- Changed the default value of `sync_unscriptable` in `SyncbackRules` from `true` to `false` to better reflect intended behavior.
- Adjusted the property filter logic to align with the updated default, ensuring consistent handling of unscriptable properties.
These changes improve the clarity and functionality of the syncback feature, enhancing its overall reliability.
- Removed the `luau` dependency from `rokit.toml`.
- Updated deduplication logic to append `~2`, `~3`, etc. instead of `~1`, `~2` for naming collisions in various files, ensuring consistent suffix management.
- Adjusted comments and documentation across multiple files to reflect the new suffix handling rules.
- Added new test cases to validate the updated deduplication behavior, ensuring accurate naming for instances with collisions.
These changes enhance the clarity and functionality of the deduplication process, improving overall consistency in naming conventions.
- Introduced a new field, `stage_paths`, to the `PatchSet` struct in `patch_compute.rs` to track filesystem paths for staging.
- Updated test cases to initialize `stage_paths` alongside other fields, ensuring consistency in patch set creation.
These changes enhance the testing framework by providing additional context for staged paths in patch sets.
- Introduced 20 new batch script files (BatchScript01.luau to BatchScript20.luau) for testing purposes.
- Enhanced the `change_processor` to consolidate git staging operations, reducing the risk of index.lock races by batching paths for staging.
- Updated the `PatchSet` struct to include a new field for pre-resolved filesystem paths to stage via git add.
- Modified the API to handle batch staging more efficiently, ensuring all files are staged correctly in mixed write scenarios.
These changes improve the testing framework and optimize git integration for batch operations.
- Updated the `compute_git_metadata` function to include canonical path resolution for changed files, improving the accuracy of ID retrieval from the tree.
- Implemented logic to check both the original and canonical paths for file IDs, ensuring comprehensive coverage of file changes.
These changes enhance the robustness of git metadata handling, ensuring more reliable synchronization behavior.
- Simplified multi-line function calls in `PatchTree.lua` and `SHA1.luau` for better clarity.
- Enhanced the `init.lua` file by formatting the `PatchTree.build` function call across multiple lines.
- Updated `change_processor.rs` and `api.rs` to improve the formatting of function calls and enhance code consistency.
- Made minor adjustments in test files to streamline assertions and improve readability.
These changes contribute to a cleaner codebase and enhance overall maintainability.
- Updated the audit documentation to clarify the handling of Ref properties, specifying the distinction between path-based and legacy ID-based attributes.
- Improved descriptions in the snapshot middleware and change batching sections to reflect recent changes in method locations and functionality.
- Added details on the new matching algorithm for forward sync, including scoring constants and its impact on instance pairing.
These changes improve the clarity of the documentation and ensure it accurately reflects the current implementation of syncback and snapshot processes.
- Renamed the `rejectWrongProtocolVersion` function to `rejectWrongVersion` for clarity.
- Enhanced version mismatch handling by providing detailed error messages when the plugin and server versions do not match.
- Updated the expected server version format in the configuration to include the patch version, ensuring accurate version checks.
These changes improve error reporting and version management in the plugin's API context.
- Introduced new tests for git-based sync direction defaults, verifying the presence of git metadata and correct mappings of changed files to instance references.
- Enhanced the `compute_git_metadata` function to include canonical paths for repository roots.
- Added multiple test cases to ensure accurate behavior of `changedIds` and `scriptCommittedHashes` based on git status.
- Created a new project structure for testing, including various script and model files to facilitate comprehensive testing of git integration.
These changes improve the robustness of the git synchronization features and ensure reliable behavior in various scenarios.
- Updated `clap` and `unicode-ident` versions in `Cargo.lock` and `Cargo.toml`.
- Added `sha1` crate to `Cargo.toml` for SHA1 hashing functionality.
- Introduced a new `git.rs` module for handling git operations, including changed files and SHA1 computation.
- Enhanced the sync confirmation dialog to auto-select defaults based on git status, improving user experience during two-way sync.
- Updated `PatchTree` to utilize git metadata for determining default selections during patch application.
These changes enhance the integration with git, providing smarter defaults and improving the overall sync process.
