def testDownloadsSingleHuntFileIfAuthorizationIsPresent(self):
hunt_urn = self._CreateHuntWithDownloadedFile()
hunt_id = hunt_urn.Basename()
results = self.GetHuntResults(hunt_urn)
self.RequestAndGrantHuntApproval(hunt_id)
self.Open("/")
self.Click("css=a[grrtarget=hunts]")
self.Click("css=td:contains('%s')" % hunt_id)
self.Click("css=li[heading=Results]")
if data_store.RelationalDBEnabled():
fd = file_store.OpenFile(
flow_export.CollectionItemToClientPath(results[0]))
else:
fd = aff4.FACTORY.Open(flow_export.CollectionItemToAff4Path(
results[0]),
token=self.token)
with mock.patch.object(fd.__class__, "Read") as mock_obj:
self.Click(
"css=grr-results-collection button:has(span.glyphicon-download)"
)
self.WaitUntil(lambda: mock_obj.called)
def Handle(self, args, context=None):
if not args.hunt_id:
raise ValueError("ApiGetHuntFileArgs.hunt_id can't be unset")
if not args.client_id:
raise ValueError("ApiGetHuntFileArgs.client_id can't be unset")
if not args.vfs_path:
raise ValueError("ApiGetHuntFileArgs.vfs_path can't be unset")
if not args.timestamp:
raise ValueError("ApiGetHuntFileArgs.timestamp can't be unset")
api_vfs.ValidateVfsPath(args.vfs_path)
path_type, components = rdf_objects.ParseCategorizedPath(args.vfs_path)
expected_client_path = db.ClientPath(
str(args.client_id), path_type, components)
results = data_store.REL_DB.ReadHuntResults(
str(args.hunt_id),
offset=0,
count=self.MAX_RECORDS_TO_CHECK,
with_timestamp=args.timestamp)
for item in results:
try:
# Do not pass the client id we got from the caller. This will
# get filled automatically from the hunt results and we check
# later that the aff4_path we get is the same as the one that
# was requested.
client_path = export.CollectionItemToClientPath(item, client_id=None)
except export.ItemNotExportableError:
continue
if client_path != expected_client_path:
continue
try:
# TODO(user): this effectively downloads the latest version of
# the file and always disregards the timestamp. Reconsider this logic
# after AFF4 implementation is gone. We also most likely don't need
# the MAX_RECORDS_TO_CHECK logic in the new implementation.
file_obj = file_store.OpenFile(client_path)
return api_call_handler_base.ApiBinaryStream(
"%s_%s" % (args.client_id, os.path.basename(file_obj.Path())),
content_generator=self._GenerateFile(file_obj),
content_length=file_obj.size)
except (file_store.FileHasNoContentError, file_store.FileNotFoundError):
break
raise HuntFileNotFoundError(
"File %s with timestamp %s and client %s "
"wasn't found among the results of hunt %s" %
(args.vfs_path, args.timestamp, args.client_id, args.hunt_id))
def Generate(self, items, token=None):
"""Generates archive from a given collection.
Iterates the collection and generates an archive by yielding contents
of every referenced AFF4Stream.
Args:
items: Iterable of rdf_client_fs.StatEntry objects
token: User's ACLToken.
Yields:
Binary chunks comprising the generated archive.
"""
del token # unused, to be removed with AFF4 code
client_ids = set()
for item_batch in collection.Batch(items, self.BATCH_SIZE):
client_paths = set()
for item in item_batch:
try:
client_path = flow_export.CollectionItemToClientPath(
item, self.client_id)
except flow_export.ItemNotExportableError:
continue
if not self.predicate(client_path):
self.ignored_files.add(client_path)
self.processed_files.add(client_path)
continue
client_ids.add(client_path.client_id)
client_paths.add(client_path)
for chunk in file_store.StreamFilesChunks(client_paths):
self.processed_files.add(chunk.client_path)
for output in self._WriteFileChunk(chunk=chunk):
yield output
self.processed_files |= client_paths - (self.ignored_files
| self.archived_files)
if client_ids:
for client_id, client_info in iteritems(
data_store.REL_DB.MultiReadClientFullInfo(client_ids)):
client = api_client.ApiClient().InitFromClientInfo(client_info)
for chunk in self._GenerateClientInfo(client_id, client):
yield chunk
for chunk in self._GenerateDescription():
yield chunk
yield self.archive_generator.Close()
def Generate(self, items, token=None):
"""Generates archive from a given collection.
Iterates the collection and generates an archive by yielding contents
of every referenced AFF4Stream.
Args:
items: Iterable of rdf_client_fs.StatEntry objects
token: User's ACLToken.
Yields:
Binary chunks comprising the generated archive.
"""
del token # unused, to be removed with AFF4 code
client_ids = set()
for item_batch in collection.Batch(items, self.BATCH_SIZE):
fds_to_write = {}
for item in item_batch:
try:
urn = flow_export.CollectionItemToAff4Path(
item, self.client_id)
client_path = flow_export.CollectionItemToClientPath(
item, self.client_id)
except flow_export.ItemNotExportableError:
continue
fd = file_store.OpenFile(client_path)
self.total_files += 1
if not self.predicate(client_path):
self.ignored_files.append(utils.SmartUnicode(urn))
continue
client_ids.add(client_path.client_id)
# content_path = os.path.join(self.prefix, *urn_components)
self.archived_files += 1
# Make sure size of the original file is passed. It's required
# when output_writer is StreamingTarWriter.
st = os.stat_result((0o644, 0, 0, 0, 0, 0, fd.size, 0, 0, 0))
fds_to_write[fd] = (client_path, urn, st)
if fds_to_write:
for fd, (client_path, urn, st) in iteritems(fds_to_write):
try:
for i, chunk in enumerate(
file_store.StreamFilesChunks([client_path])):
if i == 0:
target_path = os.path.join(
self.prefix,
urn.Path()[1:])
yield self.archive_generator.WriteFileHeader(
target_path, st=st)
yield self.archive_generator.WriteFileChunk(
chunk.data)
yield self.archive_generator.WriteFileFooter()
except Exception as exception: # pylint: disable=broad-except
logging.exception(exception)
self.archived_files -= 1
self.failed_files.append(unicode(urn))
if self.archive_generator.is_file_write_in_progress:
yield self.archive_generator.WriteFileFooter()
if client_ids:
for client_id, client_info in iteritems(
data_store.REL_DB.MultiReadClientFullInfo(client_ids)):
client = api_client.ApiClient().InitFromClientInfo(client_info)
for chunk in self._GenerateClientInfo(client_id, client):
yield chunk
for chunk in self._GenerateDescription():
yield chunk
yield self.archive_generator.Close()