def _to_cbor(args):
count = 0
ochunk = CborChunk(file_obj=sys.stdout, mode='wb', write_wrapper=to_primitives)
for fpath in args.input_path:
ichunk = Chunk(path=fpath, mode='rb')
for si in ichunk:
count += 1
ochunk.add(si)
if (args.limit is not None) and (count >= args.limit):
break
ichunk.close()
if (args.limit is not None) and (count >= args.limit):
break
ochunk.close()
sys.stderr.write('wrote {0} items\n'.format(count))
class ChunkRoller(object):
def __init__(self, chunk_dir, chunk_max=500, message=StreamItem):
self.chunk_dir = chunk_dir
self.chunk_max = chunk_max
self.t_path = os.path.join(chunk_dir, 'tmp-%d.sc.xz' % random.randint(0, 10**8))
self.o_chunk = None
self.message = message
def add(self, si_or_fc):
'''puts `si_or_fc` into the currently open chunk, which it creates if
necessary. If this item causes the chunk to cross chunk_max,
then the chunk closed after adding.
'''
if self.o_chunk is None:
if os.path.exists(self.t_path):
os.remove(self.t_path)
if self.message == StreamItem:
self.o_chunk = Chunk(self.t_path, mode='wb')
else:
logger.info('Assuming CborChunk for message=%r', type(self.message))
self.o_chunk = CborChunk(self.t_path, mode='wb')
self.o_chunk.add(si_or_fc)
logger.debug('added %d-th item to chunk', len(self.o_chunk))
if len(self.o_chunk) == self.chunk_max:
self.close()
def close(self):
if self.o_chunk:
self.o_chunk.close()
if self.message == StreamItem:
extension = 'sc'
else:
logger.warn('assuming file extension ".cbor"')
extension = 'cbor'
o_path = os.path.join(
self.chunk_dir,
'%d-%s.%s.xz' % (len(self.o_chunk), self.o_chunk.md5_hexdigest, extension)
)
os.rename(self.t_path, o_path)
self.o_chunk = None
logger.info('rolled chunk to %s', o_path)
def add(self, si_or_fc):
'''puts `si_or_fc` into the currently open chunk, which it creates if
necessary. If this item causes the chunk to cross chunk_max,
then the chunk closed after adding.
'''
if self.o_chunk is None:
if os.path.exists(self.t_path):
os.remove(self.t_path)
if self.message == StreamItem:
self.o_chunk = Chunk(self.t_path, mode='wb')
else:
logger.info('Assuming CborChunk for message=%r', type(self.message))
self.o_chunk = CborChunk(self.t_path, mode='wb')
self.o_chunk.add(si_or_fc)
logger.debug('added %d-th item to chunk', len(self.o_chunk))
if len(self.o_chunk) == self.chunk_max:
self.close()
def _to_cbor(args):
count = 0
ochunk = CborChunk(file_obj=sys.stdout, mode='wb', write_wrapper=to_primitives)
for fpath in args.input_path:
ichunk = Chunk(path=fpath, mode='rb')
for si in ichunk:
count += 1
ochunk.add(si)
if (args.limit is not None) and (count >= args.limit):
break
ichunk.close()
if (args.limit is not None) and (count >= args.limit):
break
ochunk.close()
sys.stderr.write('wrote {0} items\n'.format(count))
