Skip to main content

Tribool is an Apache2 licensed Python module that implements three-valued logic.

Suppose for a moment that you’re attempting to store a value across a network connection. You begin with a simple protocol in which the server stores the received value and then sends an acknowledgement to the client. In this design, the client experiences a delay between when the request is sent and the acknowledgement is received. In that delay, it is impossible for the client to know whether the value has been committed on the server. In such cases, it’s useful to describe the commit state of the server from the client’s perspective as True, False, or Indeterminate.

Another example occurs in database systems. Consider a record that contains a boolean field. Such a field may only be either True or False. But we want to support the notion of committing a partial record in the case that the record is large or the client does not have all relevant information. In this scenario, we wish to commit neither True nor False as the value is currently Unknown.

The Python Tribool module was designed for these cases by describing a logical data type that supports three values: True, False, and Indeterminate. The third value is best thought of as a state being either True or False. Given these three values it’s possible to define truth tables over the logical operators and, or and not and to define equality and inequality relationships.

Features

  • Pure-Python (easy to hack with)

  • Fully Documented

  • 100% Test Coverage

  • Pragmatic Design (mostly a few truth tables and thread-safe singleton pattern)

  • Developed on Python 2.7

  • Tested on CPython 2.6, 2.7, 3.2, 3.3, 3.4 and PyPy 2.5+, PyPy3 2.4+

Quickstart

Installing Tribool is simple with pip:

$ pip install tribool

You can access documentation in the interpreter with Python’s built-in help function:

>>> from tribool import Tribool
>>> help(Tribool)

Tutorial

A Python Tribool may have any of three values:

>>> from tribool import Tribool
>>> Tribool(True)          # True
>>> Tribool('False')       # False
>>> Tribool(Tribool(None)) # Indeterminate

Those three values correspond to True, False and Indeterminate. To view that value, convert the Tribool to a string:

>>> print Tribool(True), Tribool(False), Tribool(None)
True False Indeterminate

The logical operators are also defined over these values. For example, the result of negation:

>>> for value in (True, False, None):
...     print '~', Tribool(value), '=', ~Tribool(value)
~ True = False
~ False = True
~ Indeterminate = Indeterminate

Likewise for and, or, and xor the operators involving only True and False are unchanged. And mostly those involving Indeterminate result in Indeterminate. For example:

>>> True & Tribool(None)  # True and Indeterminate = Indeterminate
Tribool(None)
>>> False | Tribool(None) # False or Indeterminate = Indeterminate
Tribool(None)
>>> None ^ Tribool(None)  # Indeterminate xor Indeterminate = Indeterminate
Tribool(None)

But there are a couple cases where this is not so:

>>> True | Tribool(None)  # True or Indeterminate = True
Tribool(True)
>>> False & Tribool(None) # False and Indeterminate = False
Tribool(False)

Notice that the bitwise-operators, &|^~, have been used rather than the short-circuiting and, or, not. Python supports short-circuiting operators only for boolean values and you cannot implicitly convert a Tribool to a boolean. An attempt to do so will raise a ValueError:

>>> not Tribool(True)
Traceback (most recent call last):
  ...
ValueError: Cannot implicitly convert Tribool to bool (use the bitwise
(&, |, ^, ~) operators or insert a cast and use Tribool(...).value)

For this reason, you cannot directly use a Tribool in an if statement:

>>> if Tribool(True): pass
Traceback (most recent call last):
  ...
ValueError: Cannot implicitly convert Tribool to bool ...

To test the value of a Tribool, use the value property:

>>> print Tribool(True).value, Tribool(False).value, Tribool(None).value
True False None
>>> (Tribool(None) | True).value is True
True
>>> ready, committed = Tribool(True), Tribool(None)
>>> if (ready & committed).value is not True:
...     print 'Still waiting.'
Still waiting.

When the Tribool value is Indeterminate, the value property will be None. For example:

>>> status = Tribool(None)
>>> # Do something that will update status.
>>> while status.value is None:
...     time.sleep(1) # Busy-wait.
>>> if status.value is True:
...     print 'Success'
... else:
...     print 'Error'

Tribools also work with equality/inequality relationships. Comparing Tribools returns a Tribool because the result may be ambiguous. For the less-than and greater-than relationships, True corresponds to 1 and False to 0 just as with boolean data types. The Indeterminate value is either 0 or 1 which has some unusual implications. Some example inequalities:

>>> Tribool(False) < Tribool(True)
Tribool(True)
>>> Tribool(False) == Tribool(False)
Tribool(True)
>>> Tribool(False) > Tribool(True)
Tribool(False)

The unusual implication of the Indeterminate value is that it is not equal to itself:

>>> print Tribool(True) >= Tribool(None)
True
>>> print Tribool(False) < Tribool(None)
Indeterminate
>>> print Tribool(None) == Tribool(None)
Indeterminate

