iframe-proxy

carlos-alm · 2026-02-22T02:01:33Z

Summary

Add a native Rust parsing engine via napi-rs as an optional high-performance alternative to the existing WASM pipeline. The WASM path remains the always-available baseline; the native engine activates automatically when platform binaries are installed.

Phase 1 — Rust core (napi-rs dual-engine)

All 11 language extractors (JS/TS/TSX, Python, Go, Rust, Java, C#, Ruby, PHP, HCL) ported to Rust with a SymbolExtractor trait
Rayon-based parallel file parsing
6-level import resolution with confidence scoring
Tarjan's SCC cycle detection
Incremental parse tree cache for watch mode

Phase 1.2 — Unified API

parseFileAuto, parseFilesAuto, getActiveEngine as the single entry point in parser.js
Builder and watcher use the unified API — no direct native imports

Phase 1.3 — Incremental parsing

Native ParseTreeCache exposed via napi-rs for watch mode old-tree hints

Phase 1.4 — Native import resolution

resolveImportPath and computeConfidence extracted to src/resolve.js with native dispatch
Batch pre-resolution of all imports in a single native call

Phase 1.5 — Graceful degradation & diagnostics

--engine flag plumbed to watch command
codegraph info diagnostic command
CI workflows for cross-platform test matrix and native build pipeline

Changes

40 files changed, ~5,500 lines added
New Rust crate at crates/codegraph-core/
New JS modules: src/native.js, src/resolve.js
Refactored: src/builder.js, src/parser.js, src/cli.js, src/watcher.js, src/cycles.js
CI: build-native.yml (4-platform matrix), ci.yml
Tests: engine parity, unified API, import resolution parity, build parity, incremental cache

Test plan

npm test — full WASM test suite passes (no native required)
Cross-engine parity tests pass when native binary is available
Build parity test confirms identical graph output from both engines
codegraph info reports correct engine status
--engine wasm forces WASM even when native is available
Watch mode falls back gracefully when native unavailable

🤖 Generated with Claude Code

Move CPU-intensive parsing, import resolution, and cycle detection to Rust via napi-rs while keeping JS for CLI, SQLite, MCP, and embeddings. The WASM path remains as a fallback for unsupported platforms. Rust crate (crates/codegraph-core): - All 9 language extractors (JS/TS/TSX, Python, Go, Rust, Java, C#, Ruby, PHP, HCL) ported with SymbolExtractor trait - Rayon-based parallel file parsing - 6-level import resolution with confidence scoring - Tarjan's SCC cycle detection - Incremental parse tree cache for watch mode - napi-rs API: parseFile, parseFiles, resolveImport, resolveImports, computeConfidence, detectCycles, engineName, engineVersion JS integration: - src/native.js: platform-aware addon loader with graceful fallback - src/builder.js: dual-engine path (native fast path + WASM fallback) - src/cycles.js: dispatches to native detectCycles when available - src/watcher.js: uses native parseFile in watch mode - src/cli.js: --engine <native|wasm|auto> global option - src/index.js: exports isNativeAvailable CI & packaging: - GitHub Actions workflow for 4-platform matrix build - Platform optionalDependencies in package.json Tests: - Rust unit tests for JS, Python, Go extractors and cycle detection - Cross-engine parity tests for all 11 languages (skip when native N/A) - Full build parity test comparing SQLite output - Extended JS cycle tests for findCyclesJS

…se 1.2) Consolidate dual-engine dispatch into parser.js so builder.js and watcher.js never import native.js directly. The new parseFileAuto, parseFilesAuto, and getActiveEngine exports handle engine detection, delegation, and snake_case→camelCase normalization in one place. - parser.js: add unified API (parseFileAuto, parseFilesAuto, getActiveEngine) - builder.js: replace ~150 lines of dual-engine code with single parseFilesAuto call - watcher.js: replace dual-path updateFile with parseFileAuto call - index.js: export new unified API functions - tests/parsers/unified.test.js: 12 new tests for the unified API

Expose the Rust ParseTreeCache to JS via napi-rs so watch mode can reuse cached parse trees (old-tree hint) instead of full re-parses. WASM falls back to parseFileAuto when native is unavailable.

