PEP 570: grammar and wording fixes (GH-923)
This commit is contained in:
parent
bbdcedb71b
commit
c0927428bf
166
pep-0570.rst
166
pep-0570.rst
|
@ -3,8 +3,8 @@ Title: Python Positional-Only Parameters
|
|||
Version: $Revision$
|
||||
Last-Modified: $Date$
|
||||
Author: Larry Hastings <larry@hastings.org>,
|
||||
Pablo Galindo <pablogsal@gmail.com>,
|
||||
Mario Corchero <mariocj89@gmail.com>
|
||||
Pablo Galindo <pablogsal@gmail.com>,
|
||||
Mario Corchero <mariocj89@gmail.com>
|
||||
Discussions-To: Python-Dev <python-dev@python.org>
|
||||
Status: Draft
|
||||
Type: Standards Track
|
||||
|
@ -33,49 +33,49 @@ positional-only. This changed around Python 1.0, when
|
|||
all parameters suddenly became positional-or-keyword.
|
||||
This allowed users to provide arguments to a function both
|
||||
positionally or referencing the keyword used in the definition
|
||||
of it. But, this is not always desired nor even available as
|
||||
of it. But, this is not always desired nor available as
|
||||
even in current versions of Python, many CPython
|
||||
"builtin" functions still only accept positional-only arguments.
|
||||
|
||||
Users might want to restrict their API to not allow for parameters
|
||||
to be referenced via keyword, as that exposes the name of the
|
||||
to be referenced via keywords, as that exposes the name of the
|
||||
parameter as part of the API. If a user of said API starts using the
|
||||
argument by keyword when calling it and then the name of the parameter
|
||||
is changed, it will be a breaking change. By using positional-only
|
||||
parameters the developer can later change the name of an argument or
|
||||
argument by keyword when calling it and then the parameter
|
||||
gets renamed, it will be a breaking change. By using positional-only
|
||||
parameters the developer can later change the name of an arguments or
|
||||
transform them to ``*args`` without breaking the API.
|
||||
|
||||
Even if positional arguments only in a function can be achieved
|
||||
via using ``*args`` parameters and extracting them one by one,
|
||||
Even if making arguments positional-only in a function can be achieved
|
||||
by using ``*args`` parameters and extracting them one by one,
|
||||
the solution is far from ideal and not as expressive as the one
|
||||
proposed in this PEP, which targets to provide syntax to specify
|
||||
proposed in this PEP, which targets providing syntax to specify
|
||||
accepting a specific number of positional-only parameters. Also,
|
||||
it makes the signature of the function ambiguous as users won't
|
||||
know how many parameters the function takes by looking at help()
|
||||
know how many parameters the function takes by looking at ``help()``
|
||||
or auto-generated documentation.
|
||||
|
||||
Additionally, this will bridge the gap we currently find between
|
||||
builtin functions that today allows to specify positional-only
|
||||
builtin functions that can specify positional-only
|
||||
parameters and pure Python implementations that lack the
|
||||
syntax for it. The '/' syntax is already exposed in the
|
||||
documentation for some builtins and interfaces generated by
|
||||
the argument clinic. Making positional only arguments a possibility
|
||||
in Python will bring consistency and will remove confusion from
|
||||
users that are not familiarized with the fact that positional only
|
||||
arguments are allowed in builtins and argument clinic C interfaces.
|
||||
documentation of some builtins and interfaces generated by
|
||||
the argument clinic. Making positional-only arguments a possibility
|
||||
in Python will make the language more consistent and make it clearer
|
||||
to users that positional-only arguments are allowed in builtins and argument
|
||||
clinic C interfaces.
|
||||
|
||||
We find this useful in multiple situation, say for example we
|
||||
are looking at creating a function that converts from one type to
|
||||
We can find positional-only arguments useful in several situations. Example:
|
||||
we want to create a function that converts from one type to
|
||||
another::
|
||||
|
||||
def as_my_type(x):
|
||||
...
|
||||
|
||||
The name of the parameter provides no value whatsoever and forces
|
||||
the user to maintain it's name forever as user might rely on it
|
||||
being used as a keyword only.
|
||||
The name of the parameter provides no value whatsoever, and forces
|
||||
the developer to maintain its name forever, as users might pass ``x`` as a
|
||||
keyword.
