Myth Hash is a Python package designed to create human-readable hashes that are not only functional but also cognitively engaging. A key feature of this tool is its multilingual support, enabling the generation of hashes in multiple languages. The concept behind these human-readable hashes is to enhance communication and recognition by creating hashes that form vivid, memorable mental images. The goal is to make the hash as human-compatible as possible, ensuring it effectively serves its purpose in an intuitive and user-friendly manner.
- Character Name Generation: Generates a unique fantasy name based on an input string.
- Supported Languages: Currently supports English (
en) and German (de). - CLI Support: Easily generate names via the command line.
- Library Usage: Integrate
Myth Hashinto your Python projects. - Customizable Data: Modify the included JSON files to customize the generated names.
- Python: 3.11, 3.12, 3.13, 3.14, 3.15 (including free-threaded 3.14t and 3.15t)
- Platforms: Linux, macOS, Windows
You can install Myth Hash using pip:
pip install myth-hashTo run the CLI without installing it, use uv:
uvx myth-hash "The moon whispered secrets, but only the cats understood."After installing the package, you can use the myth-hash command to generate fantasy names.
Basic Example:
myth-hash "The moon whispered secrets, but only the cats understood." Example Output:
exotic-thoughtful-GriffinYou can also use Myth Hash within your Python code:
Example:
import myth_hash
# Generate a character name using the hash_name function
name_parts = myth_hash.hash_name(
"The moon whispered secrets, but only the cats understood.", "en"
)
print(name_parts) # Output: ('exotic', 'thoughtful', 'Griffin')
# Access individual parts of the name
physical_attr, personality_attr, mystical_character = name_parts
print(f"Physical Attribute: {physical_attr}")
print(f"Personality Trait: {personality_attr}")
print(f"Mystical Character: {mystical_character}")A name is built from one of 259 physical attributes, one of 239 personality attributes and one of 153 creatures, so there are 9,470,853 possible names. That is a small space compared to a cryptographic hash, and different inputs will regularly map to the same name.
The slow test (uv run nox -s test_slow) hashes 1,000,000 distinct inputs and counts duplicates:
| Generated names | Duplicates | Uniqueness |
|---|---|---|
| 1,000,000 | 119,949 | 88.0% |
By the birthday problem, the chance of at least one collision is about 5% for 1,000 inputs and above 99% for 10,000 inputs.
Use Myth Hash to make values easy to recognize and talk about, for example in logs, dashboards or UIs, but not as a unique identifier. Keep the original hash around when you need to tell values apart reliably.
These numbers apply to the bundled word lists; adding entries to the JSON files enlarges the name space and lowers the collision rate.
- Improved Uniqueness: Add more attributes and characters to reduce the likelihood of name collisions.
- Additional Languages: Add support for more languages with help from the community.
The project uses uv for dependency management and Nox as task runner.
uv sync # create .venv with all dev dependencies
uv run nox -s check # ruff (format + lint), pyrefly (strict), bandit
uv run nox -s test-3.14 # tests with coverage on a single Python version
uv run nox -s test # tests on all supported Python versions
uv run nox -s format_files # auto-format and apply safe lint fixesContributions are welcome! Please fork the repository, create a new branch for your feature or bug fix, and submit a pull request.
This project is licensed under the MIT License. See the LICENSE file for details.