When an object is not equal to itself, strange things can happen. Fortunately Python defines two notions of equality. The first is defined by the is relationship and may not be overriden. The second is defined by the __eq__ method. To behave as value types, Tribool objects are singletons. Threrefore two Tribools with the same value will have matching id values. For example:

>>> (id(Tribool(True)), id(Tribool(True)), id(Tribool(True)))
(4426760848, 4426760848, 4426760848)
>>> (id(Tribool(None)), id(Tribool(None)), id(Tribool(None)))
(4426719568, 4426719568, 4426719568)

This is accomplished by overriding the __new__ constructor and implementing a thread-safe singleton pattern. As singletons, Tribool objects are immutable and comparable using the is operator. Judicious use often results in code that is more readable:

>>> Succeeded, TryAgain = Tribool(True), Tribool(None)
>>> status = Tribool(None)
>>> while status is TryAgain:
...     status = try_something()
>>> if status is Succeeded:
...     print 'Success!'

Tribool objects are also hashable and work inside dict and map-like types:

>>> display = {
...     Tribool(True): 'Success',
...     Tribool(False): 'Error',
...     Tribool(None): 'Try Again',
... }
>>> print display[Tribool(None)]
Try Again

A surprising result occurs however with containers. When using the in operator, objects are tested for membership using equality. But this occurs in several steps, the first of which is using the is operator followed by the __eq__ method. In case the __eq__ method fails to return a boolean- typed value, an implicit conversion occurs which Tribool does not permit. For example:

>>> Success, Error, Unknown = map(Tribool, (True, False, None))
>>> Success in [Success, Error, Unknown] # Works!
True
>>> Error in [Success, Error, Unknown]   # Fails
Traceback (most recent call last):
  ...
ValueError: Cannot implicitly convert Tribool to bool ...

The latter attempt fails because Error is Success returns False and so Error == Success is tried. That returns Tribool(False) which does not have type bool and so an implicit conversion occurs. To achieve the affect of the in operator use the any built-in and a generator expression like so:

>>> statuses = [Success, Success, Unknown, Error]
>>> any(status is Error for status in statuses)
True

To obey the singleton pattern, Tribool also implements the __copy__ and __deepcopy__ methods as part of the copy module protocol. Pickling is another method of copying objects and so __reduce__ is implemented as part of the pickle protocol. Note also that Tribool inherits directly from tuple to prevent mutation of its internal state.

The Python Tribool module has many uses but it was originally designed to support the notion of three-valued logic as found in SQL. SQL defines similar rules for its Null value type in logical expressions. Django’s NullBooleanField is an example where these ideas intersect.

Some readers will be familiar with Boost.Tribool, an implementation of the Tribool datatype in C++. While the semantics of both packages are the same, the design of the Boost implementation differs a great deal. In particular, Boost defines a new indeterminate keyword rather than using the null value in C++. An Indeterminate object was considered in the design of this module but discarded in favor of Python’s built-in None.

Reference and Indices

License

Copyright 2015 Grant Jenks

Licensed under the Apache License, Version 2.0 (the “License”); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an “AS IS” BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

Release files for tribool 0.7.3

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for tribool 0.7.3
File Size Uploaded
tribool-0.7.3.tar.gz 7.7 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for tribool 0.7.3
File Interpreter ABI Platform
tribool-0.7.3-py2.py3-none-any.whl Python 2, Python 3 none any Details

Total release size: 19.2 kB

Release files / tribool-0.7.3.tar.gz

Download URL tribool-0.7.3.tar.gz
Size 7.7 kB
Tags Source
SHA-256 checksum
How to use checksums
201ebea42e996c921b10b34f07f76bde0c62d2605b610587a2bc24b134bcdb13
BLAKE2b-256 checksum
How to use checksums
c4109442d674824e0f8b6b895cd5836870afcdda53f09e9120a8a7e5c2bce27e
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No

Release files / tribool-0.7.3-py2.py3-none-any.whl

Download URL tribool-0.7.3-py2.py3-none-any.whl
Size 11.5 kB
Tags Python 2 Python 3
SHA-256 checksum
How to use checksums
fde13951bf571ee9e1f9136461e11c686f0247d44ce005e49fb32c126ca9a617
BLAKE2b-256 checksum
How to use checksums
aa32cd997ace62eafa0d8a1f69d9163447a68ef09b807700948bbb5e2c87770d
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No

Release history Release notifications | RSS feed

This release

0.7.3 This release

2 release files

0.7.2

2 release files

0.7.1

2 release files

0.6.2

1 release file

0.6.1

1 release file

0.5.4

1 release file

0.5.3

1 release file

0.5.2

1 release file

0.5.1

0.0.5

1 release file

0.0.4

1 release file

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page