e8f923ec86
chain: move Memo into transaction
2020-08-17 11:46:34 -07:00
312c66264a
chain: extract sprout code into sprout module.
2020-08-17 11:46:34 -07:00
e06f59ee21
chain: extract sapling code to sapling module
2020-08-17 11:46:34 -07:00
c5a8cb0c91
chain: create primitives module.
...
This contains definitions of primitive types used in other structures
and re-exports of component libraries.
2020-08-17 11:46:34 -07:00
196e841cd9
chain: rename LightClientRootHash to RootHash
...
It's not accurate to call it a LightClientRootHash, because it's not
always a root has for a light client -- sometimes it's a different kind
of root hash.
2020-08-17 11:46:34 -07:00
b296d1e2a3
chain: move Block into a leaf module.
...
This might make things a little easier to rearrange. In the future it
would probably be good to change to block::Hash, block::Header, etc.
2020-08-17 11:46:34 -07:00
948b067808
chain: move Network, NetworkUpgrade to parameters
...
Also, avoid using star-imports of the enum variants, which pollutes the
namespace.
2020-08-17 11:46:34 -07:00
64d9d55992
chain: organize block tests
...
This moves the tests::generate module into the block tests. Because
this whole set of changes is just focused on reorganization, the
generate code is unchanged, but in the future, the code should be
rewritten as a collection of proptest strategies.
2020-08-17 11:46:34 -07:00
855b89dec4
chain: create a new work module for proof-of-work
...
This extracts the `difficulty` module from `block` and the
`equihash_solution` module from the crate root. The PoW calculations
are significantly more complicated than the other block code and pretty
dissimilar from it, so it makes more sense to create a common proof of
work module.
The `EquihashSolution` and `EQUIHASH_SOLUTION_SIZE` are renamed to
`equihash::Solution` and `equihash::SOLUTION_SIZE` and imported that
way, except in `block/header.rs`, to avoid a conflict with the
`equihash` crate. In the future it would be better to encapsulate the
equihash solution check into the `equihash::Solution` type so that
callers only need to import our `work::equihash`.
The test organization leaves a little to be desired but I think that
this can be improved as we fill out the proof of work implementation.
2020-08-17 11:46:34 -07:00
dad6340cd3
chain: move BlockHeight into block
2020-08-17 11:46:34 -07:00
5f71bcd0d1
chain: move LockTime into transaction
2020-08-17 11:46:34 -07:00
8e9a239687
chain: clean transaction tests module.
...
This moves the transaction strategies to `arbitrary`, to live with the
other strategy impls (`Arbitrary` is just a default type-associated
strategy), splits the proptests into a new `prop` module, and splits the
test vector checks into a `vector`s module.
This ensures that we keep code in leaf modules and only have
organization (use statements) in non-leaf modules.
2020-08-17 11:46:34 -07:00
a35604aef3
chain: move amount to a top-level module.
2020-08-17 11:46:34 -07:00
cee7d0b8eb
chain: move serde_helpers into serialization.
2020-08-17 11:46:34 -07:00
b36fe8f937
chain: move sha256d to serialization module.
...
This extracts the SHA256d code from being split across two modules and puts it
in one module, under serialization.
The code is unchanged except for three deleted tests:
* `sha256d_flush` in `sha256d_writer` (not a meaningful test);
* `transactionhash_debug` (constructs an invalid transaction hash, and the
behavior is tested in the next test);
* `decode_state_debug` (we do not need to test the Debug output of
DecodeState);
2020-08-17 11:46:34 -07:00
9f31e551c9
chain: split serialization.rs into files
...
The code is unchanged except for removing an unneeded clippy annotation in proptests.rs.
2020-08-17 11:46:34 -07:00
0d1f56ad2f
chain: remove utils module
...
A catch-all utils module can really easily slip into being a place to stash
miscellaneous functions that don't really belong anywhere in particular.
2020-08-17 11:46:34 -07:00
4bbefeb68e
Implement Add for Work ( #894 )
...
