One less abstraction layer... about heroprotocoljs HOT 2 CLOSED

I don't really understand what you mean. Things are already cached as long as you keep references to opened archives so things don't get parsed twice.

Could you provide usescases that you think are problematic with the current API and explain how they are problematic?

from heroprotocoljs.

The problem is with the extract. The mpyqjs library is too slow.

I would like to extract the file with ANOTHER library, but use this library to parse it. My only option to interface with this library currently is sent it a replay, and a variable.

const heroprotocol = require('../');
const details = heroprotocol.get(heroprotocol.DETAILS, process.argv[2]);

I would LIKE to be able to send it data I've already extracted.

const heroprotocol = require('../');
const header = heroprotocol.parseHeader(buffer);
const details = heroprotocol.parseFile(buffer, header.build);

Why? Because I can extract files 60% faster using StormLib rather than mpyqs. There's little point to rewriting heroprotocoljs when it does everything I need, it just does not expose it.

Extracting 28 replays with mpyqs

real 1m43.637s
user 1m35.931s
sys  0m6.731s

Extracting 28 replays with storm-replay

real 0m37.774s
user 0m32.481s
sys  0m1.858s

from heroprotocoljs.