Infatoshi
Featured
netherite
A from-scratch C/CUDA reimplementation of Minecraft 1.11.2 that is bit-verified against the original Java game and includes batched CUDA reinforcement learning support.
AI repository intelligence
FR-AI / ANALYSISWhy it is worth attention
It delivers approximately 2× lower memory usage and 3.5× faster CSV writing than Python's built-in csv module, making it highly suitable for memory-sensitive and performance-critical data workloads in Lua environments.
Who it is for
- Lua developers working with CSV data
- Data engineers building ETL pipelines
- Backend engineers handling CSV-heavy services
- Machine learning practitioners preprocessing tabular data
Use cases
- Reading and writing large CSV files with low memory footprint
- Auto-detecting CSV dialects and headers via built-in sniffer
- Transforming tabular data using filter, map, sort, group by, and transpose
- Integrating into Lua-based analytics or preprocessing pipelines
Strengths
- Hybrid C/Lua architecture provides native performance with flexible API
- Significantly lower peak memory usage (~644 MB) compared to Python csv (~1261 MB) on a 1M-row benchmark
- Approximately 3.5× faster CSV writing than Python csv (0.17s vs 0.59s)
- Rich utility set including DictReader/Writer, dialects, sniffer, and data transforms
Considerations
- Full in-memory parsing is ~37% slower than Python's highly optimized C parser (9.2s vs 6.7s)
- No true streaming parser yet; currently loads entire dataset into memory for some operations
- Requires Lua 5.2+ and a C compiler for manual build, limiting portability to older Lua versions
README quick start
Installation
Description
A fast, memory-efficient CSV and tabular data toolkit for Lua.
Related repositories
Similar projects matched by category, topics, and programming language.
nsdkinx
Featured
twell
Twell is a tiny, embeddable, zero-allocation physics engine for interruptible, Apple-style UI animations, available as a single-header C library with Python bindings.
sqliteai
Featured
waste
WASTE is a C-based inference engine that streams MoE experts from disk to run a 2.78 trillion-parameter model on a consumer laptop with 64 GB RAM.