* feature: Implement CompactDifficulty to Work
* fix: Test Work on all difficulty test cases
* fix: Add Bitcoin test vectors for difficulty
* feature: Cumulative Work
* impl Add, AddAssign for Work
* derive Default for Work
* test cumulative work
* fix: comments
* fix: More comments
* Fix comment links
* Remove unused import
Co-authored-by: Deirdre Connolly <durumcrustulum@gmail.com>
2020-08-14 20:48:56 -04:00
616d82faaf
Use the jubjub::AffinePoint in the canonical, encoded little-endian byte form in
...
Pedersen tests
Keeps the librustzcash test vectors with their big-endian byte order
as comments, for greppability.
2020-08-14 02:04:12 -04:00
a2c9011e06
Impl ZcashDeserialize for jubjub::Fq, use in transaction deserialization
2020-08-14 02:04:12 -04:00
1255cec6a4
Serialize Amount to bytes using byteorder::LittleEndian
2020-08-14 02:04:12 -04:00
646b506e7a
Whoops forgot to add sprout::NoteTreeRootHash
2020-08-14 02:04:12 -04:00
4914b80cdb
Add a TODO about caching Pedersen hash generator points
2020-08-14 02:04:12 -04:00
f753e63f3b
Clarify docstring on NoteCommitment::new()
2020-08-14 02:04:12 -04:00
437f40d3c8
Refine JoinSplit::anchor to be a sprout::NoteTreeRootHash type
2020-08-14 02:04:12 -04:00
5a5882700c
Tidy JoinSplit nullifiers deserialization
2020-08-14 02:04:12 -04:00
6b13ce3e83
Add Sapling EphemeralPublicKey type that wraps jubjub::AffinePoint
2020-08-14 02:04:12 -04:00
52a10d2837
Add note about whether sign checks in Amount -> Fr conversion matter
...
Co-authored-by: Henry de Valence <hdevalence@hdevalence.ca>
2020-08-14 02:04:12 -04:00
75cad3bb0a
Impl TryFrom's for Diversifier and use those to construct Sapling NoteCommitments
2020-08-14 02:04:12 -04:00
c8771ef620
Defer to From/Into [u8; 32] vs Zcash(De)Serialize for sprout::NoteCommitments
2020-08-14 02:04:12 -04:00
619afffa16
Use convenience Reader/Writer methods to only use From/Into for nullifer serialization
...
And thus remove duplicative Zcash(De)Serialization impls
2020-08-14 02:04:12 -04:00
13b6ff1c65
Add 32byte and 64 bytes convenience writer methods to WriteExt
2020-08-14 02:04:12 -04:00
e8fdd0e1f3
Impl TryFrom for Commitments, call from impl ZcashDeserialize
2020-08-14 02:04:12 -04:00
5d13880461
Remove Zcash(De)Serialize impls for *Commitments
2020-08-14 02:04:12 -04:00
014afd8e4a
Move Pedersen hash functions to their own submodule
2020-08-14 02:04:12 -04:00
be7ea200c8
Accept an Amount<NonNegative> for the value arg of a ValueCommitment constructor
2020-08-14 02:04:12 -04:00
e024e43896
Support nullifier derivation from borrowed Spending, NullifierDeriving keys
2020-08-14 02:04:12 -04:00
40b5a73a15
impl From<&NullifierDerivingKey>, use that to derive Sapling Nullifier
2020-08-14 02:04:12 -04:00
a8625b7b84
Remove empty doc comment
2020-08-14 02:04:12 -04:00
a3530b0eed
Move pedersen_hash() to go with the other Sapling hash functions, for now
2020-08-14 02:04:12 -04:00
0242df30cb
A space
2020-08-14 02:04:12 -04:00
9d9491c4e8
Index at 1, not 0, the main loop of PedersenHashToPoint, to match spec math
2020-08-14 02:04:12 -04:00
e5c07721b4
Typo fix
...
Co-authored-by: Henry de Valence <hdevalence@hdevalence.ca>
2020-08-14 02:04:12 -04:00
ca81e48843
Remove an extra line
2020-08-14 02:04:12 -04:00
60ae9d74e7
Typo
...
