def iterator(self, start=None, stop=None):
if start:
try:
index = self._tidindex
except AttributeError:
logging.info("Building index for faster lookup of"
" transactions in the FileStorage DB.")
# Cache a sorted list of all the file pos from oid index.
# To reduce memory usage, the list is splitted in arrays of
# low order 32-bit words.
tindex = defaultdict(lambda: array(typecode))
for x in self._index.itervalues():
tindex[x >> 32].append(x & 0xffffffff)
index = self._tidindex = []
for h, l in sorted(tindex.iteritems()):
l = array(typecode, sorted(l))
x = self._read_data_header(h << 32 | l[0])
index.append((x.tid, h, l))
logging.info("... index built")
x = bisect(index, (start,)) - 1
if x >= 0:
x, h, index = index[x]
x = self._read_data_header
h = x(h << 32 | index[bisect(index, Start(x, h, start)) - 1])
return FileIterator(self._file_name, start, stop, h.tloc)
return FileIterator(self._file_name, start, stop)
def run(path, tid):
f = open(path, "rb")
f.seek(0, 2)
th = prev_txn(f)
while (str(TimeStamp(th.tid)) != tid):
th = th.prev_txn()
reader = Reader()
iterator = FileIterator(path, pos=th._pos)
header = TxnHeader(f, th._pos)
for i in iterator:
if (str(TimeStamp(i.tid)) == tid):
print "\nTRANSACTION: ", TimeStamp(
i.tid), i.user, i.description, pretty_size(header.length), "\n"
header = header.next_txn()
object_types = {}
for o in i:
ot = reader.getIdentity(o.data)
print " - ", ot, pretty_size(len(o.data))
ob = cPickle.loads(o.data)
# Not sure why some objects are stored as tuple (object, ())
if type(ob) == tuple and len(ob) == 2:
ob = ob[0]
if hasattr(ob, "__dict__"):
for i in ob.__dict__.items():
if str(i[0]) == "__doc__":
print "\t('__doc__',", i[1], ")"
elif not callable(i[1]):
print "\t", i
else:
print "can't extract:" + str(ob)
break
f.close()
def run(path, days, notPacked):
f = open(path, "rb")
f.seek(0, 2)
now = datetime.date.today()
#day->size
stats = {}
th = prev_txn(f)
bool = True
while bool:
ts = TimeStamp(th.tid)
then = datetime.date(int(ts.year()), int(ts.month()), int(ts.day()))
delta = timedelta(days=int(days))
if( not(now - then < delta)):
bool = False
th = th.prev_txn()
reader = Reader()
iterator = FileIterator(path, pos=th._pos)
for i in iterator:
object_types = {}
for o in i:
ot = reader.getIdentity(o.data)
try:
stats[ot] = stats[ot] + 1
except KeyError:
stats[ot] = 1
f.close()
for (o,n) in sorted(stats.items(), key=lambda (k,v): v, reverse=True):
print "%6d: %s" % (n,o)
def run(path, ntxn, orderTransactions):
f = open(path, "rb")
f.seek(0, 2)
th = prev_txn(f)
for i in range(ntxn - 1):
th = th.prev_txn()
reader = Reader()
iterator = FileIterator(path, pos=th._pos)
header = TxnHeader(f, th._pos)
transactions = []
for i in iterator:
transactions.append({
"tid": TimeStamp(i.tid),
"user": i.user,
"desc": i.description,
"len": header.length,
"objs": None
})
header = header.next_txn()
object_types = {}
for o in i:
ot = reader.getIdentity(o.data)
if ot in object_types:
size, count = object_types[ot]
object_types[ot] = (size + len(o.data), count + 1)
else:
object_types[ot] = (len(o.data), 1)
keys = object_types.keys()
transactions[-1]["objs"] = object_types
f.close()
if orderTransactions:
transactions = sorted(transactions,
key=lambda (d): d["len"],
reverse=True)
for tr in transactions:
print "\n\nTRANSACTION: ", tr["tid"], tr["user"], tr[
"desc"], pretty_size(tr["len"])
object_types = tr["objs"]
keys = object_types.keys()
for k in sorted(keys, key=lambda (k): object_types[k][0],
reverse=True):
# count, class, size (aggregate)
print " - ", object_types[k][1], k, pretty_size(object_types[k][0])
def iterator(self, start=None, stop=None):
if start:
try:
index = self._tidindex
except AttributeError:
# Cache a sorted list of all the file pos from oid index.
# To reduce memory usage, the list is splitted in arrays of
# low order 32-bit words.
tindex = defaultdict(lambda: array(typecode))
for x in self._index.itervalues():
tindex[x >> 32].append(x & 0xffffffff)
index = self._tidindex = []
for h, l in sorted(tindex.iteritems()):
x = array('I')
x.fromlist(sorted(l))
l = self._read_data_header(h << 32 | x[0])
index.append((l.tid, h, x))
x = bisect(index, (start, )) - 1
if x >= 0:
x, h, index = index[x]
x = self._read_data_header
h = x(h << 32 | index[bisect(index, Start(x, h, start)) - 1])
return FileIterator(self._file_name, start, stop, h.tloc)
return FileIterator(self._file_name, start, stop)
