def getTextualHash(self, driver): """ returns a textual hash of the url passed """ #get the html source html_source = driver.page_source #create and update our nilsimsa object with the source nilsimsaObj = Nilsimsa() nilsimsaObj.update(html_source) return nilsimsaObj.hexdigest()
def test_nilsimsa():
"""
tests the nilsimsa hash by choosing a random test file
computes the nilsimsa digest and compares to the true
value stored in the pickled sid_to_nil dictionary
"""
fname = random.choice(dircache.listdir(test_data_dir))
f = open(os.path.join(test_data_dir, fname), "rb")
nil = Nilsimsa(f.read())
f.close()
assert nil.hexdigest() == sid_to_nil[fname.split(".")[0]]
def test_nilsimsa():
"""
tests the nilsimsa hash by choosing a random test file
computes the nilsimsa digest and compares to the true
value stored in the pickled sid_to_nil dictionary
"""
fname = random.choice(dircache.listdir(test_data_dir))
f = open(os.path.join(test_data_dir, fname), "rb")
nil = Nilsimsa(f.read())
f.close()
assert nil.hexdigest() == sid_to_nil[fname.split(".")[0]]
def ToHash(h1,h2, classes):
try:
th = tlsh.hash(h1)
except:
th = 'None'
try:
sh = fuzzyhashlib.sdhash(h1).hexdigest().rstrip()
except:
sh = 'None'
try:
nil = Nilsimsa(h1).hexdigest()
except:
nil = 'None'
try:
ss = fuzzyhashlib.ssdeep(h1).hexdigest()
except:
ss = 'None'
ch = []
if classes!=None:
for c in classes:
name = c[0]
content = c[1]
try:
cnil = Nilsimsa(content).hexdigest()
except:
cnil = 'None'
try:
css = fuzzyhashlib.ssdeep(content).hexdigest()
except:
css = 'None'
try:
csh = 'None'
if len(content) >= 512:
csh = fuzzyhashlib.sdhash(content).hexdigest().rstrip()
except:
csh = 'None'
try:
cth = 'None'
if len(content) >= 256:
cth = tlsh.hash(content)
except:
cth = 'None'
ch.append((name,cth,csh,cnil,css))
return th,sh,nil,ss,ch
def __init__(self,string,parent,parent_filename,lines,type):
self.string=string # string dump of TC/KW from parser
self.parent=parent # TC/KW name
self.parent_filename=parent_filename # parent filename where TC/KW is located
self.lines=lines # line numbers of TC/KW
self.type=type # type - KW or TC
self.digest=Nilsimsa(self.string) # hash
def getNilsimsaHash(self, url, call_phantom=True):
if call_phantom: self.setUpGetter(url)
# if not output file exists, then the page failed to load
if not os.path.isfile("{}-output.txt".format(self.id)):
return -1
#create and update our nilsimsa object with the source
try:
with open("{}-output.txt".format(self.id), "rb") as f:
nilsimsaObj = Nilsimsa(f.read())
#nilsimsaObj.from_file("output.txt")
self.nilsimsa_hash = nilsimsaObj.hexdigest()
except Exception as e:
print(e)
finally:
# always remove the old file even if an exception is thrown
os.remove('{}-output.txt'.format(self.id))
#test = True
return self.nilsimsa_hash
def get_hash(text, hash_function="ssdeep"):
"""
Generates hashed text using one of several available hashing functions.
:param text: The string to hash
:type text: str
:param hash_function: The specific algorithm to use; options are ``'nilsimsa'``, ``'md5'``, and ``'ssdeep'`` \
(default)
:type hash_function: str
:return: A hashed representation of the provided string
:rtype: str
.. note:: The string will be passed through :py:func:`pewtils.decode_text` and the returned value will be used \
instead of the original value if it runs successfully, in order to ensure consistent hashing in both Python 2 and \
3. By default the function uses the :py:mod:`ssdeep` algorithm, which generates context-sensitive hashes that are \
useful for computing document similarities at scale.
.. note:: Using `hash_function='ssdeep'` requires the :py:mod:`ssdeep` library, which is not installed by default \
because it requires the installation of additional system libraries on certain operating systems. For help \
installing ssdeep, refer to the pewtils documentation installation section, which provides OS-specific instructions.
