This repository holds example code that demos some of the new features in Python 3.11.
You need Python 3.11 installed to run these examples. See the following tutorial instructions:
You can learn more about Python 3.11's new features in the following Real Python tutorials:
- Python 3.11 Preview: Even Better Error Messages
- Python 3.11 Preview: Task and Exception Groups
- Python 3.11 Preview: TOML and
tomllib - Python 3.11: Cool New Features for You to Try
You'll find examples from all these tutorials in this repository.
Create and activate a virtual environment:
$ python -m venv venv
$ source venv/bin/activateInstall necessary dependencies for the examples (see requirements.in):
(venv) $ python -m pip install aiohttp colorama parse python-magicAlternatively, you can install dependencies from requirements.txt if you want to ensure that you're using the same versions of the third-party packages:
(venv) $ python -m pip install -r requirements.txtNote:
python-magicdepends on a C library that you may need to install on your system. See the documentation for details.
These examples have been run with Python 3.11.0rc1, the first release candidate of Python 3.11.
This section only contains brief instructions on how you can run the examples. See the tutorials for technical details.
Run `inverse.py for a quick look at the improved tracebacks:
(venv) $ python inverse.py
Traceback (most recent call last):
File "/home/realpython/inverse.py", line 5, in <module>
print(inverse(0))
^^^^^^^^^^
File "/home/realpython/inverse.py", line 2, in inverse
return 1 / number
~~^~~~~~~~
ZeroDivisionError: division by zeroThe scientists.py and programmers.py show two different versions of the same kind of example.
For example, load scientists.py into your interactive REPL:
(venv) $ python -i scientists.py You can then experiment with dict_to_person() and convert_pair():
>>> dict_to_person(scientists[1])
Traceback (most recent call last):
...
File "/home/realpython/scientists.py", line 37, in dict_to_person
name=f"{info['name']['first']} {info['name']['last']}",
~~~~~~~~~~~~^^^^^^^^
KeyError: 'last'
>>> convert_pair(scientists[0], scientists[2])
Traceback (most recent call last):
...
File "/home/realpython/scientists.py", line 44, in convert_pair
return dict_to_person(first), dict_to_person(second)
^^^^^^^^^^^^^^^^^^^^^^
File "/home/realpython/scientists.py", line 38, in dict_to_person
life_span=(info["birth"]["year"], info["death"]["year"]),
~~~~~~~~~~~~~^^^^^^^^
TypeError: 'NoneType' object is not subscriptableSee Even Better Error Messages in Python 3.11, More Informative Error Tracebacks, and PEP 657.
The interpreter will specialize bytecodes that are run often. This is part of the Faster CPython project. Note how the bytecodes change as the program is running:
(venv) $ python specialized.py
============================ Before any invocation ============================
8 0 RESUME 0
9 2 LOAD_CONST 1 (0.3048)
4 LOAD_FAST 0 (feet)
6 BINARY_OP 5 (*)
10 RETURN_VALUE
========================== After 7 float invocations ==========================
8 0 RESUME 0
9 2 LOAD_CONST 1 (0.3048)
4 LOAD_FAST 0 (feet)
6 BINARY_OP 5 (*)
10 RETURN_VALUE
========================== After 8 float invocations ==========================
8 0 RESUME_QUICK 0
9 2 LOAD_CONST__LOAD_FAST 1 (0.3048)
4 LOAD_FAST 0 (feet)
6 BINARY_OP_MULTIPLY_FLOAT 5 (*)
10 RETURN_VALUE
===================== After 8 float and 52 int invocations ====================
8 0 RESUME_QUICK 0
9 2 LOAD_CONST__LOAD_FAST 1 (0.3048)
4 LOAD_FAST 0 (feet)
6 BINARY_OP_MULTIPLY_FLOAT 5 (*)
10 RETURN_VALUE
===================== After 8 float and 53 int invocations ====================
8 0 RESUME_QUICK 0
9 2 LOAD_CONST__LOAD_FAST 1 (0.3048)
4 LOAD_FAST 0 (feet)
6 BINARY_OP_ADAPTIVE 5 (*)
10 RETURN_VALUESee Faster Code Execution and PEP 659.
Use ExceptionGroup and except* to handle several errors at once:
(venv) $ python exception_group.py
Handling ValueErrors: (ValueError(654),)
+ Exception Group Traceback (most recent call last):
| File "/home/realpython/exception_group.py", line 2, in <module>
| raise ExceptionGroup(
| ExceptionGroup: group (2 sub-exceptions)
+-+---------------- 1 ----------------
| TypeError: str
+---------------- 2 ----------------
| TypeError: int
+------------------------------------See Exception Groups and except* in Python 3.11, Exception Groups, and PEP 654.
