def check_external_relationships(self):
"""
Check whether this file has external relationships (remote template, OLE object, etc).
:returns: :py:class:`Indicator`
"""
ext_rels = Indicator(
'ext_rels',
0,
name='External Relationships',
_type=int,
risk=RISK.NONE,
description=
'External relationships such as remote templates, remote OLE objects, etc',
hide_if_false=False)
self.indicators.append(ext_rels)
# this check only works for OpenXML files
if not self.ftg.is_openxml():
return ext_rels
# to collect relationship types:
rel_types = set()
# open an XmlParser, using a BytesIO instead of filename (to work in memory)
xmlparser = ooxml.XmlParser(self.data_bytesio)
for rel_type, target in oleobj.find_external_relationships(xmlparser):
log.debug('External relationship: type={} target={}'.format(
rel_type, target))
rel_types.add(rel_type)
ext_rels.value += 1
if ext_rels.value > 0:
ext_rels.description = 'External relationships found: {} - use oleobj for details'.format(
', '.join(rel_types))
ext_rels.risk = RISK.HIGH
return ext_rels
def test_iter_subfiles(self):
""" test that limitation on few subfiles works """
testfile = join(DATA_BASE_DIR, 'msodde', 'dde-test.xlsx')
subfiles = ['xl/theme/theme1.xml', 'docProps/app.xml']
parser = ooxml.XmlParser(testfile)
for subfile, elem, depth in parser.iter_xml(subfiles):
if self.DO_DEBUG:
print(u'{0} {1}{2}'.format(subfile, ' ' * depth,
ooxml.debug_str(elem)))
if subfile not in subfiles:
self.fail('should have been skipped: {0}'.format(subfile))
if depth == 0:
subfiles.remove(subfile)
self.assertEqual(subfiles, [],
'missed subfile(s) {0}'.format(subfiles))
def process_xlsx(filepath):
""" process an OOXML excel file (e.g. .xlsx or .xlsb or .xlsm) """
dde_links = []
parser = ooxml.XmlParser(filepath)
for subfile, elem, _ in parser.iter_xml():
tag = elem.tag.lower()
if tag == 'ddelink' or tag.endswith('}ddelink'):
# we have found a dde link. Try to get more info about it
link_info = ['DDE-Link']
if 'ddeService' in elem.attrib:
link_info.append(elem.attrib['ddeService'])
if 'ddeTopic' in elem.attrib:
link_info.append(elem.attrib['ddeTopic'])
dde_links.append(u' '.join(link_info))
# binary parts, e.g. contained in .xlsb
for subfile, content_type, handle in parser.iter_non_xml():
try:
logging.info(
'Parsing non-xml subfile {0} with content type {1}'.format(
subfile, content_type))
for record in xls_parser.parse_xlsb_part(handle, content_type,
subfile):
logging.debug('{0}: {1}'.format(subfile, record))
if isinstance(record, xls_parser.XlsbBeginSupBook) and \
record.link_type == \
xls_parser.XlsbBeginSupBook.LINK_TYPE_DDE:
dde_links.append('DDE-Link ' + record.string1 + ' ' +
record.string2)
except Exception:
if content_type.startswith('application/vnd.ms-excel.') or \
content_type.startswith('application/vnd.ms-office.'): # pylint: disable=bad-indentation
# should really be able to parse these either as xml or records
log_func = logging.warning
elif content_type.startswith('image/') or content_type == \
'application/vnd.openxmlformats-officedocument.' + \
'spreadsheetml.printerSettings':
# understandable that these are not record-base
log_func = logging.debug
else: # default
log_func = logging.info
log_func('Failed to parse {0} of content type {1}'.format(
subfile, content_type))
# in any case: continue with next
return u'\n'.join(dde_links)
def process_xlsx(filepath):
"""process an OOXML excel file (e.g. .xlsx or .xlsb or .xlsm)"""
dde_links = []
parser = ooxml.XmlParser(filepath)
for _, elem, _ in parser.iter_xml():
tag = elem.tag.lower()
if tag == "ddelink" or tag.endswith("}ddelink"):
# we have found a dde link. Try to get more info about it
link_info = []
if "ddeService" in elem.attrib:
link_info.append(elem.attrib["ddeService"])
if "ddeTopic" in elem.attrib:
link_info.append(elem.attrib["ddeTopic"])
dde_links.append(u" ".join(link_info))
# binary parts, e.g. contained in .xlsb
for subfile, content_type, handle in parser.iter_non_xml():
try:
logger.info("Parsing non-xml subfile {0} with content type {1}".format(subfile, content_type))
for record in xls_parser.parse_xlsb_part(handle, content_type, subfile):
logger.debug("{0}: {1}".format(subfile, record))
if isinstance(record, xls_parser.XlsbBeginSupBook) and record.link_type == xls_parser.XlsbBeginSupBook.LINK_TYPE_DDE:
dde_links.append(record.string1 + " " + record.string2)
except Exception as exc:
if content_type.startswith("application/vnd.ms-excel.") or content_type.startswith(
"application/vnd.ms-office."
