python-peps/peps/pep-0680.rst

PEP: 680
Title: tomllib: Support for Parsing TOML in the Standard Library
Author: Taneli Hukkinen, Shantanu Jain <hauntsaninja at gmail.com>
Sponsor: Petr Viktorin <encukou@gmail.com>
Discussions-To: https://discuss.python.org/t/13040
Status: Final
Type: Standards Track
Content-Type: text/x-rst
Created: 01-Jan-2022
Python-Version: 3.11
Post-History: `09-Dec-2021 <https://mail.python.org/archives/list/python-ideas@python.org/thread/IWJ3I32A4TY6CIVQ6ONPEBPWP4TOV2V7/>`__,
              `27-Jan-2022 <https://discuss.python.org/t/pep-680-tomllib-support-for-parsing-toml-in-the-standard-library/13040>`__,
Resolution: https://mail.python.org/archives/list/python-dev@python.org/thread/3AHGWYY562HHO55L4Z2OVYUFZP5W73IS/

.. canonical-doc:: :mod:`python:tomllib`

Abstract
========

This PEP proposes adding the ``tomllib`` module to the standard library for
parsing TOML (Tom's Obvious Minimal Language,
`https://toml.io <https://toml.io/en/>`_).


Motivation
==========

TOML is the format of choice for Python packaging, as evidenced by
:pep:`517`, :pep:`518` and :pep:`621`. This creates a bootstrapping
problem for Python build tools, forcing them to vendor a TOML parsing
package or employ other undesirable workarounds, and causes serious issues
for repackagers and other downstream consumers. Including TOML support in
the standard library would neatly solve all of these issues.

Further, many Python tools are now configurable via TOML, such as
``black``, ``mypy``, ``pytest``, ``tox``, ``pylint`` and ``isort``.
Many that are not, such as ``flake8``, cite the lack of standard library
support as a `main reason why
<https://github.com/PyCQA/flake8/issues/234#issuecomment-812800722>`__.
Given the special place TOML already has in the Python ecosystem, it makes sense
for it to be an included battery.

Finally, TOML as a format is increasingly popular (for the reasons
outlined in :pep:`518`), with various Python TOML libraries having about
2000 reverse dependencies on PyPI (for comparison, ``requests`` has about
28000 reverse dependencies). Hence, this is likely to be a generally useful
addition, even looking beyond the needs of Python packaging and related tools.


Rationale
=========

This PEP proposes basing the standard library support for reading TOML on the
third-party library ``tomli``
(`github.com/hukkin/tomli <https://github.com/hukkin/tomli>`_).

Many projects have recently switched to using ``tomli``, such as ``pip``,
``build``, ``pytest``, ``mypy``, ``black``, ``flit``, ``coverage``,
``setuptools-scm`` and ``cibuildwheel``.

``tomli`` is actively maintained and well-tested. It is about 800 lines
of code with 100% test coverage, and passes all tests in the
`proposed official TOML compliance test suite
<https://github.com/toml-lang/compliance/pull/8>`_, as well as
`the more established BurntSushi/toml-test suite
<https://github.com/BurntSushi/toml-test>`_.


Specification
=============

A new module ``tomllib`` will be added to the Python standard library,
exposing the following public functions:

.. code-block::

   def load(
       fp: SupportsRead[bytes],
       /,
       *,
       parse_float: Callable[[str], Any] = ...,
    ) -> dict[str, Any]: ...

   def loads(
       s: str,
       /,
       *,
       parse_float: Callable[[str], Any] = ...,
   ) -> dict[str, Any]: ...

``tomllib.load`` deserializes a binary file-like object containing a
TOML document to a Python ``dict``.
The ``fp`` argument must have a ``read()`` method with the same API as
``io.RawIOBase.read()``.

``tomllib.loads`` deserializes a ``str`` instance containing a TOML document
to a Python ``dict``.

The ``parse_float`` argument is a callable object that takes as input the
original string representation of a TOML float, and returns a corresponding
Python object (similar to ``parse_float`` in ``json.load``).
For example, the user may pass a function returning a ``decimal.Decimal``,
for use cases where exact precision is important. By default, TOML floats
are parsed as instances of the Python ``float`` type.

