optimize launcher_id_to_p2_puzzle_hash()

[arvidn]

Purpose:

the wallet spends a fair amount of time generating these puzzle hashes repeatedly.

function	before	after	after/before
launcher_id_to_p2_puzzle_hash()	32.48%	1.47%	0.0453

Current Behavior:

launcher_id_to_p2_puzzle_hash() generates the full puzzle, just to tree-hash it.

New Behavior:

launcher_id_to_p2_puzzle_hash() generates the puzzle hash directly.

profile before:

profile after:

[coveralls-official]

Pull Request Test Coverage Report for Build 8937938907

Details

• 6 of 6 (100.0%) changed or added relevant lines in 1 file are covered.
• 44 unchanged lines in 11 files lost coverage.
• Overall coverage decreased (-0.03%) to 90.659%

---

Files with Coverage Reduction	New Missed Lines	%
chia/daemon/keychain_proxy.py	1	64.98%
chia/harvester/harvester_api.py	1	73.02%
chia/full_node/full_node_api.py	1	80.92%
chia/farmer/farmer.py	1	72.56%
chia/wallet/util/wallet_sync_utils.py	1	86.54%
chia/daemon/client.py	1	73.33%
chia/data_layer/data_layer.py	2	85.26%
chia/server/node_discovery.py	4	79.96%
chia/wallet/wallet_node.py	5	88.39%
chia/full_node/full_node.py	5	85.19%

Totals
Change from base Build 8931619722:	-0.03%
Covered Lines:	97981
Relevant Lines:	108020

---

💛 - Coveralls

[richardkiss]

Seems reasonable.

The nice thing about this sort of optimization is if it produces the correct value once, it almost certainly will work every time.

A trick that I like to do for changes like this is to calculate the value both the old and new way and assert they are the same. Then run the test suite (or run it on a wallet with a lot of these addresses). If you do this, there's a pretty high confidence it works correctly.

[richardkiss]

Seems reasonable.

[wjblanke]

aok