chore: optimize build

[0xmad]

Explanation

Get rid of clean scripts and replace source maps to get rid of eval code + improve build speed a little bit.

Details are bellow:

• Remove clean scripts
• Update source maps

More Information

Fixes #107

Screenshots/Screencaps

N/A

Manual Testing Steps

Just try to run npm run dev and npm run build to check if extension loads properly.

Pre-Merge Checklist

• PR template is filled out
• IF this PR fixes a bug, a test that would have caught the bug has been added
• PR is linked to the appropriate GitHub issue
• PR has been added to the appropriate release Milestone

+ If there are functional changes:

• Manual testing complete & passed
• "Extension QA Board" label has been applied

PR template source from github.com/MetaMask

[0xisk]

Well I still cant approve this change since the eval code is still there not in the dist/js/backgroundPage.js script but in the dist/js/injected.js script in both builds the prod and the dev. That's because of moving the proof generation part from the background to the contentscript.

[0xmad]

@iskdrews what's the exact error there? I can't see any eval errors and the demo works as expected.

[0xisk]

@iskdrews what's the exact error there? I can't see any eval errors and the demo works as expected.

There isn't any error, because the eval code is just moved the dist/js/injected.js still not sure if that would be accepted this way to be approved from chrome reviewers or not.

[AtHeartEngineer]

Ya I'm not seeing any errors on the cryptkeeper side, except for a ObjectMultiplex - orphaned data for stream "publicConfig" warning.

[0xmad]

@iskdrews I think it's ok if we have generated code with new Function(...). It doesn't use any user input and it's just hardcoded.
I think it was an error on background script before because of inline source maps. For injected, nothing is changed. It probably had the same code before moving proof generation (it's common generated code for most of the files).

[AtHeartEngineer]

This is building and working good for me. I am not understanding the injection being an issue @iskdrews, could you explain that to me when you can?

[0xisk]

This is building and working good for me. I am not understanding the injection being an issue @iskdrews, could you explain that to me when you can?

I think all is good now, as @0xmad said it is just an hardcoded value and it doesn't take any inputs so no security risk.

[0xisk]

Ya I'm not seeing any errors on the cryptkeeper side, except for a ObjectMultiplex - orphaned data for stream "publicConfig" warning.

There is an issue for tracking resolving this warning already, but not resolved yet.
#111