|
||||
|
||||
Another good example is APIs that want to transmit the feeling
|
||||
Another good example is an API that wants to transmit the feeling
|
||||
of ownership through positional arguments, see::
|
||||
|
||||
class MyDecorator:
|
||||
|
@ -83,20 +83,20 @@ of ownership through positional arguments, see::
|
|||
...
|
||||
|
||||
Again we get no value from using keyword arguments here and it can limit
|
||||
future evolutions of the API. Say at a later time we want the decorator
|
||||
to be able to take multiple functions, we will be forced to keep
|
||||
the original argument always or we'd potentially break users.
|
||||
By being able to define positional only arguments we can change the
|
||||
name of those at will or even change them by ``*args``.
|
||||
future evolution of the API. Say at a later time we want the decorator
|
||||
to be able to take multiple functions, we will be forced to always keep
|
||||
the original argument or we would potentially break users.
|
||||
By being able to define positional-only arguments we can change the
|
||||
name of those at will or even change them to ``*args``.
|
||||
|
||||
-----------------------------------------------------
|
||||
Positional-Only Parameter Semantics In Current Python
|
||||
-----------------------------------------------------
|
||||
---------------------------------------------------
|
||||
Positional-Only Parameter Semantics In Python Today
|
||||
---------------------------------------------------
|
||||
|
||||
There are many, many examples of builtins that only
|
||||
accept positional-only parameters. The resulting
|
||||
semantics are easily experienced by the Python
|
||||
programmer--just try calling one, specifying its
|
||||
programmer -- just try calling one, specifying its
|
||||
arguments by name::
|
||||
|
||||
|
||||
|
@ -109,8 +109,8 @@ arguments by name::
|
|||
File "<stdin>", line 1, in <module>
|
||||
TypeError: pow() takes no keyword arguments
|
||||
|
||||
Pow clearly expresses that its arguments are only positional
|
||||
via the ``/`` marker, but this is at the moment only documentational,
|
||||
``pow()`` clearly expresses that its arguments are only positional
|
||||
via the ``/`` marker, but this at the moment is only documentational,
|
||||
Python developers cannot write such syntax.
|
||||
|
||||
In addition, there are some functions with particularly
|
||||
|
@ -128,9 +128,9 @@ interesting semantics:
|
|||
Obviously one can simulate any of these in pure Python code
|
||||
by accepting ``(*args, **kwargs)`` and parsing the arguments
|
||||
by hand. But this results in a disconnect between the
|
||||
Python function signature and what it actually accepts,
|
||||
Python function signature and what the function actually accepts,
|
||||
not to mention the work of implementing said argument parsing
|
||||
and the lack of clarity that generates in the signature.
|
||||
and the lack of clarity in the resulting signature.
|
||||
|
||||
==========
|
||||
Motivation
|
||||
|
@ -138,28 +138,28 @@ Motivation
|
|||
|
||||
The new syntax will allow developers to further control how their
|
||||
API can be consumed. It will allow restricting the usage of keyword
|
||||
Specify arguments by adding the new type of positional-only ones.
|
||||
arguments by adding the new type of positional-only ones.
|
||||
|
||||
A similar PEP with a broader scope (PEP 457) was proposed
|
||||
to define the syntax. This PEP builds on top of part of it
|
||||
to define and provide an implementation for the ``/`` syntax on
|
||||
A similar PEP with a broader scope (PEP 457) was proposed earlier
|
||||
to define the syntax. This PEP builds partially on top of that,
|
||||
to define and provide an implementation for the ``/`` syntax in
|
||||
function signatures.
|
||||
|
||||
Providing positional only arguments will allow for maintaining the
|
||||
Providing positional-only arguments will allow for maintaining the
|
||||
interface when creating pure Python implementation of C modules, which
|
||||
provides not only the API benefits outlined in this document but it is
|
||||
also faster, see this thread about converting keyword arguments to positional
|
||||
also faster. See this thread about converting keyword arguments to positional:
|
||||
[#thread-keyword-to-positional]_ and PEP-399 [#PEP399]_, which requires the
|
||||
same API for C accelerators than the Python implementation.
|
||||
same API for C accelerators as the Python implementation.
