metadata

tags:
  - code-understanding
  - semantic-analysis
  - rust
  - rust-analyzer
  - compiler
  - language-server
  - ai
  - dataset
license: agpl-3.0
size_categories:
  - 100K<n<1M
task_categories:
  - text-classification
  - feature-extraction
  - text-retrieval
language:
  - en

Rust-Analyzer Semantic Analysis Dataset

This dataset contains comprehensive semantic analysis data extracted from the rust-analyzer codebase using our custom rust-analyzer integration. It captures the step-by-step processing phases that rust-analyzer performs when analyzing Rust code.

Dataset Overview

This dataset provides unprecedented insight into how rust-analyzer (the most advanced Rust language server) processes its own codebase. It contains 500K+ records across multiple semantic analysis phases.

What's Included

Parsing Phase: Syntax tree generation, tokenization, and parse error handling
Name Resolution Phase: Symbol binding, scope analysis, and import resolution
Type Inference Phase: Type checking, inference decisions, and type error detection

Dataset Statistics

Total Records: ~533,000 semantic analysis events
Source Files: 1,307 Rust files from rust-analyzer codebase
Data Size: ~450MB in efficient Parquet format
Processing Phases: 3 major compiler phases captured

Dataset Structure

Each record contains:

id: Unique identifier for the analysis event
file_path: Source file being analyzed
line, column: Location in source code
phase: Processing phase (parsing, name_resolution, type_inference)
element_type: Type of code element (function, struct, variable, etc.)
element_name: Name of the element (if applicable)
syntax_data: JSON-serialized syntax tree information
symbol_data: JSON-serialized symbol resolution data
type_data: JSON-serialized type inference information
source_snippet: The actual source code being analyzed
context_before/context_after: Surrounding code context
processing_time_ms: Time taken for analysis
rust_version, analyzer_version: Tool versions used

Use Cases

Machine Learning Applications

Code completion models: Train on parsing and name resolution patterns
Type inference models: Learn from rust-analyzer's type inference decisions
Bug detection models: Identify patterns in diagnostic data
Code understanding models: Learn semantic analysis patterns

Research Applications

Compiler optimization: Analyze compilation patterns across large codebases
Language design: Study how developers use Rust language features
IDE improvement: Understand common semantic analysis patterns
Static analysis: Develop better code analysis tools

Educational Applications

Rust learning: Understand how code is processed step-by-step
Compiler education: Visualize semantic analysis phases
Code analysis tutorials: Interactive examples of language server internals

Data Quality

✅ Schema validated: All records follow consistent structure
✅ Data integrity: No corrupted or malformed records
✅ Completeness: All processed files represented
✅ Self-referential: rust-analyzer analyzing its own codebase

Technical Details

Format: Parquet files for efficient storage and fast loading
Compression: Snappy compression for optimal performance
Chunking: Files split to stay under 10MB for Git LFS compatibility
Schema: Strongly typed with proper null handling

Source

This dataset was generated by analyzing the rust-analyzer codebase (version 0.3.2000) using our custom integration that captures semantic analysis at multiple processing phases.

Source Project: /home/mdupont/2025/06/27/rust-analyzer Generated: August 2025 Tool: Custom rust-analyzer semantic extractor

Citation

If you use this dataset in your research, please cite:

@dataset{rust_analyzer_semantic_2025,
  title={Rust-Analyzer Semantic Analysis Dataset},
  author={Dupont, J. Mike},
  year={2025},
  publisher={Hugging Face},
  url={https://huggingface.co/datasets/introspector/rust-analyser}
}

License

This dataset is released under the AGPL-3.0 license, consistent with the rust-analyzer project.

Acknowledgments

Built using the rust-analyzer project
Generated with custom semantic analysis extraction tools
Optimized for machine learning and research applications