DAWG-Python

Common errors

• FileNotFoundError: [Errno 2] No such file or directory: 'my_dictionary.dawg'

  cause The DAWG file you are trying to load does not exist at the specified path, or the path is incorrect.
  fix Double-check the file path. Ensure the `.dawg` file exists and is accessible from where your script is running. Use an absolute path if unsure.

• TypeError: 'str' object cannot be interpreted as an integer

  cause You are attempting to use `IntDAWG` with payloads that are not integers (e.g., strings, floats).
  fix Convert all payloads to integers before constructing the `IntDAWG`. For example: `IntDAWG([('word', int(value)) for word, value in data])`.

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

  cause You are attempting to modify a `DAWG` object after it has been created or loaded. DAWG structures are immutable.
  fix To 'update' a DAWG, you must build a completely new `DAWG` object from the desired set of words. There is no in-place modification.

Warnings

• gotcha The `dawg-python` library is primarily a *reader* for DAWG files. While it can build DAWGs from Python lists, for very large dictionaries, the C++ `dawgdic` library or the `DAWG-Python C extension` are recommended for efficient DAWG construction.
  Fix: For large-scale DAWG construction, consider using `dawgdic` or its Python C extension bindings, and then loading the resulting `.dawg` files with `dawg-python`.
• gotcha DAWG objects (both `DAWG` and `IntDAWG`) are immutable once created or loaded from a file. You cannot add, remove, or modify words/payloads in place.
  Fix: To modify a DAWG, you must build a new DAWG object from the desired set of words and payloads.
• gotcha When building a DAWG using `DAWG(iterable_of_words)` or `IntDAWG(iterable_of_tuples)`, the input iterable should be *sorted alphabetically* for optimal performance. If not sorted, the library will sort it internally, which can be slow for large inputs.
  Fix: Pre-sort your list of words/tuples before passing them to the DAWG constructor: `DAWG(sorted(my_words_list))`.
• gotcha `IntDAWG` is specifically designed for string keys with *integer* payloads. Passing non-integer values as payloads will result in a `TypeError`.
  Fix: Ensure that all payloads provided to `IntDAWG` are integers.

Install

• pip install dawg-python Install stable version

Imports

• DAWG
  wrong:

  from dawg_python import DAWG

  correct:

  from dawg import DAWG

  The PyPI package is 'dawg-python', but the main module to import from is 'dawg'.
• IntDAWG
  wrong:

  from dawg_python import IntDAWG

  correct:

  from dawg import IntDAWG

  The PyPI package is 'dawg-python', but the main module to import from is 'dawg'.

Quickstart

This quickstart demonstrates how to create a DAWG (for illustration), save it to a file, and then load and query it. It also shows basic usage of `IntDAWG` for words with associated integer payloads. The core use case is loading and querying, with file creation typically handled by other, more performant tools for large datasets.

import os
from dawg import DAWG, IntDAWG

# 1. Create a sample DAWG file (in a real scenario, this might come from dawgdic)
words_to_build = ['apple', 'apricot', 'banana', 'cat', 'dog']
# For large sets, words should be pre-sorted for performance.
temp_dawg = DAWG(words_to_build)
dawg_file_path = 'sample_data.dawg'
temp_dawg.save(dawg_file_path)

# 2. Load the DAWG from a file (primary use case)
loaded_dawg = DAWG().load(dawg_file_path)

# 3. Query the loaded DAWG
print(f"Is 'apple' in DAWG? {'apple' in loaded_dawg}")
print(f"Words starting with 'a': {list(loaded_dawg.keys('a'))}")
print(f"Longest prefix for 'apricot': {loaded_dawg.longest_prefix('apricot')}")

# Clean up the temporary file
os.remove(dawg_file_path)

# Example with IntDAWG for words with integer payloads
int_data = [('hello', 10), ('world', 20)]
int_dawg_obj = IntDAWG(int_data)
print(f"Value for 'world': {int_dawg_obj['world']}")