#!/usr/bin/env python3
+# © Copyright 2021-2022, Scott Gasch
+
+"""A fast word unscrambler library."""
+
import logging
from typing import Dict, Mapping, Optional
logger = logging.getLogger(__name__)
letters_bits = 32
-letters_mask = 2 ** letters_bits - 1
+letters_mask = 2**letters_bits - 1
fprint_bits = 52
-fprint_mask = (2 ** fprint_bits - 1) << letters_bits
+fprint_mask = (2**fprint_bits - 1) << letters_bits
fprint_feature_bit = {
'e': 0,
lookup methods support a "fuzzy match" argument that can be set to
request similar words that do not match exactly in addition to any
exact matches.
-
"""
def __init__(self, indexfile: Optional[str] = None):
+ """
+ Constructs an unscrambler.
+
+ Args:
+ indexfile: overrides the default indexfile location if provided
+ """
+
# Cached index per instance.
self.sigs = []
self.words = []
- filename = self.get_indexfile(indexfile)
+ filename = Unscrambler.get_indexfile(indexfile)
with open(filename, 'r') as rf:
lines = rf.readlines()
for line in lines:
self.sigs.append(isig)
self.words.append(word)
- def get_indexfile(self, indexfile: Optional[str]) -> str:
+ @staticmethod
+ def get_indexfile(indexfile: Optional[str]) -> str:
+ """Returns the current indexfile location."""
if indexfile is None:
if 'unscrambler_default_indexfile' in config.config:
- indexfile = config.config['unscramble_indexfile']
+ indexfile = config.config['unscrambler_default_indexfile']
else:
indexfile = "/usr/share/dict/sparse_index"
else:
the word and their frequencies. We try to cluster "similar"
words close to each other in the signature space.
+ Args:
+ word: the word to compute a signature for
+
+ Returns:
+ The word's signature.
+
>>> train = Unscrambler.compute_word_sig('train')
>>> train
23178969883741
@staticmethod
def repopulate(
- lsigs: Dict[str, int],
dictfile: str = '/usr/share/dict/words',
indexfile: str = '/usr/share/dict/sparse_index',
) -> None:
- """Before calling this method, change letter_sigs from the default above
- unless you want to populate the same exact files.
+ """
+ Repopulates the indexfile.
+
+ .. warning::
+ Before calling this method, change letter_sigs from the
+ default above unless you want to populate the same exact
+ files.
"""
words_by_sigs: Dict[int, str] = {}
seen = set()
"""Looks up a potentially scrambled word optionally including near
"fuzzy" matches.
+ Args:
+ word: the word to lookup
+ window_size: the number of nearby fuzzy matches to return
+
+ Returns:
+ A dict of word -> bool containing unscrambled words with (close
+ to or precisely) the same letters as the input word. The bool
+ values in this dict indicate whether the key word is an exact
+ or near match. The count of entries in this dict is controlled
+ by the window_size param.
+
>>> u = Unscrambler()
>>> u.lookup('eanycleocipd', window_size=0)
{'encyclopedia': True}
-
"""
sig = Unscrambler.compute_word_sig(word)
return self.lookup_by_sig(sig, window_size=window_size)
a previous call to Unscrambler.compute_word_sig. Optionally returns
near "fuzzy" matches.
+ Args:
+ sig: the signature of the word to lookup (see :meth:`compute_word_sig`
+ to generate these signatures).
+ window_size: the number of nearby fuzzy matches to return
+
+ Returns:
+ A dict of word -> bool containing unscrambled words with (close
+ to or precisely) the same letters as the input word. The bool
+ values in this dict indicate whether the key word is an exact
+ or near match. The count of entries in this dict is controlled
+ by the window_size param.
+
>>> sig = Unscrambler.compute_word_sig('sunepsapetuargiarin')
>>> sig
18491949645300288339
>>> u = Unscrambler()
>>> u.lookup_by_sig(sig)
{'pupigerous': False, 'pupigenous': False, 'unpurposing': False, 'superpurgation': False, 'unsupporting': False, 'superseptuaginarian': True, 'purpurogallin': False, 'scuppaug': False, 'purpurigenous': False, 'purpurogenous': False, 'proppage': False}
-
"""
ret = {}
(_, location) = list_utils.binary_search(self.sigs, sig)