As much as it hurts the C hacker inside me we have settled on using
`_internal` to mark private function names that clash with a public
function of the same name.
Introduce a policy section and rename one instance, I did not grep the
codebase looking for other violations.
This came up because I had to look at what `_inner` implied when reading
the function name somewhere else.
2969b032f9 Push up the Default bound on HashEngine (Nick Johnson)
Pull request description:
Backstory for this patch is in #3084, and this might be enough to close that issue, but I think some follow up work might be required.
I took Kixunil's advice and pushed up the `Default` bound on `HashEngine` all the way to the default methods themselves on the `GeneralHash` trait. I initially placed the bound just on the assocated type `Engine` of `GeneralHash`, which is a tad cleaner and keeps the bound from "leaking" downstream into things like `Hmac` and `Hkdf` implementations. However, that restricts `GeneralHash` implementations to just unkeyed hash functions and I believe there will be value in having `Poly1305` and `SipHash24` still leverage the interface. I struggled a bit on the best spots to put these bounds (even found the syntax a little jarring at times, I found [RFC2289](https://rust-lang.github.io/rfcs/2289-associated-type-bounds.html) helpful if others do too).
Refactored the existing keyed hash function, `SipHash24`, to no longer have `Default` functions with zero-value keys. I kept the test coverage though by just hardcoding the zero-value keys over in the tests.
Refactoring the `hash_type` macro for keyed hashes was getting a little hairy, so backed off and just wrote the bare minimum for `SipHash24` inline. Once `Poly1305` lands there will be two keyed hash functions and I think it will make more sense to then generalize over them.
ACKs for top commit:
Kixunil:
ACK 2969b032f9
apoelstra:
ACK 2969b032f9
Tree-SHA512: 9ca8f8baa6d60c36627eb3564f5faafcddd0a69fed5e29965404e405d3d09cb08bec2e05f14e177b8bfa2a488efba9947c0e0c158db6b00b44c8c11869328f5a
3bd59a5497 Automated update to Github CI to rustc nightly-2024-07-31 (Update Nightly Rustc Bot)
Pull request description:
Automated update to Github CI workflow `rust.yml` by [create-pull-request](https://github.com/peter-evans/create-pull-request) GitHub action
ACKs for top commit:
tcharding:
ACK 3bd59a5497
Tree-SHA512: 713a576da3050759e186b275aa49f0106765161f4dc78283721af96dce5a6cfd5317b0493c994512ba8de0c3621687306cf31c4348b33b09848f4b27d45746d0
* The Default bound only makes sense for unkeyed hash functions which
can fire up a new engine without a key. Keyed hash functions, like
SipHash24 or Poly1305 require a secret key to be initialized and
should not implement a default engine generator.
* SipHash24 tests updated to the previous default key "0".
298b96c579 Add an extension trait for script validation (Tobin C. Harding)
c1aa33ed89 Use impl syntax instead of generic (Tobin C. Harding)
Pull request description:
Add an extension trait for the validation logic in preparation for moving the `Script` type to `primitives`.
ACKs for top commit:
Kixunil:
ACK 298b96c579
apoelstra:
ACK 298b96c579
Tree-SHA512: 6282bf7bd5657f0ec68e1369150969daf51f97dc6ff72a419fe823d60ab8a993f1e6d56b1cffa114580d388b36fe2bcbf7b9865776f98c46d68b7368168a07ee
A single trait bound can be expressed using the `impl` style. This is a
breaking change because callers can no longer use turbofish. In this
case that probably does not matter because users are likely just passing
an integer in and letting the compiler infer the type.