… 1.4) Extract resolveImportPath, computeConfidence into src/resolve.js with native dispatch and JS fallback. Batch pre-resolve all imports in a single native call before edge building. Add parity tests for import resolution, confidence scoring, and cycle detection.

…ation Proposal for (when, do, what) triplets as a complement to session-graph's (subject, predicate, object) knowledge graph, enabling scoped behavioral rules attached to codegraph's function-level dependency nodes.

- Plumb --engine flag to watch command (was hardcoded to 'auto') - Add `codegraph info` diagnostic command for engine/platform reporting - Stage CI workflows: cross-platform test matrix and native build pipeline

The 0.4 version constraint had no matching crate on crates.io — the published version is 1.1.0.

greptile-apps · 2026-02-22T02:04:32Z

Greptile Summary

This PR introduces a Rust-based native parsing engine as an optional high-performance alternative to the existing WASM pipeline, maintaining backward compatibility through a dual-engine architecture.

Key architecture decisions:

Native engine activates automatically when platform binaries are installed via optional dependencies
WASM remains the always-available baseline requiring no native compilation
All 11 language extractors ported to Rust with a unified SymbolExtractor trait
Rayon-based parallel file parsing for performance improvements
Unified API (parseFileAuto, parseFilesAuto) abstracts engine selection
Import resolution and cycle detection also ported to Rust with JS fallbacks
Incremental parsing via ParseTreeCache for watch mode with old-tree hints

Implementation quality:

Clean separation of concerns with src/native.js handling graceful loading
Comprehensive test coverage including cross-engine parity tests for all languages
Build parity test ensures identical SQLite graph output from both engines
CI matrix tests on 3 platforms × 2 Node versions without requiring native builds
Platform-specific npm packages published separately via build-native.yml workflow

Deployment strategy:

4 platform-specific packages as optionalDependencies (@optave/codegraph-{platform})
Native binaries distributed separately, main package works without them
--engine flag allows explicit selection (native/wasm/auto)
New codegraph info command for diagnostics

Confidence Score: 5/5

This PR is safe to merge with excellent architecture and comprehensive testing
The implementation demonstrates exceptional quality: (1) clean dual-engine architecture with graceful degradation, (2) comprehensive parity tests ensuring WASM and native produce identical output, (3) proper use of optional dependencies for platform binaries, (4) extensive cross-platform CI coverage, (5) well-structured Rust code mirroring the JS extractors, (6) proper error handling and fallbacks throughout. The incremental approach across phases shows careful planning, and the unified API prevents breaking changes.
No files require special attention

Important Files Changed

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A[User: npm install] --> B{Native binary<br/>available?}
    B -->|Yes| C[Load platform package<br/>via native.js]
    B -->|No| D[Use WASM fallback]
    
    C --> E[parseFilesAuto]
    D --> E
    
    E --> F{engine option}
    F -->|'native'| G[Native Rust Parser<br/>Rayon parallel]
    F -->|'wasm'| H[WASM tree-sitter]
    F -->|'auto'| I{Native available?}
    
    I -->|Yes| G
    I -->|No| H
    
    G --> J[Rust extractors<br/>11 languages]
    H --> K[JS extractors<br/>11 languages]
    
    J --> L[normalizeNativeSymbols]
    K --> L
    
    L --> M[builder.js]
    M --> N[Import resolution<br/>resolve.js + native]
    N --> O[Cycle detection<br/>cycles.js + native]
    O --> P[SQLite graph.db]
    
    Q[Watch mode] --> R[ParseTreeCache<br/>incremental.rs]
    R --> G

_{Last reviewed commit: 151fceb}

greptile-apps

_{40 files reviewed, no comments}

_{Edit Code Review Agent Settings | Greptile}

tree-sitter-php 0.23 exports LANGUAGE_PHP directly without a feature gate. The "php" feature doesn't exist in this crate.

