Versions follow CalVer with a strict backwards compatibility policy. The third digit is only for regressions.
Changes for the upcoming release can be found in the "changelog.d" directory in our repository.
Comparing subclasses using
<
,>
,<=
, and>=
is now deprecated. The docs always claimed that instances are only compared if the types are identical, so this is a first step to conform to the docs.Equality operators (
==
and!=
) were always strict in this regard. #394
attrs
now ships its own PEP 484 type hints. Together with mypy'sattrs
plugin, you've got all you need for writing statically typed code in both Python 2 and 3!At that occasion, we've also added narrative docs about type annotations in
attrs
. #238Added kw_only arguments to
attr.ib
andattr.s
, and a corresponding kw_only attribute toattr.Attribute
. This change makes it possible to have a generated__init__
with keyword-only arguments on Python 3, relaxing the required ordering of default and non-default valued attributes. #281, #411The test suite now runs with
hypothesis.HealthCheck.too_slow
disabled to prevent CI breakage on slower computers. #364, #396attr.validators.in_()
now raises aValueError
with a useful message even if the options are a string and the value is not a string. #383attr.asdict()
now properly handles deeply nested lists and dictionaries. #395Added
attr.converters.default_if_none()
that allows to replaceNone
values in attributes. For exampleattr.ib(converter=default_if_none(""))
replacesNone
by empty strings. #400, #414Fixed a reference leak where the original class would remain live after being replaced when
slots=True
is set. #407Slotted classes can now be made weakly referenceable by passing
@attr.s(weakref_slot=True)
. #420Added cache_hash option to
@attr.s
which causes the hash code to be computed once and stored on the object. #425Attributes can be named
property
anditemgetter
now. #430It is now possible to override a base class' class variable using only class annotations. #431
x=X(); x.cycle = x; repr(x)
will no longer raise aRecursionError
, and will instead show asX(x=...)
.attr.ib(factory=f)
is now syntactic sugar for the common case ofattr.ib(default=attr.Factory(f))
.Added
attr.field_dict()
to return an ordered dictionary ofattrs
attributes for a class, whose keys are the attribute names.The order of attributes that are passed into
attr.make_class()
or the these argument of@attr.s()
is now retained if the dictionary is ordered (i.e.dict
on Python 3.6 and later,collections.OrderedDict
otherwise).Before, the order was always determined by the order in which the attributes have been defined which may not be desirable when creating classes programatically.
In slotted classes,
__getstate__
and__setstate__
now ignore the__weakref__
attribute.Setting the cell type is now completely best effort. This fixes
attrs
on Jython.We cannot make any guarantees regarding Jython though, because our test suite cannot run due to dependency incompatabilities.
If
attr.s
is passed a these argument, it will no longer attempt to remove attributes with the same name from the class body.The hash of
attr.NOTHING
is now vegan and faster on 32bit Python builds.The overhead of instantiating frozen dict classes is virtually eliminated. #336
Generated
__init__
methods now have an__annotations__
attribute derived from the types of the fields.We have restructured the documentation a bit to account for
attrs
' growth in scope. Instead of putting everything into the examples page, we have started to extract narrative chapters.So far, we've added chapters on initialization and hashing.
Expect more to come!
The traversal of MROs when using multiple inheritance was backward: If you defined a class
C
that subclassesA
andB
likeC(A, B)
,attrs
would have collected the attributes fromB
before those ofA
.This is now fixed and means that in classes that employ multiple inheritance, the output of
__repr__
and the order of positional arguments in__init__
changes. Because of the nature of this bug, a proper deprecation cycle was unfortunately impossible.Generally speaking, it's advisable to prefer
kwargs
-based initialization anyways – especially if you employ multiple inheritance and diamond-shaped hierarchies.The
__repr__
set byattrs
no longer produces anAttributeError
when the instance is missing some of the specified attributes (either through deleting or after usinginit=False
on some attributes).This can break code that relied on
repr(attr_cls_instance)
raisingAttributeError
to check if anyattrs
-specified members were unset.If you were using this, you can implement a custom method for checking this:
def has_unset_members(self): for field in attr.fields(type(self)): try: getattr(self, field.name) except AttributeError: return True return False
The
attr.ib(convert=callable)
option is now deprecated in favor ofattr.ib(converter=callable)
.This is done to achieve consistency with other noun-based arguments like validator.
convert will keep working until at least January 2019 while raising a
DeprecationWarning
.
Generated
__hash__
methods now hash the class type along with the attribute values. Until now the hashes of two classes with the same values were identical which was a bug.The generated method is also much faster now.
attr.ib
’s metadata argument now defaults to a unique emptydict
instance instead of sharing a common emptydict
for all. The singleton emptydict
is still enforced.ctypes
is optional now however if it's missing, a baresuper()
will not work in slotted classes. This should only happen in special environments like Google App Engine.The attribute redefinition feature introduced in 17.3.0 now takes into account if an attribute is redefined via multiple inheritance. In that case, the definition that is closer to the base of the class hierarchy wins.
Subclasses of
auto_attribs=True
can be empty now.Equality tests are much faster now.
All generated methods now have correct
__module__
,__name__
, and (on Python 3)__qualname__
attributes.
Attributes are no longer defined on the class body.
This means that if you define a class
C
with an attributex
, the class will not have an attributex
for introspection. Instead ofC.x
, useattr.fields(C).x
or look atC.__attrs_attrs__
. The old behavior has been deprecated since version 16.1. (#253)
super()
and__class__
now work with slotted classes on Python 3. (#102, #226, #269, #270, #272)Added type argument to
attr.ib()
and correspondingtype
attribute toattr.Attribute
.This change paves the way for automatic type checking and serialization (though as of this release
attrs
does not make use of it). In Python 3.6 or higher, the value ofattr.Attribute.type
can alternately be set using variable type annotations (see PEP 526). (#151, #214, #215, #239)The combination of
str=True
andslots=True
now works on Python 2. (#198)attr.Factory
is hashable again. (#204)Subclasses now can overwrite attribute definitions of their base classes.
That means that you can -- for example -- change the default value for an attribute by redefining it. (#221, #229)
Added new option auto_attribs to
@attr.s
that allows to collect annotated fields without setting them toattr.ib()
.Setting a field to an
attr.ib()
is still possible to supply options like validators. Setting it to any other value is treated like it was passed asattr.ib(default=value)
-- passing an instance ofattr.Factory
also works as expected. (#262, #277)Instances of classes created using
attr.make_class()
can now be pickled. (#282)
- Validators are hashable again. Note that validators may become frozen in the future, pending availability of no-overhead frozen classes. #192
To encourage more participation, the project has also been moved into a dedicated GitHub organization and everyone is most welcome to join!
attrs
also has a logo now!
attrs
will set the__hash__()
method toNone
by default now. The way hashes were handled before was in conflict with Python's specification. This may break some software although this breakage is most likely just surfacing of latent bugs. You can always makeattrs
create the__hash__()
method using@attr.s(hash=True)
. See #136 for the rationale of this change.Warning
Please do not upgrade blindly and do test your software! Especially if you use instances as dict keys or put them into sets!
Correspondingly,
attr.ib
's hash argument isNone
by default too and mirrors the cmp argument as it should.
attr.assoc()
is now deprecated in favor ofattr.evolve()
and will stop working in 2018.
- Fix default hashing behavior. Now hash mirrors the value of cmp and classes are unhashable by default. #136 #142
- Added
attr.evolve()
that, given an instance of anattrs
class and field changes as keyword arguments, will instantiate a copy of the given instance with the changes applied.evolve()
replacesassoc()
, which is now deprecated.evolve()
is significantly faster thanassoc()
, and requires the class have an initializer that can take the field values as keyword arguments (likeattrs
itself can generate). #116 #124 #135 FrozenInstanceError
is now raised when trying to delete an attribute from a frozen class. #118- Frozen-ness of classes is now inherited. #128
__attrs_post_init__()
is now run if validation is disabled. #130- Added
attr.validators.in_(options)
that, given the allowed options, checks whether the attribute value is in it. This can be used to check constants, enums, mappings, etc. #181 - Added
attr.validators.and_()
that composes multiple validators into one. #161 - For convenience, the validator argument of
@attr.s
now can take a list of validators that are wrapped usingand_()
. #138 - Accordingly,
attr.validators.optional()
now can take a list of validators too. #161 - Validators can now be defined conveniently inline by using the attribute as a decorator. Check out the validator examples to see it in action! #143
attr.Factory()
now has a takes_self argument that makes the initializer to pass the partially initialized instance into the factory. In other words you can define attribute defaults based on other attributes. #165 #189- Default factories can now also be defined inline using decorators. They are always passed the partially initialized instance. #165
- Conversion can now be made optional using
attr.converters.optional()
. #105 #173 attr.make_class()
now accepts the keyword argumentbases
which allows for subclassing. #152- Metaclasses are now preserved with
slots=True
. #155
Attributes now can have user-defined metadata which greatly improves
attrs
's extensibility. #96Allow for a
__attrs_post_init__()
method that -- if defined -- will get called at the end of theattrs
-generated__init__()
method. #111Added
@attr.s(str=True)
that will optionally create a__str__()
method that is identical to__repr__()
. This is mainly useful withException
s and other classes that rely on a useful__str__()
implementation but overwrite the default one through a poor own one. Default Python class behavior is to use__repr__()
as__str__()
anyways.If you tried using
attrs
withException
s and were puzzled by the tracebacks: this option is for you.__name__
is no longer overwritten with__qualname__
forattr.s(slots=True)
classes. #99
- Added
attr.astuple()
that -- similarly toattr.asdict()
-- returns the instance as a tuple. #77 - Converters now work with frozen classes. #76
- Instantiation of
attrs
classes with converters is now significantly faster. #80 - Pickling now works with slotted classes. #81
attr.assoc()
now works with slotted classes. #84- The tuple returned by
attr.fields()
now also allows to access theAttribute
instances by name. Yes, we've subclassedtuple
so you don't have to! Thereforeattr.fields(C).x
is equivalent to the deprecatedC.x
and works with slotted classes. #88
- All instances where function arguments were called
cl
have been changed to the more Pythoniccls
. Since it was always the first argument, it's doubtful anyone ever called those function with in the keyword form. If so, sorry for any breakage but there's no practical deprecation path to solve this ugly wart.
Accessing
Attribute
instances on class objects is now deprecated and will stop working in 2017. If you need introspection please use the__attrs_attrs__
attribute or theattr.fields()
function that carry them too. In the future, the attributes that are defined on the class body and are usually overwritten in your__init__
method are simply removed after@attr.s
has been applied.This will remove the confusing error message if you write your own
__init__
and forget to initialize some attribute. Instead you will get a straightforwardAttributeError
. In other words: decorated classes will work more like plain Python classes which was alwaysattrs
's goal.The serious business aliases
attr.attributes
andattr.attr
have been deprecated in favor ofattr.attrs
andattr.attrib
which are much more consistent and frankly obvious in hindsight. They will be purged from documentation immediately but there are no plans to actually remove them.
attr.asdict()
'sdict_factory
arguments is now propagated on recursion. #45attr.asdict()
,attr.has()
andattr.fields()
are significantly faster. #48 #51- Add
attr.attrs
andattr.attrib
as a more consistent aliases forattr.s
andattr.ib
. - Add frozen option to
attr.s
that will make instances best-effort immutable. #60 attr.asdict()
now takesretain_collection_types
as an argument. IfTrue
, it does not convert attributes of typetuple
orset
tolist
. #69
Python 3.3 and 2.6 are no longer supported. They may work by chance but any effort to keep them working has ceased.
The last Python 2.6 release was on October 29, 2013 and is no longer supported by the CPython core team. Major Python packages like Django and Twisted dropped Python 2.6 a while ago already.
Python 3.3 never had a significant user base and wasn't part of any distribution's LTS release.
__slots__
have arrived! Classes now can automatically be slotted-style (and save your precious memory) just by passingslots=True
. #35- Allow the case of initializing attributes that are set to
init=False
. This allows for clean initializer parameter lists while being able to initialize attributes to default values. #32 attr.asdict()
can now produce arbitrary mappings instead of Pythondict
s when provided with adict_factory
argument. #40- Multiple performance improvements.
- Added a
convert
argument toattr.ib
, which allows specifying a function to run on arguments. This allows for simple type conversions, e.g. withattr.ib(convert=int)
. #26 - Speed up object creation when attribute validators are used. #28
- Added
attr.validators.optional()
that wraps other validators allowing attributes to beNone
. #16 - Multi-level inheritance now works. #24
__repr__()
now works with non-redecorated subclasses. #20
Initial release.