v0.3.0-beta.1 feat: move Sentrux DSM analysis to Rust - #10
Conversation
|
Important Review skippedAuto reviews are disabled on base/target branches other than the default branch. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Plus Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
📝 WalkthroughWalkthroughAdds a Rust-native Sentrux DSM analyzer with source, Git, function, dependency, module, and risk metrics. Exposes it through the CLI and orchestration, and updates the PowerShell pipeline to prefer Rust with PowerShell fallback. ChangesSentrux DSM Analysis
Estimated code review effort: 4 (Complex) | ~60 minutes Sequence Diagram(s)sequenceDiagram
participant Pipeline as run-code-intel.ps1
participant CLI as code-intel sentrux dsm
participant Analyzer as sentrux_analysis::analyze
participant Fallback as Invoke-SentruxAgentTool.ps1
Pipeline->>CLI: invoke Rust DSM command
CLI->>Analyzer: analyze repository path
Analyzer-->>CLI: return JSON snapshot
CLI-->>Pipeline: print JSON
Pipeline->>Fallback: invoke PowerShell DSM on Rust failure
Fallback-->>Pipeline: return DSM snapshot
Poem
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
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 |
|
@coderabbitai review Please focus on Rust/PowerShell DSM contract parity, source inventory exclusions, stable ID generation, dependency/risk calculations, and failure-safe PowerShell fallback routing. |
|
I'll follow up with detailed findings shortly. ✅ Action performedReview finished.
|
![gemini-code-assist[bot]](https://avatars.github.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request ports the performance-sensitive Dependency Structure Matrix (DSM) analysis from PowerShell to a native Rust implementation within the code-intel CLI, updating the orchestration pipeline to prefer the Rust executable with a PowerShell fallback. The review feedback identifies several critical issues: cross-platform compatibility is broken due to a hardcoded .exe extension in the PowerShell script; the Rust analysis suffers from high memory allocations in a hot loop during Git key resolution; recursive file traversal is vulnerable to stack overflows from symlinks; PowerShell function parsing is fragile to multiple spaces; and complexity calculations are skewed by trailing comments.
Important
The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.
![coderabbitai[bot]](https://avatars.github.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 9
🧹 Nitpick comments (2)
run-code-intel.ps1 (1)
4022-4026: 🎯 Functional Correctness | 🔵 Trivial | ⚡ Quick winFix cross-platform pathing for the Rust CLI.
Using a hardcoded backslash (
\) causes path resolution to fail on non-Windows platforms (e.g., Linux/macOS), where PowerShell treatstarget\debug\code-intel.exeas a single literal file name rather than a directory path. Use a forward slash (/) instead, which PowerShell correctly resolves on all platforms.(Note: standard Rust builds on Linux also omit the
.exeextension, so you may need platform-specific extension logic if this is intended to run locally on Linux withoutCODE_INTEL_RUST_CLIset).♻️ Proposed fix for cross-platform pathing
$sentruxDsmRustCli = if ([string]::IsNullOrWhiteSpace($env:CODE_INTEL_RUST_CLI)) { - Join-Path $PSScriptRoot "target\debug\code-intel.exe" + Join-Path $PSScriptRoot "target/debug/code-intel.exe" } else { [IO.Path]::GetFullPath($env:CODE_INTEL_RUST_CLI) }🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@run-code-intel.ps1` around lines 4022 - 4026, Update the default path in the $sentruxDsmRustCli assignment to use forward-slash separators so Join-Path resolves target/debug/code-intel.exe correctly across platforms. Preserve the CODE_INTEL_RUST_CLI override behavior; only add platform-specific executable-extension handling if the default path must support local non-Windows Rust builds.crates/code-intel-cli/tests/sentrux_analysis.rs (1)
117-125: 📐 Maintainability & Code Quality | 🔵 Trivial | 🏗️ Heavy liftExercise provider routing instead of matching script text.
This test passes if the commands appear in comments or unreachable branches. It does not prove that Rust is the default, the environment variable selects PowerShell, or a Rust failure triggers fallback. Test an extracted routing function or invoke the script with stub executables and assert the selected provider and failure path.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@crates/code-intel-cli/tests/sentrux_analysis.rs` around lines 117 - 125, Replace the text-presence assertions in production_pipeline_prefers_rust_dsm_with_explicit_powershell_rollback with behavioral coverage of the routing logic in run-code-intel.ps1. Exercise the default path to verify Rust is selected, set CODE_INTEL_SENTRUX_DSM_PROVIDER to verify PowerShell selection, and make the Rust stub fail to verify fallback invokes PowerShell; assert the resulting provider and command outcomes rather than script text.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@crates/code-intel-cli/src/sentrux_analysis.rs`:
- Around line 759-775: Replace the repeated depth-relaxation loop around depths
initialization with SCC condensation: identify strongly connected components
from modules and edges, build the component DAG, and calculate each component’s
execution depth once from its dependencies. Map component depths back to member
modules while preserving the existing depth cap and handling disconnected
modules.