- Pin @napi-rs/cli to v3 to match napi crate dependencies - Remove invalid --platform flag (v2 only) from napi build command - Bump napi-build from v2 to v3 to align with napi/napi-derive v3 - Replace find with shell glob for Windows runner compatibility

The review action completed successfully but couldn't post comments because it only had read access to pull requests.

…sion The claude-code-review action requires the workflow file to be identical on the PR branch and default branch. Also napi-build v3 does not exist on crates.io — latest is 2.x.

feat: Rust core engine via napi-rs (Phase 1) Add a native Rust parsing engine as an optional high-performance alternative to the existing WASM pipeline. The WASM path remains the always-available baseline; the native engine activates automatically when platform binaries are installed. Rust crate (crates/codegraph-core): - All 11 language extractors ported with SymbolExtractor trait - Rayon-based parallel file parsing - 6-level import resolution with confidence scoring - Tarjan's SCC cycle detection - Incremental parse tree cache for watch mode - napi-rs bindings: parseFile, parseFiles, resolveImport, resolveImports, computeConfidence, detectCycles JS integration: - Unified API via parseFileAuto/parseFilesAuto/getActiveEngine - Builder and watcher use unified API — no direct native imports - Import resolution extracted to src/resolve.js with native dispatch - --engine <native|wasm|auto> CLI flag and `codegraph info` command - Graceful fallback when native binary is unavailable CI & packaging: - 4-platform matrix build workflow (build-native.yml) - Cross-platform test matrix (ci.yml) 40 files changed, ~5200 additions

Re-rank codegraph from #8 (4.0) to #5 (4.5) reflecting v3.2.0 features: 41 CLI commands, 32 MCP tools, dataflow across all 11 languages, CFG, sequence diagrams, architecture boundaries, unified graph model. Add new competitors: GitNexus (#1, 18k stars), DeusData/codebase-memory-mcp (#6, 793 stars in 25 days). Update star counts and feature status across all 85+ ranked projects. Mark 7 roadmap items as DONE. Flag stagnant projects. Update joern.md (3,021 stars, 75 contributors, 4 community MCP wrappers) and narsil-mcp.md (129 stars, SPA frontend, +36 security rules, development paused since Feb 25).

…#559) * docs: update competitive analysis for v3.2.0 and March 2026 landscape Re-rank codegraph from #8 (4.0) to #5 (4.5) reflecting v3.2.0 features: 41 CLI commands, 32 MCP tools, dataflow across all 11 languages, CFG, sequence diagrams, architecture boundaries, unified graph model. Add new competitors: GitNexus (#1, 18k stars), DeusData/codebase-memory-mcp (#6, 793 stars in 25 days). Update star counts and feature status across all 85+ ranked projects. Mark 7 roadmap items as DONE. Flag stagnant projects. Update joern.md (3,021 stars, 75 contributors, 4 community MCP wrappers) and narsil-mcp.md (129 stars, SPA frontend, +36 security rules, development paused since Feb 25). * docs: fix narsil SPA version attribution in competitive analysis overview Line 18 incorrectly stated "v1.6.1" as the version when the SPA feature was introduced. The SPA frontend was added in v1.6.0; v1.6.1 is the current release. Updated to "added v1.6.0, current v1.6.1" to match the detailed narsil-mcp.md entry. * docs: remove hardcoded star count from joern comparison table The "32 stars, growing" value in the Community & maturity row hardcodes a stale star count. Other comparison tables use "Growing" consistently for codegraph's community status. Updated to match. * fix: correct GitNexus score, Tier 2 rank numbering, and jelly star count - GitNexus overall score corrected from 4.7 to 4.5 to match the arithmetic mean of its six sub-scores (5+5+4+4+4+5)/6 = 4.5 - Tier 2 renumbered starting at #38 (was duplicating #37 with Tier 1); also resolves the pre-existing duplicate #43 (Bikach/ChrisRoyse now #44/#45), with all subsequent entries incremented accordingly - jelly section header updated from 417 to 423 stars to match the ranking table * fix: correct aider rank and codegraph star count per review feedback * fix: align scoring breakdown sub-scores with overall rankings for stagnant projects glimpse: Community 4→2 (stagnant since Jan 2026), avg now 3.83≈3.8 matching ranking. autodev-codebase: Community 3→1 (stagnant since Jan 2026), avg now 3.33, ranking updated 3.4→3.3. * fix: align ranking scores with sub-score averages for colbymchenry and axon * fix: correct ranking inversion at positions #23/#24 (#559) autodev-codebase (3.3) was ranked #23 above Claude-code-memory (3.4) at #24. Swapped to maintain descending score order. * fix: correct score mismatches for code-graph-rag (4.5→4.2) and arbor (3.7→4.2) (#559) * fix: sync breakdown table row order with ranking table for #23/#24 (#559) * fix: correct ranking inversions and stale rank references (#559) * fix: correct sub-score/overall-score mismatches for codexray, loregrep, MATE * fix: correct score mismatches and aider header rank * fix: update narsil-mcp Key Metrics to reflect development stagnation (#559) * fix: add missing "vs arbor" comparison section (#559) * fix: remove duplicate vs-glimpse section and correct role names in vs-arbor (#559) The duplicate vs-glimpse block (stale rank #10) was left behind when vs-arbor was inserted. Removed it — the correct version exists at #11. Also fixed role vocabulary in vs-arbor: bridge → adapter, added entry.

