JSONL：大模型时代的结构化数据格式

{"name": "Alice", "age": 25, "city": "Beijing"}
{"name": "Bob", "age": 30, "city": "Shanghai"}
{"name": "Charlie", "age": 35, "city": "Shenzhen"}

[
  {"name": "Alice", "age": 25, "city": "Beijing"},
  {"name": "Bob", "age": 30, "city": "Shanghai"},
  {"name": "Charlie", "age": 35, "city": "Shenzhen"}
]

{"messages": [{"role": "system", "content": "你是一个 helpful 的助手"}, {"role": "user", "content": "你好"}, {"role": "assistant", "content": "你好！有什么可以帮你的吗？"}]}
{"messages": [{"role": "user", "content": "讲个笑话"}, {"role": "assistant", "content": "为什么程序员总是分不清圣诞节和万圣节？因为 Oct 31 == Dec 25"}]}

{"timestamp": "2026-03-31T10:00:00Z", "level": "INFO", "message": "服务启动", "service": "api"}
{"timestamp": "2026-03-31T10:00:05Z", "level": "ERROR", "message": "数据库连接失败", "service": "db", "error": "timeout"}
{"timestamp": "2026-03-31T10:00:10Z", "level": "WARN", "message": "请求过多", "service": "api", "ip": "192.168.1.1"}

{"url": "https://example.com/page1", "title": "示例页面1", "content": "页面内容...", "crawled_at": "2026-03-31T10:00:00Z"}
{"url": "https://example.com/page2", "title": "示例页面2", "content": "页面内容...", "crawled_at": "2026-03-31T10:00:05Z"}

{"id": "doc_001", "text": "人工智能是计算机科学的一个分支", "embedding": [0.1, 0.2, 0.3, ...], "metadata": {"source": "wiki", "lang": "zh"}}
{"id": "doc_002", "text": "机器学习是AI的核心技术之一", "embedding": [0.2, 0.3, 0.4, ...], "metadata": {"source": "wiki", "lang": "zh"}}

import json

# 读取 JSONL
with open('data.jsonl', 'r', encoding='utf-8') as f:
    for line in f:
        data = json.loads(line.strip())
        print(data['name'])

# 写入 JSONL
with open('output.jsonl', 'w', encoding='utf-8') as f:
    for item in data_list:
        f.write(json.dumps(item, ensure_ascii=False) + '\n')

import pandas as pd

# 读取
df = pd.read_json('data.jsonl', lines=True)

# 写入
df.to_json('output.jsonl', orient='records', lines=True, force_ascii=False)

const fs = require('fs');
const readline = require('readline');

// 读取
const fileStream = fs.createReadStream('data.jsonl');
const rl = readline.createInterface({
  input: fileStream,
  crlfDelay: Infinity
});

for await (const line of rl) {
  const data = JSON.parse(line);
  console.log(data);
}

// 写入
const data = { name: 'Alice', age: 25 };
fs.appendFileSync('output.jsonl', JSON.stringify(data) + '\n');