- Around line 630-647: Update c_like_end to avoid extending brace-less
declarations into subsequent functions: recognize declaration terminators such
as Rust semicolons and expression-bodied JavaScript functions, and return the
declaration’s end without waiting for a later `{`. When scanning braces, mask or
otherwise ignore braces inside strings and comments so only structural braces
affect depth and range termination.
- Around line 650-666: Update complexity so the arrow-syntax contribution in
complexity is not applied to JavaScript arrow functions, while preserving
complexity scoring for supported control-flow constructs and other languages.
Use the available language/context information or an appropriate context-aware
check rather than counting every trimmed line containing " => ".
- Around line 694-717: Update import_targets and its callers to resolve each
target relative to the importing source file’s package or crate root. Preserve
Python module filenames such as b.py and Cargo crate paths such as
crates/foo/src/a.rs instead of collapsing them to directory names; resolve use
crate:: imports within that crate without synthetic src-only targets. Ensure
generated targets match source module names and avoid self-edges for flat
projects and workspace crates.
- Around line 685-688: Update the edge accumulation in the import-target loop to
increment the existing count for each repeated (from, target) module pair
instead of resetting it to 1. Preserve the current filtering for self-imports
and known modules, and ensure new pairs still start with count 1 so downstream
edge, coupling, and risk metrics reflect all matching imports.
- Around line 616-625: Update python_end so a decorator line at the current
function’s sibling indentation is treated as the next declaration boundary
rather than skipped as part of the current function. Preserve skipping
decorators that are indented within the current function, and keep the existing
end-index behavior for comments, blank lines, and ordinary sibling code.
- Around line 112-125: Unify the file-discovery policy in source_inventory so
results do not depend on whether rg is installed. Centralize inventory discovery
or apply identical explicit exclusions and path normalization to both the rg
--files path and recursive_files fallback, preserving the existing Inventory
construction.
- Around line 66-68: Update the source-reading logic in the inventory file loop
to propagate fs::read_to_string failures instead of converting them to empty
content. Ensure the enclosing analysis flow returns the error so provider
fallback can run, while preserving normal module processing for successfully
read files.
In `@run-code-intel.ps1`:
- Around line 4031-4039: Add a catch block to the try/finally execution flow
around the $sentruxDsmRustCli invocation, capturing launch exceptions such as
UnauthorizedAccessException and setting the existing DSM result variables to the
values required to trigger the PowerShell compatibility fallback. Preserve the
finally block so $ErrorActionPreference is always restored.
---
Nitpick comments:
In `@crates/code-intel-cli/tests/sentrux_analysis.rs`:
- Around line 117-125: Replace the text-presence assertions in
production_pipeline_prefers_rust_dsm_with_explicit_powershell_rollback with
behavioral coverage of the routing logic in run-code-intel.ps1. Exercise the
default path to verify Rust is selected, set CODE_INTEL_SENTRUX_DSM_PROVIDER to
verify PowerShell selection, and make the Rust stub fail to verify fallback
invokes PowerShell; assert the resulting provider and command outcomes rather
than script text.
In `@run-code-intel.ps1`:
- Around line 4022-4026: Update the default path in the $sentruxDsmRustCli
assignment to use forward-slash separators so Join-Path resolves
target/debug/code-intel.exe correctly across platforms. Preserve the
CODE_INTEL_RUST_CLI override behavior; only add platform-specific
executable-extension handling if the default path must support local non-Windows
Rust builds.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: bd986e1c-08a1-4c78-b843-fb8cb031f582
📒 Files selected for processing (7)
crates/code-intel-cli/src/main.rscrates/code-intel-cli/src/sentrux.rscrates/code-intel-cli/src/sentrux_analysis.rscrates/code-intel-cli/tests/sentrux_analysis.rsdocs/plans/sentrux-rust-analysis-kernel-idea.mdorchestration/integrations.jsonrun-code-intel.ps1
|
@coderabbitai review |
✅ Action performedReview finished.
|
|
@coderabbitai review |
✅ Action performedReview finished.
|
2233admin
changed the title
1 participant
What changed
code-intel sentrux dsm <repo>Why
DSM source inventory, function extraction, dependency graph construction, and risk aggregation are CPU-heavy. Keeping this path in PowerShell made normal pipeline runs unnecessarily slow.
Impact
CODE_INTEL_SENTRUX_DSM_PROVIDER=powershellValidation
cargo test --manifest-path crates/code-intel-cli/Cargo.toml --quiet: 24 + 3 tests passedgit diff --cached --checkpassedstage=land,project-fast)Sentrux DSM provider: rustBeta release candidate