complexity for searching

[flackbash]

The Readme suggests that this algorithm has a O(1) time complexity for searching.
However, every other source I read so far about time complexity of tries claims that the complexity is O(M) for searching where M is the maximum length of all inserted words. Which also does make more sense to me intuitively because in my imagination, the algorithm works like this: for every letter in the prefix I enter, the algorithm takes the corresponding path in the tree. This means it has to follow at least the number of letters in my prefix and at most M edges.
Can you explain to me where (or if) I'm wrong here?

[thep]

Surely you are not wrong. The intention of O(1) statement is to claim that it's not dependent on the database size, just like how hash table does similarly on non-colliding cases, despite the fact that it still takes O(m) on hash function evaluation.

Anyway, I agree to add some more precision to the description. Let me revise it like this:

Before:

Trie is a kind of digital search tree, an efficient indexing method with O(1) time complexity for searching.

After:

Trie is a kind of digital search tree, an efficient indexing method in which search time is independent of database size. It only takes O(m) search time, where m is the length of the search string.

[flackbash]

Thanks a lot for the clarification!