Approved enhancements to consider after the core v2.0 Rust backend ships. Not blockers for the v2.0 release. **Wiki:** [Nice-to-Haves](https://github.com/fsecada01/TextSpitter/wiki/TextSpitter-2.0-Rust-Roadmap#nice-to-haves) ## Backlog - [ ] Memory-mapped file processing for very large PDFs (`memmap2` crate) — avoids loading entire file into RAM - [ ] SIMD-accelerated separator detection — opt-in via `[features] simd = []` / `pip install "textspitter[simd]"` - [ ] Streaming iterator API — yield chunks instead of collecting all; enables processing before extraction completes - [ ] `cargo bench` integration with [criterion](https://github.com/bheisler/criterion.rs) — reproducible micro-benchmarks replacing the ad-hoc `bench_splitting.py` - [ ] PyPI publish job for manylinux wheels (already wired in Phase 6 CI — just needs the publish trigger on release) ## Notes These are all approved and on the roadmap. Pick up after v2.0.0 ships to `main`.
Approved enhancements to consider after the core v2.0 Rust backend ships. Not blockers for the v2.0 release.
Wiki: Nice-to-Haves
Backlog
memmap2crate) — avoids loading entire file into RAM[features] simd = []/pip install "textspitter[simd]"cargo benchintegration with criterion — reproducible micro-benchmarks replacing the ad-hocbench_splitting.pyNotes
These are all approved and on the roadmap. Pick up after v2.0.0 ships to
main.