import requests
import pandas as pd
import nltk
from google.cloud import language_v1
import matplotlib.pyplot as plt
import os
from datetime import datetime, timedelta

nltk.download('punkt')
nltk.download('stopwords')

# 设置Google Cloud服务账号密钥文件路径
os.environ['GOOGLE_APPLICATION_CREDENTIALS'] = "D:\PYcharm\pycharm projects\Life\\blibliCrawler\paris-olympic-436002-8fa93022f012.json"

# 第一步：使用Google Custom Search API获取新闻数据
def fetch_news(api_key, cse_id, query, num_results=50):
    news_data = []
    for start in range(1, num_results, 10):
        url = 'https://www.googleapis.com/customsearch/v1'
        params = {
            'key': '8fa93022f012e9d617de9ca25d49dcc258f17ebd',
            'cx': 'paris-olympic-436002',
            'q': query,
            'start': start,
            'lr': 'lang_en',  # 搜索英文内容；如需中文，请改为 'lang_zh-CN'
            'sort': 'date'    # 按日期排序
        }
        response = requests.get(url, params=params)
        data = response.json()
        if 'items' not in data:
            print("未获取到更多数据。")
            break
        news_data.extend(data['items'])
    return pd.DataFrame(news_data)

# 第二步：数据预处理
def preprocess_text(text):
    # 转换为小写
    text = text.lower()
    # 分词
    tokens = nltk.word_tokenize(text)
    # 去除停用词和非字母字符
    tokens = [word for word in tokens if word.isalpha() and word not in nltk.corpus.stopwords.words('english')]
    return ' '.join(tokens)

# 第三步：情感分析
def analyze_sentiment(text_content):
    client = language_v1.LanguageServiceClient()
    document = language_v1.Document(content=text_content, type_=language_v1.Document.Type.PLAIN_TEXT)
    sentiment = client.analyze_sentiment(request={'document': document}).document_sentiment
    return sentiment.score

# 第四步：可视化
def plot_sentiment_over_time(df):
    df['date'] = pd.to_datetime(df['date'])
    df.set_index('date', inplace=True)
    df.resample('D')['sentiment'].mean().plot()
    plt.title('情感得分随时间的变化')
    plt.xlabel('日期')
    plt.ylabel('平均情感得分')
    plt.show()

# 主函数
if __name__ == "__main__":
    # 替换为您的API密钥和搜索引擎ID
    api_key = 'YOUR_GOOGLE_CUSTOM_SEARCH_API_KEY'
    cse_id = 'YOUR_CUSTOM_SEARCH_ENGINE_ID'
    query = 'Paris Olympics'  # 或 '巴黎奥运会' 获取中文内容
    # 获取新闻数据
    df = fetch_news(api_key, cse_id, query)
    print(f"共获取到 {len(df)} 篇文章。")

    # 提取必要的字段
    if 'snippet' in df.columns and 'title' in df.columns:
        df['text'] = df['title'] + ' ' + df['snippet']
    else:
        print("缺少必要的文本字段。")
        exit()

    # 处理日期字段
    if 'pagemap' in df.columns:
        dates = []
        for item in df['pagemap']:
            if 'metatags' in item and 'og:updated_time' in item['metatags'][0]:
                dates.append(item['metatags'][0]['og:updated_time'])
            else:
                dates.append(datetime.now().isoformat())
        df['date'] = dates
    else:
        df['date'] = datetime.now().isoformat()

    # 数据预处理
    df['cleaned_text'] = df['text'].apply(preprocess_text)

    # 情感分析
    df['sentiment'] = df['cleaned_text'].apply(analyze_sentiment)

    # 可视化
    plot_sentiment_over_time(df)