Uh oh!
There was an error while loading. Please reload this page.
Add DataModel.RebuildCommitHashes() for post-bootstrap rehash - #67
Closed
myieye wants to merge 2 commits into
Closed
Add DataModel.RebuildCommitHashes() for post-bootstrap rehash#67myieye wants to merge 2 commits into
myieye wants to merge 2 commits into
Conversation
Previously the duplicate-commit check ran before acquiring the repo lock, leaving a TOCTOU window where two concurrent Add calls could both see the commit as missing and proceed to insert it. https://claude.ai/code/session_01TziA3LymLmdxXxUgXywq4B
Contributor
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (3)
🚧 Files skipped from review as they are similar to previous changes (2)
📝 WalkthroughWalkthroughThis PR adds ChangesCommit Hash Rebuild Feature
Estimated code review effort🎯 2 (Simple) | ⏱️ ~12 minutes Possibly related PRs
Suggested reviewers
Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Contributor
Actionable comments posted: 0 |
Walks every commit in chain order and recomputes Hash/ParentHash against the current Commit.Ids. The intended caller is FwLite's SQL-template- based project creation path: it seeds commits whose Ids get substituted at apply time, leaving the persisted hashes (computed against template- source Ids) stale. Calling this before the first sync brings the chain back in line. Local-only by contract — once any peer has observed a commit's hash, a rebuild would break sync. Takes the repo lock and wraps in a transaction to match other write paths. Test covers the recovery scenario: corrupt a mid-chain hash, watch AddChange reject further writes, then verify RebuildCommitHashes restores every Hash to GenerateHash(parentHash) and the chain accepts writes again. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
myieyeforce-pushed
the
rebuild-commit-hashes-api
branch
from
May 29, 2026 06:51
af27140 to
bf9698dComparemyieye added a commit
to sillsdev/languageforge-lexbox
that referenced
this pull request
May 29, 2026
Templated projects are local-only today (UploadProject still throws), so the per-project Guid_N hydration we shipped was protecting against a threat that can't manifest until cross-client template sync exists. The LexBox composite (ProjectId, Id) PK already covers the loud failure case (silent commit-Id collision on server insert). Static templates are concretely simpler: no projectId plumbing through ApplyAsync, no morph-types-commit-Id placeholder substitution, no rehash trigger, no Harmony submodule API surface needed. Easy to revert if either of the deferred triggers in TEMPLATE-FOLLOWUPS.md fires: - Sharing templated projects across clients - A LexBox feature that keys on Commit.Id alone What changed: - ProjectTemplate.ApplyAsync drops the projectId param and HydrateGuids/ placeholder-substitution/rehash steps; only WS placeholders are substituted at apply time. - template.sql is baked: Guid_N tokens replaced with deterministic UUIDv5s (namespace Guid.Empty), morph-types placeholder resolved to MorphTypesSeedCommitId(Guid.Empty), hash chain recomputed. - GenerateTemplate (dev tool) no longer tokenizes — drops TemplateGuidScrubbing.cs. - BakeStaticTemplate added as a skipped dev tool for re-baking if the template is regenerated and Guid_N tokens reappear. - CreateProjectRequest gains SeedMorphTypes (default true). Download path sets false: morph-types arrive via the server sync that runs inside AfterCreate, so the pre-seed would add a redundant commit that the inbound sync would just dup. Other call sites keep the default (import path needs morph-types before AfterCreate runs). - Harmony submodule pin reverted to develop's commit (96a75b26b); the RebuildCommitHashes API on PR sillsdev/harmony#67 is no longer called from lexbox. PR stays open as a known-working primitive for if/when ② in TEMPLATE-FOLLOWUPS.md fires. - TemplatedProjects_HaveDisjointCommitIds test removed (now they intentionally share Ids). - CreateProjectFromTemplateAppliesRequestedIdentity drops its per-project-Id morph-types-commit-Id assertion (no longer pertinent). TEMPLATE-FOLLOWUPS.md lays out the staged simplifications and the events that should trigger each next step. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
myieye added a commit
