batch_analyze.py

# pip install pandas matplotlib requests
# pandas：处理表格数据
# matplotlib：画图
# requests：调用你自己的 API（就像其他服务会做的那样）
# 启动服务1
# python -m uvicorn api:app --reload --port 8000
# 启动服务2
# python batch_analyze.py
# 启动服务3
# python generate_report.py

import os
os.environ['HF_ENDPOINT'] = 'https://hf-mirror.com'
import pandas as pd
import requests
import matplotlib.pyplot as plt
import time

# 1. 读取评论数据
df = pd.read_csv("reviews.csv")
print(f"共加载 {len(df)} 条评论")

# 2. 调用本地 API 进行情感分析
results = []
api_url = "http://localhost:8000/analyze"

for idx, row in df.iterrows():
    try:
        response = requests.post(api_url, json={"text": row["review"]})
        if response.status_code == 200:
            data = response.json()
            results.append({
                "id": row["id"],
                "review": row["review"],
                "sentiment": data["sentiment"],
                "confidence": data["confidence"]
            })
        else:
            print(f"❌ 分析失败 (ID={row['id']}): {response.text}")
            results.append({
                "id": row["id"],
                "review": row["review"],
                "sentiment": "未知",
                "confidence": 0.0
            })
        # 避免请求过快（非必需，但友好）
        time.sleep(0.1)
    except Exception as e:
        print(f"⚠️ 请求异常 (ID={row['id']}): {e}")
        results.append({
            "id": row["id"],
            "review": row["review"],
            "sentiment": "错误",
            "confidence": 0.0
        })

# 转为 DataFrame
result_df = pd.DataFrame(results)
result_df.to_csv("analyzed_reviews.csv", index=False, encoding="utf-8-sig")
print("✅ 分析结果已保存到 analyzed_reviews.csv")