Add transaction-builder suport for TZE-bearing transactions.

[nuttycom]

Closes zcash/zcash#4506.

[codecov-commenter]

Codecov Report

Merging #12 into zip-tzes will decrease coverage by 30.24%.
The diff coverage is 10.73%.

@@              Coverage Diff              @@
##           zip-tzes      #12       +/-   ##
=============================================
- Coverage     64.62%   34.38%   -30.25%     
=============================================
  Files           109       97       -12     
  Lines         15428    11663     -3765     
=============================================
- Hits           9970     4010     -5960     
- Misses         5458     7653     +2195     

Impacted Files	Coverage Δ
zcash_extensions/src/consensus/transparent.rs	0.00% <0.00%> (ø)
zcash_extensions/src/transparent/demo.rs	0.00% <0.00%> (ø)
zcash_primitives/src/consensus.rs	11.32% <0.00%> (-0.68%)	⬇️
zcash_primitives/src/extensions/transparent.rs	0.00% <0.00%> (ø)
zcash_primitives/src/transaction/components.rs	28.32% <0.00%> (-5.76%)	⬇️
zcash_primitives/src/transaction/mod.rs	41.00% <ø> (-1.50%)	⬇️
zcash_primitives/src/transaction/builder.rs	43.69% <8.62%> (-7.03%)	⬇️
zcash_primitives/src/transaction/sighash.rs	79.74% <54.54%> (-9.96%)	⬇️
zcash_primitives/src/transaction/tests.rs	100.00% <100.00%> (ø)
bellman/src/groth16/generator.rs	0.00% <0.00%> (-90.37%)	⬇️
... and 69 more

---

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 dbe713a...70e66c5. Read the comment docs.

[str4d]

Concept ACK.

[str4d]

I think the approach we will end up taking is to serialize the TZE data in place of the transparent data, so this would become Option<InputData> with:

enum InputData {
    Transparent {
        index: usize,
        script_pubkey: Script,
        value: Amount,
    },
    Tze {
        index: usize,
        precondition: Vec<u8>,
        value: Amount,
    },
}

and we would domain-separate the transparent and TZE data types in the signature hash.

[nuttycom]

This is the one piece that I think we should perhaps pair on; does this require a change to the ZIP?

[str4d]

Yes, the ZIP will need to be updated to describe how it changes ZIP 243.

[str4d]

Canopy has a consensus branch ID now, but let's leave these as placeholders so that anything attached to the Canopy NU in this branch is explicitly domain-separated from anything that might end up on mainnet.

[str4d]

There has to be a nicer way to handle this, but we can figure it out in a future refactor of the Rust crates.