rust-bitcoin

Use TxInEncoder for TxIn encoding

TxIn has a TxInEncoder newtype for consensus encoding TxIn
objects, but the Encode implementation currently uses a bare
Encoder3.

This fails to properly encapsulate the inner state of the
TxIn type and its encoding behavior. Switch the implementation
to use TxInEncoder.

Use BlockEncoder for Block encoding

Block has a BlockEncoder newtype for consensus encoding Block
objects, but the Encode implementation currently uses a bare
Encoder3.

This fails to properly encapsulate the inner state of the
Block type and its encoding behavior. Switch the implementation
to use BlockEncoder.

Update API files

Update API files

Close FromHexError

The FromHexError is currently unreleased but is necessary for hex
functionality. Since it is not ready for 1.0 as an enum, it should be
enclosed in a private struct as is typical for error types from units,
primitives, etc.

Rename FromHexError enum to FromHexErrorInner and add a new
FromHexError struct to privately wrap the inner enum.

consensus_encoding: Adjust ConsensusHex impl in serde_as_consensus

The ConsensusHex Display impl in serde_as_consensus currently uses the
EncoderByteIter. While this functions correctly, it is inefficient as
it repeatedly calls current_chunk internally to fetch successive chars
as well as writing to the Formatter one char at a time. Instead, the
DisplayHex trait can be used to dump entire chunks at a time.

Adjust ConsensusHex Display impl in serde_as_consensus to write using
DisplayHex rather than char-by-char with EncoderByteIter.

Add ExactSizeEncoder impl for Option<T: ExactSizeEncoder>

Currently, there exists an Encoder impl for Option<T: Encoder>. Since
ExactSizeEncoder can be useful, there is no reason not to include an
ExactSizeEncoder impl for Option<T: ExactSizeEncoder> also.

Add ExactSizeEncoder impl for Option<T: ExactSizeEncoder>.

Replace uses of bitcoin::consensus with consensus_encoding

Throughout the bitcoin code, there are various uses of the old encoding
traits used to implement functionality. With the introduction of the
new encoding and decoding implementations, all of these can be replaced
with the new implementations to remove the dependency on bitcoin.

Adjust fuzz targets to use consensus_encoding

Various fuzz targets use the old consensus code for serialisation and
deserialisation. Since we want to remove this consensus code, the fuzz
targets must be updated.

Replace all uses of old consensus code in fuzz targets with
consensus_encoding.

Replace uses of consensus in tests

As with the uses of consensus::encode in the main code, the uses in the
tests and doctest examples can also be replaced.

Replace use of con_serde in sighash test

The bip_341_sighash_tests test in sighash relies on con_serde to hex
decode a transaction in the test data. Since we want to remove old
consensus logic, con_serde will not be available. As such, this test
should be modified to use consensus_encoding's decode_from_hex.

Add hex feature to consensus_encoding dependency in bitcoin.
Replace use of con_serde in bip_341_sighash_tests test with
decode_from_hex.

Remove old consensus code from examples

Various examples in bitcoin use old consensus code but should instead
make use of the modern consensus_encoding functionality. This also
provides users with a demonstration on how to use the new logic as
opposed to the old.

Replace all uses of consensus in examples with new consensus_encoding
functionality.

Add XOnlyPublicKey::verify

At present, users who wish to verify a message with a Schnorr signature
need to use the secp256k1 library directly. Since Legacy/FullPublicKey
already provide verify functions for ECDSA, XOnlyPublicKey should
provide a similar function for Schnorr.

Add XOnlyPublicKey::verify to Schnorr verify messages with a
taproot::Signature.

Move TapTweak trait to taproot-primitives

The TapTweak trait is necessary for WitnessProgram::p2tr and thus
Address::p2tr. With from_key_and_merkle_root moved back to
taproot-primitives, this trait can now be moved alongside it and
and later used in addresses.

Move TapTweak trait to taproot-primitives.

primitives: Implement ExactSizeEncoder for WitnessEncoder

WitnessEncoder wraps Encoder2<CompactSizeEncoder, BytesEncoder>. Both inner
encoders implement ExactSizeEncoder, but WitnessEncoder only manually
implemented Encoder and was therefore missing an ExactSizeEncoder impl for
no reason.

The manual struct definition and Encoder impl is replaced with the
encoder_newtype_exact! macro, consistent with the other encoder newtypes in
the codebase. This adds the missing ExactSizeEncoder implementation.

Also add a test asserting that the encoder's reported length matches the
actual encoded length, including the CompactSize one-byte boundary for both
element length and element count.

hashes: update API files

hashes: Add with_input method to HashEngine

this is the same as input but takes and returns self,
which allows calls to be chained without a mutable variable.

Replace dyn fmt::Write with fmt::Formatter

In Rust, the dyn keyword for a function parameter provides generic-like
functionality for a parameter. However, unlike impl or a generic, dyn
enables dynamic dispatch for the function, preventing inlining by the
compiler. Since all such uses of dyn fmt::Write are used with formatter
anyway, the dynamic dispatch cost can be eliminated by replacing the
dyn param with a fmt::Formatter type.

Replace dyn fmt::Write with fmt::Formatter in all non-deprecated
functions.

fuzz: rename old_bitcoin to bitcoin_0_32

fuzz: migrate 0.32.x fuzz targets to master

Centralize all fuzz infra on the master branch, including targets from
the LTS 0.32.x branch. This should make the infra and targets easier to
maintain, and they will probably be run more often in practice here.

The 0.32.x targets were migrated from hongfuzz to master's libfuzzer.

consensus_encoding: always enable alloc for serde

Since there is no "collect_bytes" exposed in serde, an allocation is
unavoidable for non-human-readable serialization. Make the alloc
dependency explicit in the feature declaration.

Add recent 0.32.x releases to relevant changelogs

The `[YANKED]` tag is following the keep-a-changelog convention.

hashes: fix tag name in sha256t_tag macro docs

The sha256t_tag! macro referred to `$hash_name`, which does not exist, so every
generated tag showed "The tag used for [`$hash_name`]" in its docs.

eg: https://docs.rs/bitcoin-taproot-primitives/0.1.0/bitcoin_taproot_primitives/

Add tests to verify 0 return for overflow

Following the previous change to Target::from_compact, the tests should
be updated to confirm the 0 for overflow, and also to check the
boundary values for the overflow.

Add and adjust pow unit tests to check the overflow logic if
Target::from_compact.

Return 0 on Target::from_compact overflow

Currently, the Target::from_compact function treats a negative mantissa
as a failure mode and returns a Target value of zero. This lines up
with Core's SetCompact, which sets a pfNegative flag in the same case.
In Core, a pfOverflow flag is similarly set in overflow cases, and in
every case where a negative flag is treated as a failure, the overflow
is too. As such, our implementation should also return Target::ZERO in
the overflow case as it does for the negative case.

Return Target::ZERO for mantissa/exponent values which overflow in
Target::from_compact.