to sillsdev/languageforge-lexbox
that referenced
this pull request
Jun 1, 2026
Templated projects are local-only today (UploadProject still throws), so the per-project Guid_N hydration we shipped was protecting against a threat that can't manifest until cross-client template sync exists. The LexBox composite (ProjectId, Id) PK already covers the loud failure case (silent commit-Id collision on server insert). Static templates are concretely simpler: no projectId plumbing through ApplyAsync, no morph-types-commit-Id placeholder substitution, no rehash trigger, no Harmony submodule API surface needed. Easy to revert if either of the deferred triggers in TEMPLATE-FOLLOWUPS.md fires: - Sharing templated projects across clients - A LexBox feature that keys on Commit.Id alone What changed: - ProjectTemplate.ApplyAsync drops the projectId param and HydrateGuids/ placeholder-substitution/rehash steps; only WS placeholders are substituted at apply time. - template.sql is baked: Guid_N tokens replaced with deterministic UUIDv5s (namespace Guid.Empty), morph-types placeholder resolved to MorphTypesSeedCommitId(Guid.Empty), hash chain recomputed. - GenerateTemplate (dev tool) no longer tokenizes — drops TemplateGuidScrubbing.cs. - BakeStaticTemplate added as a skipped dev tool for re-baking if the template is regenerated and Guid_N tokens reappear. - CreateProjectRequest gains SeedMorphTypes (default true). Download path sets false: morph-types arrive via the server sync that runs inside AfterCreate, so the pre-seed would add a redundant commit that the inbound sync would just dup. Other call sites keep the default (import path needs morph-types before AfterCreate runs). - Harmony submodule pin reverted to develop's commit (96a75b26b); the RebuildCommitHashes API on PR sillsdev/harmony#67 is no longer called from lexbox. PR stays open as a known-working primitive for if/when ② in TEMPLATE-FOLLOWUPS.md fires. - TemplatedProjects_HaveDisjointCommitIds test removed (now they intentionally share Ids). - CreateProjectFromTemplateAppliesRequestedIdentity drops its per-project-Id morph-types-commit-Id assertion (no longer pertinent). TEMPLATE-FOLLOWUPS.md lays out the staged simplifications and the events that should trigger each next step. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
myieye added a commit
to sillsdev/languageforge-lexbox
that referenced
this pull request
Jun 1, 2026
Templated projects are local-only today (UploadProject still throws), so the per-project Guid_N hydration we shipped was protecting against a threat that can't manifest until cross-client template sync exists. The LexBox composite (ProjectId, Id) PK already covers the loud failure case (silent commit-Id collision on server insert). Static templates are concretely simpler: no projectId plumbing through ApplyAsync, no morph-types-commit-Id placeholder substitution, no rehash trigger, no Harmony submodule API surface needed. Easy to revert if either of the deferred triggers in TEMPLATE-FOLLOWUPS.md fires: - Sharing templated projects across clients - A LexBox feature that keys on Commit.Id alone What changed: - ProjectTemplate.ApplyAsync drops the projectId param and HydrateGuids/ placeholder-substitution/rehash steps; only WS placeholders are substituted at apply time. - template.sql is baked: Guid_N tokens replaced with deterministic UUIDv5s (namespace Guid.Empty), morph-types placeholder resolved to MorphTypesSeedCommitId(Guid.Empty), hash chain recomputed. - GenerateTemplate (dev tool) no longer tokenizes — drops TemplateGuidScrubbing.cs. - BakeStaticTemplate added as a skipped dev tool for re-baking if the template is regenerated and Guid_N tokens reappear. - CreateProjectRequest gains SeedMorphTypes (default true). Download path sets false: morph-types arrive via the server sync that runs inside AfterCreate, so the pre-seed would add a redundant commit that the inbound sync would just dup. Other call sites keep the default (import path needs morph-types before AfterCreate runs). - Harmony submodule pin reverted to develop's commit (96a75b26b); the RebuildCommitHashes API on PR sillsdev/harmony#67 is no longer called from lexbox. PR stays open as a known-working primitive for if/when ② in TEMPLATE-FOLLOWUPS.md fires. - TemplatedProjects_HaveDisjointCommitIds test removed (now they intentionally share Ids). - CreateProjectFromTemplateAppliesRequestedIdentity drops its per-project-Id morph-types-commit-Id assertion (no longer pertinent). TEMPLATE-FOLLOWUPS.md lays out the staged simplifications and the events that should trigger each next step. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
