@1-/csv
v0.1.4
Published
A minimalist and ultra-fast CSV encoder and decoder package. 极简、极速的 CSV 编码与解码工具包。
Maintainers
i18n-nowReadme
@1-/csv : A minimalist and ultra-fast CSV encoder and decoder package
Features
Zero-dependency, high-performance CSV serialization/deserialization. Supports both single-line and multi-line CSV processing with full RFC 4180 compliance and real-world fault tolerance.
Usage
Installation
bun add @1-/csvSingle-line encode
import csvE from "@1-/csv/csvE.js";
const data = ["Name", "Age", "20"];
const csv = csvE(data);
// Output: Name,Age,20Single-line decode
import csvD from "@1-/csv/csvD.js";
const csv = "Name,Age,20";
const data = csvD(csv);
// Output: ["Name", "Age", "20"]Multi-line encode
import dumps from "@1-/csv/dumps.js";
const data = [
["Name", "Age"],
["John", "25"],
["Jane", "30"]
];
const csv = dumps(data);
// Output: Name,Age\nJohn,25\nJane,30Multi-line decode
import loads from "@1-/csv/loads.js";
const csv = "Name,Age\nJohn,25\nJane,30";
const data = loads(csv);
// Output: [["Name", "Age"], ["John", "25"], ["Jane", "30"]]Design Principles
Pure functional implementation using character code operations. Supports RFC 4180 and real-world variants:
- Null/undefined handling (converted to empty string)
- Quote escaping (
""→") - Cross-platform line endings (
\n,\r,\r\n) - Fault-tolerant parsing (handles incomplete lines, trailing commas, etc.)
- Multi-line CSV batch processing
Tech Stack
- Runtime: ECMAScript 2023+
- Build: No build step
- Testing:
mitata - License: MulanPSL-2.0
Code Structure
src/
├── csvD.js # Single-line decoder, RFC 4180 compatible, fault-tolerant
├── csvE.js # Single-line encoder, auto quoting/escaping
├── dumps.js # Multi-line encoder, convert array of arrays to CSV string
├── loads.js # Multi-line decoder, parse CSV string to array of arrays
├── dump.js # File writer, write data to CSV file
├── load.js # File reader, read data from CSV file
├── csvD.d.ts # Type declarations
├── csvE.d.ts # Type declarations
├── dumps.d.ts # Type declarations
├── loads.d.ts # Type declarations
├── dump.d.ts # Type declarations
└── load.d.ts # Type declarationsHistorical Note
CSV originated in 1970s IBM System/360. Lotus 1-2-3 (1983) established it as de facto standard. RFC 4180 (2005) attempted standardization, but permissiveness remains a challenge. This project achieves dual compatibility through extreme minimalism while providing both single-line and multi-line processing capabilities.
About
This library is developed by WebC.site.
WebC.site: A new paradigm of web development for AI
@1-/csv : 极简、极速的 CSV 编码与解码工具包
功能介绍
提供零依赖、高性能的 CSV 序列化与反序列化功能。支持单行与多行 CSV 处理，完全兼容 RFC 4180 标准，同时具备现实世界容错能力。
使用演示
安装
bun add @1-/csv单行编码
import csvE from "@1-/csv/csvE.js";
const data = ["姓名", "年龄", "20"];
const csv = csvE(data);
// 输出: 姓名,年龄,20单行解码
import csvD from "@1-/csv/csvD.js";
const csv = "姓名,年龄,20";
const data = csvD(csv);
// 输出: ["姓名", "年龄", "20"]多行编码
import dumps from "@1-/csv/dumps.js";
const data = [
["姓名", "年龄"],
["张三", "25"],
["李四", "30"]
];
const csv = dumps(data);
// 输出: 姓名,年龄\n张三,25\n李四,30多行解码
import loads from "@1-/csv/loads.js";
const csv = "姓名,年龄\n张三,25\n李四,30";
const data = loads(csv);
// 输出: [["姓名", "年龄"], ["张三", "25"], ["李四", "30"]]设计思路
采用纯函数式实现，基于字符码直接操作。支持 RFC 4180 及常见变体：
- 空值与 null/undefined 处理（转换为空字符串）
- 双引号转义（
""→"） - 跨平台换行符（
\n,\r,\r\n） - 容错解析（处理不完整行、尾随逗号等）
- 多行 CSV 批量处理
技术栈
- 运行时：ECMAScript 2023+
- 构建：无构建步骤
- 测试：
mitata - 许可证：MulanPSL-2.0
代码结构
src/
├── csvD.js # 单行解码器，RFC 4180 兼容，容错
├── csvE.js # 单行编码器，自动 quoting/escaping
├── dumps.js # 多行编码器，将数组的数组转换为 CSV 字符串
├── loads.js # 多行解码器，将 CSV 字符串解析为数组的数组
├── dump.js # 文件写入器，将数据写入 CSV 文件
├── load.js # 文件读取器，从 CSV 文件读取数据
├── csvD.d.ts # 类型声明
├── csvE.d.ts # 类型声明
├── dumps.d.ts # 类型声明
├── loads.d.ts # 类型声明
├── dump.d.ts # 类型声明
└── load.d.ts # 类型声明历史故事
CSV 格式源于 1970 年代 IBM System/360。1983 年 Lotus 1-2-3 确立其为事实标准。RFC 4180（2005）尝试标准化，但宽松性仍是挑战。本项目以极致精简实现双重兼容，同时提供单行与多行处理能力。
关于
本库由 WebC.site 开发。
WebC.site : 面向人工智能的网站开发新范式