Co-authored-by: Henry de Valence <hdevalence@hdevalence.ca>
2020-08-14 02:04:12 -04:00
fe45c6402a
Remove clippy::unit_arg allowance
2020-08-14 02:04:12 -04:00
2942bad6ff
Make diversify_hash() private again
2020-08-14 02:04:12 -04:00
5a249f8da5
A space
...
Co-authored-by: Henry de Valence <hdevalence@hdevalence.ca>
2020-08-14 02:04:12 -04:00
5ba453c2a1
Remove unused SaplingNoteTreeRootHash types re: Heartwood
2020-08-14 02:04:12 -04:00
68871c5bd4
From<Nullifier> for [u8; 32]
2020-08-14 02:04:12 -04:00
b2584c4a8f
A space
2020-08-14 02:04:12 -04:00
0c299972ec
Clarify docstring about the Jubjub prime-order subgroup on generate_trapdoor()
2020-08-14 02:04:12 -04:00
f816761a5c
The pedersen hash to point test vectors now pass
2020-08-14 02:04:12 -04:00
425275ad35
Fix exclusive range, shift up enumerate() indexes with map()
2020-08-14 02:04:12 -04:00
be22ef64e5
At least fewer M_i() rounds for test cases
2020-08-14 02:04:12 -04:00
41c8076dab
Add test vectors that aren't being consumed right now
2020-08-14 02:04:12 -04:00
14d502a54b
Compute pedersen hash to point in scalar field
2020-08-14 02:04:12 -04:00
8ececebea7
A space
...
Co-authored-by: teor <teor@riseup.net>
2020-08-14 02:04:12 -04:00
3b01dc50ef
Use unicode scalar field characters instead of LaTeX
...
Co-authored-by: Daira Hopwood <daira@jacaranda.org>
2020-08-14 02:04:12 -04:00
1d494a2b7a
Docstring wrap
2020-08-14 02:04:12 -04:00
e22afa3c8f
Generate uniformly random scalar w/ Fr::from_bytes_wide()
...
Which reduces mod r always.
2020-08-14 02:04:12 -04:00
c46cda920f
Add primitive merkle_crh_sapling function
2020-08-14 02:04:12 -04:00
10a9aa5844
Add prf_nf and use in From impl to derive Sapling Nullifier
2020-08-14 02:04:12 -04:00
eb731a6e2f
Add note commitment tree Position type
2020-08-14 02:04:12 -04:00
65f8f2c613
Add mixing_pedersen_hash
...
Tidy constructors for NoteCommitment
2020-08-14 02:04:12 -04:00
ed888b38ef
Correctly generate a sapling NoteCommitment from the NoteCommitmentRandomness
2020-08-14 02:04:12 -04:00
b6385ca36f
Move commitment types into commitments module
...
Docs, tidy
2020-08-14 02:04:12 -04:00
4ccec84768
Add explicit pedersen_hash function
2020-08-14 02:04:12 -04:00
83c41869e1
Improve docs
2020-08-14 02:04:12 -04:00
caef323f9b
pedersen_hash_to_point() works, with submethods using bitvec
...
Also tidy, document
2020-08-14 02:04:12 -04:00
919557c20c
My BitSlice doesn't have a known size at compile time?
...
error[E0277]: the size for values of type cannot be known at compilation time
2020-08-14 02:04:12 -04:00
579f1ed134
Flesh out windowed_pedersen_commitment_r
2020-08-14 02:04:12 -04:00
9d87f30d18
Start of, but currently unfinished, Sapling note commitment / Windowed Pedersen Commitment
2020-08-14 02:04:12 -04:00
f64e0c4bc5
Sapling value commitment, half done Sapling note commitment
2020-08-14 02:04:12 -04:00
823b06b219
Break out subtypes into modules, including arbitrary's
2020-08-14 02:04:12 -04:00
dfb10ab10f
Add NoteCommitment.extract_u()
2020-08-14 02:04:12 -04:00
33730a05cb
Do not confuse a NoteCommitment for U(NoteCommitment)
2020-08-14 02:04:12 -04:00
0f46a9b6a8
NoteCommitments, renames, ValueCommitment, strategies, stub commit()
2020-08-14 02:04:12 -04:00
fa053e2b0b
Derive a Sprout nullifier from a nullifier seed (rho) and spending key
2020-08-14 02:04:12 -04:00
0d618a3abf
Start treestate module
2020-08-14 02:04:12 -04:00
28a7420079
Remove unused imports
2020-08-14 02:04:12 -04:00
15899bc6d3
More Note refinements
2020-08-14 02:04:12 -04:00
2c285985c8
Move nullifiers to respective note modules, pull in existing types from keys
2020-08-14 02:04:12 -04:00
8b78a55c71
Start work implementing sprout note commitment function
2020-08-14 02:04:12 -04:00
0e21a70b88
feature: Implement CompactDifficulty to Work ( #838 )
...
* Implement CompactDifficulty to Work
* Add Bitcoin test vectors for difficulty
2020-08-14 09:48:15 +10:00
07917421cb
Correct coinbase check ( #898 )
...
* chain: add Transaction::is_coinbase()
This matches the check in zcashd/src/primitives/transaction.h:682
(CTransaction::IsCoinBase).
* chain: correct Block::is_coinbase_first
This matches zcashd/src/main.cpp:3968-3974 in CheckBlock.
Previously, the check allowed the first transaction to have multiple coinbase inputs.
* chain: return slices from Transaction::inputs()/outputs()
They're slices internally so we might as well just expose them that way.
2020-08-13 14:04:43 -07:00
36093b4721
Tweak light client root hash definition.
...
This changes the `light_client_root_hash` field to `light_client_root_bytes` to
hint that it's unparsed, and makes it public to match the rest of the
`BlockHeader` fields. The `LightClientRootHash` serialization methods are
hidden from the public API, so that the `LightClientRootHash` has to be
constructed by the method on the `Block`.
2020-08-11 19:13:50 -04:00
f8a24f9c11
Rename script fields to lock/unlock scripts. ( #866 )
...
This is slightly clearer and more correct than the previous pk_script.
2020-08-10 11:49:38 -07:00
7afd76f5fb
feature: Add light_client_root_hash(network) to Block
...
* use the right variant in LightClientRootHash::from_bytes()
* make block.header.light_client_root_hash pub(super)
* add tests for LightClientRootHash and block.light_client_root_hash
2020-08-10 18:54:42 +10:00
ee6f0de14d
refactor: Move NetworkUpgrade to zebra-chain
2020-08-10 18:54:42 +10:00
c4eb136426
feature: Add a LightClientRootHash type
2020-08-10 18:54:42 +10:00
53ce01fd23
fix: Rename the light client hash field
2020-08-10 18:54:42 +10:00
a987394782
Add KaTeX to rendered docs. ( #832 )
...
This adds a single test formula to the `zebra-chain` docs. We can remove it
after merging this and running it through CI to confirm that it works.
2020-08-05 17:34:30 -07:00
d49c11ee70
fix: Allow PROPTEST_CASES to override the lower block case default
2020-08-05 02:07:05 -07:00
ac7a4ae517
fix: Speed up the block round trip proptest
...
Reduce the number of cases run by the block round trip proptest, to
speed up the Zebra tests.
2020-08-04 22:35:01 -07:00
78b5bf5e9a
refactor: Move the difficulty tests into their own file
2020-08-04 15:24:52 -04:00
efb9313460
fix: Resolve a type ambiguity in a proptest
...
For some reason, implementing
`PartialOrd<BlockHeaderHash> for ExpandedDifficulty` (or its converse)
makes this unrelated code ambiguous.
2020-08-04 12:29:07 +10:00
c8b502a5da
feature: Compare ExpandedDifficulty and BlockHeaderHash
2020-08-04 12:29:07 +10:00
b7fac7b3bc
feature: Make ExpandedDifficulty use U256 internally
...
Adds a dependency on the primitive-types crate.
Also adds custom hex debug formatting for compact and expanded
difficulties.
2020-08-04 12:29:07 +10:00
c04d1b7b8f
doc: Consensus-critical difficulty encodings
2020-08-04 12:29:07 +10:00
Henry de Valence
teor
Deirdre Connolly
Jane Lusby