feat: add generic Digest256 and Digest512 structs by Farukest · Pull Request #777 · 0xMiden/crypto

Farukest · 2026-01-14T22:11:07Z

Introduces generic Digest256 and Digest512 structs to reduce code duplication across hash function implementations.

Changes:

New digest.rs module: Contains reusable Digest256 (32-byte) and Digest512 (64-byte) digest types with full trait implementations (Default, Deref, From/Into, Serializable, Deserializable, serde support, hex conversion)
Refactored SHA-2 module: Sha256Digest and Sha512Digest are now type aliases to the generic types, removing ~130 lines of duplicate code
Updated tests: Modified to use ::from() constructor syntax

This lays the groundwork for other hash functions (Blake3, Keccak) to potentially reuse these generic digest types in the future.

Closes #697

Checklist before requesting a review

Repo forked and branch created from next according to naming convention.
Commit messages and codestyle follow conventions.
Relevant issues are linked in the PR description.
Tests added for new functionality.
Documentation/comments updated according to changes.

Introduce reusable generic digest types for binary hash functions to reduce code duplication. - Add digest.rs with Digest256 (32-byte) and Digest512 (64-byte) types - Refactor Sha256Digest and Sha512Digest to type aliases - Remove ~130 lines of duplicate struct implementations - Update tests to use new constructor syntax Closes 0xMiden#697

- Remove square bracket links for Digest256 and Digest512 in module docs to fix rustdoc broken-intra-doc-links errors - Add PR 0xMiden#777 entry to CHANGELOG.md

plafer · 2026-01-15T16:08:34Z

+    /// # Safety
+    /// This function uses unsafe code to reinterpret the slice of digests as bytes.
+    /// This is safe because `Digest256` is `#[repr(transparent)]` over `[u8; 32]`.


nit: the safety argument is an implementation detail, and is redundant with the existing one on line 58, so I would just remove it

plafer · 2026-01-15T16:27:29Z

+#[cfg_attr(feature = "serde", derive(serde::Deserialize, serde::Serialize))]
+#[cfg_attr(feature = "serde", serde(into = "String", try_from = "&str"))]
+#[repr(transparent)]
+pub struct Digest256([u8; DIGEST256_BYTES]);


We could instead have just a single const-generic Digest struct defined as

// defaults to 32 bytes pub struct Digest<const N: usize = 32>([u8; N]); // example usage pub type Sha256Digest = Digest; pub type Sha512Digest = Digest<64>;

I believe this is also what was alluded to in #692 (review).

Address review feedback: - Replace separate Digest256 and Digest512 structs with single const-generic Digest<N> - Add Digest256 and Digest512 as type aliases for backward compatibility - Remove redundant safety documentation comment - Update CHANGELOG entry

Farukest · 2026-01-15T19:17:42Z

nit: the safety argument is an implementation detail, and is redundant with the existing one on line 58, so I would just remove it

Thank you for detaileed guide @plafer . All done for now. Kindly needs check.

plafer

LGTM (modulo small suggestions in this review), thank you!

Will still wait for @huitseeker to take a look before we merge though

Farukest · 2026-01-20T23:14:46Z

Done @plafer 👍

huitseeker

This is generally in good shape, but leaves a few simplifications on the table, see comments.

huitseeker · 2026-01-21T16:34:38Z

+    }
+
+    /// Converts a slice of digests into a contiguous byte slice.
+    pub fn digests_as_bytes(digests: &[Digest<N>]) -> &[u8] {


This prepare_merge function is duplicated in sha2, keccak, and blake modules. Consider moving it here and exporting it so all three can share it.

huitseeker · 2026-01-21T16:34:44Z

+// ================================================================================================
+
+#[cfg(test)]
+mod tests {


Nit: I'd add a test_memory_layout_assumptions test here to verify size_of and align_of match the inner array. The unsafe code in digests_as_bytes and prepare_merge relies on this.

huitseeker · 2026-01-21T16:34:51Z

-}
+///
+/// This is a type alias to the generic [`Digest256`] type.
+pub type Sha256Digest = Digest256;


The same pattern works for Keccak256Digest. You can make it a type alias to Digest256 and remove ~80 lines of duplicate code from keccak/mod.rs.

…ests - Move prepare_merge helper function from sha2, blake, and keccak modules to the shared digest module - Add test_memory_layout_assumptions test to verify Digest<N> has same size and alignment as [u8; N] - Update keccak module to use Digest256 type alias instead of custom Keccak256Digest struct - Remove redundant DIGEST_BYTES constant in favor of DIGEST256_BYTES

Farukest · 2026-01-22T09:30:51Z

The same pattern works for Keccak256Digest. You can make it a type alias to Digest256 and remove ~80 lines of duplicate code from keccak/mod.rs.