- Added `#[allow(clippy::too_many_arguments)]` annotations in `change_processor.rs` and `matching.rs` to suppress warnings for functions with multiple parameters.
- Updated `serve_session.rs` to include `clippy::type_complexity` in the `#[allow(dead_code)]` annotation for better code quality.
- Simplified array initialization in `file_names.rs` for consistency.
- Enhanced texture filtering logic in `matching_fixtures.rs` by replacing `map_or` with `is_some_and` for improved readability.
These changes contribute to a cleaner codebase and address static analysis warnings, ensuring better maintainability.
- Introduced new scripts for auto-fixing formatting and performing static analysis across the project, including support for Lua, Rust, and TypeScript.
- Created `format.md` documentation to guide users on running the formatting scripts.
- Updated CI scripts to utilize the number of processor threads for improved performance during linting and testing.
- Enhanced logging in the `unwatch` function to provide clearer feedback on path watching status.
These changes streamline the formatting and analysis process, ensuring consistent code quality across the project.
- Implemented a 3-pass matching algorithm in the Reconciler to handle ambiguous paths, ensuring accurate reconciliation of instances with duplicate names.
- Added default property caching and refined property comparison logic for better instance matching.
- Introduced recursive change-count matching as a tiebreaker for ambiguous results.
- Refactored equality checks and path handling for robustness with duplicate-named siblings.
- Updated descendant search functions for `WeakDom` and `RojoTree` to support the new matching algorithm.
- Exposed `snapshot`, `syncback`, and `variant_eq` modules publicly, along with integration tests for the matching algorithm.
- Enhanced deduplication processes and improved code readability.
These changes significantly improve the reconciliation process, addressing previous ambiguities and enhancing overall performance.
- Simplified function calls in `Reconciler/matching.lua` by consolidating multi-line arguments into single lines, enhancing code clarity.
- Updated `matching.rs` to improve formatting of method calls and assertions, ensuring a more consistent style throughout the test suite.
- Enhanced the `matching_fixtures.rs` file by streamlining the loading of input data and improving the readability of child copying logic.
These changes contribute to a cleaner and more maintainable codebase, facilitating easier future modifications.
- Renamed `find_by_name` to `find_descendant` for clarity and updated its implementation to use recursion for improved search capabilities in `WeakDom`.
- Introduced a new function `find_descendant_by_name` for recursively finding descendants in `RojoTree`, enhancing the matching logic for nested structures.
- Updated references in the `syncback_matching_ufowave_textures` and `forward_matching_ufowave_textures` functions to utilize the new descendant search functions, improving code readability and maintainability.
- Updated `lib.rs` to make the `snapshot`, `syncback`, and `variant_eq` modules publicly accessible.
- Introduced a new integration test file `matching_fixtures.rs` to validate the matching algorithm using real-world fixtures, ensuring accurate pairing of instances despite property discrepancies.
- Enhanced the test suite by adding the new `matching_fixtures` module to `mod.rs`, improving overall test coverage for the project.
- Introduced a new function `getClassComparisonKeys` in `rbx_dom_lua/init.lua` to pre-compute and cache comparison keys for classes, improving efficiency in property matching.
- Updated the `Matching` module in `Reconciler/matching.lua` to utilize pre-computed caches for virtual and Studio instances, optimizing the reconciliation process.
- Enhanced handling of excluded properties and types during comparisons, ensuring accurate matching while reducing unnecessary computations.
These changes significantly improve the performance and accuracy of property reconciliation, addressing previous inefficiencies in the matching logic.
- Implemented a caching mechanism for default properties per class in `rbx_dom_lua/init.lua`, allowing for efficient retrieval of encoded default values, including inherited defaults.
- Updated the matching logic in `Reconciler/matching.lua` to account for properties present only on the Studio side, ensuring accurate comparisons against default values during reconciliation.
- Added new test cases to validate the handling of default-omitted properties, enhancing the robustness of the reconciliation process.
These changes improve the efficiency and accuracy of property matching, addressing issues related to default values in the reconciliation workflow.
- Updated the `rbx-dom` subproject to the latest commit for improved functionality.
