|
| 1 | +// Compile Time Evaluated Boyer Moore Horspool String Search Algorithm |
| 2 | +// Author: David Kanekanian |
| 3 | +// |
| 4 | +// Requires C++17. |
| 5 | + |
| 6 | +#include <string_view> // std::string_view |
| 7 | +#include <optional> // std::optional |
| 8 | +#include <array> // std::array |
| 9 | +#include <iostream> // std::cout |
| 10 | + |
| 11 | + |
| 12 | +struct SearchPattern { |
| 13 | + constexpr static std::size_t num_chars{ 256 }; |
| 14 | + const std::string_view needle; |
| 15 | + std::array<char, num_chars> bad_char_table{}; |
| 16 | + |
| 17 | + constexpr SearchPattern(const std::string_view needle) noexcept |
| 18 | + : needle(needle) { |
| 19 | + // compute bad char table |
| 20 | + |
| 21 | + // Rule 1 |
| 22 | + // Characters not in the needle jump the length of the needle. |
| 23 | + for (std::size_t i = 0; i < num_chars; i++) { |
| 24 | + bad_char_table[i] = needle.length(); |
| 25 | + } |
| 26 | + |
| 27 | + // Rule 2 |
| 28 | + // Characters in the needle except last character jump: |
| 29 | + // length of needle - last position of character - 1 |
| 30 | + for (std::size_t i = 0; i < needle.length(); i++) { |
| 31 | + bad_char_table[needle[i]] = needle.length() - i - 1; |
| 32 | + } |
| 33 | + } |
| 34 | +}; |
| 35 | + |
| 36 | +/** Returns index of first occurrance of pattern in haystack. */ |
| 37 | +[[nodiscard]] constexpr std::optional<std::size_t> boyer_moore_horspool ( |
| 38 | + const SearchPattern& pattern, |
| 39 | + const std::string_view haystack) noexcept { |
| 40 | + for (std::size_t h = 0; h < haystack.length() - pattern.needle.length() + 1;) { |
| 41 | + // Search this placement of the needle. |
| 42 | + bool found = true; |
| 43 | + for (std::size_t n = pattern.needle.length() - 1; ; n--) { |
| 44 | + // Check the needle in reverse. |
| 45 | + if (haystack[h + n] != pattern.needle[n]) { |
| 46 | + // Bad character found. |
| 47 | + found = false; |
| 48 | + // Jump based on the bad character table. |
| 49 | + h += pattern.bad_char_table[haystack[h + n]]; |
| 50 | + break; |
| 51 | + } |
| 52 | + if (n == 0) { |
| 53 | + break; |
| 54 | + } |
| 55 | + } |
| 56 | + if (found) { |
| 57 | + return h; |
| 58 | + } |
| 59 | + } |
| 60 | + |
| 61 | + // Not found in the haystack. |
| 62 | + return std::nullopt; |
| 63 | +} |
| 64 | + |
| 65 | +int main() { |
| 66 | + constexpr SearchPattern needle{ "abc" }; |
| 67 | + constexpr auto haystack = "aa abc ddef"; |
| 68 | + constexpr auto index = boyer_moore_horspool(needle, haystack); |
| 69 | + static_assert(index == 3, "index is supposed to be 3!"); |
| 70 | + |
| 71 | + if (index) { |
| 72 | + std::cout << "found at index " << index.value() << '\n'; |
| 73 | + } else { |
| 74 | + std::cout << "not found" << '\n'; |
| 75 | + } |
| 76 | +} |
0 commit comments