Урок 3

#include <algorithm>
#include <iostream>
#include <set>
#include <string>
#include <utility>
#include <vector>
#include <map>
#include <cmath>
#include <cassert>
#include <optional>

using namespace std;

const int MAX_RESULT_DOCUMENT_COUNT = 5;

string ReadLine() {
    string s;
    getline(cin, s);
    return s;
}

int ReadLineWithNumber() {
    int result = 0;
    cin >> result;
    ReadLine();
    return result;
}

vector<string> SplitIntoWords(const string& text) {
    vector<string> words;
    string word;
    for (const char c : text) {
        if (c == ' ') {
            if (!word.empty()) {
                words.push_back(word);
                word.clear();
            }
        }
        else {
            word += c;
        }
    }
    if (!word.empty()) {
        words.push_back(word);
    }
    return words;
}

struct Document {
    Document() = default;

    Document(int id_, double relevance_, int rating_)
        : id(id_), relevance(relevance_), rating(rating_)
    {}

    int id = 0;
    double relevance = 0.0;
    int rating = 0;
};

enum class DocumentStatus {
    ACTUAL,
    IRRELEVANT,
    BANNED,
    REMOVED
};

class SearchServer {
public:
    // Defines an invalid document id
    // You can refer to this constant as SearchServer::INVALID_DOCUMENT_ID
    inline static constexpr int INVALID_DOCUMENT_ID = -1;

    SearchServer() = default;

    explicit SearchServer(const string& stop_words) {
        for (const string& word : SplitIntoWords(stop_words)) {
            stop_words_.insert(word);
        }
    }

    template <typename StringCollection>
    explicit SearchServer(const StringCollection& stop_words) {
        string stop_words_text;
        for (const auto& text : stop_words) {
            stop_words_text += " "s;
            stop_words_text += text;
        }
        for (const auto& word : SplitIntoWords(stop_words_text)) {
            stop_words_.insert(word);
        }
    }

    [[nodiscard]] bool AddDocument(int document_id, const string& document, DocumentStatus status, const vector<int>& ratings) {
        if (document_id < 0) {
            return false;
        }
        if (documents_.find(document_id) != documents_.end()) {
            return false;
        }

        const vector<string> words = SplitIntoWordsNoStop(document);
        for (auto& word : words) {
            if (!IsValidWord(word)) {
                return false;
            }
        }
        for (auto& word : words) {
            word_to_document_freqs_[word][document_id] += 1.0 / words.size();
        }
        documents_.emplace(document_id, DocumentData { ComputeAverageRating(ratings), status });
        documents_index_.push_back(document_id);

        return true;
    }

    int GetDocumentCount() const {
        return static_cast<int>(documents_.size());
    }

    template <typename DocumentPredicate>
    optional<vector<Document>> FindTopDocuments(const string& raw_query, DocumentPredicate document_predicate) const {
        Query query;
        if (!ParseQuery(raw_query, query)) {
            return nullopt;
        }

        auto matched_documents = FindAllDocuments(query, document_predicate);

        sort(matched_documents.begin(), matched_documents.end(),
            [](const Document& lhs, const Document& rhs) {
                const double EPSILON = 1e-6;
                if (abs(lhs.relevance - rhs.relevance) < EPSILON) {
                    return lhs.rating > rhs.rating;
                }
                else {
                    return lhs.relevance > rhs.relevance;
                }
            });
        if (matched_documents.size() > MAX_RESULT_DOCUMENT_COUNT) {
            matched_documents.resize(MAX_RESULT_DOCUMENT_COUNT);
        }

        return matched_documents;
    }

    optional<vector<Document>> FindTopDocuments(const string& raw_query, DocumentStatus status) const {
        return FindTopDocuments(raw_query,
            [&status](int document_id, DocumentStatus new_status, int rating) {
                return new_status == status;
            });
    }

    optional<vector<Document>> FindTopDocuments(const string& raw_query) const {
        return FindTopDocuments(raw_query, DocumentStatus::ACTUAL);
    }

    optional<tuple<vector<string>, DocumentStatus>> MatchDocument(const string& raw_query, int document_id) const {
        Query query;
        if (!ParseQuery(raw_query, query)) {
            return nullopt;
        }

        vector<string> matched_words;
        for (const string& word : query.plus_words) {
            if (word_to_document_freqs_.count(word) == 0) {
                continue;
            }
            if (word_to_document_freqs_.at(word).count(document_id)) {
                matched_words.push_back(word);
            }
        }
        for (const string& word : query.minus_words) {
            if (word_to_document_freqs_.count(word) == 0) {
                continue;
            }
            if (word_to_document_freqs_.at(word).count(document_id)) {
                matched_words.clear();
                break;
            }
        }

        return { { matched_words, documents_.at(document_id).status } };
    }

    int GetDocumentId(int index) const {
        if ((index >= 0) && (index < static_cast<int>(documents_.size()))) {
            return documents_index_[index];
        }
        return SearchServer::INVALID_DOCUMENT_ID;
    }

