trie-again 0.2.3

Trie data structure for prefix search and text completion

Trie Again: Python Trie implementation optimized for T9 completion

Installation

pip install trie-again

# in some cases you might want to adjust your compiler (the same is applicable for `poetry install`)
CC='gcc' CFLAGS='-march=native' pip install trie-again

Usage

# create an instance
from trie_again import Trie

# if you want to use faster version
from trie_again import CyTrie

trie = Trie()

# insert a single word
trie.insert('boy')

# insert a list of words
trie.extend(['bondage', 'coverage'])

# insert a list of words with multipliers (useful when parsing json)
data = {
    'bondage': 10,
    'coverage': 20,
}
trie.extend(data.keys(), data.values())

# check key in trie
print('bondage' in trie)
# True

# list all keys, sorted by usage
print(list(trie))
# ['coverage', 'bondage', 'boy']

# complete simple, sorted by usage
print(list(trie.complete('b')))
# ['bondage', 'boy']

T9 Like completion

# complete with t9 like approach
print(list(trie.complete(['bc', 'o', 'vn'])))
# ['coverage', 'bondage']

How it works?

b o y
b o n d a g e
c o v e r a g e
^ ^ ^
1 2 3

We use these groups to complete: bc, o, vn. It means that at position 1 it the letter may be b or c, at position 2 only o, at position 3 v or n.

Test

# test behavior
poetry run pytest

# test performance
poetry run pytest --benchmark

Dev

# very start (adjust compiler options if needed)
poetry install

# install pre commit
poetry run pre-commit install

# lint
poetry run black .
poetry run flake8 .
poetry run mypy .

# coverage
poetry run coverage run -m pytest && poetry run coverage report -m

# build package: limiting to sdist to compile it on install
poetry build -f sdist

Read an article about the trie, friends!

https://blagovdaryu.hashnode.dev/ok-lets-trie-t9-in-python