- Enhanced the `/audit` command documentation to emphasize the importance of entering Plan mode before executing any actions.
- Improved the `atlas-dedup.mdc` file by adding details on user-initiated renames and ensuring ref path integrity during deduplication.
- Cleared out deprecated material color entries in `allValues.json` for better clarity.
- Updated `database.json` to reflect new properties and classes, including `AssetQualityService` and `BaseCoreGuiConfiguration`.
- Introduced new scripts for updating the reflection database, streamlining the process for generating necessary files.
These changes improve the overall functionality and maintainability of the project, ensuring better handling of deduplication and reflection updates.
- Simplified the formatting of path handling in `change_processor.rs` and `fs_snapshot.rs` by consolidating multiple lines into single lines for better clarity.
- Updated test assertions in `two_way_sync.rs` to streamline the code and enhance maintainability.
These changes enhance the overall readability of the codebase, making it easier to manage and understand.
- Streamlined the `globs` entries in multiple MDC files by removing unnecessary line breaks and consolidating patterns into single lines.
- Updated `atlas-api.mdc`, `atlas-dedup.mdc`, `atlas-project.mdc`, `atlas-src.mdc`, `atlas-syncback.mdc`, `plugin.mdc`, `python.mdc`, and `rbx-dom.mdc` to enhance readability and maintainability.
These changes improve the organization of the MDC files, making it easier to manage and understand the file patterns used across the project.
- Updated the audit plan for the ambiguous paths follow-up, marking all critical fixes as completed, including dedup cleanup, added paths, and ref path finalization.
- Introduced a new `atlas-dedup.mdc` file detailing the deduplication system, including matching algorithms, suffix cleanup, and integration for duplicate-named instances.
- Enhanced the `syncback` process to ensure proper handling of ref paths and deduplication, improving metadata consistency and overall system reliability.
- Implemented attribute-scoped path fixes in the `FsSnapshot` to ensure only relevant lines are modified during ref path updates.
These changes improve the clarity and effectiveness of the audit process while enhancing the deduplication and synchronization mechanisms within the project.
- Updated the audit plan to include detailed todos for five critical fixes identified during the audit, addressing issues related to dedup cleanup, added paths, and ref-path processing.
- Documented the status of each fix, ensuring clarity on implementation priorities and testing requirements.
- Expanded findings to include additional critical issues discovered during the audit, improving the overall comprehensiveness of the plan.
These changes provide a structured approach to resolving identified issues, enhancing the reliability and maintainability of the project.
- Introduced two new plan files: one detailing fixes for issues identified during the ambiguous-paths branch audit, including recursive matching, ref path formatting, and dedup cleanup improvements.
- The second plan outlines follow-up actions required for production audit, addressing critical round-trip issues, metadata consistency, and necessary regression tests.
- Documented todos for each plan to ensure clarity on implementation steps and testing requirements.
These additions enhance the auditing process and provide a structured approach to resolving identified issues, improving overall project reliability and maintainability.
- Introduced a new `trueEquals` function in `trueEquals.lua` for shared value equality checks, improving consistency between `diff.lua` and `matching.lua`.
- Removed redundant equality logic from `diff.lua` and `matching.lua`, streamlining the codebase.
- Updated path handling in various test snapshots to ensure correct reference values, enhancing the accuracy of reconciliation tests.
- Implemented a post-processing step in `syncback` to fix `Rojo_Ref_*` attribute paths with correct dedup suffixes, improving path resolution.
These changes enhance the clarity and efficiency of the Reconciler's equality checks and path handling, paving the way for improved project management and user experience.
- Renamed test case to clarify that children with duplicate names are now encoded rather than skipped.
- Adjusted assertions to ensure all children are included in the encoded output, aligning with the updated encoding logic.
- Refactored related test cases for improved clarity and maintainability.
These changes enhance the accuracy of the ChangeBatcher tests, ensuring they correctly validate the new encoding behavior.
- Updated test descriptions in `diff.spec.lua` and `diff.stress.spec.lua` to reflect changes in behavior regarding duplicate-named children in both real and virtual DOM contexts.
- Adjusted assertions to ensure that duplicate-named children are correctly removed or added based on the new matching logic.
