Add string search utility

Author: Andersama

include/ctre/atoms.hpp

@@ -19,6 +19,7 @@ struct any { };
 
 // actual AST of regexp
 template <auto... Str> struct string { };
+template <auto... Str> struct string_search { };
 template <typename... Opts> struct select { };
 template <typename... Content> struct sequence { };
 struct empty { };

include/ctre/evaluation.hpp

@@ -115,10 +115,17 @@ template <typename CharT, typename Iterator, typename EndIterator> constexpr CTR
 }
 
 template <auto... String, size_t... Idx, typename Iterator, typename EndIterator> constexpr CTRE_FORCE_INLINE string_match_result<Iterator> evaluate_match_string(Iterator current, [[maybe_unused]] const EndIterator end, std::index_sequence<Idx...>) noexcept {
-
-	bool same = (compare_character(String, current, end) && ... && true);
-
-	return {current, same};
+	if constexpr (!std::is_same_v<Iterator, utf8_iterator> && is_random_accessible(typename std::iterator_traits<Iterator>::iterator_category{})) {
+		bool same = (::std::distance(current, end) >= sizeof...(String)) && ((String == *(current + Idx)) & ...);
+		if (same) {
+			return {current+=sizeof...(String), same};
+		} else {
+			return {current, same};
+		}
+	} else {
+		bool same = (compare_character(String, current, end) && ... && true);
+		return { current, same };
+	}
 }
 
 template <typename R, typename Iterator, typename EndIterator, auto... String, typename... Tail> 
@@ -132,6 +139,107 @@ constexpr CTRE_FORCE_INLINE R evaluate(const Iterator begin, Iterator current, c
 	return evaluate(begin, result.position, end, consumed_something(f, sizeof...(String) > 0), captures, ctll::list<Tail...>());
 }
 
+template<typename Ty>
+constexpr bool is_prefix(Ty* word, size_t wordlen, ptrdiff_t pos) {
+	ptrdiff_t suffixlen = wordlen - pos;
+	for (int i = 0; i < suffixlen; i++) {
+		if (word[i] != word[pos + i]) {
+			return false;
+		}
+	}
+	return true;
+}
+
+template<typename Ty>
+constexpr size_t suffix_length(Ty* word, size_t wordlen, ptrdiff_t pos) {
+	size_t i = 0;
+	// increment suffix length i to the first mismatch or beginning of the word
+	for (; (word[pos - i] == word[wordlen - 1 - i]) && (i < pos); i++);
+	return i;
+}
+//MSVC workaround, array operator[] blows up in face if constexpr, use pointers instead
+template<typename Ty, auto... String>
+constexpr auto make_delta_2(string<String...>) {
+	std::array<Ty, sizeof...(String)> chars{ String... };
+	std::array<ptrdiff_t, sizeof...(String)> table;
+	constexpr size_t patlen = sizeof...(String);
+	size_t p = 0;
+	size_t last_prefix_index = patlen - 1;
+
+	for (p = patlen - 1; p < patlen; p--) {
+		if (is_prefix(chars.data(), patlen, p + 1)) {
+			last_prefix_index = p + 1;
+		}
+		table.data()[p] = last_prefix_index + (patlen - 1 - p);
+	}
+
+	for (p = 0; p < patlen - 1; p++) {
+		size_t slen = suffix_length(chars.data(), patlen, p);
+		if (chars.data()[p - slen] != chars.data()[patlen - 1 - slen]) {
+			table.data()[patlen - 1 - slen] = patlen - 1 - p + slen;
+		}
+	}
+
+	return table;
+}
+
+template <typename Iterator, typename EndIterator, auto... String>
+constexpr CTRE_FORCE_INLINE string_match_result<Iterator> evaluate_search_string(Iterator current, const EndIterator end, string<String...>) {
+	if constexpr (sizeof...(String) > 2 && !std::is_same_v<Iterator, utf8_iterator> && is_random_accessible(typename std::iterator_traits<Iterator>::iterator_category{})) {
+		constexpr std::array<typename ::std::iterator_traits<Iterator>::value_type, sizeof...(String)> chars{ String... };
+		constexpr std::array<ptrdiff_t, sizeof...(String)> delta_2 = make_delta_2<typename ::std::iterator_traits<Iterator>::value_type>(string<String...>());
+		
+		size_t str_size = std::distance(current, end);
+		if (str_size < sizeof...(String)) { //quick exit no way to match
+			return { current, false };
+		}
+
+		size_t i = sizeof...(String) - 1; //index over to the starting location
+		for (; i < str_size;) {
+			size_t j = sizeof...(String) - 1;
+			size_t m = i + 1;
+			for (; *(current + i) == *(chars.data() + j); --i, --j) { //match string in reverse
+				if (j == 0) {
+					return { current + m, true };
+				}
+			}
+			size_t shift = enumeration<String...>::match_char(*(current + i)) ? static_cast<size_t>(*(delta_2.data() + j)) : sizeof...(String);
+			i += shift;
+		}
+
+		return { current + str_size, false };
+	} else if (sizeof...(String)) {
+		//fallback to plain string matching
+		constexpr std::array<typename ::std::iterator_traits<Iterator>::value_type, sizeof...(String)> chars{ String... };
+		constexpr typename ::std::iterator_traits<Iterator>::value_type first_char = chars.data()[0];
+		while (current != end) {
+			while (current != end && *current != first_char) {
+				current++;
+			}
+			auto result = evaluate_match_string<String...>(current, end, std::make_index_sequence<sizeof...(String)>());
+			if (result.match) {
+				return result;
+			} else {
+				++current;
+			}
+		}
+		return { current, false };
+	} else {
+		return { current, true };
+	}
+}
+
+template <typename R, typename Iterator, typename EndIterator, auto... String, typename... Tail>
+constexpr CTRE_FORCE_INLINE R evaluate(const Iterator begin, Iterator current, const EndIterator end, [[maybe_unused]] const flags& f, R captures, ctll::list<string_search<String...>, Tail...>) noexcept {
+	auto result = evaluate_search_string(current, end, string<String...>());
+	
+	if (!result.matched) {
+		return not_matched;
+	}
+
+	return evaluate(begin, std::advance(result.position, sizeof...(String)), end, consumed_something(f, sizeof...(String) > 0), captures, ctll::list<Tail...>());
+}
+
 // matching select in patterns
 template <typename R, typename Iterator, typename EndIterator, typename HeadOptions, typename... TailOptions, typename... Tail> 
 constexpr CTRE_FORCE_INLINE R evaluate(const Iterator begin, Iterator current, const EndIterator end, const flags & f, R captures, ctll::list<select<HeadOptions, TailOptions...>, Tail...>) noexcept {