feat: enhance code performance

[stanleyyconsensys]

the PR is to enhance the code performance by not doing N square time complexity loop, instead using set data structure to improve to O(N) time complexity

e.g:

• in deserialize method, when execute filtering, it compare if the account address in accountDetails by using Object.keys(this.accountDetails).includes(XXX) , which is a O(n*n) time complexity, and Object.keys will create extra memory for every filter loop
  • instead, we can use using set to reduce the complexity
• the same case as migrateAccountDetails method
• in removeAccount method, it is using a redundant loop to search where the address is exist or not

[legobeat]

It looks to me like this part is actually a fix unrelated to the performance improvement - does it warrant breaking out and a test-case? Also: Do we know that this is always fine?

[stanleyyconsensys]

It looks to me like this part is actually a fix unrelated to the performance improvement - does it warrant breaking out and a test-case? Also: Do we know that this is always fine?

actually i find that is no necessary, or it will not cause any issue if any edge case happen
e.g
accounts = [acc1, acc1, acc2] , accountDetails = {acc2}

if we take a dry run on the Prev code

1. filter accounts where those not in accountDetails, accounts become [acc1, acc1]
2. acc1 not in accountDetails, acc1 add to accountDetails
3. acc1 not in accountDetails, acc1 add to accountDetails , but the content will be same
4. at the end result will become accountDetails = {acc2, acc1}

if we take a dry run on the Updated code

1. acc2 in set, skip
2. acc1 not in set, acc1 add to accountDetails , add to set
3. acc1 in set, skip
4. at the end result will become accountDetails = {acc2, acc1}

[legobeat]

LGTM!

[stanleyyconsensys]

@legobeat can you merge for me, i dont have the permission