All done @huitseeker . kindly need your check again 👍

plafer

LGTM after latest changes, thank you!

bobbinth

Looks good! Thank you! I left a few comments inline.

bobbinth · 2026-01-25T03:10:27Z

 #[cfg_attr(feature = "serde", derive(serde::Deserialize, serde::Serialize))]
 #[cfg_attr(feature = "serde", serde(into = "String", try_from = "&str"))]
 #[repr(transparent)]
 pub struct Blake3Digest<const N: usize>([u8; N]);


Maybe for another PR, but would be good to use the Digest struct for BLAKE3 digest as well. For this, we may need to add the 24-byte variant there.

- Move changelog entry to 0.22.0 section - Replace prepare_merge function with Digest::digests_as_bytes() - Make digest module crate-private (pub(crate)) - Use type-specific digests_as_bytes in merge implementations

Farukest · 2026-01-25T14:26:20Z

@bobbinth thanks. handled all but last, for another PR.

Remove brackets from doc links to `Digest256` and `Digest512` since these types are not public, causing rustdoc to fail with `rustdoc::private-intra-doc-links` errors.

bobbinth

All looks good! Thank you!

Farukest and others added 2 commits January 15, 2026 00:37

fix: resolve broken doc links and add CHANGELOG entry

10f47ec

- Remove square bracket links for Digest256 and Digest512 in module docs to fix rustdoc broken-intra-doc-links errors - Add PR 0xMiden#777 entry to CHANGELOG.md

plafer reviewed Jan 15, 2026

View reviewed changes

Merge branch 'next' into feat/generic-digest-structs

6ce04bc

plafer approved these changes Jan 20, 2026

View reviewed changes

Comment thread miden-crypto/src/hash/digest.rs Outdated

Comment thread miden-crypto/src/hash/digest.rs Outdated

Comment thread miden-crypto/src/hash/digest.rs Outdated

refactor: use named constants for digest type aliases

3c8e396

Merge branch 'next' into feat/generic-digest-structs

746e9ed

bobbinth requested a review from huitseeker January 21, 2026 06:27

huitseeker reviewed Jan 21, 2026

View reviewed changes

plafer reviewed Jan 22, 2026

View reviewed changes

Comment thread miden-crypto/src/hash/sha2/mod.rs Outdated

refactor: implement HasherExt for Sha512 and remove outdated comments

b6ab471

plafer approved these changes Jan 23, 2026

View reviewed changes

Merge branch 'next' into feat/generic-digest-structs

a68a82f

bobbinth reviewed Jan 25, 2026

View reviewed changes

refactor: address PR review comments

892dc58

- Move changelog entry to 0.22.0 section - Replace prepare_merge function with Digest::digests_as_bytes() - Make digest module crate-private (pub(crate)) - Use type-specific digests_as_bytes in merge implementations

fix: remove intra-doc links to private digest types

efcacbd

Remove brackets from doc links to `Digest256` and `Digest512` since these types are not public, causing rustdoc to fail with `rustdoc::private-intra-doc-links` errors.

bobbinth approved these changes Jan 25, 2026

View reviewed changes

bobbinth merged commit 282940b into 0xMiden:next Jan 25, 2026
24 checks passed

plafer mentioned this pull request Jan 26, 2026

Use Digest struct for BLAKE3 digest #801

Closed

Uh oh!

Conversation

Farukest commented Jan 14, 2026

Changes:

Checklist before requesting a review

Uh oh!

plafer Jan 15, 2026

Choose a reason for hiding this comment

Uh oh!

plafer Jan 15, 2026

Choose a reason for hiding this comment

Uh oh!

Farukest commented Jan 15, 2026

Uh oh!

plafer left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Farukest commented Jan 20, 2026

Uh oh!

huitseeker left a comment

Choose a reason for hiding this comment

Uh oh!

huitseeker Jan 21, 2026

Choose a reason for hiding this comment

Uh oh!

huitseeker Jan 21, 2026

Choose a reason for hiding this comment

Uh oh!

huitseeker Jan 21, 2026

Choose a reason for hiding this comment

Uh oh!

Farukest commented Jan 22, 2026

Uh oh!

Uh oh!

plafer left a comment

Choose a reason for hiding this comment

Uh oh!

bobbinth left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

bobbinth Jan 25, 2026

Choose a reason for hiding this comment

Uh oh!

Farukest commented Jan 25, 2026

Uh oh!

bobbinth left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants