A trie structure supporting fuzzy string searches.
Project description
vtrie
Trie structure supporting approximate string matching (substitutions only) for Python (2.x and 3.x).
Installation
pip install vtrie
Features
It is similar to a dict() in general usage, and supports much of the dict() interface:
__len__(): number of items in the trie
__contains()__
__getitem()__
__setitem()__
__delitem()__
__sizeof()__: size of the trie in bytes
__repr()__: dict-like output, showing contents of the trie
has_key()
setdefault()
get()
pop()
popitem()
keys()
values()
items()
iterkeys()
itervalues()
iteritems()
Trie-specific methods:
num_nodes(): number of nodes
longest_prefix(k): find longest key matching the beginning of k, returning (key, value) pair as a 2-tuple. None is returned if no match.
suffixes(k): iterate over all (suffix, value) pairs as 2-tuples, that have k as a prefix.
neighbors(key = k, maxhd = n): iterate over all (Hamming distance, key, value) triples, as 3 tuples, where key and k differ by at least 1, but maximally n characters. Note, one can only search for neighbors of existing keys.
pairs(keylen = l, maxhd = n): iterate over ALL (Hamming distance, key1, value1, key2, value2) 5-tuples where key1 and key2 differ by at least 1, but maximally n characters. Note, pairs() returns a dirty iterator, meaning that nodes in the trie are modified while the iterator is running. An exception will be thrown when iterating with more than one dirty iterator.
Pickling
Usage
Create a trie:
>>> from vtrie import Trie >>> t = Trie()
Add strings to the trie. Currently, only ascii strings are supported:
>>> t[b"Hello"] = 123 >>> t[b"World"] = {"my": "dict"} >>> t[b"foo"] = None
Check if “Hello” is in the trie:
>>> b"Hello" in t True
Show all inserted strings sharing the same prefix:
>>> t[b"foo"] = 0 >>> t[b"foobar"] = 1 >>> t[b"fooo"] = 2 >>> t[b"hello"] = 3 >>> list(t.suffixes(b"fo")) [('o', 0), ('obar', 1), ('oo', 2)]
Search for keys that differ by less than a given number of substitutions from the provided key. The results are tuples with first the Hamming distance between the given key and the found key, then the found key and its value:
>>> t[b"hello world"] = 0 >>> t[b"*ello world"] = "a" >>> t[b"*ell* world"] = "b" >>> t[b"*ell* w*rld"] = "c" >>> t[b"hell* w*rl*"] = "d" >>> list(t.neighbors(b"hello world", maxhd = 1)) [(1, '*ello world', 'a')] >>> list(t.neighbors(b"hello world", maxhd = 2)) [(1, '*ello world', 'a'), (2, '*ell* world', 'b')] >>> print("\n".join(map(str,list(t.neighbors(b"hello world", 3))))) (3, 'hell* w*rl*', 'd') (1, '*ello world', 'a') (2, '*ell* world', 'b') (3, '*ell* w*rld', 'c')
Search for all keys of a certain length that are within a certain Hamming of each other. The results are tuples with first the Hamming distance between the found keys, then the first key and its value, and then the second key and its value:
>>> print("\n".join(map(str,list(t.pairs(keylen = 11, maxhd = 2))))) (1, 'hello world', 0, '*ello world', 'a') (2, 'hello world', 0, '*ell* world', 'b') (2, 'hell* w*rl*', 'd', '*ell* w*rld', 'c') (1, '*ello world', 'a', '*ell* world', 'b') (2, '*ello world', 'a', '*ell* w*rld', 'c') (1, '*ell* world', 'b', '*ell* w*rld', 'c')
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.