Thanks for the feedback Gregory. You raise a lot of good points; this is going to help me write a clearer pep. (0) Pretty much. They can be used as refinement for more advanced type checkers (e.g.: for linear types). (1a) I knew about the postponed evaluation but hadn't read PEP-563 yet. Thx for the heads up. (1b) I think you think you meant `Intersection` type rather than `Union` type. A value of type `Intersection[A, B]` is both of type `A` and of type `B`. If we had Intersection and allowed to passed arguments decorated with NoTypeCheck then we good do without `Annotation`. This could be a bit messy though because you'd probably want to make sure that NoTypeCheck only appears in `Intersection`. Another advantage of `Annotated` is that there's a clear "principal" type. So you can make calls to constructors transparent. e.g.: class A: .... A_with_info = Annotated[A, ...] A_with_info(5) # create the value A(5) (2a) and (2b): I don't have any strong feelings when it comes to syntax; I tried to be consistent with the standard library (and maybe I got it wrong). My understanding [] is used to create a new type whereas () is used to create a new value:

Deque(range(2)) deque([0, 1])

Deque[int] typing.Deque[int]

On Thu, 17 Jan 2019 at 19:35 Gregory P. Smith wrote:

On Thu, Jan 17, 2019 at 2:34 PM Till wrote:

...

We started a discussion in https://github.com/python/typing/issues/600 about adding support for extra annotations in the typing module.

Since this is probably going to turn into a PEP I'm transferring the discussion here to have more visibility.

The document below has been modified a bit from the one in GH to reflect the feedback I got:

+ Added a small blurb about how ``Annotated`` should support being used as an alias

Things that were raised but are not reflected in this document:

+ The dataclass example is confusing. I kept it for now because dataclasses often come up in conversations about why we might want to support annotations in the typing module. Maybe I should rework the section.

+ `...` as a valid parameter for the first argument (if you want to add an annotation but use the type inferred by your type checker). This is an interesting idea, it's probably worth adding support for it if and only if we decide to support in other places. (c.f.: https://github.com/python/typing/issues/276)

Thanks,

Add support for external annotations in the typing module ==========================================================

We propose adding an ``Annotated`` type to the typing module to decorate existing types with context-specific metadata. Specifically, a type ``T`` can be annotated with metadata ``x`` via the typehint ``Annotated[T, x]``. This metadata can be used for either static analysis or at runtime. If a library (or tool) encounters a typehint ``Annotated[T, x]`` and has no special logic for metadata ``x``, it should ignore it and simply treat the type as ``T``. Unlike the `no_type_check` functionality that current exists in the ``typing`` module which completely disables typechecking annotations on a function or a class, the ``Annotated`` type allows for both static typechecking of ``T`` (e.g., via MyPy or Pyre, which can safely ignore ``x``) together with runtime access to ``x`` within a specific application. We believe that the introduction of this type would address a diverse set of use cases of interest to the broader Python community.

Motivating examples: ~~~~~~~~~~~~~~~~~~~~

reading binary data +++++++++++++++++++

The ``struct`` module provides a way to read and write C structs directly from their byte representation. It currently relies on a string representation of the C type to read in values::

record = b'raymond \x32\x12\x08\x01\x08' name, serialnum, school, gradelevel = unpack('<10sHHb', record)

The struct documentation [struct-examples]_ suggests using a named tuple to unpack the values and make this a bit more tractable::

from collections import namedtuple Student = namedtuple('Student', 'name serialnum school gradelevel') Student._make(unpack('<10sHHb', record)) # Student(name=b'raymond ', serialnum=4658, school=264, gradelevel=8)

However, this recommendation is somewhat problematic; as we add more fields, it's going to get increasingly tedious to match the properties in the named tuple with the arguments in ``unpack``.