private:
    struct Query {
        set<string> plus_words;
        set<string> minus_words;
    };

    struct DocumentData {
        int rating;
        DocumentStatus status;
    };

    map<string, map<int, double>> word_to_document_freqs_;
    map<int, DocumentData> documents_;
    set<string> stop_words_;
    vector<int> documents_index_;

    bool IsStopWord(const string& word) const {
        return stop_words_.count(word) > 0;
    }

    static bool IsValidWord(const string& word) {
        // A valid word must not contain special characters
        return none_of(word.begin(), word.end(), [](char c) {
            return c >= '\0' && c < ' ';
            });
    }

    vector<string> SplitIntoWordsNoStop(const string& text) const {
        vector<string> words;
        for (const string& word : SplitIntoWords(text)) {
            if (!IsStopWord(word)) {
                words.push_back(word);
            }
        }
        return words;
    }

    static int ComputeAverageRating(const vector<int>& ratings) {
        if (ratings.empty()) {
            return 0;
        }
        int rating_sum = 0;
        for (const auto rating : ratings) {
            rating_sum += rating;
        }
        return rating_sum / static_cast<int>(ratings.size());
    }

    struct QueryWord {
        string data;
        bool is_minus;
        bool is_stop;
    };

    [[nodiscard]] bool ParseQueryWord(string text, QueryWord& result) const {
        result = {};

        if (text.empty()) {
            return false;
        }
        bool is_minus = false;
        if (text[0] == '-') {
            is_minus = true;
            text = text.substr(1);
        }
        if (text.empty() || text[0] == '-' || !IsValidWord(text)) {
            return false;
        }

        result = QueryWord{ text, is_minus, IsStopWord(text) };
        return true;
    }

    [[nodiscard]] bool ParseQuery(const string& text, Query& result) const {
        result = {};
        for (const string& word : SplitIntoWords(text)) {
            QueryWord query_word;
            if (!ParseQueryWord(word, query_word)) {
                return false;
            }
            if (!query_word.is_stop) {
                if (query_word.is_minus) {
                    result.minus_words.insert(query_word.data);
                }
                else {
                    result.plus_words.insert(query_word.data);
                }
            }
        }
        return true;
    }

    template<typename DocumentPredicate>
    vector<Document> FindAllDocuments(const Query& query, DocumentPredicate predicate) const {
        map<int, double> document_to_relevance;
        for (const string& word : query.plus_words) {
            if (word_to_document_freqs_.count(word) == 0) {
                continue;
            }
            for (const auto& [document_id, term_freq] : word_to_document_freqs_.at(word)) {
                const DocumentData documents_data = documents_.at(document_id);
                if (predicate(document_id, documents_data.status, documents_data.rating)) {
                    const double IDF = log(GetDocumentCount() * 1.0 / word_to_document_freqs_.at(word).size());
                    document_to_relevance[document_id] += IDF * term_freq;
                }
            }
        }
        for (const string& word : query.minus_words) {
            if (word_to_document_freqs_.count(word) == 0) {
                continue;
            }
            for (const auto& [document_id, term_freq] : word_to_document_freqs_.at(word)) {
                document_to_relevance.erase(document_id);
            }
        }

        vector<Document> matched_documents;
        for (const auto [document_id, relevance] : document_to_relevance) {
            matched_documents.push_back({ document_id, relevance, documents_.at(document_id).rating });
        }
        return matched_documents;
    }
};

void PrintDocument(const Document& document) {
    cout << "{ "s
        << "document_id = "s << document.id << ", "s
        << "relevance = "s << document.relevance << ", "s
        << "rating = "s << document.rating
        << " }"s << endl;
}

int main() {
    setlocale(LC_ALL, "Russian");

    SearchServer search_server("и в на"s);

    // Явно игнорируем результат метода AddDocument, чтобы избежать предупреждения
    // о неиспользуемом результате его вызова
    (void)search_server.AddDocument(1, "пушистый кот пушистый хвост"s, DocumentStatus::ACTUAL, { 7, 2, 7 });

    if (!search_server.AddDocument(1, "пушистый пёс и модный ошейник"s, DocumentStatus::ACTUAL, { 1, 2 })) {
        cout << "Документ не был добавлен, так его id совпадает с уже имеющимся"s << endl;
    }

    if (!search_server.AddDocument(-1, "пушистый пёс и модный ошейник"s, DocumentStatus::ACTUAL, { 1, 2 })) {
        cout << "Документ не был добавлен, так его id отрицательный"s << endl;
    }

    if (!search_server.AddDocument(3, "большой пёс скво\x12рец"s, DocumentStatus::ACTUAL, { 1, 3, 2 })) {
        cout << "Документ не был добавлен, так как содержит спецсимволы"s << endl;
    }

    if (const auto documents = search_server.FindTopDocuments("--пушистый"s)) {
        for (const Document& document : *documents) {
            PrintDocument(document);
        }
    }
    else {
        cout << "Ошибка в поисковом запросе"s << endl;
    }
}