Run count.py, count_gather.py, and count_taskgroup.py and compare their behaviors. For example:
(venv) $ python count_taskgroup.py scientists.py rot13.txt count.py
scientists.py □□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□ (44)
Files with thirteen lines are too scary!
count.py □□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□□ (32)See Asynchronous Task Groups in Python 3.11 and gh-31270.
For a different example, have a look at download_peps_gather.py and `download_peps_taskgroup.py. For example:
(venv) $ python download_peps_taskgroup.py
Downloading PEP 492
Downloading PEP 525
Downloading PEP 530
Downloading PEP 3148
Downloading PEP 3156
Downloaded PEP 492: Coroutines with async and await syntax
Downloaded PEP 530: Asynchronous Comprehensions
Downloaded PEP 525: Asynchronous Generators
Downloaded PEP 3148: futures - execute computations asynchronously
Downloaded PEP 3156: Asynchronous IO Support Rebooted: the "asyncio" ModuleSee Nicer Syntax for Asynchronous Tasks.
python_info.toml, tomli_pyproject.toml, and units.toml show three examples of TOML files.
Use read_toml.py to read them:
(venv) $ python read_toml.py python_info.toml tomli_pyproject.toml units.toml
======================python_info.toml======================
{'python': {'version': 3.11,
'release_manager': 'Pablo Galindo Salgado',
'is_beta': True,
'beta_release': 3,
'release_date': datetime.date(2022, 6, 16),
'peps': [657, 654, 678, 680, 673, 675, 646, 659]}}
====================tomli_pyproject.toml====================
{'build-system': {'requires': ['flit_core>=3.2.0,<4'],
'build-backend': 'flit_core.buildapi'},
'project': {'name': 'tomli',
'version': '2.0.1',
'description': "A lil' TOML parser",
'requires-python': '>=3.7',
'readme': 'README.md',
'keywords': ['toml'],
'urls': {'Homepage': 'https://github.com/hukkin/tomli',
'PyPI': 'https://pypi.org/project/tomli'}}}
=========================units.toml=========================
{'second': {'label': {'singular': 'second', 'plural': 'seconds'},
'aliases': ['s', 'sec', 'seconds']},
'minute': {'label': {'singular': 'minute', 'plural': 'minutes'},
'aliases': ['min', 'minutes'],
'multiplier': 60,
'to_unit': 'second'},
'hour': {'label': {'singular': 'hour', 'plural': 'hours'},
'aliases': ['h', 'hr', 'hours'],
'multiplier': 60,
'to_unit': 'minute'},
'day': {'label': {'singular': 'day', 'plural': 'days'},
'aliases': ['d', 'days'],
'multiplier': 24,
'to_unit': 'hour'},
'year': {'label': {'singular': 'year', 'plural': 'years'},
'aliases': ['y', 'yr', 'years', 'julian_year', 'julian years'],
'multiplier': 365.25,
'to_unit': 'day'}}You can use tomllib_w.py to write simplified TOML files. write_toml.py demonstrates how you can use it:
(venv) $ python write_toml.py
url = "https://realpython.com/python311-tomllib/"
[author]
name = "Geir Arne Hjelle"
email = "[email protected]"You can see an example of using a TOML file to set up a flexible unit converter in units.py.
See tomllib TOML Parser in Python 3.11, Support for TOML Configuration Parsing, and PEP 680.
polar_point.py and programmers.py uses Self for annotation:
(venv) $ python polar_point.py
PolarPoint(r=5.0, φ=0.9272952180016122)See Self Type, Improved Type Variables, and PEP 673.
execute_sql.py shows an example of LiteralString:
(venv) $ python execute_sql.py
Pretending to execute: SELECT * FROM users
Pretending to execute: SELECT * FROM users
Enter table name: users; DROP TABLE users; --
Pretending to execute: SELECT * FROM users; DROP TABLE users; --See Arbitrary Literal String Type and PEP 675.
ndarray.py and tuple_order.py show examples of using TypeVarTuple.
See Variadic Generic Types, Improved Type Variables, and PEP 646.