The returned object contains only basic Python objects (``str``, ``int``,
``bool``, ``float``, ``datetime.{datetime,date,time}``, ``list``, ``dict`` with
string keys), and the results of ``parse_float``.

``tomllib.TOMLDecodeError`` is raised in the case of invalid TOML.

Note that this PEP does not propose ``tomllib.dump`` or ``tomllib.dumps``
functions; see `Including an API for writing TOML`_ for details.


Maintenance Implications
========================

Stability of TOML
-----------------

The release of TOML 1.0.0 in January 2021 indicates the TOML format should
now be officially considered stable. Empirically, TOML has proven to be a
stable format even prior to the release of TOML 1.0.0. From the
`changelog <https://github.com/toml-lang/toml/blob/master/CHANGELOG.md>`__, we
can see that TOML has had no major changes since April 2020, and has had
two releases in the past five years (2017-2021).

In the event of changes to the TOML specification, we can treat minor
revisions as bug fixes and update the implementation in place. In the event of
major breaking changes, we should preserve support for TOML 1.x.


Maintainability of proposed implementation
------------------------------------------

The proposed implementation (``tomli``) is pure Python, well tested and
weighs in at under 1000 lines of code. It is minimalist, offering a smaller API
surface area than other TOML implementations.

The author of ``tomli`` is willing to help integrate ``tomli`` into the standard
library and help maintain it, `as per this post
<https://github.com/hukkin/tomli/issues/141#issuecomment-998018972>`__.
Furthermore, Python core developer Petr Viktorin has indicated a willingness
to maintain a read API, `as per this post
<https://discuss.python.org/t/adopting-recommending-a-toml-parser/4068/88>`__.

Rewriting the parser in C is not deemed necessary at this time. It is rare for
TOML parsing to be a bottleneck in applications, and users with higher performance
needs can use a third-party library (as is already often the case with JSON,
despite Python offering a standard library C-extension module).


TOML support a slippery slope for other things
----------------------------------------------

As discussed in the `Motivation`_ section, TOML holds a special place in the
Python ecosystem, for reading :pep:`518` ``pyproject.toml`` packaging
and tool configuration files.
This chief reason to include TOML in the standard library does not apply to
other formats, such as YAML or MessagePack.

In addition, the simplicity of TOML distinguishes it from other formats like
YAML, which are highly complicated to construct and parse.

An API for writing TOML may, however, be added in a future PEP.


Backwards Compatibility
=======================

This proposal has no backwards compatibility issues within the standard
library, as it describes a new module.
Any existing third-party module named ``tomllib`` will break, as
``import tomllib`` will import the standard library module.
However, ``tomllib`` is not registered on PyPI, so it is unlikely that any
module with this name is widely used.

Note that we avoid using the more straightforward name ``toml`` to avoid
backwards compatibility implications for users who have pinned versions of the
current ``toml`` PyPI package.
For more details, see the `Alternative names for the module`_ section.


Security Implications
=====================

Errors in the implementation could cause potential security issues.
However, the parser's output is limited to simple data types; inability to load
arbitrary classes avoids security issues common in more "powerful" formats like
pickle and YAML. Also, the implementation will be in pure Python, which reduces
security issues endemic to C, such as buffer overflows.


How to Teach This
=================

The API of ``tomllib`` mimics that of other well-established file format
libraries, such as ``json`` and ``pickle``. The lack of a ``dump`` function will
be explained in the documentation, with a link to relevant third-party libraries
(e.g. ``tomlkit``, ``tomli-w``, ``pytomlpp``).


Reference Implementation
========================

The proposed implementation can be found at https://github.com/hukkin/tomli


Rejected Ideas
==============

Basing on another TOML implementation
-------------------------------------

Several potential alternative implementations exist:

* ``tomlkit`` is well established, actively maintained and supports TOML 1.0.0.
  An important difference is that ``tomlkit`` supports style roundtripping. As a
  result, it has a more complex API and implementation (about 5x as much code as
  ``tomli``). Its author does not believe that ``tomlkit`` is a good choice for
  the standard library.

* ``toml`` is a very widely used library. However, it is not actively
  maintained, does not support TOML 1.0.0 and has a number of known bugs. Its
  API is more complex than that of ``tomli``. It allows customising output style
  through a complicated encoder API, and some very limited and mostly unused
  functionality to preserve input style through an undocumented decoder API.
  For more details on its API differences from this PEP, refer to `Appendix A
  <PEP 680 Appendix A_>`_.

