Support importing one or more private keys with indexing

[emkman]

This is inspired by some previous open PRs: #2, #25, and #32. It enables support for importing one or more wallet addresses using private keys instead of a mnemonic. It supports a single key as a string, or an array of keys. The difference between this PRs and previous is that this fully implements the current HookedSubprovider interface and supports both the address_index and num_addresses options when passed an array. I have added updated documentation, both of this feature, and num_addresses generally, which is still undocumented in master. Also includes new and updated unit tests.

[julientregoat]

this would be really awesome

[cgewecke]

@emkman Agree with @julientregoat, this looks really nice. Will circle back here this week and pull this down for a more a careful review.

Sincere apologies for the delay in getting to it and thank you for opening.

Material to resolving the underlying problem in truffle 1003

[jsvisa]

any update on this PR?

[emkman]

I don't think @cgewecke had time to review it over the last couple of months. Now it looks like master has moved forward with new nonce provider functionality so I would have to resolve those conflicts and retest. I could possibly do that this weekend but I would want to hear from @cgewecke first to make sure the update could get looked at. Otherwise, you can pull from my branch in your package.json, I can confirm it has been running in production since May.

[cgewecke]

Hi @emkman, @jsvisa. Apologies, really dropped the ball on this. This module gets treated like a step-child and needs more concentrated engineering resources. At the moment we're really only able address issues that are the source of chronic bugs at Truffle main repo, hence the nonce PR. . .

It's always possible to use someone's fork of the wallet if it adds features that are helpful. But I feel like we need to add some infrastructure here (testing and review wise) to really process PRs that extend the wallet's functionality in part because this is a sensitive piece of the stack. For example we should have tests that actually send transactions to Infura and run through a variety of scenarios.

I will talk to the rest of the team about this.

[emkman]

Agreed, I added some tests in my branch but they are still thin overall. Totally understand the need to focus resources. Let me know if I can help in any way!

[cgewecke]

@emkman Thanks so much, really sweet of you.

[benjamincburns]

Thanks for the awesome work @emkman!

[slyfox42]

Hi guys! When will these changes be published? Really looking forward to them 😁