🔍 pytest-deepassert

Enhanced pytest assertions with detailed diffs powered by DeepDiff

Stop hunting through massive data dumps. Get precise, readable assertion failures.

---

Table of Contents

• What is it for?
• How it works
• Installation
• Key Features
• Comparison with Standard Assertions
• Usage
• Configuration
• API Reference
• Limitations
• License

---

What is it for?

When testing that one complex data structure equals another, you may have situations when only some small details differ, but it is really hard to see which ones because of the cluttered comparison report. That's when pytest-deepassert helps:

• Quicker identification of differences in large & nested objects
• Focused output that highlights what matters
• Direct navigation to specific changes
• Structured diffs for complex data

If you've ever struggled with understanding WHAT EXACTLY is mismatching in your assert a == b statement, pytest-deepassert is what you need!

How it works

pytest-deepassert is a pytest plugin built on powerful deepdiff library. It provides clear, detailed difference reports for various data types:

• Basic types: int, string, float, bool
• Collections: dict, list, tuple, set, frozenset
• Advanced types: OrderedDict, NamedTuple, custom objects

🚀 Installation

From PyPI (Recommended)

pip install pytest-deepassert

From Source

git clone https://github.com/alexkuzmik/pytest-deepassert.git
cd pytest-deepassert
pip install -e .

Requirements: Python 3.8+ and pytest 6.0+

✨ Key Features

Precise Error Location

• Pinpoints exact paths where differences occur
• No more hunting through massive data dumps
• Shows specific field names, values, and array indices

Clear Change Description

• Shows left value → right value for each difference
• Categorizes changes (values changed, items added/removed)
• Human-readable change descriptions

Smart Comparison Helpers

• Works seamlessly with pytest.approx(), mock.ANY and any custom comparator utilities
• Handles complex nested structures intelligently

Simple Opt-In

• Enable with --deepassert flag - disabled by default to avoid interference
• No additional configuration needed
• Drop-in replacement for standard assertions when enabled

---

📊 Comparison with Standard Assertions

TL;DR: Standard pytest assertions work great for simple cases, but their string comparison reports are not optimized for complex data structures and utility comparator objects. That's where pytest-deepassert shines.

Example 1: Complex Dictionary Comparison

Consider this realistic test with nested dictionaries:

Click to see the test code

import pytest
from unittest.mock import ANY

def test_user_profile_comparison():
    expected = {
        "user": {
            "id": 123,
            "name": "John Doe",
            "email": "[email protected]",
            "preferences": {
                "theme": "dark",
                "notifications": True,
                "language": "en"
            },
            "metadata": {
                "created_at": ANY,  # We don't care about exact timestamp
                "last_login": "2023-12-01",
                "login_count": 42,
                "score": pytest.approx(85.5, abs=0.1)  # Approximate float comparison
            }
        },
        "permissions": ["read", "write", "admin"]
    }

    actual = {
        "user": {
            "id": 123,
            "name": "Jane Doe",  # Different name
            "email": "[email protected]",  # Different email
            "preferences": {
                "theme": "light",  # Different theme
                "notifications": True,
                "language": "es"  # Different language
            },
            "metadata": {
                "created_at": "2023-01-01T10:30:00Z",  # This will match ANY
                "last_login": "2023-11-30",  # Different date
                "login_count": 45,  # Different count
                "score": 85.52  # Close enough to match approx
            }
        },
        "permissions": ["read", "write", "admin", "delete"]  # Extra "delete" permission
    }

    assert expected == actual

📄 Standard pytest output

Click to see the standard output

example_test1.py::test_user_profile_comparison FAILED

======================================== FAILURES ========================================
_________________________ test_user_profile_comparison _________________________
example_test1.py:45: in test_user_profile_comparison
    assert expected == actual
E   AssertionError: assert {'user': {'id': 123, 'name': 'John Doe', 'email': '[email protected]', 'preferences': {'theme': 'dark', 'notifications': True, 'language': 'en'}, 'metadata': {'created_at': <ANY>, 'last_login': '2023-12-01', 'login_count': 42, 'score': 85.5 ± 0.1}}, 'permissions': ['read', 'write', 'admin']} == {'user': {'id': 123, 'name': 'Jane Doe', 'email': '[email protected]', 'preferences': {'theme': 'light', 'notifications': True, 'language': 'es'}, 'metadata': {'created_at': '2023-01-01T10:30:00Z', 'last_login': '2023-11-30', 'login_count': 45, 'score': 85.52}}, 'permissions': ['read', 'write', 'admin', 'delete']}