): # pylint: disable=bad-indentation
# should really be able to parse these either as xml or records
log_func = logger.warning
elif (
content_type.startswith("image/")
or content_type == "application/vnd.openxmlformats-officedocument." + "spreadsheetml.printerSettings"
):
# understandable that these are not record-base
log_func = logger.debug
else: # default
log_func = logger.info
log_func('Failed to parse {0} of content type {1} ("{2}")'.format(subfile, content_type, str(exc)))
# in any case: continue with next
return u"\n".join(dde_links)
def test_iter_tags(self):
""" test that limitation to tags works """
testfile = join(DATA_BASE_DIR, 'msodde', 'harmless-clean.docm')
nmspc = 'http://schemas.openxmlformats.org/wordprocessingml/2006/main'
tag = '{' + nmspc + '}p'
parser = ooxml.XmlParser(testfile)
n_found = 0
for subfile, elem, depth in parser.iter_xml(tags=tag):
n_found += 1
self.assertEqual(elem.tag, tag)
# also check that children are present
n_children = 0
for child in elem:
n_children += 1
self.assertFalse(child.tag == '')
self.assertTrue(
n_children > 0,
'no children for elem {0}'.format(ooxml.debug_str(elem)))
self.assertEqual(n_found, 7)
def process_excel_xml(filepath):
""" find dde links in xml files created with excel 2003 or excel 2007+
TODO: did not manage to create dde-link in the 2007+-xml-format. Find out
whether this is possible at all. If so, extend this function
"""
dde_links = []
parser = ooxml.XmlParser(filepath)
for _, elem, _ in parser.iter_xml():
tag = elem.tag.lower()
if tag != 'cell' and not tag.endswith('}cell'):
continue # we are only interested in cells
formula = None
for key in elem.keys():
if key.lower() == 'formula' or key.lower().endswith('}formula'):
formula = elem.get(key)
break
if formula is None:
continue
log.debug('found cell with formula {0}'.format(formula))
match = re.match(XML_DDE_FORMAT, formula)
if match:
dde_links.append(u' '.join(match.groups()[:2]))
return u'\n'.join(dde_links)
def test_iter_all(self):
""" test iter_xml without args """
expect_subfiles = dict([
('[Content_Types].xml', 11),
('_rels/.rels', 4),
('word/_rels/document.xml.rels', 6),
('word/document.xml', 102),
('word/theme/theme1.xml', 227),
('word/settings.xml', 40),
('word/fontTable.xml', 25),
('word/webSettings.xml', 3),
('docProps/app.xml', 26),
('docProps/core.xml', 10),
('word/styles.xml', 441),
])
n_elems = 0
testfile = join(DATA_BASE_DIR, 'msodde', 'harmless-clean.docx')
for subfile, elem, depth in ooxml.XmlParser(testfile).iter_xml():
n_elems += 1
if depth > 0:
continue
# now depth == 0; should occur once at end of every subfile
if subfile not in expect_subfiles:
self.fail('Subfile {0} not expected'.format(subfile))
self.assertEqual(
n_elems, expect_subfiles[subfile],
'wrong number of elems ({0}) yielded from {1}'.format(
n_elems, subfile))
_ = expect_subfiles.pop(subfile)
n_elems = 0
self.assertEqual(
len(expect_subfiles), 0,
'Forgot to iterate through subfile(s) {0}'.format(
expect_subfiles.keys()))
def process_xlsx(filepath, filed_filter_mode=None):
""" process an OOXML excel file (e.g. .xlsx or .xlsb or .xlsm) """
dde_links = []
parser = ooxml.XmlParser(filepath)
for subfile, elem, _ in parser.iter_xml():
tag = elem.tag.lower()
if tag == 'ddelink' or tag.endswith('}ddelink'):
