Rewritting Bags from scratch using a namedtuple approach, along with other (less major) updates.

New bag implementation improves a lot how bonobo works, even if this is
highly backward incompatible (sorry, that's needed, and better sooner
than later).

* New implementation uses the same approach as python's namedtuple,
  by dynamically creating the python type's code. This has drawbacks, as
  it feels like not the right way, but also a lot of benefits that
  cannot be achieved using a regular approach, especially the
  constructor parameter order, hardcoded.
* Memory usage is now much more efficient. The "keys" memory space will
  be used only once per "io type", being spent in the underlying type
  definition instead of in the actual instances.
* Transformations now needs to use tuples as output, which will be bound
  to its "output type". The output type can be infered from the tuple
  length, or explicitely set by the user using either
  `context.set_output_type(...)` or `context.set_output_fields(...)` (to
  build a bag type from a list of field names).

Jupyter/Graphviz integration is more tight, allowing to easily display
graphs in a notebook, or displaying the live transformation status in an
html table instead of a simple <div>.

For now, context processors were hacked to stay working as before but
the current API is not satisfactory, and should be replaced. This new
big change being unreasonable without some time to work on it properly,
it is postponed for next versions (0.7, 0.8, ...). Maybe the best idea
is to have some kind of "local services", that would use the same
dependency injection mechanism as the execution-wide services.

Services are now passed by keywoerd arguments only, to avoid confusion
with data-arguments.
This commit is contained in:
Romain Dorgueil
2017-11-27 00:04:51 +01:00
parent 52ea29afcb
commit 5e0b6567cd
96 changed files with 2958 additions and 1870 deletions

View File

@ -1,170 +0,0 @@
import pickle
from unittest.mock import Mock
import pytest
from bonobo import Bag
from bonobo.constants import INHERIT_INPUT, BEGIN
from bonobo.structs import Token
args = (
'foo',
'bar',
)
kwargs = dict(acme='corp')
def test_basic():
my_callable1 = Mock()
my_callable2 = Mock()
bag = Bag(*args, **kwargs)
assert not my_callable1.called
result1 = bag.apply(my_callable1)
assert my_callable1.called and result1 is my_callable1.return_value
assert not my_callable2.called
result2 = bag.apply(my_callable2)
assert my_callable2.called and result2 is my_callable2.return_value
assert result1 is not result2
my_callable1.assert_called_once_with(*args, **kwargs)
my_callable2.assert_called_once_with(*args, **kwargs)
def test_constructor_empty():
a, b = Bag(), Bag()
assert a == b
assert a.args is ()
assert a.kwargs == {}
@pytest.mark.parametrize(('arg_in', 'arg_out'), (
((), ()),
({}, ()),
(('a', 'b', 'c'), None),
))
def test_constructor_shorthand(arg_in, arg_out):
if arg_out is None:
arg_out = arg_in
assert Bag(arg_in) == arg_out
def test_constructor_kwargs_only():
assert Bag(foo='bar') == {'foo': 'bar'}
def test_constructor_identity():
assert Bag(BEGIN) is BEGIN
def test_inherit():
bag = Bag('a', a=1)
bag2 = Bag.inherit('b', b=2, _parent=bag)
bag3 = bag.extend('c', c=3)
bag4 = Bag('d', d=4)
assert bag.args == ('a', )
assert bag.kwargs == {'a': 1}
assert bag.flags is ()
assert bag2.args == (
'a',
'b',
)
assert bag2.kwargs == {'a': 1, 'b': 2}
assert INHERIT_INPUT in bag2.flags
assert bag3.args == (
'a',
'c',
)
assert bag3.kwargs == {'a': 1, 'c': 3}
assert bag3.flags is ()
assert bag4.args == ('d', )
assert bag4.kwargs == {'d': 4}
assert bag4.flags is ()
bag4.set_parent(bag)
assert bag4.args == (
'a',
'd',
)
assert bag4.kwargs == {'a': 1, 'd': 4}
assert bag4.flags is ()
bag4.set_parent(bag3)
assert bag4.args == (
'a',
'c',
'd',
)
assert bag4.kwargs == {'a': 1, 'c': 3, 'd': 4}
assert bag4.flags is ()
def test_pickle():
bag1 = Bag('a', a=1)
bag2 = Bag.inherit('b', b=2, _parent=bag1)
bag3 = bag1.extend('c', c=3)
bag4 = Bag('d', d=4)
# XXX todo this probably won't work with inheriting bags if parent is not there anymore? maybe that's not true
# because the parent may be in the serialization output but we need to verify this assertion.
for bag in bag1, bag2, bag3, bag4:
pickled = pickle.dumps(bag)
unpickled = pickle.loads(pickled)
assert unpickled == bag
def test_eq_operator_bag():
assert Bag('foo') == Bag('foo')
assert Bag('foo') != Bag('bar')
assert Bag('foo') is not Bag('foo')
assert Bag('foo') != Token('foo')
assert Token('foo') != Bag('foo')
def test_eq_operator_tuple_mixed():
assert Bag('foo', bar='baz') == ('foo', {'bar': 'baz'})
assert Bag('foo') == ('foo', {})
assert Bag() == ({}, )
def test_eq_operator_tuple_not_mixed():
assert Bag('foo', 'bar') == ('foo', 'bar')
assert Bag('foo') == ('foo', )
assert Bag() == ()
def test_eq_operator_dict():
assert Bag(foo='bar') == {'foo': 'bar'}
assert Bag(
foo='bar', corp='acme'
) == {
'foo': 'bar',
'corp': 'acme',
}
assert Bag(
foo='bar', corp='acme'
) == {
'corp': 'acme',
'foo': 'bar',
}
assert Bag() == {}
def test_repr():
bag = Bag('a', a=1)
assert repr(bag) == "Bag('a', a=1)"
def test_iterator():
bag = Bag()
assert list(bag.apply([1, 2, 3])) == [1, 2, 3]
assert list(bag.apply((1, 2, 3))) == [1, 2, 3]
assert list(bag.apply(range(5))) == [0, 1, 2, 3, 4]
assert list(bag.apply('azerty')) == ['a', 'z', 'e', 'r', 't', 'y']

View File

@ -1,4 +1,4 @@
from bonobo.structs import Token
from bonobo.constants import Token
def test_token_repr():