Done in preparation for moving logic into an extension trait so that the
functions can be parsed by the `define_extension_trait` macro.
ref: https://doc.rust-lang.org/reference/types/impl-trait.html
842fc9b24d Automated update to Github CI to rustc nightly-2024-07-28 (Update Nightly Rustc Bot)
Pull request description:
Automated update to Github CI workflow `rust.yml` by [create-pull-request](https://github.com/peter-evans/create-pull-request) GitHub action
ACKs for top commit:
tcharding:
ACK 842fc9b24d
Tree-SHA512: 3763f634bf6209207a45564ce1e7ac8ee042eb1fe7bfe32721daebcc191aaff0928dbb0a21827f83c4ac5b7a87f420b56524effd0829e9c7a678ce8fdc221167
c72069e921 Bump MSRV to 1.63 (Martin Habovstiak)
Pull request description:
The version 1.63 satisfies our requirements for MSRV and provides significant benefits so this commit bumps it. This commit also starts using some advantages of the new MSRV, namely namespaced features, weak dependencies and the ability to use trait bounds in `const` context.
This however does not yet migrade the `rand-std` feature because that requires a release of `secp256k1` with the same kind of change - bumping MSRV to 1.63 and removing `rand-std` in favor of weak dependency. (Accompanying PR to secp256k1: https://github.com/rust-bitcoin/rust-secp256k1/pull/709 )
Suggested plan:
* merge both PRs
* at some point release `hashes` and `secp256k`
* remove `rand-std` from `bitcoin`
* release the rest of the crates
ACKs for top commit:
apoelstra:
ACK c72069e921
tcharding:
ACK c72069e921
Tree-SHA512: 0b301ef8145f01967318d3ed1c738d33e6cf9e44f835f3762122b460a536f926916dbd6ea39d6f80b4f95402cd845e924401e75427dbb0731ca5b12b4fa6915e
f76f68217b ci: pin cargo-semver-checks version and cron job (Jose Storopoli)
8a91015769 ci: harden zip command with -n (Jose Storopoli)
4edd504cb6 ci: pin stable in semver-checks and updates weekly (Jose Storopoli)
1948995443 ci: semver-check for non-additive cargo features (Jose Storopoli)
b5180732e6 io: add not_unwind_safe as PhantomData (Jose Storopoli)
Pull request description:
Closes#3001.
Closes#3023.
## Summary
Adds new CI Job that checks for semver breaks in `no-default-features` against `all-features`.
A fail would represent that we are inserting non-additive cargo features.
Credits to @Kixunil for the amazing idea.
This PR does:
1. fix a non-additive feature in `bitcoin-io`;
1. move the current `semver-checks` to `semver-checks-pr` since it checks if the PR is introducing classical semver-checks;
1. adds a new `semver-checks-feature`;
1. Add `-n` to `unzip` in `semver-checks`;
1. pins `rustc` stable and updates in a cron job weekly on Fri; and
1. pins `cargo-semver-checks` to `0.33.0` and updates in a cron job weekly on Sat.
## Implementation notes
We don't need nightly (and somehow it fails) but if we incorporate the ENV [`RUSTC_BOOTSTRAP=1` as hardcoded in the cargo-semver-checks codebase](50b93599df/src/rustdoc_cmd.rs (L110C17-L110C39)) and run `cargo` with the stable toolchain it works.
To get the `cargo-semver-checks` version we use the [`crates.io` API](https://crates.io/api/v1/crates/cargo-semver-checks) along with some quite intelligible `jg` magic (NOTE: `jq` is available in GH's `ubuntu-latest` image).
ACKs for top commit:
Kixunil:
ACK f76f68217b
apoelstra:
ACK f76f68217b At some point we should pull the stable-version to the root with the nightly-version and use it everywhere; but can be in a separate PR
Tree-SHA512: 9949350fb7d14853ed9b24c43c1b8c28d1178cc623fd4bb1a71b40f0e32a56052596f40ab55b1c26d17f4f4538a7a321b34cae2a65c110ed1939cc5cf0fc55ce
353af24ffe Set `diff=rust` for `*.rs` in `.gitattributes` (Martin Habovstiak)
Pull request description:
This slightly improves the diffs shown by git commands.
