A JavaScript library for connecting to an HEAVY.AI GPU database and running queries.
Install with npm install @heavyai/connector
. Then import @heavyai/connector
in your JavaScript program.
Visit our API Docs
There have been some potentially breaking changes in v6:
window
. v6 will continue to do this if you import the script using a standard html <script>
tag. But, if you are using a bundler or module system (such as webpack), that is no longer the case. These types will be exported instead (so you can, for example, import { TPixel } from "@heavyai/connector/dist/browser-connector"
)xAsync
functions return a Promise - that hasn't changed. The corresponding non-Async functions (ie, query
vs queryAsync
) now also return Promises. The callback argument on these non-Async versions is optional since you can just use then
or catch
or the Promise. It's impossible to run any function synchronously now, however. In other words, if you call query
without a callback, it's going to run asynchronously and return a Promise.Open example.html in your browser for a basic query call.
Quick Start for Developersnpm ci npm run build npm run docs # opens API docs in your browser.
Everything in connector should be unit-tested and linted. You can find these tests in /test
.
The linter and all tests run on
It's our eventual goal to fully lint the files in heavyai-con/src
. Try to write libraries/heavyai-con
using the projects/dashboard-v2
Es6/7 style to make this goal easier to achieve.
npm run build
Creates /dist
folder and runs webpack
script npm run clean
Removes node modules, dist, and docs npm run docs
Creates and opens docs npm run lint
Runs lint npm run test
Runs linting and unit tests npm run test:unit
Runs mocha unit tests
Interested in contributing? We'd love for you to help! Check out Contributing.MD
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4