[Indexer] Implements state indexer

[baichuan3]

Summary

Part of #1028, Implements state indexer

1. Create a global Object table: (object_id, owner, flag, JSON (V), key_type), and store V according to JSON. Global states table only store table objects and raw objects.

CREATE TABLE global_states
(
  object_id          VARCHAR        NOT NULL       PRIMARY KEY,
  owner              VARCHAR        NOT NULL,
  flag               SMALLINT       NOT NULL,
  value              VARCHAR        NOT NULL,
  key_type           VARCHAR        NOT NULL,
  size               BIGINT         NOT NULL,
  created_at         BIGINT         NOT NULL,
  updated_at         BIGINT         NOT NULL
);

2. Create a leaf table: (object_id, key_hash, JSON (V), key_type), and store V according to JSON.

CREATE TABLE leaf_states
(
  id                 VARCHAR        NOT NULL      PRIMARY KEY,
  object_id          VARCHAR        NOT NULL,
  key_hash           VARCHAR        NOT NULL,
  value              VARCHAR        NOT NULL,
  value_type         VARCHAR        NOT NULL,
  created_at         BIGINT         NOT NULL,
  updated_at         BIGINT         NOT NULL,
  UNIQUE (object_id, key_hash)
);

3. The state indexer currently does not provide RPC query, and will provide query through GraphQL later.
4. Diesel for SQLite don't support batch update yet, so implements batch update directly via raw SQL.

TODO

1. Since table creation may be lazy, create global table if the table does not exist.
2. Update table size if ObjectID is table hanlde.

• Closes [Indexer] Implement state indexer  #1036

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

1 Ignored Deployment

Name	Status	Preview	Updated (UTC)
rooch	⬜️ Ignored (Inspect)		Nov 29, 2023 1:53pm

[jolestar]

The key here is not hash but the key's bcs binary. So, the field name should be key.

[baichuan3]

key in SQLite is a keyword. Or use key_str?