https://github.com/txstate-etc/canvas-client

Typescript library to make accessing the Canvas API more convenient.
https://github.com/txstate-etc/canvas-client

Last synced: about 1 month ago
JSON representation

Typescript library to make accessing the Canvas API more convenient.

• Host: GitHub
• URL: https://github.com/txstate-etc/canvas-client
• Owner: txstate-etc
• License: mit
• Created: 2019-12-18T16:52:03.000Z (about 5 years ago)
• Default Branch: master
• Last Pushed: 2023-12-08T17:11:24.000Z (about 1 year ago)
• Last Synced: 2024-04-27T02:03:37.718Z (9 months ago)
• Language: TypeScript
• Homepage:
• Size: 571 KB
• Stars: 0
• Watchers: 7
• Forks: 1
• Open Issues: 7
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# canvas-client

Typescript library to make accessing the Canvas API more convenient.

# Features

## Typescript

The canvas client provided by this library has a lot of pre-built methods where we have carefully observed the behavior of the Canvas RESTful API and provided types that accurately reflect what it accepts and returns. The pre-built methods are not complete but they should cover the most common endpoints for server-to-server integrations like course, user, and assignment management.

## Rate-limiting and token-splitting

Canvas rate-limits incoming requests to about 20 simultaneous requests per account. This can be rather problematic if you are performing server-to-server operations on behalf of the entire campus via a single service account. This library's solution is two-fold:

* Rate-limit on the client side to ensure canvas never gets overloaded. This way requests to Canvas never fail simply for rate-limiting.

  * The default client-side rate limit is a conservative 10 simultaneous requests. It's configurable. If multiple instances of your app share the same token, you may want to reduce this number.

* Support multiple tokens (each MUST come from a separate user account) and load-balance requests across the tokens. This increases your total capacity.

  * Note that if you run multiple instances, you STILL want to reduce the max simultaneous requests per token, to avoid getting errors when you have a backlog of requests overwhelming the capacity of your tokens.

## GraphQL

We've provided a very simple method for making queries against Canvas' GraphQL API. It only takes care of the fundamentals of GraphQL. We do not provide types for

graphql queries, but the `graphql` method does accept a generic so that you can provide the expected return type yourself. GraphQL requests also benefit from token splitting and rate limiting. Tokens and rate limits are shared with RESTful requests.

# Testing

Duplicate `.env.example`, rename it to `.env` and fill in good values for each of the variables. This will allow you to run `npm test` successfully.

# 2.0

The breaking change for 2.0 is node 18+ is required since I got rid of axios in favor of the native fetch api.