Usage::
from pewtils import get_hash
>>> text = 'test_string'
>>> get_hash(text)
'3:HI2:Hl'
"""
decoded_text = decode_text(text).encode("utf8").strip()
if decoded_text == "":
decoded_text = text
text = decoded_text
if hash_function == "nilsimsa":
from nilsimsa import Nilsimsa
hashed = Nilsimsa(text).hexdigest()
elif hash_function == "md5":
hashed = md5(text).hexdigest()
else:
try:
import ssdeep
except ImportError:
raise Exception("""
To use get_hash with hash_function='ssdeep' you need to install the ssdeep package. Try running:
>> BUILD_LIB=1 pip install ssdeep
If you encounter installation problems, refer to the pewtils documentation for troubleshooting help.
""")
hashed = ssdeep.hash(text)
return hashed
def test_compatability():
"""
testing compat with deprecated version by comparing nilsimsa
scores of 5 randomly selected documents from the test corpus
and asserting that both give the same hexdigest
"""
names = dircache.listdir(test_data_dir)
fnames = set([random.choice(names) for i in range(5)])
for fname in fnames:
f = open(os.path.join(test_data_dir, fname), "rb")
text = f.read()
f.close()
if not (Nilsimsa(text).hexdigest() == orig_Nilsimsa(text).hexdigest()):
assert False
assert True
def test_nilsimsa_speed():
"""
computes nilsimsa hash for all test files and prints speed
"""
corpus = []
for fname in listdir(test_data_dir):
f = open(os.path.join(test_data_dir, fname), "rb")
corpus.append(f.read())
f.close()
start = time.time()
for text in corpus:
Nilsimsa(text)
elapsed = time.time() - start
print("%d in %f --> %f per second" % (
len(corpus), elapsed, len(corpus)/elapsed))
def get_nilsimsa():
dbconn = MySQLdb.connect('10.141.221.73', 'root', 'root', 'fdroid')
dbcursor = dbconn.cursor()
# sql = 'select block_id,block_code from fdroid.cc_block where detection_id=1 and detection_tp = "20150101--20150131"'
sql = 'select detection_tp,block_id,block_code from fdroid.cc_block where detection_id=1'
dbcursor.execute(sql)
f = open('nilsimsa1.txt', 'w')
start = time.clock()
for i in dbcursor.fetchall():
f.write(
str(i[0]) + ' ' + str(i[1]) + ' ' +
Nilsimsa(i[2]).hexdigest() + '\n')
print i[1]
end = time.clock()
f.write(str(end - start))
f.close()
dbcursor.close()
dbconn.close()
def get_hash(text, hash_function="ssdeep"):
"""
Generates hashed text using one of several available hashing functions.
:param text: The string to hash
:type text: str
:param hash_function: The specific algorithm to use; options are ``'nilsimsa'``, ``'md5'``, and ``'ssdeep'`` \
(default)
:type hash_function: str
:return: A hashed representation of the provided string
:rtype: str
.. note:: The string will be passed through :py:func:`pewtils.decode_text` and the returned value will be used \
instead of the original value if it runs successfully, in order to ensure consistent hashing in both Python 2 and \
3. By default the function uses the :py:mod:`ssdeep` algorithm, which generates context-sensitive hashes that are \
useful for computing document similarities at scale.
Usage::
from pewtils import get_hash
>>> text = 'test_string'
>>> get_hash(text)
'3:HI2:Hl'
"""
decoded_text = decode_text(text).encode("utf8").strip()
if decoded_text == "":
decoded_text = text
text = decoded_text
if hash_function == "nilsimsa":
from nilsimsa import Nilsimsa
hashed = Nilsimsa(text).hexdigest()
elif hash_function == "md5":
hashed = md5(text).hexdigest()
else:
import ssdeep
hashed = ssdeep.hash(text)
return hashed
def nilsimsa_hash(text):
if isinstance(text, unicode):
text = text.encode('utf8')
return Nilsimsa(text).hexdigest()
import extractToken
import getCodeFragment
import os
try:
import cPickle as pickle
except ImportError:
import pickle
test_data_dir = os.path.join(os.path.dirname(__file__),
"nilsimsa\\test_data\\")
test_data = "test_dict.p"
test_dict = os.path.join(test_data_dir, test_data)
sid_to_nil = pickle.load(open(test_dict, "rb"))
# print sid_to_nil
nil = Nilsimsa('0' * 64)
s1 = nil.hexdigest()
