The `ToHex` trait was replaced by either simple `Display`/`LowerHex`
where appropriate or `DisplayHex` from `bitcoin_internals` which is
faster.
This change replaces the usages and removes the trait.
3e520f9094 Use hex from internals rather than hashes (Martin Habovstiak)
Pull request description:
`bitcoin-internals` contains a more performant implementation of hex encoding than what `bitcoin_hashes` uses internally. This switches the implementations for formatting trait implementations as a step towards moving over completely.
The public macros are also changed to delegate to inner type which is technically a breaking change but we will break the API anyway and the consuers should only call the macro on the actual hash newtypes where the inner types already have the appropriate implementations.
Apart from removing reliance on internal hex from public API this reduces duplicated code generated and compiled. E.g. if you created 10 hash newtypes of SHA256 the formatting implementation would be instantiated 11 times despite being the same.
To do all this some other changes were required to the hex infrastructure. Mainly modifying `put_bytes` to accept iterator (so that `iter().rev()` can be used) and adding a new `DisplayArray` type. The iterator idea was invented by Tobin C. Harding, this commit just adds a bound check and generalizes over `u8` and `&u8` returning iterators.
While it may seem that `DisplayByteSlice` would suffice it'd create and initialize a large array even for small arrays wasting performance. Knowing the exact length `DisplayArray` fixes this.
Another part of refactoring is changing from returning `impl Display` to return `impl LowerHex + UpperHex`. This makes selecting casing less annoying since the consumer no longer needs to import `Case` without cluttering the API with convenience methods.
ACKs for top commit:
tcharding:
ACK 3e520f9094
apoelstra:
ACK 3e520f9094
Tree-SHA512: 62988cec17550ed35990386e572c0d32dc7107e1c36b7c9099080747e15167e6d66497fb300178afbd22481c0360a6b7a1228fd09402d4ce5d295a8594c02aa6
This fixes several API bugs:
* Use `TryFrom` instead of `From` for fallible conversions
* Move byte conversion methods from `impl_array_newtype` to
`impl_bytes_newtype`
* Add missing trait impls like `AsRef`, `Borrow`, their mutable versions
and infallible conversions from arrays
Closes#1336
`bitcoin-internals` contains a more performant implementation of hex
encoding than what `bitcoin_hashes` uses internally. This switches the
implementations for formatting trait implementations as a step towards
moving over completely.
The public macros are also changed to delegate to inner type which is
technically a breaking change but we will break the API anyway and the
consuers should only call the macro on the actual hash newtypes where
the inner types already have the appropriate implementations.
Apart from removing reliance on internal hex from public API this
reduces duplicated code generated and compiled. E.g. if you created 10
hash newtypes of SHA256 the formatting implementation would be
instantiated 11 times despite being the same.
To do all this some other changes were required to the hex
infrastructure. Mainly modifying `put_bytes` to accept iterator (so that
`iter().rev()` can be used) and adding a new `DisplayArray` type. The
iterator idea was invented by Tobin C. Harding, this commit just adds a
bound check and generalizes over `u8` and `&u8` returning iterators.
While it may seem that `DisplayByteSlice` would suffice it'd create and
initialize a large array even for small arrays wasting performance.
Knowing the exact length `DisplayArray` fixes this.
Another part of refactoring is changing from returning `impl Display` to
return `impl LowerHex + UpperHex`. This makes selecting casing less
annoying since the consumer no longer needs to import `Case` without
cluttering the API with convenience methods.
Pnicking on oversized slice is useful to catch errors in code that's
supposed to know the exact sizes but this is undesirable in code that
doesn't. These two methods help with handling the case when `buf.len()`
is not known upfront.
All the types that we define with `impl_array_newtype` are
`Copy` so the correct conversion method to get the underlying byte array
is `to_bytes`. We currently provide `into_bytes` as well as `to_bytes`,
with one of them calling `clone` - this is unnecessary and against
convention.
- Remove `into_bytes` and for `to_bytes` just return the inner field.
- Add a method that causes build to fail if `Copy` is not implemented.
`impl_array_newtype` is an internal macro, move it to a new, ever so
meaningfully named, `macros` module.
Use `#[macro_export]`, no other changes to the macro.
1bf885550e Redesign `hex::BufEncoder` to accept owned arrays (Martin Habovstiak)
Pull request description:
Not being able to create an owned `BufEncoder` prevented returning it from functions which need to allocate the buffer on stack. Such is the case in WIP serde via consensus serialzation.
This change refactors `OutBytes` to be unsized, adds an `AsOutBytes` trait and uses that one instead of `Into` to perform the conversion.
Closes#1270
This is meant as potentially mergeable demonstration. Interestingly, it was easier than I expected.
ACKs for top commit:
tcharding:
Except for the question about the `out_bytes` module, ACK 1bf885550e
apoelstra:
ACK 1bf885550e
Tree-SHA512: 39bc48e54ae0e66b988ba5ad6ea7ecbe03e5f4c71792df0f8e2b03aa2e97e2c0fac1cb03e84ecac12ec6f13649554b57e1000710c34c638d17d9bb575d0ac0a1
Not being able to create an owned `BufEncoder` prevented returning it
from functions which need to allocate the buffer on stack. Such is the
case in WIP serde via consensus serialzation.
This change refactors `OutBytes` to be unsized, adds an `AsOutBytes`
trait and uses that one instead of `Into` to perform the conversion.
This adds a build script that automatically activates the cfg if the
version is sufficient. We don't make API changes here, just improve
debugging by enabling `track_caller`.
There's a `hex` module in `bitcoin_hashes` which is a bit out of place
and not very fast - it passes each *digit* through dynamic dispatch not
only adding overhead but also not allowing `String` to make better
allocations.
This change adds basic support for hex encoding using a stack-allocated
buffer which minimizes the overhead of dynamic dispatch. It also
provides a new `DisplayHex` trait designed to replace `ToHex` found in
`bitcoin_hashes`.
Add a new crate `bitcoin-internals` to be used for internal code needed
by multiple soon-to-be-created crates.
Add the `write_err` macro to `bitcoin-internals`, nothing else.
This patch uses a `path` dependency which means `rust-bitcoin` cannot be
released in its current state, will need to be changed once we release
the `bitcoin-internals` crate on `crates.io`.
Martin Habovstiak
Andrew Poelstra
Tobin C. Harding