* ``pytomlpp`` is a Python wrapper for the C++ project ``toml++``. Pure Python
  libraries are easier to maintain than extension modules.

* ``rtoml`` is a Python wrapper for the Rust project ``toml-rs`` and hence has
  similar shortcomings to ``pytomlpp``.
  In addition, it does not support TOML 1.0.0.

* Writing an implementation from scratch. It's unclear what we would get from
  this; ``tomli`` meets our needs and the author is willing to help with its
  inclusion in the standard library.


Including an API for writing TOML
---------------------------------

There are several reasons to not include an API for writing TOML.

The ability to write TOML is not needed for the use cases that motivate this
PEP: core Python packaging tools, and projects that need to read TOML
configuration files.

Use cases that involve editing an existing TOML file (as opposed to writing a
brand new one) are better served by a style preserving library. TOML is
intended as a human-readable and -editable configuration format, so it's
important to preserve comments, formatting and other markup. This requires
a parser whose output includes style-related metadata, making it impractical
to output plain Python types like ``str`` and ``dict``. Furthermore, it
substantially complicates the design of the API.

Even without considering style preservation, there are too many degrees of
freedom in how to design a write API. For example, what default style
(indentation, vertical and horizontal spacing, quotes, etc) should the library
use for the output, and how much control should users be given over it?
How should the library handle input and output validation? Should it support
serialization of custom types, and if so, how? While there are reasonable
options for resolving these issues, the nature of the standard library is such
that we only get "one chance to get it right".

Currently, no CPython core developers have expressed willingness to maintain a
write API, or sponsor a PEP that includes one. Since it is hard to change
or remove something in the standard library, it is safer to err on the side of
exclusion for now, and potentially revisit this later.

Therefore, writing TOML is left to third-party libraries. If a good API and
relevant use cases for it are found later, write support can be added in a
future PEP.


Assorted API details
--------------------

Types accepted as the first argument of ``tomllib.load``
''''''''''''''''''''''''''''''''''''''''''''''''''''''''

The ``toml`` library on PyPI allows passing paths (and lists of path-like
objects, ignoring missing files and merging the documents into a single object)
to its ``load`` function. However, allowing this here would be inconsistent
with the behavior of ``json.load``, ``pickle.load`` and other standard library
functions. If we agree that consistency here is desirable,
allowing paths is out of scope for this PEP. This can easily and explicitly
be worked around in user code, or by using a third-party library.

The proposed API takes a binary file, while ``toml.load`` takes a text file and
``json.load`` takes either. Using a binary file allows us to ensure UTF-8 is
the encoding used (ensuring correct parsing on platforms with other default
encodings, such as Windows), and avoid incorrectly parsing files containing
single carriage returns as valid TOML due to universal newlines in text mode.


Type accepted as the first argument of ``tomllib.loads``
''''''''''''''''''''''''''''''''''''''''''''''''''''''''

While ``tomllib.load`` takes a binary file, ``tomllib.loads`` takes
a text string. This may seem inconsistent at first.

Quoting the `TOML v1.0.0 specification <https://toml.io/en/v1.0.0#spec>`_:

    A TOML file must be a valid UTF-8 encoded Unicode document.

``tomllib.loads`` does not intend to load a TOML file, but rather the
document that the file stores. The most natural representation of
a Unicode document in Python is ``str``, not ``bytes``.

It is possible to add ``bytes`` support in the future if needed, but
we are not aware of any use cases for it.


Controlling the type of mappings returned by ``tomllib.load[s]``
----------------------------------------------------------------

The ``toml`` library on PyPI accepts a ``_dict`` argument in its ``load[s]``
functions, which works similarly to the ``object_hook`` argument in
``json.load[s]``. There are several uses of ``_dict`` found on
https://grep.app; however, almost all of them are passing
``_dict=OrderedDict``, which should be unnecessary as of Python 3.7.
We found two instances of relevant use: in one case, a custom class was passed
for friendlier KeyErrors; in the other, the custom class had several
additional lookup and mutation methods (e.g. to help resolve dotted keys).