- Refactored `hydrate.lua` and `matching.lua` to streamline the matching process by removing unnecessary parameters, enhancing clarity and maintainability.
- Updated metadata references in various test files to include the `.model.json5` extension for consistency in path handling.
These changes improve the accuracy of reconciliation tests and enhance the overall robustness of the Reconciler's path handling logic.
- Introduced a new plan file detailing the integration of ref path handling and deduplication cleanup processes.
- Implemented a `tentative_fs_path()` function to build filesystem-compatible paths, addressing issues with slugified and deduplicated instance names.
- Enhanced the `handle_tree_event()` method in the change processor to trigger dedup cleanup actions upon deletions, ensuring proper management of dedup groups.
- Updated relevant files to replace legacy path handling with the new system, improving overall path resolution and cleanup efficiency.
These changes enhance the robustness of path handling and deduplication processes, paving the way for improved project management and user experience.
- Introduced a comprehensive rewrite of the matching algorithm, replacing the previous multi-pass approach with a recursive change-count scoring function and greedy assignment strategy.
- Updated the `hydrate.lua` and `matching.lua` files to utilize the new algorithm, enhancing the efficiency of instance matching during hydration.
- Documented the new algorithm's structure, including fast-paths for Ref pins and (Name, ClassName) grouping, which significantly improve matching accuracy and performance.
- Ensured that the new implementation maintains compatibility with existing systems while providing a more robust solution for handling ambiguous instances.
These changes enhance the clarity and efficiency of the reconciliation process, paving the way for improved project management and user experience.
- Introduced a comprehensive audit plan detailing critical issues, code quality items, and test coverage gaps identified in the Ambiguous Path Handling feature.
- Documented specific todos addressing various correctness and performance issues, including case sensitivity, path formatting, and dead code removal.
- Established standards for round-trip identity and CLI syncback parity to ensure consistent behavior across the system.
These changes enhance the clarity and accountability of the auditing process, paving the way for improved feature reliability and maintainability.
- Updated the `encodeInstance.stress.spec.lua` tests to eliminate checks for skipped counts, as duplicates are now included alongside unique children.
- Adjusted expectations in the tests to reflect the new behavior of the encoding process, ensuring accurate validation of encoded instances.
- Removed legacy assertions related to skipped counts, streamlining the test cases for better clarity and maintainability.
- Finalized the transition from a duplicate-name blacklist system to a comprehensive 3-pass matching algorithm for both syncback and forward operations, enhancing the handling of ambiguous paths.
- Integrated a new `matching.lua` module to facilitate instance matching during hydration, utilizing unique name matching, class name narrowing, and structural fingerprinting.
- Updated various components, including `ChangeBatcher`, `Reconciler`, and `snapshot` modules, to leverage the new matching logic, ensuring robust file management and improved performance.
- Removed legacy duplicate handling code and streamlined the encoding process for instances with duplicate names.
These changes significantly enhance the clarity and efficiency of file operations, paving the way for improved project management and user experience.
- Introduced a new plan file detailing the transition from a duplicate-name blacklist system to a robust ambiguous path handling system, featuring a 3-pass matching algorithm and filesystem-collision-based deduplication.
- Outlined phases for implementation, including the removal of the duplicate-skip blacklist, restructuring syncback processes, and enhancing ref path resolution.
- Established a performance target to ensure efficient handling of large datasets while maintaining low latency.
These changes lay the groundwork for improved file management and clarity in handling ambiguous paths.
- Introduced a new file detailing the implementation of Ambiguous Path Handling, including naming conventions for duplicate file names and the distinction between matching and dedup groups.
- Established guidelines for suffix cleanup, base-name assignment, and VFS event handling to ensure robust file management.
- Implemented a new ref path resolution system to maintain unique paths among siblings, enhancing compatibility with existing projects.
These changes significantly improve the handling of ambiguous paths, ensuring clarity and consistency in file operations.
- Added a new section in audit.md to enforce that all incoming changes must pass through the review dialogue, ensuring user approval for every modification.