…optave#559) * docs: update competitive analysis for v3.2.0 and March 2026 landscape Re-rank codegraph from optave#8 (4.0) to optave#5 (4.5) reflecting v3.2.0 features: 41 CLI commands, 32 MCP tools, dataflow across all 11 languages, CFG, sequence diagrams, architecture boundaries, unified graph model. Add new competitors: GitNexus (optave#1, 18k stars), DeusData/codebase-memory-mcp (optave#6, 793 stars in 25 days). Update star counts and feature status across all 85+ ranked projects. Mark 7 roadmap items as DONE. Flag stagnant projects. Update joern.md (3,021 stars, 75 contributors, 4 community MCP wrappers) and narsil-mcp.md (129 stars, SPA frontend, +36 security rules, development paused since Feb 25). * docs: fix narsil SPA version attribution in competitive analysis overview Line 18 incorrectly stated "v1.6.1" as the version when the SPA feature was introduced. The SPA frontend was added in v1.6.0; v1.6.1 is the current release. Updated to "added v1.6.0, current v1.6.1" to match the detailed narsil-mcp.md entry. * docs: remove hardcoded star count from joern comparison table The "32 stars, growing" value in the Community & maturity row hardcodes a stale star count. Other comparison tables use "Growing" consistently for codegraph's community status. Updated to match. * fix: correct GitNexus score, Tier 2 rank numbering, and jelly star count - GitNexus overall score corrected from 4.7 to 4.5 to match the arithmetic mean of its six sub-scores (5+5+4+4+4+5)/6 = 4.5 - Tier 2 renumbered starting at optave#38 (was duplicating optave#37 with Tier 1); also resolves the pre-existing duplicate optave#43 (Bikach/ChrisRoyse now optave#44/optave#45), with all subsequent entries incremented accordingly - jelly section header updated from 417 to 423 stars to match the ranking table * fix: correct aider rank and codegraph star count per review feedback * fix: align scoring breakdown sub-scores with overall rankings for stagnant projects glimpse: Community 4→2 (stagnant since Jan 2026), avg now 3.83≈3.8 matching ranking. autodev-codebase: Community 3→1 (stagnant since Jan 2026), avg now 3.33, ranking updated 3.4→3.3. * fix: align ranking scores with sub-score averages for colbymchenry and axon * fix: correct ranking inversion at positions optave#23/optave#24 (optave#559) autodev-codebase (3.3) was ranked optave#23 above Claude-code-memory (3.4) at optave#24. Swapped to maintain descending score order. * fix: correct score mismatches for code-graph-rag (4.5→4.2) and arbor (3.7→4.2) (optave#559) * fix: sync breakdown table row order with ranking table for optave#23/optave#24 (optave#559) * fix: correct ranking inversions and stale rank references (optave#559) * fix: correct sub-score/overall-score mismatches for codexray, loregrep, MATE * fix: correct score mismatches and aider header rank * fix: update narsil-mcp Key Metrics to reflect development stagnation (optave#559) * fix: add missing "vs arbor" comparison section (optave#559) * fix: remove duplicate vs-glimpse section and correct role names in vs-arbor (optave#559) The duplicate vs-glimpse block (stale rank optave#10) was left behind when vs-arbor was inserted. Removed it — the correct version exists at optave#11. Also fixed role vocabulary in vs-arbor: bridge → adapter, added entry.

