import time
import random
import csv
import os
import sys
import matplotlib.pyplot as plt

sys.setrecursionlimit(20000)

BASE_PATH = r"C:\Users\andre\2026-rff_mp\smirnovad\lab1"
DOCS_PATH = os.path.join(BASE_PATH, "docs")
DATA_PATH = os.path.join(DOCS_PATH, "data")

for p in [DOCS_PATH, DATA_PATH]:
    if not os.path.exists(p): 
        os.makedirs(p)

def ll_insert(head, name, phone):
    return {'name': name, 'phone': phone, 'next': head}

def ll_find(head, name):
    curr = head
    while curr:
        if curr['name'] == name: return curr['phone']
        curr = curr['next']
    return None

def ll_delete(head, name):
    if not head: return None
    if head['name'] == name: return head['next']
    curr = head
    while curr['next']:
        if curr['next']['name'] == name:
            curr['next'] = curr['next']['next']
            return head
        curr = curr['next']
    return head

def ll_list_all(head):
    res = []
    curr = head
    while curr:
        res.append((curr['name'], curr['phone']))
        curr = curr['next']
    return sorted(res)

def ht_insert(buckets, name, phone):
    idx = hash(name) % len(buckets)
    buckets[idx] = ll_insert(buckets[idx], name, phone)

def ht_find(buckets, name):
    idx = hash(name) % len(buckets)
    return ll_find(buckets[idx], name)

def ht_delete(buckets, name):
    idx = hash(name) % len(buckets)
    buckets[idx] = ll_delete(buckets[idx], name)

def ht_list_all(buckets):
    all_recs = []
    for b in buckets:
        curr = b
        while curr:
            all_recs.append((curr['name'], curr['phone']))
            curr = curr['next']
    return sorted(all_recs)

def bst_insert(root, name, phone):
    if not root:
        return {'name': name, 'phone': phone, 'left': None, 'right': None}
    if name < root['name']:
        root['left'] = bst_insert(root['left'], name, phone)
    elif name > root['name']:
        root['right'] = bst_insert(root['right'], name, phone)
    else:
        root['phone'] = phone
    return root

def bst_find(root, name):
    if not root: return None
    if root['name'] == name: return root['phone']
    if name < root['name']: return bst_find(root['left'], name)
    return bst_find(root['right'], name)

def bst_delete(root, name):
    if not root: return None
    if name < root['name']:
        root['left'] = bst_delete(root['left'], name)
    elif name > root['name']:
        root['right'] = bst_delete(root['right'], name)
    else:
        if not root['left']: return root['right']
        if not root['right']: return root['left']
        temp = root['right']
        while temp['left']: temp = temp['left']
        root['name'], root['phone'] = temp['name'], temp['phone']
        root['right'] = bst_delete(root['right'], temp['name'])
    return root

def bst_list_all(root):
    res = []
    def _inorder(node):
        if node:
            _inorder(node['left'])
            res.append((node['name'], node['phone']))
            _inorder(node['right'])
    _inorder(root)
    return res

all_results_csv = []
summary_for_report = []

def run_experiment(struct_type, mode, data):
    print(f"Processing: {struct_type} ({mode})")
    ins_times, find_times, del_times = [], [], []

    for i in range(5):
        container = [None]*1000 if struct_type == "HashTable" else None
        
        start = time.perf_counter()
        for n, p in data:
            if struct_type == "LinkedList": container = ll_insert(container, n, p)
            elif struct_type == "HashTable": ht_insert(container, n, p)
            elif struct_type == "BST": container = bst_insert(container, n, p)
        ins_times.append(time.perf_counter() - start)

        search_list = [d[0] for d in random.sample(data, 100)] + [f"None_{j}" for j in range(10)]
        start = time.perf_counter()
        for s_name in search_list:
            if struct_type == "LinkedList": ll_find(container, s_name)
            elif struct_type == "HashTable": ht_find(container, s_name)
            elif struct_type == "BST": bst_find(container, s_name)
        find_times.append(time.perf_counter() - start)

        del_list = [d[0] for d in random.sample(data, 50)]
        start = time.perf_counter()
        for d_name in del_list:
            if struct_type == "LinkedList": container = ll_delete(container, d_name)
            elif struct_type == "HashTable": ht_delete(container, d_name)
            elif struct_type == "BST": container = bst_delete(container, d_name)
        del_times.append(time.perf_counter() - start)

        all_results_csv.append([struct_type, mode, f"Run {i+1}", ins_times[-1], find_times[-1], del_times[-1]])

    avg_ins = sum(ins_times) / 5
    avg_find = sum(find_times) / 5
    avg_del = sum(del_times) / 5
    
    all_results_csv.append([struct_type, mode, "AVERAGE", avg_ins, avg_find, avg_del])
    summary_for_report.append({"name": struct_type, "mode": mode, "ins": avg_ins, "find": avg_find, "del": avg_del})