|
||||
|
||||
There has been multiple changes in builtin functions that moved away
|
||||
There have been multiple changes in builtin functions that moved away
|
||||
from keyword arguments, like ``bool``, ``float``, ``list``, ``int``, ``tuple``
|
||||
which is a non-backward compatible. By having proper support for
|
||||
positional only arguments, this kind of APIs where it is clear that
|
||||
passing a keyword argument provides no clarity would be able to
|
||||
follow a similar approach that those builtins without breaking users.
|
||||
which is a non-backward compatible change. By having proper support for
|
||||
positional-only arguments, these kind of APIs, where it is clear that
|
||||
passing a keyword argument provides no clarity, it would be possible to
|
||||
follow a similar approach as these builtins, without breaking users.
|
||||
|
||||
This is a well discussed recurring topic in the Python mailing lists:
|
||||
This is a well discussed, recurring topic on the Python mailing lists:
|
||||
|
||||
* September 2018: `Anders Hovmöller: [Python-ideas] Positional-only
|
||||
parameters
|
||||
|
@ -193,10 +193,10 @@ and inconsistent:
|
|||
* Some functions use *both* of the above approaches. [#RANGE]_ [#ADDCH]_
|
||||
|
||||
One more important idea to consider: currently in the documentation
|
||||
there's no way to tell whether a function takes positional-only
|
||||
there is no way to tell whether a function takes positional-only
|
||||
parameters. ``open()`` accepts keyword arguments, ``ord()`` does
|
||||
not, but there is no way of telling just by reading the
|
||||
documentation that this is true.
|
||||
documentation.
|
||||
|
||||
====================
|
||||
Syntax And Semantics
|
||||
|
@ -217,10 +217,10 @@ All parameters before the ``/`` are positional-only. If ``/`` is
|
|||
not specified in a function signature, that function does not
|
||||
accept any positional-only parameters.
|
||||
The logic around optional values for positional-only argument
|
||||
Remains the same as the one for positional-or-keyword. Once
|
||||
remains the same as the one for positional-or-keyword. Once
|
||||
a positional-only argument is provided with a default,
|
||||
the following positional-only and positional-or-keyword argument
|
||||
need to have a default as well. Positional-only parameters that
|
||||
needs to have a default as well. Positional-only parameters that
|
||||
don’t have a default value are "required" positional-only parameters.
|
||||
Therefore the following are valid signatures::
|
||||
|
||||
|
@ -249,11 +249,13 @@ A draft of the proposed grammar specification is::
|
|||
new_varargslist:
|
||||
vfpdef ['=' test] (',' vfpdef ['=' test])* ',' '/' [',' [varargslist]] | varargslist
|
||||
|
||||
It will be added to the actual typedargslist and varargslist but for easier discussion is
|
||||
presented as new_typedargslist and new_varargslist. Also, notice that using a construction
|
||||
using two new rules (new_varargslist and new_varargslist) is not possible with the current
|
||||
parser as the rule is not LL(1). This is the reason the rule needs to be include in the
|
||||
existing typedargslist and varargslist (in the same way keyword-only arguments were introduced).
|
||||
It will be added to the actual typedargslist and varargslist but for easier
|
||||
discussion it is presented as new_typedargslist and new_varargslist. Also,
|
||||
notice that using a construction with two new rules (new_varargslist and
|
||||
new_varargslist) is not possible with the current parser as the rule is not
|
||||
LL(1). This is the reason the rule needs to be included in the existing
|
||||
typedargslist and varargslist (in the same way keyword-only arguments were
|
||||
introduced).
|
||||
|
||||
|
||||
==============
|
||||
|
@ -285,21 +287,21 @@ After marker proposal
|
|||
---------------------
|
||||
|
||||
A complaint against the proposal is the fact that the modifier of
|
||||
the signature impacts the "already passed" tokens.
|
||||
the signature impacts the tokens already passed.
|
||||
|
||||
This might make confusing to "human parsers" to read functions
|
||||
This might make it confusing to users to read functions
|
||||
with many arguments. Example::
|
||||
|
||||
def really_bad_example_of_a_python_function(fist_long_argument, second_long_argument,
|
||||
third_long_argument, /):
|
||||
|
||||
It is not until you reach the end of the signature that the reader
|
||||
realized the ``/`` and therefore the fact that the arguments are
|
||||
It is not until reaching the end of the signature that the reader
|
||||
realizes the ``/``, and therefore the fact that the arguments are
|
||||
position-only. This deviates from how the keyword-only marker works.