myieye
commented
Jun 2, 2026
ContributorAuthor
Superseded by #68 ( |
myieye added a commit
to sillsdev/languageforge-lexbox
that referenced
this pull request
Jun 17, 2026
Templated projects are local-only today (UploadProject still throws), so the per-project Guid_N hydration we shipped was protecting against a threat that can't manifest until cross-client template sync exists. The LexBox composite (ProjectId, Id) PK already covers the loud failure case (silent commit-Id collision on server insert). Static templates are concretely simpler: no projectId plumbing through ApplyAsync, no morph-types-commit-Id placeholder substitution, no rehash trigger, no Harmony submodule API surface needed. Easy to revert if either of the deferred triggers in TEMPLATE-FOLLOWUPS.md fires: - Sharing templated projects across clients - A LexBox feature that keys on Commit.Id alone What changed: - ProjectTemplate.ApplyAsync drops the projectId param and HydrateGuids/ placeholder-substitution/rehash steps; only WS placeholders are substituted at apply time. - template.sql is baked: Guid_N tokens replaced with deterministic UUIDv5s (namespace Guid.Empty), morph-types placeholder resolved to MorphTypesSeedCommitId(Guid.Empty), hash chain recomputed. - GenerateTemplate (dev tool) no longer tokenizes — drops TemplateGuidScrubbing.cs. - BakeStaticTemplate added as a skipped dev tool for re-baking if the template is regenerated and Guid_N tokens reappear. - CreateProjectRequest gains SeedMorphTypes (default true). Download path sets false: morph-types arrive via the server sync that runs inside AfterCreate, so the pre-seed would add a redundant commit that the inbound sync would just dup. Other call sites keep the default (import path needs morph-types before AfterCreate runs). - Harmony submodule pin reverted to develop's commit (96a75b26b); the RebuildCommitHashes API on PR sillsdev/harmony#67 is no longer called from lexbox. PR stays open as a known-working primitive for if/when ② in TEMPLATE-FOLLOWUPS.md fires. - TemplatedProjects_HaveDisjointCommitIds test removed (now they intentionally share Ids). - CreateProjectFromTemplateAppliesRequestedIdentity drops its per-project-Id morph-types-commit-Id assertion (no longer pertinent). TEMPLATE-FOLLOWUPS.md lays out the staged simplifications and the events that should trigger each next step. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Adds a public
DataModel.RebuildCommitHashes()API that walks every commit in chain order and recomputesHash/ParentHashagainst the current Commit.Ids. Intended for callers that have staged commits whose Ids changed after creation (e.g. applying a Guid-substituted SQL template), so the persisted hashes can be brought back in line before any sync validates them.Local-only by contract — once any peer has observed a commit's hash, the rebuild would break sync. Takes the repo lock + transaction to match other write paths.
Also pins
Verify.EntityFrameworkto the last v13 (v14+ pulls EF Core 10 abstractions, which collide with the EF 9.0.16 SQLite provider this repo stays on per the .NET 10 bump). Regenerates theDbContextTests.VerifyModelverified snapshot to reflect EF 9 (the previously-committed file hadProductVersion: 10.0.7baked in from the EF 10 churn before the rollback).Test plan
RebuildCommitHashesTests.RebuildCommitHashes_RestoresChainAfterHashesAreCorrupted— corrupts a mid-chain hash, watches AddChange reject further writes, then verifies RebuildCommitHashes restores every Hash toGenerateHash(parentHash)and the chain accepts writes again.RebuildCommitHashesTests.RebuildCommitHashes_IsNoOpOnEmptyChain— guards against an empty-chain regression.Downstream caller is sillsdev/languageforge-lexbox PR #2281, where it replaces a JSON-roundtrip hack that fabricated a synthetic Commit just to trigger the same internal rehash path.
🤖 Generated with Claude Code
Summary by CodeRabbit
Release Notes
New Features
Bug Fixes