- Introduced a comprehensive plan for the Reconciler overhaul, addressing issues with MeshPart creation, MeshId swapping, and CSG reconstruction, modeled after the rbxsync implementation.
- Detailed todos for verifying Tags/Attributes, handling MeshPart creation via AssetService, and ensuring proper CSG part handling during synchronization.
These changes improve the stability and user control of the reconciliation process, enhancing overall plugin functionality.
- Upgraded `luau-lsp` dependency to version 1.62.0 for improved features and bug fixes.
- Modified sourcemap generator command in VSCode settings for better output handling.
- Enabled `fillCallArguments` in project template settings to enhance code completion.
- Added new configuration options for fragment autocomplete and quote conversion in project template settings.
- Introduced a `sourcemap` flag in `SyncbackCommand` to allow optional sourcemap generation during syncback operations.
These changes enhance the development experience and improve the functionality of the Luau language server integration.
- Introduced a new function `write_sourcemap_from_syncback` to generate sourcemaps directly from in-memory data, improving performance by eliminating the need for file reads during sourcemap creation.
- Updated the `syncback_loop` to return a `SyncbackResult` struct, encapsulating filesystem operations and instance paths for better post-processing.
- Modified the `SyncbackCommand` to utilize the new sourcemap generation method, allowing for parallel execution of file writes and sourcemap creation.
These changes streamline the syncback process and enhance the efficiency of sourcemap generation.
- Updated the ServeSession struct to allow for optional ChangeProcessor and related fields, enabling lightweight oneshot sessions that do not require live filesystem monitoring.
- Modified the new and new_oneshot methods to initialize sessions accordingly, improving flexibility for commands that only need a snapshot of the tree.
- Adjusted related CLI commands to utilize the new oneshot session functionality, enhancing performance for operations like plugin installation and syncback.
These changes streamline session management and improve the efficiency of operations that do not require continuous updates.
- Deleted the gitconfig.txt template file from the project templates.
- Updated the InitCommand to set git configuration options directly using the `git config` command, enhancing the initialization process by eliminating the need for a separate configuration file.
These changes streamline the git initialization process and improve the overall setup experience.
- Canonicalized the project_root in ChangeProcessor to ensure consistent path comparisons on Windows, addressing issues with the \\?\ prefix.
- Updated snapshot_project to push canonicalized paths to relevant_paths, improving path accuracy in project snapshots.
These changes enhance the reliability of path handling in the project and improve overall consistency in file operations.
- Adjusted the settle duration from 300ms to 100ms to ensure it is shorter than the 200ms reconciliation timer. This change prevents capturing tree-correction messages as part of the same batch, enhancing the accuracy of message processing in the test environment.
This update improves the reliability of message handling during test sessions.
- Replaced numeric messageCursor values with a placeholder "[message-cursor]" across multiple snapshot files in the connected mode tests.
- This change ensures consistency in the handling of message cursors, facilitating easier updates and maintenance of test snapshots.
These updates improve the clarity and reliability of the snapshot files used in end-to-end testing.
- Created a new .gitignore file to exclude .txt files in the logs directory.
- Updated the `ChangeProcessor` to skip directory-level WRITE events on Windows, preventing stale data issues in the VFS cache and ensuring correct patch generation.
These changes enhance the handling of file events and improve the organization of log files.
- Implemented two-way sync for Ref properties using `Rojo_Ref_*` attributes, enabling detection of changes in Studio and path-based encoding.
- Developed `RefPathIndex` for efficient resolution of attributes during forward sync, including handling ambiguous paths.
- Enhanced `ChangeProcessor` and `ChangeBatcher` to manage Ref property changes during live sync sessions.
- Introduced `InstanceMap:getIdByInstance` for reverse lookups and added comprehensive integration tests for various edge cases.
- Optimized CI scripts for improved parallel execution of Rust tasks.
These updates significantly enhance the synchronization system's handling of Ref properties and improve overall performance.
- Changed the `luau-lsp.sourcemap.autogenerate` setting from `true` to `false` in the VSCode configuration, altering the behavior of sourcemap generation for the Luau language server.
This change ensures that sourcemaps are not automatically generated, allowing for more controlled management of source mapping in the development environment.
