1、前言

本项目是基于requests爬虫+flask+echarts搭建的网易云评论的可视化大屏，主要涉及到的技术有爬虫、数据库操作，flask框架，echarts图表。

最终效果如下：

2、实现

2.1 挑选想要采集的歌曲评论

为了爬取歌曲评论，一想到的是网易云音乐，故从中挑选评论比较多的五首歌，分别是：

富士山下
爱情转移
孤独患者
葡萄成熟时
任我行

这五首歌，对应的网易云 id 是：

music_list = {65766: '富士山下', 65536: '爱情转移', 64093: '孤独患者', 66285: '葡萄成熟时', 27483202: '任我行'}

2.2 构建爬虫

2.2.1 采集歌曲评论

url = 'https://music.163.com/weapi/comment/resource/comments/get?csrf_token='
data = {
    'params': encText,
    'encSecKey': encSecKey
}
respond = requests.post(url, headers=headers, data=data)
json_data = json.loads(respond.text)
comments = json_data['data']['comments']
for per_comment in comments:
    comment = per_comment['content']
    user_name = per_comment['user']['nickname']
    user_id = per_comment['user']['userId']
    comment_time = per_comment['time']
    timestamp = comment_time / 1000
    dt_object = datetime.datetime.fromtimestamp(timestamp)
    phone = per_comment.get('extInfo', {}).get('endpoint', {}).get('CLIENT_TYPE', '')
    ip = per_comment['ipLocation']['location']
    per_line = pd.DataFrame(
        {'user_id': [user_id], 'comment': [comment], 'user_name': [user_name], 'time': [dt_object], 'ip': [ip],
         'phone': [phone], 'music_id': [music_id], 'music_name': [music_list[music_id]]})
    result = pd.concat([result, per_line])
    last_time = comment_time

2.2.2 清洗数据入库

def process_csv(file_path):
    # 读取 CSV 文件
    df = pd.read_csv(file_path)
    df = df.drop_duplicates()
    # 数据库连接
    connection = pymysql.connect(
        host='127.0.0.1',
        user='root',
        password='123456',
        database='music_comments'
    )
    try:
        # 创建表（如果不存在）
        create_table_if_not_exists(connection)
        insert_query = """
        INSERT INTO comments (user_id, comment, user_name, time, ip, phone, music_id, music_name)
        VALUES (%s, %s, %s, %s, %s, %s, %s, %s)
        """
        with connection.cursor() as cursor:
            for _, row in df.iterrows():
                cursor.execute(insert_query, (
                    row['user_id'] if pd.notna(row['user_id']) else None,
                    row['comment'] if pd.notna(row['comment']) else None,
                    row['user_name'] if pd.notna(row['user_name']) else None,
                    row['time'] if pd.notna(row['time']) else None,
                    row['ip'] if pd.notna(row['ip']) else None,
                    row['phone'] if pd.notna(row['phone']) else None,
                    row['music_id'] if pd.notna(row['music_id']) else None,
                    row['music_name'] if pd.notna(row['music_name']) else None
                ))
        connection.commit()
        print("数据写入成功")
    except Exception as e:
        print(f"错误: {e}")
    finally:
        connection.close()

最后数据库中数据效果如下：

2.3 搭建flask框架

from flask import Flask,render_template
 
app = Flask(__name__)
app.config['JSON_AS_ASCII'] = False
 
@app.route('/')
def index():
    return render_template('index.html')
 
if __name__ == '__main__':
    app.run(debug=True)

2.4 数据传值

# 左上图
data_music_name = ['富士山下', '爱情转移', '孤独患者', '葡萄成熟时', '任我行']
data_count = [216651, 142228, 109656, 102065, 26275]
list_1_data = []
for i in range(0, len(data_music_name)):
    music_name = data_music_name[i]
    count = data_count[i]
    list_1_data.append([music_name, count])
# 中上图
data2 = df['phone'].value_counts().reset_index()
data2 = data2[~data2['phone'].isnull()]
data_phone = data2['phone'].tolist()
data_count = data2['count'].tolist()
list_2_name = []
list_2_value = []
for i in range(0, len(data_phone)):
    list_2_name.append(data_phone[i])
    list_2_value.append(data_count[i])
# 右上图
text_df = df[~df['comment'].isnull()]
text = '。'.join(text_df['comment'].tolist())
words = jieba.cut(text)
word_counts = Counter(words)
filtered_counts = {word: count for word, count in word_counts.items()
                   if len(word) > 1 and re.match(r'^[\u4e00-\u9fa5]+$', word)}
most_common_words = Counter(filtered_counts).most_common(10)
most_common_words4 = Counter(filtered_counts).most_common(100)
list_3_name = []
list_3_value = []
for per_word in most_common_words:
    list_3_name.append(per_word[0])
    list_3_value.append(per_word[1])
# 左下图
data4 = df['ip'].value_counts().reset_index()
data4 = data4[~data4['ip'].isnull()]
data_ip = data4['ip'].tolist()
data_count = data4['count'].tolist()
list_4_name = []
list_4_value = []
# 显示全部
# for i in range(0, len(data_ip)):
# 显示前十
for i in range(0, 10):
    list_4_name.append(data_ip[i])
    list_4_value.append(data_count[i])
# 词云
wordcloud_data = [{"name": word, "value": count} for word, count in most_common_words4]

2.5 完整代码&数据集获取

完整代码&数据集可以点击最上面链接下载或者扫描下方👇🏻👇🏻👇🏻二维码获取，还有更多可视化大屏等着你：

001 服务大数据可视化监管平台
002 水质情况实时监测预警系统
003 联心菜市场数据中心
004 政务大数据共享交换平台
005 可视化监控管理
006 全国疫情实时监控
007 惠民服务平台
008 兰州智慧消防大数据平台
......