def test_listDirectory_skip(self):
from string import ascii_uppercase
SNAPSHOT_ID = 'listDirectory_nested'
SUBDIR = 'subdir'
FILENAME1 = 'nested.txt'
FILENAME2 = 'another.txt'
FILENAME3 = 'another.bak'
site, tool, ctx = self._makeOne(SNAPSHOT_ID)
self._makeFile(tool,
SNAPSHOT_ID,
FILENAME1,
printable_bytes,
subdir=SUBDIR)
self._makeFile(tool,
SNAPSHOT_ID,
FILENAME2,
ascii_uppercase.encode('utf-8'),
subdir=SUBDIR)
self._makeFile(tool, SNAPSHOT_ID, FILENAME3, b'abc', subdir=SUBDIR)
names = ctx.listDirectory(SUBDIR,
skip=(FILENAME1, ),
skip_suffixes=('.bak', ))
self.assertEqual(len(names), 1)
self.assertFalse(FILENAME1 in names)
self.assertTrue(FILENAME2 in names)
self.assertFalse(FILENAME3 in names)
def test_listDirectory_multiple(self):
from string import ascii_uppercase
SNAPSHOT_ID = 'listDirectory_nested'
SUBDIR = 'subdir'
FILENAME1 = 'nested.txt'
FILENAME2 = 'another.txt'
site, tool, ctx = self._makeOne(SNAPSHOT_ID)
self._makeFile(tool,
SNAPSHOT_ID,
FILENAME1,
printable_bytes,
subdir=SUBDIR)
self._makeFile(tool,
SNAPSHOT_ID,
FILENAME2,
ascii_uppercase.encode('utf-8'),
subdir=SUBDIR)
names = ctx.listDirectory(SUBDIR)
self.assertEqual(len(names), 2)
self.assertTrue(FILENAME1 in names)
self.assertTrue(FILENAME2 in names)
class Sanitizer(ProcessChunk):
'''Take a TAQ file and make it fake while preserving structure'''
# These could be overriden as desired
fudge_columns = ['Bid_Price', 'Bid_Size', 'Ask_Price', 'Ask_Size']
# This will preserve the fake symbol across chunks
symbol_map = {}
ascii_bytes = ascii_uppercase.encode('ascii')
def _process_chunks(self, iterator_in):
'''Return chunks with changed symbols and fudged times and values.
For now, successive calls will result in a dropped chunk.'''
# last_symbol = None
for chunk in iterator_in:
# XXX a little annoying AND undocumented that split makes
# thing unwriteable. Should double-check.
chunk.flags.writeable = True
self.fake_symbol_replace(chunk)
self.fudge_up(chunk)
yield chunk
def fake_symbol_replace(self, chunk, symbol_column='Symbol_root'):
'''Make a new fake symbol if we don't have it yet, and return it'''
real_symbol = chunk[symbol_column][0]
new_fake_symbol = bytes(sample(self.ascii_bytes, len(real_symbol)))
fake_symbol = self.symbol_map.setdefault(real_symbol, new_fake_symbol)
chunk[symbol_column] = fake_symbol
def fudge_up(self, chunk):
'''Increase each entry in column by some random increment.
Make sure the values stay monotonic, and don't get bigger than
max_value.'''
for col in self.fudge_columns:
# Note that we don't worry about decimal place here - just treating
# everything as an integer is fine for this purpose
data = chunk[col].astype(np.int64)
mean_val = np.mean(data)
std_val = np.std(data)
fake_data = (np.random.standard_normal(len(data)) * std_val +
mean_val).astype(np.int64)
# np.min wasn't working here
fake_data[fake_data < 0] = 0
num_bytes = len(chunk[0][col])
fake_bytes = np.char.zfill(
fake_data.astype('S{}'.format(num_bytes)), num_bytes)
# this is where the side-effects happen
chunk[col] = fake_bytes
def test_listDirectory_multiple(self):
from string import ascii_uppercase
SNAPSHOT_ID = 'listDirectory_nested'
SUBDIR = 'subdir'
FILENAME1 = 'nested.txt'
FILENAME2 = 'another.txt'
site, tool, ctx = self._makeOne(SNAPSHOT_ID)
self._makeFile(tool, SNAPSHOT_ID, FILENAME1, printable_bytes,
subdir=SUBDIR)
self._makeFile(tool, SNAPSHOT_ID, FILENAME2,
ascii_uppercase.encode('utf-8'), subdir=SUBDIR)
names = ctx.listDirectory(SUBDIR)
self.assertEqual(len(names), 2)
self.assertTrue(FILENAME1 in names)
self.assertTrue(FILENAME2 in names)
def test_listDirectory_skip(self):
from string import ascii_uppercase
SNAPSHOT_ID = 'listDirectory_nested'
SUBDIR = 'subdir'
FILENAME1 = 'nested.txt'
FILENAME2 = 'another.txt'
FILENAME3 = 'another.bak'
site, tool, ctx = self._makeOne(SNAPSHOT_ID)
self._makeFile(tool, SNAPSHOT_ID, FILENAME1, printable_bytes,
subdir=SUBDIR)
self._makeFile(tool, SNAPSHOT_ID, FILENAME2,
ascii_uppercase.encode('utf-8'), subdir=SUBDIR)
self._makeFile(tool, SNAPSHOT_ID, FILENAME3, b'abc', subdir=SUBDIR)
names = ctx.listDirectory(SUBDIR, skip=(FILENAME1,),
skip_suffixes=('.bak',))
self.assertEqual(len(names), 1)
self.assertFalse(FILENAME1 in names)
self.assertTrue(FILENAME2 in names)
self.assertFalse(FILENAME3 in names)
if c in ignore:
continue
if len(l) and c ^ l[-1] == 0x20:
l.pop()
else:
l.append(c)
return l
advent.setup(2018, 5)
fin = advent.get_input(mode='rb')
polymer = fin.read().rstrip()
trimmed = react_fast(polymer)
reacted_len = len(trimmed)
advent.print_answer(1, reacted_len)
best_reacted_len = reacted_len
for l, L in zip(ascii_lowercase.encode(), ascii_uppercase.encode()):
reacted_len = len(react_fast(trimmed, {l, L}))
if reacted_len < best_reacted_len:
best_reacted_len = reacted_len
advent.print_answer(2, best_reacted_len)
def isupper(b):
return all(x in ascii_uppercase.encode() for x in b)