gix: allow sha256-only builds by 10ne1 · Pull Request #2627 · GitoxideLabs/gitoxide

Adrian Ratiu (10ne1) · 2026-05-28T13:00:46Z

What

This is part of #281 and fixes the following open item in that ticket:

 Make SHA256-only builds compile where supported by feature flags.

Evidence: cargo check -p gix --no-default-features --features sha256 fails because
gix/src/config/cache/incubate.rs and gix/src/config/tree/sections/core.rs still name gix_hash::Kind::Sha1 
unconditionally.

It's 4 commits in total after addressing all feedback (including both Codex and Claude feedback). The most important concerns are:

To allow such builds
To behave like the git.git project when it comes to repository format version and extensions.

AI disclosure
This PR was written with the help of Claude Code, but was not vibe-coded. I'm not a fan of vibe-coding. I used AI to better understand the codebase myself (asked it a lot of questions), review both my code and the code generated by AI, used it especially to detect bugs and corner-cases and suggest fixes, however I very carefully reviewed & edited each code/comment lines, the commit messages and so on, until I was myself satisfied with the result.

Three call sites in gix/src/config referenced gix_hash::Kind::Sha1 unconditionally, which fails to compile when the Sha1 variant is feature-gated out of gix_hash::Kind. Replace the hardcoded Kind::Sha1 with the Default impl which is feature-aware (sha1 when available, otherwise sha256), so this preserves Sha1 as the default for normal builds while also letting sha256-only builds compile. For the core.rs Abbrev validator use Kind::longest() which accepts the widest range of valid lenghts. This only enfources a shape bound, the authoritative repo-hash check happens elsewhere (see comment). Add a Justfile check to verify the gix build. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com> Signed-off-by: Adrian Ratiu <adrian.ratiu@collabora.com>

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 62de1ecc75

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "Codex (@codex) review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "Codex (@codex) address that feedback".

Copilot

Pull request overview

Note

Copilot was unable to run its full agentic suite in this review.

This PR tightens repository object format handling to better match git behavior, particularly around extensions.objectFormat vs repository format versions, and improves build/test coverage for sha256-only configurations.

Changes:

Reject extensions.objectFormat on repository format version 0, and add a dedicated config error for that case.
Add fixtures + tests to assert v0 repositories with extensions.objectFormat are rejected.
Adjust config validation/build checks to better support sha256-only builds.

Reviewed changes

Copilot reviewed 6 out of 7 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
`justfile`	Adds a `cargo check` entry for `--features sha256` to exercise sha256-only builds.
`gix/tests/gix/repository/open.rs`	Adds a regression test ensuring `extensions.objectFormat` on v0 repos is rejected.
`gix/tests/fixtures/make_config_repos.sh`	Creates new fixture repos with `extensions.objectFormat` set on v0 format repos.
`gix/src/config/tree/sections/core.rs`	Makes `core.abbrev` validation hash-kind-agnostic via `Kind::longest()`.
`gix/src/config/mod.rs`	Introduces `ObjectFormatRequiresV1` config error variant.
`gix/src/config/cache/incubate.rs`	Implements new `objectFormat`/format-version logic and adds `legacy_object_hash()`.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot

Pull request overview

Copilot reviewed 6 out of 8 changed files in this pull request and generated 3 comments.

A missing extensions.objectFormat means legacy Sha1. In sha256-only builds, Kind::default() is Sha256, so such repos were silently mislabeled as Sha256. Resolve the implicit case to Sha1 when supported, else error out to avoid any mislabeling. Also reject extensions.objectFormat when repositoryFormatVersion is 0, matching git, which treats it as an invalid v1-only extension. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com> Signed-off-by: Adrian Ratiu <adrian.ratiu@collabora.com>

Christoph Rüßler (cruessler)

Just two minor suggestions, otherwise this is good to go from my side!

In comments we usually stick to SHA1 and SHA256 when referring to the hash kind. We also tend to use backticks for things like objectFormat and gix_hash::Kind::Sha1 (the latter could even be turned into a proper link).

