正则返回完整字符串而非捕获组，如何仅提取目标捕获组值？

首先得说，其实处理JSON格式的数据，最稳妥的方式是用专门的JSON解析器，不过既然你问的是正则捕获组的用法，我先给你解决正则的问题，再补充更优方案。

正则捕获组的正确写法 #

你的问题大概率是正则表达式没精准定位到amount的数值，或者你没正确提取捕获组的内容。针对你给出的服务器响应数据，可以用这个正则来精准捕获金额：

"amount":"([0-9.]+)"

如果需要同时捕获币种和对应的金额，就用这个更完整的正则：

"base":"([A-Z]+)","currency":"USD","amount":"([0-9.]+)"

这里的括号就是捕获组：

• 第一个捕获组([A-Z]+)会匹配BTC、BCH这类币种代码
• 第二个捕获组([0-9.]+)就是你要的纯金额数值

怎么获取捕获组的值 #

不同编程语言里提取捕获组的方式略有不同，给你举几个常见场景的例子：

Python 示例

import re

response = '{"data":[{"base":"BTC","currency":"USD","amount":"8199.91"},{"base":"BCH","currency":"USD","amount":"1012.22"},{"base":"ETH","currency":"USD","amount":"825.94"},{"base":"LTC","currency":"USD","amount":"150.11"}]}'

# 只提取所有金额
amounts = re.findall(r'"amount":"([0-9.]+)"', response)
print(amounts)  # 输出: ['8199.91', '1012.22', '825.94', '150.11']

# 同时提取币种和金额
matches = re.findall(r'"base":"([A-Z]+)","currency":"USD","amount":"([0-9.]+)"', response)
for coin, price in matches:
    print(f"币种: {coin}, 价格: {price}")

JavaScript 示例

const response = '{"data":[{"base":"BTC","currency":"USD","amount":"8199.91"},{"base":"BCH","currency":"USD","amount":"1012.22"},{"base":"ETH","currency":"USD","amount":"825.94"},{"base":"LTC","currency":"USD","amount":"150.11"}]}';
const regex = /"amount":"([0-9.]+)"/g;
let match;
const amounts = [];
while ((match = regex.exec(response)) !== null) {
    amounts.push(match[1]); // match[0]是整个匹配的完整字符串，match[1]才是捕获组的数值
}
console.log(amounts);

更推荐的JSON解析方案 #

其实正则处理JSON挺脆弱的——如果服务器返回的字段顺序变了、多了个空格，或者新增了其他字段，正则很可能直接失效。用JSON解析器才是长期可靠的方案，比如Python的示例：

import json

response = '{"data":[{"base":"BTC","currency":"USD","amount":"8199.91"},{"base":"BCH","currency":"USD","amount":"1012.22"},{"base":"ETH","currency":"USD","amount":"825.94"},{"base":"LTC","currency":"USD","amount":"150.11"}]}'
data = json.loads(response)

for item in data['data']:
    print(f"币种: {item['base']}, 价格: {item['amount']}")

这样不管字段顺序怎么变，都能准确拿到你要的数值，代码可读性也更高。

内容的提问来源于stack exchange，提问作者j Doe