Crate vsdb

source · []
Expand description

GitHub top language

VSDB is a ‘Git’ in the form of a KV database.

Based on the powerful version control function of VSDB, you can easily give your data structure the ability to version management.

Make everything versioned !!


  • Support Git-like verison operations, such as:
    • Create countless branches and merge them to their parents
    • Rolling back a ‘branch’ to a specified historical ‘version’
    • Querying the historical value of a key on the specified ‘branch’
  • Most APIs is similar as the coresponding data structures in the standard library
    • Use Vecx just like Vec
    • Use Mapx just like HashMap
    • Use MapxOrd just like BTreeMap


Suppose you have a great algorithm like this:

struct GreatAlgo {
    a: Vec<...>,
    b: BTreeMap<...>,
    c: u128,
    d: HashMap<...>,
    e: ...

Simply replace the original structure with the corresponding VSDB data structure, and your algorithm get the powerful version control ability at once!

#[dervive(Vs, Default)]
struct GreatAlgo {
    a: VecxVs<...>,
    b: MapxOrdVs<...>,
    c: OrphanVs<u128>,
    d: MapxVs<...>,
    e: ...

let algo = GreatAlgo.default();



the #[derive(Vs)] macro can be applied to structures whose internal fields are all types defined in VSDB (primitive types and their collections are also supported), but can not be applied to nesting wrapper among VSDB-types, you should implement the VsMgmt trait(or a part of it) manually.

This data structure can be handled correctly by #[derive(Vs)]:

use std::collections::{
    HashMap, HashSet, BTreeSet, VecDeque, LinkedList
use std::sync::atomic::AtomicU64;
use std::marker::PhantomData;
use ruc::*;
use vsdb::{
    BranchName, VersionName, ParentBranchName,
    MapxVs, MapxOrdVs, OrphanVs, VsMgmt, Vs,
    VecxVs, impl_vs_methods_nope

struct GoodCase<K, T> {
    a: VecxVs<i64>,
    b: SubItem0,
    c: SubItem1,
    d: SubItem2,
    e: u8,
    f: Vec<i16>,
    g: VecDeque<i64>,
    h: BTreeSet<u16>,
    i: HashMap<K, AtomicU64>,
    j: HashSet<i32>,
    k: LinkedList<()>,
    l: Box<dyn AsRef<bool>>,
    m: Box<dyn AsRef<[Vec<u128>]>>,
    n: PhantomData<T>,

struct SubItem0(MapxVs<u8, u8>, VecxVs<u8>);

struct SubItem1 {
    a: OrphanVs<i16>,
    b: MapxOrdVs<String, u8>

struct SubItem2 {
    a: i8,
    b: u128

// // A nope implementation of `VsMgmt` for custom stateless types.
// // the `#[derive(Vs)]` on 'SubItem2' is same as this implementation.
// impl VsMgmt for SubItem2 {
//     impl_vs_methods_nope!();
// }

But this one can NOT be handled correctly by #[derive(Vs)]:

use vsdb::{Vs, VecxVs, MapxVs};
use ruc::*;

// It can be compiled, but the result is wrong !
// The versioned methods of the inner 'MapxVs<u8, u8>' will missing,
// you should implement the 'VsMgmt' trait(or a part of it) manually.
struct BadCase {
    a: VecxVs<MapxVs<u8, u8>>,

Some complete examples:

Compilation features

  • [default] sled_engine, use sled as the backend database
    • Faster compilation speed
    • Support for compiling into a statically linked binary
  • rocks_engine, use rocksdb as the backend database
    • Faster running speed
    • Can not be compiled into a statically linked binary
  • [default] cbor_codec, use cbor as the codec
    • Faster running speed
  • bcs_codec, use bcs as the codec
    • Created by the ‘Libre’ project of Facebook
    • Security reinforcement for blockchain scenarios

Low-level design

Based on the underlying one-dimensional linear storage structure (native kv-database, such as sled/rocksdb, etc.), multiple different namespaces are divided, and then abstract each dimension in the multi-dimensional logical structure based on these divided namespaces.

In the category of kv-database, namespaces can be expressed as different key ranges, or different key prefix.

This is the same as expressing complex data structures in computer memory(the memory itself is just a one-dimensional linear structure).

User data will be divided into two dimensions: ‘branch’ and ‘version’, the functions of the ‘basic’ category are stateless, and the functions of the ‘versioned’ category are stateful. In the internal implementation, each stateful function is implemented based on its corresponding stateless function, all stateful data has two additional identification dimensions (‘branch’ and ‘version’), somewhat like the logic in Git. Stateless functions do not have the feature of ‘version’ management, but they have higher performance.


pub use basic::mapx::Mapx;
pub use basic::mapx_ord::MapxOrd;
pub use basic::vecx::Vecx;
pub use versioned::mapx::MapxVs;
pub use versioned::mapx_ord::MapxOrdVs;
pub use versioned::orphan::OrphanVs;
pub use versioned::vecx::VecxVs;
pub use versioned::VsMgmt;
pub use merkle::MerkleTree;


Unversioned functions.

A simple ‘Merkle-Tree’ ported from solana project.

Versioned functions.


A helper for implementing VsMgmt for collection types, struct NewType(HashMap), struct NewType(BTreeMap), etc.

Add nope implementations of VsMgmt for types that are not defined in VSDB.

Parse bytes to a specified integer type.

Parse bytes to a Prefix type.


Avoid making mistakes between branch name and version name.

Avoid making mistakes between branch name and version name.

Avoid making mistakes between branch name and version name.


Methods used to encode and decode the KEY.

For keys that their serialized order keep consistent with their original format. When using this kind of keys, we can do some ordered operations, such as: get_le/get_be ...

Methods used to encode and decode the VALUE.


Flush data to disk, may take a long time.

Set ${VSDB_BASE_DIR} manually.

Derive Macros