DAWG2 Python Library

Common errors

• AttributeError: 'DAWG' object has no attribute 'save'

  cause Attempting to save a DAWG created or loaded with `dawg_python` (dawg2-python). This library is read-only.
  fix Use the original `dawg` library (e.g., `import dawg; d = dawg.DAWG(words); d.save('file.dawg')`) to create and save DAWG files. `dawg_python` is only for loading and querying.

• TypeError: DAWG() takes no arguments

  cause Attempting to initialize `dawg_python.DAWG` with a list of words or other data. The constructor is intended for internal use or loading, not direct data input for building a DAWG.
  fix Initialize the `DAWG` object without arguments (e.g., `d = dawg_python.DAWG()`) and then call `.load('filepath.dawg')` to load an existing DAWG file. Creating DAWGs from data requires the original `dawg` library.

• ModuleNotFoundError: No module named 'dawg'

  cause This error likely occurs if you are following documentation that refers to the *original* `dawg` library (the C-extension version) for creating `.dawg` files, but you have only installed `dawg2-python`.
  fix If you intend to *create* `.dawg` files, you must install the `dawg` library (note the missing `_python`): `pip install dawg`. If you only intend to *read* existing `.dawg` files, ensure you are importing `dawg_python` correctly and loading an already saved file.

Warnings

• gotcha DAWG2-Python is a pure-Python library for *reading* existing DAWG files. It does not support creating, modifying, or saving DAWG structures. Attempting to use methods like `save()` or passing data to the constructor will result in errors.
  Fix: Use the original `dawg` library (which has C extensions) or `dawgdic` C++ tools to create and save `.dawg` files. Then, use `dawg2-python` for read-only access.
• breaking Starting from version 0.8.0, the minimal Python version required has been updated to 3.8.
  Fix: Ensure your Python environment is running Python 3.8 or a newer compatible version (less than 4.0).
• gotcha The `setup.py` build system was migrated to `poetry` in version 0.8.0. This primarily affects maintainers and those building from source, but it indicates a change in development tooling.
  Fix: No direct action required for end-users installing via `pip`. Developers should be aware of the shift to `poetry` for project management.

Install

• pip install dawg2-python Install stable version

Imports

• DAWG

  from dawg_python import DAWG

• BytesDAWG

  from dawg_python import BytesDAWG

• RecordDAWG

  from dawg_python import RecordDAWG

Quickstart

This quickstart demonstrates how to load and query a `.dawg` file using `dawg2-python`. Note that `dawg2-python` is a pure-Python *reader* and cannot create or save `.dawg` files. For creating the `.dawg` file in this example, the `dawg` (C-extension based) library is used. You would typically use existing `.dawg` files generated by `dawg` or `dawgdic`.

import os
import tempfile
import dawg # This is the *original* DAWG library, for creating the .dawg file
import dawg_python # This is dawg2-python, for reading the .dawg file

# Create a dummy .dawg file using the original 'dawg' library
# (dawg2-python is for reading, not creating/saving)
def create_dummy_dawg(filepath):
    words = ['apple', 'apricot', 'banana', 'bandana', 'cat']
    d = dawg.DAWG(words)
    d.save(filepath)

# Use a temporary file for demonstration
with tempfile.NamedTemporaryFile(suffix='.dawg', delete=False) as tmp_file:
    dawg_filepath = tmp_file.name

try:
    # 1. Create a DAWG file (requires the 'dawg' library installed)
    print(f"Creating a temporary DAWG file at {dawg_filepath}...")
    create_dummy_dawg(dawg_filepath)
    print("DAWG file created.")

    # 2. Load and query the DAWG file using dawg2-python
    print("Loading DAWG using dawg2-python...")
    reader = dawg_python.DAWG().load(dawg_filepath)
    print("DAWG loaded successfully.")

    # Querying words
    print(f"'apple' in DAWG: {'apple' in reader}")
    print(f"'banana' in DAWG: {'banana' in reader}")
    print(f"'orange' in DAWG: {'orange' in reader}")

    # Getting prefixes
    print(f"Prefixes for 'bandana': {list(reader.prefixes('bandana'))}")

    # Iterating through all words
    print("First 5 words:")
    for i, word in enumerate(reader.iterkeys()):
        if i >= 5: break
        print(f"- {word}")

finally:
    # Clean up the temporary file
    if os.path.exists(dawg_filepath):
        os.remove(dawg_filepath)
        print(f"Cleaned up temporary file: {dawg_filepath}")