E
E     Differing items:
E     {'permissions': ['read', 'write', 'admin']} != {'permissions': ['read', 'write', 'admin', 'delete']}
E     {'user': {'email': '[email protected]', 'id': 123, 'metadata': {'created_at': <ANY>, 'last_login': '2023-12-01', 'login_count': 42, 'score': 85.5 ± 0.1}, 'name': 'John Doe', ...}} != {'user': {'email': '[email protected]', 'id': 123, 'metadata': {'created_at': '2023-01-01T10:30:00Z', 'last_login': '2023-11-30', 'login_count': 45, 'score': 85.52}, 'name': 'Jane Doe', ...}}

E
E     Full diff:
E       {
E           'permissions': [
E               'read',
E               'write',
E               'admin',
E     -         'delete',
E           ],
E           'user': {
E     -         'email': '[email protected]',
E     ?                    ^ -
E     +         'email': '[email protected]',
E     ?                    ^^
E               'id': 123,
E               'metadata': {
E     -             'created_at': '2023-01-01T10:30:00Z',
E     +             'created_at': <ANY>,
E     -             'last_login': '2023-11-30',
E     ?                                   ---
E     +             'last_login': '2023-12-01',
E     ?                                  +++
E     -             'login_count': 45,
E     ?                             ^
E     +             'login_count': 42,
E     ?                             ^
E     -             'score': 85.52,
E     ?                          ^
E     +             'score': 85.5 ± 0.1,
E     ?                          ^^^^^^
E               },
E     -         'name': 'Jane Doe',
E     ?                   ^ -
E     +         'name': 'John Doe',
E     ?                   ^^
E               'preferences': {
E     -             'language': 'es',
E     ?                           ^
E     +             'language': 'en',
E     ?                           ^
E                   'notifications': True,
E     -             'theme': 'light',
E     ?                       ^^^^^
E     +             'theme': 'dark',
E     ?                       ^^^^
E               },
E           },
E       }

✨ With pytest-deepassert (with pytest --deepassert -vv)

example_test1.py::test_user_profile_comparison FAILED

======================================== FAILURES ========================================
_________________________ test_user_profile_comparison _________________________
example_test1.py:45: in test_user_profile_comparison
    assert expected == actual
E   assert
E     DeepAssert detailed comparison:
E         Item root['permissions'][3] ("delete") added to iterable.
E         Value of root['user']['name'] changed from "John Doe" to "Jane Doe".
E         Value of root['user']['email'] changed from "[email protected]" to "[email protected]".
E         Value of root['user']['preferences']['theme'] changed from "dark" to "light".
E         Value of root['user']['preferences']['language'] changed from "en" to "es".
E         Value of root['user']['metadata']['last_login'] changed from "2023-12-01" to "2023-11-30".
E         Value of root['user']['metadata']['login_count'] changed from 42 to 45.
E
E     [... standard pytest diff continues below ...]

🎯 Key Improvements

Feature	Standard pytest	pytest-deepassert
Smart filtering	Shows all field comparisons	✅ Ignores created_at (matches ANY)
Precision	Comprehensive diff coverage	✅ Ignores score (within pytest.approx tolerance)
Focus	Complete context provided	🎯 Highlights actual differences
Format	String-based comparison	📋 Structured, categorized output

📋 Example 2: Smart Comparison Helpers

pytest-deepassert seamlessly handles special comparison helpers:

Click to see the test code

import pytest
from unittest.mock import ANY

def test_with_special_comparisons():
    expected = {
        "timestamp": ANY,  # We don't care about exact timestamp
        "value": pytest.approx(3.14159, abs=0.001),  # Approximate float comparison
        "metadata": {
            "version": "1.0.0",
            "debug": False
        }
    }

    actual = {
        "timestamp": "2023-12-01T10:30:00Z",
        "value": 3.14160,  # Close enough
        "metadata": {
            "version": "1.0.1",  # Different version
            "debug": False
        }
    }

    assert expected == actual

📄 Standard pytest output

Click to see the standard output

example_test2.py::test_with_special_comparisons FAILED

======================================== FAILURES ========================================
______________________________________ test_with_special_comparisons _______________________________________
example_test2.py:23: in test_with_special_comparisons
    assert expected == actual
E   AssertionError: assert {'timestamp': <ANY>, 'value': 3.14159 ± 0.001, 'metadata': {'version': '1.0.0', 'debug': False}} == {'timestamp': '2023-12-01T10:30:00Z', 'value': 3.1416, 'metadata': {'version': '1.0.1', 'debug': False}}

E
E     Common items:
E     {'timestamp': <ANY>, 'value': 3.14159 ± 0.001}
E     Differing items:
E     {'metadata': {'debug': False, 'version': '1.0.0'}} != {'metadata': {'debug': False, 'version': '1.0.1'}}

E
E     Full diff:
E       {
E           'metadata': {
E               'debug': False,
E     -         'version': '1.0.1',
E     ?                         ^
E     +         'version': '1.0.0',
E     ?                         ^
E           },
E     -     'timestamp': '2023-12-01T10:30:00Z',
E     +     'timestamp': <ANY>,
E     -     'value': 3.1416,
E     ?                   ^
E     +     'value': 3.14159 ± 0.001,
E     ?                   ^^^^^^^^^^
E       }

✨ With pytest-deepassert

example_test2.py::test_with_special_comparisons FAILED

======================================== FAILURES ========================================
______________________________________ test_with_special_comparisons _______________________________________
example_test2.py:23: in test_with_special_comparisons
    assert expected == actual
E   assert
E     DeepAssert detailed comparison:
E         Value of root['metadata']['version'] changed from "1.0.0" to "1.0.1".
E
E     [... standard pytest diff continues below ...]

Notice: Only the actual difference (version) is shown. The timestamp and value fields are correctly ignored!

---

💡 Usage

After installation, enable pytest-deepassert by passing the --deepassert flag when running pytest:

pytest --deepassert

This will enhance all your == assertions inside the tests. No code changes required!

---

Configuration

pytest-deepassert is disabled by default to avoid any interference with your existing test suite. Enable it when needed:

Command Line Options

Option	Description
--deepassert	Enable pytest-deepassert for this test run

---

API Reference

pytest_deepassert.equal(left, right, verbose_level=2)

For use cases where you need enhanced assertions outside of test modules (e.g., in helper functions), you can use the equal() function directly.

Parameters:

• left (Any): The expected object
• right (Any): The actual object
• verbose_level (int, optional): Controls the verbosity of the diff report. Default: 2
  • 0: Minimal output (only reports if objects are different)
  • 1: Standard output (shows changes with brief details)
  • 2: Detailed output (shows full changes with all details and types)

Example:

import pytest_deepassert

def helper_function_for_assertion(actual, expected):
    # Use default verbose_level=2 for detailed output
    pytest_deepassert.equal(expected, actual)

def another_helper(actual, expected):
    # Use verbose_level=1 for less detailed output
    pytest_deepassert.equal(expected, actual, verbose_level=1)

Note: The traceback will automatically hide the internal frames of the equal() function, showing only where it was called from.

---

Limitations

The tool only enhances assertions inside the test modules (pytest limitation). If you want to have deep assertion reports in the other modules of your project (e.g. some helper functions for your testlib), consider using pytest_deepassert.equal(left, right) function as described in the API Reference section.

---

License

This project is licensed under the MIT License - see the LICENSE file for details.

---

Links

• PyPI Package
• GitHub Repository
• Issue Tracker
• DeepDiff Library

---

Made by Alexander Kuzmik

If this project helped you, please consider giving it a ⭐ on GitHub!