Searching

With one algorithm under our belt, let’s learn two more: searching.

A search algorithm takes a needle and a haystack: the needle is something to look for, and the haystack is the thing we’re looking in. For now, we’ll focus on searching for values in lists, i.e., we’ll ask questions of the form “is value x in the list l?”. Before reading on… what else might we want to search for?

What should a search algorithm return? Sometimes, it’s enough to return a boolean, saying, “yes, it’s in there”. More generally, we want to know where the value is, returning an index or other way of finding the value; if the value can’t be found, we’ll signal the problem using a ValueError or a sentinel value that indicates “not found”.

Alternative searches

We might also want to search for values that meet some predicate (i.e., if f takes values to booleans, we want to find a value x in l such that f(x) is True), or we might want to collect every value in some range of values (e.g., numbers between 0 and 15). It’s not hard to adapt a search to work with a predicate; range queries often depend on particular data structures.