ACKs for top commit:
tcharding:
utACK 353af24ffe
apoelstra:
ACK 353af24ffe
Tree-SHA512: 2b74efa362943e91823e0ad03efaa92f577ab73aabb0e0eecc28cbe1841d2e4a3ce14e94d2a6664e472f7bc5a2d6174f848ee2bd2dbb54a1b992b7874254c51d
feef34fdea Make ScriptBuf::p2wpkh_script_code stand alone (Tobin C. Harding)
Pull request description:
We would like to move the `Script` type to `primitives` without moving any key stuff, including pubkey hashes. We may later, before releasing `primitives`, move the `WPubkeyHash` at which time this patch can be reverted or re-implemented on `ScriptBuf`.
The `ScriptBuf::p2wpkh_script_code` function does not take `self` as a parameter but it does return `Self` - this can trivially be made into a standalone function.
Make `ScriptBuf::p2wpkh_script_code` a standalone function.
ACKs for top commit:
Kixunil:
ACK feef34fdea
apoelstra:
ACK feef34fdea
Tree-SHA512: 9de43bb274480b85d328ddd4cb6d79b501a525837307ca399dea1c3d39e9a7c0d8719dbb0e5d852629351d2b1fe09ce0244df82c53ff9ef2342be85a71887329
98fe6179db Use "unfinalized" in header of Midstate (Tobin C. Harding)
86de586898 Use const to construct Midstate (Tobin C. Harding)
dcb18bfa7a Add length to sha256::Midstate (Tobin C. Harding)
db0502d3cd Use third person in rustdoc (Tobin C. Harding)
34dd95f909 Debug Midstate forwards (Tobin C. Harding)
1d0e70b1da Add regression test for Midstate debug output (Tobin C. Harding)
ca823945fc Manually implement AsRef (remove Borrow) (Tobin C. Harding)
7dc68b62e9 Remove serde from sha256::Midstate (Tobin C. Harding)
07e8e5d3a6 Stop using macro for Midstate (Tobin C. Harding)
37b54dd54c Move from_midstate function (Tobin C. Harding)
9efe4cea9d Move impl block under struct (Tobin C. Harding)
5941008733 Import sha256t in docs builds (Tobin C. Harding)
Pull request description:
"overhaul the midstate API" might be a better description. PR is broken up into many patches to the extreme to make sure all API the changes are easily noticed.
Resolve: #2918
ACKs for top commit:
Kixunil:
ACK 98fe6179db
apoelstra:
ACK 98fe6179db
Tree-SHA512: 1384cdeea52f6a0d5b41bcfdc9127f9b983f57bb4b15c7749af2a4a40388598a9fe1ab07d0a6507ffd2f38983459f9dd400c5d3f927f6d37c1f9d80a5c7f9a6e
a7fa237d32 githooks: Added post-merge githook to check if hooks changed (Ryan Breen)
Pull request description:
Referring back to [this post](https://github.com/rust-bitcoin/rust-bitcoin/pull/2980#discussion_r1673140172) in #2980, there was discussion that a post-merge githook should be added to check if the user's copied githooks have changed.
In order to distinguish between a githook that a user installed themselves and a rust-bitcoin githook, I chose to add a tag at the beginning of the project's hooks.
ACKs for top commit:
Kixunil:
ACK a7fa237d32
apoelstra:
ACK a7fa237d32 I have never used `git pull` but I also have not ever used githooks, so I guess this is good
Tree-SHA512: 90314fea81a93e5a2c698fad4837f7661dd90f9fb2ecd015cbe6472fbef8647e06a1595e2aab4f557a1cc770050b828d430eabda04500954c0e4c9995098237c
The version 1.63 satisfies our requirements for MSRV and provides
significant benefits so this commit bumps it. This commit also starts
using some advantages of the new MSRV, namely namespaced features, weak
dependencies and the ability to use trait bounds in `const` context.
This however does not yet migrade the `rand-std` feature because that
requires a release of `secp256k1` with the same kind of change - bumping
MSRV to 1.63 and removing `rand-std` in favor of weak dependency.
82a0656b57 Direct link to ./CONTRIBUTING.md (Joao Leal)
Pull request description:
For convenience.
