import bonobo
from bonobo import examples
from bonobo.examples.files._services import get_services
def skip_comments(line):
line = line.strip()
if not line.startswith('#'):
yield line
def get_graph(*, _limit=(), _print=()):
return bonobo.Graph(
bonobo.FileReader('datasets/passwd.txt'),
skip_comments,
*_limit,
lambda s: s.split(':')[0],
*_print,
bonobo.FileWriter('usernames.txt', fs='fs.output'),
)
if __name__ == '__main__':
parser = examples.get_argument_parser()
with bonobo.parse_args(parser) as options:
bonobo.run(get_graph(**examples.get_graph_options(options)), services=get_services())
bonobo.CsvReader('datasets/coffeeshops.txt'),
*((bonobo.Limit(_limit), ) if _limit else ()),
*((bonobo.PrettyPrinter(), ) if _print else ()),
bonobo.CsvWriter('coffeeshops.csv', fs='fs.output')
)
if __name__ == '__main__':
parser = bonobo.get_argument_parser()
parser.add_argument(
'--limit',
'-l',
type=int,
default=None,
help='If set, limits the number of processed lines.'
)
parser.add_argument(
'--print',
'-p',
action='store_true',
default=False,
help='If set, pretty prints before writing to output file.'
)
with bonobo.parse_args(parser) as options:
bonobo.run(
get_graph(_limit=options['limit'], _print=options['print']),
services=get_services()
)
import bonobo
from bonobo.examples.files._services import get_services
def get_graph(*, _limit=None, _print=False):
graph = bonobo.Graph()
trunk = graph.add_chain(bonobo.JsonReader('datasets/theaters.json'), *((bonobo.Limit(_limit),) if _limit else ()))
if _print:
graph.add_chain(bonobo.PrettyPrinter(), _input=trunk.output)
graph.add_chain(bonobo.JsonWriter('theaters.json', fs='fs.output'), _input=trunk.output)
graph.add_chain(bonobo.LdjsonWriter('theaters.ldjson', fs='fs.output'), _input=trunk.output)
return graph
if __name__ == '__main__':
parser = bonobo.get_argument_parser()
parser.add_argument('--limit', '-l', type=int, default=None, help='If set, limits the number of processed lines.')
parser.add_argument(
'--print', '-p', action='store_true', default=False, help='If set, pretty prints before writing to output file.'
)
with bonobo.parse_args(parser) as options:
bonobo.run(get_graph(_limit=options['limit'], _print=options['print']), services=get_services())
import bonobo
from bonobo.examples.files._services import get_services
def get_graph(*, _limit=None, _print=False):
return bonobo.Graph(bonobo.CsvReader('datasets/coffeeshops.txt'),
*((bonobo.Limit(_limit), ) if _limit else ()),
*((bonobo.PrettyPrinter(), ) if _print else ()),
bonobo.CsvWriter('coffeeshops.csv', fs='fs.output'))
if __name__ == '__main__':
parser = bonobo.get_argument_parser()
parser.add_argument('--limit',
'-l',
type=int,
default=None,
help='If set, limits the number of processed lines.')
parser.add_argument(
'--print',
'-p',
action='store_true',
default=False,
help='If set, pretty prints before writing to output file.')
with bonobo.parse_args(parser) as options:
bonobo.run(get_graph(_limit=options['limit'], _print=options['print']),
services=get_services())