News

The Ruby gem and the Rust crate now expose one engine across every supported runtime. This post enumerates the supported formats, archives, runtimes, and platforms.

Each prebuilt platform gem now carries its own pure-portable cdylib next to the extension — the ffi tier works everywhere from MRI fallback to TruffleRuby — while the source gem stays binary-free.

Both engines now produce byte-identical parslet trees — the CanFlatten fold is ported exactly, dynamic blocks can read AND write captures across the FFI boundary, and TreeDiff.why explains any divergence in one call.

parsanol-rs 0.6–0.8 ships a bytecode VM with lead-byte dispatch and selective rule memoization, SIMD scan kernels, compiled-program artifact caching, and tree-sitter-style incremental parsing sessions.

Native Rust parsing is now the default on every mode — cause-tree errors, prefix and reporter parsing, incremental sessions — with a pure-Ruby engine always underneath as the portable floor.

Single parse() API, zero-overhead position info, Ruby 4.0 support, and WASM build fixes.

Ruby bindings now support native capture extraction, isolated scopes, and context-sensitive parsing via FFI callbacks.

Major release adding capture extraction, isolated scopes, and context-sensitive parsing.

Added Slice objects for lazy extraction and streaming parser support for large files.

Introduced backend abstraction with trait-based selection, Bytecode VM for faster parsing, and pre-commit hooks.

Improved release infrastructure with automated gem builds and cross-platform support.

First release of Parsanol Ruby bindings with native extension support and Parslet-compatible API.

First public release of Parsanol, a modern parser combinator library for Rust and Ruby.