Add fixtures with sha1 and sha256 variants and an open() test asserting that a version-0 repository setting extensions.objectFormat is rejected. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Signed-off-by: Adrian Ratiu <adrian.ratiu@collabora.com>

legacy_object_hash() has a feature-gated split: it returns Sha1 when the build supports it and an error otherwise. Add a unit test asserting the branch for the current build, and run the gix lib tests under --no-default-features --features sha256 so the error path is actually exercised in CI. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Signed-off-by: Adrian Ratiu <adrian.ratiu@collabora.com>

Copilot

Pull request overview

Copilot reviewed 6 out of 8 changed files in this pull request and generated no new comments.

Sebastian Thiel (Byron)

Thanks so much, and sorry for the long wait! Also thanks to Christoph Rüßler (@cruessler) for the initial review, it's much appreciated.

The most notable change I made here is to… change the prefixes of the commit message. There is know that we only do conventional commit messages for the purpose of controlling the changelog. In this case, a gix: prefix would probably have done nothing, but it might be confusing to look at. One of the commits is a genuine fix: and I made it so.

Sebastian Thiel (Byron) · 2026-06-18T05:11:49Z

+    #[error(
+        "extensions.objectFormat is a v1-only extension, but the repository format version is 0; \
+         set core.repositoryFormatVersion=1 to use it, or remove extensions.objectFormat to fall back to the default Sha1 format (if supported by this build)"
+    )]
+    ObjectFormatRequiresV1,


A wonderful error message!

Sebastian Thiel (Byron) · 2026-06-18T05:16:26Z

+        // objectFormat is a "v1-only" extension: git refuses to operate on a version-0 repo that
+        // sets it, even for sha1 (unlike grandfathered extensions like preciousObjects, which v0
+        // still honours). This rejection was introduced in git 2.29.0 (2020). We match it.


I like the research that went into this.

Sebastian Thiel (Byron) · 2026-06-18T05:24:51Z

+    #[cfg(not(feature = "sha1"))]
+    {
+        Err(Error::UnsupportedObjectFormat { name: "sha1".into() })
+    }


I was a bit stuck here and decided to go with something else.

Initially, instead of saying this isn't possible, I asked myself what would Git do if it would allow to not compile SHA-1 in. Then I think it should just be like SHA-1 never happened, hence SHA-256 is returned.

However, Git doesn't do that and it gix tooling that relies on SHA-256 only builds will simply have to declare it properly so both Git and the tool can open it.

An interesting journey, leading to nothing but this comment 😁.

Sebastian Thiel (Byron) · 2026-06-18T05:34:49Z

 }
+
+#[cfg(test)]
+mod tests {


This test felt redundant in the light of compile-time safety and docs of legacy_object_hash_reflects_build_features.

Sebastian Thiel (Byron) · 2026-06-18T05:36:14Z

+    git config extensions.objectFormat "$format"
+    git config core.repositoryFormatVersion 0
+  )
+done


I removed the 'dance' in favor of something more direct, for better or worse.

Sebastian Thiel (Byron) · 2026-06-18T05:47:34Z

Adrian Ratiu (@10ne1) I wasn't able to push my changes and review commit back into this branch which is an inconvenience, so I have to close this PR and let it be superseded by #2651.
Thanks for your understanding.

