nim-by-example/content/bitsets.md
2014-12-11 17:04:49 -05:00

24 lines
1.6 KiB
Markdown

---
title: Bitsets
---
# Bitsets
Nim comes with a built in way to build a set of ordinal types. In order for a type to be usable in a bitset, it must be an ordinal and <<\texttt{high(T)} < 2^{16}>>. For sets of non-ordinal types, see the [sets module](http://nim-lang.org/sets.html), which contains hashsets.
However, best practice is to keep bitset size significantly smaller since each possible element in the set consumes one bit, therefore a bitset of <<2^{16}>> elements will consume 64KiB.
Bitsets have all the useful operations of mathematical sets:
|Operator | Description | Example Code |
|-------------|-------------------------------|----------------------------------------------|
| `a in B` | is a an element of B? | `'d' in {'a'..'z'}` |
| `a notin B` | is a not an element of B? | `40 notin {2..20} ` |
| `A + B` | union of A with B | `{'a'..'m'} + {'n'..'z'} == {'a'..'z'}` |
| `A - B` | relative complement of A in B | `{'a'..'z'} - {'b'..'d'} == {'a', 'e'..'z'}` |
| `A + b` | add element b to set A | `{'b'..'z'} + 'a' == {'a'..'z'}` |
| `A - b` | remove element b from set A | `{'a'..'z'} - 'a' == {'b'..'z'}` |
| `A * B` | intersection of A with B | `{'a'..'m'} * {'c'..'z} == {'c'..'m'}` |
| `A <= B` | is A a subset of B? | `{'a'..'c'} <= {'a'..'z'}` |
| `A < B` | is A a strict subset of B? | `{'b'..'c'} < {'a'..'z'}` |