# we have found a dde link. Try to get more info about it
link_info = ['DDE-Link']
if 'ddeService' in elem.attrib:
link_info.append(elem.attrib['ddeService'])
if 'ddeTopic' in elem.attrib:
link_info.append(elem.attrib['ddeTopic'])
dde_links.append(u' '.join(link_info))
# binary parts, e.g. contained in .xlsb
for subfile, content_type, handle in parser.iter_non_xml():
if content_type == 'application/vnd.openxmlformats-officedocument.' + \
'spreadsheetml.printerSettings':
continue # printer settings
if not content_type.startswith('application/vnd.ms-excel.') and \
not content_type.startswith('application/vnd.ms-office.'): # pylint: disable=bad-indentation
logging.warning('Unexpected content type: ' + content_type)
# try parsing anyway
logging.info('Parsing non-xml subfile {0} with content type {1}'
.format(subfile, content_type))
for record in xls_parser.parse_xlsb_part(handle, content_type, subfile):
logging.debug('{0}: {1}'.format(subfile, record))
if isinstance(record, xls_parser.XlsbBeginSupBook) and \
record.link_type == \
xls_parser.XlsbBeginSupBook.LINK_TYPE_DDE:
dde_links.append('DDE-Link ' + record.string1 + ' ' +
record.string2)
return u'\n'.join(dde_links)
def process_docx(filepath, field_filter_mode=None):
""" find dde-links (and other fields) in Word 2007+ files """
parser = ooxml.XmlParser(filepath)
all_fields = []
level = 0
ddetext = u''
for _, subs, depth in parser.iter_xml(tags=TAG_W_P + TAG_W_FLDSIMPLE):
if depth == 0: # at end of subfile:
level = 0 # reset
if subs.tag in TAG_W_FLDSIMPLE:
# concatenate the attribute of the field, if present:
attrib_instr = subs.attrib.get(ATTR_W_INSTR[0]) or \
subs.attrib.get(ATTR_W_INSTR[1])
if attrib_instr is not None:
all_fields.append(unquote(attrib_instr))
continue
# have a TAG_W_P
for curr_elem in subs:
# check if w:r; parse children to pull out first FLDCHAR/INSTRTEXT
elem = None
if curr_elem.tag in TAG_W_R:
for child in curr_elem:
if child.tag in TAG_W_FLDCHAR or \
child.tag in TAG_W_INSTRTEXT:
elem = child
break
if elem is None:
continue # no fldchar or instrtext in this w:r
else:
elem = curr_elem
if elem is None:
raise BadOOXML(filepath, 'Got "None"-Element from iter_xml')
# check if FLDCHARTYPE and whether "begin" or "end" tag
attrib_type = elem.attrib.get(ATTR_W_FLDCHARTYPE[0]) or \
elem.attrib.get(ATTR_W_FLDCHARTYPE[1])
if attrib_type is not None:
if attrib_type == "begin":
level += 1
if attrib_type == "end":
level -= 1
if level == 0 or level == -1: # edge-case; level gets -1
all_fields.append(ddetext)
ddetext = u''
level = 0 # reset edge-case
# concatenate the text of the field, if present:
if elem.tag in TAG_W_INSTRTEXT and elem.text is not None:
# expand field code if QUOTED
ddetext += unquote(elem.text)
# apply field command filter
log.debug('filtering with mode "{0}"'.format(field_filter_mode))
if field_filter_mode in (FIELD_FILTER_ALL, None):
clean_fields = all_fields
elif field_filter_mode == FIELD_FILTER_DDE:
clean_fields = [field for field in all_fields
if FIELD_DDE_REGEX.match(field)]
elif field_filter_mode == FIELD_FILTER_BLACKLIST:
# check if fields are acceptable and should not be returned
clean_fields = [field for field in all_fields
if not field_is_blacklisted(field.strip())]
else:
raise ValueError('Unexpected field_filter_mode: "{0}"'
.format(field_filter_mode))
return u'\n'.join(clean_fields)
