2, objs = _get_all_objs(conjV, is_pas)
for sub in subs:
for obj in objs:
objNegated = _is_negated(obj)
if is_pas: # reverse object / subject for passive
svos.append((to_str(expand(obj, tokens, visited)),
"!" + v.lemma_ if verbNegated or objNegated else v.lemma_, to_str(expand(sub, tokens, visited))))
svos.append((to_str(expand(obj, tokens, visited)),
"!" + v2.lemma_ if verbNegated or objNegated else v2.lemma_, to_str(expand(sub, tokens, visited))))
else:
svos.append((to_str(expand(sub, tokens, visited)),
"!" + v.lower_ if verbNegated or objNegated else v.lower_, to_str(expand(obj, tokens, visited))))
svos.append((to_str(expand(sub, tokens, visited)),
"!" + v2.lower_ if verbNegated or objNegated else v2.lower_, to_str(expand(obj, tokens, visited))))
else:
v, objs = _get_all_objs(v, is_pas)
for sub in subs:
for obj in objs:
objNegated = _is_negated(obj)
if is_pas: # reverse object / subject for passive
svos.append((to_str(expand(obj, tokens, visited)),
"!" + v.lemma_ if verbNegated or objNegated else v.lemma_, to_str(expand(sub, tokens, visited))))
else:
svos.append((to_str(expand(sub, tokens, visited)),
"!" + v.lower_ if verbNegated or objNegated else v.lower_, to_str(expand(obj, tokens, visited))))
return svos
def modification_():
final_txt = ''
c = fl = 0
pattern = load_lists(fpath)['MDF']
pattern = pattern.replace("'", "").strip('][').split(', ')
sentences = sent_tokenize(cc)
for sentence in sentences:
c += 1
for value in pattern:
if value in sentence:
sentence1 = sentence.split(
value)[0] + ' modifies ' + sentence.split(value)[1]
final_txt += " " + sentence1 + " "
fl += 2
break
if c > fl:
final_txt += " " + sentence
fl += 1
return final_txt.strip()
def communicate_to_sr(stri):
final_txt = ''
c = fl = 0
pattern = load_lists(fpath)['COMU']
pattern = pattern.replace("'", "").strip('][').split(', ')
sentences = sent_tokenize(stri)
for sentence in sentences:
c += 1
for value in pattern:
if value in sentence:
sentence1 = sentence.split(
value)[0] + ' receives from' + sentence.split(value)[1]
sentence2 = sentence.split(
value)[0] + ' sends to' + sentence.split(value)[1]
final_txt += " " + sentence1 + " " + sentence2
fl += 2
c += 1
break
if c > fl:
final_txt += " " + sentence
fl += 1
return final_txt
def ellipsis_subject(stri):
ellipsis_verbs = load_lists(fpath)['verbs']
ellipsis_verbs = ellipsis_verbs.replace("'", "").strip('][').split(', ')
sent_text = nltk.sent_tokenize(stri)
result = ""
for sentence in sent_text:
token = nltk.word_tokenize(sentence)
doc = nlp(sentence)
if nltk.pos_tag(token)[0][1] == "VB" or nltk.pos_tag(
token)[0][1] == "VBZ" or doc[0].pos_ == "VERB" or doc[
0].text.lower() in ellipsis_verbs:
new_sentence = " It " + nltk.pos_tag(
token)[0][0].lower() + " " + " ".join(sentence.split(" ")[1:])
result += " " + new_sentence
elif doc[0].dep_ == "ROOT":
if doc[0].text.lower in ellipsis_verbs:
new_sentence = " It " + doc[0].text.lower() + " " + " ".join(
sentence.split(" ")[1:])
result += " " + new_sentence
elif doc[0].text.lower() in ellipsis_verbs and doc[0].dep_ != "ROOT":
result += " " + doc.text
else:
result += " " + sentence
return result
def coref_the_following_middle(stri):
final_txt = ''
list2 = load_lists(fpath)['TFL']
list2 = list2.replace("'", "").strip('][').split(', ')
sentences = sent_tokenize(stri)
c = 0
fl = len(final_txt)
for sentence in sentences:
c += 1
for value in list2:
if value in sentence:
sentence.strip()
if sentence[-1] == "." and ":" in sentence:
sentence = sentence[:-1]
sentence2 = sentence.split(":")[0].replace(
value,
sentence.split(":", 1)[1]) + ". "
final_txt += ' ' + sentence2
fl += 1
break
if c > fl:
final_txt += ' ' + sentence
fl += 1
return final_txt
import spacy
from nltk import sent_tokenize
nlp = spacy.load("en_core_web_lg")
from lists_patterns import load_lists, fpath
import main
if not main.args.input_file:
raise ValueError(
"usage: main.py [-h] [--asterisk ASTERISK] [--crf CRF] [--rmdup RMDUP] [--gname GNAME] [--input_file INPUT_FILE]"
)
else:
f = open(main.args.input_file, encoding='iso-8859-1')
txt = f.readlines()
txt = " ".join(txt)
txt = txt.replace('\n', ' ')
titles_list = load_lists(fpath)['MS_TITLES']
titles_list = titles_list.replace("'", "").strip('][').split(', ')
main_verbs = load_lists(fpath)['verbs']
main_verbs = main_verbs.replace("'", "").strip('][').split(', ')
def delete_brackets(stri):
stri = stri.replace("[", "")
stri = stri.replace("]", "")
stri = stri.replace("<", "")
stri = stri.replace(">", "")
return stri
txt = delete_brackets(txt)
txt = txt.strip(" ")
import re
from list_iocs import iocs
from allennlp.predictors.predictor import Predictor
from lists_patterns import load_lists, fpath
from nltk import sent_tokenize
my_svo_triplet, all_nodes = [], []
main_verbs = load_lists(fpath)['verbs']
main_verbs = main_verbs.replace("'", "").strip('][').split(', ')
sentences = r''' '''
# Abstractive/ To be added
def ats():
for sentence in range(sent_tokenize(sentences)):
predictor = Predictor.from_path("srl-model.tar.gz")
predictions = predictor.predict(sentence)
lst = []
nodes = []
for k in predictions['verbs']:
if k['description'].count('[') > 1:
lst.append(k['description'])
for jj in range(len(lst)):
nodes.append([])
for j in re.findall(r"[^[]*\[([^]]*)\]", lst[jj]):
nodes[jj].append(j)
print("*****sentence:", sentence, '*****nodes: ', nodes)
for lis_ in nodes:
for indx in range(len(lis_)):
if lis_[0].split(":", 1)[0].lower().strip() == "v" and lis_[0].split(":", 1)[
1].lower().strip() in main_verbs:
def CـC(txt):
pattern = load_lists(fpath)['C_C']
pattern = pattern.replace("'", "").strip('][').split(', ')
big_regex = re.compile('|'.join(map(re.escape, pattern)), re.IGNORECASE)
sentence = big_regex.sub('remote ip:*', str(txt))
return sentence
