Skip to content

Atom Accelerator Question about Milestone 2v3

danb edited this page Aug 3, 2023 · 1 revision

Milestone 2: Generalized Indexer Implementation

Summary

Milestone 2 focuses on implementing a generalized indexer that can accommodate various data points consistently. This system will index information into tables to manage and query data effectively. However, it does not offer fine-grained control over data shaping.

Introduction

Every Block has a consistent set of data points. This milestone is intended to index these consistent data points into a general data format.

Indexing Details

For example:

  • Blocks: Will be indexed into a Block table.
  • Transactions: Will be indexed into a Transaction table.
  • Messages: Will be indexed into a Message table.
  • Block Events: Block events and their attributes will be indexed into a BlockEvent and BlockEventAttribute table.
  • Message Events: Message events and their attributes will be indexed into a MessageEvent and MessageEventAttribute table.

Milestone 3: Custom Block Event and Transaction Message Parser Interface

Summary

Milestone 3 enables users to index data in a custom format, allowing for better querying and organization. This caters to specific user needs, offering more flexibility compared to the generalized indexer in Milestone 2.

Introduction

When a Block event or Transaction message is encountered, the user of the indexer may want to index the data into a custom format for better querying/organization of the data. This milestone is intended to provide users the ability to implement custom parsers for their own particular use-cases.

Custom Parser Example

For example, say a user is interested in voting patterns and wants to index the MsgVote transaction message into a custom database table format.

This user would create a custom parser that executes when a MsgVote message is encountered. The custom parser would transform the message into the desired data format and insert it into their custom table. They would then be able to more easily query for voting patterns rather than relying on the generalized indexer dataset as described in Milestone 2.