def get_instance(self, db):
"""
Decompresses the instance blob if necessary and returns it as string.
EDACC can store compressed and uncompressed instances. To distinguish
between them, we prepend the ASCII characters "LZMA" to a compressed instance.
"""
table = db.metadata.tables['Instances']
c_instance = table.c['instance']
c_id = table.c['idInstance']
# get prefix
instance_header = db.session.connection().execute(select([func.substring(c_instance, 1, 4)],
c_id == self.idInstance).select_from(
table)).first()[0]
data_length = db.session.connection().execute(select([func.length(c_instance)],
c_id == self.idInstance).select_from(
table)).first()[0]
if data_length > 32 * 1024 * 1024:
return "Instance too large for processing. Please use the EDACC GUI application."
if instance_header == 'LZMA': # compressed instance?
# get blob without LZMA prefix
instance_blob = db.session.connection().execute(select([func.substring(c_instance, 5)],
c_id == self.idInstance).select_from(
table)).first()[0]
return utils.lzma_decompress(instance_blob)
else:
return self.instance
def test_lzma_compression(self):
from edacc import utils
uncompressed_data = "TestData with \x12\x65 weird bytes \n and everything"
compressed_data = utils.lzma_compress(uncompressed_data)
decompressed_data = utils.lzma_decompress(compressed_data)
assert decompressed_data == uncompressed_data
assert len(compressed_data) > 13 # there should always be a 13 bytes header
len_bytes = struct.unpack('<Q', compressed_data[5:13])[0]
assert len_bytes == len(uncompressed_data)
def test_lzma_compression(self):
from edacc import utils
uncompressed_data = "TestData with \x12\x65 weird bytes \n and everything"
compressed_data = utils.lzma_compress(uncompressed_data)
decompressed_data = utils.lzma_decompress(compressed_data)
assert decompressed_data == uncompressed_data
assert len(
compressed_data) > 13 # there should always be a 13 bytes header
len_bytes = struct.unpack('<Q', compressed_data[5:13])[0]
assert len_bytes == len(uncompressed_data)