|
||||
|
||||
That said we could not find an implementation that would modify the
|
||||
arguments after the marker, as that will force the one before the
|
||||
marker to be position only as well. Example::
|
||||
marker to be position-only as well. Example::
|
||||
|
||||
def (x, y, /, z):
|
||||
|
||||
|
@ -307,48 +309,50 @@ If we define that ``/`` makes only z position-only it won't be possible
|
|||
to call x and y via keyword argument. Finding a way to work around it
|
||||
will add confusion given that at the moment keyword arguments cannot be
|
||||
followed by positional arguments. ``/`` will therefore make both the
|
||||
preceding and following position-only.
|
||||
preceding and following parameters position-only.
|
||||
|
||||
-------------------
|
||||
Per-argument marker
|
||||
-------------------
|
||||
|
||||
Using a per argument marker might be an option as well. The approach
|
||||
Using a per-argument marker might be an option as well. The approach
|
||||
basically adds a token to each of the arguments that are position only
|
||||
and requires those to be placed together. Example::
|
||||
|
||||
def (.arg1, .arg2, arg3):
|
||||
|
||||
Note the dot on arg1 and arg2. Even if this approach might look easier
|
||||
to read it has been discarded as ``/`` goes further inline with the
|
||||
keyword-only approach and is less error prone.
|
||||
to read, it has been discarded as ``/`` goes further in line with the
|
||||
keyword-only approach and is less error-prone.
|
||||
|
||||
There are some libraries that use leading underscore[#leading-underscore]_
|
||||
to mark those arguments as positional only.
|
||||
to mark those arguments as positional-only.
|
||||
|
||||
----------------
|
||||
Using decorators
|
||||
----------------
|
||||
|
||||
It has been suggested on python-ideas [#python-ideas-decorator-based]_ to provide
|
||||
a decorator written in Python as an implementation for this feature. This approach
|
||||
has the advantage that keeps parameter declaration more easy to read but also
|
||||
introduces an asymmetry on how parameter behaviour is declared. Also, as the ``/``
|
||||
syntax is already introduced for C functions, this inconsistency will make more
|
||||
difficult to implement all tools and modules that deal with this syntax including
|
||||
but not limited to, the argument clinic, the inspect module and the ast module.
|
||||
Another disadvantage of this approach is that calling the decorated functions
|
||||
will be slower than the functions generated if the feature was implemented directly
|
||||
in C.
|
||||
It has been suggested on python-ideas [#python-ideas-decorator-based]_ to
|
||||
provide a decorator written in Python as an implementation for this feature.
|
||||
This approach has the advantage that keeps parameter declaration more easy to
|
||||
read but also introduces an asymmetry on how parameter behaviour is declared.
|
||||
Also, as the ``/`` syntax is already introduced for C functions, this
|
||||
inconsistency will make it more difficult to implement all tools and modules
|
||||
that deal with this syntax including but not limited to, the argument clinic,
|
||||
the inspect module and the ast module. Another disadvantage of this approach
|
||||
is that calling the decorated functions will be slower than the functions
|
||||
generated if the feature was implemented directly in C.
|
||||
|
||||
======
|
||||
Thanks
|
||||
======
|
||||
|
||||
Credit for most of the content of this PEP is contained in Larry Hastings’s PEP 457.
|
||||
Credit for most of the content of this PEP is contained in Larry Hastings’s
|
||||
PEP 457.
|
||||
|
||||
Credit for the use of '/' as the separator between positional-only and positional-or-keyword
|
||||
parameters go to Guido van Rossum, in a proposal from 2012. [#GUIDO]_
|
||||
Credit for the use of '/' as the separator between positional-only and
|
||||
positional-or-keyword parameters go to Guido van Rossum, in a proposal from
|
||||
2012. [#GUIDO]_
|
||||
|
||||
Credit for discussion about the simplification of the grammar goes to
|
||||
Braulio Valdivieso.
|
||||
|
|
Loading…
Reference in New Issue