N = 10000
records_raw = [(f"User_{i:05d}", f"8-900-{random.randint(100, 999)}") for i in range(N)]
records_shuffled = records_raw[:]
random.shuffle(records_shuffled)
records_sorted = sorted(records_raw)

for m_name, d_set in [("случайный", records_shuffled), ("сортированный", records_sorted)]:
    for s_type in ["LinkedList", "HashTable", "BST"]:
        run_experiment(s_type, m_name, d_set)

with open(os.path.join(DATA_PATH, "results.csv"), "w", newline="", encoding="utf-8") as f:
    writer = csv.writer(f)
    writer.writerow(["Структура", "Режим", "Итерация", "Вставка", "Поиск", "Удаление"])
    writer.writerows(all_results_csv)

def create_plots():
    labels = ["insert", "find", "delete"]
    structs = ["LinkedList", "HashTable", "BST"]
    colors = ['#5dade2', '#e67e22', '#58d68d']

    fig1, axs = plt.subplots(1, 3, figsize=(18, 6))
    fig1.suptitle("Влияние порядка данных на время операций", fontsize=16, fontweight='bold')

    for i, s_name in enumerate(structs):
        rand_data = next(r for r in summary_for_report if r['name'] == s_name and r['mode'] == "случайный")
        sort_data = next(r for r in summary_for_report if r['name'] == s_name and r['mode'] == "сортированный")
        
        x = [0, 1, 2]
        width = 0.35
        axs[i].bar([p - width/2 for p in x], [rand_data['ins'], rand_data['find'], rand_data['del']], width, label='случайный', color=colors[0])
        axs[i].bar([p + width/2 for p in x], [sort_data['ins'], sort_data['find'], sort_data['del']], width, label='сортированный', color='#e74c3c', alpha=0.8)
        
        axs[i].set_title(s_name, fontweight='bold')
        axs[i].set_xticks(x)
        axs[i].set_xticklabels(labels)
        axs[i].set_ylabel("Время (с)")
        axs[i].legend()
        axs[i].grid(axis='y', linestyle='--', alpha=0.3)

    plt.tight_layout(rect=[0, 0.03, 1, 0.95])
    plt.savefig(os.path.join(DATA_PATH, "order_impact.png"))

    fig2, axs2 = plt.subplots(1, 3, figsize=(18, 6))
    fig2.suptitle(f"Сравнение структур данных (N={N})", fontsize=16, fontweight='bold')
    
    op_keys = ['ins', 'find', 'del']
    op_names = ['insert', 'find', 'delete']

    for i, op in enumerate(op_keys):
        plot_labels = []
        plot_values = []
        plot_colors = []
        
        for r in summary_for_report:
            plot_labels.append(f"{r['name']}\n({r['mode'][:4]})")
            plot_values.append(r[op])
            if r['name'] == "LinkedList": plot_colors.append(colors[0])
            elif r['name'] == "HashTable": plot_colors.append(colors[1])
            else: plot_colors.append(colors[2])
            
        bars = axs2[i].bar(plot_labels, plot_values, color=plot_colors)
        axs2[i].set_title(f"Операция: {op_names[i]}", fontweight='bold')
        axs2[i].set_ylabel("Время (с)")
        axs2[i].tick_params(axis='x', rotation=15)
        
        for bar in bars:
            height = bar.get_height()
            axs2[i].text(bar.get_x() + bar.get_width()/2., height, f'{height:.4f}', ha='center', va='bottom', fontsize=8)

    plt.tight_layout(rect=[0, 0.03, 1, 0.95])
    plt.savefig(os.path.join(DATA_PATH, "struct_comparison.png"))

create_plots()

with open(os.path.join(DOCS_PATH, "report.md"), "w", encoding="utf-8") as f:
    f.write("# Технический отчет: Сравнительный анализ структур данных\n\n")
    f.write("## 1. Вводные данные\n")
    f.write(f"Целью теста является оценка производительности LinkedList, HashTable и BST на массиве из {N} элементов. ")
    f.write("Анализировались сценарии со случайным распределением и предварительной сортировкой ключей.\n\n")
    
    f.write("## 2. Результаты измерений (AVG)\n")
    f.write("| Алгоритм | Входные данные | Вставка (с) | Поиск (с) | Удаление (с) |\n")
    f.write("| :--- | :--- | :--- | :--- | :--- |\n")
    for r in summary_for_report:
        f.write(f"| {r['name']} | {r['mode']} | {r['ins']:.6f} | {r['find']:.6f} | {r['del']:.6f} |\n")
    
    f.write("\n## 3. Визуальный анализ\n")
    f.write("### Сравнение по типам операций\n![Сравнение](data/struct_comparison.png)\n\n")
    f.write("### Влияние упорядоченности на производительность\n![Влияние порядка](data/order_impact.png)\n\n")
    
    f.write("## 4. Экспертные выводы\n")
    f.write("- **Эффект вырождения BST:** На отсортированных последовательностях BST демонстрирует критический рост времени выполнения (деградация до $O(N)$). ")
    f.write("Это связано с отсутствием балансировки, превращающим дерево в линейный список.\n")
    f.write("- **Инвариантность HashTable:** Хеш-таблица показывает наиболее стабильные результаты. Скорость доступа не коррелирует с порядком входных данных.\n")
    f.write("- **Линейная сложность LinkedList:** Связный список предсказуемо неэффективен при поиске, так как требует итерации по всей глубине структуры.\n")
    f.write("- **Итоговая оценка:** Для систем с высокой интенсивностью поиска и вставки оптимальным выбором является HashTable.")

print("Готово.")