Instead, annotations can provide better interoperability with a type checker or an IDE without adding any special logic outside of the ``struct`` module::

from typing import NamedTuple UnsignedShort = Annotated[int, struct.ctype('H')] SignedChar = Annotated[int, struct.ctype('b')]

@struct.packed class Student(NamedTuple): # MyPy typechecks 'name' field as 'str' name: Annotated[str, struct.ctype("<10s")] serialnum: UnsignedShort school: SignedChar gradelevel: SignedChar

# 'unpack' only uses the metadata within the type annotations Student.unpack(record)) # Student(name=b'raymond ', serialnum=4658, school=264, gradelevel=8)

dataclasses ++++++++++++

Here's an example with dataclasses [dataclass]_ that is a problematic from the typechecking standpoint::

from dataclasses import dataclass, field

@dataclass class C: myint: int = 0 # the field tells the @dataclass decorator that the default action in the # constructor of this class is to set "self.mylist = list()" mylist: List[int] = field(default_factory=list)

Even though one might expect that ``mylist`` is a class attribute accessible via ``C.mylist`` (like ``C.myint`` is) due to the assignment syntax, that is not the case. Instead, the ``@dataclass`` decorator strips out the assignment to this attribute, leading to an ``AttributeError`` upon access::

C.myint # Ok: 0 C.mylist # AttributeError: type object 'C' has no attribute 'mylist'

This can lead to confusion for newcomers to the library who may not expect this behavior. Furthermore, the typechecker needs to understand the semantics of dataclasses and know to not treat the above example as an assignment operation in (which translates to additional complexity).

It makes more sense to move the information contained in ``field`` to an annotation::

@dataclass class C: myint: int = 0 mylist: Annotated[List[int], field(default_factory=list)]

# now, the AttributeError is more intuitive because there is no assignment operator C.mylist # AttributeError

# the constructor knows how to use the annotations to set the 'mylist' attribute c = C() c.mylist # []

The main benefit of writing annotations like this is that it provides a way for clients to gracefully degrade when they don't know what to do with the extra annotations (by just ignoring them). If you used a typechecker that didn't have any special handling for dataclasses and the ``field`` annotation, you would still be able to run checks as though the type were simply::

class C: myint: int = 0 mylist: List[int]

lowering barriers to developing new types +++++++++++++++++++++++++++++++++++++++++

Typically when adding a new type, we need to upstream that type to the typing module and change MyPy [MyPy]_, PyCharm [PyCharm]_, Pyre [Pyre]_, pytype [pytype]_, etc. This is particularly important when working on open-source code that makes use of our new types, seeing as the code would not be immediately transportable to other developers' tools without additional logic (this is a limitation of MyPy plugins [MyPy-plugins]_), which allow for extending MyPy but would require a consumer of new typehints to be using MyPy and have the same plugin installed). As a result, there is a high cost to developing and trying out new types in a codebase. Ideally, we should be able to introduce new types in a manner that allows for graceful degradation when clients do not have a custom MyPy plugin, which would lower the barrier to development and ensure some degree of backward compatibility.

For example, suppose that we wanted to add support for tagged unions [tagged-unions]_ to Python. One way to accomplish would be to annotate ``TypedDict`` in Python such that only one field is allowed to be set::

Currency = Annotated( TypedDict('Currency', {'dollars': float, 'pounds': float}, total=False), TaggedUnion, )

This is a somewhat cumbersome syntax but it allows us to iterate on this proof-of-concept and have people with non-patched IDEs work in a codebase with tagged unions. We could easily test this proposal and iron out the kinks before trying to upstream tagged union to `typing`, MyPy, etc. Moreover, tools that do not have support for parsing the ``TaggedUnion`` annotation would still be able able to treat `Currency` as a ``TypedDict``, which is still a close approximation (slightly less strict).

Details of proposed changes to ``typing`` ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

syntax ++++++

``Annotated`` is parameterized with a type and an arbitrary list of Python values that represent the annotations. Here are the specific details of the syntax:

* The first argument to ``Annotated`` must be a valid ``typing`` type or ``...`` (to use the infered type).

