fix: Handle unhashable BooleanField representation value

[emfpdlzj]

This is a small consistency fix for BooleanField.to_representation(), matching the existing TypeError handling used by to_internal_value() while preserving the existing bool(value) fallback.

Summary

BooleanField.to_representation() now handles unhashable values consistently with to_internal_value().

Previously, values like [], {}, or set() could raise TypeError during membership checks before reaching the existing bool(value) fallback. This change suppresses TypeError around the explicit true/false/null token checks, allowing unrecognized unhashable values to be represented by their truthiness.

Test

• Added BooleanField representation coverage for empty and non-empty unhashable collection values.
• Ran:
  uv run --no-project --with 'django==5.2.*' --with 'pytest==9.*' --with pytest-django --with dj-database-url --with importlib-metadata --with pytz -m pytest tests/test_fields.py -k BooleanField

[Copilot]

Pull request overview

This PR makes BooleanField.to_representation() resilient to unhashable values (e.g., [], {}, set()) by suppressing TypeError during membership checks, aligning its behavior with to_internal_value() while preserving the existing bool(value) fallback.

Changes:

• Wrap BooleanField.to_representation()’s TRUE/FALSE/NULL token membership checks in contextlib.suppress(TypeError).
• Update BooleanField output tests to use a list of (value, expected) pairs (allowing unhashable test inputs) and add coverage for unhashable collections.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
rest_framework/fields.py	Suppresses TypeError during set-membership checks in BooleanField.to_representation() so unhashable values fall back to bool(value).
tests/test_fields.py	Extends BooleanField representation coverage to include unhashable collections by switching outputs to tuple-pairs and adding new cases.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Natgho]

PR seems reasonable and appropriate.

Only a potential edge-case scenario might be cause for concern;

Let's say that, either by mistake or due to a custom library integration, a complex custom object that cannot be converted to a string or whose type cannot be checked was passed to the to_representation method:

Old Behavior: The code raised a TypeError and crashed (the API returned a 500 error).

New Behavior: The code does not crash; instead, it passes the object through the bool(object) filter. Since all non-empty or non-None custom objects in Python return True by default, the API now returns the True value as JSON without crashing.

However, this is simply a bug fix that needed to be addressed anyway. That’s why I think it’s acceptable.

[browniebroke]

This is a small consistency fix for BooleanField.to_representation(), matching the existing TypeError handling used by to_internal_value() while preserving the existing bool(value) fallback.

Taking a step back away from the model field itself, what concrete example would cause that from being used? I'm thinking of a minimal example of a serializer with the data/Python object that would make use of this. I think we would need to map a complex field (e.g. dict) to a BooleanField?

[Natgho]

This is a small consistency fix for BooleanField.to_representation(), matching the existing TypeError handling used by to_internal_value() while preserving the existing bool(value) fallback.

Taking a step back away from the model field itself, what concrete example would cause that from being used? I'm thinking of a minimal example of a serializer with the data/Python object that would make use of this. I think we would need to map a complex field (e.g. dict) to a BooleanField?

Yes, actually, this is more of an improvement aimed at preventing the “representation” step from crashing. As it stands now, the system crashes when dealing with a complex variable type because there is no mechanism to handle this situation. This code prevents that error and manages this part using a boolean.