security-and-privacy/mat2
security-and-privacy/mat2
1
0
Fork 0
mirror of synced 2024-11-22 09:14:23 +01:00
Code Releases Activity

Improve the reliability of the office parser

Browse Source
This commit is contained in:
jvoisin 2018-06-21 23:18:50 +02:00
parent 846a261465
commit 5b38bd7ccd
2 changed files with 12 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
libmat2/office.py
View File

@ -16,6 +16,13 @@ class ArchiveBasedAbstractParser(abstract.AbstractParser):
files_to_keep = set() # type: Set[str] files_to_keep = set() # type: Set[str]
files_to_omit = set() # type: Set[Pattern] files_to_omit = set() # type: Set[Pattern]
def __init__(self, filename):
super().__init__(filename)
try: # better fail here than later
zipfile.ZipFile(self.filename)
except zipfile.BadZipFile:
raise ValueError
def _clean_zipinfo(self, zipinfo: zipfile.ZipInfo) -> zipfile.ZipInfo: def _clean_zipinfo(self, zipinfo: zipfile.ZipInfo) -> zipfile.ZipInfo:
zipinfo.create_system = 3 # Linux zipinfo.create_system = 3 # Linux
zipinfo.comment = b'' zipinfo.comment = b''

5
tests/test_libmat2.py
View File

@ -90,6 +90,11 @@ class TestCorruptedFiles(unittest.TestCase):
os.remove('./tests/data/clean.torrent') os.remove('./tests/data/clean.torrent')
def test_odg(self):
shutil.copy('./tests/data/dirty.png', './tests/data/clean.odg')
with self.assertRaises(ValueError):
office.LibreOfficeParser('./tests/data/clean.odg')
class TestGetMeta(unittest.TestCase): class TestGetMeta(unittest.TestCase):
def test_pdf(self): def test_pdf(self):
p = pdf.PDFParser('./tests/data/dirty.pdf') p = pdf.PDFParser('./tests/data/dirty.pdf')