Adrian Ratiu (10ne1) · 2026-06-18T07:10:37Z

Sebastian Thiel (@Byron) I enabled Allow edits and access to secrets by maintainers for all my branches going forward. Thanks!

chatgpt-codex-connector Bot reviewed May 28, 2026

View reviewed changes

Comment thread gix/src/config/cache/incubate.rs Outdated

Adrian Ratiu (10ne1) mentioned this pull request May 28, 2026

Sha 256 support #281

Open

50 tasks

Christoph Rüßler (cruessler) reviewed Jun 3, 2026

View reviewed changes

Comment thread gix/src/config/cache/incubate.rs Outdated

Adrian Ratiu (10ne1) force-pushed the dev/aratiu/allow-sha256-only-builds branch from bfaa2bc to 99c75c8 Compare June 8, 2026 15:01

Copilot AI review requested due to automatic review settings June 8, 2026 15:01

Copilot AI reviewed Jun 8, 2026

View reviewed changes

Comment thread gix/src/config/cache/incubate.rs

Comment thread gix/src/config/mod.rs Outdated

Comment thread gix/tests/gix/repository/open.rs Outdated

Adrian Ratiu (10ne1) force-pushed the dev/aratiu/allow-sha256-only-builds branch from 99c75c8 to 047458b Compare June 8, 2026 15:16

Adrian Ratiu (10ne1) requested review from Copilot and Christoph Rüßler (cruessler) June 8, 2026 15:25

Adrian Ratiu (10ne1) force-pushed the dev/aratiu/allow-sha256-only-builds branch from 047458b to 7a3e783 Compare June 8, 2026 15:44

Copilot AI reviewed Jun 8, 2026

View reviewed changes

Comment thread gix/src/config/mod.rs

Comment thread gix/src/config/cache/incubate.rs

Comment thread gix/tests/fixtures/make_config_repos.sh Outdated

Adrian Ratiu (10ne1) force-pushed the dev/aratiu/allow-sha256-only-builds branch from 7a3e783 to edf9793 Compare June 8, 2026 16:19

Christoph Rüßler (cruessler) reviewed Jun 13, 2026

View reviewed changes

Comment thread gix/tests/fixtures/make_config_repos.sh Outdated

Comment thread gix/tests/gix/repository/open.rs Outdated

Adrian Ratiu (10ne1) and others added 2 commits June 15, 2026 16:39

Adrian Ratiu (10ne1) force-pushed the dev/aratiu/allow-sha256-only-builds branch from edf9793 to 1b3ca28 Compare June 15, 2026 13:44

Copilot AI review requested due to automatic review settings June 15, 2026 13:44

Copilot AI reviewed Jun 15, 2026

View reviewed changes

Adrian Ratiu (10ne1) requested a review from Christoph Rüßler (cruessler) June 16, 2026 16:23

Sebastian Thiel (Byron) self-assigned this Jun 18, 2026

Sebastian Thiel (Byron) approved these changes Jun 18, 2026

View reviewed changes

Sebastian Thiel (Byron) mentioned this pull request Jun 18, 2026

allow sha256 only builds #2651

Merged

1 task

Sebastian Thiel (Byron) closed this Jun 18, 2026

Adrian Ratiu (10ne1) deleted the dev/aratiu/allow-sha256-only-builds branch June 18, 2026 11:22

Uh oh!

Uh oh!

Conversation

Adrian Ratiu (10ne1) commented May 28, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

chatgpt-codex-connector Bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Christoph Rüßler (cruessler) left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Sebastian Thiel (Byron) left a comment

Choose a reason for hiding this comment

Uh oh!

Sebastian Thiel (Byron) Jun 18, 2026

Choose a reason for hiding this comment

Uh oh!

Sebastian Thiel (Byron) Jun 18, 2026

Choose a reason for hiding this comment

Uh oh!

Sebastian Thiel (Byron) Jun 18, 2026

Choose a reason for hiding this comment

Uh oh!

Sebastian Thiel (Byron) Jun 18, 2026

Choose a reason for hiding this comment

Uh oh!

Sebastian Thiel (Byron) Jun 18, 2026

Choose a reason for hiding this comment

Uh oh!

Sebastian Thiel (Byron) commented Jun 18, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Adrian Ratiu (10ne1) commented Jun 18, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Adrian Ratiu (10ne1) commented May 28, 2026 •

edited

Loading

Christoph Rüßler (cruessler) left a comment •

edited

Loading

Sebastian Thiel (Byron) commented Jun 18, 2026 •

edited

Loading