* Multiple type annotations are supported (Annotated supports variadic arguments): ``Annotated[int, ValueRange(3, 10), ctype("char")]``

* ``Annotated`` must be called with at least two arguments (``Annotated[int]`` is not valid)

* The order of the annotations is preserved and matters for equality checks::

Annotated[int, ValueRange(3, 10), ctype("char")] != \ Annotated[int, ctype("char"), ValueRange(3, 10)]

* Nested ``Annotated`` types are flattened, with metadata ordered starting with the innermost annotation::

Annotated[Annotated[int, ValueRange(3, 10)], ctype("char")] ==\ Annotated[int, ValueRange(3, 10), ctype("char")]

* Duplicated annotations are not removed: ``Annotated[int, ValueRange(3, 10)] != Annotated[int, ValueRange(3, 10), ValueRange(3, 10)]``

* ``Annotation`` can be used a higher order aliases::

Typevar T = ... Vec = Annotated[List[Tuple[T, T]], MaxLen(10)] # Vec[int] == `Annotated[List[Tuple[int, int]], MaxLen(10)]

consuming annotations ++++++++++++++++++++++

Ultimately, the responsibility of how to interpret the annotations (if at all) is the responsibility of the tool or library encountering the `Annotated` type. A tool or library encountering an `Annotated` type can scan through the annotations to determine if they are of interest (e.g., using `isinstance`).

**Unknown annotations** When a tool or a library does not support annotations or encounters an unknown annotation it should just ignore it and treat annotated type as the underlying type. For example, if we were to add an annotation that is not an instance of `struct.ctype` to the annotation for name (e.g., `Annotated[str, 'foo', struct.ctype("<10s")]`), the unpack method should ignore it.

**Namespacing annotations** We do not need namespaces for annotations since the class used by the annotations acts as a namespace.

**Multiple annotations** It's up to the tool consuming the annotations to decide whether the client is allowed to have several annotations on one type and how to merge those annotations.

Since the ``Annotated`` type allows you to put several annotations of the same (or different) type(s) on any node, the tools or libraries consuming those annotations are in charge of dealing with potential duplicates. For example, if you are doing value range analysis you might allow this::

T1 = Annotated[int, ValueRange(-10, 5)] T2 = Annotated[T1, ValueRange(-20, 3)]

Flattening nested annotations, this translates to::

T2 = Annotated[int, ValueRange(-10, 5), ValueRange(-20, 3)]

An application consuming this type might choose to reduce these annotations via an intersection of the ranges, in which case ``T2`` would be treated equivalently to ``Annotated[int, ValueRange(-10, 3)]``.

An alternative application might reduce these via a union, in which case ``T2`` would be treated equivalently to ``Annotated[int, ValueRange(-20, 5)]``.

Other applications may decide to not support multiple annotations and throw an exception.

(0) Observaton / TL;DR - This PEP really seems to be more of a way to declare multiple different arbitrary purposes annotations all attached to a single callable/parameter/return/variable. So that static checkers continue to work, but runtime user of annotations for whatever purpose can also work at the same time.

(1a) A struct.unpack supporting this will then need to evaluate annotations in the outer scope at runtime due to our desired long term PEP-563 `from __future__ import annotations` behavior. But that becomes true of anything else wanting to use annotations at runtime so we should really make a typing library function that does this for everyone to use.

(1b) This proposal potentially expands the burden of type checkers... but it shouldn't. They should be free to take the first type listed in an Annotated[] block as the type of the variable, raising an error if someone has listed multiple types (telling them to use Union[] for that). a static checker *could* do useful things with multiple annotations it knows how to handle, but I think it'd be unwise to implement that in any manner where Annotated and Union could both be used for the same purpose.

It makes me wonder if Annotated[] is meaningfully different from Union at all.

(2a) At first glance I don't like that the `T1 = Annotated[int, SomeOtherInfo(23)]` syntax uses [] rather than () as it really is constructing a runtime type. It isn't clear what should use [] and what should use () so I'd suggest using () for everything there.

(2b) Ask yourself: Why should SomeOtherInfo and ValueRange and struct.ctype be () calls yet none of `Annotated[Union[List[bytes], Dict[bytes, Optional[float]]]]` be calls? If you can come up with an answer to that, why _should_ anyone need to know that?

-gps