Lib.rs

Serde Buckets

'Buckets' for Serde Deserialisation

The Bucket stores data (and it's type) according to the serde data model, allowing for format analogous representation of any self-describing type.

This type is intended for use in the deserialisation process, where the structure is not known until a given field is parsed. This type can store that data without copying or taking ownership of heap allocated types (unlike serde_value).

Under the hood, the Bucket type is a vec of 'nodes' that depict the type, it's value and other attributes. This makes it easier to work with and removes some of the heap allocation that you find with recursive data structures.

This is a crate created for my projects, and as such is unlikely to be maintained. It might be worth saying 'bucket' and use serde_value instead.

Example

The following examples use serde_json as the format

use serde_bucket::Bucket;

// parse an input using your favourite serde library
// deserialise it into the `Bucket` type.
let input = r#"{"a": 10, "b": false}"#;
let mut  bucket: Bucket = serde_json::from_str(&input).unwrap();

// our example structure
#[derive(Deserialize)]
struct Example {
    a: u8,
    b: bool
}

// use `deserialize_into` to "deserialise into" a given type.
// the error type (in this `serde_json::Error`) must implement `serde::de::Error`
let value = bucket.deserialize_into::<Example, serde_json::Error>().unwrap();
assert_eq!(value, Example { a: 10, b: false });

Thanks

Merci Buckets for stopping by <3 - here is My Twitter if you want to say hello