Python Twitter Bot本地正常但Heroku部署崩溃的排查修复咨询

1. 文件编码不匹配：本地生成的quotes3.json采用Windows系统默认编码（如Windows-1252），而Heroku的Linux环境默认以UTF-8编码读取文件，导致解码时遇到无法识别的字节（0x93），抛出UnicodeDecodeError。
2. 依赖本地预生成文件：代码未在Heroku运行时动态生成JSON文件，而是直接读取本地提交的编码不兼容的文件。

方案1：强制指定文件编码读取（快速修复） #

修改get_quote()函数中读取JSON文件的代码，明确指定与本地文件匹配的编码：

def get_quote():
    # 指定编码为Windows-1252，匹配本地生成文件的编码
    with open('quotes3.json', 'r', encoding='windows-1252') as f:
        read_f = json.load(f)
    # 后续代码保持不变
    random_quote = random.choice(read_f)
    author = random_quote['author']
    quote_text = f"{random_quote['quote']}"
    tweet = re.sub("[^A-Za-z0-9.,:;' ]","",quote_text)
    ftweet= tweet +" - " +author +" #Inspiration_of_the_day"
        
    while True:
        if len(ftweet) <=280:
            return ftweet

方案2：在Heroku上动态生成JSON文件（推荐，彻底解决编码问题） #

避免依赖本地文件，让应用在Heroku启动时自动爬取生成UTF-8编码的JSON文件：

1. 修改quotes_scraper()函数，写文件时明确指定UTF-8编码：

def quotes_scraper():
    collection = list()
    TAGS = ["science","humor","life", "knowledge","motivational", "inspirational", "truth", "philosophy"  ]

    for tag in TAGS:
        url = f"https://www.goodreads.com/quotes/tag/{tag}" 
        page = requests.get(url)
        # 改用page.text让requests自动处理响应编码
        soup = BeautifulSoup(page.text, "html.parser")
        raw_data = soup.find_all(class_ ="quoteText")

        for quote in raw_data:
            quote_text = quote.contents[0].strip()
            author = quote.find(class_="authorOrTitle").text.strip()
            data = {
                "quote": quote_text, 
                "author": author,
                "genre": tag
            }
            collection.append(data)

    # 明确指定UTF-8编码写入文件
    with open("quotes3.json", "w", encoding='utf-8') as qfile:
        json.dump(collection, qfile, ensure_ascii=False)
    return collection

2. 修改run()函数，启动时先调用quotes_scraper()生成文件：

def run():
    # 启动时先爬取生成UTF-8编码的quotes3.json
    quotes_scraper()
    while True:
        tweet()
        time.sleep(10)
        like_and_retweet()
        time.sleep(run_tm(60*60*4,60*60*10))

3. 将quotes3.json加入.gitignore文件，避免提交本地编码的文件到Heroku：

# .gitignore内容
quotes3.json

额外优化（可选） #

添加异常处理，避免爬取失败导致应用崩溃：

def quotes_scraper():
    collection = list()
    TAGS = ["science","humor","life", "knowledge","motivational", "inspirational", "truth", "philosophy"  ]

    for tag in TAGS:
        url = f"https://www.goodreads.com/quotes/tag/{tag}" 
        try:
            page = requests.get(url, timeout=10)
            page.raise_for_status()  # 抛出HTTP错误
            soup = BeautifulSoup(page.text, "html.parser")
            raw_data = soup.find_all(class_ ="quoteText")

            for quote in raw_data:
                quote_text = quote.contents[0].strip()
                author = quote.find(class_="authorOrTitle").text.strip()
                data = {
                    "quote": quote_text, 
                    "author": author,
                    "genre": tag
                }
                collection.append(data)
        except Exception as e:
            print(f"爬取标签 {tag} 失败: {e}")
            continue

    with open("quotes3.json", "w", encoding='utf-8') as qfile:
        json.dump(collection, qfile, ensure_ascii=False)
    return collection

内容的提问来源于stack exchange，提问作者kana natasha