nil = Nilsimsa('0' * 63 + '1')
s2 = nil.hexdigest()
print s1, s2
print compare_digests(s1, s2)
# for i in range(1,30):
# cloneGroup = getCodeFragment.getCloneClass('1.2.txt', i)
# s1 = Nilsimsa(cloneGroup[0]).hexdigest()
# s2 = Nilsimsa(cloneGroup[1]).hexdigest()
# #print s1,s2
# print compare_digests(s1,s2)
# if compare_digests(s1,s2) <0:
# getCodeFragment.printCloneClass('1.2.txt', i)
def get_ads(base_url):
c = conn.cursor()
page = download.get(base_url + "/search/cpg")
for p in page.select(".row"):
pid = p['data-pid']
a_tag = p.find('a', class_='hdrlnk')
ad_href = a_tag['href']
ad_title = a_tag.text
dt = p.find('time')['datetime']
dt = datetime.datetime.strptime(dt, "%Y-%m-%d %H:%M")
dt = int(dt.strftime("%s"))
c.execute("SELECT * FROM ad WHERE id = ?", (pid, ))
row = c.fetchone()
if row is None:
url = ad_href
if not ad_href.startswith('http'):
url = base_url + ad_href
time.sleep(0.5)
ad = download.get(url)
print url
ad_text = ad.find(id='postingbody')
if ad_text is None:
if ad.find(id='has_been_removed'):
continue
else:
raise "malformed body"
ad_text = ad_text.text.strip()
ad_text = filter(lambda x: x in string.printable, ad_text)
nilsimsa = Nilsimsa(ad_text)
lshash = nilsimsa.hexdigest()
# c.execute("SELECT * FROM ad")
# row = c.fetchone()
# while row:
# diff = nilsimsa.compare(row[4], True)
# if diff < 10:
# print diff
# print cache.get("text:" + row[0])
# print "----"
# print ad_text
# sys.exit()
seen = generate_word_counts(ad_text)
cache.write("text:" + pid, ad_text)
row = (pid, url, ad_title, dt, lshash)
c.execute(
"INSERT INTO ad (id, url, title, posted, lshash) " +
" VALUES (?,?,?,?,?)", row)
for word in seen:
if word not in stopwords:
row = (pid, word, seen[word])
c.execute(
"INSERT INTO tf (id, word, cnt) " + "VALUES (?,?,?)",
row)
conn.commit()
},
}
if data['last-modified']:
try:
last_modified = int(
datetime.datetime(
*eut.parsedate(data['last-modified'])[:6]).strftime('%s'))
except Exception, exc:
logger.info('failed to parse last-modified=%r',
data['last-modified'])
last_modified = 0
else:
last_modified = 0
doc_id = md5(data['content-location']).hexdigest()
content_hash = Nilsimsa(data['body']).hexdigest()
file_id = (doc_id, last_modified, content_hash)
file_id_str = '%s-%d-%s' % file_id
kvlclient.setup_namespace(highlights_kvlayer_tables)
if data['store'] is False:
kvlclient.delete('files', (file_id[0], ))
kvlclient.delete('highlights', (file_id[0], ))
logger.info('cleared all store records related to doc_id=%r',
file_id[0])
else: # storing is allowed
payload_strs = list(kvlclient.get('highlights', file_id))
if payload_strs and payload_strs[0][1]:
payload_str = payload_strs[0][1]
try:
payload = json.loads(payload_str)
def calc_nilsimsa(self, gold_surface_form, comp_surface_form):
nil_0 = Nilsimsa(gold_surface_form)
nil_1 = Nilsimsa(comp_surface_form)
nil = compare_digests(nil_0.hexdigest(), nil_1.hexdigest())
return nil
def compute_hash(self, text):
from nilsimsa import Nilsimsa
result = Nilsimsa(data=text)
result = result.hexdigest()
return str(result)
def test_unicode():
"""
ensures that feeding unicode to Nilsimsa behaves gracefully
"""
nil = Nilsimsa(u'\u1F631')
assert nil.hexdigest()
def get_nilsimsa(string):
return Nilsimsa(string).hexdigest()
[] # ssdeep
]
compare_results = [
[], # nilsimsa
[], # tlsh
[] # ssdeep
]
for filename in os.listdir('.'):
if filename.startswith("prog"):
with open(filename, "rb") as file:
file_data = file.read()
# nilsimsa
hashes[0].append(Nilsimsa(file_data))
# tlsh
hashes[1].append((filename, tlsh.hash(file_data)))
# ssdeep
output = subprocess.Popen(["ssdeep.exe", filename],
stdout=subprocess.PIPE).communicate()[0]
hashes[2].append((filename, output))
#hashes[2].append((filename, str(output.splitlines()[2]).split("\'", 1)[1].split(",", 1)[0]))
print("")
print("nilsimsa (different 0 - 128 similar)")
for e in hashes[0]:
print(str(e.hexdigest()))
compare_results[0].append(hashes[0][0].compare(e.hexdigest(), True))
