Conversation
There was a problem hiding this comment.
Pull request overview
This pull request introduces a comprehensive semantic-heuristic search ranking system that significantly enhances code search relevance by incorporating structural analysis of source code. The PR adds a new "structural" ranker that extends BM25 with code-aware features: weighting matches by whether they appear in code vs comments vs strings, applying complexity gravity to boost algorithmically dense files, penalizing low-information-density data files, dampening common language keywords, and adjusting test file rankings based on query intent.
Changes:
- New structural ranking algorithm with configurable weights for code/comment/string matches
- Complexity gravity system (brain/logic/default/low/off) to boost complex logic files
- Noise penalty to demote large low-complexity data files (JSON, logs, minified JS)
- Language-specific stopword dampening (e.g., "func" in Go, "def" in Python)
- Test file detection with adaptive ranking based on query intent
- Content-type filters (--only-code, --only-comments, --only-strings) for precise searching
- Path filter with glob pattern support (path:/search/, file:*.go)
- TUI function key controls (F1-F4) for live ranker/filter adjustment
- HTTP server real-time search with JSON API endpoint
- MCP server integration with new parameters
Reviewed changes
Copilot reviewed 27 out of 88 changed files in this pull request and generated no comments.
Show a summary per file
| File | Description |
|---|---|
| vendor/* | Formatting updates (whitespace, indentation, build tags) across multiple vendored libraries |
| pkg/ranker/ranker.go | Core structural ranker implementation with gravity, noise penalty, test dampening |
| pkg/ranker/stopwords.go | Language-specific keyword dampening system for 18 languages |
| pkg/ranker/*_test.go | Comprehensive test coverage for all new ranking features |
| pkg/search/executor.go | Path filter with glob support, post-evaluation metadata filters |
| pkg/search/parser.go, lexer.go | Path filter parsing with slash-in-value handling |
| pkg/search/*_test.go | 150+ new test cases for path/glob filters and metadata evaluation |
| search.go, search_filter_test.go | Content-type filtering with per-byte classification from scc |
| config.go | New configuration fields with intent-based parameter resolution |
| tui.go, tui_test.go | Function key controls for live ranker/filter cycling |
| http.go, asset/templates/* | Real-time search UI with JSON API and dropdown controls |
| mcp.go | Enhanced MCP tool descriptions with new parameters |
| main.go | CLI integration with mutual exclusivity checks and auto-selection logic |
| language.go | scc integration for per-byte content classification |
| console.go | Updated ranking call with new parameters |
| README.md | Comprehensive documentation of new features |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
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 free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.
This adds in a new ranking algorithm which expands on BM25 in the following ways
It also adds in options to control all of these via the interfaces, including the TUI and HTTP.