ACKs for top commit:
tcharding:
ACK 82a0656b57
Kixunil:
ACK 82a0656b57
Tree-SHA512: ef2e0941dfb5d9c341be2f8f80286e8d8d654728262a6324a8bfdef81cfa62cc1b3f864b8926d90205b9d1deac4a7805cf70fe4a9ebd68dc58200586cec8a204
29b213daca Move validation module to consensus_validation (Tobin C. Harding)
Pull request description:
The `consensus` module is currently doing two things, validation and encoding. These two things are orthogonal.
Move the `consensus::validation` module to `consensus_validation`. Remove the function re-exports from `consensus`.
This was originally discussed here: https://github.com/rust-bitcoin/rust-bitcoin/issues/2779
ACKs for top commit:
Kixunil:
ACK 29b213daca
apoelstra:
ACK 29b213daca
Tree-SHA512: 3bd0e43c220b0d89a47e9df0e0c92b776ccc65f5f60d57f413db834acc8e86269379bc9fdd688f8c4f0138db22f8eb8983770afa2d7d53d51acf063f2302121c
991c73cdf9 primitives: Re-export everything from units (Tobin C. Harding)
7a44908aee primitives: Use wildard re-export in locktimes (Tobin C. Harding)
d1c876eda5 Remove rustfmt attribute (Tobin C. Harding)
e3d9376a9b units: Remove serde re-export (Tobin C. Harding)
Pull request description:
Re-export `units` types by doing:
- Patch 1: Remove the public re-export of `serde` so that it does not shadow the private one in `primitives`
- Patch2: Fix formatting (remove attribute and format)
- Patch 3 and 4: Use wildcard re-exports
ACKs for top commit:
Kixunil:
ACK 991c73cdf9
apoelstra:
ACK 991c73cdf9
Tree-SHA512: dd71aa97fe5718a1229bfe26b776a72ebdd745a15a0e8ce8dae421c3a4c49f0f1e27d676056b24c35318cc3643a2ce712b4a69cc09f42b7c43259680bea931f3
33d0a95b8c Automated update to Github CI to rustc nightly-2024-07-24 (Update Nightly Rustc Bot)
Pull request description:
Automated update to Github CI workflow `rust.yml` by [create-pull-request](https://github.com/peter-evans/create-pull-request) GitHub action
ACKs for top commit:
tcharding:
ACK 33d0a95b8c
Tree-SHA512: a33ed623396222f73f145187297a7f6f44532790328d4d6d6821067e6d5fa3a21d7d285e420aec5101af9e5b54c778f12812cfff418bc1526d63000f007e3c26
We would like to move the `Script` type to `primitives` without moving
any key stuff, including pubkey hashes. We may later, before releasing
`primitives`, move the `WPubkeyHash` at which time this patch can be
reverted or re-implemented on `ScriptBuf`.
The `ScriptBuf::p2wpkh_script_code` function does not take `self` as a
parameter but it does return `Self` - this can trivially be made into a
standalone function.
Make `ScriptBuf::p2wpkh_script_code` a standalone function.
3b15ef6d27 Fix rustdocs in `blockdata` (Jamil Lambert, PhD)
Pull request description:
Following up on the [comment](https://github.com/rust-bitcoin/rust-bitcoin/pull/2646#discussion_r1548848611) in #2646 the rustdocs formatting was fixed in `blockdata`.
ACKs for top commit:
Kixunil:
ACK 3b15ef6d27
tcharding:
ACK 3b15ef6d27
Tree-SHA512: 509aa2b8ae559f5a7f8230c016c0866f468cc147773238e226b9a975784d7a424c41a5d966ddcf9b3f8f8d4fe197a4f272b5777c61f3cc53051803abb520b95e
We recently decided to make everything in `units` available at the same
path as in `primitives` and not re-export the actual `units` crate.
Re-export everything from the `units` crate root at the `primitives`
crate root using a wildcard.
We recently decided to use wildcard re-exports when re-exporting things
from an identically named module in a sub crate, ie. to mirror the
directory structure structure ala core/std.
