def _parse_serializers(self):
ser_conf = self._get_env_conf(SERIALIZATION_CONF_PICKE_FILE_PATH) or {}
serializers = {
'input': get_protocol_from_name(ser_conf.get('input', DEFAULT_INPUT_SERIALIZED)),
'output': get_protocol_from_name(ser_conf.get('output', DEFAULT_OUTPUT_SERIALIZED)),
'inter': get_protocol_from_name(ser_conf.get('inter', DEFAULT_INTER_SERIALIZED)),
}
self._set_serializers(serializers)
def cat(self, path, serializer='raw', tab_separated=False):
"""
Returns a generator over files defined by path
:param path: path to the files
:param serializer: input serializer. Options are json, pickle and raw(default)
:param tab_seperated: boolean if input is tab separated
"""
job = self._hdpm._run_hadoop_cmd('fs', ('-cat', path))
output = job.yield_stdout()
output_serializer = get_protocol_from_name(serializer)
for line in output:
line = line.rstrip()
if tab_separated:
ls = line.split('\t')
if len(ls) > 1:
yield tuple(output_serializer.decode(part) for part in line.split('\t'))
else:
yield output_serializer.decode(line)
else:
yield output_serializer.decode(line)
job.join()
def cat(self, path, serializer='raw', tab_separated=False):
"""
Returns a generator over files defined by path
:param path: path to the files
:param serializer: input serializer. Options are json, pickle and raw(default)
:param tab_seperated: boolean if input is tab separated
"""
job = self._hdpm._run_hadoop_cmd('fs', ('-cat', path))
output = job.yield_stdout()
output_serializer = get_protocol_from_name(serializer)
for line in output:
line = line.rstrip()
if tab_separated:
ls = line.split('\t')
if len(ls) > 1:
yield tuple(
output_serializer.decode(part)
for part in line.split('\t'))
else:
yield output_serializer.decode(line)
else:
yield output_serializer.decode(line)
job.join()
def _parse_serializers(self):
ser_conf = self._get_env_conf(SERIALIZATION_CONF_PICKE_FILE_PATH) or {}
serializers = {
'input':
get_protocol_from_name(
ser_conf.get('input', DEFAULT_INPUT_SERIALIZED)),
'output':
get_protocol_from_name(
ser_conf.get('output', DEFAULT_OUTPUT_SERIALIZED)),
'inter':
get_protocol_from_name(
ser_conf.get('inter', DEFAULT_INTER_SERIALIZED)),
}
self._set_serializers(serializers)