使用Tweepy获取特定推文线程评论时遇AttributeError求助

错误原因 #

你碰到的AttributeError: 'Client' object has no attribute 'conversation'是因为Tweepy v2的Client类没有conversation这个方法，属于误用了不存在的接口。

修正方案 #

要获取指定推文线程的所有内容，需要用Tweepy v2的search_recent_tweets接口，结合conversation_id作为查询条件，同时要请求用户信息字段来获取推文作者名称。以下是修正后的代码：

import tweepy

# 注意移除TOKEN前后的空格
BEARER_TOKEN = '你的Bearer Token'
# 注意移除ID前后的空格
conversation_id = '目标推文的conversation_id'

client = tweepy.Client(BEARER_TOKEN)

# 构造查询条件，指定对话ID
query = f"conversation_id:{conversation_id}"
# 请求对话内推文，同时获取作者信息
response = client.search_recent_tweets(
    query=query,
    tweet_fields=['created_at'],
    user_fields=['name', 'username'],
    expansions='author_id',
    max_results=100  # 单次请求最多返回100条，内容更多需分页处理
)

# 将作者信息存入字典，方便通过author_id快速匹配
authors = {user.id: user for user in response.includes['users']}

# 保存对话内容到TXT文件
with open(f"conversation_{conversation_id}.txt", "w", encoding='utf-8') as f:
    for tweet in response.data:
        author = authors[tweet.author_id]
        f.write(f"【{author.name} (@{author.username})】\n发布时间: {tweet.created_at}\n内容:\n{tweet.text}\n\n")

print(f"对话已保存到conversation_{conversation_id}.txt")

分页补充（针对长对话） #

如果对话线程的推文数量超过100条，需要通过分页获取全部内容，可使用Paginator实现：

from tweepy import Paginator

all_tweets = []
authors = {}

# 遍历分页结果，收集所有推文和作者信息
for response in Paginator(client.search_recent_tweets,
                          query=query,
                          tweet_fields=['created_at'],
                          user_fields=['name', 'username'],
                          expansions='author_id',
                          max_results=100):
    all_tweets.extend(response.data)
    for user in response.includes['users']:
        if user.id not in authors:
            authors[user.id] = user

# 保存逻辑不变，遍历all_tweets即可
with open(f"conversation_{conversation_id}.txt", "w", encoding='utf-8') as f:
    for tweet in all_tweets:
        author = authors[tweet.author_id]
        f.write(f"【{author.name} (@{author.username})】\n发布时间: {tweet.created_at}\n内容:\n{tweet.text}\n\n")

print(f"完整对话已保存到conversation_{conversation_id}.txt")

内容的提问来源于stack exchange，提问作者Marc Genenger