Balance proof as a namedtuple instead of a tuple

[pirapira]

What this PR does

This PR changes create_balance_proof()s return value from a tuple to a dictionary.

Why I'm making this PR

• accessing tuple elements like balance_proof[0] balance_proof[3] was less readable than balance_proof["balance_hash"].
• in order to use **balance_proof in other function calls, balance_proof has to be a dictionary.

What's tricky about this PR (if any)

The fields of the dictionary have the same name as the parameters of create_balance_proof_countersignature().

---

Any reviewer can check these:

• If the PR is fixing a bug or adding a feature, add an entry to CHANGELOG.md.
• If the PR changed a Solidity source, run make compile_contracts and add the resulting raiden_contracts/data/contracts.json in the PR.
• If the PR is changing documentation only, add [skip ci] in the commit message so Travis does not waste time.
  • But, if the PR changes comments in a Solidity source, do not add [skip ci] and let Travis check the hash of the source.
• In Python, use keyword arguments
• Squash unnecessary commits
• Comment commits
• Follow naming conventions
  • solidityFunction
  • _solidity_argument
  • solidity_variable
  • python_variable
  • PYTHON_CONSTANT
• Follow the Signature Convention in CONTRIBUTING.md
• For each new contract
  • The deployment script deploys the new contract.
  • etherscan_verify.py runs on the new contract.
• Bookkeep
  • The gas cost of new functions are stored in gas.json.
• Solidity specific conventions
  • Document arguments of functions in natspec
  • Care reentrancy problems
• When you catch a require() failure in Solidity, look for a specific error message like pytest.raises(TransactionFailed, match="error message"):

And before "merge" all checkboxes have to be checked. If you find redundant points, remove them.

[codecov]

Codecov Report

Merging #1272 into master will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master    #1272   +/-   ##
=======================================
  Coverage   81.66%   81.66%           
=======================================
  Files          21       21           
  Lines        1456     1456           
  Branches      190      190           
=======================================
  Hits         1189     1189           
  Misses        228      228           
  Partials       39       39

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update e66852f...f21cbf8. Read the comment docs.

[LefterisJP]

Hey @pirapira I will review the PR as requested but before I do I have a suggestion to tweak what this PR is doing. If you like it please implement it, if not then I will review the PR as it already is.

In the "Why I am making this PR" section you stated:

1. accessing tuple elements like balance_proof[0] balance_proof[3] was less readable than balance_proof["balance_hash"].

I think that what's even more readable is a named tuple. But apart from readability named tuples are also very well typed and can be checked with mypy while dictionaries not so much unless we use the typed dicts.

I know this is tests-only code so that's why I am leaving it up to you to decide what to do.

So it would be something like

class BalanceProof(NamedTuple):
    balance_hash: TypeOfBalanceHash
    nonce: int
    additional_hash: TypeOfAddHash
    signature: TypeOfSignature

2. in order to use **balance_proof in other function calls, balance_proof has to be a dictionary.

You can do that with named tuples too! You just need to unpack them as a dict. So if balance_proof is a named tuple it would be: **balance_proof._asdict().

[pirapira]

@LefterisJP then I'll do namedtuple. I guess the additional change is not big.

[LefterisJP]

thank you. LGTM

[LefterisJP]

nitpick: add kwargs

[pirapira]

This will be done here, after Alderaan: #1292

[LefterisJP]

Hmm ... that's not what I mean. I meant to specify the name of the arguments such as

token_network.functions.updateNonClosingBalanceProof(
    channel_identifier=channel_identifier,
    participant1=A,
    ....
)

[pirapira]

Positional arguments cannot come after kwargs.

[LefterisJP]

Ah I just noticed *balance_proof_A._asdict().values(). All right then, disregard.