In the `primitives::locktime` modules re-export everything from the
`units::locktime` modules using a wildcard.
There are no other use statements so we do not need to keep the public
ones separate with a `rustfmt` attribute.
Remove the attribute and run the formatter.
We re-export to help users keep their dependencies in sync but `serde`
is at `v1.0` so this is not really a problem.
Remove the public re-export of `serde` (and with current MSRV we don't
need the `extern crate` at all).
dbb5917128 Automated update to Github CI to rustc nightly-2024-07-21 (Update Nightly Rustc Bot)
Pull request description:
Automated update to Github CI workflow `rust.yml` by [create-pull-request](https://github.com/peter-evans/create-pull-request) GitHub action
ACKs for top commit:
tcharding:
ACK dbb5917128
Tree-SHA512: 6f9a7939ff2eea6d07338cb75e393e319b2d38009dbad35e408bd23534e6240ef8c95a268b2316f7b81045eba7031e4269918a3841fe3353da966605ea752407
0beefda7ea CONTRIBUTING: error/expect messages (Jose Storopoli)
Pull request description:
- adds an item that all error messages
should be lower case, except for proper nouns and variable names.
- adds a section on expect messages,
following discussion in #3019 and #3053.
Closes#3053.
ACKs for top commit:
tcharding:
ACK 0beefda7ea
Kixunil:
ACK 0beefda7ea
apoelstra:
ACK 0beefda7ea
Tree-SHA512: b924245e5d18d4f627d4998073422003def789c87502bca121d5db443f5bbe387bfe951664458c9495a6ea0995b6ccf6e28d1784e19dfb876d7eddf3c6209727
ab581a90f8 Remove re-export of ParseIntError (Tobin C. Harding)
Pull request description:
In d242125 I claimed that `ParseIntError` was somehow special, I no longer thing this is the case. As we pin down the re-export policy (for errors and other types) it is hard if we have one non-typical re-export.
We have https://github.com/rust-bitcoin/rust-bitcoin/issues/3068 to discuss the policy, for now just remove the unusual re-export.
ACKs for top commit:
Kixunil:
ACK ab581a90f8
shinghim:
ACK ab581a90f8
apoelstra:
ACK ab581a90f8
Tree-SHA512: 5ac4123aeb27c8cee78e5760f21e70be8035d526ba7e14e72759cba27f98b51cc2cba9b2bf0eeb99e0f6b7210ec4a750986bb6c5dc0725ed892730fdec8a7e06
54c30556a2 Move params to network module (Tobin C. Harding)
045a661ebe Create network directory (Tobin C. Harding)
Pull request description:
Discussed in #2779. Patch one moves `network.rs` to `network/mod.rs`, and patch 2 moves the `params` module over there.
ACKs for top commit:
apoelstra:
ACK 54c30556a2 Yeah, this seems like a good place for it
Kixunil:
ACK 54c30556a2
Tree-SHA512: 134813419db21323d303d465b12fcbf37fd61dc1baf3305e156d324eafd822379e63dede02877ee99dce41540193a29e6e13acd13f9121f3e2fe11096524aa5e
The `consensus` module is currently doing two things, validation and
encoding. These two things are orthogonal.
Move the `consensus::validation` module to `consensus_validation`.
Remove the function re-exports from `consensus`.
04bcb7950d Automated update to Github CI to rustc nightly-2024-07-17 (Update Nightly Rustc Bot)
Pull request description:
Automated update to Github CI workflow `rust.yml` by [create-pull-request](https://github.com/peter-evans/create-pull-request) GitHub action
ACKs for top commit:
tcharding:
ACK 04bcb7950d
Tree-SHA512: 2b2a02fd5eb5dfcc1e8b1551ef66180d76019655262a5eebeb6e849885db5e37c495cd9d1108ce550b9401c446a5abf2ca30508eb3d1139a807188b668d232ef
Tobin C. Harding
merge-script
Update Nightly Rustc Bot
Nick Johnson
Shing Him Ng
Martin Habovstiak
Jose Storopoli
Ryan Breen
Joao Leal