Such a parameter is not necessary for the core use cases outlined in the
`Motivation`_ section. The absence of this can be pretty easily worked around
using a wrapper class, transformer function, or a third-party library. Finally,
support could be added later in a backward-compatible way.


Removing support for ``parse_float`` in ``tomllib.load[s]``
-----------------------------------------------------------

This option is not strictly necessary, since TOML floats should be implemented
as "IEEE 754 binary64 values", which is equivalent to a Python ``float`` on most
architectures.

The TOML specification uses the word "SHOULD", however, implying a
recommendation that can be ignored for valid reasons. Parsing floats
differently, such as to ``decimal.Decimal``, allows users extra precision beyond
that promised by the TOML format. In the author of ``tomli``'s experience, this
is particularly useful in scientific and financial applications. This is also
useful for other cases that need greater precision, or where end-users include
non-developers who may not be aware of the limits of binary64 floats.

There are also niche architectures where the Python ``float`` is not a IEEE 754
binary64 value. The ``parse_float`` argument allows users to achieve correct
TOML semantics even on such architectures.


Alternative names for the module
--------------------------------

Ideally, we would be able to use the ``toml`` module name.

However, the ``toml`` package on PyPI is widely used, so there are backward
compatibility concerns. Since the standard library takes precedence over third
party packages, libraries and applications who current depend on the ``toml``
package would likely break when upgrading Python versions due to the many
API incompatibilities listed in `Appendix A <PEP 680 Appendix A_>`_,
even if they pin their dependency versions.

To further clarify, applications with pinned dependencies are of greatest
concern here. Even if we were able to obtain control of the ``toml`` PyPI
package name and repurpose it for a backport of the proposed new module,
we would still break users on new Python versions that included it in the
standard library, regardless of whether they have pinned an older version of
the existing ``toml`` package. This is unfortunate, since pinning
would likely be a common response to breaking changes introduced by repurposing
the ``toml`` package as a backport (that is incompatible with today's ``toml``).

Finally, the ``toml`` package on PyPI is not actively maintained, but as of
yet, efforts to request that the author add other maintainers
`have been unsuccessful <https://github.com/uiri/toml/issues/361>`__,
so action here would likely have to be taken without the author's consent.

Instead, this PEP proposes the name ``tomllib``. This mirrors ``plistlib``
and ``xdrlib``, two other file format modules in the standard library, as well
as other modules, such as ``pathlib``, ``contextlib`` and ``graphlib``.

Other names considered but rejected include:

* ``tomlparser``. This mirrors ``configparser``, but is perhaps somewhat less
  appropriate if we include a write API in the future.
* ``tomli``. This assumes we use ``tomli`` as the basis for implementation.
* ``toml`` under some namespace, such as ``parser.toml``. However, this is
  awkward, especially so since existing parsing libraries like ``json``,
  ``pickle``, ``xml``, ``html`` etc. would not be included in the namespace.


Previous Discussion
===================

* `bpo-40059: Provide a toml module in the standard library
  <https://bugs.python.org/issue40059>`_
* `[Python-Dev] Adding a toml module to the standard lib?
  <https://mail.python.org/pipermail/python-dev/2019-May/157405.html>`_
* `[Python-Ideas] Python standard library TOML module
  <https://mail.python.org/archives/list/python-ideas@python.org/thread/IWJ3I32A4TY6CIVQ6ONPEBPWP4TOV2V7/>`_
* `[Packaging] Adopting/recommending a toml parser?
  <https://discuss.python.org/t/adopting-recommending-a-toml-parser/4068>`_
* `hukkin/tomli#141: Please consider pushing tomli into the stdlib
  <https://github.com/hukkin/tomli/issues/141>`_


.. _PEP 680 Appendix A:

Appendix A: Differences between proposed API and ``toml``
=========================================================

This appendix covers the differences between the API proposed in this PEP and
that of the third-party package ``toml``. These differences are relevant to
understanding the amount of breakage we could expect if we used the ``toml``
name for the standard library module, as well as to better understand the design
space. Note that this list might not be exhaustive.

#. No proposed inclusion of a write API (no ``toml.dump[s]``)

   This PEP currently proposes not including a write API; that is, there will
   be no equivalent of ``toml.dump`` or ``toml.dumps``, as discussed at
   `Including an API for writing TOML`_.

   If we included a write API, it would be relatively straightforward to
   convert most code that uses ``toml`` to the new standard library module
   (acknowledging that this is very different from a compatible API, as it
   would still require code changes).

   A significant fraction of ``toml`` users rely on this, based on comparing
   `occurrences of "toml.load" <https://grep.app/search?q=toml.load&filter[lang][0]=Python>`__
   to `occurrences of "toml.dump" <https://grep.app/search?q=toml.dump&filter[lang][0]=Python>`__.

#. Different first argument of ``toml.load``

   ``toml.load`` has the following signature:

   .. code-block::

       def load(
           f: Union[SupportsRead[str], str, bytes, list[PathLike | str | bytes]],
           _dict: Type[MutableMapping[str, Any]] = ...,
           decoder: TomlDecoder = ...,
       ) -> MutableMapping[str, Any]: ...

   This is quite different from the first argument proposed in this PEP:
   ``SupportsRead[bytes]``.

   Recapping the reasons for this, previously mentioned at
   `Types accepted as the first argument of tomllib.load`_:

   * Allowing paths (and even lists of paths) as arguments is inconsistent with
     other similar functions in the standard library.
   * Using ``SupportsRead[bytes]`` allows us to ensure UTF-8 is the encoding used,
     and avoid incorrectly parsing single carriage returns as valid TOML.

   A significant fraction of ``toml`` users rely on this, based on manual
   inspection of `occurrences of "toml.load"
   <https://grep.app/search?q=toml.load&filter[lang][0]=Python>`__.

#. Errors

   ``toml`` raises ``TomlDecodeError``, vs. the proposed :pep:`8`-compliant
   ``TOMLDecodeError``.

   A significant fraction of ``toml`` users rely on this, based on
   `occurrences of "TomlDecodeError"
   <https://grep.app/search?q=TomlDecodeError&case=true&filter[lang][0]=Python>`__.

#. ``toml.load[s]`` accepts a ``_dict`` argument

   Discussed at `Controlling the type of mappings returned by tomllib.load[s]`_.

   As mentioned there, almost all usage consists of ``_dict=OrderedDict``,
   which is not necessary in Python 3.7 and later.

#. ``toml.load[s]`` support an undocumented ``decoder`` argument

   It seems the intended use case is for an implementation of comment
   preservation. The information recorded is not sufficient to roundtrip the
   TOML document preserving style, the implementation has known bugs, the
   feature is undocumented and we could only find one instance of its use on
   https://grep.app.

   The `toml.TomlDecoder interface
   <https://github.com/uiri/toml/blob/3f637dba5f68db63d4b30967fedda51c82459471/toml/decoder.pyi#L36>`__
   exposed is far from simple, containing nine methods.

   Users are likely better served by a more complete implementation of
   style-preserving parsing and writing.

#. ``toml.dump[s]`` support an ``encoder`` argument

   Note that we currently propose to not include a write API; however, if that
   were to change, these differences would likely become relevant.

   The ``encoder`` argument enables two use cases:

   * control over how custom types should be serialized, and
   * control over how output should be formatted.

   The first is reasonable; however, we could only find two instances of
   this on https://grep.app. One of these two used this ability to add
   support for dumping ``decimal.Decimal``, which a potential standard library
   implementation would support out of the box.
   If needed for other types, this use case could be well served by the
   equivalent of the ``default`` argument in ``json.dump``.

   The second use case is enabled by allowing users to specify subclasses of
   `toml.TomlEncoder
   <https://github.com/uiri/toml/blob/3f637dba5f68db63d4b30967fedda51c82459471/toml/encoder.pyi#L9>`__
   and overriding methods to specify parts of the TOML writing process. The API
   consists of five methods and exposes substantial implementation detail.

   There is some usage of the ``encoder`` API on https://grep.app; however, it
   appears to account for a tiny fraction of the overall usage of ``toml``.

#. Timezones

   ``toml`` uses and exposes custom ``toml.tz.TomlTz`` timezone objects. The
   proposed implementation uses ``datetime.timezone`` objects from the standard
   library.


Copyright
=========

This document is placed in the public domain or under the
CC0-1.0-Universal license, whichever is more permissive.