The internal handling of exceptions has been restructured. Exception handling is now done with a jump-table:
(venv) $ python zero_cost_exceptions.py
4 0 RESUME 0
5 2 NOP
6 4 LOAD_CONST 1 (1)
6 LOAD_FAST 0 (number)
8 BINARY_OP 11 (/)
12 RETURN_VALUE
>> 14 PUSH_EXC_INFO
7 16 LOAD_GLOBAL 0 (ZeroDivisionError)
28 CHECK_EXC_MATCH
30 POP_JUMP_FORWARD_IF_FALSE 19 (to 70)
32 POP_TOP
8 34 LOAD_GLOBAL 3 (NULL + print)
46 LOAD_CONST 2 ('0 has no inverse')
48 PRECALL 1
52 CALL 1
62 POP_TOP
64 POP_EXCEPT
66 LOAD_CONST 0 (None)
68 RETURN_VALUE
7 >> 70 RERAISE 0
>> 72 COPY 3
74 POP_EXCEPT
76 RERAISE 1
ExceptionTable:
4 to 10 -> 14 [0]
14 to 62 -> 72 [1] lasti
70 to 70 -> 72 [1] lastiSee Zero-Cost Exceptions and gh-84403.
Use .add_notes() to annotate exceptions with custom notes:
(venv) $ python exception_notes.py
err.__notes__ = ['Enriching Exceptions with Notes', 'Python 3.11']
--------------------- Loop over notes ----------------------
Enriching Exceptions with Notes
Python 3.11
--------------- Notes are added to traceback ---------------
Traceback (most recent call last):
File "/home/realpython/exception_notes.py", line 12, in <module>
raise err
ValueError: 678
Enriching Exceptions with Notes
Python 3.11You can use exception notes to add context to errors raised by your program. For example, timestamped_errors.py adds a timestamp to an error:
(venv) python timestamped_errors.py
Traceback (most recent call last):
...
ZeroDivisionError: division by zero
Raised at 2022-10-18 12:28:55.447764See Annotate Exceptions With Custom Notes, Exception Notes, and PEP 678.
You can use sys.exception() to access the active exception:
(venv) $ python active_exception.py
Handling gh-90486
Handling gh-90486Note that this is typically not necessary in regular code. You can use the except ValueError as err syntax instead.
See Reference the Active Exception With sys.exception() and gh-90486.
traceback_demo.py shows that tracebacks can be consistently accessed through the exception object:
(venv) $ python traceback_demo.py
tb_last(exc_value.__traceback__) = 'bad_calculation:13'
tb_last(exc_tb) = 'bad_calculation:13'See Reference the Active Traceback Consistently and gh-89874.
You can use math.cbrt() to calculate cube roots:
(venv) $ python cube_root.py
math.cbrt(729) = 9.000000000000002
729 ** (1 / 3) = 8.999999999999998
math.pow(729, 1 / 3) = 8.999999999999998You can use math.exp2() to calculate powers of two:
(venv) $ python power_of_two.py
math.exp2(16) = 65536.0
2**16 = 65536
math.pow(2, 16) = 65536.0See Cube Roots and Powers of Two, gh-88523 and gh-90075.
You can use underscores when defining fractions from strings:
(venv) $ python underscore.py
Fraction('6_024/1_729') = 6024/1729See Underscores in Fractions and gh-88424.
The Norwegian calculator implemented in kalkulator.py uses operator.call():
(venv) $ python kalkulator.py
20 pluss 22 = 42.0
2022 minus 1991 = 31.0
45 ganger 45 = 2025.0
11 delt på 3 = 3.6666666666666665See Flexible Calling of Objects and gh-88185.
Some of the less used standard libraries have started raising deprecation warnings. If you happen to be using them, you should look for more modern alternatives.
For example, dead_imghdr.py shows the warning raised by importing imghdr and how to use python-magic as an alternative:
(venv) $ python dead_imghdr.py
/home/realpython/dead_imghdr.py:1: DeprecationWarning: 'imghdr' is deprecated and slated for removal in Python 3.13
import imghdr
jpeg
JPEG image data, JFIF standard 1.02, aspect ratio, density 100x100, segment length 16, progressive, precision 8, 1920x1080, components 3See Dead Batteries and PEP 594.
- Geir Arne Hjelle, E-mail: [email protected]
Distributed under the MIT license. See LICENSE for more information.