serde-csv-core

为无_std crate提供的CSV序列化和反序列化。

serde-csv-core基于csv-core crate构建。它不需要任何内存分配，这意味着它非常适合嵌入式环境。

序列化

Writer一次序列化一条记录。

use heapless::String;
use serde::Serialize;

#[derive(Serialize)]
struct Record {
    pub country: String<32>,
    pub city: String<32>,
    pub population: u32,
}

let records = [
    Record {
        country: "Poland".into(),
        city: "Cracow".into(),
        population: 766_683,
    },
    Record {
        country: "Japan".into(),
        city: "Tokyo".into(),
        population: 13_515_271,
    },
];

let mut writer = serde_csv_core::Writer::new();
let mut csv = [0; 128];
let mut nwritten = 0;
for record in &records {
    nwritten += writer.serialize(&record, &mut csv[nwritten..])?;
}

assert_eq!(&csv[..nwritten], b"Poland,Cracow,766683\nJapan,Tokyo,13515271\n");

反序列化

Reader<N>一次反序列化一条记录。 N 是用于临时存储未转义的字段的内部缓冲区的容量。

use heapless::{String, Vec};
use serde::Deserialize;

#[derive(Debug, PartialEq, Eq, Deserialize)]
struct Record {
    pub country: String<32>,
    pub city: String<32>,
    pub population: u32,
}

let csv = b"Poland,Cracow,766683\nJapan,Tokyo,13515271\n";

let mut reader = serde_csv_core::Reader::<32>::new();
let mut records: Vec<Record, 2> = Vec::new();
let mut nread = 0;
while nread < csv.len() {
    let (record, n) = reader.deserialize::<Record>(&csv[nread..])?;
    records.push(record);
    nread += n;
}

assert_eq!(records, &[
    Record {
        country: "Poland".into(),
        city: "Cracow".into(),
        population: 766_683,
    },
    Record {
        country: "Japan".into(),
        city: "Tokyo".into(),
        population: 13_515_271,
    },
]);

许可证

根据您的选择，许可受Apache License，版本2.0或MIT许可。

除非您明确声明，否则根据Apache-2.0许可定义，您有意提交的任何贡献，都应按上述方式双重许可，没有任何附加条款或条件。