carlos-alm added 7 commits February 21, 2026 17:43

feat: wire up incremental parsing via native ParseTreeCache (Phase 1.3)

f372674

Expose the Rust ParseTreeCache to JS via napi-rs so watch mode can reuse cached parse trees (old-tree hint) instead of full re-parses. WASM falls back to parseFileAuto when native is unavailable.

feat: graceful degradation & diagnostics (Phase 1.5)

a9ad51e

- Plumb --engine flag to watch command (was hardcoded to 'auto') - Add `codegraph info` diagnostic command for engine/platform reporting - Stage CI workflows: cross-platform test matrix and native build pipeline

fix: bump tree-sitter-hcl to ^1 to match available crate versions

151fceb

The 0.4 version constraint had no matching crate on crates.io — the published version is 1.1.0.

greptile-apps Bot reviewed Feb 22, 2026

View reviewed changes

carlos-alm and others added 5 commits February 21, 2026 19:11

fix: remove non-existent php feature from tree-sitter-php dep

a8339f2

tree-sitter-php 0.23 exports LANGUAGE_PHP directly without a feature gate. The "php" feature doesn't exist in this crate.

fix: grant pull-requests write permission to code review workflow

5b2de16

The review action completed successfully but couldn't post comments because it only had read access to pull requests.

Merge branch 'main' into feat/rust-core

6ba5f19

fix: revert code review workflow to match main and fix napi-build ver…

f36b5e4

…sion The claude-code-review action requires the workflow file to be identical on the PR branch and default branch. Also napi-build v3 does not exist on crates.io — latest is 2.x.

carlos-alm merged commit 9944f37 into main Feb 22, 2026
9 checks passed

carlos-alm deleted the feat/rust-core branch February 22, 2026 02:58

carlos-alm mentioned this pull request Feb 22, 2026

feat: registry hardening, structural analysis, single-repo MCP default #19

Merged

6 tasks

claude Bot mentioned this pull request Feb 24, 2026

docs: score backlog and integrate LLM features into roadmap #73

Merged

5 tasks

This was referenced Feb 25, 2026

feat: node role classification + dead code detection #91

Merged

docs: move roadmap files into docs/roadmap/ #92

Merged

greptile-apps Bot mentioned this pull request Mar 3, 2026

docs: competitive deep-dive vs Joern #260

Merged

3 tasks

carlos-alm mentioned this pull request Mar 21, 2026

docs: update competitive analysis for v3.2.0 and March 2026 landscape #559

Merged

3 tasks

carlos-alm mentioned this pull request May 27, 2026

refactor(domain): decompose parser, analysis, and search modules #1236

Merged

2 tasks

Sunbelt Computer Software

PL/B Language Development and Support

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

feat: Rust core engine via napi-rs (Phase 1)#8

feat: Rust core engine via napi-rs (Phase 1)#8
carlos-alm merged 12 commits into
mainfrom
feat/rust-core

carlos-alm commented Feb 22, 2026

Uh oh!

greptile-apps Bot commented Feb 22, 2026

Uh oh!

greptile-apps Bot left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Sunbelt Computer Software

PL/B Language Development and Support

Uh oh!

Conversation

carlos-alm commented Feb 22, 2026

Summary

Changes

Test plan

Uh oh!

greptile-apps Bot commented Feb 22, 2026

Greptile Summary

Confidence Score: 5/5

Important Files Changed

Flowchart

Uh oh!

greptile-apps Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant